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 Free Software Foundation, Inc.
49 @c *** Converted to texinfo by Dean Wakerley, dean@wakerley.com
51 @c IMPORTANT: whenever you modify this file, run `install.texi2html' to
52 @c test the generation of HTML documents for the gcc.gnu.org web pages.
54 @c Do not use @footnote{} in this file as it breaks install.texi2html!
56 @c Include everything if we're not making html
60 @set prerequisiteshtml
71 @c Part 2 Summary Description and Copyright
73 Copyright @copyright{} 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997, 1998,
74 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008 Free Software Foundation, Inc.
76 Permission is granted to copy, distribute and/or modify this document
77 under the terms of the GNU Free Documentation License, Version 1.2 or
78 any later version published by the Free Software Foundation; with no
79 Invariant Sections, the Front-Cover texts being (a) (see below), and
80 with the Back-Cover Texts being (b) (see below). A copy of the
81 license is included in the section entitled ``@uref{./gfdl.html,,GNU
82 Free Documentation License}''.
84 (a) The FSF's Front-Cover Text is:
88 (b) The FSF's Back-Cover Text is:
90 You have freedom to copy and modify this GNU Manual, like GNU
91 software. Copies published by the Free Software Foundation raise
92 funds for GNU development.
97 @dircategory Software development
99 * gccinstall: (gccinstall). Installing the GNU Compiler Collection.
102 @c Part 3 Titlepage and Copyright
104 @title Installing GCC
107 @c The following two commands start the copyright page.
109 @vskip 0pt plus 1filll
113 @c Part 4 Top node, Master Menu, and/or Table of Contents
116 @comment node-name, next, Previous, up
119 * Installing GCC:: This document describes the generic installation
120 procedure for GCC as well as detailing some target
121 specific installation instructions.
123 * Specific:: Host/target specific installation notes for GCC.
124 * Binaries:: Where to get pre-compiled binaries.
126 * Old:: Old installation documentation.
128 * GNU Free Documentation License:: How you can copy and share this manual.
129 * Concept Index:: This index has two entries.
137 @c Part 5 The Body of the Document
138 @c ***Installing GCC**********************************************************
140 @comment node-name, next, previous, up
141 @node Installing GCC, Binaries, , Top
145 @chapter Installing GCC
148 The latest version of this document is always available at
149 @uref{http://gcc.gnu.org/install/,,http://gcc.gnu.org/install/}.
151 This document describes the generic installation procedure for GCC as well
152 as detailing some target specific installation instructions.
154 GCC includes several components that previously were separate distributions
155 with their own installation instructions. This document supersedes all
156 package specific installation instructions.
158 @emph{Before} starting the build/install procedure please check the
160 @ref{Specific, host/target specific installation notes}.
163 @uref{specific.html,,host/target specific installation notes}.
165 We recommend you browse the entire generic installation instructions before
168 Lists of successful builds for released versions of GCC are
169 available at @uref{http://gcc.gnu.org/buildstat.html}.
170 These lists are updated as new information becomes available.
172 The installation procedure itself is broken into five steps.
177 * Downloading the source::
180 * Testing:: (optional)
187 @uref{prerequisites.html,,Prerequisites}
189 @uref{download.html,,Downloading the source}
191 @uref{configure.html,,Configuration}
193 @uref{build.html,,Building}
195 @uref{test.html,,Testing} (optional)
197 @uref{finalinstall.html,,Final install}
201 Please note that GCC does not support @samp{make uninstall} and probably
202 won't do so in the near future as this would open a can of worms. Instead,
203 we suggest that you install GCC into a directory of its own and simply
204 remove that directory when you do not need that specific version of GCC
205 any longer, and, if shared libraries are installed there as well, no
206 more binaries exist that use them.
209 There are also some @uref{old.html,,old installation instructions},
210 which are mostly obsolete but still contain some information which has
211 not yet been merged into the main part of this manual.
219 @uref{./index.html,,Return to the GCC Installation page}
225 @c ***Prerequisites**************************************************
227 @comment node-name, next, previous, up
228 @node Prerequisites, Downloading the source, , Installing GCC
230 @ifset prerequisiteshtml
232 @chapter Prerequisites
234 @cindex Prerequisites
236 GCC requires that various tools and packages be available for use in the
237 build procedure. Modifying GCC sources requires additional tools
240 @heading Tools/packages necessary for building GCC
242 @item ISO C90 compiler
243 Necessary to bootstrap GCC, although versions of GCC prior
244 to 3.4 also allow bootstrapping with a traditional (K&R) C compiler.
246 To build all languages in a cross-compiler or other configuration where
247 3-stage bootstrap is not performed, you need to start with an existing
248 GCC binary (version 2.95 or later) because source code for language
249 frontends other than C might use GCC extensions.
253 In order to build the Ada compiler (GNAT) you must already have GNAT
254 installed because portions of the Ada frontend are written in Ada (with
255 GNAT extensions.) Refer to the Ada installation instructions for more
256 specific information.
258 @item A ``working'' POSIX compatible shell, or GNU bash
260 Necessary when running @command{configure} because some
261 @command{/bin/sh} shells have bugs and may crash when configuring the
262 target libraries. In other cases, @command{/bin/sh} or @command{ksh}
263 have disastrous corner-case performance problems. This
264 can cause target @command{configure} runs to literally take days to
265 complete in some cases.
267 So on some platforms @command{/bin/ksh} is sufficient, on others it
268 isn't. See the host/target specific instructions for your platform, or
269 use @command{bash} to be sure. Then set @env{CONFIG_SHELL} in your
270 environment to your ``good'' shell prior to running
271 @command{configure}/@command{make}.
273 @command{zsh} is not a fully compliant POSIX shell and will not
274 work when configuring GCC@.
278 Necessary in some circumstances, optional in others. See the
279 host/target specific instructions for your platform for the exact
282 @item gzip version 1.2.4 (or later) or
283 @itemx bzip2 version 1.0.2 (or later)
285 Necessary to uncompress GCC @command{tar} files when source code is
286 obtained via FTP mirror sites.
288 @item GNU make version 3.80 (or later)
290 You must have GNU make installed to build GCC@.
292 @item GNU tar version 1.14 (or later)
294 Necessary (only on some platforms) to untar the source code. Many
295 systems' @command{tar} programs will also work, only try GNU
296 @command{tar} if you have problems.
298 @item GNU Multiple Precision Library (GMP) version 4.1 (or later)
300 Necessary to build GCC@. If you do not have it installed in your
301 library search path, you will have to configure with the
302 @option{--with-gmp} configure option. See also
303 @option{--with-gmp-lib} and @option{--with-gmp-include}.
305 @item MPFR Library version 2.3.0 (or later)
307 Necessary to build GCC@. It can be downloaded from
308 @uref{http://www.mpfr.org/}. The version of MPFR that is bundled with
309 GMP 4.1.x contains numerous bugs. Although GCC may appear to function
310 with the buggy versions of MPFR, there are a few bugs that will not be
311 fixed when using this version. It is strongly recommended to upgrade
312 to the recommended version of MPFR.
314 The @option{--with-mpfr} configure option should be used if your MPFR
315 Library is not installed in your default library search path. See
316 also @option{--with-mpfr-lib} and @option{--with-mpfr-include}.
318 @item @command{jar}, or InfoZIP (@command{zip} and @command{unzip})
320 Necessary to build libgcj, the GCJ runtime.
325 @heading Tools/packages necessary for modifying GCC
327 @item autoconf version 2.59
328 @itemx GNU m4 version 1.4 (or later)
330 Necessary when modifying @file{configure.ac}, @file{aclocal.m4}, etc.@:
331 to regenerate @file{configure} and @file{config.in} files.
333 @item automake version 1.9.6
335 Necessary when modifying a @file{Makefile.am} file to regenerate its
336 associated @file{Makefile.in}.
338 Much of GCC does not use automake, so directly edit the @file{Makefile.in}
339 file. Specifically this applies to the @file{gcc}, @file{intl},
340 @file{libcpp}, @file{libiberty}, @file{libobjc} directories as well
341 as any of their subdirectories.
343 For directories that use automake, GCC requires the latest release in
344 the 1.9.x series, which is currently 1.9.6. When regenerating a directory
345 to a newer version, please update all the directories using an older 1.9.x
346 to the latest released version.
348 @item gettext version 0.14.5 (or later)
350 Needed to regenerate @file{gcc.pot}.
352 @item gperf version 2.7.2 (or later)
354 Necessary when modifying @command{gperf} input files, e.g.@:
355 @file{gcc/cp/cfns.gperf} to regenerate its associated header file, e.g.@:
356 @file{gcc/cp/cfns.h}.
362 Necessary to run the GCC testsuite; see the section on testing for details.
364 @item autogen version 5.5.4 (or later) and
365 @itemx guile version 1.4.1 (or later)
367 Necessary to regenerate @file{fixinc/fixincl.x} from
368 @file{fixinc/inclhack.def} and @file{fixinc/*.tpl}.
370 Necessary to run @samp{make check} for @file{fixinc}.
372 Necessary to regenerate the top level @file{Makefile.in} file from
373 @file{Makefile.tpl} and @file{Makefile.def}.
375 @item Flex version 2.5.4 (or later)
377 Necessary when modifying @file{*.l} files.
379 Necessary to build GCC during development because the generated output
380 files are not included in the SVN repository. They are included in
383 @item Texinfo version 4.4 (or later)
385 Necessary for running @command{makeinfo} when modifying @file{*.texi}
386 files to test your changes.
388 Necessary for running @command{make dvi} or @command{make pdf} to
389 create printable documentation in DVI or PDF format. Texinfo version
390 4.8 or later is required for @command{make pdf}.
392 Necessary to build GCC documentation during development because the
393 generated output files are not included in the SVN repository. They are
394 included in releases.
396 @item @TeX{} (any working version)
398 Necessary for running @command{texi2dvi} and @command{texi2pdf}, which
399 are used when running @command{make dvi} or @command{make pdf} to create
400 DVI or PDF files, respectively.
402 @item SVN (any version)
403 @itemx SSH (any version)
405 Necessary to access the SVN repository. Public releases and weekly
406 snapshots of the development sources are also available via FTP@.
408 @item Perl version 5.6.1 (or later)
410 Necessary when regenerating @file{Makefile} dependencies in libiberty.
411 Necessary when regenerating @file{libiberty/functions.texi}.
412 Necessary when generating manpages from Texinfo manuals.
413 Necessary when targetting Darwin, building libstdc++,
414 and not using @option{--disable-symvers}.
415 Used by various scripts to generate some files included in SVN (mainly
416 Unicode-related and rarely changing) from source tables.
418 @item GNU diffutils version 2.7 (or later)
420 Useful when submitting patches for the GCC source code.
422 @item patch version 2.5.4 (or later)
424 Necessary when applying patches, created with @command{diff}, to one's
430 If you wish to modify @file{.java} files in libjava, you will need to
431 configure with @option{--enable-java-maintainer-mode}, and you will need
432 to have executables named @command{ecj1} and @command{gjavah} in your path.
433 The @command{ecj1} executable should run the Eclipse Java compiler via
434 the GCC-specific entry point. You can download a suitable jar from
435 @uref{ftp://sourceware.org/pub/java/}, or by running the script
436 @command{contrib/download_ecj}.
445 @uref{./index.html,,Return to the GCC Installation page}
449 @c ***Downloading the source**************************************************
451 @comment node-name, next, previous, up
452 @node Downloading the source, Configuration, Prerequisites, Installing GCC
456 @chapter Downloading GCC
458 @cindex Downloading GCC
459 @cindex Downloading the Source
461 GCC is distributed via @uref{http://gcc.gnu.org/svn.html,,SVN} and FTP
462 tarballs compressed with @command{gzip} or
463 @command{bzip2}. It is possible to download a full distribution or specific
466 Please refer to the @uref{http://gcc.gnu.org/releases.html,,releases web page}
467 for information on how to obtain GCC@.
469 The full distribution includes the C, C++, Objective-C, Fortran, Java,
470 and Ada (in the case of GCC 3.1 and later) compilers. The full
471 distribution also includes runtime libraries for C++, Objective-C,
472 Fortran, and Java. In GCC 3.0 and later versions, the GNU compiler
473 testsuites are also included in the full distribution.
475 If you choose to download specific components, you must download the core
476 GCC distribution plus any language specific distributions you wish to
477 use. The core distribution includes the C language front end as well as the
478 shared components. Each language has a tarball which includes the language
479 front end as well as the language runtime (when appropriate).
481 Unpack the core distribution as well as any language specific
482 distributions in the same directory.
484 If you also intend to build binutils (either to upgrade an existing
485 installation or for use in place of the corresponding tools of your
486 OS), unpack the binutils distribution either in the same directory or
487 a separate one. In the latter case, add symbolic links to any
488 components of the binutils you intend to build alongside the compiler
489 (@file{bfd}, @file{binutils}, @file{gas}, @file{gprof}, @file{ld},
490 @file{opcodes}, @dots{}) to the directory containing the GCC sources.
497 @uref{./index.html,,Return to the GCC Installation page}
501 @c ***Configuration***********************************************************
503 @comment node-name, next, previous, up
504 @node Configuration, Building, Downloading the source, Installing GCC
508 @chapter Installing GCC: Configuration
510 @cindex Configuration
511 @cindex Installing GCC: Configuration
513 Like most GNU software, GCC must be configured before it can be built.
514 This document describes the recommended configuration procedure
515 for both native and cross targets.
517 We use @var{srcdir} to refer to the toplevel source directory for
518 GCC; we use @var{objdir} to refer to the toplevel build/object directory.
520 If you obtained the sources via SVN, @var{srcdir} must refer to the top
521 @file{gcc} directory, the one where the @file{MAINTAINERS} can be found,
522 and not its @file{gcc} subdirectory, otherwise the build will fail.
524 If either @var{srcdir} or @var{objdir} is located on an automounted NFS
525 file system, the shell's built-in @command{pwd} command will return
526 temporary pathnames. Using these can lead to various sorts of build
527 problems. To avoid this issue, set the @env{PWDCMD} environment
528 variable to an automounter-aware @command{pwd} command, e.g.,
529 @command{pawd} or @samp{amq -w}, during the configuration and build
532 First, we @strong{highly} recommend that GCC be built into a
533 separate directory than the sources which does @strong{not} reside
534 within the source tree. This is how we generally build GCC; building
535 where @var{srcdir} == @var{objdir} should still work, but doesn't
536 get extensive testing; building where @var{objdir} is a subdirectory
537 of @var{srcdir} is unsupported.
539 If you have previously built GCC in the same directory for a
540 different target machine, do @samp{make distclean} to delete all files
541 that might be invalid. One of the files this deletes is @file{Makefile};
542 if @samp{make distclean} complains that @file{Makefile} does not exist
543 or issues a message like ``don't know how to make distclean'' it probably
544 means that the directory is already suitably clean. However, with the
545 recommended method of building in a separate @var{objdir}, you should
546 simply use a different @var{objdir} for each target.
548 Second, when configuring a native system, either @command{cc} or
549 @command{gcc} must be in your path or you must set @env{CC} in
550 your environment before running configure. Otherwise the configuration
554 Note that the bootstrap compiler and the resulting GCC must be link
555 compatible, else the bootstrap will fail with linker errors about
556 incompatible object file formats. Several multilibed targets are
557 affected by this requirement, see
559 @ref{Specific, host/target specific installation notes}.
562 @uref{specific.html,,host/target specific installation notes}.
571 % @var{srcdir}/configure [@var{options}] [@var{target}]
574 @heading Distributor options
576 If you will be distributing binary versions of GCC, with modifications
577 to the source code, you should use the options described in this
578 section to make clear that your version contains modifications.
581 @item --with-pkgversion=@var{version}
582 Specify a string that identifies your package. You may wish
583 to include a build number or build date. This version string will be
584 included in the output of @command{gcc --version}. This suffix does
585 not replace the default version string, only the @samp{GCC} part.
587 The default value is @samp{GCC}.
589 @item --with-bugurl=@var{url}
590 Specify the URL that users should visit if they wish to report a bug.
591 You are of course welcome to forward bugs reported to you to the FSF,
592 if you determine that they are not bugs in your modifications.
594 The default value refers to the FSF's GCC bug tracker.
598 @heading Target specification
601 GCC has code to correctly determine the correct value for @var{target}
602 for nearly all native systems. Therefore, we highly recommend you not
603 provide a configure target when configuring a native compiler.
606 @var{target} must be specified as @option{--target=@var{target}}
607 when configuring a cross compiler; examples of valid targets would be
608 m68k-coff, sh-elf, etc.
611 Specifying just @var{target} instead of @option{--target=@var{target}}
612 implies that the host defaults to @var{target}.
616 @heading Options specification
618 Use @var{options} to override several configure time options for
619 GCC@. A list of supported @var{options} follows; @samp{configure
620 --help} may list other options, but those not listed below may not
621 work and should not normally be used.
623 Note that each @option{--enable} option has a corresponding
624 @option{--disable} option and that each @option{--with} option has a
625 corresponding @option{--without} option.
628 @item --prefix=@var{dirname}
629 Specify the toplevel installation
630 directory. This is the recommended way to install the tools into a directory
631 other than the default. The toplevel installation directory defaults to
634 We @strong{highly} recommend against @var{dirname} being the same or a
635 subdirectory of @var{objdir} or vice versa. If specifying a directory
636 beneath a user's home directory tree, some shells will not expand
637 @var{dirname} correctly if it contains the @samp{~} metacharacter; use
640 The following standard @command{autoconf} options are supported. Normally you
641 should not need to use these options.
643 @item --exec-prefix=@var{dirname}
644 Specify the toplevel installation directory for architecture-dependent
645 files. The default is @file{@var{prefix}}.
647 @item --bindir=@var{dirname}
648 Specify the installation directory for the executables called by users
649 (such as @command{gcc} and @command{g++}). The default is
650 @file{@var{exec-prefix}/bin}.
652 @item --libdir=@var{dirname}
653 Specify the installation directory for object code libraries and
654 internal data files of GCC@. The default is @file{@var{exec-prefix}/lib}.
656 @item --libexecdir=@var{dirname}
657 Specify the installation directory for internal executables of GCC@.
658 The default is @file{@var{exec-prefix}/libexec}.
660 @item --with-slibdir=@var{dirname}
661 Specify the installation directory for the shared libgcc library. The
662 default is @file{@var{libdir}}.
664 @item --infodir=@var{dirname}
665 Specify the installation directory for documentation in info format.
666 The default is @file{@var{prefix}/info}.
668 @item --datadir=@var{dirname}
669 Specify the installation directory for some architecture-independent
670 data files referenced by GCC@. The default is @file{@var{prefix}/share}.
672 @item --mandir=@var{dirname}
673 Specify the installation directory for manual pages. The default is
674 @file{@var{prefix}/man}. (Note that the manual pages are only extracts from
675 the full GCC manuals, which are provided in Texinfo format. The manpages
676 are derived by an automatic conversion process from parts of the full
679 @item --with-gxx-include-dir=@var{dirname}
681 the installation directory for G++ header files. The default is
682 @file{@var{prefix}/include/c++/@var{version}}.
686 @item --program-prefix=@var{prefix}
687 GCC supports some transformations of the names of its programs when
688 installing them. This option prepends @var{prefix} to the names of
689 programs to install in @var{bindir} (see above). For example, specifying
690 @option{--program-prefix=foo-} would result in @samp{gcc}
691 being installed as @file{/usr/local/bin/foo-gcc}.
693 @item --program-suffix=@var{suffix}
694 Appends @var{suffix} to the names of programs to install in @var{bindir}
695 (see above). For example, specifying @option{--program-suffix=-3.1}
696 would result in @samp{gcc} being installed as
697 @file{/usr/local/bin/gcc-3.1}.
