1 \input texinfo.tex @c -*-texinfo-*-
4 @setfilename gccinstall.info
5 @settitle Installing GCC
10 @include gcc-common.texi
12 @c Specify title for specific html page
14 @settitle Installing GCC
17 @settitle Host/Target specific installation notes for GCC
19 @ifset prerequisiteshtml
20 @settitle Prerequisites for GCC
23 @settitle Downloading GCC
26 @settitle Installing GCC: Configuration
29 @settitle Installing GCC: Building
32 @settitle Installing GCC: Testing
34 @ifset finalinstallhtml
35 @settitle Installing GCC: Final installation
38 @settitle Installing GCC: Binaries
41 @settitle Installing GCC: Old documentation
44 @settitle Installing GCC: GNU Free Documentation License
47 @c Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997, 1998,
48 @c 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009 Free Software Foundation, Inc.
49 @c *** Converted to texinfo by Dean Wakerley, dean@wakerley.com
51 @c IMPORTANT: whenever you modify this file, run `install.texi2html' to
52 @c test the generation of HTML documents for the gcc.gnu.org web pages.
54 @c Do not use @footnote{} in this file as it breaks install.texi2html!
56 @c Include everything if we're not making html
60 @set prerequisiteshtml
71 @c Part 2 Summary Description and Copyright
73 Copyright @copyright{} 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997,
74 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007,
75 2008 Free Software Foundation, Inc.
77 Permission is granted to copy, distribute and/or modify this document
78 under the terms of the GNU Free Documentation License, Version 1.2 or
79 any later version published by the Free Software Foundation; with no
80 Invariant Sections, the Front-Cover texts being (a) (see below), and
81 with the Back-Cover Texts being (b) (see below). A copy of the
82 license is included in the section entitled ``@uref{./gfdl.html,,GNU
83 Free Documentation License}''.
85 (a) The FSF's Front-Cover Text is:
89 (b) The FSF's Back-Cover Text is:
91 You have freedom to copy and modify this GNU Manual, like GNU
92 software. Copies published by the Free Software Foundation raise
93 funds for GNU development.
98 @dircategory Software development
100 * gccinstall: (gccinstall). Installing the GNU Compiler Collection.
103 @c Part 3 Titlepage and Copyright
105 @title Installing GCC
108 @c The following two commands start the copyright page.
110 @vskip 0pt plus 1filll
114 @c Part 4 Top node, Master Menu, and/or Table of Contents
117 @comment node-name, next, Previous, up
120 * Installing GCC:: This document describes the generic installation
121 procedure for GCC as well as detailing some target
122 specific installation instructions.
124 * Specific:: Host/target specific installation notes for GCC.
125 * Binaries:: Where to get pre-compiled binaries.
127 * Old:: Old installation documentation.
129 * GNU Free Documentation License:: How you can copy and share this manual.
130 * Concept Index:: This index has two entries.
138 @c Part 5 The Body of the Document
139 @c ***Installing GCC**********************************************************
141 @comment node-name, next, previous, up
142 @node Installing GCC, Binaries, , Top
146 @chapter Installing GCC
149 The latest version of this document is always available at
150 @uref{http://gcc.gnu.org/install/,,http://gcc.gnu.org/install/}.
152 This document describes the generic installation procedure for GCC as well
153 as detailing some target specific installation instructions.
155 GCC includes several components that previously were separate distributions
156 with their own installation instructions. This document supersedes all
157 package specific installation instructions.
159 @emph{Before} starting the build/install procedure please check the
161 @ref{Specific, host/target specific installation notes}.
164 @uref{specific.html,,host/target specific installation notes}.
166 We recommend you browse the entire generic installation instructions before
169 Lists of successful builds for released versions of GCC are
170 available at @uref{http://gcc.gnu.org/buildstat.html}.
171 These lists are updated as new information becomes available.
173 The installation procedure itself is broken into five steps.
178 * Downloading the source::
181 * Testing:: (optional)
188 @uref{prerequisites.html,,Prerequisites}
190 @uref{download.html,,Downloading the source}
192 @uref{configure.html,,Configuration}
194 @uref{build.html,,Building}
196 @uref{test.html,,Testing} (optional)
198 @uref{finalinstall.html,,Final install}
202 Please note that GCC does not support @samp{make uninstall} and probably
203 won't do so in the near future as this would open a can of worms. Instead,
204 we suggest that you install GCC into a directory of its own and simply
205 remove that directory when you do not need that specific version of GCC
206 any longer, and, if shared libraries are installed there as well, no
207 more binaries exist that use them.
210 There are also some @uref{old.html,,old installation instructions},
211 which are mostly obsolete but still contain some information which has
212 not yet been merged into the main part of this manual.
220 @uref{./index.html,,Return to the GCC Installation page}
226 @c ***Prerequisites**************************************************
228 @comment node-name, next, previous, up
229 @node Prerequisites, Downloading the source, , Installing GCC
231 @ifset prerequisiteshtml
233 @chapter Prerequisites
235 @cindex Prerequisites
237 GCC requires that various tools and packages be available for use in the
238 build procedure. Modifying GCC sources requires additional tools
241 @heading Tools/packages necessary for building GCC
243 @item ISO C90 compiler
244 Necessary to bootstrap GCC, although versions of GCC prior
245 to 3.4 also allow bootstrapping with a traditional (K&R) C compiler.
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 2.95 or later) because source code for language
250 frontends other than C might use GCC extensions.
254 In order to build the Ada compiler (GNAT) you must already have GNAT
255 installed because portions of the Ada frontend are written in Ada (with
256 GNAT extensions.) Refer to the Ada installation instructions for more
257 specific information.
259 @item A ``working'' POSIX compatible shell, or GNU bash
261 Necessary when running @command{configure} because some
262 @command{/bin/sh} shells have bugs and may crash when configuring the
263 target libraries. In other cases, @command{/bin/sh} or @command{ksh}
264 have disastrous corner-case performance problems. This
265 can cause target @command{configure} runs to literally take days to
266 complete in some cases.
268 So on some platforms @command{/bin/ksh} is sufficient, on others it
269 isn't. See the host/target specific instructions for your platform, or
270 use @command{bash} to be sure. Then set @env{CONFIG_SHELL} in your
271 environment to your ``good'' shell prior to running
272 @command{configure}/@command{make}.
274 @command{zsh} is not a fully compliant POSIX shell and will not
275 work when configuring GCC@.
277 @item A POSIX or SVR4 awk
279 Necessary for creating some of the generated source files for GCC@.
280 If in doubt, use a recent GNU awk version, as some of the older ones
281 are broken. GNU awk version 3.1.5 is known to work.
285 Necessary in some circumstances, optional in others. See the
286 host/target specific instructions for your platform for the exact
289 @item gzip version 1.2.4 (or later) or
290 @itemx bzip2 version 1.0.2 (or later)
292 Necessary to uncompress GCC @command{tar} files when source code is
293 obtained via FTP mirror sites.
295 @item GNU make version 3.80 (or later)
297 You must have GNU make installed to build GCC@.
299 @item GNU tar version 1.14 (or later)
301 Necessary (only on some platforms) to untar the source code. Many
302 systems' @command{tar} programs will also work, only try GNU
303 @command{tar} if you have problems.
305 @item GNU Multiple Precision Library (GMP) version 4.2 (or later)
307 Necessary to build GCC@. If you do not have it installed in your
308 library search path, you will have to configure with the
309 @option{--with-gmp} configure option. See also @option{--with-gmp-lib}
310 and @option{--with-gmp-include}. Alternatively, if a GMP source
311 distribution is found in a subdirectory of your GCC sources named
312 @file{gmp}, it will be built together with GCC@.
314 @item MPFR Library version 2.3.2 (or later)
316 Necessary to build GCC@. It can be downloaded from
317 @uref{http://www.mpfr.org/}. The @option{--with-mpfr} configure
318 option should be used if your MPFR Library is not installed in your
319 default library search path. See also @option{--with-mpfr-lib} and
320 @option{--with-mpfr-include}. Alternatively, if a MPFR source
321 distribution is found in a subdirectory of your GCC sources named
322 @file{mpfr}, it will be built together with GCC@.
324 @item Parma Polyhedra Library (PPL) version 0.10
326 Necessary to build GCC with the Graphite loop optimizations.
327 It can be downloaded from @uref{http://www.cs.unipr.it/ppl/Download/}.
329 The @option{--with-ppl} configure option should be used if PPL is not
330 installed in your default library search path.
332 @item CLooG-PPL version 0.15
334 Necessary to build GCC with the Graphite loop optimizations. It can
335 be downloaded from @uref{ftp://gcc.gnu.org/pub/gcc/infrastructure/}.
336 The code in @file{cloog-ppl-0.15.tar.gz} comes from a branch of CLooG
337 available from @uref{http://repo.or.cz/w/cloog-ppl.git}. CLooG-PPL
338 should be configured with @option{--with-ppl}.
340 The @option{--with-cloog} configure option should be used if CLooG is
341 not installed in your default library search path.
343 @item @command{jar}, or InfoZIP (@command{zip} and @command{unzip})
345 Necessary to build libgcj, the GCJ runtime.
347 @item MPC Library version 0.6.0 (or later)
349 Optional when building GCC@. Having this library will enable
350 additional optimizations on complex numbers. It can be downloaded
351 from @uref{http://www.multiprecision.org/mpc/}. The
352 @option{--with-mpc} configure option should be used if your MPC
353 Library is not installed in your default library search path. See
354 also @option{--with-mpc-lib} and @option{--with-mpc-include}.
355 Alternatively, if an MPC source distribution is found in a
356 subdirectory of your GCC sources named @file{mpc}, it will be built
361 @heading Tools/packages necessary for modifying GCC
363 @item autoconf version 2.59
364 @itemx GNU m4 version 1.4 (or later)
366 Necessary when modifying @file{configure.ac}, @file{aclocal.m4}, etc.@:
367 to regenerate @file{configure} and @file{config.in} files.
369 @item automake version 1.9.6
371 Necessary when modifying a @file{Makefile.am} file to regenerate its
372 associated @file{Makefile.in}.
374 Much of GCC does not use automake, so directly edit the @file{Makefile.in}
375 file. Specifically this applies to the @file{gcc}, @file{intl},
376 @file{libcpp}, @file{libiberty}, @file{libobjc} directories as well
377 as any of their subdirectories.
379 For directories that use automake, GCC requires the latest release in
380 the 1.9.x series, which is currently 1.9.6. When regenerating a directory
381 to a newer version, please update all the directories using an older 1.9.x
382 to the latest released version.
384 @item gettext version 0.14.5 (or later)
386 Needed to regenerate @file{gcc.pot}.
388 @item gperf version 2.7.2 (or later)
390 Necessary when modifying @command{gperf} input files, e.g.@:
391 @file{gcc/cp/cfns.gperf} to regenerate its associated header file, e.g.@:
392 @file{gcc/cp/cfns.h}.
398 Necessary to run the GCC testsuite; see the section on testing for details.
400 @item autogen version 5.5.4 (or later) and
401 @itemx guile version 1.4.1 (or later)
403 Necessary to regenerate @file{fixinc/fixincl.x} from
404 @file{fixinc/inclhack.def} and @file{fixinc/*.tpl}.
406 Necessary to run @samp{make check} for @file{fixinc}.
408 Necessary to regenerate the top level @file{Makefile.in} file from
409 @file{Makefile.tpl} and @file{Makefile.def}.
411 @item Flex version 2.5.4 (or later)
413 Necessary when modifying @file{*.l} files.
415 Necessary to build GCC during development because the generated output
416 files are not included in the SVN repository. They are included in
419 @item Texinfo version 4.7 (or later)
421 Necessary for running @command{makeinfo} when modifying @file{*.texi}
422 files to test your changes.
424 Necessary for running @command{make dvi} or @command{make pdf} to
425 create printable documentation in DVI or PDF format. Texinfo version
426 4.8 or later is required for @command{make pdf}.
428 Necessary to build GCC documentation during development because the
429 generated output files are not included in the SVN repository. They are
430 included in releases.
432 @item @TeX{} (any working version)
434 Necessary for running @command{texi2dvi} and @command{texi2pdf}, which
435 are used when running @command{make dvi} or @command{make pdf} to create
436 DVI or PDF files, respectively.
438 @item SVN (any version)
439 @itemx SSH (any version)
441 Necessary to access the SVN repository. Public releases and weekly
442 snapshots of the development sources are also available via FTP@.
444 @item Perl version 5.6.1 (or later)
446 Necessary when regenerating @file{Makefile} dependencies in libiberty.
447 Necessary when regenerating @file{libiberty/functions.texi}.
448 Necessary when generating manpages from Texinfo manuals.
449 Necessary when targetting Darwin, building libstdc++,
450 and not using @option{--disable-symvers}.
451 Used by various scripts to generate some files included in SVN (mainly
452 Unicode-related and rarely changing) from source tables.
454 @item GNU diffutils version 2.7 (or later)
456 Useful when submitting patches for the GCC source code.
458 @item patch version 2.5.4 (or later)
460 Necessary when applying patches, created with @command{diff}, to one's
466 If you wish to modify @file{.java} files in libjava, you will need to
467 configure with @option{--enable-java-maintainer-mode}, and you will need
468 to have executables named @command{ecj1} and @command{gjavah} in your path.
469 The @command{ecj1} executable should run the Eclipse Java compiler via
470 the GCC-specific entry point. You can download a suitable jar from
471 @uref{ftp://sourceware.org/pub/java/}, or by running the script
472 @command{contrib/download_ecj}.
474 @item antlr.jar version 2.7.1 (or later)
477 If you wish to build the @command{gjdoc} binary in libjava, you will
478 need to have an @file{antlr.jar} library available. The library is
479 searched in system locations but can be configured with
480 @option{--with-antlr-jar=} instead. When configuring with
481 @option{--enable-java-maintainer-mode}, you will need to have one of
482 the executables named @command{cantlr}, @command{runantlr} or
483 @command{antlr} in your path.
492 @uref{./index.html,,Return to the GCC Installation page}
496 @c ***Downloading the source**************************************************
498 @comment node-name, next, previous, up
499 @node Downloading the source, Configuration, Prerequisites, Installing GCC
503 @chapter Downloading GCC
505 @cindex Downloading GCC
506 @cindex Downloading the Source
508 GCC is distributed via @uref{http://gcc.gnu.org/svn.html,,SVN} and FTP
509 tarballs compressed with @command{gzip} or
510 @command{bzip2}. It is possible to download a full distribution or specific
513 Please refer to the @uref{http://gcc.gnu.org/releases.html,,releases web page}
514 for information on how to obtain GCC@.
516 The full distribution includes the C, C++, Objective-C, Fortran, Java,
517 and Ada (in the case of GCC 3.1 and later) compilers. The full
518 distribution also includes runtime libraries for C++, Objective-C,
519 Fortran, and Java. In GCC 3.0 and later versions, the GNU compiler
520 testsuites are also included in the full distribution.
522 If you choose to download specific components, you must download the core
523 GCC distribution plus any language specific distributions you wish to
524 use. The core distribution includes the C language front end as well as the
525 shared components. Each language has a tarball which includes the language
526 front end as well as the language runtime (when appropriate).
528 Unpack the core distribution as well as any language specific
529 distributions in the same directory.
531 If you also intend to build binutils (either to upgrade an existing
532 installation or for use in place of the corresponding tools of your
533 OS), unpack the binutils distribution either in the same directory or
534 a separate one. In the latter case, add symbolic links to any
535 components of the binutils you intend to build alongside the compiler
536 (@file{bfd}, @file{binutils}, @file{gas}, @file{gprof}, @file{ld},
537 @file{opcodes}, @dots{}) to the directory containing the GCC sources.
539 Likewise the GMP, MPFR and MPC libraries can be automatically built
540 together with GCC. Unpack the GMP, MPFR and/or MPC source
541 distributions in the directory containing the GCC sources and rename
542 their directories to @file{gmp}, @file{mpfr} and @file{mpc},
543 respectively (or use symbolic links with the same name).
550 @uref{./index.html,,Return to the GCC Installation page}
554 @c ***Configuration***********************************************************
556 @comment node-name, next, previous, up
557 @node Configuration, Building, Downloading the source, Installing GCC
561 @chapter Installing GCC: Configuration
563 @cindex Configuration
564 @cindex Installing GCC: Configuration
566 Like most GNU software, GCC must be configured before it can be built.
567 This document describes the recommended configuration procedure
568 for both native and cross targets.
570 We use @var{srcdir} to refer to the toplevel source directory for
571 GCC; we use @var{objdir} to refer to the toplevel build/object directory.
573 If you obtained the sources via SVN, @var{srcdir} must refer to the top
574 @file{gcc} directory, the one where the @file{MAINTAINERS} can be found,
575 and not its @file{gcc} subdirectory, otherwise the build will fail.
577 If either @var{srcdir} or @var{objdir} is located on an automounted NFS
578 file system, the shell's built-in @command{pwd} command will return
579 temporary pathnames. Using these can lead to various sorts of build
580 problems. To avoid this issue, set the @env{PWDCMD} environment
581 variable to an automounter-aware @command{pwd} command, e.g.,
582 @command{pawd} or @samp{amq -w}, during the configuration and build
585 First, we @strong{highly} recommend that GCC be built into a
586 separate directory than the sources which does @strong{not} reside
587 within the source tree. This is how we generally build GCC; building
588 where @var{srcdir} == @var{objdir} should still work, but doesn't
589 get extensive testing; building where @var{objdir} is a subdirectory
590 of @var{srcdir} is unsupported.
592 If you have previously built GCC in the same directory for a
593 different target machine, do @samp{make distclean} to delete all files
594 that might be invalid. One of the files this deletes is @file{Makefile};
595 if @samp{make distclean} complains that @file{Makefile} does not exist
596 or issues a message like ``don't know how to make distclean'' it probably
597 means that the directory is already suitably clean. However, with the
598 recommended method of building in a separate @var{objdir}, you should
599 simply use a different @var{objdir} for each target.
601 Second, when configuring a native system, either @command{cc} or
602 @command{gcc} must be in your path or you must set @env{CC} in
603 your environment before running configure. Otherwise the configuration
607 Note that the bootstrap compiler and the resulting GCC must be link
608 compatible, else the bootstrap will fail with linker errors about
609 incompatible object file formats. Several multilibed targets are
610 affected by this requirement, see
612 @ref{Specific, host/target specific installation notes}.
615 @uref{specific.html,,host/target specific installation notes}.
624 % @var{srcdir}/configure [@var{options}] [@var{target}]
627 @heading Distributor options
629 If you will be distributing binary versions of GCC, with modifications
630 to the source code, you should use the options described in this
631 section to make clear that your version contains modifications.
634 @item --with-pkgversion=@var{version}
635 Specify a string that identifies your package. You may wish
636 to include a build number or build date. This version string will be
637 included in the output of @command{gcc --version}. This suffix does
638 not replace the default version string, only the @samp{GCC} part.
640 The default value is @samp{GCC}.
642 @item --with-bugurl=@var{url}
643 Specify the URL that users should visit if they wish to report a bug.
644 You are of course welcome to forward bugs reported to you to the FSF,
645 if you determine that they are not bugs in your modifications.
647 The default value refers to the FSF's GCC bug tracker.
651 @heading Target specification
654 GCC has code to correctly determine the correct value for @var{target}
655 for nearly all native systems. Therefore, we highly recommend you not
656 provide a configure target when configuring a native compiler.
659 @var{target} must be specified as @option{--target=@var{target}}
660 when configuring a cross compiler; examples of valid targets would be
661 m68k-elf, sh-elf, etc.
664 Specifying just @var{target} instead of @option{--target=@var{target}}
665 implies that the host defaults to @var{target}.
669 @heading Options specification
671 Use @var{options} to override several configure time options for
672 GCC@. A list of supported @var{options} follows; @samp{configure
673 --help} may list other options, but those not listed below may not
674 work and should not normally be used.
676 Note that each @option{--enable} option has a corresponding
677 @option{--disable} option and that each @option{--with} option has a
678 corresponding @option{--without} option.
681 @item --prefix=@var{dirname}
682 Specify the toplevel installation
683 directory. This is the recommended way to install the tools into a directory
684 other than the default. The toplevel installation directory defaults to
687 We @strong{highly} recommend against @var{dirname} being the same or a
688 subdirectory of @var{objdir} or vice versa. If specifying a directory
689 beneath a user's home directory tree, some shells will not expand
690 @var{dirname} correctly if it contains the @samp{~} metacharacter; use
693 The following standard @command{autoconf} options are supported. Normally you
694 should not need to use these options.
696 @item --exec-prefix=@var{dirname}
697 Specify the toplevel installation directory for architecture-dependent
698 files. The default is @file{@var{prefix}}.
700 @item --bindir=@var{dirname}
701 Specify the installation directory for the executables called by users
702 (such as @command{gcc} and @command{g++}). The default is
703 @file{@var{exec-prefix}/bin}.
705 @item --libdir=@var{dirname}
706 Specify the installation directory for object code libraries and
707 internal data files of GCC@. The default is @file{@var{exec-prefix}/lib}.
709 @item --libexecdir=@var{dirname}
710 Specify the installation directory for internal executables of GCC@.
711 The default is @file{@var{exec-prefix}/libexec}.
713 @item --with-slibdir=@var{dirname}
714 Specify the installation directory for the shared libgcc library. The
715 default is @file{@var{libdir}}.
