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,
49 @c 2010 Free Software Foundation, Inc.
50 @c *** Converted to texinfo by Dean Wakerley, dean@wakerley.com
52 @c IMPORTANT: whenever you modify this file, run `install.texi2html' to
53 @c test the generation of HTML documents for the gcc.gnu.org web pages.
55 @c Do not use @footnote{} in this file as it breaks install.texi2html!
57 @c Include everything if we're not making html
61 @set prerequisiteshtml
72 @c Part 2 Summary Description and Copyright
74 Copyright @copyright{} 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997,
75 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007,
76 2008 Free Software Foundation, Inc.
78 Permission is granted to copy, distribute and/or modify this document
79 under the terms of the GNU Free Documentation License, Version 1.2 or
80 any later version published by the Free Software Foundation; with no
81 Invariant Sections, the Front-Cover texts being (a) (see below), and
82 with the Back-Cover Texts being (b) (see below). A copy of the
83 license is included in the section entitled ``@uref{./gfdl.html,,GNU
84 Free Documentation License}''.
86 (a) The FSF's Front-Cover Text is:
90 (b) The FSF's Back-Cover Text is:
92 You have freedom to copy and modify this GNU Manual, like GNU
93 software. Copies published by the Free Software Foundation raise
94 funds for GNU development.
99 @dircategory Software development
101 * gccinstall: (gccinstall). Installing the GNU Compiler Collection.
104 @c Part 3 Titlepage and Copyright
106 @title Installing GCC
109 @c The following two commands start the copyright page.
111 @vskip 0pt plus 1filll
115 @c Part 4 Top node, Master Menu, and/or Table of Contents
118 @comment node-name, next, Previous, up
121 * Installing GCC:: This document describes the generic installation
122 procedure for GCC as well as detailing some target
123 specific installation instructions.
125 * Specific:: Host/target specific installation notes for GCC.
126 * Binaries:: Where to get pre-compiled binaries.
128 * Old:: Old installation documentation.
130 * GNU Free Documentation License:: How you can copy and share this manual.
131 * Concept Index:: This index has two entries.
139 @c Part 5 The Body of the Document
140 @c ***Installing GCC**********************************************************
142 @comment node-name, next, previous, up
143 @node Installing GCC, Binaries, , Top
147 @chapter Installing GCC
150 The latest version of this document is always available at
151 @uref{http://gcc.gnu.org/install/,,http://gcc.gnu.org/install/}.
153 This document describes the generic installation procedure for GCC as well
154 as detailing some target specific installation instructions.
156 GCC includes several components that previously were separate distributions
157 with their own installation instructions. This document supersedes all
158 package specific installation instructions.
160 @emph{Before} starting the build/install procedure please check the
162 @ref{Specific, host/target specific installation notes}.
165 @uref{specific.html,,host/target specific installation notes}.
167 We recommend you browse the entire generic installation instructions before
170 Lists of successful builds for released versions of GCC are
171 available at @uref{http://gcc.gnu.org/buildstat.html}.
172 These lists are updated as new information becomes available.
174 The installation procedure itself is broken into five steps.
179 * Downloading the source::
182 * Testing:: (optional)
189 @uref{prerequisites.html,,Prerequisites}
191 @uref{download.html,,Downloading the source}
193 @uref{configure.html,,Configuration}
195 @uref{build.html,,Building}
197 @uref{test.html,,Testing} (optional)
199 @uref{finalinstall.html,,Final install}
203 Please note that GCC does not support @samp{make uninstall} and probably
204 won't do so in the near future as this would open a can of worms. Instead,
205 we suggest that you install GCC into a directory of its own and simply
206 remove that directory when you do not need that specific version of GCC
207 any longer, and, if shared libraries are installed there as well, no
208 more binaries exist that use them.
211 There are also some @uref{old.html,,old installation instructions},
212 which are mostly obsolete but still contain some information which has
213 not yet been merged into the main part of this manual.
221 @uref{./index.html,,Return to the GCC Installation page}
227 @c ***Prerequisites**************************************************
229 @comment node-name, next, previous, up
230 @node Prerequisites, Downloading the source, , Installing GCC
232 @ifset prerequisiteshtml
234 @chapter Prerequisites
236 @cindex Prerequisites
238 GCC requires that various tools and packages be available for use in the
239 build procedure. Modifying GCC sources requires additional tools
242 @heading Tools/packages necessary for building GCC
244 @item ISO C90 compiler
245 Necessary to bootstrap GCC, although versions of GCC prior
246 to 3.4 also allow bootstrapping with a traditional (K&R) C compiler.
248 To build all languages in a cross-compiler or other configuration where
249 3-stage bootstrap is not performed, you need to start with an existing
250 GCC binary (version 2.95 or later) because source code for language
251 frontends other than C might use GCC extensions.
255 In order to build the Ada compiler (GNAT) you must already have GNAT
256 installed because portions of the Ada frontend are written in Ada (with
257 GNAT extensions.) Refer to the Ada installation instructions for more
258 specific information.
260 @item A ``working'' POSIX compatible shell, or GNU bash
262 Necessary when running @command{configure} because some
263 @command{/bin/sh} shells have bugs and may crash when configuring the
264 target libraries. In other cases, @command{/bin/sh} or @command{ksh}
265 have disastrous corner-case performance problems. This
266 can cause target @command{configure} runs to literally take days to
267 complete in some cases.
269 So on some platforms @command{/bin/ksh} is sufficient, on others it
270 isn't. See the host/target specific instructions for your platform, or
271 use @command{bash} to be sure. Then set @env{CONFIG_SHELL} in your
272 environment to your ``good'' shell prior to running
273 @command{configure}/@command{make}.
275 @command{zsh} is not a fully compliant POSIX shell and will not
276 work when configuring GCC@.
278 @item A POSIX or SVR4 awk
280 Necessary for creating some of the generated source files for GCC@.
281 If in doubt, use a recent GNU awk version, as some of the older ones
282 are broken. GNU awk version 3.1.5 is known to work.
286 Necessary in some circumstances, optional in others. See the
287 host/target specific instructions for your platform for the exact
290 @item gzip version 1.2.4 (or later) or
291 @itemx bzip2 version 1.0.2 (or later)
293 Necessary to uncompress GCC @command{tar} files when source code is
294 obtained via FTP mirror sites.
296 @item GNU make version 3.80 (or later)
298 You must have GNU make installed to build GCC@.
300 @item GNU tar version 1.14 (or later)
302 Necessary (only on some platforms) to untar the source code. Many
303 systems' @command{tar} programs will also work, only try GNU
304 @command{tar} if you have problems.
306 @item GNU Multiple Precision Library (GMP) version 4.2 (or later)
308 Necessary to build GCC@. If you do not have it installed in your
309 library search path, you will have to configure with the
310 @option{--with-gmp} configure option. See also @option{--with-gmp-lib}
311 and @option{--with-gmp-include}. Alternatively, if a GMP source
312 distribution is found in a subdirectory of your GCC sources named
313 @file{gmp}, it will be built together with GCC@.
315 @item MPFR Library version 2.3.2 (or later)
317 Necessary to build GCC@. It can be downloaded from
318 @uref{http://www.mpfr.org/}. The @option{--with-mpfr} configure
319 option should be used if your MPFR Library is not installed in your
320 default library search path. See also @option{--with-mpfr-lib} and
321 @option{--with-mpfr-include}. Alternatively, if a MPFR source
322 distribution is found in a subdirectory of your GCC sources named
323 @file{mpfr}, it will be built together with GCC@.
325 @item MPC Library version 0.8.0 (or later)
327 Necessary to build GCC@. It can be downloaded from
328 @uref{http://www.multiprecision.org/}. The @option{--with-mpc}
329 configure option should be used if your MPC Library is not installed
330 in your default library search path. See also @option{--with-mpc-lib}
331 and @option{--with-mpc-include}. Alternatively, if an MPC source
332 distribution is found in a subdirectory of your GCC sources named
333 @file{mpc}, it will be built together with GCC@.
335 @item Parma Polyhedra Library (PPL) version 0.10
337 Necessary to build GCC with the Graphite loop optimizations.
338 It can be downloaded from @uref{http://www.cs.unipr.it/ppl/Download/}.
340 The @option{--with-ppl} configure option should be used if PPL is not
341 installed in your default library search path.
343 @item CLooG-PPL version 0.15
345 Necessary to build GCC with the Graphite loop optimizations. It can
346 be downloaded from @uref{ftp://gcc.gnu.org/pub/gcc/infrastructure/}.
347 The code in @file{cloog-ppl-0.15.tar.gz} comes from a branch of CLooG
348 available from @uref{http://repo.or.cz/w/cloog-ppl.git}. CLooG-PPL
349 should be configured with @option{--with-ppl}.
351 The @option{--with-cloog} configure option should be used if CLooG is
352 not installed in your default library search path.
354 @item @command{jar}, or InfoZIP (@command{zip} and @command{unzip})
356 Necessary to build libgcj, the GCJ runtime.
358 @item libelf version 0.8.12 (or later)
360 Necessary to build link-time optimization (LTO) support. It can be
361 downloaded from @uref{http://www.mr511.de/software/libelf-0.8.12.tar.gz},
362 though it is commonly available in several systems.
364 The @option{--with-libelf} configure option should be used if libelf is
365 not installed in your default library search patch.
369 @heading Tools/packages necessary for modifying GCC
371 @item autoconf version 2.64
372 @itemx GNU m4 version 1.4.6 (or later)
374 Necessary when modifying @file{configure.ac}, @file{aclocal.m4}, etc.@:
375 to regenerate @file{configure} and @file{config.in} files.
377 @item automake version 1.11
379 Necessary when modifying a @file{Makefile.am} file to regenerate its
380 associated @file{Makefile.in}.
382 Much of GCC does not use automake, so directly edit the @file{Makefile.in}
383 file. Specifically this applies to the @file{gcc}, @file{intl},
384 @file{libcpp}, @file{libiberty}, @file{libobjc} directories as well
385 as any of their subdirectories.
387 For directories that use automake, GCC requires the latest release in
388 the 1.11 series, which is currently 1.11. When regenerating a directory
389 to a newer version, please update all the directories using an older 1.11
390 to the latest released version.
392 @item gettext version 0.14.5 (or later)
394 Needed to regenerate @file{gcc.pot}.
396 @item gperf version 2.7.2 (or later)
398 Necessary when modifying @command{gperf} input files, e.g.@:
399 @file{gcc/cp/cfns.gperf} to regenerate its associated header file, e.g.@:
400 @file{gcc/cp/cfns.h}.
406 Necessary to run the GCC testsuite; see the section on testing for details.
408 @item autogen version 5.5.4 (or later) and
409 @itemx guile version 1.4.1 (or later)
411 Necessary to regenerate @file{fixinc/fixincl.x} from
412 @file{fixinc/inclhack.def} and @file{fixinc/*.tpl}.
414 Necessary to run @samp{make check} for @file{fixinc}.
416 Necessary to regenerate the top level @file{Makefile.in} file from
417 @file{Makefile.tpl} and @file{Makefile.def}.
419 @item Flex version 2.5.4 (or later)
421 Necessary when modifying @file{*.l} files.
423 Necessary to build GCC during development because the generated output
424 files are not included in the SVN repository. They are included in
427 @item Texinfo version 4.7 (or later)
429 Necessary for running @command{makeinfo} when modifying @file{*.texi}
430 files to test your changes.
432 Necessary for running @command{make dvi} or @command{make pdf} to
433 create printable documentation in DVI or PDF format. Texinfo version
434 4.8 or later is required for @command{make pdf}.
436 Necessary to build GCC documentation during development because the
437 generated output files are not included in the SVN repository. They are
438 included in releases.
440 @item @TeX{} (any working version)
442 Necessary for running @command{texi2dvi} and @command{texi2pdf}, which
443 are used when running @command{make dvi} or @command{make pdf} to create
444 DVI or PDF files, respectively.
446 @item SVN (any version)
447 @itemx SSH (any version)
449 Necessary to access the SVN repository. Public releases and weekly
450 snapshots of the development sources are also available via FTP@.
452 @item Perl version 5.6.1 (or later)
454 Necessary when regenerating @file{Makefile} dependencies in libiberty.
455 Necessary when regenerating @file{libiberty/functions.texi}.
456 Necessary when generating manpages from Texinfo manuals.
457 Necessary when targetting Darwin, building libstdc++,
458 and not using @option{--disable-symvers}.
459 Used by various scripts to generate some files included in SVN (mainly
460 Unicode-related and rarely changing) from source tables.
462 @item GNU diffutils version 2.7 (or later)
464 Useful when submitting patches for the GCC source code.
466 @item patch version 2.5.4 (or later)
468 Necessary when applying patches, created with @command{diff}, to one's
474 If you wish to modify @file{.java} files in libjava, you will need to
475 configure with @option{--enable-java-maintainer-mode}, and you will need
476 to have executables named @command{ecj1} and @command{gjavah} in your path.
477 The @command{ecj1} executable should run the Eclipse Java compiler via
478 the GCC-specific entry point. You can download a suitable jar from
479 @uref{ftp://sourceware.org/pub/java/}, or by running the script
480 @command{contrib/download_ecj}.
482 @item antlr.jar version 2.7.1 (or later)
485 If you wish to build the @command{gjdoc} binary in libjava, you will
486 need to have an @file{antlr.jar} library available. The library is
487 searched in system locations but can be configured with
488 @option{--with-antlr-jar=} instead. When configuring with
489 @option{--enable-java-maintainer-mode}, you will need to have one of
490 the executables named @command{cantlr}, @command{runantlr} or
491 @command{antlr} in your path.
500 @uref{./index.html,,Return to the GCC Installation page}
504 @c ***Downloading the source**************************************************
506 @comment node-name, next, previous, up
507 @node Downloading the source, Configuration, Prerequisites, Installing GCC
511 @chapter Downloading GCC
513 @cindex Downloading GCC
514 @cindex Downloading the Source
516 GCC is distributed via @uref{http://gcc.gnu.org/svn.html,,SVN} and FTP
517 tarballs compressed with @command{gzip} or
518 @command{bzip2}. It is possible to download a full distribution or specific
521 Please refer to the @uref{http://gcc.gnu.org/releases.html,,releases web page}
522 for information on how to obtain GCC@.
524 The full distribution includes the C, C++, Objective-C, Fortran, Java,
525 and Ada (in the case of GCC 3.1 and later) compilers. The full
526 distribution also includes runtime libraries for C++, Objective-C,
527 Fortran, and Java. In GCC 3.0 and later versions, the GNU compiler
528 testsuites are also included in the full distribution.
530 If you choose to download specific components, you must download the core
531 GCC distribution plus any language specific distributions you wish to
532 use. The core distribution includes the C language front end as well as the
533 shared components. Each language has a tarball which includes the language
534 front end as well as the language runtime (when appropriate).
536 Unpack the core distribution as well as any language specific
537 distributions in the same directory.
539 If you also intend to build binutils (either to upgrade an existing
540 installation or for use in place of the corresponding tools of your
541 OS), unpack the binutils distribution either in the same directory or
542 a separate one. In the latter case, add symbolic links to any
543 components of the binutils you intend to build alongside the compiler
544 (@file{bfd}, @file{binutils}, @file{gas}, @file{gprof}, @file{ld},
545 @file{opcodes}, @dots{}) to the directory containing the GCC sources.
547 Likewise the GMP, MPFR and MPC libraries can be automatically built
548 together with GCC. Unpack the GMP, MPFR and/or MPC source
549 distributions in the directory containing the GCC sources and rename
550 their directories to @file{gmp}, @file{mpfr} and @file{mpc},
551 respectively (or use symbolic links with the same name).
558 @uref{./index.html,,Return to the GCC Installation page}
562 @c ***Configuration***********************************************************
564 @comment node-name, next, previous, up
565 @node Configuration, Building, Downloading the source, Installing GCC
569 @chapter Installing GCC: Configuration
571 @cindex Configuration
572 @cindex Installing GCC: Configuration
574 Like most GNU software, GCC must be configured before it can be built.
575 This document describes the recommended configuration procedure
576 for both native and cross targets.
578 We use @var{srcdir} to refer to the toplevel source directory for
579 GCC; we use @var{objdir} to refer to the toplevel build/object directory.
581 If you obtained the sources via SVN, @var{srcdir} must refer to the top
582 @file{gcc} directory, the one where the @file{MAINTAINERS} file can be
583 found, and not its @file{gcc} subdirectory, otherwise the build will fail.
585 If either @var{srcdir} or @var{objdir} is located on an automounted NFS
586 file system, the shell's built-in @command{pwd} command will return
587 temporary pathnames. Using these can lead to various sorts of build
588 problems. To avoid this issue, set the @env{PWDCMD} environment
589 variable to an automounter-aware @command{pwd} command, e.g.,
590 @command{pawd} or @samp{amq -w}, during the configuration and build
593 First, we @strong{highly} recommend that GCC be built into a
594 separate directory from the sources which does @strong{not} reside
595 within the source tree. This is how we generally build GCC; building
596 where @var{srcdir} == @var{objdir} should still work, but doesn't
597 get extensive testing; building where @var{objdir} is a subdirectory
598 of @var{srcdir} is unsupported.
600 If you have previously built GCC in the same directory for a
601 different target machine, do @samp{make distclean} to delete all files
602 that might be invalid. One of the files this deletes is @file{Makefile};
603 if @samp{make distclean} complains that @file{Makefile} does not exist
604 or issues a message like ``don't know how to make distclean'' it probably
605 means that the directory is already suitably clean. However, with the
606 recommended method of building in a separate @var{objdir}, you should
607 simply use a different @var{objdir} for each target.
609 Second, when configuring a native system, either @command{cc} or
610 @command{gcc} must be in your path or you must set @env{CC} in
611 your environment before running configure. Otherwise the configuration
615 Note that the bootstrap compiler and the resulting GCC must be link
616 compatible, else the bootstrap will fail with linker errors about
617 incompatible object file formats. Several multilibed targets are
618 affected by this requirement, see
620 @ref{Specific, host/target specific installation notes}.
623 @uref{specific.html,,host/target specific installation notes}.
632 % @var{srcdir}/configure [@var{options}] [@var{target}]
635 @heading Distributor options
637 If you will be distributing binary versions of GCC, with modifications
638 to the source code, you should use the options described in this
639 section to make clear that your version contains modifications.
642 @item --with-pkgversion=@var{version}
643 Specify a string that identifies your package. You may wish
644 to include a build number or build date. This version string will be
645 included in the output of @command{gcc --version}. This suffix does
646 not replace the default version string, only the @samp{GCC} part.
648 The default value is @samp{GCC}.
650 @item --with-bugurl=@var{url}
651 Specify the URL that users should visit if they wish to report a bug.
652 You are of course welcome to forward bugs reported to you to the FSF,
653 if you determine that they are not bugs in your modifications.
655 The default value refers to the FSF's GCC bug tracker.
659 @heading Target specification
662 GCC has code to correctly determine the correct value for @var{target}
663 for nearly all native systems. Therefore, we highly recommend you do
664 not provide a configure target when configuring a native compiler.
667 @var{target} must be specified as @option{--target=@var{target}}
668 when configuring a cross compiler; examples of valid targets would be
669 m68k-elf, sh-elf, etc.
672 Specifying just @var{target} instead of @option{--target=@var{target}}
673 implies that the host defaults to @var{target}.
677 @heading Options specification
679 Use @var{options} to override several configure time options for
680 GCC@. A list of supported @var{options} follows; @samp{configure
681 --help} may list other options, but those not listed below may not
682 work and should not normally be used.
684 Note that each @option{--enable} option has a corresponding
685 @option{--disable} option and that each @option{--with} option has a
686 corresponding @option{--without} option.
689 @item --prefix=@var{dirname}
690 Specify the toplevel installation
691 directory. This is the recommended way to install the tools into a directory
692 other than the default. The toplevel installation directory defaults to
695 We @strong{highly} recommend against @var{dirname} being the same or a
696 subdirectory of @var{objdir} or vice versa. If specifying a directory
697 beneath a user's home directory tree, some shells will not expand
698 @var{dirname} correctly if it contains the @samp{~} metacharacter; use
701 The following standard @command{autoconf} options are supported. Normally you
702 should not need to use these options.
704 @item --exec-prefix=@var{dirname}
705 Specify the toplevel installation directory for architecture-dependent
706 files. The default is @file{@var{prefix}}.
708 @item --bindir=@var{dirname}
709 Specify the installation directory for the executables called by users
710 (such as @command{gcc} and @command{g++}). The default is
711 @file{@var{exec-prefix}/bin}.
713 @item --libdir=@var{dirname}
714 Specify the installation directory for object code libraries and
715 internal data files of GCC@. The default is @file{@var{exec-prefix}/lib}.
717 @item --libexecdir=@var{dirname}
718 Specify the installation directory for internal executables of GCC@.
719 The default is @file{@var{exec-prefix}/libexec}.
721 @item --with-slibdir=@var{dirname}
722 Specify the installation directory for the shared libgcc library. The
723 default is @file{@var{libdir}}.
725 @item --datarootdir=@var{dirname}
726 Specify the root of the directory tree for read-only architecture-independent
727 data files referenced by GCC@. The default is @file{@var{prefix}/share}.
729 @item --infodir=@var{dirname}
730 Specify the installation directory for documentation in info format.
731 The default is @file{@var{datarootdir}/info}.
733 @item --datadir=@var{dirname}
734 Specify the installation directory for some architecture-independent
735 data files referenced by GCC@. The default is @file{@var{datarootdir}}.
737 @item --docdir=@var{dirname}
738 Specify the installation directory for documentation files (other
739 than Info) for GCC@. The default is @file{@var{datarootdir}/doc}.