699 @item --program-transform-name=@var{pattern}
700 Applies the @samp{sed} script @var{pattern} to be applied to the names
701 of programs to install in @var{bindir} (see above). @var{pattern} has to
702 consist of one or more basic @samp{sed} editing commands, separated by
703 semicolons. For example, if you want the @samp{gcc} program name to be
704 transformed to the installed program @file{/usr/local/bin/myowngcc} and
705 the @samp{g++} program name to be transformed to
706 @file{/usr/local/bin/gspecial++} without changing other program names,
707 you could use the pattern
708 @option{--program-transform-name='s/^gcc$/myowngcc/; s/^g++$/gspecial++/'}
709 to achieve this effect.
711 All three options can be combined and used together, resulting in more
712 complex conversion patterns. As a basic rule, @var{prefix} (and
713 @var{suffix}) are prepended (appended) before further transformations
714 can happen with a special transformation script @var{pattern}.
716 As currently implemented, this option only takes effect for native
717 builds; cross compiler binaries' names are not transformed even when a
718 transformation is explicitly asked for by one of these options.
720 For native builds, some of the installed programs are also installed
721 with the target alias in front of their name, as in
722 @samp{i686-pc-linux-gnu-gcc}. All of the above transformations happen
723 before the target alias is prepended to the name---so, specifying
724 @option{--program-prefix=foo-} and @option{program-suffix=-3.1}, the
725 resulting binary would be installed as
726 @file{/usr/local/bin/i686-pc-linux-gnu-foo-gcc-3.1}.
728 As a last shortcoming, none of the installed Ada programs are
729 transformed yet, which will be fixed in some time.
731 @item --with-local-prefix=@var{dirname}
733 installation directory for local include files. The default is
734 @file{/usr/local}. Specify this option if you want the compiler to
735 search directory @file{@var{dirname}/include} for locally installed
736 header files @emph{instead} of @file{/usr/local/include}.
738 You should specify @option{--with-local-prefix} @strong{only} if your
739 site has a different convention (not @file{/usr/local}) for where to put
742 The default value for @option{--with-local-prefix} is @file{/usr/local}
743 regardless of the value of @option{--prefix}. Specifying
744 @option{--prefix} has no effect on which directory GCC searches for
745 local header files. This may seem counterintuitive, but actually it is
748 The purpose of @option{--prefix} is to specify where to @emph{install
749 GCC}. The local header files in @file{/usr/local/include}---if you put
750 any in that directory---are not part of GCC@. They are part of other
751 programs---perhaps many others. (GCC installs its own header files in
752 another directory which is based on the @option{--prefix} value.)
754 Both the local-prefix include directory and the GCC-prefix include
755 directory are part of GCC's ``system include'' directories. Although these
756 two directories are not fixed, they need to be searched in the proper
757 order for the correct processing of the include_next directive. The
758 local-prefix include directory is searched before the GCC-prefix
759 include directory. Another characteristic of system include directories
760 is that pedantic warnings are turned off for headers in these directories.
762 Some autoconf macros add @option{-I @var{directory}} options to the
763 compiler command line, to ensure that directories containing installed
764 packages' headers are searched. When @var{directory} is one of GCC's
765 system include directories, GCC will ignore the option so that system
766 directories continue to be processed in the correct order. This
767 may result in a search order different from what was specified but the
768 directory will still be searched.
770 GCC automatically searches for ordinary libraries using
771 @env{GCC_EXEC_PREFIX}. Thus, when the same installation prefix is
772 used for both GCC and packages, GCC will automatically search for
773 both headers and libraries. This provides a configuration that is
774 easy to use. GCC behaves in a manner similar to that when it is
775 installed as a system compiler in @file{/usr}.
777 Sites that need to install multiple versions of GCC may not want to
778 use the above simple configuration. It is possible to use the
779 @option{--program-prefix}, @option{--program-suffix} and
780 @option{--program-transform-name} options to install multiple versions
781 into a single directory, but it may be simpler to use different prefixes
782 and the @option{--with-local-prefix} option to specify the location of the
783 site-specific files for each version. It will then be necessary for
784 users to specify explicitly the location of local site libraries
785 (e.g., with @env{LIBRARY_PATH}).
787 The same value can be used for both @option{--with-local-prefix} and
788 @option{--prefix} provided it is not @file{/usr}. This can be used
789 to avoid the default search of @file{/usr/local/include}.
791 @strong{Do not} specify @file{/usr} as the @option{--with-local-prefix}!
792 The directory you use for @option{--with-local-prefix} @strong{must not}
793 contain any of the system's standard header files. If it did contain
794 them, certain programs would be miscompiled (including GNU Emacs, on
795 certain targets), because this would override and nullify the header
796 file corrections made by the @command{fixincludes} script.
798 Indications are that people who use this option use it based on mistaken
799 ideas of what it is for. People use it as if it specified where to
800 install part of GCC@. Perhaps they make this assumption because
801 installing GCC creates the directory.
803 @item --enable-shared[=@var{package}[,@dots{}]]
804 Build shared versions of libraries, if shared libraries are supported on
805 the target platform. Unlike GCC 2.95.x and earlier, shared libraries
806 are enabled by default on all platforms that support shared libraries.
808 If a list of packages is given as an argument, build shared libraries
809 only for the listed packages. For other packages, only static libraries
810 will be built. Package names currently recognized in the GCC tree are
811 @samp{libgcc} (also known as @samp{gcc}), @samp{libstdc++} (not
812 @samp{libstdc++-v3}), @samp{libffi}, @samp{zlib}, @samp{boehm-gc},
813 @samp{ada}, @samp{libada}, @samp{libjava} and @samp{libobjc}.
814 Note @samp{libiberty} does not support shared libraries at all.
816 Use @option{--disable-shared} to build only static libraries. Note that
817 @option{--disable-shared} does not accept a list of package names as
818 argument, only @option{--enable-shared} does.
820 @item @anchor{with-gnu-as}--with-gnu-as
821 Specify that the compiler should assume that the
822 assembler it finds is the GNU assembler. However, this does not modify
823 the rules to find an assembler and will result in confusion if the
824 assembler found is not actually the GNU assembler. (Confusion may also
825 result if the compiler finds the GNU assembler but has not been
826 configured with @option{--with-gnu-as}.) If you have more than one
827 assembler installed on your system, you may want to use this option in
828 connection with @option{--with-as=@var{pathname}} or
829 @option{--with-build-time-tools=@var{pathname}}.
831 The following systems are the only ones where it makes a difference
832 whether you use the GNU assembler. On any other system,
833 @option{--with-gnu-as} has no effect.
836 @item @samp{hppa1.0-@var{any}-@var{any}}
837 @item @samp{hppa1.1-@var{any}-@var{any}}
838 @item @samp{sparc-sun-solaris2.@var{any}}
839 @item @samp{sparc64-@var{any}-solaris2.@var{any}}
842 On the systems listed above (except for the HP-PA, the SPARC, for ISC on
843 the 386, if you use the GNU assembler, you should also use the GNU linker
844 (and specify @option{--with-gnu-ld}).
846 @item @anchor{with-as}--with-as=@var{pathname}
847 Specify that the compiler should use the assembler pointed to by
848 @var{pathname}, rather than the one found by the standard rules to find
849 an assembler, which are:
852 Unless GCC is being built with a cross compiler, check the
853 @file{@var{libexec}/gcc/@var{target}/@var{version}} directory.
854 @var{libexec} defaults to @file{@var{exec-prefix}/libexec};
855 @var{exec-prefix} defaults to @var{prefix}, which
856 defaults to @file{/usr/local} unless overridden by the
857 @option{--prefix=@var{pathname}} switch described above. @var{target}
858 is the target system triple, such as @samp{sparc-sun-solaris2.7}, and
859 @var{version} denotes the GCC version, such as 3.0.
862 If the target system is the same that you are building on, check
863 operating system specific directories (e.g.@: @file{/usr/ccs/bin} on
867 Check in the @env{PATH} for a tool whose name is prefixed by the
868 target system triple.
871 Check in the @env{PATH} for a tool whose name is not prefixed by the
872 target system triple, if the host and target system triple are
873 the same (in other words, we use a host tool if it can be used for
877 You may want to use @option{--with-as} if no assembler
878 is installed in the directories listed above, or if you have multiple
879 assemblers installed and want to choose one that is not found by the
882 @item @anchor{with-gnu-ld}--with-gnu-ld
883 Same as @uref{#with-gnu-as,,@option{--with-gnu-as}}
886 @item --with-ld=@var{pathname}
887 Same as @uref{#with-as,,@option{--with-as}}
891 Specify that stabs debugging
892 information should be used instead of whatever format the host normally
893 uses. Normally GCC uses the same debug format as the host system.
895 On MIPS based systems and on Alphas, you must specify whether you want
896 GCC to create the normal ECOFF debugging format, or to use BSD-style
897 stabs passed through the ECOFF symbol table. The normal ECOFF debug
898 format cannot fully handle languages other than C@. BSD stabs format can
899 handle other languages, but it only works with the GNU debugger GDB@.
901 Normally, GCC uses the ECOFF debugging format by default; if you
902 prefer BSD stabs, specify @option{--with-stabs} when you configure GCC@.
904 No matter which default you choose when you configure GCC, the user
905 can use the @option{-gcoff} and @option{-gstabs+} options to specify explicitly
906 the debug format for a particular compilation.
908 @option{--with-stabs} is meaningful on the ISC system on the 386, also, if
909 @option{--with-gas} is used. It selects use of stabs debugging
910 information embedded in COFF output. This kind of debugging information
911 supports C++ well; ordinary COFF debugging information does not.
913 @option{--with-stabs} is also meaningful on 386 systems running SVR4. It
914 selects use of stabs debugging information embedded in ELF output. The
915 C++ compiler currently (2.6.0) does not support the DWARF debugging
916 information normally used on 386 SVR4 platforms; stabs provide a
917 workable alternative. This requires gas and gdb, as the normal SVR4
918 tools can not generate or interpret stabs.
920 @item --disable-multilib
921 Specify that multiple target
922 libraries to support different target variants, calling
923 conventions, etc.@: should not be built. The default is to build a
924 predefined set of them.
926 Some targets provide finer-grained control over which multilibs are built
927 (e.g., @option{--disable-softfloat}):
933 fpu, 26bit, underscore, interwork, biendian, nofmult.
936 softfloat, m68881, m68000, m68020.
939 single-float, biendian, softfloat.
941 @item powerpc*-*-*, rs6000*-*-*
942 aix64, pthread, softfloat, powercpu, powerpccpu, powerpcos, biendian,
947 @item --enable-threads
948 Specify that the target
949 supports threads. This affects the Objective-C compiler and runtime
950 library, and exception handling for other languages like C++ and Java.
951 On some systems, this is the default.
953 In general, the best (and, in many cases, the only known) threading
954 model available will be configured for use. Beware that on some
955 systems, GCC has not been taught what threading models are generally
956 available for the system. In this case, @option{--enable-threads} is an
957 alias for @option{--enable-threads=single}.
959 @item --disable-threads
960 Specify that threading support should be disabled for the system.
961 This is an alias for @option{--enable-threads=single}.
963 @item --enable-threads=@var{lib}
965 @var{lib} is the thread support library. This affects the Objective-C
966 compiler and runtime library, and exception handling for other languages
967 like C++ and Java. The possibilities for @var{lib} are:
975 Ada tasking support. For non-Ada programs, this setting is equivalent
976 to @samp{single}. When used in conjunction with the Ada run time, it
977 causes GCC to use the same thread primitives as Ada uses. This option
978 is necessary when using both Ada and the back end exception handling,
979 which is the default for most Ada targets.
981 Generic MACH thread support, known to work on NeXTSTEP@. (Please note
982 that the file needed to support this configuration, @file{gthr-mach.h}, is
983 missing and thus this setting will cause a known bootstrap failure.)
985 This is an alias for @samp{single}.
987 Generic POSIX/Unix98 thread support.
989 Generic POSIX/Unix95 thread support.
991 RTEMS thread support.
993 Disable thread support, should work for all platforms.
995 Sun Solaris 2 thread support.
997 VxWorks thread support.
999 Microsoft Win32 API thread support.
1001 Novell Kernel Services thread support.
1005 Specify that the target supports TLS (Thread Local Storage). Usually
1006 configure can correctly determine if TLS is supported. In cases where
1007 it guesses incorrectly, TLS can be explicitly enabled or disabled with
1008 @option{--enable-tls} or @option{--disable-tls}. This can happen if
1009 the assembler supports TLS but the C library does not, or if the
1010 assumptions made by the configure test are incorrect.
1013 Specify that the target does not support TLS.
1014 This is an alias for @option{--enable-tls=no}.
1016 @item --with-cpu=@var{cpu}
1017 Specify which cpu variant the compiler should generate code for by default.
1018 @var{cpu} will be used as the default value of the @option{-mcpu=} switch.
1019 This option is only supported on some targets, including ARM, i386, M68k,
1020 PowerPC, and SPARC@.
1022 @item --with-schedule=@var{cpu}
1023 @itemx --with-arch=@var{cpu}
1024 @itemx --with-tune=@var{cpu}
1025 @itemx --with-abi=@var{abi}
1026 @itemx --with-fpu=@var{type}
1027 @itemx --with-float=@var{type}
1028 These configure options provide default values for the @option{-mschedule=},
1029 @option{-march=}, @option{-mtune=}, @option{-mabi=}, and @option{-mfpu=}
1030 options and for @option{-mhard-float} or @option{-msoft-float}. As with
1031 @option{--with-cpu}, which switches will be accepted and acceptable values
1032 of the arguments depend on the target.
1034 @item --with-mode=@var{mode}
1035 Specify if the compiler should default to @option{-marm} or @option{-mthumb}.
1036 This option is only supported on ARM targets.
1038 @item --with-divide=@var{type}
1039 Specify how the compiler should generate code for checking for
1040 division by zero. This option is only supported on the MIPS target.
1041 The possibilities for @var{type} are:
1044 Division by zero checks use conditional traps (this is the default on
1045 systems that support conditional traps).
1047 Division by zero checks use the break instruction.
1050 @c If you make --with-llsc the default for additional targets,
1051 @c update the --with-llsc description in the MIPS section below.
1054 On MIPS targets, make @option{-mllsc} the default when no
1055 @option{-mno-lsc} option is passed. This is the default for
1056 Linux-based targets, as the kernel will emulate them if the ISA does
1059 @item --without-llsc
1060 On MIPS targets, make @option{-mno-llsc} the default when no
1061 @option{-mllsc} option is passed.
1063 @item --enable-__cxa_atexit
1064 Define if you want to use __cxa_atexit, rather than atexit, to
1065 register C++ destructors for local statics and global objects.
1066 This is essential for fully standards-compliant handling of
1067 destructors, but requires __cxa_atexit in libc. This option is currently
1068 only available on systems with GNU libc. When enabled, this will cause
1069 @option{-fuse-cxa-atexit} to be passed by default.
1071 @item --enable-target-optspace
1073 libraries should be optimized for code space instead of code speed.
1074 This is the default for the m32r platform.
1077 Specify that a user visible @command{cpp} program should not be installed.
1079 @item --with-cpp-install-dir=@var{dirname}
1080 Specify that the user visible @command{cpp} program should be installed
1081 in @file{@var{prefix}/@var{dirname}/cpp}, in addition to @var{bindir}.
1083 @item --enable-initfini-array
1084 Force the use of sections @code{.init_array} and @code{.fini_array}
1085 (instead of @code{.init} and @code{.fini}) for constructors and
1086 destructors. Option @option{--disable-initfini-array} has the
1087 opposite effect. If neither option is specified, the configure script
1088 will try to guess whether the @code{.init_array} and
1089 @code{.fini_array} sections are supported and, if they are, use them.
1091 @item --enable-maintainer-mode
1092 The build rules that
1093 regenerate the GCC master message catalog @file{gcc.pot} are normally
1094 disabled. This is because it can only be rebuilt if the complete source
1095 tree is present. If you have changed the sources and want to rebuild the
1096 catalog, configuring with @option{--enable-maintainer-mode} will enable
1097 this. Note that you need a recent version of the @code{gettext} tools
1100 @item --disable-bootstrap
1101 For a native build, the default configuration is to perform
1102 a 3-stage bootstrap of the compiler when @samp{make} is invoked,
1103 testing that GCC can compile itself correctly. If you want to disable
1104 this process, you can configure with @option{--disable-bootstrap}.
1106 @item --enable-bootstrap
1107 In special cases, you may want to perform a 3-stage build
1108 even if the target and host triplets are different.
1109 This could happen when the host can run code compiled for
1110 the target (e.g.@: host is i686-linux, target is i486-linux).
1111 Starting from GCC 4.2, to do this you have to configure explicitly
1112 with @option{--enable-bootstrap}.
1114 @item --enable-generated-files-in-srcdir
1115 Neither the .c and .h files that are generated from Bison and flex nor the
1116 info manuals and man pages that are built from the .texi files are present
1117 in the SVN development tree. When building GCC from that development tree,
1118 or from one of our snapshots, those generated files are placed in your
1119 build directory, which allows for the source to be in a readonly
1122 If you configure with @option{--enable-generated-files-in-srcdir} then those
1123 generated files will go into the source directory. This is mainly intended
1124 for generating release or prerelease tarballs of the GCC sources, since it
1125 is not a requirement that the users of source releases to have flex, Bison,
1128 @item --enable-version-specific-runtime-libs
1130 that runtime libraries should be installed in the compiler specific
1131 subdirectory (@file{@var{libdir}/gcc}) rather than the usual places. In
1132 addition, @samp{libstdc++}'s include files will be installed into
1133 @file{@var{libdir}} unless you overruled it by using
1134 @option{--with-gxx-include-dir=@var{dirname}}. Using this option is
1135 particularly useful if you intend to use several versions of GCC in
1136 parallel. This is currently supported by @samp{libgfortran},
1137 @samp{libjava}, @samp{libmudflap}, @samp{libstdc++}, and @samp{libobjc}.
1139 @item --enable-languages=@var{lang1},@var{lang2},@dots{}
1140 Specify that only a particular subset of compilers and
1141 their runtime libraries should be built. For a list of valid values for
1142 @var{langN} you can issue the following command in the
1143 @file{gcc} directory of your GCC source tree:@*
1145 grep language= */config-lang.in
1147 Currently, you can use any of the following:
1148 @code{all}, @code{ada}, @code{c}, @code{c++}, @code{fortran}, @code{java},
1149 @code{objc}, @code{obj-c++}.
1150 Building the Ada compiler has special requirements, see below.
1151 If you do not pass this flag, or specify the option @code{all}, then all
1152 default languages available in the @file{gcc} sub-tree will be configured.
1153 Ada and Objective-C++ are not default languages; the rest are.
1154 Re-defining @code{LANGUAGES} when calling @samp{make} @strong{does not}
1155 work anymore, as those language sub-directories might not have been
1158 @item --enable-stage1-languages=@var{lang1},@var{lang2},@dots{}
1159 Specify that a particular subset of compilers and their runtime
1160 libraries should be built with the system C compiler during stage 1 of
1161 the bootstrap process, rather than only in later stages with the
1162 bootstrapped C compiler. The list of valid values is the same as for
1163 @option{--enable-languages}, and the option @code{all} will select all
1164 of the languages enabled by @option{--enable-languages}. This option is
1165 primarily useful for GCC development; for instance, when a development
1166 version of the compiler cannot bootstrap due to compiler bugs, or when
1167 one is debugging front ends other than the C front end. When this
1168 option is used, one can then build the target libraries for the
1169 specified languages with the stage-1 compiler by using @command{make
1170 stage1-bubble all-target}, or run the testsuite on the stage-1 compiler
1171 for the specified languages using @command{make stage1-start check-gcc}.
