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 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 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 3.79.1 (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 2.2.1 (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 GNU Bison version 1.28 (or later)
377 Necessary when modifying @file{*.y} files. Necessary to build the
378 @code{treelang} front end (which is not enabled by default) from a
379 checkout of the SVN repository; the generated files are not in the
380 repository. They are included in releases.
382 Berkeley @command{yacc} (@command{byacc}) has been reported to work
385 @item Flex version 2.5.4 (or later)
387 Necessary when modifying @file{*.l} files.
389 Necessary to build GCC during development because the generated output
390 files are not included in the SVN repository. They are included in
393 @item Texinfo version 4.4 (or later)
395 Necessary for running @command{makeinfo} when modifying @file{*.texi}
396 files to test your changes.
398 Necessary for running @command{make dvi} or @command{make pdf} to
399 create printable documentation in DVI or PDF format. Texinfo version
400 4.8 or later is required for @command{make pdf}.
402 Necessary to build GCC documentation during development because the
403 generated output files are not included in the SVN repository. They are
404 included in releases.
406 @item @TeX{} (any working version)
408 Necessary for running @command{texi2dvi} and @command{texi2pdf}, which
409 are used when running @command{make dvi} or @command{make pdf} to create
410 DVI or PDF files, respectively.
412 @item SVN (any version)
413 @itemx SSH (any version)
415 Necessary to access the SVN repository. Public releases and weekly
416 snapshots of the development sources are also available via FTP@.
418 @item Perl version 5.6.1 (or later)
420 Necessary when regenerating @file{Makefile} dependencies in libiberty.
421 Necessary when regenerating @file{libiberty/functions.texi}.
422 Necessary when generating manpages from Texinfo manuals.
423 Necessary when targetting Darwin, building libstdc++,
424 and not using @option{--disable-symvers}.
425 Used by various scripts to generate some files included in SVN (mainly
426 Unicode-related and rarely changing) from source tables.
428 @item GNU diffutils version 2.7 (or later)
430 Useful when submitting patches for the GCC source code.
432 @item patch version 2.5.4 (or later)
434 Necessary when applying patches, created with @command{diff}, to one's
440 If you wish to modify @file{.java} files in libjava, you will need to
441 configure with @option{--enable-java-maintainer-mode}, and you will need
442 to have executables named @command{ecj1} and @command{gjavah} in your path.
443 The @command{ecj1} executable should run the Eclipse Java compiler via
444 the GCC-specific entry point. You can download a suitable jar from
445 @uref{ftp://sourceware.org/pub/java/}, or by running the script
446 @command{contrib/download_ecj}.
455 @uref{./index.html,,Return to the GCC Installation page}
459 @c ***Downloading the source**************************************************
461 @comment node-name, next, previous, up
462 @node Downloading the source, Configuration, Prerequisites, Installing GCC
466 @chapter Downloading GCC
468 @cindex Downloading GCC
469 @cindex Downloading the Source
471 GCC is distributed via @uref{http://gcc.gnu.org/svn.html,,SVN} and FTP
472 tarballs compressed with @command{gzip} or
473 @command{bzip2}. It is possible to download a full distribution or specific
476 Please refer to the @uref{http://gcc.gnu.org/releases.html,,releases web page}
477 for information on how to obtain GCC@.
479 The full distribution includes the C, C++, Objective-C, Fortran, Java,
480 and Ada (in the case of GCC 3.1 and later) compilers. The full
481 distribution also includes runtime libraries for C++, Objective-C,
482 Fortran, and Java. In GCC 3.0 and later versions, the GNU compiler
483 testsuites are also included in the full distribution.
485 If you choose to download specific components, you must download the core
486 GCC distribution plus any language specific distributions you wish to
487 use. The core distribution includes the C language front end as well as the
488 shared components. Each language has a tarball which includes the language
489 front end as well as the language runtime (when appropriate).
491 Unpack the core distribution as well as any language specific
492 distributions in the same directory.
494 If you also intend to build binutils (either to upgrade an existing
495 installation or for use in place of the corresponding tools of your
496 OS), unpack the binutils distribution either in the same directory or
497 a separate one. In the latter case, add symbolic links to any
498 components of the binutils you intend to build alongside the compiler
499 (@file{bfd}, @file{binutils}, @file{gas}, @file{gprof}, @file{ld},
500 @file{opcodes}, @dots{}) to the directory containing the GCC sources.
507 @uref{./index.html,,Return to the GCC Installation page}
511 @c ***Configuration***********************************************************
513 @comment node-name, next, previous, up
514 @node Configuration, Building, Downloading the source, Installing GCC
518 @chapter Installing GCC: Configuration
520 @cindex Configuration
521 @cindex Installing GCC: Configuration
523 Like most GNU software, GCC must be configured before it can be built.
524 This document describes the recommended configuration procedure
525 for both native and cross targets.
527 We use @var{srcdir} to refer to the toplevel source directory for
528 GCC; we use @var{objdir} to refer to the toplevel build/object directory.
530 If you obtained the sources via SVN, @var{srcdir} must refer to the top
531 @file{gcc} directory, the one where the @file{MAINTAINERS} can be found,
532 and not its @file{gcc} subdirectory, otherwise the build will fail.
534 If either @var{srcdir} or @var{objdir} is located on an automounted NFS
535 file system, the shell's built-in @command{pwd} command will return
536 temporary pathnames. Using these can lead to various sorts of build
537 problems. To avoid this issue, set the @env{PWDCMD} environment
538 variable to an automounter-aware @command{pwd} command, e.g.,
539 @command{pawd} or @samp{amq -w}, during the configuration and build
542 First, we @strong{highly} recommend that GCC be built into a
543 separate directory than the sources which does @strong{not} reside
544 within the source tree. This is how we generally build GCC; building
545 where @var{srcdir} == @var{objdir} should still work, but doesn't
546 get extensive testing; building where @var{objdir} is a subdirectory
547 of @var{srcdir} is unsupported.
549 If you have previously built GCC in the same directory for a
550 different target machine, do @samp{make distclean} to delete all files
551 that might be invalid. One of the files this deletes is @file{Makefile};
552 if @samp{make distclean} complains that @file{Makefile} does not exist
553 or issues a message like ``don't know how to make distclean'' it probably
554 means that the directory is already suitably clean. However, with the
555 recommended method of building in a separate @var{objdir}, you should
556 simply use a different @var{objdir} for each target.
558 Second, when configuring a native system, either @command{cc} or
559 @command{gcc} must be in your path or you must set @env{CC} in
560 your environment before running configure. Otherwise the configuration
564 Note that the bootstrap compiler and the resulting GCC must be link
565 compatible, else the bootstrap will fail with linker errors about
566 incompatible object file formats. Several multilibed targets are
567 affected by this requirement, see
569 @ref{Specific, host/target specific installation notes}.
572 @uref{specific.html,,host/target specific installation notes}.
581 % @var{srcdir}/configure [@var{options}] [@var{target}]
584 @heading Distributor options
586 If you will be distributing binary versions of GCC, with modifications
587 to the source code, you should use the options described in this
588 section to make clear that your version contains modifications.
591 @item --with-pkgversion=@var{version}
592 Specify a string that identifies your package. You may wish
593 to include a build number or build date. This version string will be
594 included in the output of @command{gcc --version}. This suffix does
595 not replace the default version string, only the @samp{GCC} part.
597 The default value is @samp{GCC}.
599 @item --with-bugurl=@var{url}
600 Specify the URL that users should visit if they wish to report a bug.
601 You are of course welcome to forward bugs reported to you to the FSF,
602 if you determine that they are not bugs in your modifications.
604 The default value refers to the FSF's GCC bug tracker.
608 @heading Target specification
611 GCC has code to correctly determine the correct value for @var{target}
612 for nearly all native systems. Therefore, we highly recommend you not
613 provide a configure target when configuring a native compiler.
616 @var{target} must be specified as @option{--target=@var{target}}
617 when configuring a cross compiler; examples of valid targets would be
618 m68k-coff, sh-elf, etc.
621 Specifying just @var{target} instead of @option{--target=@var{target}}
622 implies that the host defaults to @var{target}.
626 @heading Options specification
628 Use @var{options} to override several configure time options for
629 GCC@. A list of supported @var{options} follows; @samp{configure
630 --help} may list other options, but those not listed below may not
631 work and should not normally be used.
633 Note that each @option{--enable} option has a corresponding
634 @option{--disable} option and that each @option{--with} option has a
635 corresponding @option{--without} option.
638 @item --prefix=@var{dirname}
639 Specify the toplevel installation
640 directory. This is the recommended way to install the tools into a directory
641 other than the default. The toplevel installation directory defaults to
644 We @strong{highly} recommend against @var{dirname} being the same or a
645 subdirectory of @var{objdir} or vice versa. If specifying a directory
646 beneath a user's home directory tree, some shells will not expand
647 @var{dirname} correctly if it contains the @samp{~} metacharacter; use
650 The following standard @command{autoconf} options are supported. Normally you
651 should not need to use these options.
653 @item --exec-prefix=@var{dirname}
654 Specify the toplevel installation directory for architecture-dependent
655 files. The default is @file{@var{prefix}}.
657 @item --bindir=@var{dirname}
658 Specify the installation directory for the executables called by users
659 (such as @command{gcc} and @command{g++}). The default is
660 @file{@var{exec-prefix}/bin}.
662 @item --libdir=@var{dirname}
663 Specify the installation directory for object code libraries and
664 internal data files of GCC@. The default is @file{@var{exec-prefix}/lib}.
666 @item --libexecdir=@var{dirname}
667 Specify the installation directory for internal executables of GCC@.
668 The default is @file{@var{exec-prefix}/libexec}.
670 @item --with-slibdir=@var{dirname}
671 Specify the installation directory for the shared libgcc library. The
672 default is @file{@var{libdir}}.
674 @item --infodir=@var{dirname}
675 Specify the installation directory for documentation in info format.
676 The default is @file{@var{prefix}/info}.
678 @item --datadir=@var{dirname}
679 Specify the installation directory for some architecture-independent
680 data files referenced by GCC@. The default is @file{@var{prefix}/share}.
682 @item --mandir=@var{dirname}
683 Specify the installation directory for manual pages. The default is
684 @file{@var{prefix}/man}. (Note that the manual pages are only extracts from
685 the full GCC manuals, which are provided in Texinfo format. The manpages
686 are derived by an automatic conversion process from parts of the full
689 @item --with-gxx-include-dir=@var{dirname}
691 the installation directory for G++ header files. The default is
692 @file{@var{prefix}/include/c++/@var{version}}.
696 @item --program-prefix=@var{prefix}
697 GCC supports some transformations of the names of its programs when
698 installing them. This option prepends @var{prefix} to the names of
699 programs to install in @var{bindir} (see above). For example, specifying
700 @option{--program-prefix=foo-} would result in @samp{gcc}
701 being installed as @file{/usr/local/bin/foo-gcc}.
703 @item --program-suffix=@var{suffix}
704 Appends @var{suffix} to the names of programs to install in @var{bindir}
705 (see above). For example, specifying @option{--program-suffix=-3.1}
706 would result in @samp{gcc} being installed as
707 @file{/usr/local/bin/gcc-3.1}.
709 @item --program-transform-name=@var{pattern}
710 Applies the @samp{sed} script @var{pattern} to be applied to the names
711 of programs to install in @var{bindir} (see above). @var{pattern} has to
712 consist of one or more basic @samp{sed} editing commands, separated by
713 semicolons. For example, if you want the @samp{gcc} program name to be
714 transformed to the installed program @file{/usr/local/bin/myowngcc} and
715 the @samp{g++} program name to be transformed to
716 @file{/usr/local/bin/gspecial++} without changing other program names,
717 you could use the pattern
718 @option{--program-transform-name='s/^gcc$/myowngcc/; s/^g++$/gspecial++/'}
719 to achieve this effect.
721 All three options can be combined and used together, resulting in more
722 complex conversion patterns. As a basic rule, @var{prefix} (and
723 @var{suffix}) are prepended (appended) before further transformations
724 can happen with a special transformation script @var{pattern}.
726 As currently implemented, this option only takes effect for native
727 builds; cross compiler binaries' names are not transformed even when a
728 transformation is explicitly asked for by one of these options.
730 For native builds, some of the installed programs are also installed
731 with the target alias in front of their name, as in
732 @samp{i686-pc-linux-gnu-gcc}. All of the above transformations happen
733 before the target alias is prepended to the name---so, specifying
734 @option{--program-prefix=foo-} and @option{program-suffix=-3.1}, the
735 resulting binary would be installed as
736 @file{/usr/local/bin/i686-pc-linux-gnu-foo-gcc-3.1}.
738 As a last shortcoming, none of the installed Ada programs are
739 transformed yet, which will be fixed in some time.
741 @item --with-local-prefix=@var{dirname}
743 installation directory for local include files. The default is
744 @file{/usr/local}. Specify this option if you want the compiler to
745 search directory @file{@var{dirname}/include} for locally installed
746 header files @emph{instead} of @file{/usr/local/include}.
748 You should specify @option{--with-local-prefix} @strong{only} if your
749 site has a different convention (not @file{/usr/local}) for where to put
752 The default value for @option{--with-local-prefix} is @file{/usr/local}
753 regardless of the value of @option{--prefix}. Specifying
754 @option{--prefix} has no effect on which directory GCC searches for
755 local header files. This may seem counterintuitive, but actually it is
758 The purpose of @option{--prefix} is to specify where to @emph{install
759 GCC}. The local header files in @file{/usr/local/include}---if you put
760 any in that directory---are not part of GCC@. They are part of other
761 programs---perhaps many others. (GCC installs its own header files in
762 another directory which is based on the @option{--prefix} value.)
764 Both the local-prefix include directory and the GCC-prefix include
765 directory are part of GCC's ``system include'' directories. Although these
766 two directories are not fixed, they need to be searched in the proper
767 order for the correct processing of the include_next directive. The
768 local-prefix include directory is searched before the GCC-prefix
769 include directory. Another characteristic of system include directories
770 is that pedantic warnings are turned off for headers in these directories.
772 Some autoconf macros add @option{-I @var{directory}} options to the
773 compiler command line, to ensure that directories containing installed
774 packages' headers are searched. When @var{directory} is one of GCC's
775 system include directories, GCC will ignore the option so that system
776 directories continue to be processed in the correct order. This
777 may result in a search order different from what was specified but the
778 directory will still be searched.
780 GCC automatically searches for ordinary libraries using
781 @env{GCC_EXEC_PREFIX}. Thus, when the same installation prefix is
782 used for both GCC and packages, GCC will automatically search for
783 both headers and libraries. This provides a configuration that is
784 easy to use. GCC behaves in a manner similar to that when it is
785 installed as a system compiler in @file{/usr}.
787 Sites that need to install multiple versions of GCC may not want to
788 use the above simple configuration. It is possible to use the
789 @option{--program-prefix}, @option{--program-suffix} and
790 @option{--program-transform-name} options to install multiple versions
791 into a single directory, but it may be simpler to use different prefixes
792 and the @option{--with-local-prefix} option to specify the location of the
793 site-specific files for each version. It will then be necessary for
794 users to specify explicitly the location of local site libraries
795 (e.g., with @env{LIBRARY_PATH}).
797 The same value can be used for both @option{--with-local-prefix} and
798 @option{--prefix} provided it is not @file{/usr}. This can be used
799 to avoid the default search of @file{/usr/local/include}.
801 @strong{Do not} specify @file{/usr} as the @option{--with-local-prefix}!