741 @item --htmldir=@var{dirname}
742 Specify the installation directory for HTML documentation files.
743 The default is @file{@var{docdir}}.
745 @item --pdfdir=@var{dirname}
746 Specify the installation directory for PDF documentation files.
747 The default is @file{@var{docdir}}.
749 @item --mandir=@var{dirname}
750 Specify the installation directory for manual pages. The default is
751 @file{@var{datarootdir}/man}. (Note that the manual pages are only extracts
752 from the full GCC manuals, which are provided in Texinfo format. The manpages
753 are derived by an automatic conversion process from parts of the full
756 @item --with-gxx-include-dir=@var{dirname}
758 the installation directory for G++ header files. The default depends
759 on other configuration options, and differs between cross and native
764 @item --program-prefix=@var{prefix}
765 GCC supports some transformations of the names of its programs when
766 installing them. This option prepends @var{prefix} to the names of
767 programs to install in @var{bindir} (see above). For example, specifying
768 @option{--program-prefix=foo-} would result in @samp{gcc}
769 being installed as @file{/usr/local/bin/foo-gcc}.
771 @item --program-suffix=@var{suffix}
772 Appends @var{suffix} to the names of programs to install in @var{bindir}
773 (see above). For example, specifying @option{--program-suffix=-3.1}
774 would result in @samp{gcc} being installed as
775 @file{/usr/local/bin/gcc-3.1}.
777 @item --program-transform-name=@var{pattern}
778 Applies the @samp{sed} script @var{pattern} to be applied to the names
779 of programs to install in @var{bindir} (see above). @var{pattern} has to
780 consist of one or more basic @samp{sed} editing commands, separated by
781 semicolons. For example, if you want the @samp{gcc} program name to be
782 transformed to the installed program @file{/usr/local/bin/myowngcc} and
783 the @samp{g++} program name to be transformed to
784 @file{/usr/local/bin/gspecial++} without changing other program names,
785 you could use the pattern
786 @option{--program-transform-name='s/^gcc$/myowngcc/; s/^g++$/gspecial++/'}
787 to achieve this effect.
789 All three options can be combined and used together, resulting in more
790 complex conversion patterns. As a basic rule, @var{prefix} (and
791 @var{suffix}) are prepended (appended) before further transformations
792 can happen with a special transformation script @var{pattern}.
794 As currently implemented, this option only takes effect for native
795 builds; cross compiler binaries' names are not transformed even when a
796 transformation is explicitly asked for by one of these options.
798 For native builds, some of the installed programs are also installed
799 with the target alias in front of their name, as in
800 @samp{i686-pc-linux-gnu-gcc}. All of the above transformations happen
801 before the target alias is prepended to the name---so, specifying
802 @option{--program-prefix=foo-} and @option{program-suffix=-3.1}, the
803 resulting binary would be installed as
804 @file{/usr/local/bin/i686-pc-linux-gnu-foo-gcc-3.1}.
806 As a last shortcoming, none of the installed Ada programs are
807 transformed yet, which will be fixed in some time.
809 @item --with-local-prefix=@var{dirname}
811 installation directory for local include files. The default is
812 @file{/usr/local}. Specify this option if you want the compiler to
813 search directory @file{@var{dirname}/include} for locally installed
814 header files @emph{instead} of @file{/usr/local/include}.
816 You should specify @option{--with-local-prefix} @strong{only} if your
817 site has a different convention (not @file{/usr/local}) for where to put
820 The default value for @option{--with-local-prefix} is @file{/usr/local}
821 regardless of the value of @option{--prefix}. Specifying
822 @option{--prefix} has no effect on which directory GCC searches for
823 local header files. This may seem counterintuitive, but actually it is
826 The purpose of @option{--prefix} is to specify where to @emph{install
827 GCC}. The local header files in @file{/usr/local/include}---if you put
828 any in that directory---are not part of GCC@. They are part of other
829 programs---perhaps many others. (GCC installs its own header files in
830 another directory which is based on the @option{--prefix} value.)
832 Both the local-prefix include directory and the GCC-prefix include
833 directory are part of GCC's ``system include'' directories. Although these
834 two directories are not fixed, they need to be searched in the proper
835 order for the correct processing of the include_next directive. The
836 local-prefix include directory is searched before the GCC-prefix
837 include directory. Another characteristic of system include directories
838 is that pedantic warnings are turned off for headers in these directories.
840 Some autoconf macros add @option{-I @var{directory}} options to the
841 compiler command line, to ensure that directories containing installed
842 packages' headers are searched. When @var{directory} is one of GCC's
843 system include directories, GCC will ignore the option so that system
844 directories continue to be processed in the correct order. This
845 may result in a search order different from what was specified but the
846 directory will still be searched.
848 GCC automatically searches for ordinary libraries using
849 @env{GCC_EXEC_PREFIX}. Thus, when the same installation prefix is
850 used for both GCC and packages, GCC will automatically search for
851 both headers and libraries. This provides a configuration that is
852 easy to use. GCC behaves in a manner similar to that when it is
853 installed as a system compiler in @file{/usr}.
855 Sites that need to install multiple versions of GCC may not want to
856 use the above simple configuration. It is possible to use the
857 @option{--program-prefix}, @option{--program-suffix} and
858 @option{--program-transform-name} options to install multiple versions
859 into a single directory, but it may be simpler to use different prefixes
860 and the @option{--with-local-prefix} option to specify the location of the
861 site-specific files for each version. It will then be necessary for
862 users to specify explicitly the location of local site libraries
863 (e.g., with @env{LIBRARY_PATH}).
865 The same value can be used for both @option{--with-local-prefix} and
866 @option{--prefix} provided it is not @file{/usr}. This can be used
867 to avoid the default search of @file{/usr/local/include}.
869 @strong{Do not} specify @file{/usr} as the @option{--with-local-prefix}!
870 The directory you use for @option{--with-local-prefix} @strong{must not}
871 contain any of the system's standard header files. If it did contain
872 them, certain programs would be miscompiled (including GNU Emacs, on
873 certain targets), because this would override and nullify the header
874 file corrections made by the @command{fixincludes} script.
876 Indications are that people who use this option use it based on mistaken
877 ideas of what it is for. People use it as if it specified where to
878 install part of GCC@. Perhaps they make this assumption because
879 installing GCC creates the directory.
881 @item --enable-shared[=@var{package}[,@dots{}]]
882 Build shared versions of libraries, if shared libraries are supported on
883 the target platform. Unlike GCC 2.95.x and earlier, shared libraries
884 are enabled by default on all platforms that support shared libraries.
886 If a list of packages is given as an argument, build shared libraries
887 only for the listed packages. For other packages, only static libraries
888 will be built. Package names currently recognized in the GCC tree are
889 @samp{libgcc} (also known as @samp{gcc}), @samp{libstdc++} (not
890 @samp{libstdc++-v3}), @samp{libffi}, @samp{zlib}, @samp{boehm-gc},
891 @samp{ada}, @samp{libada}, @samp{libjava} and @samp{libobjc}.
892 Note @samp{libiberty} does not support shared libraries at all.
894 Use @option{--disable-shared} to build only static libraries. Note that
895 @option{--disable-shared} does not accept a list of package names as
896 argument, only @option{--enable-shared} does.
898 @item @anchor{with-gnu-as}--with-gnu-as
899 Specify that the compiler should assume that the
900 assembler it finds is the GNU assembler. However, this does not modify
901 the rules to find an assembler and will result in confusion if the
902 assembler found is not actually the GNU assembler. (Confusion may also
903 result if the compiler finds the GNU assembler but has not been
904 configured with @option{--with-gnu-as}.) If you have more than one
905 assembler installed on your system, you may want to use this option in
906 connection with @option{--with-as=@var{pathname}} or
907 @option{--with-build-time-tools=@var{pathname}}.
909 The following systems are the only ones where it makes a difference
910 whether you use the GNU assembler. On any other system,
911 @option{--with-gnu-as} has no effect.
914 @item @samp{hppa1.0-@var{any}-@var{any}}
915 @item @samp{hppa1.1-@var{any}-@var{any}}
916 @item @samp{sparc-sun-solaris2.@var{any}}
917 @item @samp{sparc64-@var{any}-solaris2.@var{any}}
920 @item @anchor{with-as}--with-as=@var{pathname}
921 Specify that the compiler should use the assembler pointed to by
922 @var{pathname}, rather than the one found by the standard rules to find
923 an assembler, which are:
926 Unless GCC is being built with a cross compiler, check the
927 @file{@var{libexec}/gcc/@var{target}/@var{version}} directory.
928 @var{libexec} defaults to @file{@var{exec-prefix}/libexec};
929 @var{exec-prefix} defaults to @var{prefix}, which
930 defaults to @file{/usr/local} unless overridden by the
931 @option{--prefix=@var{pathname}} switch described above. @var{target}
932 is the target system triple, such as @samp{sparc-sun-solaris2.7}, and
933 @var{version} denotes the GCC version, such as 3.0.
936 If the target system is the same that you are building on, check
937 operating system specific directories (e.g.@: @file{/usr/ccs/bin} on
941 Check in the @env{PATH} for a tool whose name is prefixed by the
942 target system triple.
945 Check in the @env{PATH} for a tool whose name is not prefixed by the
946 target system triple, if the host and target system triple are
947 the same (in other words, we use a host tool if it can be used for
951 You may want to use @option{--with-as} if no assembler
952 is installed in the directories listed above, or if you have multiple
953 assemblers installed and want to choose one that is not found by the
956 @item @anchor{with-gnu-ld}--with-gnu-ld
957 Same as @uref{#with-gnu-as,,@option{--with-gnu-as}}
960 @item --with-ld=@var{pathname}
961 Same as @uref{#with-as,,@option{--with-as}}
965 Specify that stabs debugging
966 information should be used instead of whatever format the host normally
967 uses. Normally GCC uses the same debug format as the host system.
969 On MIPS based systems and on Alphas, you must specify whether you want
970 GCC to create the normal ECOFF debugging format, or to use BSD-style
971 stabs passed through the ECOFF symbol table. The normal ECOFF debug
972 format cannot fully handle languages other than C@. BSD stabs format can
973 handle other languages, but it only works with the GNU debugger GDB@.
975 Normally, GCC uses the ECOFF debugging format by default; if you
976 prefer BSD stabs, specify @option{--with-stabs} when you configure GCC@.
978 No matter which default you choose when you configure GCC, the user
979 can use the @option{-gcoff} and @option{-gstabs+} options to specify explicitly
980 the debug format for a particular compilation.
982 @option{--with-stabs} is meaningful on the ISC system on the 386, also, if
983 @option{--with-gas} is used. It selects use of stabs debugging
984 information embedded in COFF output. This kind of debugging information
985 supports C++ well; ordinary COFF debugging information does not.
987 @option{--with-stabs} is also meaningful on 386 systems running SVR4. It
988 selects use of stabs debugging information embedded in ELF output. The
989 C++ compiler currently (2.6.0) does not support the DWARF debugging
990 information normally used on 386 SVR4 platforms; stabs provide a
991 workable alternative. This requires gas and gdb, as the normal SVR4
992 tools can not generate or interpret stabs.
994 @item --disable-multilib
995 Specify that multiple target
996 libraries to support different target variants, calling
997 conventions, etc.@: should not be built. The default is to build a
998 predefined set of them.
1000 Some targets provide finer-grained control over which multilibs are built
1001 (e.g., @option{--disable-softfloat}):
1007 fpu, 26bit, underscore, interwork, biendian, nofmult.
1010 softfloat, m68881, m68000, m68020.
1013 single-float, biendian, softfloat.
1015 @item powerpc*-*-*, rs6000*-*-*
1016 aix64, pthread, softfloat, powercpu, powerpccpu, powerpcos, biendian,
1021 @item --with-multilib-list=@var{list}
1022 @itemx --without-multilib-list
1023 Specify what multilibs to build.
1024 Currently only implemented for sh*-*-*.
1026 @var{list} is a comma separated list of CPU names. These must be of the
1027 form @code{sh*} or @code{m*} (in which case they match the compiler option
1028 for that processor). The list should not contain any endian options -
1029 these are handled by @option{--with-endian}.
1031 If @var{list} is empty, then there will be no multilibs for extra
1032 processors. The multilib for the secondary endian remains enabled.
1034 As a special case, if an entry in the list starts with a @code{!}
1035 (exclamation point), then it is added to the list of excluded multilibs.
1036 Entries of this sort should be compatible with @samp{MULTILIB_EXCLUDES}
1037 (once the leading @code{!} has been stripped).
1039 If @option{--with-multilib-list} is not given, then a default set of
1040 multilibs is selected based on the value of @option{--target}. This is
1041 usually the complete set of libraries, but some targets imply a more
1044 Example 1: to configure a compiler for SH4A only, but supporting both
1045 endians, with little endian being the default:
1047 --with-cpu=sh4a --with-endian=little,big --with-multilib-list=
1050 Example 2: to configure a compiler for both SH4A and SH4AL-DSP, but with
1051 only little endian SH4AL:
1053 --with-cpu=sh4a --with-endian=little,big --with-multilib-list=sh4al,!mb/m4al
1056 @item --with-endian=@var{endians}
1057 Specify what endians to use.
1058 Currently only implemented for sh*-*-*.
1060 @var{endians} may be one of the following:
1063 Use big endian exclusively.
1065 Use little endian exclusively.
1067 Use big endian by default. Provide a multilib for little endian.
1069 Use little endian by default. Provide a multilib for big endian.
1072 @item --enable-threads
1073 Specify that the target
1074 supports threads. This affects the Objective-C compiler and runtime
1075 library, and exception handling for other languages like C++ and Java.
1076 On some systems, this is the default.
1078 In general, the best (and, in many cases, the only known) threading
1079 model available will be configured for use. Beware that on some
1080 systems, GCC has not been taught what threading models are generally
1081 available for the system. In this case, @option{--enable-threads} is an
1082 alias for @option{--enable-threads=single}.
1084 @item --disable-threads
1085 Specify that threading support should be disabled for the system.
1086 This is an alias for @option{--enable-threads=single}.
1088 @item --enable-threads=@var{lib}
1090 @var{lib} is the thread support library. This affects the Objective-C
1091 compiler and runtime library, and exception handling for other languages
1092 like C++ and Java. The possibilities for @var{lib} are:
1100 Ada tasking support. For non-Ada programs, this setting is equivalent
1101 to @samp{single}. When used in conjunction with the Ada run time, it
1102 causes GCC to use the same thread primitives as Ada uses. This option
1103 is necessary when using both Ada and the back end exception handling,
1104 which is the default for most Ada targets.
1106 Generic MACH thread support, known to work on NeXTSTEP@. (Please note
1107 that the file needed to support this configuration, @file{gthr-mach.h}, is
1108 missing and thus this setting will cause a known bootstrap failure.)
1110 This is an alias for @samp{single}.
1112 Generic POSIX/Unix98 thread support.
1114 Generic POSIX/Unix95 thread support.
1116 RTEMS thread support.
1118 Disable thread support, should work for all platforms.
1120 Sun Solaris 2/Unix International thread support. Only use this if you
1121 really need to use this legacy API instead of the default, @samp{posix}.
1123 VxWorks thread support.
1125 Microsoft Win32 API thread support.
1127 Novell Kernel Services thread support.
1131 Specify that the target supports TLS (Thread Local Storage). Usually
1132 configure can correctly determine if TLS is supported. In cases where
1133 it guesses incorrectly, TLS can be explicitly enabled or disabled with
1134 @option{--enable-tls} or @option{--disable-tls}. This can happen if
1135 the assembler supports TLS but the C library does not, or if the
1136 assumptions made by the configure test are incorrect.
1139 Specify that the target does not support TLS.
1140 This is an alias for @option{--enable-tls=no}.
1142 @item --with-cpu=@var{cpu}
1143 @itemx --with-cpu-32=@var{cpu}
1144 @itemx --with-cpu-64=@var{cpu}
1145 Specify which cpu variant the compiler should generate code for by default.
1146 @var{cpu} will be used as the default value of the @option{-mcpu=} switch.
1147 This option is only supported on some targets, including ARM, i386, M68k,
1148 PowerPC, and SPARC@. The @option{--with-cpu-32} and
1149 @option{--with-cpu-64} options specify separate default CPUs for
1150 32-bit and 64-bit modes; these options are only supported for i386,
1153 @item --with-schedule=@var{cpu}
1154 @itemx --with-arch=@var{cpu}
1155 @itemx --with-arch-32=@var{cpu}
1156 @itemx --with-arch-64=@var{cpu}
1157 @itemx --with-tune=@var{cpu}
1158 @itemx --with-tune-32=@var{cpu}
1159 @itemx --with-tune-64=@var{cpu}
1160 @itemx --with-abi=@var{abi}
1161 @itemx --with-fpu=@var{type}
1162 @itemx --with-float=@var{type}
1163 These configure options provide default values for the @option{-mschedule=},
1164 @option{-march=}, @option{-mtune=}, @option{-mabi=}, and @option{-mfpu=}
1165 options and for @option{-mhard-float} or @option{-msoft-float}. As with
1166 @option{--with-cpu}, which switches will be accepted and acceptable values
1167 of the arguments depend on the target.
1169 @item --with-mode=@var{mode}
1170 Specify if the compiler should default to @option{-marm} or @option{-mthumb}.
1171 This option is only supported on ARM targets.
1173 @item --with-divide=@var{type}
1174 Specify how the compiler should generate code for checking for
1175 division by zero. This option is only supported on the MIPS target.
1176 The possibilities for @var{type} are:
1179 Division by zero checks use conditional traps (this is the default on
1180 systems that support conditional traps).
1182 Division by zero checks use the break instruction.
1185 @c If you make --with-llsc the default for additional targets,
1186 @c update the --with-llsc description in the MIPS section below.
1189 On MIPS targets, make @option{-mllsc} the default when no
1190 @option{-mno-lsc} option is passed. This is the default for
1191 Linux-based targets, as the kernel will emulate them if the ISA does
1194 @item --without-llsc
1195 On MIPS targets, make @option{-mno-llsc} the default when no
1196 @option{-mllsc} option is passed.
1199 On MIPS targets, make @option{-msynci} the default when no
1200 @option{-mno-synci} option is passed.
1202 @item --without-synci
1203 On MIPS targets, make @option{-mno-synci} the default when no
1204 @option{-msynci} option is passed. This is the default.
1206 @item --with-mips-plt
1207 On MIPS targets, make use of copy relocations and PLTs.
1208 These features are extensions to the traditional
1209 SVR4-based MIPS ABIs and require support from GNU binutils
1210 and the runtime C library.
1212 @item --enable-__cxa_atexit
1213 Define if you want to use __cxa_atexit, rather than atexit, to
1214 register C++ destructors for local statics and global objects.
1215 This is essential for fully standards-compliant handling of
1216 destructors, but requires __cxa_atexit in libc. This option is currently
1217 only available on systems with GNU libc. When enabled, this will cause
1218 @option{-fuse-cxa-atexit} to be passed by default.
1220 @item --enable-target-optspace
1222 libraries should be optimized for code space instead of code speed.
1223 This is the default for the m32r platform.
1225 @item --with-cpp-install-dir=@var{dirname}
1226 Specify that the user visible @command{cpp} program should be installed
1227 in @file{@var{prefix}/@var{dirname}/cpp}, in addition to @var{bindir}.
1229 @item --enable-initfini-array
1230 Force the use of sections @code{.init_array} and @code{.fini_array}
1231 (instead of @code{.init} and @code{.fini}) for constructors and
1232 destructors. Option @option{--disable-initfini-array} has the
1233 opposite effect. If neither option is specified, the configure script
1234 will try to guess whether the @code{.init_array} and
1235 @code{.fini_array} sections are supported and, if they are, use them.
1237 @item --enable-build-with-cxx
1238 Build GCC using a C++ compiler rather than a C compiler. This is an
1239 experimental option which may become the default in a later release.
1241 @item --enable-maintainer-mode
1242 The build rules that regenerate the Autoconf and Automake output files as
1243 well as the GCC master message catalog @file{gcc.pot} are normally
1244 disabled. This is because it can only be rebuilt if the complete source
1245 tree is present. If you have changed the sources and want to rebuild the
1246 catalog, configuring with @option{--enable-maintainer-mode} will enable
1247 this. Note that you need a recent version of the @code{gettext} tools
1250 @item --disable-bootstrap
1251 For a native build, the default configuration is to perform
1252 a 3-stage bootstrap of the compiler when @samp{make} is invoked,
1253 testing that GCC can compile itself correctly. If you want to disable
1254 this process, you can configure with @option{--disable-bootstrap}.
1256 @item --enable-bootstrap
1257 In special cases, you may want to perform a 3-stage build
1258 even if the target and host triplets are different.
1259 This is possible when the host can run code compiled for
1260 the target (e.g.@: host is i686-linux, target is i486-linux).
1261 Starting from GCC 4.2, to do this you have to configure explicitly
1262 with @option{--enable-bootstrap}.
1264 @item --enable-generated-files-in-srcdir
1265 Neither the .c and .h files that are generated from Bison and flex nor the
1266 info manuals and man pages that are built from the .texi files are present
1267 in the SVN development tree. When building GCC from that development tree,
1268 or from one of our snapshots, those generated files are placed in your
1269 build directory, which allows for the source to be in a readonly
1272 If you configure with @option{--enable-generated-files-in-srcdir} then those
1273 generated files will go into the source directory. This is mainly intended
1274 for generating release or prerelease tarballs of the GCC sources, since it
1275 is not a requirement that the users of source releases to have flex, Bison,
1278 @item --enable-version-specific-runtime-libs
1280 that runtime libraries should be installed in the compiler specific
1281 subdirectory (@file{@var{libdir}/gcc}) rather than the usual places. In
1282 addition, @samp{libstdc++}'s include files will be installed into
1283 @file{@var{libdir}} unless you overruled it by using
1284 @option{--with-gxx-include-dir=@var{dirname}}. Using this option is
1285 particularly useful if you intend to use several versions of GCC in
1286 parallel. This is currently supported by @samp{libgfortran},
1287 @samp{libjava}, @samp{libmudflap}, @samp{libstdc++}, and @samp{libobjc}.
