12021-08-30 07:46:52 nwclark 2 version 3.25 3 * No changes from previous version 4 52021-08-25 08:05:16 nwclark 6 version 3.24_50 7 * Remove code and tests only present to support perls before 5.6.1 8 Storable implicitly needs >= v5.6.1. 9 * Use SvPVCLEAR from ppport.h 10 * Remove XS code "commented" out with #if 0 11 * Refactor store_lhash() to remove some code duplication 12 * Avoid calling hv_iterval() twice for each hash entry 13 * Fix a bug in the recursion depth check in store_lhash() 14 15unreleased 16 version 3.24 17 * Fix a (possible) typo in Sntohl 18 19unreleased 20 version 3.23 21 * Fix typos 22 * avoid stderr noise in t/canonical.t 23 242020-07-31 19:36:37 atoomic 25 version 3.22 26 * use PERL_COMPARE macros 27 282020-04-23 13:33:05 ilmari 29 version 3.21 30 * fix repeated-word typos 31 * fix t/huge.t PERL_TEST_MEMORY diagnostic messages 32 332020-01-27 10:27:00 TonyC 34 version 3.20 35 * fix a format string and arguments for some debugging text 36 * linkify references to alternatives to Storable 37 382020-01-27 11:01:00 TonyC 39 version 3.19 40 * add casts to match some I32 parameters to "%d" formats (#17339) 41 * fix dependencies in Makefile.PL -> META (#17422) 42 * make use of note() optional, this requires a newer version of 43 Test::More and there's a circular dependency between later 44 versions of Test::More and Storable (#17422) 45 462019-11-19 07:59:39 TonyC 47 version 3.18 48 * update bug tracker to point at github (#17298) 49 * disallow vstring magic strings over 2GB-1 (#17306) 50 * mark some ASCII dependent tests as ASCII platform only 51 522019-08-08 11:48:00 TonyC 53 version 3.17 54 * correct a data type to ensure the check for too large results from 55 STORABLE_freeze() are detected correctly (detected by Coverity) 56 * removed remains of stack size detection from the build process. 57 * moved CAN_FLOCK detection into XS to simplify the build process. 58 592019-06-11 10:43:00 TonyC 60 version 3.16 61 * (perl #134179) fix self-referencing structures that include regexps 62 * bless regexps to preserve bless qr//, "Foo" 63 642019-04-23 16:00:00 xsawyerx 65 version 3.15 66 * Fix leaking. 67 68unreleased 69 version 3.14 70 * (perl #133708) don't build-time probe for stack limits at all 71 72unreleased 73 version 3.12 74 * (perl #133411) don't probe for stack limits with -Dusecrosscompile 75 762018-04-27 20:40:00 xsawyerx 77 version 3.11 78 * Fix Strawberry Perl build failures. 79 802018-04-21 22:00:00 xsawyerx 81 Version 3.10 82 * Fix binary artifacts from distribution. 83 842018-04-21 16:49:00 xsawyerx 85 Version 3.09 86 * Fix "provides" in metadata (META.yml/META.json) to use the Storable 87 template instead of a small other file (which also didn't exist). 88 892018-04-21 11:23:00 xsawyerx 90 Version 3.08 91 * (perl #132849) try to disable core files when deliberatly segfaulting. 92 * (perl #127743) don't probe Storable limits so much. 93 * (perl #132893) don't probe for Storable recursion limits on old Win32. 94 * (perl #132870) workaround VC2017 compiler bug. 95 * (perl #127743) re-work for debugging builds with MSVC. 96 * (perl #133039) dont build a Storable.so/.dll with a static perl build. 97 982018-02-07 15:08:00 tonyc 99 Version 3.06 100 101 * support large object ids. The code in theory supported arrays 102 with more than 2**32 elements, but references to the elements 103 emitted at the end of the array with be retrieved as references to 104 the wrong elements. 105 * 32-bit object ids over 2**31-1 weren't correctly handled. 106 * hook object id generation now supports 64-bit ids where needed 107 * writing 64-bit lengths in network order now works 108 * reading 64-bit lengths in network order now reads the components 109 in the correct order. 110 * retrieving large object tags are now only handled on 64-bit 111 platforms, large object tags should only be emitted for objects 112 that are too large for the 32-bit address space, so it was only 113 wasted code. 114 * reading 32-bit lengths for LSCALAR and LUTF8STR as unsigned 115 (perl #131990) 116 * reading flagged large object hashes didn't read the flags 117 * treat the 32-bit size of hook data as unsigned, values over 2GB 118 were treated as large (close to 2**64) parameters to NEWSV(). 119 (perl #131999) 120 * added support for hook data over 4GB in size 121 * zero length data receievd from STORABLE_freeze() no longer 122 results in an invalid SV being passed to STORABLE_thaw/_attach() 123 (perl #118551) 124 * where practical, padding is now cleared when emitting a long 125 double (perl #131136) 126 * cache the value of $Storable::DEBUGME (since cperl enabled 127 Storable TRACEME builds for all -DDEBUGGING builds) 128 * no longer discard exceptions thrown by 129 STORABLE_freeze/_thaw/attach() (perl #25933) 130 * fix dependencies used to build Storable.pm from __Storable__.pm 131 * add experimental support for freezing/thawing regular 132 expressions (perl #50608) 133 * re-work recursion limiting to store the limit in a perl variable 134 instead of baked into Storable.$so. This allows static Storable 135 builds to work, and avoids the kind of circular reference on 136 Storable.$so. 137 1382017-07-24 13:57:13 rurban 139 Version 3.05_13 140 141 * mingw fix: use safe defaults, not segfaulting defaults. 