1This is gccinstall.info, produced by makeinfo version 6.7 from 2install.texi. 3 4Copyright (C) 1988-2019 Free Software Foundation, Inc. 5 6 Permission is granted to copy, distribute and/or modify this document 7under the terms of the GNU Free Documentation License, Version 1.3 or 8any later version published by the Free Software Foundation; with no 9Invariant Sections, the Front-Cover texts being (a) (see below), and 10with the Back-Cover Texts being (b) (see below). A copy of the license 11is included in the section entitled "GNU Free Documentation License". 12 13 (a) The FSF's Front-Cover Text is: 14 15 A GNU Manual 16 17 (b) The FSF's Back-Cover Text is: 18 19 You have freedom to copy and modify this GNU Manual, like GNU 20software. Copies published by the Free Software Foundation raise funds 21for GNU development. 22INFO-DIR-SECTION Software development 23START-INFO-DIR-ENTRY 24* gccinstall: (gccinstall). Installing the GNU Compiler Collection. 25END-INFO-DIR-ENTRY 26 27 Copyright (C) 1988-2019 Free Software Foundation, Inc. 28 29 Permission is granted to copy, distribute and/or modify this document 30under the terms of the GNU Free Documentation License, Version 1.3 or 31any later version published by the Free Software Foundation; with no 32Invariant Sections, the Front-Cover texts being (a) (see below), and 33with the Back-Cover Texts being (b) (see below). A copy of the license 34is included in the section entitled "GNU Free Documentation License". 35 36 (a) The FSF's Front-Cover Text is: 37 38 A GNU Manual 39 40 (b) The FSF's Back-Cover Text is: 41 42 You have freedom to copy and modify this GNU Manual, like GNU 43software. Copies published by the Free Software Foundation raise funds 44for GNU development. 45 46 47File: gccinstall.info, Node: Top, Up: (dir) 48 49* Menu: 50 51* Installing GCC:: This document describes the generic installation 52 procedure for GCC as well as detailing some target 53 specific installation instructions. 54 55* Specific:: Host/target specific installation notes for GCC. 56* Binaries:: Where to get pre-compiled binaries. 57 58* Old:: Old installation documentation. 59 60* GNU Free Documentation License:: How you can copy and share this manual. 61* Concept Index:: This index has two entries. 62 63 64File: gccinstall.info, Node: Installing GCC, Next: Binaries, Up: Top 65 661 Installing GCC 67**************** 68 69The latest version of this document is always available at 70http://gcc.gnu.org/install/. It refers to the current development 71sources, instructions for specific released versions are included with 72the sources. 73 74 This document describes the generic installation procedure for GCC as 75well as detailing some target specific installation instructions. 76 77 GCC includes several components that previously were separate 78distributions with their own installation instructions. This document 79supersedes all package-specific installation instructions. 80 81 _Before_ starting the build/install procedure please check the *note 82host/target specific installation notes: Specific. We recommend you 83browse the entire generic installation instructions before you proceed. 84 85 Lists of successful builds for released versions of GCC are available 86at <http://gcc.gnu.org/buildstat.html>. These lists are updated as new 87information becomes available. 88 89 The installation procedure itself is broken into five steps. 90 91* Menu: 92 93* Prerequisites:: 94* Downloading the source:: 95* Configuration:: 96* Building:: 97* Testing:: (optional) 98* Final install:: 99 100 Please note that GCC does not support 'make uninstall' and probably 101won't do so in the near future as this would open a can of worms. 102Instead, we suggest that you install GCC into a directory of its own and 103simply remove that directory when you do not need that specific version 104of GCC any longer, and, if shared libraries are installed there as well, 105no more binaries exist that use them. 106 107 108File: gccinstall.info, Node: Prerequisites, Next: Downloading the source, Up: Installing GCC 109 1102 Prerequisites 111*************** 112 113GCC requires that various tools and packages be available for use in the 114build procedure. Modifying GCC sources requires additional tools 115described below. 116 117Tools/packages necessary for building GCC 118========================================= 119 120ISO C++98 compiler 121 Necessary to bootstrap GCC, although versions of GCC prior to 4.8 122 also allow bootstrapping with a ISO C89 compiler and versions of 123 GCC prior to 3.4 also allow bootstrapping with a traditional (K&R) 124 C compiler. 125 126 To build all languages in a cross-compiler or other configuration 127 where 3-stage bootstrap is not performed, you need to start with an 128 existing GCC binary (version 3.4 or later) because source code for 129 language frontends other than C might use GCC extensions. 130 131 Note that to bootstrap GCC with versions of GCC earlier than 3.4, 132 you may need to use '--disable-stage1-checking', though 133 bootstrapping the compiler with such earlier compilers is strongly 134 discouraged. 135 136C standard library and headers 137 138 In order to build GCC, the C standard library and headers must be 139 present for all target variants for which target libraries will be 140 built (and not only the variant of the host C++ compiler). 141 142 This affects the popular 'x86_64-pc-linux-gnu' platform (among 143 other multilib targets), for which 64-bit ('x86_64') and 32-bit 144 ('i386') libc headers are usually packaged separately. If you do a 145 build of a native compiler on 'x86_64-pc-linux-gnu', make sure you 146 either have the 32-bit libc developer package properly installed 147 (the exact name of the package depends on your distro) or you must 148 build GCC as a 64-bit only compiler by configuring with the option 149 '--disable-multilib'. Otherwise, you may encounter an error such 150 as 'fatal error: gnu/stubs-32.h: No such file' 151 152GNAT 153 154 In order to build the Ada compiler (GNAT) you must already have 155 GNAT installed because portions of the Ada frontend are written in 156 Ada (with GNAT extensions.) Refer to the Ada installation 157 instructions for more specific information. 158 159A "working" POSIX compatible shell, or GNU bash 160 161 Necessary when running 'configure' because some '/bin/sh' shells 162 have bugs and may crash when configuring the target libraries. In 163 other cases, '/bin/sh' or 'ksh' have disastrous corner-case 164 performance problems. This can cause target 'configure' runs to 165 literally take days to complete in some cases. 166 167 So on some platforms '/bin/ksh' is sufficient, on others it isn't. 168 See the host/target specific instructions for your platform, or use 169 'bash' to be sure. Then set 'CONFIG_SHELL' in your environment to 170 your "good" shell prior to running 'configure'/'make'. 171 172 'zsh' is not a fully compliant POSIX shell and will not work when 173 configuring GCC. 174 175A POSIX or SVR4 awk 176 177 Necessary for creating some of the generated source files for GCC. 178 If in doubt, use a recent GNU awk version, as some of the older 179 ones are broken. GNU awk version 3.1.5 is known to work. 180 181GNU binutils 182 183 Necessary in some circumstances, optional in others. See the 184 host/target specific instructions for your platform for the exact 185 requirements. 186 187gzip version 1.2.4 (or later) or 188bzip2 version 1.0.2 (or later) 189 190 Necessary to uncompress GCC 'tar' files when source code is 191 obtained via FTP mirror sites. 192 193GNU make version 3.80 (or later) 194 195 You must have GNU make installed to build GCC. 196 197GNU tar version 1.14 (or later) 198 199 Necessary (only on some platforms) to untar the source code. Many 200 systems' 'tar' programs will also work, only try GNU 'tar' if you 201 have problems. 202 203Perl version between 5.6.1 and 5.6.24 204 205 Necessary when targeting Darwin, building 'libstdc++', and not 206 using '--disable-symvers'. Necessary when targeting Solaris 2 with 207 Sun 'ld' and not using '--disable-symvers'. The bundled 'perl' in 208 Solaris 8 and up works. 209 210 Necessary when regenerating 'Makefile' dependencies in libiberty. 211 Necessary when regenerating 'libiberty/functions.texi'. Necessary 212 when generating manpages from Texinfo manuals. Used by various 213 scripts to generate some files included in the source repository 214 (mainly Unicode-related and rarely changing) from source tables. 215 216 Used by 'automake'. 217 218 Several support libraries are necessary to build GCC, some are 219required, others optional. While any sufficiently new version of 220required tools usually work, library requirements are generally 221stricter. Newer versions may work in some cases, but it's safer to use 222the exact versions documented. We appreciate bug reports about problems 223with newer versions, though. If your OS vendor provides packages for 224the support libraries then using those packages may be the simplest way 225to install the libraries. 226 227GNU Multiple Precision Library (GMP) version 4.3.2 (or later) 228 229 Necessary to build GCC. If a GMP source distribution is found in a 230 subdirectory of your GCC sources named 'gmp', it will be built 231 together with GCC. Alternatively, if GMP is already installed but 232 it is not in your library search path, you will have to configure 233 with the '--with-gmp' configure option. See also '--with-gmp-lib' 234 and '--with-gmp-include'. The in-tree build is only supported with 235 the GMP version that download_prerequisites installs. 236 237MPFR Library version 2.4.2 (or later) 238 239 Necessary to build GCC. It can be downloaded from 240 <https://www.mpfr.org>. If an MPFR source distribution is found in 241 a subdirectory of your GCC sources named 'mpfr', it will be built 242 together with GCC. Alternatively, if MPFR is already installed but 243 it is not in your default library search path, the '--with-mpfr' 244 configure option should be used. See also '--with-mpfr-lib' and 245 '--with-mpfr-include'. The in-tree build is only supported with 246 the MPFR version that download_prerequisites installs. 247 248MPC Library version 0.8.1 (or later) 249 250 Necessary to build GCC. It can be downloaded from 251 <http://www.multiprecision.org/mpc/>. If an MPC source 252 distribution is found in a subdirectory of your GCC sources named 253 'mpc', it will be built together with GCC. Alternatively, if MPC is 254 already installed but it is not in your default library search 255 path, the '--with-mpc' configure option should be used. See also 256 '--with-mpc-lib' and '--with-mpc-include'. The in-tree build is 257 only supported with the MPC version that download_prerequisites 258 installs. 259 260isl Library version 0.15 or later. 261 262 Necessary to build GCC with the Graphite loop optimizations. It 263 can be downloaded from <ftp://gcc.gnu.org/pub/gcc/infrastructure/>. 264 If an isl source distribution is found in a subdirectory of your 265 GCC sources named 'isl', it will be built together with GCC. 266 Alternatively, the '--with-isl' configure option should be used if 267 isl is not installed in your default library search path. 268 269Tools/packages necessary for modifying GCC 270========================================== 271 272autoconf version 2.69 273GNU m4 version 1.4.6 (or later) 274 275 Necessary when modifying 'configure.ac', 'aclocal.m4', etc. to 276 regenerate 'configure' and 'config.in' files. 277 278automake version 1.15.1 279 280 Necessary when modifying a 'Makefile.am' file to regenerate its 281 associated 'Makefile.in'. 282 283 Much of GCC does not use automake, so directly edit the 284 'Makefile.in' file. Specifically this applies to the 'gcc', 285 'intl', 'libcpp', 'libiberty', 'libobjc' directories as well as any 286 of their subdirectories. 287 288 For directories that use automake, GCC requires the latest release 289 in the 1.15 series, which is currently 1.15.1. When regenerating a 290 directory to a newer version, please update all the directories 291 using an older 1.15 to the latest released version. 292 293gettext version 0.14.5 (or later) 294 295 Needed to regenerate 'gcc.pot'. 296 297gperf version 2.7.2 (or later) 298 299 Necessary when modifying 'gperf' input files, e.g. 300 'gcc/cp/cfns.gperf' to regenerate its associated header file, e.g. 301 'gcc/cp/cfns.h'. 302 303DejaGnu 1.4.4 304Expect 305Tcl 306 307 Necessary to run the GCC testsuite; see the section on testing for 308 details. Tcl 8.6 has a known regression in RE pattern handling 309 that make parts of the testsuite fail. See 310 <http://core.tcl.tk/tcl/tktview/267b7e2334ee2e9de34c4b00d6e72e2f1997085f> 311 for more information. This bug has been fixed in 8.6.1. 312 313autogen version 5.5.4 (or later) and 314guile version 1.4.1 (or later) 315 316 Necessary to regenerate 'fixinc/fixincl.x' from 317 'fixinc/inclhack.def' and 'fixinc/*.tpl'. 318 319 Necessary to run 'make check' for 'fixinc'. 320 321 Necessary to regenerate the top level 'Makefile.in' file from 322 'Makefile.tpl' and 'Makefile.def'. 323 324Flex version 2.5.4 (or later) 325 326 Necessary when modifying '*.l' files. 327 328 Necessary to build GCC during development because the generated 329 output files are not included in the version-controlled source 330 repository. They are included in releases. 331 332Texinfo version 4.7 (or later) 333 334 Necessary for running 'makeinfo' when modifying '*.texi' files to 335 test your changes. 336 337 Necessary for running 'make dvi' or 'make pdf' to create printable 338 documentation in DVI or PDF format. Texinfo version 4.8 or later 339 is required for 'make pdf'. 340 341 Necessary to build GCC documentation during development because the 342 generated output files are not included in the repository. They 343 are included in releases. 344 345TeX (any working version) 346 347 Necessary for running 'texi2dvi' and 'texi2pdf', which are used 348 when running 'make dvi' or 'make pdf' to create DVI or PDF files, 349 respectively. 350 351Sphinx version 1.0 (or later) 352 353 Necessary to regenerate 'jit/docs/_build/texinfo' from the '.rst' 354 files in the directories below 'jit/docs'. 355 356git (any version) 357SSH (any version) 358 359 Necessary to access the source repository. Public releases and 360 weekly snapshots of the development sources are also available via 361 HTTPS. 362 363GNU diffutils version 2.7 (or later) 364 365 Useful when submitting patches for the GCC source code. 366 367patch version 2.5.4 (or later) 368 369 Necessary when applying patches, created with 'diff', to one's own 370 sources. 371 372 373File: gccinstall.info, Node: Downloading the source, Next: Configuration, Prev: Prerequisites, Up: Installing GCC 374 3753 Downloading GCC 376***************** 377 378GCC is distributed via git and via HTTPS as tarballs compressed with 379'gzip' or 'bzip2'. 380 381 Please refer to the releases web page for information on how to 382obtain GCC. 383 384 The source distribution includes the C, C++, Objective-C, Fortran, 385and Ada (in the case of GCC 3.1 and later) compilers, as well as runtime 386libraries for C++, Objective-C, and Fortran. For previous versions 387these were downloadable as separate components such as the core GCC 388distribution, which included the C language front end and shared 389components, and language-specific distributions including the language 390front end and the language runtime (where appropriate). 391 392 If you also intend to build binutils (either to upgrade an existing 393installation or for use in place of the corresponding tools of your OS), 394unpack the binutils distribution either in the same directory or a 395separate one. In the latter case, add symbolic links to any components 396of the binutils you intend to build alongside the compiler ('bfd', 397'binutils', 'gas', 'gprof', 'ld', 'opcodes', ...) to the directory 398containing the GCC sources. 399 400 Likewise the GMP, MPFR and MPC libraries can be automatically built 401together with GCC. You may simply run the 402'contrib/download_prerequisites' script in the GCC source directory to 403set up everything. Otherwise unpack the GMP, MPFR and/or MPC source 404distributions in the directory containing the GCC sources and rename 405their directories to 'gmp', 'mpfr' and 'mpc', respectively (or use 406symbolic links with the same name). 407 408 409File: gccinstall.info, Node: Configuration, Next: Building, Prev: Downloading the source, Up: Installing GCC 410 4114 Installing GCC: Configuration 412******************************* 413 414Like most GNU software, GCC must be configured before it can be built. 415This document describes the recommended configuration procedure for both 416native and cross targets. 417 418 We use SRCDIR to refer to the toplevel source directory for GCC; we 419use OBJDIR to refer to the toplevel build/object directory. 420 421 If you obtained the sources by cloning the repository, SRCDIR must 422refer to the top 'gcc' directory, the one where the 'MAINTAINERS' file 423can be found, and not its 'gcc' subdirectory, otherwise the build will 424fail. 425 426 If either SRCDIR or OBJDIR is located on an automounted NFS file 427system, the shell's built-in 'pwd' command will return temporary 428pathnames. Using these can lead to various sorts of build problems. To 429avoid this issue, set the 'PWDCMD' environment variable to an 430automounter-aware 'pwd' command, e.g., 'pawd' or 'amq -w', during the 431configuration and build phases. 432 433 First, we *highly* recommend that GCC be built into a separate 434directory from the sources which does *not* reside within the source 435tree. This is how we generally build GCC; building where SRCDIR == 436OBJDIR should still work, but doesn't get extensive testing; building 437where OBJDIR is a subdirectory of SRCDIR is unsupported. 438 439 If you have previously built GCC in the same directory for a 440different target machine, do 'make distclean' to delete all files that 441might be invalid. One of the files this deletes is 'Makefile'; if 'make 442distclean' complains that 'Makefile' does not exist or issues a message 443like "don't know how to make distclean" it probably means that the 444directory is already suitably clean. However, with the recommended 445method of building in a separate OBJDIR, you should simply use a 446different OBJDIR for each target. 447 448 Second, when configuring a native system, either 'cc' or 'gcc' must 449be in your path or you must set 'CC' in your environment before running 450configure. Otherwise the configuration scripts may fail. 451 452 To configure GCC: 453 454 % mkdir OBJDIR 455 % cd OBJDIR 456 % SRCDIR/configure [OPTIONS] [TARGET] 457 458Distributor options 459=================== 460 461If you will be distributing binary versions of GCC, with modifications 462to the source code, you should use the options described in this section 463to make clear that your version contains modifications. 464 465'--with-pkgversion=VERSION' 466 Specify a string that identifies your package. You may wish to 467 include a build number or build date. This version string will be 468 included in the output of 'gcc --version'. This suffix does not 469 replace the default version string, only the 'GCC' part. 470 471 The default value is 'GCC'. 472 473'--with-bugurl=URL' 474 Specify the URL that users should visit if they wish to report a 475 bug. You are of course welcome to forward bugs reported to you to 476 the FSF, if you determine that they are not bugs in your 477 modifications. 478 479 The default value refers to the FSF's GCC bug tracker. 480 481Target specification 482==================== 483 484 * GCC has code to correctly determine the correct value for TARGET 485 for nearly all native systems. Therefore, we highly recommend you 486 do not provide a configure target when configuring a native 487 compiler. 488 489 * TARGET must be specified as '--target=TARGET' when configuring a 490 cross compiler; examples of valid targets would be m68k-elf, 491 sh-elf, etc. 492 493 * Specifying just TARGET instead of '--target=TARGET' implies that 494 the host defaults to TARGET. 495 496Options specification 497===================== 498 499Use OPTIONS to override several configure time options for GCC. A list 500of supported OPTIONS follows; 'configure --help' may list other options, 501but those not listed below may not work and should not normally be used. 502 503 Note that each '--enable' option has a corresponding '--disable' 504option and that each '--with' option has a corresponding '--without' 505option. 506 507'--prefix=DIRNAME' 508 Specify the toplevel installation directory. This is the 509 recommended way to install the tools into a directory other than 510 the default. The toplevel installation directory defaults to 511 '/usr/local'. 512 513 We *highly* recommend against DIRNAME being the same or a 514 subdirectory of OBJDIR or vice versa. If specifying a directory 515 beneath a user's home directory tree, some shells will not expand 516 DIRNAME correctly if it contains the '~' metacharacter; use '$HOME' 517 instead. 518 519 The following standard 'autoconf' options are supported. Normally 520 you should not need to use these options. 521 '--exec-prefix=DIRNAME' 522 Specify the toplevel installation directory for 523 architecture-dependent files. The default is 'PREFIX'. 524 525 '--bindir=DIRNAME' 526 Specify the installation directory for the executables called 527 by users (such as 'gcc' and 'g++'). The default is 528 'EXEC-PREFIX/bin'. 529 530 '--libdir=DIRNAME' 531 Specify the installation directory for object code libraries 532 and internal data files of GCC. The default is 533 'EXEC-PREFIX/lib'. 534 535 '--libexecdir=DIRNAME' 536 Specify the installation directory for internal executables of 537 GCC. The default is 'EXEC-PREFIX/libexec'. 538 539 '--with-slibdir=DIRNAME' 540 Specify the installation directory for the shared libgcc 541 library. The default is 'LIBDIR'. 542 543 '--datarootdir=DIRNAME' 544 Specify the root of the directory tree for read-only 545 architecture-independent data files referenced by GCC. The 546 default is 'PREFIX/share'. 547 548 '--infodir=DIRNAME' 549 Specify the installation directory for documentation in info 550 format. The default is 'DATAROOTDIR/info'. 551 552 '--datadir=DIRNAME' 553 Specify the installation directory for some 554 architecture-independent data files referenced by GCC. The 555 default is 'DATAROOTDIR'. 556 557 '--docdir=DIRNAME' 558 Specify the installation directory for documentation files 559 (other than Info) for GCC. The default is 'DATAROOTDIR/doc'. 560 561 '--htmldir=DIRNAME' 562 Specify the installation directory for HTML documentation 563 files. The default is 'DOCDIR'. 564 565 '--pdfdir=DIRNAME' 566 Specify the installation directory for PDF documentation 567 files. The default is 'DOCDIR'. 568 569 '--mandir=DIRNAME' 570 Specify the installation directory for manual pages. The 571 default is 'DATAROOTDIR/man'. (Note that the manual pages are 572 only extracts from the full GCC manuals, which are provided in 573 Texinfo format. The manpages are derived by an automatic 574 conversion process from parts of the full manual.) 575 576 '--with-gxx-include-dir=DIRNAME' 577 Specify the installation directory for G++ header files. The 578 default depends on other configuration options, and differs 579 between cross and native configurations. 580 581 '--with-specs=SPECS' 582 Specify additional command line driver SPECS. This can be 583 useful if you need to turn on a non-standard feature by 584 default without modifying the compiler's source code, for 585 instance 586 '--with-specs=%{!fcommon:%{!fno-common:-fno-common}}'. *Note 587 Specifying subprocesses and the switches to pass to them: 588 (gcc)Spec Files, 589 590'--program-prefix=PREFIX' 591 GCC supports some transformations of the names of its programs when 592 installing them. This option prepends PREFIX to the names of 593 programs to install in BINDIR (see above). For example, specifying 594 '--program-prefix=foo-' would result in 'gcc' being installed as 595 '/usr/local/bin/foo-gcc'. 596 597'--program-suffix=SUFFIX' 598 Appends SUFFIX to the names of programs to install in BINDIR (see 599 above). For example, specifying '--program-suffix=-3.1' would 600 result in 'gcc' being installed as '/usr/local/bin/gcc-3.1'. 601 602'--program-transform-name=PATTERN' 603 Applies the 'sed' script PATTERN to be applied to the names of 604 programs to install in BINDIR (see above). PATTERN has to consist 605 of one or more basic 'sed' editing commands, separated by 606 semicolons. For example, if you want the 'gcc' program name to be 607 transformed to the installed program '/usr/local/bin/myowngcc' and 608 the 'g++' program name to be transformed to 609 '/usr/local/bin/gspecial++' without changing other program names, 610 you could use the pattern 611 '--program-transform-name='s/^gcc$/myowngcc/; s/^g++$/gspecial++/'' 612 to achieve this effect. 613 614 All three options can be combined and used together, resulting in 615 more complex conversion patterns. As a basic rule, PREFIX (and 616 SUFFIX) are prepended (appended) before further transformations can 617 happen with a special transformation script PATTERN. 618 619 As currently implemented, this option only takes effect for native 620 builds; cross compiler binaries' names are not transformed even 621 when a transformation is explicitly asked for by one of these 622 options. 623 624 For native builds, some of the installed programs are also 625 installed with the target alias in front of their name, as in 626 'i686-pc-linux-gnu-gcc'. All of the above transformations happen 627 before the target alias is prepended to the name--so, specifying 628 '--program-prefix=foo-' and 'program-suffix=-3.1', the resulting 629 binary would be installed as 630 '/usr/local/bin/i686-pc-linux-gnu-foo-gcc-3.1'. 631 632 As a last shortcoming, none of the installed Ada programs are 633 transformed yet, which will be fixed in some time. 634 635'--with-local-prefix=DIRNAME' 636 Specify the installation directory for local include files. The 637 default is '/usr/local'. Specify this option if you want the 638 compiler to search directory 'DIRNAME/include' for locally 639 installed header files _instead_ of '/usr/local/include'. 640 641 You should specify '--with-local-prefix' *only* if your site has a 642 different convention (not '/usr/local') for where to put 643 site-specific files. 644 645 The default value for '--with-local-prefix' is '/usr/local' 646 regardless of the value of '--prefix'. Specifying '--prefix' has 647 no effect on which directory GCC searches for local header files. 648 This may seem counterintuitive, but actually it is logical. 649 650 The purpose of '--prefix' is to specify where to _install GCC_. The 651 local header files in '/usr/local/include'--if you put any in that 652 directory--are not part of GCC. They are part of other 653 programs--perhaps many others. (GCC installs its own header files 654 in another directory which is based on the '--prefix' value.) 