1289 @item --enable-languages=@var{lang1},@var{lang2},@dots{}
1290 Specify that only a particular subset of compilers and
1291 their runtime libraries should be built. For a list of valid values for
1292 @var{langN} you can issue the following command in the
1293 @file{gcc} directory of your GCC source tree:@*
1295 grep language= */config-lang.in
1297 Currently, you can use any of the following:
1298 @code{all}, @code{ada}, @code{c}, @code{c++}, @code{fortran}, @code{java},
1299 @code{objc}, @code{obj-c++}.
1300 Building the Ada compiler has special requirements, see below.
1301 If you do not pass this flag, or specify the option @code{all}, then all
1302 default languages available in the @file{gcc} sub-tree will be configured.
1303 Ada and Objective-C++ are not default languages; the rest are.
1304 Re-defining @code{LANGUAGES} when calling @samp{make} @strong{does not}
1305 work anymore, as those language sub-directories might not have been
1308 @item --enable-stage1-languages=@var{lang1},@var{lang2},@dots{}
1309 Specify that a particular subset of compilers and their runtime
1310 libraries should be built with the system C compiler during stage 1 of
1311 the bootstrap process, rather than only in later stages with the
1312 bootstrapped C compiler. The list of valid values is the same as for
1313 @option{--enable-languages}, and the option @code{all} will select all
1314 of the languages enabled by @option{--enable-languages}. This option is
1315 primarily useful for GCC development; for instance, when a development
1316 version of the compiler cannot bootstrap due to compiler bugs, or when
1317 one is debugging front ends other than the C front end. When this
1318 option is used, one can then build the target libraries for the
1319 specified languages with the stage-1 compiler by using @command{make
1320 stage1-bubble all-target}, or run the testsuite on the stage-1 compiler
1321 for the specified languages using @command{make stage1-start check-gcc}.
1323 @item --disable-libada
1324 Specify that the run-time libraries and tools used by GNAT should not
1325 be built. This can be useful for debugging, or for compatibility with
1326 previous Ada build procedures, when it was required to explicitly
1327 do a @samp{make -C gcc gnatlib_and_tools}.
1329 @item --disable-libssp
1330 Specify that the run-time libraries for stack smashing protection
1331 should not be built.
1333 @item --disable-libgomp
1334 Specify that the run-time libraries used by GOMP should not be built.
1337 Specify that the compiler should
1338 use DWARF 2 debugging information as the default.
1340 @item --enable-targets=all
1341 @itemx --enable-targets=@var{target_list}
1342 Some GCC targets, e.g.@: powerpc64-linux, build bi-arch compilers.
1343 These are compilers that are able to generate either 64-bit or 32-bit
1344 code. Typically, the corresponding 32-bit target, e.g.@:
1345 powerpc-linux for powerpc64-linux, only generates 32-bit code. This
1346 option enables the 32-bit target to be a bi-arch compiler, which is
1347 useful when you want a bi-arch compiler that defaults to 32-bit, and
1348 you are building a bi-arch or multi-arch binutils in a combined tree.
1349 On mips-linux, this will build a tri-arch compiler (ABI o32/n32/64),
1351 Currently, this option only affects sparc-linux, powerpc-linux, x86-linux
1354 @item --enable-secureplt
1355 This option enables @option{-msecure-plt} by default for powerpc-linux.
1357 @xref{RS/6000 and PowerPC Options,, RS/6000 and PowerPC Options, gcc,
1358 Using the GNU Compiler Collection (GCC)},
1361 See ``RS/6000 and PowerPC Options'' in the main manual
1365 This option enables @option{-mcld} by default for 32-bit x86 targets.
1367 @xref{i386 and x86-64 Options,, i386 and x86-64 Options, gcc,
1368 Using the GNU Compiler Collection (GCC)},
1371 See ``i386 and x86-64 Options'' in the main manual
1374 @item --enable-win32-registry
1375 @itemx --enable-win32-registry=@var{key}
1376 @itemx --disable-win32-registry
1377 The @option{--enable-win32-registry} option enables Microsoft Windows-hosted GCC
1378 to look up installations paths in the registry using the following key:
1381 @code{HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\@var{key}}
1384 @var{key} defaults to GCC version number, and can be overridden by the
1385 @option{--enable-win32-registry=@var{key}} option. Vendors and distributors
1386 who use custom installers are encouraged to provide a different key,
1387 perhaps one comprised of vendor name and GCC version number, to
1388 avoid conflict with existing installations. This feature is enabled
1389 by default, and can be disabled by @option{--disable-win32-registry}
1390 option. This option has no effect on the other hosts.
1393 Specify that the machine does not have a floating point unit. This
1394 option only applies to @samp{m68k-sun-sunos@var{n}}. On any other
1395 system, @option{--nfp} has no effect.
1397 @item --enable-werror
1398 @itemx --disable-werror
1399 @itemx --enable-werror=yes
1400 @itemx --enable-werror=no
1401 When you specify this option, it controls whether certain files in the
1402 compiler are built with @option{-Werror} in bootstrap stage2 and later.
1403 If you don't specify it, @option{-Werror} is turned on for the main
1404 development trunk. However it defaults to off for release branches and
1405 final releases. The specific files which get @option{-Werror} are
1406 controlled by the Makefiles.
1408 @item --enable-checking
1409 @itemx --enable-checking=@var{list}
1410 When you specify this option, the compiler is built to perform internal
1411 consistency checks of the requested complexity. This does not change the
1412 generated code, but adds error checking within the compiler. This will
1413 slow down the compiler and may only work properly if you are building
1414 the compiler with GCC@. This is @samp{yes} by default when building
1415 from SVN or snapshots, but @samp{release} for releases. The default
1416 for building the stage1 compiler is @samp{yes}. More control
1417 over the checks may be had by specifying @var{list}. The categories of
1418 checks available are @samp{yes} (most common checks
1419 @samp{assert,misc,tree,gc,rtlflag,runtime}), @samp{no} (no checks at
1420 all), @samp{all} (all but @samp{valgrind}), @samp{release} (cheapest
1421 checks @samp{assert,runtime}) or @samp{none} (same as @samp{no}).
1422 Individual checks can be enabled with these flags @samp{assert},
1423 @samp{df}, @samp{fold}, @samp{gc}, @samp{gcac} @samp{misc}, @samp{rtl},
1424 @samp{rtlflag}, @samp{runtime}, @samp{tree}, and @samp{valgrind}.
1426 The @samp{valgrind} check requires the external @command{valgrind}
1427 simulator, available from @uref{http://valgrind.org/}. The
1428 @samp{df}, @samp{rtl}, @samp{gcac} and @samp{valgrind} checks are very expensive.
1429 To disable all checking, @samp{--disable-checking} or
1430 @samp{--enable-checking=none} must be explicitly requested. Disabling
1431 assertions will make the compiler and runtime slightly faster but
1432 increase the risk of undetected internal errors causing wrong code to be
1435 @item --disable-stage1-checking
1436 @itemx --enable-stage1-checking
1437 @itemx --enable-stage1-checking=@var{list}
1438 If no @option{--enable-checking} option is specified the stage1
1439 compiler will be built with @samp{yes} checking enabled, otherwise
1440 the stage1 checking flags are the same as specified by
1441 @option{--enable-checking}. To build the stage1 compiler with
1442 different checking options use @option{--enable-stage1-checking}.
1443 The list of checking options is the same as for @option{--enable-checking}.
1444 If your system is too slow or too small to bootstrap a released compiler
1445 with checking for stage1 enabled, you can use @samp{--disable-stage1-checking}
1446 to disable checking for the stage1 compiler.
1448 @item --enable-coverage
1449 @itemx --enable-coverage=@var{level}
1450 With this option, the compiler is built to collect self coverage
1451 information, every time it is run. This is for internal development
1452 purposes, and only works when the compiler is being built with gcc. The
1453 @var{level} argument controls whether the compiler is built optimized or
1454 not, values are @samp{opt} and @samp{noopt}. For coverage analysis you
1455 want to disable optimization, for performance analysis you want to
1456 enable optimization. When coverage is enabled, the default level is
1457 without optimization.
1459 @item --enable-gather-detailed-mem-stats
1460 When this option is specified more detailed information on memory
1461 allocation is gathered. This information is printed when using
1462 @option{-fmem-report}.
1465 @itemx --with-gc=@var{choice}
1466 With this option you can specify the garbage collector implementation
1467 used during the compilation process. @var{choice} can be one of
1468 @samp{page} and @samp{zone}, where @samp{page} is the default.
1471 @itemx --disable-nls
1472 The @option{--enable-nls} option enables Native Language Support (NLS),
1473 which lets GCC output diagnostics in languages other than American
1474 English. Native Language Support is enabled by default if not doing a
1475 canadian cross build. The @option{--disable-nls} option disables NLS@.
1477 @item --with-included-gettext
1478 If NLS is enabled, the @option{--with-included-gettext} option causes the build
1479 procedure to prefer its copy of GNU @command{gettext}.
1481 @item --with-catgets
1482 If NLS is enabled, and if the host lacks @code{gettext} but has the
1483 inferior @code{catgets} interface, the GCC build procedure normally
1484 ignores @code{catgets} and instead uses GCC's copy of the GNU
1485 @code{gettext} library. The @option{--with-catgets} option causes the
1486 build procedure to use the host's @code{catgets} in this situation.
1488 @item --with-libiconv-prefix=@var{dir}
1489 Search for libiconv header files in @file{@var{dir}/include} and
1490 libiconv library files in @file{@var{dir}/lib}.
1492 @item --enable-obsolete
1493 Enable configuration for an obsoleted system. If you attempt to
1494 configure GCC for a system (build, host, or target) which has been
1495 obsoleted, and you do not specify this flag, configure will halt with an
1498 All support for systems which have been obsoleted in one release of GCC
1499 is removed entirely in the next major release, unless someone steps
1500 forward to maintain the port.
1502 @item --enable-decimal-float
1503 @itemx --enable-decimal-float=yes
1504 @itemx --enable-decimal-float=no
1505 @itemx --enable-decimal-float=bid
1506 @itemx --enable-decimal-float=dpd
1507 @itemx --disable-decimal-float
1508 Enable (or disable) support for the C decimal floating point extension
1509 that is in the IEEE 754-2008 standard. This is enabled by default only
1510 on PowerPC, i386, and x86_64 GNU/Linux systems. Other systems may also
1511 support it, but require the user to specifically enable it. You can
1512 optionally control which decimal floating point format is used (either
1513 @samp{bid} or @samp{dpd}). The @samp{bid} (binary integer decimal)
1514 format is default on i386 and x86_64 systems, and the @samp{dpd}
1515 (densely packed decimal) format is default on PowerPC systems.
1517 @item --enable-fixed-point
1518 @itemx --disable-fixed-point
1519 Enable (or disable) support for C fixed-point arithmetic.
1520 This option is enabled by default for some targets (such as MIPS) which
1521 have hardware-support for fixed-point operations. On other targets, you
1522 may enable this option manually.
1524 @item --with-long-double-128
1525 Specify if @code{long double} type should be 128-bit by default on selected
1526 GNU/Linux architectures. If using @code{--without-long-double-128},
1527 @code{long double} will be by default 64-bit, the same as @code{double} type.
1528 When neither of these configure options are used, the default will be
1529 128-bit @code{long double} when built against GNU C Library 2.4 and later,
1530 64-bit @code{long double} otherwise.
1532 @item --with-gmp=@var{pathname}
1533 @itemx --with-gmp-include=@var{pathname}
1534 @itemx --with-gmp-lib=@var{pathname}
1535 @itemx --with-mpfr=@var{pathname}
1536 @itemx --with-mpfr-include=@var{pathname}
1537 @itemx --with-mpfr-lib=@var{pathname}
1538 @itemx --with-mpc=@var{pathname}
1539 @itemx --with-mpc-include=@var{pathname}
1540 @itemx --with-mpc-lib=@var{pathname}
1541 If you do not have GMP (the GNU Multiple Precision library), the MPFR
1542 library and/or the MPC library installed in a standard location and
1543 you want to build GCC, you can explicitly specify the directory where
1544 they are installed (@samp{--with-gmp=@var{gmpinstalldir}},
1545 @samp{--with-mpfr=@var{mpfrinstalldir}},
1546 @samp{--with-mpc=@var{mpcinstalldir}}). The
1547 @option{--with-gmp=@var{gmpinstalldir}} option is shorthand for
1548 @option{--with-gmp-lib=@var{gmpinstalldir}/lib} and
1549 @option{--with-gmp-include=@var{gmpinstalldir}/include}. Likewise the
1550 @option{--with-mpfr=@var{mpfrinstalldir}} option is shorthand for
1551 @option{--with-mpfr-lib=@var{mpfrinstalldir}/lib} and
1552 @option{--with-mpfr-include=@var{mpfrinstalldir}/include}, also the
1553 @option{--with-mpc=@var{mpcinstalldir}} option is shorthand for
1554 @option{--with-mpc-lib=@var{mpcinstalldir}/lib} and
1555 @option{--with-mpc-include=@var{mpcinstalldir}/include}. If these
1556 shorthand assumptions are not correct, you can use the explicit
1557 include and lib options directly.
1559 @item --with-ppl=@var{pathname}
1560 @itemx --with-ppl-include=@var{pathname}
1561 @itemx --with-ppl-lib=@var{pathname}
1562 @itemx --with-cloog=@var{pathname}
1563 @itemx --with-cloog-include=@var{pathname}
1564 @itemx --with-cloog-lib=@var{pathname}
1565 If you do not have PPL (the Parma Polyhedra Library) and the CLooG
1566 libraries installed in a standard location and you want to build GCC,
1567 you can explicitly specify the directory where they are installed
1568 (@samp{--with-ppl=@var{pplinstalldir}},
1569 @samp{--with-cloog=@var{clooginstalldir}}). The
1570 @option{--with-ppl=@var{pplinstalldir}} option is shorthand for
1571 @option{--with-ppl-lib=@var{pplinstalldir}/lib} and
1572 @option{--with-ppl-include=@var{pplinstalldir}/include}. Likewise the
1573 @option{--with-cloog=@var{clooginstalldir}} option is shorthand for
1574 @option{--with-cloog-lib=@var{clooginstalldir}/lib} and
1575 @option{--with-cloog-include=@var{clooginstalldir}/include}. If these
1576 shorthand assumptions are not correct, you can use the explicit
1577 include and lib options directly.
1579 @item --with-host-libstdcxx=@var{linker-args}
1580 If you are linking with a static copy of PPL, you can use this option
1581 to specify how the linker should find the standard C++ library used
1582 internally by PPL. Typical values of @var{linker-args} might be
1583 @samp{-lstdc++} or @samp{-Wl,-Bstatic,-lstdc++,-Bdynamic -lm}. If you are
1584 linking with a shared copy of PPL, you probably do not need this
1585 option; shared library dependencies will cause the linker to search
1586 for the standard C++ library automatically.
1588 @item --with-stage1-ldflags=@var{flags}
1589 This option may be used to set linker flags to be used when linking
1590 stage 1 of GCC. These are also used when linking GCC if configured with
1591 @option{--disable-bootstrap}. By default no special flags are used.
1593 @item --with-stage1-libs=@var{libs}
1594 This option may be used to set libraries to be used when linking stage 1
1595 of GCC. These are also used when linking GCC if configured with
1596 @option{--disable-bootstrap}. The default is the argument to
1597 @option{--with-host-libstdcxx}, if specified.
1599 @item --with-boot-ldflags=@var{flags}
1600 This option may be used to set linker flags to be used when linking
1601 stage 2 and later when bootstrapping GCC. By default no special flags
1604 @item --with-boot-libs=@var{libs}
1605 This option may be used to set libraries to be used when linking stage 2
1606 and later when bootstrapping GCC. The default is the argument to
1607 @option{--with-host-libstdcxx}, if specified.
1609 @item --with-debug-prefix-map=@var{map}
1610 Convert source directory names using @option{-fdebug-prefix-map} when
1611 building runtime libraries. @samp{@var{map}} is a space-separated
1612 list of maps of the form @samp{@var{old}=@var{new}}.
1614 @item --enable-linker-build-id
1615 Tells GCC to pass @option{--build-id} option to the linker for all final
1616 links (links performed without the @option{-r} or @option{--relocatable}
1617 option), if the linker supports it. If you specify
1618 @option{--enable-linker-build-id}, but your linker does not
1619 support @option{--build-id} option, a warning is issued and the
1620 @option{--enable-linker-build-id} option is ignored. The default is off.
1622 @item --enable-gnu-unique-object
1623 @itemx --disable-gnu-unique-object
1624 Tells GCC to use the gnu_unique_object relocation for C++ template
1625 static data members and inline function local statics. Enabled by
1626 default for a native toolchain with an assembler that accepts it and
1627 GLIBC 2.11 or above, otherwise disabled.
1631 @subheading Cross-Compiler-Specific Options
1632 The following options only apply to building cross compilers.
1635 @item --with-sysroot
1636 @itemx --with-sysroot=@var{dir}
1637 Tells GCC to consider @var{dir} as the root of a tree that contains a
1638 (subset of) the root filesystem of the target operating system.
1639 Target system headers, libraries and run-time object files will be
1640 searched in there. More specifically, this acts as if
1641 @option{--sysroot=@var{dir}} was added to the default options of the built
1642 compiler. The specified directory is not copied into the
1643 install tree, unlike the options @option{--with-headers} and
1644 @option{--with-libs} that this option obsoletes. The default value,
1645 in case @option{--with-sysroot} is not given an argument, is
1646 @option{$@{gcc_tooldir@}/sys-root}. If the specified directory is a
1647 subdirectory of @option{$@{exec_prefix@}}, then it will be found relative to
1648 the GCC binaries if the installation tree is moved.
1650 This option affects the system root for the compiler used to build
1651 target libraries (which runs on the build system) and the compiler newly
1652 installed with @code{make install}; it does not affect the compiler which is
1653 used to build GCC itself.
1655 @item --with-build-sysroot
1656 @itemx --with-build-sysroot=@var{dir}
1657 Tells GCC to consider @var{dir} as the system root (see
1658 @option{--with-sysroot}) while building target libraries, instead of
1659 the directory specified with @option{--with-sysroot}. This option is
1660 only useful when you are already using @option{--with-sysroot}. You
1661 can use @option{--with-build-sysroot} when you are configuring with
1662 @option{--prefix} set to a directory that is different from the one in
1663 which you are installing GCC and your target libraries.
1665 This option affects the system root for the compiler used to build
1666 target libraries (which runs on the build system); it does not affect
1667 the compiler which is used to build GCC itself.
1669 @item --with-headers
1670 @itemx --with-headers=@var{dir}
1671 Deprecated in favor of @option{--with-sysroot}.
1672 Specifies that target headers are available when building a cross compiler.
1673 The @var{dir} argument specifies a directory which has the target include
1674 files. These include files will be copied into the @file{gcc} install
1675 directory. @emph{This option with the @var{dir} argument is required} when
1676 building a cross compiler, if @file{@var{prefix}/@var{target}/sys-include}
1677 doesn't pre-exist. If @file{@var{prefix}/@var{target}/sys-include} does
1678 pre-exist, the @var{dir} argument may be omitted. @command{fixincludes}
1679 will be run on these files to make them compatible with GCC@.
1681 @item --without-headers
1682 Tells GCC not use any target headers from a libc when building a cross
1683 compiler. When crossing to GNU/Linux, you need the headers so GCC
1684 can build the exception handling for libgcc.
1687 @itemx --with-libs="@var{dir1} @var{dir2} @dots{} @var{dirN}"
1688 Deprecated in favor of @option{--with-sysroot}.
1689 Specifies a list of directories which contain the target runtime
1690 libraries. These libraries will be copied into the @file{gcc} install
1691 directory. If the directory list is omitted, this option has no
1695 Specifies that @samp{newlib} is
1696 being used as the target C library. This causes @code{__eprintf} to be
1697 omitted from @file{libgcc.a} on the assumption that it will be provided by
1700 @item --with-build-time-tools=@var{dir}
1701 Specifies where to find the set of target tools (assembler, linker, etc.)
1702 that will be used while building GCC itself. This option can be useful
1703 if the directory layouts are different between the system you are building
1704 GCC on, and the system where you will deploy it.
1706 For example, on an @samp{ia64-hp-hpux} system, you may have the GNU
1707 assembler and linker in @file{/usr/bin}, and the native tools in a
1708 different path, and build a toolchain that expects to find the
1709 native tools in @file{/usr/bin}.
1711 When you use this option, you should ensure that @var{dir} includes
1712 @command{ar}, @command{as}, @command{ld}, @command{nm},
1713 @command{ranlib} and @command{strip} if necessary, and possibly
1714 @command{objdump}. Otherwise, GCC may use an inconsistent set of
1718 @subheading Java-Specific Options
1720 The following option applies to the build of the Java front end.
1723 @item --disable-libgcj
1724 Specify that the run-time libraries
1725 used by GCJ should not be built. This is useful in case you intend
1726 to use GCJ with some other run-time, or you're going to install it
1727 separately, or it just happens not to build on your particular
1728 machine. In general, if the Java front end is enabled, the GCJ
1729 libraries will be enabled too, unless they're known to not work on
1730 the target platform. If GCJ is enabled but @samp{libgcj} isn't built, you
1731 may need to port it; in this case, before modifying the top-level
1732 @file{configure.in} so that @samp{libgcj} is enabled by default on this platform,
1733 you may use @option{--enable-libgcj} to override the default.
1737 The following options apply to building @samp{libgcj}.
1739 @subsubheading General Options
1742 @item --enable-java-maintainer-mode
1743 By default the @samp{libjava} build will not attempt to compile the
1744 @file{.java} source files to @file{.class}. Instead, it will use the
1745 @file{.class} files from the source tree. If you use this option you
1746 must have executables named @command{ecj1} and @command{gjavah} in your path
1747 for use by the build. You must use this option if you intend to
1748 modify any @file{.java} files in @file{libjava}.