142 mingw fails on the stacksize binary search, leaving it empty. 143 144Wed Apr 19 09:11:07 2017 +0200 Reini Urban <rurban@cpan.org> 145 Version 3.05_12 146 147 * enhance stack reserve from 8 to 16 148 * fix LD_LIBRARY_PATH usage for CORE 149 * fixed some coverity warnings and leaks 150 * added a release make target 151 152Wed Mar 29 21:04:28 2017 +0200 Reini Urban <rurban@cpan.org> 153 Version 3.05_11 154 155 * croak on sizes read > I32_MAX 156 * simplify last_op_in_netorder 157 * protect from empty retrieve_vstring 158 * protect store_other error buf, potential static 159 buffer overflow. 160 161Tue Mar 14 09:52:20 2017 +0100 Reini Urban <rurban@cpan.org> 162 Version 3.05_10 163 164 * CORE-only improvements to stacksize 165 166Thu Mar 9 19:20:19 2017 +0100 Reini Urban <rurban@cpan.org> 167 Version 3.05_09 168 169 * compute the stacksizes, improve cleanup within croak 170 from stack exhaustion. 171 * added stack_depth and stack_depth_hash getters. 172 173Wed Mar 8 21:03:43 CET 2017 Reini Urban <rurban@cpan.org> 174 Version 3.05_08 175 176 * finetune the max stack limit, for C++, DEBUGGING and 32bit. 177 * fix t/blessed.t for cperl5.22 178 179Sun Mar 5 13:36:47 2017 +0100 Reini Urban <rurban@cpan.org> 180 Version 3.05_07 181 182 * Fixed a podchecker issue 183 184Sun Mar 5 11:42:04 2017 +0100 Reini Urban <rurban@cpan.org> 185 Version 3.05_06 186 187 * Fixed wrong recursion depth error with large arrays containing 188 another array. 189 L<[cperl #257]|https://github.com/perl11/cperl/issues/257> 190 191Thu Feb 2 12:40:44 2017 +0100 Reini Urban <rurban@cpanel.net> 192 Version 3.05_05 193 194 * Add leak tests for [cpan #97316], [perl #121928] 195 * Limit the max recursion depth to 1200 on 32bit systems. 196 We have no max_depth option yet, as in JSON::XS. 197 198Thu Feb 2 11:59:21 2017 +0100 Reini Urban <rurban@cpanel.net> 199 Version 3.05_04 200 201 * Fix retrieve_tied_array which fails since 5.16 202 [cpan #84705] 203 * Improve t/blessed.t in the creation of sv_yes/sv_no 204 with threaded perls. 205 206Tue Jan 31 02:55:30 2017 +0100 Reini Urban <rurban@cpanel.net> 207 Version 3.05_03 208 209 * Tune t/recurse.t stack-overflow limit more. 210 211Mon Jan 30 19:50:29 2017 +0100 Reini Urban <rurban@cpanel.net> 212 Version 3.05_02 213 214 * Tune t/recurse.t stack-overflow limit. Small 64bit systems overflow 215 even with depth 3000, where 32bit are ok. 216 217Mon Jan 30 15:13:38 2017 +0100 Reini Urban <rurban@cpanel.net> 218 Version 3.05_01 219 220 * Protect against stack overflows with nested arrays and hashes 221 [cpan #97526]. This imposes a new limit to your nested structures, 222 but JSON::XS has a limit of 512. We use a max_depth of 3000 for the 223 typical stack limit of 8k. 224 225 226Sun Jan 29 11:36:43 2017 +0100 Reini Urban <rurban@cpanel.net> 227 Version 3.05 228 229 * Protect against classname len overflow on the stack 230 and 2x on the heap with retrieve_bless and retrieve_hook. 231 A serious security issue with malcrafted storable files or buffers, 232 but p5p accepts no CVE on Storable attacks. See RT #130635 233 (reported by JD). 234 * Fix NULL ptr SEGVs with retrieve_code and retrieve_other. 235 See RT #130098 (reported and fixed by JD) 236 * Fix wrong huge LOBJECT support, broken since 3.00c. 237 Repro with `export PERL_TEST_MEMORY=8` 238 * Fix the few remaining 2-arg open calls. 239 * Portability and backport fixes back to 5.6.2 240 241Sat Jan 7 09:01:29 2017 +0100 Reini Urban <rurban@cpanel.net> 242 Version 3.04c 243 244 * fix printf types and warnings, esp. for 32bit use64bitint 245 * Change sv_setpvn(…, "…", …) to sv_setpvs(…, "…") 246 247Tue Jul 26 11:49:33 2016 +1000 Tony Cook <tony@develop-help.com> 248 Version 3.03c 249 250 * remove . from @INC when loading optional modules 251 252Sun Nov 20 18:06:45 2016 +0100 Reini Urban <rurban@cpanel.net> 253 Version 3.02c 254 255 * Fix -Wc++11-compat warnings, fix -Wchar-subscripts 256 257Fri Sep 16 01:32:59 2016 +0200 Reini Urban <rurban@cpanel.net> 258 Version 3.01c 259 260 * Added warn_security("Movable-Type CVE-2015-1592 Storable metasploit attack") 261 when detecting the third destructive metasploit vector, 262 thawing bless \"mt-config.cgi", "CGITempFile". 263 264Thu Mar 31 17:10:27 2016 +0200 Reini Urban <rurban@cpanel.net> 265 Version 3.00c 266 267 * Added support for u64 strings, arrays and hashes >2G 268 via a new LOBJECT tag. This is for 32bit systems and lengths 269 between 2GB and 4GB (I32-U32), and 64bit (>I32). 270 * Bumped STORABLE_BIN_MINOR and STORABLE_BIN_WRITE_MINOR from 10 to 11 271 * fix parallel tests, use unique filenames. 