1 \input texinfo.tex @c -*-texinfo-*-
4 @setfilename gccinstall.info
9 @include gcc-common.texi
11 @c Specify title for specific html page
13 @settitle Installing GCC
16 @settitle Host/Target specific installation notes for GCC
18 @ifset prerequisiteshtml
19 @settitle Prerequisites for GCC
22 @settitle Downloading GCC
25 @settitle Installing GCC: Configuration
28 @settitle Installing GCC: Building
31 @settitle Installing GCC: Testing
33 @ifset finalinstallhtml
34 @settitle Installing GCC: Final installation
37 @settitle Installing GCC: Binaries
40 @settitle Installing GCC: Old documentation
43 @settitle Installing GCC: GNU Free Documentation License
46 @c Copyright (C) 1988-2020 Free Software Foundation, Inc.
47 @c *** Converted to texinfo by Dean Wakerley, dean@wakerley.com
49 @c IMPORTANT: whenever you modify this file, run `install.texi2html' to
50 @c test the generation of HTML documents for the gcc.gnu.org web pages.
52 @c Do not use @footnote{} in this file as it breaks install.texi2html!
54 @c Include everything if we're not making html
58 @set prerequisiteshtml
69 @c Part 2 Summary Description and Copyright
71 Copyright @copyright{} 1988-2020 Free Software Foundation, Inc.
73 Permission is granted to copy, distribute and/or modify this document
74 under the terms of the GNU Free Documentation License, Version 1.3 or
75 any later version published by the Free Software Foundation; with no
76 Invariant Sections, the Front-Cover texts being (a) (see below), and
77 with the Back-Cover Texts being (b) (see below). A copy of the
78 license is included in the section entitled ``@uref{./gfdl.html,,GNU
79 Free Documentation License}''.
81 (a) The FSF's Front-Cover Text is:
85 (b) The FSF's Back-Cover Text is:
87 You have freedom to copy and modify this GNU Manual, like GNU
88 software. Copies published by the Free Software Foundation raise
89 funds for GNU development.
94 @dircategory Software development
96 * gccinstall: (gccinstall). Installing the GNU Compiler Collection.
99 @c Part 3 Titlepage and Copyright
101 @title Installing GCC
104 @c The following two commands start the copyright page.
106 @vskip 0pt plus 1filll
110 @c Part 4 Top node, Master Menu, and/or Table of Contents
113 @comment node-name, next, Previous, up
116 * Installing GCC:: This document describes the generic installation
117 procedure for GCC as well as detailing some target
118 specific installation instructions.
120 * Specific:: Host/target specific installation notes for GCC.
121 * Binaries:: Where to get pre-compiled binaries.
123 * Old:: Old installation documentation.
125 * GNU Free Documentation License:: How you can copy and share this manual.
126 * Concept Index:: This index has two entries.
134 @c Part 5 The Body of the Document
135 @c ***Installing GCC**********************************************************
137 @comment node-name, next, previous, up
138 @node Installing GCC, Binaries, , Top
142 @chapter Installing GCC
145 The latest version of this document is always available at
146 @uref{http://gcc.gnu.org/install/,,http://gcc.gnu.org/install/}.
147 It refers to the current development sources, instructions for
148 specific released versions are included with the sources.
150 This document describes the generic installation procedure for GCC as well
151 as detailing some target specific installation instructions.
153 GCC includes several components that previously were separate distributions
154 with their own installation instructions. This document supersedes all
155 package-specific installation instructions.
157 @emph{Before} starting the build/install procedure please check the
159 @ref{Specific, host/target specific installation notes}.
162 @uref{specific.html,,host/target specific installation notes}.
164 We recommend you browse the entire generic installation instructions before
167 Lists of successful builds for released versions of GCC are
168 available at @uref{http://gcc.gnu.org/buildstat.html}.
169 These lists are updated as new information becomes available.
171 The installation procedure itself is broken into five steps.
176 * Downloading the source::
179 * Testing:: (optional)
186 @uref{prerequisites.html,,Prerequisites}
188 @uref{download.html,,Downloading the source}
190 @uref{configure.html,,Configuration}
192 @uref{build.html,,Building}
194 @uref{test.html,,Testing} (optional)
196 @uref{finalinstall.html,,Final install}
200 Please note that GCC does not support @samp{make uninstall} and probably
201 won't do so in the near future as this would open a can of worms. Instead,
202 we suggest that you install GCC into a directory of its own and simply
203 remove that directory when you do not need that specific version of GCC
204 any longer, and, if shared libraries are installed there as well, no
205 more binaries exist that use them.
208 There are also some @uref{old.html,,old installation instructions},
209 which are mostly obsolete but still contain some information which has
210 not yet been merged into the main part of this manual.
218 @uref{./index.html,,Return to the GCC Installation page}
224 @c ***Prerequisites**************************************************
226 @comment node-name, next, previous, up
227 @node Prerequisites, Downloading the source, , Installing GCC
229 @ifset prerequisiteshtml
231 @chapter Prerequisites
233 @cindex Prerequisites
235 GCC requires that various tools and packages be available for use in the
236 build procedure. Modifying GCC sources requires additional tools
239 @heading Tools/packages necessary for building GCC
241 @item ISO C++98 compiler
242 Necessary to bootstrap GCC, although versions of GCC prior
243 to 4.8 also allow bootstrapping with a ISO C89 compiler and versions
244 of GCC prior to 3.4 also allow bootstrapping with a traditional
247 To build all languages in a cross-compiler or other configuration where
248 3-stage bootstrap is not performed, you need to start with an existing
249 GCC binary (version 3.4 or later) because source code for language
250 frontends other than C might use GCC extensions.
252 Note that to bootstrap GCC with versions of GCC earlier than 3.4, you
253 may need to use @option{--disable-stage1-checking}, though
254 bootstrapping the compiler with such earlier compilers is strongly
257 @item C standard library and headers
259 In order to build GCC, the C standard library and headers must be present
260 for all target variants for which target libraries will be built (and not
261 only the variant of the host C++ compiler).
263 This affects the popular @samp{x86_64-pc-linux-gnu} platform (among
264 other multilib targets), for which 64-bit (@samp{x86_64}) and 32-bit
265 (@samp{i386}) libc headers are usually packaged separately. If you do a
266 build of a native compiler on @samp{x86_64-pc-linux-gnu}, make sure you
267 either have the 32-bit libc developer package properly installed (the exact
268 name of the package depends on your distro) or you must build GCC as a
269 64-bit only compiler by configuring with the option
270 @option{--disable-multilib}. Otherwise, you may encounter an error such as
271 @samp{fatal error: gnu/stubs-32.h: No such file}
273 @item @anchor{GNAT-prerequisite}GNAT
275 In order to build GNAT, the Ada compiler, you need a working GNAT
276 compiler (GCC version 4.7 or later).
278 This includes GNAT tools such as @command{gnatmake} and
279 @command{gnatlink}, since the Ada front end is written in Ada and
280 uses some GNAT-specific extensions.
282 In order to build a cross compiler, it is strongly recommended to install
283 the new compiler as native first, and then use it to build the cross
284 compiler. Other native compiler versions may work but this is not guaranteed and
285 will typically fail with hard to understand compilation errors during the
288 Similarly, it is strongly recommended to use an older version of GNAT to build
289 GNAT. More recent versions of GNAT than the version built are not guaranteed
290 to work and will often fail during the build with compilation errors.
292 Note that @command{configure} does not test whether the GNAT installation works
293 and has a sufficiently recent version; if too old a GNAT version is
294 installed and @option{--enable-languages=ada} is used, the build will fail.
296 @env{ADA_INCLUDE_PATH} and @env{ADA_OBJECT_PATH} environment variables
297 must not be set when building the Ada compiler, the Ada tools, or the
298 Ada runtime libraries. You can check that your build environment is clean
299 by verifying that @samp{gnatls -v} lists only one explicit path in each
302 @item A ``working'' POSIX compatible shell, or GNU bash
304 Necessary when running @command{configure} because some
305 @command{/bin/sh} shells have bugs and may crash when configuring the
306 target libraries. In other cases, @command{/bin/sh} or @command{ksh}
307 have disastrous corner-case performance problems. This
308 can cause target @command{configure} runs to literally take days to
309 complete in some cases.
311 So on some platforms @command{/bin/ksh} is sufficient, on others it
312 isn't. See the host/target specific instructions for your platform, or
313 use @command{bash} to be sure. Then set @env{CONFIG_SHELL} in your
314 environment to your ``good'' shell prior to running
315 @command{configure}/@command{make}.
317 @command{zsh} is not a fully compliant POSIX shell and will not
318 work when configuring GCC@.
320 @item A POSIX or SVR4 awk
322 Necessary for creating some of the generated source files for GCC@.
323 If in doubt, use a recent GNU awk version, as some of the older ones
324 are broken. GNU awk version 3.1.5 is known to work.
328 Necessary in some circumstances, optional in others. See the
329 host/target specific instructions for your platform for the exact
332 @item gzip version 1.2.4 (or later) or
333 @itemx bzip2 version 1.0.2 (or later)
335 Necessary to uncompress GCC @command{tar} files when source code is
336 obtained via HTTPS mirror sites.
338 @item GNU make version 3.80 (or later)
340 You must have GNU make installed to build GCC@.
342 @item GNU tar version 1.14 (or later)
344 Necessary (only on some platforms) to untar the source code. Many
345 systems' @command{tar} programs will also work, only try GNU
346 @command{tar} if you have problems.
348 @item Perl version between 5.6.1 and 5.6.24
350 Necessary when targeting Darwin, building @samp{libstdc++},
351 and not using @option{--disable-symvers}.
352 Necessary when targeting Solaris 2 with Sun @command{ld} and not using
353 @option{--disable-symvers}. The bundled @command{perl} in Solaris@tie{}8
356 Necessary when regenerating @file{Makefile} dependencies in libiberty.
357 Necessary when regenerating @file{libiberty/functions.texi}.
358 Necessary when generating manpages from Texinfo manuals.
359 Used by various scripts to generate some files included in the source
360 repository (mainly Unicode-related and rarely changing) from source
363 Used by @command{automake}.
367 Several support libraries are necessary to build GCC, some are required,
368 others optional. While any sufficiently new version of required tools
369 usually work, library requirements are generally stricter. Newer
370 versions may work in some cases, but it's safer to use the exact
371 versions documented. We appreciate bug reports about problems with
372 newer versions, though. If your OS vendor provides packages for the
373 support libraries then using those packages may be the simplest way to
374 install the libraries.
377 @item GNU Multiple Precision Library (GMP) version 4.3.2 (or later)
379 Necessary to build GCC@. If a GMP source distribution is found in a
380 subdirectory of your GCC sources named @file{gmp}, it will be built
381 together with GCC. Alternatively, if GMP is already installed but it
382 is not in your library search path, you will have to configure with the
383 @option{--with-gmp} configure option. See also @option{--with-gmp-lib}
384 and @option{--with-gmp-include}.
385 The in-tree build is only supported with the GMP version that
386 download_prerequisites installs.
388 @item MPFR Library version 3.1.0 (or later)
390 Necessary to build GCC@. It can be downloaded from
391 @uref{https://www.mpfr.org}. If an MPFR source distribution is found
392 in a subdirectory of your GCC sources named @file{mpfr}, it will be
393 built together with GCC. Alternatively, if MPFR is already installed
394 but it is not in your default library search path, the
395 @option{--with-mpfr} configure option should be used. See also
396 @option{--with-mpfr-lib} and @option{--with-mpfr-include}.
397 The in-tree build is only supported with the MPFR version that
398 download_prerequisites installs.
400 @item MPC Library version 1.0.1 (or later)
402 Necessary to build GCC@. It can be downloaded from
403 @uref{http://www.multiprecision.org/mpc/}. If an MPC source distribution
404 is found in a subdirectory of your GCC sources named @file{mpc}, it
405 will be built together with GCC. Alternatively, if MPC is already
406 installed but it is not in your default library search path, the
407 @option{--with-mpc} configure option should be used. See also
408 @option{--with-mpc-lib} and @option{--with-mpc-include}.
409 The in-tree build is only supported with the MPC version that
410 download_prerequisites installs.
412 @item isl Library version 0.15 or later.
414 Necessary to build GCC with the Graphite loop optimizations.
415 It can be downloaded from @uref{https://gcc.gnu.org/pub/gcc/infrastructure/}.
416 If an isl source distribution is found
417 in a subdirectory of your GCC sources named @file{isl}, it will be
418 built together with GCC. Alternatively, the @option{--with-isl} configure
419 option should be used if isl is not installed in your default library
424 Necessary to build GCC with zstd compression used for LTO bytecode.
425 The library is searched in your default library patch search.
426 Alternatively, the @option{--with-zstd} configure option should be used.
430 @heading Tools/packages necessary for modifying GCC
432 @item autoconf version 2.69
433 @itemx GNU m4 version 1.4.6 (or later)
435 Necessary when modifying @file{configure.ac}, @file{aclocal.m4}, etc.@:
436 to regenerate @file{configure} and @file{config.in} files.
438 @item automake version 1.15.1
440 Necessary when modifying a @file{Makefile.am} file to regenerate its
441 associated @file{Makefile.in}.
443 Much of GCC does not use automake, so directly edit the @file{Makefile.in}
444 file. Specifically this applies to the @file{gcc}, @file{intl},
445 @file{libcpp}, @file{libiberty}, @file{libobjc} directories as well
446 as any of their subdirectories.
448 For directories that use automake, GCC requires the latest release in
449 the 1.15 series, which is currently 1.15.1. When regenerating a directory
450 to a newer version, please update all the directories using an older 1.15
451 to the latest released version.
453 @item gettext version 0.14.5 (or later)
455 Needed to regenerate @file{gcc.pot}.
457 @item gperf version 2.7.2 (or later)
459 Necessary when modifying @command{gperf} input files, e.g.@:
460 @file{gcc/cp/cfns.gperf} to regenerate its associated header file, e.g.@:
461 @file{gcc/cp/cfns.h}.
467 Necessary to run the GCC testsuite; see the section on testing for
470 @item autogen version 5.5.4 (or later) and
471 @itemx guile version 1.4.1 (or later)
473 Necessary to regenerate @file{fixinc/fixincl.x} from
474 @file{fixinc/inclhack.def} and @file{fixinc/*.tpl}.
476 Necessary to run @samp{make check} for @file{fixinc}.
478 Necessary to regenerate the top level @file{Makefile.in} file from
479 @file{Makefile.tpl} and @file{Makefile.def}.
481 @item Flex version 2.5.4 (or later)
483 Necessary when modifying @file{*.l} files.
485 Necessary to build GCC during development because the generated output
486 files are not included in the version-controlled source repository.
487 They are included in releases.
489 @item Texinfo version 4.7 (or later)
491 Necessary for running @command{makeinfo} when modifying @file{*.texi}
492 files to test your changes.
494 Necessary for running @command{make dvi} or @command{make pdf} to
495 create printable documentation in DVI or PDF format. Texinfo version
496 4.8 or later is required for @command{make pdf}.
498 Necessary to build GCC documentation during development because the
499 generated output files are not included in the repository. They are
500 included in releases.
502 @item @TeX{} (any working version)
504 Necessary for running @command{texi2dvi} and @command{texi2pdf}, which
505 are used when running @command{make dvi} or @command{make pdf} to create
506 DVI or PDF files, respectively.
508 @item Sphinx version 1.0 (or later)
510 Necessary to regenerate @file{jit/docs/_build/texinfo} from the @file{.rst}
511 files in the directories below @file{jit/docs}.
513 @item git (any version)
514 @itemx SSH (any version)
516 Necessary to access the source repository. Public releases and weekly
517 snapshots of the development sources are also available via HTTPS@.
519 @item GNU diffutils version 2.7 (or later)
521 Useful when submitting patches for the GCC source code.
523 @item patch version 2.5.4 (or later)
525 Necessary when applying patches, created with @command{diff}, to one's
535 @uref{./index.html,,Return to the GCC Installation page}
539 @c ***Downloading the source**************************************************
541 @comment node-name, next, previous, up
542 @node Downloading the source, Configuration, Prerequisites, Installing GCC
546 @chapter Downloading GCC
548 @cindex Downloading GCC
549 @cindex Downloading the Source
551 GCC is distributed via @uref{http://gcc.gnu.org/git.html,,git} and via
552 HTTPS as tarballs compressed with @command{gzip} or @command{bzip2}.
554 Please refer to the @uref{http://gcc.gnu.org/releases.html,,releases web page}
555 for information on how to obtain GCC@.
557 The source distribution includes the C, C++, Objective-C, Fortran,
558 and Ada (in the case of GCC 3.1 and later) compilers, as well as
559 runtime libraries for C++, Objective-C, and Fortran.
560 For previous versions these were downloadable as separate components such
561 as the core GCC distribution, which included the C language front end and
562 shared components, and language-specific distributions including the
563 language front end and the language runtime (where appropriate).
565 If you also intend to build binutils (either to upgrade an existing
566 installation or for use in place of the corresponding tools of your
567 OS), unpack the binutils distribution either in the same directory or
568 a separate one. In the latter case, add symbolic links to any
569 components of the binutils you intend to build alongside the compiler
570 (@file{bfd}, @file{binutils}, @file{gas}, @file{gprof}, @file{ld},
571 @file{opcodes}, @dots{}) to the directory containing the GCC sources.
573 Likewise the GMP, MPFR and MPC libraries can be automatically built
574 together with GCC. You may simply run the
575 @command{contrib/download_prerequisites} script in the GCC source directory
576 to set up everything.
577 Otherwise unpack the GMP, MPFR and/or MPC source
578 distributions in the directory containing the GCC sources and rename
579 their directories to @file{gmp}, @file{mpfr} and @file{mpc},
580 respectively (or use symbolic links with the same name).
587 @uref{./index.html,,Return to the GCC Installation page}
591 @c ***Configuration***********************************************************
593 @comment node-name, next, previous, up
594 @node Configuration, Building, Downloading the source, Installing GCC
598 @chapter Installing GCC: Configuration
600 @cindex Configuration
601 @cindex Installing GCC: Configuration
603 Like most GNU software, GCC must be configured before it can be built.
604 This document describes the recommended configuration procedure
605 for both native and cross targets.
607 We use @var{srcdir} to refer to the toplevel source directory for
608 GCC; we use @var{objdir} to refer to the toplevel build/object directory.
610 If you obtained the sources by cloning the repository, @var{srcdir}
611 must refer to the top @file{gcc} directory, the one where the
612 @file{MAINTAINERS} file can be found, and not its @file{gcc}
613 subdirectory, otherwise the build will fail.
615 If either @var{srcdir} or @var{objdir} is located on an automounted NFS
616 file system, the shell's built-in @command{pwd} command will return
617 temporary pathnames. Using these can lead to various sorts of build
618 problems. To avoid this issue, set the @env{PWDCMD} environment
619 variable to an automounter-aware @command{pwd} command, e.g.,
620 @command{pawd} or @samp{amq -w}, during the configuration and build
623 First, we @strong{highly} recommend that GCC be built into a
624 separate directory from the sources which does @strong{not} reside
625 within the source tree. This is how we generally build GCC; building
626 where @var{srcdir} == @var{objdir} should still work, but doesn't
627 get extensive testing; building where @var{objdir} is a subdirectory
628 of @var{srcdir} is unsupported.
630 If you have previously built GCC in the same directory for a
631 different target machine, do @samp{make distclean} to delete all files
632 that might be invalid. One of the files this deletes is @file{Makefile};
633 if @samp{make distclean} complains that @file{Makefile} does not exist
634 or issues a message like ``don't know how to make distclean'' it probably
635 means that the directory is already suitably clean. However, with the
636 recommended method of building in a separate @var{objdir}, you should
637 simply use a different @var{objdir} for each target.
639 Second, when configuring a native system, either @command{cc} or
640 @command{gcc} must be in your path or you must set @env{CC} in
641 your environment before running configure. Otherwise the configuration
645 Note that the bootstrap compiler and the resulting GCC must be link
646 compatible, else the bootstrap will fail with linker errors about
647 incompatible object file formats. Several multilibed targets are
648 affected by this requirement, see
650 @ref{Specific, host/target specific installation notes}.
653 @uref{specific.html,,host/target specific installation notes}.
662 % @var{srcdir}/configure [@var{options}] [@var{target}]
665 @heading Distributor options
667 If you will be distributing binary versions of GCC, with modifications
668 to the source code, you should use the options described in this
669 section to make clear that your version contains modifications.
672 @item --with-pkgversion=@var{version}
673 Specify a string that identifies your package. You may wish
674 to include a build number or build date. This version string will be
675 included in the output of @command{gcc --version}. This suffix does
676 not replace the default version string, only the @samp{GCC} part.
678 The default value is @samp{GCC}.
680 @item --with-bugurl=@var{url}
681 Specify the URL that users should visit if they wish to report a bug.
682 You are of course welcome to forward bugs reported to you to the FSF,
683 if you determine that they are not bugs in your modifications.
685 The default value refers to the FSF's GCC bug tracker.
689 @heading Target specification
692 GCC has code to correctly determine the correct value for @var{target}
693 for nearly all native systems. Therefore, we highly recommend you do
694 not provide a configure target when configuring a native compiler.
697 @var{target} must be specified as @option{--target=@var{target}}
698 when configuring a cross compiler; examples of valid targets would be
699 m68k-elf, sh-elf, etc.
702 Specifying just @var{target} instead of @option{--target=@var{target}}
703 implies that the host defaults to @var{target}.
707 @heading Options specification
709 Use @var{options} to override several configure time options for
710 GCC@. A list of supported @var{options} follows; @samp{configure
711 --help} may list other options, but those not listed below may not
712 work and should not normally be used.
714 Note that each @option{--enable} option has a corresponding
715 @option{--disable} option and that each @option{--with} option has a
716 corresponding @option{--without} option.
719 @item --prefix=@var{dirname}
720 Specify the toplevel installation
721 directory. This is the recommended way to install the tools into a directory
722 other than the default. The toplevel installation directory defaults to
725 We @strong{highly} recommend against @var{dirname} being the same or a
726 subdirectory of @var{objdir} or vice versa. If specifying a directory
727 beneath a user's home directory tree, some shells will not expand
728 @var{dirname} correctly if it contains the @samp{~} metacharacter; use
731 The following standard @command{autoconf} options are supported. Normally you
732 should not need to use these options.
734 @item --exec-prefix=@var{dirname}
735 Specify the toplevel installation directory for architecture-dependent
736 files. The default is @file{@var{prefix}}.
738 @item --bindir=@var{dirname}
739 Specify the installation directory for the executables called by users
740 (such as @command{gcc} and @command{g++}). The default is
741 @file{@var{exec-prefix}/bin}.
743 @item --libdir=@var{dirname}
744 Specify the installation directory for object code libraries and
745 internal data files of GCC@. The default is @file{@var{exec-prefix}/lib}.
747 @item --libexecdir=@var{dirname}
748 Specify the installation directory for internal executables of GCC@.
749 The default is @file{@var{exec-prefix}/libexec}.
751 @item --with-slibdir=@var{dirname}
752 Specify the installation directory for the shared libgcc library. The
753 default is @file{@var{libdir}}.
755 @item --datarootdir=@var{dirname}
756 Specify the root of the directory tree for read-only architecture-independent
757 data files referenced by GCC@. The default is @file{@var{prefix}/share}.
759 @item --infodir=@var{dirname}
760 Specify the installation directory for documentation in info format.
761 The default is @file{@var{datarootdir}/info}.
763 @item --datadir=@var{dirname}
764 Specify the installation directory for some architecture-independent
765 data files referenced by GCC@. The default is @file{@var{datarootdir}}.
767 @item --docdir=@var{dirname}
768 Specify the installation directory for documentation files (other
769 than Info) for GCC@. The default is @file{@var{datarootdir}/doc}.