1750 @item --with-java-home=@var{dirname}
1751 This @samp{libjava} option overrides the default value of the
1752 @samp{java.home} system property. It is also used to set
1753 @samp{sun.boot.class.path} to @file{@var{dirname}/lib/rt.jar}. By
1754 default @samp{java.home} is set to @file{@var{prefix}} and
1755 @samp{sun.boot.class.path} to
1756 @file{@var{datadir}/java/libgcj-@var{version}.jar}.
1758 @item --with-ecj-jar=@var{filename}
1759 This option can be used to specify the location of an external jar
1760 file containing the Eclipse Java compiler. A specially modified
1761 version of this compiler is used by @command{gcj} to parse
1762 @file{.java} source files. If this option is given, the
1763 @samp{libjava} build will create and install an @file{ecj1} executable
1764 which uses this jar file at runtime.
1766 If this option is not given, but an @file{ecj.jar} file is found in
1767 the topmost source tree at configure time, then the @samp{libgcj}
1768 build will create and install @file{ecj1}, and will also install the
1769 discovered @file{ecj.jar} into a suitable place in the install tree.
1771 If @file{ecj1} is not installed, then the user will have to supply one
1772 on his path in order for @command{gcj} to properly parse @file{.java}
1773 source files. A suitable jar is available from
1774 @uref{ftp://sourceware.org/pub/java/}.
1776 @item --disable-getenv-properties
1777 Don't set system properties from @env{GCJ_PROPERTIES}.
1779 @item --enable-hash-synchronization
1780 Use a global hash table for monitor locks. Ordinarily,
1781 @samp{libgcj}'s @samp{configure} script automatically makes
1782 the correct choice for this option for your platform. Only use
1783 this if you know you need the library to be configured differently.
1785 @item --enable-interpreter
1786 Enable the Java interpreter. The interpreter is automatically
1787 enabled by default on all platforms that support it. This option
1788 is really only useful if you want to disable the interpreter
1789 (using @option{--disable-interpreter}).
1791 @item --disable-java-net
1792 Disable java.net. This disables the native part of java.net only,
1793 using non-functional stubs for native method implementations.
1795 @item --disable-jvmpi
1796 Disable JVMPI support.
1798 @item --disable-libgcj-bc
1799 Disable BC ABI compilation of certain parts of libgcj. By default,
1800 some portions of libgcj are compiled with @option{-findirect-dispatch}
1801 and @option{-fno-indirect-classes}, allowing them to be overridden at
1804 If @option{--disable-libgcj-bc} is specified, libgcj is built without
1805 these options. This allows the compile-time linker to resolve
1806 dependencies when statically linking to libgcj. However it makes it
1807 impossible to override the affected portions of libgcj at run-time.
1809 @item --enable-reduced-reflection
1810 Build most of libgcj with @option{-freduced-reflection}. This reduces
1811 the size of libgcj at the expense of not being able to do accurate
1812 reflection on the classes it contains. This option is safe if you
1813 know that code using libgcj will never use reflection on the standard
1814 runtime classes in libgcj (including using serialization, RMI or CORBA).
1817 Enable runtime eCos target support.
1819 @item --without-libffi
1820 Don't use @samp{libffi}. This will disable the interpreter and JNI
1821 support as well, as these require @samp{libffi} to work.
1823 @item --enable-libgcj-debug
1824 Enable runtime debugging code.
1826 @item --enable-libgcj-multifile
1827 If specified, causes all @file{.java} source files to be
1828 compiled into @file{.class} files in one invocation of
1829 @samp{gcj}. This can speed up build time, but is more
1830 resource-intensive. If this option is unspecified or
1831 disabled, @samp{gcj} is invoked once for each @file{.java}
1832 file to compile into a @file{.class} file.
1834 @item --with-libiconv-prefix=DIR
1835 Search for libiconv in @file{DIR/include} and @file{DIR/lib}.
1837 @item --enable-sjlj-exceptions
1838 Force use of the @code{setjmp}/@code{longjmp}-based scheme for exceptions.
1839 @samp{configure} ordinarily picks the correct value based on the platform.
1840 Only use this option if you are sure you need a different setting.
1842 @item --with-system-zlib
1843 Use installed @samp{zlib} rather than that included with GCC@.
1845 @item --with-win32-nlsapi=ansi, unicows or unicode
1846 Indicates how MinGW @samp{libgcj} translates between UNICODE
1847 characters and the Win32 API@.
1849 @item --enable-java-home
1850 If enabled, this creates a JPackage compatible SDK environment during install.
1851 Note that if --enable-java-home is used, --with-arch-directory=ARCH must also
1854 @item --with-arch-directory=ARCH
1855 Specifies the name to use for the @file{jre/lib/ARCH} directory in the SDK
1856 environment created when --enable-java-home is passed. Typical names for this
1857 directory include i386, amd64, ia64, etc.
1859 @item --with-os-directory=DIR
1860 Specifies the OS directory for the SDK include directory. This is set to auto
1861 detect, and is typically 'linux'.
1863 @item --with-origin-name=NAME
1864 Specifies the JPackage origin name. This defaults to the 'gcj' in
1867 @item --with-arch-suffix=SUFFIX
1868 Specifies the suffix for the sdk directory. Defaults to the empty string.
1869 Examples include '.x86_64' in 'java-1.5.0-gcj-1.5.0.0.x86_64'.
1871 @item --with-jvm-root-dir=DIR
1872 Specifies where to install the SDK. Default is $(prefix)/lib/jvm.
1874 @item --with-jvm-jar-dir=DIR
1875 Specifies where to install jars. Default is $(prefix)/lib/jvm-exports.
1877 @item --with-python-dir=DIR
1878 Specifies where to install the Python modules used for aot-compile. DIR should
1879 not include the prefix used in installation. For example, if the Python modules
1880 are to be installed in /usr/lib/python2.5/site-packages, then
1881 --with-python-dir=/lib/python2.5/site-packages should be passed. If this is
1882 not specified, then the Python modules are installed in $(prefix)/share/python.
1884 @item --enable-aot-compile-rpm
1885 Adds aot-compile-rpm to the list of installed scripts.
1887 @item --enable-browser-plugin
1888 Build the gcjwebplugin web browser plugin.
1892 Use the single-byte @code{char} and the Win32 A functions natively,
1893 translating to and from UNICODE when using these functions. If
1894 unspecified, this is the default.
1897 Use the @code{WCHAR} and Win32 W functions natively. Adds
1898 @code{-lunicows} to @file{libgcj.spec} to link with @samp{libunicows}.
1899 @file{unicows.dll} needs to be deployed on Microsoft Windows 9X machines
1900 running built executables. @file{libunicows.a}, an open-source
1901 import library around Microsoft's @code{unicows.dll}, is obtained from
1902 @uref{http://libunicows.sourceforge.net/}, which also gives details
1903 on getting @file{unicows.dll} from Microsoft.
1906 Use the @code{WCHAR} and Win32 W functions natively. Does @emph{not}
1907 add @code{-lunicows} to @file{libgcj.spec}. The built executables will
1908 only run on Microsoft Windows NT and above.
1912 Enable support for link-time optimization (LTO). This is enabled by
1913 default if a working libelf implementation is found (see
1914 @option{--with-libelf}).
1916 @item --with-libelf=@var{pathname}
1917 @itemx --with-libelf-include=@var{pathname}
1918 @itemx --with-libelf-lib=@var{pathname}
1919 If you do not have libelf installed in a standard location and you
1920 want to enable support for link-time optimization (LTO), you can
1921 explicitly specify the directory where libelf is installed
1922 (@samp{--with-libelf=@var{libelfinstalldir}}). The
1923 @option{--with-libelf=@var{libelfinstalldir}} option is shorthand for
1924 @option{--with-libelf-include=@var{libelfinstalldir}/include}
1925 @option{--with-libelf-lib=@var{libelfinstalldir}/lib}.
1928 Enable support for using @command{gold} as the linker. If gold support is
1929 enabled together with @option{--enable-lto}, an additional directory
1930 @file{lto-plugin} will be built. The code in this directory is a
1931 plugin for gold that allows the link-time optimizer to extract object
1932 files with LTO information out of library archives. See
1933 @option{-flto} and @option{-fwhopr} for details.
1936 @subsubheading AWT-Specific Options
1940 Use the X Window System.
1942 @item --enable-java-awt=PEER(S)
1943 Specifies the AWT peer library or libraries to build alongside
1944 @samp{libgcj}. If this option is unspecified or disabled, AWT
1945 will be non-functional. Current valid values are @option{gtk} and
1946 @option{xlib}. Multiple libraries should be separated by a
1947 comma (i.e.@: @option{--enable-java-awt=gtk,xlib}).
1949 @item --enable-gtk-cairo
1950 Build the cairo Graphics2D implementation on GTK@.
1952 @item --enable-java-gc=TYPE
1953 Choose garbage collector. Defaults to @option{boehm} if unspecified.
1955 @item --disable-gtktest
1956 Do not try to compile and run a test GTK+ program.
1958 @item --disable-glibtest
1959 Do not try to compile and run a test GLIB program.
1961 @item --with-libart-prefix=PFX
1962 Prefix where libart is installed (optional).
1964 @item --with-libart-exec-prefix=PFX
1965 Exec prefix where libart is installed (optional).
1967 @item --disable-libarttest
1968 Do not try to compile and run a test libart program.
1977 @uref{./index.html,,Return to the GCC Installation page}
1981 @c ***Building****************************************************************
1983 @comment node-name, next, previous, up
1984 @node Building, Testing, Configuration, Installing GCC
1990 @cindex Installing GCC: Building
1992 Now that GCC is configured, you are ready to build the compiler and
1995 Some commands executed when making the compiler may fail (return a
1996 nonzero status) and be ignored by @command{make}. These failures, which
1997 are often due to files that were not found, are expected, and can safely
2000 It is normal to have compiler warnings when compiling certain files.
2001 Unless you are a GCC developer, you can generally ignore these warnings
2002 unless they cause compilation to fail. Developers should attempt to fix
2003 any warnings encountered, however they can temporarily continue past
2004 warnings-as-errors by specifying the configure flag
2005 @option{--disable-werror}.
2007 On certain old systems, defining certain environment variables such as
2008 @env{CC} can interfere with the functioning of @command{make}.
2010 If you encounter seemingly strange errors when trying to build the
2011 compiler in a directory other than the source directory, it could be
2012 because you have previously configured the compiler in the source
2013 directory. Make sure you have done all the necessary preparations.
2015 If you build GCC on a BSD system using a directory stored in an old System
2016 V file system, problems may occur in running @command{fixincludes} if the
2017 System V file system doesn't support symbolic links. These problems
2018 result in a failure to fix the declaration of @code{size_t} in
2019 @file{sys/types.h}. If you find that @code{size_t} is a signed type and
2020 that type mismatches occur, this could be the cause.
2022 The solution is not to use such a directory for building GCC@.
2024 Similarly, when building from SVN or snapshots, or if you modify
2025 @file{*.l} files, you need the Flex lexical analyzer generator
2026 installed. If you do not modify @file{*.l} files, releases contain
2027 the Flex-generated files and you do not need Flex installed to build
2028 them. There is still one Flex-based lexical analyzer (part of the
2029 build machinery, not of GCC itself) that is used even if you only
2030 build the C front end.
2032 When building from SVN or snapshots, or if you modify Texinfo
2033 documentation, you need version 4.7 or later of Texinfo installed if you
2034 want Info documentation to be regenerated. Releases contain Info
2035 documentation pre-built for the unmodified documentation in the release.
2037 @section Building a native compiler
2039 For a native build, the default configuration is to perform
2040 a 3-stage bootstrap of the compiler when @samp{make} is invoked.
2041 This will build the entire GCC system and ensure that it compiles
2042 itself correctly. It can be disabled with the @option{--disable-bootstrap}
2043 parameter to @samp{configure}, but bootstrapping is suggested because
2044 the compiler will be tested more completely and could also have
2047 The bootstrapping process will complete the following steps:
2051 Build tools necessary to build the compiler.
2054 Perform a 3-stage bootstrap of the compiler. This includes building
2055 three times the target tools for use by the compiler such as binutils
2056 (bfd, binutils, gas, gprof, ld, and opcodes) if they have been
2057 individually linked or moved into the top level GCC source tree before
2061 Perform a comparison test of the stage2 and stage3 compilers.
2064 Build runtime libraries using the stage3 compiler from the previous step.
2068 If you are short on disk space you might consider @samp{make
2069 bootstrap-lean} instead. The sequence of compilation is the
2070 same described above, but object files from the stage1 and
2071 stage2 of the 3-stage bootstrap of the compiler are deleted as
2072 soon as they are no longer needed.
2074 If you wish to use non-default GCC flags when compiling the stage2
2075 and stage3 compilers, set @code{BOOT_CFLAGS} on the command line when
2076 doing @samp{make}. For example, if you want to save additional space
2077 during the bootstrap and in the final installation as well, you can
2078 build the compiler binaries without debugging information as in the
2079 following example. This will save roughly 40% of disk space both for
2080 the bootstrap and the final installation. (Libraries will still contain
2081 debugging information.)
2084 make BOOT_CFLAGS='-O' bootstrap
2087 You can place non-default optimization flags into @code{BOOT_CFLAGS}; they
2088 are less well tested here than the default of @samp{-g -O2}, but should
2089 still work. In a few cases, you may find that you need to specify special
2090 flags such as @option{-msoft-float} here to complete the bootstrap; or,
2091 if the native compiler miscompiles the stage1 compiler, you may need
2092 to work around this, by choosing @code{BOOT_CFLAGS} to avoid the parts
2093 of the stage1 compiler that were miscompiled, or by using @samp{make
2094 bootstrap4} to increase the number of stages of bootstrap.
2096 @code{BOOT_CFLAGS} does not apply to bootstrapped target libraries.
2097 Since these are always compiled with the compiler currently being
2098 bootstrapped, you can use @code{CFLAGS_FOR_TARGET} to modify their
2099 compilation flags, as for non-bootstrapped target libraries.
2100 Again, if the native compiler miscompiles the stage1 compiler, you may
2101 need to work around this by avoiding non-working parts of the stage1
2102 compiler. Use @code{STAGE1_TFLAGS} to this end.
2104 If you used the flag @option{--enable-languages=@dots{}} to restrict
2105 the compilers to be built, only those you've actually enabled will be
2106 built. This will of course only build those runtime libraries, for
2107 which the particular compiler has been built. Please note,
2108 that re-defining @env{LANGUAGES} when calling @samp{make}
2109 @strong{does not} work anymore!
2111 If the comparison of stage2 and stage3 fails, this normally indicates
2112 that the stage2 compiler has compiled GCC incorrectly, and is therefore
2113 a potentially serious bug which you should investigate and report. (On
2114 a few systems, meaningful comparison of object files is impossible; they
2115 always appear ``different''. If you encounter this problem, you will
2116 need to disable comparison in the @file{Makefile}.)
2118 If you do not want to bootstrap your compiler, you can configure with
2119 @option{--disable-bootstrap}. In particular cases, you may want to
2120 bootstrap your compiler even if the target system is not the same as
2121 the one you are building on: for example, you could build a
2122 @code{powerpc-unknown-linux-gnu} toolchain on a
2123 @code{powerpc64-unknown-linux-gnu} host. In this case, pass
2124 @option{--enable-bootstrap} to the configure script.
2126 @code{BUILD_CONFIG} can be used to bring in additional customization
2127 to the build. It can be set to a whitespace-separated list of names.
2128 For each such @code{NAME}, top-level @file{config/@code{NAME}.mk} will
2129 be included by the top-level @file{Makefile}, bringing in any settings
2130 it contains. The default @code{BUILD_CONFIG} can be set using the
2131 configure option @option{--with-build-config=@code{NAME}...}. Some
2132 examples of supported build configurations are:
2135 @item @samp{bootstrap-O1}
2136 Removes any @option{-O}-started option from @code{BOOT_CFLAGS}, and adds
2137 @option{-O1} to it. @samp{BUILD_CONFIG=bootstrap-O1} is equivalent to
2138 @samp{BOOT_CFLAGS='-g -O1'}.
2140 @item @samp{bootstrap-O3}
2141 Analogous to @code{bootstrap-O1}.
2143 @item @samp{bootstrap-debug}
2144 Verifies that the compiler generates the same executable code, whether
2145 or not it is asked to emit debug information. To this end, this
2146 option builds stage2 host programs without debug information, and uses
2147 @file{contrib/compare-debug} to compare them with the stripped stage3
2148 object files. If @code{BOOT_CFLAGS} is overridden so as to not enable
2149 debug information, stage2 will have it, and stage3 won't. This option
2150 is enabled by default when GCC bootstrapping is enabled, if
2151 @code{strip} can turn object files compiled with and without debug
2152 info into identical object files. In addition to better test
2153 coverage, this option makes default bootstraps faster and leaner.
2155 @item @samp{bootstrap-debug-big}
2156 Rather than comparing stripped object files, as in
2157 @code{bootstrap-debug}, this option saves internal compiler dumps
2158 during stage2 and stage3 and compares them as well, which helps catch
2159 additional potential problems, but at a great cost in terms of disk
2160 space. It can be specified in addition to @samp{bootstrap-debug}.
2162 @item @samp{bootstrap-debug-lean}
2163 This option saves disk space compared with @code{bootstrap-debug-big},
2164 but at the expense of some recompilation. Instead of saving the dumps
2165 of stage2 and stage3 until the final compare, it uses
2166 @option{-fcompare-debug} to generate, compare and remove the dumps
2167 during stage3, repeating the compilation that already took place in
2168 stage2, whose dumps were not saved.
2170 @item @samp{bootstrap-debug-lib}
2171 This option tests executable code invariance over debug information
2172 generation on target libraries, just like @code{bootstrap-debug-lean}
2173 tests it on host programs. It builds stage3 libraries with
2174 @option{-fcompare-debug}, and it can be used along with any of the
2175 @code{bootstrap-debug} options above.
2177 There aren't @code{-lean} or @code{-big} counterparts to this option
2178 because most libraries are only build in stage3, so bootstrap compares
2179 would not get significant coverage. Moreover, the few libraries built
2180 in stage2 are used in stage3 host programs, so we wouldn't want to
2181 compile stage2 libraries with different options for comparison purposes.
2183 @item @samp{bootstrap-debug-ckovw}
2184 Arranges for error messages to be issued if the compiler built on any
2185 stage is run without the option @option{-fcompare-debug}. This is
2186 useful to verify the full @option{-fcompare-debug} testing coverage. It
2187 must be used along with @code{bootstrap-debug-lean} and
2188 @code{bootstrap-debug-lib}.
2190 @item @samp{bootstrap-time}
2191 Arranges for the run time of each program started by the GCC driver,
2192 built in any stage, to be logged to @file{time.log}, in the top level of
2197 @section Building a cross compiler
2199 When building a cross compiler, it is not generally possible to do a
2200 3-stage bootstrap of the compiler. This makes for an interesting problem
2201 as parts of GCC can only be built with GCC@.
2203 To build a cross compiler, we recommend first building and installing a
2204 native compiler. You can then use the native GCC compiler to build the
2205 cross compiler. The installed native compiler needs to be GCC version
2208 If the cross compiler is to be built with support for the Java
2209 programming language and the ability to compile .java source files is
2210 desired, the installed native compiler used to build the cross
2211 compiler needs to be the same GCC version as the cross compiler. In
2212 addition the cross compiler needs to be configured with
2213 @option{--with-ecj-jar=@dots{}}.
2215 Assuming you have already installed a native copy of GCC and configured
2216 your cross compiler, issue the command @command{make}, which performs the
2221 Build host tools necessary to build the compiler.
2224 Build target tools for use by the compiler such as binutils (bfd,
2225 binutils, gas, gprof, ld, and opcodes)
2226 if they have been individually linked or moved into the top level GCC source
2227 tree before configuring.
2230 Build the compiler (single stage only).
2233 Build runtime libraries using the compiler from the previous step.
2236 Note that if an error occurs in any step the make process will exit.
2238 If you are not building GNU binutils in the same source tree as GCC,
2239 you will need a cross-assembler and cross-linker installed before
2240 configuring GCC@. Put them in the directory
2241 @file{@var{prefix}/@var{target}/bin}. Here is a table of the tools
2242 you should put in this directory:
2246 This should be the cross-assembler.
2249 This should be the cross-linker.
2252 This should be the cross-archiver: a program which can manipulate
2253 archive files (linker libraries) in the target machine's format.
2256 This should be a program to construct a symbol table in an archive file.
2259 The installation of GCC will find these programs in that directory,
2260 and copy or link them to the proper place to for the cross-compiler to
2261 find them when run later.
2263 The easiest way to provide these files is to build the Binutils package.
2264 Configure it with the same @option{--host} and @option{--target}
2265 options that you use for configuring GCC, then build and install
2266 them. They install their executables automatically into the proper
2267 directory. Alas, they do not support all the targets that GCC
2270 If you are not building a C library in the same source tree as GCC,
2271 you should also provide the target libraries and headers before
2272 configuring GCC, specifying the directories with
2273 @option{--with-sysroot} or @option{--with-headers} and
2274 @option{--with-libs}. Many targets also require ``start files'' such
2275 as @file{crt0.o} and
2276 @file{crtn.o} which are linked into each executable. There may be several
2277 alternatives for @file{crt0.o}, for use with profiling or other
2278 compilation options. Check your target's definition of
2279 @code{STARTFILE_SPEC} to find out what start files it uses.