655 656 Both the local-prefix include directory and the GCC-prefix include 657 directory are part of GCC's "system include" directories. Although 658 these two directories are not fixed, they need to be searched in 659 the proper order for the correct processing of the include_next 660 directive. The local-prefix include directory is searched before 661 the GCC-prefix include directory. Another characteristic of system 662 include directories is that pedantic warnings are turned off for 663 headers in these directories. 664 665 Some autoconf macros add '-I DIRECTORY' options to the compiler 666 command line, to ensure that directories containing installed 667 packages' headers are searched. When DIRECTORY is one of GCC's 668 system include directories, GCC will ignore the option so that 669 system directories continue to be processed in the correct order. 670 This may result in a search order different from what was specified 671 but the directory will still be searched. 672 673 GCC automatically searches for ordinary libraries using 674 'GCC_EXEC_PREFIX'. Thus, when the same installation prefix is used 675 for both GCC and packages, GCC will automatically search for both 676 headers and libraries. This provides a configuration that is easy 677 to use. GCC behaves in a manner similar to that when it is 678 installed as a system compiler in '/usr'. 679 680 Sites that need to install multiple versions of GCC may not want to 681 use the above simple configuration. It is possible to use the 682 '--program-prefix', '--program-suffix' and 683 '--program-transform-name' options to install multiple versions 684 into a single directory, but it may be simpler to use different 685 prefixes and the '--with-local-prefix' option to specify the 686 location of the site-specific files for each version. It will then 687 be necessary for users to specify explicitly the location of local 688 site libraries (e.g., with 'LIBRARY_PATH'). 689 690 The same value can be used for both '--with-local-prefix' and 691 '--prefix' provided it is not '/usr'. This can be used to avoid 692 the default search of '/usr/local/include'. 693 694 *Do not* specify '/usr' as the '--with-local-prefix'! The 695 directory you use for '--with-local-prefix' *must not* contain any 696 of the system's standard header files. If it did contain them, 697 certain programs would be miscompiled (including GNU Emacs, on 698 certain targets), because this would override and nullify the 699 header file corrections made by the 'fixincludes' script. 700 701 Indications are that people who use this option use it based on 702 mistaken ideas of what it is for. People use it as if it specified 703 where to install part of GCC. Perhaps they make this assumption 704 because installing GCC creates the directory. 705 706'--with-gcc-major-version-only' 707 Specifies that GCC should use only the major number rather than 708 MAJOR.MINOR.PATCHLEVEL in filesystem paths. 709 710'--with-native-system-header-dir=DIRNAME' 711 Specifies that DIRNAME is the directory that contains native system 712 header files, rather than '/usr/include'. This option is most 713 useful if you are creating a compiler that should be isolated from 714 the system as much as possible. It is most commonly used with the 715 '--with-sysroot' option and will cause GCC to search DIRNAME inside 716 the system root specified by that option. 717 718'--enable-shared[=PACKAGE[,...]]' 719 Build shared versions of libraries, if shared libraries are 720 supported on the target platform. Unlike GCC 2.95.x and earlier, 721 shared libraries are enabled by default on all platforms that 722 support shared libraries. 723 724 If a list of packages is given as an argument, build shared 725 libraries only for the listed packages. For other packages, only 726 static libraries will be built. Package names currently recognized 727 in the GCC tree are 'libgcc' (also known as 'gcc'), 'libstdc++' 728 (not 'libstdc++-v3'), 'libffi', 'zlib', 'boehm-gc', 'ada', 729 'libada', 'libgo', 'libobjc', and 'libphobos'. Note 'libiberty' 730 does not support shared libraries at all. 731 732 Use '--disable-shared' to build only static libraries. Note that 733 '--disable-shared' does not accept a list of package names as 734 argument, only '--enable-shared' does. 735 736 Contrast with '--enable-host-shared', which affects _host_ code. 737 738'--enable-host-shared' 739 Specify that the _host_ code should be built into 740 position-independent machine code (with -fPIC), allowing it to be 741 used within shared libraries, but yielding a slightly slower 742 compiler. 743 744 This option is required when building the libgccjit.so library. 745 746 Contrast with '--enable-shared', which affects _target_ libraries. 747 748'--with-gnu-as' 749 Specify that the compiler should assume that the assembler it finds 750 is the GNU assembler. However, this does not modify the rules to 751 find an assembler and will result in confusion if the assembler 752 found is not actually the GNU assembler. (Confusion may also 753 result if the compiler finds the GNU assembler but has not been 754 configured with '--with-gnu-as'.) If you have more than one 755 assembler installed on your system, you may want to use this option 756 in connection with '--with-as=PATHNAME' or 757 '--with-build-time-tools=PATHNAME'. 758 759 The following systems are the only ones where it makes a difference 760 whether you use the GNU assembler. On any other system, 761 '--with-gnu-as' has no effect. 762 763 * 'hppa1.0-ANY-ANY' 764 * 'hppa1.1-ANY-ANY' 765 * 'sparc-sun-solaris2.ANY' 766 * 'sparc64-ANY-solaris2.ANY' 767 768'--with-as=PATHNAME' 769 Specify that the compiler should use the assembler pointed to by 770 PATHNAME, rather than the one found by the standard rules to find 771 an assembler, which are: 772 * Unless GCC is being built with a cross compiler, check the 773 'LIBEXEC/gcc/TARGET/VERSION' directory. LIBEXEC defaults to 774 'EXEC-PREFIX/libexec'; EXEC-PREFIX defaults to PREFIX, which 775 defaults to '/usr/local' unless overridden by the 776 '--prefix=PATHNAME' switch described above. TARGET is the 777 target system triple, such as 'sparc-sun-solaris2.7', and 778 VERSION denotes the GCC version, such as 3.0. 779 780 * If the target system is the same that you are building on, 781 check operating system specific directories (e.g. 782 '/usr/ccs/bin' on Sun Solaris 2). 783 784 * Check in the 'PATH' for a tool whose name is prefixed by the 785 target system triple. 786 787 * Check in the 'PATH' for a tool whose name is not prefixed by 788 the target system triple, if the host and target system triple 789 are the same (in other words, we use a host tool if it can be 790 used for the target as well). 791 792 You may want to use '--with-as' if no assembler is installed in the 793 directories listed above, or if you have multiple assemblers 794 installed and want to choose one that is not found by the above 795 rules. 796 797'--with-gnu-ld' 798 Same as '--with-gnu-as' but for the linker. 799 800'--with-ld=PATHNAME' 801 Same as '--with-as' but for the linker. 802 803'--with-stabs' 804 Specify that stabs debugging information should be used instead of 805 whatever format the host normally uses. Normally GCC uses the same 806 debug format as the host system. 807 808'--with-tls=DIALECT' 809 Specify the default TLS dialect, for systems were there is a 810 choice. For ARM targets, possible values for DIALECT are 'gnu' or 811 'gnu2', which select between the original GNU dialect and the GNU 812 TLS descriptor-based dialect. 813 814'--enable-multiarch' 815 Specify whether to enable or disable multiarch support. The 816 default is to check for glibc start files in a multiarch location, 817 and enable it if the files are found. The auto detection is 818 enabled for native builds, and for cross builds configured with 819 '--with-sysroot', and without '--with-native-system-header-dir'. 820 More documentation about multiarch can be found at 821 <https://wiki.debian.org/Multiarch>. 822 823'--enable-sjlj-exceptions' 824 Force use of the 'setjmp'/'longjmp'-based scheme for exceptions. 825 'configure' ordinarily picks the correct value based on the 826 platform. Only use this option if you are sure you need a 827 different setting. 828 829'--enable-vtable-verify' 830 Specify whether to enable or disable the vtable verification 831 feature. Enabling this feature causes libstdc++ to be built with 832 its virtual calls in verifiable mode. This means that, when linked 833 with libvtv, every virtual call in libstdc++ will verify the vtable 834 pointer through which the call will be made before actually making 835 the call. If not linked with libvtv, the verifier will call stub 836 functions (in libstdc++ itself) and do nothing. If vtable 837 verification is disabled, then libstdc++ is not built with its 838 virtual calls in verifiable mode at all. However the libvtv 839 library will still be built (see '--disable-libvtv' to turn off 840 building libvtv). '--disable-vtable-verify' is the default. 841 842'--disable-gcov' 843 Specify that the run-time library used for coverage analysis and 844 associated host tools should not be built. 845 846'--disable-multilib' 847 Specify that multiple target libraries to support different target 848 variants, calling conventions, etc. should not be built. The 849 default is to build a predefined set of them. 850 851 Some targets provide finer-grained control over which multilibs are 852 built (e.g., '--disable-softfloat'): 853 'arm-*-*' 854 fpu, 26bit, underscore, interwork, biendian, nofmult. 855 856 'm68*-*-*' 857 softfloat, m68881, m68000, m68020. 858 859 'mips*-*-*' 860 single-float, biendian, softfloat. 861 862 'powerpc*-*-*, rs6000*-*-*' 863 aix64, pthread, softfloat, powercpu, powerpccpu, powerpcos, 864 biendian, sysv, aix. 865 866'--with-multilib-list=LIST' 867'--without-multilib-list' 868 Specify what multilibs to build. LIST is a comma separated list of 869 values, possibly consisting of a single value. Currently only 870 implemented for aarch64*-*-*, arm*-*-*, riscv*-*-*, sh*-*-* and 871 x86-64-*-linux*. The accepted values and meaning for each target 872 is given below. 873 874 'aarch64*-*-*' 875 LIST is a comma separated list of 'ilp32', and 'lp64' to 876 enable ILP32 and LP64 run-time libraries, respectively. If 877 LIST is empty, then there will be no multilibs and only the 878 default run-time library will be built. If LIST is 'default' 879 or -with-multilib-list= is not specified, then the default set 880 of libraries is selected based on the value of '--target'. 881 882 'arm*-*-*' 883 LIST is a comma separated list of 'aprofile' and 'rmprofile' 884 to build multilibs for A or R and M architecture profiles 885 respectively. Note that, due to some limitation of the 886 current multilib framework, using the combined 887 'aprofile,rmprofile' multilibs selects in some cases a less 888 optimal multilib than when using the multilib profile for the 889 architecture targetted. The special value 'default' is also 890 accepted and is equivalent to omitting the option, i.e., only 891 the default run-time library will be enabled. 892 893 LIST may instead contain '@name', to use the multilib 894 configuration Makefile fragment 'name' in 'gcc/config/arm' in 895 the source tree (it is part of the corresponding sources, 896 after all). It is recommended, but not required, that files 897 used for this purpose to be named starting with 't-ml-', to 898 make their intended purpose self-evident, in line with GCC 899 conventions. Such files enable custom, user-chosen multilib 900 lists to be configured. Whether multiple such files can be 901 used together depends on the contents of the supplied files. 902 See 'gcc/config/arm/t-multilib' and its supplementary 903 'gcc/config/arm/t-*profile' files for an example of what such 904 Makefile fragments might look like for this version of GCC. 905 The macros expected to be defined in these fragments are not 906 stable across GCC releases, so make sure they define the 907 'MULTILIB'-related macros expected by the version of GCC you 908 are building. *Note Target Makefile Fragments: (gccint)Target 909 Fragment. 910 911 The table below gives the combination of ISAs, architectures, 912 FPUs and floating-point ABIs for which multilibs are built for 913 each predefined profile. The union of these options is 914 considered when specifying both 'aprofile' and 'rmprofile'. 915 916 Option aprofile rmprofile 917 ISAs '-marm' and '-mthumb' 918 '-mthumb' 919 Architecturesdefault default architecture 920 architecture '-march=armv6s-m' 921 '-march=armv7-a' '-march=armv7-m' 922 '-march=armv7ve' '-march=armv7e-m' 923 '-march=armv8-a' '-march=armv8-m.base' 924 '-march=armv8-m.main' 925 '-march=armv7' 926 FPUs none none 927 '-mfpu=vfpv3-d16' '-mfpu=vfpv3-d16' 928 '-mfpu=neon' '-mfpu=fpv4-sp-d16' 929 '-mfpu=vfpv4-d16' '-mfpu=fpv5-sp-d16' 930 '-mfpu=neon-vfpv4' '-mfpu=fpv5-d16' 931 '-mfpu=neon-fp-armv8' 932 floating-point'-mfloat-abi=soft' '-mfloat-abi=soft' 933 ABIs '-mfloat-abi=softfp' '-mfloat-abi=softfp' 934 '-mfloat-abi=hard' '-mfloat-abi=hard' 935 936 'riscv*-*-*' 937 LIST is a single ABI name. The target architecture must be 938 either 'rv32gc' or 'rv64gc'. This will build a single 939 multilib for the specified architecture and ABI pair. If 940 '--with-multilib-list' is not given, then a default set of 941 multilibs is selected based on the value of '--target'. This 942 is usually a large set of multilibs. 943 944 'sh*-*-*' 945 LIST is a comma separated list of CPU names. These must be of 946 the form 'sh*' or 'm*' (in which case they match the compiler 947 option for that processor). The list should not contain any 948 endian options - these are handled by '--with-endian'. 949 950 If LIST is empty, then there will be no multilibs for extra 951 processors. The multilib for the secondary endian remains 952 enabled. 953 954 As a special case, if an entry in the list starts with a '!' 955 (exclamation point), then it is added to the list of excluded 956 multilibs. Entries of this sort should be compatible with 957 'MULTILIB_EXCLUDES' (once the leading '!' has been stripped). 958 959 If '--with-multilib-list' is not given, then a default set of 960 multilibs is selected based on the value of '--target'. This 961 is usually the complete set of libraries, but some targets 962 imply a more specialized subset. 963 964 Example 1: to configure a compiler for SH4A only, but 965 supporting both endians, with little endian being the default: 966 --with-cpu=sh4a --with-endian=little,big --with-multilib-list= 967 968 Example 2: to configure a compiler for both SH4A and 969 SH4AL-DSP, but with only little endian SH4AL: 970 --with-cpu=sh4a --with-endian=little,big \ 971 --with-multilib-list=sh4al,!mb/m4al 972 973 'x86-64-*-linux*' 974 LIST is a comma separated list of 'm32', 'm64' and 'mx32' to 975 enable 32-bit, 64-bit and x32 run-time libraries, 976 respectively. If LIST is empty, then there will be no 977 multilibs and only the default run-time library will be 978 enabled. 979 980 If '--with-multilib-list' is not given, then only 32-bit and 981 64-bit run-time libraries will be enabled. 982 983'--with-endian=ENDIANS' 984 Specify what endians to use. Currently only implemented for 985 sh*-*-*. 986 987 ENDIANS may be one of the following: 988 'big' 989 Use big endian exclusively. 990 'little' 991 Use little endian exclusively. 992 'big,little' 993 Use big endian by default. Provide a multilib for little 994 endian. 995 'little,big' 996 Use little endian by default. Provide a multilib for big 997 endian. 998 999'--enable-threads' 1000 Specify that the target supports threads. This affects the 1001 Objective-C compiler and runtime library, and exception handling 1002 for other languages like C++. On some systems, this is the 1003 default. 1004 1005 In general, the best (and, in many cases, the only known) threading 1006 model available will be configured for use. Beware that on some 1007 systems, GCC has not been taught what threading models are 1008 generally available for the system. In this case, 1009 '--enable-threads' is an alias for '--enable-threads=single'. 1010 1011'--disable-threads' 1012 Specify that threading support should be disabled for the system. 1013 This is an alias for '--enable-threads=single'. 1014 1015'--enable-threads=LIB' 1016 Specify that LIB is the thread support library. This affects the 1017 Objective-C compiler and runtime library, and exception handling 1018 for other languages like C++. The possibilities for LIB are: 1019 1020 'aix' 1021 AIX thread support. 1022 'dce' 1023 DCE thread support. 1024 'lynx' 1025 LynxOS thread support. 1026 'mipssde' 1027 MIPS SDE thread support. 1028 'no' 1029 This is an alias for 'single'. 1030 'posix' 1031 Generic POSIX/Unix98 thread support. 1032 'rtems' 1033 RTEMS thread support. 1034 'single' 1035 Disable thread support, should work for all platforms. 1036 'tpf' 1037 TPF thread support. 1038 'vxworks' 1039 VxWorks thread support. 1040 'win32' 1041 Microsoft Win32 API thread support. 1042 1043'--enable-tls' 1044 Specify that the target supports TLS (Thread Local Storage). 1045 Usually configure can correctly determine if TLS is supported. In 1046 cases where it guesses incorrectly, TLS can be explicitly enabled 1047 or disabled with '--enable-tls' or '--disable-tls'. This can 1048 happen if the assembler supports TLS but the C library does not, or 1049 if the assumptions made by the configure test are incorrect. 1050 1051'--disable-tls' 1052 Specify that the target does not support TLS. This is an alias for 1053 '--enable-tls=no'. 1054 1055'--with-cpu=CPU' 1056'--with-cpu-32=CPU' 1057'--with-cpu-64=CPU' 1058 Specify which cpu variant the compiler should generate code for by 1059 default. CPU will be used as the default value of the '-mcpu=' 1060 switch. This option is only supported on some targets, including 1061 ARC, ARM, i386, M68k, PowerPC, and SPARC. It is mandatory for ARC. 1062 The '--with-cpu-32' and '--with-cpu-64' options specify separate 1063 default CPUs for 32-bit and 64-bit modes; these options are only 1064 supported for i386, x86-64, PowerPC, and SPARC. 1065 1066'--with-schedule=CPU' 1067'--with-arch=CPU' 1068'--with-arch-32=CPU' 1069'--with-arch-64=CPU' 1070'--with-tune=CPU' 1071'--with-tune-32=CPU' 1072'--with-tune-64=CPU' 1073'--with-abi=ABI' 1074'--with-fpu=TYPE' 1075'--with-float=TYPE' 1076 These configure options provide default values for the 1077 '-mschedule=', '-march=', '-mtune=', '-mabi=', and '-mfpu=' options 1078 and for '-mhard-float' or '-msoft-float'. As with '--with-cpu', 1079 which switches will be accepted and acceptable values of the 1080 arguments depend on the target. 1081 1082'--with-mode=MODE' 1083 Specify if the compiler should default to '-marm' or '-mthumb'. 1084 This option is only supported on ARM targets. 1085 1086'--with-stack-offset=NUM' 1087 This option sets the default for the -mstack-offset=NUM option, and 1088 will thus generally also control the setting of this option for 1089 libraries. This option is only supported on Epiphany targets. 1090 1091'--with-fpmath=ISA' 1092 This options sets '-mfpmath=sse' by default and specifies the 1093 default ISA for floating-point arithmetics. You can select either 1094 'sse' which enables '-msse2' or 'avx' which enables '-mavx' by 1095 default. This option is only supported on i386 and x86-64 targets. 1096 1097'--with-fp-32=MODE' 1098 On MIPS targets, set the default value for the '-mfp' option when 1099 using the o32 ABI. The possibilities for MODE are: 1100 '32' 1101 Use the o32 FP32 ABI extension, as with the '-mfp32' 1102 command-line option. 1103 'xx' 1104 Use the o32 FPXX ABI extension, as with the '-mfpxx' 1105 command-line option. 1106 '64' 1107 Use the o32 FP64 ABI extension, as with the '-mfp64' 1108 command-line option. 1109 In the absence of this configuration option the default is to use 1110 the o32 FP32 ABI extension. 1111 1112'--with-odd-spreg-32' 1113 On MIPS targets, set the '-modd-spreg' option by default when using 1114 the o32 ABI. 1115 1116'--without-odd-spreg-32' 1117 On MIPS targets, set the '-mno-odd-spreg' option by default when 1118 using the o32 ABI. This is normally used in conjunction with 1119 '--with-fp-32=64' in order to target the o32 FP64A ABI extension. 1120 1121'--with-nan=ENCODING' 1122 On MIPS targets, set the default encoding convention to use for the 1123 special not-a-number (NaN) IEEE 754 floating-point data. The 1124 possibilities for ENCODING are: 1125 'legacy' 1126 Use the legacy encoding, as with the '-mnan=legacy' 1127 command-line option. 1128 '2008' 1129 Use the 754-2008 encoding, as with the '-mnan=2008' 1130 command-line option. 1131 To use this configuration option you must have an assembler version 1132 installed that supports the '-mnan=' command-line option too. In 1133 the absence of this configuration option the default convention is 1134 the legacy encoding, as when neither of the '-mnan=2008' and 1135 '-mnan=legacy' command-line options has been used. 1136 1137'--with-divide=TYPE' 1138 Specify how the compiler should generate code for checking for 1139 division by zero. This option is only supported on the MIPS 1140 target. The possibilities for TYPE are: 1141 'traps' 1142 Division by zero checks use conditional traps (this is the 1143 default on systems that support conditional traps). 1144 'breaks' 1145 Division by zero checks use the break instruction. 1146 1147'--with-llsc' 1148 On MIPS targets, make '-mllsc' the default when no '-mno-llsc' 1149 option is passed. This is the default for Linux-based targets, as 1150 the kernel will emulate them if the ISA does not provide them. 1151 1152'--without-llsc' 1153 On MIPS targets, make '-mno-llsc' the default when no '-mllsc' 1154 option is passed. 1155 1156'--with-synci' 1157 On MIPS targets, make '-msynci' the default when no '-mno-synci' 1158 option is passed. 1159 1160'--without-synci' 1161 On MIPS targets, make '-mno-synci' the default when no '-msynci' 1162 option is passed. This is the default. 1163 1164'--with-lxc1-sxc1' 1165 On MIPS targets, make '-mlxc1-sxc1' the default when no 1166 '-mno-lxc1-sxc1' option is passed. This is the default. 1167 1168'--without-lxc1-sxc1' 1169 On MIPS targets, make '-mno-lxc1-sxc1' the default when no 1170 '-mlxc1-sxc1' option is passed. The indexed load/store 1171 instructions are not directly a problem but can lead to unexpected 1172 behaviour when deployed in an application intended for a 32-bit 1173 address space but run on a 64-bit processor. The issue is seen 1174 because all known MIPS 64-bit Linux kernels execute o32 and n32 1175 applications with 64-bit addressing enabled which affects the 1176 overflow behaviour of the indexed addressing mode. GCC will assume 1177 that ordinary 32-bit arithmetic overflow behaviour is the same 1178 whether performed as an 'addu' instruction or as part of the 1179 address calculation in 'lwxc1' type instructions. This assumption 1180 holds true in a pure 32-bit environment and can hold true in a 1181 64-bit environment if the address space is accurately set to be 1182 32-bit for o32 and n32. 1183 1184'--with-madd4' 1185 On MIPS targets, make '-mmadd4' the default when no '-mno-madd4' 1186 option is passed. This is the default. 1187 1188'--without-madd4' 1189 On MIPS targets, make '-mno-madd4' the default when no '-mmadd4' 1190 option is passed. The 'madd4' instruction family can be 1191 problematic when targeting a combination of cores that implement 1192 these instructions differently. There are two known cores that 1193 implement these as fused operations instead of unfused (where 1194 unfused is normally expected). Disabling these instructions is the 1195 only way to ensure compatible code is generated; this will incur a 1196 performance penalty. 1197 1198'--with-mips-plt' 1199 On MIPS targets, make use of copy relocations and PLTs. These 1200 features are extensions to the traditional SVR4-based MIPS ABIs and 1201 require support from GNU binutils and the runtime C library. 1202 1203'--with-stack-clash-protection-guard-size=SIZE' 1204 On certain targets this option sets the default stack clash 1205 protection guard size as a power of two in bytes. On AArch64 SIZE 1206 is required to be either 12 (4KB) or 16 (64KB). 1207 1208'--enable-__cxa_atexit' 1209 Define if you want to use __cxa_atexit, rather than atexit, to 1210 register C++ destructors for local statics and global objects. 1211 This is essential for fully standards-compliant handling of 1212 destructors, but requires __cxa_atexit in libc. This option is 1213 currently only available on systems with GNU libc. When enabled, 1214 this will cause '-fuse-cxa-atexit' to be passed by default. 1215 1216'--enable-gnu-indirect-function' 1217 Define if you want to enable the 'ifunc' attribute. This option is 1218 currently only available on systems with GNU libc on certain 1219 targets. 1220 1221'--enable-target-optspace' 1222 Specify that target libraries should be optimized for code space 1223 instead of code speed. This is the default for the m32r platform. 1224 1225'--with-cpp-install-dir=DIRNAME' 1226 Specify that the user visible 'cpp' program should be installed in 1227 'PREFIX/DIRNAME/cpp', in addition to BINDIR. 1228 1229'--enable-comdat' 1230 Enable COMDAT group support. This is primarily used to override 1231 the automatically detected value. 1232 1233'--enable-initfini-array' 1234 Force the use of sections '.init_array' and '.fini_array' (instead 1235 of '.init' and '.fini') for constructors and destructors. Option 1236 '--disable-initfini-array' has the opposite effect. If neither 1237 option is specified, the configure script will try to guess whether 1238 the '.init_array' and '.fini_array' sections are supported and, if 1239 they are, use them. 1240 1241'--enable-link-mutex' 1242 When building GCC, use a mutex to avoid linking the compilers for 1243 multiple languages at the same time, to avoid thrashing on build 1244 systems with limited free memory. The default is not to use such a 1245 mutex. 1246 1247'--enable-maintainer-mode' 1248 The build rules that regenerate the Autoconf and Automake output 1249 files as well as the GCC master message catalog 'gcc.pot' are 1250 normally disabled. This is because it can only be rebuilt if the 1251 complete source tree is present. If you have changed the sources 1252 and want to rebuild the catalog, configuring with 1253 '--enable-maintainer-mode' will enable this. Note that you need a 1254 recent version of the 'gettext' tools to do so. 1255 1256'--disable-bootstrap' 1257 For a native build, the default configuration is to perform a 1258 3-stage bootstrap of the compiler when 'make' is invoked, testing 1259 that GCC can compile itself correctly. If you want to disable this 1260 process, you can configure with '--disable-bootstrap'. 1261 1262'--enable-bootstrap' 1263 In special cases, you may want to perform a 3-stage build even if 1264 the target and host triplets are different. This is possible when 1265 the host can run code compiled for the target (e.g. host is 1266 i686-linux, target is i486-linux). Starting from GCC 4.2, to do 1267 this you have to configure explicitly with '--enable-bootstrap'. 