717 @item --infodir=@var{dirname}
718 Specify the installation directory for documentation in info format.
719 The default is @file{@var{prefix}/info}.
721 @item --datadir=@var{dirname}
722 Specify the installation directory for some architecture-independent
723 data files referenced by GCC@. The default is @file{@var{prefix}/share}.
725 @item --mandir=@var{dirname}
726 Specify the installation directory for manual pages. The default is
727 @file{@var{prefix}/man}. (Note that the manual pages are only extracts from
728 the full GCC manuals, which are provided in Texinfo format. The manpages
729 are derived by an automatic conversion process from parts of the full
732 @item --with-gxx-include-dir=@var{dirname}
734 the installation directory for G++ header files. The default depends
735 on other configuration options, and differs between cross and native
740 @item --program-prefix=@var{prefix}
741 GCC supports some transformations of the names of its programs when
742 installing them. This option prepends @var{prefix} to the names of
743 programs to install in @var{bindir} (see above). For example, specifying
744 @option{--program-prefix=foo-} would result in @samp{gcc}
745 being installed as @file{/usr/local/bin/foo-gcc}.
747 @item --program-suffix=@var{suffix}
748 Appends @var{suffix} to the names of programs to install in @var{bindir}
749 (see above). For example, specifying @option{--program-suffix=-3.1}
750 would result in @samp{gcc} being installed as
751 @file{/usr/local/bin/gcc-3.1}.
753 @item --program-transform-name=@var{pattern}
754 Applies the @samp{sed} script @var{pattern} to be applied to the names
755 of programs to install in @var{bindir} (see above). @var{pattern} has to
756 consist of one or more basic @samp{sed} editing commands, separated by
757 semicolons. For example, if you want the @samp{gcc} program name to be
758 transformed to the installed program @file{/usr/local/bin/myowngcc} and
759 the @samp{g++} program name to be transformed to
760 @file{/usr/local/bin/gspecial++} without changing other program names,
761 you could use the pattern
762 @option{--program-transform-name='s/^gcc$/myowngcc/; s/^g++$/gspecial++/'}
763 to achieve this effect.
765 All three options can be combined and used together, resulting in more
766 complex conversion patterns. As a basic rule, @var{prefix} (and
767 @var{suffix}) are prepended (appended) before further transformations
768 can happen with a special transformation script @var{pattern}.
770 As currently implemented, this option only takes effect for native
771 builds; cross compiler binaries' names are not transformed even when a
772 transformation is explicitly asked for by one of these options.
774 For native builds, some of the installed programs are also installed
775 with the target alias in front of their name, as in
776 @samp{i686-pc-linux-gnu-gcc}. All of the above transformations happen
777 before the target alias is prepended to the name---so, specifying
778 @option{--program-prefix=foo-} and @option{program-suffix=-3.1}, the
779 resulting binary would be installed as
780 @file{/usr/local/bin/i686-pc-linux-gnu-foo-gcc-3.1}.
782 As a last shortcoming, none of the installed Ada programs are
783 transformed yet, which will be fixed in some time.
785 @item --with-local-prefix=@var{dirname}
787 installation directory for local include files. The default is
788 @file{/usr/local}. Specify this option if you want the compiler to
789 search directory @file{@var{dirname}/include} for locally installed
790 header files @emph{instead} of @file{/usr/local/include}.
792 You should specify @option{--with-local-prefix} @strong{only} if your
793 site has a different convention (not @file{/usr/local}) for where to put
796 The default value for @option{--with-local-prefix} is @file{/usr/local}
797 regardless of the value of @option{--prefix}. Specifying
798 @option{--prefix} has no effect on which directory GCC searches for
799 local header files. This may seem counterintuitive, but actually it is
802 The purpose of @option{--prefix} is to specify where to @emph{install
803 GCC}. The local header files in @file{/usr/local/include}---if you put
804 any in that directory---are not part of GCC@. They are part of other
805 programs---perhaps many others. (GCC installs its own header files in
806 another directory which is based on the @option{--prefix} value.)
808 Both the local-prefix include directory and the GCC-prefix include
809 directory are part of GCC's ``system include'' directories. Although these
810 two directories are not fixed, they need to be searched in the proper
811 order for the correct processing of the include_next directive. The
812 local-prefix include directory is searched before the GCC-prefix
813 include directory. Another characteristic of system include directories
814 is that pedantic warnings are turned off for headers in these directories.
816 Some autoconf macros add @option{-I @var{directory}} options to the
817 compiler command line, to ensure that directories containing installed
818 packages' headers are searched. When @var{directory} is one of GCC's
819 system include directories, GCC will ignore the option so that system
820 directories continue to be processed in the correct order. This
821 may result in a search order different from what was specified but the
822 directory will still be searched.
824 GCC automatically searches for ordinary libraries using
825 @env{GCC_EXEC_PREFIX}. Thus, when the same installation prefix is
826 used for both GCC and packages, GCC will automatically search for
827 both headers and libraries. This provides a configuration that is
828 easy to use. GCC behaves in a manner similar to that when it is
829 installed as a system compiler in @file{/usr}.
831 Sites that need to install multiple versions of GCC may not want to
832 use the above simple configuration. It is possible to use the
833 @option{--program-prefix}, @option{--program-suffix} and
834 @option{--program-transform-name} options to install multiple versions
835 into a single directory, but it may be simpler to use different prefixes
836 and the @option{--with-local-prefix} option to specify the location of the
837 site-specific files for each version. It will then be necessary for
838 users to specify explicitly the location of local site libraries
839 (e.g., with @env{LIBRARY_PATH}).
841 The same value can be used for both @option{--with-local-prefix} and
842 @option{--prefix} provided it is not @file{/usr}. This can be used
843 to avoid the default search of @file{/usr/local/include}.
845 @strong{Do not} specify @file{/usr} as the @option{--with-local-prefix}!
846 The directory you use for @option{--with-local-prefix} @strong{must not}
847 contain any of the system's standard header files. If it did contain
848 them, certain programs would be miscompiled (including GNU Emacs, on
849 certain targets), because this would override and nullify the header
850 file corrections made by the @command{fixincludes} script.
852 Indications are that people who use this option use it based on mistaken
853 ideas of what it is for. People use it as if it specified where to
854 install part of GCC@. Perhaps they make this assumption because
855 installing GCC creates the directory.
857 @item --enable-shared[=@var{package}[,@dots{}]]
858 Build shared versions of libraries, if shared libraries are supported on
859 the target platform. Unlike GCC 2.95.x and earlier, shared libraries
860 are enabled by default on all platforms that support shared libraries.
862 If a list of packages is given as an argument, build shared libraries
863 only for the listed packages. For other packages, only static libraries
864 will be built. Package names currently recognized in the GCC tree are
865 @samp{libgcc} (also known as @samp{gcc}), @samp{libstdc++} (not
866 @samp{libstdc++-v3}), @samp{libffi}, @samp{zlib}, @samp{boehm-gc},
867 @samp{ada}, @samp{libada}, @samp{libjava} and @samp{libobjc}.
868 Note @samp{libiberty} does not support shared libraries at all.
870 Use @option{--disable-shared} to build only static libraries. Note that
871 @option{--disable-shared} does not accept a list of package names as
872 argument, only @option{--enable-shared} does.
874 @item @anchor{with-gnu-as}--with-gnu-as
875 Specify that the compiler should assume that the
876 assembler it finds is the GNU assembler. However, this does not modify
877 the rules to find an assembler and will result in confusion if the
878 assembler found is not actually the GNU assembler. (Confusion may also
879 result if the compiler finds the GNU assembler but has not been
880 configured with @option{--with-gnu-as}.) If you have more than one
881 assembler installed on your system, you may want to use this option in
882 connection with @option{--with-as=@var{pathname}} or
883 @option{--with-build-time-tools=@var{pathname}}.
885 The following systems are the only ones where it makes a difference
886 whether you use the GNU assembler. On any other system,
887 @option{--with-gnu-as} has no effect.
890 @item @samp{hppa1.0-@var{any}-@var{any}}
891 @item @samp{hppa1.1-@var{any}-@var{any}}
892 @item @samp{sparc-sun-solaris2.@var{any}}
893 @item @samp{sparc64-@var{any}-solaris2.@var{any}}
896 @item @anchor{with-as}--with-as=@var{pathname}
897 Specify that the compiler should use the assembler pointed to by
898 @var{pathname}, rather than the one found by the standard rules to find
899 an assembler, which are:
902 Unless GCC is being built with a cross compiler, check the
903 @file{@var{libexec}/gcc/@var{target}/@var{version}} directory.
904 @var{libexec} defaults to @file{@var{exec-prefix}/libexec};
905 @var{exec-prefix} defaults to @var{prefix}, which
906 defaults to @file{/usr/local} unless overridden by the
907 @option{--prefix=@var{pathname}} switch described above. @var{target}
908 is the target system triple, such as @samp{sparc-sun-solaris2.7}, and
909 @var{version} denotes the GCC version, such as 3.0.
912 If the target system is the same that you are building on, check
913 operating system specific directories (e.g.@: @file{/usr/ccs/bin} on
917 Check in the @env{PATH} for a tool whose name is prefixed by the
918 target system triple.
921 Check in the @env{PATH} for a tool whose name is not prefixed by the
922 target system triple, if the host and target system triple are
923 the same (in other words, we use a host tool if it can be used for
927 You may want to use @option{--with-as} if no assembler
928 is installed in the directories listed above, or if you have multiple
929 assemblers installed and want to choose one that is not found by the
932 @item @anchor{with-gnu-ld}--with-gnu-ld
933 Same as @uref{#with-gnu-as,,@option{--with-gnu-as}}
936 @item --with-ld=@var{pathname}
937 Same as @uref{#with-as,,@option{--with-as}}
941 Specify that stabs debugging
942 information should be used instead of whatever format the host normally
943 uses. Normally GCC uses the same debug format as the host system.
945 On MIPS based systems and on Alphas, you must specify whether you want
946 GCC to create the normal ECOFF debugging format, or to use BSD-style
947 stabs passed through the ECOFF symbol table. The normal ECOFF debug
948 format cannot fully handle languages other than C@. BSD stabs format can
949 handle other languages, but it only works with the GNU debugger GDB@.
951 Normally, GCC uses the ECOFF debugging format by default; if you
952 prefer BSD stabs, specify @option{--with-stabs} when you configure GCC@.
954 No matter which default you choose when you configure GCC, the user
955 can use the @option{-gcoff} and @option{-gstabs+} options to specify explicitly
956 the debug format for a particular compilation.
958 @option{--with-stabs} is meaningful on the ISC system on the 386, also, if
959 @option{--with-gas} is used. It selects use of stabs debugging
960 information embedded in COFF output. This kind of debugging information
961 supports C++ well; ordinary COFF debugging information does not.
963 @option{--with-stabs} is also meaningful on 386 systems running SVR4. It
964 selects use of stabs debugging information embedded in ELF output. The
965 C++ compiler currently (2.6.0) does not support the DWARF debugging
966 information normally used on 386 SVR4 platforms; stabs provide a
967 workable alternative. This requires gas and gdb, as the normal SVR4
968 tools can not generate or interpret stabs.
970 @item --disable-multilib
971 Specify that multiple target
972 libraries to support different target variants, calling
973 conventions, etc.@: should not be built. The default is to build a
974 predefined set of them.
976 Some targets provide finer-grained control over which multilibs are built
977 (e.g., @option{--disable-softfloat}):
983 fpu, 26bit, underscore, interwork, biendian, nofmult.
986 softfloat, m68881, m68000, m68020.
989 single-float, biendian, softfloat.
991 @item powerpc*-*-*, rs6000*-*-*
992 aix64, pthread, softfloat, powercpu, powerpccpu, powerpcos, biendian,
997 @item --with-multilib-list=@var{list}
998 @itemx --without-multilib-list
999 Specify what multilibs to build.
1000 Currently only implemented for sh*-*-*.
1002 @var{list} is a comma separated list of CPU names. These must be of the
1003 form @code{sh*} or @code{m*} (in which case they match the compiler option
1004 for that processor). The list should not contain any endian options -
1005 these are handled by @option{--with-endian}.
1007 If @var{list} is empty, then there will be no multilibs for extra
1008 processors. The multilib for the secondary endian remains enabled.
1010 As a special case, if an entry in the list starts with a @code{!}
1011 (exclamation point), then it is added to the list of excluded multilibs.
1012 Entries of this sort should be compatible with @samp{MULTILIB_EXCLUDES}
1013 (once the leading @code{!} has been stripped).
1015 If @option{--with-multilib-list} is not given, then a default set of
1016 multilibs is selected based on the value of @option{--target}. This is
1017 usually the complete set of libraries, but some targets imply a more
1020 Example 1: to configure a compiler for SH4A only, but supporting both
1021 endians, with little endian being the default:
1023 --with-cpu=sh4a --with-endian=little,big --with-multilib-list=
1026 Example 2: to configure a compiler for both SH4A and SH4AL-DSP, but with
1027 only little endian SH4AL:
1029 --with-cpu=sh4a --with-endian=little,big --with-multilib-list=sh4al,!mb/m4al
1032 @item --with-endian=@var{endians}
1033 Specify what endians to use.
1034 Currently only implemented for sh*-*-*.
1036 @var{endians} may be one of the following:
1039 Use big endian exclusively.
1041 Use little endian exclusively.
1043 Use big endian by default. Provide a multilib for little endian.
1045 Use little endian by default. Provide a multilib for big endian.
1048 @item --enable-threads
1049 Specify that the target
1050 supports threads. This affects the Objective-C compiler and runtime
1051 library, and exception handling for other languages like C++ and Java.
1052 On some systems, this is the default.
1054 In general, the best (and, in many cases, the only known) threading
1055 model available will be configured for use. Beware that on some
1056 systems, GCC has not been taught what threading models are generally
1057 available for the system. In this case, @option{--enable-threads} is an
1058 alias for @option{--enable-threads=single}.
1060 @item --disable-threads
1061 Specify that threading support should be disabled for the system.
1062 This is an alias for @option{--enable-threads=single}.
1064 @item --enable-threads=@var{lib}
1066 @var{lib} is the thread support library. This affects the Objective-C
1067 compiler and runtime library, and exception handling for other languages
1068 like C++ and Java. The possibilities for @var{lib} are:
1076 Ada tasking support. For non-Ada programs, this setting is equivalent
1077 to @samp{single}. When used in conjunction with the Ada run time, it
1078 causes GCC to use the same thread primitives as Ada uses. This option
1079 is necessary when using both Ada and the back end exception handling,
1080 which is the default for most Ada targets.
1082 Generic MACH thread support, known to work on NeXTSTEP@. (Please note
1083 that the file needed to support this configuration, @file{gthr-mach.h}, is
1084 missing and thus this setting will cause a known bootstrap failure.)
1086 This is an alias for @samp{single}.
1088 Generic POSIX/Unix98 thread support.
1090 Generic POSIX/Unix95 thread support.
1092 RTEMS thread support.
1094 Disable thread support, should work for all platforms.
1096 Sun Solaris 2 thread support.
1098 VxWorks thread support.
1100 Microsoft Win32 API thread support.
1102 Novell Kernel Services thread support.
1106 Specify that the target supports TLS (Thread Local Storage). Usually
1107 configure can correctly determine if TLS is supported. In cases where
1108 it guesses incorrectly, TLS can be explicitly enabled or disabled with
1109 @option{--enable-tls} or @option{--disable-tls}. This can happen if
1110 the assembler supports TLS but the C library does not, or if the
1111 assumptions made by the configure test are incorrect.
1114 Specify that the target does not support TLS.
1115 This is an alias for @option{--enable-tls=no}.
1117 @item --with-cpu=@var{cpu}
1118 @itemx --with-cpu-32=@var{cpu}
1119 @itemx --with-cpu-64=@var{cpu}
1120 Specify which cpu variant the compiler should generate code for by default.
1121 @var{cpu} will be used as the default value of the @option{-mcpu=} switch.
1122 This option is only supported on some targets, including ARM, i386, M68k,
1123 PowerPC, and SPARC@. The @option{--with-cpu-32} and
1124 @option{--with-cpu-64} options specify separate default CPUs for
1125 32-bit and 64-bit modes; these options are only supported for i386,
1128 @item --with-schedule=@var{cpu}
1129 @itemx --with-arch=@var{cpu}
1130 @itemx --with-arch-32=@var{cpu}
1131 @itemx --with-arch-64=@var{cpu}
1132 @itemx --with-tune=@var{cpu}
1133 @itemx --with-tune-32=@var{cpu}
1134 @itemx --with-tune-64=@var{cpu}
1135 @itemx --with-abi=@var{abi}
1136 @itemx --with-fpu=@var{type}
1137 @itemx --with-float=@var{type}
1138 These configure options provide default values for the @option{-mschedule=},
1139 @option{-march=}, @option{-mtune=}, @option{-mabi=}, and @option{-mfpu=}
1140 options and for @option{-mhard-float} or @option{-msoft-float}. As with
1141 @option{--with-cpu}, which switches will be accepted and acceptable values
1142 of the arguments depend on the target.
1144 @item --with-mode=@var{mode}
1145 Specify if the compiler should default to @option{-marm} or @option{-mthumb}.
1146 This option is only supported on ARM targets.
1148 @item --with-divide=@var{type}
1149 Specify how the compiler should generate code for checking for
1150 division by zero. This option is only supported on the MIPS target.
1151 The possibilities for @var{type} are:
1154 Division by zero checks use conditional traps (this is the default on
1155 systems that support conditional traps).
1157 Division by zero checks use the break instruction.
1160 @c If you make --with-llsc the default for additional targets,
1161 @c update the --with-llsc description in the MIPS section below.
1164 On MIPS targets, make @option{-mllsc} the default when no
1165 @option{-mno-lsc} option is passed. This is the default for
1166 Linux-based targets, as the kernel will emulate them if the ISA does
1169 @item --without-llsc
1170 On MIPS targets, make @option{-mno-llsc} the default when no
1171 @option{-mllsc} option is passed.
1173 @item --with-mips-plt
1174 On MIPS targets, make use of copy relocations and PLTs.
1175 These features are extensions to the traditional
1176 SVR4-based MIPS ABIs and require support from GNU binutils
1177 and the runtime C library.
1179 @item --enable-__cxa_atexit
1180 Define if you want to use __cxa_atexit, rather than atexit, to
1181 register C++ destructors for local statics and global objects.
1182 This is essential for fully standards-compliant handling of
1183 destructors, but requires __cxa_atexit in libc. This option is currently
1184 only available on systems with GNU libc. When enabled, this will cause
1185 @option{-fuse-cxa-atexit} to be passed by default.
1187 @item --enable-target-optspace
1189 libraries should be optimized for code space instead of code speed.
1190 This is the default for the m32r platform.
1193 Specify that a user visible @command{cpp} program should not be installed.
1195 @item --with-cpp-install-dir=@var{dirname}
1196 Specify that the user visible @command{cpp} program should be installed
1197 in @file{@var{prefix}/@var{dirname}/cpp}, in addition to @var{bindir}.
1199 @item --enable-initfini-array
1200 Force the use of sections @code{.init_array} and @code{.fini_array}
1201 (instead of @code{.init} and @code{.fini}) for constructors and
1202 destructors. Option @option{--disable-initfini-array} has the
1203 opposite effect. If neither option is specified, the configure script
1204 will try to guess whether the @code{.init_array} and
1205 @code{.fini_array} sections are supported and, if they are, use them.
1207 @item --enable-build-with-cxx
1208 Build GCC using a C++ compiler rather than a C compiler. This is an
1209 experimental option which may become the default in a later release.
1211 @item --enable-maintainer-mode
1212 The build rules that
1213 regenerate the GCC master message catalog @file{gcc.pot} are normally
1214 disabled. This is because it can only be rebuilt if the complete source
1215 tree is present. If you have changed the sources and want to rebuild the
1216 catalog, configuring with @option{--enable-maintainer-mode} will enable
1217 this. Note that you need a recent version of the @code{gettext} tools
1220 @item --disable-bootstrap
1221 For a native build, the default configuration is to perform
1222 a 3-stage bootstrap of the compiler when @samp{make} is invoked,
1223 testing that GCC can compile itself correctly. If you want to disable
1224 this process, you can configure with @option{--disable-bootstrap}.
1226 @item --enable-bootstrap
1227 In special cases, you may want to perform a 3-stage build
1228 even if the target and host triplets are different.
1229 This could happen when the host can run code compiled for
1230 the target (e.g.@: host is i686-linux, target is i486-linux).
1231 Starting from GCC 4.2, to do this you have to configure explicitly
1232 with @option{--enable-bootstrap}.