802 The directory you use for @option{--with-local-prefix} @strong{must not}
803 contain any of the system's standard header files. If it did contain
804 them, certain programs would be miscompiled (including GNU Emacs, on
805 certain targets), because this would override and nullify the header
806 file corrections made by the @command{fixincludes} script.
808 Indications are that people who use this option use it based on mistaken
809 ideas of what it is for. People use it as if it specified where to
810 install part of GCC@. Perhaps they make this assumption because
811 installing GCC creates the directory.
813 @item --enable-shared[=@var{package}[,@dots{}]]
814 Build shared versions of libraries, if shared libraries are supported on
815 the target platform. Unlike GCC 2.95.x and earlier, shared libraries
816 are enabled by default on all platforms that support shared libraries.
818 If a list of packages is given as an argument, build shared libraries
819 only for the listed packages. For other packages, only static libraries
820 will be built. Package names currently recognized in the GCC tree are
821 @samp{libgcc} (also known as @samp{gcc}), @samp{libstdc++} (not
822 @samp{libstdc++-v3}), @samp{libffi}, @samp{zlib}, @samp{boehm-gc},
823 @samp{ada}, @samp{libada}, @samp{libjava} and @samp{libobjc}.
824 Note @samp{libiberty} does not support shared libraries at all.
826 Use @option{--disable-shared} to build only static libraries. Note that
827 @option{--disable-shared} does not accept a list of package names as
828 argument, only @option{--enable-shared} does.
830 @item @anchor{with-gnu-as}--with-gnu-as
831 Specify that the compiler should assume that the
832 assembler it finds is the GNU assembler. However, this does not modify
833 the rules to find an assembler and will result in confusion if the
834 assembler found is not actually the GNU assembler. (Confusion may also
835 result if the compiler finds the GNU assembler but has not been
836 configured with @option{--with-gnu-as}.) If you have more than one
837 assembler installed on your system, you may want to use this option in
838 connection with @option{--with-as=@var{pathname}} or
839 @option{--with-build-time-tools=@var{pathname}}.
841 The following systems are the only ones where it makes a difference
842 whether you use the GNU assembler. On any other system,
843 @option{--with-gnu-as} has no effect.
846 @item @samp{hppa1.0-@var{any}-@var{any}}
847 @item @samp{hppa1.1-@var{any}-@var{any}}
848 @item @samp{i386-@var{any}-sysv}
849 @item @samp{m68k-bull-sysv}
850 @item @samp{m68k-hp-hpux}
851 @item @samp{m68000-hp-hpux}
852 @item @samp{m68000-att-sysv}
853 @item @samp{sparc-sun-solaris2.@var{any}}
854 @item @samp{sparc64-@var{any}-solaris2.@var{any}}
857 On the systems listed above (except for the HP-PA, the SPARC, for ISC on
858 the 386, if you use the GNU assembler, you should also use the GNU linker
859 (and specify @option{--with-gnu-ld}).
861 @item @anchor{with-as}--with-as=@var{pathname}
862 Specify that the compiler should use the assembler pointed to by
863 @var{pathname}, rather than the one found by the standard rules to find
864 an assembler, which are:
867 Unless GCC is being built with a cross compiler, check the
868 @file{@var{libexec}/gcc/@var{target}/@var{version}} directory.
869 @var{libexec} defaults to @file{@var{exec-prefix}/libexec};
870 @var{exec-prefix} defaults to @var{prefix}, which
871 defaults to @file{/usr/local} unless overridden by the
872 @option{--prefix=@var{pathname}} switch described above. @var{target}
873 is the target system triple, such as @samp{sparc-sun-solaris2.7}, and
874 @var{version} denotes the GCC version, such as 3.0.
877 If the target system is the same that you are building on, check
878 operating system specific directories (e.g.@: @file{/usr/ccs/bin} on
882 Check in the @env{PATH} for a tool whose name is prefixed by the
883 target system triple.
886 Check in the @env{PATH} for a tool whose name is not prefixed by the
887 target system triple, if the host and target system triple are
888 the same (in other words, we use a host tool if it can be used for
892 You may want to use @option{--with-as} if no assembler
893 is installed in the directories listed above, or if you have multiple
894 assemblers installed and want to choose one that is not found by the
897 @item @anchor{with-gnu-ld}--with-gnu-ld
898 Same as @uref{#with-gnu-as,,@option{--with-gnu-as}}
901 @item --with-ld=@var{pathname}
902 Same as @uref{#with-as,,@option{--with-as}}
906 Specify that stabs debugging
907 information should be used instead of whatever format the host normally
908 uses. Normally GCC uses the same debug format as the host system.
910 On MIPS based systems and on Alphas, you must specify whether you want
911 GCC to create the normal ECOFF debugging format, or to use BSD-style
912 stabs passed through the ECOFF symbol table. The normal ECOFF debug
913 format cannot fully handle languages other than C@. BSD stabs format can
914 handle other languages, but it only works with the GNU debugger GDB@.
916 Normally, GCC uses the ECOFF debugging format by default; if you
917 prefer BSD stabs, specify @option{--with-stabs} when you configure GCC@.
919 No matter which default you choose when you configure GCC, the user
920 can use the @option{-gcoff} and @option{-gstabs+} options to specify explicitly
921 the debug format for a particular compilation.
923 @option{--with-stabs} is meaningful on the ISC system on the 386, also, if
924 @option{--with-gas} is used. It selects use of stabs debugging
925 information embedded in COFF output. This kind of debugging information
926 supports C++ well; ordinary COFF debugging information does not.
928 @option{--with-stabs} is also meaningful on 386 systems running SVR4. It
929 selects use of stabs debugging information embedded in ELF output. The
930 C++ compiler currently (2.6.0) does not support the DWARF debugging
931 information normally used on 386 SVR4 platforms; stabs provide a
932 workable alternative. This requires gas and gdb, as the normal SVR4
933 tools can not generate or interpret stabs.
935 @item --disable-multilib
936 Specify that multiple target
937 libraries to support different target variants, calling
938 conventions, etc.@: should not be built. The default is to build a
939 predefined set of them.
941 Some targets provide finer-grained control over which multilibs are built
942 (e.g., @option{--disable-softfloat}):
948 fpu, 26bit, underscore, interwork, biendian, nofmult.
951 softfloat, m68881, m68000, m68020.
954 single-float, biendian, softfloat.
956 @item powerpc*-*-*, rs6000*-*-*
957 aix64, pthread, softfloat, powercpu, powerpccpu, powerpcos, biendian,
962 @item --enable-threads
963 Specify that the target
964 supports threads. This affects the Objective-C compiler and runtime
965 library, and exception handling for other languages like C++ and Java.
966 On some systems, this is the default.
968 In general, the best (and, in many cases, the only known) threading
969 model available will be configured for use. Beware that on some
970 systems, GCC has not been taught what threading models are generally
971 available for the system. In this case, @option{--enable-threads} is an
972 alias for @option{--enable-threads=single}.
974 @item --disable-threads
975 Specify that threading support should be disabled for the system.
976 This is an alias for @option{--enable-threads=single}.
978 @item --enable-threads=@var{lib}
980 @var{lib} is the thread support library. This affects the Objective-C
981 compiler and runtime library, and exception handling for other languages
982 like C++ and Java. The possibilities for @var{lib} are:
990 Ada tasking support. For non-Ada programs, this setting is equivalent
991 to @samp{single}. When used in conjunction with the Ada run time, it
992 causes GCC to use the same thread primitives as Ada uses. This option
993 is necessary when using both Ada and the back end exception handling,
994 which is the default for most Ada targets.
996 Generic MACH thread support, known to work on NeXTSTEP@. (Please note
997 that the file needed to support this configuration, @file{gthr-mach.h}, is
998 missing and thus this setting will cause a known bootstrap failure.)
1000 This is an alias for @samp{single}.
1002 Generic POSIX/Unix98 thread support.
1004 Generic POSIX/Unix95 thread support.
1006 RTEMS thread support.
1008 Disable thread support, should work for all platforms.
1010 Sun Solaris 2 thread support.
1012 VxWorks thread support.
1014 Microsoft Win32 API thread support.
1016 Novell Kernel Services thread support.
1020 Specify that the target supports TLS (Thread Local Storage). Usually
1021 configure can correctly determine if TLS is supported. In cases where
1022 it guesses incorrectly, TLS can be explicitly enabled or disabled with
1023 @option{--enable-tls} or @option{--disable-tls}. This can happen if
1024 the assembler supports TLS but the C library does not, or if the
1025 assumptions made by the configure test are incorrect.
1028 Specify that the target does not support TLS.
1029 This is an alias for @option{--enable-tls=no}.
1031 @item --with-cpu=@var{cpu}
1032 Specify which cpu variant the compiler should generate code for by default.
1033 @var{cpu} will be used as the default value of the @option{-mcpu=} switch.
1034 This option is only supported on some targets, including ARM, i386, M68k,
1035 PowerPC, and SPARC@.
1037 @item --with-schedule=@var{cpu}
1038 @itemx --with-arch=@var{cpu}
1039 @itemx --with-tune=@var{cpu}
1040 @itemx --with-abi=@var{abi}
1041 @itemx --with-fpu=@var{type}
1042 @itemx --with-float=@var{type}
1043 These configure options provide default values for the @option{-mschedule=},
1044 @option{-march=}, @option{-mtune=}, @option{-mabi=}, and @option{-mfpu=}
1045 options and for @option{-mhard-float} or @option{-msoft-float}. As with
1046 @option{--with-cpu}, which switches will be accepted and acceptable values
1047 of the arguments depend on the target.
1049 @item --with-mode=@var{mode}
1050 Specify if the compiler should default to @option{-marm} or @option{-mthumb}.
1051 This option is only supported on ARM targets.
1053 @item --with-divide=@var{type}
1054 Specify how the compiler should generate code for checking for
1055 division by zero. This option is only supported on the MIPS target.
1056 The possibilities for @var{type} are:
1059 Division by zero checks use conditional traps (this is the default on
1060 systems that support conditional traps).
1062 Division by zero checks use the break instruction.
1065 @item --enable-__cxa_atexit
1066 Define if you want to use __cxa_atexit, rather than atexit, to
1067 register C++ destructors for local statics and global objects.
1068 This is essential for fully standards-compliant handling of
1069 destructors, but requires __cxa_atexit in libc. This option is currently
1070 only available on systems with GNU libc. When enabled, this will cause
1071 @option{-fuse-cxa-exit} to be passed by default.
1073 @item --enable-target-optspace
1075 libraries should be optimized for code space instead of code speed.
1076 This is the default for the m32r platform.
1079 Specify that a user visible @command{cpp} program should not be installed.
1081 @item --with-cpp-install-dir=@var{dirname}
1082 Specify that the user visible @command{cpp} program should be installed
1083 in @file{@var{prefix}/@var{dirname}/cpp}, in addition to @var{bindir}.
1085 @item --enable-initfini-array
1086 Force the use of sections @code{.init_array} and @code{.fini_array}
1087 (instead of @code{.init} and @code{.fini}) for constructors and
1088 destructors. Option @option{--disable-initfini-array} has the
1089 opposite effect. If neither option is specified, the configure script
1090 will try to guess whether the @code{.init_array} and
1091 @code{.fini_array} sections are supported and, if they are, use them.
1093 @item --enable-maintainer-mode
1094 The build rules that
1095 regenerate the GCC master message catalog @file{gcc.pot} are normally
1096 disabled. This is because it can only be rebuilt if the complete source
1097 tree is present. If you have changed the sources and want to rebuild the
1098 catalog, configuring with @option{--enable-maintainer-mode} will enable
1099 this. Note that you need a recent version of the @code{gettext} tools
1102 @item --disable-bootstrap
1103 For a native build, the default configuration is to perform
1104 a 3-stage bootstrap of the compiler when @samp{make} is invoked,
1105 testing that GCC can compile itself correctly. If you want to disable
1106 this process, you can configure with @option{--disable-bootstrap}.
1108 @item --enable-bootstrap
1109 In special cases, you may want to perform a 3-stage build
1110 even if the target and host triplets are different.
1111 This could happen when the host can run code compiled for
1112 the target (e.g.@: host is i686-linux, target is i486-linux).
1113 Starting from GCC 4.2, to do this you have to configure explicitly
1114 with @option{--enable-bootstrap}.
1116 @item --enable-generated-files-in-srcdir
1117 Neither the .c and .h files that are generated from Bison and flex nor the
1118 info manuals and man pages that are built from the .texi files are present
1119 in the SVN development tree. When building GCC from that development tree,
1120 or from one of our snapshots, those generated files are placed in your
1121 build directory, which allows for the source to be in a readonly
1124 If you configure with @option{--enable-generated-files-in-srcdir} then those
1125 generated files will go into the source directory. This is mainly intended
1126 for generating release or prerelease tarballs of the GCC sources, since it
1127 is not a requirement that the users of source releases to have flex, Bison,
1130 @item --enable-version-specific-runtime-libs
1132 that runtime libraries should be installed in the compiler specific
1133 subdirectory (@file{@var{libdir}/gcc}) rather than the usual places. In
1134 addition, @samp{libstdc++}'s include files will be installed into
1135 @file{@var{libdir}} unless you overruled it by using
1136 @option{--with-gxx-include-dir=@var{dirname}}. Using this option is
1137 particularly useful if you intend to use several versions of GCC in
1138 parallel. This is currently supported by @samp{libgfortran},
1139 @samp{libjava}, @samp{libmudflap}, @samp{libstdc++}, and @samp{libobjc}.
1141 @item --enable-languages=@var{lang1},@var{lang2},@dots{}
1142 Specify that only a particular subset of compilers and
1143 their runtime libraries should be built. For a list of valid values for
1144 @var{langN} you can issue the following command in the
1145 @file{gcc} directory of your GCC source tree:@*
1147 grep language= */config-lang.in
1149 Currently, you can use any of the following:
1150 @code{all}, @code{ada}, @code{c}, @code{c++}, @code{fortran}, @code{java},
1151 @code{objc}, @code{obj-c++}, @code{treelang}.
1152 Building the Ada compiler has special requirements, see below.
1153 If you do not pass this flag, or specify the option @code{all}, then all
1154 default languages available in the @file{gcc} sub-tree will be configured.
1155 Ada, Objective-C++, and treelang are not default languages; the rest are.
1156 Re-defining @code{LANGUAGES} when calling @samp{make} @strong{does not}
1157 work anymore, as those language sub-directories might not have been
1160 @item --enable-stage1-languages=@var{lang1},@var{lang2},@dots{}
1161 Specify that a particular subset of compilers and their runtime
1162 libraries should be built with the system C compiler during stage 1 of
1163 the bootstrap process, rather than only in later stages with the
1164 bootstrapped C compiler. The list of valid values is the same as for
1165 @option{--enable-languages}, and the option @code{all} will select all
1166 of the languages enabled by @option{--enable-languages}. This option is
1167 primarily useful for GCC development; for instance, when a development
1168 version of the compiler cannot bootstrap due to compiler bugs, or when
1169 one is debugging front ends other than the C front end. When this
1170 option is used, one can then build the target libraries for the
1171 specified languages with the stage-1 compiler by using @command{make
1172 stage1-bubble all-target}, or run the testsuite on the stage-1 compiler
1173 for the specified languages using @command{make stage1-start check-gcc}.
1175 @item --disable-libada
1176 Specify that the run-time libraries and tools used by GNAT should not
1177 be built. This can be useful for debugging, or for compatibility with
1178 previous Ada build procedures, when it was required to explicitly
1179 do a @samp{make -C gcc gnatlib_and_tools}.