771 @item --htmldir=@var{dirname}
772 Specify the installation directory for HTML documentation files.
773 The default is @file{@var{docdir}}.
775 @item --pdfdir=@var{dirname}
776 Specify the installation directory for PDF documentation files.
777 The default is @file{@var{docdir}}.
779 @item --mandir=@var{dirname}
780 Specify the installation directory for manual pages. The default is
781 @file{@var{datarootdir}/man}. (Note that the manual pages are only extracts
782 from the full GCC manuals, which are provided in Texinfo format. The manpages
783 are derived by an automatic conversion process from parts of the full
786 @item --with-gxx-include-dir=@var{dirname}
788 the installation directory for G++ header files. The default depends
789 on other configuration options, and differs between cross and native
792 @item --with-specs=@var{specs}
793 Specify additional command line driver SPECS.
794 This can be useful if you need to turn on a non-standard feature by
795 default without modifying the compiler's source code, for instance
796 @option{--with-specs=%@{!fcommon:%@{!fno-common:-fno-common@}@}}.
798 @xref{Spec Files,, Specifying subprocesses and the switches to pass to them,
799 gcc, Using the GNU Compiler Collection (GCC)},
802 See ``Spec Files'' in the main manual
807 @item --program-prefix=@var{prefix}
808 GCC supports some transformations of the names of its programs when
809 installing them. This option prepends @var{prefix} to the names of
810 programs to install in @var{bindir} (see above). For example, specifying
811 @option{--program-prefix=foo-} would result in @samp{gcc}
812 being installed as @file{/usr/local/bin/foo-gcc}.
814 @item --program-suffix=@var{suffix}
815 Appends @var{suffix} to the names of programs to install in @var{bindir}
816 (see above). For example, specifying @option{--program-suffix=-3.1}
817 would result in @samp{gcc} being installed as
818 @file{/usr/local/bin/gcc-3.1}.
820 @item --program-transform-name=@var{pattern}
821 Applies the @samp{sed} script @var{pattern} to be applied to the names
822 of programs to install in @var{bindir} (see above). @var{pattern} has to
823 consist of one or more basic @samp{sed} editing commands, separated by
824 semicolons. For example, if you want the @samp{gcc} program name to be
825 transformed to the installed program @file{/usr/local/bin/myowngcc} and
826 the @samp{g++} program name to be transformed to
827 @file{/usr/local/bin/gspecial++} without changing other program names,
828 you could use the pattern
829 @option{--program-transform-name='s/^gcc$/myowngcc/; s/^g++$/gspecial++/'}
830 to achieve this effect.
832 All three options can be combined and used together, resulting in more
833 complex conversion patterns. As a basic rule, @var{prefix} (and
834 @var{suffix}) are prepended (appended) before further transformations
835 can happen with a special transformation script @var{pattern}.
837 As currently implemented, this option only takes effect for native
838 builds; cross compiler binaries' names are not transformed even when a
839 transformation is explicitly asked for by one of these options.
841 For native builds, some of the installed programs are also installed
842 with the target alias in front of their name, as in
843 @samp{i686-pc-linux-gnu-gcc}. All of the above transformations happen
844 before the target alias is prepended to the name---so, specifying
845 @option{--program-prefix=foo-} and @option{program-suffix=-3.1}, the
846 resulting binary would be installed as
847 @file{/usr/local/bin/i686-pc-linux-gnu-foo-gcc-3.1}.
849 As a last shortcoming, none of the installed Ada programs are
850 transformed yet, which will be fixed in some time.
852 @item --with-local-prefix=@var{dirname}
854 installation directory for local include files. The default is
855 @file{/usr/local}. Specify this option if you want the compiler to
856 search directory @file{@var{dirname}/include} for locally installed
857 header files @emph{instead} of @file{/usr/local/include}.
859 You should specify @option{--with-local-prefix} @strong{only} if your
860 site has a different convention (not @file{/usr/local}) for where to put
863 The default value for @option{--with-local-prefix} is @file{/usr/local}
864 regardless of the value of @option{--prefix}. Specifying
865 @option{--prefix} has no effect on which directory GCC searches for
866 local header files. This may seem counterintuitive, but actually it is
869 The purpose of @option{--prefix} is to specify where to @emph{install
870 GCC}. The local header files in @file{/usr/local/include}---if you put
871 any in that directory---are not part of GCC@. They are part of other
872 programs---perhaps many others. (GCC installs its own header files in
873 another directory which is based on the @option{--prefix} value.)
875 Both the local-prefix include directory and the GCC-prefix include
876 directory are part of GCC's ``system include'' directories. Although these
877 two directories are not fixed, they need to be searched in the proper
878 order for the correct processing of the include_next directive. The
879 local-prefix include directory is searched before the GCC-prefix
880 include directory. Another characteristic of system include directories
881 is that pedantic warnings are turned off for headers in these directories.
883 Some autoconf macros add @option{-I @var{directory}} options to the
884 compiler command line, to ensure that directories containing installed
885 packages' headers are searched. When @var{directory} is one of GCC's
886 system include directories, GCC will ignore the option so that system
887 directories continue to be processed in the correct order. This
888 may result in a search order different from what was specified but the
889 directory will still be searched.
891 GCC automatically searches for ordinary libraries using
892 @env{GCC_EXEC_PREFIX}. Thus, when the same installation prefix is
893 used for both GCC and packages, GCC will automatically search for
894 both headers and libraries. This provides a configuration that is
895 easy to use. GCC behaves in a manner similar to that when it is
896 installed as a system compiler in @file{/usr}.
898 Sites that need to install multiple versions of GCC may not want to
899 use the above simple configuration. It is possible to use the
900 @option{--program-prefix}, @option{--program-suffix} and
901 @option{--program-transform-name} options to install multiple versions
902 into a single directory, but it may be simpler to use different prefixes
903 and the @option{--with-local-prefix} option to specify the location of the
904 site-specific files for each version. It will then be necessary for
905 users to specify explicitly the location of local site libraries
906 (e.g., with @env{LIBRARY_PATH}).
908 The same value can be used for both @option{--with-local-prefix} and
909 @option{--prefix} provided it is not @file{/usr}. This can be used
910 to avoid the default search of @file{/usr/local/include}.
912 @strong{Do not} specify @file{/usr} as the @option{--with-local-prefix}!
913 The directory you use for @option{--with-local-prefix} @strong{must not}
914 contain any of the system's standard header files. If it did contain
915 them, certain programs would be miscompiled (including GNU Emacs, on
916 certain targets), because this would override and nullify the header
917 file corrections made by the @command{fixincludes} script.
919 Indications are that people who use this option use it based on mistaken
920 ideas of what it is for. People use it as if it specified where to
921 install part of GCC@. Perhaps they make this assumption because
922 installing GCC creates the directory.
924 @item --with-gcc-major-version-only
925 Specifies that GCC should use only the major number rather than
926 @var{major}.@var{minor}.@var{patchlevel} in filesystem paths.
928 @item --with-native-system-header-dir=@var{dirname}
929 Specifies that @var{dirname} is the directory that contains native system
930 header files, rather than @file{/usr/include}. This option is most useful
931 if you are creating a compiler that should be isolated from the system
932 as much as possible. It is most commonly used with the
933 @option{--with-sysroot} option and will cause GCC to search
934 @var{dirname} inside the system root specified by that option.
936 @item --enable-shared[=@var{package}[,@dots{}]]
937 Build shared versions of libraries, if shared libraries are supported on
938 the target platform. Unlike GCC 2.95.x and earlier, shared libraries
939 are enabled by default on all platforms that support shared libraries.
941 If a list of packages is given as an argument, build shared libraries
942 only for the listed packages. For other packages, only static libraries
943 will be built. Package names currently recognized in the GCC tree are
944 @samp{libgcc} (also known as @samp{gcc}), @samp{libstdc++} (not
945 @samp{libstdc++-v3}), @samp{libffi}, @samp{zlib}, @samp{boehm-gc},
946 @samp{ada}, @samp{libada}, @samp{libgo}, @samp{libobjc}, and @samp{libphobos}.
947 Note @samp{libiberty} does not support shared libraries at all.
949 Use @option{--disable-shared} to build only static libraries. Note that
950 @option{--disable-shared} does not accept a list of package names as
951 argument, only @option{--enable-shared} does.
953 Contrast with @option{--enable-host-shared}, which affects @emph{host}
956 @item --enable-host-shared
957 Specify that the @emph{host} code should be built into position-independent
958 machine code (with -fPIC), allowing it to be used within shared libraries,
959 but yielding a slightly slower compiler.
961 This option is required when building the libgccjit.so library.
963 Contrast with @option{--enable-shared}, which affects @emph{target}
966 @item @anchor{with-gnu-as}--with-gnu-as
967 Specify that the compiler should assume that the
968 assembler it finds is the GNU assembler. However, this does not modify
969 the rules to find an assembler and will result in confusion if the
970 assembler found is not actually the GNU assembler. (Confusion may also
971 result if the compiler finds the GNU assembler but has not been
972 configured with @option{--with-gnu-as}.) If you have more than one
973 assembler installed on your system, you may want to use this option in
974 connection with @option{--with-as=@var{pathname}} or
975 @option{--with-build-time-tools=@var{pathname}}.
977 The following systems are the only ones where it makes a difference
978 whether you use the GNU assembler. On any other system,
979 @option{--with-gnu-as} has no effect.
982 @item @samp{hppa1.0-@var{any}-@var{any}}
983 @item @samp{hppa1.1-@var{any}-@var{any}}
984 @item @samp{sparc-sun-solaris2.@var{any}}
985 @item @samp{sparc64-@var{any}-solaris2.@var{any}}
988 @item @anchor{with-as}--with-as=@var{pathname}
989 Specify that the compiler should use the assembler pointed to by
990 @var{pathname}, rather than the one found by the standard rules to find
991 an assembler, which are:
994 Unless GCC is being built with a cross compiler, check the
995 @file{@var{libexec}/gcc/@var{target}/@var{version}} directory.
996 @var{libexec} defaults to @file{@var{exec-prefix}/libexec};
997 @var{exec-prefix} defaults to @var{prefix}, which
998 defaults to @file{/usr/local} unless overridden by the
999 @option{--prefix=@var{pathname}} switch described above. @var{target}
1000 is the target system triple, such as @samp{sparc-sun-solaris2.7}, and
1001 @var{version} denotes the GCC version, such as 3.0.
1004 If the target system is the same that you are building on, check
1005 operating system specific directories (e.g.@: @file{/usr/ccs/bin} on
1009 Check in the @env{PATH} for a tool whose name is prefixed by the
1010 target system triple.
1013 Check in the @env{PATH} for a tool whose name is not prefixed by the
1014 target system triple, if the host and target system triple are
1015 the same (in other words, we use a host tool if it can be used for
1016 the target as well).
1019 You may want to use @option{--with-as} if no assembler
1020 is installed in the directories listed above, or if you have multiple
1021 assemblers installed and want to choose one that is not found by the
1024 @item @anchor{with-gnu-ld}--with-gnu-ld
1025 Same as @uref{#with-gnu-as,,@option{--with-gnu-as}}
1028 @item --with-ld=@var{pathname}
1029 Same as @uref{#with-as,,@option{--with-as}}
1033 Specify that stabs debugging
1034 information should be used instead of whatever format the host normally
1035 uses. Normally GCC uses the same debug format as the host system.
1037 @item --with-tls=@var{dialect}
1038 Specify the default TLS dialect, for systems were there is a choice.
1039 For ARM targets, possible values for @var{dialect} are @code{gnu} or
1040 @code{gnu2}, which select between the original GNU dialect and the GNU TLS
1041 descriptor-based dialect.
1043 @item --enable-multiarch
1044 Specify whether to enable or disable multiarch support. The default is
1045 to check for glibc start files in a multiarch location, and enable it
1046 if the files are found. The auto detection is enabled for native builds,
1047 and for cross builds configured with @option{--with-sysroot}, and without
1048 @option{--with-native-system-header-dir}.
1049 More documentation about multiarch can be found at
1050 @uref{https://wiki.debian.org/Multiarch}.
1052 @item --enable-sjlj-exceptions
1053 Force use of the @code{setjmp}/@code{longjmp}-based scheme for exceptions.
1054 @samp{configure} ordinarily picks the correct value based on the platform.
1055 Only use this option if you are sure you need a different setting.
1057 @item --enable-vtable-verify
1058 Specify whether to enable or disable the vtable verification feature.
1059 Enabling this feature causes libstdc++ to be built with its virtual calls
1060 in verifiable mode. This means that, when linked with libvtv, every
1061 virtual call in libstdc++ will verify the vtable pointer through which the
1062 call will be made before actually making the call. If not linked with libvtv,
1063 the verifier will call stub functions (in libstdc++ itself) and do nothing.
1064 If vtable verification is disabled, then libstdc++ is not built with its
1065 virtual calls in verifiable mode at all. However the libvtv library will
1066 still be built (see @option{--disable-libvtv} to turn off building libvtv).
1067 @option{--disable-vtable-verify} is the default.
1069 @item --disable-gcov
1070 Specify that the run-time library used for coverage analysis
1071 and associated host tools should not be built.
1073 @item --disable-multilib
1074 Specify that multiple target
1075 libraries to support different target variants, calling
1076 conventions, etc.@: should not be built. The default is to build a
1077 predefined set of them.
1079 Some targets provide finer-grained control over which multilibs are built
1080 (e.g., @option{--disable-softfloat}):
1083 fpu, 26bit, underscore, interwork, biendian, nofmult.
1086 softfloat, m68881, m68000, m68020.
1089 single-float, biendian, softfloat.
1094 @item powerpc*-*-*, rs6000*-*-*
1095 aix64, pthread, softfloat, powercpu, powerpccpu, powerpcos, biendian,
1100 @item --with-multilib-list=@var{list}
1101 @itemx --without-multilib-list
1102 Specify what multilibs to build. @var{list} is a comma separated list of
1103 values, possibly consisting of a single value. Currently only implemented
1104 for aarch64*-*-*, arm*-*-*, riscv*-*-*, sh*-*-* and x86-64-*-linux*. The
1105 accepted values and meaning for each target is given below.
1109 @var{list} is a comma separated list of @code{ilp32}, and @code{lp64}
1110 to enable ILP32 and LP64 run-time libraries, respectively. If
1111 @var{list} is empty, then there will be no multilibs and only the
1112 default run-time library will be built. If @var{list} is
1113 @code{default} or --with-multilib-list= is not specified, then the
1114 default set of libraries is selected based on the value of
1118 @var{list} is a comma separated list of @code{aprofile} and
1119 @code{rmprofile} to build multilibs for A or R and M architecture
1120 profiles respectively. Note that, due to some limitation of the current
1121 multilib framework, using the combined @code{aprofile,rmprofile}
1122 multilibs selects in some cases a less optimal multilib than when using
1123 the multilib profile for the architecture targetted. The special value
1124 @code{default} is also accepted and is equivalent to omitting the
1125 option, i.e., only the default run-time library will be enabled.
1127 @var{list} may instead contain @code{@@name}, to use the multilib
1128 configuration Makefile fragment @file{name} in @file{gcc/config/arm} in
1129 the source tree (it is part of the corresponding sources, after all).
1130 It is recommended, but not required, that files used for this purpose to
1131 be named starting with @file{t-ml-}, to make their intended purpose
1132 self-evident, in line with GCC conventions. Such files enable custom,
1133 user-chosen multilib lists to be configured. Whether multiple such
1134 files can be used together depends on the contents of the supplied
1135 files. See @file{gcc/config/arm/t-multilib} and its supplementary
1136 @file{gcc/config/arm/t-*profile} files for an example of what such
1137 Makefile fragments might look like for this version of GCC. The macros
1138 expected to be defined in these fragments are not stable across GCC
1139 releases, so make sure they define the @code{MULTILIB}-related macros
1140 expected by the version of GCC you are building.
1142 @xref{Target Fragment,, Target Makefile Fragments, gccint, GNU Compiler
1143 Collection (GCC) Internals}.
1146 See ``Target Makefile Fragments'' in the internals manual.
1149 The table below gives the combination of ISAs, architectures, FPUs and
1150 floating-point ABIs for which multilibs are built for each predefined
1151 profile. The union of these options is considered when specifying both
1152 @code{aprofile} and @code{rmprofile}.
1154 @multitable @columnfractions .15 .28 .30
1155 @item Option @tab aprofile @tab rmprofile
1157 @tab @code{-marm} and @code{-mthumb}
1159 @item Architectures@*@*@*@*@*@*
1160 @tab default architecture@*
1161 @code{-march=armv7-a}@*
1162 @code{-march=armv7ve}@*
1163 @code{-march=armv8-a}@*@*@*
1164 @tab default architecture@*
1165 @code{-march=armv6s-m}@*
1166 @code{-march=armv7-m}@*
1167 @code{-march=armv7e-m}@*
1168 @code{-march=armv8-m.base}@*
1169 @code{-march=armv8-m.main}@*
1171 @item FPUs@*@*@*@*@*
1173 @code{-mfpu=vfpv3-d16}@*
1175 @code{-mfpu=vfpv4-d16}@*
1176 @code{-mfpu=neon-vfpv4}@*
1177 @code{-mfpu=neon-fp-armv8}
1179 @code{-mfpu=vfpv3-d16}@*
1180 @code{-mfpu=fpv4-sp-d16}@*
1181 @code{-mfpu=fpv5-sp-d16}@*
1182 @code{-mfpu=fpv5-d16}@*
1183 @item floating-point@/ ABIs@*@*
1184 @tab @code{-mfloat-abi=soft}@*
1185 @code{-mfloat-abi=softfp}@*
1186 @code{-mfloat-abi=hard}
1187 @tab @code{-mfloat-abi=soft}@*
1188 @code{-mfloat-abi=softfp}@*
1189 @code{-mfloat-abi=hard}
1193 @var{list} is a single ABI name. The target architecture must be either
1194 @code{rv32gc} or @code{rv64gc}. This will build a single multilib for the
1195 specified architecture and ABI pair. If @code{--with-multilib-list} is not
1196 given, then a default set of multilibs is selected based on the value of
1197 @option{--target}. This is usually a large set of multilibs.
1200 @var{list} is a comma separated list of CPU names. These must be of the
1201 form @code{sh*} or @code{m*} (in which case they match the compiler option
1202 for that processor). The list should not contain any endian options -
1203 these are handled by @option{--with-endian}.
1205 If @var{list} is empty, then there will be no multilibs for extra
1206 processors. The multilib for the secondary endian remains enabled.
1208 As a special case, if an entry in the list starts with a @code{!}
1209 (exclamation point), then it is added to the list of excluded multilibs.
1210 Entries of this sort should be compatible with @samp{MULTILIB_EXCLUDES}
1211 (once the leading @code{!} has been stripped).
1213 If @option{--with-multilib-list} is not given, then a default set of
1214 multilibs is selected based on the value of @option{--target}. This is
1215 usually the complete set of libraries, but some targets imply a more
1218 Example 1: to configure a compiler for SH4A only, but supporting both
1219 endians, with little endian being the default:
1221 --with-cpu=sh4a --with-endian=little,big --with-multilib-list=
1224 Example 2: to configure a compiler for both SH4A and SH4AL-DSP, but with
1225 only little endian SH4AL:
1227 --with-cpu=sh4a --with-endian=little,big \
1228 --with-multilib-list=sh4al,!mb/m4al
1231 @item x86-64-*-linux*
1232 @var{list} is a comma separated list of @code{m32}, @code{m64} and
1233 @code{mx32} to enable 32-bit, 64-bit and x32 run-time libraries,
1234 respectively. If @var{list} is empty, then there will be no multilibs
1235 and only the default run-time library will be enabled.
1237 If @option{--with-multilib-list} is not given, then only 32-bit and
1238 64-bit run-time libraries will be enabled.
1241 @item --with-endian=@var{endians}
1242 Specify what endians to use.
1243 Currently only implemented for sh*-*-*.
1245 @var{endians} may be one of the following:
1248 Use big endian exclusively.
1250 Use little endian exclusively.
1252 Use big endian by default. Provide a multilib for little endian.
1254 Use little endian by default. Provide a multilib for big endian.
1257 @item --enable-threads
1258 Specify that the target
1259 supports threads. This affects the Objective-C compiler and runtime
1260 library, and exception handling for other languages like C++.
1261 On some systems, this is the default.
1263 In general, the best (and, in many cases, the only known) threading
1264 model available will be configured for use. Beware that on some
1265 systems, GCC has not been taught what threading models are generally
1266 available for the system. In this case, @option{--enable-threads} is an
1267 alias for @option{--enable-threads=single}.
1269 @item --disable-threads
1270 Specify that threading support should be disabled for the system.
1271 This is an alias for @option{--enable-threads=single}.
1273 @item --enable-threads=@var{lib}
1275 @var{lib} is the thread support library. This affects the Objective-C
1276 compiler and runtime library, and exception handling for other languages
1277 like C++. The possibilities for @var{lib} are:
1285 LynxOS thread support.
1287 MIPS SDE thread support.
1289 This is an alias for @samp{single}.
1291 Generic POSIX/Unix98 thread support.
1293 RTEMS thread support.
1295 Disable thread support, should work for all platforms.
1299 VxWorks thread support.
1301 Microsoft Win32 API thread support.
1305 Specify that the target supports TLS (Thread Local Storage). Usually
1306 configure can correctly determine if TLS is supported. In cases where
1307 it guesses incorrectly, TLS can be explicitly enabled or disabled with
1308 @option{--enable-tls} or @option{--disable-tls}. This can happen if
1309 the assembler supports TLS but the C library does not, or if the
1310 assumptions made by the configure test are incorrect.
1313 Specify that the target does not support TLS.
1314 This is an alias for @option{--enable-tls=no}.
1316 @item --disable-tm-clone-registry
1317 Disable TM clone registry in libgcc. It is enabled in libgcc by default.
1318 This option helps to reduce code size for embedded targets which do
1319 not use transactional memory.
1321 @item --with-cpu=@var{cpu}
1322 @itemx --with-cpu-32=@var{cpu}
1323 @itemx --with-cpu-64=@var{cpu}
1324 Specify which cpu variant the compiler should generate code for by default.
1325 @var{cpu} will be used as the default value of the @option{-mcpu=} switch.
1326 This option is only supported on some targets, including ARC, ARM, i386, M68k,
1327 PowerPC, and SPARC@. It is mandatory for ARC@. The @option{--with-cpu-32} and
1328 @option{--with-cpu-64} options specify separate default CPUs for
1329 32-bit and 64-bit modes; these options are only supported for i386,
1330 x86-64, PowerPC, and SPARC@.
1332 @item --with-schedule=@var{cpu}
1333 @itemx --with-arch=@var{cpu}
1334 @itemx --with-arch-32=@var{cpu}
1335 @itemx --with-arch-64=@var{cpu}
1336 @itemx --with-tune=@var{cpu}
1337 @itemx --with-tune-32=@var{cpu}
1338 @itemx --with-tune-64=@var{cpu}
1339 @itemx --with-abi=@var{abi}
1340 @itemx --with-fpu=@var{type}
1341 @itemx --with-float=@var{type}
1342 These configure options provide default values for the @option{-mschedule=},
1343 @option{-march=}, @option{-mtune=}, @option{-mabi=}, and @option{-mfpu=}
1344 options and for @option{-mhard-float} or @option{-msoft-float}. As with
1345 @option{--with-cpu}, which switches will be accepted and acceptable values
1346 of the arguments depend on the target.
1348 @item --with-mode=@var{mode}
1349 Specify if the compiler should default to @option{-marm} or @option{-mthumb}.
1350 This option is only supported on ARM targets.
1352 @item --with-stack-offset=@var{num}
1353 This option sets the default for the -mstack-offset=@var{num} option,
1354 and will thus generally also control the setting of this option for
1355 libraries. This option is only supported on Epiphany targets.
1357 @item --with-fpmath=@var{isa}
1358 This options sets @option{-mfpmath=sse} by default and specifies the default
1359 ISA for floating-point arithmetics. You can select either @samp{sse} which
1360 enables @option{-msse2} or @samp{avx} which enables @option{-mavx} by default.