1234 @item --enable-generated-files-in-srcdir
1235 Neither the .c and .h files that are generated from Bison and flex nor the
1236 info manuals and man pages that are built from the .texi files are present
1237 in the SVN development tree. When building GCC from that development tree,
1238 or from one of our snapshots, those generated files are placed in your
1239 build directory, which allows for the source to be in a readonly
1242 If you configure with @option{--enable-generated-files-in-srcdir} then those
1243 generated files will go into the source directory. This is mainly intended
1244 for generating release or prerelease tarballs of the GCC sources, since it
1245 is not a requirement that the users of source releases to have flex, Bison,
1248 @item --enable-version-specific-runtime-libs
1250 that runtime libraries should be installed in the compiler specific
1251 subdirectory (@file{@var{libdir}/gcc}) rather than the usual places. In
1252 addition, @samp{libstdc++}'s include files will be installed into
1253 @file{@var{libdir}} unless you overruled it by using
1254 @option{--with-gxx-include-dir=@var{dirname}}. Using this option is
1255 particularly useful if you intend to use several versions of GCC in
1256 parallel. This is currently supported by @samp{libgfortran},
1257 @samp{libjava}, @samp{libmudflap}, @samp{libstdc++}, and @samp{libobjc}.
1259 @item --enable-languages=@var{lang1},@var{lang2},@dots{}
1260 Specify that only a particular subset of compilers and
1261 their runtime libraries should be built. For a list of valid values for
1262 @var{langN} you can issue the following command in the
1263 @file{gcc} directory of your GCC source tree:@*
1265 grep language= */config-lang.in
1267 Currently, you can use any of the following:
1268 @code{all}, @code{ada}, @code{c}, @code{c++}, @code{fortran}, @code{java},
1269 @code{objc}, @code{obj-c++}.
1270 Building the Ada compiler has special requirements, see below.
1271 If you do not pass this flag, or specify the option @code{all}, then all
1272 default languages available in the @file{gcc} sub-tree will be configured.
1273 Ada and Objective-C++ are not default languages; the rest are.
1274 Re-defining @code{LANGUAGES} when calling @samp{make} @strong{does not}
1275 work anymore, as those language sub-directories might not have been
1278 @item --enable-stage1-languages=@var{lang1},@var{lang2},@dots{}
1279 Specify that a particular subset of compilers and their runtime
1280 libraries should be built with the system C compiler during stage 1 of
1281 the bootstrap process, rather than only in later stages with the
1282 bootstrapped C compiler. The list of valid values is the same as for
1283 @option{--enable-languages}, and the option @code{all} will select all
1284 of the languages enabled by @option{--enable-languages}. This option is
1285 primarily useful for GCC development; for instance, when a development
1286 version of the compiler cannot bootstrap due to compiler bugs, or when
1287 one is debugging front ends other than the C front end. When this
1288 option is used, one can then build the target libraries for the
1289 specified languages with the stage-1 compiler by using @command{make
1290 stage1-bubble all-target}, or run the testsuite on the stage-1 compiler
1291 for the specified languages using @command{make stage1-start check-gcc}.
1293 @item --disable-libada
1294 Specify that the run-time libraries and tools used by GNAT should not
1295 be built. This can be useful for debugging, or for compatibility with
1296 previous Ada build procedures, when it was required to explicitly
1297 do a @samp{make -C gcc gnatlib_and_tools}.
1299 @item --disable-libssp
1300 Specify that the run-time libraries for stack smashing protection
1301 should not be built.
1303 @item --disable-libgomp
1304 Specify that the run-time libraries used by GOMP should not be built.
1307 Specify that the compiler should
1308 use DWARF 2 debugging information as the default.
1310 @item --enable-targets=all
1311 @itemx --enable-targets=@var{target_list}
1312 Some GCC targets, e.g.@: powerpc64-linux, build bi-arch compilers.
1313 These are compilers that are able to generate either 64-bit or 32-bit
1314 code. Typically, the corresponding 32-bit target, e.g.@:
1315 powerpc-linux for powerpc64-linux, only generates 32-bit code. This
1316 option enables the 32-bit target to be a bi-arch compiler, which is
1317 useful when you want a bi-arch compiler that defaults to 32-bit, and
1318 you are building a bi-arch or multi-arch binutils in a combined tree.
1319 Currently, this option only affects sparc-linux, powerpc-linux and
1322 @item --enable-secureplt
1323 This option enables @option{-msecure-plt} by default for powerpc-linux.
1325 @xref{RS/6000 and PowerPC Options,, RS/6000 and PowerPC Options, gcc,
1326 Using the GNU Compiler Collection (GCC)},
1329 See ``RS/6000 and PowerPC Options'' in the main manual
1333 This option enables @option{-mcld} by default for 32-bit x86 targets.
1335 @xref{i386 and x86-64 Options,, i386 and x86-64 Options, gcc,
1336 Using the GNU Compiler Collection (GCC)},
1339 See ``i386 and x86-64 Options'' in the main manual
1342 @item --enable-win32-registry
1343 @itemx --enable-win32-registry=@var{key}
1344 @itemx --disable-win32-registry
1345 The @option{--enable-win32-registry} option enables Microsoft Windows-hosted GCC
1346 to look up installations paths in the registry using the following key:
1349 @code{HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\@var{key}}
1352 @var{key} defaults to GCC version number, and can be overridden by the
1353 @option{--enable-win32-registry=@var{key}} option. Vendors and distributors
1354 who use custom installers are encouraged to provide a different key,
1355 perhaps one comprised of vendor name and GCC version number, to
1356 avoid conflict with existing installations. This feature is enabled
1357 by default, and can be disabled by @option{--disable-win32-registry}
1358 option. This option has no effect on the other hosts.
1361 Specify that the machine does not have a floating point unit. This
1362 option only applies to @samp{m68k-sun-sunos@var{n}}. On any other
1363 system, @option{--nfp} has no effect.
1365 @item --enable-werror
1366 @itemx --disable-werror
1367 @itemx --enable-werror=yes
1368 @itemx --enable-werror=no
1369 When you specify this option, it controls whether certain files in the
1370 compiler are built with @option{-Werror} in bootstrap stage2 and later.
1371 If you don't specify it, @option{-Werror} is turned on for the main
1372 development trunk. However it defaults to off for release branches and
1373 final releases. The specific files which get @option{-Werror} are
1374 controlled by the Makefiles.
1376 @item --enable-checking
1377 @itemx --enable-checking=@var{list}
1378 When you specify this option, the compiler is built to perform internal
1379 consistency checks of the requested complexity. This does not change the
1380 generated code, but adds error checking within the compiler. This will
1381 slow down the compiler and may only work properly if you are building
1382 the compiler with GCC@. This is @samp{yes} by default when building
1383 from SVN or snapshots, but @samp{release} for releases. The default
1384 for building the stage1 compiler is @samp{yes}. More control
1385 over the checks may be had by specifying @var{list}. The categories of
1386 checks available are @samp{yes} (most common checks
1387 @samp{assert,misc,tree,gc,rtlflag,runtime}), @samp{no} (no checks at
1388 all), @samp{all} (all but @samp{valgrind}), @samp{release} (cheapest
1389 checks @samp{assert,runtime}) or @samp{none} (same as @samp{no}).
1390 Individual checks can be enabled with these flags @samp{assert},
1391 @samp{df}, @samp{fold}, @samp{gc}, @samp{gcac} @samp{misc}, @samp{rtl},
1392 @samp{rtlflag}, @samp{runtime}, @samp{tree}, and @samp{valgrind}.
1394 The @samp{valgrind} check requires the external @command{valgrind}
1395 simulator, available from @uref{http://valgrind.org/}. The
1396 @samp{df}, @samp{rtl}, @samp{gcac} and @samp{valgrind} checks are very expensive.
1397 To disable all checking, @samp{--disable-checking} or
1398 @samp{--enable-checking=none} must be explicitly requested. Disabling
1399 assertions will make the compiler and runtime slightly faster but
1400 increase the risk of undetected internal errors causing wrong code to be
1403 @item --disable-stage1-checking
1404 @item --enable-stage1-checking
1405 @itemx --enable-stage1-checking=@var{list}
1406 If no @option{--enable-checking} option is specified the stage1
1407 compiler will be built with @samp{yes} checking enabled, otherwise
1408 the stage1 checking flags are the same as specified by
1409 @option{--enable-checking}. To build the stage1 compiler with
1410 different checking options use @option{--enable-stage1-checking}.
1411 The list of checking options is the same as for @option{--enable-checking}.
1412 If your system is too slow or too small to bootstrap a released compiler
1413 with checking for stage1 enabled, you can use @samp{--disable-stage1-checking}
1414 to disable checking for the stage1 compiler.
1416 @item --enable-coverage
1417 @itemx --enable-coverage=@var{level}
1418 With this option, the compiler is built to collect self coverage
1419 information, every time it is run. This is for internal development
1420 purposes, and only works when the compiler is being built with gcc. The
1421 @var{level} argument controls whether the compiler is built optimized or
1422 not, values are @samp{opt} and @samp{noopt}. For coverage analysis you
1423 want to disable optimization, for performance analysis you want to
1424 enable optimization. When coverage is enabled, the default level is
1425 without optimization.
1427 @item --enable-gather-detailed-mem-stats
1428 When this option is specified more detailed information on memory
1429 allocation is gathered. This information is printed when using
1430 @option{-fmem-report}.
1433 @itemx --with-gc=@var{choice}
1434 With this option you can specify the garbage collector implementation
1435 used during the compilation process. @var{choice} can be one of
1436 @samp{page} and @samp{zone}, where @samp{page} is the default.
1439 @itemx --disable-nls
1440 The @option{--enable-nls} option enables Native Language Support (NLS),
1441 which lets GCC output diagnostics in languages other than American
1442 English. Native Language Support is enabled by default if not doing a
1443 canadian cross build. The @option{--disable-nls} option disables NLS@.
1445 @item --with-included-gettext
1446 If NLS is enabled, the @option{--with-included-gettext} option causes the build
1447 procedure to prefer its copy of GNU @command{gettext}.
1449 @item --with-catgets
1450 If NLS is enabled, and if the host lacks @code{gettext} but has the
1451 inferior @code{catgets} interface, the GCC build procedure normally
1452 ignores @code{catgets} and instead uses GCC's copy of the GNU
1453 @code{gettext} library. The @option{--with-catgets} option causes the
1454 build procedure to use the host's @code{catgets} in this situation.
1456 @item --with-libiconv-prefix=@var{dir}
1457 Search for libiconv header files in @file{@var{dir}/include} and
1458 libiconv library files in @file{@var{dir}/lib}.
1460 @item --enable-obsolete
1461 Enable configuration for an obsoleted system. If you attempt to
1462 configure GCC for a system (build, host, or target) which has been
1463 obsoleted, and you do not specify this flag, configure will halt with an
1466 All support for systems which have been obsoleted in one release of GCC
1467 is removed entirely in the next major release, unless someone steps
1468 forward to maintain the port.
1470 @item --enable-decimal-float
1471 @itemx --enable-decimal-float=yes
1472 @itemx --enable-decimal-float=no
1473 @itemx --enable-decimal-float=bid
1474 @itemx --enable-decimal-float=dpd
1475 @itemx --disable-decimal-float
1476 Enable (or disable) support for the C decimal floating point extension
1477 that is in the IEEE 754-2008 standard. This is enabled by default only
1478 on PowerPC, i386, and x86_64 GNU/Linux systems. Other systems may also
1479 support it, but require the user to specifically enable it. You can
1480 optionally control which decimal floating point format is used (either
1481 @samp{bid} or @samp{dpd}). The @samp{bid} (binary integer decimal)
1482 format is default on i386 and x86_64 systems, and the @samp{dpd}
1483 (densely packed decimal) format is default on PowerPC systems.
1485 @item --enable-fixed-point
1486 @itemx --disable-fixed-point
1487 Enable (or disable) support for C fixed-point arithmetic.
1488 This option is enabled by default for some targets (such as MIPS) which
1489 have hardware-support for fixed-point operations. On other targets, you
1490 may enable this option manually.
1492 @item --with-long-double-128
1493 Specify if @code{long double} type should be 128-bit by default on selected
1494 GNU/Linux architectures. If using @code{--without-long-double-128},
1495 @code{long double} will be by default 64-bit, the same as @code{double} type.
1496 When neither of these configure options are used, the default will be
1497 128-bit @code{long double} when built against GNU C Library 2.4 and later,
1498 64-bit @code{long double} otherwise.
1500 @item --with-gmp=@var{pathname}
1501 @itemx --with-gmp-include=@var{pathname}
1502 @itemx --with-gmp-lib=@var{pathname}
1503 @itemx --with-mpfr=@var{pathname}
1504 @itemx --with-mpfr-include=@var{pathname}
1505 @itemx --with-mpfr-lib=@var{pathname}
1506 @itemx --with-mpc=@var{pathname}
1507 @itemx --with-mpc-include=@var{pathname}
1508 @itemx --with-mpc-lib=@var{pathname}
1509 If you do not have GMP (the GNU Multiple Precision library), the MPFR
1510 library and/or the MPC library installed in a standard location and
1511 you want to build GCC, you can explicitly specify the directory where
1512 they are installed (@samp{--with-gmp=@var{gmpinstalldir}},
1513 @samp{--with-mpfr=@var{mpfrinstalldir}},
1514 @samp{--with-mpc=@var{mpcinstalldir}}). The
1515 @option{--with-gmp=@var{gmpinstalldir}} option is shorthand for
1516 @option{--with-gmp-lib=@var{gmpinstalldir}/lib} and
1517 @option{--with-gmp-include=@var{gmpinstalldir}/include}. Likewise the
1518 @option{--with-mpfr=@var{mpfrinstalldir}} option is shorthand for
1519 @option{--with-mpfr-lib=@var{mpfrinstalldir}/lib} and
1520 @option{--with-mpfr-include=@var{mpfrinstalldir}/include}, also the
1521 @option{--with-mpc=@var{mpcinstalldir}} option is shorthand for
1522 @option{--with-mpc-lib=@var{mpcinstalldir}/lib} and
1523 @option{--with-mpc-include=@var{mpcinstalldir}/include}. If these
1524 shorthand assumptions are not correct, you can use the explicit
1525 include and lib options directly.
1527 @item --with-ppl=@var{pathname}
1528 @itemx --with-ppl-include=@var{pathname}
1529 @itemx --with-ppl-lib=@var{pathname}
1530 @itemx --with-cloog=@var{pathname}
1531 @itemx --with-cloog-include=@var{pathname}
1532 @itemx --with-cloog-lib=@var{pathname}
1533 If you do not have PPL (the Parma Polyhedra Library) and the CLooG
1534 libraries installed in a standard location and you want to build GCC,
1535 you can explicitly specify the directory where they are installed
1536 (@samp{--with-ppl=@var{pplinstalldir}},
1537 @samp{--with-cloog=@var{clooginstalldir}}). The
1538 @option{--with-ppl=@var{pplinstalldir}} option is shorthand for
1539 @option{--with-ppl-lib=@var{pplinstalldir}/lib} and
1540 @option{--with-ppl-include=@var{pplinstalldir}/include}. Likewise the
1541 @option{--with-cloog=@var{clooginstalldir}} option is shorthand for
1542 @option{--with-cloog-lib=@var{clooginstalldir}/lib} and
1543 @option{--with-cloog-include=@var{clooginstalldir}/include}. If these
1544 shorthand assumptions are not correct, you can use the explicit
1545 include and lib options directly.
1547 @item --with-host-libstdcxx=@var{linker-args}
1548 If you are linking with a static copy of PPL, you can use this option
1549 to specify how the linker should find the standard C++ library used
1550 internally by PPL. Typical values of @var{linker-args} might be
1551 @samp{-lstdc++} or @samp{-Wl,-Bstatic,-lstdc++,-Bdynamic -lm}. If you are
1552 linking with a shared copy of PPL, you probably do not need this
1553 option; shared library dependencies will cause the linker to search
1554 for the standard C++ library automatically.
1556 @item --with-stage1-ldflags=@var{flags}
1557 This option may be used to set linker flags to be used when linking
1558 stage 1 of GCC. These are also used when linking GCC if configured with
1559 @option{--disable-bootstrap}. By default no special flags are used.
1561 @item --with-stage1-libs=@var{libs}
1562 This option may be used to set libraries to be used when linking stage 1
1563 of GCC. These are also used when linking GCC if configured with
1564 @option{--disable-bootstrap}. The default is the argument to
1565 @option{--with-host-libstdcxx}, if specified.
1567 @item --with-boot-ldflags=@var{flags}
1568 This option may be used to set linker flags to be used when linking
1569 stage 2 and later when bootstrapping GCC. By default no special flags
1572 @item --with-boot-libs=@var{libs}
1573 This option may be used to set libraries to be used when linking stage 2
1574 and later when bootstrapping GCC. The default is the argument to
1575 @option{--with-host-libstdcxx}, if specified.
1577 @item --with-debug-prefix-map=@var{map}
1578 Convert source directory names using @option{-fdebug-prefix-map} when
1579 building runtime libraries. @samp{@var{map}} is a space-separated
1580 list of maps of the form @samp{@var{old}=@var{new}}.
1582 @item --enable-linker-build-id
1583 Tells GCC to pass @option{--build-id} option to the linker for all final
1584 links (links performed without the @option{-r} or @option{--relocatable}
1585 option), if the linker supports it. If you specify
1586 @option{--enable-linker-build-id}, but your linker does not
1587 support @option{--build-id} option, a warning is issued and the
1588 @option{--enable-linker-build-id} option is ignored. The default is off.
1592 @subheading Cross-Compiler-Specific Options
1593 The following options only apply to building cross compilers.
1595 @item --with-sysroot
1596 @itemx --with-sysroot=@var{dir}
1597 Tells GCC to consider @var{dir} as the root of a tree that contains a
1598 (subset of) the root filesystem of the target operating system.
1599 Target system headers, libraries and run-time object files will be
1600 searched in there. The specified directory is not copied into the
1601 install tree, unlike the options @option{--with-headers} and
1602 @option{--with-libs} that this option obsoletes. The default value,
1603 in case @option{--with-sysroot} is not given an argument, is
1604 @option{$@{gcc_tooldir@}/sys-root}. If the specified directory is a
1605 subdirectory of @option{$@{exec_prefix@}}, then it will be found relative to
1606 the GCC binaries if the installation tree is moved.
1608 @item --with-build-sysroot
1609 @itemx --with-build-sysroot=@var{dir}
1610 Tells GCC to consider @var{dir} as the system root (see
1611 @option{--with-sysroot}) while building target libraries, instead of
1612 the directory specified with @option{--with-sysroot}. This option is
1613 only useful when you are already using @option{--with-sysroot}. You
1614 can use @option{--with-build-sysroot} when you are configuring with
1615 @option{--prefix} set to a directory that is different from the one in
1616 which you are installing GCC and your target libraries.
1618 This option affects the system root for the compiler used to build
1619 target libraries (which runs on the build system); it does not affect
1620 the compiler which is used to build GCC itself.
1622 @item --with-headers
1623 @itemx --with-headers=@var{dir}
1624 Deprecated in favor of @option{--with-sysroot}.
1625 Specifies that target headers are available when building a cross compiler.
1626 The @var{dir} argument specifies a directory which has the target include
1627 files. These include files will be copied into the @file{gcc} install
1628 directory. @emph{This option with the @var{dir} argument is required} when
1629 building a cross compiler, if @file{@var{prefix}/@var{target}/sys-include}
1630 doesn't pre-exist. If @file{@var{prefix}/@var{target}/sys-include} does
1631 pre-exist, the @var{dir} argument may be omitted. @command{fixincludes}
1632 will be run on these files to make them compatible with GCC@.
1634 @item --without-headers
1635 Tells GCC not use any target headers from a libc when building a cross
1636 compiler. When crossing to GNU/Linux, you need the headers so GCC
1637 can build the exception handling for libgcc.
1640 @itemx --with-libs=``@var{dir1} @var{dir2} @dots{} @var{dirN}''
1641 Deprecated in favor of @option{--with-sysroot}.
1642 Specifies a list of directories which contain the target runtime
1643 libraries. These libraries will be copied into the @file{gcc} install
1644 directory. If the directory list is omitted, this option has no
1648 Specifies that @samp{newlib} is
1649 being used as the target C library. This causes @code{__eprintf} to be
1650 omitted from @file{libgcc.a} on the assumption that it will be provided by
1653 @item --with-build-time-tools=@var{dir}
1654 Specifies where to find the set of target tools (assembler, linker, etc.)
1655 that will be used while building GCC itself. This option can be useful
1656 if the directory layouts are different between the system you are building
1657 GCC on, and the system where you will deploy it.
1659 For example, on an @samp{ia64-hp-hpux} system, you may have the GNU
1660 assembler and linker in @file{/usr/bin}, and the native tools in a
1661 different path, and build a toolchain that expects to find the
1662 native tools in @file{/usr/bin}.
1664 When you use this option, you should ensure that @var{dir} includes
1665 @command{ar}, @command{as}, @command{ld}, @command{nm},
1666 @command{ranlib} and @command{strip} if necessary, and possibly
1667 @command{objdump}. Otherwise, GCC may use an inconsistent set of
1671 @subheading Java-Specific Options
1673 The following option applies to the build of the Java front end.
1676 @item --disable-libgcj
1677 Specify that the run-time libraries
1678 used by GCJ should not be built. This is useful in case you intend
1679 to use GCJ with some other run-time, or you're going to install it
1680 separately, or it just happens not to build on your particular
1681 machine. In general, if the Java front end is enabled, the GCJ
1682 libraries will be enabled too, unless they're known to not work on
1683 the target platform. If GCJ is enabled but @samp{libgcj} isn't built, you
1684 may need to port it; in this case, before modifying the top-level
1685 @file{configure.in} so that @samp{libgcj} is enabled by default on this platform,
1686 you may use @option{--enable-libgcj} to override the default.