1181 @item --disable-libssp
1182 Specify that the run-time libraries for stack smashing protection
1183 should not be built.
1185 @item --disable-libgomp
1186 Specify that the run-time libraries used by GOMP should not be built.
1189 Specify that the compiler should
1190 use DWARF 2 debugging information as the default.
1192 @item --enable-targets=all
1193 @itemx --enable-targets=@var{target_list}
1194 Some GCC targets, e.g.@: powerpc64-linux, build bi-arch compilers.
1195 These are compilers that are able to generate either 64-bit or 32-bit
1196 code. Typically, the corresponding 32-bit target, e.g.@:
1197 powerpc-linux for powerpc64-linux, only generates 32-bit code. This
1198 option enables the 32-bit target to be a bi-arch compiler, which is
1199 useful when you want a bi-arch compiler that defaults to 32-bit, and
1200 you are building a bi-arch or multi-arch binutils in a combined tree.
1201 Currently, this option only affects powerpc-linux and x86-linux.
1203 @item --enable-secureplt
1204 This option enables @option{-msecure-plt} by default for powerpc-linux.
1206 @xref{RS/6000 and PowerPC Options,, RS/6000 and PowerPC Options, gcc,
1207 Using the GNU Compiler Collection (GCC)},
1210 See ``RS/6000 and PowerPC Options'' in the main manual
1213 @item --enable-win32-registry
1214 @itemx --enable-win32-registry=@var{key}
1215 @itemx --disable-win32-registry
1216 The @option{--enable-win32-registry} option enables Microsoft Windows-hosted GCC
1217 to look up installations paths in the registry using the following key:
1220 @code{HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\@var{key}}
1223 @var{key} defaults to GCC version number, and can be overridden by the
1224 @option{--enable-win32-registry=@var{key}} option. Vendors and distributors
1225 who use custom installers are encouraged to provide a different key,
1226 perhaps one comprised of vendor name and GCC version number, to
1227 avoid conflict with existing installations. This feature is enabled
1228 by default, and can be disabled by @option{--disable-win32-registry}
1229 option. This option has no effect on the other hosts.
1232 Specify that the machine does not have a floating point unit. This
1233 option only applies to @samp{m68k-sun-sunos@var{n}}. On any other
1234 system, @option{--nfp} has no effect.
1236 @item --enable-werror
1237 @itemx --disable-werror
1238 @itemx --enable-werror=yes
1239 @itemx --enable-werror=no
1240 When you specify this option, it controls whether certain files in the
1241 compiler are built with @option{-Werror} in bootstrap stage2 and later.
1242 If you don't specify it, @option{-Werror} is turned on for the main
1243 development trunk. However it defaults to off for release branches and
1244 final releases. The specific files which get @option{-Werror} are
1245 controlled by the Makefiles.
1247 @item --enable-checking
1248 @itemx --enable-checking=@var{list}
1249 When you specify this option, the compiler is built to perform internal
1250 consistency checks of the requested complexity. This does not change the
1251 generated code, but adds error checking within the compiler. This will
1252 slow down the compiler and may only work properly if you are building
1253 the compiler with GCC@. This is @samp{yes} by default when building
1254 from SVN or snapshots, but @samp{release} for releases. More control
1255 over the checks may be had by specifying @var{list}. The categories of
1256 checks available are @samp{yes} (most common checks
1257 @samp{assert,misc,tree,gc,rtlflag,runtime}), @samp{no} (no checks at
1258 all), @samp{all} (all but @samp{valgrind}), @samp{release} (cheapest
1259 checks @samp{assert,runtime}) or @samp{none} (same as @samp{no}).
1260 Individual checks can be enabled with these flags @samp{assert},
1261 @samp{fold}, @samp{gc}, @samp{gcac} @samp{misc}, @samp{rtl},
1262 @samp{rtlflag}, @samp{runtime}, @samp{tree}, and @samp{valgrind}.
1264 The @samp{valgrind} check requires the external @command{valgrind}
1265 simulator, available from @uref{http://valgrind.org/}. The
1266 @samp{rtl}, @samp{gcac} and @samp{valgrind} checks are very expensive.
1267 To disable all checking, @samp{--disable-checking} or
1268 @samp{--enable-checking=none} must be explicitly requested. Disabling
1269 assertions will make the compiler and runtime slightly faster but
1270 increase the risk of undetected internal errors causing wrong code to be
1273 @item --enable-coverage
1274 @itemx --enable-coverage=@var{level}
1275 With this option, the compiler is built to collect self coverage
1276 information, every time it is run. This is for internal development
1277 purposes, and only works when the compiler is being built with gcc. The
1278 @var{level} argument controls whether the compiler is built optimized or
1279 not, values are @samp{opt} and @samp{noopt}. For coverage analysis you
1280 want to disable optimization, for performance analysis you want to
1281 enable optimization. When coverage is enabled, the default level is
1282 without optimization.
1284 @item --enable-gather-detailed-mem-stats
1285 When this option is specified more detailed information on memory
1286 allocation is gathered. This information is printed when using
1287 @option{-fmem-report}.
1290 @itemx --with-gc=@var{choice}
1291 With this option you can specify the garbage collector implementation
1292 used during the compilation process. @var{choice} can be one of
1293 @samp{page} and @samp{zone}, where @samp{page} is the default.
1296 @itemx --disable-nls
1297 The @option{--enable-nls} option enables Native Language Support (NLS),
1298 which lets GCC output diagnostics in languages other than American
1299 English. Native Language Support is enabled by default if not doing a
1300 canadian cross build. The @option{--disable-nls} option disables NLS@.
1302 @item --with-included-gettext
1303 If NLS is enabled, the @option{--with-included-gettext} option causes the build
1304 procedure to prefer its copy of GNU @command{gettext}.
1306 @item --with-catgets
1307 If NLS is enabled, and if the host lacks @code{gettext} but has the
1308 inferior @code{catgets} interface, the GCC build procedure normally
1309 ignores @code{catgets} and instead uses GCC's copy of the GNU
1310 @code{gettext} library. The @option{--with-catgets} option causes the
1311 build procedure to use the host's @code{catgets} in this situation.
1313 @item --with-libiconv-prefix=@var{dir}
1314 Search for libiconv header files in @file{@var{dir}/include} and
1315 libiconv library files in @file{@var{dir}/lib}.
1317 @item --enable-obsolete
1318 Enable configuration for an obsoleted system. If you attempt to
1319 configure GCC for a system (build, host, or target) which has been
1320 obsoleted, and you do not specify this flag, configure will halt with an
1323 All support for systems which have been obsoleted in one release of GCC
1324 is removed entirely in the next major release, unless someone steps
1325 forward to maintain the port.
1327 @item --enable-decimal-float
1328 @itemx --enable-decimal-float=yes
1329 @itemx --enable-decimal-float=no
1330 @itemx --enable-decimal-float=bid
1331 @itemx --enable-decimal-float=dpd
1332 @itemx --disable-decimal-float
1333 Enable (or disable) support for the C decimal floating point extension
1334 that is in the IEEE 754R extension to the IEEE754 floating point
1335 standard. This is enabled by default only on PowerPC, i386, and
1336 x86_64 GNU/Linux systems. Other systems may also support it, but
1337 require the user to specifically enable it. You can optionally
1338 control which decimal floating point format is used (either @samp{bid}
1339 or @samp{dpd}). The @samp{bid} (binary integer decimal) format is
1340 default on i386 and x86_64 systems, and the @samp{dpd} (densely packed
1341 decimal) format is default on PowerPC systems.
1343 @item --enable-fixed-point
1344 @itemx --disable-fixed-point
1345 Enable (or disable) support for C fixed-point arithmetic.
1346 This option is enabled by default for some targets (such as MIPS) which
1347 have hardware-support for fixed-point operations. On other targets, you
1348 may enable this option manually.
1350 @item --with-long-double-128
1351 Specify if @code{long double} type should be 128-bit by default on selected
1352 GNU/Linux architectures. If using @code{--without-long-double-128},
1353 @code{long double} will be by default 64-bit, the same as @code{double} type.
1354 When neither of these configure options are used, the default will be
1355 128-bit @code{long double} when built against GNU C Library 2.4 and later,
1356 64-bit @code{long double} otherwise.
1358 @item --with-gmp=@var{pathname}
1359 @itemx --with-gmp-include=@var{pathname}
1360 @itemx --with-gmp-lib=@var{pathname}
1361 @itemx --with-mpfr=@var{pathname}
1362 @itemx --with-mpfr-include=@var{pathname}
1363 @itemx --with-mpfr-lib=@var{pathname}
1364 If you do not have GMP (the GNU Multiple Precision library) and the
1365 MPFR Libraries installed in a standard location and you want to build
1366 GCC, you can explicitly specify the directory where they are installed
1367 (@samp{--with-gmp=@var{gmpinstalldir}},
1368 @samp{--with-mpfr=@var{mpfrinstalldir}}). The
1369 @option{--with-gmp=@var{gmpinstalldir}} option is shorthand for
1370 @option{--with-gmp-lib=@var{gmpinstalldir}/lib} and
1371 @option{--with-gmp-include=@var{gmpinstalldir}/include}. Likewise the
1372 @option{--with-mpfr=@var{mpfrinstalldir}} option is shorthand for
1373 @option{--with-mpfr-lib=@var{mpfrinstalldir}/lib} and
1374 @option{--with-mpfr-include=@var{mpfrinstalldir}/include}. If these
1375 shorthand assumptions are not correct, you can use the explicit
1376 include and lib options directly.
1378 @item --with-debug-prefix-map=@var{map}
1379 Convert source directory names using @option{-fdebug-prefix-map} when
1380 building runtime libraries. @samp{@var{map}} is a space-separated
1381 list of maps of the form @samp{@var{old}=@var{new}}.
1385 @subheading Cross-Compiler-Specific Options
1386 The following options only apply to building cross compilers.
1388 @item --with-sysroot
1389 @itemx --with-sysroot=@var{dir}
1390 Tells GCC to consider @var{dir} as the root of a tree that contains a
1391 (subset of) the root filesystem of the target operating system.
1392 Target system headers, libraries and run-time object files will be
1393 searched in there. The specified directory is not copied into the
1394 install tree, unlike the options @option{--with-headers} and
1395 @option{--with-libs} that this option obsoletes. The default value,
1396 in case @option{--with-sysroot} is not given an argument, is
1397 @option{$@{gcc_tooldir@}/sys-root}. If the specified directory is a
1398 subdirectory of @option{$@{exec_prefix@}}, then it will be found relative to
1399 the GCC binaries if the installation tree is moved.
1401 @item --with-build-sysroot
1402 @itemx --with-build-sysroot=@var{dir}
1403 Tells GCC to consider @var{dir} as the system root (see
1404 @option{--with-sysroot}) while building target libraries, instead of
1405 the directory specified with @option{--with-sysroot}. This option is
1406 only useful when you are already using @option{--with-sysroot}. You
1407 can use @option{--with-build-sysroot} when you are configuring with
1408 @option{--prefix} set to a directory that is different from the one in
1409 which you are installing GCC and your target libraries.
1411 This option affects the system root for the compiler used to build
1412 target libraries (which runs on the build system); it does not affect
1413 the compiler which is used to build GCC itself.
1415 @item --with-headers
1416 @itemx --with-headers=@var{dir}
1417 Deprecated in favor of @option{--with-sysroot}.
1418 Specifies that target headers are available when building a cross compiler.
1419 The @var{dir} argument specifies a directory which has the target include
1420 files. These include files will be copied into the @file{gcc} install
1421 directory. @emph{This option with the @var{dir} argument is required} when
1422 building a cross compiler, if @file{@var{prefix}/@var{target}/sys-include}
1423 doesn't pre-exist. If @file{@var{prefix}/@var{target}/sys-include} does
1424 pre-exist, the @var{dir} argument may be omitted. @command{fixincludes}
1425 will be run on these files to make them compatible with GCC@.
1427 @item --without-headers
1428 Tells GCC not use any target headers from a libc when building a cross
1429 compiler. When crossing to GNU/Linux, you need the headers so GCC
1430 can build the exception handling for libgcc.
1433 @itemx --with-libs=``@var{dir1} @var{dir2} @dots{} @var{dirN}''
1434 Deprecated in favor of @option{--with-sysroot}.
1435 Specifies a list of directories which contain the target runtime
1436 libraries. These libraries will be copied into the @file{gcc} install
1437 directory. If the directory list is omitted, this option has no
1441 Specifies that @samp{newlib} is
1442 being used as the target C library. This causes @code{__eprintf} to be
1443 omitted from @file{libgcc.a} on the assumption that it will be provided by
1446 @item --with-build-time-tools=@var{dir}
1447 Specifies where to find the set of target tools (assembler, linker, etc.)
1448 that will be used while building GCC itself. This option can be useful
1449 if the directory layouts are different between the system you are building
1450 GCC on, and the system where you will deploy it.
1452 For example, on a @option{ia64-hp-hpux} system, you may have the GNU
1453 assembler and linker in @file{/usr/bin}, and the native tools in a
1454 different path, and build a toolchain that expects to find the
1455 native tools in @file{/usr/bin}.
1457 When you use this option, you should ensure that @var{dir} includes
1458 @command{ar}, @command{as}, @command{ld}, @command{nm},
1459 @command{ranlib} and @command{strip} if necessary, and possibly
1460 @command{objdump}. Otherwise, GCC may use an inconsistent set of
1464 @subheading Java-Specific Options
1466 The following option applies to the build of the Java front end.
1469 @item --disable-libgcj
1470 Specify that the run-time libraries
1471 used by GCJ should not be built. This is useful in case you intend
1472 to use GCJ with some other run-time, or you're going to install it
1473 separately, or it just happens not to build on your particular
1474 machine. In general, if the Java front end is enabled, the GCJ
1475 libraries will be enabled too, unless they're known to not work on
1476 the target platform. If GCJ is enabled but @samp{libgcj} isn't built, you
1477 may need to port it; in this case, before modifying the top-level
1478 @file{configure.in} so that @samp{libgcj} is enabled by default on this platform,
1479 you may use @option{--enable-libgcj} to override the default.
1483 The following options apply to building @samp{libgcj}.
1485 @subsubheading General Options
1488 @item --enable-java-maintainer-mode
1489 By default the @samp{libjava} build will not attempt to compile the
1490 @file{.java} source files to @file{.class}. Instead, it will use the
1491 @file{.class} files from the source tree. If you use this option you
1492 must have executables named @command{ecj1} and @command{gjavah} in your path
1493 for use by the build. You must use this option if you intend to
1494 modify any @file{.java} files in @file{libjava}.
1496 @item --with-java-home=@var{dirname}
1497 This @samp{libjava} option overrides the default value of the
1498 @samp{java.home} system property. It is also used to set
1499 @samp{sun.boot.class.path} to @file{@var{dirname}/lib/rt.jar}. By
1500 default @samp{java.home} is set to @file{@var{prefix}} and
1501 @samp{sun.boot.class.path} to
1502 @file{@var{datadir}/java/libgcj-@var{version}.jar}.
1504 @item --with-ecj-jar=@var{filename}
1505 This option can be used to specify the location of an external jar
1506 file containing the Eclipse Java compiler. A specially modified
1507 version of this compiler is used by @command{gcj} to parse
1508 @file{.java} source files. If this option is given, the
1509 @samp{libjava} build will create and install an @file{ecj1} executable
1510 which uses this jar file at runtime.