272 * fixed 2 instances of 2arg open, 273 * added optional flag arguments to skip tie and bless on retrieve/thaw, 274 * added SECURITY WARNING and Large data support to docs 275 * compute CAN_FLOCK at compile-time 276 * reformat everything consistently 277 * enable DEBUGME tracing and asserts with -DDEBUGGING 278 * fix all 64 bit compiler warnings 279 * added some abstraction methods to avoid code duplication 280 281?????? p5p <perl5-porters@perl.org> 282 Version 2.65 283 284 * Replace multiple 'use vars' by 'our' 285 * remove Config dependency 286 287Wed Jul 2 16:25:25 IST 2014 Abhijit Menon-Sen <ams@toroid.org> 288 Version 2.51 289 290 * [perl #121928] Fix memory leak for dclone inside freeze hook 291 (Alex Solovey) 292 * Do not call DESTROY for empty objects 293 (Vladimir Timofeev) 294 * Other bugfixes 295 296Sat Jul 13 18:34:27 IST 2013 Abhijit Menon-Sen <ams@toroid.org> 297 Version 2.45 298 299 * [perl #118829] Memory leaks in STORABLE_attach 300 (Vladimir Timofeev) 301 * [perl #118139] Don't SEGV during global destruction 302 (Nicholas Clark, report/test from Reini Urban) 303 * Added security warnings section (Steffen Mueller) 304 * Update INSTALLDIRS to favour installation in 'site' 305 (James E Keenan) 306 307Tue 11 Sep 06:51:11 IST 2012 Abhijit Menon-Sen <ams@toroid.org> 308 Version 2.39 309 310 Various bugfixes, including compatibility fixes for older 311 versions of Perl and vstring handling. 312 313Sun 3 Jul 09:10:11 IST 2011 Abhijit Menon-Sen <ams@toroid.org> 314 Version 2.29 315 316 Various bugfixes, notably including preventing nfreeze from 317 incorrectly stringifying integers. 318 319Fri 3 Dec 14:12:32 GMT 2010 David Leadbeater <dgl@dgl.cx> 320 Version 2.25 321 322 Support for serializing coderefs containing UTF-8. 323 324Fri Nov 12 10:52:19 IST 2010 Abhijit Menon-Sen <ams@toroid.org> 325 326 Version 2.24 327 328 Performance improvement for overloaded classes from Benjamin 329 Holzman. 330 331Fri Nov 12 10:36:22 IST 2010 Abhijit Menon-Sen <ams@toroid.org> 332 333 Version 2.23 334 335 Release the latest version from the Perl repository. 336 337Thu Aug 6 10:55:50 IST 2009 Abhijit Menon-Sen <ams@toroid.org> 338 339 Version 2.21 340 341 Includes hints/hpux.pl that was inadvertently left out of 2.20. 342 343Mon May 18 09:38:20 IST 2009 Abhijit Menon-Sen <ams@toroid.org> 344 345 Version 2.20 346 347 Fix bug handling blessed references to overloaded objects, plus 348 other miscellaneous fixes. 349 350 (Version 2.19 was released with 5.8.9.) 351 352Thu Nov 22 13:24:18 IST 2007 Abhijit Menon-Sen <ams@toroid.org> 353 354 Version 2.18 355 356 Compile fixes for older Perls. (No functional changes.) 357 358Sat Nov 17 02:12:12 IST 2007 Abhijit Menon-Sen <ams@toroid.org> 359 360 Version 2.17 361 362 Various broken tests fixed. (No functional changes.) 363 364Sat Mar 31 06:11:06 IST 2007 Abhijit Menon-Sen <ams@toroid.org> 365 366 Version 2.16 367 368 1. Fixes to Storable::dclone, read_magic, retrieve_lscalar 369 2. Storable 0.1 compatibility 370 3. Miscellaneous compile/leak/test/portability fixes 371 372Mon May 23 22:48:49 IST 2005 Abhijit Menon-Sen <ams@wiw.org> 373 374 Version 2.15 375 376 Minor changes to address a couple of compile problems. 377 378Mon Apr 25 07:29:14 IST 2005 Abhijit Menon-Sen <ams@wiw.org> 379 380 Version 2.14 381 382 1. Store weak references 383 2. Add STORABLE_attach hook. 384 385Thu Jun 17 12:26:43 BST 2004 Nicholas Clark <nick@ccl4.org> 386 387 Version 2.13 388 389 1. Don't change the type of top level overloaded references to RV - 390 they are perfectly correct as PVMG 391 2. Storable needs to cope with incoming frozen data that happens to be 392 utf8 encoded. 393 394Wed Mar 17 15:40:29 GMT 2004 Nicholas Clark <nick@ccl4.org> 395 396 Version 2.12 397 398 1. Add regression tests for the auto-require of STORABLE_thaw 399 2. Add auto-require of modules to restore overloading (and tests) 400 3. Change to no context (should give speedup with ithreads) 401 402Sat Mar 13 20:11:03 GMT 2004 Nicholas Clark <nick@ccl4.org> 403 404 Version 2.11 405 406 1. Storing restricted hashes in canonical order would SEGV. Fixed. 407 2. It was impossible to retrieve references to PL_sv_no and 408 PL_sv_undef from STORABLE_thaw hooks. 409 3. restrict.t was failing on 5.8.0, due to 5.8.0's unique 410 implementation of restricted hashes using PL_sv_undef 411 4. These changes allow a space optimisation for restricted hashes. 412 413Sat Jan 24 16:22:32 IST 2004 Abhijit Menon-Sen <ams@wiw.org> 414 415 Version 2.10 416 417 1. Thread safety: Storable::CLONE/init_perlinterp() now create 418 a new Perl context for each new ithread. 419 (From Stas Bekman and Jan Dubois.) 420 2. Fix a tag count mismatch with $Storable::Deparse that caused 421 all back-references after a stored sub to be off-by-N (where 422 N was the number of code references in between). 423 (From Sam Vilain.) 424 3. Prevent CODE references from turning into SCALAR references. 425 (From Slaven Rezic.) 426 427Sat Jan 3 18:49:18 GMT 2004 Nicholas Clark <nick@ccl4.org> 428 429 Version 2.09 430 431 Fix minor problems with the CPAN release 432 1: Make Storable.xs work on 5.8.2 and later (already in the core) 433 2: Ship the linux hints file 434 3: Ship Test::More for the benefit of Perls pre 5.6.2 435 4: Correct Makefile.PL to only install in core for 5.8.0 and later 436 437Sat Sep 6 01:08:20 IST 2003 Abhijit Menon-Sen <ams@wiw.org> 438 439 Version 2.08 440 441 This release works around a 5.8.0 bug which caused hashes to not 442 be marked as having key flags even though an HEK had HEK_WASUTF8 443 set. (Note that the only reasonable solution is to silently drop 444 the flag from the affected key.) 445 446 Users of RT 3 who were seeing assertion failures should upgrade. 447 (Perl 5.8.1 will have the bug fixed.) 