1268 1269'--enable-generated-files-in-srcdir' 1270 Neither the .c and .h files that are generated from Bison and flex 1271 nor the info manuals and man pages that are built from the .texi 1272 files are present in the repository development tree. When 1273 building GCC from that development tree, or from one of our 1274 snapshots, those generated files are placed in your build 1275 directory, which allows for the source to be in a readonly 1276 directory. 1277 1278 If you configure with '--enable-generated-files-in-srcdir' then 1279 those generated files will go into the source directory. This is 1280 mainly intended for generating release or prerelease tarballs of 1281 the GCC sources, since it is not a requirement that the users of 1282 source releases to have flex, Bison, or makeinfo. 1283 1284'--enable-version-specific-runtime-libs' 1285 Specify that runtime libraries should be installed in the compiler 1286 specific subdirectory ('LIBDIR/gcc') rather than the usual places. 1287 In addition, 'libstdc++''s include files will be installed into 1288 'LIBDIR' unless you overruled it by using 1289 '--with-gxx-include-dir=DIRNAME'. Using this option is 1290 particularly useful if you intend to use several versions of GCC in 1291 parallel. This is currently supported by 'libgfortran', 1292 'libstdc++', and 'libobjc'. 1293 1294'--with-aix-soname='aix', 'svr4' or 'both'' 1295 Traditional AIX shared library versioning (versioned 'Shared 1296 Object' files as members of unversioned 'Archive Library' files 1297 named 'lib.a') causes numerous headaches for package managers. 1298 However, 'Import Files' as members of 'Archive Library' files allow 1299 for *filename-based versioning* of shared libraries as seen on 1300 Linux/SVR4, where this is called the "SONAME". But as they prevent 1301 static linking, 'Import Files' may be used with 'Runtime Linking' 1302 only, where the linker does search for 'libNAME.so' before 1303 'libNAME.a' library filenames with the '-lNAME' linker flag. 1304 1305 For detailed information please refer to the AIX ld Command 1306 reference. 1307 1308 As long as shared library creation is enabled, upon: 1309 '--with-aix-soname=aix' 1310 '--with-aix-soname=both' 1311 A (traditional AIX) 'Shared Archive Library' file is created: 1312 * using the 'libNAME.a' filename scheme 1313 * with the 'Shared Object' file as archive member named 1314 'libNAME.so.V' (except for 'libgcc_s', where the 'Shared 1315 Object' file is named 'shr.o' for backwards 1316 compatibility), which 1317 - is used for runtime loading from inside the 1318 'libNAME.a' file 1319 - is used for dynamic loading via 1320 'dlopen("libNAME.a(libNAME.so.V)", RTLD_MEMBER)' 1321 - is used for shared linking 1322 - is used for static linking, so no separate 'Static 1323 Archive Library' file is needed 1324 '--with-aix-soname=both' 1325 '--with-aix-soname=svr4' 1326 A (second) 'Shared Archive Library' file is created: 1327 * using the 'libNAME.so.V' filename scheme 1328 * with the 'Shared Object' file as archive member named 1329 'shr.o', which 1330 - is created with the '-G linker flag' 1331 - has the 'F_LOADONLY' flag set 1332 - is used for runtime loading from inside the 1333 'libNAME.so.V' file 1334 - is used for dynamic loading via 1335 'dlopen("libNAME.so.V(shr.o)", RTLD_MEMBER)' 1336 * with the 'Import File' as archive member named 'shr.imp', 1337 which 1338 - refers to 'libNAME.so.V(shr.o)' as the "SONAME", to 1339 be recorded in the 'Loader Section' of subsequent 1340 binaries 1341 - indicates whether 'libNAME.so.V(shr.o)' is 32 or 64 1342 bit 1343 - lists all the public symbols exported by 1344 'lib.so.V(shr.o)', eventually decorated with the 1345 ''weak' Keyword' 1346 - is necessary for shared linking against 1347 'lib.so.V(shr.o)' 1348 A symbolic link using the 'libNAME.so' filename scheme is 1349 created: 1350 * pointing to the 'libNAME.so.V' 'Shared Archive Library' 1351 file 1352 * to permit the 'ld Command' to find 'lib.so.V(shr.imp)' 1353 via the '-lNAME' argument (requires 'Runtime Linking' to 1354 be enabled) 1355 * to permit dynamic loading of 'lib.so.V(shr.o)' without 1356 the need to specify the version number via 1357 'dlopen("libNAME.so(shr.o)", RTLD_MEMBER)' 1358 1359 As long as static library creation is enabled, upon: 1360 '--with-aix-soname=svr4' 1361 A 'Static Archive Library' is created: 1362 * using the 'libNAME.a' filename scheme 1363 * with all the 'Static Object' files as archive members, 1364 which 1365 - are used for static linking 1366 1367 While the aix-soname='svr4' option does not create 'Shared Object' 1368 files as members of unversioned 'Archive Library' files any more, 1369 package managers still are responsible to transfer 'Shared Object' 1370 files found as member of a previously installed unversioned 1371 'Archive Library' file into the newly installed 'Archive Library' 1372 file with the same filename. 1373 1374 _WARNING:_ Creating 'Shared Object' files with 'Runtime Linking' 1375 enabled may bloat the TOC, eventually leading to 'TOC overflow' 1376 errors, requiring the use of either the '-Wl,-bbigtoc' linker flag 1377 (seen to break with the 'GDB' debugger) or some of the TOC-related 1378 compiler flags, *Note RS/6000 and PowerPC Options: (gcc)RS/6000 and 1379 PowerPC Options. 1380 1381 '--with-aix-soname' is currently supported by 'libgcc_s' only, so 1382 this option is still experimental and not for normal use yet. 1383 1384 Default is the traditional behavior '--with-aix-soname='aix''. 1385 1386'--enable-languages=LANG1,LANG2,...' 1387 Specify that only a particular subset of compilers and their 1388 runtime libraries should be built. For a list of valid values for 1389 LANGN you can issue the following command in the 'gcc' directory of 1390 your GCC source tree: 1391 grep ^language= */config-lang.in 1392 Currently, you can use any of the following: 'all', 'default', 1393 'ada', 'c', 'c++', 'd', 'fortran', 'go', 'jit', 'lto', 'objc', 1394 'obj-c++'. Building the Ada compiler has special requirements, see 1395 below. If you do not pass this flag, or specify the option 1396 'default', then the default languages available in the 'gcc' 1397 sub-tree will be configured. Ada, D, Go, Jit, and Objective-C++ 1398 are not default languages. LTO is not a default language, but is 1399 built by default because '--enable-lto' is enabled by default. The 1400 other languages are default languages. If 'all' is specified, then 1401 all available languages are built. An exception is 'jit' language, 1402 which requires '--enable-host-shared' to be included with 'all'. 1403 1404'--enable-stage1-languages=LANG1,LANG2,...' 1405 Specify that a particular subset of compilers and their runtime 1406 libraries should be built with the system C compiler during stage 1 1407 of the bootstrap process, rather than only in later stages with the 1408 bootstrapped C compiler. The list of valid values is the same as 1409 for '--enable-languages', and the option 'all' will select all of 1410 the languages enabled by '--enable-languages'. This option is 1411 primarily useful for GCC development; for instance, when a 1412 development version of the compiler cannot bootstrap due to 1413 compiler bugs, or when one is debugging front ends other than the C 1414 front end. When this option is used, one can then build the target 1415 libraries for the specified languages with the stage-1 compiler by 1416 using 'make stage1-bubble all-target', or run the testsuite on the 1417 stage-1 compiler for the specified languages using 'make 1418 stage1-start check-gcc'. 1419 1420'--disable-libada' 1421 Specify that the run-time libraries and tools used by GNAT should 1422 not be built. This can be useful for debugging, or for 1423 compatibility with previous Ada build procedures, when it was 1424 required to explicitly do a 'make -C gcc gnatlib_and_tools'. 1425 1426'--disable-libsanitizer' 1427 Specify that the run-time libraries for the various sanitizers 1428 should not be built. 1429 1430'--disable-libssp' 1431 Specify that the run-time libraries for stack smashing protection 1432 should not be built or linked against. On many targets library 1433 support is provided by the C library instead. 1434 1435'--disable-libquadmath' 1436 Specify that the GCC quad-precision math library should not be 1437 built. On some systems, the library is required to be linkable 1438 when building the Fortran front end, unless 1439 '--disable-libquadmath-support' is used. 1440 1441'--disable-libquadmath-support' 1442 Specify that the Fortran front end and 'libgfortran' do not add 1443 support for 'libquadmath' on systems supporting it. 1444 1445'--disable-libgomp' 1446 Specify that the GNU Offloading and Multi Processing Runtime 1447 Library should not be built. 1448 1449'--disable-libvtv' 1450 Specify that the run-time libraries used by vtable verification 1451 should not be built. 1452 1453'--with-dwarf2' 1454 Specify that the compiler should use DWARF 2 debugging information 1455 as the default. 1456 1457'--with-advance-toolchain=AT' 1458 On 64-bit PowerPC Linux systems, configure the compiler to use the 1459 header files, library files, and the dynamic linker from the 1460 Advance Toolchain release AT instead of the default versions that 1461 are provided by the Linux distribution. In general, this option is 1462 intended for the developers of GCC, and it is not intended for 1463 general use. 1464 1465'--enable-targets=all' 1466'--enable-targets=TARGET_LIST' 1467 Some GCC targets, e.g. powerpc64-linux, build bi-arch compilers. 1468 These are compilers that are able to generate either 64-bit or 1469 32-bit code. Typically, the corresponding 32-bit target, e.g. 1470 powerpc-linux for powerpc64-linux, only generates 32-bit code. 1471 This option enables the 32-bit target to be a bi-arch compiler, 1472 which is useful when you want a bi-arch compiler that defaults to 1473 32-bit, and you are building a bi-arch or multi-arch binutils in a 1474 combined tree. On mips-linux, this will build a tri-arch compiler 1475 (ABI o32/n32/64), defaulted to o32. Currently, this option only 1476 affects sparc-linux, powerpc-linux, x86-linux, mips-linux and 1477 s390-linux. 1478 1479'--enable-default-pie' 1480 Turn on '-fPIE' and '-pie' by default. 1481 1482'--enable-secureplt' 1483 This option enables '-msecure-plt' by default for powerpc-linux. 1484 *Note RS/6000 and PowerPC Options: (gcc)RS/6000 and PowerPC 1485 Options, 1486 1487'--enable-default-ssp' 1488 Turn on '-fstack-protector-strong' by default. 1489 1490'--enable-cld' 1491 This option enables '-mcld' by default for 32-bit x86 targets. 1492 *Note i386 and x86-64 Options: (gcc)i386 and x86-64 Options, 1493 1494'--enable-large-address-aware' 1495 The '--enable-large-address-aware' option arranges for MinGW 1496 executables to be linked using the '--large-address-aware' option, 1497 that enables the use of more than 2GB of memory. If GCC is 1498 configured with this option, its effects can be reversed by passing 1499 the '-Wl,--disable-large-address-aware' option to the so-configured 1500 compiler driver. 1501 1502'--enable-win32-registry' 1503'--enable-win32-registry=KEY' 1504'--disable-win32-registry' 1505 The '--enable-win32-registry' option enables Microsoft 1506 Windows-hosted GCC to look up installations paths in the registry 1507 using the following key: 1508 1509 HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\KEY 1510 1511 KEY defaults to GCC version number, and can be overridden by the 1512 '--enable-win32-registry=KEY' option. Vendors and distributors who 1513 use custom installers are encouraged to provide a different key, 1514 perhaps one comprised of vendor name and GCC version number, to 1515 avoid conflict with existing installations. This feature is 1516 enabled by default, and can be disabled by 1517 '--disable-win32-registry' option. This option has no effect on 1518 the other hosts. 1519 1520'--nfp' 1521 Specify that the machine does not have a floating point unit. This 1522 option only applies to 'm68k-sun-sunosN'. On any other system, 1523 '--nfp' has no effect. 1524 1525'--enable-werror' 1526'--disable-werror' 1527'--enable-werror=yes' 1528'--enable-werror=no' 1529 When you specify this option, it controls whether certain files in 1530 the compiler are built with '-Werror' in bootstrap stage2 and 1531 later. If you don't specify it, '-Werror' is turned on for the 1532 main development trunk. However it defaults to off for release 1533 branches and final releases. The specific files which get 1534 '-Werror' are controlled by the Makefiles. 1535 1536'--enable-checking' 1537'--disable-checking' 1538'--enable-checking=LIST' 1539 This option controls performing internal consistency checks in the 1540 compiler. It does not change the generated code, but adds error 1541 checking of the requested complexity. This slows down the compiler 1542 and may only work properly if you are building the compiler with 1543 GCC. 1544 1545 When the option is not specified, the active set of checks depends 1546 on context. Namely, bootstrap stage 1 defaults to 1547 '--enable-checking=yes', builds from release branches or release 1548 archives default to '--enable-checking=release', and otherwise 1549 '--enable-checking=yes,extra' is used. When the option is 1550 specified without a LIST, the result is the same as 1551 '--enable-checking=yes'. Likewise, '--disable-checking' is 1552 equivalent to '--enable-checking=no'. 1553 1554 The categories of checks available in LIST are 'yes' (most common 1555 checks 'assert,misc,gc,gimple,rtlflag,runtime,tree,types'), 'no' 1556 (no checks at all), 'all' (all but 'valgrind'), 'release' (cheapest 1557 checks 'assert,runtime') or 'none' (same as 'no'). 'release' 1558 checks are always on and to disable them '--disable-checking' or 1559 '--enable-checking=no[,<other checks>]' must be explicitly 1560 requested. Disabling assertions makes the compiler and runtime 1561 slightly faster but increases the risk of undetected internal 1562 errors causing wrong code to be generated. 1563 1564 Individual checks can be enabled with these flags: 'assert', 'df', 1565 'extra', 'fold', 'gc', 'gcac', 'gimple', 'misc', 'rtl', 'rtlflag', 1566 'runtime', 'tree', 'types' and 'valgrind'. 'extra' extends 'misc' 1567 checking with extra checks that might affect code generation and 1568 should therefore not differ between stage1 and later stages in 1569 bootstrap. 1570 1571 The 'valgrind' check requires the external 'valgrind' simulator, 1572 available from <http://valgrind.org/>. The 'rtl' checks are 1573 expensive and the 'df', 'gcac' and 'valgrind' checks are very 1574 expensive. 1575 1576'--disable-stage1-checking' 1577'--enable-stage1-checking' 1578'--enable-stage1-checking=LIST' 1579 This option affects only bootstrap build. If no 1580 '--enable-checking' option is specified the stage1 compiler is 1581 built with 'yes' checking enabled, otherwise the stage1 checking 1582 flags are the same as specified by '--enable-checking'. To build 1583 the stage1 compiler with different checking options use 1584 '--enable-stage1-checking'. The list of checking options is the 1585 same as for '--enable-checking'. If your system is too slow or too 1586 small to bootstrap a released compiler with checking for stage1 1587 enabled, you can use '--disable-stage1-checking' to disable 1588 checking for the stage1 compiler. 1589 1590'--enable-coverage' 1591'--enable-coverage=LEVEL' 1592 With this option, the compiler is built to collect self coverage 1593 information, every time it is run. This is for internal 1594 development purposes, and only works when the compiler is being 1595 built with gcc. The LEVEL argument controls whether the compiler 1596 is built optimized or not, values are 'opt' and 'noopt'. For 1597 coverage analysis you want to disable optimization, for performance 1598 analysis you want to enable optimization. When coverage is 1599 enabled, the default level is without optimization. 1600 1601'--enable-gather-detailed-mem-stats' 1602 When this option is specified more detailed information on memory 1603 allocation is gathered. This information is printed when using 1604 '-fmem-report'. 1605 1606'--enable-valgrind-annotations' 1607 Mark selected memory related operations in the compiler when run 1608 under valgrind to suppress false positives. 1609 1610'--enable-nls' 1611'--disable-nls' 1612 The '--enable-nls' option enables Native Language Support (NLS), 1613 which lets GCC output diagnostics in languages other than American 1614 English. Native Language Support is enabled by default if not 1615 doing a canadian cross build. The '--disable-nls' option disables 1616 NLS. 1617 1618'--with-included-gettext' 1619 If NLS is enabled, the '--with-included-gettext' option causes the 1620 build procedure to prefer its copy of GNU 'gettext'. 1621 1622'--with-catgets' 1623 If NLS is enabled, and if the host lacks 'gettext' but has the 1624 inferior 'catgets' interface, the GCC build procedure normally 1625 ignores 'catgets' and instead uses GCC's copy of the GNU 'gettext' 1626 library. The '--with-catgets' option causes the build procedure to 1627 use the host's 'catgets' in this situation. 1628 1629'--with-libiconv-prefix=DIR' 1630 Search for libiconv header files in 'DIR/include' and libiconv 1631 library files in 'DIR/lib'. 1632 1633'--enable-obsolete' 1634 Enable configuration for an obsoleted system. If you attempt to 1635 configure GCC for a system (build, host, or target) which has been 1636 obsoleted, and you do not specify this flag, configure will halt 1637 with an error message. 1638 1639 All support for systems which have been obsoleted in one release of 1640 GCC is removed entirely in the next major release, unless someone 1641 steps forward to maintain the port. 1642 1643'--enable-decimal-float' 1644'--enable-decimal-float=yes' 1645'--enable-decimal-float=no' 1646'--enable-decimal-float=bid' 1647'--enable-decimal-float=dpd' 1648'--disable-decimal-float' 1649 Enable (or disable) support for the C decimal floating point 1650 extension that is in the IEEE 754-2008 standard. This is enabled 1651 by default only on PowerPC, i386, and x86_64 GNU/Linux systems. 1652 Other systems may also support it, but require the user to 1653 specifically enable it. You can optionally control which decimal 1654 floating point format is used (either 'bid' or 'dpd'). The 'bid' 1655 (binary integer decimal) format is default on i386 and x86_64 1656 systems, and the 'dpd' (densely packed decimal) format is default 1657 on PowerPC systems. 1658 1659'--enable-fixed-point' 1660'--disable-fixed-point' 1661 Enable (or disable) support for C fixed-point arithmetic. This 1662 option is enabled by default for some targets (such as MIPS) which 1663 have hardware-support for fixed-point operations. On other 1664 targets, you may enable this option manually. 1665 1666'--with-long-double-128' 1667 Specify if 'long double' type should be 128-bit by default on 1668 selected GNU/Linux architectures. If using 1669 '--without-long-double-128', 'long double' will be by default 1670 64-bit, the same as 'double' type. When neither of these configure 1671 options are used, the default will be 128-bit 'long double' when 1672 built against GNU C Library 2.4 and later, 64-bit 'long double' 1673 otherwise. 1674 1675'--with-long-double-format=ibm' 1676'--with-long-double-format=ieee' 1677 Specify whether 'long double' uses the IBM extended double format 1678 or the IEEE 128-bit floating point format on PowerPC Linux systems. 1679 This configuration switch will only work on little endian PowerPC 1680 Linux systems and on big endian 64-bit systems where the default 1681 cpu is at least power7 (i.e. '--with-cpu=power7', 1682 '--with-cpu=power8', or '--with-cpu=power9' is used). 1683 1684 If you use the '--with-long-double-64' configuration option, the 1685 '--with-long-double-format=ibm' and 1686 '--with-long-double-format=ieee' options are ignored. 1687 1688 The default 'long double' format is to use IBM extended double. 1689 Until all of the libraries are converted to use IEEE 128-bit 1690 floating point, it is not recommended to use 1691 '--with-long-double-format=ieee'. 1692 1693 On little endian PowerPC Linux systems, if you explicitly set the 1694 'long double' type, it will build multilibs to allow you to select 1695 either 'long double' format, unless you disable multilibs with the 1696 '--disable-multilib' option. At present, 'long double' multilibs 1697 are not built on big endian PowerPC Linux systems. If you are 1698 building multilibs, you will need to configure the compiler using 1699 the '--with-system-zlib' option. 1700 1701 If you do not set the 'long double' type explicitly, no multilibs 1702 will be generated. 1703 1704'--enable-fdpic' 1705 On SH Linux systems, generate ELF FDPIC code. 1706 1707'--with-gmp=PATHNAME' 1708'--with-gmp-include=PATHNAME' 1709'--with-gmp-lib=PATHNAME' 1710'--with-mpfr=PATHNAME' 1711'--with-mpfr-include=PATHNAME' 1712'--with-mpfr-lib=PATHNAME' 1713'--with-mpc=PATHNAME' 1714'--with-mpc-include=PATHNAME' 1715'--with-mpc-lib=PATHNAME' 1716 If you want to build GCC but do not have the GMP library, the MPFR 1717 library and/or the MPC library installed in a standard location and 1718 do not have their sources present in the GCC source tree then you 1719 can explicitly specify the directory where they are installed 1720 ('--with-gmp=GMPINSTALLDIR', '--with-mpfr=MPFRINSTALLDIR', 1721 '--with-mpc=MPCINSTALLDIR'). The '--with-gmp=GMPINSTALLDIR' option 1722 is shorthand for '--with-gmp-lib=GMPINSTALLDIR/lib' and 1723 '--with-gmp-include=GMPINSTALLDIR/include'. Likewise the 1724 '--with-mpfr=MPFRINSTALLDIR' option is shorthand for 1725 '--with-mpfr-lib=MPFRINSTALLDIR/lib' and 1726 '--with-mpfr-include=MPFRINSTALLDIR/include', also the 1727 '--with-mpc=MPCINSTALLDIR' option is shorthand for 1728 '--with-mpc-lib=MPCINSTALLDIR/lib' and 1729 '--with-mpc-include=MPCINSTALLDIR/include'. If these shorthand 1730 assumptions are not correct, you can use the explicit include and 1731 lib options directly. You might also need to ensure the shared 1732 libraries can be found by the dynamic linker when building and 1733 using GCC, for example by setting the runtime shared library path 1734 variable ('LD_LIBRARY_PATH' on GNU/Linux and Solaris systems). 1735 1736 These flags are applicable to the host platform only. When 1737 building a cross compiler, they will not be used to configure 1738 target libraries. 1739 1740'--with-isl=PATHNAME' 1741'--with-isl-include=PATHNAME' 1742'--with-isl-lib=PATHNAME' 1743 If you do not have the isl library installed in a standard location 1744 and you want to build GCC, you can explicitly specify the directory 1745 where it is installed ('--with-isl=ISLINSTALLDIR'). The 1746 '--with-isl=ISLINSTALLDIR' option is shorthand for 1747 '--with-isl-lib=ISLINSTALLDIR/lib' and 1748 '--with-isl-include=ISLINSTALLDIR/include'. If this shorthand 1749 assumption is not correct, you can use the explicit include and lib 1750 options directly. 1751 1752 These flags are applicable to the host platform only. When 1753 building a cross compiler, they will not be used to configure 1754 target libraries. 1755 1756'--with-stage1-ldflags=FLAGS' 1757 This option may be used to set linker flags to be used when linking 1758 stage 1 of GCC. These are also used when linking GCC if configured 1759 with '--disable-bootstrap'. If '--with-stage1-libs' is not set to 1760 a value, then the default is '-static-libstdc++ -static-libgcc', if 1761 supported. 1762 1763'--with-stage1-libs=LIBS' 1764 This option may be used to set libraries to be used when linking 1765 stage 1 of GCC. These are also used when linking GCC if configured 1766 with '--disable-bootstrap'. 1767 1768'--with-boot-ldflags=FLAGS' 1769 This option may be used to set linker flags to be used when linking 1770 stage 2 and later when bootstrapping GCC. If -with-boot-libs is not 1771 is set to a value, then the default is '-static-libstdc++ 1772 -static-libgcc'. 1773 1774'--with-boot-libs=LIBS' 1775 This option may be used to set libraries to be used when linking 1776 stage 2 and later when bootstrapping GCC. 1777 1778'--with-debug-prefix-map=MAP' 1779 Convert source directory names using '-fdebug-prefix-map' when 1780 building runtime libraries. 'MAP' is a space-separated list of 1781 maps of the form 'OLD=NEW'. 1782 1783'--enable-linker-build-id' 1784 Tells GCC to pass '--build-id' option to the linker for all final 1785 links (links performed without the '-r' or '--relocatable' option), 1786 if the linker supports it. If you specify 1787 '--enable-linker-build-id', but your linker does not support 1788 '--build-id' option, a warning is issued and the 1789 '--enable-linker-build-id' option is ignored. The default is off. 1790 1791'--with-linker-hash-style=CHOICE' 1792 Tells GCC to pass '--hash-style=CHOICE' option to the linker for 1793 all final links. CHOICE can be one of 'sysv', 'gnu', and 'both' 1794 where 'sysv' is the default. 1795 1796'--enable-gnu-unique-object' 1797'--disable-gnu-unique-object' 1798 Tells GCC to use the gnu_unique_object relocation for C++ template 1799 static data members and inline function local statics. Enabled by 1800 default for a toolchain with an assembler that accepts it and GLIBC 1801 2.11 or above, otherwise disabled. 1802 1803'--with-diagnostics-color=CHOICE' 1804 Tells GCC to use CHOICE as the default for '-fdiagnostics-color=' 1805 option (if not used explicitly on the command line). CHOICE can be 1806 one of 'never', 'auto', 'always', and 'auto-if-env' where 'auto' is 1807 the default. 'auto-if-env' means that '-fdiagnostics-color=auto' 1808 will be the default if 'GCC_COLORS' is present and non-empty in the 1809 environment, and '-fdiagnostics-color=never' otherwise. 1810 1811'--enable-lto' 1812'--disable-lto' 1813 Enable support for link-time optimization (LTO). This is enabled by 1814 default, and may be disabled using '--disable-lto'. 1815 1816'--enable-linker-plugin-configure-flags=FLAGS' 1817'--enable-linker-plugin-flags=FLAGS' 1818 By default, linker plugins (such as the LTO plugin) are built for 1819 the host system architecture. For the case that the linker has a 1820 different (but run-time compatible) architecture, these flags can 1821 be specified to build plugins that are compatible to the linker. 1822 For example, if you are building GCC for a 64-bit x86_64 1823 ('x86_64-pc-linux-gnu') host system, but have a 32-bit x86 1824 GNU/Linux ('i686-pc-linux-gnu') linker executable (which is 1825 executable on the former system), you can configure GCC as follows 1826 for getting compatible linker plugins: 1827 1828 % SRCDIR/configure \ 1829 --host=x86_64-pc-linux-gnu \ 1830 --enable-linker-plugin-configure-flags=--host=i686-pc-linux-gnu \ 1831 --enable-linker-plugin-flags='CC=gcc\ -m32\ -Wl,-rpath,[...]/i686-pc-linux-gnu/lib' 1832 1833'--with-plugin-ld=PATHNAME' 1834 Enable an alternate linker to be used at link-time optimization 1835 (LTO) link time when '-fuse-linker-plugin' is enabled. This linker 1836 should have plugin support such as gold starting with version 2.20 1837 or GNU ld starting with version 2.21. See '-fuse-linker-plugin' 1838 for details. 