1690 The following options apply to building @samp{libgcj}.
1692 @subsubheading General Options
1695 @item --enable-java-maintainer-mode
1696 By default the @samp{libjava} build will not attempt to compile the
1697 @file{.java} source files to @file{.class}. Instead, it will use the
1698 @file{.class} files from the source tree. If you use this option you
1699 must have executables named @command{ecj1} and @command{gjavah} in your path
1700 for use by the build. You must use this option if you intend to
1701 modify any @file{.java} files in @file{libjava}.
1703 @item --with-java-home=@var{dirname}
1704 This @samp{libjava} option overrides the default value of the
1705 @samp{java.home} system property. It is also used to set
1706 @samp{sun.boot.class.path} to @file{@var{dirname}/lib/rt.jar}. By
1707 default @samp{java.home} is set to @file{@var{prefix}} and
1708 @samp{sun.boot.class.path} to
1709 @file{@var{datadir}/java/libgcj-@var{version}.jar}.
1711 @item --with-ecj-jar=@var{filename}
1712 This option can be used to specify the location of an external jar
1713 file containing the Eclipse Java compiler. A specially modified
1714 version of this compiler is used by @command{gcj} to parse
1715 @file{.java} source files. If this option is given, the
1716 @samp{libjava} build will create and install an @file{ecj1} executable
1717 which uses this jar file at runtime.
1719 If this option is not given, but an @file{ecj.jar} file is found in
1720 the topmost source tree at configure time, then the @samp{libgcj}
1721 build will create and install @file{ecj1}, and will also install the
1722 discovered @file{ecj.jar} into a suitable place in the install tree.
1724 If @file{ecj1} is not installed, then the user will have to supply one
1725 on his path in order for @command{gcj} to properly parse @file{.java}
1726 source files. A suitable jar is available from
1727 @uref{ftp://sourceware.org/pub/java/}.
1729 @item --disable-getenv-properties
1730 Don't set system properties from @env{GCJ_PROPERTIES}.
1732 @item --enable-hash-synchronization
1733 Use a global hash table for monitor locks. Ordinarily,
1734 @samp{libgcj}'s @samp{configure} script automatically makes
1735 the correct choice for this option for your platform. Only use
1736 this if you know you need the library to be configured differently.
1738 @item --enable-interpreter
1739 Enable the Java interpreter. The interpreter is automatically
1740 enabled by default on all platforms that support it. This option
1741 is really only useful if you want to disable the interpreter
1742 (using @option{--disable-interpreter}).
1744 @item --disable-java-net
1745 Disable java.net. This disables the native part of java.net only,
1746 using non-functional stubs for native method implementations.
1748 @item --disable-jvmpi
1749 Disable JVMPI support.
1751 @item --disable-libgcj-bc
1752 Disable BC ABI compilation of certain parts of libgcj. By default,
1753 some portions of libgcj are compiled with @option{-findirect-dispatch}
1754 and @option{-fno-indirect-classes}, allowing them to be overridden at
1757 If @option{--disable-libgcj-bc} is specified, libgcj is built without
1758 these options. This allows the compile-time linker to resolve
1759 dependencies when statically linking to libgcj. However it makes it
1760 impossible to override the affected portions of libgcj at run-time.
1762 @item --enable-reduced-reflection
1763 Build most of libgcj with @option{-freduced-reflection}. This reduces
1764 the size of libgcj at the expense of not being able to do accurate
1765 reflection on the classes it contains. This option is safe if you
1766 know that code using libgcj will never use reflection on the standard
1767 runtime classes in libgcj (including using serialization, RMI or CORBA).
1770 Enable runtime eCos target support.
1772 @item --without-libffi
1773 Don't use @samp{libffi}. This will disable the interpreter and JNI
1774 support as well, as these require @samp{libffi} to work.
1776 @item --enable-libgcj-debug
1777 Enable runtime debugging code.
1779 @item --enable-libgcj-multifile
1780 If specified, causes all @file{.java} source files to be
1781 compiled into @file{.class} files in one invocation of
1782 @samp{gcj}. This can speed up build time, but is more
1783 resource-intensive. If this option is unspecified or
1784 disabled, @samp{gcj} is invoked once for each @file{.java}
1785 file to compile into a @file{.class} file.
1787 @item --with-libiconv-prefix=DIR
1788 Search for libiconv in @file{DIR/include} and @file{DIR/lib}.
1790 @item --enable-sjlj-exceptions
1791 Force use of the @code{setjmp}/@code{longjmp}-based scheme for exceptions.
1792 @samp{configure} ordinarily picks the correct value based on the platform.
1793 Only use this option if you are sure you need a different setting.
1795 @item --with-system-zlib
1796 Use installed @samp{zlib} rather than that included with GCC@.
1798 @item --with-win32-nlsapi=ansi, unicows or unicode
1799 Indicates how MinGW @samp{libgcj} translates between UNICODE
1800 characters and the Win32 API@.
1802 @item --enable-java-home
1803 If enabled, this creates a JPackage compatible SDK environment during install.
1804 Note that if --enable-java-home is used, --with-arch-directory=ARCH must also
1807 @item --with-arch-directory=ARCH
1808 Specifies the name to use for the @file{jre/lib/ARCH} directory in the SDK
1809 environment created when --enable-java-home is passed. Typical names for this
1810 directory include i386, amd64, ia64, etc.
1812 @item --with-os-directory=DIR
1813 Specifies the OS directory for the SDK include directory. This is set to auto
1814 detect, and is typically 'linux'.
1816 @item --with-origin-name=NAME
1817 Specifies the JPackage origin name. This defaults to the 'gcj' in
1820 @item --with-arch-suffix=SUFFIX
1821 Specifies the suffix for the sdk directory. Defaults to the empty string.
1822 Examples include '.x86_64' in 'java-1.5.0-gcj-1.5.0.0.x86_64'.
1824 @item --with-jvm-root-dir=DIR
1825 Specifies where to install the SDK. Default is $(prefix)/lib/jvm.
1827 @item --with-jvm-jar-dir=DIR
1828 Specifies where to install jars. Default is $(prefix)/lib/jvm-exports.
1830 @item --with-python-dir=DIR
1831 Specifies where to install the Python modules used for aot-compile. DIR should
1832 not include the prefix used in installation. For example, if the Python modules
1833 are to be installed in /usr/lib/python2.5/site-packages, then
1834 --with-python-dir=/lib/python2.5/site-packages should be passed. If this is
1835 not specified, then the Python modules are installed in $(prefix)/share/python.
1837 @item --enable-aot-compile-rpm
1838 Adds aot-compile-rpm to the list of installed scripts.
1842 Use the single-byte @code{char} and the Win32 A functions natively,
1843 translating to and from UNICODE when using these functions. If
1844 unspecified, this is the default.
1847 Use the @code{WCHAR} and Win32 W functions natively. Adds
1848 @code{-lunicows} to @file{libgcj.spec} to link with @samp{libunicows}.
1849 @file{unicows.dll} needs to be deployed on Microsoft Windows 9X machines
1850 running built executables. @file{libunicows.a}, an open-source
1851 import library around Microsoft's @code{unicows.dll}, is obtained from
1852 @uref{http://libunicows.sourceforge.net/}, which also gives details
1853 on getting @file{unicows.dll} from Microsoft.
1856 Use the @code{WCHAR} and Win32 W functions natively. Does @emph{not}
1857 add @code{-lunicows} to @file{libgcj.spec}. The built executables will
1858 only run on Microsoft Windows NT and above.
1862 @subsubheading AWT-Specific Options
1866 Use the X Window System.
1868 @item --enable-java-awt=PEER(S)
1869 Specifies the AWT peer library or libraries to build alongside
1870 @samp{libgcj}. If this option is unspecified or disabled, AWT
1871 will be non-functional. Current valid values are @option{gtk} and
1872 @option{xlib}. Multiple libraries should be separated by a
1873 comma (i.e.@: @option{--enable-java-awt=gtk,xlib}).
1875 @item --enable-gtk-cairo
1876 Build the cairo Graphics2D implementation on GTK@.
1878 @item --enable-java-gc=TYPE
1879 Choose garbage collector. Defaults to @option{boehm} if unspecified.
1881 @item --disable-gtktest
1882 Do not try to compile and run a test GTK+ program.
1884 @item --disable-glibtest
1885 Do not try to compile and run a test GLIB program.
1887 @item --with-libart-prefix=PFX
1888 Prefix where libart is installed (optional).
1890 @item --with-libart-exec-prefix=PFX
1891 Exec prefix where libart is installed (optional).
1893 @item --disable-libarttest
1894 Do not try to compile and run a test libart program.
1903 @uref{./index.html,,Return to the GCC Installation page}
1907 @c ***Building****************************************************************
1909 @comment node-name, next, previous, up
1910 @node Building, Testing, Configuration, Installing GCC
1916 @cindex Installing GCC: Building
1918 Now that GCC is configured, you are ready to build the compiler and
1921 Some commands executed when making the compiler may fail (return a
1922 nonzero status) and be ignored by @command{make}. These failures, which
1923 are often due to files that were not found, are expected, and can safely
1926 It is normal to have compiler warnings when compiling certain files.
1927 Unless you are a GCC developer, you can generally ignore these warnings
1928 unless they cause compilation to fail. Developers should attempt to fix
1929 any warnings encountered, however they can temporarily continue past
1930 warnings-as-errors by specifying the configure flag
1931 @option{--disable-werror}.
1933 On certain old systems, defining certain environment variables such as
1934 @env{CC} can interfere with the functioning of @command{make}.
1936 If you encounter seemingly strange errors when trying to build the
1937 compiler in a directory other than the source directory, it could be
1938 because you have previously configured the compiler in the source
1939 directory. Make sure you have done all the necessary preparations.
1941 If you build GCC on a BSD system using a directory stored in an old System
1942 V file system, problems may occur in running @command{fixincludes} if the
1943 System V file system doesn't support symbolic links. These problems
1944 result in a failure to fix the declaration of @code{size_t} in
1945 @file{sys/types.h}. If you find that @code{size_t} is a signed type and
1946 that type mismatches occur, this could be the cause.
1948 The solution is not to use such a directory for building GCC@.
1950 Similarly, when building from SVN or snapshots, or if you modify
1951 @file{*.l} files, you need the Flex lexical analyzer generator
1952 installed. If you do not modify @file{*.l} files, releases contain
1953 the Flex-generated files and you do not need Flex installed to build
1954 them. There is still one Flex-based lexical analyzer (part of the
1955 build machinery, not of GCC itself) that is used even if you only
1956 build the C front end.
1958 When building from SVN or snapshots, or if you modify Texinfo
1959 documentation, you need version 4.7 or later of Texinfo installed if you
1960 want Info documentation to be regenerated. Releases contain Info
1961 documentation pre-built for the unmodified documentation in the release.
1963 @section Building a native compiler
1965 For a native build, the default configuration is to perform
1966 a 3-stage bootstrap of the compiler when @samp{make} is invoked.
1967 This will build the entire GCC system and ensure that it compiles
1968 itself correctly. It can be disabled with the @option{--disable-bootstrap}
1969 parameter to @samp{configure}, but bootstrapping is suggested because
1970 the compiler will be tested more completely and could also have
1973 The bootstrapping process will complete the following steps:
1977 Build tools necessary to build the compiler.
1980 Perform a 3-stage bootstrap of the compiler. This includes building
1981 three times the target tools for use by the compiler such as binutils
1982 (bfd, binutils, gas, gprof, ld, and opcodes) if they have been
1983 individually linked or moved into the top level GCC source tree before
1987 Perform a comparison test of the stage2 and stage3 compilers.
1990 Build runtime libraries using the stage3 compiler from the previous step.
1994 If you are short on disk space you might consider @samp{make
1995 bootstrap-lean} instead. The sequence of compilation is the
1996 same described above, but object files from the stage1 and
1997 stage2 of the 3-stage bootstrap of the compiler are deleted as
1998 soon as they are no longer needed.
2000 If you wish to use non-default GCC flags when compiling the stage2
2001 and stage3 compilers, set @code{BOOT_CFLAGS} on the command line when
2002 doing @samp{make}. For example, if you want to save additional space
2003 during the bootstrap and in the final installation as well, you can
2004 build the compiler binaries without debugging information as in the
2005 following example. This will save roughly 40% of disk space both for
2006 the bootstrap and the final installation. (Libraries will still contain
2007 debugging information.)
2010 make BOOT_CFLAGS='-O' bootstrap
2013 You can place non-default optimization flags into @code{BOOT_CFLAGS}; they
2014 are less well tested here than the default of @samp{-g -O2}, but should
2015 still work. In a few cases, you may find that you need to specify special
2016 flags such as @option{-msoft-float} here to complete the bootstrap; or,
2017 if the native compiler miscompiles the stage1 compiler, you may need
2018 to work around this, by choosing @code{BOOT_CFLAGS} to avoid the parts
2019 of the stage1 compiler that were miscompiled, or by using @samp{make
2020 bootstrap4} to increase the number of stages of bootstrap.
2022 @code{BOOT_CFLAGS} does not apply to bootstrapped target libraries.
2023 Since these are always compiled with the compiler currently being
2024 bootstrapped, you can use @code{CFLAGS_FOR_TARGET} to modify their
2025 compilation flags, as for non-bootstrapped target libraries.
2026 Again, if the native compiler miscompiles the stage1 compiler, you may
2027 need to work around this by avoiding non-working parts of the stage1
2028 compiler. Use @code{STAGE1_TFLAGS} to this end.
2030 If you used the flag @option{--enable-languages=@dots{}} to restrict
2031 the compilers to be built, only those you've actually enabled will be
2032 built. This will of course only build those runtime libraries, for
2033 which the particular compiler has been built. Please note,
2034 that re-defining @env{LANGUAGES} when calling @samp{make}
2035 @strong{does not} work anymore!
2037 If the comparison of stage2 and stage3 fails, this normally indicates
2038 that the stage2 compiler has compiled GCC incorrectly, and is therefore
2039 a potentially serious bug which you should investigate and report. (On
2040 a few systems, meaningful comparison of object files is impossible; they
2041 always appear ``different''. If you encounter this problem, you will
2042 need to disable comparison in the @file{Makefile}.)
2044 If you do not want to bootstrap your compiler, you can configure with
2045 @option{--disable-bootstrap}. In particular cases, you may want to
2046 bootstrap your compiler even if the target system is not the same as
2047 the one you are building on: for example, you could build a
2048 @code{powerpc-unknown-linux-gnu} toolchain on a
2049 @code{powerpc64-unknown-linux-gnu} host. In this case, pass
2050 @option{--enable-bootstrap} to the configure script.
2052 @code{BUILD_CONFIG} can be used to bring in additional customization to
2053 the build. It can be set to a whitespace-separated list of names. For
2054 each such @code{NAME}, top-level @file{config/@code{NAME}.mk} will be
2055 included by the top-level @file{Makefile}, bringing in any settings it
2056 contains. Some examples are:
2059 @item @samp{bootstrap-O1}
2060 Removes any @option{-O}-started option from @code{BOOT_CFLAGS}, and adds
2061 @option{-O1} to it. @samp{BUILD_CONFIG=bootstrap-O1} is equivalent to
2062 @samp{BOOT_CFLAGS='-g -O1'}.
2064 @item @samp{bootstrap-O3}
2065 Analogous to @code{bootstrap-O1}.
2067 @item @samp{bootstrap-debug}
2068 Builds stage2 without debug information, and uses
2069 @file{contrib/compare-debug} to compare object files.
2073 @section Building a cross compiler
2075 When building a cross compiler, it is not generally possible to do a
2076 3-stage bootstrap of the compiler. This makes for an interesting problem
2077 as parts of GCC can only be built with GCC@.
2079 To build a cross compiler, we first recommend building and installing a
2080 native compiler. You can then use the native GCC compiler to build the
2081 cross compiler. The installed native compiler needs to be GCC version
2084 If the cross compiler is to be built with support for the Java
2085 programming language and the ability to compile .java source files is
2086 desired, the installed native compiler used to build the cross
2087 compiler needs to be the same GCC version as the cross compiler. In
2088 addition the cross compiler needs to be configured with
2089 @option{--with-ecj-jar=@dots{}}.
2091 Assuming you have already installed a native copy of GCC and configured
2092 your cross compiler, issue the command @command{make}, which performs the
2097 Build host tools necessary to build the compiler.
2100 Build target tools for use by the compiler such as binutils (bfd,
2101 binutils, gas, gprof, ld, and opcodes)
2102 if they have been individually linked or moved into the top level GCC source
2103 tree before configuring.
2106 Build the compiler (single stage only).
2109 Build runtime libraries using the compiler from the previous step.
2112 Note that if an error occurs in any step the make process will exit.
2114 If you are not building GNU binutils in the same source tree as GCC,
2115 you will need a cross-assembler and cross-linker installed before
2116 configuring GCC@. Put them in the directory
2117 @file{@var{prefix}/@var{target}/bin}. Here is a table of the tools
2118 you should put in this directory:
2122 This should be the cross-assembler.
2125 This should be the cross-linker.
2128 This should be the cross-archiver: a program which can manipulate
2129 archive files (linker libraries) in the target machine's format.
2132 This should be a program to construct a symbol table in an archive file.
2135 The installation of GCC will find these programs in that directory,
2136 and copy or link them to the proper place to for the cross-compiler to
2137 find them when run later.
2139 The easiest way to provide these files is to build the Binutils package.
2140 Configure it with the same @option{--host} and @option{--target}
2141 options that you use for configuring GCC, then build and install
2142 them. They install their executables automatically into the proper
2143 directory. Alas, they do not support all the targets that GCC
2146 If you are not building a C library in the same source tree as GCC,
2147 you should also provide the target libraries and headers before
2148 configuring GCC, specifying the directories with
2149 @option{--with-sysroot} or @option{--with-headers} and
2150 @option{--with-libs}. Many targets also require ``start files'' such
2151 as @file{crt0.o} and
2152 @file{crtn.o} which are linked into each executable. There may be several
2153 alternatives for @file{crt0.o}, for use with profiling or other
2154 compilation options. Check your target's definition of
2155 @code{STARTFILE_SPEC} to find out what start files it uses.
2157 @section Building in parallel
2159 GNU Make 3.79 and above, which is necessary to build GCC, support
2160 building in parallel. To activate this, you can use @samp{make -j 2}
2161 instead of @samp{make}. You can also specify a bigger number, and
2162 in most cases using a value greater than the number of processors in
2163 your machine will result in fewer and shorter I/O latency hits, thus
2164 improving overall throughput; this is especially true for slow drives
2165 and network filesystems.
2167 @section Building the Ada compiler
2169 In order to build GNAT, the Ada compiler, you need a working GNAT
2170 compiler (GCC version 3.4 or later).
2171 This includes GNAT tools such as @command{gnatmake} and
2172 @command{gnatlink}, since the Ada front end is written in Ada and
2173 uses some GNAT-specific extensions.
2175 In order to build a cross compiler, it is suggested to install
2176 the new compiler as native first, and then use it to build the cross
2179 @command{configure} does not test whether the GNAT installation works
2180 and has a sufficiently recent version; if too old a GNAT version is
2181 installed, the build will fail unless @option{--enable-languages} is
2182 used to disable building the Ada front end.
2184 @env{ADA_INCLUDE_PATH} and @env{ADA_OBJECT_PATH} environment variables
2185 must not be set when building the Ada compiler, the Ada tools, or the
2186 Ada runtime libraries. You can check that your build environment is clean
2187 by verifying that @samp{gnatls -v} lists only one explicit path in each
2190 @section Building with profile feedback
2192 It is possible to use profile feedback to optimize the compiler itself. This
2193 should result in a faster compiler binary. Experiments done on x86 using gcc
2194 3.3 showed approximately 7 percent speedup on compiling C programs. To
2195 bootstrap the compiler with profile feedback, use @code{make profiledbootstrap}.
2197 When @samp{make profiledbootstrap} is run, it will first build a @code{stage1}
2198 compiler. This compiler is used to build a @code{stageprofile} compiler
2199 instrumented to collect execution counts of instruction and branch
2200 probabilities. Then runtime libraries are compiled with profile collected.
2201 Finally a @code{stagefeedback} compiler is built using the information collected.
2203 Unlike standard bootstrap, several additional restrictions apply. The
2204 compiler used to build @code{stage1} needs to support a 64-bit integral type.
2205 It is recommended to only use GCC for this. Also parallel make is currently
2206 not supported since collisions in profile collecting may occur.
2213 @uref{./index.html,,Return to the GCC Installation page}
2217 @c ***Testing*****************************************************************
2219 @comment node-name, next, previous, up
2220 @node Testing, Final install, Building, Installing GCC
2224 @chapter Installing GCC: Testing
2227 @cindex Installing GCC: Testing
2230 Before you install GCC, we encourage you to run the testsuites and to
2231 compare your results with results from a similar configuration that have
2232 been submitted to the
2233 @uref{http://gcc.gnu.org/ml/gcc-testresults/,,gcc-testresults mailing list}.