448 449Mon May 5 10:24:16 IST 2003 Abhijit Menon-Sen <ams@wiw.org> 450 451 Version 2.07 452 453 Minor bugfixes (self-tied objects are now correctly stored, as 454 are the results of additions larger than INT_MAX). 455 456Mon Oct 7 21:56:38 BST 2002 Nicholas Clark <nick@ccl4.org> 457 458 Version 2.06 459 460 Remove qr// from t/downgrade.t so that it will run on 5.004 461 Mention $File::Spec::VERSION a second time in t/forgive.t so that it 462 runs without warnings in 5.004 (this may be a 5.00405 bug I'm working 463 round) 464 Fix t/integer.t initialisation to actually generate 64 bits of 9c 465 Fix comparison tests to use eval to get around 64 bit IV conversion 466 issues on 5.6.x, following my t/integer.t ^ precedence bug found by 467 Rafael Garcia-Suarez 468 Alter t/malice.t to work with Test/More.pm in t/, and skip individual 469 subtests that use $Config{ptrsize}, so that the rest of the test can 470 now be run with 5.004 471 Change t/malice.t and the error message in check_magic in Storable.xs 472 from "Pointer integer size" to "Pointer size" 473 Remove prerequisite of Test::More from Makefile.PL 474 Ship Test::Builder, Test::Simple and Test::More in t 475 476Thu Oct 3 08:57:22 IST 2002 Abhijit Menon-Sen <ams@wiw.org> 477 478 Version 2.05 479 480 Adds support for CODE references from Slaven Rezic 481 <slaven.rezic@berlin.de>. 482 483Fri Jun 7 23:55:41 BST 2002 Nicholas Clark 484 485 Version 2.04 486 487 Bug fix from Radu Greab <radu@netsoft.ro> (plus regression test) 488 to fix a recently introduced bug detected by Dave Rolsky. 489 Bug was that for a non threaded build, the class information was 490 being lost at freeze time on the first object with a STORABLE_freeze 491 hook. Consequentially the object was not blessed at all when thawed. 492 (The presence (or lack) of STORABLE_thaw was irrelevant; this was 493 a store-time data lost bug, caused by failure to initialize internal 494 context) 495 The bug was introduced as development perl change 16442 (on 496 2002/05/07), so has been present since 2.00. 497 Patches to introduce more regression tests to reduce the chance of 498 a reoccurrence of this sort of goof are always welcome. 499 500Thu May 30 20:31:08 BST 2002 Nicholas Clark <nick@ccl4.org> 501 502 Version 2.03 Header changes on 5.6.x on Unix where IV is long long 503 504 5.6.x introduced the ability to have IVs as long long. However, 505 Configure still defined BYTEORDER based on the size of a long. 506 Storable uses the BYTEORDER value as part of the header, but 507 doesn't explicitly store sizeof(IV) anywhere in the header. 508 Hence on 5.6.x built with IV as long long on a platform that 509 uses Configure (ie most things except VMS and Windows) headers 510 are identical for the different IV sizes, despite the files 511 containing some fields based on sizeof(IV) 512 513 5.8.0 is consistent; all platforms have BYTEORDER in config.h 514 based on sizeof(IV) rather than sizeof(long). This means that 515 the value of BYTEORDER will change from (say) 4321 to 87654321 516 between 5.6.1 and 5.8.0 built with the same options to Configure 517 on the same machine. This means that the Storable header will 518 differ, and the two versions will wrongly thing that they are 519 incompatible. 520 521 For the benefit of long term consistency, Storable now 522 implements the 5.8.0 BYTEORDER policy on 5.6.x. This means that 523 2.03 onwards default to be incompatible with 2.02 and earlier 524 (ie the large 1.0.x installed base) on the same 5.6.x perl. 525 526 To allow interworking, a new variable 527 $Storable::interwork_56_64bit is introduced. It defaults to 528 false. Set it to true to read and write old format files. Don't 529 use it unless you have existing stored data written with 5.6.x 530 that you couldn't otherwise read, or you need to interwork with 531 a machine running older Storable on a 5.6.x with long long IVs 532 (i.e., you probably don't need to use it). 533 534Sat May 25 22:38:39 BST 2002 Nicholas Clark <nick@ccl4.org> 535 536 Version 2.02 537 538 Rewrite Storable.xs so that the file header structure for write_magic 539 is built at compile time, and check_magic attempts to the header in 540 blocks rather than byte per byte. These changes make the compiled 541 extension 2.25% smaller, but are not significant enough to give a 542 noticeable speed up. 543 544Thu May 23 22:50:41 BST 2002 Nicholas Clark <nick@ccl4.org> 545 546 Version 2.01 547 548 - New regression tests integer.t 549 - Add code to safely store large unsigned integers. 550 - Change code not to attempt to store large integers (ie > 32 bits) 551 in network order as 32 bits. 552 553 *Never* underestimate the value of a pathological test suite carefully 554 crafted with maximum malice before writing a line of real code. It 555 prevents crafty bugs from stowing away in your released code. 556 It's much less embarrassing to find them before you ship. 557 (Well, never underestimate it if you ever want to work for me) 558 559Fri May 17 22:48:59 BST 2002 Nicholas Clark <nick@ccl4.org> 560 561 Version 2.0, binary format 2.5 (but writes format 2.4 on pre 5.7.3) 562 563 The perl5 porters have decided to make sure that Storable still 564 builds on pre-5.8 perls, and make the 5.8 version available on CPAN. 565 The VERSION is now 2.0, and it passes all tests on 5.005_03, 5.6.1 566 and 5.6.1 with threads. On 5.6.0 t/downgrade.t fails tests 34 and 37, 567 due to a bug in 5.6.0 - upgrade to 5.6.1. 