1512 If this option is not given, but an @file{ecj.jar} file is found in
1513 the topmost source tree at configure time, then the @samp{libgcj}
1514 build will create and install @file{ecj1}, and will also install the
1515 discovered @file{ecj.jar} into a suitable place in the install tree.
1517 If @file{ecj1} is not installed, then the user will have to supply one
1518 on his path in order for @command{gcj} to properly parse @file{.java}
1519 source files. A suitable jar is available from
1520 @uref{ftp://sourceware.org/pub/java/}.
1522 @item --disable-getenv-properties
1523 Don't set system properties from @env{GCJ_PROPERTIES}.
1525 @item --enable-hash-synchronization
1526 Use a global hash table for monitor locks. Ordinarily,
1527 @samp{libgcj}'s @samp{configure} script automatically makes
1528 the correct choice for this option for your platform. Only use
1529 this if you know you need the library to be configured differently.
1531 @item --enable-interpreter
1532 Enable the Java interpreter. The interpreter is automatically
1533 enabled by default on all platforms that support it. This option
1534 is really only useful if you want to disable the interpreter
1535 (using @option{--disable-interpreter}).
1537 @item --disable-java-net
1538 Disable java.net. This disables the native part of java.net only,
1539 using non-functional stubs for native method implementations.
1541 @item --disable-jvmpi
1542 Disable JVMPI support.
1545 Enable runtime eCos target support.
1547 @item --without-libffi
1548 Don't use @samp{libffi}. This will disable the interpreter and JNI
1549 support as well, as these require @samp{libffi} to work.
1551 @item --enable-libgcj-debug
1552 Enable runtime debugging code.
1554 @item --enable-libgcj-multifile
1555 If specified, causes all @file{.java} source files to be
1556 compiled into @file{.class} files in one invocation of
1557 @samp{gcj}. This can speed up build time, but is more
1558 resource-intensive. If this option is unspecified or
1559 disabled, @samp{gcj} is invoked once for each @file{.java}
1560 file to compile into a @file{.class} file.
1562 @item --with-libiconv-prefix=DIR
1563 Search for libiconv in @file{DIR/include} and @file{DIR/lib}.
1565 @item --enable-sjlj-exceptions
1566 Force use of the @code{setjmp}/@code{longjmp}-based scheme for exceptions.
1567 @samp{configure} ordinarily picks the correct value based on the platform.
1568 Only use this option if you are sure you need a different setting.
1570 @item --with-system-zlib
1571 Use installed @samp{zlib} rather than that included with GCC@.
1573 @item --with-win32-nlsapi=ansi, unicows or unicode
1574 Indicates how MinGW @samp{libgcj} translates between UNICODE
1575 characters and the Win32 API@.
1578 Use the single-byte @code{char} and the Win32 A functions natively,
1579 translating to and from UNICODE when using these functions. If
1580 unspecified, this is the default.
1583 Use the @code{WCHAR} and Win32 W functions natively. Adds
1584 @code{-lunicows} to @file{libgcj.spec} to link with @samp{libunicows}.
1585 @file{unicows.dll} needs to be deployed on Microsoft Windows 9X machines
1586 running built executables. @file{libunicows.a}, an open-source
1587 import library around Microsoft's @code{unicows.dll}, is obtained from
1588 @uref{http://libunicows.sourceforge.net/}, which also gives details
1589 on getting @file{unicows.dll} from Microsoft.
1592 Use the @code{WCHAR} and Win32 W functions natively. Does @emph{not}
1593 add @code{-lunicows} to @file{libgcj.spec}. The built executables will
1594 only run on Microsoft Windows NT and above.
1598 @subsubheading AWT-Specific Options
1602 Use the X Window System.
1604 @item --enable-java-awt=PEER(S)
1605 Specifies the AWT peer library or libraries to build alongside
1606 @samp{libgcj}. If this option is unspecified or disabled, AWT
1607 will be non-functional. Current valid values are @option{gtk} and
1608 @option{xlib}. Multiple libraries should be separated by a
1609 comma (i.e.@: @option{--enable-java-awt=gtk,xlib}).
1611 @item --enable-gtk-cairo
1612 Build the cairo Graphics2D implementation on GTK@.
1614 @item --enable-java-gc=TYPE
1615 Choose garbage collector. Defaults to @option{boehm} if unspecified.
1617 @item --disable-gtktest
1618 Do not try to compile and run a test GTK+ program.
1620 @item --disable-glibtest
1621 Do not try to compile and run a test GLIB program.
1623 @item --with-libart-prefix=PFX
1624 Prefix where libart is installed (optional).
1626 @item --with-libart-exec-prefix=PFX
1627 Exec prefix where libart is installed (optional).
1629 @item --disable-libarttest
1630 Do not try to compile and run a test libart program.
1639 @uref{./index.html,,Return to the GCC Installation page}
1643 @c ***Building****************************************************************
1645 @comment node-name, next, previous, up
1646 @node Building, Testing, Configuration, Installing GCC
1652 @cindex Installing GCC: Building
1654 Now that GCC is configured, you are ready to build the compiler and
1657 Some commands executed when making the compiler may fail (return a
1658 nonzero status) and be ignored by @command{make}. These failures, which
1659 are often due to files that were not found, are expected, and can safely
1662 It is normal to have compiler warnings when compiling certain files.
1663 Unless you are a GCC developer, you can generally ignore these warnings
1664 unless they cause compilation to fail. Developers should attempt to fix
1665 any warnings encountered, however they can temporarily continue past
1666 warnings-as-errors by specifying the configure flag
1667 @option{--disable-werror}.
1669 On certain old systems, defining certain environment variables such as
1670 @env{CC} can interfere with the functioning of @command{make}.
1672 If you encounter seemingly strange errors when trying to build the
1673 compiler in a directory other than the source directory, it could be
1674 because you have previously configured the compiler in the source
1675 directory. Make sure you have done all the necessary preparations.
1677 If you build GCC on a BSD system using a directory stored in an old System
1678 V file system, problems may occur in running @command{fixincludes} if the
1679 System V file system doesn't support symbolic links. These problems
1680 result in a failure to fix the declaration of @code{size_t} in
1681 @file{sys/types.h}. If you find that @code{size_t} is a signed type and
1682 that type mismatches occur, this could be the cause.
1684 The solution is not to use such a directory for building GCC@.
1686 When building from SVN or snapshots and enabling the @code{treelang}
1687 front end, or if you modify @file{*.y} files, you need the Bison parser
1688 generator installed. If you do not modify @file{*.y} files, releases
1689 contain the Bison-generated files and you do not need Bison installed
1690 to build them. Note that most front ends now use hand-written parsers,
1691 which can be modified with no need for Bison.
1693 Similarly, when building from SVN or snapshots, or if you modify
1694 @file{*.l} files, you need the Flex lexical analyzer generator installed.
1695 There is still one Flex-based lexical analyzer (part of the build
1696 machinery, not of GCC itself) that is used even if you only build the
1699 When building from SVN or snapshots, or if you modify Texinfo
1700 documentation, you need version 4.4 or later of Texinfo installed if you
1701 want Info documentation to be regenerated. Releases contain Info
1702 documentation pre-built for the unmodified documentation in the release.
1704 @section Building a native compiler
1706 For a native build, the default configuration is to perform
1707 a 3-stage bootstrap of the compiler when @samp{make} is invoked.
1708 This will build the entire GCC system and ensure that it compiles
1709 itself correctly. It can be disabled with the @option{--disable-bootstrap}
1710 parameter to @samp{configure}, but bootstrapping is suggested because
1711 the compiler will be tested more completely and could also have
1714 The bootstrapping process will complete the following steps:
1718 Build tools necessary to build the compiler.
1721 Perform a 3-stage bootstrap of the compiler. This includes building
1722 three times the target tools for use by the compiler such as binutils
1723 (bfd, binutils, gas, gprof, ld, and opcodes) if they have been
1724 individually linked or moved into the top level GCC source tree before
1728 Perform a comparison test of the stage2 and stage3 compilers.
1731 Build runtime libraries using the stage3 compiler from the previous step.
1735 If you are short on disk space you might consider @samp{make
1736 bootstrap-lean} instead. The sequence of compilation is the
1737 same described above, but object files from the stage1 and
1738 stage2 of the 3-stage bootstrap of the compiler are deleted as
1739 soon as they are no longer needed.
1741 If you want to save additional space during the bootstrap and in
1742 the final installation as well, you can build the compiler binaries
1743 without debugging information as in the following example. This will save
1744 roughly 40% of disk space both for the bootstrap and the final installation.
1745 (Libraries will still contain debugging information.)
1748 make CFLAGS='-O' LIBCFLAGS='-g -O2' \
1749 LIBCXXFLAGS='-g -O2 -fno-implicit-templates' bootstrap
1752 If you wish to use non-default GCC flags when compiling the stage2 and
1753 stage3 compilers, set @code{BOOT_CFLAGS} on the command line when doing
1754 @samp{make}. Non-default optimization flags are less well
1755 tested here than the default of @samp{-g -O2}, but should still work.
1756 In a few cases, you may find that you need to specify special flags such
1757 as @option{-msoft-float} here to complete the bootstrap; or, if the
1758 native compiler miscompiles the stage1 compiler, you may need to work
1759 around this, by choosing @code{BOOT_CFLAGS} to avoid the parts of the
1760 stage1 compiler that were miscompiled, or by using @samp{make
1761 bootstrap4} to increase the number of stages of bootstrap.
1763 Note that using non-standard @code{CFLAGS} can cause bootstrap to fail
1764 if these trigger a warning with the new compiler. For example using
1765 @samp{-O2 -g -mcpu=i686} on @code{i686-pc-linux-gnu} will cause bootstrap
1766 failure as @option{-mcpu=} is deprecated in 3.4.0 and above.
1769 If you used the flag @option{--enable-languages=@dots{}} to restrict
1770 the compilers to be built, only those you've actually enabled will be
1771 built. This will of course only build those runtime libraries, for
1772 which the particular compiler has been built. Please note,
1773 that re-defining @env{LANGUAGES} when calling @samp{make}
1774 @strong{does not} work anymore!
1776 If the comparison of stage2 and stage3 fails, this normally indicates
1777 that the stage2 compiler has compiled GCC incorrectly, and is therefore
1778 a potentially serious bug which you should investigate and report. (On
1779 a few systems, meaningful comparison of object files is impossible; they
1780 always appear ``different''. If you encounter this problem, you will
1781 need to disable comparison in the @file{Makefile}.)
1783 If you do not want to bootstrap your compiler, you can configure with
1784 @option{--disable-bootstrap}. In particular cases, you may want to
1785 bootstrap your compiler even if the target system is not the same as
1786 the one you are building on: for example, you could build a
1787 @code{powerpc-unknown-linux-gnu} toolchain on a
1788 @code{powerpc64-unknown-linux-gnu} host. In this case, pass
1789 @option{--enable-bootstrap} to the configure script.
1792 @section Building a cross compiler
1794 When building a cross compiler, it is not generally possible to do a
1795 3-stage bootstrap of the compiler. This makes for an interesting problem
1796 as parts of GCC can only be built with GCC@.
1798 To build a cross compiler, we first recommend building and installing a
1799 native compiler. You can then use the native GCC compiler to build the
1800 cross compiler. The installed native compiler needs to be GCC version
1803 If the cross compiler is to be built with support for the Java
1804 programming language and the ability to compile .java source files is
1805 desired, the installed native compiler used to build the cross
1806 compiler needs to be the same GCC version as the cross compiler. In
1807 addition the cross compiler needs to be configured with
1808 @option{--with-ecj-jar=@dots{}}.
1810 Assuming you have already installed a native copy of GCC and configured
1811 your cross compiler, issue the command @command{make}, which performs the
1816 Build host tools necessary to build the compiler.
1819 Build target tools for use by the compiler such as binutils (bfd,
1820 binutils, gas, gprof, ld, and opcodes)
1821 if they have been individually linked or moved into the top level GCC source
1822 tree before configuring.
1825 Build the compiler (single stage only).
1828 Build runtime libraries using the compiler from the previous step.
1831 Note that if an error occurs in any step the make process will exit.
1833 If you are not building GNU binutils in the same source tree as GCC,
1834 you will need a cross-assembler and cross-linker installed before
1835 configuring GCC@. Put them in the directory
1836 @file{@var{prefix}/@var{target}/bin}. Here is a table of the tools
1837 you should put in this directory:
1841 This should be the cross-assembler.
1844 This should be the cross-linker.
1847 This should be the cross-archiver: a program which can manipulate
1848 archive files (linker libraries) in the target machine's format.
1851 This should be a program to construct a symbol table in an archive file.
1854 The installation of GCC will find these programs in that directory,
1855 and copy or link them to the proper place to for the cross-compiler to
1856 find them when run later.
1858 The easiest way to provide these files is to build the Binutils package.
1859 Configure it with the same @option{--host} and @option{--target}
1860 options that you use for configuring GCC, then build and install
1861 them. They install their executables automatically into the proper
1862 directory. Alas, they do not support all the targets that GCC
1865 If you are not building a C library in the same source tree as GCC,
1866 you should also provide the target libraries and headers before
1867 configuring GCC, specifying the directories with
1868 @option{--with-sysroot} or @option{--with-headers} and
1869 @option{--with-libs}. Many targets also require ``start files'' such
1870 as @file{crt0.o} and
1871 @file{crtn.o} which are linked into each executable. There may be several
1872 alternatives for @file{crt0.o}, for use with profiling or other
1873 compilation options. Check your target's definition of
1874 @code{STARTFILE_SPEC} to find out what start files it uses.
1876 @section Building in parallel
1878 GNU Make 3.79 and above, which is necessary to build GCC, support
1879 building in parallel. To activate this, you can use @samp{make -j 2}
1880 instead of @samp{make}. You can also specify a bigger number, and
1881 in most cases using a value greater than the number of processors in
1882 your machine will result in fewer and shorter I/O latency hits, thus
1883 improving overall throughput; this is especially true for slow drives
1884 and network filesystems.
1886 @section Building the Ada compiler
1888 In order to build GNAT, the Ada compiler, you need a working GNAT
1889 compiler (GNAT version 3.14 or later, or GCC version 3.1 or later).
1890 This includes GNAT tools such as @command{gnatmake} and
1891 @command{gnatlink}, since the Ada front end is written in Ada and
1892 uses some GNAT-specific extensions.
1894 In order to build a cross compiler, it is suggested to install
1895 the new compiler as native first, and then use it to build the cross
1898 @command{configure} does not test whether the GNAT installation works
1899 and has a sufficiently recent version; if too old a GNAT version is
1900 installed, the build will fail unless @option{--enable-languages} is
1901 used to disable building the Ada front end.
1903 @section Building with profile feedback
1905 It is possible to use profile feedback to optimize the compiler itself. This
1906 should result in a faster compiler binary. Experiments done on x86 using gcc
1907 3.3 showed approximately 7 percent speedup on compiling C programs. To
1908 bootstrap the compiler with profile feedback, use @code{make profiledbootstrap}.
1910 When @samp{make profiledbootstrap} is run, it will first build a @code{stage1}
1911 compiler. This compiler is used to build a @code{stageprofile} compiler
1912 instrumented to collect execution counts of instruction and branch
1913 probabilities. Then runtime libraries are compiled with profile collected.
1914 Finally a @code{stagefeedback} compiler is built using the information collected.
1916 Unlike standard bootstrap, several additional restrictions apply. The
1917 compiler used to build @code{stage1} needs to support a 64-bit integral type.