2234 Some of these archived results are linked from the build status lists
2235 at @uref{http://gcc.gnu.org/buildstat.html}, although not everyone who
2236 reports a successful build runs the testsuites and submits the results.
2237 This step is optional and may require you to download additional software,
2238 but it can give you confidence in your new GCC installation or point out
2239 problems before you install and start using your new GCC@.
2241 First, you must have @uref{download.html,,downloaded the testsuites}.
2242 These are part of the full distribution, but if you downloaded the
2243 ``core'' compiler plus any front ends, you must download the testsuites
2246 Second, you must have the testing tools installed. This includes
2247 @uref{http://www.gnu.org/software/dejagnu/,,DejaGnu}, Tcl, and Expect;
2248 the DejaGnu site has links to these.
2250 If the directories where @command{runtest} and @command{expect} were
2251 installed are not in the @env{PATH}, you may need to set the following
2252 environment variables appropriately, as in the following example (which
2253 assumes that DejaGnu has been installed under @file{/usr/local}):
2256 TCL_LIBRARY = /usr/local/share/tcl8.0
2257 DEJAGNULIBS = /usr/local/share/dejagnu
2260 (On systems such as Cygwin, these paths are required to be actual
2261 paths, not mounts or links; presumably this is due to some lack of
2262 portability in the DejaGnu code.)
2265 Finally, you can run the testsuite (which may take a long time):
2267 cd @var{objdir}; make -k check
2270 This will test various components of GCC, such as compiler
2271 front ends and runtime libraries. While running the testsuite, DejaGnu
2272 might emit some harmless messages resembling
2273 @samp{WARNING: Couldn't find the global config file.} or
2274 @samp{WARNING: Couldn't find tool init file} that can be ignored.
2276 If you are testing a cross-compiler, you may want to run the testsuite
2277 on a simulator as described at @uref{http://gcc.gnu.org/simtest-howto.html}.
2279 @section How can you run the testsuite on selected tests?
2281 In order to run sets of tests selectively, there are targets
2282 @samp{make check-gcc} and @samp{make check-g++}
2283 in the @file{gcc} subdirectory of the object directory. You can also
2284 just run @samp{make check} in a subdirectory of the object directory.
2287 A more selective way to just run all @command{gcc} execute tests in the
2291 make check-gcc RUNTESTFLAGS="execute.exp @var{other-options}"
2294 Likewise, in order to run only the @command{g++} ``old-deja'' tests in
2295 the testsuite with filenames matching @samp{9805*}, you would use
2298 make check-g++ RUNTESTFLAGS="old-deja.exp=9805* @var{other-options}"
2301 The @file{*.exp} files are located in the testsuite directories of the GCC
2302 source, the most important ones being @file{compile.exp},
2303 @file{execute.exp}, @file{dg.exp} and @file{old-deja.exp}.
2304 To get a list of the possible @file{*.exp} files, pipe the
2305 output of @samp{make check} into a file and look at the
2306 @samp{Running @dots{} .exp} lines.
2308 @section Passing options and running multiple testsuites
2310 You can pass multiple options to the testsuite using the
2311 @samp{--target_board} option of DejaGNU, either passed as part of
2312 @samp{RUNTESTFLAGS}, or directly to @command{runtest} if you prefer to
2313 work outside the makefiles. For example,
2316 make check-g++ RUNTESTFLAGS="--target_board=unix/-O3/-fmerge-constants"
2319 will run the standard @command{g++} testsuites (``unix'' is the target name
2320 for a standard native testsuite situation), passing
2321 @samp{-O3 -fmerge-constants} to the compiler on every test, i.e.,
2322 slashes separate options.
2324 You can run the testsuites multiple times using combinations of options
2325 with a syntax similar to the brace expansion of popular shells:
2328 @dots{}"--target_board=arm-sim\@{-mhard-float,-msoft-float\@}\@{-O1,-O2,-O3,\@}"
2331 (Note the empty option caused by the trailing comma in the final group.)
2332 The following will run each testsuite eight times using the @samp{arm-sim}
2333 target, as if you had specified all possible combinations yourself:
2336 --target_board=arm-sim/-mhard-float/-O1
2337 --target_board=arm-sim/-mhard-float/-O2
2338 --target_board=arm-sim/-mhard-float/-O3
2339 --target_board=arm-sim/-mhard-float
2340 --target_board=arm-sim/-msoft-float/-O1
2341 --target_board=arm-sim/-msoft-float/-O2
2342 --target_board=arm-sim/-msoft-float/-O3
2343 --target_board=arm-sim/-msoft-float
2346 They can be combined as many times as you wish, in arbitrary ways. This
2350 @dots{}"--target_board=unix/-Wextra\@{-O3,-fno-strength\@}\@{-fomit-frame,\@}"
2353 will generate four combinations, all involving @samp{-Wextra}.
2355 The disadvantage to this method is that the testsuites are run in serial,
2356 which is a waste on multiprocessor systems. For users with GNU Make and
2357 a shell which performs brace expansion, you can run the testsuites in
2358 parallel by having the shell perform the combinations and @command{make}
2359 do the parallel runs. Instead of using @samp{--target_board}, use a
2360 special makefile target:
2363 make -j@var{N} check-@var{testsuite}//@var{test-target}/@var{option1}/@var{option2}/@dots{}
2369 make -j3 check-gcc//sh-hms-sim/@{-m1,-m2,-m3,-m3e,-m4@}/@{,-nofpu@}
2372 will run three concurrent ``make-gcc'' testsuites, eventually testing all
2373 ten combinations as described above. Note that this is currently only
2374 supported in the @file{gcc} subdirectory. (To see how this works, try
2375 typing @command{echo} before the example given here.)
2378 @section Additional testing for Java Class Libraries
2380 The Java runtime tests can be executed via @samp{make check}
2381 in the @file{@var{target}/libjava/testsuite} directory in
2384 The @uref{http://sourceware.org/mauve/,,Mauve Project} provides
2385 a suite of tests for the Java Class Libraries. This suite can be run
2386 as part of libgcj testing by placing the Mauve tree within the libjava
2387 testsuite at @file{libjava/testsuite/libjava.mauve/mauve}, or by
2388 specifying the location of that tree when invoking @samp{make}, as in
2389 @samp{make MAUVEDIR=~/mauve check}.
2391 @section How to interpret test results
2393 The result of running the testsuite are various @file{*.sum} and @file{*.log}
2394 files in the testsuite subdirectories. The @file{*.log} files contain a
2395 detailed log of the compiler invocations and the corresponding
2396 results, the @file{*.sum} files summarize the results. These summaries
2397 contain status codes for all tests:
2401 PASS: the test passed as expected
2403 XPASS: the test unexpectedly passed
2405 FAIL: the test unexpectedly failed
2407 XFAIL: the test failed as expected
2409 UNSUPPORTED: the test is not supported on this platform
2411 ERROR: the testsuite detected an error
2413 WARNING: the testsuite detected a possible problem
2416 It is normal for some tests to report unexpected failures. At the
2417 current time the testing harness does not allow fine grained control
2418 over whether or not a test is expected to fail. This problem should
2419 be fixed in future releases.
2422 @section Submitting test results
2424 If you want to report the results to the GCC project, use the
2425 @file{contrib/test_summary} shell script. Start it in the @var{objdir} with
2428 @var{srcdir}/contrib/test_summary -p your_commentary.txt \
2429 -m gcc-testresults@@gcc.gnu.org |sh
2432 This script uses the @command{Mail} program to send the results, so
2433 make sure it is in your @env{PATH}. The file @file{your_commentary.txt} is
2434 prepended to the testsuite summary and should contain any special
2435 remarks you have on your results or your build environment. Please
2436 do not edit the testsuite result block or the subject line, as these
2437 messages may be automatically processed.
2444 @uref{./index.html,,Return to the GCC Installation page}
2448 @c ***Final install***********************************************************
2450 @comment node-name, next, previous, up
2451 @node Final install, , Testing, Installing GCC
2453 @ifset finalinstallhtml
2455 @chapter Installing GCC: Final installation
2458 Now that GCC has been built (and optionally tested), you can install it with
2460 cd @var{objdir}; make install
2463 We strongly recommend to install into a target directory where there is
2464 no previous version of GCC present. Also, the GNAT runtime should not
2465 be stripped, as this would break certain features of the debugger that
2466 depend on this debugging information (catching Ada exceptions for
2469 That step completes the installation of GCC; user level binaries can
2470 be found in @file{@var{prefix}/bin} where @var{prefix} is the value
2471 you specified with the @option{--prefix} to configure (or
2472 @file{/usr/local} by default). (If you specified @option{--bindir},
2473 that directory will be used instead; otherwise, if you specified
2474 @option{--exec-prefix}, @file{@var{exec-prefix}/bin} will be used.)
2475 Headers for the C++ and Java libraries are installed in
2476 @file{@var{prefix}/include}; libraries in @file{@var{libdir}}
2477 (normally @file{@var{prefix}/lib}); internal parts of the compiler in
2478 @file{@var{libdir}/gcc} and @file{@var{libexecdir}/gcc}; documentation
2479 in info format in @file{@var{infodir}} (normally
2480 @file{@var{prefix}/info}).
2482 When installing cross-compilers, GCC's executables
2483 are not only installed into @file{@var{bindir}}, that
2484 is, @file{@var{exec-prefix}/bin}, but additionally into
2485 @file{@var{exec-prefix}/@var{target-alias}/bin}, if that directory
2486 exists. Typically, such @dfn{tooldirs} hold target-specific
2487 binutils, including assembler and linker.
2489 Installation into a temporary staging area or into a @command{chroot}
2490 jail can be achieved with the command
2493 make DESTDIR=@var{path-to-rootdir} install
2496 @noindent where @var{path-to-rootdir} is the absolute path of
2497 a directory relative to which all installation paths will be
2498 interpreted. Note that the directory specified by @code{DESTDIR}
2499 need not exist yet; it will be created if necessary.
2501 There is a subtle point with tooldirs and @code{DESTDIR}:
2502 If you relocate a cross-compiler installation with
2503 e.g.@: @samp{DESTDIR=@var{rootdir}}, then the directory
2504 @file{@var{rootdir}/@var{exec-prefix}/@var{target-alias}/bin} will
2505 be filled with duplicated GCC executables only if it already exists,
2506 it will not be created otherwise. This is regarded as a feature,
2507 not as a bug, because it gives slightly more control to the packagers
2508 using the @code{DESTDIR} feature.
2510 If you are bootstrapping a released version of GCC then please
2511 quickly review the build status page for your release, available from
2512 @uref{http://gcc.gnu.org/buildstat.html}.
2513 If your system is not listed for the version of GCC that you built,
2515 @email{gcc@@gcc.gnu.org} indicating
2516 that you successfully built and installed GCC@.
2517 Include the following information:
2521 Output from running @file{@var{srcdir}/config.guess}. Do not send
2522 that file itself, just the one-line output from running it.
2525 The output of @samp{gcc -v} for your newly installed @command{gcc}.
2526 This tells us which version of GCC you built and the options you passed to
2530 Whether you enabled all languages or a subset of them. If you used a
2531 full distribution then this information is part of the configure
2532 options in the output of @samp{gcc -v}, but if you downloaded the
2533 ``core'' compiler plus additional front ends then it isn't apparent
2534 which ones you built unless you tell us about it.
2537 If the build was for GNU/Linux, also include:
2540 The distribution name and version (e.g., Red Hat 7.1 or Debian 2.2.3);
2541 this information should be available from @file{/etc/issue}.
2544 The version of the Linux kernel, available from @samp{uname --version}
2548 The version of glibc you used; for RPM-based systems like Red Hat,
2549 Mandrake, and SuSE type @samp{rpm -q glibc} to get the glibc version,
2550 and on systems like Debian and Progeny use @samp{dpkg -l libc6}.
2552 For other systems, you can include similar information if you think it is
2556 Any other information that you think would be useful to people building
2557 GCC on the same configuration. The new entry in the build status list
2558 will include a link to the archived copy of your message.
2561 We'd also like to know if the
2563 @ref{Specific, host/target specific installation notes}
2566 @uref{specific.html,,host/target specific installation notes}
2568 didn't include your host/target information or if that information is
2569 incomplete or out of date. Send a note to
2570 @email{gcc@@gcc.gnu.org} detailing how the information should be changed.
2572 If you find a bug, please report it following the
2573 @uref{../bugs.html,,bug reporting guidelines}.
2575 If you want to print the GCC manuals, do @samp{cd @var{objdir}; make
2576 dvi}. You will need to have @command{texi2dvi} (version at least 4.7)
2577 and @TeX{} installed. This creates a number of @file{.dvi} files in
2578 subdirectories of @file{@var{objdir}}; these may be converted for
2579 printing with programs such as @command{dvips}. Alternately, by using
2580 @samp{make pdf} in place of @samp{make dvi}, you can create documentation
2581 in the form of @file{.pdf} files; this requires @command{texi2pdf}, which
2582 is included with Texinfo version 4.8 and later. You can also
2583 @uref{http://www.gnu.org/order/order.html,,buy printed manuals from the
2584 Free Software Foundation}, though such manuals may not be for the most
2585 recent version of GCC@.
2587 If you would like to generate online HTML documentation, do @samp{cd
2588 @var{objdir}; make html} and HTML will be generated for the gcc manuals in
2589 @file{@var{objdir}/gcc/HTML}.
2596 @uref{./index.html,,Return to the GCC Installation page}
2600 @c ***Binaries****************************************************************
2602 @comment node-name, next, previous, up
2603 @node Binaries, Specific, Installing GCC, Top
2607 @chapter Installing GCC: Binaries
2610 @cindex Installing GCC: Binaries
2612 We are often asked about pre-compiled versions of GCC@. While we cannot
2613 provide these for all platforms, below you'll find links to binaries for
2614 various platforms where creating them by yourself is not easy due to various
2617 Please note that we did not create these binaries, nor do we
2618 support them. If you have any problems installing them, please
2619 contact their makers.
2626 @uref{http://www.bullfreeware.com,,Bull's Freeware and Shareware Archive for AIX};
2629 @uref{http://pware.hvcc.edu,,Hudson Valley Community College Open Source Software for IBM System p};
2632 @uref{http://www.perzl.org/aix,,AIX 5L and 6 Open Source Packages}.
2636 DOS---@uref{http://www.delorie.com/djgpp/,,DJGPP}.
2639 Renesas H8/300[HS]---@uref{http://h8300-hms.sourceforge.net/,,GNU
2640 Development Tools for the Renesas H8/300[HS] Series}.
2646 @uref{http://hpux.cs.utah.edu/,,HP-UX Porting Center};
2649 @uref{ftp://sunsite.informatik.rwth-aachen.de/pub/packages/gcc_hpux/,,Binaries for HP-UX 11.00 at Aachen University of Technology}.
2653 Motorola 68HC11/68HC12---@uref{http://www.gnu-m68hc11.org,,GNU
2654 Development Tools for the Motorola 68HC11/68HC12}.
2657 @uref{http://www.sco.com/skunkware/devtools/index.html#gcc,,SCO
2658 OpenServer/Unixware}.
2661 Solaris 2 (SPARC, Intel)---@uref{http://www.sunfreeware.com/,,Sunfreeware}.
2664 SGI---@uref{http://freeware.sgi.com/,,SGI Freeware}.
2670 The @uref{http://sourceware.org/cygwin/,,Cygwin} project;
2672 The @uref{http://www.mingw.org/,,MinGW} project.
2676 @uref{ftp://ftp.thewrittenword.com/packages/by-name/,,The
2677 Written Word} offers binaries for
2678 AIX 4.3.3, 5.1 and 5.2,
2680 Tru64 UNIX 4.0D and 5.1,
2682 HP-UX 10.20, 11.00, and 11.11, and
2683 Solaris/SPARC 2.5.1, 2.6, 7, 8, 9 and 10.
2686 @uref{http://www.openpkg.org/,,OpenPKG} offers binaries for quite a
2687 number of platforms.
2690 The @uref{http://gcc.gnu.org/wiki/GFortranBinaries,,GFortran Wiki} has
2691 links to GNU Fortran binaries for several platforms.
2694 In addition to those specific offerings, you can get a binary
2695 distribution CD-ROM from the
2696 @uref{http://www.gnu.org/order/order.html,,Free Software Foundation}.
2697 It contains binaries for a number of platforms, and
2698 includes not only GCC, but other stuff as well. The current CD does
2699 not contain the latest version of GCC, but it should allow
2700 bootstrapping the compiler. An updated version of that disk is in the
2708 @uref{./index.html,,Return to the GCC Installation page}
2712 @c ***Specific****************************************************************
2714 @comment node-name, next, previous, up
2715 @node Specific, Old, Binaries, Top
2719 @chapter Host/target specific installation notes for GCC
2722 @cindex Specific installation notes
2723 @cindex Target specific installation
2724 @cindex Host specific installation
2725 @cindex Target specific installation notes
2727 Please read this document carefully @emph{before} installing the
2728 GNU Compiler Collection on your machine.
2730 Note that this list of install notes is @emph{not} a list of supported
2731 hosts or targets. Not all supported hosts and targets are listed
2732 here, only the ones that require host-specific or target-specific
2738 @uref{#alpha-x-x,,alpha*-*-*}
2740 @uref{#alpha-dec-osf,,alpha*-dec-osf*}
2742 @uref{#arc-x-elf,,arc-*-elf}
2744 @uref{#arm-x-elf,,arm-*-elf}
2748 @uref{#bfin,,Blackfin}
2752 @uref{#x-x-freebsd,,*-*-freebsd*}
2754 @uref{#h8300-hms,,h8300-hms}
2756 @uref{#hppa-hp-hpux,,hppa*-hp-hpux*}
2758 @uref{#hppa-hp-hpux10,,hppa*-hp-hpux10}
2760 @uref{#hppa-hp-hpux11,,hppa*-hp-hpux11}
2762 @uref{#x-x-linux-gnu,,*-*-linux-gnu}
2764 @uref{#ix86-x-linux,,i?86-*-linux*}
2766 @uref{#ix86-x-solaris210,,i?86-*-solaris2.10}
2768 @uref{#ia64-x-linux,,ia64-*-linux}
2770 @uref{#ia64-x-hpux,,ia64-*-hpux*}
2772 @uref{#x-ibm-aix,,*-ibm-aix*}
2774 @uref{#iq2000-x-elf,,iq2000-*-elf}
2776 @uref{#m32c-x-elf,,m32c-*-elf}
2778 @uref{#m32r-x-elf,,m32r-*-elf}
2780 @uref{#m6811-elf,,m6811-elf}
2782 @uref{#m6812-elf,,m6812-elf}
2784 @uref{#m68k-x-x,,m68k-*-*}
2786 @uref{#m68k-uclinux,,m68k-uclinux}
2788 @uref{#mep-x-elf,,mep-*-elf}
2790 @uref{#mips-x-x,,mips-*-*}
2792 @uref{#mips-sgi-irix5,,mips-sgi-irix5}
2794 @uref{#mips-sgi-irix6,,mips-sgi-irix6}
2796 @uref{#powerpc-x-x,,powerpc*-*-*}
2798 @uref{#powerpc-x-darwin,,powerpc-*-darwin*}
2800 @uref{#powerpc-x-elf,,powerpc-*-elf}
2802 @uref{#powerpc-x-linux-gnu,,powerpc*-*-linux-gnu*}
2804 @uref{#powerpc-x-netbsd,,powerpc-*-netbsd*}
2806 @uref{#powerpc-x-eabisim,,powerpc-*-eabisim}
2808 @uref{#powerpc-x-eabi,,powerpc-*-eabi}
2810 @uref{#powerpcle-x-elf,,powerpcle-*-elf}
2812 @uref{#powerpcle-x-eabisim,,powerpcle-*-eabisim}
2814 @uref{#powerpcle-x-eabi,,powerpcle-*-eabi}
2816 @uref{#s390-x-linux,,s390-*-linux*}
2818 @uref{#s390x-x-linux,,s390x-*-linux*}
2820 @uref{#s390x-ibm-tpf,,s390x-ibm-tpf*}
2822 @uref{#x-x-solaris2,,*-*-solaris2*}
2824 @uref{#sparc-sun-solaris2,,sparc-sun-solaris2*}
2826 @uref{#sparc-sun-solaris27,,sparc-sun-solaris2.7}
2828 @uref{#sparc-x-linux,,sparc-*-linux*}
2830 @uref{#sparc64-x-solaris2,,sparc64-*-solaris2*}
2832 @uref{#sparcv9-x-solaris2,,sparcv9-*-solaris2*}
2834 @uref{#x-x-vxworks,,*-*-vxworks*}
2836 @uref{#x86-64-x-x,,x86_64-*-*, amd64-*-*}
2838 @uref{#xtensa-x-elf,,xtensa*-*-elf}
2840 @uref{#xtensa-x-linux,,xtensa*-*-linux*}
2842 @uref{#windows,,Microsoft Windows}
2844 @uref{#x-x-cygwin,,*-*-cygwin}
2846 @uref{#x-x-interix,,*-*-interix}
2848 @uref{#x-x-mingw32,,*-*-mingw32}
2852 @uref{#older,,Older systems}
2857 @uref{#elf,,all ELF targets} (SVR4, Solaris 2, etc.)