1173 @item --disable-libada
1174 Specify that the run-time libraries and tools used by GNAT should not
1175 be built. This can be useful for debugging, or for compatibility with
1176 previous Ada build procedures, when it was required to explicitly
1177 do a @samp{make -C gcc gnatlib_and_tools}.
1179 @item --disable-libssp
1180 Specify that the run-time libraries for stack smashing protection
1181 should not be built.
1183 @item --disable-libgomp
1184 Specify that the run-time libraries used by GOMP should not be built.
1187 Specify that the compiler should
1188 use DWARF 2 debugging information as the default.
1190 @item --enable-targets=all
1191 @itemx --enable-targets=@var{target_list}
1192 Some GCC targets, e.g.@: powerpc64-linux, build bi-arch compilers.
1193 These are compilers that are able to generate either 64-bit or 32-bit
1194 code. Typically, the corresponding 32-bit target, e.g.@:
1195 powerpc-linux for powerpc64-linux, only generates 32-bit code. This
1196 option enables the 32-bit target to be a bi-arch compiler, which is
1197 useful when you want a bi-arch compiler that defaults to 32-bit, and
1198 you are building a bi-arch or multi-arch binutils in a combined tree.
1199 Currently, this option only affects sparc-linux, powerpc-linux and
1202 @item --enable-secureplt
1203 This option enables @option{-msecure-plt} by default for powerpc-linux.
1205 @xref{RS/6000 and PowerPC Options,, RS/6000 and PowerPC Options, gcc,
1206 Using the GNU Compiler Collection (GCC)},
1209 See ``RS/6000 and PowerPC Options'' in the main manual
1213 This option enables @option{-mcld} by default for 32-bit x86 targets.
1215 @xref{i386 and x86-64 Options,, i386 and x86-64 Options, gcc,
1216 Using the GNU Compiler Collection (GCC)},
1219 See ``i386 and x86-64 Options'' in the main manual
1222 @item --enable-win32-registry
1223 @itemx --enable-win32-registry=@var{key}
1224 @itemx --disable-win32-registry
1225 The @option{--enable-win32-registry} option enables Microsoft Windows-hosted GCC
1226 to look up installations paths in the registry using the following key:
1229 @code{HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\@var{key}}
1232 @var{key} defaults to GCC version number, and can be overridden by the
1233 @option{--enable-win32-registry=@var{key}} option. Vendors and distributors
1234 who use custom installers are encouraged to provide a different key,
1235 perhaps one comprised of vendor name and GCC version number, to
1236 avoid conflict with existing installations. This feature is enabled
1237 by default, and can be disabled by @option{--disable-win32-registry}
1238 option. This option has no effect on the other hosts.
1241 Specify that the machine does not have a floating point unit. This
1242 option only applies to @samp{m68k-sun-sunos@var{n}}. On any other
1243 system, @option{--nfp} has no effect.
1245 @item --enable-werror
1246 @itemx --disable-werror
1247 @itemx --enable-werror=yes
1248 @itemx --enable-werror=no
1249 When you specify this option, it controls whether certain files in the
1250 compiler are built with @option{-Werror} in bootstrap stage2 and later.
1251 If you don't specify it, @option{-Werror} is turned on for the main
1252 development trunk. However it defaults to off for release branches and
1253 final releases. The specific files which get @option{-Werror} are
1254 controlled by the Makefiles.
1256 @item --enable-checking
1257 @itemx --enable-checking=@var{list}
1258 When you specify this option, the compiler is built to perform internal
1259 consistency checks of the requested complexity. This does not change the
1260 generated code, but adds error checking within the compiler. This will
1261 slow down the compiler and may only work properly if you are building
1262 the compiler with GCC@. This is @samp{yes} by default when building
1263 from SVN or snapshots, but @samp{release} for releases. More control
1264 over the checks may be had by specifying @var{list}. The categories of
1265 checks available are @samp{yes} (most common checks
1266 @samp{assert,misc,tree,gc,rtlflag,runtime}), @samp{no} (no checks at
1267 all), @samp{all} (all but @samp{valgrind}), @samp{release} (cheapest
1268 checks @samp{assert,runtime}) or @samp{none} (same as @samp{no}).
1269 Individual checks can be enabled with these flags @samp{assert},
1270 @samp{df}, @samp{fold}, @samp{gc}, @samp{gcac} @samp{misc}, @samp{rtl},
1271 @samp{rtlflag}, @samp{runtime}, @samp{tree}, and @samp{valgrind}.
1273 The @samp{valgrind} check requires the external @command{valgrind}
1274 simulator, available from @uref{http://valgrind.org/}. The
1275 @samp{df}, @samp{rtl}, @samp{gcac} and @samp{valgrind} checks are very expensive.
1276 To disable all checking, @samp{--disable-checking} or
1277 @samp{--enable-checking=none} must be explicitly requested. Disabling
1278 assertions will make the compiler and runtime slightly faster but
1279 increase the risk of undetected internal errors causing wrong code to be
1282 @item --enable-coverage
1283 @itemx --enable-coverage=@var{level}
1284 With this option, the compiler is built to collect self coverage
1285 information, every time it is run. This is for internal development
1286 purposes, and only works when the compiler is being built with gcc. The
1287 @var{level} argument controls whether the compiler is built optimized or
1288 not, values are @samp{opt} and @samp{noopt}. For coverage analysis you
1289 want to disable optimization, for performance analysis you want to
1290 enable optimization. When coverage is enabled, the default level is
1291 without optimization.
1293 @item --enable-gather-detailed-mem-stats
1294 When this option is specified more detailed information on memory
1295 allocation is gathered. This information is printed when using
1296 @option{-fmem-report}.
1299 @itemx --with-gc=@var{choice}
1300 With this option you can specify the garbage collector implementation
1301 used during the compilation process. @var{choice} can be one of
1302 @samp{page} and @samp{zone}, where @samp{page} is the default.
1305 @itemx --disable-nls
1306 The @option{--enable-nls} option enables Native Language Support (NLS),
1307 which lets GCC output diagnostics in languages other than American
1308 English. Native Language Support is enabled by default if not doing a
1309 canadian cross build. The @option{--disable-nls} option disables NLS@.
1311 @item --with-included-gettext
1312 If NLS is enabled, the @option{--with-included-gettext} option causes the build
1313 procedure to prefer its copy of GNU @command{gettext}.
1315 @item --with-catgets
1316 If NLS is enabled, and if the host lacks @code{gettext} but has the
1317 inferior @code{catgets} interface, the GCC build procedure normally
1318 ignores @code{catgets} and instead uses GCC's copy of the GNU
1319 @code{gettext} library. The @option{--with-catgets} option causes the
1320 build procedure to use the host's @code{catgets} in this situation.
1322 @item --with-libiconv-prefix=@var{dir}
1323 Search for libiconv header files in @file{@var{dir}/include} and
1324 libiconv library files in @file{@var{dir}/lib}.
1326 @item --enable-obsolete
1327 Enable configuration for an obsoleted system. If you attempt to
1328 configure GCC for a system (build, host, or target) which has been
1329 obsoleted, and you do not specify this flag, configure will halt with an
1332 All support for systems which have been obsoleted in one release of GCC
1333 is removed entirely in the next major release, unless someone steps
1334 forward to maintain the port.
1336 @item --enable-decimal-float
1337 @itemx --enable-decimal-float=yes
1338 @itemx --enable-decimal-float=no
1339 @itemx --enable-decimal-float=bid
1340 @itemx --enable-decimal-float=dpd
1341 @itemx --disable-decimal-float
1342 Enable (or disable) support for the C decimal floating point extension
1343 that is in the IEEE 754R extension to the IEEE754 floating point
1344 standard. This is enabled by default only on PowerPC, i386, and
1345 x86_64 GNU/Linux systems. Other systems may also support it, but
1346 require the user to specifically enable it. You can optionally
1347 control which decimal floating point format is used (either @samp{bid}
1348 or @samp{dpd}). The @samp{bid} (binary integer decimal) format is
1349 default on i386 and x86_64 systems, and the @samp{dpd} (densely packed
1350 decimal) format is default on PowerPC systems.
1352 @item --enable-fixed-point
1353 @itemx --disable-fixed-point
1354 Enable (or disable) support for C fixed-point arithmetic.
1355 This option is enabled by default for some targets (such as MIPS) which
1356 have hardware-support for fixed-point operations. On other targets, you
1357 may enable this option manually.
1359 @item --with-long-double-128
1360 Specify if @code{long double} type should be 128-bit by default on selected
1361 GNU/Linux architectures. If using @code{--without-long-double-128},
1362 @code{long double} will be by default 64-bit, the same as @code{double} type.
1363 When neither of these configure options are used, the default will be
1364 128-bit @code{long double} when built against GNU C Library 2.4 and later,
1365 64-bit @code{long double} otherwise.
1367 @item --with-gmp=@var{pathname}
1368 @itemx --with-gmp-include=@var{pathname}
1369 @itemx --with-gmp-lib=@var{pathname}
1370 @itemx --with-mpfr=@var{pathname}
1371 @itemx --with-mpfr-include=@var{pathname}
1372 @itemx --with-mpfr-lib=@var{pathname}
1373 If you do not have GMP (the GNU Multiple Precision library) and the
1374 MPFR Libraries installed in a standard location and you want to build
1375 GCC, you can explicitly specify the directory where they are installed
1376 (@samp{--with-gmp=@var{gmpinstalldir}},
1377 @samp{--with-mpfr=@var{mpfrinstalldir}}). The
1378 @option{--with-gmp=@var{gmpinstalldir}} option is shorthand for
1379 @option{--with-gmp-lib=@var{gmpinstalldir}/lib} and
1380 @option{--with-gmp-include=@var{gmpinstalldir}/include}. Likewise the
1381 @option{--with-mpfr=@var{mpfrinstalldir}} option is shorthand for
1382 @option{--with-mpfr-lib=@var{mpfrinstalldir}/lib} and
1383 @option{--with-mpfr-include=@var{mpfrinstalldir}/include}. If these
1384 shorthand assumptions are not correct, you can use the explicit
1385 include and lib options directly.
1387 @item --with-debug-prefix-map=@var{map}
1388 Convert source directory names using @option{-fdebug-prefix-map} when
1389 building runtime libraries. @samp{@var{map}} is a space-separated
1390 list of maps of the form @samp{@var{old}=@var{new}}.
1394 @subheading Cross-Compiler-Specific Options
1395 The following options only apply to building cross compilers.
1397 @item --with-sysroot
1398 @itemx --with-sysroot=@var{dir}
1399 Tells GCC to consider @var{dir} as the root of a tree that contains a
1400 (subset of) the root filesystem of the target operating system.
1401 Target system headers, libraries and run-time object files will be
1402 searched in there. The specified directory is not copied into the
1403 install tree, unlike the options @option{--with-headers} and
1404 @option{--with-libs} that this option obsoletes. The default value,
1405 in case @option{--with-sysroot} is not given an argument, is
1406 @option{$@{gcc_tooldir@}/sys-root}. If the specified directory is a
1407 subdirectory of @option{$@{exec_prefix@}}, then it will be found relative to
1408 the GCC binaries if the installation tree is moved.
1410 @item --with-build-sysroot
1411 @itemx --with-build-sysroot=@var{dir}
1412 Tells GCC to consider @var{dir} as the system root (see
1413 @option{--with-sysroot}) while building target libraries, instead of
1414 the directory specified with @option{--with-sysroot}. This option is
1415 only useful when you are already using @option{--with-sysroot}. You
1416 can use @option{--with-build-sysroot} when you are configuring with
1417 @option{--prefix} set to a directory that is different from the one in
1418 which you are installing GCC and your target libraries.
1420 This option affects the system root for the compiler used to build
1421 target libraries (which runs on the build system); it does not affect
1422 the compiler which is used to build GCC itself.
1424 @item --with-headers
1425 @itemx --with-headers=@var{dir}
1426 Deprecated in favor of @option{--with-sysroot}.
1427 Specifies that target headers are available when building a cross compiler.
1428 The @var{dir} argument specifies a directory which has the target include
1429 files. These include files will be copied into the @file{gcc} install
1430 directory. @emph{This option with the @var{dir} argument is required} when
1431 building a cross compiler, if @file{@var{prefix}/@var{target}/sys-include}
1432 doesn't pre-exist. If @file{@var{prefix}/@var{target}/sys-include} does
1433 pre-exist, the @var{dir} argument may be omitted. @command{fixincludes}
1434 will be run on these files to make them compatible with GCC@.
1436 @item --without-headers
1437 Tells GCC not use any target headers from a libc when building a cross
1438 compiler. When crossing to GNU/Linux, you need the headers so GCC
1439 can build the exception handling for libgcc.
1442 @itemx --with-libs=``@var{dir1} @var{dir2} @dots{} @var{dirN}''
1443 Deprecated in favor of @option{--with-sysroot}.
1444 Specifies a list of directories which contain the target runtime
1445 libraries. These libraries will be copied into the @file{gcc} install
1446 directory. If the directory list is omitted, this option has no
1450 Specifies that @samp{newlib} is
1451 being used as the target C library. This causes @code{__eprintf} to be
1452 omitted from @file{libgcc.a} on the assumption that it will be provided by
1455 @item --with-build-time-tools=@var{dir}
1456 Specifies where to find the set of target tools (assembler, linker, etc.)
1457 that will be used while building GCC itself. This option can be useful
1458 if the directory layouts are different between the system you are building
1459 GCC on, and the system where you will deploy it.
1461 For example, on a @option{ia64-hp-hpux} system, you may have the GNU
1462 assembler and linker in @file{/usr/bin}, and the native tools in a
1463 different path, and build a toolchain that expects to find the
1464 native tools in @file{/usr/bin}.
1466 When you use this option, you should ensure that @var{dir} includes
1467 @command{ar}, @command{as}, @command{ld}, @command{nm},
1468 @command{ranlib} and @command{strip} if necessary, and possibly
1469 @command{objdump}. Otherwise, GCC may use an inconsistent set of
1473 @subheading Java-Specific Options
1475 The following option applies to the build of the Java front end.
1478 @item --disable-libgcj
1479 Specify that the run-time libraries
1480 used by GCJ should not be built. This is useful in case you intend
1481 to use GCJ with some other run-time, or you're going to install it
1482 separately, or it just happens not to build on your particular
1483 machine. In general, if the Java front end is enabled, the GCJ
1484 libraries will be enabled too, unless they're known to not work on
1485 the target platform. If GCJ is enabled but @samp{libgcj} isn't built, you
1486 may need to port it; in this case, before modifying the top-level
1487 @file{configure.in} so that @samp{libgcj} is enabled by default on this platform,
1488 you may use @option{--enable-libgcj} to override the default.
1492 The following options apply to building @samp{libgcj}.
1494 @subsubheading General Options
1497 @item --enable-java-maintainer-mode
1498 By default the @samp{libjava} build will not attempt to compile the
1499 @file{.java} source files to @file{.class}. Instead, it will use the
1500 @file{.class} files from the source tree. If you use this option you
1501 must have executables named @command{ecj1} and @command{gjavah} in your path
1502 for use by the build. You must use this option if you intend to
1503 modify any @file{.java} files in @file{libjava}.
1505 @item --with-java-home=@var{dirname}
1506 This @samp{libjava} option overrides the default value of the
1507 @samp{java.home} system property. It is also used to set
1508 @samp{sun.boot.class.path} to @file{@var{dirname}/lib/rt.jar}. By
1509 default @samp{java.home} is set to @file{@var{prefix}} and
1510 @samp{sun.boot.class.path} to
1511 @file{@var{datadir}/java/libgcj-@var{version}.jar}.
1513 @item --with-ecj-jar=@var{filename}
1514 This option can be used to specify the location of an external jar
1515 file containing the Eclipse Java compiler. A specially modified
1516 version of this compiler is used by @command{gcj} to parse
1517 @file{.java} source files. If this option is given, the
1518 @samp{libjava} build will create and install an @file{ecj1} executable
1519 which uses this jar file at runtime.
1521 If this option is not given, but an @file{ecj.jar} file is found in
1522 the topmost source tree at configure time, then the @samp{libgcj}
1523 build will create and install @file{ecj1}, and will also install the
1524 discovered @file{ecj.jar} into a suitable place in the install tree.
1526 If @file{ecj1} is not installed, then the user will have to supply one
1527 on his path in order for @command{gcj} to properly parse @file{.java}
1528 source files. A suitable jar is available from
1529 @uref{ftp://sourceware.org/pub/java/}.
1531 @item --disable-getenv-properties
1532 Don't set system properties from @env{GCJ_PROPERTIES}.
1534 @item --enable-hash-synchronization
1535 Use a global hash table for monitor locks. Ordinarily,
1536 @samp{libgcj}'s @samp{configure} script automatically makes
1537 the correct choice for this option for your platform. Only use
1538 this if you know you need the library to be configured differently.
1540 @item --enable-interpreter
1541 Enable the Java interpreter. The interpreter is automatically
1542 enabled by default on all platforms that support it. This option
1543 is really only useful if you want to disable the interpreter
1544 (using @option{--disable-interpreter}).
1546 @item --disable-java-net
1547 Disable java.net. This disables the native part of java.net only,
1548 using non-functional stubs for native method implementations.
1550 @item --disable-jvmpi
1551 Disable JVMPI support.
1553 @item --disable-libgcj-bc
1554 Disable BC ABI compilation of certain parts of libgcj. By default,
1555 some portions of libgcj are compiled with @option{-findirect-dispatch}
1556 and @option{-fno-indirect-classes}, allowing them to be overridden at
1559 If @option{--disable-libgcj-bc} is specified, libgcj is built without
1560 these options. This allows the compile-time linker to resolve
1561 dependencies when statically linking to libgcj. However it makes it
1562 impossible to override the affected portions of libgcj at run-time.
1565 Enable runtime eCos target support.
1567 @item --without-libffi
1568 Don't use @samp{libffi}. This will disable the interpreter and JNI
1569 support as well, as these require @samp{libffi} to work.
1571 @item --enable-libgcj-debug
1572 Enable runtime debugging code.
1574 @item --enable-libgcj-multifile
1575 If specified, causes all @file{.java} source files to be
1576 compiled into @file{.class} files in one invocation of
1577 @samp{gcj}. This can speed up build time, but is more
1578 resource-intensive. If this option is unspecified or
1579 disabled, @samp{gcj} is invoked once for each @file{.java}
1580 file to compile into a @file{.class} file.
1582 @item --with-libiconv-prefix=DIR
1583 Search for libiconv in @file{DIR/include} and @file{DIR/lib}.
1585 @item --enable-sjlj-exceptions
1586 Force use of the @code{setjmp}/@code{longjmp}-based scheme for exceptions.
1587 @samp{configure} ordinarily picks the correct value based on the platform.
1588 Only use this option if you are sure you need a different setting.
1590 @item --with-system-zlib
1591 Use installed @samp{zlib} rather than that included with GCC@.
1593 @item --with-win32-nlsapi=ansi, unicows or unicode
1594 Indicates how MinGW @samp{libgcj} translates between UNICODE
1595 characters and the Win32 API@.
1598 Use the single-byte @code{char} and the Win32 A functions natively,
1599 translating to and from UNICODE when using these functions. If
1600 unspecified, this is the default.
1603 Use the @code{WCHAR} and Win32 W functions natively. Adds
1604 @code{-lunicows} to @file{libgcj.spec} to link with @samp{libunicows}.