1918 It is recommended to only use GCC for this. Also parallel make is currently
1919 not supported since collisions in profile collecting may occur.
1926 @uref{./index.html,,Return to the GCC Installation page}
1930 @c ***Testing*****************************************************************
1932 @comment node-name, next, previous, up
1933 @node Testing, Final install, Building, Installing GCC
1937 @chapter Installing GCC: Testing
1940 @cindex Installing GCC: Testing
1943 Before you install GCC, we encourage you to run the testsuites and to
1944 compare your results with results from a similar configuration that have
1945 been submitted to the
1946 @uref{http://gcc.gnu.org/ml/gcc-testresults/,,gcc-testresults mailing list}.
1947 Some of these archived results are linked from the build status lists
1948 at @uref{http://gcc.gnu.org/buildstat.html}, although not everyone who
1949 reports a successful build runs the testsuites and submits the results.
1950 This step is optional and may require you to download additional software,
1951 but it can give you confidence in your new GCC installation or point out
1952 problems before you install and start using your new GCC@.
1954 First, you must have @uref{download.html,,downloaded the testsuites}.
1955 These are part of the full distribution, but if you downloaded the
1956 ``core'' compiler plus any front ends, you must download the testsuites
1959 Second, you must have the testing tools installed. This includes
1960 @uref{http://www.gnu.org/software/dejagnu/,,DejaGnu}, Tcl, and Expect;
1961 the DejaGnu site has links to these.
1963 If the directories where @command{runtest} and @command{expect} were
1964 installed are not in the @env{PATH}, you may need to set the following
1965 environment variables appropriately, as in the following example (which
1966 assumes that DejaGnu has been installed under @file{/usr/local}):
1969 TCL_LIBRARY = /usr/local/share/tcl8.0
1970 DEJAGNULIBS = /usr/local/share/dejagnu
1973 (On systems such as Cygwin, these paths are required to be actual
1974 paths, not mounts or links; presumably this is due to some lack of
1975 portability in the DejaGnu code.)
1978 Finally, you can run the testsuite (which may take a long time):
1980 cd @var{objdir}; make -k check
1983 This will test various components of GCC, such as compiler
1984 front ends and runtime libraries. While running the testsuite, DejaGnu
1985 might emit some harmless messages resembling
1986 @samp{WARNING: Couldn't find the global config file.} or
1987 @samp{WARNING: Couldn't find tool init file} that can be ignored.
1989 If you are testing a cross-compiler, you may want to run the testsuite
1990 on a simulator as described at @uref{http://gcc.gnu.org/simtest-howto.html}.
1992 @section How can you run the testsuite on selected tests?
1994 In order to run sets of tests selectively, there are targets
1995 @samp{make check-gcc} and @samp{make check-g++}
1996 in the @file{gcc} subdirectory of the object directory. You can also
1997 just run @samp{make check} in a subdirectory of the object directory.
2000 A more selective way to just run all @command{gcc} execute tests in the
2004 make check-gcc RUNTESTFLAGS="execute.exp @var{other-options}"
2007 Likewise, in order to run only the @command{g++} ``old-deja'' tests in
2008 the testsuite with filenames matching @samp{9805*}, you would use
2011 make check-g++ RUNTESTFLAGS="old-deja.exp=9805* @var{other-options}"
2014 The @file{*.exp} files are located in the testsuite directories of the GCC
2015 source, the most important ones being @file{compile.exp},
2016 @file{execute.exp}, @file{dg.exp} and @file{old-deja.exp}.
2017 To get a list of the possible @file{*.exp} files, pipe the
2018 output of @samp{make check} into a file and look at the
2019 @samp{Running @dots{} .exp} lines.
2021 @section Passing options and running multiple testsuites
2023 You can pass multiple options to the testsuite using the
2024 @samp{--target_board} option of DejaGNU, either passed as part of
2025 @samp{RUNTESTFLAGS}, or directly to @command{runtest} if you prefer to
2026 work outside the makefiles. For example,
2029 make check-g++ RUNTESTFLAGS="--target_board=unix/-O3/-fmerge-constants"
2032 will run the standard @command{g++} testsuites (``unix'' is the target name
2033 for a standard native testsuite situation), passing
2034 @samp{-O3 -fmerge-constants} to the compiler on every test, i.e.,
2035 slashes separate options.
2037 You can run the testsuites multiple times using combinations of options
2038 with a syntax similar to the brace expansion of popular shells:
2041 @dots{}"--target_board=arm-sim/@{-mhard-float,-msoft-float@}@{-O1,-O2,-O3,@}"
2044 (Note the empty option caused by the trailing comma in the final group.)
2045 The following will run each testsuite eight times using the @samp{arm-sim}
2046 target, as if you had specified all possible combinations yourself:
2049 --target_board=arm-sim/-mhard-float/-O1
2050 --target_board=arm-sim/-mhard-float/-O2
2051 --target_board=arm-sim/-mhard-float/-O3
2052 --target_board=arm-sim/-mhard-float
2053 --target_board=arm-sim/-msoft-float/-O1
2054 --target_board=arm-sim/-msoft-float/-O2
2055 --target_board=arm-sim/-msoft-float/-O3
2056 --target_board=arm-sim/-msoft-float
2059 They can be combined as many times as you wish, in arbitrary ways. This
2063 @dots{}"--target_board=unix/-Wextra@{-O3,-fno-strength-reduce@}@{-fomit-frame-pointer,@}"
2066 will generate four combinations, all involving @samp{-Wextra}.
2068 The disadvantage to this method is that the testsuites are run in serial,
2069 which is a waste on multiprocessor systems. For users with GNU Make and
2070 a shell which performs brace expansion, you can run the testsuites in
2071 parallel by having the shell perform the combinations and @command{make}
2072 do the parallel runs. Instead of using @samp{--target_board}, use a
2073 special makefile target:
2076 make -j@var{N} check-@var{testsuite}//@var{test-target}/@var{option1}/@var{option2}/@dots{}
2082 make -j3 check-gcc//sh-hms-sim/@{-m1,-m2,-m3,-m3e,-m4@}/@{,-nofpu@}
2085 will run three concurrent ``make-gcc'' testsuites, eventually testing all
2086 ten combinations as described above. Note that this is currently only
2087 supported in the @file{gcc} subdirectory. (To see how this works, try
2088 typing @command{echo} before the example given here.)
2091 @section Additional testing for Java Class Libraries
2093 The Java runtime tests can be executed via @samp{make check}
2094 in the @file{@var{target}/libjava/testsuite} directory in
2097 The @uref{http://sourceware.org/mauve/,,Mauve Project} provides
2098 a suite of tests for the Java Class Libraries. This suite can be run
2099 as part of libgcj testing by placing the Mauve tree within the libjava
2100 testsuite at @file{libjava/testsuite/libjava.mauve/mauve}, or by
2101 specifying the location of that tree when invoking @samp{make}, as in
2102 @samp{make MAUVEDIR=~/mauve check}.
2104 @section How to interpret test results
2106 The result of running the testsuite are various @file{*.sum} and @file{*.log}
2107 files in the testsuite subdirectories. The @file{*.log} files contain a
2108 detailed log of the compiler invocations and the corresponding
2109 results, the @file{*.sum} files summarize the results. These summaries
2110 contain status codes for all tests:
2114 PASS: the test passed as expected
2116 XPASS: the test unexpectedly passed
2118 FAIL: the test unexpectedly failed
2120 XFAIL: the test failed as expected
2122 UNSUPPORTED: the test is not supported on this platform
2124 ERROR: the testsuite detected an error
2126 WARNING: the testsuite detected a possible problem
2129 It is normal for some tests to report unexpected failures. At the
2130 current time the testing harness does not allow fine grained control
2131 over whether or not a test is expected to fail. This problem should
2132 be fixed in future releases.
2135 @section Submitting test results
2137 If you want to report the results to the GCC project, use the
2138 @file{contrib/test_summary} shell script. Start it in the @var{objdir} with
2141 @var{srcdir}/contrib/test_summary -p your_commentary.txt \
2142 -m gcc-testresults@@gcc.gnu.org |sh
2145 This script uses the @command{Mail} program to send the results, so
2146 make sure it is in your @env{PATH}. The file @file{your_commentary.txt} is
2147 prepended to the testsuite summary and should contain any special
2148 remarks you have on your results or your build environment. Please
2149 do not edit the testsuite result block or the subject line, as these
2150 messages may be automatically processed.
2157 @uref{./index.html,,Return to the GCC Installation page}
2161 @c ***Final install***********************************************************
2163 @comment node-name, next, previous, up
2164 @node Final install, , Testing, Installing GCC
2166 @ifset finalinstallhtml
2168 @chapter Installing GCC: Final installation
2171 Now that GCC has been built (and optionally tested), you can install it with
2173 cd @var{objdir}; make install
2176 We strongly recommend to install into a target directory where there is
2177 no previous version of GCC present. Also, the GNAT runtime should not
2178 be stripped, as this would break certain features of the debugger that
2179 depend on this debugging information (catching Ada exceptions for
2182 That step completes the installation of GCC; user level binaries can
2183 be found in @file{@var{prefix}/bin} where @var{prefix} is the value
2184 you specified with the @option{--prefix} to configure (or
2185 @file{/usr/local} by default). (If you specified @option{--bindir},
2186 that directory will be used instead; otherwise, if you specified
2187 @option{--exec-prefix}, @file{@var{exec-prefix}/bin} will be used.)
2188 Headers for the C++ and Java libraries are installed in
2189 @file{@var{prefix}/include}; libraries in @file{@var{libdir}}
2190 (normally @file{@var{prefix}/lib}); internal parts of the compiler in
2191 @file{@var{libdir}/gcc} and @file{@var{libexecdir}/gcc}; documentation
2192 in info format in @file{@var{infodir}} (normally
2193 @file{@var{prefix}/info}).
2195 When installing cross-compilers, GCC's executables
2196 are not only installed into @file{@var{bindir}}, that
2197 is, @file{@var{exec-prefix}/bin}, but additionally into
2198 @file{@var{exec-prefix}/@var{target-alias}/bin}, if that directory
2199 exists. Typically, such @dfn{tooldirs} hold target-specific
2200 binutils, including assembler and linker.
2202 Installation into a temporary staging area or into a @command{chroot}
2203 jail can be achieved with the command
2206 make DESTDIR=@var{path-to-rootdir} install
2209 @noindent where @var{path-to-rootdir} is the absolute path of
2210 a directory relative to which all installation paths will be
2211 interpreted. Note that the directory specified by @code{DESTDIR}
2212 need not exist yet; it will be created if necessary.
2214 There is a subtle point with tooldirs and @code{DESTDIR}:
2215 If you relocate a cross-compiler installation with
2216 e.g.@: @samp{DESTDIR=@var{rootdir}}, then the directory
2217 @file{@var{rootdir}/@var{exec-prefix}/@var{target-alias}/bin} will
2218 be filled with duplicated GCC executables only if it already exists,
2219 it will not be created otherwise. This is regarded as a feature,
2220 not as a bug, because it gives slightly more control to the packagers
2221 using the @code{DESTDIR} feature.
2223 If you are bootstrapping a released version of GCC then please
2224 quickly review the build status page for your release, available from
2225 @uref{http://gcc.gnu.org/buildstat.html}.
2226 If your system is not listed for the version of GCC that you built,
2228 @email{gcc@@gcc.gnu.org} indicating
2229 that you successfully built and installed GCC@.
2230 Include the following information:
2234 Output from running @file{@var{srcdir}/config.guess}. Do not send
2235 that file itself, just the one-line output from running it.
2238 The output of @samp{gcc -v} for your newly installed @command{gcc}.
2239 This tells us which version of GCC you built and the options you passed to
2243 Whether you enabled all languages or a subset of them. If you used a
2244 full distribution then this information is part of the configure
2245 options in the output of @samp{gcc -v}, but if you downloaded the
2246 ``core'' compiler plus additional front ends then it isn't apparent
2247 which ones you built unless you tell us about it.
2250 If the build was for GNU/Linux, also include:
2253 The distribution name and version (e.g., Red Hat 7.1 or Debian 2.2.3);
2254 this information should be available from @file{/etc/issue}.
2257 The version of the Linux kernel, available from @samp{uname --version}
2261 The version of glibc you used; for RPM-based systems like Red Hat,
2262 Mandrake, and SuSE type @samp{rpm -q glibc} to get the glibc version,
2263 and on systems like Debian and Progeny use @samp{dpkg -l libc6}.
2265 For other systems, you can include similar information if you think it is
2269 Any other information that you think would be useful to people building
2270 GCC on the same configuration. The new entry in the build status list
2271 will include a link to the archived copy of your message.
2274 We'd also like to know if the
2276 @ref{Specific, host/target specific installation notes}
2279 @uref{specific.html,,host/target specific installation notes}
2281 didn't include your host/target information or if that information is
2282 incomplete or out of date. Send a note to
2283 @email{gcc@@gcc.gnu.org} detailing how the information should be changed.
2285 If you find a bug, please report it following the
2286 @uref{../bugs.html,,bug reporting guidelines}.
2288 If you want to print the GCC manuals, do @samp{cd @var{objdir}; make
2289 dvi}. You will need to have @command{texi2dvi} (version at least 4.4)
2290 and @TeX{} installed. This creates a number of @file{.dvi} files in
2291 subdirectories of @file{@var{objdir}}; these may be converted for
2292 printing with programs such as @command{dvips}. Alternately, by using
2293 @samp{make pdf} in place of @samp{make dvi}, you can create documentation
2294 in the form of @file{.pdf} files; this requires @command{texi2pdf}, which
2295 is included with Texinfo version 4.8 and later. You can also
2296 @uref{http://www.gnu.org/order/order.html,,buy printed manuals from the
2297 Free Software Foundation}, though such manuals may not be for the most
2298 recent version of GCC@.
2300 If you would like to generate online HTML documentation, do @samp{cd
2301 @var{objdir}; make html} and HTML will be generated for the gcc manuals in
2302 @file{@var{objdir}/gcc/HTML}.
2309 @uref{./index.html,,Return to the GCC Installation page}
2313 @c ***Binaries****************************************************************
2315 @comment node-name, next, previous, up
2316 @node Binaries, Specific, Installing GCC, Top
2320 @chapter Installing GCC: Binaries
2323 @cindex Installing GCC: Binaries
2325 We are often asked about pre-compiled versions of GCC@. While we cannot
2326 provide these for all platforms, below you'll find links to binaries for
2327 various platforms where creating them by yourself is not easy due to various
2330 Please note that we did not create these binaries, nor do we
2331 support them. If you have any problems installing them, please
2332 contact their makers.
2339 @uref{http://www.bullfreeware.com,,Bull's Freeware and Shareware Archive for AIX};
2342 @uref{http://aixpdslib.seas.ucla.edu,,UCLA Software Library for AIX}.
2346 DOS---@uref{http://www.delorie.com/djgpp/,,DJGPP}.
2349 Renesas H8/300[HS]---@uref{http://h8300-hms.sourceforge.net/,,GNU
2350 Development Tools for the Renesas H8/300[HS] Series}.
2356 @uref{http://hpux.cs.utah.edu/,,HP-UX Porting Center};
2359 @uref{ftp://sunsite.informatik.rwth-aachen.de/pub/packages/gcc_hpux/,,Binaries for HP-UX 11.00 at Aachen University of Technology}.