1839 1840'--enable-canonical-system-headers' 1841'--disable-canonical-system-headers' 1842 Enable system header path canonicalization for 'libcpp'. This can 1843 produce shorter header file paths in diagnostics and dependency 1844 output files, but these changed header paths may conflict with some 1845 compilation environments. Enabled by default, and may be disabled 1846 using '--disable-canonical-system-headers'. 1847 1848'--with-glibc-version=MAJOR.MINOR' 1849 Tell GCC that when the GNU C Library (glibc) is used on the target 1850 it will be version MAJOR.MINOR or later. Normally this can be 1851 detected from the C library's header files, but this option may be 1852 needed when bootstrapping a cross toolchain without the header 1853 files available for building the initial bootstrap compiler. 1854 1855 If GCC is configured with some multilibs that use glibc and some 1856 that do not, this option applies only to the multilibs that use 1857 glibc. However, such configurations may not work well as not all 1858 the relevant configuration in GCC is on a per-multilib basis. 1859 1860'--enable-as-accelerator-for=TARGET' 1861 Build as offload target compiler. Specify offload host triple by 1862 TARGET. 1863 1864'--enable-offload-targets=TARGET1[=PATH1],...,TARGETN[=PATHN]' 1865 Enable offloading to targets TARGET1, ..., TARGETN. Offload 1866 compilers are expected to be already installed. Default search 1867 path for them is 'EXEC-PREFIX', but it can be changed by specifying 1868 paths PATH1, ..., PATHN. 1869 1870 % SRCDIR/configure \ 1871 --enable-offload-targets=x86_64-intelmicemul-linux-gnu=/path/to/x86_64/compiler,nvptx-none,hsa 1872 1873 If 'hsa' is specified as one of the targets, the compiler will be 1874 built with support for HSA GPU accelerators. Because the same 1875 compiler will emit the accelerator code, no path should be 1876 specified. 1877 1878'--with-hsa-runtime=PATHNAME' 1879'--with-hsa-runtime-include=PATHNAME' 1880'--with-hsa-runtime-lib=PATHNAME' 1881 1882 If you configure GCC with HSA offloading but do not have the HSA 1883 run-time library installed in a standard location then you can 1884 explicitly specify the directory where they are installed. The 1885 '--with-hsa-runtime=HSAINSTALLDIR' option is a shorthand for 1886 '--with-hsa-runtime-lib=HSAINSTALLDIR/lib' and 1887 '--with-hsa-runtime-include=HSAINSTALLDIR/include'. 1888 1889'--enable-cet' 1890'--disable-cet' 1891 Enable building target run-time libraries with control-flow 1892 instrumentation, see '-fcf-protection' option. When '--enable-cet' 1893 is specified target libraries are configured to add 1894 '-fcf-protection' and, if needed, other target specific options to 1895 a set of building options. 1896 1897 The option is disabled by default. When '--enable-cet=auto' is 1898 used, it is enabled on Linux/x86 if target binutils supports 'Intel 1899 CET' instructions and disabled otherwise. In this case the target 1900 libraries are configured to get additional '-fcf-protection' 1901 option. 1902 1903'--with-riscv-attribute='yes', 'no' or 'default'' 1904 Generate RISC-V attribute by default, in order to record extra 1905 build information in object. 1906 1907 The option is disabled by default. It is enabled on RISC-V/ELF 1908 (bare-metal) target if target binutils supported. 1909 1910Cross-Compiler-Specific Options 1911------------------------------- 1912 1913The following options only apply to building cross compilers. 1914 1915'--with-sysroot' 1916'--with-sysroot=DIR' 1917 Tells GCC to consider DIR as the root of a tree that contains (a 1918 subset of) the root filesystem of the target operating system. 1919 Target system headers, libraries and run-time object files will be 1920 searched for in there. More specifically, this acts as if 1921 '--sysroot=DIR' was added to the default options of the built 1922 compiler. The specified directory is not copied into the install 1923 tree, unlike the options '--with-headers' and '--with-libs' that 1924 this option obsoletes. The default value, in case '--with-sysroot' 1925 is not given an argument, is '${gcc_tooldir}/sys-root'. If the 1926 specified directory is a subdirectory of '${exec_prefix}', then it 1927 will be found relative to the GCC binaries if the installation tree 1928 is moved. 1929 1930 This option affects the system root for the compiler used to build 1931 target libraries (which runs on the build system) and the compiler 1932 newly installed with 'make install'; it does not affect the 1933 compiler which is used to build GCC itself. 1934 1935 If you specify the '--with-native-system-header-dir=DIRNAME' option 1936 then the compiler will search that directory within DIRNAME for 1937 native system headers rather than the default '/usr/include'. 1938 1939'--with-build-sysroot' 1940'--with-build-sysroot=DIR' 1941 Tells GCC to consider DIR as the system root (see '--with-sysroot') 1942 while building target libraries, instead of the directory specified 1943 with '--with-sysroot'. This option is only useful when you are 1944 already using '--with-sysroot'. You can use '--with-build-sysroot' 1945 when you are configuring with '--prefix' set to a directory that is 1946 different from the one in which you are installing GCC and your 1947 target libraries. 1948 1949 This option affects the system root for the compiler used to build 1950 target libraries (which runs on the build system); it does not 1951 affect the compiler which is used to build GCC itself. 1952 1953 If you specify the '--with-native-system-header-dir=DIRNAME' option 1954 then the compiler will search that directory within DIRNAME for 1955 native system headers rather than the default '/usr/include'. 1956 1957'--with-headers' 1958'--with-headers=DIR' 1959 Deprecated in favor of '--with-sysroot'. Specifies that target 1960 headers are available when building a cross compiler. The DIR 1961 argument specifies a directory which has the target include files. 1962 These include files will be copied into the 'gcc' install 1963 directory. _This option with the DIR argument is required_ when 1964 building a cross compiler, if 'PREFIX/TARGET/sys-include' doesn't 1965 pre-exist. If 'PREFIX/TARGET/sys-include' does pre-exist, the DIR 1966 argument may be omitted. 'fixincludes' will be run on these files 1967 to make them compatible with GCC. 1968 1969'--without-headers' 1970 Tells GCC not use any target headers from a libc when building a 1971 cross compiler. When crossing to GNU/Linux, you need the headers 1972 so GCC can build the exception handling for libgcc. 1973 1974'--with-libs' 1975'--with-libs="DIR1 DIR2 ... DIRN"' 1976 Deprecated in favor of '--with-sysroot'. Specifies a list of 1977 directories which contain the target runtime libraries. These 1978 libraries will be copied into the 'gcc' install directory. If the 1979 directory list is omitted, this option has no effect. 1980 1981'--with-newlib' 1982 Specifies that 'newlib' is being used as the target C library. 1983 This causes '__eprintf' to be omitted from 'libgcc.a' on the 1984 assumption that it will be provided by 'newlib'. 1985 1986'--with-avrlibc' 1987 Specifies that 'AVR-Libc' is being used as the target C library. 1988 This causes float support functions like '__addsf3' to be omitted 1989 from 'libgcc.a' on the assumption that it will be provided by 1990 'libm.a'. For more technical details, cf. PR54461. This option 1991 is only supported for the AVR target. It is not supported for 1992 RTEMS configurations, which currently use newlib. The option is 1993 supported since version 4.7.2 and is the default in 4.8.0 and 1994 newer. 1995 1996'--with-nds32-lib=LIBRARY' 1997 Specifies that LIBRARY setting is used for building 'libgcc.a'. 1998 Currently, the valid LIBRARY is 'newlib' or 'mculib'. This option 1999 is only supported for the NDS32 target. 2000 2001'--with-build-time-tools=DIR' 2002 Specifies where to find the set of target tools (assembler, linker, 2003 etc.) that will be used while building GCC itself. This option 2004 can be useful if the directory layouts are different between the 2005 system you are building GCC on, and the system where you will 2006 deploy it. 2007 2008 For example, on an 'ia64-hp-hpux' system, you may have the GNU 2009 assembler and linker in '/usr/bin', and the native tools in a 2010 different path, and build a toolchain that expects to find the 2011 native tools in '/usr/bin'. 2012 2013 When you use this option, you should ensure that DIR includes 'ar', 2014 'as', 'ld', 'nm', 'ranlib' and 'strip' if necessary, and possibly 2015 'objdump'. Otherwise, GCC may use an inconsistent set of tools. 2016 2017Overriding 'configure' test results 2018................................... 2019 2020Sometimes, it might be necessary to override the result of some 2021'configure' test, for example in order to ease porting to a new system 2022or work around a bug in a test. The toplevel 'configure' script 2023provides three variables for this: 2024 2025'build_configargs' 2026 The contents of this variable is passed to all build 'configure' 2027 scripts. 2028 2029'host_configargs' 2030 The contents of this variable is passed to all host 'configure' 2031 scripts. 2032 2033'target_configargs' 2034 The contents of this variable is passed to all target 'configure' 2035 scripts. 2036 2037 In order to avoid shell and 'make' quoting issues for complex 2038overrides, you can pass a setting for 'CONFIG_SITE' and set variables in 2039the site file. 2040 2041Objective-C-Specific Options 2042---------------------------- 2043 2044The following options apply to the build of the Objective-C runtime 2045library. 2046 2047'--enable-objc-gc' 2048 Specify that an additional variant of the GNU Objective-C runtime 2049 library is built, using an external build of the 2050 Boehm-Demers-Weiser garbage collector 2051 (<http://www.hboehm.info/gc/>). This library needs to be available 2052 for each multilib variant, unless configured with 2053 '--enable-objc-gc='auto'' in which case the build of the additional 2054 runtime library is skipped when not available and the build 2055 continues. 2056 2057'--with-target-bdw-gc=LIST' 2058'--with-target-bdw-gc-include=LIST' 2059'--with-target-bdw-gc-lib=LIST' 2060 Specify search directories for the garbage collector header files 2061 and libraries. LIST is a comma separated list of key value pairs 2062 of the form 'MULTILIBDIR=PATH', where the default multilib key is 2063 named as '.' (dot), or is omitted (e.g. 2064 '--with-target-bdw-gc=/opt/bdw-gc,32=/opt-bdw-gc32'). 2065 2066 The options '--with-target-bdw-gc-include' and 2067 '--with-target-bdw-gc-lib' must always be specified together for 2068 each multilib variant and they take precedence over 2069 '--with-target-bdw-gc'. If '--with-target-bdw-gc-include' is 2070 missing values for a multilib, then the value for the default 2071 multilib is used (e.g. 2072 '--with-target-bdw-gc-include=/opt/bdw-gc/include' 2073 '--with-target-bdw-gc-lib=/opt/bdw-gc/lib64,32=/opt-bdw-gc/lib32'). 2074 If none of these options are specified, the library is assumed in 2075 default locations. 2076 2077D-Specific Options 2078------------------ 2079 2080The following options apply to the build of the D runtime library. 2081 2082'--with-target-system-zlib' 2083 Use installed 'zlib' rather than that included with GCC. This 2084 needs to be available for each multilib variant, unless configured 2085 with '--with-target-system-zlib='auto'' in which case the 2086 GCC included 'zlib' is only used when the system installed library 2087 is not available. 2088 2089 2090File: gccinstall.info, Node: Building, Next: Testing, Prev: Configuration, Up: Installing GCC 2091 20925 Building 2093********** 2094 2095Now that GCC is configured, you are ready to build the compiler and 2096runtime libraries. 2097 2098 Some commands executed when making the compiler may fail (return a 2099nonzero status) and be ignored by 'make'. These failures, which are 2100often due to files that were not found, are expected, and can safely be 2101ignored. 2102 2103 It is normal to have compiler warnings when compiling certain files. 2104Unless you are a GCC developer, you can generally ignore these warnings 2105unless they cause compilation to fail. Developers should attempt to fix 2106any warnings encountered, however they can temporarily continue past 2107warnings-as-errors by specifying the configure flag '--disable-werror'. 2108 2109 On certain old systems, defining certain environment variables such 2110as 'CC' can interfere with the functioning of 'make'. 2111 2112 If you encounter seemingly strange errors when trying to build the 2113compiler in a directory other than the source directory, it could be 2114because you have previously configured the compiler in the source 2115directory. Make sure you have done all the necessary preparations. 2116 2117 If you build GCC on a BSD system using a directory stored in an old 2118System V file system, problems may occur in running 'fixincludes' if the 2119System V file system doesn't support symbolic links. These problems 2120result in a failure to fix the declaration of 'size_t' in 'sys/types.h'. 2121If you find that 'size_t' is a signed type and that type mismatches 2122occur, this could be the cause. 2123 2124 The solution is not to use such a directory for building GCC. 2125 2126 Similarly, when building from the source repository or snapshots, or 2127if you modify '*.l' files, you need the Flex lexical analyzer generator 2128installed. If you do not modify '*.l' files, releases contain the 2129Flex-generated files and you do not need Flex installed to build them. 2130There is still one Flex-based lexical analyzer (part of the build 2131machinery, not of GCC itself) that is used even if you only build the C 2132front end. 2133 2134 When building from the source repository or snapshots, or if you 2135modify Texinfo documentation, you need version 4.7 or later of Texinfo 2136installed if you want Info documentation to be regenerated. Releases 2137contain Info documentation pre-built for the unmodified documentation in 2138the release. 2139 21405.1 Building a native compiler 2141============================== 2142 2143For a native build, the default configuration is to perform a 3-stage 2144bootstrap of the compiler when 'make' is invoked. This will build the 2145entire GCC system and ensure that it compiles itself correctly. It can 2146be disabled with the '--disable-bootstrap' parameter to 'configure', but 2147bootstrapping is suggested because the compiler will be tested more 2148completely and could also have better performance. 2149 2150 The bootstrapping process will complete the following steps: 2151 2152 * Build tools necessary to build the compiler. 2153 2154 * Perform a 3-stage bootstrap of the compiler. This includes 2155 building three times the target tools for use by the compiler such 2156 as binutils (bfd, binutils, gas, gprof, ld, and opcodes) if they 2157 have been individually linked or moved into the top level GCC 2158 source tree before configuring. 2159 2160 * Perform a comparison test of the stage2 and stage3 compilers. 2161 2162 * Build runtime libraries using the stage3 compiler from the previous 2163 step. 2164 2165 If you are short on disk space you might consider 'make 2166bootstrap-lean' instead. The sequence of compilation is the same 2167described above, but object files from the stage1 and stage2 of the 21683-stage bootstrap of the compiler are deleted as soon as they are no 2169longer needed. 2170 2171 If you wish to use non-default GCC flags when compiling the stage2 2172and stage3 compilers, set 'BOOT_CFLAGS' on the command line when doing 2173'make'. For example, if you want to save additional space during the 2174bootstrap and in the final installation as well, you can build the 2175compiler binaries without debugging information as in the following 2176example. This will save roughly 40% of disk space both for the 2177bootstrap and the final installation. (Libraries will still contain 2178debugging information.) 2179 2180 make BOOT_CFLAGS='-O' bootstrap 2181 2182 You can place non-default optimization flags into 'BOOT_CFLAGS'; they 2183are less well tested here than the default of '-g -O2', but should still 2184work. In a few cases, you may find that you need to specify special 2185flags such as '-msoft-float' here to complete the bootstrap; or, if the 2186native compiler miscompiles the stage1 compiler, you may need to work 2187around this, by choosing 'BOOT_CFLAGS' to avoid the parts of the stage1 2188compiler that were miscompiled, or by using 'make bootstrap4' to 2189increase the number of stages of bootstrap. 2190 2191 'BOOT_CFLAGS' does not apply to bootstrapped target libraries. Since 2192these are always compiled with the compiler currently being 2193bootstrapped, you can use 'CFLAGS_FOR_TARGET' to modify their 2194compilation flags, as for non-bootstrapped target libraries. Again, if 2195the native compiler miscompiles the stage1 compiler, you may need to 2196work around this by avoiding non-working parts of the stage1 compiler. 2197Use 'STAGE1_TFLAGS' to this end. 2198 2199 If you used the flag '--enable-languages=...' to restrict the 2200compilers to be built, only those you've actually enabled will be built. 2201This will of course only build those runtime libraries, for which the 2202particular compiler has been built. Please note, that re-defining 2203'LANGUAGES' when calling 'make' *does not* work anymore! 2204 2205 If the comparison of stage2 and stage3 fails, this normally indicates 2206that the stage2 compiler has compiled GCC incorrectly, and is therefore 2207a potentially serious bug which you should investigate and report. (On 2208a few systems, meaningful comparison of object files is impossible; they 2209always appear "different". If you encounter this problem, you will need 2210to disable comparison in the 'Makefile'.) 2211 2212 If you do not want to bootstrap your compiler, you can configure with 2213'--disable-bootstrap'. In particular cases, you may want to bootstrap 2214your compiler even if the target system is not the same as the one you 2215are building on: for example, you could build a 2216'powerpc-unknown-linux-gnu' toolchain on a 'powerpc64-unknown-linux-gnu' 2217host. In this case, pass '--enable-bootstrap' to the configure script. 2218 2219 'BUILD_CONFIG' can be used to bring in additional customization to 2220the build. It can be set to a whitespace-separated list of names. For 2221each such 'NAME', top-level 'config/NAME.mk' will be included by the 2222top-level 'Makefile', bringing in any settings it contains. The default 2223'BUILD_CONFIG' can be set using the configure option 2224'--with-build-config=NAME...'. Some examples of supported build 2225configurations are: 2226 2227'bootstrap-O1' 2228 Removes any '-O'-started option from 'BOOT_CFLAGS', and adds '-O1' 2229 to it. 'BUILD_CONFIG=bootstrap-O1' is equivalent to 2230 'BOOT_CFLAGS='-g -O1''. 2231 2232'bootstrap-O3' 2233 Analogous to 'bootstrap-O1'. 2234 2235'bootstrap-lto' 2236 Enables Link-Time Optimization for host tools during bootstrapping. 2237 'BUILD_CONFIG=bootstrap-lto' is equivalent to adding '-flto' to 2238 'BOOT_CFLAGS'. This option assumes that the host supports the 2239 linker plugin (e.g. GNU ld version 2.21 or later or GNU gold 2240 version 2.21 or later). 2241 2242'bootstrap-lto-noplugin' 2243 This option is similar to 'bootstrap-lto', but is intended for 2244 hosts that do not support the linker plugin. Without the linker 2245 plugin static libraries are not compiled with link-time 2246 optimizations. Since the GCC middle end and back end are in 2247 'libbackend.a' this means that only the front end is actually LTO 2248 optimized. 2249 2250'bootstrap-lto-lean' 2251 This option is similar to 'bootstrap-lto', but is intended for 2252 faster build by only using LTO in the final bootstrap stage. With 2253 'make profiledbootstrap' the LTO frontend is trained only on 2254 generator files. 2255 2256'bootstrap-debug' 2257 Verifies that the compiler generates the same executable code, 2258 whether or not it is asked to emit debug information. To this end, 2259 this option builds stage2 host programs without debug information, 2260 and uses 'contrib/compare-debug' to compare them with the stripped 2261 stage3 object files. If 'BOOT_CFLAGS' is overridden so as to not 2262 enable debug information, stage2 will have it, and stage3 won't. 2263 This option is enabled by default when GCC bootstrapping is 2264 enabled, if 'strip' can turn object files compiled with and without 2265 debug info into identical object files. In addition to better test 2266 coverage, this option makes default bootstraps faster and leaner. 2267 2268'bootstrap-debug-big' 2269 Rather than comparing stripped object files, as in 2270 'bootstrap-debug', this option saves internal compiler dumps during 2271 stage2 and stage3 and compares them as well, which helps catch 2272 additional potential problems, but at a great cost in terms of disk 2273 space. It can be specified in addition to 'bootstrap-debug'. 2274 2275'bootstrap-debug-lean' 2276 This option saves disk space compared with 'bootstrap-debug-big', 2277 but at the expense of some recompilation. Instead of saving the 2278 dumps of stage2 and stage3 until the final compare, it uses 2279 '-fcompare-debug' to generate, compare and remove the dumps during 2280 stage3, repeating the compilation that already took place in 2281 stage2, whose dumps were not saved. 2282 2283'bootstrap-debug-lib' 2284 This option tests executable code invariance over debug information 2285 generation on target libraries, just like 'bootstrap-debug-lean' 2286 tests it on host programs. It builds stage3 libraries with 2287 '-fcompare-debug', and it can be used along with any of the 2288 'bootstrap-debug' options above. 2289 2290 There aren't '-lean' or '-big' counterparts to this option because 2291 most libraries are only build in stage3, so bootstrap compares 2292 would not get significant coverage. Moreover, the few libraries 2293 built in stage2 are used in stage3 host programs, so we wouldn't 2294 want to compile stage2 libraries with different options for 2295 comparison purposes. 2296 2297'bootstrap-debug-ckovw' 2298 Arranges for error messages to be issued if the compiler built on 2299 any stage is run without the option '-fcompare-debug'. This is 2300 useful to verify the full '-fcompare-debug' testing coverage. It 2301 must be used along with 'bootstrap-debug-lean' and 2302 'bootstrap-debug-lib'. 2303 2304'bootstrap-cet' 2305 This option enables Intel CET for host tools during bootstrapping. 2306 'BUILD_CONFIG=bootstrap-cet' is equivalent to adding 2307 '-fcf-protection' to 'BOOT_CFLAGS'. This option assumes that the 2308 host supports Intel CET (e.g. GNU assembler version 2.30 or later). 2309 2310'bootstrap-time' 2311 Arranges for the run time of each program started by the GCC 2312 driver, built in any stage, to be logged to 'time.log', in the top 2313 level of the build tree. 2314 23155.2 Building a cross compiler 2316============================= 2317 2318When building a cross compiler, it is not generally possible to do a 23193-stage bootstrap of the compiler. This makes for an interesting 2320problem as parts of GCC can only be built with GCC. 2321 2322 To build a cross compiler, we recommend first building and installing 2323a native compiler. You can then use the native GCC compiler to build 2324the cross compiler. The installed native compiler needs to be GCC 2325version 2.95 or later. 2326 2327 Assuming you have already installed a native copy of GCC and 2328configured your cross compiler, issue the command 'make', which performs 2329the following steps: 2330 2331 * Build host tools necessary to build the compiler. 2332 2333 * Build target tools for use by the compiler such as binutils (bfd, 2334 binutils, gas, gprof, ld, and opcodes) if they have been 2335 individually linked or moved into the top level GCC source tree 2336 before configuring. 2337 2338 * Build the compiler (single stage only). 2339 2340 * Build runtime libraries using the compiler from the previous step. 2341 2342 Note that if an error occurs in any step the make process will exit. 2343 2344 If you are not building GNU binutils in the same source tree as GCC, 2345you will need a cross-assembler and cross-linker installed before 2346configuring GCC. Put them in the directory 'PREFIX/TARGET/bin'. Here 2347is a table of the tools you should put in this directory: 2348 2349'as' 2350 This should be the cross-assembler. 2351 2352'ld' 2353 This should be the cross-linker. 2354 2355'ar' 2356 This should be the cross-archiver: a program which can manipulate 2357 archive files (linker libraries) in the target machine's format. 2358 2359'ranlib' 2360 This should be a program to construct a symbol table in an archive 2361 file. 2362 2363 The installation of GCC will find these programs in that directory, 2364and copy or link them to the proper place to for the cross-compiler to 2365find them when run later. 2366 2367 The easiest way to provide these files is to build the Binutils 2368package. Configure it with the same '--host' and '--target' options 2369that you use for configuring GCC, then build and install them. They 2370install their executables automatically into the proper directory. 2371Alas, they do not support all the targets that GCC supports. 2372 2373 If you are not building a C library in the same source tree as GCC, 2374you should also provide the target libraries and headers before 2375configuring GCC, specifying the directories with '--with-sysroot' or 2376'--with-headers' and '--with-libs'. Many targets also require "start 2377files" such as 'crt0.o' and 'crtn.o' which are linked into each 2378executable. There may be several alternatives for 'crt0.o', for use 2379with profiling or other compilation options. Check your target's 2380definition of 'STARTFILE_SPEC' to find out what start files it uses. 2381 23825.3 Building in parallel 2383======================== 2384 2385GNU Make 3.80 and above, which is necessary to build GCC, support 2386building in parallel. To activate this, you can use 'make -j 2' instead 2387of 'make'. You can also specify a bigger number, and in most cases 2388using a value greater than the number of processors in your machine will 2389result in fewer and shorter I/O latency hits, thus improving overall 2390throughput; this is especially true for slow drives and network 2391filesystems. 2392 23935.4 Building the Ada compiler 2394============================= 2395 2396In order to build GNAT, the Ada compiler, you need a working GNAT 2397compiler (GCC version 4.0 or later). This includes GNAT tools such as 2398'gnatmake' and 'gnatlink', since the Ada front end is written in Ada and 2399uses some GNAT-specific extensions. 2400 2401 In order to build a cross compiler, it is suggested to install the 2402new compiler as native first, and then use it to build the cross 2403compiler. 