1605 @file{unicows.dll} needs to be deployed on Microsoft Windows 9X machines
1606 running built executables. @file{libunicows.a}, an open-source
1607 import library around Microsoft's @code{unicows.dll}, is obtained from
1608 @uref{http://libunicows.sourceforge.net/}, which also gives details
1609 on getting @file{unicows.dll} from Microsoft.
1612 Use the @code{WCHAR} and Win32 W functions natively. Does @emph{not}
1613 add @code{-lunicows} to @file{libgcj.spec}. The built executables will
1614 only run on Microsoft Windows NT and above.
1618 @subsubheading AWT-Specific Options
1622 Use the X Window System.
1624 @item --enable-java-awt=PEER(S)
1625 Specifies the AWT peer library or libraries to build alongside
1626 @samp{libgcj}. If this option is unspecified or disabled, AWT
1627 will be non-functional. Current valid values are @option{gtk} and
1628 @option{xlib}. Multiple libraries should be separated by a
1629 comma (i.e.@: @option{--enable-java-awt=gtk,xlib}).
1631 @item --enable-gtk-cairo
1632 Build the cairo Graphics2D implementation on GTK@.
1634 @item --enable-java-gc=TYPE
1635 Choose garbage collector. Defaults to @option{boehm} if unspecified.
1637 @item --disable-gtktest
1638 Do not try to compile and run a test GTK+ program.
1640 @item --disable-glibtest
1641 Do not try to compile and run a test GLIB program.
1643 @item --with-libart-prefix=PFX
1644 Prefix where libart is installed (optional).
1646 @item --with-libart-exec-prefix=PFX
1647 Exec prefix where libart is installed (optional).
1649 @item --disable-libarttest
1650 Do not try to compile and run a test libart program.
1659 @uref{./index.html,,Return to the GCC Installation page}
1663 @c ***Building****************************************************************
1665 @comment node-name, next, previous, up
1666 @node Building, Testing, Configuration, Installing GCC
1672 @cindex Installing GCC: Building
1674 Now that GCC is configured, you are ready to build the compiler and
1677 Some commands executed when making the compiler may fail (return a
1678 nonzero status) and be ignored by @command{make}. These failures, which
1679 are often due to files that were not found, are expected, and can safely
1682 It is normal to have compiler warnings when compiling certain files.
1683 Unless you are a GCC developer, you can generally ignore these warnings
1684 unless they cause compilation to fail. Developers should attempt to fix
1685 any warnings encountered, however they can temporarily continue past
1686 warnings-as-errors by specifying the configure flag
1687 @option{--disable-werror}.
1689 On certain old systems, defining certain environment variables such as
1690 @env{CC} can interfere with the functioning of @command{make}.
1692 If you encounter seemingly strange errors when trying to build the
1693 compiler in a directory other than the source directory, it could be
1694 because you have previously configured the compiler in the source
1695 directory. Make sure you have done all the necessary preparations.
1697 If you build GCC on a BSD system using a directory stored in an old System
1698 V file system, problems may occur in running @command{fixincludes} if the
1699 System V file system doesn't support symbolic links. These problems
1700 result in a failure to fix the declaration of @code{size_t} in
1701 @file{sys/types.h}. If you find that @code{size_t} is a signed type and
1702 that type mismatches occur, this could be the cause.
1704 The solution is not to use such a directory for building GCC@.
1706 Similarly, when building from SVN or snapshots, or if you modify
1707 @file{*.l} files, you need the Flex lexical analyzer generator
1708 installed. If you do not modify @file{*.l} files, releases contain
1709 the Flex-generated files and you do not need Flex installed to build
1710 them. There is still one Flex-based lexical analyzer (part of the
1711 build machinery, not of GCC itself) that is used even if you only
1712 build the C front end.
1714 When building from SVN or snapshots, or if you modify Texinfo
1715 documentation, you need version 4.4 or later of Texinfo installed if you
1716 want Info documentation to be regenerated. Releases contain Info
1717 documentation pre-built for the unmodified documentation in the release.
1719 @section Building a native compiler
1721 For a native build, the default configuration is to perform
1722 a 3-stage bootstrap of the compiler when @samp{make} is invoked.
1723 This will build the entire GCC system and ensure that it compiles
1724 itself correctly. It can be disabled with the @option{--disable-bootstrap}
1725 parameter to @samp{configure}, but bootstrapping is suggested because
1726 the compiler will be tested more completely and could also have
1729 The bootstrapping process will complete the following steps:
1733 Build tools necessary to build the compiler.
1736 Perform a 3-stage bootstrap of the compiler. This includes building
1737 three times the target tools for use by the compiler such as binutils
1738 (bfd, binutils, gas, gprof, ld, and opcodes) if they have been
1739 individually linked or moved into the top level GCC source tree before
1743 Perform a comparison test of the stage2 and stage3 compilers.
1746 Build runtime libraries using the stage3 compiler from the previous step.
1750 If you are short on disk space you might consider @samp{make
1751 bootstrap-lean} instead. The sequence of compilation is the
1752 same described above, but object files from the stage1 and
1753 stage2 of the 3-stage bootstrap of the compiler are deleted as
1754 soon as they are no longer needed.
1756 If you wish to use non-default GCC flags when compiling the stage2
1757 and stage3 compilers, set @code{BOOT_CFLAGS} on the command line when
1758 doing @samp{make}. For example, if you want to save additional space
1759 during the bootstrap and in the final installation as well, you can
1760 build the compiler binaries without debugging information as in the
1761 following example. This will save roughly 40% of disk space both for
1762 the bootstrap and the final installation. (Libraries will still contain
1763 debugging information.)
1766 make BOOT_CFLAGS='-O' bootstrap
1769 You can place non-default optimization flags into @code{BOOT_CFLAGS}; they
1770 are less well tested here than the default of @samp{-g -O2}, but should
1771 still work. In a few cases, you may find that you need to specify special
1772 flags such as @option{-msoft-float} here to complete the bootstrap; or,
1773 if the native compiler miscompiles the stage1 compiler, you may need
1774 to work around this, by choosing @code{BOOT_CFLAGS} to avoid the parts
1775 of the stage1 compiler that were miscompiled, or by using @samp{make
1776 bootstrap4} to increase the number of stages of bootstrap.
1778 @code{BOOT_CFLAGS} does not apply to bootstrapped target libraries.
1779 Since these are always compiled with the compiler currently being
1780 bootstrapped, you can use @code{CFLAGS_FOR_TARGET} to modify their
1781 compilation flags, as for non-bootstrapped target libraries.
1782 Again, if the native compiler miscompiles the stage1 compiler, you may
1783 need to work around this by avoiding non-working parts of the stage1
1784 compiler. Use @code{STAGE1_LIBCFLAGS} to this end.
1786 If you used the flag @option{--enable-languages=@dots{}} to restrict
1787 the compilers to be built, only those you've actually enabled will be
1788 built. This will of course only build those runtime libraries, for
1789 which the particular compiler has been built. Please note,
1790 that re-defining @env{LANGUAGES} when calling @samp{make}
1791 @strong{does not} work anymore!
1793 If the comparison of stage2 and stage3 fails, this normally indicates
1794 that the stage2 compiler has compiled GCC incorrectly, and is therefore
1795 a potentially serious bug which you should investigate and report. (On
1796 a few systems, meaningful comparison of object files is impossible; they
1797 always appear ``different''. If you encounter this problem, you will
1798 need to disable comparison in the @file{Makefile}.)
1800 If you do not want to bootstrap your compiler, you can configure with
1801 @option{--disable-bootstrap}. In particular cases, you may want to
1802 bootstrap your compiler even if the target system is not the same as
1803 the one you are building on: for example, you could build a
1804 @code{powerpc-unknown-linux-gnu} toolchain on a
1805 @code{powerpc64-unknown-linux-gnu} host. In this case, pass
1806 @option{--enable-bootstrap} to the configure script.
1809 @section Building a cross compiler
1811 When building a cross compiler, it is not generally possible to do a
1812 3-stage bootstrap of the compiler. This makes for an interesting problem
1813 as parts of GCC can only be built with GCC@.
1815 To build a cross compiler, we first recommend building and installing a
1816 native compiler. You can then use the native GCC compiler to build the
1817 cross compiler. The installed native compiler needs to be GCC version
1820 If the cross compiler is to be built with support for the Java
1821 programming language and the ability to compile .java source files is
1822 desired, the installed native compiler used to build the cross
1823 compiler needs to be the same GCC version as the cross compiler. In
1824 addition the cross compiler needs to be configured with
1825 @option{--with-ecj-jar=@dots{}}.
1827 Assuming you have already installed a native copy of GCC and configured
1828 your cross compiler, issue the command @command{make}, which performs the
1833 Build host tools necessary to build the compiler.
1836 Build target tools for use by the compiler such as binutils (bfd,
1837 binutils, gas, gprof, ld, and opcodes)
1838 if they have been individually linked or moved into the top level GCC source
1839 tree before configuring.
1842 Build the compiler (single stage only).
1845 Build runtime libraries using the compiler from the previous step.
1848 Note that if an error occurs in any step the make process will exit.
1850 If you are not building GNU binutils in the same source tree as GCC,
1851 you will need a cross-assembler and cross-linker installed before
1852 configuring GCC@. Put them in the directory
1853 @file{@var{prefix}/@var{target}/bin}. Here is a table of the tools
1854 you should put in this directory:
1858 This should be the cross-assembler.
1861 This should be the cross-linker.
1864 This should be the cross-archiver: a program which can manipulate
1865 archive files (linker libraries) in the target machine's format.
1868 This should be a program to construct a symbol table in an archive file.
1871 The installation of GCC will find these programs in that directory,
1872 and copy or link them to the proper place to for the cross-compiler to
1873 find them when run later.
1875 The easiest way to provide these files is to build the Binutils package.
1876 Configure it with the same @option{--host} and @option{--target}
1877 options that you use for configuring GCC, then build and install
1878 them. They install their executables automatically into the proper
1879 directory. Alas, they do not support all the targets that GCC
1882 If you are not building a C library in the same source tree as GCC,
1883 you should also provide the target libraries and headers before
1884 configuring GCC, specifying the directories with
1885 @option{--with-sysroot} or @option{--with-headers} and
1886 @option{--with-libs}. Many targets also require ``start files'' such
1887 as @file{crt0.o} and
1888 @file{crtn.o} which are linked into each executable. There may be several
1889 alternatives for @file{crt0.o}, for use with profiling or other
1890 compilation options. Check your target's definition of
1891 @code{STARTFILE_SPEC} to find out what start files it uses.
1893 @section Building in parallel
1895 GNU Make 3.79 and above, which is necessary to build GCC, support
1896 building in parallel. To activate this, you can use @samp{make -j 2}
1897 instead of @samp{make}. You can also specify a bigger number, and
1898 in most cases using a value greater than the number of processors in
1899 your machine will result in fewer and shorter I/O latency hits, thus
1900 improving overall throughput; this is especially true for slow drives
1901 and network filesystems.
1903 @section Building the Ada compiler
1905 In order to build GNAT, the Ada compiler, you need a working GNAT
1906 compiler (GNAT version 3.14 or later, or GCC version 3.1 or later).
1907 This includes GNAT tools such as @command{gnatmake} and
1908 @command{gnatlink}, since the Ada front end is written in Ada and
1909 uses some GNAT-specific extensions.
1911 In order to build a cross compiler, it is suggested to install
1912 the new compiler as native first, and then use it to build the cross
1915 @command{configure} does not test whether the GNAT installation works
1916 and has a sufficiently recent version; if too old a GNAT version is
1917 installed, the build will fail unless @option{--enable-languages} is
1918 used to disable building the Ada front end.
1920 @env{ADA_INCLUDE_PATH} and @env{ADA_OBJECT_PATH} environment variables
1921 must not be set when building the Ada compiler, the Ada tools, or the
1922 Ada runtime libraries. You can check that your build environment is clean
1923 by verifying that @samp{gnatls -v} lists only one explicit path in each
1926 @section Building with profile feedback
1928 It is possible to use profile feedback to optimize the compiler itself. This
1929 should result in a faster compiler binary. Experiments done on x86 using gcc
1930 3.3 showed approximately 7 percent speedup on compiling C programs. To
1931 bootstrap the compiler with profile feedback, use @code{make profiledbootstrap}.
1933 When @samp{make profiledbootstrap} is run, it will first build a @code{stage1}
1934 compiler. This compiler is used to build a @code{stageprofile} compiler
1935 instrumented to collect execution counts of instruction and branch
1936 probabilities. Then runtime libraries are compiled with profile collected.
1937 Finally a @code{stagefeedback} compiler is built using the information collected.
1939 Unlike standard bootstrap, several additional restrictions apply. The
1940 compiler used to build @code{stage1} needs to support a 64-bit integral type.
1941 It is recommended to only use GCC for this. Also parallel make is currently
1942 not supported since collisions in profile collecting may occur.
1949 @uref{./index.html,,Return to the GCC Installation page}
1953 @c ***Testing*****************************************************************
1955 @comment node-name, next, previous, up
1956 @node Testing, Final install, Building, Installing GCC
1960 @chapter Installing GCC: Testing
1963 @cindex Installing GCC: Testing
1966 Before you install GCC, we encourage you to run the testsuites and to
1967 compare your results with results from a similar configuration that have
1968 been submitted to the
1969 @uref{http://gcc.gnu.org/ml/gcc-testresults/,,gcc-testresults mailing list}.
1970 Some of these archived results are linked from the build status lists
1971 at @uref{http://gcc.gnu.org/buildstat.html}, although not everyone who
1972 reports a successful build runs the testsuites and submits the results.
1973 This step is optional and may require you to download additional software,
1974 but it can give you confidence in your new GCC installation or point out
1975 problems before you install and start using your new GCC@.
1977 First, you must have @uref{download.html,,downloaded the testsuites}.
1978 These are part of the full distribution, but if you downloaded the
1979 ``core'' compiler plus any front ends, you must download the testsuites
1982 Second, you must have the testing tools installed. This includes
1983 @uref{http://www.gnu.org/software/dejagnu/,,DejaGnu}, Tcl, and Expect;
1984 the DejaGnu site has links to these.
1986 If the directories where @command{runtest} and @command{expect} were
1987 installed are not in the @env{PATH}, you may need to set the following
1988 environment variables appropriately, as in the following example (which
1989 assumes that DejaGnu has been installed under @file{/usr/local}):
1992 TCL_LIBRARY = /usr/local/share/tcl8.0
1993 DEJAGNULIBS = /usr/local/share/dejagnu
1996 (On systems such as Cygwin, these paths are required to be actual
1997 paths, not mounts or links; presumably this is due to some lack of
1998 portability in the DejaGnu code.)
2001 Finally, you can run the testsuite (which may take a long time):
2003 cd @var{objdir}; make -k check
2006 This will test various components of GCC, such as compiler
2007 front ends and runtime libraries. While running the testsuite, DejaGnu
2008 might emit some harmless messages resembling
2009 @samp{WARNING: Couldn't find the global config file.} or
2010 @samp{WARNING: Couldn't find tool init file} that can be ignored.
2012 If you are testing a cross-compiler, you may want to run the testsuite
2013 on a simulator as described at @uref{http://gcc.gnu.org/simtest-howto.html}.
2015 @section How can you run the testsuite on selected tests?
2017 In order to run sets of tests selectively, there are targets
2018 @samp{make check-gcc} and @samp{make check-g++}
2019 in the @file{gcc} subdirectory of the object directory. You can also
2020 just run @samp{make check} in a subdirectory of the object directory.
2023 A more selective way to just run all @command{gcc} execute tests in the
2027 make check-gcc RUNTESTFLAGS="execute.exp @var{other-options}"
2030 Likewise, in order to run only the @command{g++} ``old-deja'' tests in
2031 the testsuite with filenames matching @samp{9805*}, you would use
2034 make check-g++ RUNTESTFLAGS="old-deja.exp=9805* @var{other-options}"
2037 The @file{*.exp} files are located in the testsuite directories of the GCC
2038 source, the most important ones being @file{compile.exp},
2039 @file{execute.exp}, @file{dg.exp} and @file{old-deja.exp}.
2040 To get a list of the possible @file{*.exp} files, pipe the
2041 output of @samp{make check} into a file and look at the
2042 @samp{Running @dots{} .exp} lines.
2044 @section Passing options and running multiple testsuites
2046 You can pass multiple options to the testsuite using the
2047 @samp{--target_board} option of DejaGNU, either passed as part of
2048 @samp{RUNTESTFLAGS}, or directly to @command{runtest} if you prefer to
2049 work outside the makefiles. For example,
2052 make check-g++ RUNTESTFLAGS="--target_board=unix/-O3/-fmerge-constants"
2055 will run the standard @command{g++} testsuites (``unix'' is the target name
2056 for a standard native testsuite situation), passing
2057 @samp{-O3 -fmerge-constants} to the compiler on every test, i.e.,
2058 slashes separate options.
2060 You can run the testsuites multiple times using combinations of options
2061 with a syntax similar to the brace expansion of popular shells:
2064 @dots{}"--target_board=arm-sim\@{-mhard-float,-msoft-float\@}\@{-O1,-O2,-O3,\@}"
2067 (Note the empty option caused by the trailing comma in the final group.)
2068 The following will run each testsuite eight times using the @samp{arm-sim}
2069 target, as if you had specified all possible combinations yourself:
2072 --target_board=arm-sim/-mhard-float/-O1
2073 --target_board=arm-sim/-mhard-float/-O2
2074 --target_board=arm-sim/-mhard-float/-O3
2075 --target_board=arm-sim/-mhard-float
2076 --target_board=arm-sim/-msoft-float/-O1
2077 --target_board=arm-sim/-msoft-float/-O2
2078 --target_board=arm-sim/-msoft-float/-O3
2079 --target_board=arm-sim/-msoft-float
2082 They can be combined as many times as you wish, in arbitrary ways. This
2086 @dots{}"--target_board=unix/-Wextra\@{-O3,-fno-strength\@}\@{-fomit-frame,\@}"
2089 will generate four combinations, all involving @samp{-Wextra}.
2091 The disadvantage to this method is that the testsuites are run in serial,
2092 which is a waste on multiprocessor systems. For users with GNU Make and
2093 a shell which performs brace expansion, you can run the testsuites in
2094 parallel by having the shell perform the combinations and @command{make}
2095 do the parallel runs. Instead of using @samp{--target_board}, use a
2096 special makefile target:
2099 make -j@var{N} check-@var{testsuite}//@var{test-target}/@var{option1}/@var{option2}/@dots{}
2105 make -j3 check-gcc//sh-hms-sim/@{-m1,-m2,-m3,-m3e,-m4@}/@{,-nofpu@}
2108 will run three concurrent ``make-gcc'' testsuites, eventually testing all
2109 ten combinations as described above. Note that this is currently only
2110 supported in the @file{gcc} subdirectory. (To see how this works, try
2111 typing @command{echo} before the example given here.)
2114 @section Additional testing for Java Class Libraries
2116 The Java runtime tests can be executed via @samp{make check}
2117 in the @file{@var{target}/libjava/testsuite} directory in
2120 The @uref{http://sourceware.org/mauve/,,Mauve Project} provides
2121 a suite of tests for the Java Class Libraries. This suite can be run
2122 as part of libgcj testing by placing the Mauve tree within the libjava
2123 testsuite at @file{libjava/testsuite/libjava.mauve/mauve}, or by
2124 specifying the location of that tree when invoking @samp{make}, as in
2125 @samp{make MAUVEDIR=~/mauve check}.