2281 @section Building in parallel
2283 GNU Make 3.80 and above, which is necessary to build GCC, support
2284 building in parallel. To activate this, you can use @samp{make -j 2}
2285 instead of @samp{make}. You can also specify a bigger number, and
2286 in most cases using a value greater than the number of processors in
2287 your machine will result in fewer and shorter I/O latency hits, thus
2288 improving overall throughput; this is especially true for slow drives
2289 and network filesystems.
2291 @section Building the Ada compiler
2293 In order to build GNAT, the Ada compiler, you need a working GNAT
2294 compiler (GCC version 3.4 or later).
2295 This includes GNAT tools such as @command{gnatmake} and
2296 @command{gnatlink}, since the Ada front end is written in Ada and
2297 uses some GNAT-specific extensions.
2299 In order to build a cross compiler, it is suggested to install
2300 the new compiler as native first, and then use it to build the cross
2303 @command{configure} does not test whether the GNAT installation works
2304 and has a sufficiently recent version; if too old a GNAT version is
2305 installed, the build will fail unless @option{--enable-languages} is
2306 used to disable building the Ada front end.
2308 @env{ADA_INCLUDE_PATH} and @env{ADA_OBJECT_PATH} environment variables
2309 must not be set when building the Ada compiler, the Ada tools, or the
2310 Ada runtime libraries. You can check that your build environment is clean
2311 by verifying that @samp{gnatls -v} lists only one explicit path in each
2314 @section Building with profile feedback
2316 It is possible to use profile feedback to optimize the compiler itself. This
2317 should result in a faster compiler binary. Experiments done on x86 using gcc
2318 3.3 showed approximately 7 percent speedup on compiling C programs. To
2319 bootstrap the compiler with profile feedback, use @code{make profiledbootstrap}.
2321 When @samp{make profiledbootstrap} is run, it will first build a @code{stage1}
2322 compiler. This compiler is used to build a @code{stageprofile} compiler
2323 instrumented to collect execution counts of instruction and branch
2324 probabilities. Then runtime libraries are compiled with profile collected.
2325 Finally a @code{stagefeedback} compiler is built using the information collected.
2327 Unlike standard bootstrap, several additional restrictions apply. The
2328 compiler used to build @code{stage1} needs to support a 64-bit integral type.
2329 It is recommended to only use GCC for this. Also parallel make is currently
2330 not supported since collisions in profile collecting may occur.
2337 @uref{./index.html,,Return to the GCC Installation page}
2341 @c ***Testing*****************************************************************
2343 @comment node-name, next, previous, up
2344 @node Testing, Final install, Building, Installing GCC
2348 @chapter Installing GCC: Testing
2351 @cindex Installing GCC: Testing
2354 Before you install GCC, we encourage you to run the testsuites and to
2355 compare your results with results from a similar configuration that have
2356 been submitted to the
2357 @uref{http://gcc.gnu.org/ml/gcc-testresults/,,gcc-testresults mailing list}.
2358 Some of these archived results are linked from the build status lists
2359 at @uref{http://gcc.gnu.org/buildstat.html}, although not everyone who
2360 reports a successful build runs the testsuites and submits the results.
2361 This step is optional and may require you to download additional software,
2362 but it can give you confidence in your new GCC installation or point out
2363 problems before you install and start using your new GCC@.
2365 First, you must have @uref{download.html,,downloaded the testsuites}.
2366 These are part of the full distribution, but if you downloaded the
2367 ``core'' compiler plus any front ends, you must download the testsuites
2370 Second, you must have the testing tools installed. This includes
2371 @uref{http://www.gnu.org/software/dejagnu/,,DejaGnu}, Tcl, and Expect;
2372 the DejaGnu site has links to these.
2374 If the directories where @command{runtest} and @command{expect} were
2375 installed are not in the @env{PATH}, you may need to set the following
2376 environment variables appropriately, as in the following example (which
2377 assumes that DejaGnu has been installed under @file{/usr/local}):
2380 TCL_LIBRARY = /usr/local/share/tcl8.0
2381 DEJAGNULIBS = /usr/local/share/dejagnu
2384 (On systems such as Cygwin, these paths are required to be actual
2385 paths, not mounts or links; presumably this is due to some lack of
2386 portability in the DejaGnu code.)
2389 Finally, you can run the testsuite (which may take a long time):
2391 cd @var{objdir}; make -k check
2394 This will test various components of GCC, such as compiler
2395 front ends and runtime libraries. While running the testsuite, DejaGnu
2396 might emit some harmless messages resembling
2397 @samp{WARNING: Couldn't find the global config file.} or
2398 @samp{WARNING: Couldn't find tool init file} that can be ignored.
2400 If you are testing a cross-compiler, you may want to run the testsuite
2401 on a simulator as described at @uref{http://gcc.gnu.org/simtest-howto.html}.
2403 @section How can you run the testsuite on selected tests?
2405 In order to run sets of tests selectively, there are targets
2406 @samp{make check-gcc} and @samp{make check-g++}
2407 in the @file{gcc} subdirectory of the object directory. You can also
2408 just run @samp{make check} in a subdirectory of the object directory.
2411 A more selective way to just run all @command{gcc} execute tests in the
2415 make check-gcc RUNTESTFLAGS="execute.exp @var{other-options}"
2418 Likewise, in order to run only the @command{g++} ``old-deja'' tests in
2419 the testsuite with filenames matching @samp{9805*}, you would use
2422 make check-g++ RUNTESTFLAGS="old-deja.exp=9805* @var{other-options}"
2425 The @file{*.exp} files are located in the testsuite directories of the GCC
2426 source, the most important ones being @file{compile.exp},
2427 @file{execute.exp}, @file{dg.exp} and @file{old-deja.exp}.
2428 To get a list of the possible @file{*.exp} files, pipe the
2429 output of @samp{make check} into a file and look at the
2430 @samp{Running @dots{} .exp} lines.
2432 @section Passing options and running multiple testsuites
2434 You can pass multiple options to the testsuite using the
2435 @samp{--target_board} option of DejaGNU, either passed as part of
2436 @samp{RUNTESTFLAGS}, or directly to @command{runtest} if you prefer to
2437 work outside the makefiles. For example,
2440 make check-g++ RUNTESTFLAGS="--target_board=unix/-O3/-fmerge-constants"
2443 will run the standard @command{g++} testsuites (``unix'' is the target name
2444 for a standard native testsuite situation), passing
2445 @samp{-O3 -fmerge-constants} to the compiler on every test, i.e.,
2446 slashes separate options.
2448 You can run the testsuites multiple times using combinations of options
2449 with a syntax similar to the brace expansion of popular shells:
2452 @dots{}"--target_board=arm-sim\@{-mhard-float,-msoft-float\@}\@{-O1,-O2,-O3,\@}"
2455 (Note the empty option caused by the trailing comma in the final group.)
2456 The following will run each testsuite eight times using the @samp{arm-sim}
2457 target, as if you had specified all possible combinations yourself:
2460 --target_board=arm-sim/-mhard-float/-O1
2461 --target_board=arm-sim/-mhard-float/-O2
2462 --target_board=arm-sim/-mhard-float/-O3
2463 --target_board=arm-sim/-mhard-float
2464 --target_board=arm-sim/-msoft-float/-O1
2465 --target_board=arm-sim/-msoft-float/-O2
2466 --target_board=arm-sim/-msoft-float/-O3
2467 --target_board=arm-sim/-msoft-float
2470 They can be combined as many times as you wish, in arbitrary ways. This
2474 @dots{}"--target_board=unix/-Wextra\@{-O3,-fno-strength\@}\@{-fomit-frame,\@}"
2477 will generate four combinations, all involving @samp{-Wextra}.
2479 The disadvantage to this method is that the testsuites are run in serial,
2480 which is a waste on multiprocessor systems. For users with GNU Make and
2481 a shell which performs brace expansion, you can run the testsuites in
2482 parallel by having the shell perform the combinations and @command{make}
2483 do the parallel runs. Instead of using @samp{--target_board}, use a
2484 special makefile target:
2487 make -j@var{N} check-@var{testsuite}//@var{test-target}/@var{option1}/@var{option2}/@dots{}
2493 make -j3 check-gcc//sh-hms-sim/@{-m1,-m2,-m3,-m3e,-m4@}/@{,-nofpu@}
2496 will run three concurrent ``make-gcc'' testsuites, eventually testing all
2497 ten combinations as described above. Note that this is currently only
2498 supported in the @file{gcc} subdirectory. (To see how this works, try
2499 typing @command{echo} before the example given here.)
2502 @section Additional testing for Java Class Libraries
2504 The Java runtime tests can be executed via @samp{make check}
2505 in the @file{@var{target}/libjava/testsuite} directory in
2508 The @uref{http://sourceware.org/mauve/,,Mauve Project} provides
2509 a suite of tests for the Java Class Libraries. This suite can be run
2510 as part of libgcj testing by placing the Mauve tree within the libjava
2511 testsuite at @file{libjava/testsuite/libjava.mauve/mauve}, or by
2512 specifying the location of that tree when invoking @samp{make}, as in
2513 @samp{make MAUVEDIR=~/mauve check}.
2515 @section How to interpret test results
2517 The result of running the testsuite are various @file{*.sum} and @file{*.log}
2518 files in the testsuite subdirectories. The @file{*.log} files contain a
2519 detailed log of the compiler invocations and the corresponding
2520 results, the @file{*.sum} files summarize the results. These summaries
2521 contain status codes for all tests:
2525 PASS: the test passed as expected
2527 XPASS: the test unexpectedly passed
2529 FAIL: the test unexpectedly failed
2531 XFAIL: the test failed as expected
2533 UNSUPPORTED: the test is not supported on this platform
2535 ERROR: the testsuite detected an error
2537 WARNING: the testsuite detected a possible problem
2540 It is normal for some tests to report unexpected failures. At the
2541 current time the testing harness does not allow fine grained control
2542 over whether or not a test is expected to fail. This problem should
2543 be fixed in future releases.
2546 @section Submitting test results
2548 If you want to report the results to the GCC project, use the
2549 @file{contrib/test_summary} shell script. Start it in the @var{objdir} with
2552 @var{srcdir}/contrib/test_summary -p your_commentary.txt \
2553 -m gcc-testresults@@gcc.gnu.org |sh
2556 This script uses the @command{Mail} program to send the results, so
2557 make sure it is in your @env{PATH}. The file @file{your_commentary.txt} is
2558 prepended to the testsuite summary and should contain any special
2559 remarks you have on your results or your build environment. Please
2560 do not edit the testsuite result block or the subject line, as these
2561 messages may be automatically processed.
2568 @uref{./index.html,,Return to the GCC Installation page}
2572 @c ***Final install***********************************************************
2574 @comment node-name, next, previous, up
2575 @node Final install, , Testing, Installing GCC
2577 @ifset finalinstallhtml
2579 @chapter Installing GCC: Final installation
2582 Now that GCC has been built (and optionally tested), you can install it with
2584 cd @var{objdir}; make install
2587 We strongly recommend to install into a target directory where there is
2588 no previous version of GCC present. Also, the GNAT runtime should not
2589 be stripped, as this would break certain features of the debugger that
2590 depend on this debugging information (catching Ada exceptions for
2593 That step completes the installation of GCC; user level binaries can
2594 be found in @file{@var{prefix}/bin} where @var{prefix} is the value
2595 you specified with the @option{--prefix} to configure (or
2596 @file{/usr/local} by default). (If you specified @option{--bindir},
2597 that directory will be used instead; otherwise, if you specified
2598 @option{--exec-prefix}, @file{@var{exec-prefix}/bin} will be used.)
2599 Headers for the C++ and Java libraries are installed in
2600 @file{@var{prefix}/include}; libraries in @file{@var{libdir}}
2601 (normally @file{@var{prefix}/lib}); internal parts of the compiler in
2602 @file{@var{libdir}/gcc} and @file{@var{libexecdir}/gcc}; documentation
2603 in info format in @file{@var{infodir}} (normally
2604 @file{@var{prefix}/info}).
2606 When installing cross-compilers, GCC's executables
2607 are not only installed into @file{@var{bindir}}, that
2608 is, @file{@var{exec-prefix}/bin}, but additionally into
2609 @file{@var{exec-prefix}/@var{target-alias}/bin}, if that directory
2610 exists. Typically, such @dfn{tooldirs} hold target-specific
2611 binutils, including assembler and linker.
2613 Installation into a temporary staging area or into a @command{chroot}
2614 jail can be achieved with the command
2617 make DESTDIR=@var{path-to-rootdir} install
2620 @noindent where @var{path-to-rootdir} is the absolute path of
2621 a directory relative to which all installation paths will be
2622 interpreted. Note that the directory specified by @code{DESTDIR}
2623 need not exist yet; it will be created if necessary.
2625 There is a subtle point with tooldirs and @code{DESTDIR}:
2626 If you relocate a cross-compiler installation with
2627 e.g.@: @samp{DESTDIR=@var{rootdir}}, then the directory
2628 @file{@var{rootdir}/@var{exec-prefix}/@var{target-alias}/bin} will
2629 be filled with duplicated GCC executables only if it already exists,
2630 it will not be created otherwise. This is regarded as a feature,
2631 not as a bug, because it gives slightly more control to the packagers
2632 using the @code{DESTDIR} feature.
2634 If you are bootstrapping a released version of GCC then please
2635 quickly review the build status page for your release, available from
2636 @uref{http://gcc.gnu.org/buildstat.html}.
2637 If your system is not listed for the version of GCC that you built,
2639 @email{gcc@@gcc.gnu.org} indicating
2640 that you successfully built and installed GCC@.
2641 Include the following information:
2645 Output from running @file{@var{srcdir}/config.guess}. Do not send
2646 that file itself, just the one-line output from running it.
2649 The output of @samp{gcc -v} for your newly installed @command{gcc}.
2650 This tells us which version of GCC you built and the options you passed to
2654 Whether you enabled all languages or a subset of them. If you used a
2655 full distribution then this information is part of the configure
2656 options in the output of @samp{gcc -v}, but if you downloaded the
2657 ``core'' compiler plus additional front ends then it isn't apparent
2658 which ones you built unless you tell us about it.
2661 If the build was for GNU/Linux, also include:
2664 The distribution name and version (e.g., Red Hat 7.1 or Debian 2.2.3);
2665 this information should be available from @file{/etc/issue}.
2668 The version of the Linux kernel, available from @samp{uname --version}
2672 The version of glibc you used; for RPM-based systems like Red Hat,
2673 Mandrake, and SuSE type @samp{rpm -q glibc} to get the glibc version,
2674 and on systems like Debian and Progeny use @samp{dpkg -l libc6}.
2676 For other systems, you can include similar information if you think it is
2680 Any other information that you think would be useful to people building
2681 GCC on the same configuration. The new entry in the build status list
2682 will include a link to the archived copy of your message.
2685 We'd also like to know if the
2687 @ref{Specific, host/target specific installation notes}
2690 @uref{specific.html,,host/target specific installation notes}
2692 didn't include your host/target information or if that information is
2693 incomplete or out of date. Send a note to
2694 @email{gcc@@gcc.gnu.org} detailing how the information should be changed.
2696 If you find a bug, please report it following the
2697 @uref{../bugs/,,bug reporting guidelines}.
2699 If you want to print the GCC manuals, do @samp{cd @var{objdir}; make
2700 dvi}. You will need to have @command{texi2dvi} (version at least 4.7)
2701 and @TeX{} installed. This creates a number of @file{.dvi} files in
2702 subdirectories of @file{@var{objdir}}; these may be converted for
2703 printing with programs such as @command{dvips}. Alternately, by using
2704 @samp{make pdf} in place of @samp{make dvi}, you can create documentation
2705 in the form of @file{.pdf} files; this requires @command{texi2pdf}, which
2706 is included with Texinfo version 4.8 and later. You can also
2707 @uref{http://shop.fsf.org/,,buy printed manuals from the
2708 Free Software Foundation}, though such manuals may not be for the most
2709 recent version of GCC@.
2711 If you would like to generate online HTML documentation, do @samp{cd
2712 @var{objdir}; make html} and HTML will be generated for the gcc manuals in
2713 @file{@var{objdir}/gcc/HTML}.
2720 @uref{./index.html,,Return to the GCC Installation page}
2724 @c ***Binaries****************************************************************
2726 @comment node-name, next, previous, up
2727 @node Binaries, Specific, Installing GCC, Top
2731 @chapter Installing GCC: Binaries
2734 @cindex Installing GCC: Binaries
2736 We are often asked about pre-compiled versions of GCC@. While we cannot
2737 provide these for all platforms, below you'll find links to binaries for
2738 various platforms where creating them by yourself is not easy due to various
2741 Please note that we did not create these binaries, nor do we
2742 support them. If you have any problems installing them, please
2743 contact their makers.
2750 @uref{http://www.bullfreeware.com,,Bull's Freeware and Shareware Archive for AIX};
2753 @uref{http://pware.hvcc.edu,,Hudson Valley Community College Open Source Software for IBM System p};
2756 @uref{http://www.perzl.org/aix/,,AIX 5L and 6 Open Source Packages}.
2760 DOS---@uref{http://www.delorie.com/djgpp/,,DJGPP}.
2763 Renesas H8/300[HS]---@uref{http://h8300-hms.sourceforge.net/,,GNU
2764 Development Tools for the Renesas H8/300[HS] Series}.
2770 @uref{http://hpux.cs.utah.edu/,,HP-UX Porting Center};
2773 @uref{ftp://sunsite.informatik.rwth-aachen.de/pub/packages/gcc_hpux/,,Binaries for HP-UX 11.00 at Aachen University of Technology}.
2777 Motorola 68HC11/68HC12---@uref{http://www.gnu-m68hc11.org,,GNU
2778 Development Tools for the Motorola 68HC11/68HC12}.
2781 @uref{http://www.sco.com/skunkware/devtools/index.html#gcc,,SCO
2782 OpenServer/Unixware}.
2785 Solaris 2 (SPARC, Intel)---@uref{http://www.sunfreeware.com/,,Sunfreeware}.
2788 SGI---@uref{http://freeware.sgi.com/,,SGI Freeware}.
2794 The @uref{http://sourceware.org/cygwin/,,Cygwin} project;
2796 The @uref{http://www.mingw.org/,,MinGW} project.
2800 @uref{ftp://ftp.thewrittenword.com/packages/by-name/,,The
2801 Written Word} offers binaries for
2802 AIX 4.3.3, 5.1 and 5.2,
2804 Tru64 UNIX 4.0D and 5.1,
2806 HP-UX 10.20, 11.00, and 11.11, and
2807 Solaris/SPARC 2.5.1, 2.6, 7, 8, 9 and 10.
2810 @uref{http://www.openpkg.org/,,OpenPKG} offers binaries for quite a
2811 number of platforms.
2814 The @uref{http://gcc.gnu.org/wiki/GFortranBinaries,,GFortran Wiki} has
2815 links to GNU Fortran binaries for several platforms.
2823 @uref{./index.html,,Return to the GCC Installation page}
2827 @c ***Specific****************************************************************
2829 @comment node-name, next, previous, up
2830 @node Specific, Old, Binaries, Top
2834 @chapter Host/target specific installation notes for GCC
2837 @cindex Specific installation notes
2838 @cindex Target specific installation
2839 @cindex Host specific installation
2840 @cindex Target specific installation notes
2842 Please read this document carefully @emph{before} installing the
2843 GNU Compiler Collection on your machine.
2845 Note that this list of install notes is @emph{not} a list of supported
2846 hosts or targets. Not all supported hosts and targets are listed
2847 here, only the ones that require host-specific or target-specific
2853 @uref{#alpha-x-x,,alpha*-*-*}
2855 @uref{#alpha-dec-osf,,alpha*-dec-osf*}
2857 @uref{#arc-x-elf,,arc-*-elf}
2859 @uref{#arm-x-elf,,arm-*-elf}
2863 @uref{#bfin,,Blackfin}
2867 @uref{#x-x-freebsd,,*-*-freebsd*}
2869 @uref{#h8300-hms,,h8300-hms}
2871 @uref{#hppa-hp-hpux,,hppa*-hp-hpux*}
2873 @uref{#hppa-hp-hpux10,,hppa*-hp-hpux10}
2875 @uref{#hppa-hp-hpux11,,hppa*-hp-hpux11}
2877 @uref{#x-x-linux-gnu,,*-*-linux-gnu}
2879 @uref{#ix86-x-linux,,i?86-*-linux*}
2881 @uref{#ix86-x-solaris210,,i?86-*-solaris2.10}
2883 @uref{#ia64-x-linux,,ia64-*-linux}
2885 @uref{#ia64-x-hpux,,ia64-*-hpux*}
2887 @uref{#x-ibm-aix,,*-ibm-aix*}
2889 @uref{#iq2000-x-elf,,iq2000-*-elf}
2891 @uref{#lm32-x-elf,,lm32-*-elf}
2893 @uref{#lm32-x-uclinux,,lm32-*-uclinux}
2895 @uref{#m32c-x-elf,,m32c-*-elf}
2897 @uref{#m32r-x-elf,,m32r-*-elf}
2899 @uref{#m6811-elf,,m6811-elf}
2901 @uref{#m6812-elf,,m6812-elf}
2903 @uref{#m68k-x-x,,m68k-*-*}
2905 @uref{#m68k-uclinux,,m68k-uclinux}
2907 @uref{#mep-x-elf,,mep-*-elf}
2909 @uref{#mips-x-x,,mips-*-*}
2911 @uref{#mips-sgi-irix5,,mips-sgi-irix5}
2913 @uref{#mips-sgi-irix6,,mips-sgi-irix6}
2915 @uref{#powerpc-x-x,,powerpc*-*-*}
2917 @uref{#powerpc-x-darwin,,powerpc-*-darwin*}
2919 @uref{#powerpc-x-elf,,powerpc-*-elf}
2921 @uref{#powerpc-x-linux-gnu,,powerpc*-*-linux-gnu*}
2923 @uref{#powerpc-x-netbsd,,powerpc-*-netbsd*}
2925 @uref{#powerpc-x-eabisim,,powerpc-*-eabisim}
2927 @uref{#powerpc-x-eabi,,powerpc-*-eabi}
2929 @uref{#powerpcle-x-elf,,powerpcle-*-elf}
2931 @uref{#powerpcle-x-eabisim,,powerpcle-*-eabisim}
2933 @uref{#powerpcle-x-eabi,,powerpcle-*-eabi}
2935 @uref{#s390-x-linux,,s390-*-linux*}
2937 @uref{#s390x-x-linux,,s390x-*-linux*}
2939 @uref{#s390x-ibm-tpf,,s390x-ibm-tpf*}
2941 @uref{#x-x-solaris2,,*-*-solaris2*}
2943 @uref{#sparc-sun-solaris2,,sparc-sun-solaris2*}
2945 @uref{#sparc-sun-solaris27,,sparc-sun-solaris2.7}
2947 @uref{#sparc-x-linux,,sparc-*-linux*}
2949 @uref{#sparc64-x-solaris2,,sparc64-*-solaris2*}
2951 @uref{#sparcv9-x-solaris2,,sparcv9-*-solaris2*}
2953 @uref{#x-x-vxworks,,*-*-vxworks*}
2955 @uref{#x86-64-x-x,,x86_64-*-*, amd64-*-*}
2957 @uref{#xtensa-x-elf,,xtensa*-*-elf}
2959 @uref{#xtensa-x-linux,,xtensa*-*-linux*}
2961 @uref{#windows,,Microsoft Windows}
2963 @uref{#x-x-cygwin,,*-*-cygwin}
2965 @uref{#x-x-interix,,*-*-interix}
2967 @uref{#x-x-mingw32,,*-*-mingw32}
2971 @uref{#older,,Older systems}
2976 @uref{#elf,,all ELF targets} (SVR4, Solaris 2, etc.)