1361 This option is only supported on i386 and x86-64 targets.
1363 @item --with-fp-32=@var{mode}
1364 On MIPS targets, set the default value for the @option{-mfp} option when using
1365 the o32 ABI. The possibilities for @var{mode} are:
1368 Use the o32 FP32 ABI extension, as with the @option{-mfp32} command-line
1371 Use the o32 FPXX ABI extension, as with the @option{-mfpxx} command-line
1374 Use the o32 FP64 ABI extension, as with the @option{-mfp64} command-line
1377 In the absence of this configuration option the default is to use the o32
1380 @item --with-odd-spreg-32
1381 On MIPS targets, set the @option{-modd-spreg} option by default when using
1384 @item --without-odd-spreg-32
1385 On MIPS targets, set the @option{-mno-odd-spreg} option by default when using
1386 the o32 ABI. This is normally used in conjunction with
1387 @option{--with-fp-32=64} in order to target the o32 FP64A ABI extension.
1389 @item --with-nan=@var{encoding}
1390 On MIPS targets, set the default encoding convention to use for the
1391 special not-a-number (NaN) IEEE 754 floating-point data. The
1392 possibilities for @var{encoding} are:
1395 Use the legacy encoding, as with the @option{-mnan=legacy} command-line
1398 Use the 754-2008 encoding, as with the @option{-mnan=2008} command-line
1401 To use this configuration option you must have an assembler version
1402 installed that supports the @option{-mnan=} command-line option too.
1403 In the absence of this configuration option the default convention is
1404 the legacy encoding, as when neither of the @option{-mnan=2008} and
1405 @option{-mnan=legacy} command-line options has been used.
1407 @item --with-divide=@var{type}
1408 Specify how the compiler should generate code for checking for
1409 division by zero. This option is only supported on the MIPS target.
1410 The possibilities for @var{type} are:
1413 Division by zero checks use conditional traps (this is the default on
1414 systems that support conditional traps).
1416 Division by zero checks use the break instruction.
1419 @c If you make --with-llsc the default for additional targets,
1420 @c update the --with-llsc description in the MIPS section below.
1423 On MIPS targets, make @option{-mllsc} the default when no
1424 @option{-mno-llsc} option is passed. This is the default for
1425 Linux-based targets, as the kernel will emulate them if the ISA does
1428 @item --without-llsc
1429 On MIPS targets, make @option{-mno-llsc} the default when no
1430 @option{-mllsc} option is passed.
1433 On MIPS targets, make @option{-msynci} the default when no
1434 @option{-mno-synci} option is passed.
1436 @item --without-synci
1437 On MIPS targets, make @option{-mno-synci} the default when no
1438 @option{-msynci} option is passed. This is the default.
1440 @item --with-lxc1-sxc1
1441 On MIPS targets, make @option{-mlxc1-sxc1} the default when no
1442 @option{-mno-lxc1-sxc1} option is passed. This is the default.
1444 @item --without-lxc1-sxc1
1445 On MIPS targets, make @option{-mno-lxc1-sxc1} the default when no
1446 @option{-mlxc1-sxc1} option is passed. The indexed load/store
1447 instructions are not directly a problem but can lead to unexpected
1448 behaviour when deployed in an application intended for a 32-bit address
1449 space but run on a 64-bit processor. The issue is seen because all
1450 known MIPS 64-bit Linux kernels execute o32 and n32 applications
1451 with 64-bit addressing enabled which affects the overflow behaviour
1452 of the indexed addressing mode. GCC will assume that ordinary
1453 32-bit arithmetic overflow behaviour is the same whether performed
1454 as an @code{addu} instruction or as part of the address calculation
1455 in @code{lwxc1} type instructions. This assumption holds true in a
1456 pure 32-bit environment and can hold true in a 64-bit environment if
1457 the address space is accurately set to be 32-bit for o32 and n32.
1460 On MIPS targets, make @option{-mmadd4} the default when no
1461 @option{-mno-madd4} option is passed. This is the default.
1463 @item --without-madd4
1464 On MIPS targets, make @option{-mno-madd4} the default when no
1465 @option{-mmadd4} option is passed. The @code{madd4} instruction
1466 family can be problematic when targeting a combination of cores that
1467 implement these instructions differently. There are two known cores
1468 that implement these as fused operations instead of unfused (where
1469 unfused is normally expected). Disabling these instructions is the
1470 only way to ensure compatible code is generated; this will incur
1471 a performance penalty.
1473 @item --with-mips-plt
1474 On MIPS targets, make use of copy relocations and PLTs.
1475 These features are extensions to the traditional
1476 SVR4-based MIPS ABIs and require support from GNU binutils
1477 and the runtime C library.
1479 @item --with-stack-clash-protection-guard-size=@var{size}
1480 On certain targets this option sets the default stack clash protection guard
1481 size as a power of two in bytes. On AArch64 @var{size} is required to be either
1482 12 (4KB) or 16 (64KB).
1484 @item --enable-__cxa_atexit
1485 Define if you want to use __cxa_atexit, rather than atexit, to
1486 register C++ destructors for local statics and global objects.
1487 This is essential for fully standards-compliant handling of
1488 destructors, but requires __cxa_atexit in libc. This option is currently
1489 only available on systems with GNU libc. When enabled, this will cause
1490 @option{-fuse-cxa-atexit} to be passed by default.
1492 @item --enable-gnu-indirect-function
1493 Define if you want to enable the @code{ifunc} attribute. This option is
1494 currently only available on systems with GNU libc on certain targets.
1496 @item --enable-target-optspace
1498 libraries should be optimized for code space instead of code speed.
1499 This is the default for the m32r platform.
1501 @item --with-cpp-install-dir=@var{dirname}
1502 Specify that the user visible @command{cpp} program should be installed
1503 in @file{@var{prefix}/@var{dirname}/cpp}, in addition to @var{bindir}.
1505 @item --enable-comdat
1506 Enable COMDAT group support. This is primarily used to override the
1507 automatically detected value.
1509 @item --enable-initfini-array
1510 Force the use of sections @code{.init_array} and @code{.fini_array}
1511 (instead of @code{.init} and @code{.fini}) for constructors and
1512 destructors. Option @option{--disable-initfini-array} has the
1513 opposite effect. If neither option is specified, the configure script
1514 will try to guess whether the @code{.init_array} and
1515 @code{.fini_array} sections are supported and, if they are, use them.
1517 @item --enable-link-mutex
1518 When building GCC, use a mutex to avoid linking the compilers for
1519 multiple languages at the same time, to avoid thrashing on build
1520 systems with limited free memory. The default is not to use such a mutex.
1522 @item --enable-maintainer-mode
1523 The build rules that regenerate the Autoconf and Automake output files as
1524 well as the GCC master message catalog @file{gcc.pot} are normally
1525 disabled. This is because it can only be rebuilt if the complete source
1526 tree is present. If you have changed the sources and want to rebuild the
1527 catalog, configuring with @option{--enable-maintainer-mode} will enable
1528 this. Note that you need a recent version of the @code{gettext} tools
1531 @item --disable-bootstrap
1532 For a native build, the default configuration is to perform
1533 a 3-stage bootstrap of the compiler when @samp{make} is invoked,
1534 testing that GCC can compile itself correctly. If you want to disable
1535 this process, you can configure with @option{--disable-bootstrap}.
1537 @item --enable-bootstrap
1538 In special cases, you may want to perform a 3-stage build
1539 even if the target and host triplets are different.
1540 This is possible when the host can run code compiled for
1541 the target (e.g.@: host is i686-linux, target is i486-linux).
1542 Starting from GCC 4.2, to do this you have to configure explicitly
1543 with @option{--enable-bootstrap}.
1545 @item --enable-generated-files-in-srcdir
1546 Neither the .c and .h files that are generated from Bison and flex nor the
1547 info manuals and man pages that are built from the .texi files are present
1548 in the repository development tree. When building GCC from that development tree,
1549 or from one of our snapshots, those generated files are placed in your
1550 build directory, which allows for the source to be in a readonly
1553 If you configure with @option{--enable-generated-files-in-srcdir} then those
1554 generated files will go into the source directory. This is mainly intended
1555 for generating release or prerelease tarballs of the GCC sources, since it
1556 is not a requirement that the users of source releases to have flex, Bison,
1559 @item --enable-version-specific-runtime-libs
1561 that runtime libraries should be installed in the compiler specific
1562 subdirectory (@file{@var{libdir}/gcc}) rather than the usual places. In
1563 addition, @samp{libstdc++}'s include files will be installed into
1564 @file{@var{libdir}} unless you overruled it by using
1565 @option{--with-gxx-include-dir=@var{dirname}}. Using this option is
1566 particularly useful if you intend to use several versions of GCC in
1567 parallel. The default is @samp{yes} for @samp{libada}, and @samp{no} for
1568 the remaining libraries.
1570 @item @anchor{WithAixSoname}--with-aix-soname=@samp{aix}, @samp{svr4} or @samp{both}
1571 Traditional AIX shared library versioning (versioned @code{Shared Object}
1572 files as members of unversioned @code{Archive Library} files named
1573 @samp{lib.a}) causes numerous headaches for package managers. However,
1574 @code{Import Files} as members of @code{Archive Library} files allow for
1575 @strong{filename-based versioning} of shared libraries as seen on Linux/SVR4,
1576 where this is called the "SONAME". But as they prevent static linking,
1577 @code{Import Files} may be used with @code{Runtime Linking} only, where the
1578 linker does search for @samp{libNAME.so} before @samp{libNAME.a} library
1579 filenames with the @samp{-lNAME} linker flag.
1581 @anchor{AixLdCommand}For detailed information please refer to the AIX
1582 @uref{https://www.ibm.com/support/knowledgecenter/search/%22the%20ld%20command%2C%20also%20called%20the%20linkage%20editor%20or%20binder%22,,ld
1585 As long as shared library creation is enabled, upon:
1587 @item --with-aix-soname=aix
1588 @item --with-aix-soname=both
1589 A (traditional AIX) @code{Shared Archive Library} file is created:
1591 @item using the @samp{libNAME.a} filename scheme
1592 @item with the @code{Shared Object} file as archive member named
1593 @samp{libNAME.so.V} (except for @samp{libgcc_s}, where the @code{Shared
1594 Object} file is named @samp{shr.o} for backwards compatibility), which
1596 @item is used for runtime loading from inside the @samp{libNAME.a} file
1597 @item is used for dynamic loading via
1598 @code{dlopen("libNAME.a(libNAME.so.V)", RTLD_MEMBER)}
1599 @item is used for shared linking
1600 @item is used for static linking, so no separate @code{Static Archive
1601 Library} file is needed
1604 @item --with-aix-soname=both
1605 @item --with-aix-soname=svr4
1606 A (second) @code{Shared Archive Library} file is created:
1608 @item using the @samp{libNAME.so.V} filename scheme
1609 @item with the @code{Shared Object} file as archive member named
1612 @item is created with the @code{-G linker flag}
1613 @item has the @code{F_LOADONLY} flag set
1614 @item is used for runtime loading from inside the @samp{libNAME.so.V} file
1615 @item is used for dynamic loading via @code{dlopen("libNAME.so.V(shr.o)",
1618 @item with the @code{Import File} as archive member named @samp{shr.imp},
1621 @item refers to @samp{libNAME.so.V(shr.o)} as the "SONAME", to be recorded
1622 in the @code{Loader Section} of subsequent binaries
1623 @item indicates whether @samp{libNAME.so.V(shr.o)} is 32 or 64 bit
1624 @item lists all the public symbols exported by @samp{lib.so.V(shr.o)},
1625 eventually decorated with the @code{@samp{weak} Keyword}
1626 @item is necessary for shared linking against @samp{lib.so.V(shr.o)}
1629 A symbolic link using the @samp{libNAME.so} filename scheme is created:
1631 @item pointing to the @samp{libNAME.so.V} @code{Shared Archive Library} file
1632 @item to permit the @code{ld Command} to find @samp{lib.so.V(shr.imp)} via
1633 the @samp{-lNAME} argument (requires @code{Runtime Linking} to be enabled)
1634 @item to permit dynamic loading of @samp{lib.so.V(shr.o)} without the need
1635 to specify the version number via @code{dlopen("libNAME.so(shr.o)",
1640 As long as static library creation is enabled, upon:
1642 @item --with-aix-soname=svr4
1643 A @code{Static Archive Library} is created:
1645 @item using the @samp{libNAME.a} filename scheme
1646 @item with all the @code{Static Object} files as archive members, which
1648 @item are used for static linking
1653 While the aix-soname=@samp{svr4} option does not create @code{Shared Object}
1654 files as members of unversioned @code{Archive Library} files any more, package
1655 managers still are responsible to
1656 @uref{./specific.html#TransferAixShobj,,transfer} @code{Shared Object} files
1657 found as member of a previously installed unversioned @code{Archive Library}
1658 file into the newly installed @code{Archive Library} file with the same
1661 @emph{WARNING:} Creating @code{Shared Object} files with @code{Runtime Linking}
1662 enabled may bloat the TOC, eventually leading to @code{TOC overflow} errors,
1663 requiring the use of either the @option{-Wl,-bbigtoc} linker flag (seen to
1664 break with the @code{GDB} debugger) or some of the TOC-related compiler flags,
1666 @xref{RS/6000 and PowerPC Options,, RS/6000 and PowerPC Options, gcc,
1667 Using the GNU Compiler Collection (GCC)}.
1670 see ``RS/6000 and PowerPC Options'' in the main manual.
1673 @option{--with-aix-soname} is currently supported by @samp{libgcc_s} only, so
1674 this option is still experimental and not for normal use yet.
1676 Default is the traditional behavior @option{--with-aix-soname=@samp{aix}}.
1678 @item --enable-languages=@var{lang1},@var{lang2},@dots{}
1679 Specify that only a particular subset of compilers and
1680 their runtime libraries should be built. For a list of valid values for
1681 @var{langN} you can issue the following command in the
1682 @file{gcc} directory of your GCC source tree:@*
1684 grep ^language= */config-lang.in
1686 Currently, you can use any of the following:
1687 @code{all}, @code{default}, @code{ada}, @code{c}, @code{c++}, @code{d},
1688 @code{fortran}, @code{go}, @code{jit}, @code{lto}, @code{objc}, @code{obj-c++}.
1689 Building the Ada compiler has special requirements, see below.
1690 If you do not pass this flag, or specify the option @code{default}, then the
1691 default languages available in the @file{gcc} sub-tree will be configured.
1692 Ada, D, Go, Jit, and Objective-C++ are not default languages. LTO is not a
1693 default language, but is built by default because @option{--enable-lto} is
1694 enabled by default. The other languages are default languages. If
1695 @code{all} is specified, then all available languages are built. An
1696 exception is @code{jit} language, which requires
1697 @option{--enable-host-shared} to be included with @code{all}.
1699 @item --enable-stage1-languages=@var{lang1},@var{lang2},@dots{}
1700 Specify that a particular subset of compilers and their runtime
1701 libraries should be built with the system C compiler during stage 1 of
1702 the bootstrap process, rather than only in later stages with the
1703 bootstrapped C compiler. The list of valid values is the same as for
1704 @option{--enable-languages}, and the option @code{all} will select all
1705 of the languages enabled by @option{--enable-languages}. This option is
1706 primarily useful for GCC development; for instance, when a development
1707 version of the compiler cannot bootstrap due to compiler bugs, or when
1708 one is debugging front ends other than the C front end. When this
1709 option is used, one can then build the target libraries for the
1710 specified languages with the stage-1 compiler by using @command{make
1711 stage1-bubble all-target}, or run the testsuite on the stage-1 compiler
1712 for the specified languages using @command{make stage1-start check-gcc}.
1714 @item --disable-libada
1715 Specify that the run-time libraries and tools used by GNAT should not
1716 be built. This can be useful for debugging, or for compatibility with
1717 previous Ada build procedures, when it was required to explicitly
1718 do a @samp{make -C gcc gnatlib_and_tools}.
1720 @item --disable-libsanitizer
1721 Specify that the run-time libraries for the various sanitizers should
1724 @item --disable-libssp
1725 Specify that the run-time libraries for stack smashing protection
1726 should not be built or linked against. On many targets library support
1727 is provided by the C library instead.
1729 @item --disable-libquadmath
1730 Specify that the GCC quad-precision math library should not be built.
1731 On some systems, the library is required to be linkable when building
1732 the Fortran front end, unless @option{--disable-libquadmath-support}
1735 @item --disable-libquadmath-support
1736 Specify that the Fortran front end and @code{libgfortran} do not add
1737 support for @code{libquadmath} on systems supporting it.
1739 @item --disable-libgomp
1740 Specify that the GNU Offloading and Multi Processing Runtime Library
1741 should not be built.
1743 @item --disable-libvtv
1744 Specify that the run-time libraries used by vtable verification
1745 should not be built.
1748 Specify that the compiler should
1749 use DWARF 2 debugging information as the default.
1751 @item --with-advance-toolchain=@var{at}
1752 On 64-bit PowerPC Linux systems, configure the compiler to use the
1753 header files, library files, and the dynamic linker from the Advance
1754 Toolchain release @var{at} instead of the default versions that are
1755 provided by the Linux distribution. In general, this option is
1756 intended for the developers of GCC, and it is not intended for general
1759 @item --enable-targets=all
1760 @itemx --enable-targets=@var{target_list}
1761 Some GCC targets, e.g.@: powerpc64-linux, build bi-arch compilers.
1762 These are compilers that are able to generate either 64-bit or 32-bit
1763 code. Typically, the corresponding 32-bit target, e.g.@:
1764 powerpc-linux for powerpc64-linux, only generates 32-bit code. This
1765 option enables the 32-bit target to be a bi-arch compiler, which is
1766 useful when you want a bi-arch compiler that defaults to 32-bit, and
1767 you are building a bi-arch or multi-arch binutils in a combined tree.
1768 On mips-linux, this will build a tri-arch compiler (ABI o32/n32/64),
1770 Currently, this option only affects sparc-linux, powerpc-linux, x86-linux,
1771 mips-linux and s390-linux.
1773 @item --enable-default-pie
1774 Turn on @option{-fPIE} and @option{-pie} by default.
1776 @item --enable-secureplt
1777 This option enables @option{-msecure-plt} by default for powerpc-linux.
1779 @xref{RS/6000 and PowerPC Options,, RS/6000 and PowerPC Options, gcc,
1780 Using the GNU Compiler Collection (GCC)},
1783 See ``RS/6000 and PowerPC Options'' in the main manual
1786 @item --enable-default-ssp
1787 Turn on @option{-fstack-protector-strong} by default.
1790 This option enables @option{-mcld} by default for 32-bit x86 targets.
1792 @xref{i386 and x86-64 Options,, i386 and x86-64 Options, gcc,
1793 Using the GNU Compiler Collection (GCC)},
1796 See ``i386 and x86-64 Options'' in the main manual
1799 @item --enable-large-address-aware
1800 The @option{--enable-large-address-aware} option arranges for MinGW
1801 executables to be linked using the @option{--large-address-aware}
1802 option, that enables the use of more than 2GB of memory. If GCC is
1803 configured with this option, its effects can be reversed by passing the
1804 @option{-Wl,--disable-large-address-aware} option to the so-configured
1807 @item --enable-win32-registry
1808 @itemx --enable-win32-registry=@var{key}
1809 @itemx --disable-win32-registry
1810 The @option{--enable-win32-registry} option enables Microsoft Windows-hosted GCC
1811 to look up installations paths in the registry using the following key:
1814 @code{HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\@var{key}}
1817 @var{key} defaults to GCC version number, and can be overridden by the
1818 @option{--enable-win32-registry=@var{key}} option. Vendors and distributors
1819 who use custom installers are encouraged to provide a different key,
1820 perhaps one comprised of vendor name and GCC version number, to
1821 avoid conflict with existing installations. This feature is enabled
1822 by default, and can be disabled by @option{--disable-win32-registry}
1823 option. This option has no effect on the other hosts.
1826 Specify that the machine does not have a floating point unit. This
1827 option only applies to @samp{m68k-sun-sunos@var{n}}. On any other
1828 system, @option{--nfp} has no effect.
1830 @item --enable-werror
1831 @itemx --disable-werror
1832 @itemx --enable-werror=yes
1833 @itemx --enable-werror=no
1834 When you specify this option, it controls whether certain files in the
1835 compiler are built with @option{-Werror} in bootstrap stage2 and later.
1836 If you don't specify it, @option{-Werror} is turned on for the main
1837 development trunk. However it defaults to off for release branches and
1838 final releases. The specific files which get @option{-Werror} are
1839 controlled by the Makefiles.
1841 @item --enable-checking
1842 @itemx --disable-checking
1843 @itemx --enable-checking=@var{list}
1844 This option controls performing internal consistency checks in the compiler.
1845 It does not change the generated code, but adds error checking of the
1846 requested complexity. This slows down the compiler and may only work
1847 properly if you are building the compiler with GCC@.
1849 When the option is not specified, the active set of checks depends on context.
1850 Namely, bootstrap stage 1 defaults to @samp{--enable-checking=yes}, builds
1851 from release branches or release archives default to
1852 @samp{--enable-checking=release}, and otherwise
1853 @samp{--enable-checking=yes,extra} is used. When the option is
1854 specified without a @var{list}, the result is the same as
1855 @samp{--enable-checking=yes}. Likewise, @samp{--disable-checking} is
1856 equivalent to @samp{--enable-checking=no}.
1858 The categories of checks available in @var{list} are @samp{yes} (most common
1859 checks @samp{assert,misc,gc,gimple,rtlflag,runtime,tree,types}), @samp{no}
1860 (no checks at all), @samp{all} (all but @samp{valgrind}), @samp{release}
1861 (cheapest checks @samp{assert,runtime}) or @samp{none} (same as @samp{no}).
1862 @samp{release} checks are always on and to disable them
1863 @samp{--disable-checking} or @samp{--enable-checking=no[,<other checks>]}
1864 must be explicitly requested. Disabling assertions makes the compiler and
1865 runtime slightly faster but increases the risk of undetected internal errors
1866 causing wrong code to be generated.
1868 Individual checks can be enabled with these flags: @samp{assert}, @samp{df},
1869 @samp{extra}, @samp{fold}, @samp{gc}, @samp{gcac}, @samp{gimple},
1870 @samp{misc}, @samp{rtl}, @samp{rtlflag}, @samp{runtime}, @samp{tree},
1871 @samp{types} and @samp{valgrind}. @samp{extra} extends @samp{misc}
1872 checking with extra checks that might affect code generation and should
1873 therefore not differ between stage1 and later stages in bootstrap.
1875 The @samp{valgrind} check requires the external @command{valgrind} simulator,
1876 available from @uref{http://valgrind.org/}. The @samp{rtl} checks are
1877 expensive and the @samp{df}, @samp{gcac} and @samp{valgrind} checks are very
1880 @item --disable-stage1-checking
1881 @itemx --enable-stage1-checking
1882 @itemx --enable-stage1-checking=@var{list}
1883 This option affects only bootstrap build. If no @option{--enable-checking}
1884 option is specified the stage1 compiler is built with @samp{yes} checking
1885 enabled, otherwise the stage1 checking flags are the same as specified by
1886 @option{--enable-checking}. To build the stage1 compiler with
1887 different checking options use @option{--enable-stage1-checking}.
1888 The list of checking options is the same as for @option{--enable-checking}.
1889 If your system is too slow or too small to bootstrap a released compiler
1890 with checking for stage1 enabled, you can use @samp{--disable-stage1-checking}
1891 to disable checking for the stage1 compiler.
1893 @item --enable-coverage
1894 @itemx --enable-coverage=@var{level}
1895 With this option, the compiler is built to collect self coverage
1896 information, every time it is run. This is for internal development
1897 purposes, and only works when the compiler is being built with gcc. The
1898 @var{level} argument controls whether the compiler is built optimized or
1899 not, values are @samp{opt} and @samp{noopt}. For coverage analysis you
1900 want to disable optimization, for performance analysis you want to
1901 enable optimization. When coverage is enabled, the default level is
1902 without optimization.
1904 @item --enable-gather-detailed-mem-stats
1905 When this option is specified more detailed information on memory
1906 allocation is gathered. This information is printed when using
1907 @option{-fmem-report}.
1909 @item --enable-valgrind-annotations
1910 Mark selected memory related operations in the compiler when run under
1911 valgrind to suppress false positives.