2363 Motorola 68HC11/68HC12---@uref{http://www.gnu-m68hc11.org,,GNU
2364 Development Tools for the Motorola 68HC11/68HC12}.
2367 @uref{http://www.sco.com/skunkware/devtools/index.html#gcc,,SCO
2368 OpenServer/Unixware}.
2371 Solaris 2 (SPARC, Intel)---@uref{http://www.sunfreeware.com/,,Sunfreeware}.
2374 SGI---@uref{http://freeware.sgi.com/,,SGI Freeware}.
2380 The @uref{http://sourceware.org/cygwin/,,Cygwin} project;
2382 The @uref{http://www.mingw.org/,,MinGW} project.
2386 @uref{ftp://ftp.thewrittenword.com/packages/by-name/,,The
2387 Written Word} offers binaries for
2390 Digital UNIX 4.0D and 5.1,
2392 HP-UX 10.20, 11.00, and 11.11, and
2393 Solaris/SPARC 2.5.1, 2.6, 7, 8, and 9.
2396 @uref{http://www.openpkg.org/,,OpenPKG} offers binaries for quite a
2397 number of platforms.
2400 The @uref{http://gcc.gnu.org/wiki/GFortranBinaries,,GFortran Wiki} has
2401 links to GNU Fortran binaries for several platforms.
2404 In addition to those specific offerings, you can get a binary
2405 distribution CD-ROM from the
2406 @uref{http://www.gnu.org/order/order.html,,Free Software Foundation}.
2407 It contains binaries for a number of platforms, and
2408 includes not only GCC, but other stuff as well. The current CD does
2409 not contain the latest version of GCC, but it should allow
2410 bootstrapping the compiler. An updated version of that disk is in the
2418 @uref{./index.html,,Return to the GCC Installation page}
2422 @c ***Specific****************************************************************
2424 @comment node-name, next, previous, up
2425 @node Specific, Old, Binaries, Top
2429 @chapter Host/target specific installation notes for GCC
2432 @cindex Specific installation notes
2433 @cindex Target specific installation
2434 @cindex Host specific installation
2435 @cindex Target specific installation notes
2437 Please read this document carefully @emph{before} installing the
2438 GNU Compiler Collection on your machine.
2440 Note that this list of install notes is @emph{not} a list of supported
2441 hosts or targets. Not all supported hosts and targets are listed
2442 here, only the ones that require host-specific or target-specific
2448 @uref{#alpha-x-x,,alpha*-*-*}
2450 @uref{#alpha-dec-osf,,alpha*-dec-osf*}
2452 @uref{#alphaev5-cray-unicosmk,,alphaev5-cray-unicosmk*}
2454 @uref{#arc-x-elf,,arc-*-elf}
2456 @uref{#arm-x-elf,,arm-*-elf}
2457 @uref{#arm-x-coff,,arm-*-coff}
2458 @uref{#arm-x-aout,,arm-*-aout}
2460 @uref{#xscale-x-x,,xscale-*-*}
2464 @uref{#bfin,,Blackfin}
2470 @uref{#x-x-freebsd,,*-*-freebsd*}
2472 @uref{#h8300-hms,,h8300-hms}
2474 @uref{#hppa-hp-hpux,,hppa*-hp-hpux*}
2476 @uref{#hppa-hp-hpux10,,hppa*-hp-hpux10}
2478 @uref{#hppa-hp-hpux11,,hppa*-hp-hpux11}
2480 @uref{#x-x-linux-gnu,,*-*-linux-gnu}
2482 @uref{#ix86-x-linuxaout,,i?86-*-linux*aout}
2484 @uref{#ix86-x-linux,,i?86-*-linux*}
2486 @uref{#ix86-x-sco32v5,,i?86-*-sco3.2v5*}
2488 @uref{#ix86-x-solaris210,,i?86-*-solaris2.10}
2490 @uref{#ix86-x-udk,,i?86-*-udk}
2492 @uref{#ia64-x-linux,,ia64-*-linux}
2494 @uref{#ia64-x-hpux,,ia64-*-hpux*}
2496 @uref{#x-ibm-aix,,*-ibm-aix*}
2498 @uref{#iq2000-x-elf,,iq2000-*-elf}
2500 @uref{#m32c-x-elf,,m32c-*-elf}
2502 @uref{#m32r-x-elf,,m32r-*-elf}
2504 @uref{#m6811-elf,,m6811-elf}
2506 @uref{#m6812-elf,,m6812-elf}
2508 @uref{#m68k-x-x,,m68k-*-*}
2510 @uref{#m68k-hp-hpux,,m68k-hp-hpux}
2512 @uref{#m68k-uclinux,,m68k-uclinux}
2514 @uref{#mips-x-x,,mips-*-*}
2516 @uref{#mips-sgi-irix5,,mips-sgi-irix5}
2518 @uref{#mips-sgi-irix6,,mips-sgi-irix6}
2520 @uref{#powerpc-x-x,,powerpc*-*-*, powerpc-*-sysv4}
2522 @uref{#powerpc-x-darwin,,powerpc-*-darwin*}
2524 @uref{#powerpc-x-elf,,powerpc-*-elf, powerpc-*-sysv4}
2526 @uref{#powerpc-x-linux-gnu,,powerpc*-*-linux-gnu*}
2528 @uref{#powerpc-x-netbsd,,powerpc-*-netbsd*}
2530 @uref{#powerpc-x-eabisim,,powerpc-*-eabisim}
2532 @uref{#powerpc-x-eabi,,powerpc-*-eabi}
2534 @uref{#powerpcle-x-elf,,powerpcle-*-elf, powerpcle-*-sysv4}
2536 @uref{#powerpcle-x-eabisim,,powerpcle-*-eabisim}
2538 @uref{#powerpcle-x-eabi,,powerpcle-*-eabi}
2540 @uref{#s390-x-linux,,s390-*-linux*}
2542 @uref{#s390x-x-linux,,s390x-*-linux*}
2544 @uref{#s390x-ibm-tpf,,s390x-ibm-tpf*}
2546 @uref{#x-x-solaris2,,*-*-solaris2*}
2548 @uref{#sparc-sun-solaris2,,sparc-sun-solaris2*}
2550 @uref{#sparc-sun-solaris27,,sparc-sun-solaris2.7}
2552 @uref{#sparc-x-linux,,sparc-*-linux*}
2554 @uref{#sparc64-x-solaris2,,sparc64-*-solaris2*}
2556 @uref{#sparcv9-x-solaris2,,sparcv9-*-solaris2*}
2558 @uref{#x-x-sysv,,*-*-sysv*}
2560 @uref{#vax-dec-ultrix,,vax-dec-ultrix}
2562 @uref{#x-x-vxworks,,*-*-vxworks*}
2564 @uref{#x86-64-x-x,,x86_64-*-*, amd64-*-*}
2566 @uref{#xtensa-x-elf,,xtensa-*-elf}
2568 @uref{#xtensa-x-linux,,xtensa-*-linux*}
2570 @uref{#windows,,Microsoft Windows}
2574 @uref{#older,,Older systems}
2579 @uref{#elf,,all ELF targets} (SVR4, Solaris 2, etc.)
2585 <!-- -------- host/target specific issues start here ---------------- -->
2588 @heading @anchor{alpha-x-x}alpha*-*-*
2590 This section contains general configuration information for all
2591 alpha-based platforms using ELF (in particular, ignore this section for
2592 DEC OSF/1, Digital UNIX and Tru64 UNIX)@. In addition to reading this
2593 section, please read all other sections that match your target.
2595 We require binutils 2.11.2 or newer.
2596 Previous binutils releases had a number of problems with DWARF 2
2597 debugging information, not the least of which is incorrect linking of
2603 @heading @anchor{alpha-dec-osf}alpha*-dec-osf*
2604 Systems using processors that implement the DEC Alpha architecture and
2605 are running the DEC/Compaq Unix (DEC OSF/1, Digital UNIX, or Compaq
2606 Tru64 UNIX) operating system, for example the DEC Alpha AXP systems.
2608 As of GCC 3.2, versions before @code{alpha*-dec-osf4} are no longer
2609 supported. (These are the versions which identify themselves as DEC
2612 In Digital Unix V4.0, virtual memory exhausted bootstrap failures
2613 may be fixed by configuring with @option{--with-gc=simple},
2614 reconfiguring Kernel Virtual Memory and Swap parameters
2615 per the @command{/usr/sbin/sys_check} Tuning Suggestions,
2616 or applying the patch in
2617 @uref{http://gcc.gnu.org/ml/gcc/2002-08/msg00822.html}.
2619 In Tru64 UNIX V5.1, Compaq introduced a new assembler that does not
2620 currently (2001-06-13) work with @command{mips-tfile}. As a workaround,
2621 we need to use the old assembler, invoked via the barely documented
2622 @option{-oldas} option. To bootstrap GCC, you either need to use the
2626 % CC=cc @var{srcdir}/configure [@var{options}] [@var{target}]
2629 or you can use a copy of GCC 2.95.3 or higher built on Tru64 UNIX V4.0:
2632 % CC=gcc -Wa,-oldas @var{srcdir}/configure [@var{options}] [@var{target}]
2635 As of GNU binutils 2.11.2, neither GNU @command{as} nor GNU @command{ld}
2636 are supported on Tru64 UNIX, so you must not configure GCC with
2637 @option{--with-gnu-as} or @option{--with-gnu-ld}.
2639 GCC writes a @samp{.verstamp} directive to the assembler output file
2640 unless it is built as a cross-compiler. It gets the version to use from
2641 the system header file @file{/usr/include/stamp.h}. If you install a
2642 new version of DEC Unix, you should rebuild GCC to pick up the new version
2645 Note that since the Alpha is a 64-bit architecture, cross-compilers from
2646 32-bit machines will not generate code as efficient as that generated
2647 when the compiler is running on a 64-bit machine because many
2648 optimizations that depend on being able to represent a word on the
2649 target in an integral value on the host cannot be performed. Building
2650 cross-compilers on the Alpha for 32-bit machines has only been tested in
2651 a few cases and may not work properly.
2653 @samp{make compare} may fail on old versions of DEC Unix unless you add
2654 @option{-save-temps} to @code{CFLAGS}. On these systems, the name of the
2655 assembler input file is stored in the object file, and that makes
2656 comparison fail if it differs between the @code{stage1} and
2657 @code{stage2} compilations. The option @option{-save-temps} forces a
2658 fixed name to be used for the assembler input file, instead of a
2659 randomly chosen name in @file{/tmp}. Do not add @option{-save-temps}
2660 unless the comparisons fail without that option. If you add
2661 @option{-save-temps}, you will have to manually delete the @samp{.i} and
2662 @samp{.s} files after each series of compilations.
2664 GCC now supports both the native (ECOFF) debugging format used by DBX
2665 and GDB and an encapsulated STABS format for use only with GDB@. See the
2666 discussion of the @option{--with-stabs} option of @file{configure} above
2667 for more information on these formats and how to select them.
2669 There is a bug in DEC's assembler that produces incorrect line numbers
2670 for ECOFF format when the @samp{.align} directive is used. To work
2671 around this problem, GCC will not emit such alignment directives
2672 while writing ECOFF format debugging information even if optimization is
2673 being performed. Unfortunately, this has the very undesirable
2674 side-effect that code addresses when @option{-O} is specified are
2675 different depending on whether or not @option{-g} is also specified.
2677 To avoid this behavior, specify @option{-gstabs+} and use GDB instead of
2678 DBX@. DEC is now aware of this problem with the assembler and hopes to
2679 provide a fix shortly.
2684 @heading @anchor{alphaev5-cray-unicosmk}alphaev5-cray-unicosmk*
2685 Cray T3E systems running Unicos/Mk.
2687 This port is incomplete and has many known bugs. We hope to improve the
2688 support for this target soon. Currently, only the C front end is supported,
2689 and it is not possible to build parallel applications. Cray modules are not
2690 supported; in particular, Craylibs are assumed to be in
2691 @file{/opt/ctl/craylibs/craylibs}.
2693 On this platform, you need to tell GCC where to find the assembler and
2694 the linker. The simplest way to do so is by providing @option{--with-as}
2695 and @option{--with-ld} to @file{configure}, e.g.@:
2698 configure --with-as=/opt/ctl/bin/cam --with-ld=/opt/ctl/bin/cld \
2699 --enable-languages=c
2702 The comparison test at the end of the bootstrapping process fails on Unicos/Mk
2703 because the assembler inserts timestamps into object files. You should
2704 be able to work around this by doing @samp{make all} after getting this
2710 @heading @anchor{arc-x-elf}arc-*-elf
2711 Argonaut ARC processor.
2712 This configuration is intended for embedded systems.
2717 @heading @anchor{arm-x-elf}arm-*-elf
2718 @heading @anchor{xscale-x-x}xscale-*-*
2719 ARM-family processors. Subtargets that use the ELF object format
2720 require GNU binutils 2.13 or newer. Such subtargets include:
2721 @code{arm-*-freebsd}, @code{arm-*-netbsdelf}, @code{arm-*-*linux},
2722 @code{arm-*-rtems} and @code{arm-*-kaos}.
2727 @heading @anchor{arm-x-coff}arm-*-coff
2728 ARM-family processors. Note that there are two different varieties
2729 of PE format subtarget supported: @code{arm-wince-pe} and
2730 @code{arm-pe} as well as a standard COFF target @code{arm-*-coff}.
2735 @heading @anchor{arm-x-aout}arm-*-aout
2736 ARM-family processors. These targets support the AOUT file format:
2737 @code{arm-*-aout}, @code{arm-*-netbsd}.
2742 @heading @anchor{avr}avr
2744 ATMEL AVR-family micro controllers. These are used in embedded
2745 applications. There are no standard Unix configurations.
2747 @xref{AVR Options,, AVR Options, gcc, Using the GNU Compiler
2751 See ``AVR Options'' in the main manual
2753 for the list of supported MCU types.
2755 Use @samp{configure --target=avr --enable-languages="c"} to configure GCC@.
2757 Further installation notes and other useful information about AVR tools
2758 can also be obtained from:
2762 @uref{http://www.nongnu.org/avr/,,http://www.nongnu.org/avr/}
2764 @uref{http://home.overta.ru/users/denisc/,,http://home.overta.ru/users/denisc/}
2766 @uref{http://www.amelek.gda.pl/avr/,,http://www.amelek.gda.pl/avr/}
2769 We @emph{strongly} recommend using binutils 2.13 or newer.
2771 The following error:
2773 Error: register required
2776 indicates that you should upgrade to a newer version of the binutils.
2781 @heading @anchor{bfin}Blackfin
2783 The Blackfin processor, an Analog Devices DSP.
2785 @xref{Blackfin Options,, Blackfin Options, gcc, Using the GNU Compiler
2789 See ``Blackfin Options'' in the main manual
2792 More information, and a version of binutils with support for this processor,
2793 is available at @uref{http://blackfin.uclinux.org}
2798 @heading @anchor{c4x}c4x
2800 Texas Instruments TMS320C3x and TMS320C4x Floating Point Digital Signal
2801 Processors. These are used in embedded applications. There are no
2802 standard Unix configurations.
2804 @xref{TMS320C3x/C4x Options,, TMS320C3x/C4x Options, gcc, Using the
2805 GNU Compiler Collection (GCC)},
2808 See ``TMS320C3x/C4x Options'' in the main manual
2810 for the list of supported MCU types.
2812 GCC can be configured as a cross compiler for both the C3x and C4x
2813 architectures on the same system. Use @samp{configure --target=c4x
2814 --enable-languages="c,c++"} to configure.