2982 <!-- -------- host/target specific issues start here ---------------- -->
2985 @heading @anchor{alpha-x-x}alpha*-*-*
2987 This section contains general configuration information for all
2988 alpha-based platforms using ELF (in particular, ignore this section for
2989 DEC OSF/1, Digital UNIX and Tru64 UNIX)@. In addition to reading this
2990 section, please read all other sections that match your target.
2992 We require binutils 2.11.2 or newer.
2993 Previous binutils releases had a number of problems with DWARF 2
2994 debugging information, not the least of which is incorrect linking of
3000 @heading @anchor{alpha-dec-osf}alpha*-dec-osf*
3001 Systems using processors that implement the DEC Alpha architecture and
3002 are running the DEC/Compaq Unix (DEC OSF/1, Digital UNIX, or Compaq
3003 Tru64 UNIX) operating system, for example the DEC Alpha AXP systems.
3005 As of GCC 3.2, versions before @code{alpha*-dec-osf4} are no longer
3006 supported. (These are the versions which identify themselves as DEC
3009 In Digital Unix V4.0, virtual memory exhausted bootstrap failures
3010 may be fixed by configuring with @option{--with-gc=simple},
3011 reconfiguring Kernel Virtual Memory and Swap parameters
3012 per the @command{/usr/sbin/sys_check} Tuning Suggestions,
3013 or applying the patch in
3014 @uref{http://gcc.gnu.org/ml/gcc/2002-08/msg00822.html}.
3016 In Tru64 UNIX V5.1, Compaq introduced a new assembler that does not
3017 currently (2001-06-13) work with @command{mips-tfile}. As a workaround,
3018 we need to use the old assembler, invoked via the barely documented
3019 @option{-oldas} option. To bootstrap GCC, you either need to use the
3023 % CC=cc @var{srcdir}/configure [@var{options}] [@var{target}]
3026 or you can use a copy of GCC 2.95.3 or higher built on Tru64 UNIX V4.0:
3029 % CC=gcc -Wa,-oldas @var{srcdir}/configure [@var{options}] [@var{target}]
3032 As of GNU binutils 2.11.2, neither GNU @command{as} nor GNU @command{ld}
3033 are supported on Tru64 UNIX, so you must not configure GCC with
3034 @option{--with-gnu-as} or @option{--with-gnu-ld}.
3036 GCC writes a @samp{.verstamp} directive to the assembler output file
3037 unless it is built as a cross-compiler. It gets the version to use from
3038 the system header file @file{/usr/include/stamp.h}. If you install a
3039 new version of DEC Unix, you should rebuild GCC to pick up the new version
3042 @samp{make compare} may fail on old versions of DEC Unix unless you add
3043 @option{-save-temps} to @code{BOOT_CFLAGS}. On these systems, the name
3044 of the assembler input file is stored in the object file, and that makes
3045 comparison fail if it differs between the @code{stage1} and
3046 @code{stage2} compilations. The option @option{-save-temps} forces a
3047 fixed name to be used for the assembler input file, instead of a
3048 randomly chosen name in @file{/tmp}. Do not add @option{-save-temps}
3049 unless the comparisons fail without that option. If you add
3050 @option{-save-temps}, you will have to manually delete the @samp{.i} and
3051 @samp{.s} files after each series of compilations.
3053 GCC now supports both the native (ECOFF) debugging format used by DBX
3054 and GDB and an encapsulated STABS format for use only with GDB@. See the
3055 discussion of the @option{--with-stabs} option of @file{configure} above
3056 for more information on these formats and how to select them.
3058 There is a bug in DEC's assembler that produces incorrect line numbers
3059 for ECOFF format when the @samp{.align} directive is used. To work
3060 around this problem, GCC will not emit such alignment directives
3061 while writing ECOFF format debugging information even if optimization is
3062 being performed. Unfortunately, this has the very undesirable
3063 side-effect that code addresses when @option{-O} is specified are
3064 different depending on whether or not @option{-g} is also specified.
3066 To avoid this behavior, specify @option{-gstabs+} and use GDB instead of
3067 DBX@. DEC is now aware of this problem with the assembler and hopes to
3068 provide a fix shortly.
3073 @heading @anchor{arc-x-elf}arc-*-elf
3074 Argonaut ARC processor.
3075 This configuration is intended for embedded systems.
3080 @heading @anchor{arm-x-elf}arm-*-elf
3081 ARM-family processors. Subtargets that use the ELF object format
3082 require GNU binutils 2.13 or newer. Such subtargets include:
3083 @code{arm-*-freebsd}, @code{arm-*-netbsdelf}, @code{arm-*-*linux}
3084 and @code{arm-*-rtems}.
3089 @heading @anchor{avr}avr
3091 ATMEL AVR-family micro controllers. These are used in embedded
3092 applications. There are no standard Unix configurations.
3094 @xref{AVR Options,, AVR Options, gcc, Using the GNU Compiler
3098 See ``AVR Options'' in the main manual
3100 for the list of supported MCU types.
3102 Use @samp{configure --target=avr --enable-languages="c"} to configure GCC@.
3104 Further installation notes and other useful information about AVR tools
3105 can also be obtained from:
3109 @uref{http://www.nongnu.org/avr/,,http://www.nongnu.org/avr/}
3111 @uref{http://www.amelek.gda.pl/avr/,,http://www.amelek.gda.pl/avr/}
3114 We @emph{strongly} recommend using binutils 2.13 or newer.
3116 The following error:
3118 Error: register required
3121 indicates that you should upgrade to a newer version of the binutils.
3126 @heading @anchor{bfin}Blackfin
3128 The Blackfin processor, an Analog Devices DSP.
3130 @xref{Blackfin Options,, Blackfin Options, gcc, Using the GNU Compiler
3134 See ``Blackfin Options'' in the main manual
3137 More information, and a version of binutils with support for this processor,
3138 is available at @uref{http://blackfin.uclinux.org}
3143 @heading @anchor{cris}CRIS
3145 CRIS is the CPU architecture in Axis Communications ETRAX system-on-a-chip
3146 series. These are used in embedded applications.
3149 @xref{CRIS Options,, CRIS Options, gcc, Using the GNU Compiler
3153 See ``CRIS Options'' in the main manual
3155 for a list of CRIS-specific options.
3157 There are a few different CRIS targets:
3160 Mainly for monolithic embedded systems. Includes a multilib for the
3161 @samp{v10} core used in @samp{ETRAX 100 LX}.
3162 @item cris-axis-linux-gnu
3163 A GNU/Linux port for the CRIS architecture, currently targeting
3164 @samp{ETRAX 100 LX} by default.
3167 For @code{cris-axis-elf} you need binutils 2.11
3168 or newer. For @code{cris-axis-linux-gnu} you need binutils 2.12 or newer.
3170 Pre-packaged tools can be obtained from
3171 @uref{ftp://ftp.axis.com/pub/axis/tools/cris/compiler-kit/}. More
3172 information about this platform is available at
3173 @uref{http://developer.axis.com/}.
3178 @heading @anchor{crx}CRX
3180 The CRX CompactRISC architecture is a low-power 32-bit architecture with
3181 fast context switching and architectural extensibility features.
3184 @xref{CRX Options,, CRX Options, gcc, Using and Porting the GNU Compiler
3189 See ``CRX Options'' in the main manual for a list of CRX-specific options.
3192 Use @samp{configure --target=crx-elf --enable-languages=c,c++} to configure
3193 GCC@ for building a CRX cross-compiler. The option @samp{--target=crx-elf}
3194 is also used to build the @samp{newlib} C library for CRX.
3196 It is also possible to build libstdc++-v3 for the CRX architecture. This
3197 needs to be done in a separate step with the following configure settings:
3198 @samp{gcc/libstdc++-v3/configure --host=crx-elf --with-newlib
3199 --enable-sjlj-exceptions --enable-cxx-flags='-fexceptions -frtti'}
3204 @heading @anchor{dos}DOS
3206 Please have a look at the @uref{binaries.html,,binaries page}.
3208 You cannot install GCC by itself on MSDOS; it will not compile under
3209 any MSDOS compiler except itself. You need to get the complete
3210 compilation package DJGPP, which includes binaries as well as sources,
3211 and includes all the necessary compilation tools and libraries.
3216 @heading @anchor{x-x-freebsd}*-*-freebsd*
3218 Support for FreeBSD 1 was discontinued in GCC 3.2. Support for
3219 FreeBSD 2 (and any mutant a.out variants of FreeBSD 3) was
3220 discontinued in GCC 4.0.
3222 In GCC 4.5, we enabled the use of @code{dl_iterate_phdr} inside boehm-gc on
3223 FreeBSD 7 or later. In order to better match the configuration of the
3224 FreeBSD system compiler: We also enabled the check to see if libc
3225 provides SSP support (which it does on FreeBSD 7), the use of
3226 @code{dl_iterate_phdr} inside @file{libgcc_s.so.1} (on FreeBSD 7 or later)
3227 and the use of @code{__cxa_atexit} by default (on FreeBSD 6 or later).
3229 We support FreeBSD using the ELF file format with DWARF 2 debugging
3230 for all CPU architectures. You may use @option{-gstabs} instead of
3231 @option{-g}, if you really want the old debugging format. There are
3232 no known issues with mixing object files and libraries with different
3233 debugging formats. Otherwise, this release of GCC should now match
3234 more of the configuration used in the stock FreeBSD configuration of
3235 GCC@. In particular, @option{--enable-threads} is now configured by
3236 default. However, as a general user, do not attempt to replace the
3237 system compiler with this release. Known to bootstrap and check with
3238 good results on FreeBSD 7.2-STABLE@. In the past, known to bootstrap
3239 and check with good results on FreeBSD 3.0, 3.4, 4.0, 4.2, 4.3, 4.4,
3240 4.5, 4.8, 4.9 and 5-CURRENT@.
3242 The version of binutils installed in @file{/usr/bin} probably works
3243 with this release of GCC@. Bootstrapping against the latest GNU
3244 binutils and/or the version found in @file{/usr/ports/devel/binutils} has
3245 been known to enable additional features and improve overall testsuite
3246 results. However, it is currently known that boehm-gc (which itself
3247 is required for java) may not configure properly on FreeBSD prior to
3248 the FreeBSD 7.0 release with GNU binutils after 2.16.1.
3253 @heading @anchor{h8300-hms}h8300-hms
3254 Renesas H8/300 series of processors.
3256 Please have a look at the @uref{binaries.html,,binaries page}.
3258 The calling convention and structure layout has changed in release 2.6.
3259 All code must be recompiled. The calling convention now passes the
3260 first three arguments in function calls in registers. Structures are no
3261 longer a multiple of 2 bytes.
3266 @heading @anchor{hppa-hp-hpux}hppa*-hp-hpux*
3267 Support for HP-UX version 9 and older was discontinued in GCC 3.4.
3269 We require using gas/binutils on all hppa platforms. Version 2.19 or
3270 later is recommended.
3272 It may be helpful to configure GCC with the
3273 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}} and
3274 @option{--with-as=@dots{}} options to ensure that GCC can find GAS@.
3276 The HP assembler should not be used with GCC. It is rarely tested and may
3277 not work. It shouldn't be used with any languages other than C due to its
3280 Specifically, @option{-g} does not work (HP-UX uses a peculiar debugging
3281 format which GCC does not know about). It also inserts timestamps
3282 into each object file it creates, causing the 3-stage comparison test to
3283 fail during a bootstrap. You should be able to continue by saying
3284 @samp{make all-host all-target} after getting the failure from @samp{make}.
3286 Various GCC features are not supported. For example, it does not support weak
3287 symbols or alias definitions. As a result, explicit template instantiations
3288 are required when using C++. This makes it difficult if not impossible to
3289 build many C++ applications.
3291 There are two default scheduling models for instructions. These are
3292 PROCESSOR_7100LC and PROCESSOR_8000. They are selected from the pa-risc
3293 architecture specified for the target machine when configuring.
3294 PROCESSOR_8000 is the default. PROCESSOR_7100LC is selected when
3295 the target is a @samp{hppa1*} machine.
3297 The PROCESSOR_8000 model is not well suited to older processors. Thus,
3298 it is important to completely specify the machine architecture when
3299 configuring if you want a model other than PROCESSOR_8000. The macro
3300 TARGET_SCHED_DEFAULT can be defined in BOOT_CFLAGS if a different
3301 default scheduling model is desired.
3303 As of GCC 4.0, GCC uses the UNIX 95 namespace for HP-UX 10.10
3304 through 11.00, and the UNIX 98 namespace for HP-UX 11.11 and later.
3305 This namespace change might cause problems when bootstrapping with
3306 an earlier version of GCC or the HP compiler as essentially the same
3307 namespace is required for an entire build. This problem can be avoided
3308 in a number of ways. With HP cc, @env{UNIX_STD} can be set to @samp{95}
3309 or @samp{98}. Another way is to add an appropriate set of predefines
3310 to @env{CC}. The description for the @option{munix=} option contains
3311 a list of the predefines used with each standard.
3313 More specific information to @samp{hppa*-hp-hpux*} targets follows.
3318 @heading @anchor{hppa-hp-hpux10}hppa*-hp-hpux10
3320 For hpux10.20, we @emph{highly} recommend you pick up the latest sed patch
3321 @code{PHCO_19798} from HP@. HP has two sites which provide patches free of
3327 <a href="http://us.itrc.hp.com/service/home/home.do">US, Canada, Asia-Pacific, and
3331 @uref{http://us.itrc.hp.com/service/home/home.do,,} US, Canada, Asia-Pacific,
3335 @uref{http://europe.itrc.hp.com/service/home/home.do,,} Europe.
3338 The C++ ABI has changed incompatibly in GCC 4.0. COMDAT subspaces are
3339 used for one-only code and data. This resolves many of the previous
3340 problems in using C++ on this target. However, the ABI is not compatible
3341 with the one implemented under HP-UX 11 using secondary definitions.
3346 @heading @anchor{hppa-hp-hpux11}hppa*-hp-hpux11
3348 GCC 3.0 and up support HP-UX 11. GCC 2.95.x is not supported and cannot
3349 be used to compile GCC 3.0 and up.
3351 The libffi and libjava libraries haven't been ported to 64-bit HP-UX@
3354 Refer to @uref{binaries.html,,binaries} for information about obtaining
3355 precompiled GCC binaries for HP-UX@. Precompiled binaries must be obtained
3356 to build the Ada language as it can't be bootstrapped using C@. Ada is
3357 only available for the 32-bit PA-RISC runtime.
3359 Starting with GCC 3.4 an ISO C compiler is required to bootstrap. The
3360 bundled compiler supports only traditional C; you will need either HP's
3361 unbundled compiler, or a binary distribution of GCC@.
3363 It is possible to build GCC 3.3 starting with the bundled HP compiler,
3364 but the process requires several steps. GCC 3.3 can then be used to
3365 build later versions. The fastjar program contains ISO C code and
3366 can't be built with the HP bundled compiler. This problem can be
3367 avoided by not building the Java language. For example, use the
3368 @option{--enable-languages="c,c++,f77,objc"} option in your configure
3371 There are several possible approaches to building the distribution.
3372 Binutils can be built first using the HP tools. Then, the GCC
3373 distribution can be built. The second approach is to build GCC
3374 first using the HP tools, then build binutils, then rebuild GCC@.
3375 There have been problems with various binary distributions, so it
3376 is best not to start from a binary distribution.
3378 On 64-bit capable systems, there are two distinct targets. Different
3379 installation prefixes must be used if both are to be installed on
3380 the same system. The @samp{hppa[1-2]*-hp-hpux11*} target generates code
3381 for the 32-bit PA-RISC runtime architecture and uses the HP linker.
3382 The @samp{hppa64-hp-hpux11*} target generates 64-bit code for the
3383 PA-RISC 2.0 architecture.
3385 The script config.guess now selects the target type based on the compiler
3386 detected during configuration. You must define @env{PATH} or @env{CC} so
3387 that configure finds an appropriate compiler for the initial bootstrap.
3388 When @env{CC} is used, the definition should contain the options that are
3389 needed whenever @env{CC} is used.
3391 Specifically, options that determine the runtime architecture must be
3392 in @env{CC} to correctly select the target for the build. It is also
3393 convenient to place many other compiler options in @env{CC}. For example,
3394 @env{CC="cc -Ac +DA2.0W -Wp,-H16376 -D_CLASSIC_TYPES -D_HPUX_SOURCE"}
3395 can be used to bootstrap the GCC 3.3 branch with the HP compiler in
3396 64-bit K&R/bundled mode. The @option{+DA2.0W} option will result in
3397 the automatic selection of the @samp{hppa64-hp-hpux11*} target. The
3398 macro definition table of cpp needs to be increased for a successful
3399 build with the HP compiler. _CLASSIC_TYPES and _HPUX_SOURCE need to
3400 be defined when building with the bundled compiler, or when using the
3401 @option{-Ac} option. These defines aren't necessary with @option{-Ae}.
3403 It is best to explicitly configure the @samp{hppa64-hp-hpux11*} target
3404 with the @option{--with-ld=@dots{}} option. This overrides the standard
3405 search for ld. The two linkers supported on this target require different
3406 commands. The default linker is determined during configuration. As a
3407 result, it's not possible to switch linkers in the middle of a GCC build.
3408 This has been reported to sometimes occur in unified builds of binutils
3411 A recent linker patch must be installed for the correct operation of
3412 GCC 3.3 and later. @code{PHSS_26559} and @code{PHSS_24304} are the
3413 oldest linker patches that are known to work. They are for HP-UX
3414 11.00 and 11.11, respectively. @code{PHSS_24303}, the companion to
3415 @code{PHSS_24304}, might be usable but it hasn't been tested. These
3416 patches have been superseded. Consult the HP patch database to obtain
3417 the currently recommended linker patch for your system.
3419 The patches are necessary for the support of weak symbols on the
3420 32-bit port, and for the running of initializers and finalizers. Weak
3421 symbols are implemented using SOM secondary definition symbols. Prior
3422 to HP-UX 11, there are bugs in the linker support for secondary symbols.
3423 The patches correct a problem of linker core dumps creating shared
3424 libraries containing secondary symbols, as well as various other
3425 linking issues involving secondary symbols.
3427 GCC 3.3 uses the ELF DT_INIT_ARRAY and DT_FINI_ARRAY capabilities to
3428 run initializers and finalizers on the 64-bit port. The 32-bit port
3429 uses the linker @option{+init} and @option{+fini} options for the same
3430 purpose. The patches correct various problems with the +init/+fini
3431 options, including program core dumps. Binutils 2.14 corrects a
3432 problem on the 64-bit port resulting from HP's non-standard use of
3433 the .init and .fini sections for array initializers and finalizers.
3435 Although the HP and GNU linkers are both supported for the
3436 @samp{hppa64-hp-hpux11*} target, it is strongly recommended that the
3437 HP linker be used for link editing on this target.
3439 At this time, the GNU linker does not support the creation of long
3440 branch stubs. As a result, it can't successfully link binaries
3441 containing branch offsets larger than 8 megabytes. In addition,
3442 there are problems linking shared libraries, linking executables
3443 with @option{-static}, and with dwarf2 unwind and exception support.
3444 It also doesn't provide stubs for internal calls to global functions
3445 in shared libraries, so these calls can't be overloaded.
3447 The HP dynamic loader does not support GNU symbol versioning, so symbol
3448 versioning is not supported. It may be necessary to disable symbol
3449 versioning with @option{--disable-symvers} when using GNU ld.
3451 POSIX threads are the default. The optional DCE thread library is not
3452 supported, so @option{--enable-threads=dce} does not work.
3457 @heading @anchor{x-x-linux-gnu}*-*-linux-gnu
3459 Versions of libstdc++-v3 starting with 3.2.1 require bug fixes present
3460 in glibc 2.2.5 and later. More information is available in the
3461 libstdc++-v3 documentation.