2404 2405 'configure' does not test whether the GNAT installation works and has 2406a sufficiently recent version; if too old a GNAT version is installed, 2407the build will fail unless '--enable-languages' is used to disable 2408building the Ada front end. 2409 2410 'ADA_INCLUDE_PATH' and 'ADA_OBJECT_PATH' environment variables must 2411not be set when building the Ada compiler, the Ada tools, or the Ada 2412runtime libraries. You can check that your build environment is clean 2413by verifying that 'gnatls -v' lists only one explicit path in each 2414section. 2415 24165.5 Building with profile feedback 2417================================== 2418 2419It is possible to use profile feedback to optimize the compiler itself. 2420This should result in a faster compiler binary. Experiments done on x86 2421using gcc 3.3 showed approximately 7 percent speedup on compiling C 2422programs. To bootstrap the compiler with profile feedback, use 'make 2423profiledbootstrap'. 2424 2425 When 'make profiledbootstrap' is run, it will first build a 'stage1' 2426compiler. This compiler is used to build a 'stageprofile' compiler 2427instrumented to collect execution counts of instruction and branch 2428probabilities. Training run is done by building 'stagetrain' compiler. 2429Finally a 'stagefeedback' compiler is built using the information 2430collected. 2431 2432 Unlike standard bootstrap, several additional restrictions apply. 2433The compiler used to build 'stage1' needs to support a 64-bit integral 2434type. It is recommended to only use GCC for this. 2435 2436 On Linux/x86_64 hosts with some restrictions (no virtualization) it 2437is also possible to do autofdo build with 'make autoprofiledback'. This 2438uses Linux perf to sample branches in the binary and then rebuild it 2439with feedback derived from the profile. Linux perf and the 'autofdo' 2440toolkit needs to be installed for this. 2441 2442 Only the profile from the current build is used, so when an error 2443occurs it is recommended to clean before restarting. Otherwise the code 2444quality may be much worse. 2445 2446 2447File: gccinstall.info, Node: Testing, Next: Final install, Prev: Building, Up: Installing GCC 2448 24496 Installing GCC: Testing 2450************************* 2451 2452Before you install GCC, we encourage you to run the testsuites and to 2453compare your results with results from a similar configuration that have 2454been submitted to the gcc-testresults mailing list. Some of these 2455archived results are linked from the build status lists at 2456<http://gcc.gnu.org/buildstat.html>, although not everyone who reports a 2457successful build runs the testsuites and submits the results. This step 2458is optional and may require you to download additional software, but it 2459can give you confidence in your new GCC installation or point out 2460problems before you install and start using your new GCC. 2461 2462 First, you must have downloaded the testsuites. These are part of 2463the full distribution, but if you downloaded the "core" compiler plus 2464any front ends, you must download the testsuites separately. 2465 2466 Second, you must have the testing tools installed. This includes 2467DejaGnu, Tcl, and Expect; the DejaGnu site has links to these. For 2468running the BRIG frontend tests, a tool to assemble the binary BRIGs 2469from HSAIL text, HSAILasm must be installed. 2470 2471 If the directories where 'runtest' and 'expect' were installed are 2472not in the 'PATH', you may need to set the following environment 2473variables appropriately, as in the following example (which assumes that 2474DejaGnu has been installed under '/usr/local'): 2475 2476 TCL_LIBRARY = /usr/local/share/tcl8.0 2477 DEJAGNULIBS = /usr/local/share/dejagnu 2478 2479 (On systems such as Cygwin, these paths are required to be actual 2480paths, not mounts or links; presumably this is due to some lack of 2481portability in the DejaGnu code.) 2482 2483 Finally, you can run the testsuite (which may take a long time): 2484 cd OBJDIR; make -k check 2485 2486 This will test various components of GCC, such as compiler front ends 2487and runtime libraries. While running the testsuite, DejaGnu might emit 2488some harmless messages resembling 'WARNING: Couldn't find the global 2489config file.' or 'WARNING: Couldn't find tool init file' that can be 2490ignored. 2491 2492 If you are testing a cross-compiler, you may want to run the 2493testsuite on a simulator as described at 2494<http://gcc.gnu.org/simtest-howto.html>. 2495 24966.1 How can you run the testsuite on selected tests? 2497==================================================== 2498 2499In order to run sets of tests selectively, there are targets 'make 2500check-gcc' and language specific 'make check-c', 'make check-c++', 'make 2501check-d' 'make check-fortran', 'make check-ada', 'make check-objc', 2502'make check-obj-c++', 'make check-lto' in the 'gcc' subdirectory of the 2503object directory. You can also just run 'make check' in a subdirectory 2504of the object directory. 2505 2506 A more selective way to just run all 'gcc' execute tests in the 2507testsuite is to use 2508 2509 make check-gcc RUNTESTFLAGS="execute.exp OTHER-OPTIONS" 2510 2511 Likewise, in order to run only the 'g++' "old-deja" tests in the 2512testsuite with filenames matching '9805*', you would use 2513 2514 make check-g++ RUNTESTFLAGS="old-deja.exp=9805* OTHER-OPTIONS" 2515 2516 The file-matching expression following FILENAME'.exp=' is treated as 2517a series of whitespace-delimited glob expressions so that multiple 2518patterns may be passed, although any whitespace must either be escaped 2519or surrounded by single quotes if multiple expressions are desired. For 2520example, 2521 2522 make check-g++ RUNTESTFLAGS="old-deja.exp=9805*\ virtual2.c OTHER-OPTIONS" 2523 make check-g++ RUNTESTFLAGS="'old-deja.exp=9805* virtual2.c' OTHER-OPTIONS" 2524 2525 The '*.exp' files are located in the testsuite directories of the GCC 2526source, the most important ones being 'compile.exp', 'execute.exp', 2527'dg.exp' and 'old-deja.exp'. To get a list of the possible '*.exp' 2528files, pipe the output of 'make check' into a file and look at the 2529'Running ... .exp' lines. 2530 25316.2 Passing options and running multiple testsuites 2532=================================================== 2533 2534You can pass multiple options to the testsuite using the 2535'--target_board' option of DejaGNU, either passed as part of 2536'RUNTESTFLAGS', or directly to 'runtest' if you prefer to work outside 2537the makefiles. For example, 2538 2539 make check-g++ RUNTESTFLAGS="--target_board=unix/-O3/-fmerge-constants" 2540 2541 will run the standard 'g++' testsuites ("unix" is the target name for 2542a standard native testsuite situation), passing '-O3 -fmerge-constants' 2543to the compiler on every test, i.e., slashes separate options. 2544 2545 You can run the testsuites multiple times using combinations of 2546options with a syntax similar to the brace expansion of popular shells: 2547 2548 ..."--target_board=arm-sim\{-mhard-float,-msoft-float\}\{-O1,-O2,-O3,\}" 2549 2550 (Note the empty option caused by the trailing comma in the final 2551group.) The following will run each testsuite eight times using the 2552'arm-sim' target, as if you had specified all possible combinations 2553yourself: 2554 2555 --target_board='arm-sim/-mhard-float/-O1 \ 2556 arm-sim/-mhard-float/-O2 \ 2557 arm-sim/-mhard-float/-O3 \ 2558 arm-sim/-mhard-float \ 2559 arm-sim/-msoft-float/-O1 \ 2560 arm-sim/-msoft-float/-O2 \ 2561 arm-sim/-msoft-float/-O3 \ 2562 arm-sim/-msoft-float' 2563 2564 They can be combined as many times as you wish, in arbitrary ways. 2565This list: 2566 2567 ..."--target_board=unix/-Wextra\{-O3,-fno-strength\}\{-fomit-frame,\}" 2568 2569 will generate four combinations, all involving '-Wextra'. 2570 2571 The disadvantage to this method is that the testsuites are run in 2572serial, which is a waste on multiprocessor systems. For users with GNU 2573Make and a shell which performs brace expansion, you can run the 2574testsuites in parallel by having the shell perform the combinations and 2575'make' do the parallel runs. Instead of using '--target_board', use a 2576special makefile target: 2577 2578 make -jN check-TESTSUITE//TEST-TARGET/OPTION1/OPTION2/... 2579 2580 For example, 2581 2582 make -j3 check-gcc//sh-hms-sim/{-m1,-m2,-m3,-m3e,-m4}/{,-nofpu} 2583 2584 will run three concurrent "make-gcc" testsuites, eventually testing 2585all ten combinations as described above. Note that this is currently 2586only supported in the 'gcc' subdirectory. (To see how this works, try 2587typing 'echo' before the example given here.) 2588 25896.3 How to interpret test results 2590================================= 2591 2592The result of running the testsuite are various '*.sum' and '*.log' 2593files in the testsuite subdirectories. The '*.log' files contain a 2594detailed log of the compiler invocations and the corresponding results, 2595the '*.sum' files summarize the results. These summaries contain status 2596codes for all tests: 2597 2598 * PASS: the test passed as expected 2599 * XPASS: the test unexpectedly passed 2600 * FAIL: the test unexpectedly failed 2601 * XFAIL: the test failed as expected 2602 * UNSUPPORTED: the test is not supported on this platform 2603 * ERROR: the testsuite detected an error 2604 * WARNING: the testsuite detected a possible problem 2605 2606 It is normal for some tests to report unexpected failures. At the 2607current time the testing harness does not allow fine grained control 2608over whether or not a test is expected to fail. This problem should be 2609fixed in future releases. 2610 26116.4 Submitting test results 2612=========================== 2613 2614If you want to report the results to the GCC project, use the 2615'contrib/test_summary' shell script. Start it in the OBJDIR with 2616 2617 SRCDIR/contrib/test_summary -p your_commentary.txt \ 2618 -m gcc-testresults@gcc.gnu.org |sh 2619 2620 This script uses the 'Mail' program to send the results, so make sure 2621it is in your 'PATH'. The file 'your_commentary.txt' is prepended to 2622the testsuite summary and should contain any special remarks you have on 2623your results or your build environment. Please do not edit the 2624testsuite result block or the subject line, as these messages may be 2625automatically processed. 2626 2627 2628File: gccinstall.info, Node: Final install, Prev: Testing, Up: Installing GCC 2629 26307 Installing GCC: Final installation 2631************************************ 2632 2633Now that GCC has been built (and optionally tested), you can install it 2634with 2635 cd OBJDIR && make install 2636 2637 We strongly recommend to install into a target directory where there 2638is no previous version of GCC present. Also, the GNAT runtime should 2639not be stripped, as this would break certain features of the debugger 2640that depend on this debugging information (catching Ada exceptions for 2641instance). 2642 2643 That step completes the installation of GCC; user level binaries can 2644be found in 'PREFIX/bin' where PREFIX is the value you specified with 2645the '--prefix' to configure (or '/usr/local' by default). (If you 2646specified '--bindir', that directory will be used instead; otherwise, if 2647you specified '--exec-prefix', 'EXEC-PREFIX/bin' will be used.) Headers 2648for the C++ library are installed in 'PREFIX/include'; libraries in 2649'LIBDIR' (normally 'PREFIX/lib'); internal parts of the compiler in 2650'LIBDIR/gcc' and 'LIBEXECDIR/gcc'; documentation in info format in 2651'INFODIR' (normally 'PREFIX/info'). 2652 2653 When installing cross-compilers, GCC's executables are not only 2654installed into 'BINDIR', that is, 'EXEC-PREFIX/bin', but additionally 2655into 'EXEC-PREFIX/TARGET-ALIAS/bin', if that directory exists. 2656Typically, such "tooldirs" hold target-specific binutils, including 2657assembler and linker. 2658 2659 Installation into a temporary staging area or into a 'chroot' jail 2660can be achieved with the command 2661 2662 make DESTDIR=PATH-TO-ROOTDIR install 2663 2664where PATH-TO-ROOTDIR is the absolute path of a directory relative to 2665which all installation paths will be interpreted. Note that the 2666directory specified by 'DESTDIR' need not exist yet; it will be created 2667if necessary. 2668 2669 There is a subtle point with tooldirs and 'DESTDIR': If you relocate 2670a cross-compiler installation with e.g. 'DESTDIR=ROOTDIR', then the 2671directory 'ROOTDIR/EXEC-PREFIX/TARGET-ALIAS/bin' will be filled with 2672duplicated GCC executables only if it already exists, it will not be 2673created otherwise. This is regarded as a feature, not as a bug, because 2674it gives slightly more control to the packagers using the 'DESTDIR' 2675feature. 2676 2677 You can install stripped programs and libraries with 2678 2679 make install-strip 2680 2681 If you are bootstrapping a released version of GCC then please 2682quickly review the build status page for your release, available from 2683<http://gcc.gnu.org/buildstat.html>. If your system is not listed for 2684the version of GCC that you built, send a note to <gcc@gcc.gnu.org> 2685indicating that you successfully built and installed GCC. Include the 2686following information: 2687 2688 * Output from running 'SRCDIR/config.guess'. Do not send that file 2689 itself, just the one-line output from running it. 2690 2691 * The output of 'gcc -v' for your newly installed 'gcc'. This tells 2692 us which version of GCC you built and the options you passed to 2693 configure. 2694 2695 * Whether you enabled all languages or a subset of them. If you used 2696 a full distribution then this information is part of the configure 2697 options in the output of 'gcc -v', but if you downloaded the "core" 2698 compiler plus additional front ends then it isn't apparent which 2699 ones you built unless you tell us about it. 2700 2701 * If the build was for GNU/Linux, also include: 2702 * The distribution name and version (e.g., Red Hat 7.1 or Debian 2703 2.2.3); this information should be available from 2704 '/etc/issue'. 2705 2706 * The version of the Linux kernel, available from 'uname 2707 --version' or 'uname -a'. 2708 2709 * The version of glibc you used; for RPM-based systems like Red 2710 Hat, Mandrake, and SuSE type 'rpm -q glibc' to get the glibc 2711 version, and on systems like Debian and Progeny use 'dpkg -l 2712 libc6'. 2713 For other systems, you can include similar information if you think 2714 it is relevant. 2715 2716 * Any other information that you think would be useful to people 2717 building GCC on the same configuration. The new entry in the build 2718 status list will include a link to the archived copy of your 2719 message. 2720 2721 We'd also like to know if the *note host/target specific installation 2722notes: Specific. didn't include your host/target information or if that 2723information is incomplete or out of date. Send a note to 2724<gcc@gcc.gnu.org> detailing how the information should be changed. 2725 2726 If you find a bug, please report it following the bug reporting 2727guidelines. 2728 2729 If you want to print the GCC manuals, do 'cd OBJDIR; make dvi'. You 2730will need to have 'texi2dvi' (version at least 4.7) and TeX installed. 2731This creates a number of '.dvi' files in subdirectories of 'OBJDIR'; 2732these may be converted for printing with programs such as 'dvips'. 2733Alternately, by using 'make pdf' in place of 'make dvi', you can create 2734documentation in the form of '.pdf' files; this requires 'texi2pdf', 2735which is included with Texinfo version 4.8 and later. You can also buy 2736printed manuals from the Free Software Foundation, though such manuals 2737may not be for the most recent version of GCC. 2738 2739 If you would like to generate online HTML documentation, do 'cd 2740OBJDIR; make html' and HTML will be generated for the gcc manuals in 2741'OBJDIR/gcc/HTML'. 2742 2743 2744File: gccinstall.info, Node: Binaries, Next: Specific, Prev: Installing GCC, Up: Top 2745 27468 Installing GCC: Binaries 2747************************** 2748 2749We are often asked about pre-compiled versions of GCC. While we cannot 2750provide these for all platforms, below you'll find links to binaries for 2751various platforms where creating them by yourself is not easy due to 2752various reasons. 2753 2754 Please note that we did not create these binaries, nor do we support 2755them. If you have any problems installing them, please contact their 2756makers. 2757 2758 * AIX: 2759 * Bull's Open Source Software Archive for for AIX 5L and AIX 6; 2760 2761 * AIX Open Source Packages (AIX5L AIX 6.1 AIX 7.1). 2762 2763 * DOS--DJGPP. 2764 2765 * HP-UX: 2766 * HP-UX Porting Center; 2767 2768 * Solaris 2 (SPARC, Intel): 2769 * OpenCSW 2770 2771 * TGCware 2772 2773 * macOS: 2774 * The Homebrew package manager; 2775 * MacPorts. 2776 2777 * Microsoft Windows: 2778 * The Cygwin project; 2779 * The MinGW and mingw-w64 projects. 2780 2781 * OpenPKG offers binaries for quite a number of platforms. 2782 2783 * The GFortran Wiki has links to GNU Fortran binaries for several 2784 platforms. 2785 2786 2787File: gccinstall.info, Node: Specific, Next: Old, Prev: Binaries, Up: Top 2788 27899 Host/target specific installation notes for GCC 2790************************************************* 2791 2792Please read this document carefully _before_ installing the GNU Compiler 2793Collection on your machine. 2794 2795 Note that this list of install notes is _not_ a list of supported 2796hosts or targets. Not all supported hosts and targets are listed here, 2797only the ones that require host-specific or target-specific information 2798have to. 2799 2800aarch64*-*-* 2801============ 2802 2803Binutils pre 2.24 does not have support for selecting '-mabi' and does 2804not support ILP32. If it is used to build GCC 4.9 or later, GCC will 2805not support option '-mabi=ilp32'. 2806 2807 To enable a workaround for the Cortex-A53 erratum number 835769 by 2808default (for all CPUs regardless of -mcpu option given) at configure 2809time use the '--enable-fix-cortex-a53-835769' option. This will enable 2810the fix by default and can be explicitly disabled during compilation by 2811passing the '-mno-fix-cortex-a53-835769' option. Conversely, 2812'--disable-fix-cortex-a53-835769' will disable the workaround by 2813default. The workaround is disabled by default if neither of 2814'--enable-fix-cortex-a53-835769' or '--disable-fix-cortex-a53-835769' is 2815given at configure time. 2816 2817 To enable a workaround for the Cortex-A53 erratum number 843419 by 2818default (for all CPUs regardless of -mcpu option given) at configure 2819time use the '--enable-fix-cortex-a53-843419' option. This workaround 2820is applied at link time. Enabling the workaround will cause GCC to pass 2821the relevant option to the linker. It can be explicitly disabled during 2822compilation by passing the '-mno-fix-cortex-a53-843419' option. 2823Conversely, '--disable-fix-cortex-a53-843419' will disable the 2824workaround by default. The workaround is disabled by default if neither 2825of '--enable-fix-cortex-a53-843419' or '--disable-fix-cortex-a53-843419' 2826is given at configure time. 2827 2828 To enable Branch Target Identification Mechanism and Return Address 2829Signing by default at configure time use the 2830'--enable-standard-branch-protection' option. This is equivalent to 2831having '-mbranch-protection=standard' during compilation. This can be 2832explicitly disabled during compilation by passing the 2833'-mbranch-protection=none' option which turns off all types of branch 2834protections. Conversely, '--disable-standard-branch-protection' will 2835disable both the protections by default. This mechanism is turned off 2836by default if neither of the options are given at configure time. 2837 2838alpha*-*-* 2839========== 2840 2841This section contains general configuration information for all 2842Alpha-based platforms using ELF. In addition to reading this section, 2843please read all other sections that match your target. 2844 2845 We require binutils 2.11.2 or newer. Previous binutils releases had 2846a number of problems with DWARF 2 debugging information, not the least 2847of which is incorrect linking of shared libraries. 2848 2849amd64-*-solaris2.1[0-9]* 2850======================== 2851 2852This is a synonym for 'x86_64-*-solaris2.1[0-9]*'. 2853 2854amdgcn-unknown-amdhsa 2855===================== 2856 2857AMD GCN GPU target. 2858 2859 Instead of GNU Binutils, you will need to install LLVM 6, or later, 2860and copy 'bin/llvm-mc' to 'amdgcn-unknown-amdhsa/bin/as', 'bin/lld' to 2861'amdgcn-unknown-amdhsa/bin/ld', 'bin/llvm-nm' to 2862'amdgcn-unknown-amdhsa/bin/nm', and 'bin/llvm-ar' to both 2863'bin/amdgcn-unknown-amdhsa-ar' and 'bin/amdgcn-unknown-amdhsa-ranlib'. 2864 2865 Use Newlib (2019-01-16, or newer). 2866 2867 To run the binaries, install the HSA Runtime from the ROCm Platform, 2868and use 'libexec/gcc/amdhsa-unknown-amdhsa/VERSION/gcn-run' to launch 2869them on the GPU. 2870 2871arc-*-elf32 2872=========== 2873 2874Use 'configure --target=arc-elf32 --with-cpu=CPU 2875--enable-languages="c,c++"' to configure GCC, with CPU being one of 2876'arc600', 'arc601', or 'arc700'. 2877 2878arc-linux-uclibc 2879================ 2880 2881Use 'configure --target=arc-linux-uclibc --with-cpu=arc700 2882--enable-languages="c,c++"' to configure GCC. 2883 2884arm-*-eabi 2885========== 2886 2887ARM-family processors. 2888 2889 Building the Ada frontend commonly fails (an infinite loop executing 2890'xsinfo') if the host compiler is GNAT 4.8. Host compilers built from 2891the GNAT 4.6, 4.9 or 5 release branches are known to succeed. 2892 2893avr 2894=== 2895 2896ATMEL AVR-family micro controllers. These are used in embedded 2897applications. There are no standard Unix configurations. *Note AVR 2898Options: (gcc)AVR Options, for the list of supported MCU types. 2899 2900 Use 'configure --target=avr --enable-languages="c"' to configure GCC. 2901 2902 Further installation notes and other useful information about AVR 2903tools can also be obtained from: 2904 2905 * http://www.nongnu.org/avr/ 2906 * http://www.amelek.gda.pl/avr/ 2907 2908 The following error: 2909 Error: register required 2910 2911 indicates that you should upgrade to a newer version of the binutils. 2912 2913Blackfin 2914======== 2915 2916The Blackfin processor, an Analog Devices DSP. *Note Blackfin Options: 2917(gcc)Blackfin Options, 2918 2919 More information, and a version of binutils with support for this 2920processor, is available at <https://blackfin.uclinux.org> 2921 2922CR16 2923==== 2924 2925The CR16 CompactRISC architecture is a 16-bit architecture. This 2926architecture is used in embedded applications. 2927 2928 *Note CR16 Options: (gcc)CR16 Options, 2929 2930 Use 'configure --target=cr16-elf --enable-languages=c,c++' to 2931configure GCC for building a CR16 elf cross-compiler. 2932 2933 Use 'configure --target=cr16-uclinux --enable-languages=c,c++' to 2934configure GCC for building a CR16 uclinux cross-compiler. 2935 2936CRIS 2937==== 2938 2939CRIS is the CPU architecture in Axis Communications ETRAX 2940system-on-a-chip series. These are used in embedded applications. 2941 2942 *Note CRIS Options: (gcc)CRIS Options, for a list of CRIS-specific 2943options. 2944 2945 There are a few different CRIS targets: 2946'cris-axis-elf' 2947 Mainly for monolithic embedded systems. Includes a multilib for 2948 the 'v10' core used in 'ETRAX 100 LX'. 2949'cris-axis-linux-gnu' 2950 A GNU/Linux port for the CRIS architecture, currently targeting 2951 'ETRAX 100 LX' by default. 2952 2953 Pre-packaged tools can be obtained from 2954<ftp://ftp.axis.com/pub/axis/tools/cris/compiler-kit/>. More 2955information about this platform is available at 2956<http://developer.axis.com/>. 2957 2958DOS 2959=== 2960 2961Please have a look at the binaries page. 2962 2963 You cannot install GCC by itself on MSDOS; it will not compile under 2964any MSDOS compiler except itself. You need to get the complete 2965compilation package DJGPP, which includes binaries as well as sources, 2966and includes all the necessary compilation tools and libraries. 2967 2968epiphany-*-elf 2969============== 2970 2971Adapteva Epiphany. This configuration is intended for embedded systems. 2972 2973*-*-freebsd* 2974============ 2975 2976Support for FreeBSD 1 was discontinued in GCC 3.2. Support for FreeBSD 29772 (and any mutant a.out variants of FreeBSD 3) was discontinued in GCC 29784.0. 2979 2980 In order to better utilize FreeBSD base system functionality and 2981match the configuration of the system compiler, GCC 4.5 and above as 2982well as GCC 4.4 past 2010-06-20 leverage SSP support in libc (which is 2983present on FreeBSD 7 or later) and the use of '__cxa_atexit' by default 2984(on FreeBSD 6 or later). The use of 'dl_iterate_phdr' inside 2985'libgcc_s.so.1' and boehm-gc (on FreeBSD 7 or later) is enabled by GCC 29864.5 and above. 2987 2988 We support FreeBSD using the ELF file format with DWARF 2 debugging 2989for all CPU architectures. You may use '-gstabs' instead of '-g', if 2990you really want the old debugging format. There are no known issues 2991with mixing object files and libraries with different debugging formats. 2992Otherwise, this release of GCC should now match more of the 2993configuration used in the stock FreeBSD configuration of GCC. In 2994particular, '--enable-threads' is now configured by default. However, 2995as a general user, do not attempt to replace the system compiler with 2996this release. Known to bootstrap and check with good results on FreeBSD 29977.2-STABLE. In the past, known to bootstrap and check with good results 2998on FreeBSD 3.0, 3.4, 4.0, 4.2, 4.3, 4.4, 4.5, 4.8, 4.9 and 5-CURRENT. 2999 3000 The version of binutils installed in '/usr/bin' probably works with 3001this release of GCC. Bootstrapping against the latest GNU binutils 3002and/or the version found in '/usr/ports/devel/binutils' has been known 3003to enable additional features and improve overall testsuite results. 3004However, it is currently known that boehm-gc may not configure properly 3005on FreeBSD prior to the FreeBSD 7.0 release with GNU binutils after 30062.16.1. 3007 3008ft32-*-elf 3009========== 3010 3011The FT32 processor. This configuration is intended for embedded 3012systems. 3013 3014h8300-hms 3015========= 3016 3017Renesas H8/300 series of processors. 3018 3019 Please have a look at the binaries page. 3020 3021 The calling convention and structure layout has changed in release 30222.6. All code must be recompiled. The calling convention now passes 3023the first three arguments in function calls in registers. Structures 3024are no longer a multiple of 2 bytes. 3025 3026hppa*-hp-hpux* 3027============== 3028 3029Support for HP-UX version 9 and older was discontinued in GCC 3.4. 3030 3031 We require using gas/binutils on all hppa platforms. Version 2.19 or 3032later is recommended. 3033 3034 It may be helpful to configure GCC with the '--with-gnu-as' and 3035'--with-as=...' options to ensure that GCC can find GAS. 3036 3037 The HP assembler should not be used with GCC. It is rarely tested and 3038may not work. It shouldn't be used with any languages other than C due 3039to its many limitations. 