568 569 Jarkko and I have collated the list of changes the perl5 porters have 570 from the perl5 Changes file: 571 572 - data features of upcoming perl 5.8.0 are supported: Unicode hash 573 keys (Unicode hash values have been supported since Storable 1.0.1) 574 and "restricted hashes" (readonly hashes and hash entries) 575 - a newer version of perl can now be used to serialize data which is 576 not supported in earlier perls: Storable will attempt to do the 577 right thing for as long as possible, croaking only when safe data 578 conversion simply isn't possible. Alternatively earlier perls can 579 opt to have a lossy downgrade data instead of croaking 580 - when built with perls pre 5.7.3 this Storable writes out files 581 with binary format 2.4, the same format as Storable 1.0.8 onwards. 582 This should mean that this Storable will inter-operate seamlessly 583 with any Storable 1.0.8 or newer on perls pre 5.7.3 584 - dclone() now works with empty string scalar objects 585 - retrieving of large hashes is now more efficient 586 - more routines autosplit out of the main module, so Storable should 587 load slightly more quickly 588 - better documentation 589 - the internal context objects are now freed explicitly, rather than 590 relying on thread or process exit 591 - bugs fixed in debugging trace code affecting builds made with 64 bit 592 IVs 593 - code tidy-ups to allow clean compiles with more warning options 594 turned on avoid problems with $@ getting corrupted on 5.005_03 if 595 Carp wasn't already loaded 596 - added &show_file_magic, so you can add to /etc/magic and teach 597 Unix's file command about Storable files 598 599 We plan to keep Storable on CPAN in sync with the Perl core, so 600 if you encounter bugs or other problems building or using Storable, 601 please let us know at perl5-porters@perl.org 602 Patches welcome! 603 604Sat Dec 1 14:37:54 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com> 605 606 This is the LAST maintenance release of the Storable module. 607 Indeed, Storable is now part of perl 5.8, and will be maintained 608 as part of Perl. The CPAN module will remain available there 609 for people running pre-5.8 perls. 610 611 Avoid requiring Fcntl upfront, useful to embedded runtimes. 612 Use an eval {} for testing, instead of making Storable.pm 613 simply fail its compilation in the BEGIN block. 614 615 store_fd() will now correctly autoflush file if needed. 616 617Tue Aug 28 23:53:20 MEST 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com> 618 619 Fixed truncation race with lock_retrieve() in lock_store(). 620 The file has to be truncated only once the exclusive lock is held. 621 622 Removed spurious debugging messages in .xs file. 623 624Sun Jul 1 13:27:32 MEST 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com> 625 626 Systematically use "=over 4" for POD linters. 627 Apparently, POD linters are much stricter than would 628 otherwise be needed, but that's OK. 629 630 Fixed memory corruption on croaks during thaw(). Thanks 631 to Claudio Garcia for reproducing this bug and providing the 632 code to exercise it. Added test cases for this bug, adapted 633 from Claudio's code. 634 635 Made code compile cleanly with -Wall (from Jarkko Hietaniemi). 636 637 Changed tagnum and classnum from I32 to IV in context. Also 638 from Jarkko. 639 640Thu Mar 15 01:22:32 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com> 641 642 Last version was wrongly compiling with assertions on, due 643 to an edit glitch. That did not cause any problem (apart from 644 a slight performance loss) excepted on Win* platforms, where the 645 assertion code does not compile. 646 647Sat Feb 17 13:37:37 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com> 648 649 Version 1.0.10. 650 651 Forgot to increase version number at previous patch (there were 652 two of them, which is why we jump from 1.0.8 to 1.0.10). 653 654Sat Feb 17 13:35:00 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com> 655 656 Version 1.0.8, binary format 2.4. 657 658 Fixed incorrect error message. 659 660 Now bless objects ASAP at retrieve time, which is meant to fix 661 two bugs: 662 663 * Indirect references to overloaded object were not able to 664 restore overloading if the object was not blessed yet, 665 which was possible since blessing occurred only after the 666 recursive retrieval. 667 668 * Storable hooks asking for serialization of blessed ref could 669 get un-blessed refs at retrieval time, for the very same 670 reason. 671 672 The fix implemented here was suggested by Nick Ing-Simmons. 673 674 Added support for blessed ref to tied structures. This is the 675 cause for the binary format change. 676 677 Added EBCDIC version of the compatibility test with 0.6.11, 678 from Peter Prymmer 679 680 Added tests for the new features, and to make sure the bugs they 681 are meant to fix are indeed fixed. 682 683Wed Jan 3 10:43:18 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com> 684 685 Removed spurious 'clean' entry in Makefile.PL. 686 687 Added CAN_FLOCK to determine whether we can flock() or not, 688 by inspecting Perl's configuration parameters, as determined 689 by Configure. 