3466 @heading @anchor{ix86-x-linux}i?86-*-linux*
3468 As of GCC 3.3, binutils 2.13.1 or later is required for this platform.
3469 See @uref{http://gcc.gnu.org/PR10877,,bug 10877} for more information.
3471 If you receive Signal 11 errors when building on GNU/Linux, then it is
3472 possible you have a hardware problem. Further information on this can be
3473 found on @uref{http://www.bitwizard.nl/sig11/,,www.bitwizard.nl}.
3478 @heading @anchor{ix86-x-solaris210}i?86-*-solaris2.10
3479 Use this for Solaris 10 or later on x86 and x86-64 systems. This
3480 configuration is supported by GCC 4.0 and later versions only.
3482 It is recommended that you configure GCC to use the GNU assembler in
3483 @file{/usr/sfw/bin/gas} but the Sun linker, using the options
3484 @option{--with-gnu-as --with-as=/usr/sfw/bin/gas --without-gnu-ld
3485 --with-ld=/usr/ccs/bin/ld}.
3490 @heading @anchor{ia64-x-linux}ia64-*-linux
3491 IA-64 processor (also known as IPF, or Itanium Processor Family)
3494 If you are using the installed system libunwind library with
3495 @option{--with-system-libunwind}, then you must use libunwind 0.98 or
3498 None of the following versions of GCC has an ABI that is compatible
3499 with any of the other versions in this list, with the exception that
3500 Red Hat 2.96 and Trillian 000171 are compatible with each other:
3501 3.1, 3.0.2, 3.0.1, 3.0, Red Hat 2.96, and Trillian 000717.
3502 This primarily affects C++ programs and programs that create shared libraries.
3503 GCC 3.1 or later is recommended for compiling linux, the kernel.
3504 As of version 3.1 GCC is believed to be fully ABI compliant, and hence no
3505 more major ABI changes are expected.
3510 @heading @anchor{ia64-x-hpux}ia64-*-hpux*
3511 Building GCC on this target requires the GNU Assembler. The bundled HP
3512 assembler will not work. To prevent GCC from using the wrong assembler,
3513 the option @option{--with-gnu-as} may be necessary.
3515 The GCC libunwind library has not been ported to HPUX@. This means that for
3516 GCC versions 3.2.3 and earlier, @option{--enable-libunwind-exceptions}
3517 is required to build GCC@. For GCC 3.3 and later, this is the default.
3518 For gcc 3.4.3 and later, @option{--enable-libunwind-exceptions} is
3519 removed and the system libunwind library will always be used.
3523 <!-- rs6000-ibm-aix*, powerpc-ibm-aix* -->
3525 @heading @anchor{x-ibm-aix}*-ibm-aix*
3526 Support for AIX version 3 and older was discontinued in GCC 3.4.
3527 Support for AIX version 4.2 and older was discontinued in GCC 4.5.
3529 ``out of memory'' bootstrap failures may indicate a problem with
3530 process resource limits (ulimit). Hard limits are configured in the
3531 @file{/etc/security/limits} system configuration file.
3533 GCC can bootstrap with recent versions of IBM XLC, but bootstrapping
3534 with an earlier release of GCC is recommended. Bootstrapping with XLC
3535 requires a larger data segment, which can be enabled through the
3536 @var{LDR_CNTRL} environment variable, e.g.,
3539 % LDR_CNTRL=MAXDATA=0x50000000
3543 One can start with a pre-compiled version of GCC to build from
3544 sources. One may delete GCC's ``fixed'' header files when starting
3545 with a version of GCC built for an earlier release of AIX.
3547 To speed up the configuration phases of bootstrapping and installing GCC,
3548 one may use GNU Bash instead of AIX @command{/bin/sh}, e.g.,
3551 % CONFIG_SHELL=/opt/freeware/bin/bash
3552 % export CONFIG_SHELL
3555 and then proceed as described in @uref{build.html,,the build
3556 instructions}, where we strongly recommend specifying an absolute path
3557 to invoke @var{srcdir}/configure.
3559 Because GCC on AIX is built as a 32-bit executable by default,
3560 (although it can generate 64-bit programs) the GMP and MPFR libraries
3561 required by gfortran must be 32-bit libraries. Building GMP and MPFR
3562 as static archive libraries works better than shared libraries.
3564 Errors involving @code{alloca} when building GCC generally are due
3565 to an incorrect definition of @code{CC} in the Makefile or mixing files
3566 compiled with the native C compiler and GCC@. During the stage1 phase of
3567 the build, the native AIX compiler @strong{must} be invoked as @command{cc}
3568 (not @command{xlc}). Once @command{configure} has been informed of
3569 @command{xlc}, one needs to use @samp{make distclean} to remove the
3570 configure cache files and ensure that @env{CC} environment variable
3571 does not provide a definition that will confuse @command{configure}.
3572 If this error occurs during stage2 or later, then the problem most likely
3573 is the version of Make (see above).
3575 The native @command{as} and @command{ld} are recommended for bootstrapping
3576 on AIX@. The GNU Assembler, GNU Linker, and GNU Binutils version 2.20
3577 is required to bootstrap on AIX 5@. The native AIX tools do
3578 interoperate with GCC@.
3580 Building @file{libstdc++.a} requires a fix for an AIX Assembler bug
3581 APAR IY26685 (AIX 4.3) or APAR IY25528 (AIX 5.1). It also requires a
3582 fix for another AIX Assembler bug and a co-dependent AIX Archiver fix
3583 referenced as APAR IY53606 (AIX 5.2) or as APAR IY54774 (AIX 5.1)
3585 @samp{libstdc++} in GCC 3.4 increments the major version number of the
3586 shared object and GCC installation places the @file{libstdc++.a}
3587 shared library in a common location which will overwrite the and GCC
3588 3.3 version of the shared library. Applications either need to be
3589 re-linked against the new shared library or the GCC 3.1 and GCC 3.3
3590 versions of the @samp{libstdc++} shared object needs to be available
3591 to the AIX runtime loader. The GCC 3.1 @samp{libstdc++.so.4}, if
3592 present, and GCC 3.3 @samp{libstdc++.so.5} shared objects can be
3593 installed for runtime dynamic loading using the following steps to set
3594 the @samp{F_LOADONLY} flag in the shared object for @emph{each}
3595 multilib @file{libstdc++.a} installed:
3597 Extract the shared objects from the currently installed
3598 @file{libstdc++.a} archive:
3600 % ar -x libstdc++.a libstdc++.so.4 libstdc++.so.5
3603 Enable the @samp{F_LOADONLY} flag so that the shared object will be
3604 available for runtime dynamic loading, but not linking:
3606 % strip -e libstdc++.so.4 libstdc++.so.5
3609 Archive the runtime-only shared object in the GCC 3.4
3610 @file{libstdc++.a} archive:
3612 % ar -q libstdc++.a libstdc++.so.4 libstdc++.so.5
3615 Linking executables and shared libraries may produce warnings of
3616 duplicate symbols. The assembly files generated by GCC for AIX always
3617 have included multiple symbol definitions for certain global variable
3618 and function declarations in the original program. The warnings should
3619 not prevent the linker from producing a correct library or runnable
3622 AIX 4.3 utilizes a ``large format'' archive to support both 32-bit and
3623 64-bit object modules. The routines provided in AIX 4.3.0 and AIX 4.3.1
3624 to parse archive libraries did not handle the new format correctly.
3625 These routines are used by GCC and result in error messages during
3626 linking such as ``not a COFF file''. The version of the routines shipped
3627 with AIX 4.3.1 should work for a 32-bit environment. The @option{-g}
3628 option of the archive command may be used to create archives of 32-bit
3629 objects using the original ``small format''. A correct version of the
3630 routines is shipped with AIX 4.3.2 and above.
3632 Some versions of the AIX binder (linker) can fail with a relocation
3633 overflow severe error when the @option{-bbigtoc} option is used to link
3634 GCC-produced object files into an executable that overflows the TOC@. A fix
3635 for APAR IX75823 (OVERFLOW DURING LINK WHEN USING GCC AND -BBIGTOC) is
3636 available from IBM Customer Support and from its
3637 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
3638 website as PTF U455193.
3640 The AIX 4.3.2.1 linker (bos.rte.bind_cmds Level 4.3.2.1) will dump core
3641 with a segmentation fault when invoked by any version of GCC@. A fix for
3642 APAR IX87327 is available from IBM Customer Support and from its
3643 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
3644 website as PTF U461879. This fix is incorporated in AIX 4.3.3 and above.
3646 The initial assembler shipped with AIX 4.3.0 generates incorrect object
3647 files. A fix for APAR IX74254 (64BIT DISASSEMBLED OUTPUT FROM COMPILER FAILS
3648 TO ASSEMBLE/BIND) is available from IBM Customer Support and from its
3649 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
3650 website as PTF U453956. This fix is incorporated in AIX 4.3.1 and above.
3652 AIX provides National Language Support (NLS)@. Compilers and assemblers
3653 use NLS to support locale-specific representations of various data
3654 formats including floating-point numbers (e.g., @samp{.} vs @samp{,} for
3655 separating decimal fractions). There have been problems reported where
3656 GCC does not produce the same floating-point formats that the assembler
3657 expects. If one encounters this problem, set the @env{LANG}
3658 environment variable to @samp{C} or @samp{En_US}.
3660 A default can be specified with the @option{-mcpu=@var{cpu_type}}
3661 switch and using the configure option @option{--with-cpu-@var{cpu_type}}.
3666 @heading @anchor{iq2000-x-elf}iq2000-*-elf
3667 Vitesse IQ2000 processors. These are used in embedded
3668 applications. There are no standard Unix configurations.
3673 @heading @anchor{lm32-x-elf}lm32-*-elf
3674 Lattice Mico32 processor.
3675 This configuration is intended for embedded systems.
3680 @heading @anchor{lm32-x-uclinux}lm32-*-uclinux
3681 Lattice Mico32 processor.
3682 This configuration is intended for embedded systems running uClinux.
3687 @heading @anchor{m32c-x-elf}m32c-*-elf
3688 Renesas M32C processor.
3689 This configuration is intended for embedded systems.
3694 @heading @anchor{m32r-x-elf}m32r-*-elf
3695 Renesas M32R processor.
3696 This configuration is intended for embedded systems.
3701 @heading @anchor{m6811-elf}m6811-elf
3702 Motorola 68HC11 family micro controllers. These are used in embedded
3703 applications. There are no standard Unix configurations.
3708 @heading @anchor{m6812-elf}m6812-elf
3709 Motorola 68HC12 family micro controllers. These are used in embedded
3710 applications. There are no standard Unix configurations.
3715 @heading @anchor{m68k-x-x}m68k-*-*
3717 @samp{m68k-*-elf*}, @samp{m68k-*-rtems}, @samp{m68k-*-uclinux} and
3719 build libraries for both M680x0 and ColdFire processors. If you only
3720 need the M680x0 libraries, you can omit the ColdFire ones by passing
3721 @option{--with-arch=m68k} to @command{configure}. Alternatively, you
3722 can omit the M680x0 libraries by passing @option{--with-arch=cf} to
3723 @command{configure}. These targets default to 5206 or 5475 code as
3724 appropriate for the target system when
3725 configured with @option{--with-arch=cf} and 68020 code otherwise.
3727 The @samp{m68k-*-netbsd} and
3728 @samp{m68k-*-openbsd} targets also support the @option{--with-arch}
3729 option. They will generate ColdFire CFV4e code when configured with
3730 @option{--with-arch=cf} and 68020 code otherwise.
3732 You can override the default processors listed above by configuring
3733 with @option{--with-cpu=@var{target}}. This @var{target} can either
3734 be a @option{-mcpu} argument or one of the following values:
3735 @samp{m68000}, @samp{m68010}, @samp{m68020}, @samp{m68030},
3736 @samp{m68040}, @samp{m68060}, @samp{m68020-40} and @samp{m68020-60}.
3741 @heading @anchor{m68k-x-uclinux}m68k-*-uclinux
3742 GCC 4.3 changed the uClinux configuration so that it uses the
3743 @samp{m68k-linux-gnu} ABI rather than the @samp{m68k-elf} ABI.
3744 It also added improved support for C++ and flat shared libraries,
3745 both of which were ABI changes. However, you can still use the
3746 original ABI by configuring for @samp{m68k-uclinuxoldabi} or
3747 @samp{m68k-@var{vendor}-uclinuxoldabi}.
3753 @heading @anchor{mep-x-elf}mep-*-elf
3754 Toshiba Media embedded Processor.
3755 This configuration is intended for embedded systems.
3760 @heading @anchor{mips-x-x}mips-*-*
3761 If on a MIPS system you get an error message saying ``does not have gp
3762 sections for all it's [sic] sectons [sic]'', don't worry about it. This
3763 happens whenever you use GAS with the MIPS linker, but there is not
3764 really anything wrong, and it is okay to use the output file. You can
3765 stop such warnings by installing the GNU linker.
3767 It would be nice to extend GAS to produce the gp tables, but they are
3768 optional, and there should not be a warning about their absence.
3770 The libstdc++ atomic locking routines for MIPS targets requires MIPS II
3771 and later. A patch went in just after the GCC 3.3 release to
3772 make @samp{mips*-*-*} use the generic implementation instead. You can also
3773 configure for @samp{mipsel-elf} as a workaround. The
3774 @samp{mips*-*-linux*} target continues to use the MIPS II routines. More
3775 work on this is expected in future releases.
3777 @c If you make --with-llsc the default for another target, please also
3778 @c update the description of the --with-llsc option.
3780 The built-in @code{__sync_*} functions are available on MIPS II and
3781 later systems and others that support the @samp{ll}, @samp{sc} and
3782 @samp{sync} instructions. This can be overridden by passing
3783 @option{--with-llsc} or @option{--without-llsc} when configuring GCC.
3784 Since the Linux kernel emulates these instructions if they are
3785 missing, the default for @samp{mips*-*-linux*} targets is
3786 @option{--with-llsc}. The @option{--with-llsc} and
3787 @option{--without-llsc} configure options may be overridden at compile
3788 time by passing the @option{-mllsc} or @option{-mno-llsc} options to
3791 MIPS systems check for division by zero (unless
3792 @option{-mno-check-zero-division} is passed to the compiler) by
3793 generating either a conditional trap or a break instruction. Using
3794 trap results in smaller code, but is only supported on MIPS II and
3795 later. Also, some versions of the Linux kernel have a bug that
3796 prevents trap from generating the proper signal (@code{SIGFPE}). To enable
3797 the use of break, use the @option{--with-divide=breaks}
3798 @command{configure} option when configuring GCC@. The default is to
3799 use traps on systems that support them.
3801 Cross-compilers for the MIPS as target using the MIPS assembler
3802 currently do not work, because the auxiliary programs
3803 @file{mips-tdump.c} and @file{mips-tfile.c} can't be compiled on
3804 anything but a MIPS@. It does work to cross compile for a MIPS
3805 if you use the GNU assembler and linker.
3807 The assembler from GNU binutils 2.17 and earlier has a bug in the way
3808 it sorts relocations for REL targets (o32, o64, EABI). This can cause
3809 bad code to be generated for simple C++ programs. Also the linker
3810 from GNU binutils versions prior to 2.17 has a bug which causes the
3811 runtime linker stubs in very large programs, like @file{libgcj.so}, to
3812 be incorrectly generated. GNU Binutils 2.18 and later (and snapshots
3813 made after Nov. 9, 2006) should be free from both of these problems.
3818 @heading @anchor{mips-sgi-irix5}mips-sgi-irix5
3820 In order to compile GCC on an SGI running IRIX 5, the @samp{compiler_dev.hdr}
3821 subsystem must be installed from the IDO CD-ROM supplied by SGI@.
3822 It is also available for download from
3823 @uref{ftp://ftp.sgi.com/sgi/IRIX5.3/iris-development-option-5.3.tardist}.
3825 If you use the MIPS C compiler to bootstrap, it may be necessary
3826 to increase its table size for switch statements with the
3827 @option{-Wf,-XNg1500} option. If you use the @option{-O2}
3828 optimization option, you also need to use @option{-Olimit 3000}.
3830 To enable debugging under IRIX 5, you must use GNU binutils 2.15 or
3831 later, and use the @option{--with-gnu-ld} @command{configure} option
3832 when configuring GCC@. You need to use GNU @command{ar} and @command{nm},
3833 also distributed with GNU binutils.
3835 Some users have reported that @command{/bin/sh} will hang during bootstrap.
3836 This problem can be avoided by running the commands:
3839 % CONFIG_SHELL=/bin/ksh
3840 % export CONFIG_SHELL
3844 before starting the build.
3849 @heading @anchor{mips-sgi-irix6}mips-sgi-irix6
3851 If you are using SGI's MIPSpro @command{cc} as your bootstrap compiler, you must
3852 ensure that the N32 ABI is in use. To test this, compile a simple C
3853 file with @command{cc} and then run @command{file} on the
3854 resulting object file. The output should look like:
3857 test.o: ELF N32 MSB @dots{}
3864 test.o: ELF 32-bit MSB @dots{}
3871 test.o: ELF 64-bit MSB @dots{}
3875 then your version of @command{cc} uses the O32 or N64 ABI by default. You
3876 should set the environment variable @env{CC} to @samp{cc -n32}
3877 before configuring GCC@.
3879 If you want the resulting @command{gcc} to run on old 32-bit systems
3880 with the MIPS R4400 CPU, you need to ensure that only code for the @samp{mips3}
3881 instruction set architecture (ISA) is generated. While GCC 3.x does
3882 this correctly, both GCC 2.95 and SGI's MIPSpro @command{cc} may change
3883 the ISA depending on the machine where GCC is built. Using one of them
3884 as the bootstrap compiler may result in @samp{mips4} code, which won't run at
3885 all on @samp{mips3}-only systems. For the test program above, you should see:
3888 test.o: ELF N32 MSB mips-3 @dots{}
3895 test.o: ELF N32 MSB mips-4 @dots{}
3899 instead, you should set the environment variable @env{CC} to @samp{cc
3900 -n32 -mips3} or @samp{gcc -mips3} respectively before configuring GCC@.
3902 MIPSpro C 7.4 may cause bootstrap failures, due to a bug when inlining
3903 @code{memcmp}. Either add @code{-U__INLINE_INTRINSICS} to the @env{CC}
3904 environment variable as a workaround or upgrade to MIPSpro C 7.4.1m.
3906 GCC on IRIX 6 is usually built to support the N32, O32 and N64 ABIs. If
3907 you build GCC on a system that doesn't have the N64 libraries installed
3908 or cannot run 64-bit binaries,
3909 you need to configure with @option{--disable-multilib} so GCC doesn't
3910 try to use them. This will disable building the O32 libraries, too.
3911 Look for @file{/usr/lib64/libc.so.1} to see if you
3912 have the 64-bit libraries installed.
3914 To enable debugging for the O32 ABI, you must use GNU @command{as} from
3915 GNU binutils 2.15 or later. You may also use GNU @command{ld}, but
3916 this is not required and currently causes some problems with Ada.
3918 The @option{--enable-libgcj}
3919 option is disabled by default: IRIX 6 uses a very low default limit
3920 (20480) for the command line length. Although @command{libtool} contains a
3921 workaround for this problem, at least the N64 @samp{libgcj} is known not
3922 to build despite this, running into an internal error of the native
3923 @command{ld}. A sure fix is to increase this limit (@samp{ncargs}) to
3924 its maximum of 262144 bytes. If you have root access, you can use the
3925 @command{systune} command to do this.
3927 @code{wchar_t} support in @samp{libstdc++} is not available for old
3928 IRIX 6.5.x releases, @math{x < 19}. The problem cannot be autodetected
3929 and in order to build GCC for such targets you need to configure with
3930 @option{--disable-wchar_t}.
3932 See @uref{http://freeware.sgi.com/} for more
3933 information about using GCC on IRIX platforms.
3938 @heading @anchor{moxie-x-elf}moxie-*-elf
3939 The moxie processor. See @uref{http://moxielogic.org/} for more
3940 information about this processor.
3945 @heading @anchor{powerpc-x-x}powerpc-*-*
3947 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
3948 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
3951 @uref{ftp://ftp.kernel.org/pub/linux/devel/binutils,,binutils 2.15}
3952 or newer for a working GCC@.
3957 @heading @anchor{powerpc-x-darwin}powerpc-*-darwin*
3958 PowerPC running Darwin (Mac OS X kernel).
3960 Pre-installed versions of Mac OS X may not include any developer tools,
3961 meaning that you will not be able to build GCC from source. Tool
3962 binaries are available at
3963 @uref{http://developer.apple.com/darwin/projects/compiler/} (free
3964 registration required).
3966 This version of GCC requires at least cctools-590.36. The
3967 cctools-590.36 package referenced from
3968 @uref{http://gcc.gnu.org/ml/gcc/2006-03/msg00507.html} will not work
3969 on systems older than 10.3.9 (aka darwin7.9.0).
3974 @heading @anchor{powerpc-x-elf}powerpc-*-elf
3975 PowerPC system in big endian mode, running System V.4.
3980 @heading @anchor{powerpc-x-linux-gnu}powerpc*-*-linux-gnu*
3982 PowerPC system in big endian mode running Linux.
3987 @heading @anchor{powerpc-x-netbsd}powerpc-*-netbsd*
3988 PowerPC system in big endian mode running NetBSD@.
3993 @heading @anchor{powerpc-x-eabisim}powerpc-*-eabisim
3994 Embedded PowerPC system in big endian mode for use in running under the
4000 @heading @anchor{powerpc-x-eabi}powerpc-*-eabi
4001 Embedded PowerPC system in big endian mode.
4006 @heading @anchor{powerpcle-x-elf}powerpcle-*-elf
4007 PowerPC system in little endian mode, running System V.4.
4012 @heading @anchor{powerpcle-x-eabisim}powerpcle-*-eabisim
4013 Embedded PowerPC system in little endian mode for use in running under
4019 @heading @anchor{powerpcle-x-eabi}powerpcle-*-eabi
4020 Embedded PowerPC system in little endian mode.
4025 @heading @anchor{rx-x-elf}rx-*-elf
4026 The Renesas RX processor. See
4027 @uref{http://eu.renesas.com/fmwk.jsp?cnt=rx600_series_landing.jsp&fp=/products/mpumcu/rx_family/rx600_series}
4028 for more information about this processor.