2127 @section How to interpret test results
2129 The result of running the testsuite are various @file{*.sum} and @file{*.log}
2130 files in the testsuite subdirectories. The @file{*.log} files contain a
2131 detailed log of the compiler invocations and the corresponding
2132 results, the @file{*.sum} files summarize the results. These summaries
2133 contain status codes for all tests:
2137 PASS: the test passed as expected
2139 XPASS: the test unexpectedly passed
2141 FAIL: the test unexpectedly failed
2143 XFAIL: the test failed as expected
2145 UNSUPPORTED: the test is not supported on this platform
2147 ERROR: the testsuite detected an error
2149 WARNING: the testsuite detected a possible problem
2152 It is normal for some tests to report unexpected failures. At the
2153 current time the testing harness does not allow fine grained control
2154 over whether or not a test is expected to fail. This problem should
2155 be fixed in future releases.
2158 @section Submitting test results
2160 If you want to report the results to the GCC project, use the
2161 @file{contrib/test_summary} shell script. Start it in the @var{objdir} with
2164 @var{srcdir}/contrib/test_summary -p your_commentary.txt \
2165 -m gcc-testresults@@gcc.gnu.org |sh
2168 This script uses the @command{Mail} program to send the results, so
2169 make sure it is in your @env{PATH}. The file @file{your_commentary.txt} is
2170 prepended to the testsuite summary and should contain any special
2171 remarks you have on your results or your build environment. Please
2172 do not edit the testsuite result block or the subject line, as these
2173 messages may be automatically processed.
2180 @uref{./index.html,,Return to the GCC Installation page}
2184 @c ***Final install***********************************************************
2186 @comment node-name, next, previous, up
2187 @node Final install, , Testing, Installing GCC
2189 @ifset finalinstallhtml
2191 @chapter Installing GCC: Final installation
2194 Now that GCC has been built (and optionally tested), you can install it with
2196 cd @var{objdir}; make install
2199 We strongly recommend to install into a target directory where there is
2200 no previous version of GCC present. Also, the GNAT runtime should not
2201 be stripped, as this would break certain features of the debugger that
2202 depend on this debugging information (catching Ada exceptions for
2205 That step completes the installation of GCC; user level binaries can
2206 be found in @file{@var{prefix}/bin} where @var{prefix} is the value
2207 you specified with the @option{--prefix} to configure (or
2208 @file{/usr/local} by default). (If you specified @option{--bindir},
2209 that directory will be used instead; otherwise, if you specified
2210 @option{--exec-prefix}, @file{@var{exec-prefix}/bin} will be used.)
2211 Headers for the C++ and Java libraries are installed in
2212 @file{@var{prefix}/include}; libraries in @file{@var{libdir}}
2213 (normally @file{@var{prefix}/lib}); internal parts of the compiler in
2214 @file{@var{libdir}/gcc} and @file{@var{libexecdir}/gcc}; documentation
2215 in info format in @file{@var{infodir}} (normally
2216 @file{@var{prefix}/info}).
2218 When installing cross-compilers, GCC's executables
2219 are not only installed into @file{@var{bindir}}, that
2220 is, @file{@var{exec-prefix}/bin}, but additionally into
2221 @file{@var{exec-prefix}/@var{target-alias}/bin}, if that directory
2222 exists. Typically, such @dfn{tooldirs} hold target-specific
2223 binutils, including assembler and linker.
2225 Installation into a temporary staging area or into a @command{chroot}
2226 jail can be achieved with the command
2229 make DESTDIR=@var{path-to-rootdir} install
2232 @noindent where @var{path-to-rootdir} is the absolute path of
2233 a directory relative to which all installation paths will be
2234 interpreted. Note that the directory specified by @code{DESTDIR}
2235 need not exist yet; it will be created if necessary.
2237 There is a subtle point with tooldirs and @code{DESTDIR}:
2238 If you relocate a cross-compiler installation with
2239 e.g.@: @samp{DESTDIR=@var{rootdir}}, then the directory
2240 @file{@var{rootdir}/@var{exec-prefix}/@var{target-alias}/bin} will
2241 be filled with duplicated GCC executables only if it already exists,
2242 it will not be created otherwise. This is regarded as a feature,
2243 not as a bug, because it gives slightly more control to the packagers
2244 using the @code{DESTDIR} feature.
2246 If you are bootstrapping a released version of GCC then please
2247 quickly review the build status page for your release, available from
2248 @uref{http://gcc.gnu.org/buildstat.html}.
2249 If your system is not listed for the version of GCC that you built,
2251 @email{gcc@@gcc.gnu.org} indicating
2252 that you successfully built and installed GCC@.
2253 Include the following information:
2257 Output from running @file{@var{srcdir}/config.guess}. Do not send
2258 that file itself, just the one-line output from running it.
2261 The output of @samp{gcc -v} for your newly installed @command{gcc}.
2262 This tells us which version of GCC you built and the options you passed to
2266 Whether you enabled all languages or a subset of them. If you used a
2267 full distribution then this information is part of the configure
2268 options in the output of @samp{gcc -v}, but if you downloaded the
2269 ``core'' compiler plus additional front ends then it isn't apparent
2270 which ones you built unless you tell us about it.
2273 If the build was for GNU/Linux, also include:
2276 The distribution name and version (e.g., Red Hat 7.1 or Debian 2.2.3);
2277 this information should be available from @file{/etc/issue}.
2280 The version of the Linux kernel, available from @samp{uname --version}
2284 The version of glibc you used; for RPM-based systems like Red Hat,
2285 Mandrake, and SuSE type @samp{rpm -q glibc} to get the glibc version,
2286 and on systems like Debian and Progeny use @samp{dpkg -l libc6}.
2288 For other systems, you can include similar information if you think it is
2292 Any other information that you think would be useful to people building
2293 GCC on the same configuration. The new entry in the build status list
2294 will include a link to the archived copy of your message.
2297 We'd also like to know if the
2299 @ref{Specific, host/target specific installation notes}
2302 @uref{specific.html,,host/target specific installation notes}
2304 didn't include your host/target information or if that information is
2305 incomplete or out of date. Send a note to
2306 @email{gcc@@gcc.gnu.org} detailing how the information should be changed.
2308 If you find a bug, please report it following the
2309 @uref{../bugs.html,,bug reporting guidelines}.
2311 If you want to print the GCC manuals, do @samp{cd @var{objdir}; make
2312 dvi}. You will need to have @command{texi2dvi} (version at least 4.4)
2313 and @TeX{} installed. This creates a number of @file{.dvi} files in
2314 subdirectories of @file{@var{objdir}}; these may be converted for
2315 printing with programs such as @command{dvips}. Alternately, by using
2316 @samp{make pdf} in place of @samp{make dvi}, you can create documentation
2317 in the form of @file{.pdf} files; this requires @command{texi2pdf}, which
2318 is included with Texinfo version 4.8 and later. You can also
2319 @uref{http://www.gnu.org/order/order.html,,buy printed manuals from the
2320 Free Software Foundation}, though such manuals may not be for the most
2321 recent version of GCC@.
2323 If you would like to generate online HTML documentation, do @samp{cd
2324 @var{objdir}; make html} and HTML will be generated for the gcc manuals in
2325 @file{@var{objdir}/gcc/HTML}.
2332 @uref{./index.html,,Return to the GCC Installation page}
2336 @c ***Binaries****************************************************************
2338 @comment node-name, next, previous, up
2339 @node Binaries, Specific, Installing GCC, Top
2343 @chapter Installing GCC: Binaries
2346 @cindex Installing GCC: Binaries
2348 We are often asked about pre-compiled versions of GCC@. While we cannot
2349 provide these for all platforms, below you'll find links to binaries for
2350 various platforms where creating them by yourself is not easy due to various
2353 Please note that we did not create these binaries, nor do we
2354 support them. If you have any problems installing them, please
2355 contact their makers.
2362 @uref{http://www.bullfreeware.com,,Bull's Freeware and Shareware Archive for AIX};
2365 @uref{http://pware.hvcc.edu,,Hudson Valley Community College Open Source Softeware for IBM System p};
2368 @uref{http://www.perzl.org/aix,,AIX 5L and 6 Open Source Packages}.
2372 DOS---@uref{http://www.delorie.com/djgpp/,,DJGPP}.
2375 Renesas H8/300[HS]---@uref{http://h8300-hms.sourceforge.net/,,GNU
2376 Development Tools for the Renesas H8/300[HS] Series}.
2382 @uref{http://hpux.cs.utah.edu/,,HP-UX Porting Center};
2385 @uref{ftp://sunsite.informatik.rwth-aachen.de/pub/packages/gcc_hpux/,,Binaries for HP-UX 11.00 at Aachen University of Technology}.
2389 Motorola 68HC11/68HC12---@uref{http://www.gnu-m68hc11.org,,GNU
2390 Development Tools for the Motorola 68HC11/68HC12}.
2393 @uref{http://www.sco.com/skunkware/devtools/index.html#gcc,,SCO
2394 OpenServer/Unixware}.
2397 Solaris 2 (SPARC, Intel)---@uref{http://www.sunfreeware.com/,,Sunfreeware}.
2400 SGI---@uref{http://freeware.sgi.com/,,SGI Freeware}.
2406 The @uref{http://sourceware.org/cygwin/,,Cygwin} project;
2408 The @uref{http://www.mingw.org/,,MinGW} project.
2412 @uref{ftp://ftp.thewrittenword.com/packages/by-name/,,The
2413 Written Word} offers binaries for
2414 AIX 4.3.3, 5.1 and 5.2,
2416 Tru64 UNIX 4.0D and 5.1,
2418 HP-UX 10.20, 11.00, and 11.11, and
2419 Solaris/SPARC 2.5.1, 2.6, 7, 8, 9 and 10.
2422 @uref{http://www.openpkg.org/,,OpenPKG} offers binaries for quite a
2423 number of platforms.
2426 The @uref{http://gcc.gnu.org/wiki/GFortranBinaries,,GFortran Wiki} has
2427 links to GNU Fortran binaries for several platforms.
2430 In addition to those specific offerings, you can get a binary
2431 distribution CD-ROM from the
2432 @uref{http://www.gnu.org/order/order.html,,Free Software Foundation}.
2433 It contains binaries for a number of platforms, and
2434 includes not only GCC, but other stuff as well. The current CD does
2435 not contain the latest version of GCC, but it should allow
2436 bootstrapping the compiler. An updated version of that disk is in the
2444 @uref{./index.html,,Return to the GCC Installation page}
2448 @c ***Specific****************************************************************
2450 @comment node-name, next, previous, up
2451 @node Specific, Old, Binaries, Top
2455 @chapter Host/target specific installation notes for GCC
2458 @cindex Specific installation notes
2459 @cindex Target specific installation
2460 @cindex Host specific installation
2461 @cindex Target specific installation notes
2463 Please read this document carefully @emph{before} installing the
2464 GNU Compiler Collection on your machine.
2466 Note that this list of install notes is @emph{not} a list of supported
2467 hosts or targets. Not all supported hosts and targets are listed
2468 here, only the ones that require host-specific or target-specific
2474 @uref{#alpha-x-x,,alpha*-*-*}
2476 @uref{#alpha-dec-osf,,alpha*-dec-osf*}
2478 @uref{#arc-x-elf,,arc-*-elf}
2480 @uref{#arm-x-elf,,arm-*-elf}
2481 @uref{#arm-x-coff,,arm-*-coff}
2482 @uref{#arm-x-aout,,arm-*-aout}
2486 @uref{#bfin,,Blackfin}
2490 @uref{#x-x-freebsd,,*-*-freebsd*}
2492 @uref{#h8300-hms,,h8300-hms}
2494 @uref{#hppa-hp-hpux,,hppa*-hp-hpux*}
2496 @uref{#hppa-hp-hpux10,,hppa*-hp-hpux10}
2498 @uref{#hppa-hp-hpux11,,hppa*-hp-hpux11}
2500 @uref{#x-x-linux-gnu,,*-*-linux-gnu}
2502 @uref{#ix86-x-linux,,i?86-*-linux*}
2504 @uref{#ix86-x-solaris210,,i?86-*-solaris2.10}
2506 @uref{#ia64-x-linux,,ia64-*-linux}
2508 @uref{#ia64-x-hpux,,ia64-*-hpux*}
2510 @uref{#x-ibm-aix,,*-ibm-aix*}
2512 @uref{#iq2000-x-elf,,iq2000-*-elf}
2514 @uref{#m32c-x-elf,,m32c-*-elf}
2516 @uref{#m32r-x-elf,,m32r-*-elf}
2518 @uref{#m6811-elf,,m6811-elf}
2520 @uref{#m6812-elf,,m6812-elf}
2522 @uref{#m68k-x-x,,m68k-*-*}
2524 @uref{#m68k-uclinux,,m68k-uclinux}
2526 @uref{#mips-x-x,,mips-*-*}
2528 @uref{#mips-sgi-irix5,,mips-sgi-irix5}
2530 @uref{#mips-sgi-irix6,,mips-sgi-irix6}
2532 @uref{#powerpc-x-x,,powerpc*-*-*}
2534 @uref{#powerpc-x-darwin,,powerpc-*-darwin*}
2536 @uref{#powerpc-x-elf,,powerpc-*-elf}
2538 @uref{#powerpc-x-linux-gnu,,powerpc*-*-linux-gnu*}
2540 @uref{#powerpc-x-netbsd,,powerpc-*-netbsd*}
2542 @uref{#powerpc-x-eabisim,,powerpc-*-eabisim}
2544 @uref{#powerpc-x-eabi,,powerpc-*-eabi}
2546 @uref{#powerpcle-x-elf,,powerpcle-*-elf}
2548 @uref{#powerpcle-x-eabisim,,powerpcle-*-eabisim}
2550 @uref{#powerpcle-x-eabi,,powerpcle-*-eabi}
2552 @uref{#s390-x-linux,,s390-*-linux*}
2554 @uref{#s390x-x-linux,,s390x-*-linux*}
2556 @uref{#s390x-ibm-tpf,,s390x-ibm-tpf*}
2558 @uref{#x-x-solaris2,,*-*-solaris2*}
2560 @uref{#sparc-sun-solaris2,,sparc-sun-solaris2*}
2562 @uref{#sparc-sun-solaris27,,sparc-sun-solaris2.7}
2564 @uref{#sparc-x-linux,,sparc-*-linux*}
2566 @uref{#sparc64-x-solaris2,,sparc64-*-solaris2*}
2568 @uref{#sparcv9-x-solaris2,,sparcv9-*-solaris2*}
2570 @uref{#x-x-vxworks,,*-*-vxworks*}
2572 @uref{#x86-64-x-x,,x86_64-*-*, amd64-*-*}
2574 @uref{#xtensa-x-elf,,xtensa-*-elf}
2576 @uref{#xtensa-x-linux,,xtensa-*-linux*}
2578 @uref{#windows,,Microsoft Windows}
2582 @uref{#older,,Older systems}
2587 @uref{#elf,,all ELF targets} (SVR4, Solaris 2, etc.)
2593 <!-- -------- host/target specific issues start here ---------------- -->
2596 @heading @anchor{alpha-x-x}alpha*-*-*
2598 This section contains general configuration information for all
2599 alpha-based platforms using ELF (in particular, ignore this section for
2600 DEC OSF/1, Digital UNIX and Tru64 UNIX)@. In addition to reading this
2601 section, please read all other sections that match your target.
2603 We require binutils 2.11.2 or newer.
2604 Previous binutils releases had a number of problems with DWARF 2
2605 debugging information, not the least of which is incorrect linking of
2611 @heading @anchor{alpha-dec-osf}alpha*-dec-osf*
2612 Systems using processors that implement the DEC Alpha architecture and
2613 are running the DEC/Compaq Unix (DEC OSF/1, Digital UNIX, or Compaq
2614 Tru64 UNIX) operating system, for example the DEC Alpha AXP systems.
2616 As of GCC 3.2, versions before @code{alpha*-dec-osf4} are no longer
2617 supported. (These are the versions which identify themselves as DEC
2620 In Digital Unix V4.0, virtual memory exhausted bootstrap failures
2621 may be fixed by configuring with @option{--with-gc=simple},
2622 reconfiguring Kernel Virtual Memory and Swap parameters
2623 per the @command{/usr/sbin/sys_check} Tuning Suggestions,
2624 or applying the patch in
2625 @uref{http://gcc.gnu.org/ml/gcc/2002-08/msg00822.html}.
2627 In Tru64 UNIX V5.1, Compaq introduced a new assembler that does not
2628 currently (2001-06-13) work with @command{mips-tfile}. As a workaround,
2629 we need to use the old assembler, invoked via the barely documented
2630 @option{-oldas} option. To bootstrap GCC, you either need to use the
2634 % CC=cc @var{srcdir}/configure [@var{options}] [@var{target}]
2637 or you can use a copy of GCC 2.95.3 or higher built on Tru64 UNIX V4.0:
2640 % CC=gcc -Wa,-oldas @var{srcdir}/configure [@var{options}] [@var{target}]
2643 As of GNU binutils 2.11.2, neither GNU @command{as} nor GNU @command{ld}
2644 are supported on Tru64 UNIX, so you must not configure GCC with
2645 @option{--with-gnu-as} or @option{--with-gnu-ld}.
2647 GCC writes a @samp{.verstamp} directive to the assembler output file
2648 unless it is built as a cross-compiler. It gets the version to use from
2649 the system header file @file{/usr/include/stamp.h}. If you install a
2650 new version of DEC Unix, you should rebuild GCC to pick up the new version
2653 Note that since the Alpha is a 64-bit architecture, cross-compilers from
2654 32-bit machines will not generate code as efficient as that generated
2655 when the compiler is running on a 64-bit machine because many
2656 optimizations that depend on being able to represent a word on the
2657 target in an integral value on the host cannot be performed. Building
2658 cross-compilers on the Alpha for 32-bit machines has only been tested in
2659 a few cases and may not work properly.
2661 @samp{make compare} may fail on old versions of DEC Unix unless you add
2662 @option{-save-temps} to @code{BOOT_CFLAGS}. On these systems, the name
2663 of the assembler input file is stored in the object file, and that makes
2664 comparison fail if it differs between the @code{stage1} and
2665 @code{stage2} compilations. The option @option{-save-temps} forces a
2666 fixed name to be used for the assembler input file, instead of a
2667 randomly chosen name in @file{/tmp}. Do not add @option{-save-temps}
2668 unless the comparisons fail without that option. If you add
2669 @option{-save-temps}, you will have to manually delete the @samp{.i} and
2670 @samp{.s} files after each series of compilations.
2672 GCC now supports both the native (ECOFF) debugging format used by DBX
2673 and GDB and an encapsulated STABS format for use only with GDB@. See the
2674 discussion of the @option{--with-stabs} option of @file{configure} above
2675 for more information on these formats and how to select them.
2677 There is a bug in DEC's assembler that produces incorrect line numbers
2678 for ECOFF format when the @samp{.align} directive is used. To work
2679 around this problem, GCC will not emit such alignment directives
2680 while writing ECOFF format debugging information even if optimization is
2681 being performed. Unfortunately, this has the very undesirable
2682 side-effect that code addresses when @option{-O} is specified are
2683 different depending on whether or not @option{-g} is also specified.
2685 To avoid this behavior, specify @option{-gstabs+} and use GDB instead of
2686 DBX@. DEC is now aware of this problem with the assembler and hopes to
2687 provide a fix shortly.
2692 @heading @anchor{arc-x-elf}arc-*-elf
2693 Argonaut ARC processor.
2694 This configuration is intended for embedded systems.
2699 @heading @anchor{arm-x-elf}arm-*-elf
2700 ARM-family processors. Subtargets that use the ELF object format
2701 require GNU binutils 2.13 or newer. Such subtargets include:
2702 @code{arm-*-freebsd}, @code{arm-*-netbsdelf}, @code{arm-*-*linux}
2703 and @code{arm-*-rtems}.