2817 Further installation notes and other useful information about C4x tools
2818 can also be obtained from:
2822 @uref{http://www.elec.canterbury.ac.nz/c4x/,,http://www.elec.canterbury.ac.nz/c4x/}
2828 @heading @anchor{cris}CRIS
2830 CRIS is the CPU architecture in Axis Communications ETRAX system-on-a-chip
2831 series. These are used in embedded applications.
2834 @xref{CRIS Options,, CRIS Options, gcc, Using the GNU Compiler
2838 See ``CRIS Options'' in the main manual
2840 for a list of CRIS-specific options.
2842 There are a few different CRIS targets:
2844 @item cris-axis-aout
2845 Old target. Includes a multilib for the @samp{elinux} a.out-based
2846 target. No multilibs for newer architecture variants.
2848 Mainly for monolithic embedded systems. Includes a multilib for the
2849 @samp{v10} core used in @samp{ETRAX 100 LX}.
2850 @item cris-axis-linux-gnu
2851 A GNU/Linux port for the CRIS architecture, currently targeting
2852 @samp{ETRAX 100 LX} by default.
2855 For @code{cris-axis-aout} and @code{cris-axis-elf} you need binutils 2.11
2856 or newer. For @code{cris-axis-linux-gnu} you need binutils 2.12 or newer.
2858 Pre-packaged tools can be obtained from
2859 @uref{ftp://ftp.axis.com/pub/axis/tools/cris/compiler-kit/}. More
2860 information about this platform is available at
2861 @uref{http://developer.axis.com/}.
2866 @heading @anchor{crx}CRX
2868 The CRX CompactRISC architecture is a low-power 32-bit architecture with
2869 fast context switching and architectural extensibility features.
2872 @xref{CRX Options,, CRX Options, gcc, Using and Porting the GNU Compiler
2877 See ``CRX Options'' in the main manual for a list of CRX-specific options.
2880 Use @samp{configure --target=crx-elf --enable-languages=c,c++} to configure
2881 GCC@ for building a CRX cross-compiler. The option @samp{--target=crx-elf}
2882 is also used to build the @samp{newlib} C library for CRX.
2884 It is also possible to build libstdc++-v3 for the CRX architecture. This
2885 needs to be done in a separate step with the following configure settings:
2886 @samp{gcc/libstdc++-v3/configure --host=crx-elf --with-newlib
2887 --enable-sjlj-exceptions --enable-cxx-flags='-fexceptions -frtti'}
2892 @heading @anchor{dos}DOS
2894 Please have a look at the @uref{binaries.html,,binaries page}.
2896 You cannot install GCC by itself on MSDOS; it will not compile under
2897 any MSDOS compiler except itself. You need to get the complete
2898 compilation package DJGPP, which includes binaries as well as sources,
2899 and includes all the necessary compilation tools and libraries.
2904 @heading @anchor{x-x-freebsd}*-*-freebsd*
2906 The version of binutils installed in @file{/usr/bin} probably works with
2907 this release of GCC@. However, on FreeBSD 4, bootstrapping against the
2908 latest FSF binutils is known to improve overall testsuite results; and,
2909 on FreeBSD/alpha, using binutils 2.14 or later is required to build libjava.
2911 Support for FreeBSD 1 was discontinued in GCC 3.2.
2913 Support for FreeBSD 2 will be discontinued after GCC 3.4. The
2914 following was true for GCC 3.1 but the current status is unknown.
2915 For FreeBSD 2 or any mutant a.out versions of FreeBSD 3: All
2916 configuration support and files as shipped with GCC 2.95 are still in
2917 place. FreeBSD 2.2.7 has been known to bootstrap completely; however,
2918 it is unknown which version of binutils was used (it is assumed that it
2919 was the system copy in @file{/usr/bin}) and C++ EH failures were noted.
2921 For FreeBSD using the ELF file format: DWARF 2 debugging is now the
2922 default for all CPU architectures. It had been the default on
2923 FreeBSD/alpha since its inception. You may use @option{-gstabs} instead
2924 of @option{-g}, if you really want the old debugging format. There are
2925 no known issues with mixing object files and libraries with different
2926 debugging formats. Otherwise, this release of GCC should now match more
2927 of the configuration used in the stock FreeBSD configuration of GCC@. In
2928 particular, @option{--enable-threads} is now configured by default.
2929 However, as a general user, do not attempt to replace the system
2930 compiler with this release. Known to bootstrap and check with good
2931 results on FreeBSD 4.9-STABLE and 5-CURRENT@. In the past, known to
2932 bootstrap and check with good results on FreeBSD 3.0, 3.4, 4.0, 4.2,
2933 4.3, 4.4, 4.5, 4.8-STABLE@.
2935 In principle, @option{--enable-threads} is now compatible with
2936 @option{--enable-libgcj} on FreeBSD@. However, it has only been built
2937 and tested on @samp{i386-*-freebsd[45]} and @samp{alpha-*-freebsd[45]}.
2939 library may be incorrectly built (symbols are missing at link time).
2940 There is a rare timing-based startup hang (probably involves an
2941 assumption about the thread library). Multi-threaded boehm-gc (required for
2942 libjava) exposes severe threaded signal-handling bugs on FreeBSD before
2943 4.5-RELEASE@. Other CPU architectures
2944 supported by FreeBSD will require additional configuration tuning in, at
2945 the very least, both boehm-gc and libffi.
2947 Shared @file{libgcc_s.so} is now built and installed by default.
2952 @heading @anchor{h8300-hms}h8300-hms
2953 Renesas H8/300 series of processors.
2955 Please have a look at the @uref{binaries.html,,binaries page}.
2957 The calling convention and structure layout has changed in release 2.6.
2958 All code must be recompiled. The calling convention now passes the
2959 first three arguments in function calls in registers. Structures are no
2960 longer a multiple of 2 bytes.
2965 @heading @anchor{hppa-hp-hpux}hppa*-hp-hpux*
2966 Support for HP-UX version 9 and older was discontinued in GCC 3.4.
2968 We require using gas/binutils on all hppa platforms;
2969 you may encounter a variety of problems if you try to use the HP assembler.
2971 Specifically, @option{-g} does not work on HP-UX (since that system
2972 uses a peculiar debugging format which GCC does not know about), unless
2973 you use GAS and GDB@. It may be helpful to configure GCC with the
2974 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}} and
2975 @option{--with-as=@dots{}} options to ensure that GCC can find GAS@.
2977 If you wish to use the pa-risc 2.0 architecture support with a 32-bit
2978 runtime, you must use gas/binutils 2.11 or newer.
2980 There are two default scheduling models for instructions. These are
2981 PROCESSOR_7100LC and PROCESSOR_8000. They are selected from the pa-risc
2982 architecture specified for the target machine when configuring.
2983 PROCESSOR_8000 is the default. PROCESSOR_7100LC is selected when
2984 the target is a @samp{hppa1*} machine.
2986 The PROCESSOR_8000 model is not well suited to older processors. Thus,
2987 it is important to completely specify the machine architecture when
2988 configuring if you want a model other than PROCESSOR_8000. The macro
2989 TARGET_SCHED_DEFAULT can be defined in BOOT_CFLAGS if a different
2990 default scheduling model is desired.
2992 As of GCC 4.0, GCC uses the UNIX 95 namespace for HP-UX 10.10
2993 through 11.00, and the UNIX 98 namespace for HP-UX 11.11 and later.
2994 This namespace change might cause problems when bootstrapping with
2995 an earlier version of GCC or the HP compiler as essentially the same
2996 namespace is required for an entire build. This problem can be avoided
2997 in a number of ways. With HP cc, @env{UNIX_STD} can be set to @samp{95}
2998 or @samp{98}. Another way is to add an appropriate set of predefines
2999 to @env{CC}. The description for the @option{munix=} option contains
3000 a list of the predefines used with each standard.
3002 As of GCC 4.1, @env{DWARF2} exception handling is available on HP-UX.
3003 It is now the default. This exposed a bug in the handling of data
3004 relocations in the GAS assembler. The handling of 64-bit data relocations
3005 was seriously broken, affecting debugging and exception support on all
3006 @samp{hppa64-*-*} targets. Under some circumstances, 32-bit data relocations
3007 could also be handled incorrectly. This problem is fixed in GAS version
3010 GCC versions prior to 4.1 incorrectly passed and returned complex
3011 values. They are now passed in the same manner as aggregates.
3013 More specific information to @samp{hppa*-hp-hpux*} targets follows.
3018 @heading @anchor{hppa-hp-hpux10}hppa*-hp-hpux10
3020 For hpux10.20, we @emph{highly} recommend you pick up the latest sed patch
3021 @code{PHCO_19798} from HP@. HP has two sites which provide patches free of
3027 <a href="http://us.itrc.hp.com/service/home/home.do">US, Canada, Asia-Pacific, and
3031 @uref{http://us.itrc.hp.com/service/home/home.do,,} US, Canada, Asia-Pacific,
3035 @uref{http://europe.itrc.hp.com/service/home/home.do,,} Europe.
3038 The HP assembler on these systems has some problems. Most notably the
3039 assembler inserts timestamps into each object file it creates, causing
3040 the 3-stage comparison test to fail during a bootstrap.
3041 You should be able to continue by saying @samp{make all-host all-target}
3042 after getting the failure from @samp{make}.
3044 GCC 4.0 requires CVS binutils as of April 28, 2004 or later. Earlier
3045 versions require binutils 2.8 or later.
3047 The C++ ABI has changed incompatibly in GCC 4.0. COMDAT subspaces are
3048 used for one-only code and data. This resolves many of the previous
3049 problems in using C++ on this target. However, the ABI is not compatible
3050 with the one implemented under HP-UX 11 using secondary definitions.
3055 @heading @anchor{hppa-hp-hpux11}hppa*-hp-hpux11
3057 GCC 3.0 and up support HP-UX 11. GCC 2.95.x is not supported and cannot
3058 be used to compile GCC 3.0 and up.
3060 Refer to @uref{binaries.html,,binaries} for information about obtaining
3061 precompiled GCC binaries for HP-UX@. Precompiled binaries must be obtained
3062 to build the Ada language as it can't be bootstrapped using C@. Ada is
3063 only available for the 32-bit PA-RISC runtime. The libffi and libjava
3064 haven't been ported to HP-UX and don't build.
3066 Starting with GCC 3.4 an ISO C compiler is required to bootstrap. The
3067 bundled compiler supports only traditional C; you will need either HP's
3068 unbundled compiler, or a binary distribution of GCC@.
3070 It is possible to build GCC 3.3 starting with the bundled HP compiler,
3071 but the process requires several steps. GCC 3.3 can then be used to
3072 build later versions. The fastjar program contains ISO C code and
3073 can't be built with the HP bundled compiler. This problem can be
3074 avoided by not building the Java language. For example, use the
3075 @option{--enable-languages="c,c++,f77,objc"} option in your configure
3078 There are several possible approaches to building the distribution.
3079 Binutils can be built first using the HP tools. Then, the GCC
3080 distribution can be built. The second approach is to build GCC
3081 first using the HP tools, then build binutils, then rebuild GCC@.
3082 There have been problems with various binary distributions, so it
3083 is best not to start from a binary distribution.
3085 On 64-bit capable systems, there are two distinct targets. Different
3086 installation prefixes must be used if both are to be installed on
3087 the same system. The @samp{hppa[1-2]*-hp-hpux11*} target generates code
3088 for the 32-bit PA-RISC runtime architecture and uses the HP linker.
3089 The @samp{hppa64-hp-hpux11*} target generates 64-bit code for the
3090 PA-RISC 2.0 architecture. The HP and GNU linkers are both supported
3093 The script config.guess now selects the target type based on the compiler
3094 detected during configuration. You must define @env{PATH} or @env{CC} so
3095 that configure finds an appropriate compiler for the initial bootstrap.
3096 When @env{CC} is used, the definition should contain the options that are
3097 needed whenever @env{CC} is used.
3099 Specifically, options that determine the runtime architecture must be
3100 in @env{CC} to correctly select the target for the build. It is also
3101 convenient to place many other compiler options in @env{CC}. For example,
3102 @env{CC="cc -Ac +DA2.0W -Wp,-H16376 -D_CLASSIC_TYPES -D_HPUX_SOURCE"}
3103 can be used to bootstrap the GCC 3.3 branch with the HP compiler in
3104 64-bit K&R/bundled mode. The @option{+DA2.0W} option will result in
3105 the automatic selection of the @samp{hppa64-hp-hpux11*} target. The
3106 macro definition table of cpp needs to be increased for a successful
3107 build with the HP compiler. _CLASSIC_TYPES and _HPUX_SOURCE need to
3108 be defined when building with the bundled compiler, or when using the
3109 @option{-Ac} option. These defines aren't necessary with @option{-Ae}.
3111 It is best to explicitly configure the @samp{hppa64-hp-hpux11*} target
3112 with the @option{--with-ld=@dots{}} option. This overrides the standard
3113 search for ld. The two linkers supported on this target require different
3114 commands. The default linker is determined during configuration. As a
3115 result, it's not possible to switch linkers in the middle of a GCC build.
3116 This has been been reported to sometimes occur in unified builds of
3119 GCC 3.0 through 3.2 require binutils 2.11 or above. GCC 3.3 through
3120 GCC 4.0 require binutils 2.14 or later.
3122 Although the HP assembler can be used for an initial build, it shouldn't
3123 be used with any languages other than C and perhaps Fortran due to its
3124 many limitations. For example, it does not support weak symbols or alias
3125 definitions. As a result, explicit template instantiations are required
3126 when using C++. This makes it difficult if not impossible to build many
3127 C++ applications. You can't generate debugging information when using
3128 the HP assembler. Finally, bootstrapping fails in the final
3129 comparison of object modules due to the time stamps that it inserts into
3130 the modules. The bootstrap can be continued from this point with
3131 @samp{make all-host all-target}.
3133 A recent linker patch must be installed for the correct operation of
3134 GCC 3.3 and later. @code{PHSS_26559} and @code{PHSS_24304} are the
3135 oldest linker patches that are known to work. They are for HP-UX
3136 11.00 and 11.11, respectively. @code{PHSS_24303}, the companion to
3137 @code{PHSS_24304}, might be usable but it hasn't been tested. These
3138 patches have been superseded. Consult the HP patch database to obtain
3139 the currently recommended linker patch for your system.
3141 The patches are necessary for the support of weak symbols on the
3142 32-bit port, and for the running of initializers and finalizers. Weak
3143 symbols are implemented using SOM secondary definition symbols. Prior
3144 to HP-UX 11, there are bugs in the linker support for secondary symbols.
3145 The patches correct a problem of linker core dumps creating shared
3146 libraries containing secondary symbols, as well as various other
3147 linking issues involving secondary symbols.
3149 GCC 3.3 uses the ELF DT_INIT_ARRAY and DT_FINI_ARRAY capabilities to
3150 run initializers and finalizers on the 64-bit port. The 32-bit port
3151 uses the linker @option{+init} and @option{+fini} options for the same
3152 purpose. The patches correct various problems with the +init/+fini
3153 options, including program core dumps. Binutils 2.14 corrects a
3154 problem on the 64-bit port resulting from HP's non-standard use of
3155 the .init and .fini sections for array initializers and finalizers.
3157 There are a number of issues to consider in selecting which linker to
3158 use with the 64-bit port. The GNU 64-bit linker can only create dynamic
3159 binaries. The @option{-static} option causes linking with archive
3160 libraries but doesn't produce a truly static binary. Dynamic binaries
3161 still require final binding by the dynamic loader to resolve a set of
3162 dynamic-loader-defined symbols. The default behavior of the HP linker
3163 is the same as the GNU linker. However, it can generate true 64-bit
3164 static binaries using the @option{+compat} option.