4033 @heading @anchor{s390-x-linux}s390-*-linux*
4034 S/390 system running GNU/Linux for S/390@.
4039 @heading @anchor{s390x-x-linux}s390x-*-linux*
4040 zSeries system (64-bit) running GNU/Linux for zSeries@.
4045 @heading @anchor{s390x-ibm-tpf}s390x-ibm-tpf*
4046 zSeries system (64-bit) running TPF@. This platform is
4047 supported as cross-compilation target only.
4052 @c Please use Solaris 2 to refer to all release of Solaris, starting
4053 @c with 2.0 until 2.6, 7, 8, etc. Solaris 1 was a marketing name for
4054 @c SunOS 4 releases which we don't use to avoid confusion. Solaris
4055 @c alone is too unspecific and must be avoided.
4056 @heading @anchor{x-x-solaris2}*-*-solaris2*
4058 Sun does not ship a C compiler with Solaris 2. To bootstrap and install
4059 GCC you first have to install a pre-built compiler, see the
4060 @uref{binaries.html,,binaries page} for details.
4062 The Solaris 2 @command{/bin/sh} will often fail to configure
4063 @file{libstdc++-v3}, @file{boehm-gc} or @file{libjava}. We therefore
4064 recommend using the following initial sequence of commands
4067 % CONFIG_SHELL=/bin/ksh
4068 % export CONFIG_SHELL
4071 and proceed as described in @uref{configure.html,,the configure instructions}.
4072 In addition we strongly recommend specifying an absolute path to invoke
4073 @var{srcdir}/configure.
4075 Solaris 2 comes with a number of optional OS packages. Some of these
4076 are needed to use GCC fully, namely @code{SUNWarc},
4077 @code{SUNWbtool}, @code{SUNWesu}, @code{SUNWhea}, @code{SUNWlibm},
4078 @code{SUNWsprot}, and @code{SUNWtoo}. If you did not install all
4079 optional packages when installing Solaris 2, you will need to verify that
4080 the packages that GCC needs are installed.
4082 To check whether an optional package is installed, use
4083 the @command{pkginfo} command. To add an optional package, use the
4084 @command{pkgadd} command. For further details, see the Solaris 2
4087 Trying to use the linker and other tools in
4088 @file{/usr/ucb} to install GCC has been observed to cause trouble.
4089 For example, the linker may hang indefinitely. The fix is to remove
4090 @file{/usr/ucb} from your @env{PATH}.
4092 The build process works more smoothly with the legacy Sun tools so, if you
4093 have @file{/usr/xpg4/bin} in your @env{PATH}, we recommend that you place
4094 @file{/usr/bin} before @file{/usr/xpg4/bin} for the duration of the build.
4096 We recommend the use of GNU binutils 2.14 or later, or the vendor tools
4097 (Sun @command{as}, Sun @command{ld}). Note that your mileage may vary
4098 if you use a combination of the GNU tools and the Sun tools: while the
4099 combination GNU @command{as} + Sun @command{ld} should reasonably work,
4100 the reverse combination Sun @command{as} + GNU @command{ld} is known to
4101 cause memory corruption at runtime in some cases for C++ programs.
4103 The stock GNU binutils 2.15 release is broken on this platform because of a
4104 single bug. It has been fixed on the 2.15 branch in the CVS repository.
4105 You can obtain a working version by checking out the binutils-2_15-branch
4106 from the CVS repository or applying the patch
4107 @uref{http://sourceware.org/ml/binutils-cvs/2004-09/msg00036.html} to the
4110 We recommend the use of GNU binutils 2.16 or later in conjunction with GCC
4111 4.x, or the vendor tools (Sun @command{as}, Sun @command{ld}). However,
4112 for Solaris 10 and above, an additional patch is required in order for the
4113 GNU linker to be able to cope with a new flavor of shared libraries. You
4114 can obtain a working version by checking out the binutils-2_16-branch from
4115 the CVS repository or applying the patch
4116 @uref{http://sourceware.org/ml/binutils-cvs/2005-07/msg00122.html} to the
4119 Sun bug 4296832 turns up when compiling X11 headers with GCC 2.95 or
4120 newer: @command{g++} will complain that types are missing. These headers
4121 assume that omitting the type means @code{int}; this assumption worked for
4122 C89 but is wrong for C++, and is now wrong for C99 also.
4124 @command{g++} accepts such (invalid) constructs with the option
4125 @option{-fpermissive}; it will assume that any missing type is @code{int}
4126 (as defined by C89).
4128 There are patches for Solaris 7 (108376-21 or newer for SPARC,
4129 108377-20 for Intel), and Solaris 8 (108652-24 or newer for SPARC,
4130 108653-22 for Intel) that fix this bug.
4132 Sun bug 4927647 sometimes causes random spurious testsuite failures
4133 related to missing diagnostic output. This bug doesn't affect GCC
4134 itself, rather it is a kernel bug triggered by the @command{expect}
4135 program which is used only by the GCC testsuite driver. When the bug
4136 causes the @command{expect} program to miss anticipated output, extra
4137 testsuite failures appear.
4139 There are patches for Solaris 8 (117350-12 or newer for SPARC,
4140 117351-12 or newer for Intel) and Solaris 9 (117171-11 or newer for
4141 SPARC, 117172-11 or newer for Intel) that address this problem.
4146 @heading @anchor{sparc-sun-solaris2}sparc-sun-solaris2*
4148 When GCC is configured to use binutils 2.14 or later the binaries
4149 produced are smaller than the ones produced using Sun's native tools;
4150 this difference is quite significant for binaries containing debugging
4153 Starting with Solaris 7, the operating system is capable of executing
4154 64-bit SPARC V9 binaries. GCC 3.1 and later properly supports
4155 this; the @option{-m64} option enables 64-bit code generation.
4156 However, if all you want is code tuned for the UltraSPARC CPU, you
4157 should try the @option{-mtune=ultrasparc} option instead, which produces
4158 code that, unlike full 64-bit code, can still run on non-UltraSPARC
4161 When configuring on a Solaris 7 or later system that is running a kernel
4162 that supports only 32-bit binaries, one must configure with
4163 @option{--disable-multilib}, since we will not be able to build the
4164 64-bit target libraries.
4166 GCC 3.3 and GCC 3.4 trigger code generation bugs in earlier versions of
4167 the GNU compiler (especially GCC 3.0.x versions), which lead to the
4168 miscompilation of the stage1 compiler and the subsequent failure of the
4169 bootstrap process. A workaround is to use GCC 3.2.3 as an intermediary
4170 stage, i.e.@: to bootstrap that compiler with the base compiler and then
4171 use it to bootstrap the final compiler.
4173 GCC 3.4 triggers a code generation bug in versions 5.4 (Sun ONE Studio 7)
4174 and 5.5 (Sun ONE Studio 8) of the Sun compiler, which causes a bootstrap
4175 failure in form of a miscompilation of the stage1 compiler by the Sun
4176 compiler. This is Sun bug 4974440. This is fixed with patch 112760-07.
4178 GCC 3.4 changed the default debugging format from STABS to DWARF-2 for
4179 32-bit code on Solaris 7 and later. If you use the Sun assembler, this
4180 change apparently runs afoul of Sun bug 4910101 (which is referenced as
4181 an x86-only problem by Sun, probably because they do not use DWARF-2).
4182 A symptom of the problem is that you cannot compile C++ programs like
4183 @command{groff} 1.19.1 without getting messages similar to the following:
4186 ld: warning: relocation error: R_SPARC_UA32: @dots{}
4187 external symbolic relocation against non-allocatable section
4188 .debug_info cannot be processed at runtime: relocation ignored.
4191 To work around this problem, compile with @option{-gstabs+} instead of
4194 When configuring the GNU Multiple Precision Library (GMP) or the MPFR
4195 library on a Solaris 7 or later system, the canonical target triplet
4196 must be specified as the @command{build} parameter on the configure
4197 line. This triplet can be obtained by invoking ./config.guess in
4198 the toplevel source directory of GCC (and not that of GMP or MPFR).
4199 For example on a Solaris 7 system:
4202 % ./configure --build=sparc-sun-solaris2.7 --prefix=xxx
4208 @heading @anchor{sparc-sun-solaris27}sparc-sun-solaris2.7
4210 Sun patch 107058-01 (1999-01-13) for Solaris 7/SPARC triggers a bug in
4211 the dynamic linker. This problem (Sun bug 4210064) affects GCC 2.8
4212 and later, including all EGCS releases. Sun formerly recommended
4213 107058-01 for all Solaris 7 users, but around 1999-09-01 it started to
4214 recommend it only for people who use Sun's compilers.
4216 Here are some workarounds to this problem:
4219 Do not install Sun patch 107058-01 until after Sun releases a
4220 complete patch for bug 4210064. This is the simplest course to take,
4221 unless you must also use Sun's C compiler. Unfortunately 107058-01
4222 is preinstalled on some new Solaris 7-based hosts, so you may have to
4226 Copy the original, unpatched Solaris 7
4227 @command{/usr/ccs/bin/as} into
4228 @command{/usr/local/libexec/gcc/sparc-sun-solaris2.7/3.4/as},
4229 adjusting the latter name to fit your local conventions and software
4233 Install Sun patch 106950-03 (1999-05-25) or later. Nobody with
4234 both 107058-01 and 106950-03 installed has reported the bug with GCC
4235 and Sun's dynamic linker. This last course of action is riskiest,
4236 for two reasons. First, you must install 106950 on all hosts that
4237 run code generated by GCC; it doesn't suffice to install it only on
4238 the hosts that run GCC itself. Second, Sun says that 106950-03 is
4239 only a partial fix for bug 4210064, but Sun doesn't know whether the
4240 partial fix is adequate for GCC@. Revision -08 or later should fix
4241 the bug. The current (as of 2004-05-23) revision is -24, and is included in
4242 the Solaris 7 Recommended Patch Cluster.
4245 GCC 3.3 triggers a bug in version 5.0 Alpha 03/27/98 of the Sun assembler,
4246 which causes a bootstrap failure when linking the 64-bit shared version of
4247 libgcc. A typical error message is:
4250 ld: fatal: relocation error: R_SPARC_32: file libgcc/sparcv9/_muldi3.o:
4251 symbol <unknown>: offset 0xffffffff7ec133e7 is non-aligned.
4254 This bug has been fixed in the final 5.0 version of the assembler.
4256 A similar problem was reported for version Sun WorkShop 6 99/08/18 of the
4257 Sun assembler, which causes a bootstrap failure with GCC 4.0.0:
4260 ld: fatal: relocation error: R_SPARC_DISP32:
4261 file .libs/libstdc++.lax/libsupc++convenience.a/vterminate.o:
4262 symbol <unknown>: offset 0xfccd33ad is non-aligned
4265 This bug has been fixed in more recent revisions of the assembler.
4270 @heading @anchor{sparc-x-linux}sparc-*-linux*
4272 GCC versions 3.0 and higher require binutils 2.11.2 and glibc 2.2.4
4273 or newer on this platform. All earlier binutils and glibc
4274 releases mishandled unaligned relocations on @code{sparc-*-*} targets.
4280 @heading @anchor{sparc64-x-solaris2}sparc64-*-solaris2*
4282 When configuring the GNU Multiple Precision Library (GMP) or the
4283 MPFR library, the canonical target triplet must be specified as
4284 the @command{build} parameter on the configure line. For example
4285 on a Solaris 7 system:
4288 % ./configure --build=sparc64-sun-solaris2.7 --prefix=xxx
4291 The following compiler flags must be specified in the configure
4292 step in order to bootstrap this target with the Sun compiler:
4295 % CC="cc -xarch=v9 -xildoff" @var{srcdir}/configure [@var{options}] [@var{target}]
4298 @option{-xarch=v9} specifies the SPARC-V9 architecture to the Sun toolchain
4299 and @option{-xildoff} turns off the incremental linker.
4304 @heading @anchor{sparcv9-x-solaris2}sparcv9-*-solaris2*
4306 This is a synonym for sparc64-*-solaris2*.
4311 @heading @anchor{x-x-vxworks}*-*-vxworks*
4312 Support for VxWorks is in flux. At present GCC supports @emph{only} the
4313 very recent VxWorks 5.5 (aka Tornado 2.2) release, and only on PowerPC@.
4314 We welcome patches for other architectures supported by VxWorks 5.5.
4315 Support for VxWorks AE would also be welcome; we believe this is merely
4316 a matter of writing an appropriate ``configlette'' (see below). We are
4317 not interested in supporting older, a.out or COFF-based, versions of
4320 VxWorks comes with an older version of GCC installed in
4321 @file{@var{$WIND_BASE}/host}; we recommend you do not overwrite it.
4322 Choose an installation @var{prefix} entirely outside @var{$WIND_BASE}.
4323 Before running @command{configure}, create the directories @file{@var{prefix}}
4324 and @file{@var{prefix}/bin}. Link or copy the appropriate assembler,
4325 linker, etc.@: into @file{@var{prefix}/bin}, and set your @var{PATH} to
4326 include that directory while running both @command{configure} and
4329 You must give @command{configure} the
4330 @option{--with-headers=@var{$WIND_BASE}/target/h} switch so that it can
4331 find the VxWorks system headers. Since VxWorks is a cross compilation
4332 target only, you must also specify @option{--target=@var{target}}.
4333 @command{configure} will attempt to create the directory
4334 @file{@var{prefix}/@var{target}/sys-include} and copy files into it;
4335 make sure the user running @command{configure} has sufficient privilege
4338 GCC's exception handling runtime requires a special ``configlette''
4339 module, @file{contrib/gthr_supp_vxw_5x.c}. Follow the instructions in
4340 that file to add the module to your kernel build. (Future versions of
4341 VxWorks will incorporate this module.)
4346 @heading @anchor{x86-64-x-x}x86_64-*-*, amd64-*-*
4348 GCC supports the x86-64 architecture implemented by the AMD64 processor
4349 (amd64-*-* is an alias for x86_64-*-*) on GNU/Linux, FreeBSD and NetBSD@.
4350 On GNU/Linux the default is a bi-arch compiler which is able to generate
4351 both 64-bit x86-64 and 32-bit x86 code (via the @option{-m32} switch).
4356 @heading @anchor{xtensa-x-elf}xtensa*-*-elf
4358 This target is intended for embedded Xtensa systems using the
4359 @samp{newlib} C library. It uses ELF but does not support shared
4360 objects. Designed-defined instructions specified via the
4361 Tensilica Instruction Extension (TIE) language are only supported
4362 through inline assembly.
4364 The Xtensa configuration information must be specified prior to
4365 building GCC@. The @file{include/xtensa-config.h} header
4366 file contains the configuration information. If you created your
4367 own Xtensa configuration with the Xtensa Processor Generator, the
4368 downloaded files include a customized copy of this header file,
4369 which you can use to replace the default header file.
4374 @heading @anchor{xtensa-x-linux}xtensa*-*-linux*
4376 This target is for Xtensa systems running GNU/Linux. It supports ELF
4377 shared objects and the GNU C library (glibc). It also generates
4378 position-independent code (PIC) regardless of whether the
4379 @option{-fpic} or @option{-fPIC} options are used. In other
4380 respects, this target is the same as the
4381 @uref{#xtensa*-*-elf,,@samp{xtensa*-*-elf}} target.
4386 @heading @anchor{windows}Microsoft Windows
4388 @subheading Intel 16-bit versions
4389 The 16-bit versions of Microsoft Windows, such as Windows 3.1, are not
4392 However, the 32-bit port has limited support for Microsoft
4393 Windows 3.11 in the Win32s environment, as a target only. See below.
4395 @subheading Intel 32-bit versions
4397 The 32-bit versions of Windows, including Windows 95, Windows NT, Windows
4398 XP, and Windows Vista, are supported by several different target
4399 platforms. These targets differ in which Windows subsystem they target
4400 and which C libraries are used.
4403 @item Cygwin @uref{#x-x-cygwin,,*-*-cygwin}: Cygwin provides a user-space
4404 Linux API emulation layer in the Win32 subsystem.
4405 @item Interix @uref{#x-x-interix,,*-*-interix}: The Interix subsystem
4406 provides native support for POSIX.
4407 @item MinGW @uref{#x-x-mingw32,,*-*-mingw32}: MinGW is a native GCC port for
4408 the Win32 subsystem that provides a subset of POSIX.
4409 @item MKS i386-pc-mks: NuTCracker from MKS. See
4410 @uref{http://www.mkssoftware.com/} for more information.
4413 @subheading Intel 64-bit versions
4415 GCC contains support for x86-64 using the mingw-w64
4416 runtime library, available from @uref{http://mingw-w64.sourceforge.net/}.
4417 This library should be used with the target triple x86_64-pc-mingw32.
4419 Presently Windows for Itanium is not supported.
4421 @subheading Windows CE
4423 Windows CE is supported as a target only on ARM (arm-wince-pe), Hitachi
4424 SuperH (sh-wince-pe), and MIPS (mips-wince-pe).
4426 @subheading Other Windows Platforms
4428 GCC no longer supports Windows NT on the Alpha or PowerPC.
4430 GCC no longer supports the Windows POSIX subsystem. However, it does
4431 support the Interix subsystem. See above.
4433 Old target names including *-*-winnt and *-*-windowsnt are no longer used.
4435 PW32 (i386-pc-pw32) support was never completed, and the project seems to
4436 be inactive. See @uref{http://pw32.sourceforge.net/} for more information.
4438 UWIN support has been removed due to a lack of maintenance.
4443 @heading @anchor{x-x-cygwin}*-*-cygwin
4445 Ports of GCC are included with the
4446 @uref{http://www.cygwin.com/,,Cygwin environment}.
4448 GCC will build under Cygwin without modification; it does not build
4449 with Microsoft's C++ compiler and there are no plans to make it do so.
4451 The Cygwin native compiler can be configured to target any 32-bit x86
4452 cpu architecture desired; the default is i686-pc-cygwin. It should be
4453 used with as up-to-date a version of binutils as possible; use either
4454 the latest official GNU binutils release in the Cygwin distribution,
4455 or version 2.20 or above if building your own.
4460 @heading @anchor{x-x-interix}*-*-interix
4462 The Interix target is used by OpenNT, Interix, Services For UNIX (SFU),
4463 and Subsystem for UNIX-based Applications (SUA). Applications compiled
4464 with this target run in the Interix subsystem, which is separate from
4465 the Win32 subsystem. This target was last known to work in GCC 3.3.
4467 For more information, see @uref{http://www.interix.com/}.
4472 @heading @anchor{x-x-mingw32}*-*-mingw32
4474 GCC will build with and support only MinGW runtime 3.12 and later.
4475 Earlier versions of headers are incompatible with the new default semantics
4476 of @code{extern inline} in @code{-std=c99} and @code{-std=gnu99} modes.
4481 @heading @anchor{older}Older systems
4483 GCC contains support files for many older (1980s and early
4484 1990s) Unix variants. For the most part, support for these systems
4485 has not been deliberately removed, but it has not been maintained for
4486 several years and may suffer from bitrot.
4488 Starting with GCC 3.1, each release has a list of ``obsoleted'' systems.
4489 Support for these systems is still present in that release, but
4490 @command{configure} will fail unless the @option{--enable-obsolete}
4491 option is given. Unless a maintainer steps forward, support for these
4492 systems will be removed from the next release of GCC@.
4494 Support for old systems as hosts for GCC can cause problems if the
4495 workarounds for compiler, library and operating system bugs affect the
4496 cleanliness or maintainability of the rest of GCC@. In some cases, to
4497 bring GCC up on such a system, if still possible with current GCC, may
4498 require first installing an old version of GCC which did work on that
4499 system, and using it to compile a more recent GCC, to avoid bugs in the
4500 vendor compiler. Old releases of GCC 1 and GCC 2 are available in the
4501 @file{old-releases} directory on the @uref{../mirrors.html,,GCC mirror
4502 sites}. Header bugs may generally be avoided using
4503 @command{fixincludes}, but bugs or deficiencies in libraries and the
4504 operating system may still cause problems.
4506 Support for older systems as targets for cross-compilation is less
4507 problematic than support for them as hosts for GCC; if an enthusiast
4508 wishes to make such a target work again (including resurrecting any of
4509 the targets that never worked with GCC 2, starting from the last
4510 version before they were removed), patches
4511 @uref{../contribute.html,,following the usual requirements} would be
4512 likely to be accepted, since they should not affect the support for more
4515 For some systems, old versions of GNU binutils may also be useful,
4516 and are available from @file{pub/binutils/old-releases} on
4517 @uref{http://sourceware.org/mirrors.html,,sourceware.org mirror sites}.
4519 Some of the information on specific systems above relates to
4520 such older systems, but much of the information
4521 about GCC on such systems (which may no longer be applicable to
4522 current GCC) is to be found in the GCC texinfo manual.
4527 @heading @anchor{elf}all ELF targets (SVR4, Solaris 2, etc.)
4529 C++ support is significantly better on ELF targets if you use the
4530 @uref{./configure.html#with-gnu-ld,,GNU linker}; duplicate copies of
4531 inlines, vtables and template instantiations will be discarded
4540 @uref{./index.html,,Return to the GCC Installation page}
4544 @c ***Old documentation******************************************************
4546 @include install-old.texi
4552 @uref{./index.html,,Return to the GCC Installation page}
4556 @c ***GFDL********************************************************************
4564 @uref{./index.html,,Return to the GCC Installation page}
4568 @c ***************************************************************************
4569 @c Part 6 The End of the Document
4571 @comment node-name, next, previous, up
4572 @node Concept Index, , GNU Free Documentation License, Top
4576 @unnumbered Concept Index