1914 @itemx --disable-nls
1915 The @option{--enable-nls} option enables Native Language Support (NLS),
1916 which lets GCC output diagnostics in languages other than American
1917 English. Native Language Support is enabled by default if not doing a
1918 canadian cross build. The @option{--disable-nls} option disables NLS@.
1920 @item --with-included-gettext
1921 If NLS is enabled, the @option{--with-included-gettext} option causes the build
1922 procedure to prefer its copy of GNU @command{gettext}.
1924 @item --with-catgets
1925 If NLS is enabled, and if the host lacks @code{gettext} but has the
1926 inferior @code{catgets} interface, the GCC build procedure normally
1927 ignores @code{catgets} and instead uses GCC's copy of the GNU
1928 @code{gettext} library. The @option{--with-catgets} option causes the
1929 build procedure to use the host's @code{catgets} in this situation.
1931 @item --with-libiconv-prefix=@var{dir}
1932 Search for libiconv header files in @file{@var{dir}/include} and
1933 libiconv library files in @file{@var{dir}/lib}.
1935 @item --enable-obsolete
1936 Enable configuration for an obsoleted system. If you attempt to
1937 configure GCC for a system (build, host, or target) which has been
1938 obsoleted, and you do not specify this flag, configure will halt with an
1941 All support for systems which have been obsoleted in one release of GCC
1942 is removed entirely in the next major release, unless someone steps
1943 forward to maintain the port.
1945 @item --enable-decimal-float
1946 @itemx --enable-decimal-float=yes
1947 @itemx --enable-decimal-float=no
1948 @itemx --enable-decimal-float=bid
1949 @itemx --enable-decimal-float=dpd
1950 @itemx --disable-decimal-float
1951 Enable (or disable) support for the C decimal floating point extension
1952 that is in the IEEE 754-2008 standard. This is enabled by default only
1953 on PowerPC, i386, and x86_64 GNU/Linux systems. Other systems may also
1954 support it, but require the user to specifically enable it. You can
1955 optionally control which decimal floating point format is used (either
1956 @samp{bid} or @samp{dpd}). The @samp{bid} (binary integer decimal)
1957 format is default on i386 and x86_64 systems, and the @samp{dpd}
1958 (densely packed decimal) format is default on PowerPC systems.
1960 @item --enable-fixed-point
1961 @itemx --disable-fixed-point
1962 Enable (or disable) support for C fixed-point arithmetic.
1963 This option is enabled by default for some targets (such as MIPS) which
1964 have hardware-support for fixed-point operations. On other targets, you
1965 may enable this option manually.
1967 @item --with-long-double-128
1968 Specify if @code{long double} type should be 128-bit by default on selected
1969 GNU/Linux architectures. If using @code{--without-long-double-128},
1970 @code{long double} will be by default 64-bit, the same as @code{double} type.
1971 When neither of these configure options are used, the default will be
1972 128-bit @code{long double} when built against GNU C Library 2.4 and later,
1973 64-bit @code{long double} otherwise.
1975 @item --with-long-double-format=ibm
1976 @itemx --with-long-double-format=ieee
1977 Specify whether @code{long double} uses the IBM extended double format
1978 or the IEEE 128-bit floating point format on PowerPC Linux systems.
1979 This configuration switch will only work on little endian PowerPC
1980 Linux systems and on big endian 64-bit systems where the default cpu
1981 is at least power7 (i.e.@: @option{--with-cpu=power7},
1982 @option{--with-cpu=power8}, or @option{--with-cpu=power9} is used).
1984 If you use the @option{--with-long-double-64} configuration option,
1985 the @option{--with-long-double-format=ibm} and
1986 @option{--with-long-double-format=ieee} options are ignored.
1988 The default @code{long double} format is to use IBM extended double.
1989 Until all of the libraries are converted to use IEEE 128-bit floating
1990 point, it is not recommended to use
1991 @option{--with-long-double-format=ieee}.
1993 On little endian PowerPC Linux systems, if you explicitly set the
1994 @code{long double} type, it will build multilibs to allow you to
1995 select either @code{long double} format, unless you disable multilibs
1996 with the @code{--disable-multilib} option. At present,
1997 @code{long double} multilibs are not built on big endian PowerPC Linux
1998 systems. If you are building multilibs, you will need to configure
1999 the compiler using the @option{--with-system-zlib} option.
2001 If you do not set the @code{long double} type explicitly, no multilibs
2004 @item --enable-fdpic
2005 On SH Linux systems, generate ELF FDPIC code.
2007 @item --with-gmp=@var{pathname}
2008 @itemx --with-gmp-include=@var{pathname}
2009 @itemx --with-gmp-lib=@var{pathname}
2010 @itemx --with-mpfr=@var{pathname}
2011 @itemx --with-mpfr-include=@var{pathname}
2012 @itemx --with-mpfr-lib=@var{pathname}
2013 @itemx --with-mpc=@var{pathname}
2014 @itemx --with-mpc-include=@var{pathname}
2015 @itemx --with-mpc-lib=@var{pathname}
2016 If you want to build GCC but do not have the GMP library, the MPFR
2017 library and/or the MPC library installed in a standard location and
2018 do not have their sources present in the GCC source tree then you
2019 can explicitly specify the directory where they are installed
2020 (@samp{--with-gmp=@var{gmpinstalldir}},
2021 @samp{--with-mpfr=@/@var{mpfrinstalldir}},
2022 @samp{--with-mpc=@/@var{mpcinstalldir}}). The
2023 @option{--with-gmp=@/@var{gmpinstalldir}} option is shorthand for
2024 @option{--with-gmp-lib=@/@var{gmpinstalldir}/lib} and
2025 @option{--with-gmp-include=@/@var{gmpinstalldir}/include}. Likewise the
2026 @option{--with-mpfr=@/@var{mpfrinstalldir}} option is shorthand for
2027 @option{--with-mpfr-lib=@/@var{mpfrinstalldir}/lib} and
2028 @option{--with-mpfr-include=@/@var{mpfrinstalldir}/include}, also the
2029 @option{--with-mpc=@/@var{mpcinstalldir}} option is shorthand for
2030 @option{--with-mpc-lib=@/@var{mpcinstalldir}/lib} and
2031 @option{--with-mpc-include=@/@var{mpcinstalldir}/include}. If these
2032 shorthand assumptions are not correct, you can use the explicit
2033 include and lib options directly. You might also need to ensure the
2034 shared libraries can be found by the dynamic linker when building and
2035 using GCC, for example by setting the runtime shared library path
2036 variable (@env{LD_LIBRARY_PATH} on GNU/Linux and Solaris systems).
2038 These flags are applicable to the host platform only. When building
2039 a cross compiler, they will not be used to configure target libraries.
2041 @item --with-isl=@var{pathname}
2042 @itemx --with-isl-include=@var{pathname}
2043 @itemx --with-isl-lib=@var{pathname}
2044 If you do not have the isl library installed in a standard location and you
2045 want to build GCC, you can explicitly specify the directory where it is
2046 installed (@samp{--with-isl=@/@var{islinstalldir}}). The
2047 @option{--with-isl=@/@var{islinstalldir}} option is shorthand for
2048 @option{--with-isl-lib=@/@var{islinstalldir}/lib} and
2049 @option{--with-isl-include=@/@var{islinstalldir}/include}. If this
2050 shorthand assumption is not correct, you can use the explicit
2051 include and lib options directly.
2053 These flags are applicable to the host platform only. When building
2054 a cross compiler, they will not be used to configure target libraries.
2056 @item --with-stage1-ldflags=@var{flags}
2057 This option may be used to set linker flags to be used when linking
2058 stage 1 of GCC. These are also used when linking GCC if configured with
2059 @option{--disable-bootstrap}. If @option{--with-stage1-libs} is not set to a
2060 value, then the default is @samp{-static-libstdc++ -static-libgcc}, if
2063 @item --with-stage1-libs=@var{libs}
2064 This option may be used to set libraries to be used when linking stage 1
2065 of GCC. These are also used when linking GCC if configured with
2066 @option{--disable-bootstrap}.
2068 @item --with-boot-ldflags=@var{flags}
2069 This option may be used to set linker flags to be used when linking
2070 stage 2 and later when bootstrapping GCC. If --with-boot-libs
2071 is not is set to a value, then the default is
2072 @samp{-static-libstdc++ -static-libgcc}.
2074 @item --with-boot-libs=@var{libs}
2075 This option may be used to set libraries to be used when linking stage 2
2076 and later when bootstrapping GCC.
2078 @item --with-debug-prefix-map=@var{map}
2079 Convert source directory names using @option{-fdebug-prefix-map} when
2080 building runtime libraries. @samp{@var{map}} is a space-separated
2081 list of maps of the form @samp{@var{old}=@var{new}}.
2083 @item --enable-linker-build-id
2084 Tells GCC to pass @option{--build-id} option to the linker for all final
2085 links (links performed without the @option{-r} or @option{--relocatable}
2086 option), if the linker supports it. If you specify
2087 @option{--enable-linker-build-id}, but your linker does not
2088 support @option{--build-id} option, a warning is issued and the
2089 @option{--enable-linker-build-id} option is ignored. The default is off.
2091 @item --with-linker-hash-style=@var{choice}
2092 Tells GCC to pass @option{--hash-style=@var{choice}} option to the
2093 linker for all final links. @var{choice} can be one of
2094 @samp{sysv}, @samp{gnu}, and @samp{both} where @samp{sysv} is the default.
2096 @item --enable-gnu-unique-object
2097 @itemx --disable-gnu-unique-object
2098 Tells GCC to use the gnu_unique_object relocation for C++ template
2099 static data members and inline function local statics. Enabled by
2100 default for a toolchain with an assembler that accepts it and
2101 GLIBC 2.11 or above, otherwise disabled.
2103 @item --with-diagnostics-color=@var{choice}
2104 Tells GCC to use @var{choice} as the default for @option{-fdiagnostics-color=}
2105 option (if not used explicitly on the command line). @var{choice}
2106 can be one of @samp{never}, @samp{auto}, @samp{always}, and @samp{auto-if-env}
2107 where @samp{auto} is the default. @samp{auto-if-env} makes
2108 @option{-fdiagnostics-color=auto} the default if @env{GCC_COLORS}
2109 is present and non-empty in the environment of the compiler, and
2110 @option{-fdiagnostics-color=never} otherwise.
2112 @item --with-diagnostics-urls=@var{choice}
2113 Tells GCC to use @var{choice} as the default for @option{-fdiagnostics-urls=}
2114 option (if not used explicitly on the command line). @var{choice}
2115 can be one of @samp{never}, @samp{auto}, @samp{always}, and @samp{auto-if-env}
2116 where @samp{auto} is the default. @samp{auto-if-env} makes
2117 @option{-fdiagnostics-urls=auto} the default if @env{GCC_URLS}
2118 or @env{TERM_URLS} is present and non-empty in the environment of the
2119 compiler, and @option{-fdiagnostics-urls=never} otherwise.
2122 @itemx --disable-lto
2123 Enable support for link-time optimization (LTO). This is enabled by
2124 default, and may be disabled using @option{--disable-lto}.
2126 @item --enable-linker-plugin-configure-flags=FLAGS
2127 @itemx --enable-linker-plugin-flags=FLAGS
2128 By default, linker plugins (such as the LTO plugin) are built for the
2129 host system architecture. For the case that the linker has a
2130 different (but run-time compatible) architecture, these flags can be
2131 specified to build plugins that are compatible to the linker. For
2132 example, if you are building GCC for a 64-bit x86_64
2133 (@samp{x86_64-pc-linux-gnu}) host system, but have a 32-bit x86
2134 GNU/Linux (@samp{i686-pc-linux-gnu}) linker executable (which is
2135 executable on the former system), you can configure GCC as follows for
2136 getting compatible linker plugins:
2139 % @var{srcdir}/configure \
2140 --host=x86_64-pc-linux-gnu \
2141 --enable-linker-plugin-configure-flags=--host=i686-pc-linux-gnu \
2142 --enable-linker-plugin-flags='CC=gcc\ -m32\ -Wl,-rpath,[...]/i686-pc-linux-gnu/lib'
2145 @item --with-plugin-ld=@var{pathname}
2146 Enable an alternate linker to be used at link-time optimization (LTO)
2147 link time when @option{-fuse-linker-plugin} is enabled.
2148 This linker should have plugin support such as gold starting with
2149 version 2.20 or GNU ld starting with version 2.21.
2150 See @option{-fuse-linker-plugin} for details.
2152 @item --enable-canonical-system-headers
2153 @itemx --disable-canonical-system-headers
2154 Enable system header path canonicalization for @file{libcpp}. This can
2155 produce shorter header file paths in diagnostics and dependency output
2156 files, but these changed header paths may conflict with some compilation
2157 environments. Enabled by default, and may be disabled using
2158 @option{--disable-canonical-system-headers}.
2160 @item --with-glibc-version=@var{major}.@var{minor}
2161 Tell GCC that when the GNU C Library (glibc) is used on the target it
2162 will be version @var{major}.@var{minor} or later. Normally this can
2163 be detected from the C library's header files, but this option may be
2164 needed when bootstrapping a cross toolchain without the header files
2165 available for building the initial bootstrap compiler.
2167 If GCC is configured with some multilibs that use glibc and some that
2168 do not, this option applies only to the multilibs that use glibc.
2169 However, such configurations may not work well as not all the relevant
2170 configuration in GCC is on a per-multilib basis.
2172 @item --enable-as-accelerator-for=@var{target}
2173 Build as offload target compiler. Specify offload host triple by @var{target}.
2175 @item --enable-offload-targets=@var{target1}[=@var{path1}],@dots{},@var{targetN}[=@var{pathN}]
2176 Enable offloading to targets @var{target1}, @dots{}, @var{targetN}.
2177 Offload compilers are expected to be already installed. Default search
2178 path for them is @file{@var{exec-prefix}}, but it can be changed by
2179 specifying paths @var{path1}, @dots{}, @var{pathN}.
2182 % @var{srcdir}/configure \
2183 --enable-offload-targets=x86_64-intelmicemul-linux-gnu=/path/to/x86_64/compiler,nvptx-none,hsa
2186 If @samp{hsa} is specified as one of the targets, the compiler will be
2187 built with support for HSA GPU accelerators. Because the same
2188 compiler will emit the accelerator code, no path should be specified.
2190 @item --with-hsa-runtime=@var{pathname}
2191 @itemx --with-hsa-runtime-include=@var{pathname}
2192 @itemx --with-hsa-runtime-lib=@var{pathname}
2194 If you configure GCC with HSA offloading but do not have the HSA
2195 run-time library installed in a standard location then you can
2196 explicitly specify the directory where they are installed. The
2197 @option{--with-hsa-runtime=@/@var{hsainstalldir}} option is a
2199 @option{--with-hsa-runtime-lib=@/@var{hsainstalldir}/lib} and
2200 @option{--with-hsa-runtime-include=@/@var{hsainstalldir}/include}.
2203 @itemx --disable-cet
2204 Enable building target run-time libraries with control-flow
2205 instrumentation, see @option{-fcf-protection} option. When
2206 @code{--enable-cet} is specified target libraries are configured
2207 to add @option{-fcf-protection} and, if needed, other target
2208 specific options to a set of building options.
2210 The option is disabled by default. When @code{--enable-cet=auto}
2211 is used, it is enabled on Linux/x86 if target binutils
2212 supports @code{Intel CET} instructions and disabled otherwise.
2213 In this case the target libraries are configured to get additional
2214 @option{-fcf-protection} option.
2216 @item --with-riscv-attribute=@samp{yes}, @samp{no} or @samp{default}
2217 Generate RISC-V attribute by default, in order to record extra build
2218 information in object.
2220 The option is disabled by default. It is enabled on RISC-V/ELF (bare-metal)
2221 target if target binutils supported.
2224 @subheading Cross-Compiler-Specific Options
2225 The following options only apply to building cross compilers.
2228 @item --with-toolexeclibdir=@var{dir}
2229 Specify the installation directory for libraries built with a cross compiler.
2230 The default is @option{$@{gcc_tooldir@}/lib}.
2232 @item --with-sysroot
2233 @itemx --with-sysroot=@var{dir}
2234 Tells GCC to consider @var{dir} as the root of a tree that contains
2235 (a subset of) the root filesystem of the target operating system.
2236 Target system headers, libraries and run-time object files will be
2237 searched for in there. More specifically, this acts as if
2238 @option{--sysroot=@var{dir}} was added to the default options of the built
2239 compiler. The specified directory is not copied into the
2240 install tree, unlike the options @option{--with-headers} and
2241 @option{--with-libs} that this option obsoletes. The default value,
2242 in case @option{--with-sysroot} is not given an argument, is
2243 @option{$@{gcc_tooldir@}/sys-root}. If the specified directory is a
2244 subdirectory of @option{$@{exec_prefix@}}, then it will be found relative to
2245 the GCC binaries if the installation tree is moved.
2247 This option affects the system root for the compiler used to build
2248 target libraries (which runs on the build system) and the compiler newly
2249 installed with @code{make install}; it does not affect the compiler which is
2250 used to build GCC itself.
2252 If you specify the @option{--with-native-system-header-dir=@var{dirname}}
2253 option then the compiler will search that directory within @var{dirname} for
2254 native system headers rather than the default @file{/usr/include}.
2256 @item --with-build-sysroot
2257 @itemx --with-build-sysroot=@var{dir}
2258 Tells GCC to consider @var{dir} as the system root (see
2259 @option{--with-sysroot}) while building target libraries, instead of
2260 the directory specified with @option{--with-sysroot}. This option is
2261 only useful when you are already using @option{--with-sysroot}. You
2262 can use @option{--with-build-sysroot} when you are configuring with
2263 @option{--prefix} set to a directory that is different from the one in
2264 which you are installing GCC and your target libraries.
2266 This option affects the system root for the compiler used to build
2267 target libraries (which runs on the build system); it does not affect
2268 the compiler which is used to build GCC itself.
2270 If you specify the @option{--with-native-system-header-dir=@var{dirname}}
2271 option then the compiler will search that directory within @var{dirname} for
2272 native system headers rather than the default @file{/usr/include}.
2274 @item --with-headers
2275 @itemx --with-headers=@var{dir}
2276 Deprecated in favor of @option{--with-sysroot}.
2277 Specifies that target headers are available when building a cross compiler.
2278 The @var{dir} argument specifies a directory which has the target include
2279 files. These include files will be copied into the @file{gcc} install
2280 directory. @emph{This option with the @var{dir} argument is required} when
2281 building a cross compiler, if @file{@var{prefix}/@var{target}/sys-include}
2282 doesn't pre-exist. If @file{@var{prefix}/@var{target}/sys-include} does
2283 pre-exist, the @var{dir} argument may be omitted. @command{fixincludes}
2284 will be run on these files to make them compatible with GCC@.
2286 @item --without-headers
2287 Tells GCC not use any target headers from a libc when building a cross
2288 compiler. When crossing to GNU/Linux, you need the headers so GCC
2289 can build the exception handling for libgcc.
2292 @itemx --with-libs="@var{dir1} @var{dir2} @dots{} @var{dirN}"
2293 Deprecated in favor of @option{--with-sysroot}.
2294 Specifies a list of directories which contain the target runtime
2295 libraries. These libraries will be copied into the @file{gcc} install
2296 directory. If the directory list is omitted, this option has no
2300 Specifies that @samp{newlib} is
2301 being used as the target C library. This causes @code{__eprintf} to be
2302 omitted from @file{libgcc.a} on the assumption that it will be provided by
2308 @item --with-avrlibc
2309 Only supported for the AVR target. Specifies that @samp{AVR-Libc} is
2310 being used as the target C@tie{} library. This causes float support
2311 functions like @code{__addsf3} to be omitted from @file{libgcc.a} on
2312 the assumption that it will be provided by @file{libm.a}. For more
2313 technical details, cf. @uref{http://gcc.gnu.org/PR54461,,PR54461}.
2314 It is not supported for
2315 RTEMS configurations, which currently use newlib. The option is
2316 supported since version 4.7.2 and is the default in 4.8.0 and newer.
2318 @item --with-double=@{32|64|32,64|64,32@}
2319 @itemx --with-long-double=@{32|64|32,64|64,32|double@}
2320 Only supported for the AVR target since version@tie{}10.
2321 Specify the default layout available for the C/C++ @samp{double}
2322 and @samp{long double} type, respectively. The following rules apply:
2325 The first value after the @samp{=} specifies the default layout (in bits)
2326 of the type and also the default for the @option{-mdouble=} resp.
2327 @option{-mlong-double=} compiler option.
2329 If more than one value is specified, respective multilib variants are
2330 available, and @option{-mdouble=} resp. @option{-mlong-double=} acts
2331 as a multilib option.
2333 If @option{--with-long-double=double} is specified, @samp{double} and
2334 @samp{long double} will have the same layout.
2336 The defaults are @option{--with-long-double=64,32} and
2337 @option{--with-double=32,64}. The default @samp{double} layout imposed by
2338 the latter is compatible with older versions of the compiler that implement
2339 @samp{double} as a 32-bit type, which does not comply to the language standard.
2341 Not all combinations of @option{--with-double=} and
2342 @option{--with-long-double=} are valid. For example, the combination
2343 @option{--with-double=32,64} @option{--with-long-double=32} will be
2344 rejected because the first option specifies the availability of
2345 multilibs for @samp{double}, whereas the second option implies
2346 that @samp{long double} --- and hence also @samp{double} --- is always
2349 @item --with-double-comparison=@{tristate|bool|libf7@}
2350 Only supported for the AVR target since version@tie{}10.
2351 Specify what result format is returned by library functions that
2352 compare 64-bit floating point values (@code{DFmode}).
2353 The GCC default is @samp{tristate}. If the floating point
2354 implementation returns a boolean instead, set it to @samp{bool}.
2356 @item --with-libf7=@{libgcc|math|math-symbols|no@}
2357 Only supported for the AVR target since version@tie{}10.
2358 Specify to which degree code from LibF7 is included in libgcc.
2359 LibF7 is an ad-hoc, AVR-specific, 64-bit floating point emulation
2360 written in C and (inline) assembly. @samp{libgcc} adds support
2361 for functions that one would usually expect in libgcc like double addition,
2362 double comparisons and double conversions. @samp{math} also adds routines
2363 that one would expect in @file{libm.a}, but with @code{__} (two underscores)
2364 prepended to the symbol names as specified by @file{math.h}.
2365 @samp{math-symbols} also defines weak aliases for the functions
2366 declared in @file{math.h}. However, @code{--with-libf7} won't
2367 install no @file{math.h} header file whatsoever, this file must come
2368 from elsewhere. This option sets @option{--with-double-comparison}
2371 @item --with-nds32-lib=@var{library}
2372 Specifies that @var{library} setting is used for building @file{libgcc.a}.
2373 Currently, the valid @var{library} is @samp{newlib} or @samp{mculib}.
2374 This option is only supported for the NDS32 target.
2376 @item --with-build-time-tools=@var{dir}
2377 Specifies where to find the set of target tools (assembler, linker, etc.)
2378 that will be used while building GCC itself. This option can be useful
2379 if the directory layouts are different between the system you are building
2380 GCC on, and the system where you will deploy it.
2382 For example, on an @samp{ia64-hp-hpux} system, you may have the GNU
2383 assembler and linker in @file{/usr/bin}, and the native tools in a
2384 different path, and build a toolchain that expects to find the
2385 native tools in @file{/usr/bin}.
2387 When you use this option, you should ensure that @var{dir} includes
2388 @command{ar}, @command{as}, @command{ld}, @command{nm},
2389 @command{ranlib} and @command{strip} if necessary, and possibly
2390 @command{objdump}. Otherwise, GCC may use an inconsistent set of
2394 @subsubheading Overriding @command{configure} test results
2396 Sometimes, it might be necessary to override the result of some
2397 @command{configure} test, for example in order to ease porting to a new
2398 system or work around a bug in a test. The toplevel @command{configure}
2399 script provides three variables for this:
2403 @item build_configargs
2404 @cindex @code{build_configargs}
2405 The contents of this variable is passed to all build @command{configure}
2408 @item host_configargs
2409 @cindex @code{host_configargs}
2410 The contents of this variable is passed to all host @command{configure}
2413 @item target_configargs
2414 @cindex @code{target_configargs}
2415 The contents of this variable is passed to all target @command{configure}
2420 In order to avoid shell and @command{make} quoting issues for complex
2421 overrides, you can pass a setting for @env{CONFIG_SITE} and set
2422 variables in the site file.