2863 <!-- -------- host/target specific issues start here ---------------- -->
2866 @heading @anchor{alpha-x-x}alpha*-*-*
2868 This section contains general configuration information for all
2869 alpha-based platforms using ELF (in particular, ignore this section for
2870 DEC OSF/1, Digital UNIX and Tru64 UNIX)@. In addition to reading this
2871 section, please read all other sections that match your target.
2873 We require binutils 2.11.2 or newer.
2874 Previous binutils releases had a number of problems with DWARF 2
2875 debugging information, not the least of which is incorrect linking of
2881 @heading @anchor{alpha-dec-osf}alpha*-dec-osf*
2882 Systems using processors that implement the DEC Alpha architecture and
2883 are running the DEC/Compaq Unix (DEC OSF/1, Digital UNIX, or Compaq
2884 Tru64 UNIX) operating system, for example the DEC Alpha AXP systems.
2886 As of GCC 3.2, versions before @code{alpha*-dec-osf4} are no longer
2887 supported. (These are the versions which identify themselves as DEC
2890 In Digital Unix V4.0, virtual memory exhausted bootstrap failures
2891 may be fixed by configuring with @option{--with-gc=simple},
2892 reconfiguring Kernel Virtual Memory and Swap parameters
2893 per the @command{/usr/sbin/sys_check} Tuning Suggestions,
2894 or applying the patch in
2895 @uref{http://gcc.gnu.org/ml/gcc/2002-08/msg00822.html}.
2897 In Tru64 UNIX V5.1, Compaq introduced a new assembler that does not
2898 currently (2001-06-13) work with @command{mips-tfile}. As a workaround,
2899 we need to use the old assembler, invoked via the barely documented
2900 @option{-oldas} option. To bootstrap GCC, you either need to use the
2904 % CC=cc @var{srcdir}/configure [@var{options}] [@var{target}]
2907 or you can use a copy of GCC 2.95.3 or higher built on Tru64 UNIX V4.0:
2910 % CC=gcc -Wa,-oldas @var{srcdir}/configure [@var{options}] [@var{target}]
2913 As of GNU binutils 2.11.2, neither GNU @command{as} nor GNU @command{ld}
2914 are supported on Tru64 UNIX, so you must not configure GCC with
2915 @option{--with-gnu-as} or @option{--with-gnu-ld}.
2917 GCC writes a @samp{.verstamp} directive to the assembler output file
2918 unless it is built as a cross-compiler. It gets the version to use from
2919 the system header file @file{/usr/include/stamp.h}. If you install a
2920 new version of DEC Unix, you should rebuild GCC to pick up the new version
2923 @samp{make compare} may fail on old versions of DEC Unix unless you add
2924 @option{-save-temps} to @code{BOOT_CFLAGS}. On these systems, the name
2925 of the assembler input file is stored in the object file, and that makes
2926 comparison fail if it differs between the @code{stage1} and
2927 @code{stage2} compilations. The option @option{-save-temps} forces a
2928 fixed name to be used for the assembler input file, instead of a
2929 randomly chosen name in @file{/tmp}. Do not add @option{-save-temps}
2930 unless the comparisons fail without that option. If you add
2931 @option{-save-temps}, you will have to manually delete the @samp{.i} and
2932 @samp{.s} files after each series of compilations.
2934 GCC now supports both the native (ECOFF) debugging format used by DBX
2935 and GDB and an encapsulated STABS format for use only with GDB@. See the
2936 discussion of the @option{--with-stabs} option of @file{configure} above
2937 for more information on these formats and how to select them.
2939 There is a bug in DEC's assembler that produces incorrect line numbers
2940 for ECOFF format when the @samp{.align} directive is used. To work
2941 around this problem, GCC will not emit such alignment directives
2942 while writing ECOFF format debugging information even if optimization is
2943 being performed. Unfortunately, this has the very undesirable
2944 side-effect that code addresses when @option{-O} is specified are
2945 different depending on whether or not @option{-g} is also specified.
2947 To avoid this behavior, specify @option{-gstabs+} and use GDB instead of
2948 DBX@. DEC is now aware of this problem with the assembler and hopes to
2949 provide a fix shortly.
2954 @heading @anchor{arc-x-elf}arc-*-elf
2955 Argonaut ARC processor.
2956 This configuration is intended for embedded systems.
2961 @heading @anchor{arm-x-elf}arm-*-elf
2962 ARM-family processors. Subtargets that use the ELF object format
2963 require GNU binutils 2.13 or newer. Such subtargets include:
2964 @code{arm-*-freebsd}, @code{arm-*-netbsdelf}, @code{arm-*-*linux}
2965 and @code{arm-*-rtems}.
2970 @heading @anchor{avr}avr
2972 ATMEL AVR-family micro controllers. These are used in embedded
2973 applications. There are no standard Unix configurations.
2975 @xref{AVR Options,, AVR Options, gcc, Using the GNU Compiler
2979 See ``AVR Options'' in the main manual
2981 for the list of supported MCU types.
2983 Use @samp{configure --target=avr --enable-languages="c"} to configure GCC@.
2985 Further installation notes and other useful information about AVR tools
2986 can also be obtained from:
2990 @uref{http://www.nongnu.org/avr/,,http://www.nongnu.org/avr/}
2992 @uref{http://home.overta.ru/users/denisc/,,http://home.overta.ru/users/denisc/}
2994 @uref{http://www.amelek.gda.pl/avr/,,http://www.amelek.gda.pl/avr/}
2997 We @emph{strongly} recommend using binutils 2.13 or newer.
2999 The following error:
3001 Error: register required
3004 indicates that you should upgrade to a newer version of the binutils.
3009 @heading @anchor{bfin}Blackfin
3011 The Blackfin processor, an Analog Devices DSP.
3013 @xref{Blackfin Options,, Blackfin Options, gcc, Using the GNU Compiler
3017 See ``Blackfin Options'' in the main manual
3020 More information, and a version of binutils with support for this processor,
3021 is available at @uref{http://blackfin.uclinux.org}
3026 @heading @anchor{cris}CRIS
3028 CRIS is the CPU architecture in Axis Communications ETRAX system-on-a-chip
3029 series. These are used in embedded applications.
3032 @xref{CRIS Options,, CRIS Options, gcc, Using the GNU Compiler
3036 See ``CRIS Options'' in the main manual
3038 for a list of CRIS-specific options.
3040 There are a few different CRIS targets:
3043 Mainly for monolithic embedded systems. Includes a multilib for the
3044 @samp{v10} core used in @samp{ETRAX 100 LX}.
3045 @item cris-axis-linux-gnu
3046 A GNU/Linux port for the CRIS architecture, currently targeting
3047 @samp{ETRAX 100 LX} by default.
3050 For @code{cris-axis-elf} you need binutils 2.11
3051 or newer. For @code{cris-axis-linux-gnu} you need binutils 2.12 or newer.
3053 Pre-packaged tools can be obtained from
3054 @uref{ftp://ftp.axis.com/pub/axis/tools/cris/compiler-kit/}. More
3055 information about this platform is available at
3056 @uref{http://developer.axis.com/}.
3061 @heading @anchor{crx}CRX
3063 The CRX CompactRISC architecture is a low-power 32-bit architecture with
3064 fast context switching and architectural extensibility features.
3067 @xref{CRX Options,, CRX Options, gcc, Using and Porting the GNU Compiler
3072 See ``CRX Options'' in the main manual for a list of CRX-specific options.
3075 Use @samp{configure --target=crx-elf --enable-languages=c,c++} to configure
3076 GCC@ for building a CRX cross-compiler. The option @samp{--target=crx-elf}
3077 is also used to build the @samp{newlib} C library for CRX.
3079 It is also possible to build libstdc++-v3 for the CRX architecture. This
3080 needs to be done in a separate step with the following configure settings:
3081 @samp{gcc/libstdc++-v3/configure --host=crx-elf --with-newlib
3082 --enable-sjlj-exceptions --enable-cxx-flags='-fexceptions -frtti'}
3087 @heading @anchor{dos}DOS
3089 Please have a look at the @uref{binaries.html,,binaries page}.
3091 You cannot install GCC by itself on MSDOS; it will not compile under
3092 any MSDOS compiler except itself. You need to get the complete
3093 compilation package DJGPP, which includes binaries as well as sources,
3094 and includes all the necessary compilation tools and libraries.
3099 @heading @anchor{x-x-freebsd}*-*-freebsd*
3101 The version of binutils installed in @file{/usr/bin} probably works with
3102 this release of GCC@. However, on FreeBSD 4, bootstrapping against the
3103 latest FSF binutils is known to improve overall testsuite results; and,
3104 on FreeBSD/alpha, using binutils 2.14 or later is required to build libjava.
3106 Support for FreeBSD 1 was discontinued in GCC 3.2.
3108 Support for FreeBSD 2 will be discontinued after GCC 3.4. The
3109 following was true for GCC 3.1 but the current status is unknown.
3110 For FreeBSD 2 or any mutant a.out versions of FreeBSD 3: All
3111 configuration support and files as shipped with GCC 2.95 are still in
3112 place. FreeBSD 2.2.7 has been known to bootstrap completely; however,
3113 it is unknown which version of binutils was used (it is assumed that it
3114 was the system copy in @file{/usr/bin}) and C++ EH failures were noted.
3116 For FreeBSD using the ELF file format: DWARF 2 debugging is now the
3117 default for all CPU architectures. It had been the default on
3118 FreeBSD/alpha since its inception. You may use @option{-gstabs} instead
3119 of @option{-g}, if you really want the old debugging format. There are
3120 no known issues with mixing object files and libraries with different
3121 debugging formats. Otherwise, this release of GCC should now match more
3122 of the configuration used in the stock FreeBSD configuration of GCC@. In
3123 particular, @option{--enable-threads} is now configured by default.
3124 However, as a general user, do not attempt to replace the system
3125 compiler with this release. Known to bootstrap and check with good
3126 results on FreeBSD 4.9-STABLE and 5-CURRENT@. In the past, known to
3127 bootstrap and check with good results on FreeBSD 3.0, 3.4, 4.0, 4.2,
3128 4.3, 4.4, 4.5, 4.8-STABLE@.
3130 In principle, @option{--enable-threads} is now compatible with
3131 @option{--enable-libgcj} on FreeBSD@. However, it has only been built
3132 and tested on @samp{i386-*-freebsd[45]} and @samp{alpha-*-freebsd[45]}.
3134 library may be incorrectly built (symbols are missing at link time).
3135 There is a rare timing-based startup hang (probably involves an
3136 assumption about the thread library). Multi-threaded boehm-gc (required for
3137 libjava) exposes severe threaded signal-handling bugs on FreeBSD before
3138 4.5-RELEASE@. Other CPU architectures
3139 supported by FreeBSD will require additional configuration tuning in, at
3140 the very least, both boehm-gc and libffi.
3142 Shared @file{libgcc_s.so} is now built and installed by default.
3147 @heading @anchor{h8300-hms}h8300-hms
3148 Renesas H8/300 series of processors.
3150 Please have a look at the @uref{binaries.html,,binaries page}.
3152 The calling convention and structure layout has changed in release 2.6.
3153 All code must be recompiled. The calling convention now passes the
3154 first three arguments in function calls in registers. Structures are no
3155 longer a multiple of 2 bytes.
3160 @heading @anchor{hppa-hp-hpux}hppa*-hp-hpux*
3161 Support for HP-UX version 9 and older was discontinued in GCC 3.4.
3163 We require using gas/binutils on all hppa platforms. Version 2.19 or
3164 later is recommended.
3166 It may be helpful to configure GCC with the
3167 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}} and
3168 @option{--with-as=@dots{}} options to ensure that GCC can find GAS@.
3170 The HP assembler should not be used with GCC. It is rarely tested and may
3171 not work. It shouldn't be used with any languages other than C due to its
3174 Specifically, @option{-g} does not work (HP-UX uses a peculiar debugging
3175 format which GCC does not know about). It also inserts timestamps
3176 into each object file it creates, causing the 3-stage comparison test to
3177 fail during a bootstrap. You should be able to continue by saying
3178 @samp{make all-host all-target} after getting the failure from @samp{make}.
3180 Various GCC features are not supported. For example, it does not support weak
3181 symbols or alias definitions. As a result, explicit template instantiations
3182 are required when using C++. This makes it difficult if not impossible to
3183 build many C++ applications.
3185 There are two default scheduling models for instructions. These are
3186 PROCESSOR_7100LC and PROCESSOR_8000. They are selected from the pa-risc
3187 architecture specified for the target machine when configuring.
3188 PROCESSOR_8000 is the default. PROCESSOR_7100LC is selected when
3189 the target is a @samp{hppa1*} machine.
3191 The PROCESSOR_8000 model is not well suited to older processors. Thus,
3192 it is important to completely specify the machine architecture when
3193 configuring if you want a model other than PROCESSOR_8000. The macro
3194 TARGET_SCHED_DEFAULT can be defined in BOOT_CFLAGS if a different
3195 default scheduling model is desired.
3197 As of GCC 4.0, GCC uses the UNIX 95 namespace for HP-UX 10.10
3198 through 11.00, and the UNIX 98 namespace for HP-UX 11.11 and later.
3199 This namespace change might cause problems when bootstrapping with
3200 an earlier version of GCC or the HP compiler as essentially the same
3201 namespace is required for an entire build. This problem can be avoided
3202 in a number of ways. With HP cc, @env{UNIX_STD} can be set to @samp{95}
3203 or @samp{98}. Another way is to add an appropriate set of predefines
3204 to @env{CC}. The description for the @option{munix=} option contains
3205 a list of the predefines used with each standard.
3207 More specific information to @samp{hppa*-hp-hpux*} targets follows.
3212 @heading @anchor{hppa-hp-hpux10}hppa*-hp-hpux10
3214 For hpux10.20, we @emph{highly} recommend you pick up the latest sed patch
3215 @code{PHCO_19798} from HP@. HP has two sites which provide patches free of
3221 <a href="http://us.itrc.hp.com/service/home/home.do">US, Canada, Asia-Pacific, and
3225 @uref{http://us.itrc.hp.com/service/home/home.do,,} US, Canada, Asia-Pacific,
3229 @uref{http://europe.itrc.hp.com/service/home/home.do,,} Europe.
3232 The C++ ABI has changed incompatibly in GCC 4.0. COMDAT subspaces are
3233 used for one-only code and data. This resolves many of the previous
3234 problems in using C++ on this target. However, the ABI is not compatible
3235 with the one implemented under HP-UX 11 using secondary definitions.
3240 @heading @anchor{hppa-hp-hpux11}hppa*-hp-hpux11
3242 GCC 3.0 and up support HP-UX 11. GCC 2.95.x is not supported and cannot
3243 be used to compile GCC 3.0 and up.
3245 The libffi and libjava libraries haven't been ported to 64-bit HP-UX@
3248 Refer to @uref{binaries.html,,binaries} for information about obtaining
3249 precompiled GCC binaries for HP-UX@. Precompiled binaries must be obtained
3250 to build the Ada language as it can't be bootstrapped using C@. Ada is
3251 only available for the 32-bit PA-RISC runtime.
3253 Starting with GCC 3.4 an ISO C compiler is required to bootstrap. The
3254 bundled compiler supports only traditional C; you will need either HP's
3255 unbundled compiler, or a binary distribution of GCC@.
3257 It is possible to build GCC 3.3 starting with the bundled HP compiler,
3258 but the process requires several steps. GCC 3.3 can then be used to
3259 build later versions. The fastjar program contains ISO C code and
3260 can't be built with the HP bundled compiler. This problem can be
3261 avoided by not building the Java language. For example, use the
3262 @option{--enable-languages="c,c++,f77,objc"} option in your configure
3265 There are several possible approaches to building the distribution.
3266 Binutils can be built first using the HP tools. Then, the GCC
3267 distribution can be built. The second approach is to build GCC
3268 first using the HP tools, then build binutils, then rebuild GCC@.
3269 There have been problems with various binary distributions, so it
3270 is best not to start from a binary distribution.
3272 On 64-bit capable systems, there are two distinct targets. Different
3273 installation prefixes must be used if both are to be installed on
3274 the same system. The @samp{hppa[1-2]*-hp-hpux11*} target generates code
3275 for the 32-bit PA-RISC runtime architecture and uses the HP linker.
3276 The @samp{hppa64-hp-hpux11*} target generates 64-bit code for the
3277 PA-RISC 2.0 architecture.
3279 The script config.guess now selects the target type based on the compiler
3280 detected during configuration. You must define @env{PATH} or @env{CC} so
3281 that configure finds an appropriate compiler for the initial bootstrap.
3282 When @env{CC} is used, the definition should contain the options that are
3283 needed whenever @env{CC} is used.
3285 Specifically, options that determine the runtime architecture must be
3286 in @env{CC} to correctly select the target for the build. It is also
3287 convenient to place many other compiler options in @env{CC}. For example,
3288 @env{CC="cc -Ac +DA2.0W -Wp,-H16376 -D_CLASSIC_TYPES -D_HPUX_SOURCE"}
3289 can be used to bootstrap the GCC 3.3 branch with the HP compiler in
3290 64-bit K&R/bundled mode. The @option{+DA2.0W} option will result in
3291 the automatic selection of the @samp{hppa64-hp-hpux11*} target. The
3292 macro definition table of cpp needs to be increased for a successful
3293 build with the HP compiler. _CLASSIC_TYPES and _HPUX_SOURCE need to
3294 be defined when building with the bundled compiler, or when using the
3295 @option{-Ac} option. These defines aren't necessary with @option{-Ae}.
3297 It is best to explicitly configure the @samp{hppa64-hp-hpux11*} target
3298 with the @option{--with-ld=@dots{}} option. This overrides the standard
3299 search for ld. The two linkers supported on this target require different
3300 commands. The default linker is determined during configuration. As a
3301 result, it's not possible to switch linkers in the middle of a GCC build.
3302 This has been reported to sometimes occur in unified builds of binutils
3305 A recent linker patch must be installed for the correct operation of
3306 GCC 3.3 and later. @code{PHSS_26559} and @code{PHSS_24304} are the
3307 oldest linker patches that are known to work. They are for HP-UX
3308 11.00 and 11.11, respectively. @code{PHSS_24303}, the companion to
3309 @code{PHSS_24304}, might be usable but it hasn't been tested. These
3310 patches have been superseded. Consult the HP patch database to obtain
3311 the currently recommended linker patch for your system.
3313 The patches are necessary for the support of weak symbols on the
3314 32-bit port, and for the running of initializers and finalizers. Weak
3315 symbols are implemented using SOM secondary definition symbols. Prior
3316 to HP-UX 11, there are bugs in the linker support for secondary symbols.
3317 The patches correct a problem of linker core dumps creating shared
3318 libraries containing secondary symbols, as well as various other
3319 linking issues involving secondary symbols.
3321 GCC 3.3 uses the ELF DT_INIT_ARRAY and DT_FINI_ARRAY capabilities to
3322 run initializers and finalizers on the 64-bit port. The 32-bit port
3323 uses the linker @option{+init} and @option{+fini} options for the same
3324 purpose. The patches correct various problems with the +init/+fini
3325 options, including program core dumps. Binutils 2.14 corrects a
3326 problem on the 64-bit port resulting from HP's non-standard use of
3327 the .init and .fini sections for array initializers and finalizers.
3329 Although the HP and GNU linkers are both supported for the
3330 @samp{hppa64-hp-hpux11*} target, it is strongly recommended that the
3331 HP linker be used for link editing on this target.
3333 At this time, the GNU linker does not support the creation of long
3334 branch stubs. As a result, it can't successfully link binaries
3335 containing branch offsets larger than 8 megabytes. In addition,
3336 there are problems linking shared libraries, linking executables
3337 with @option{-static}, and with dwarf2 unwind and exception support.
3338 It also doesn't provide stubs for internal calls to global functions
3339 in shared libraries, so these calls can't be overloaded.
3341 The HP dynamic loader does not support GNU symbol versioning, so symbol
3342 versioning is not supported. It may be necessary to disable symbol
3343 versioning with @option{--disable-symvers} when using GNU ld.
3345 POSIX threads are the default. The optional DCE thread library is not
3346 supported, so @option{--enable-threads=dce} does not work.
3351 @heading @anchor{x-x-linux-gnu}*-*-linux-gnu
3353 Versions of libstdc++-v3 starting with 3.2.1 require bug fixes present
3354 in glibc 2.2.5 and later. More information is available in the
3355 libstdc++-v3 documentation.
3360 @heading @anchor{ix86-x-linux}i?86-*-linux*
3362 As of GCC 3.3, binutils 2.13.1 or later is required for this platform.
3363 See @uref{http://gcc.gnu.org/PR10877,,bug 10877} for more information.
3365 If you receive Signal 11 errors when building on GNU/Linux, then it is
3366 possible you have a hardware problem. Further information on this can be
3367 found on @uref{http://www.bitwizard.nl/sig11/,,www.bitwizard.nl}.
3372 @heading @anchor{ix86-x-solaris210}i?86-*-solaris2.10
3373 Use this for Solaris 10 or later on x86 and x86-64 systems. This
3374 configuration is supported by GCC 4.0 and later versions only.
3376 It is recommended that you configure GCC to use the GNU assembler in
3377 @file{/usr/sfw/bin/gas} but the Sun linker, using the options
3378 @option{--with-gnu-as --with-as=/usr/sfw/bin/gas --without-gnu-ld
3379 --with-ld=/usr/ccs/bin/ld}.