2708 @heading @anchor{arm-x-coff}arm-*-coff
2709 ARM-family processors. Note that there are two different varieties
2710 of PE format subtarget supported: @code{arm-wince-pe} and
2711 @code{arm-pe} as well as a standard COFF target @code{arm-*-coff}.
2716 @heading @anchor{arm-x-aout}arm-*-aout
2717 ARM-family processors. These targets support the AOUT file format:
2718 @code{arm-*-aout}, @code{arm-*-netbsd}.
2723 @heading @anchor{avr}avr
2725 ATMEL AVR-family micro controllers. These are used in embedded
2726 applications. There are no standard Unix configurations.
2728 @xref{AVR Options,, AVR Options, gcc, Using the GNU Compiler
2732 See ``AVR Options'' in the main manual
2734 for the list of supported MCU types.
2736 Use @samp{configure --target=avr --enable-languages="c"} to configure GCC@.
2738 Further installation notes and other useful information about AVR tools
2739 can also be obtained from:
2743 @uref{http://www.nongnu.org/avr/,,http://www.nongnu.org/avr/}
2745 @uref{http://home.overta.ru/users/denisc/,,http://home.overta.ru/users/denisc/}
2747 @uref{http://www.amelek.gda.pl/avr/,,http://www.amelek.gda.pl/avr/}
2750 We @emph{strongly} recommend using binutils 2.13 or newer.
2752 The following error:
2754 Error: register required
2757 indicates that you should upgrade to a newer version of the binutils.
2762 @heading @anchor{bfin}Blackfin
2764 The Blackfin processor, an Analog Devices DSP.
2766 @xref{Blackfin Options,, Blackfin Options, gcc, Using the GNU Compiler
2770 See ``Blackfin Options'' in the main manual
2773 More information, and a version of binutils with support for this processor,
2774 is available at @uref{http://blackfin.uclinux.org}
2779 @heading @anchor{cris}CRIS
2781 CRIS is the CPU architecture in Axis Communications ETRAX system-on-a-chip
2782 series. These are used in embedded applications.
2785 @xref{CRIS Options,, CRIS Options, gcc, Using the GNU Compiler
2789 See ``CRIS Options'' in the main manual
2791 for a list of CRIS-specific options.
2793 There are a few different CRIS targets:
2796 Mainly for monolithic embedded systems. Includes a multilib for the
2797 @samp{v10} core used in @samp{ETRAX 100 LX}.
2798 @item cris-axis-linux-gnu
2799 A GNU/Linux port for the CRIS architecture, currently targeting
2800 @samp{ETRAX 100 LX} by default.
2803 For @code{cris-axis-elf} you need binutils 2.11
2804 or newer. For @code{cris-axis-linux-gnu} you need binutils 2.12 or newer.
2806 Pre-packaged tools can be obtained from
2807 @uref{ftp://ftp.axis.com/pub/axis/tools/cris/compiler-kit/}. More
2808 information about this platform is available at
2809 @uref{http://developer.axis.com/}.
2814 @heading @anchor{crx}CRX
2816 The CRX CompactRISC architecture is a low-power 32-bit architecture with
2817 fast context switching and architectural extensibility features.
2820 @xref{CRX Options,, CRX Options, gcc, Using and Porting the GNU Compiler
2825 See ``CRX Options'' in the main manual for a list of CRX-specific options.
2828 Use @samp{configure --target=crx-elf --enable-languages=c,c++} to configure
2829 GCC@ for building a CRX cross-compiler. The option @samp{--target=crx-elf}
2830 is also used to build the @samp{newlib} C library for CRX.
2832 It is also possible to build libstdc++-v3 for the CRX architecture. This
2833 needs to be done in a separate step with the following configure settings:
2834 @samp{gcc/libstdc++-v3/configure --host=crx-elf --with-newlib
2835 --enable-sjlj-exceptions --enable-cxx-flags='-fexceptions -frtti'}
2840 @heading @anchor{dos}DOS
2842 Please have a look at the @uref{binaries.html,,binaries page}.
2844 You cannot install GCC by itself on MSDOS; it will not compile under
2845 any MSDOS compiler except itself. You need to get the complete
2846 compilation package DJGPP, which includes binaries as well as sources,
2847 and includes all the necessary compilation tools and libraries.
2852 @heading @anchor{x-x-freebsd}*-*-freebsd*
2854 The version of binutils installed in @file{/usr/bin} probably works with
2855 this release of GCC@. However, on FreeBSD 4, bootstrapping against the
2856 latest FSF binutils is known to improve overall testsuite results; and,
2857 on FreeBSD/alpha, using binutils 2.14 or later is required to build libjava.
2859 Support for FreeBSD 1 was discontinued in GCC 3.2.
2861 Support for FreeBSD 2 will be discontinued after GCC 3.4. The
2862 following was true for GCC 3.1 but the current status is unknown.
2863 For FreeBSD 2 or any mutant a.out versions of FreeBSD 3: All
2864 configuration support and files as shipped with GCC 2.95 are still in
2865 place. FreeBSD 2.2.7 has been known to bootstrap completely; however,
2866 it is unknown which version of binutils was used (it is assumed that it
2867 was the system copy in @file{/usr/bin}) and C++ EH failures were noted.
2869 For FreeBSD using the ELF file format: DWARF 2 debugging is now the
2870 default for all CPU architectures. It had been the default on
2871 FreeBSD/alpha since its inception. You may use @option{-gstabs} instead
2872 of @option{-g}, if you really want the old debugging format. There are
2873 no known issues with mixing object files and libraries with different
2874 debugging formats. Otherwise, this release of GCC should now match more
2875 of the configuration used in the stock FreeBSD configuration of GCC@. In
2876 particular, @option{--enable-threads} is now configured by default.
2877 However, as a general user, do not attempt to replace the system
2878 compiler with this release. Known to bootstrap and check with good
2879 results on FreeBSD 4.9-STABLE and 5-CURRENT@. In the past, known to
2880 bootstrap and check with good results on FreeBSD 3.0, 3.4, 4.0, 4.2,
2881 4.3, 4.4, 4.5, 4.8-STABLE@.
2883 In principle, @option{--enable-threads} is now compatible with
2884 @option{--enable-libgcj} on FreeBSD@. However, it has only been built
2885 and tested on @samp{i386-*-freebsd[45]} and @samp{alpha-*-freebsd[45]}.
2887 library may be incorrectly built (symbols are missing at link time).
2888 There is a rare timing-based startup hang (probably involves an
2889 assumption about the thread library). Multi-threaded boehm-gc (required for
2890 libjava) exposes severe threaded signal-handling bugs on FreeBSD before
2891 4.5-RELEASE@. Other CPU architectures
2892 supported by FreeBSD will require additional configuration tuning in, at
2893 the very least, both boehm-gc and libffi.
2895 Shared @file{libgcc_s.so} is now built and installed by default.
2900 @heading @anchor{h8300-hms}h8300-hms
2901 Renesas H8/300 series of processors.
2903 Please have a look at the @uref{binaries.html,,binaries page}.
2905 The calling convention and structure layout has changed in release 2.6.
2906 All code must be recompiled. The calling convention now passes the
2907 first three arguments in function calls in registers. Structures are no
2908 longer a multiple of 2 bytes.
2913 @heading @anchor{hppa-hp-hpux}hppa*-hp-hpux*
2914 Support for HP-UX version 9 and older was discontinued in GCC 3.4.
2916 We require using gas/binutils on all hppa platforms;
2917 you may encounter a variety of problems if you try to use the HP assembler.
2919 Specifically, @option{-g} does not work on HP-UX (since that system
2920 uses a peculiar debugging format which GCC does not know about), unless
2921 you use GAS and GDB@. It may be helpful to configure GCC with the
2922 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}} and
2923 @option{--with-as=@dots{}} options to ensure that GCC can find GAS@.
2925 If you wish to use the pa-risc 2.0 architecture support with a 32-bit
2926 runtime, you must use gas/binutils 2.11 or newer.
2928 There are two default scheduling models for instructions. These are
2929 PROCESSOR_7100LC and PROCESSOR_8000. They are selected from the pa-risc
2930 architecture specified for the target machine when configuring.
2931 PROCESSOR_8000 is the default. PROCESSOR_7100LC is selected when
2932 the target is a @samp{hppa1*} machine.
2934 The PROCESSOR_8000 model is not well suited to older processors. Thus,
2935 it is important to completely specify the machine architecture when
2936 configuring if you want a model other than PROCESSOR_8000. The macro
2937 TARGET_SCHED_DEFAULT can be defined in BOOT_CFLAGS if a different
2938 default scheduling model is desired.
2940 As of GCC 4.0, GCC uses the UNIX 95 namespace for HP-UX 10.10
2941 through 11.00, and the UNIX 98 namespace for HP-UX 11.11 and later.
2942 This namespace change might cause problems when bootstrapping with
2943 an earlier version of GCC or the HP compiler as essentially the same
2944 namespace is required for an entire build. This problem can be avoided
2945 in a number of ways. With HP cc, @env{UNIX_STD} can be set to @samp{95}
2946 or @samp{98}. Another way is to add an appropriate set of predefines
2947 to @env{CC}. The description for the @option{munix=} option contains
2948 a list of the predefines used with each standard.
2950 As of GCC 4.1, @env{DWARF2} exception handling is available on HP-UX@.
2951 It is now the default. This exposed a bug in the handling of data
2952 relocations in the GAS assembler. The handling of 64-bit data relocations
2953 was seriously broken, affecting debugging and exception support on all
2954 @samp{hppa64-*-*} targets. Under some circumstances, 32-bit data relocations
2955 could also be handled incorrectly. This problem is fixed in GAS version
2958 GCC versions prior to 4.1 incorrectly passed and returned complex
2959 values. They are now passed in the same manner as aggregates.
2961 More specific information to @samp{hppa*-hp-hpux*} targets follows.
2966 @heading @anchor{hppa-hp-hpux10}hppa*-hp-hpux10
2968 For hpux10.20, we @emph{highly} recommend you pick up the latest sed patch
2969 @code{PHCO_19798} from HP@. HP has two sites which provide patches free of
2975 <a href="http://us.itrc.hp.com/service/home/home.do">US, Canada, Asia-Pacific, and
2979 @uref{http://us.itrc.hp.com/service/home/home.do,,} US, Canada, Asia-Pacific,
2983 @uref{http://europe.itrc.hp.com/service/home/home.do,,} Europe.
2986 The HP assembler on these systems has some problems. Most notably the
2987 assembler inserts timestamps into each object file it creates, causing
2988 the 3-stage comparison test to fail during a bootstrap.
2989 You should be able to continue by saying @samp{make all-host all-target}
2990 after getting the failure from @samp{make}.
2992 GCC 4.0 requires CVS binutils as of April 28, 2004 or later. Earlier
2993 versions require binutils 2.8 or later.
2995 The C++ ABI has changed incompatibly in GCC 4.0. COMDAT subspaces are
2996 used for one-only code and data. This resolves many of the previous
2997 problems in using C++ on this target. However, the ABI is not compatible
2998 with the one implemented under HP-UX 11 using secondary definitions.
3003 @heading @anchor{hppa-hp-hpux11}hppa*-hp-hpux11
3005 GCC 3.0 and up support HP-UX 11. GCC 2.95.x is not supported and cannot
3006 be used to compile GCC 3.0 and up.
3008 Refer to @uref{binaries.html,,binaries} for information about obtaining
3009 precompiled GCC binaries for HP-UX@. Precompiled binaries must be obtained
3010 to build the Ada language as it can't be bootstrapped using C@. Ada is
3011 only available for the 32-bit PA-RISC runtime. The libffi and libjava
3012 haven't been ported to HP-UX and don't build.
3014 Starting with GCC 3.4 an ISO C compiler is required to bootstrap. The
3015 bundled compiler supports only traditional C; you will need either HP's
3016 unbundled compiler, or a binary distribution of GCC@.
3018 It is possible to build GCC 3.3 starting with the bundled HP compiler,
3019 but the process requires several steps. GCC 3.3 can then be used to
3020 build later versions. The fastjar program contains ISO C code and
3021 can't be built with the HP bundled compiler. This problem can be
3022 avoided by not building the Java language. For example, use the
3023 @option{--enable-languages="c,c++,f77,objc"} option in your configure
3026 There are several possible approaches to building the distribution.
3027 Binutils can be built first using the HP tools. Then, the GCC
3028 distribution can be built. The second approach is to build GCC
3029 first using the HP tools, then build binutils, then rebuild GCC@.
3030 There have been problems with various binary distributions, so it
3031 is best not to start from a binary distribution.
3033 On 64-bit capable systems, there are two distinct targets. Different
3034 installation prefixes must be used if both are to be installed on
3035 the same system. The @samp{hppa[1-2]*-hp-hpux11*} target generates code
3036 for the 32-bit PA-RISC runtime architecture and uses the HP linker.
3037 The @samp{hppa64-hp-hpux11*} target generates 64-bit code for the
3038 PA-RISC 2.0 architecture. The HP and GNU linkers are both supported
3041 The script config.guess now selects the target type based on the compiler
3042 detected during configuration. You must define @env{PATH} or @env{CC} so
3043 that configure finds an appropriate compiler for the initial bootstrap.
3044 When @env{CC} is used, the definition should contain the options that are
3045 needed whenever @env{CC} is used.
3047 Specifically, options that determine the runtime architecture must be
3048 in @env{CC} to correctly select the target for the build. It is also
3049 convenient to place many other compiler options in @env{CC}. For example,
3050 @env{CC="cc -Ac +DA2.0W -Wp,-H16376 -D_CLASSIC_TYPES -D_HPUX_SOURCE"}
3051 can be used to bootstrap the GCC 3.3 branch with the HP compiler in
3052 64-bit K&R/bundled mode. The @option{+DA2.0W} option will result in
3053 the automatic selection of the @samp{hppa64-hp-hpux11*} target. The
3054 macro definition table of cpp needs to be increased for a successful
3055 build with the HP compiler. _CLASSIC_TYPES and _HPUX_SOURCE need to
3056 be defined when building with the bundled compiler, or when using the
3057 @option{-Ac} option. These defines aren't necessary with @option{-Ae}.
3059 It is best to explicitly configure the @samp{hppa64-hp-hpux11*} target
3060 with the @option{--with-ld=@dots{}} option. This overrides the standard
3061 search for ld. The two linkers supported on this target require different
3062 commands. The default linker is determined during configuration. As a
3063 result, it's not possible to switch linkers in the middle of a GCC build.
3064 This has been been reported to sometimes occur in unified builds of
3067 GCC 3.0 through 3.2 require binutils 2.11 or above. GCC 3.3 through
3068 GCC 4.0 require binutils 2.14 or later.
3070 Although the HP assembler can be used for an initial build, it shouldn't
3071 be used with any languages other than C and perhaps Fortran due to its
3072 many limitations. For example, it does not support weak symbols or alias
3073 definitions. As a result, explicit template instantiations are required
3074 when using C++. This makes it difficult if not impossible to build many
3075 C++ applications. You can't generate debugging information when using
3076 the HP assembler. Finally, bootstrapping fails in the final
3077 comparison of object modules due to the time stamps that it inserts into
3078 the modules. The bootstrap can be continued from this point with
3079 @samp{make all-host all-target}.
3081 A recent linker patch must be installed for the correct operation of
3082 GCC 3.3 and later. @code{PHSS_26559} and @code{PHSS_24304} are the
3083 oldest linker patches that are known to work. They are for HP-UX
3084 11.00 and 11.11, respectively. @code{PHSS_24303}, the companion to
3085 @code{PHSS_24304}, might be usable but it hasn't been tested. These
3086 patches have been superseded. Consult the HP patch database to obtain
3087 the currently recommended linker patch for your system.
3089 The patches are necessary for the support of weak symbols on the
3090 32-bit port, and for the running of initializers and finalizers. Weak
3091 symbols are implemented using SOM secondary definition symbols. Prior
3092 to HP-UX 11, there are bugs in the linker support for secondary symbols.
3093 The patches correct a problem of linker core dumps creating shared
3094 libraries containing secondary symbols, as well as various other
3095 linking issues involving secondary symbols.
3097 GCC 3.3 uses the ELF DT_INIT_ARRAY and DT_FINI_ARRAY capabilities to
3098 run initializers and finalizers on the 64-bit port. The 32-bit port
3099 uses the linker @option{+init} and @option{+fini} options for the same
3100 purpose. The patches correct various problems with the +init/+fini
3101 options, including program core dumps. Binutils 2.14 corrects a
3102 problem on the 64-bit port resulting from HP's non-standard use of
3103 the .init and .fini sections for array initializers and finalizers.
3105 There are a number of issues to consider in selecting which linker to
3106 use with the 64-bit port. The GNU 64-bit linker can only create dynamic
3107 binaries. The @option{-static} option causes linking with archive
3108 libraries but doesn't produce a truly static binary. Dynamic binaries
3109 still require final binding by the dynamic loader to resolve a set of
3110 dynamic-loader-defined symbols. The default behavior of the HP linker
3111 is the same as the GNU linker. However, it can generate true 64-bit
3112 static binaries using the @option{+compat} option.
3114 The HP 64-bit linker doesn't support linkonce semantics. As a
3115 result, C++ programs have many more sections than they should.
3117 The GNU 64-bit linker has some issues with shared library support
3118 and exceptions. As a result, we only support libgcc in archive
3119 format. For similar reasons, dwarf2 unwind and exception support
3120 are disabled. The GNU linker also has problems creating binaries
3121 with @option{-static}. It doesn't provide stubs for internal
3122 calls to global functions in shared libraries, so these calls
3123 can't be overloaded.
3125 Thread support is not implemented in GCC 3.0 through 3.2, so the
3126 @option{--enable-threads} configure option does not work. In 3.3
3127 and later, POSIX threads are supported. The optional DCE thread
3128 library is not supported.
3130 This port still is undergoing significant development.
3135 @heading @anchor{x-x-linux-gnu}*-*-linux-gnu
3137 Versions of libstdc++-v3 starting with 3.2.1 require bug fixes present
3138 in glibc 2.2.5 and later. More information is available in the
3139 libstdc++-v3 documentation.
3144 @heading @anchor{ix86-x-linux}i?86-*-linux*
3146 As of GCC 3.3, binutils 2.13.1 or later is required for this platform.
3147 See @uref{http://gcc.gnu.org/PR10877,,bug 10877} for more information.
3149 If you receive Signal 11 errors when building on GNU/Linux, then it is
3150 possible you have a hardware problem. Further information on this can be
3151 found on @uref{http://www.bitwizard.nl/sig11/,,www.bitwizard.nl}.
3156 @heading @anchor{ix86-x-solaris210}i?86-*-solaris2.10
3157 Use this for Solaris 10 or later on x86 and x86-64 systems. This
3158 configuration is supported by GCC 4.0 and later versions only.
3160 It is recommended that you configure GCC to use the GNU assembler in
3161 @file{/usr/sfw/bin/gas} but the Sun linker, using the options
3162 @option{--with-gnu-as --with-as=/usr/sfw/bin/gas --without-gnu-ld
3163 --with-ld=/usr/ccs/bin/ld}.
3168 @heading @anchor{ia64-x-linux}ia64-*-linux
3169 IA-64 processor (also known as IPF, or Itanium Processor Family)
3172 If you are using the installed system libunwind library with
3173 @option{--with-system-libunwind}, then you must use libunwind 0.98 or
3176 None of the following versions of GCC has an ABI that is compatible
3177 with any of the other versions in this list, with the exception that
3178 Red Hat 2.96 and Trillian 000171 are compatible with each other:
3179 3.1, 3.0.2, 3.0.1, 3.0, Red Hat 2.96, and Trillian 000717.