2424 @subheading Objective-C-Specific Options
2426 The following options apply to the build of the Objective-C runtime library.
2429 @item --enable-objc-gc
2430 Specify that an additional variant of the GNU Objective-C runtime library
2431 is built, using an external build of the Boehm-Demers-Weiser garbage
2432 collector (@uref{https://www.hboehm.info/gc/}). This library needs to be
2433 available for each multilib variant, unless configured with
2434 @option{--enable-objc-gc=@samp{auto}} in which case the build of the
2435 additional runtime library is skipped when not available and the build
2438 @item --with-target-bdw-gc=@var{list}
2439 @itemx --with-target-bdw-gc-include=@var{list}
2440 @itemx --with-target-bdw-gc-lib=@var{list}
2441 Specify search directories for the garbage collector header files and
2442 libraries. @var{list} is a comma separated list of key value pairs of the
2443 form @samp{@var{multilibdir}=@var{path}}, where the default multilib key
2444 is named as @samp{.} (dot), or is omitted (e.g.@:
2445 @samp{--with-target-bdw-gc=/opt/bdw-gc,32=/opt-bdw-gc32}).
2447 The options @option{--with-target-bdw-gc-include} and
2448 @option{--with-target-bdw-gc-lib} must always be specified together
2449 for each multilib variant and they take precedence over
2450 @option{--with-target-bdw-gc}. If @option{--with-target-bdw-gc-include}
2451 is missing values for a multilib, then the value for the default
2452 multilib is used (e.g.@: @samp{--with-target-bdw-gc-include=/opt/bdw-gc/include}
2453 @samp{--with-target-bdw-gc-lib=/opt/bdw-gc/lib64,32=/opt-bdw-gc/lib32}).
2454 If none of these options are specified, the library is assumed in
2458 @subheading D-Specific Options
2460 The following options apply to the build of the D runtime library.
2463 @item --with-target-system-zlib
2464 Use installed @samp{zlib} rather than that included with GCC@. This needs
2465 to be available for each multilib variant, unless configured with
2466 @option{--with-target-system-zlib=@samp{auto}} in which case the GCC@ included
2467 @samp{zlib} is only used when the system installed library is not available.
2475 @uref{./index.html,,Return to the GCC Installation page}
2479 @c ***Building****************************************************************
2481 @comment node-name, next, previous, up
2482 @node Building, Testing, Configuration, Installing GCC
2488 @cindex Installing GCC: Building
2490 Now that GCC is configured, you are ready to build the compiler and
2493 Some commands executed when making the compiler may fail (return a
2494 nonzero status) and be ignored by @command{make}. These failures, which
2495 are often due to files that were not found, are expected, and can safely
2498 It is normal to have compiler warnings when compiling certain files.
2499 Unless you are a GCC developer, you can generally ignore these warnings
2500 unless they cause compilation to fail. Developers should attempt to fix
2501 any warnings encountered, however they can temporarily continue past
2502 warnings-as-errors by specifying the configure flag
2503 @option{--disable-werror}.
2505 On certain old systems, defining certain environment variables such as
2506 @env{CC} can interfere with the functioning of @command{make}.
2508 If you encounter seemingly strange errors when trying to build the
2509 compiler in a directory other than the source directory, it could be
2510 because you have previously configured the compiler in the source
2511 directory. Make sure you have done all the necessary preparations.
2513 If you build GCC on a BSD system using a directory stored in an old System
2514 V file system, problems may occur in running @command{fixincludes} if the
2515 System V file system doesn't support symbolic links. These problems
2516 result in a failure to fix the declaration of @code{size_t} in
2517 @file{sys/types.h}. If you find that @code{size_t} is a signed type and
2518 that type mismatches occur, this could be the cause.
2520 The solution is not to use such a directory for building GCC@.
2522 Similarly, when building from the source repository or snapshots, or if you modify
2523 @file{*.l} files, you need the Flex lexical analyzer generator
2524 installed. If you do not modify @file{*.l} files, releases contain
2525 the Flex-generated files and you do not need Flex installed to build
2526 them. There is still one Flex-based lexical analyzer (part of the
2527 build machinery, not of GCC itself) that is used even if you only
2528 build the C front end.
2530 When building from the source repository or snapshots, or if you modify Texinfo
2531 documentation, you need version 4.7 or later of Texinfo installed if you
2532 want Info documentation to be regenerated. Releases contain Info
2533 documentation pre-built for the unmodified documentation in the release.
2535 @section Building a native compiler
2537 For a native build, the default configuration is to perform
2538 a 3-stage bootstrap of the compiler when @samp{make} is invoked.
2539 This will build the entire GCC system and ensure that it compiles
2540 itself correctly. It can be disabled with the @option{--disable-bootstrap}
2541 parameter to @samp{configure}, but bootstrapping is suggested because
2542 the compiler will be tested more completely and could also have
2545 The bootstrapping process will complete the following steps:
2549 Build tools necessary to build the compiler.
2552 Perform a 3-stage bootstrap of the compiler. This includes building
2553 three times the target tools for use by the compiler such as binutils
2554 (bfd, binutils, gas, gprof, ld, and opcodes) if they have been
2555 individually linked or moved into the top level GCC source tree before
2559 Perform a comparison test of the stage2 and stage3 compilers.
2562 Build runtime libraries using the stage3 compiler from the previous step.
2566 If you are short on disk space you might consider @samp{make
2567 bootstrap-lean} instead. The sequence of compilation is the
2568 same described above, but object files from the stage1 and
2569 stage2 of the 3-stage bootstrap of the compiler are deleted as
2570 soon as they are no longer needed.
2572 If you wish to use non-default GCC flags when compiling the stage2
2573 and stage3 compilers, set @code{BOOT_CFLAGS} on the command line when
2574 doing @samp{make}. For example, if you want to save additional space
2575 during the bootstrap and in the final installation as well, you can
2576 build the compiler binaries without debugging information as in the
2577 following example. This will save roughly 40% of disk space both for
2578 the bootstrap and the final installation. (Libraries will still contain
2579 debugging information.)
2582 make BOOT_CFLAGS='-O' bootstrap
2585 You can place non-default optimization flags into @code{BOOT_CFLAGS}; they
2586 are less well tested here than the default of @samp{-g -O2}, but should
2587 still work. In a few cases, you may find that you need to specify special
2588 flags such as @option{-msoft-float} here to complete the bootstrap; or,
2589 if the native compiler miscompiles the stage1 compiler, you may need
2590 to work around this, by choosing @code{BOOT_CFLAGS} to avoid the parts
2591 of the stage1 compiler that were miscompiled, or by using @samp{make
2592 bootstrap4} to increase the number of stages of bootstrap.
2594 @code{BOOT_CFLAGS} does not apply to bootstrapped target libraries.
2595 Since these are always compiled with the compiler currently being
2596 bootstrapped, you can use @code{CFLAGS_FOR_TARGET} to modify their
2597 compilation flags, as for non-bootstrapped target libraries.
2598 Again, if the native compiler miscompiles the stage1 compiler, you may
2599 need to work around this by avoiding non-working parts of the stage1
2600 compiler. Use @code{STAGE1_TFLAGS} to this end.
2602 If you used the flag @option{--enable-languages=@dots{}} to restrict
2603 the compilers to be built, only those you've actually enabled will be
2604 built. This will of course only build those runtime libraries, for
2605 which the particular compiler has been built. Please note,
2606 that re-defining @env{LANGUAGES} when calling @samp{make}
2607 @strong{does not} work anymore!
2609 If the comparison of stage2 and stage3 fails, this normally indicates
2610 that the stage2 compiler has compiled GCC incorrectly, and is therefore
2611 a potentially serious bug which you should investigate and report. (On
2612 a few systems, meaningful comparison of object files is impossible; they
2613 always appear ``different''. If you encounter this problem, you will
2614 need to disable comparison in the @file{Makefile}.)
2616 If you do not want to bootstrap your compiler, you can configure with
2617 @option{--disable-bootstrap}. In particular cases, you may want to
2618 bootstrap your compiler even if the target system is not the same as
2619 the one you are building on: for example, you could build a
2620 @code{powerpc-unknown-linux-gnu} toolchain on a
2621 @code{powerpc64-unknown-linux-gnu} host. In this case, pass
2622 @option{--enable-bootstrap} to the configure script.
2624 @code{BUILD_CONFIG} can be used to bring in additional customization
2625 to the build. It can be set to a whitespace-separated list of names.
2626 For each such @code{NAME}, top-level @file{config/@code{NAME}.mk} will
2627 be included by the top-level @file{Makefile}, bringing in any settings
2628 it contains. The default @code{BUILD_CONFIG} can be set using the
2629 configure option @option{--with-build-config=@code{NAME}...}. Some
2630 examples of supported build configurations are:
2633 @item @samp{bootstrap-O1}
2634 Removes any @option{-O}-started option from @code{BOOT_CFLAGS}, and adds
2635 @option{-O1} to it. @samp{BUILD_CONFIG=bootstrap-O1} is equivalent to
2636 @samp{BOOT_CFLAGS='-g -O1'}.
2638 @item @samp{bootstrap-O3}
2639 @itemx @samp{bootstrap-Og}
2640 Analogous to @code{bootstrap-O1}.
2642 @item @samp{bootstrap-lto}
2643 Enables Link-Time Optimization for host tools during bootstrapping.
2644 @samp{BUILD_CONFIG=bootstrap-lto} is equivalent to adding
2645 @option{-flto} to @samp{BOOT_CFLAGS}. This option assumes that the host
2646 supports the linker plugin (e.g.@: GNU ld version 2.21 or later or GNU gold
2647 version 2.21 or later).
2649 @item @samp{bootstrap-lto-noplugin}
2650 This option is similar to @code{bootstrap-lto}, but is intended for
2651 hosts that do not support the linker plugin. Without the linker plugin
2652 static libraries are not compiled with link-time optimizations. Since
2653 the GCC middle end and back end are in @file{libbackend.a} this means
2654 that only the front end is actually LTO optimized.
2656 @item @samp{bootstrap-lto-lean}
2657 This option is similar to @code{bootstrap-lto}, but is intended for
2658 faster build by only using LTO in the final bootstrap stage.
2659 With @samp{make profiledbootstrap} the LTO frontend
2660 is trained only on generator files.
2662 @item @samp{bootstrap-debug}
2663 Verifies that the compiler generates the same executable code, whether
2664 or not it is asked to emit debug information. To this end, this
2665 option builds stage2 host programs without debug information, and uses
2666 @file{contrib/compare-debug} to compare them with the stripped stage3
2667 object files. If @code{BOOT_CFLAGS} is overridden so as to not enable
2668 debug information, stage2 will have it, and stage3 won't. This option
2669 is enabled by default when GCC bootstrapping is enabled, if
2670 @code{strip} can turn object files compiled with and without debug
2671 info into identical object files. In addition to better test
2672 coverage, this option makes default bootstraps faster and leaner.
2674 @item @samp{bootstrap-debug-big}
2675 Rather than comparing stripped object files, as in
2676 @code{bootstrap-debug}, this option saves internal compiler dumps
2677 during stage2 and stage3 and compares them as well, which helps catch
2678 additional potential problems, but at a great cost in terms of disk
2679 space. It can be specified in addition to @samp{bootstrap-debug}.
2681 @item @samp{bootstrap-debug-lean}
2682 This option saves disk space compared with @code{bootstrap-debug-big},
2683 but at the expense of some recompilation. Instead of saving the dumps
2684 of stage2 and stage3 until the final compare, it uses
2685 @option{-fcompare-debug} to generate, compare and remove the dumps
2686 during stage3, repeating the compilation that already took place in
2687 stage2, whose dumps were not saved.
2689 @item @samp{bootstrap-debug-lib}
2690 This option tests executable code invariance over debug information
2691 generation on target libraries, just like @code{bootstrap-debug-lean}
2692 tests it on host programs. It builds stage3 libraries with
2693 @option{-fcompare-debug}, and it can be used along with any of the
2694 @code{bootstrap-debug} options above.
2696 There aren't @code{-lean} or @code{-big} counterparts to this option
2697 because most libraries are only build in stage3, so bootstrap compares
2698 would not get significant coverage. Moreover, the few libraries built
2699 in stage2 are used in stage3 host programs, so we wouldn't want to
2700 compile stage2 libraries with different options for comparison purposes.
2702 @item @samp{bootstrap-debug-ckovw}
2703 Arranges for error messages to be issued if the compiler built on any
2704 stage is run without the option @option{-fcompare-debug}. This is
2705 useful to verify the full @option{-fcompare-debug} testing coverage. It
2706 must be used along with @code{bootstrap-debug-lean} and
2707 @code{bootstrap-debug-lib}.
2709 @item @samp{bootstrap-cet}
2710 This option enables Intel CET for host tools during bootstrapping.
2711 @samp{BUILD_CONFIG=bootstrap-cet} is equivalent to adding
2712 @option{-fcf-protection} to @samp{BOOT_CFLAGS}. This option
2713 assumes that the host supports Intel CET (e.g.@: GNU assembler version
2716 @item @samp{bootstrap-time}
2717 Arranges for the run time of each program started by the GCC driver,
2718 built in any stage, to be logged to @file{time.log}, in the top level of
2723 @section Building a cross compiler
2725 When building a cross compiler, it is not generally possible to do a
2726 3-stage bootstrap of the compiler. This makes for an interesting problem
2727 as parts of GCC can only be built with GCC@.
2729 To build a cross compiler, we recommend first building and installing a
2730 native compiler. You can then use the native GCC compiler to build the
2731 cross compiler. The installed native compiler needs to be GCC version
2734 Assuming you have already installed a native copy of GCC and configured
2735 your cross compiler, issue the command @command{make}, which performs the
2740 Build host tools necessary to build the compiler.
2743 Build target tools for use by the compiler such as binutils (bfd,
2744 binutils, gas, gprof, ld, and opcodes)
2745 if they have been individually linked or moved into the top level GCC source
2746 tree before configuring.
2749 Build the compiler (single stage only).
2752 Build runtime libraries using the compiler from the previous step.
2755 Note that if an error occurs in any step the make process will exit.
2757 If you are not building GNU binutils in the same source tree as GCC,
2758 you will need a cross-assembler and cross-linker installed before
2759 configuring GCC@. Put them in the directory
2760 @file{@var{prefix}/@var{target}/bin}. Here is a table of the tools
2761 you should put in this directory:
2765 This should be the cross-assembler.
2768 This should be the cross-linker.
2771 This should be the cross-archiver: a program which can manipulate
2772 archive files (linker libraries) in the target machine's format.
2775 This should be a program to construct a symbol table in an archive file.
2778 The installation of GCC will find these programs in that directory,
2779 and copy or link them to the proper place to for the cross-compiler to
2780 find them when run later.
2782 The easiest way to provide these files is to build the Binutils package.
2783 Configure it with the same @option{--host} and @option{--target}
2784 options that you use for configuring GCC, then build and install
2785 them. They install their executables automatically into the proper
2786 directory. Alas, they do not support all the targets that GCC
2789 If you are not building a C library in the same source tree as GCC,
2790 you should also provide the target libraries and headers before
2791 configuring GCC, specifying the directories with
2792 @option{--with-sysroot} or @option{--with-headers} and
2793 @option{--with-libs}. Many targets also require ``start files'' such
2794 as @file{crt0.o} and
2795 @file{crtn.o} which are linked into each executable. There may be several
2796 alternatives for @file{crt0.o}, for use with profiling or other
2797 compilation options. Check your target's definition of
2798 @code{STARTFILE_SPEC} to find out what start files it uses.
2800 @section Building in parallel
2802 GNU Make 3.80 and above, which is necessary to build GCC, support
2803 building in parallel. To activate this, you can use @samp{make -j 2}
2804 instead of @samp{make}. You can also specify a bigger number, and
2805 in most cases using a value greater than the number of processors in
2806 your machine will result in fewer and shorter I/O latency hits, thus
2807 improving overall throughput; this is especially true for slow drives
2808 and network filesystems.
2810 @section Building the Ada compiler
2813 @ref{GNAT-prerequisite}.
2816 @uref{prerequisites.html#GNAT-prerequisite,,GNAT prerequisites}.
2819 @section Building with profile feedback
2821 It is possible to use profile feedback to optimize the compiler itself. This
2822 should result in a faster compiler binary. Experiments done on x86 using gcc
2823 3.3 showed approximately 7 percent speedup on compiling C programs. To
2824 bootstrap the compiler with profile feedback, use @code{make profiledbootstrap}.
2826 When @samp{make profiledbootstrap} is run, it will first build a @code{stage1}
2827 compiler. This compiler is used to build a @code{stageprofile} compiler
2828 instrumented to collect execution counts of instruction and branch
2829 probabilities. Training run is done by building @code{stagetrain}
2830 compiler. Finally a @code{stagefeedback} compiler is built
2831 using the information collected.
2833 Unlike standard bootstrap, several additional restrictions apply. The
2834 compiler used to build @code{stage1} needs to support a 64-bit integral type.
2835 It is recommended to only use GCC for this.
2837 On Linux/x86_64 hosts with some restrictions (no virtualization) it is
2838 also possible to do autofdo build with @samp{make
2839 autoprofiledback}. This uses Linux perf to sample branches in the
2840 binary and then rebuild it with feedback derived from the profile.
2841 Linux perf and the @code{autofdo} toolkit needs to be installed for
2844 Only the profile from the current build is used, so when an error
2845 occurs it is recommended to clean before restarting. Otherwise
2846 the code quality may be much worse.
2853 @uref{./index.html,,Return to the GCC Installation page}
2857 @c ***Testing*****************************************************************
2859 @comment node-name, next, previous, up
2860 @node Testing, Final install, Building, Installing GCC
2864 @chapter Installing GCC: Testing
2867 @cindex Installing GCC: Testing
2870 Before you install GCC, we encourage you to run the testsuites and to
2871 compare your results with results from a similar configuration that have
2872 been submitted to the
2873 @uref{http://gcc.gnu.org/ml/gcc-testresults/,,gcc-testresults mailing list}.
2874 Some of these archived results are linked from the build status lists
2875 at @uref{http://gcc.gnu.org/buildstat.html}, although not everyone who
2876 reports a successful build runs the testsuites and submits the results.
2877 This step is optional and may require you to download additional software,
2878 but it can give you confidence in your new GCC installation or point out
2879 problems before you install and start using your new GCC@.
2881 First, you must have @uref{download.html,,downloaded the testsuites}.
2882 These are part of the full distribution, but if you downloaded the
2883 ``core'' compiler plus any front ends, you must download the testsuites
2886 Second, you must have the testing tools installed. This includes
2887 @uref{http://www.gnu.org/software/dejagnu/,,DejaGnu}, Tcl, and Expect;
2888 the DejaGnu site has links to these. For running the BRIG frontend
2889 tests, a tool to assemble the binary BRIGs from HSAIL text,
2890 @uref{https://github.com/HSAFoundation/HSAIL-Tools/,,HSAILasm} must
2893 If the directories where @command{runtest} and @command{expect} were
2894 installed are not in the @env{PATH}, you may need to set the following
2895 environment variables appropriately, as in the following example (which
2896 assumes that DejaGnu has been installed under @file{/usr/local}):
2899 TCL_LIBRARY = /usr/local/share/tcl8.0
2900 DEJAGNULIBS = /usr/local/share/dejagnu
2903 (On systems such as Cygwin, these paths are required to be actual
2904 paths, not mounts or links; presumably this is due to some lack of
2905 portability in the DejaGnu code.)
2908 Finally, you can run the testsuite (which may take a long time):
2910 cd @var{objdir}; make -k check
2913 This will test various components of GCC, such as compiler
2914 front ends and runtime libraries. While running the testsuite, DejaGnu
2915 might emit some harmless messages resembling
2916 @samp{WARNING: Couldn't find the global config file.} or
2917 @samp{WARNING: Couldn't find tool init file} that can be ignored.
2919 If you are testing a cross-compiler, you may want to run the testsuite
2920 on a simulator as described at @uref{http://gcc.gnu.org/simtest-howto.html}.
2922 @section How can you run the testsuite on selected tests?
2924 In order to run sets of tests selectively, there are targets
2925 @samp{make check-gcc} and language specific @samp{make check-c},
2926 @samp{make check-c++}, @samp{make check-d} @samp{make check-fortran},
2927 @samp{make check-ada}, @samp{make check-objc}, @samp{make check-obj-c++},
2928 @samp{make check-lto}
2929 in the @file{gcc} subdirectory of the object directory. You can also
2930 just run @samp{make check} in a subdirectory of the object directory.
2933 A more selective way to just run all @command{gcc} execute tests in the
2937 make check-gcc RUNTESTFLAGS="execute.exp @var{other-options}"
2940 Likewise, in order to run only the @command{g++} ``old-deja'' tests in
2941 the testsuite with filenames matching @samp{9805*}, you would use
2944 make check-g++ RUNTESTFLAGS="old-deja.exp=9805* @var{other-options}"
2947 The file-matching expression following @var{filename}@command{.exp=} is treated
2948 as a series of whitespace-delimited glob expressions so that multiple patterns
2949 may be passed, although any whitespace must either be escaped or surrounded by
2950 single quotes if multiple expressions are desired. For example,
2953 make check-g++ RUNTESTFLAGS="old-deja.exp=9805*\ virtual2.c @var{other-options}"
2954 make check-g++ RUNTESTFLAGS="'old-deja.exp=9805* virtual2.c' @var{other-options}"
2957 The @file{*.exp} files are located in the testsuite directories of the GCC
2958 source, the most important ones being @file{compile.exp},
2959 @file{execute.exp}, @file{dg.exp} and @file{old-deja.exp}.
2960 To get a list of the possible @file{*.exp} files, pipe the
2961 output of @samp{make check} into a file and look at the
2962 @samp{Running @dots{} .exp} lines.
2964 @section Passing options and running multiple testsuites
2966 You can pass multiple options to the testsuite using the
2967 @samp{--target_board} option of DejaGNU, either passed as part of
2968 @samp{RUNTESTFLAGS}, or directly to @command{runtest} if you prefer to
2969 work outside the makefiles. For example,
2972 make check-g++ RUNTESTFLAGS="--target_board=unix/-O3/-fmerge-constants"
2975 will run the standard @command{g++} testsuites (``unix'' is the target name
2976 for a standard native testsuite situation), passing
2977 @samp{-O3 -fmerge-constants} to the compiler on every test, i.e.,
2978 slashes separate options.
2980 You can run the testsuites multiple times using combinations of options
2981 with a syntax similar to the brace expansion of popular shells:
2984 @dots{}"--target_board=arm-sim\@{-mhard-float,-msoft-float\@}\@{-O1,-O2,-O3,\@}"
2987 (Note the empty option caused by the trailing comma in the final group.)
2988 The following will run each testsuite eight times using the @samp{arm-sim}
2989 target, as if you had specified all possible combinations yourself:
2992 --target_board='arm-sim/-mhard-float/-O1 \
2993 arm-sim/-mhard-float/-O2 \
2994 arm-sim/-mhard-float/-O3 \
2995 arm-sim/-mhard-float \
2996 arm-sim/-msoft-float/-O1 \
2997 arm-sim/-msoft-float/-O2 \
2998 arm-sim/-msoft-float/-O3 \
2999 arm-sim/-msoft-float'
3002 They can be combined as many times as you wish, in arbitrary ways. This
3006 @dots{}"--target_board=unix/-Wextra\@{-O3,-fno-strength\@}\@{-fomit-frame,\@}"
3009 will generate four combinations, all involving @samp{-Wextra}.