3040 3041 Specifically, '-g' does not work (HP-UX uses a peculiar debugging 3042format which GCC does not know about). It also inserts timestamps into 3043each object file it creates, causing the 3-stage comparison test to fail 3044during a bootstrap. You should be able to continue by saying 'make 3045all-host all-target' after getting the failure from 'make'. 3046 3047 Various GCC features are not supported. For example, it does not 3048support weak symbols or alias definitions. As a result, explicit 3049template instantiations are required when using C++. This makes it 3050difficult if not impossible to build many C++ applications. 3051 3052 There are two default scheduling models for instructions. These are 3053PROCESSOR_7100LC and PROCESSOR_8000. They are selected from the pa-risc 3054architecture specified for the target machine when configuring. 3055PROCESSOR_8000 is the default. PROCESSOR_7100LC is selected when the 3056target is a 'hppa1*' machine. 3057 3058 The PROCESSOR_8000 model is not well suited to older processors. 3059Thus, it is important to completely specify the machine architecture 3060when configuring if you want a model other than PROCESSOR_8000. The 3061macro TARGET_SCHED_DEFAULT can be defined in BOOT_CFLAGS if a different 3062default scheduling model is desired. 3063 3064 As of GCC 4.0, GCC uses the UNIX 95 namespace for HP-UX 10.10 through 306511.00, and the UNIX 98 namespace for HP-UX 11.11 and later. This 3066namespace change might cause problems when bootstrapping with an earlier 3067version of GCC or the HP compiler as essentially the same namespace is 3068required for an entire build. This problem can be avoided in a number 3069of ways. With HP cc, 'UNIX_STD' can be set to '95' or '98'. Another 3070way is to add an appropriate set of predefines to 'CC'. The description 3071for the 'munix=' option contains a list of the predefines used with each 3072standard. 3073 3074 More specific information to 'hppa*-hp-hpux*' targets follows. 3075 3076hppa*-hp-hpux10 3077=============== 3078 3079For hpux10.20, we _highly_ recommend you pick up the latest sed patch 3080'PHCO_19798' from HP. 3081 3082 The C++ ABI has changed incompatibly in GCC 4.0. COMDAT subspaces 3083are used for one-only code and data. This resolves many of the previous 3084problems in using C++ on this target. However, the ABI is not 3085compatible with the one implemented under HP-UX 11 using secondary 3086definitions. 3087 3088hppa*-hp-hpux11 3089=============== 3090 3091GCC 3.0 and up support HP-UX 11. GCC 2.95.x is not supported and cannot 3092be used to compile GCC 3.0 and up. 3093 3094 The libffi library haven't been ported to 64-bit HP-UX and doesn't 3095build. 3096 3097 Refer to binaries for information about obtaining precompiled GCC 3098binaries for HP-UX. Precompiled binaries must be obtained to build the 3099Ada language as it cannot be bootstrapped using C. Ada is only 3100available for the 32-bit PA-RISC runtime. 3101 3102 Starting with GCC 3.4 an ISO C compiler is required to bootstrap. 3103The bundled compiler supports only traditional C; you will need either 3104HP's unbundled compiler, or a binary distribution of GCC. 3105 3106 It is possible to build GCC 3.3 starting with the bundled HP 3107compiler, but the process requires several steps. GCC 3.3 can then be 3108used to build later versions. 3109 3110 There are several possible approaches to building the distribution. 3111Binutils can be built first using the HP tools. Then, the GCC 3112distribution can be built. The second approach is to build GCC first 3113using the HP tools, then build binutils, then rebuild GCC. There have 3114been problems with various binary distributions, so it is best not to 3115start from a binary distribution. 3116 3117 On 64-bit capable systems, there are two distinct targets. Different 3118installation prefixes must be used if both are to be installed on the 3119same system. The 'hppa[1-2]*-hp-hpux11*' target generates code for the 312032-bit PA-RISC runtime architecture and uses the HP linker. The 3121'hppa64-hp-hpux11*' target generates 64-bit code for the PA-RISC 2.0 3122architecture. 3123 3124 The script config.guess now selects the target type based on the 3125compiler detected during configuration. You must define 'PATH' or 'CC' 3126so that configure finds an appropriate compiler for the initial 3127bootstrap. When 'CC' is used, the definition should contain the options 3128that are needed whenever 'CC' is used. 3129 3130 Specifically, options that determine the runtime architecture must be 3131in 'CC' to correctly select the target for the build. It is also 3132convenient to place many other compiler options in 'CC'. For example, 3133'CC="cc -Ac +DA2.0W -Wp,-H16376 -D_CLASSIC_TYPES -D_HPUX_SOURCE"' can be 3134used to bootstrap the GCC 3.3 branch with the HP compiler in 64-bit 3135K&R/bundled mode. The '+DA2.0W' option will result in the automatic 3136selection of the 'hppa64-hp-hpux11*' target. The macro definition table 3137of cpp needs to be increased for a successful build with the HP 3138compiler. _CLASSIC_TYPES and _HPUX_SOURCE need to be defined when 3139building with the bundled compiler, or when using the '-Ac' option. 3140These defines aren't necessary with '-Ae'. 3141 3142 It is best to explicitly configure the 'hppa64-hp-hpux11*' target 3143with the '--with-ld=...' option. This overrides the standard search for 3144ld. The two linkers supported on this target require different 3145commands. The default linker is determined during configuration. As a 3146result, it's not possible to switch linkers in the middle of a GCC 3147build. This has been reported to sometimes occur in unified builds of 3148binutils and GCC. 3149 3150 A recent linker patch must be installed for the correct operation of 3151GCC 3.3 and later. 'PHSS_26559' and 'PHSS_24304' are the oldest linker 3152patches that are known to work. They are for HP-UX 11.00 and 11.11, 3153respectively. 'PHSS_24303', the companion to 'PHSS_24304', might be 3154usable but it hasn't been tested. These patches have been superseded. 3155Consult the HP patch database to obtain the currently recommended linker 3156patch for your system. 3157 3158 The patches are necessary for the support of weak symbols on the 315932-bit port, and for the running of initializers and finalizers. Weak 3160symbols are implemented using SOM secondary definition symbols. Prior 3161to HP-UX 11, there are bugs in the linker support for secondary symbols. 3162The patches correct a problem of linker core dumps creating shared 3163libraries containing secondary symbols, as well as various other linking 3164issues involving secondary symbols. 3165 3166 GCC 3.3 uses the ELF DT_INIT_ARRAY and DT_FINI_ARRAY capabilities to 3167run initializers and finalizers on the 64-bit port. The 32-bit port 3168uses the linker '+init' and '+fini' options for the same purpose. The 3169patches correct various problems with the +init/+fini options, including 3170program core dumps. Binutils 2.14 corrects a problem on the 64-bit port 3171resulting from HP's non-standard use of the .init and .fini sections for 3172array initializers and finalizers. 3173 3174 Although the HP and GNU linkers are both supported for the 3175'hppa64-hp-hpux11*' target, it is strongly recommended that the HP 3176linker be used for link editing on this target. 3177 3178 At this time, the GNU linker does not support the creation of long 3179branch stubs. As a result, it cannot successfully link binaries 3180containing branch offsets larger than 8 megabytes. In addition, there 3181are problems linking shared libraries, linking executables with 3182'-static', and with dwarf2 unwind and exception support. It also 3183doesn't provide stubs for internal calls to global functions in shared 3184libraries, so these calls cannot be overloaded. 3185 3186 The HP dynamic loader does not support GNU symbol versioning, so 3187symbol versioning is not supported. It may be necessary to disable 3188symbol versioning with '--disable-symvers' when using GNU ld. 3189 3190 POSIX threads are the default. The optional DCE thread library is 3191not supported, so '--enable-threads=dce' does not work. 3192 3193*-*-linux-gnu 3194============= 3195 3196Versions of libstdc++-v3 starting with 3.2.1 require bug fixes present 3197in glibc 2.2.5 and later. More information is available in the 3198libstdc++-v3 documentation. 3199 3200i?86-*-linux* 3201============= 3202 3203As of GCC 3.3, binutils 2.13.1 or later is required for this platform. 3204See bug 10877 for more information. 3205 3206 If you receive Signal 11 errors when building on GNU/Linux, then it 3207is possible you have a hardware problem. Further information on this 3208can be found on www.bitwizard.nl. 3209 3210i?86-*-solaris2.10 3211================== 3212 3213Use this for Solaris 10 or later on x86 and x86-64 systems. Starting 3214with GCC 4.7, there is also a 64-bit 'amd64-*-solaris2.1[0-9]*' or 3215'x86_64-*-solaris2.1[0-9]*' configuration that corresponds to 3216'sparcv9-sun-solaris2*'. 3217 3218 It is recommended that you configure GCC to use the GNU assembler. 3219The versions included in Solaris 10, from GNU binutils 2.15 (in 3220'/usr/sfw/bin/gas'), and Solaris 11, from GNU binutils 2.19 or newer 3221(also available as '/usr/bin/gas' and '/usr/gnu/bin/as'), work fine. 3222The current version, from GNU binutils 2.29, is known to work, but the 3223version from GNU binutils 2.26 must be avoided. Recent versions of the 3224Solaris assembler in '/usr/ccs/bin/as' work almost as well, though. 3225 3226 For linking, the Solaris linker, is preferred. If you want to use 3227the GNU linker instead, note that due to a packaging bug the version in 3228Solaris 10, from GNU binutils 2.15 (in '/usr/sfw/bin/gld'), cannot be 3229used, while the version in Solaris 11, from GNU binutils 2.19 or newer 3230(also in '/usr/gnu/bin/ld' and '/usr/bin/gld'), works, as does the 3231latest version, from GNU binutils 2.29. 3232 3233 To use GNU 'as', configure with the options '--with-gnu-as 3234--with-as=/usr/sfw/bin/gas'. It may be necessary to configure with 3235'--without-gnu-ld --with-ld=/usr/ccs/bin/ld' to guarantee use of Sun 3236'ld'. 3237 3238ia64-*-linux 3239============ 3240 3241IA-64 processor (also known as IPF, or Itanium Processor Family) running 3242GNU/Linux. 3243 3244 If you are using the installed system libunwind library with 3245'--with-system-libunwind', then you must use libunwind 0.98 or later. 3246 3247 None of the following versions of GCC has an ABI that is compatible 3248with any of the other versions in this list, with the exception that Red 3249Hat 2.96 and Trillian 000171 are compatible with each other: 3.1, 3.0.2, 32503.0.1, 3.0, Red Hat 2.96, and Trillian 000717. This primarily affects 3251C++ programs and programs that create shared libraries. GCC 3.1 or 3252later is recommended for compiling linux, the kernel. As of version 3.1 3253GCC is believed to be fully ABI compliant, and hence no more major ABI 3254changes are expected. 3255 3256ia64-*-hpux* 3257============ 3258 3259Building GCC on this target requires the GNU Assembler. The bundled HP 3260assembler will not work. To prevent GCC from using the wrong assembler, 3261the option '--with-gnu-as' may be necessary. 3262 3263 The GCC libunwind library has not been ported to HPUX. This means 3264that for GCC versions 3.2.3 and earlier, '--enable-libunwind-exceptions' 3265is required to build GCC. For GCC 3.3 and later, this is the default. 3266For gcc 3.4.3 and later, '--enable-libunwind-exceptions' is removed and 3267the system libunwind library will always be used. 3268 3269*-ibm-aix* 3270========== 3271 3272Support for AIX version 3 and older was discontinued in GCC 3.4. 3273Support for AIX version 4.2 and older was discontinued in GCC 4.5. 3274 3275 "out of memory" bootstrap failures may indicate a problem with 3276process resource limits (ulimit). Hard limits are configured in the 3277'/etc/security/limits' system configuration file. 3278 3279 GCC 4.9 and above require a C++ compiler for bootstrap. IBM VAC++ / 3280xlC cannot bootstrap GCC. xlc can bootstrap an older version of GCC and 3281G++ can bootstrap recent releases of GCC. 3282 3283 GCC can bootstrap with recent versions of IBM XLC, but bootstrapping 3284with an earlier release of GCC is recommended. Bootstrapping with XLC 3285requires a larger data segment, which can be enabled through the 3286LDR_CNTRL environment variable, e.g., 3287 3288 % LDR_CNTRL=MAXDATA=0x50000000 3289 % export LDR_CNTRL 3290 3291 One can start with a pre-compiled version of GCC to build from 3292sources. One may delete GCC's "fixed" header files when starting with a 3293version of GCC built for an earlier release of AIX. 3294 3295 To speed up the configuration phases of bootstrapping and installing 3296GCC, one may use GNU Bash instead of AIX '/bin/sh', e.g., 3297 3298 % CONFIG_SHELL=/opt/freeware/bin/bash 3299 % export CONFIG_SHELL 3300 3301 and then proceed as described in the build instructions, where we 3302strongly recommend specifying an absolute path to invoke 3303SRCDIR/configure. 3304 3305 Because GCC on AIX is built as a 32-bit executable by default, 3306(although it can generate 64-bit programs) the GMP and MPFR libraries 3307required by gfortran must be 32-bit libraries. Building GMP and MPFR as 3308static archive libraries works better than shared libraries. 3309 3310 Errors involving 'alloca' when building GCC generally are due to an 3311incorrect definition of 'CC' in the Makefile or mixing files compiled 3312with the native C compiler and GCC. During the stage1 phase of the 3313build, the native AIX compiler *must* be invoked as 'cc' (not 'xlc'). 3314Once 'configure' has been informed of 'xlc', one needs to use 'make 3315distclean' to remove the configure cache files and ensure that 'CC' 3316environment variable does not provide a definition that will confuse 3317'configure'. If this error occurs during stage2 or later, then the 3318problem most likely is the version of Make (see above). 3319 3320 The native 'as' and 'ld' are recommended for bootstrapping on AIX. 3321The GNU Assembler, GNU Linker, and GNU Binutils version 2.20 is the 3322minimum level that supports bootstrap on AIX 5. The GNU Assembler has 3323not been updated to support AIX 6 or AIX 7. The native AIX tools do 3324interoperate with GCC. 3325 3326 AIX 7.1 added partial support for DWARF debugging, but full support 3327requires AIX 7.1 TL03 SP7 that supports additional DWARF sections and 3328fixes a bug in the assembler. AIX 7.1 TL03 SP5 distributed a version of 3329libm.a missing important symbols; a fix for IV77796 will be included in 3330SP6. 3331 3332 AIX 5.3 TL10, AIX 6.1 TL05 and AIX 7.1 TL00 introduced an AIX 3333assembler change that sometimes produces corrupt assembly files causing 3334AIX linker errors. The bug breaks GCC bootstrap on AIX and can cause 3335compilation failures with existing GCC installations. An AIX iFix for 3336AIX 5.3 is available (APAR IZ98385 for AIX 5.3 TL10, APAR IZ98477 for 3337AIX 5.3 TL11 and IZ98134 for AIX 5.3 TL12). AIX 5.3 TL11 SP8, AIX 5.3 3338TL12 SP5, AIX 6.1 TL04 SP11, AIX 6.1 TL05 SP7, AIX 6.1 TL06 SP6, AIX 6.1 3339TL07 and AIX 7.1 TL01 should include the fix. 3340 3341 Building 'libstdc++.a' requires a fix for an AIX Assembler bug APAR 3342IY26685 (AIX 4.3) or APAR IY25528 (AIX 5.1). It also requires a fix for 3343another AIX Assembler bug and a co-dependent AIX Archiver fix referenced 3344as APAR IY53606 (AIX 5.2) or as APAR IY54774 (AIX 5.1) 3345 3346 'libstdc++' in GCC 3.4 increments the major version number of the 3347shared object and GCC installation places the 'libstdc++.a' shared 3348library in a common location which will overwrite the and GCC 3.3 3349version of the shared library. Applications either need to be re-linked 3350against the new shared library or the GCC 3.1 and GCC 3.3 versions of 3351the 'libstdc++' shared object needs to be available to the AIX runtime 3352loader. The GCC 3.1 'libstdc++.so.4', if present, and GCC 3.3 3353'libstdc++.so.5' shared objects can be installed for runtime dynamic 3354loading using the following steps to set the 'F_LOADONLY' flag in the 3355shared object for _each_ multilib 'libstdc++.a' installed: 3356 3357 Extract the shared objects from the currently installed 'libstdc++.a' 3358archive: 3359 % ar -x libstdc++.a libstdc++.so.4 libstdc++.so.5 3360 3361 Enable the 'F_LOADONLY' flag so that the shared object will be 3362available for runtime dynamic loading, but not linking: 3363 % strip -e libstdc++.so.4 libstdc++.so.5 3364 3365 Archive the runtime-only shared object in the GCC 3.4 'libstdc++.a' 3366archive: 3367 % ar -q libstdc++.a libstdc++.so.4 libstdc++.so.5 3368 3369 Eventually, the '--with-aix-soname=svr4' configure option may drop 3370the need for this procedure for libraries that support it. 3371 3372 Linking executables and shared libraries may produce warnings of 3373duplicate symbols. The assembly files generated by GCC for AIX always 3374have included multiple symbol definitions for certain global variable 3375and function declarations in the original program. The warnings should 3376not prevent the linker from producing a correct library or runnable 3377executable. 3378 3379 AIX 4.3 utilizes a "large format" archive to support both 32-bit and 338064-bit object modules. The routines provided in AIX 4.3.0 and AIX 4.3.1 3381to parse archive libraries did not handle the new format correctly. 3382These routines are used by GCC and result in error messages during 3383linking such as "not a COFF file". The version of the routines shipped 3384with AIX 4.3.1 should work for a 32-bit environment. The '-g' option of 3385the archive command may be used to create archives of 32-bit objects 3386using the original "small format". A correct version of the routines is 3387shipped with AIX 4.3.2 and above. 3388 3389 Some versions of the AIX binder (linker) can fail with a relocation 3390overflow severe error when the '-bbigtoc' option is used to link 3391GCC-produced object files into an executable that overflows the TOC. A 3392fix for APAR IX75823 (OVERFLOW DURING LINK WHEN USING GCC AND -BBIGTOC) 3393is available from IBM Customer Support and from its 3394techsupport.services.ibm.com website as PTF U455193. 3395 3396 The AIX 4.3.2.1 linker (bos.rte.bind_cmds Level 4.3.2.1) will dump 3397core with a segmentation fault when invoked by any version of GCC. A 3398fix for APAR IX87327 is available from IBM Customer Support and from its 3399techsupport.services.ibm.com website as PTF U461879. This fix is 3400incorporated in AIX 4.3.3 and above. 3401 3402 The initial assembler shipped with AIX 4.3.0 generates incorrect 3403object files. A fix for APAR IX74254 (64BIT DISASSEMBLED OUTPUT FROM 3404COMPILER FAILS TO ASSEMBLE/BIND) is available from IBM Customer Support 3405and from its techsupport.services.ibm.com website as PTF U453956. This 3406fix is incorporated in AIX 4.3.1 and above. 3407 3408 AIX provides National Language Support (NLS). Compilers and 3409assemblers use NLS to support locale-specific representations of various 3410data formats including floating-point numbers (e.g., '.' vs ',' for 3411separating decimal fractions). There have been problems reported where 3412GCC does not produce the same floating-point formats that the assembler 3413expects. If one encounters this problem, set the 'LANG' environment 3414variable to 'C' or 'En_US'. 3415 3416 A default can be specified with the '-mcpu=CPU_TYPE' switch and using 3417the configure option '--with-cpu-CPU_TYPE'. 3418 3419iq2000-*-elf 3420============ 3421 3422Vitesse IQ2000 processors. These are used in embedded applications. 3423There are no standard Unix configurations. 3424 3425lm32-*-elf 3426========== 3427 3428Lattice Mico32 processor. This configuration is intended for embedded 3429systems. 3430 3431lm32-*-uclinux 3432============== 3433 3434Lattice Mico32 processor. This configuration is intended for embedded 3435systems running uClinux. 3436 3437m32c-*-elf 3438========== 3439 3440Renesas M32C processor. This configuration is intended for embedded 3441systems. 3442 3443m32r-*-elf 3444========== 3445 3446Renesas M32R processor. This configuration is intended for embedded 3447systems. 3448 3449m68k-*-* 3450======== 3451 3452By default, 'm68k-*-elf*', 'm68k-*-rtems', 'm68k-*-uclinux' and 3453'm68k-*-linux' build libraries for both M680x0 and ColdFire processors. 3454If you only need the M680x0 libraries, you can omit the ColdFire ones by 3455passing '--with-arch=m68k' to 'configure'. Alternatively, you can omit 3456the M680x0 libraries by passing '--with-arch=cf' to 'configure'. These 3457targets default to 5206 or 5475 code as appropriate for the target 3458system when configured with '--with-arch=cf' and 68020 code otherwise. 3459 3460 The 'm68k-*-netbsd' and 'm68k-*-openbsd' targets also support the 3461'--with-arch' option. They will generate ColdFire CFV4e code when 3462configured with '--with-arch=cf' and 68020 code otherwise. 3463 3464 You can override the default processors listed above by configuring 3465with '--with-cpu=TARGET'. This TARGET can either be a '-mcpu' argument 3466or one of the following values: 'm68000', 'm68010', 'm68020', 'm68030', 3467'm68040', 'm68060', 'm68020-40' and 'm68020-60'. 3468 3469 GCC requires at least binutils version 2.17 on these targets. 3470 3471m68k-*-uclinux 3472============== 3473 3474GCC 4.3 changed the uClinux configuration so that it uses the 3475'm68k-linux-gnu' ABI rather than the 'm68k-elf' ABI. It also added 3476improved support for C++ and flat shared libraries, both of which were 3477ABI changes. 3478 3479microblaze-*-elf 3480================ 3481 3482Xilinx MicroBlaze processor. This configuration is intended for 3483embedded systems. 3484 3485mips-*-* 3486======== 3487 3488If on a MIPS system you get an error message saying "does not have gp 3489sections for all it's [sic] sectons [sic]", don't worry about it. This 3490happens whenever you use GAS with the MIPS linker, but there is not 3491really anything wrong, and it is okay to use the output file. You can 3492stop such warnings by installing the GNU linker. 3493 3494 It would be nice to extend GAS to produce the gp tables, but they are 3495optional, and there should not be a warning about their absence. 3496 3497 The libstdc++ atomic locking routines for MIPS targets requires MIPS 3498II and later. A patch went in just after the GCC 3.3 release to make 3499'mips*-*-*' use the generic implementation instead. You can also 3500configure for 'mipsel-elf' as a workaround. The 'mips*-*-linux*' target 3501continues to use the MIPS II routines. More work on this is expected in 3502future releases. 3503 3504 The built-in '__sync_*' functions are available on MIPS II and later 3505systems and others that support the 'll', 'sc' and 'sync' instructions. 3506This can be overridden by passing '--with-llsc' or '--without-llsc' when 3507configuring GCC. Since the Linux kernel emulates these instructions if 3508they are missing, the default for 'mips*-*-linux*' targets is 3509'--with-llsc'. The '--with-llsc' and '--without-llsc' configure options 3510may be overridden at compile time by passing the '-mllsc' or '-mno-llsc' 3511options to the compiler. 3512 3513 MIPS systems check for division by zero (unless 3514'-mno-check-zero-division' is passed to the compiler) by generating 3515either a conditional trap or a break instruction. Using trap results in 3516smaller code, but is only supported on MIPS II and later. Also, some 3517versions of the Linux kernel have a bug that prevents trap from 3518generating the proper signal ('SIGFPE'). To enable the use of break, 3519use the '--with-divide=breaks' 'configure' option when configuring GCC. 3520The default is to use traps on systems that support them. 3521 3522moxie-*-elf 3523=========== 3524 3525The moxie processor. 3526 3527msp430-*-elf 3528============ 3529 3530TI MSP430 processor. This configuration is intended for embedded 3531systems. 3532 3533nds32le-*-elf 3534============= 3535 3536Andes NDS32 target in little endian mode. 3537 3538nds32be-*-elf 3539============= 3540 3541Andes NDS32 target in big endian mode. 3542 3543nvptx-*-none 3544============ 3545 3546Nvidia PTX target. 3547 3548 Instead of GNU binutils, you will need to install nvptx-tools. Tell 3549GCC where to find it: 3550'--with-build-time-tools=[install-nvptx-tools]/nvptx-none/bin'. 3551 3552 You will need newlib 3.0 git revision 3553cd31fbb2aea25f94d7ecedc9db16dfc87ab0c316 or later. It can be 3554automatically built together with GCC. For this, add a symbolic link to 3555nvptx-newlib's 'newlib' directory to the directory containing the GCC 3556sources. 3557 3558 Use the '--disable-sjlj-exceptions' and 3559'--enable-newlib-io-long-long' options when configuring. 3560 3561or1k-*-elf 3562========== 3563 3564The OpenRISC 1000 32-bit processor with delay slots. This configuration 3565is intended for embedded systems. 3566 3567or1k-*-linux 3568============ 3569 3570The OpenRISC 1000 32-bit processor with delay slots. 3571 3572powerpc-*-* 3573=========== 3574 3575You can specify a default version for the '-mcpu=CPU_TYPE' switch by 3576using the configure option '--with-cpu-CPU_TYPE'. 3577 3578 You will need GNU binutils 2.15 or newer. 3579 3580powerpc-*-darwin* 3581================= 3582 3583PowerPC running Darwin (Mac OS X kernel). 3584 3585 Pre-installed versions of Mac OS X may not include any developer 3586tools, meaning that you will not be able to build GCC from source. Tool 3587binaries are available at <https://opensource.apple.com>. 3588 3589 This version of GCC requires at least cctools-590.36. The 3590cctools-590.36 package referenced from 3591<http://gcc.gnu.org/ml/gcc/2006-03/msg00507.html> will not work on 3592systems older than 10.3.9 (aka darwin7.9.0). 3593 3594powerpc-*-elf 3595============= 3596 3597PowerPC system in big endian mode, running System V.4. 3598 3599powerpc*-*-linux-gnu* 3600===================== 3601 3602PowerPC system in big endian mode running Linux. 3603 3604powerpc-*-netbsd* 3605================= 3606 3607PowerPC system in big endian mode running NetBSD. 3608 3609powerpc-*-eabisim 3610================= 3611 3612Embedded PowerPC system in big endian mode for use in running under the 3613PSIM simulator. 3614 3615powerpc-*-eabi 3616============== 3617 3618Embedded PowerPC system in big endian mode. 3619 3620powerpcle-*-elf 3621=============== 3622 3623PowerPC system in little endian mode, running System V.4. 3624 3625powerpcle-*-eabisim 3626=================== 3627 3628Embedded PowerPC system in little endian mode for use in running under 3629the PSIM simulator. 3630 3631powerpcle-*-eabi 3632================ 3633 3634Embedded PowerPC system in little endian mode. 3635 3636rl78-*-elf 3637========== 3638 3639The Renesas RL78 processor. This configuration is intended for embedded 3640systems. 3641 3642riscv32-*-elf 3643============= 3644 3645The RISC-V RV32 instruction set. This configuration is intended for 3646embedded systems. This (and all other RISC-V) targets are supported 3647upstream as of the binutils 2.28 release. 