3180 This primarily affects C++ programs and programs that create shared libraries.
3181 GCC 3.1 or later is recommended for compiling linux, the kernel.
3182 As of version 3.1 GCC is believed to be fully ABI compliant, and hence no
3183 more major ABI changes are expected.
3188 @heading @anchor{ia64-x-hpux}ia64-*-hpux*
3189 Building GCC on this target requires the GNU Assembler. The bundled HP
3190 assembler will not work. To prevent GCC from using the wrong assembler,
3191 the option @option{--with-gnu-as} may be necessary.
3193 The GCC libunwind library has not been ported to HPUX@. This means that for
3194 GCC versions 3.2.3 and earlier, @option{--enable-libunwind-exceptions}
3195 is required to build GCC@. For GCC 3.3 and later, this is the default.
3196 For gcc 3.4.3 and later, @option{--enable-libunwind-exceptions} is
3197 removed and the system libunwind library will always be used.
3201 <!-- rs6000-ibm-aix*, powerpc-ibm-aix* -->
3203 @heading @anchor{x-ibm-aix}*-ibm-aix*
3204 Support for AIX version 3 and older was discontinued in GCC 3.4.
3206 ``out of memory'' bootstrap failures may indicate a problem with
3207 process resource limits (ulimit). Hard limits are configured in the
3208 @file{/etc/security/limits} system configuration file.
3210 To speed up the configuration phases of bootstrapping and installing GCC,
3211 one may use GNU Bash instead of AIX @command{/bin/sh}, e.g.,
3214 % CONFIG_SHELL=/opt/freeware/bin/bash
3215 % export CONFIG_SHELL
3218 and then proceed as described in @uref{build.html,,the build
3219 instructions}, where we strongly recommend specifying an absolute path
3220 to invoke @var{srcdir}/configure.
3222 Because GCC on AIX is built as a 32-bit executable by default,
3223 (although it can generate 64-bit programs) the GMP and MPFR libraries
3224 required by gfortran must be 32-bit libraries. Building GMP and MPFR
3225 as static archive libraries works better than shared libraries.
3227 Errors involving @code{alloca} when building GCC generally are due
3228 to an incorrect definition of @code{CC} in the Makefile or mixing files
3229 compiled with the native C compiler and GCC@. During the stage1 phase of
3230 the build, the native AIX compiler @strong{must} be invoked as @command{cc}
3231 (not @command{xlc}). Once @command{configure} has been informed of
3232 @command{xlc}, one needs to use @samp{make distclean} to remove the
3233 configure cache files and ensure that @env{CC} environment variable
3234 does not provide a definition that will confuse @command{configure}.
3235 If this error occurs during stage2 or later, then the problem most likely
3236 is the version of Make (see above).
3238 The native @command{as} and @command{ld} are recommended for bootstrapping
3239 on AIX 4 and required for bootstrapping on AIX 5L@. The GNU Assembler
3240 reports that it supports WEAK symbols on AIX 4, which causes GCC to try to
3241 utilize weak symbol functionality although it is not supported. The GNU
3242 Assembler and Linker do not support AIX 5L sufficiently to bootstrap GCC@.
3243 The native AIX tools do interoperate with GCC@.
3245 Building @file{libstdc++.a} requires a fix for an AIX Assembler bug
3246 APAR IY26685 (AIX 4.3) or APAR IY25528 (AIX 5.1). It also requires a
3247 fix for another AIX Assembler bug and a co-dependent AIX Archiver fix
3248 referenced as APAR IY53606 (AIX 5.2) or a APAR IY54774 (AIX 5.1)
3250 @samp{libstdc++} in GCC 3.4 increments the major version number of the
3251 shared object and GCC installation places the @file{libstdc++.a}
3252 shared library in a common location which will overwrite the and GCC
3253 3.3 version of the shared library. Applications either need to be
3254 re-linked against the new shared library or the GCC 3.1 and GCC 3.3
3255 versions of the @samp{libstdc++} shared object needs to be available
3256 to the AIX runtime loader. The GCC 3.1 @samp{libstdc++.so.4}, if
3257 present, and GCC 3.3 @samp{libstdc++.so.5} shared objects can be
3258 installed for runtime dynamic loading using the following steps to set
3259 the @samp{F_LOADONLY} flag in the shared object for @emph{each}
3260 multilib @file{libstdc++.a} installed:
3262 Extract the shared objects from the currently installed
3263 @file{libstdc++.a} archive:
3265 % ar -x libstdc++.a libstdc++.so.4 libstdc++.so.5
3268 Enable the @samp{F_LOADONLY} flag so that the shared object will be
3269 available for runtime dynamic loading, but not linking:
3271 % strip -e libstdc++.so.4 libstdc++.so.5
3274 Archive the runtime-only shared object in the GCC 3.4
3275 @file{libstdc++.a} archive:
3277 % ar -q libstdc++.a libstdc++.so.4 libstdc++.so.5
3280 Linking executables and shared libraries may produce warnings of
3281 duplicate symbols. The assembly files generated by GCC for AIX always
3282 have included multiple symbol definitions for certain global variable
3283 and function declarations in the original program. The warnings should
3284 not prevent the linker from producing a correct library or runnable
3287 AIX 4.3 utilizes a ``large format'' archive to support both 32-bit and
3288 64-bit object modules. The routines provided in AIX 4.3.0 and AIX 4.3.1
3289 to parse archive libraries did not handle the new format correctly.
3290 These routines are used by GCC and result in error messages during
3291 linking such as ``not a COFF file''. The version of the routines shipped
3292 with AIX 4.3.1 should work for a 32-bit environment. The @option{-g}
3293 option of the archive command may be used to create archives of 32-bit
3294 objects using the original ``small format''. A correct version of the
3295 routines is shipped with AIX 4.3.2 and above.
3297 Some versions of the AIX binder (linker) can fail with a relocation
3298 overflow severe error when the @option{-bbigtoc} option is used to link
3299 GCC-produced object files into an executable that overflows the TOC@. A fix
3300 for APAR IX75823 (OVERFLOW DURING LINK WHEN USING GCC AND -BBIGTOC) is
3301 available from IBM Customer Support and from its
3302 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
3303 website as PTF U455193.
3305 The AIX 4.3.2.1 linker (bos.rte.bind_cmds Level 4.3.2.1) will dump core
3306 with a segmentation fault when invoked by any version of GCC@. A fix for
3307 APAR IX87327 is available from IBM Customer Support and from its
3308 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
3309 website as PTF U461879. This fix is incorporated in AIX 4.3.3 and above.
3311 The initial assembler shipped with AIX 4.3.0 generates incorrect object
3312 files. A fix for APAR IX74254 (64BIT DISASSEMBLED OUTPUT FROM COMPILER FAILS
3313 TO ASSEMBLE/BIND) is available from IBM Customer Support and from its
3314 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
3315 website as PTF U453956. This fix is incorporated in AIX 4.3.1 and above.
3317 AIX provides National Language Support (NLS)@. Compilers and assemblers
3318 use NLS to support locale-specific representations of various data
3319 formats including floating-point numbers (e.g., @samp{.} vs @samp{,} for
3320 separating decimal fractions). There have been problems reported where
3321 GCC does not produce the same floating-point formats that the assembler
3322 expects. If one encounters this problem, set the @env{LANG}
3323 environment variable to @samp{C} or @samp{En_US}.
3325 By default, GCC for AIX 4.1 and above produces code that can be used on
3326 both Power or PowerPC processors.
3328 A default can be specified with the @option{-mcpu=@var{cpu_type}}
3329 switch and using the configure option @option{--with-cpu-@var{cpu_type}}.
3334 @heading @anchor{iq2000-x-elf}iq2000-*-elf
3335 Vitesse IQ2000 processors. These are used in embedded
3336 applications. There are no standard Unix configurations.
3341 @heading @anchor{m32c-x-elf}m32c-*-elf
3342 Renesas M32C processor.
3343 This configuration is intended for embedded systems.
3348 @heading @anchor{m32r-x-elf}m32r-*-elf
3349 Renesas M32R processor.
3350 This configuration is intended for embedded systems.
3355 @heading @anchor{m6811-elf}m6811-elf
3356 Motorola 68HC11 family micro controllers. These are used in embedded
3357 applications. There are no standard Unix configurations.
3362 @heading @anchor{m6812-elf}m6812-elf
3363 Motorola 68HC12 family micro controllers. These are used in embedded
3364 applications. There are no standard Unix configurations.
3369 @heading @anchor{m68k-x-x}m68k-*-*
3370 By default, @samp{m68k-*-aout}, @samp{m68k-*-coff*},
3371 @samp{m68k-*-elf*}, @samp{m68k-*-rtems} and @samp{m68k-*-uclinux}
3372 build libraries for both M680x0 and ColdFire processors. If you only
3373 need the M680x0 libraries, you can omit the ColdFire ones by passing
3374 @option{--with-arch=m68k} to @command{configure}. Alternatively, you
3375 can omit the M680x0 libraries by passing @option{--with-arch=cf} to
3376 @command{configure}. These targets default to 5206 code when
3377 configured with @option{--with-arch=cf} and 68020 code otherwise.
3379 The @samp{m68k-*-linux-gnu}, @samp{m68k-*-netbsd} and
3380 @samp{m68k-*-openbsd} targets also support the @option{--with-arch}
3381 option. They will generate ColdFire CFV4e code when configured with
3382 @option{--with-arch=cf} and 68020 code otherwise.
3384 You can override the default processors listed above by configuring
3385 with @option{--with-cpu=@var{target}}. This @var{target} can either
3386 be a @option{-mcpu} argument or one of the following values:
3387 @samp{m68000}, @samp{m68010}, @samp{m68020}, @samp{m68030},
3388 @samp{m68040}, @samp{m68060}, @samp{m68020-40} and @samp{m68020-60}.
3393 @heading @anchor{m68k-x-uclinux}m68k-*-uclinux
3394 GCC 4.3 changed the uClinux configuration so that it uses the
3395 @samp{m68k-linux-gnu} ABI rather than the @samp{m68k-elf} ABI.
3396 It also added improved support for C++ and flat shared libraries,
3397 both of which were ABI changes. However, you can still use the
3398 original ABI by configuring for @samp{m68k-uclinuxoldabi} or
3399 @samp{m68k-@var{vendor}-uclinuxoldabi}.
3404 @heading @anchor{mips-x-x}mips-*-*
3405 If on a MIPS system you get an error message saying ``does not have gp
3406 sections for all it's [sic] sectons [sic]'', don't worry about it. This
3407 happens whenever you use GAS with the MIPS linker, but there is not
3408 really anything wrong, and it is okay to use the output file. You can
3409 stop such warnings by installing the GNU linker.
3411 It would be nice to extend GAS to produce the gp tables, but they are
3412 optional, and there should not be a warning about their absence.
3414 The libstdc++ atomic locking routines for MIPS targets requires MIPS II
3415 and later. A patch went in just after the GCC 3.3 release to
3416 make @samp{mips*-*-*} use the generic implementation instead. You can also
3417 configure for @samp{mipsel-elf} as a workaround. The
3418 @samp{mips*-*-linux*} target continues to use the MIPS II routines. More
3419 work on this is expected in future releases.
3421 @c If you make --with-llsc the default for another target, please also
3422 @c update the description of the --with-llsc option.
3424 The built-in @code{__sync_*} functions are available on MIPS II and
3425 later systems and others that support the @samp{ll}, @samp{sc} and
3426 @samp{sync} instructions. This can be overridden by passing
3427 @option{--with-llsc} or @option{--without-llsc} when configuring GCC.
3428 Since the Linux kernel emulates these instructions if they are
3429 missing, the default for @samp{mips*-*-linux*} targets is
3430 @option{--with-llsc}. The @option{--with-llsc} and
3431 @option{--without-llsc} configure options may be overridden at compile
3432 time by passing the @option{-mllsc} or @option{-mno-llsc} options to
3435 MIPS systems check for division by zero (unless
3436 @option{-mno-check-zero-division} is passed to the compiler) by
3437 generating either a conditional trap or a break instruction. Using
3438 trap results in smaller code, but is only supported on MIPS II and
3439 later. Also, some versions of the Linux kernel have a bug that
3440 prevents trap from generating the proper signal (@code{SIGFPE}). To enable
3441 the use of break, use the @option{--with-divide=breaks}
3442 @command{configure} option when configuring GCC@. The default is to
3443 use traps on systems that support them.
3445 Cross-compilers for the MIPS as target using the MIPS assembler
3446 currently do not work, because the auxiliary programs
3447 @file{mips-tdump.c} and @file{mips-tfile.c} can't be compiled on
3448 anything but a MIPS@. It does work to cross compile for a MIPS
3449 if you use the GNU assembler and linker.
3451 The assembler from GNU binutils 2.17 and earlier has a bug in the way
3452 it sorts relocations for REL targets (o32, o64, EABI). This can cause
3453 bad code to be generated for simple C++ programs. Also the linker
3454 from GNU binutils versions prior to 2.17 has a bug which causes the
3455 runtime linker stubs in very large programs, like @file{libgcj.so}, to
3456 be incorrectly generated. GNU Binutils 2.18 and later (and snapshots
3457 made after Nov. 9, 2006) should be free from both of these problems.
3462 @heading @anchor{mips-sgi-irix5}mips-sgi-irix5
3464 In order to compile GCC on an SGI running IRIX 5, the @samp{compiler_dev.hdr}
3465 subsystem must be installed from the IDO CD-ROM supplied by SGI@.
3466 It is also available for download from
3467 @uref{ftp://ftp.sgi.com/sgi/IRIX5.3/iris-development-option-5.3.tardist}.
3469 If you use the MIPS C compiler to bootstrap, it may be necessary
3470 to increase its table size for switch statements with the
3471 @option{-Wf,-XNg1500} option. If you use the @option{-O2}
3472 optimization option, you also need to use @option{-Olimit 3000}.
3474 To enable debugging under IRIX 5, you must use GNU binutils 2.15 or
3475 later, and use the @option{--with-gnu-ld} @command{configure} option
3476 when configuring GCC@. You need to use GNU @command{ar} and @command{nm},
3477 also distributed with GNU binutils.
3479 Some users have reported that @command{/bin/sh} will hang during bootstrap.
3480 This problem can be avoided by running the commands:
3483 % CONFIG_SHELL=/bin/ksh
3484 % export CONFIG_SHELL
3487 before starting the build.
3492 @heading @anchor{mips-sgi-irix6}mips-sgi-irix6
3494 If you are using SGI's MIPSpro @command{cc} as your bootstrap compiler, you must
3495 ensure that the N32 ABI is in use. To test this, compile a simple C
3496 file with @command{cc} and then run @command{file} on the
3497 resulting object file. The output should look like:
3500 test.o: ELF N32 MSB @dots{}
3506 test.o: ELF 32-bit MSB @dots{}
3512 test.o: ELF 64-bit MSB @dots{}
3515 then your version of @command{cc} uses the O32 or N64 ABI by default. You
3516 should set the environment variable @env{CC} to @samp{cc -n32}
3517 before configuring GCC@.
3519 If you want the resulting @command{gcc} to run on old 32-bit systems
3520 with the MIPS R4400 CPU, you need to ensure that only code for the @samp{mips3}
3521 instruction set architecture (ISA) is generated. While GCC 3.x does
3522 this correctly, both GCC 2.95 and SGI's MIPSpro @command{cc} may change
3523 the ISA depending on the machine where GCC is built. Using one of them
3524 as the bootstrap compiler may result in @samp{mips4} code, which won't run at
3525 all on @samp{mips3}-only systems. For the test program above, you should see:
3528 test.o: ELF N32 MSB mips-3 @dots{}
3534 test.o: ELF N32 MSB mips-4 @dots{}
3537 instead, you should set the environment variable @env{CC} to @samp{cc
3538 -n32 -mips3} or @samp{gcc -mips3} respectively before configuring GCC@.
3540 MIPSpro C 7.4 may cause bootstrap failures, due to a bug when inlining
3541 @code{memcmp}. Either add @code{-U__INLINE_INTRINSICS} to the @env{CC}
3542 environment variable as a workaround or upgrade to MIPSpro C 7.4.1m.
3544 GCC on IRIX 6 is usually built to support the N32, O32 and N64 ABIs. If
3545 you build GCC on a system that doesn't have the N64 libraries installed
3546 or cannot run 64-bit binaries,
3547 you need to configure with @option{--disable-multilib} so GCC doesn't
3548 try to use them. This will disable building the O32 libraries, too.
3549 Look for @file{/usr/lib64/libc.so.1} to see if you
3550 have the 64-bit libraries installed.
3552 To enable debugging for the O32 ABI, you must use GNU @command{as} from
3553 GNU binutils 2.15 or later. You may also use GNU @command{ld}, but
3554 this is not required and currently causes some problems with Ada.
3556 The @option{--enable-libgcj}
3557 option is disabled by default: IRIX 6 uses a very low default limit
3558 (20480) for the command line length. Although @command{libtool} contains a
3559 workaround for this problem, at least the N64 @samp{libgcj} is known not
3560 to build despite this, running into an internal error of the native
3561 @command{ld}. A sure fix is to increase this limit (@samp{ncargs}) to
3562 its maximum of 262144 bytes. If you have root access, you can use the
3563 @command{systune} command to do this.
3565 @code{wchar_t} support in @samp{libstdc++} is not available for old
3566 IRIX 6.5.x releases, @math{x < 19}. The problem cannot be autodetected
3567 and in order to build GCC for such targets you need to configure with
3568 @option{--disable-wchar_t}.
3570 See @uref{http://freeware.sgi.com/} for more
3571 information about using GCC on IRIX platforms.
3576 @heading @anchor{powerpc-x-x}powerpc-*-*
3578 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
3579 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
3584 @heading @anchor{powerpc-x-darwin}powerpc-*-darwin*
3585 PowerPC running Darwin (Mac OS X kernel).
3587 Pre-installed versions of Mac OS X may not include any developer tools,
3588 meaning that you will not be able to build GCC from source. Tool
3589 binaries are available at
3590 @uref{http://developer.apple.com/darwin/projects/compiler/} (free
3591 registration required).
3593 This version of GCC requires at least cctools-590.36. The
3594 cctools-590.36 package referenced from
3595 @uref{http://gcc.gnu.org/ml/gcc/2006-03/msg00507.html} will not work
3596 on systems older than 10.3.9 (aka darwin7.9.0).
3601 @heading @anchor{powerpc-x-elf}powerpc-*-elf
3602 PowerPC system in big endian mode, running System V.4.
3607 @heading @anchor{powerpc-x-linux-gnu}powerpc*-*-linux-gnu*
3610 @uref{ftp://ftp.kernel.org/pub/linux/devel/binutils,,binutils 2.15}
3611 or newer for a working GCC@.
3616 @heading @anchor{powerpc-x-netbsd}powerpc-*-netbsd*
3617 PowerPC system in big endian mode running NetBSD@. To build the
3618 documentation you will need Texinfo version 4.4 (NetBSD 1.5.1 included
3619 Texinfo version 3.12).
3624 @heading @anchor{powerpc-x-eabisim}powerpc-*-eabisim
3625 Embedded PowerPC system in big endian mode for use in running under the
3631 @heading @anchor{powerpc-x-eabi}powerpc-*-eabi
3632 Embedded PowerPC system in big endian mode.
3637 @heading @anchor{powerpcle-x-elf}powerpcle-*-elf
3638 PowerPC system in little endian mode, running System V.4.
3643 @heading @anchor{powerpcle-x-eabisim}powerpcle-*-eabisim
3644 Embedded PowerPC system in little endian mode for use in running under
3650 @heading @anchor{powerpcle-x-eabi}powerpcle-*-eabi
3651 Embedded PowerPC system in little endian mode.
3656 @heading @anchor{s390-x-linux}s390-*-linux*
3657 S/390 system running GNU/Linux for S/390@.
3662 @heading @anchor{s390x-x-linux}s390x-*-linux*
3663 zSeries system (64-bit) running GNU/Linux for zSeries@.