3384 @heading @anchor{ia64-x-linux}ia64-*-linux
3385 IA-64 processor (also known as IPF, or Itanium Processor Family)
3388 If you are using the installed system libunwind library with
3389 @option{--with-system-libunwind}, then you must use libunwind 0.98 or
3392 None of the following versions of GCC has an ABI that is compatible
3393 with any of the other versions in this list, with the exception that
3394 Red Hat 2.96 and Trillian 000171 are compatible with each other:
3395 3.1, 3.0.2, 3.0.1, 3.0, Red Hat 2.96, and Trillian 000717.
3396 This primarily affects C++ programs and programs that create shared libraries.
3397 GCC 3.1 or later is recommended for compiling linux, the kernel.
3398 As of version 3.1 GCC is believed to be fully ABI compliant, and hence no
3399 more major ABI changes are expected.
3404 @heading @anchor{ia64-x-hpux}ia64-*-hpux*
3405 Building GCC on this target requires the GNU Assembler. The bundled HP
3406 assembler will not work. To prevent GCC from using the wrong assembler,
3407 the option @option{--with-gnu-as} may be necessary.
3409 The GCC libunwind library has not been ported to HPUX@. This means that for
3410 GCC versions 3.2.3 and earlier, @option{--enable-libunwind-exceptions}
3411 is required to build GCC@. For GCC 3.3 and later, this is the default.
3412 For gcc 3.4.3 and later, @option{--enable-libunwind-exceptions} is
3413 removed and the system libunwind library will always be used.
3417 <!-- rs6000-ibm-aix*, powerpc-ibm-aix* -->
3419 @heading @anchor{x-ibm-aix}*-ibm-aix*
3420 Support for AIX version 3 and older was discontinued in GCC 3.4.
3421 Support for AIX version 4.2 and older was discontinued in GCC 4.5.
3423 ``out of memory'' bootstrap failures may indicate a problem with
3424 process resource limits (ulimit). Hard limits are configured in the
3425 @file{/etc/security/limits} system configuration file.
3427 GCC can bootstrap with recent versions of IBM XLC, but bootstrapping
3428 with an earlier release of GCC is recommended. Bootstrapping with XLC
3429 requires a larger data segment, which can be enabled through the
3430 @var{LDR_CNTRL} environment variable, e.g.,
3433 % LDR_CNTRL=MAXDATA=0x50000000
3437 One can start with a pre-compiled version of GCC to build from
3438 sources. One may delete GCC's ``fixed'' header files when starting
3439 with a version of GCC built for an earlier release of AIX.
3441 To speed up the configuration phases of bootstrapping and installing GCC,
3442 one may use GNU Bash instead of AIX @command{/bin/sh}, e.g.,
3445 % CONFIG_SHELL=/opt/freeware/bin/bash
3446 % export CONFIG_SHELL
3449 and then proceed as described in @uref{build.html,,the build
3450 instructions}, where we strongly recommend specifying an absolute path
3451 to invoke @var{srcdir}/configure.
3453 Because GCC on AIX is built as a 32-bit executable by default,
3454 (although it can generate 64-bit programs) the GMP and MPFR libraries
3455 required by gfortran must be 32-bit libraries. Building GMP and MPFR
3456 as static archive libraries works better than shared libraries.
3458 Errors involving @code{alloca} when building GCC generally are due
3459 to an incorrect definition of @code{CC} in the Makefile or mixing files
3460 compiled with the native C compiler and GCC@. During the stage1 phase of
3461 the build, the native AIX compiler @strong{must} be invoked as @command{cc}
3462 (not @command{xlc}). Once @command{configure} has been informed of
3463 @command{xlc}, one needs to use @samp{make distclean} to remove the
3464 configure cache files and ensure that @env{CC} environment variable
3465 does not provide a definition that will confuse @command{configure}.
3466 If this error occurs during stage2 or later, then the problem most likely
3467 is the version of Make (see above).
3469 The native @command{as} and @command{ld} are recommended for bootstrapping
3470 on AIX@. The GNU Assembler, GNU Linker, and GNU Binutils version 2.20
3471 is required to bootstrap on AIX 5@. The native AIX tools do
3472 interoperate with GCC@.
3474 Building @file{libstdc++.a} requires a fix for an AIX Assembler bug
3475 APAR IY26685 (AIX 4.3) or APAR IY25528 (AIX 5.1). It also requires a
3476 fix for another AIX Assembler bug and a co-dependent AIX Archiver fix
3477 referenced as APAR IY53606 (AIX 5.2) or as APAR IY54774 (AIX 5.1)
3479 @samp{libstdc++} in GCC 3.4 increments the major version number of the
3480 shared object and GCC installation places the @file{libstdc++.a}
3481 shared library in a common location which will overwrite the and GCC
3482 3.3 version of the shared library. Applications either need to be
3483 re-linked against the new shared library or the GCC 3.1 and GCC 3.3
3484 versions of the @samp{libstdc++} shared object needs to be available
3485 to the AIX runtime loader. The GCC 3.1 @samp{libstdc++.so.4}, if
3486 present, and GCC 3.3 @samp{libstdc++.so.5} shared objects can be
3487 installed for runtime dynamic loading using the following steps to set
3488 the @samp{F_LOADONLY} flag in the shared object for @emph{each}
3489 multilib @file{libstdc++.a} installed:
3491 Extract the shared objects from the currently installed
3492 @file{libstdc++.a} archive:
3494 % ar -x libstdc++.a libstdc++.so.4 libstdc++.so.5
3497 Enable the @samp{F_LOADONLY} flag so that the shared object will be
3498 available for runtime dynamic loading, but not linking:
3500 % strip -e libstdc++.so.4 libstdc++.so.5
3503 Archive the runtime-only shared object in the GCC 3.4
3504 @file{libstdc++.a} archive:
3506 % ar -q libstdc++.a libstdc++.so.4 libstdc++.so.5
3509 Linking executables and shared libraries may produce warnings of
3510 duplicate symbols. The assembly files generated by GCC for AIX always
3511 have included multiple symbol definitions for certain global variable
3512 and function declarations in the original program. The warnings should
3513 not prevent the linker from producing a correct library or runnable
3516 AIX 4.3 utilizes a ``large format'' archive to support both 32-bit and
3517 64-bit object modules. The routines provided in AIX 4.3.0 and AIX 4.3.1
3518 to parse archive libraries did not handle the new format correctly.
3519 These routines are used by GCC and result in error messages during
3520 linking such as ``not a COFF file''. The version of the routines shipped
3521 with AIX 4.3.1 should work for a 32-bit environment. The @option{-g}
3522 option of the archive command may be used to create archives of 32-bit
3523 objects using the original ``small format''. A correct version of the
3524 routines is shipped with AIX 4.3.2 and above.
3526 Some versions of the AIX binder (linker) can fail with a relocation
3527 overflow severe error when the @option{-bbigtoc} option is used to link
3528 GCC-produced object files into an executable that overflows the TOC@. A fix
3529 for APAR IX75823 (OVERFLOW DURING LINK WHEN USING GCC AND -BBIGTOC) is
3530 available from IBM Customer Support and from its
3531 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
3532 website as PTF U455193.
3534 The AIX 4.3.2.1 linker (bos.rte.bind_cmds Level 4.3.2.1) will dump core
3535 with a segmentation fault when invoked by any version of GCC@. A fix for
3536 APAR IX87327 is available from IBM Customer Support and from its
3537 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
3538 website as PTF U461879. This fix is incorporated in AIX 4.3.3 and above.
3540 The initial assembler shipped with AIX 4.3.0 generates incorrect object
3541 files. A fix for APAR IX74254 (64BIT DISASSEMBLED OUTPUT FROM COMPILER FAILS
3542 TO ASSEMBLE/BIND) is available from IBM Customer Support and from its
3543 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
3544 website as PTF U453956. This fix is incorporated in AIX 4.3.1 and above.
3546 AIX provides National Language Support (NLS)@. Compilers and assemblers
3547 use NLS to support locale-specific representations of various data
3548 formats including floating-point numbers (e.g., @samp{.} vs @samp{,} for
3549 separating decimal fractions). There have been problems reported where
3550 GCC does not produce the same floating-point formats that the assembler
3551 expects. If one encounters this problem, set the @env{LANG}
3552 environment variable to @samp{C} or @samp{En_US}.
3554 A default can be specified with the @option{-mcpu=@var{cpu_type}}
3555 switch and using the configure option @option{--with-cpu-@var{cpu_type}}.
3560 @heading @anchor{iq2000-x-elf}iq2000-*-elf
3561 Vitesse IQ2000 processors. These are used in embedded
3562 applications. There are no standard Unix configurations.
3567 @heading @anchor{m32c-x-elf}m32c-*-elf
3568 Renesas M32C processor.
3569 This configuration is intended for embedded systems.
3574 @heading @anchor{m32r-x-elf}m32r-*-elf
3575 Renesas M32R processor.
3576 This configuration is intended for embedded systems.
3581 @heading @anchor{m6811-elf}m6811-elf
3582 Motorola 68HC11 family micro controllers. These are used in embedded
3583 applications. There are no standard Unix configurations.
3588 @heading @anchor{m6812-elf}m6812-elf
3589 Motorola 68HC12 family micro controllers. These are used in embedded
3590 applications. There are no standard Unix configurations.
3595 @heading @anchor{m68k-x-x}m68k-*-*
3597 @samp{m68k-*-elf*}, @samp{m68k-*-rtems}, @samp{m68k-*-uclinux} and
3599 build libraries for both M680x0 and ColdFire processors. If you only
3600 need the M680x0 libraries, you can omit the ColdFire ones by passing
3601 @option{--with-arch=m68k} to @command{configure}. Alternatively, you
3602 can omit the M680x0 libraries by passing @option{--with-arch=cf} to
3603 @command{configure}. These targets default to 5206 or 5475 code as
3604 appropriate for the target system when
3605 configured with @option{--with-arch=cf} and 68020 code otherwise.
3607 The @samp{m68k-*-netbsd} and
3608 @samp{m68k-*-openbsd} targets also support the @option{--with-arch}
3609 option. They will generate ColdFire CFV4e code when configured with
3610 @option{--with-arch=cf} and 68020 code otherwise.
3612 You can override the default processors listed above by configuring
3613 with @option{--with-cpu=@var{target}}. This @var{target} can either
3614 be a @option{-mcpu} argument or one of the following values:
3615 @samp{m68000}, @samp{m68010}, @samp{m68020}, @samp{m68030},
3616 @samp{m68040}, @samp{m68060}, @samp{m68020-40} and @samp{m68020-60}.
3621 @heading @anchor{m68k-x-uclinux}m68k-*-uclinux
3622 GCC 4.3 changed the uClinux configuration so that it uses the
3623 @samp{m68k-linux-gnu} ABI rather than the @samp{m68k-elf} ABI.
3624 It also added improved support for C++ and flat shared libraries,
3625 both of which were ABI changes. However, you can still use the
3626 original ABI by configuring for @samp{m68k-uclinuxoldabi} or
3627 @samp{m68k-@var{vendor}-uclinuxoldabi}.
3633 @heading @anchor{mep-x-elf}mep-*-elf
3634 Toshiba Media embedded Processor.
3635 This configuration is intended for embedded systems.
3640 @heading @anchor{mips-x-x}mips-*-*
3641 If on a MIPS system you get an error message saying ``does not have gp
3642 sections for all it's [sic] sectons [sic]'', don't worry about it. This
3643 happens whenever you use GAS with the MIPS linker, but there is not
3644 really anything wrong, and it is okay to use the output file. You can
3645 stop such warnings by installing the GNU linker.
3647 It would be nice to extend GAS to produce the gp tables, but they are
3648 optional, and there should not be a warning about their absence.
3650 The libstdc++ atomic locking routines for MIPS targets requires MIPS II
3651 and later. A patch went in just after the GCC 3.3 release to
3652 make @samp{mips*-*-*} use the generic implementation instead. You can also
3653 configure for @samp{mipsel-elf} as a workaround. The
3654 @samp{mips*-*-linux*} target continues to use the MIPS II routines. More
3655 work on this is expected in future releases.
3657 @c If you make --with-llsc the default for another target, please also
3658 @c update the description of the --with-llsc option.
3660 The built-in @code{__sync_*} functions are available on MIPS II and
3661 later systems and others that support the @samp{ll}, @samp{sc} and
3662 @samp{sync} instructions. This can be overridden by passing
3663 @option{--with-llsc} or @option{--without-llsc} when configuring GCC.
3664 Since the Linux kernel emulates these instructions if they are
3665 missing, the default for @samp{mips*-*-linux*} targets is
3666 @option{--with-llsc}. The @option{--with-llsc} and
3667 @option{--without-llsc} configure options may be overridden at compile
3668 time by passing the @option{-mllsc} or @option{-mno-llsc} options to
3671 MIPS systems check for division by zero (unless
3672 @option{-mno-check-zero-division} is passed to the compiler) by
3673 generating either a conditional trap or a break instruction. Using
3674 trap results in smaller code, but is only supported on MIPS II and
3675 later. Also, some versions of the Linux kernel have a bug that
3676 prevents trap from generating the proper signal (@code{SIGFPE}). To enable
3677 the use of break, use the @option{--with-divide=breaks}
3678 @command{configure} option when configuring GCC@. The default is to
3679 use traps on systems that support them.
3681 Cross-compilers for the MIPS as target using the MIPS assembler
3682 currently do not work, because the auxiliary programs
3683 @file{mips-tdump.c} and @file{mips-tfile.c} can't be compiled on
3684 anything but a MIPS@. It does work to cross compile for a MIPS
3685 if you use the GNU assembler and linker.
3687 The assembler from GNU binutils 2.17 and earlier has a bug in the way
3688 it sorts relocations for REL targets (o32, o64, EABI). This can cause
3689 bad code to be generated for simple C++ programs. Also the linker
3690 from GNU binutils versions prior to 2.17 has a bug which causes the
3691 runtime linker stubs in very large programs, like @file{libgcj.so}, to
3692 be incorrectly generated. GNU Binutils 2.18 and later (and snapshots
3693 made after Nov. 9, 2006) should be free from both of these problems.
3698 @heading @anchor{mips-sgi-irix5}mips-sgi-irix5
3700 In order to compile GCC on an SGI running IRIX 5, the @samp{compiler_dev.hdr}
3701 subsystem must be installed from the IDO CD-ROM supplied by SGI@.
3702 It is also available for download from
3703 @uref{ftp://ftp.sgi.com/sgi/IRIX5.3/iris-development-option-5.3.tardist}.
3705 If you use the MIPS C compiler to bootstrap, it may be necessary
3706 to increase its table size for switch statements with the
3707 @option{-Wf,-XNg1500} option. If you use the @option{-O2}
3708 optimization option, you also need to use @option{-Olimit 3000}.
3710 To enable debugging under IRIX 5, you must use GNU binutils 2.15 or
3711 later, and use the @option{--with-gnu-ld} @command{configure} option
3712 when configuring GCC@. You need to use GNU @command{ar} and @command{nm},
3713 also distributed with GNU binutils.
3715 Some users have reported that @command{/bin/sh} will hang during bootstrap.
3716 This problem can be avoided by running the commands:
3719 % CONFIG_SHELL=/bin/ksh
3720 % export CONFIG_SHELL
3723 before starting the build.
3728 @heading @anchor{mips-sgi-irix6}mips-sgi-irix6
3730 If you are using SGI's MIPSpro @command{cc} as your bootstrap compiler, you must
3731 ensure that the N32 ABI is in use. To test this, compile a simple C
3732 file with @command{cc} and then run @command{file} on the
3733 resulting object file. The output should look like:
3736 test.o: ELF N32 MSB @dots{}
3742 test.o: ELF 32-bit MSB @dots{}
3748 test.o: ELF 64-bit MSB @dots{}
3751 then your version of @command{cc} uses the O32 or N64 ABI by default. You
3752 should set the environment variable @env{CC} to @samp{cc -n32}
3753 before configuring GCC@.
3755 If you want the resulting @command{gcc} to run on old 32-bit systems
3756 with the MIPS R4400 CPU, you need to ensure that only code for the @samp{mips3}
3757 instruction set architecture (ISA) is generated. While GCC 3.x does
3758 this correctly, both GCC 2.95 and SGI's MIPSpro @command{cc} may change
3759 the ISA depending on the machine where GCC is built. Using one of them
3760 as the bootstrap compiler may result in @samp{mips4} code, which won't run at
3761 all on @samp{mips3}-only systems. For the test program above, you should see:
3764 test.o: ELF N32 MSB mips-3 @dots{}
3770 test.o: ELF N32 MSB mips-4 @dots{}
3773 instead, you should set the environment variable @env{CC} to @samp{cc
3774 -n32 -mips3} or @samp{gcc -mips3} respectively before configuring GCC@.
3776 MIPSpro C 7.4 may cause bootstrap failures, due to a bug when inlining
3777 @code{memcmp}. Either add @code{-U__INLINE_INTRINSICS} to the @env{CC}
3778 environment variable as a workaround or upgrade to MIPSpro C 7.4.1m.
3780 GCC on IRIX 6 is usually built to support the N32, O32 and N64 ABIs. If
3781 you build GCC on a system that doesn't have the N64 libraries installed
3782 or cannot run 64-bit binaries,
3783 you need to configure with @option{--disable-multilib} so GCC doesn't
3784 try to use them. This will disable building the O32 libraries, too.
3785 Look for @file{/usr/lib64/libc.so.1} to see if you
3786 have the 64-bit libraries installed.
3788 To enable debugging for the O32 ABI, you must use GNU @command{as} from
3789 GNU binutils 2.15 or later. You may also use GNU @command{ld}, but
3790 this is not required and currently causes some problems with Ada.
3792 The @option{--enable-libgcj}
3793 option is disabled by default: IRIX 6 uses a very low default limit
3794 (20480) for the command line length. Although @command{libtool} contains a
3795 workaround for this problem, at least the N64 @samp{libgcj} is known not
3796 to build despite this, running into an internal error of the native
3797 @command{ld}. A sure fix is to increase this limit (@samp{ncargs}) to
3798 its maximum of 262144 bytes. If you have root access, you can use the
3799 @command{systune} command to do this.
3801 @code{wchar_t} support in @samp{libstdc++} is not available for old
3802 IRIX 6.5.x releases, @math{x < 19}. The problem cannot be autodetected
3803 and in order to build GCC for such targets you need to configure with
3804 @option{--disable-wchar_t}.
3806 See @uref{http://freeware.sgi.com/} for more
3807 information about using GCC on IRIX platforms.
3812 @heading @anchor{moxie-x-elf}moxie-*-elf
3813 The moxie processor. See @uref{http://moxielogic.org/} for more
3814 information about this processor.
3819 @heading @anchor{powerpc-x-x}powerpc-*-*
3821 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
3822 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
3825 @uref{ftp://ftp.kernel.org/pub/linux/devel/binutils,,binutils 2.15}
3826 or newer for a working GCC@.
3831 @heading @anchor{powerpc-x-darwin}powerpc-*-darwin*
3832 PowerPC running Darwin (Mac OS X kernel).
3834 Pre-installed versions of Mac OS X may not include any developer tools,
3835 meaning that you will not be able to build GCC from source. Tool
3836 binaries are available at
3837 @uref{http://developer.apple.com/darwin/projects/compiler/} (free
3838 registration required).
3840 This version of GCC requires at least cctools-590.36. The
3841 cctools-590.36 package referenced from
3842 @uref{http://gcc.gnu.org/ml/gcc/2006-03/msg00507.html} will not work
3843 on systems older than 10.3.9 (aka darwin7.9.0).
3848 @heading @anchor{powerpc-x-elf}powerpc-*-elf
3849 PowerPC system in big endian mode, running System V.4.
3854 @heading @anchor{powerpc-x-linux-gnu}powerpc*-*-linux-gnu*
3856 PowerPC system in big endian mode running Linux.
3861 @heading @anchor{powerpc-x-netbsd}powerpc-*-netbsd*
3862 PowerPC system in big endian mode running NetBSD@.
3867 @heading @anchor{powerpc-x-eabisim}powerpc-*-eabisim
3868 Embedded PowerPC system in big endian mode for use in running under the
3874 @heading @anchor{powerpc-x-eabi}powerpc-*-eabi
3875 Embedded PowerPC system in big endian mode.
3880 @heading @anchor{powerpcle-x-elf}powerpcle-*-elf
3881 PowerPC system in little endian mode, running System V.4.
3886 @heading @anchor{powerpcle-x-eabisim}powerpcle-*-eabisim
3887 Embedded PowerPC system in little endian mode for use in running under
3893 @heading @anchor{powerpcle-x-eabi}powerpcle-*-eabi
3894 Embedded PowerPC system in little endian mode.
3899 @heading @anchor{s390-x-linux}s390-*-linux*
3900 S/390 system running GNU/Linux for S/390@.
3905 @heading @anchor{s390x-x-linux}s390x-*-linux*
3906 zSeries system (64-bit) running GNU/Linux for zSeries@.
3911 @heading @anchor{s390x-ibm-tpf}s390x-ibm-tpf*
3912 zSeries system (64-bit) running TPF@. This platform is
3913 supported as cross-compilation target only.