3648 3649riscv32-*-linux 3650=============== 3651 3652The RISC-V RV32 instruction set running GNU/Linux. This (and all other 3653RISC-V) targets are supported upstream as of the binutils 2.28 release. 3654 3655riscv64-*-elf 3656============= 3657 3658The RISC-V RV64 instruction set. This configuration is intended for 3659embedded systems. This (and all other RISC-V) targets are supported 3660upstream as of the binutils 2.28 release. 3661 3662riscv64-*-linux 3663=============== 3664 3665The RISC-V RV64 instruction set running GNU/Linux. This (and all other 3666RISC-V) targets are supported upstream as of the binutils 2.28 release. 3667 3668rx-*-elf 3669======== 3670 3671The Renesas RX processor. 3672 3673s390-*-linux* 3674============= 3675 3676S/390 system running GNU/Linux for S/390. 3677 3678s390x-*-linux* 3679============== 3680 3681zSeries system (64-bit) running GNU/Linux for zSeries. 3682 3683s390x-ibm-tpf* 3684============== 3685 3686zSeries system (64-bit) running TPF. This platform is supported as 3687cross-compilation target only. 3688 3689*-*-solaris2* 3690============= 3691 3692Support for Solaris 10 has been obsoleted in GCC 9, but can still be 3693enabled by configuring with '--enable-obsolete'. Support will be 3694removed in GCC 10. Support for Solaris 9 has been removed in GCC 5. 3695Support for Solaris 8 has been removed in GCC 4.8. Support for Solaris 36967 has been removed in GCC 4.6. 3697 3698 Sun does not ship a C compiler with Solaris 2 before Solaris 10, 3699though you can download the Sun Studio compilers for free. In Solaris 370010 and 11, GCC 3.4.3 is available as '/usr/sfw/bin/gcc'. Solaris 11 3701also provides GCC 4.5.2, 4.7.3, and 4.8.2 as '/usr/gcc/4.5/bin/gcc' or 3702similar. Alternatively, you can install a pre-built GCC to bootstrap 3703and install GCC. See the binaries page for details. 3704 3705 The Solaris 2 '/bin/sh' will often fail to configure 'libstdc++-v3'or 3706'boehm-gc'. We therefore recommend using the following initial sequence 3707of commands 3708 3709 % CONFIG_SHELL=/bin/ksh 3710 % export CONFIG_SHELL 3711 3712and proceed as described in the configure instructions. In addition we 3713strongly recommend specifying an absolute path to invoke 3714'SRCDIR/configure'. 3715 3716 Solaris 10 comes with a number of optional OS packages. Some of 3717these are needed to use GCC fully, namely 'SUNWarc', 'SUNWbtool', 3718'SUNWesu', 'SUNWhea', 'SUNWlibm', 'SUNWsprot', and 'SUNWtoo'. If you 3719did not install all optional packages when installing Solaris 10, you 3720will need to verify that the packages that GCC needs are installed. To 3721check whether an optional package is installed, use the 'pkginfo' 3722command. To add an optional package, use the 'pkgadd' command. For 3723further details, see the Solaris 10 documentation. 3724 3725 Starting with Solaris 11, the package management has changed, so you 3726need to check for 'system/header', 'system/linker', and 3727'developer/assembler' packages. Checking for and installing packages is 3728done with the 'pkg' command now. 3729 3730 Trying to use the linker and other tools in '/usr/ucb' to install GCC 3731has been observed to cause trouble. For example, the linker may hang 3732indefinitely. The fix is to remove '/usr/ucb' from your 'PATH'. 3733 3734 The build process works more smoothly with the legacy Sun tools so, 3735if you have '/usr/xpg4/bin' in your 'PATH', we recommend that you place 3736'/usr/bin' before '/usr/xpg4/bin' for the duration of the build. 3737 3738 We recommend the use of the Solaris assembler or the GNU assembler, 3739in conjunction with the Solaris linker. The GNU 'as' versions included 3740in Solaris 10, from GNU binutils 2.15 (in '/usr/sfw/bin/gas'), and 3741Solaris 11, from GNU binutils 2.19 or newer (also in '/usr/bin/gas' and 3742'/usr/gnu/bin/as'), are known to work. The current version, from GNU 3743binutils 2.29, is known to work as well. Note that your mileage may 3744vary if you use a combination of the GNU tools and the Solaris tools: 3745while the combination GNU 'as' + Sun 'ld' should reasonably work, the 3746reverse combination Sun 'as' + GNU 'ld' may fail to build or cause 3747memory corruption at runtime in some cases for C++ programs. GNU 'ld' 3748usually works as well, although the version included in Solaris 10 3749cannot be used due to several bugs. Again, the current version (2.29) 3750is known to work, but generally lacks platform specific features, so 3751better stay with Solaris 'ld'. To use the LTO linker plugin 3752('-fuse-linker-plugin') with GNU 'ld', GNU binutils _must_ be configured 3753with '--enable-largefile'. 3754 3755 To enable symbol versioning in 'libstdc++' with the Solaris linker, 3756you need to have any version of GNU 'c++filt', which is part of GNU 3757binutils. 'libstdc++' symbol versioning will be disabled if no 3758appropriate version is found. Solaris 'c++filt' from the Solaris Studio 3759compilers does _not_ work. 3760 3761 Sun bug 4927647 sometimes causes random spurious testsuite failures 3762related to missing diagnostic output. This bug doesn't affect GCC 3763itself, rather it is a kernel bug triggered by the 'expect' program 3764which is used only by the GCC testsuite driver. When the bug causes the 3765'expect' program to miss anticipated output, extra testsuite failures 3766appear. 3767 3768sparc*-*-* 3769========== 3770 3771This section contains general configuration information for all 3772SPARC-based platforms. In addition to reading this section, please read 3773all other sections that match your target. 3774 3775 Newer versions of the GNU Multiple Precision Library (GMP), the MPFR 3776library and the MPC library are known to be miscompiled by earlier 3777versions of GCC on these platforms. We therefore recommend the use of 3778the exact versions of these libraries listed as minimal versions in the 3779prerequisites. 3780 3781sparc-sun-solaris2* 3782=================== 3783 3784When GCC is configured to use GNU binutils 2.14 or later, the binaries 3785produced are smaller than the ones produced using Sun's native tools; 3786this difference is quite significant for binaries containing debugging 3787information. 3788 3789 Starting with Solaris 7, the operating system is capable of executing 379064-bit SPARC V9 binaries. GCC 3.1 and later properly supports this; the 3791'-m64' option enables 64-bit code generation. However, if all you want 3792is code tuned for the UltraSPARC CPU, you should try the 3793'-mtune=ultrasparc' option instead, which produces code that, unlike 3794full 64-bit code, can still run on non-UltraSPARC machines. 3795 3796 When configuring the GNU Multiple Precision Library (GMP), the MPFR 3797library or the MPC library on a Solaris 7 or later system, the canonical 3798target triplet must be specified as the 'build' parameter on the 3799configure line. This target triplet can be obtained by invoking 3800'./config.guess' in the toplevel source directory of GCC (and not that 3801of GMP or MPFR or MPC). For example on a Solaris 9 system: 3802 3803 % ./configure --build=sparc-sun-solaris2.9 --prefix=xxx 3804 3805sparc-sun-solaris2.10 3806===================== 3807 3808There is a bug in older versions of the Sun assembler which breaks 3809thread-local storage (TLS). A typical error message is 3810 3811 ld: fatal: relocation error: R_SPARC_TLS_LE_HIX22: file /var/tmp//ccamPA1v.o: 3812 symbol <unknown>: bad symbol type SECT: symbol type must be TLS 3813 3814This bug is fixed in Sun patch 118683-03 or later. 3815 3816sparc-*-linux* 3817============== 3818 3819sparc64-*-solaris2* 3820=================== 3821 3822When configuring the GNU Multiple Precision Library (GMP), the MPFR 3823library or the MPC library, the canonical target triplet must be 3824specified as the 'build' parameter on the configure line. For example 3825on a Solaris 9 system: 3826 3827 % ./configure --build=sparc64-sun-solaris2.9 --prefix=xxx 3828 3829sparcv9-*-solaris2* 3830=================== 3831 3832This is a synonym for 'sparc64-*-solaris2*'. 3833 3834c6x-*-* 3835======= 3836 3837The C6X family of processors. This port requires binutils-2.22 or 3838newer. 3839 3840tilegx-*-linux* 3841=============== 3842 3843The TILE-Gx processor in little endian mode, running GNU/Linux. This 3844port requires binutils-2.22 or newer. 3845 3846tilegxbe-*-linux* 3847================= 3848 3849The TILE-Gx processor in big endian mode, running GNU/Linux. This port 3850requires binutils-2.23 or newer. 3851 3852tilepro-*-linux* 3853================ 3854 3855The TILEPro processor running GNU/Linux. This port requires 3856binutils-2.22 or newer. 3857 3858visium-*-elf 3859============ 3860 3861CDS VISIUMcore processor. This configuration is intended for embedded 3862systems. 3863 3864*-*-vxworks* 3865============ 3866 3867Support for VxWorks is in flux. At present GCC supports _only_ the very 3868recent VxWorks 5.5 (aka Tornado 2.2) release, and only on PowerPC. We 3869welcome patches for other architectures supported by VxWorks 5.5. 3870Support for VxWorks AE would also be welcome; we believe this is merely 3871a matter of writing an appropriate "configlette" (see below). We are 3872not interested in supporting older, a.out or COFF-based, versions of 3873VxWorks in GCC 3. 3874 3875 VxWorks comes with an older version of GCC installed in 3876'$WIND_BASE/host'; we recommend you do not overwrite it. Choose an 3877installation PREFIX entirely outside $WIND_BASE. Before running 3878'configure', create the directories 'PREFIX' and 'PREFIX/bin'. Link or 3879copy the appropriate assembler, linker, etc. into 'PREFIX/bin', and set 3880your PATH to include that directory while running both 'configure' and 3881'make'. 3882 3883 You must give 'configure' the '--with-headers=$WIND_BASE/target/h' 3884switch so that it can find the VxWorks system headers. Since VxWorks is 3885a cross compilation target only, you must also specify 3886'--target=TARGET'. 'configure' will attempt to create the directory 3887'PREFIX/TARGET/sys-include' and copy files into it; make sure the user 3888running 'configure' has sufficient privilege to do so. 3889 3890 GCC's exception handling runtime requires a special "configlette" 3891module, 'contrib/gthr_supp_vxw_5x.c'. Follow the instructions in that 3892file to add the module to your kernel build. (Future versions of 3893VxWorks will incorporate this module.) 3894 3895x86_64-*-*, amd64-*-* 3896===================== 3897 3898GCC supports the x86-64 architecture implemented by the AMD64 processor 3899(amd64-*-* is an alias for x86_64-*-*) on GNU/Linux, FreeBSD and NetBSD. 3900On GNU/Linux the default is a bi-arch compiler which is able to generate 3901both 64-bit x86-64 and 32-bit x86 code (via the '-m32' switch). 3902 3903x86_64-*-solaris2.1[0-9]* 3904========================= 3905 3906GCC also supports the x86-64 architecture implemented by the AMD64 3907processor ('amd64-*-*' is an alias for 'x86_64-*-*') on Solaris 10 or 3908later. Unlike other systems, without special options a bi-arch compiler 3909is built which generates 32-bit code by default, but can generate 64-bit 3910x86-64 code with the '-m64' switch. Since GCC 4.7, there is also a 3911configuration that defaults to 64-bit code, but can generate 32-bit code 3912with '-m32'. To configure and build this way, you have to provide all 3913support libraries like 'libgmp' as 64-bit code, configure with 3914'--target=x86_64-pc-solaris2.1x' and 'CC=gcc -m64'. 3915 3916xtensa*-*-elf 3917============= 3918 3919This target is intended for embedded Xtensa systems using the 'newlib' C 3920library. It uses ELF but does not support shared objects. 3921Designed-defined instructions specified via the Tensilica Instruction 3922Extension (TIE) language are only supported through inline assembly. 3923 3924 The Xtensa configuration information must be specified prior to 3925building GCC. The 'include/xtensa-config.h' header file contains the 3926configuration information. If you created your own Xtensa configuration 3927with the Xtensa Processor Generator, the downloaded files include a 3928customized copy of this header file, which you can use to replace the 3929default header file. 3930 3931xtensa*-*-linux* 3932================ 3933 3934This target is for Xtensa systems running GNU/Linux. It supports ELF 3935shared objects and the GNU C library (glibc). It also generates 3936position-independent code (PIC) regardless of whether the '-fpic' or 3937'-fPIC' options are used. In other respects, this target is the same as 3938the 'xtensa*-*-elf' target. 3939 3940Microsoft Windows 3941================= 3942 3943Intel 16-bit versions 3944--------------------- 3945 3946The 16-bit versions of Microsoft Windows, such as Windows 3.1, are not 3947supported. 3948 3949 However, the 32-bit port has limited support for Microsoft Windows 39503.11 in the Win32s environment, as a target only. See below. 3951 3952Intel 32-bit versions 3953--------------------- 3954 3955The 32-bit versions of Windows, including Windows 95, Windows NT, 3956Windows XP, and Windows Vista, are supported by several different target 3957platforms. These targets differ in which Windows subsystem they target 3958and which C libraries are used. 3959 3960 * Cygwin *-*-cygwin: Cygwin provides a user-space Linux API emulation 3961 layer in the Win32 subsystem. 3962 * MinGW *-*-mingw32: MinGW is a native GCC port for the Win32 3963 subsystem that provides a subset of POSIX. 3964 * MKS i386-pc-mks: NuTCracker from MKS. See 3965 <https://www.mkssoftware.com> for more information. 3966 3967Intel 64-bit versions 3968--------------------- 3969 3970GCC contains support for x86-64 using the mingw-w64 runtime library, 3971available from <http://mingw-w64.org/doku.php>. This library should be 3972used with the target triple x86_64-pc-mingw32. 3973 3974 Presently Windows for Itanium is not supported. 3975 3976Windows CE 3977---------- 3978 3979Windows CE is supported as a target only on Hitachi SuperH 3980(sh-wince-pe), and MIPS (mips-wince-pe). 3981 3982Other Windows Platforms 3983----------------------- 3984 3985GCC no longer supports Windows NT on the Alpha or PowerPC. 3986 3987 GCC no longer supports the Windows POSIX subsystem. However, it does 3988support the Interix subsystem. See above. 3989 3990 Old target names including *-*-winnt and *-*-windowsnt are no longer 3991used. 3992 3993 PW32 (i386-pc-pw32) support was never completed, and the project 3994seems to be inactive. See <http://pw32.sourceforge.net/> for more 3995information. 3996 3997 UWIN support has been removed due to a lack of maintenance. 3998 3999*-*-cygwin 4000========== 4001 4002Ports of GCC are included with the Cygwin environment. 4003 4004 GCC will build under Cygwin without modification; it does not build 4005with Microsoft's C++ compiler and there are no plans to make it do so. 4006 4007 The Cygwin native compiler can be configured to target any 32-bit x86 4008cpu architecture desired; the default is i686-pc-cygwin. It should be 4009used with as up-to-date a version of binutils as possible; use either 4010the latest official GNU binutils release in the Cygwin distribution, or 4011version 2.20 or above if building your own. 4012 4013*-*-mingw32 4014=========== 4015 4016GCC will build with and support only MinGW runtime 3.12 and later. 4017Earlier versions of headers are incompatible with the new default 4018semantics of 'extern inline' in '-std=c99' and '-std=gnu99' modes. 4019 4020Older systems 4021============= 4022 4023GCC contains support files for many older (1980s and early 1990s) Unix 4024variants. For the most part, support for these systems has not been 4025deliberately removed, but it has not been maintained for several years 4026and may suffer from bitrot. 4027 4028 Starting with GCC 3.1, each release has a list of "obsoleted" 4029systems. Support for these systems is still present in that release, 4030but 'configure' will fail unless the '--enable-obsolete' option is 4031given. Unless a maintainer steps forward, support for these systems 4032will be removed from the next release of GCC. 4033 4034 Support for old systems as hosts for GCC can cause problems if the 4035workarounds for compiler, library and operating system bugs affect the 4036cleanliness or maintainability of the rest of GCC. In some cases, to 4037bring GCC up on such a system, if still possible with current GCC, may 4038require first installing an old version of GCC which did work on that 4039system, and using it to compile a more recent GCC, to avoid bugs in the 4040vendor compiler. Old releases of GCC 1 and GCC 2 are available in the 4041'old-releases' directory on the GCC mirror sites. Header bugs may 4042generally be avoided using 'fixincludes', but bugs or deficiencies in 4043libraries and the operating system may still cause problems. 4044 4045 Support for older systems as targets for cross-compilation is less 4046problematic than support for them as hosts for GCC; if an enthusiast 4047wishes to make such a target work again (including resurrecting any of 4048the targets that never worked with GCC 2, starting from the last version 4049before they were removed), patches following the usual requirements 4050would be likely to be accepted, since they should not affect the support 4051for more modern targets. 4052 4053 For some systems, old versions of GNU binutils may also be useful, 4054and are available from 'pub/binutils/old-releases' on sourceware.org 4055mirror sites. 4056 4057 Some of the information on specific systems above relates to such 4058older systems, but much of the information about GCC on such systems 4059(which may no longer be applicable to current GCC) is to be found in the 4060GCC texinfo manual. 4061 4062all ELF targets (SVR4, Solaris 2, etc.) 4063======================================= 4064 4065C++ support is significantly better on ELF targets if you use the GNU 4066linker; duplicate copies of inlines, vtables and template instantiations 4067will be discarded automatically. 4068 4069 4070File: gccinstall.info, Node: Old, Next: GNU Free Documentation License, Prev: Specific, Up: Top 4071 407210 Old installation documentation 4073********************************* 4074 4075Note most of this information is out of date and superseded by the 4076previous chapters of this manual. It is provided for historical 4077reference only, because of a lack of volunteers to merge it into the 4078main manual. 4079 4080* Menu: 4081 4082* Configurations:: Configurations Supported by GCC. 4083 4084 Here is the procedure for installing GCC on a GNU or Unix system. 4085 4086 1. If you have chosen a configuration for GCC which requires other GNU 4087 tools (such as GAS or the GNU linker) instead of the standard 4088 system tools, install the required tools in the build directory 4089 under the names 'as', 'ld' or whatever is appropriate. 4090 4091 Alternatively, you can do subsequent compilation using a value of 4092 the 'PATH' environment variable such that the necessary GNU tools 4093 come before the standard system tools. 4094 4095 2. Specify the host, build and target machine configurations. You do 4096 this when you run the 'configure' script. 4097 4098 The "build" machine is the system which you are using, the "host" 4099 machine is the system where you want to run the resulting compiler 4100 (normally the build machine), and the "target" machine is the 4101 system for which you want the compiler to generate code. 4102 4103 If you are building a compiler to produce code for the machine it 4104 runs on (a native compiler), you normally do not need to specify 4105 any operands to 'configure'; it will try to guess the type of 4106 machine you are on and use that as the build, host and target 4107 machines. So you don't need to specify a configuration when 4108 building a native compiler unless 'configure' cannot figure out 4109 what your configuration is or guesses wrong. 4110 4111 In those cases, specify the build machine's "configuration name" 4112 with the '--host' option; the host and target will default to be 4113 the same as the host machine. 4114 4115 Here is an example: 4116 4117 ./configure --host=sparc-sun-sunos4.1 4118 4119 A configuration name may be canonical or it may be more or less 4120 abbreviated. 4121 4122 A canonical configuration name has three parts, separated by 4123 dashes. It looks like this: 'CPU-COMPANY-SYSTEM'. (The three 4124 parts may themselves contain dashes; 'configure' can figure out 4125 which dashes serve which purpose.) For example, 4126 'm68k-sun-sunos4.1' specifies a Sun 3. 4127 4128 You can also replace parts of the configuration by nicknames or 4129 aliases. For example, 'sun3' stands for 'm68k-sun', so 4130 'sun3-sunos4.1' is another way to specify a Sun 3. 4131 4132 You can specify a version number after any of the system types, and 4133 some of the CPU types. In most cases, the version is irrelevant, 4134 and will be ignored. So you might as well specify the version if 4135 you know it. 4136 4137 See *note Configurations::, for a list of supported configuration 4138 names and notes on many of the configurations. You should check 4139 the notes in that section before proceeding any further with the 4140 installation of GCC. 4141 4142 4143File: gccinstall.info, Node: Configurations, Up: Old 4144 414510.1 Configurations Supported by GCC 4146==================================== 4147 4148Here are the possible CPU types: 4149 4150 1750a, a29k, alpha, arm, avr, cN, clipper, dsp16xx, elxsi, fr30, 4151 h8300, hppa1.0, hppa1.1, i370, i386, i486, i586, i686, i786, i860, 4152 i960, ip2k, m32r, m68000, m68k, m88k, mcore, mips, mipsel, mips64, 4153 mips64el, mn10200, mn10300, ns32k, pdp11, powerpc, powerpcle, romp, 4154 rs6000, sh, sparc, sparclite, sparc64, v850, vax, we32k. 4155 4156 Here are the recognized company names. As you can see, customary 4157abbreviations are used rather than the longer official names. 4158 4159 acorn, alliant, altos, apollo, apple, att, bull, cbm, convergent, 4160 convex, crds, dec, dg, dolphin, elxsi, encore, harris, hitachi, hp, 4161 ibm, intergraph, isi, mips, motorola, ncr, next, ns, omron, plexus, 4162 sequent, sgi, sony, sun, tti, unicom, wrs. 4163 4164 The company name is meaningful only to disambiguate when the rest of 4165the information supplied is insufficient. You can omit it, writing just 4166'CPU-SYSTEM', if it is not needed. For example, 'vax-ultrix4.2' is 4167equivalent to 'vax-dec-ultrix4.2'. 4168 4169 Here is a list of system types: 4170 4171 386bsd, aix, acis, amigaos, aos, aout, aux, bosx, bsd, clix, coff, 4172 ctix, cxux, dgux, dynix, ebmon, ecoff, elf, esix, freebsd, hms, 4173 genix, gnu, linux, linux-gnu, hiux, hpux, iris, irix, isc, luna, 4174 lynxos, mach, minix, msdos, mvs, netbsd, newsos, nindy, ns, osf, 4175 osfrose, ptx, riscix, riscos, rtu, sco, sim, solaris, sunos, sym, 4176 sysv, udi, ultrix, unicos, uniplus, unos, vms, vsta, vxworks, 4177 winnt, xenix. 4178 4179You can omit the system type; then 'configure' guesses the operating 4180system from the CPU and company. 4181 4182 You can add a version number to the system type; this may or may not 4183make a difference. For example, you can write 'bsd4.3' or 'bsd4.4' to 4184distinguish versions of BSD. In practice, the version number is most 4185needed for 'sysv3' and 'sysv4', which are often treated differently. 4186 4187 'linux-gnu' is the canonical name for the GNU/Linux target; however 4188GCC will also accept 'linux'. The version of the kernel in use is not 4189relevant on these systems. A suffix such as 'libc1' or 'aout' 4190distinguishes major versions of the C library; all of the suffixed 4191versions are obsolete. 4192 4193 If you specify an impossible combination such as 'i860-dg-vms', then 4194you may get an error message from 'configure', or it may ignore part of 4195the information and do the best it can with the rest. 'configure' 4196always prints the canonical name for the alternative that it used. GCC 4197does not support all possible alternatives. 4198 4199 Often a particular model of machine has a name. Many machine names 4200are recognized as aliases for CPU/company combinations. Thus, the 4201machine name 'sun3', mentioned above, is an alias for 'm68k-sun'. 4202Sometimes we accept a company name as a machine name, when the name is 4203popularly used for a particular machine. Here is a table of the known 4204machine names: 4205 4206 3300, 3b1, 3bN, 7300, altos3068, altos, apollo68, att-7300, 4207 balance, convex-cN, crds, decstation-3100, decstation, delta, 4208 encore, fx2800, gmicro, hp7NN, hp8NN, hp9k2NN, hp9k3NN, hp9k7NN, 4209 hp9k8NN, iris4d, iris, isi68, m3230, magnum, merlin, miniframe, 4210 mmax, news-3600, news800, news, next, pbd, pc532, pmax, powerpc, 4211 powerpcle, ps2, risc-news, rtpc, sun2, sun386i, sun386, sun3, sun4, 4212 symmetry, tower-32, tower. 4213 4214Remember that a machine name specifies both the cpu type and the company 4215name. 4216 4217 4218File: gccinstall.info, Node: GNU Free Documentation License, Next: Concept Index, Prev: Old, Up: Top 4219 4220GNU Free Documentation License 4221****************************** 4222 4223 Version 1.3, 3 November 2008 4224 4225 Copyright (C) 2000, 2001, 2002, 2007, 2008 Free Software Foundation, Inc. 4226 <http://fsf.org/> 4227 4228 Everyone is permitted to copy and distribute verbatim copies 4229 of this license document, but changing it is not allowed. 4230 4231 0. PREAMBLE 4232 4233 The purpose of this License is to make a manual, textbook, or other 4234 functional and useful document "free" in the sense of freedom: to 4235 assure everyone the effective freedom to copy and redistribute it, 4236 with or without modifying it, either commercially or 4237 noncommercially. Secondarily, this License preserves for the 4238 author and publisher a way to get credit for their work, while not 4239 being considered responsible for modifications made by others. 4240 4241 This License is a kind of "copyleft", which means that derivative 4242 works of the document must themselves be free in the same sense. 4243 It complements the GNU General Public License, which is a copyleft 4244 license designed for free software. 4245 4246 We have designed this License in order to use it for manuals for 4247 free software, because free software needs free documentation: a 4248 free program should come with manuals providing the same freedoms 4249 that the software does. But this License is not limited to 4250 software manuals; it can be used for any textual work, regardless 4251 of subject matter or whether it is published as a printed book. We 4252 recommend this License principally for works whose purpose is 4253 instruction or reference. 4254 4255 1. APPLICABILITY AND DEFINITIONS 4256 4257 This License applies to any manual or other work, in any medium, 4258 that contains a notice placed by the copyright holder saying it can 4259 be distributed under the terms of this License. Such a notice 4260 grants a world-wide, royalty-free license, unlimited in duration, 4261 to use that work under the conditions stated herein. The 4262 "Document", below, refers to any such manual or work. Any member 4263 of the public is a licensee, and is addressed as "you". You accept 4264 the license if you copy, modify or distribute the work in a way 4265 requiring permission under copyright law. 4266 4267 A "Modified Version" of the Document means any work containing the 4268 Document or a portion of it, either copied verbatim, or with 4269 modifications and/or translated into another language. 