3668 @heading @anchor{s390x-ibm-tpf}s390x-ibm-tpf*
3669 zSeries system (64-bit) running TPF@. This platform is
3670 supported as cross-compilation target only.
3675 @c Please use Solaris 2 to refer to all release of Solaris, starting
3676 @c with 2.0 until 2.6, 7, 8, etc. Solaris 1 was a marketing name for
3677 @c SunOS 4 releases which we don't use to avoid confusion. Solaris
3678 @c alone is too unspecific and must be avoided.
3679 @heading @anchor{x-x-solaris2}*-*-solaris2*
3681 Sun does not ship a C compiler with Solaris 2. To bootstrap and install
3682 GCC you first have to install a pre-built compiler, see the
3683 @uref{binaries.html,,binaries page} for details.
3685 The Solaris 2 @command{/bin/sh} will often fail to configure
3686 @file{libstdc++-v3}, @file{boehm-gc} or @file{libjava}. We therefore
3687 recommend using the following initial sequence of commands
3690 % CONFIG_SHELL=/bin/ksh
3691 % export CONFIG_SHELL
3694 and proceed as described in @uref{configure.html,,the configure instructions}.
3695 In addition we strongly recommend specifying an absolute path to invoke
3696 @var{srcdir}/configure.
3698 Solaris 2 comes with a number of optional OS packages. Some of these
3699 are needed to use GCC fully, namely @code{SUNWarc},
3700 @code{SUNWbtool}, @code{SUNWesu}, @code{SUNWhea}, @code{SUNWlibm},
3701 @code{SUNWsprot}, and @code{SUNWtoo}. If you did not install all
3702 optional packages when installing Solaris 2, you will need to verify that
3703 the packages that GCC needs are installed.
3705 To check whether an optional package is installed, use
3706 the @command{pkginfo} command. To add an optional package, use the
3707 @command{pkgadd} command. For further details, see the Solaris 2
3710 Trying to use the linker and other tools in
3711 @file{/usr/ucb} to install GCC has been observed to cause trouble.
3712 For example, the linker may hang indefinitely. The fix is to remove
3713 @file{/usr/ucb} from your @env{PATH}.
3715 The build process works more smoothly with the legacy Sun tools so, if you
3716 have @file{/usr/xpg4/bin} in your @env{PATH}, we recommend that you place
3717 @file{/usr/bin} before @file{/usr/xpg4/bin} for the duration of the build.
3719 All releases of GNU binutils prior to 2.11.2 have known bugs on this
3720 platform. We recommend the use of GNU binutils 2.11.2 or later, or the
3721 vendor tools (Sun @command{as}, Sun @command{ld}). Note that your mileage
3722 may vary if you use a combination of the GNU tools and the Sun tools: while
3723 the combination GNU @command{as} + Sun @command{ld} should reasonably work,
3724 the reverse combination Sun @command{as} + GNU @command{ld} is known to
3725 cause memory corruption at runtime in some cases for C++ programs.
3727 The stock GNU binutils 2.15 release is broken on this platform because of a
3728 single bug. It has been fixed on the 2.15 branch in the CVS repository.
3729 You can obtain a working version by checking out the binutils-2_15-branch
3730 from the CVS repository or applying the patch
3731 @uref{http://sourceware.org/ml/binutils-cvs/2004-09/msg00036.html} to the
3734 We recommend using GNU binutils 2.16 or later in conjunction with GCC 4.x,
3735 or the vendor tools (Sun @command{as}, Sun @command{ld}). However, for
3736 Solaris 10 and above, an additional patch is required in order for the GNU
3737 linker to be able to cope with a new flavor of shared libraries. You
3738 can obtain a working version by checking out the binutils-2_16-branch from
3739 the CVS repository or applying the patch
3740 @uref{http://sourceware.org/ml/binutils-cvs/2005-07/msg00122.html} to the
3743 Sun bug 4296832 turns up when compiling X11 headers with GCC 2.95 or
3744 newer: @command{g++} will complain that types are missing. These headers assume
3745 that omitting the type means @code{int}; this assumption worked for C89 but
3746 is wrong for C++, and is now wrong for C99 also.
3748 @command{g++} accepts such (invalid) constructs with the option
3749 @option{-fpermissive}; it
3750 will assume that any missing type is @code{int} (as defined by C89).
3752 There are patches for Solaris 2.6 (105633-56 or newer for SPARC,
3753 106248-42 or newer for Intel), Solaris 7 (108376-21 or newer for SPARC,
3754 108377-20 for Intel), and Solaris 8 (108652-24 or newer for SPARC,
3755 108653-22 for Intel) that fix this bug.
3757 Sun bug 4927647 sometimes causes random spurious testsuite failures
3758 related to missing diagnostic output. This bug doesn't affect GCC
3759 itself, rather it is a kernel bug triggered by the @command{expect}
3760 program which is used only by the GCC testsuite driver. When the bug
3761 causes the @command{expect} program to miss anticipated output, extra
3762 testsuite failures appear.
3764 There are patches for Solaris 8 (117350-12 or newer for SPARC,
3765 117351-12 or newer for Intel) and Solaris 9 (117171-11 or newer for
3766 SPARC, 117172-11 or newer for Intel) that address this problem.
3771 @heading @anchor{sparc-sun-solaris2}sparc-sun-solaris2*
3773 When GCC is configured to use binutils 2.11.2 or later the binaries
3774 produced are smaller than the ones produced using Sun's native tools;
3775 this difference is quite significant for binaries containing debugging
3778 Sun @command{as} 4.x is broken in that it cannot cope with long symbol names.
3779 A typical error message might look similar to the following:
3782 /usr/ccs/bin/as: "/var/tmp/ccMsw135.s", line 11041: error:
3783 can't compute value of an expression involving an external symbol.
3786 This is Sun bug 4237974. This is fixed with patch 108908-02 for Solaris
3787 2.6 and has been fixed in later (5.x) versions of the assembler,
3788 starting with Solaris 7.
3790 Starting with Solaris 7, the operating system is capable of executing
3791 64-bit SPARC V9 binaries. GCC 3.1 and later properly supports
3792 this; the @option{-m64} option enables 64-bit code generation.
3793 However, if all you want is code tuned for the UltraSPARC CPU, you
3794 should try the @option{-mtune=ultrasparc} option instead, which produces
3795 code that, unlike full 64-bit code, can still run on non-UltraSPARC
3798 When configuring on a Solaris 7 or later system that is running a kernel
3799 that supports only 32-bit binaries, one must configure with
3800 @option{--disable-multilib}, since we will not be able to build the
3801 64-bit target libraries.
3803 GCC 3.3 and GCC 3.4 trigger code generation bugs in earlier versions of
3804 the GNU compiler (especially GCC 3.0.x versions), which lead to the
3805 miscompilation of the stage1 compiler and the subsequent failure of the
3806 bootstrap process. A workaround is to use GCC 3.2.3 as an intermediary
3807 stage, i.e.@: to bootstrap that compiler with the base compiler and then
3808 use it to bootstrap the final compiler.
3810 GCC 3.4 triggers a code generation bug in versions 5.4 (Sun ONE Studio 7)
3811 and 5.5 (Sun ONE Studio 8) of the Sun compiler, which causes a bootstrap
3812 failure in form of a miscompilation of the stage1 compiler by the Sun
3813 compiler. This is Sun bug 4974440. This is fixed with patch 112760-07.
3815 GCC 3.4 changed the default debugging format from STABS to DWARF-2 for
3816 32-bit code on Solaris 7 and later. If you use the Sun assembler, this
3817 change apparently runs afoul of Sun bug 4910101 (which is referenced as
3818 a x86-only problem by Sun, probably because they do not use DWARF-2).
3819 A symptom of the problem is that you cannot compile C++ programs like
3820 @command{groff} 1.19.1 without getting messages similar to the following:
3823 ld: warning: relocation error: R_SPARC_UA32: @dots{}
3824 external symbolic relocation against non-allocatable section
3825 .debug_info cannot be processed at runtime: relocation ignored.
3828 To work around this problem, compile with @option{-gstabs+} instead of
3831 When configuring the GNU Multiple Precision Library (GMP) or the MPFR
3832 library on a Solaris 7 or later system, the canonical target triplet
3833 must be specified as the @command{build} parameter on the configure
3834 line. This triplet can be obtained by invoking ./config.guess in
3835 the toplevel source directory of GCC (and not that of GMP or MPFR).
3836 For example on a Solaris 7 system:
3839 % ./configure --build=sparc-sun-solaris2.7 --prefix=xxx
3845 @heading @anchor{sparc-sun-solaris27}sparc-sun-solaris2.7
3847 Sun patch 107058-01 (1999-01-13) for Solaris 7/SPARC triggers a bug in
3848 the dynamic linker. This problem (Sun bug 4210064) affects GCC 2.8
3849 and later, including all EGCS releases. Sun formerly recommended
3850 107058-01 for all Solaris 7 users, but around 1999-09-01 it started to
3851 recommend it only for people who use Sun's compilers.
3853 Here are some workarounds to this problem:
3856 Do not install Sun patch 107058-01 until after Sun releases a
3857 complete patch for bug 4210064. This is the simplest course to take,
3858 unless you must also use Sun's C compiler. Unfortunately 107058-01
3859 is preinstalled on some new Solaris 7-based hosts, so you may have to
3863 Copy the original, unpatched Solaris 7
3864 @command{/usr/ccs/bin/as} into
3865 @command{/usr/local/libexec/gcc/sparc-sun-solaris2.7/3.4/as},
3866 adjusting the latter name to fit your local conventions and software
3870 Install Sun patch 106950-03 (1999-05-25) or later. Nobody with
3871 both 107058-01 and 106950-03 installed has reported the bug with GCC
3872 and Sun's dynamic linker. This last course of action is riskiest,
3873 for two reasons. First, you must install 106950 on all hosts that
3874 run code generated by GCC; it doesn't suffice to install it only on
3875 the hosts that run GCC itself. Second, Sun says that 106950-03 is
3876 only a partial fix for bug 4210064, but Sun doesn't know whether the
3877 partial fix is adequate for GCC@. Revision -08 or later should fix
3878 the bug. The current (as of 2004-05-23) revision is -24, and is included in
3879 the Solaris 7 Recommended Patch Cluster.
3882 GCC 3.3 triggers a bug in version 5.0 Alpha 03/27/98 of the Sun assembler,
3883 which causes a bootstrap failure when linking the 64-bit shared version of
3884 libgcc. A typical error message is:
3887 ld: fatal: relocation error: R_SPARC_32: file libgcc/sparcv9/_muldi3.o:
3888 symbol <unknown>: offset 0xffffffff7ec133e7 is non-aligned.
3891 This bug has been fixed in the final 5.0 version of the assembler.
3893 A similar problem was reported for version Sun WorkShop 6 99/08/18 of the
3894 Sun assembler, which causes a bootstrap failure with GCC 4.0.0:
3897 ld: fatal: relocation error: R_SPARC_DISP32:
3898 file .libs/libstdc++.lax/libsupc++convenience.a/vterminate.o:
3899 symbol <unknown>: offset 0xfccd33ad is non-aligned
3902 This bug has been fixed in more recent revisions of the assembler.
3907 @heading @anchor{sparc-x-linux}sparc-*-linux*
3909 GCC versions 3.0 and higher require binutils 2.11.2 and glibc 2.2.4
3910 or newer on this platform. All earlier binutils and glibc
3911 releases mishandled unaligned relocations on @code{sparc-*-*} targets.
3917 @heading @anchor{sparc64-x-solaris2}sparc64-*-solaris2*
3919 When configuring the GNU Multiple Precision Library (GMP) or the
3920 MPFR library, the canonical target triplet must be specified as
3921 the @command{build} parameter on the configure line. For example
3922 on a Solaris 7 system:
3925 % ./configure --build=sparc64-sun-solaris2.7 --prefix=xxx
3928 The following compiler flags must be specified in the configure
3929 step in order to bootstrap this target with the Sun compiler:
3932 % CC="cc -xarch=v9 -xildoff" @var{srcdir}/configure [@var{options}] [@var{target}]
3935 @option{-xarch=v9} specifies the SPARC-V9 architecture to the Sun toolchain
3936 and @option{-xildoff} turns off the incremental linker.
3941 @heading @anchor{sparcv9-x-solaris2}sparcv9-*-solaris2*
3943 This is a synonym for sparc64-*-solaris2*.
3948 @heading @anchor{x-x-vxworks}*-*-vxworks*
3949 Support for VxWorks is in flux. At present GCC supports @emph{only} the
3950 very recent VxWorks 5.5 (aka Tornado 2.2) release, and only on PowerPC@.
3951 We welcome patches for other architectures supported by VxWorks 5.5.
3952 Support for VxWorks AE would also be welcome; we believe this is merely
3953 a matter of writing an appropriate ``configlette'' (see below). We are
3954 not interested in supporting older, a.out or COFF-based, versions of
3957 VxWorks comes with an older version of GCC installed in
3958 @file{@var{$WIND_BASE}/host}; we recommend you do not overwrite it.
3959 Choose an installation @var{prefix} entirely outside @var{$WIND_BASE}.
3960 Before running @command{configure}, create the directories @file{@var{prefix}}
3961 and @file{@var{prefix}/bin}. Link or copy the appropriate assembler,
3962 linker, etc.@: into @file{@var{prefix}/bin}, and set your @var{PATH} to
3963 include that directory while running both @command{configure} and
3966 You must give @command{configure} the
3967 @option{--with-headers=@var{$WIND_BASE}/target/h} switch so that it can
3968 find the VxWorks system headers. Since VxWorks is a cross compilation
3969 target only, you must also specify @option{--target=@var{target}}.
3970 @command{configure} will attempt to create the directory
3971 @file{@var{prefix}/@var{target}/sys-include} and copy files into it;
3972 make sure the user running @command{configure} has sufficient privilege
3975 GCC's exception handling runtime requires a special ``configlette''
3976 module, @file{contrib/gthr_supp_vxw_5x.c}. Follow the instructions in
3977 that file to add the module to your kernel build. (Future versions of
3978 VxWorks will incorporate this module.)
3983 @heading @anchor{x86-64-x-x}x86_64-*-*, amd64-*-*
3985 GCC supports the x86-64 architecture implemented by the AMD64 processor
3986 (amd64-*-* is an alias for x86_64-*-*) on GNU/Linux, FreeBSD and NetBSD@.
3987 On GNU/Linux the default is a bi-arch compiler which is able to generate
3988 both 64-bit x86-64 and 32-bit x86 code (via the @option{-m32} switch).
3993 @heading @anchor{xtensa-x-elf}xtensa-*-elf
3995 This target is intended for embedded Xtensa systems using the
3996 @samp{newlib} C library. It uses ELF but does not support shared
3997 objects. Designed-defined instructions specified via the
3998 Tensilica Instruction Extension (TIE) language are only supported
3999 through inline assembly.
4001 The Xtensa configuration information must be specified prior to
4002 building GCC@. The @file{include/xtensa-config.h} header
4003 file contains the configuration information. If you created your
4004 own Xtensa configuration with the Xtensa Processor Generator, the
4005 downloaded files include a customized copy of this header file,
4006 which you can use to replace the default header file.
4011 @heading @anchor{xtensa-x-linux}xtensa-*-linux*
4013 This target is for Xtensa systems running GNU/Linux. It supports ELF
4014 shared objects and the GNU C library (glibc). It also generates
4015 position-independent code (PIC) regardless of whether the
4016 @option{-fpic} or @option{-fPIC} options are used. In other
4017 respects, this target is the same as the
4018 @uref{#xtensa-*-elf,,@samp{xtensa-*-elf}} target.
4023 @heading @anchor{windows}Microsoft Windows (32-bit)
4025 Ports of GCC are included with the
4026 @uref{http://www.cygwin.com/,,Cygwin environment}.
4028 GCC will build under Cygwin without modification; it does not build
4029 with Microsoft's C++ compiler and there are no plans to make it do so.
4031 For MinGW, GCC will build with and support only MinGW runtime 3.12 and later.
4032 Earlier versions of headers are incompatible with the new default semantics
4033 of @code{extern inline} in @code{-std=c99} and @code{-std=gnu99} modes.
4038 @heading @anchor{os2}OS/2
4040 GCC does not currently support OS/2. However, Andrew Zabolotny has been
4041 working on a generic OS/2 port with pgcc. The current code can be found
4042 at @uref{http://www.goof.com/pcg/os2/,,http://www.goof.com/pcg/os2/}.
4047 @heading @anchor{older}Older systems
4049 GCC contains support files for many older (1980s and early
4050 1990s) Unix variants. For the most part, support for these systems
4051 has not been deliberately removed, but it has not been maintained for
4052 several years and may suffer from bitrot.
4054 Starting with GCC 3.1, each release has a list of ``obsoleted'' systems.
4055 Support for these systems is still present in that release, but
4056 @command{configure} will fail unless the @option{--enable-obsolete}
4057 option is given. Unless a maintainer steps forward, support for these
4058 systems will be removed from the next release of GCC@.
4060 Support for old systems as hosts for GCC can cause problems if the
4061 workarounds for compiler, library and operating system bugs affect the
4062 cleanliness or maintainability of the rest of GCC@. In some cases, to
4063 bring GCC up on such a system, if still possible with current GCC, may
4064 require first installing an old version of GCC which did work on that
4065 system, and using it to compile a more recent GCC, to avoid bugs in the
4066 vendor compiler. Old releases of GCC 1 and GCC 2 are available in the
4067 @file{old-releases} directory on the @uref{../mirrors.html,,GCC mirror
4068 sites}. Header bugs may generally be avoided using
4069 @command{fixincludes}, but bugs or deficiencies in libraries and the
4070 operating system may still cause problems.
4072 Support for older systems as targets for cross-compilation is less
4073 problematic than support for them as hosts for GCC; if an enthusiast
4074 wishes to make such a target work again (including resurrecting any of
4075 the targets that never worked with GCC 2, starting from the last
4076 version before they were removed), patches
4077 @uref{../contribute.html,,following the usual requirements} would be
4078 likely to be accepted, since they should not affect the support for more
4081 For some systems, old versions of GNU binutils may also be useful,
4082 and are available from @file{pub/binutils/old-releases} on
4083 @uref{http://sourceware.org/mirrors.html,,sourceware.org mirror sites}.
4085 Some of the information on specific systems above relates to
4086 such older systems, but much of the information
4087 about GCC on such systems (which may no longer be applicable to
4088 current GCC) is to be found in the GCC texinfo manual.
4093 @heading @anchor{elf}all ELF targets (SVR4, Solaris 2, etc.)
4095 C++ support is significantly better on ELF targets if you use the
4096 @uref{./configure.html#with-gnu-ld,,GNU linker}; duplicate copies of
4097 inlines, vtables and template instantiations will be discarded
4106 @uref{./index.html,,Return to the GCC Installation page}
4110 @c ***Old documentation******************************************************
4112 @include install-old.texi
4118 @uref{./index.html,,Return to the GCC Installation page}
4122 @c ***GFDL********************************************************************
4130 @uref{./index.html,,Return to the GCC Installation page}
4134 @c ***************************************************************************
4135 @c Part 6 The End of the Document
4137 @comment node-name, next, previous, up
4138 @node Concept Index, , GNU Free Documentation License, Top
4142 @unnumbered Concept Index