690 691 Trace offending package when overloading cannot be restored 692 on a scalar. 693 694 Made context cleanup safer to avoid dup freeing, mostly in the 695 presence of repeated exceptions during store/retrieve (which can 696 cause memory leaks anyway, so it's just additional safety, not a 697 definite fix). 698 699Sun Nov 5 18:23:48 MET 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 700 701 Version 1.0.6. 702 703 Fixed severe "object lost" bug for STORABLE_freeze returns, 704 when refs to lexicals, taken within the hook, were to be 705 serialized by Storable. Enhanced the t/recurse.t test to 706 stress hook a little more with refs to lexicals. 707 708Thu Oct 26 19:14:38 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 709 710 Version 1.0.5. 711 712 Documented that store() and retrieve() can return undef. 713 That is, the error reporting is not always made via exceptions, 714 as the paragraph on error reporting was implying. 715 716 Auto requires module of blessed ref when STORABLE_thaw misses. 717 When the Storable engine looks for the STORABLE_thaw hook and 718 does not find it, it now tries to require the package into which 719 the blessed reference is. 720 721 Just check $^O, in t/lock.t: there's no need to pull the whole 722 Config module for that. 723 724Mon Oct 23 20:03:49 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 725 726 Version 1.0.4. 727 728 Protected calls to flock() for DOS platform: apparently, the 729 flock/fcnlt emulation is reported to be broken on that 730 platform. 731 732 Added logcarp emulation if they don't have Log::Agent, since 733 we now use it to carp when lock_store/lock_retrieve is used 734 on DOS. 735 736Fri Sep 29 21:52:29 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 737 738 Version 1.0.3. 739 740 Avoid using "tainted" and "dirty" since Perl remaps them via 741 cpp (i.e. #define). This is deeply harmful when threading 742 is enabled. This concerned both the context structure and 743 local variable and argument names. Brrr..., scary! 744 745Thu Sep 28 23:46:39 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 746 747 Version 1.0.2. 748 749 Fixed spelling in README. 750 751 Added lock_store, lock_nstore, and lock_retrieve (advisory locking) 752 after a proposal from Erik Haugan <erik@solbors.no>. 753 754 Perls before 5.004_04 lack newSVpvn, added remapping in XS. 755 756 Fixed stupid typo in the t/utf8.t test. 757 758Sun Sep 17 18:51:10 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 759 760 Version 1.0.1, binary format 2.3. 761 762 Documented that doubles are stored stringified by nstore(). 763 764 Added Salvador Ortiz Garcia in CREDITS section, He identified 765 a bug in the store hooks and proposed the right fix: the class 766 id was allocated too soon. His bug case was also added to 767 the regression test suite. 768 769 Now only taint retrieved data when source was tainted. A bug 770 discovered by Marc Lehmann. 771 772 Added support for UTF-8 strings, a contribution of Marc Lehmann. 773 This is normally only activated in post-5.6 perls. 774 775Thu Aug 31 23:06:06 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 776 777 First official release Storable 1.0, for inclusion in perl 5.7.0. 778 The license scheme is now compatible with Perl's. 779 780Thu Aug 24 01:02:02 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 781 782 ANSI-fied most of the code, preparing for Perl core integration. 783 The next version of Storable will be 0.8, and will be integrated 784 into the Perl core (development branch). 785 786 Dispatch tables were moved upfront to relieve some compilers, 787 especially on AIX and Windows platforms. 788 789 Merged 64-bit fixes from perl5-porters. 790 791Mon Aug 14 09:22:04 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 792 793 Added a refcnt dec in retrieve_tied_key(): sv_magic() increases 794 the refcnt on the mg_ptr as well. 795 796 Removed spurious dependency to Devel::Peek, which was used for 797 testing only in t/tied_items.t. Thanks to Conrad Heiney 798 <conrad@fringehead.org> for spotting it first. 799 800Sun Aug 13 22:12:59 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 801 802 Marc Lehmann kindly contributed code to add overloading support 803 and to handle references to tied variables. 804 805 Rewrote leading blurb about compatibility to make it clearer what 806 "backward compatibility" is about: when I say 0.7 is backward 807 compatible with 0.6, it means the revision 0.7 can read files 808 produced by 0.6. 809 810 Mention new Clone(3) extension in SEE ALSO. 811 812 Was wrongly optimizing for "undef" values in hashes by not 813 fully recursing: as a result, tied "undef" values were incorrectly 814 serialized. 815 816Sun Jul 30 12:59:17 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 817 818 First revision of Storable 0.7. 819 820 The serializing format is new, known as version 2.0. It is fully 821 backward compatible with 0.6. Earlier formats are deprecated and 822 have not even been tested: next version will drop pre-0.6 format. 823 824 Changes since 0.6@11: 825 826 - Moved interface to the "beta" status. Some tiny parts are still 827 subject to change, but nothing important enough to warrant an "alpha" 828 status any longer. 