4270 4271 A "Secondary Section" is a named appendix or a front-matter section 4272 of the Document that deals exclusively with the relationship of the 4273 publishers or authors of the Document to the Document's overall 4274 subject (or to related matters) and contains nothing that could 4275 fall directly within that overall subject. (Thus, if the Document 4276 is in part a textbook of mathematics, a Secondary Section may not 4277 explain any mathematics.) The relationship could be a matter of 4278 historical connection with the subject or with related matters, or 4279 of legal, commercial, philosophical, ethical or political position 4280 regarding them. 4281 4282 The "Invariant Sections" are certain Secondary Sections whose 4283 titles are designated, as being those of Invariant Sections, in the 4284 notice that says that the Document is released under this License. 4285 If a section does not fit the above definition of Secondary then it 4286 is not allowed to be designated as Invariant. The Document may 4287 contain zero Invariant Sections. If the Document does not identify 4288 any Invariant Sections then there are none. 4289 4290 The "Cover Texts" are certain short passages of text that are 4291 listed, as Front-Cover Texts or Back-Cover Texts, in the notice 4292 that says that the Document is released under this License. A 4293 Front-Cover Text may be at most 5 words, and a Back-Cover Text may 4294 be at most 25 words. 4295 4296 A "Transparent" copy of the Document means a machine-readable copy, 4297 represented in a format whose specification is available to the 4298 general public, that is suitable for revising the document 4299 straightforwardly with generic text editors or (for images composed 4300 of pixels) generic paint programs or (for drawings) some widely 4301 available drawing editor, and that is suitable for input to text 4302 formatters or for automatic translation to a variety of formats 4303 suitable for input to text formatters. A copy made in an otherwise 4304 Transparent file format whose markup, or absence of markup, has 4305 been arranged to thwart or discourage subsequent modification by 4306 readers is not Transparent. An image format is not Transparent if 4307 used for any substantial amount of text. A copy that is not 4308 "Transparent" is called "Opaque". 4309 4310 Examples of suitable formats for Transparent copies include plain 4311 ASCII without markup, Texinfo input format, LaTeX input format, 4312 SGML or XML using a publicly available DTD, and standard-conforming 4313 simple HTML, PostScript or PDF designed for human modification. 4314 Examples of transparent image formats include PNG, XCF and JPG. 4315 Opaque formats include proprietary formats that can be read and 4316 edited only by proprietary word processors, SGML or XML for which 4317 the DTD and/or processing tools are not generally available, and 4318 the machine-generated HTML, PostScript or PDF produced by some word 4319 processors for output purposes only. 4320 4321 The "Title Page" means, for a printed book, the title page itself, 4322 plus such following pages as are needed to hold, legibly, the 4323 material this License requires to appear in the title page. For 4324 works in formats which do not have any title page as such, "Title 4325 Page" means the text near the most prominent appearance of the 4326 work's title, preceding the beginning of the body of the text. 4327 4328 The "publisher" means any person or entity that distributes copies 4329 of the Document to the public. 4330 4331 A section "Entitled XYZ" means a named subunit of the Document 4332 whose title either is precisely XYZ or contains XYZ in parentheses 4333 following text that translates XYZ in another language. (Here XYZ 4334 stands for a specific section name mentioned below, such as 4335 "Acknowledgements", "Dedications", "Endorsements", or "History".) 4336 To "Preserve the Title" of such a section when you modify the 4337 Document means that it remains a section "Entitled XYZ" according 4338 to this definition. 4339 4340 The Document may include Warranty Disclaimers next to the notice 4341 which states that this License applies to the Document. These 4342 Warranty Disclaimers are considered to be included by reference in 4343 this License, but only as regards disclaiming warranties: any other 4344 implication that these Warranty Disclaimers may have is void and 4345 has no effect on the meaning of this License. 4346 4347 2. VERBATIM COPYING 4348 4349 You may copy and distribute the Document in any medium, either 4350 commercially or noncommercially, provided that this License, the 4351 copyright notices, and the license notice saying this License 4352 applies to the Document are reproduced in all copies, and that you 4353 add no other conditions whatsoever to those of this License. You 4354 may not use technical measures to obstruct or control the reading 4355 or further copying of the copies you make or distribute. However, 4356 you may accept compensation in exchange for copies. If you 4357 distribute a large enough number of copies you must also follow the 4358 conditions in section 3. 4359 4360 You may also lend copies, under the same conditions stated above, 4361 and you may publicly display copies. 4362 4363 3. COPYING IN QUANTITY 4364 4365 If you publish printed copies (or copies in media that commonly 4366 have printed covers) of the Document, numbering more than 100, and 4367 the Document's license notice requires Cover Texts, you must 4368 enclose the copies in covers that carry, clearly and legibly, all 4369 these Cover Texts: Front-Cover Texts on the front cover, and 4370 Back-Cover Texts on the back cover. Both covers must also clearly 4371 and legibly identify you as the publisher of these copies. The 4372 front cover must present the full title with all words of the title 4373 equally prominent and visible. You may add other material on the 4374 covers in addition. Copying with changes limited to the covers, as 4375 long as they preserve the title of the Document and satisfy these 4376 conditions, can be treated as verbatim copying in other respects. 4377 4378 If the required texts for either cover are too voluminous to fit 4379 legibly, you should put the first ones listed (as many as fit 4380 reasonably) on the actual cover, and continue the rest onto 4381 adjacent pages. 4382 4383 If you publish or distribute Opaque copies of the Document 4384 numbering more than 100, you must either include a machine-readable 4385 Transparent copy along with each Opaque copy, or state in or with 4386 each Opaque copy a computer-network location from which the general 4387 network-using public has access to download using public-standard 4388 network protocols a complete Transparent copy of the Document, free 4389 of added material. If you use the latter option, you must take 4390 reasonably prudent steps, when you begin distribution of Opaque 4391 copies in quantity, to ensure that this Transparent copy will 4392 remain thus accessible at the stated location until at least one 4393 year after the last time you distribute an Opaque copy (directly or 4394 through your agents or retailers) of that edition to the public. 4395 4396 It is requested, but not required, that you contact the authors of 4397 the Document well before redistributing any large number of copies, 4398 to give them a chance to provide you with an updated version of the 4399 Document. 4400 4401 4. MODIFICATIONS 4402 4403 You may copy and distribute a Modified Version of the Document 4404 under the conditions of sections 2 and 3 above, provided that you 4405 release the Modified Version under precisely this License, with the 4406 Modified Version filling the role of the Document, thus licensing 4407 distribution and modification of the Modified Version to whoever 4408 possesses a copy of it. In addition, you must do these things in 4409 the Modified Version: 4410 4411 A. Use in the Title Page (and on the covers, if any) a title 4412 distinct from that of the Document, and from those of previous 4413 versions (which should, if there were any, be listed in the 4414 History section of the Document). You may use the same title 4415 as a previous version if the original publisher of that 4416 version gives permission. 4417 4418 B. List on the Title Page, as authors, one or more persons or 4419 entities responsible for authorship of the modifications in 4420 the Modified Version, together with at least five of the 4421 principal authors of the Document (all of its principal 4422 authors, if it has fewer than five), unless they release you 4423 from this requirement. 4424 4425 C. State on the Title page the name of the publisher of the 4426 Modified Version, as the publisher. 4427 4428 D. Preserve all the copyright notices of the Document. 4429 4430 E. Add an appropriate copyright notice for your modifications 4431 adjacent to the other copyright notices. 4432 4433 F. Include, immediately after the copyright notices, a license 4434 notice giving the public permission to use the Modified 4435 Version under the terms of this License, in the form shown in 4436 the Addendum below. 4437 4438 G. Preserve in that license notice the full lists of Invariant 4439 Sections and required Cover Texts given in the Document's 4440 license notice. 4441 4442 H. Include an unaltered copy of this License. 4443 4444 I. Preserve the section Entitled "History", Preserve its Title, 4445 and add to it an item stating at least the title, year, new 4446 authors, and publisher of the Modified Version as given on the 4447 Title Page. If there is no section Entitled "History" in the 4448 Document, create one stating the title, year, authors, and 4449 publisher of the Document as given on its Title Page, then add 4450 an item describing the Modified Version as stated in the 4451 previous sentence. 4452 4453 J. Preserve the network location, if any, given in the Document 4454 for public access to a Transparent copy of the Document, and 4455 likewise the network locations given in the Document for 4456 previous versions it was based on. These may be placed in the 4457 "History" section. You may omit a network location for a work 4458 that was published at least four years before the Document 4459 itself, or if the original publisher of the version it refers 4460 to gives permission. 4461 4462 K. For any section Entitled "Acknowledgements" or "Dedications", 4463 Preserve the Title of the section, and preserve in the section 4464 all the substance and tone of each of the contributor 4465 acknowledgements and/or dedications given therein. 4466 4467 L. Preserve all the Invariant Sections of the Document, unaltered 4468 in their text and in their titles. Section numbers or the 4469 equivalent are not considered part of the section titles. 4470 4471 M. Delete any section Entitled "Endorsements". Such a section 4472 may not be included in the Modified Version. 4473 4474 N. Do not retitle any existing section to be Entitled 4475 "Endorsements" or to conflict in title with any Invariant 4476 Section. 4477 4478 O. Preserve any Warranty Disclaimers. 4479 4480 If the Modified Version includes new front-matter sections or 4481 appendices that qualify as Secondary Sections and contain no 4482 material copied from the Document, you may at your option designate 4483 some or all of these sections as invariant. To do this, add their 4484 titles to the list of Invariant Sections in the Modified Version's 4485 license notice. These titles must be distinct from any other 4486 section titles. 4487 4488 You may add a section Entitled "Endorsements", provided it contains 4489 nothing but endorsements of your Modified Version by various 4490 parties--for example, statements of peer review or that the text 4491 has been approved by an organization as the authoritative 4492 definition of a standard. 4493 4494 You may add a passage of up to five words as a Front-Cover Text, 4495 and a passage of up to 25 words as a Back-Cover Text, to the end of 4496 the list of Cover Texts in the Modified Version. Only one passage 4497 of Front-Cover Text and one of Back-Cover Text may be added by (or 4498 through arrangements made by) any one entity. If the Document 4499 already includes a cover text for the same cover, previously added 4500 by you or by arrangement made by the same entity you are acting on 4501 behalf of, you may not add another; but you may replace the old 4502 one, on explicit permission from the previous publisher that added 4503 the old one. 4504 4505 The author(s) and publisher(s) of the Document do not by this 4506 License give permission to use their names for publicity for or to 4507 assert or imply endorsement of any Modified Version. 4508 4509 5. COMBINING DOCUMENTS 4510 4511 You may combine the Document with other documents released under 4512 this License, under the terms defined in section 4 above for 4513 modified versions, provided that you include in the combination all 4514 of the Invariant Sections of all of the original documents, 4515 unmodified, and list them all as Invariant Sections of your 4516 combined work in its license notice, and that you preserve all 4517 their Warranty Disclaimers. 4518 4519 The combined work need only contain one copy of this License, and 4520 multiple identical Invariant Sections may be replaced with a single 4521 copy. If there are multiple Invariant Sections with the same name 4522 but different contents, make the title of each such section unique 4523 by adding at the end of it, in parentheses, the name of the 4524 original author or publisher of that section if known, or else a 4525 unique number. Make the same adjustment to the section titles in 4526 the list of Invariant Sections in the license notice of the 4527 combined work. 4528 4529 In the combination, you must combine any sections Entitled 4530 "History" in the various original documents, forming one section 4531 Entitled "History"; likewise combine any sections Entitled 4532 "Acknowledgements", and any sections Entitled "Dedications". You 4533 must delete all sections Entitled "Endorsements." 4534 4535 6. COLLECTIONS OF DOCUMENTS 4536 4537 You may make a collection consisting of the Document and other 4538 documents released under this License, and replace the individual 4539 copies of this License in the various documents with a single copy 4540 that is included in the collection, provided that you follow the 4541 rules of this License for verbatim copying of each of the documents 4542 in all other respects. 4543 4544 You may extract a single document from such a collection, and 4545 distribute it individually under this License, provided you insert 4546 a copy of this License into the extracted document, and follow this 4547 License in all other respects regarding verbatim copying of that 4548 document. 4549 4550 7. AGGREGATION WITH INDEPENDENT WORKS 4551 4552 A compilation of the Document or its derivatives with other 4553 separate and independent documents or works, in or on a volume of a 4554 storage or distribution medium, is called an "aggregate" if the 4555 copyright resulting from the compilation is not used to limit the 4556 legal rights of the compilation's users beyond what the individual 4557 works permit. When the Document is included in an aggregate, this 4558 License does not apply to the other works in the aggregate which 4559 are not themselves derivative works of the Document. 4560 4561 If the Cover Text requirement of section 3 is applicable to these 4562 copies of the Document, then if the Document is less than one half 4563 of the entire aggregate, the Document's Cover Texts may be placed 4564 on covers that bracket the Document within the aggregate, or the 4565 electronic equivalent of covers if the Document is in electronic 4566 form. Otherwise they must appear on printed covers that bracket 4567 the whole aggregate. 4568 4569 8. TRANSLATION 4570 4571 Translation is considered a kind of modification, so you may 4572 distribute translations of the Document under the terms of section 4573 4. Replacing Invariant Sections with translations requires special 4574 permission from their copyright holders, but you may include 4575 translations of some or all Invariant Sections in addition to the 4576 original versions of these Invariant Sections. You may include a 4577 translation of this License, and all the license notices in the 4578 Document, and any Warranty Disclaimers, provided that you also 4579 include the original English version of this License and the 4580 original versions of those notices and disclaimers. In case of a 4581 disagreement between the translation and the original version of 4582 this License or a notice or disclaimer, the original version will 4583 prevail. 4584 4585 If a section in the Document is Entitled "Acknowledgements", 4586 "Dedications", or "History", the requirement (section 4) to 4587 Preserve its Title (section 1) will typically require changing the 4588 actual title. 4589 4590 9. TERMINATION 4591 4592 You may not copy, modify, sublicense, or distribute the Document 4593 except as expressly provided under this License. Any attempt 4594 otherwise to copy, modify, sublicense, or distribute it is void, 4595 and will automatically terminate your rights under this License. 4596 4597 However, if you cease all violation of this License, then your 4598 license from a particular copyright holder is reinstated (a) 4599 provisionally, unless and until the copyright holder explicitly and 4600 finally terminates your license, and (b) permanently, if the 4601 copyright holder fails to notify you of the violation by some 4602 reasonable means prior to 60 days after the cessation. 4603 4604 Moreover, your license from a particular copyright holder is 4605 reinstated permanently if the copyright holder notifies you of the 4606 violation by some reasonable means, this is the first time you have 4607 received notice of violation of this License (for any work) from 4608 that copyright holder, and you cure the violation prior to 30 days 4609 after your receipt of the notice. 4610 4611 Termination of your rights under this section does not terminate 4612 the licenses of parties who have received copies or rights from you 4613 under this License. If your rights have been terminated and not 4614 permanently reinstated, receipt of a copy of some or all of the 4615 same material does not give you any rights to use it. 4616 4617 10. FUTURE REVISIONS OF THIS LICENSE 4618 4619 The Free Software Foundation may publish new, revised versions of 4620 the GNU Free Documentation License from time to time. Such new 4621 versions will be similar in spirit to the present version, but may 4622 differ in detail to address new problems or concerns. See 4623 <http://www.gnu.org/copyleft/>. 4624 4625 Each version of the License is given a distinguishing version 4626 number. If the Document specifies that a particular numbered 4627 version of this License "or any later version" applies to it, you 4628 have the option of following the terms and conditions either of 4629 that specified version or of any later version that has been 4630 published (not as a draft) by the Free Software Foundation. If the 4631 Document does not specify a version number of this License, you may 4632 choose any version ever published (not as a draft) by the Free 4633 Software Foundation. If the Document specifies that a proxy can 4634 decide which future versions of this License can be used, that 4635 proxy's public statement of acceptance of a version permanently 4636 authorizes you to choose that version for the Document. 4637 4638 11. RELICENSING 4639 4640 "Massive Multiauthor Collaboration Site" (or "MMC Site") means any 4641 World Wide Web server that publishes copyrightable works and also 4642 provides prominent facilities for anybody to edit those works. A 4643 public wiki that anybody can edit is an example of such a server. 4644 A "Massive Multiauthor Collaboration" (or "MMC") contained in the 4645 site means any set of copyrightable works thus published on the MMC 4646 site. 4647 4648 "CC-BY-SA" means the Creative Commons Attribution-Share Alike 3.0 4649 license published by Creative Commons Corporation, a not-for-profit 4650 corporation with a principal place of business in San Francisco, 4651 California, as well as future copyleft versions of that license 4652 published by that same organization. 4653 4654 "Incorporate" means to publish or republish a Document, in whole or 4655 in part, as part of another Document. 4656 4657 An MMC is "eligible for relicensing" if it is licensed under this 4658 License, and if all works that were first published under this 4659 License somewhere other than this MMC, and subsequently 4660 incorporated in whole or in part into the MMC, (1) had no cover 4661 texts or invariant sections, and (2) were thus incorporated prior 4662 to November 1, 2008. 4663 4664 The operator of an MMC Site may republish an MMC contained in the 4665 site under CC-BY-SA on the same site at any time before August 1, 4666 2009, provided the MMC is eligible for relicensing. 4667 4668ADDENDUM: How to use this License for your documents 4669==================================================== 4670 4671To use this License in a document you have written, include a copy of 4672the License in the document and put the following copyright and license 4673notices just after the title page: 4674 4675 Copyright (C) YEAR YOUR NAME. 4676 Permission is granted to copy, distribute and/or modify this document 4677 under the terms of the GNU Free Documentation License, Version 1.3 4678 or any later version published by the Free Software Foundation; 4679 with no Invariant Sections, no Front-Cover Texts, and no Back-Cover 4680 Texts. A copy of the license is included in the section entitled ``GNU 4681 Free Documentation License''. 4682 4683 If you have Invariant Sections, Front-Cover Texts and Back-Cover 4684Texts, replace the "with...Texts." line with this: 4685 4686 with the Invariant Sections being LIST THEIR TITLES, with 4687 the Front-Cover Texts being LIST, and with the Back-Cover Texts 4688 being LIST. 4689 4690 If you have Invariant Sections without Cover Texts, or some other 4691combination of the three, merge those two alternatives to suit the 4692situation. 4693 4694 If your document contains nontrivial examples of program code, we 4695recommend releasing these examples in parallel under your choice of free 4696software license, such as the GNU General Public License, to permit 4697their use in free software. 4698 4699 4700File: gccinstall.info, Node: Concept Index, Prev: GNU Free Documentation License, Up: Top 4701 4702Concept Index 4703************* 4704 4705[index] 4706* Menu: 4707 4708* Binaries: Binaries. (line 6) 4709* build_configargs: Configuration. (line 1618) 4710* Configuration: Configuration. (line 6) 4711* configurations supported by GCC: Configurations. (line 6) 4712* Downloading GCC: Downloading the source. 4713 (line 6) 4714* Downloading the Source: Downloading the source. 4715 (line 6) 4716* FDL, GNU Free Documentation License: GNU Free Documentation License. 4717 (line 6) 4718* Host specific installation: Specific. (line 6) 4719* host_configargs: Configuration. (line 1622) 4720* Installing GCC: Binaries: Binaries. (line 6) 4721* Installing GCC: Building: Building. (line 6) 4722* Installing GCC: Configuration: Configuration. (line 6) 4723* Installing GCC: Testing: Testing. (line 6) 4724* Prerequisites: Prerequisites. (line 6) 4725* Specific: Specific. (line 6) 4726* Specific installation notes: Specific. (line 6) 4727* Target specific installation: Specific. (line 6) 4728* Target specific installation notes: Specific. (line 6) 4729* target_configargs: Configuration. (line 1626) 4730* Testing: Testing. (line 6) 4731* Testsuite: Testing. (line 6) 4732 4733 4734 4735Tag Table: 4736Node: Top1696 4737Node: Installing GCC2254 4738Node: Prerequisites3888 4739Node: Downloading the source14407 4740Node: Configuration16078 4741Ref: with-gnu-as31709 4742Ref: with-as32604 4743Ref: with-gnu-ld34017 4744Ref: WithAixSoname56426 4745Ref: AixLdCommand57087 4746Node: Building93671 4747Node: Testing110280 4748Node: Final install118167 4749Node: Binaries123467 4750Node: Specific124594 4751Ref: aarch64-x-x125101 4752Ref: alpha-x-x127121 4753Ref: amd64-x-solaris210127530 4754Ref: amdgcn-unknown-amdhsa127633 4755Ref: arc-x-elf32128219 4756Ref: arc-linux-uclibc128395 4757Ref: arm-x-eabi128536 4758Ref: avr128789 4759Ref: bfin129370 4760Ref: cr16129612 4761Ref: cris130028 4762Ref: dos130726 4763Ref: epiphany-x-elf131049 4764Ref: x-x-freebsd131154 4765Ref: ft32-x-elf132954 4766Ref: h8300-hms133052 4767Ref: hppa-hp-hpux133404 4768Ref: hppa-hp-hpux10135776 4769Ref: hppa-hp-hpux11136189 4770Ref: x-x-linux-gnu141591 4771Ref: ix86-x-linux141784 4772Ref: ix86-x-solaris210142097 4773Ref: ia64-x-linux143442 4774Ref: ia64-x-hpux144212 4775Ref: x-ibm-aix144767 4776Ref: TransferAixShobj148429 4777Ref: iq2000-x-elf152239 4778Ref: lm32-x-elf152379 4779Ref: lm32-x-uclinux152483 4780Ref: m32c-x-elf152611 4781Ref: m32r-x-elf152713 4782Ref: m68k-x-x152815 4783Ref: m68k-x-uclinux153853 4784Ref: microblaze-x-elf154098 4785Ref: mips-x-x154217 4786Ref: moxie-x-elf156127 4787Ref: msp430-x-elf156174 4788Ref: nds32le-x-elf156277 4789Ref: nds32be-x-elf156349 4790Ref: nvptx-x-none156418 4791Ref: or1k-x-elf156981 4792Ref: or1k-x-linux157112 4793Ref: powerpc-x-x157193 4794Ref: powerpc-x-darwin157384 4795Ref: powerpc-x-elf157878 4796Ref: powerpc-x-linux-gnu157963 4797Ref: powerpc-x-netbsd158058 4798Ref: powerpc-x-eabisim158146 4799Ref: powerpc-x-eabi158272 4800Ref: powerpcle-x-elf158348 4801Ref: powerpcle-x-eabisim158440 4802Ref: powerpcle-x-eabi158573 4803Ref: rl78-x-elf158656 4804Ref: riscv32-x-elf158762 4805Ref: riscv32-x-linux158972 4806Ref: riscv64-x-elf159150 4807Ref: riscv64-x-linux159360 4808Ref: rx-x-elf159538 4809Ref: s390-x-linux159584 4810Ref: s390x-x-linux159656 4811Ref: s390x-ibm-tpf159743 4812Ref: x-x-solaris2159874 4813Ref: sparc-x-x163895 4814Ref: sparc-sun-solaris2164397 4815Ref: sparc-sun-solaris210165536 4816Ref: sparc-x-linux165911 4817Ref: sparc64-x-solaris2165942 4818Ref: sparcv9-x-solaris2166275 4819Ref: c6x-x-x166362 4820Ref: tilegx-*-linux166454 4821Ref: tilegxbe-*-linux166596 4822Ref: tilepro-*-linux166739 4823Ref: visium-x-elf166860 4824Ref: x-x-vxworks166968 4825Ref: x86-64-x-x168491 4826Ref: x86-64-x-solaris210168819 4827Ref: xtensa-x-elf169483 4828Ref: xtensa-x-linux170154 4829Ref: windows170495 4830Ref: x-x-cygwin172336 4831Ref: x-x-mingw32172889 4832Ref: older173115 4833Ref: elf175232 4834Node: Old175490 4835Node: Configurations178623 4836Node: GNU Free Documentation License182161 4837Node: Concept Index207289 4838 4839End Tag Table 4840 4841 4842Local Variables: 4843coding: utf-8 4844End: 4845