3011 The disadvantage to this method is that the testsuites are run in serial,
3012 which is a waste on multiprocessor systems. For users with GNU Make and
3013 a shell which performs brace expansion, you can run the testsuites in
3014 parallel by having the shell perform the combinations and @command{make}
3015 do the parallel runs. Instead of using @samp{--target_board}, use a
3016 special makefile target:
3019 make -j@var{N} check-@var{testsuite}//@var{test-target}/@var{option1}/@var{option2}/@dots{}
3025 make -j3 check-gcc//sh-hms-sim/@{-m1,-m2,-m3,-m3e,-m4@}/@{,-nofpu@}
3028 will run three concurrent ``make-gcc'' testsuites, eventually testing all
3029 ten combinations as described above. Note that this is currently only
3030 supported in the @file{gcc} subdirectory. (To see how this works, try
3031 typing @command{echo} before the example given here.)
3034 @section How to interpret test results
3036 The result of running the testsuite are various @file{*.sum} and @file{*.log}
3037 files in the testsuite subdirectories. The @file{*.log} files contain a
3038 detailed log of the compiler invocations and the corresponding
3039 results, the @file{*.sum} files summarize the results. These summaries
3040 contain status codes for all tests:
3044 PASS: the test passed as expected
3046 XPASS: the test unexpectedly passed
3048 FAIL: the test unexpectedly failed
3050 XFAIL: the test failed as expected
3052 UNSUPPORTED: the test is not supported on this platform
3054 ERROR: the testsuite detected an error
3056 WARNING: the testsuite detected a possible problem
3059 It is normal for some tests to report unexpected failures. At the
3060 current time the testing harness does not allow fine grained control
3061 over whether or not a test is expected to fail. This problem should
3062 be fixed in future releases.
3065 @section Submitting test results
3067 If you want to report the results to the GCC project, use the
3068 @file{contrib/test_summary} shell script. Start it in the @var{objdir} with
3071 @var{srcdir}/contrib/test_summary -p your_commentary.txt \
3072 -m gcc-testresults@@gcc.gnu.org |sh
3075 This script uses the @command{Mail} program to send the results, so
3076 make sure it is in your @env{PATH}. The file @file{your_commentary.txt} is
3077 prepended to the testsuite summary and should contain any special
3078 remarks you have on your results or your build environment. Please
3079 do not edit the testsuite result block or the subject line, as these
3080 messages may be automatically processed.
3087 @uref{./index.html,,Return to the GCC Installation page}
3091 @c ***Final install***********************************************************
3093 @comment node-name, next, previous, up
3094 @node Final install, , Testing, Installing GCC
3096 @ifset finalinstallhtml
3098 @chapter Installing GCC: Final installation
3101 Now that GCC has been built (and optionally tested), you can install it with
3103 cd @var{objdir} && make install
3106 We strongly recommend to install into a target directory where there is
3107 no previous version of GCC present. Also, the GNAT runtime should not
3108 be stripped, as this would break certain features of the debugger that
3109 depend on this debugging information (catching Ada exceptions for
3112 That step completes the installation of GCC; user level binaries can
3113 be found in @file{@var{prefix}/bin} where @var{prefix} is the value
3114 you specified with the @option{--prefix} to configure (or
3115 @file{/usr/local} by default). (If you specified @option{--bindir},
3116 that directory will be used instead; otherwise, if you specified
3117 @option{--exec-prefix}, @file{@var{exec-prefix}/bin} will be used.)
3118 Headers for the C++ library are installed in
3119 @file{@var{prefix}/include}; libraries in @file{@var{libdir}}
3120 (normally @file{@var{prefix}/lib}); internal parts of the compiler in
3121 @file{@var{libdir}/gcc} and @file{@var{libexecdir}/gcc}; documentation
3122 in info format in @file{@var{infodir}} (normally
3123 @file{@var{prefix}/info}).
3125 When installing cross-compilers, GCC's executables
3126 are not only installed into @file{@var{bindir}}, that
3127 is, @file{@var{exec-prefix}/bin}, but additionally into
3128 @file{@var{exec-prefix}/@var{target-alias}/bin}, if that directory
3129 exists. Typically, such @dfn{tooldirs} hold target-specific
3130 binutils, including assembler and linker.
3132 Installation into a temporary staging area or into a @command{chroot}
3133 jail can be achieved with the command
3136 make DESTDIR=@var{path-to-rootdir} install
3140 where @var{path-to-rootdir} is the absolute path of
3141 a directory relative to which all installation paths will be
3142 interpreted. Note that the directory specified by @code{DESTDIR}
3143 need not exist yet; it will be created if necessary.
3145 There is a subtle point with tooldirs and @code{DESTDIR}:
3146 If you relocate a cross-compiler installation with
3147 e.g.@: @samp{DESTDIR=@var{rootdir}}, then the directory
3148 @file{@var{rootdir}/@var{exec-prefix}/@var{target-alias}/bin} will
3149 be filled with duplicated GCC executables only if it already exists,
3150 it will not be created otherwise. This is regarded as a feature,
3151 not as a bug, because it gives slightly more control to the packagers
3152 using the @code{DESTDIR} feature.
3154 You can install stripped programs and libraries with
3160 If you are bootstrapping a released version of GCC then please
3161 quickly review the build status page for your release, available from
3162 @uref{http://gcc.gnu.org/buildstat.html}.
3163 If your system is not listed for the version of GCC that you built,
3165 @email{gcc@@gcc.gnu.org} indicating
3166 that you successfully built and installed GCC@.
3167 Include the following information:
3171 Output from running @file{@var{srcdir}/config.guess}. Do not send
3172 that file itself, just the one-line output from running it.
3175 The output of @samp{gcc -v} for your newly installed @command{gcc}.
3176 This tells us which version of GCC you built and the options you passed to
3180 Whether you enabled all languages or a subset of them. If you used a
3181 full distribution then this information is part of the configure
3182 options in the output of @samp{gcc -v}, but if you downloaded the
3183 ``core'' compiler plus additional front ends then it isn't apparent
3184 which ones you built unless you tell us about it.
3187 If the build was for GNU/Linux, also include:
3190 The distribution name and version (e.g., Red Hat 7.1 or Debian 2.2.3);
3191 this information should be available from @file{/etc/issue}.
3194 The version of the Linux kernel, available from @samp{uname --version}
3198 The version of glibc you used; for RPM-based systems like Red Hat,
3199 Mandrake, and SuSE type @samp{rpm -q glibc} to get the glibc version,
3200 and on systems like Debian and Progeny use @samp{dpkg -l libc6}.
3202 For other systems, you can include similar information if you think it is
3206 Any other information that you think would be useful to people building
3207 GCC on the same configuration. The new entry in the build status list
3208 will include a link to the archived copy of your message.
3211 We'd also like to know if the
3213 @ref{Specific, host/target specific installation notes}
3216 @uref{specific.html,,host/target specific installation notes}
3218 didn't include your host/target information or if that information is
3219 incomplete or out of date. Send a note to
3220 @email{gcc@@gcc.gnu.org} detailing how the information should be changed.
3222 If you find a bug, please report it following the
3223 @uref{../bugs/,,bug reporting guidelines}.
3225 If you want to print the GCC manuals, do @samp{cd @var{objdir}; make
3226 dvi}. You will need to have @command{texi2dvi} (version at least 4.7)
3227 and @TeX{} installed. This creates a number of @file{.dvi} files in
3228 subdirectories of @file{@var{objdir}}; these may be converted for
3229 printing with programs such as @command{dvips}. Alternately, by using
3230 @samp{make pdf} in place of @samp{make dvi}, you can create documentation
3231 in the form of @file{.pdf} files; this requires @command{texi2pdf}, which
3232 is included with Texinfo version 4.8 and later. You can also
3233 @uref{https://shop.fsf.org/,,buy printed manuals from the
3234 Free Software Foundation}, though such manuals may not be for the most
3235 recent version of GCC@.
3237 If you would like to generate online HTML documentation, do @samp{cd
3238 @var{objdir}; make html} and HTML will be generated for the gcc manuals in
3239 @file{@var{objdir}/gcc/HTML}.
3246 @uref{./index.html,,Return to the GCC Installation page}
3250 @c ***Binaries****************************************************************
3252 @comment node-name, next, previous, up
3253 @node Binaries, Specific, Installing GCC, Top
3257 @chapter Installing GCC: Binaries
3260 @cindex Installing GCC: Binaries
3262 We are often asked about pre-compiled versions of GCC@. While we cannot
3263 provide these for all platforms, below you'll find links to binaries for
3264 various platforms where creating them by yourself is not easy due to various
3267 Please note that we did not create these binaries, nor do we
3268 support them. If you have any problems installing them, please
3269 contact their makers.
3276 @uref{http://www.bullfreeware.com,,Bull's Open Source Software Archive for
3277 for AIX 5L and AIX 6};
3280 @uref{http://www.perzl.org/aix/,,AIX Open Source Packages (AIX5L AIX 6.1
3285 DOS---@uref{http://www.delorie.com/djgpp/,,DJGPP}.
3291 @uref{http://hpux.connect.org.uk/,,HP-UX Porting Center};
3295 Solaris 2 (SPARC, Intel):
3298 @uref{https://www.opencsw.org/,,OpenCSW}
3301 @uref{http://jupiterrise.com/tgcware/,,TGCware}
3308 The @uref{https://brew.sh,,Homebrew} package manager;
3310 @uref{https://www.macports.org,,MacPorts}.
3317 The @uref{https://sourceware.org/cygwin/,,Cygwin} project;
3319 The @uref{http://www.mingw.org/,,MinGW} and
3320 @uref{http://mingw-w64.org/doku.php,,mingw-w64} projects.
3324 @uref{http://www.openpkg.org/,,OpenPKG} offers binaries for quite a
3325 number of platforms.
3328 The @uref{http://gcc.gnu.org/wiki/GFortranBinaries,,GFortran Wiki} has
3329 links to GNU Fortran binaries for several platforms.
3337 @uref{./index.html,,Return to the GCC Installation page}
3341 @c ***Specific****************************************************************
3343 @comment node-name, next, previous, up
3344 @node Specific, Old, Binaries, Top
3348 @chapter Host/target specific installation notes for GCC
3351 @cindex Specific installation notes
3352 @cindex Target specific installation
3353 @cindex Host specific installation
3354 @cindex Target specific installation notes
3356 Please read this document carefully @emph{before} installing the
3357 GNU Compiler Collection on your machine.
3359 Note that this list of install notes is @emph{not} a list of supported
3360 hosts or targets. Not all supported hosts and targets are listed
3361 here, only the ones that require host-specific or target-specific
3362 information have to.
3367 @uref{#aarch64-x-x,,aarch64*-*-*}
3369 @uref{#alpha-x-x,,alpha*-*-*}
3371 @uref{#amd64-x-solaris2,,amd64-*-solaris2*}
3373 @uref{#arm-x-eabi,,arm-*-eabi}
3377 @uref{#bfin,,Blackfin}
3381 @uref{#x-x-freebsd,,*-*-freebsd*}
3383 @uref{#h8300-hms,,h8300-hms}
3385 @uref{#hppa-hp-hpux,,hppa*-hp-hpux*}
3387 @uref{#hppa-hp-hpux10,,hppa*-hp-hpux10}
3389 @uref{#hppa-hp-hpux11,,hppa*-hp-hpux11}
3391 @uref{#x-x-linux-gnu,,*-*-linux-gnu}
3393 @uref{#ix86-x-linux,,i?86-*-linux*}
3395 @uref{#ix86-x-solaris2,,i?86-*-solaris2*}
3397 @uref{#ia64-x-linux,,ia64-*-linux}
3399 @uref{#ia64-x-hpux,,ia64-*-hpux*}
3401 @uref{#x-ibm-aix,,*-ibm-aix*}
3403 @uref{#iq2000-x-elf,,iq2000-*-elf}
3405 @uref{#lm32-x-elf,,lm32-*-elf}
3407 @uref{#lm32-x-uclinux,,lm32-*-uclinux}
3409 @uref{#m32c-x-elf,,m32c-*-elf}
3411 @uref{#m32r-x-elf,,m32r-*-elf}
3413 @uref{#m68k-x-x,,m68k-*-*}
3415 @uref{#m68k-uclinux,,m68k-uclinux}
3417 @uref{#microblaze-x-elf,,microblaze-*-elf}
3419 @uref{#mips-x-x,,mips-*-*}
3421 @uref{#nds32le-x-elf,,nds32le-*-elf}
3423 @uref{#nds32be-x-elf,,nds32be-*-elf}
3425 @uref{#nvptx-x-none,,nvptx-*-none}
3427 @uref{#or1k-x-elf,,or1k-*-elf}
3429 @uref{#or1k-x-linux,,or1k-*-linux}
3431 @uref{#powerpc-x-x,,powerpc*-*-*}
3433 @uref{#powerpc-x-darwin,,powerpc-*-darwin*}
3435 @uref{#powerpc-x-elf,,powerpc-*-elf}
3437 @uref{#powerpc-x-linux-gnu,,powerpc*-*-linux-gnu*}
3439 @uref{#powerpc-x-netbsd,,powerpc-*-netbsd*}
3441 @uref{#powerpc-x-eabisim,,powerpc-*-eabisim}
3443 @uref{#powerpc-x-eabi,,powerpc-*-eabi}
3445 @uref{#powerpcle-x-elf,,powerpcle-*-elf}
3447 @uref{#powerpcle-x-eabisim,,powerpcle-*-eabisim}
3449 @uref{#powerpcle-x-eabi,,powerpcle-*-eabi}
3451 @uref{#riscv32-x-elf,,riscv32-*-elf}
3453 @uref{#riscv32-x-linux,,riscv32-*-linux}
3455 @uref{#riscv64-x-elf,,riscv64-*-elf}
3457 @uref{#riscv64-x-linux,,riscv64-*-linux}
3459 @uref{#s390-x-linux,,s390-*-linux*}
3461 @uref{#s390x-x-linux,,s390x-*-linux*}
3463 @uref{#s390x-ibm-tpf,,s390x-ibm-tpf*}
3465 @uref{#x-x-solaris2,,*-*-solaris2*}
3467 @uref{#sparc-x-x,,sparc*-*-*}
3469 @uref{#sparc-sun-solaris2,,sparc-sun-solaris2*}
3471 @uref{#sparc-x-linux,,sparc-*-linux*}
3473 @uref{#sparc64-x-solaris2,,sparc64-*-solaris2*}
3475 @uref{#sparcv9-x-solaris2,,sparcv9-*-solaris2*}
3477 @uref{#c6x-x-x,,c6x-*-*}
3479 @uref{#tilegx-x-linux,,tilegx-*-linux*}
3481 @uref{#tilegxbe-x-linux,,tilegxbe-*-linux*}
3483 @uref{#tilepro-x-linux,,tilepro-*-linux*}
3485 @uref{#visium-x-elf, visium-*-elf}
3487 @uref{#x-x-vxworks,,*-*-vxworks*}
3489 @uref{#x86-64-x-x,,x86_64-*-*, amd64-*-*}
3491 @uref{#x86-64-x-solaris2,,x86_64-*-solaris2*}
3493 @uref{#xtensa-x-elf,,xtensa*-*-elf}
3495 @uref{#xtensa-x-linux,,xtensa*-*-linux*}
3497 @uref{#windows,,Microsoft Windows}
3499 @uref{#x-x-cygwin,,*-*-cygwin}
3501 @uref{#x-x-mingw32,,*-*-mingw32}
3505 @uref{#older,,Older systems}
3510 @uref{#elf,,all ELF targets} (SVR4, Solaris 2, etc.)
3516 <!-- -------- host/target specific issues start here ---------------- -->
3519 @anchor{aarch64-x-x}
3520 @heading aarch64*-*-*
3521 Binutils pre 2.24 does not have support for selecting @option{-mabi} and
3522 does not support ILP32. If it is used to build GCC 4.9 or later, GCC will
3523 not support option @option{-mabi=ilp32}.
3525 To enable a workaround for the Cortex-A53 erratum number 835769 by default
3526 (for all CPUs regardless of -mcpu option given) at configure time use the
3527 @option{--enable-fix-cortex-a53-835769} option. This will enable the fix by
3528 default and can be explicitly disabled during compilation by passing the
3529 @option{-mno-fix-cortex-a53-835769} option. Conversely,
3530 @option{--disable-fix-cortex-a53-835769} will disable the workaround by
3531 default. The workaround is disabled by default if neither of
3532 @option{--enable-fix-cortex-a53-835769} or
3533 @option{--disable-fix-cortex-a53-835769} is given at configure time.
3535 To enable a workaround for the Cortex-A53 erratum number 843419 by default
3536 (for all CPUs regardless of -mcpu option given) at configure time use the
3537 @option{--enable-fix-cortex-a53-843419} option. This workaround is applied at
3538 link time. Enabling the workaround will cause GCC to pass the relevant option
3539 to the linker. It can be explicitly disabled during compilation by passing the
3540 @option{-mno-fix-cortex-a53-843419} option. Conversely,
3541 @option{--disable-fix-cortex-a53-843419} will disable the workaround by default.
3542 The workaround is disabled by default if neither of
3543 @option{--enable-fix-cortex-a53-843419} or
3544 @option{--disable-fix-cortex-a53-843419} is given at configure time.
3546 To enable Branch Target Identification Mechanism and Return Address Signing by
3547 default at configure time use the @option{--enable-standard-branch-protection}
3548 option. This is equivalent to having @option{-mbranch-protection=standard}
3549 during compilation. This can be explicitly disabled during compilation by
3550 passing the @option{-mbranch-protection=none} option which turns off all
3551 types of branch protections. Conversely,
3552 @option{--disable-standard-branch-protection} will disable both the
3553 protections by default. This mechanism is turned off by default if neither
3554 of the options are given at configure time.
3561 This section contains general configuration information for all
3562 Alpha-based platforms using ELF@. In addition to reading this
3563 section, please read all other sections that match your target.
3568 @anchor{amd64-x-solaris2}
3569 @heading amd64-*-solaris2*
3570 This is a synonym for @samp{x86_64-*-solaris2*}.
3575 @anchor{amdgcn-unknown-amdhsa}
3576 @heading amdgcn-unknown-amdhsa
3579 Instead of GNU Binutils, you will need to install LLVM 6, or later, and copy
3580 @file{bin/llvm-mc} to @file{amdgcn-unknown-amdhsa/bin/as},
3581 @file{bin/lld} to @file{amdgcn-unknown-amdhsa/bin/ld},
3582 @file{bin/llvm-nm} to @file{amdgcn-unknown-amdhsa/bin/nm}, and
3583 @file{bin/llvm-ar} to both @file{bin/amdgcn-unknown-amdhsa-ar} and
3584 @file{bin/amdgcn-unknown-amdhsa-ranlib}.
3586 Use Newlib (2019-01-16, or newer).
3588 To run the binaries, install the HSA Runtime from the
3589 @uref{https://rocm.github.io,,ROCm Platform}, and use
3590 @file{libexec/gcc/amdhsa-unknown-amdhsa/@var{version}/gcn-run} to launch them
3596 @anchor{arc-x-elf32}
3597 @heading arc-*-elf32
3599 Use @samp{configure --target=arc-elf32 --with-cpu=@var{cpu} --enable-languages="c,c++"}
3600 to configure GCC, with @var{cpu} being one of @samp{arc600}, @samp{arc601},
3606 @anchor{arc-linux-uclibc}
3607 @heading arc-linux-uclibc
3609 Use @samp{configure --target=arc-linux-uclibc --with-cpu=arc700 --enable-languages="c,c++"} to configure GCC@.
3616 ARM-family processors.
3618 Building the Ada frontend commonly fails (an infinite loop executing
3619 @code{xsinfo}) if the host compiler is GNAT 4.8. Host compilers built from the
3620 GNAT 4.6, 4.9 or 5 release branches are known to succeed.
3627 ATMEL AVR-family micro controllers. These are used in embedded
3628 applications. There are no standard Unix configurations.
3630 @xref{AVR Options,, AVR Options, gcc, Using the GNU Compiler
3634 See ``AVR Options'' in the main manual
3636 for the list of supported MCU types.
3638 Use @samp{configure --target=avr --enable-languages="c"} to configure GCC@.
3640 Further installation notes and other useful information about AVR tools
3641 can also be obtained from:
3645 @uref{http://www.nongnu.org/avr/,,http://www.nongnu.org/avr/}
3647 @uref{http://www.amelek.gda.pl/avr/,,http://www.amelek.gda.pl/avr/}
3650 The following error:
3652 Error: register required
3655 indicates that you should upgrade to a newer version of the binutils.
3662 The Blackfin processor, an Analog Devices DSP.
3664 @xref{Blackfin Options,, Blackfin Options, gcc, Using the GNU Compiler
3668 See ``Blackfin Options'' in the main manual
3671 More information, and a version of binutils with support for this processor,
3672 are available at @uref{https://sourceforge.net/projects/adi-toolchain/}.
3679 The CR16 CompactRISC architecture is a 16-bit architecture. This
3680 architecture is used in embedded applications.
3683 @xref{CR16 Options,, CR16 Options, gcc, Using and Porting the GNU Compiler
3688 See ``CR16 Options'' in the main manual for a list of CR16-specific options.
3691 Use @samp{configure --target=cr16-elf --enable-languages=c,c++} to configure
3692 GCC@ for building a CR16 elf cross-compiler.
3694 Use @samp{configure --target=cr16-uclinux --enable-languages=c,c++} to
3695 configure GCC@ for building a CR16 uclinux cross-compiler.
3702 CRIS is the CPU architecture in Axis Communications ETRAX system-on-a-chip
3703 series. These are used in embedded applications.
3706 @xref{CRIS Options,, CRIS Options, gcc, Using the GNU Compiler
3710 See ``CRIS Options'' in the main manual
3712 for a list of CRIS-specific options.
3714 There are a few different CRIS targets:
3717 Mainly for monolithic embedded systems. Includes a multilib for the
3718 @samp{v10} core used in @samp{ETRAX 100 LX}.
3719 @item cris-axis-linux-gnu
3720 A GNU/Linux port for the CRIS architecture, currently targeting
3721 @samp{ETRAX 100 LX} by default.
3724 Pre-packaged tools can be obtained from
3725 @uref{ftp://ftp.axis.com/@/pub/@/axis/@/tools/@/cris/@/compiler-kit/}. More
3726 information about this platform is available at
3727 @uref{http://developer.axis.com/}.
3734 Please have a look at the @uref{binaries.html,,binaries page}.
3736 You cannot install GCC by itself on MSDOS; it will not compile under
3737 any MSDOS compiler except itself. You need to get the complete
3738 compilation package DJGPP, which includes binaries as well as sources,
3739 and includes all the necessary compilation tools and libraries.
3744 @anchor{epiphany-x-elf}
3745 @heading epiphany-*-elf
3747 This configuration is intended for embedded systems.
3752 @anchor{x-x-freebsd}
3753 @heading *-*-freebsd*
3754 Support for FreeBSD 1 was discontinued in GCC 3.2. Support for
3755 FreeBSD 2 (and any mutant a.out variants of FreeBSD 3) was
3756 discontinued in GCC 4.0.
3758 In order to better utilize FreeBSD base system functionality and match
3759 the configuration of the system compiler, GCC 4.5 and above as well as
3760 GCC 4.4 past 2010-06-20 leverage SSP support in libc (which is present
3761 on FreeBSD 7 or later) and the use of @code{__cxa_atexit} by default
3762 (on FreeBSD 6 or later). The use of @code{dl_iterate_phdr} inside
3763 @file{libgcc_s.so.1} and boehm-gc (on FreeBSD 7 or later) is enabled
3764 by GCC 4.5 and above.
3766 We support FreeBSD using the ELF file format with DWARF 2 debugging
3767 for all CPU architectures. You may use @option{-gstabs} instead of
3768 @option{-g}, if you really want the old debugging format. There are
3769 no known issues with mixing object files and libraries with different
3770 debugging formats. Otherwise, this release of GCC should now match
3771 more of the configuration used in the stock FreeBSD configuration of
3772 GCC@. In particular, @option{--enable-threads} is now configured by
3773 default. However, as a general user, do not attempt to replace the
3774 system compiler with this release. Known to bootstrap and check with
3775 good results on FreeBSD 7.2-STABLE@. In the past, known to bootstrap
3776 and check with good results on FreeBSD 3.0, 3.4, 4.0, 4.2, 4.3, 4.4,
3777 4.5, 4.8, 4.9 and 5-CURRENT@.