829 830 - Slightly reduced the size of the Storable image by factorizing 831 object class names and removing final object storage notification due 832 to a redesign of the blessed object storing. 833 834 - Classes can now redefine how they wish their instances to be serialized 835 and/or deep cloned. Serializing hooks are written in Perl code. 836 837 - The engine is now fully re-entrant. 838 839Sun Apr 2 23:47:50 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 840 841 Added provision to detect more recent binary formats, since 842 the new upcoming Storable-0.7 will use a different format. 843 In order to prevent attempting the de-serialization of newer 844 formats by older versions, I'm adding this now to the 0.6 series. 845 846 I'm expecting this revision to be the last of the 0.6 series. 847 Unless it does not work with perl 5.6, which I don't use yet, 848 and therefore against which I cannot test. 849 850Wed Mar 29 19:55:21 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 851 852 Added note about format incompatibilities with old versions 853 (i.e. pre 0.5@9 formats, which cannot be understood as there 854 was no versionning information in the file by then). 855 856 Protect all $@ variables when eval {} used, to avoid corrupting 857 it when store/retrieve is called within an exception handler. 858 859 Mistakenly included "patchlevel.h" instead of <patchlevel.h>, 860 preventing Perl's patchlevel from being included, which is 861 needed starting from 5.6. 862 863Tue May 12 09:15:15 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 864 865 Fixed shared "undef" bug in hashes, which did not remain shared 866 through store/retrieve. 867 868Thu Feb 10 19:48:16 MET 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com> 869 870 added last_op_in_netorder() predicate 871 documented last_op_in_netorder() 872 added tests for the new last_op_in_netorder() predicate 873 874Wed Oct 20 19:07:36 MEST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com> 875 876 Forgot to update VERSION 877 878Tue Oct 19 21:25:02 MEST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com> 879 880 Added mention of japanese translation for the manual page. 881 882 Fixed typo in macro that made threaded code not compilable, 883 especially on Win32 platforms. 884 885 Changed detection of older perls (pre-5.005) by testing PATCHLEVEL 886 directly instead of relying on internal symbols. 887 888Tue Sep 14 22:13:28 MEST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com> 889 890 Integrated "thread-safe" patch from Murray Nesbitt. 891 Note that this may not be very efficient for threaded code, 892 see comment in the code. 893 894 Try to avoid compilation warning on 64-bit CPUs. Can't test it, 895 since I don't have access to such machines. 896 897Mon Jul 12 14:37:19 METDST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com> 898 899 changed my e-mail to pobox. 900 901 mentioned it is not thread-safe. 902 903 updated version number. 904 905 uses new internal PL_* naming convention. 906 907Fri Jul 3 13:38:16 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 908 909 Updated benchmark figures due to recent optimizations done in 910 store(): tagnums are now stored as-is in the hash table, so 911 no surrounding SV is created. And the "shared keys" mode for 912 hash table was turned off. 913 914 Fixed backward compatibility (wrt 0.5@9) for retrieval of 915 blessed refs. That old version did something wrong, but the 916 bugfix prevented correct retrieval of the old format. 917 918Mon Jun 22 11:00:48 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 919 920 Changed benchmark figures. 921 922 Adjust refcnt of tied objects after calling sv_magic() to avoid 923 memory leaks. Contributed by Jeff Gresham. 924 925Fri Jun 12 11:50:04 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 926 927 Added workaround for persistent LVALUE-ness in perl5.004. All 928 scalars tagged as being an lvalue are handled as if they were 929 not an lvalue at all. Added test for that LVALUE bug workaround. 930 931 Now handles Perl immortal scalars explicitly, by storing &sv_yes 932 as such, explicitly. 933 934 Retrieval of non-immortal undef cannot be shared. Previous 935 version was over-optimizing by not creating a separate SV for 936 all undefined scalars seen. 937 938Thu Jun 4 17:21:51 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 939 940 Baseline for Storable-0.6@0. 941 942 This version introduces a binary incompatibility in the generated 943 binary image, which is more compact than older ones by approximatively 944 15%, depending on the exact degree of sharing in your structures. 945 946 The good news is that your older images can still be retrieved with 947 this version, i.e. backward compatibility is preserved. This version 948 of Storable can only generate new binaries however. 949 950 Another good news is that the retrieval of data structure is 951 significantly quicker than before, because a Perl array is used 952 instead of a hash table to keep track of retrieved objects, and 953 also because the image being smaller, less I/O function calls are 954 made. 955 956Tue May 12 09:15:15 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 957 958 Version number now got from Storable.pm directly. 