3918 @c Please use Solaris 2 to refer to all release of Solaris, starting
3919 @c with 2.0 until 2.6, 7, 8, etc. Solaris 1 was a marketing name for
3920 @c SunOS 4 releases which we don't use to avoid confusion. Solaris
3921 @c alone is too unspecific and must be avoided.
3922 @heading @anchor{x-x-solaris2}*-*-solaris2*
3924 Sun does not ship a C compiler with Solaris 2. To bootstrap and install
3925 GCC you first have to install a pre-built compiler, see the
3926 @uref{binaries.html,,binaries page} for details.
3928 The Solaris 2 @command{/bin/sh} will often fail to configure
3929 @file{libstdc++-v3}, @file{boehm-gc} or @file{libjava}. We therefore
3930 recommend using the following initial sequence of commands
3933 % CONFIG_SHELL=/bin/ksh
3934 % export CONFIG_SHELL
3937 and proceed as described in @uref{configure.html,,the configure instructions}.
3938 In addition we strongly recommend specifying an absolute path to invoke
3939 @var{srcdir}/configure.
3941 Solaris 2 comes with a number of optional OS packages. Some of these
3942 are needed to use GCC fully, namely @code{SUNWarc},
3943 @code{SUNWbtool}, @code{SUNWesu}, @code{SUNWhea}, @code{SUNWlibm},
3944 @code{SUNWsprot}, and @code{SUNWtoo}. If you did not install all
3945 optional packages when installing Solaris 2, you will need to verify that
3946 the packages that GCC needs are installed.
3948 To check whether an optional package is installed, use
3949 the @command{pkginfo} command. To add an optional package, use the
3950 @command{pkgadd} command. For further details, see the Solaris 2
3953 Trying to use the linker and other tools in
3954 @file{/usr/ucb} to install GCC has been observed to cause trouble.
3955 For example, the linker may hang indefinitely. The fix is to remove
3956 @file{/usr/ucb} from your @env{PATH}.
3958 The build process works more smoothly with the legacy Sun tools so, if you
3959 have @file{/usr/xpg4/bin} in your @env{PATH}, we recommend that you place
3960 @file{/usr/bin} before @file{/usr/xpg4/bin} for the duration of the build.
3962 We recommend the use of GNU binutils 2.14 or later, or the vendor tools
3963 (Sun @command{as}, Sun @command{ld}). Note that your mileage may vary
3964 if you use a combination of the GNU tools and the Sun tools: while the
3965 combination GNU @command{as} + Sun @command{ld} should reasonably work,
3966 the reverse combination Sun @command{as} + GNU @command{ld} is known to
3967 cause memory corruption at runtime in some cases for C++ programs.
3969 The stock GNU binutils 2.15 release is broken on this platform because of a
3970 single bug. It has been fixed on the 2.15 branch in the CVS repository.
3971 You can obtain a working version by checking out the binutils-2_15-branch
3972 from the CVS repository or applying the patch
3973 @uref{http://sourceware.org/ml/binutils-cvs/2004-09/msg00036.html} to the
3976 We recommend the use of GNU binutils 2.16 or later in conjunction with GCC
3977 4.x, or the vendor tools (Sun @command{as}, Sun @command{ld}). However,
3978 for Solaris 10 and above, an additional patch is required in order for the
3979 GNU linker to be able to cope with a new flavor of shared libraries. You
3980 can obtain a working version by checking out the binutils-2_16-branch from
3981 the CVS repository or applying the patch
3982 @uref{http://sourceware.org/ml/binutils-cvs/2005-07/msg00122.html} to the
3985 Sun bug 4296832 turns up when compiling X11 headers with GCC 2.95 or
3986 newer: @command{g++} will complain that types are missing. These headers
3987 assume that omitting the type means @code{int}; this assumption worked for
3988 C89 but is wrong for C++, and is now wrong for C99 also.
3990 @command{g++} accepts such (invalid) constructs with the option
3991 @option{-fpermissive}; it will assume that any missing type is @code{int}
3992 (as defined by C89).
3994 There are patches for Solaris 7 (108376-21 or newer for SPARC,
3995 108377-20 for Intel), and Solaris 8 (108652-24 or newer for SPARC,
3996 108653-22 for Intel) that fix this bug.
3998 Sun bug 4927647 sometimes causes random spurious testsuite failures
3999 related to missing diagnostic output. This bug doesn't affect GCC
4000 itself, rather it is a kernel bug triggered by the @command{expect}
4001 program which is used only by the GCC testsuite driver. When the bug
4002 causes the @command{expect} program to miss anticipated output, extra
4003 testsuite failures appear.
4005 There are patches for Solaris 8 (117350-12 or newer for SPARC,
4006 117351-12 or newer for Intel) and Solaris 9 (117171-11 or newer for
4007 SPARC, 117172-11 or newer for Intel) that address this problem.
4012 @heading @anchor{sparc-sun-solaris2}sparc-sun-solaris2*
4014 When GCC is configured to use binutils 2.14 or later the binaries
4015 produced are smaller than the ones produced using Sun's native tools;
4016 this difference is quite significant for binaries containing debugging
4019 Starting with Solaris 7, the operating system is capable of executing
4020 64-bit SPARC V9 binaries. GCC 3.1 and later properly supports
4021 this; the @option{-m64} option enables 64-bit code generation.
4022 However, if all you want is code tuned for the UltraSPARC CPU, you
4023 should try the @option{-mtune=ultrasparc} option instead, which produces
4024 code that, unlike full 64-bit code, can still run on non-UltraSPARC
4027 When configuring on a Solaris 7 or later system that is running a kernel
4028 that supports only 32-bit binaries, one must configure with
4029 @option{--disable-multilib}, since we will not be able to build the
4030 64-bit target libraries.
4032 GCC 3.3 and GCC 3.4 trigger code generation bugs in earlier versions of
4033 the GNU compiler (especially GCC 3.0.x versions), which lead to the
4034 miscompilation of the stage1 compiler and the subsequent failure of the
4035 bootstrap process. A workaround is to use GCC 3.2.3 as an intermediary
4036 stage, i.e.@: to bootstrap that compiler with the base compiler and then
4037 use it to bootstrap the final compiler.
4039 GCC 3.4 triggers a code generation bug in versions 5.4 (Sun ONE Studio 7)
4040 and 5.5 (Sun ONE Studio 8) of the Sun compiler, which causes a bootstrap
4041 failure in form of a miscompilation of the stage1 compiler by the Sun
4042 compiler. This is Sun bug 4974440. This is fixed with patch 112760-07.
4044 GCC 3.4 changed the default debugging format from STABS to DWARF-2 for
4045 32-bit code on Solaris 7 and later. If you use the Sun assembler, this
4046 change apparently runs afoul of Sun bug 4910101 (which is referenced as
4047 an x86-only problem by Sun, probably because they do not use DWARF-2).
4048 A symptom of the problem is that you cannot compile C++ programs like
4049 @command{groff} 1.19.1 without getting messages similar to the following:
4052 ld: warning: relocation error: R_SPARC_UA32: @dots{}
4053 external symbolic relocation against non-allocatable section
4054 .debug_info cannot be processed at runtime: relocation ignored.
4057 To work around this problem, compile with @option{-gstabs+} instead of
4060 When configuring the GNU Multiple Precision Library (GMP) or the MPFR
4061 library on a Solaris 7 or later system, the canonical target triplet
4062 must be specified as the @command{build} parameter on the configure
4063 line. This triplet can be obtained by invoking ./config.guess in
4064 the toplevel source directory of GCC (and not that of GMP or MPFR).
4065 For example on a Solaris 7 system:
4068 % ./configure --build=sparc-sun-solaris2.7 --prefix=xxx
4074 @heading @anchor{sparc-sun-solaris27}sparc-sun-solaris2.7
4076 Sun patch 107058-01 (1999-01-13) for Solaris 7/SPARC triggers a bug in
4077 the dynamic linker. This problem (Sun bug 4210064) affects GCC 2.8
4078 and later, including all EGCS releases. Sun formerly recommended
4079 107058-01 for all Solaris 7 users, but around 1999-09-01 it started to
4080 recommend it only for people who use Sun's compilers.
4082 Here are some workarounds to this problem:
4085 Do not install Sun patch 107058-01 until after Sun releases a
4086 complete patch for bug 4210064. This is the simplest course to take,
4087 unless you must also use Sun's C compiler. Unfortunately 107058-01
4088 is preinstalled on some new Solaris 7-based hosts, so you may have to
4092 Copy the original, unpatched Solaris 7
4093 @command{/usr/ccs/bin/as} into
4094 @command{/usr/local/libexec/gcc/sparc-sun-solaris2.7/3.4/as},
4095 adjusting the latter name to fit your local conventions and software
4099 Install Sun patch 106950-03 (1999-05-25) or later. Nobody with
4100 both 107058-01 and 106950-03 installed has reported the bug with GCC
4101 and Sun's dynamic linker. This last course of action is riskiest,
4102 for two reasons. First, you must install 106950 on all hosts that
4103 run code generated by GCC; it doesn't suffice to install it only on
4104 the hosts that run GCC itself. Second, Sun says that 106950-03 is
4105 only a partial fix for bug 4210064, but Sun doesn't know whether the
4106 partial fix is adequate for GCC@. Revision -08 or later should fix
4107 the bug. The current (as of 2004-05-23) revision is -24, and is included in
4108 the Solaris 7 Recommended Patch Cluster.
4111 GCC 3.3 triggers a bug in version 5.0 Alpha 03/27/98 of the Sun assembler,
4112 which causes a bootstrap failure when linking the 64-bit shared version of
4113 libgcc. A typical error message is:
4116 ld: fatal: relocation error: R_SPARC_32: file libgcc/sparcv9/_muldi3.o:
4117 symbol <unknown>: offset 0xffffffff7ec133e7 is non-aligned.
4120 This bug has been fixed in the final 5.0 version of the assembler.
4122 A similar problem was reported for version Sun WorkShop 6 99/08/18 of the
4123 Sun assembler, which causes a bootstrap failure with GCC 4.0.0:
4126 ld: fatal: relocation error: R_SPARC_DISP32:
4127 file .libs/libstdc++.lax/libsupc++convenience.a/vterminate.o:
4128 symbol <unknown>: offset 0xfccd33ad is non-aligned
4131 This bug has been fixed in more recent revisions of the assembler.
4136 @heading @anchor{sparc-x-linux}sparc-*-linux*
4138 GCC versions 3.0 and higher require binutils 2.11.2 and glibc 2.2.4
4139 or newer on this platform. All earlier binutils and glibc
4140 releases mishandled unaligned relocations on @code{sparc-*-*} targets.
4146 @heading @anchor{sparc64-x-solaris2}sparc64-*-solaris2*
4148 When configuring the GNU Multiple Precision Library (GMP) or the
4149 MPFR library, the canonical target triplet must be specified as
4150 the @command{build} parameter on the configure line. For example
4151 on a Solaris 7 system:
4154 % ./configure --build=sparc64-sun-solaris2.7 --prefix=xxx
4157 The following compiler flags must be specified in the configure
4158 step in order to bootstrap this target with the Sun compiler:
4161 % CC="cc -xarch=v9 -xildoff" @var{srcdir}/configure [@var{options}] [@var{target}]
4164 @option{-xarch=v9} specifies the SPARC-V9 architecture to the Sun toolchain
4165 and @option{-xildoff} turns off the incremental linker.
4170 @heading @anchor{sparcv9-x-solaris2}sparcv9-*-solaris2*
4172 This is a synonym for sparc64-*-solaris2*.
4177 @heading @anchor{x-x-vxworks}*-*-vxworks*
4178 Support for VxWorks is in flux. At present GCC supports @emph{only} the
4179 very recent VxWorks 5.5 (aka Tornado 2.2) release, and only on PowerPC@.
4180 We welcome patches for other architectures supported by VxWorks 5.5.
4181 Support for VxWorks AE would also be welcome; we believe this is merely
4182 a matter of writing an appropriate ``configlette'' (see below). We are
4183 not interested in supporting older, a.out or COFF-based, versions of
4186 VxWorks comes with an older version of GCC installed in
4187 @file{@var{$WIND_BASE}/host}; we recommend you do not overwrite it.
4188 Choose an installation @var{prefix} entirely outside @var{$WIND_BASE}.
4189 Before running @command{configure}, create the directories @file{@var{prefix}}
4190 and @file{@var{prefix}/bin}. Link or copy the appropriate assembler,
4191 linker, etc.@: into @file{@var{prefix}/bin}, and set your @var{PATH} to
4192 include that directory while running both @command{configure} and
4195 You must give @command{configure} the
4196 @option{--with-headers=@var{$WIND_BASE}/target/h} switch so that it can
4197 find the VxWorks system headers. Since VxWorks is a cross compilation
4198 target only, you must also specify @option{--target=@var{target}}.
4199 @command{configure} will attempt to create the directory
4200 @file{@var{prefix}/@var{target}/sys-include} and copy files into it;
4201 make sure the user running @command{configure} has sufficient privilege
4204 GCC's exception handling runtime requires a special ``configlette''
4205 module, @file{contrib/gthr_supp_vxw_5x.c}. Follow the instructions in
4206 that file to add the module to your kernel build. (Future versions of
4207 VxWorks will incorporate this module.)
4212 @heading @anchor{x86-64-x-x}x86_64-*-*, amd64-*-*
4214 GCC supports the x86-64 architecture implemented by the AMD64 processor
4215 (amd64-*-* is an alias for x86_64-*-*) on GNU/Linux, FreeBSD and NetBSD@.
4216 On GNU/Linux the default is a bi-arch compiler which is able to generate
4217 both 64-bit x86-64 and 32-bit x86 code (via the @option{-m32} switch).
4222 @heading @anchor{xtensa-x-elf}xtensa*-*-elf
4224 This target is intended for embedded Xtensa systems using the
4225 @samp{newlib} C library. It uses ELF but does not support shared
4226 objects. Designed-defined instructions specified via the
4227 Tensilica Instruction Extension (TIE) language are only supported
4228 through inline assembly.
4230 The Xtensa configuration information must be specified prior to
4231 building GCC@. The @file{include/xtensa-config.h} header
4232 file contains the configuration information. If you created your
4233 own Xtensa configuration with the Xtensa Processor Generator, the
4234 downloaded files include a customized copy of this header file,
4235 which you can use to replace the default header file.
4240 @heading @anchor{xtensa-x-linux}xtensa*-*-linux*
4242 This target is for Xtensa systems running GNU/Linux. It supports ELF
4243 shared objects and the GNU C library (glibc). It also generates
4244 position-independent code (PIC) regardless of whether the
4245 @option{-fpic} or @option{-fPIC} options are used. In other
4246 respects, this target is the same as the
4247 @uref{#xtensa*-*-elf,,@samp{xtensa*-*-elf}} target.
4252 @heading @anchor{windows}Microsoft Windows
4254 @subheading Intel 16-bit versions
4255 The 16-bit versions of Microsoft Windows, such as Windows 3.1, are not
4258 However, the 32-bit port has limited support for Microsoft
4259 Windows 3.11 in the Win32s environment, as a target only. See below.
4261 @subheading Intel 32-bit versions
4263 The 32-bit versions of Windows, including Windows 95, Windows NT, Windows
4264 XP, and Windows Vista, are supported by several different target
4265 platforms. These targets differ in which Windows subsystem they target
4266 and which C libraries are used.
4269 @item Cygwin @uref{#x-x-cygwin,,*-*-cygwin}: Cygwin provides a user-space
4270 Linux API emulation layer in the Win32 subsystem.
4271 @item Interix @uref{#x-x-interix,,*-*-interix}: The Interix subsystem
4272 provides native support for POSIX.
4273 @item MinGW @uref{#x-x-mingw32,,*-*-mingw32}: MinGW is a native GCC port for
4274 the Win32 subsystem that provides a subset of POSIX.
4275 @item MKS i386-pc-mks: NuTCracker from MKS. See
4276 @uref{http://www.mkssoftware.com/} for more information.
4279 @subheading Intel 64-bit versions
4281 GCC contains support for x86-64 using the mingw-w64
4282 runtime library, available from @uref{http://mingw-w64.sourceforge.net/}.
4283 This library should be used with the target triple x86_64-pc-mingw32.
4285 Presently Windows for Itanium is not supported.
4287 @subheading Windows CE
4289 Windows CE is supported as a target only on ARM (arm-wince-pe), Hitachi
4290 SuperH (sh-wince-pe), and MIPS (mips-wince-pe).
4292 @subheading Other Windows Platforms
4294 GCC no longer supports Windows NT on the Alpha or PowerPC.
4296 GCC no longer supports the Windows POSIX subsystem. However, it does
4297 support the Interix subsystem. See above.
4299 Old target names including *-*-winnt and *-*-windowsnt are no longer used.
4301 PW32 (i386-pc-pw32) support was never completed, and the project seems to
4302 be inactive. See @uref{http://pw32.sourceforge.net/} for more information.
4304 UWIN support has been removed due to a lack of maintenance.
4309 @heading @anchor{x-x-cygwin}*-*-cygwin
4311 Ports of GCC are included with the
4312 @uref{http://www.cygwin.com/,,Cygwin environment}.
4314 GCC will build under Cygwin without modification; it does not build
4315 with Microsoft's C++ compiler and there are no plans to make it do so.
4317 Cygwin can be compiled with i?86-pc-cygwin.
4322 @heading @anchor{x-x-interix}*-*-interix
4324 The Interix target is used by OpenNT, Interix, Services For UNIX (SFU),
4325 and Subsystem for UNIX-based Applications (SUA). Applications compiled
4326 with this target run in the Interix subsystem, which is separate from
4327 the Win32 subsystem. This target was last known to work in GCC 3.3.
4329 For more information, see @uref{http://www.interix.com/}.
4334 @heading @anchor{x-x-mingw32}*-*-mingw32
4336 GCC will build with and support only MinGW runtime 3.12 and later.
4337 Earlier versions of headers are incompatible with the new default semantics
4338 of @code{extern inline} in @code{-std=c99} and @code{-std=gnu99} modes.
4343 @heading @anchor{os2}OS/2
4345 GCC does not currently support OS/2. However, Andrew Zabolotny has been
4346 working on a generic OS/2 port with pgcc. The current code can be found
4347 at @uref{http://www.goof.com/pcg/os2/,,http://www.goof.com/pcg/os2/}.
4352 @heading @anchor{older}Older systems
4354 GCC contains support files for many older (1980s and early
4355 1990s) Unix variants. For the most part, support for these systems
4356 has not been deliberately removed, but it has not been maintained for
4357 several years and may suffer from bitrot.
4359 Starting with GCC 3.1, each release has a list of ``obsoleted'' systems.
4360 Support for these systems is still present in that release, but
4361 @command{configure} will fail unless the @option{--enable-obsolete}
4362 option is given. Unless a maintainer steps forward, support for these
4363 systems will be removed from the next release of GCC@.
4365 Support for old systems as hosts for GCC can cause problems if the
4366 workarounds for compiler, library and operating system bugs affect the
4367 cleanliness or maintainability of the rest of GCC@. In some cases, to
4368 bring GCC up on such a system, if still possible with current GCC, may
4369 require first installing an old version of GCC which did work on that
4370 system, and using it to compile a more recent GCC, to avoid bugs in the
4371 vendor compiler. Old releases of GCC 1 and GCC 2 are available in the
4372 @file{old-releases} directory on the @uref{../mirrors.html,,GCC mirror
4373 sites}. Header bugs may generally be avoided using
4374 @command{fixincludes}, but bugs or deficiencies in libraries and the
4375 operating system may still cause problems.
4377 Support for older systems as targets for cross-compilation is less
4378 problematic than support for them as hosts for GCC; if an enthusiast
4379 wishes to make such a target work again (including resurrecting any of
4380 the targets that never worked with GCC 2, starting from the last
4381 version before they were removed), patches
4382 @uref{../contribute.html,,following the usual requirements} would be
4383 likely to be accepted, since they should not affect the support for more
4386 For some systems, old versions of GNU binutils may also be useful,
4387 and are available from @file{pub/binutils/old-releases} on
4388 @uref{http://sourceware.org/mirrors.html,,sourceware.org mirror sites}.
4390 Some of the information on specific systems above relates to
4391 such older systems, but much of the information
4392 about GCC on such systems (which may no longer be applicable to
4393 current GCC) is to be found in the GCC texinfo manual.
4398 @heading @anchor{elf}all ELF targets (SVR4, Solaris 2, etc.)
4400 C++ support is significantly better on ELF targets if you use the
4401 @uref{./configure.html#with-gnu-ld,,GNU linker}; duplicate copies of
4402 inlines, vtables and template instantiations will be discarded
4411 @uref{./index.html,,Return to the GCC Installation page}
4415 @c ***Old documentation******************************************************
4417 @include install-old.texi
4423 @uref{./index.html,,Return to the GCC Installation page}
4427 @c ***GFDL********************************************************************
4435 @uref{./index.html,,Return to the GCC Installation page}
4439 @c ***************************************************************************
4440 @c Part 6 The End of the Document
4442 @comment node-name, next, previous, up
4443 @node Concept Index, , GNU Free Documentation License, Top
4447 @unnumbered Concept Index