3779 The version of binutils installed in @file{/usr/bin} probably works
3780 with this release of GCC@. Bootstrapping against the latest GNU
3781 binutils and/or the version found in @file{/usr/ports/devel/binutils} has
3782 been known to enable additional features and improve overall testsuite
3783 results. However, it is currently known that boehm-gc may not configure
3784 properly on FreeBSD prior to the FreeBSD 7.0 release with GNU binutils
3793 This configuration is intended for embedded systems.
3800 Renesas H8/300 series of processors.
3802 Please have a look at the @uref{binaries.html,,binaries page}.
3804 The calling convention and structure layout has changed in release 2.6.
3805 All code must be recompiled. The calling convention now passes the
3806 first three arguments in function calls in registers. Structures are no
3807 longer a multiple of 2 bytes.
3812 @anchor{hppa-hp-hpux}
3813 @heading hppa*-hp-hpux*
3814 Support for HP-UX version 9 and older was discontinued in GCC 3.4.
3816 We require using gas/binutils on all hppa platforms. Version 2.19 or
3817 later is recommended.
3819 It may be helpful to configure GCC with the
3820 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}} and
3821 @option{--with-as=@dots{}} options to ensure that GCC can find GAS@.
3823 The HP assembler should not be used with GCC. It is rarely tested and may
3824 not work. It shouldn't be used with any languages other than C due to its
3827 Specifically, @option{-g} does not work (HP-UX uses a peculiar debugging
3828 format which GCC does not know about). It also inserts timestamps
3829 into each object file it creates, causing the 3-stage comparison test to
3830 fail during a bootstrap. You should be able to continue by saying
3831 @samp{make all-host all-target} after getting the failure from @samp{make}.
3833 Various GCC features are not supported. For example, it does not support weak
3834 symbols or alias definitions. As a result, explicit template instantiations
3835 are required when using C++. This makes it difficult if not impossible to
3836 build many C++ applications.
3838 There are two default scheduling models for instructions. These are
3839 PROCESSOR_7100LC and PROCESSOR_8000. They are selected from the pa-risc
3840 architecture specified for the target machine when configuring.
3841 PROCESSOR_8000 is the default. PROCESSOR_7100LC is selected when
3842 the target is a @samp{hppa1*} machine.
3844 The PROCESSOR_8000 model is not well suited to older processors. Thus,
3845 it is important to completely specify the machine architecture when
3846 configuring if you want a model other than PROCESSOR_8000. The macro
3847 TARGET_SCHED_DEFAULT can be defined in BOOT_CFLAGS if a different
3848 default scheduling model is desired.
3850 As of GCC 4.0, GCC uses the UNIX 95 namespace for HP-UX 10.10
3851 through 11.00, and the UNIX 98 namespace for HP-UX 11.11 and later.
3852 This namespace change might cause problems when bootstrapping with
3853 an earlier version of GCC or the HP compiler as essentially the same
3854 namespace is required for an entire build. This problem can be avoided
3855 in a number of ways. With HP cc, @env{UNIX_STD} can be set to @samp{95}
3856 or @samp{98}. Another way is to add an appropriate set of predefines
3857 to @env{CC}. The description for the @option{munix=} option contains
3858 a list of the predefines used with each standard.
3860 More specific information to @samp{hppa*-hp-hpux*} targets follows.
3865 @anchor{hppa-hp-hpux10}
3866 @heading hppa*-hp-hpux10
3867 For hpux10.20, we @emph{highly} recommend you pick up the latest sed patch
3868 @code{PHCO_19798} from HP@.
3870 The C++ ABI has changed incompatibly in GCC 4.0. COMDAT subspaces are
3871 used for one-only code and data. This resolves many of the previous
3872 problems in using C++ on this target. However, the ABI is not compatible
3873 with the one implemented under HP-UX 11 using secondary definitions.
3878 @anchor{hppa-hp-hpux11}
3879 @heading hppa*-hp-hpux11
3880 GCC 3.0 and up support HP-UX 11. GCC 2.95.x is not supported and cannot
3881 be used to compile GCC 3.0 and up.
3883 The libffi library haven't been ported to 64-bit HP-UX@ and doesn't build.
3885 Refer to @uref{binaries.html,,binaries} for information about obtaining
3886 precompiled GCC binaries for HP-UX@. Precompiled binaries must be obtained
3887 to build the Ada language as it cannot be bootstrapped using C@. Ada is
3888 only available for the 32-bit PA-RISC runtime.
3890 Starting with GCC 3.4 an ISO C compiler is required to bootstrap. The
3891 bundled compiler supports only traditional C; you will need either HP's
3892 unbundled compiler, or a binary distribution of GCC@.
3894 It is possible to build GCC 3.3 starting with the bundled HP compiler,
3895 but the process requires several steps. GCC 3.3 can then be used to
3896 build later versions.
3898 There are several possible approaches to building the distribution.
3899 Binutils can be built first using the HP tools. Then, the GCC
3900 distribution can be built. The second approach is to build GCC
3901 first using the HP tools, then build binutils, then rebuild GCC@.
3902 There have been problems with various binary distributions, so it
3903 is best not to start from a binary distribution.
3905 On 64-bit capable systems, there are two distinct targets. Different
3906 installation prefixes must be used if both are to be installed on
3907 the same system. The @samp{hppa[1-2]*-hp-hpux11*} target generates code
3908 for the 32-bit PA-RISC runtime architecture and uses the HP linker.
3909 The @samp{hppa64-hp-hpux11*} target generates 64-bit code for the
3910 PA-RISC 2.0 architecture.
3912 The script config.guess now selects the target type based on the compiler
3913 detected during configuration. You must define @env{PATH} or @env{CC} so
3914 that configure finds an appropriate compiler for the initial bootstrap.
3915 When @env{CC} is used, the definition should contain the options that are
3916 needed whenever @env{CC} is used.
3918 Specifically, options that determine the runtime architecture must be
3919 in @env{CC} to correctly select the target for the build. It is also
3920 convenient to place many other compiler options in @env{CC}. For example,
3921 @env{CC="cc -Ac +DA2.0W -Wp,-H16376 -D_CLASSIC_TYPES -D_HPUX_SOURCE"}
3922 can be used to bootstrap the GCC 3.3 branch with the HP compiler in
3923 64-bit K&R/bundled mode. The @option{+DA2.0W} option will result in
3924 the automatic selection of the @samp{hppa64-hp-hpux11*} target. The
3925 macro definition table of cpp needs to be increased for a successful
3926 build with the HP compiler. _CLASSIC_TYPES and _HPUX_SOURCE need to
3927 be defined when building with the bundled compiler, or when using the
3928 @option{-Ac} option. These defines aren't necessary with @option{-Ae}.
3930 It is best to explicitly configure the @samp{hppa64-hp-hpux11*} target
3931 with the @option{--with-ld=@dots{}} option. This overrides the standard
3932 search for ld. The two linkers supported on this target require different
3933 commands. The default linker is determined during configuration. As a
3934 result, it's not possible to switch linkers in the middle of a GCC build.
3935 This has been reported to sometimes occur in unified builds of binutils
3938 A recent linker patch must be installed for the correct operation of
3939 GCC 3.3 and later. @code{PHSS_26559} and @code{PHSS_24304} are the
3940 oldest linker patches that are known to work. They are for HP-UX
3941 11.00 and 11.11, respectively. @code{PHSS_24303}, the companion to
3942 @code{PHSS_24304}, might be usable but it hasn't been tested. These
3943 patches have been superseded. Consult the HP patch database to obtain
3944 the currently recommended linker patch for your system.
3946 The patches are necessary for the support of weak symbols on the
3947 32-bit port, and for the running of initializers and finalizers. Weak
3948 symbols are implemented using SOM secondary definition symbols. Prior
3949 to HP-UX 11, there are bugs in the linker support for secondary symbols.
3950 The patches correct a problem of linker core dumps creating shared
3951 libraries containing secondary symbols, as well as various other
3952 linking issues involving secondary symbols.
3954 GCC 3.3 uses the ELF DT_INIT_ARRAY and DT_FINI_ARRAY capabilities to
3955 run initializers and finalizers on the 64-bit port. The 32-bit port
3956 uses the linker @option{+init} and @option{+fini} options for the same
3957 purpose. The patches correct various problems with the +init/+fini
3958 options, including program core dumps. Binutils 2.14 corrects a
3959 problem on the 64-bit port resulting from HP's non-standard use of
3960 the .init and .fini sections for array initializers and finalizers.
3962 Although the HP and GNU linkers are both supported for the
3963 @samp{hppa64-hp-hpux11*} target, it is strongly recommended that the
3964 HP linker be used for link editing on this target.
3966 At this time, the GNU linker does not support the creation of long
3967 branch stubs. As a result, it cannot successfully link binaries
3968 containing branch offsets larger than 8 megabytes. In addition,
3969 there are problems linking shared libraries, linking executables
3970 with @option{-static}, and with dwarf2 unwind and exception support.
3971 It also doesn't provide stubs for internal calls to global functions
3972 in shared libraries, so these calls cannot be overloaded.
3974 The HP dynamic loader does not support GNU symbol versioning, so symbol
3975 versioning is not supported. It may be necessary to disable symbol
3976 versioning with @option{--disable-symvers} when using GNU ld.
3978 POSIX threads are the default. The optional DCE thread library is not
3979 supported, so @option{--enable-threads=dce} does not work.
3984 @anchor{x-x-linux-gnu}
3985 @heading *-*-linux-gnu
3986 Versions of libstdc++-v3 starting with 3.2.1 require bug fixes present
3987 in glibc 2.2.5 and later. More information is available in the
3988 libstdc++-v3 documentation.
3993 @anchor{ix86-x-linux}
3994 @heading i?86-*-linux*
3995 As of GCC 3.3, binutils 2.13.1 or later is required for this platform.
3996 See @uref{http://gcc.gnu.org/PR10877,,bug 10877} for more information.
3998 If you receive Signal 11 errors when building on GNU/Linux, then it is
3999 possible you have a hardware problem. Further information on this can be
4000 found on @uref{http://www.bitwizard.nl/sig11/,,www.bitwizard.nl}.
4005 @anchor{ix86-x-solaris2}
4006 @heading i?86-*-solaris2*
4007 Use this for Solaris 11 or later on x86 and x86-64 systems. Starting
4008 with GCC 4.7, there is also a 64-bit @samp{amd64-*-solaris2*} or
4009 @samp{x86_64-*-solaris2*} configuration that corresponds to
4010 @samp{sparcv9-sun-solaris2*}.
4012 It is recommended that you configure GCC to use the GNU assembler. The
4013 versions included in Solaris 11, from GNU binutils 2.19 or
4014 newer (also available as @file{/usr/bin/gas} and
4015 @file{/usr/gnu/bin/as}), work fine. The current version, from GNU
4016 binutils 2.32, is known to work, but the version from GNU binutils 2.26
4017 must be avoided. Recent versions of the Solaris assembler in
4018 @file{/usr/ccs/bin/as} work almost as well, though.
4019 @c FIXME: as patch requirements?
4021 For linking, the Solaris linker, is preferred. If you want to use the GNU
4022 linker instead, the version in Solaris 11, from GNU binutils 2.19 or
4023 newer (also in @file{/usr/gnu/bin/ld} and @file{/usr/bin/gld}), works,
4024 as does the latest version, from GNU binutils 2.32.
4026 To use GNU @command{as}, configure with the options
4027 @option{--with-gnu-as --with-as=@//usr/@/gnu/@/bin/@/as}. It may be necessary
4028 to configure with @option{--without-gnu-ld --with-ld=@//usr/@/ccs/@/bin/@/ld} to
4029 guarantee use of Sun @command{ld}.
4030 @c FIXME: why --without-gnu-ld --with-ld?
4035 @anchor{ia64-x-linux}
4036 @heading ia64-*-linux
4037 IA-64 processor (also known as IPF, or Itanium Processor Family)
4040 If you are using the installed system libunwind library with
4041 @option{--with-system-libunwind}, then you must use libunwind 0.98 or
4044 None of the following versions of GCC has an ABI that is compatible
4045 with any of the other versions in this list, with the exception that
4046 Red Hat 2.96 and Trillian 000171 are compatible with each other:
4047 3.1, 3.0.2, 3.0.1, 3.0, Red Hat 2.96, and Trillian 000717.
4048 This primarily affects C++ programs and programs that create shared libraries.
4049 GCC 3.1 or later is recommended for compiling linux, the kernel.
4050 As of version 3.1 GCC is believed to be fully ABI compliant, and hence no
4051 more major ABI changes are expected.
4056 @anchor{ia64-x-hpux}
4057 @heading ia64-*-hpux*
4058 Building GCC on this target requires the GNU Assembler. The bundled HP
4059 assembler will not work. To prevent GCC from using the wrong assembler,
4060 the option @option{--with-gnu-as} may be necessary.
4062 The GCC libunwind library has not been ported to HPUX@. This means that for
4063 GCC versions 3.2.3 and earlier, @option{--enable-libunwind-exceptions}
4064 is required to build GCC@. For GCC 3.3 and later, this is the default.
4065 For gcc 3.4.3 and later, @option{--enable-libunwind-exceptions} is
4066 removed and the system libunwind library will always be used.
4070 <!-- rs6000-ibm-aix*, powerpc-ibm-aix* -->
4074 Support for AIX version 3 and older was discontinued in GCC 3.4.
4075 Support for AIX version 4.2 and older was discontinued in GCC 4.5.
4077 ``out of memory'' bootstrap failures may indicate a problem with
4078 process resource limits (ulimit). Hard limits are configured in the
4079 @file{/etc/security/limits} system configuration file.
4081 GCC 4.9 and above require a C++ compiler for bootstrap. IBM VAC++ / xlC
4082 cannot bootstrap GCC. xlc can bootstrap an older version of GCC and
4083 G++ can bootstrap recent releases of GCC.
4085 GCC can bootstrap with recent versions of IBM XLC, but bootstrapping
4086 with an earlier release of GCC is recommended. Bootstrapping with XLC
4087 requires a larger data segment, which can be enabled through the
4088 @var{LDR_CNTRL} environment variable, e.g.,
4091 % LDR_CNTRL=MAXDATA=0x50000000
4095 One can start with a pre-compiled version of GCC to build from
4096 sources. One may delete GCC's ``fixed'' header files when starting
4097 with a version of GCC built for an earlier release of AIX.
4099 To speed up the configuration phases of bootstrapping and installing GCC,
4100 one may use GNU Bash instead of AIX @command{/bin/sh}, e.g.,
4103 % CONFIG_SHELL=/opt/freeware/bin/bash
4104 % export CONFIG_SHELL
4107 and then proceed as described in @uref{build.html,,the build
4108 instructions}, where we strongly recommend specifying an absolute path
4109 to invoke @var{srcdir}/configure.
4111 Because GCC on AIX is built as a 32-bit executable by default,
4112 (although it can generate 64-bit programs) the GMP and MPFR libraries
4113 required by gfortran must be 32-bit libraries. Building GMP and MPFR
4114 as static archive libraries works better than shared libraries.
4116 Errors involving @code{alloca} when building GCC generally are due
4117 to an incorrect definition of @code{CC} in the Makefile or mixing files
4118 compiled with the native C compiler and GCC@. During the stage1 phase of
4119 the build, the native AIX compiler @strong{must} be invoked as @command{cc}
4120 (not @command{xlc}). Once @command{configure} has been informed of
4121 @command{xlc}, one needs to use @samp{make distclean} to remove the
4122 configure cache files and ensure that @env{CC} environment variable
4123 does not provide a definition that will confuse @command{configure}.
4124 If this error occurs during stage2 or later, then the problem most likely
4125 is the version of Make (see above).
4127 The native @command{as} and @command{ld} are recommended for
4128 bootstrapping on AIX@. The GNU Assembler, GNU Linker, and GNU
4129 Binutils version 2.20 is the minimum level that supports bootstrap on
4130 AIX 5@. The GNU Assembler has not been updated to support AIX 6@ or
4131 AIX 7. The native AIX tools do interoperate with GCC@.
4133 AIX 7.1 added partial support for DWARF debugging, but full support
4134 requires AIX 7.1 TL03 SP7 that supports additional DWARF sections and
4135 fixes a bug in the assembler. AIX 7.1 TL03 SP5 distributed a version
4136 of libm.a missing important symbols; a fix for IV77796 will be
4139 AIX 5.3 TL10, AIX 6.1 TL05 and AIX 7.1 TL00 introduced an AIX
4140 assembler change that sometimes produces corrupt assembly files
4141 causing AIX linker errors. The bug breaks GCC bootstrap on AIX and
4142 can cause compilation failures with existing GCC installations. An
4143 AIX iFix for AIX 5.3 is available (APAR IZ98385 for AIX 5.3 TL10, APAR
4144 IZ98477 for AIX 5.3 TL11 and IZ98134 for AIX 5.3 TL12). AIX 5.3 TL11 SP8,
4145 AIX 5.3 TL12 SP5, AIX 6.1 TL04 SP11, AIX 6.1 TL05 SP7, AIX 6.1 TL06 SP6,
4146 AIX 6.1 TL07 and AIX 7.1 TL01 should include the fix.
4148 Building @file{libstdc++.a} requires a fix for an AIX Assembler bug
4149 APAR IY26685 (AIX 4.3) or APAR IY25528 (AIX 5.1). It also requires a
4150 fix for another AIX Assembler bug and a co-dependent AIX Archiver fix
4151 referenced as APAR IY53606 (AIX 5.2) or as APAR IY54774 (AIX 5.1)
4153 @anchor{TransferAixShobj}
4154 @samp{libstdc++} in GCC 3.4 increments the major version number of the
4155 shared object and GCC installation places the @file{libstdc++.a}
4156 shared library in a common location which will overwrite the and GCC
4157 3.3 version of the shared library. Applications either need to be
4158 re-linked against the new shared library or the GCC 3.1 and GCC 3.3
4159 versions of the @samp{libstdc++} shared object needs to be available
4160 to the AIX runtime loader. The GCC 3.1 @samp{libstdc++.so.4}, if
4161 present, and GCC 3.3 @samp{libstdc++.so.5} shared objects can be
4162 installed for runtime dynamic loading using the following steps to set
4163 the @samp{F_LOADONLY} flag in the shared object for @emph{each}
4164 multilib @file{libstdc++.a} installed:
4166 Extract the shared objects from the currently installed
4167 @file{libstdc++.a} archive:
4169 % ar -x libstdc++.a libstdc++.so.4 libstdc++.so.5
4172 Enable the @samp{F_LOADONLY} flag so that the shared object will be
4173 available for runtime dynamic loading, but not linking:
4175 % strip -e libstdc++.so.4 libstdc++.so.5
4178 Archive the runtime-only shared object in the GCC 3.4
4179 @file{libstdc++.a} archive:
4181 % ar -q libstdc++.a libstdc++.so.4 libstdc++.so.5
4185 @uref{./configure.html#WithAixSoname,,@option{--with-aix-soname=svr4}}
4186 configure option may drop the need for this procedure for libraries that
4189 Linking executables and shared libraries may produce warnings of
4190 duplicate symbols. The assembly files generated by GCC for AIX always
4191 have included multiple symbol definitions for certain global variable
4192 and function declarations in the original program. The warnings should
4193 not prevent the linker from producing a correct library or runnable
4196 AIX 4.3 utilizes a ``large format'' archive to support both 32-bit and
4197 64-bit object modules. The routines provided in AIX 4.3.0 and AIX 4.3.1
4198 to parse archive libraries did not handle the new format correctly.
4199 These routines are used by GCC and result in error messages during
4200 linking such as ``not a COFF file''. The version of the routines shipped
4201 with AIX 4.3.1 should work for a 32-bit environment. The @option{-g}
4202 option of the archive command may be used to create archives of 32-bit
4203 objects using the original ``small format''. A correct version of the
4204 routines is shipped with AIX 4.3.2 and above.
4206 Some versions of the AIX binder (linker) can fail with a relocation
4207 overflow severe error when the @option{-bbigtoc} option is used to link
4208 GCC-produced object files into an executable that overflows the TOC@. A fix
4209 for APAR IX75823 (OVERFLOW DURING LINK WHEN USING GCC AND -BBIGTOC) is
4210 available from IBM Customer Support and from its
4211 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
4212 website as PTF U455193.
4214 The AIX 4.3.2.1 linker (bos.rte.bind_cmds Level 4.3.2.1) will dump core
4215 with a segmentation fault when invoked by any version of GCC@. A fix for
4216 APAR IX87327 is available from IBM Customer Support and from its
4217 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
4218 website as PTF U461879. This fix is incorporated in AIX 4.3.3 and above.
4220 The initial assembler shipped with AIX 4.3.0 generates incorrect object
4221 files. A fix for APAR IX74254 (64BIT DISASSEMBLED OUTPUT FROM COMPILER FAILS
4222 TO ASSEMBLE/BIND) is available from IBM Customer Support and from its
4223 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
4224 website as PTF U453956. This fix is incorporated in AIX 4.3.1 and above.
4226 AIX provides National Language Support (NLS)@. Compilers and assemblers
4227 use NLS to support locale-specific representations of various data
4228 formats including floating-point numbers (e.g., @samp{.} vs @samp{,} for
4229 separating decimal fractions). There have been problems reported where
4230 GCC does not produce the same floating-point formats that the assembler
4231 expects. If one encounters this problem, set the @env{LANG}
4232 environment variable to @samp{C} or @samp{En_US}.
4234 A default can be specified with the @option{-mcpu=@var{cpu_type}}
4235 switch and using the configure option @option{--with-cpu-@var{cpu_type}}.
4240 @anchor{iq2000-x-elf}
4241 @heading iq2000-*-elf
4242 Vitesse IQ2000 processors. These are used in embedded
4243 applications. There are no standard Unix configurations.
4250 Lattice Mico32 processor.
4251 This configuration is intended for embedded systems.
4256 @anchor{lm32-x-uclinux}
4257 @heading lm32-*-uclinux
4258 Lattice Mico32 processor.
4259 This configuration is intended for embedded systems running uClinux.
4266 Renesas M32C processor.
4267 This configuration is intended for embedded systems.
4274 Renesas M32R processor.
4275 This configuration is intended for embedded systems.
4283 @samp{m68k-*-elf*}, @samp{m68k-*-rtems}, @samp{m68k-*-uclinux} and
4285 build libraries for both M680x0 and ColdFire processors. If you only
4286 need the M680x0 libraries, you can omit the ColdFire ones by passing
4287 @option{--with-arch=m68k} to @command{configure}. Alternatively, you
4288 can omit the M680x0 libraries by passing @option{--with-arch=cf} to
4289 @command{configure}. These targets default to 5206 or 5475 code as
4290 appropriate for the target system when
4291 configured with @option{--with-arch=cf} and 68020 code otherwise.
4293 The @samp{m68k-*-netbsd} and
4294 @samp{m68k-*-openbsd} targets also support the @option{--with-arch}
4295 option. They will generate ColdFire CFV4e code when configured with
4296 @option{--with-arch=cf} and 68020 code otherwise.
4298 You can override the default processors listed above by configuring
4299 with @option{--with-cpu=@var{target}}. This @var{target} can either
4300 be a @option{-mcpu} argument or one of the following values:
4301 @samp{m68000}, @samp{m68010}, @samp{m68020}, @samp{m68030},
4302 @samp{m68040}, @samp{m68060}, @samp{m68020-40} and @samp{m68020-60}.
4304 GCC requires at least binutils version 2.17 on these targets.
4309 @anchor{m68k-x-uclinux}
4310 @heading m68k-*-uclinux
4311 GCC 4.3 changed the uClinux configuration so that it uses the
4312 @samp{m68k-linux-gnu} ABI rather than the @samp{m68k-elf} ABI.
4313 It also added improved support for C++ and flat shared libraries,
4314 both of which were ABI changes.