959 960 Fixed overzealous sv_type() optimization, which would make 961 Storable fail when faced with an "upgraded" SV to the PVIV 962 or PVNV kind containing a reference. 963 964Thu Apr 30 15:11:30 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 965 966 Extended the SYNOPSIS section to give quick overview of the 967 routines and their signature. 968 969 Optimized sv_type() to avoid flags checking when not needed, i.e. 970 when their type makes it impossible for them to be refs or tied. 971 This slightly increases throughput by a few percents when refs 972 and tied variables are marginal occurrences in your data. 973 974 Stubs for XS now use OutputStream and InputStream file types to 975 make it work when the given file is actually a socket. Perl 976 makes a distinction for sockets in its internal I/O structures 977 by having both a read and a write structure, whereas plain files 978 share the same one. 979 980Tue Jun 3 09:41:33 METDST 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 981 982 Thanks to a contribution from Benjamin A. Holzman, Storable is now 983 able to correctly serialize tied SVs, i.e. tied arrays, hashes 984 and scalars. 985 986Thu Apr 9 18:07:51 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 987 988 I said SvPOK() had changed to SvPOKp(), but that was a lie... 989 990Wed Apr 8 13:14:29 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 991 992 Wrote sizeof(SV *) instead of sizeof(I32) when portable, which 993 in effect mangled the object tags and prevented portability 994 across 32/64 bit architectures! 995 996Wed Mar 25 14:57:02 MET 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 997 998 Added code example for store_fd() and retrieve_fd() in the 999 man page, to emphasize that file descriptors must be passed as 1000 globs, not as plain strings. 1001 1002 Cannot use SV addresses as tag when using nstore() on LP64. This 1003 was the cause of problems when creating a storable image on an 1004 LP64 machine and retrieving it on an ILP32 system, which is 1005 exactly what nstore() is meant for... 1006 1007 However, we continue to use SV addresses as tags for plain store(), 1008 because benchmarking shows that it saves up to 8% of the store 1009 time, and store() is meant to be fast at the expense of lack 1010 of portability. 1011 1012 This means there will be approximately an 8% degradation of 1013 performance for nstore(), but it's now working as expected. 1014 That cost may vary on your machine of course, since it is 1015 solely caused by the memory allocation overhead used to create 1016 unique SV tags for each distinct stored SV. 1017 1018Tue Jan 20 09:21:53 MET 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 1019 1020 Don't use any '_' in version number. 1021 1022Tue Jan 13 17:51:50 MET 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 1023 1024 Updated version number. 1025 1026 added binmode() calls for systems where it matters. 1027 1028 Be sure to pass globs, not plain file strings, to C routines, 1029 so that Storable can be used under the Perl debugger. 1030 1031Wed Nov 5 10:53:22 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 1032 1033 Fix memory leaks on seen hash table and returned SV refs. 1034 1035 Storable did not work properly when tainting enabled. 1036 1037 Fixed "Allocation too large" messages in freeze/thaw and added. 1038 proper regression test in t/freeze.t. 1039 1040Tue Jun 3 09:41:33 METDST 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 1041 1042 Updated version number 1043 1044 Added freeze/thaw interface and dclone. 1045 1046Fri May 16 10:45:47 METDST 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 1047 1048 Forgot that AutoLoader does not export its own AUTOLOAD. 1049 I could use 1050 1051 use AutoLoader 'AUTOLOAD'; 1052 1053 but that would not be backward compatible. So the export is 1054 done by hand... 1055 1056Tue Mar 25 11:21:32 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 1057 1058 Empty scalar strings are now "defined" at retrieval time. 1059 1060 New test to ensure an empty string is defined when retrieved. 1061 1062Thu Feb 27 16:32:44 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 1063 1064 Updated version number 1065 1066 Declare VERSION as being used 1067 1068 Fixed a typo in the PerlIO_putc remapping. 1069 PerlIO_read and perlIO_write inverted size/nb_items. 1070 (only relevant for pre-perl5.004 versions) 1071 1072Thu Feb 27 15:58:31 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 1073 1074 Updated version number 1075 1076 Added VERSION identification 1077 1078 Allow build with perl5.003, which is ante perlIO time 1079 1080Mon Jan 13 17:53:18 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 1081 1082 Random code fixes. 1083 1084Wed Jan 22 15:19:56 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com> 1085 1086 Updated version number in Makefile.PL. 1087 1088 Added "thanks to" section to README. 1089 1090 Documented new forgive_me variable. 1091 1092 Made 64-bit clean. 1093 1094 Added forgive_me support to allow store() of data structures 1095 containing non-storable items like CODE refs. 1096