4319 @anchor{microblaze-x-elf}
4320 @heading microblaze-*-elf
4321 Xilinx MicroBlaze processor.
4322 This configuration is intended for embedded systems.
4329 If on a MIPS system you get an error message saying ``does not have gp
4330 sections for all it's [sic] sectons [sic]'', don't worry about it. This
4331 happens whenever you use GAS with the MIPS linker, but there is not
4332 really anything wrong, and it is okay to use the output file. You can
4333 stop such warnings by installing the GNU linker.
4335 It would be nice to extend GAS to produce the gp tables, but they are
4336 optional, and there should not be a warning about their absence.
4338 The libstdc++ atomic locking routines for MIPS targets requires MIPS II
4339 and later. A patch went in just after the GCC 3.3 release to
4340 make @samp{mips*-*-*} use the generic implementation instead. You can also
4341 configure for @samp{mipsel-elf} as a workaround. The
4342 @samp{mips*-*-linux*} target continues to use the MIPS II routines. More
4343 work on this is expected in future releases.
4345 @c If you make --with-llsc the default for another target, please also
4346 @c update the description of the --with-llsc option.
4348 The built-in @code{__sync_*} functions are available on MIPS II and
4349 later systems and others that support the @samp{ll}, @samp{sc} and
4350 @samp{sync} instructions. This can be overridden by passing
4351 @option{--with-llsc} or @option{--without-llsc} when configuring GCC.
4352 Since the Linux kernel emulates these instructions if they are
4353 missing, the default for @samp{mips*-*-linux*} targets is
4354 @option{--with-llsc}. The @option{--with-llsc} and
4355 @option{--without-llsc} configure options may be overridden at compile
4356 time by passing the @option{-mllsc} or @option{-mno-llsc} options to
4359 MIPS systems check for division by zero (unless
4360 @option{-mno-check-zero-division} is passed to the compiler) by
4361 generating either a conditional trap or a break instruction. Using
4362 trap results in smaller code, but is only supported on MIPS II and
4363 later. Also, some versions of the Linux kernel have a bug that
4364 prevents trap from generating the proper signal (@code{SIGFPE}). To enable
4365 the use of break, use the @option{--with-divide=breaks}
4366 @command{configure} option when configuring GCC@. The default is to
4367 use traps on systems that support them.
4372 @anchor{moxie-x-elf}
4373 @heading moxie-*-elf
4374 The moxie processor.
4379 @anchor{msp430-x-elf}
4380 @heading msp430-*-elf*
4381 TI MSP430 processor.
4382 This configuration is intended for embedded systems.
4384 @samp{msp430-*-elf} is the standard configuration with most GCC
4385 features enabled by default.
4387 @samp{msp430-*-elfbare} is tuned for a bare-metal environment, and disables
4388 features related to shared libraries and other functionality not used for
4389 this device. This reduces code and data usage of the GCC libraries, resulting
4390 in a minimal run-time environment by default.
4392 Features disabled by default include:
4394 @item transactional memory
4401 @anchor{nds32le-x-elf}
4402 @heading nds32le-*-elf
4403 Andes NDS32 target in little endian mode.
4408 @anchor{nds32be-x-elf}
4409 @heading nds32be-*-elf
4410 Andes NDS32 target in big endian mode.
4415 @anchor{nvptx-x-none}
4416 @heading nvptx-*-none
4419 Instead of GNU binutils, you will need to install
4420 @uref{https://github.com/MentorEmbedded/nvptx-tools/,,nvptx-tools}.
4421 Tell GCC where to find it:
4422 @option{--with-build-time-tools=[install-nvptx-tools]/nvptx-none/bin}.
4424 You will need newlib 3.0 git revision
4425 cd31fbb2aea25f94d7ecedc9db16dfc87ab0c316 or later. It can be
4426 automatically built together with GCC@. For this, add a symbolic link
4427 to nvptx-newlib's @file{newlib} directory to the directory containing
4430 Use the @option{--disable-sjlj-exceptions} and
4431 @option{--enable-newlib-io-long-long} options when configuring.
4438 The OpenRISC 1000 32-bit processor with delay slots.
4439 This configuration is intended for embedded systems.
4444 @anchor{or1k-x-linux}
4445 @heading or1k-*-linux
4446 The OpenRISC 1000 32-bit processor with delay slots.
4451 @anchor{powerpc-x-x}
4452 @heading powerpc-*-*
4453 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
4454 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
4456 You will need GNU binutils 2.20 or newer.
4461 @anchor{powerpc-x-darwin}
4462 @heading powerpc-*-darwin*
4463 PowerPC running Darwin (Mac OS X kernel).
4465 Pre-installed versions of Mac OS X may not include any developer tools,
4466 meaning that you will not be able to build GCC from source. Tool
4467 binaries are available at
4468 @uref{https://opensource.apple.com}.
4470 This version of GCC requires at least cctools-590.36. The
4471 cctools-590.36 package referenced from
4472 @uref{http://gcc.gnu.org/ml/gcc/2006-03/msg00507.html} will not work
4473 on systems older than 10.3.9 (aka darwin7.9.0).
4478 @anchor{powerpc-x-elf}
4479 @heading powerpc-*-elf
4480 PowerPC system in big endian mode, running System V.4.
4485 @anchor{powerpc-x-linux-gnu}
4486 @heading powerpc*-*-linux-gnu*
4487 PowerPC system in big endian mode running Linux.
4492 @anchor{powerpc-x-netbsd}
4493 @heading powerpc-*-netbsd*
4494 PowerPC system in big endian mode running NetBSD@.
4499 @anchor{powerpc-x-eabisim}
4500 @heading powerpc-*-eabisim
4501 Embedded PowerPC system in big endian mode for use in running under the
4507 @anchor{powerpc-x-eabi}
4508 @heading powerpc-*-eabi
4509 Embedded PowerPC system in big endian mode.
4514 @anchor{powerpcle-x-elf}
4515 @heading powerpcle-*-elf
4516 PowerPC system in little endian mode, running System V.4.
4521 @anchor{powerpcle-x-eabisim}
4522 @heading powerpcle-*-eabisim
4523 Embedded PowerPC system in little endian mode for use in running under
4529 @anchor{powerpcle-x-eabi}
4530 @heading powerpcle-*-eabi
4531 Embedded PowerPC system in little endian mode.
4538 The Renesas RL78 processor.
4539 This configuration is intended for embedded systems.
4544 @anchor{riscv32-x-elf}
4545 @heading riscv32-*-elf
4546 The RISC-V RV32 instruction set.
4547 This configuration is intended for embedded systems.
4548 This (and all other RISC-V) targets are supported upstream as of the
4549 binutils 2.28 release.
4554 @anchor{riscv32-x-linux}
4555 @heading riscv32-*-linux
4556 The RISC-V RV32 instruction set running GNU/Linux.
4557 This (and all other RISC-V) targets are supported upstream as of the
4558 binutils 2.28 release.
4563 @anchor{riscv64-x-elf}
4564 @heading riscv64-*-elf
4565 The RISC-V RV64 instruction set.
4566 This configuration is intended for embedded systems.
4567 This (and all other RISC-V) targets are supported upstream as of the
4568 binutils 2.28 release.
4573 @anchor{riscv64-x-linux}
4574 @heading riscv64-*-linux
4575 The RISC-V RV64 instruction set running GNU/Linux.
4576 This (and all other RISC-V) targets are supported upstream as of the
4577 binutils 2.28 release.
4584 The Renesas RX processor.
4589 @anchor{s390-x-linux}
4590 @heading s390-*-linux*
4591 S/390 system running GNU/Linux for S/390@.
4596 @anchor{s390x-x-linux}
4597 @heading s390x-*-linux*
4598 zSeries system (64-bit) running GNU/Linux for zSeries@.
4603 @anchor{s390x-ibm-tpf}
4604 @heading s390x-ibm-tpf*
4605 zSeries system (64-bit) running TPF@. This platform is
4606 supported as cross-compilation target only.
4611 @c Please use Solaris 2 to refer to all release of Solaris, starting
4612 @c with 2.0 until 2.6, 7, 8, etc. Solaris 1 was a marketing name for
4613 @c SunOS 4 releases which we don't use to avoid confusion. Solaris
4614 @c alone is too unspecific and must be avoided.
4615 @anchor{x-x-solaris2}
4616 @heading *-*-solaris2*
4617 Support for Solaris 10 has been removed in GCC 10. Support for Solaris
4618 9 has been removed in GCC 5. Support for Solaris 8 has been removed in
4619 GCC 4.8. Support for Solaris 7 has been removed in GCC 4.6.
4621 Solaris 11 provides GCC 4.5.2, 4.7.3, and 4.8.2 as
4622 @command{/usr/gcc/4.5/bin/gcc} or similar. Alternatively,
4623 you can install a pre-built GCC to bootstrap and install GCC. See the
4624 @uref{binaries.html,,binaries page} for details.
4626 The Solaris 2 @command{/bin/sh} will often fail to configure
4627 @samp{libstdc++-v3}or @samp{boehm-gc}. We therefore recommend using the
4628 following initial sequence of commands
4631 % CONFIG_SHELL=/bin/ksh
4632 % export CONFIG_SHELL
4636 and proceed as described in @uref{configure.html,,the configure instructions}.
4637 In addition we strongly recommend specifying an absolute path to invoke
4638 @command{@var{srcdir}/configure}.
4640 In Solaris 11, you need to check for @code{system/header},
4641 @code{system/linker}, and @code{developer/assembler} packages.
4643 Trying to use the linker and other tools in
4644 @file{/usr/ucb} to install GCC has been observed to cause trouble.
4645 For example, the linker may hang indefinitely. The fix is to remove
4646 @file{/usr/ucb} from your @env{PATH}.
4648 The build process works more smoothly with the legacy Sun tools so, if you
4649 have @file{/usr/xpg4/bin} in your @env{PATH}, we recommend that you place
4650 @file{/usr/bin} before @file{/usr/xpg4/bin} for the duration of the build.
4652 We recommend the use of the Solaris assembler or the GNU assembler, in
4653 conjunction with the Solaris linker. The GNU @command{as}
4654 versions included in Solaris 11,
4655 from GNU binutils 2.19 or newer (also in @file{/usr/bin/gas} and
4656 @file{/usr/gnu/bin/as}), are known to work.
4657 The current version, from GNU binutils 2.32,
4658 is known to work as well. Note that your mileage may vary
4659 if you use a combination of the GNU tools and the Solaris tools: while the
4660 combination GNU @command{as} + Sun @command{ld} should reasonably work,
4661 the reverse combination Sun @command{as} + GNU @command{ld} may fail to
4662 build or cause memory corruption at runtime in some cases for C++ programs.
4664 GNU @command{ld} usually works as well. Again, the current
4665 version (2.32) is known to work, but generally lacks platform specific
4666 features, so better stay with Solaris @command{ld}. To use the LTO linker
4667 plugin (@option{-fuse-linker-plugin}) with GNU @command{ld}, GNU
4668 binutils @emph{must} be configured with @option{--enable-largefile}.
4670 To enable symbol versioning in @samp{libstdc++} with the Solaris linker,
4671 you need to have any version of GNU @command{c++filt}, which is part of
4672 GNU binutils. @samp{libstdc++} symbol versioning will be disabled if no
4673 appropriate version is found. Solaris @command{c++filt} from the Solaris
4674 Studio compilers does @emph{not} work.
4681 This section contains general configuration information for all
4682 SPARC-based platforms. In addition to reading this section, please
4683 read all other sections that match your target.
4685 Newer versions of the GNU Multiple Precision Library (GMP), the MPFR
4686 library and the MPC library are known to be miscompiled by earlier
4687 versions of GCC on these platforms. We therefore recommend the use
4688 of the exact versions of these libraries listed as minimal versions
4689 in @uref{prerequisites.html,,the prerequisites}.
4694 @anchor{sparc-sun-solaris2}
4695 @heading sparc-sun-solaris2*
4696 When GCC is configured to use GNU binutils 2.14 or later, the binaries
4697 produced are smaller than the ones produced using Sun's native tools;
4698 this difference is quite significant for binaries containing debugging
4701 Starting with Solaris 7, the operating system is capable of executing
4702 64-bit SPARC V9 binaries. GCC 3.1 and later properly supports
4703 this; the @option{-m64} option enables 64-bit code generation.
4704 However, if all you want is code tuned for the UltraSPARC CPU, you
4705 should try the @option{-mtune=ultrasparc} option instead, which produces
4706 code that, unlike full 64-bit code, can still run on non-UltraSPARC
4709 When configuring the GNU Multiple Precision Library (GMP), the MPFR
4710 library or the MPC library on a Solaris 7 or later system, the canonical
4711 target triplet must be specified as the @command{build} parameter on the
4712 configure line. This target triplet can be obtained by invoking @command{./config.guess} in the toplevel source directory of GCC (and
4713 not that of GMP or MPFR or MPC). For example on a Solaris 9 system:
4716 % ./configure --build=sparc-sun-solaris2.9 --prefix=xxx
4722 @anchor{sparc-x-linux}
4723 @heading sparc-*-linux*
4728 @anchor{sparc64-x-solaris2}
4729 @heading sparc64-*-solaris2*
4730 When configuring the GNU Multiple Precision Library (GMP), the MPFR
4731 library or the MPC library, the canonical target triplet must be specified
4732 as the @command{build} parameter on the configure line. For example
4733 on a Solaris 9 system:
4736 % ./configure --build=sparc64-sun-solaris2.9 --prefix=xxx
4742 @anchor{sparcv9-x-solaris2}
4743 @heading sparcv9-*-solaris2*
4744 This is a synonym for @samp{sparc64-*-solaris2*}.
4751 The C6X family of processors. This port requires binutils-2.22 or newer.
4756 @anchor{tilegx-*-linux}
4757 @heading tilegx-*-linux*
4758 The TILE-Gx processor in little endian mode, running GNU/Linux. This
4759 port requires binutils-2.22 or newer.
4764 @anchor{tilegxbe-*-linux}
4765 @heading tilegxbe-*-linux*
4766 The TILE-Gx processor in big endian mode, running GNU/Linux. This
4767 port requires binutils-2.23 or newer.
4772 @anchor{tilepro-*-linux}
4773 @heading tilepro-*-linux*
4774 The TILEPro processor running GNU/Linux. This port requires
4775 binutils-2.22 or newer.
4780 @anchor{visium-x-elf}
4781 @heading visium-*-elf
4782 CDS VISIUMcore processor.
4783 This configuration is intended for embedded systems.
4788 @anchor{x-x-vxworks}
4789 @heading *-*-vxworks*
4790 Support for VxWorks is in flux. At present GCC supports @emph{only} the
4791 very recent VxWorks 5.5 (aka Tornado 2.2) release, and only on PowerPC@.
4792 We welcome patches for other architectures supported by VxWorks 5.5.
4793 Support for VxWorks AE would also be welcome; we believe this is merely
4794 a matter of writing an appropriate ``configlette'' (see below). We are
4795 not interested in supporting older, a.out or COFF-based, versions of
4798 VxWorks comes with an older version of GCC installed in
4799 @file{@var{$WIND_BASE}/host}; we recommend you do not overwrite it.
4800 Choose an installation @var{prefix} entirely outside @var{$WIND_BASE}.
4801 Before running @command{configure}, create the directories @file{@var{prefix}}
4802 and @file{@var{prefix}/bin}. Link or copy the appropriate assembler,
4803 linker, etc.@: into @file{@var{prefix}/bin}, and set your @var{PATH} to
4804 include that directory while running both @command{configure} and
4807 You must give @command{configure} the
4808 @option{--with-headers=@var{$WIND_BASE}/target/h} switch so that it can
4809 find the VxWorks system headers. Since VxWorks is a cross compilation
4810 target only, you must also specify @option{--target=@var{target}}.
4811 @command{configure} will attempt to create the directory
4812 @file{@var{prefix}/@var{target}/sys-include} and copy files into it;
4813 make sure the user running @command{configure} has sufficient privilege
4816 GCC's exception handling runtime requires a special ``configlette''
4817 module, @file{contrib/gthr_supp_vxw_5x.c}. Follow the instructions in
4818 that file to add the module to your kernel build. (Future versions of
4819 VxWorks will incorporate this module.)
4825 @heading x86_64-*-*, amd64-*-*
4826 GCC supports the x86-64 architecture implemented by the AMD64 processor
4827 (amd64-*-* is an alias for x86_64-*-*) on GNU/Linux, FreeBSD and NetBSD@.
4828 On GNU/Linux the default is a bi-arch compiler which is able to generate
4829 both 64-bit x86-64 and 32-bit x86 code (via the @option{-m32} switch).
4834 @anchor{x86-64-x-solaris2}
4835 @heading x86_64-*-solaris2*
4836 GCC also supports the x86-64 architecture implemented by the AMD64
4837 processor (@samp{amd64-*-*} is an alias for @samp{x86_64-*-*}) on
4838 Solaris 10 or later. Unlike other systems, without special options a
4839 bi-arch compiler is built which generates 32-bit code by default, but
4840 can generate 64-bit x86-64 code with the @option{-m64} switch. Since
4841 GCC 4.7, there is also a configuration that defaults to 64-bit code, but
4842 can generate 32-bit code with @option{-m32}. To configure and build
4843 this way, you have to provide all support libraries like @file{libgmp}
4844 as 64-bit code, configure with @option{--target=x86_64-pc-solaris2.1x}
4845 and @samp{CC=gcc -m64}.
4850 @anchor{xtensa-x-elf}
4851 @heading xtensa*-*-elf
4852 This target is intended for embedded Xtensa systems using the
4853 @samp{newlib} C library. It uses ELF but does not support shared
4854 objects. Designed-defined instructions specified via the
4855 Tensilica Instruction Extension (TIE) language are only supported
4856 through inline assembly.
4858 The Xtensa configuration information must be specified prior to
4859 building GCC@. The @file{include/xtensa-config.h} header
4860 file contains the configuration information. If you created your
4861 own Xtensa configuration with the Xtensa Processor Generator, the
4862 downloaded files include a customized copy of this header file,
4863 which you can use to replace the default header file.
4868 @anchor{xtensa-x-linux}
4869 @heading xtensa*-*-linux*
4870 This target is for Xtensa systems running GNU/Linux. It supports ELF
4871 shared objects and the GNU C library (glibc). It also generates
4872 position-independent code (PIC) regardless of whether the
4873 @option{-fpic} or @option{-fPIC} options are used. In other
4874 respects, this target is the same as the
4875 @uref{#xtensa*-*-elf,,@samp{xtensa*-*-elf}} target.
4881 @heading Microsoft Windows
4883 @subheading Intel 16-bit versions
4884 The 16-bit versions of Microsoft Windows, such as Windows 3.1, are not
4887 However, the 32-bit port has limited support for Microsoft
4888 Windows 3.11 in the Win32s environment, as a target only. See below.
4890 @subheading Intel 32-bit versions
4891 The 32-bit versions of Windows, including Windows 95, Windows NT, Windows
4892 XP, and Windows Vista, are supported by several different target
4893 platforms. These targets differ in which Windows subsystem they target
4894 and which C libraries are used.
4897 @item Cygwin @uref{#x-x-cygwin,,*-*-cygwin}: Cygwin provides a user-space
4898 Linux API emulation layer in the Win32 subsystem.
4899 @item MinGW @uref{#x-x-mingw32,,*-*-mingw32}: MinGW is a native GCC port for
4900 the Win32 subsystem that provides a subset of POSIX.
4901 @item MKS i386-pc-mks: NuTCracker from MKS. See
4902 @uref{https://www.mkssoftware.com} for more information.
4905 @subheading Intel 64-bit versions
4906 GCC contains support for x86-64 using the mingw-w64
4907 runtime library, available from @uref{http://mingw-w64.org/doku.php}.
4908 This library should be used with the target triple x86_64-pc-mingw32.
4910 Presently Windows for Itanium is not supported.
4912 @subheading Windows CE
4913 Windows CE is supported as a target only on Hitachi
4914 SuperH (sh-wince-pe), and MIPS (mips-wince-pe).
4916 @subheading Other Windows Platforms
4917 GCC no longer supports Windows NT on the Alpha or PowerPC.
4919 GCC no longer supports the Windows POSIX subsystem. However, it does
4920 support the Interix subsystem. See above.
4922 Old target names including *-*-winnt and *-*-windowsnt are no longer used.
4924 PW32 (i386-pc-pw32) support was never completed, and the project seems to
4925 be inactive. See @uref{http://pw32.sourceforge.net/} for more information.
4927 UWIN support has been removed due to a lack of maintenance.
4934 Ports of GCC are included with the
4935 @uref{http://www.cygwin.com/,,Cygwin environment}.
4937 GCC will build under Cygwin without modification; it does not build
4938 with Microsoft's C++ compiler and there are no plans to make it do so.
4940 The Cygwin native compiler can be configured to target any 32-bit x86
4941 cpu architecture desired; the default is i686-pc-cygwin. It should be
4942 used with as up-to-date a version of binutils as possible; use either
4943 the latest official GNU binutils release in the Cygwin distribution,
4944 or version 2.20 or above if building your own.
4949 @anchor{x-x-mingw32}
4950 @heading *-*-mingw32
4951 GCC will build with and support only MinGW runtime 3.12 and later.
4952 Earlier versions of headers are incompatible with the new default semantics
4953 of @code{extern inline} in @code{-std=c99} and @code{-std=gnu99} modes.
4959 @heading Older systems
4960 GCC contains support files for many older (1980s and early
4961 1990s) Unix variants. For the most part, support for these systems
4962 has not been deliberately removed, but it has not been maintained for
4963 several years and may suffer from bitrot.
4965 Starting with GCC 3.1, each release has a list of ``obsoleted'' systems.
4966 Support for these systems is still present in that release, but
4967 @command{configure} will fail unless the @option{--enable-obsolete}
4968 option is given. Unless a maintainer steps forward, support for these
4969 systems will be removed from the next release of GCC@.
4971 Support for old systems as hosts for GCC can cause problems if the
4972 workarounds for compiler, library and operating system bugs affect the
4973 cleanliness or maintainability of the rest of GCC@. In some cases, to
4974 bring GCC up on such a system, if still possible with current GCC, may
4975 require first installing an old version of GCC which did work on that
4976 system, and using it to compile a more recent GCC, to avoid bugs in the
4977 vendor compiler. Old releases of GCC 1 and GCC 2 are available in the
4978 @file{old-releases} directory on the @uref{../mirrors.html,,GCC mirror
4979 sites}. Header bugs may generally be avoided using
4980 @command{fixincludes}, but bugs or deficiencies in libraries and the
4981 operating system may still cause problems.
4983 Support for older systems as targets for cross-compilation is less
4984 problematic than support for them as hosts for GCC; if an enthusiast
4985 wishes to make such a target work again (including resurrecting any of
4986 the targets that never worked with GCC 2, starting from the last
4987 version before they were removed), patches
4988 @uref{../contribute.html,,following the usual requirements} would be
4989 likely to be accepted, since they should not affect the support for more
4992 For some systems, old versions of GNU binutils may also be useful,
4993 and are available from @file{pub/binutils/old-releases} on
4994 @uref{https://sourceware.org/mirrors.html,,sourceware.org mirror sites}.
4996 Some of the information on specific systems above relates to
4997 such older systems, but much of the information
4998 about GCC on such systems (which may no longer be applicable to
4999 current GCC) is to be found in the GCC texinfo manual.
5005 @heading all ELF targets (SVR4, Solaris 2, etc.)
5006 C++ support is significantly better on ELF targets if you use the
5007 @uref{./configure.html#with-gnu-ld,,GNU linker}; duplicate copies of
5008 inlines, vtables and template instantiations will be discarded
5017 @uref{./index.html,,Return to the GCC Installation page}
5021 @c ***Old documentation******************************************************
5023 @include install-old.texi
5029 @uref{./index.html,,Return to the GCC Installation page}
5033 @c ***GFDL********************************************************************
5041 @uref{./index.html,,Return to the GCC Installation page}
5045 @c ***************************************************************************
5046 @c Part 6 The End of the Document
5048 @comment node-name, next, previous, up
5049 @node Concept Index, , GNU Free Documentation License, Top
5053 @unnumbered Concept Index