3166 The HP 64-bit linker doesn't support linkonce semantics. As a
3167 result, C++ programs have many more sections than they should.
3169 The GNU 64-bit linker has some issues with shared library support
3170 and exceptions. As a result, we only support libgcc in archive
3171 format. For similar reasons, dwarf2 unwind and exception support
3172 are disabled. The GNU linker also has problems creating binaries
3173 with @option{-static}. It doesn't provide stubs for internal
3174 calls to global functions in shared libraries, so these calls
3175 can't be overloaded.
3177 Thread support is not implemented in GCC 3.0 through 3.2, so the
3178 @option{--enable-threads} configure option does not work. In 3.3
3179 and later, POSIX threads are supported. The optional DCE thread
3180 library is not supported.
3182 This port still is undergoing significant development.
3187 @heading @anchor{x-x-linux-gnu}*-*-linux-gnu
3189 Versions of libstdc++-v3 starting with 3.2.1 require bug fixes present
3190 in glibc 2.2.5 and later. More information is available in the
3191 libstdc++-v3 documentation.
3196 @heading @anchor{ix86-x-linuxaout}i?86-*-linux*aout
3197 Use this configuration to generate @file{a.out} binaries on Linux-based
3198 GNU systems. This configuration is being superseded.
3203 @heading @anchor{ix86-x-linux}i?86-*-linux*
3205 As of GCC 3.3, binutils 2.13.1 or later is required for this platform.
3206 See @uref{http://gcc.gnu.org/PR10877,,bug 10877} for more information.
3208 If you receive Signal 11 errors when building on GNU/Linux, then it is
3209 possible you have a hardware problem. Further information on this can be
3210 found on @uref{http://www.bitwizard.nl/sig11/,,www.bitwizard.nl}.
3215 @heading @anchor{ix86-x-sco32v5}i?86-*-sco3.2v5*
3216 Use this for the SCO OpenServer Release 5 family of operating systems.
3218 Unlike earlier versions of GCC, the ability to generate COFF with this
3219 target is no longer provided.
3221 Earlier versions of GCC emitted DWARF 1 when generating ELF to allow
3222 the system debugger to be used. That support was too burdensome to
3223 maintain. GCC now emits only DWARF 2 for this target. This means you
3224 may use either the UDK debugger or GDB to debug programs built by this
3227 GCC is now only supported on releases 5.0.4 and later, and requires that
3228 you install Support Level Supplement OSS646B or later, and Support Level
3229 Supplement OSS631C or later. If you are using release 5.0.7 of
3230 OpenServer, you must have at least the first maintenance pack installed
3231 (this includes the relevant portions of OSS646). OSS646, also known as
3232 the ``Execution Environment Update'', provides updated link editors and
3233 assemblers, as well as updated standard C and math libraries. The C
3234 startup modules are also updated to support the System V gABI draft, and
3235 GCC relies on that behavior. OSS631 provides a collection of commonly
3236 used open source libraries, some of which GCC depends on (such as GNU
3237 gettext and zlib). SCO OpenServer Release 5.0.7 has all of this built
3238 in by default, but OSS631C and later also apply to that release. Please
3240 @uref{ftp://ftp.sco.com/pub/openserver5,,ftp://ftp.sco.com/pub/openserver5}
3241 for the latest versions of these (and other potentially useful)
3244 Although there is support for using the native assembler, it is
3245 recommended that you configure GCC to use the GNU assembler. You do
3246 this by using the flags
3247 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}}. You should
3248 use a modern version of GNU binutils. Version 2.13.2.1 was used for all
3249 testing. In general, only the @option{--with-gnu-as} option is tested.
3250 A modern bintuils (as well as a plethora of other development related
3251 GNU utilities) can be found in Support Level Supplement OSS658A, the
3252 ``GNU Development Tools'' package. See the SCO web and ftp sites for details.
3253 That package also contains the currently ``officially supported'' version of
3254 GCC, version 2.95.3. It is useful for bootstrapping this version.
3259 @heading @anchor{ix86-x-solaris210}i?86-*-solaris2.10
3260 Use this for Solaris 10 or later on x86 and x86-64 systems. This
3261 configuration is supported by GCC 4.0 and later versions only.
3263 It is recommended that you configure GCC to use the GNU assembler in
3264 @file{/usr/sfw/bin/gas} but the Sun linker, using the options
3265 @option{--with-gnu-as --with-as=/usr/sfw/bin/gas --without-gnu-ld
3266 --with-ld=/usr/ccs/bin/ld}.
3271 @heading @anchor{ix86-x-udk}i?86-*-udk
3273 This target emulates the SCO Universal Development Kit and requires that
3274 package be installed. (If it is installed, you will have a
3275 @file{/udk/usr/ccs/bin/cc} file present.) It's very much like the
3276 @samp{i?86-*-unixware7*} target
3277 but is meant to be used when hosting on a system where UDK isn't the
3278 default compiler such as OpenServer 5 or Unixware 2. This target will
3279 generate binaries that will run on OpenServer, Unixware 2, or Unixware 7,
3280 with the same warnings and caveats as the SCO UDK@.
3282 This target is a little tricky to build because we have to distinguish
3283 it from the native tools (so it gets headers, startups, and libraries
3284 from the right place) while making the tools not think we're actually
3285 building a cross compiler. The easiest way to do this is with a configure
3289 CC=/udk/usr/ccs/bin/cc @var{/your/path/to}/gcc/configure \
3290 --host=i686-pc-udk --target=i686-pc-udk --program-prefix=udk-
3293 @emph{You should substitute @samp{i686} in the above command with the appropriate
3294 processor for your host.}
3296 After the usual @samp{make} and
3297 @samp{make install}, you can then access the UDK-targeted GCC
3298 tools by adding @command{udk-} before the commonly known name. For
3299 example, to invoke the C compiler, you would use @command{udk-gcc}.
3300 They will coexist peacefully with any native-target GCC tools you may
3307 @heading @anchor{ia64-x-linux}ia64-*-linux
3308 IA-64 processor (also known as IPF, or Itanium Processor Family)
3311 If you are using the installed system libunwind library with
3312 @option{--with-system-libunwind}, then you must use libunwind 0.98 or
3315 None of the following versions of GCC has an ABI that is compatible
3316 with any of the other versions in this list, with the exception that
3317 Red Hat 2.96 and Trillian 000171 are compatible with each other:
3318 3.1, 3.0.2, 3.0.1, 3.0, Red Hat 2.96, and Trillian 000717.
3319 This primarily affects C++ programs and programs that create shared libraries.
3320 GCC 3.1 or later is recommended for compiling linux, the kernel.
3321 As of version 3.1 GCC is believed to be fully ABI compliant, and hence no
3322 more major ABI changes are expected.
3327 @heading @anchor{ia64-x-hpux}ia64-*-hpux*
3328 Building GCC on this target requires the GNU Assembler. The bundled HP
3329 assembler will not work. To prevent GCC from using the wrong assembler,
3330 the option @option{--with-gnu-as} may be necessary.
3332 The GCC libunwind library has not been ported to HPUX@. This means that for
3333 GCC versions 3.2.3 and earlier, @option{--enable-libunwind-exceptions}
3334 is required to build GCC@. For GCC 3.3 and later, this is the default.
3335 For gcc 3.4.3 and later, @option{--enable-libunwind-exceptions} is
3336 removed and the system libunwind library will always be used.
3340 <!-- rs6000-ibm-aix*, powerpc-ibm-aix* -->
3342 @heading @anchor{x-ibm-aix}*-ibm-aix*
3343 Support for AIX version 3 and older was discontinued in GCC 3.4.
3345 ``out of memory'' bootstrap failures may indicate a problem with
3346 process resource limits (ulimit). Hard limits are configured in the
3347 @file{/etc/security/limits} system configuration file.
3349 To speed up the configuration phases of bootstrapping and installing GCC,
3350 one may use GNU Bash instead of AIX @command{/bin/sh}, e.g.,
3353 % CONFIG_SHELL=/opt/freeware/bin/bash
3354 % export CONFIG_SHELL
3357 and then proceed as described in @uref{build.html,,the build
3358 instructions}, where we strongly recommend specifying an absolute path
3359 to invoke @var{srcdir}/configure.
3361 Because GCC on AIX is built as a 32-bit executable by default,
3362 (although it can generate 64-bit programs) the GMP and MPFR libraries
3363 required by gfortran must be 32-bit libraries. Building GMP and MPFR
3364 as static archive libraries works better than shared libraries.
3366 Errors involving @code{alloca} when building GCC generally are due
3367 to an incorrect definition of @code{CC} in the Makefile or mixing files
3368 compiled with the native C compiler and GCC@. During the stage1 phase of
3369 the build, the native AIX compiler @strong{must} be invoked as @command{cc}
3370 (not @command{xlc}). Once @command{configure} has been informed of
3371 @command{xlc}, one needs to use @samp{make distclean} to remove the
3372 configure cache files and ensure that @env{CC} environment variable
3373 does not provide a definition that will confuse @command{configure}.
3374 If this error occurs during stage2 or later, then the problem most likely
3375 is the version of Make (see above).
3377 The native @command{as} and @command{ld} are recommended for bootstrapping
3378 on AIX 4 and required for bootstrapping on AIX 5L@. The GNU Assembler
3379 reports that it supports WEAK symbols on AIX 4, which causes GCC to try to
3380 utilize weak symbol functionality although it is not supported. The GNU
3381 Assembler and Linker do not support AIX 5L sufficiently to bootstrap GCC@.
3382 The native AIX tools do interoperate with GCC@.
3384 Building @file{libstdc++.a} requires a fix for an AIX Assembler bug
3385 APAR IY26685 (AIX 4.3) or APAR IY25528 (AIX 5.1). It also requires a
3386 fix for another AIX Assembler bug and a co-dependent AIX Archiver fix
3387 referenced as APAR IY53606 (AIX 5.2) or a APAR IY54774 (AIX 5.1)
3389 @samp{libstdc++} in GCC 3.4 increments the major version number of the
3390 shared object and GCC installation places the @file{libstdc++.a}
3391 shared library in a common location which will overwrite the and GCC
3392 3.3 version of the shared library. Applications either need to be
3393 re-linked against the new shared library or the GCC 3.1 and GCC 3.3
3394 versions of the @samp{libstdc++} shared object needs to be available
3395 to the AIX runtime loader. The GCC 3.1 @samp{libstdc++.so.4}, if
3396 present, and GCC 3.3 @samp{libstdc++.so.5} shared objects can be
3397 installed for runtime dynamic loading using the following steps to set
3398 the @samp{F_LOADONLY} flag in the shared object for @emph{each}
3399 multilib @file{libstdc++.a} installed:
3401 Extract the shared objects from the currently installed
3402 @file{libstdc++.a} archive:
3404 % ar -x libstdc++.a libstdc++.so.4 libstdc++.so.5
3407 Enable the @samp{F_LOADONLY} flag so that the shared object will be
3408 available for runtime dynamic loading, but not linking:
3410 % strip -e libstdc++.so.4 libstdc++.so.5
3413 Archive the runtime-only shared object in the GCC 3.4
3414 @file{libstdc++.a} archive:
3416 % ar -q libstdc++.a libstdc++.so.4 libstdc++.so.5
3419 Linking executables and shared libraries may produce warnings of
3420 duplicate symbols. The assembly files generated by GCC for AIX always
3421 have included multiple symbol definitions for certain global variable
3422 and function declarations in the original program. The warnings should
3423 not prevent the linker from producing a correct library or runnable
3426 AIX 4.3 utilizes a ``large format'' archive to support both 32-bit and
3427 64-bit object modules. The routines provided in AIX 4.3.0 and AIX 4.3.1
3428 to parse archive libraries did not handle the new format correctly.
3429 These routines are used by GCC and result in error messages during
3430 linking such as ``not a COFF file''. The version of the routines shipped
3431 with AIX 4.3.1 should work for a 32-bit environment. The @option{-g}
3432 option of the archive command may be used to create archives of 32-bit
3433 objects using the original ``small format''. A correct version of the
3434 routines is shipped with AIX 4.3.2 and above.
3436 Some versions of the AIX binder (linker) can fail with a relocation
3437 overflow severe error when the @option{-bbigtoc} option is used to link
3438 GCC-produced object files into an executable that overflows the TOC@. A fix
3439 for APAR IX75823 (OVERFLOW DURING LINK WHEN USING GCC AND -BBIGTOC) is
3440 available from IBM Customer Support and from its
3441 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
3442 website as PTF U455193.
3444 The AIX 4.3.2.1 linker (bos.rte.bind_cmds Level 4.3.2.1) will dump core
3445 with a segmentation fault when invoked by any version of GCC@. A fix for
3446 APAR IX87327 is available from IBM Customer Support and from its
3447 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
3448 website as PTF U461879. This fix is incorporated in AIX 4.3.3 and above.
3450 The initial assembler shipped with AIX 4.3.0 generates incorrect object
3451 files. A fix for APAR IX74254 (64BIT DISASSEMBLED OUTPUT FROM COMPILER FAILS
3452 TO ASSEMBLE/BIND) is available from IBM Customer Support and from its
3453 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
3454 website as PTF U453956. This fix is incorporated in AIX 4.3.1 and above.
3456 AIX provides National Language Support (NLS)@. Compilers and assemblers
3457 use NLS to support locale-specific representations of various data
3458 formats including floating-point numbers (e.g., @samp{.} vs @samp{,} for
3459 separating decimal fractions). There have been problems reported where
3460 GCC does not produce the same floating-point formats that the assembler
3461 expects. If one encounters this problem, set the @env{LANG}
3462 environment variable to @samp{C} or @samp{En_US}.
3464 By default, GCC for AIX 4.1 and above produces code that can be used on
3465 both Power or PowerPC processors.
3467 A default can be specified with the @option{-mcpu=@var{cpu_type}}
3468 switch and using the configure option @option{--with-cpu-@var{cpu_type}}.
3473 @heading @anchor{iq2000-x-elf}iq2000-*-elf
3474 Vitesse IQ2000 processors. These are used in embedded
3475 applications. There are no standard Unix configurations.
3480 @heading @anchor{m32c-x-elf}m32c-*-elf
3481 Renesas M32C processor.
3482 This configuration is intended for embedded systems.
3487 @heading @anchor{m32r-x-elf}m32r-*-elf
3488 Renesas M32R processor.
3489 This configuration is intended for embedded systems.
3494 @heading @anchor{m6811-elf}m6811-elf
3495 Motorola 68HC11 family micro controllers. These are used in embedded
3496 applications. There are no standard Unix configurations.
3501 @heading @anchor{m6812-elf}m6812-elf
3502 Motorola 68HC12 family micro controllers. These are used in embedded
3503 applications. There are no standard Unix configurations.
3508 @heading @anchor{m68k-x-x}m68k-*-*
3509 By default, @samp{m68k-*-aout}, @samp{m68k-*-coff*},
3510 @samp{m68k-*-elf*}, @samp{m68k-*-rtems} and @samp{m68k-*-uclinux}
3511 build libraries for both M680x0 and ColdFire processors. If you only
3512 need the M680x0 libraries, you can omit the ColdFire ones by passing
3513 @option{--with-arch=m68k} to @command{configure}. Alternatively, you
3514 can omit the M680x0 libraries by passing @option{--with-arch=cf} to
3515 @command{configure}. These targets default to 5206 code when
3516 configured with @option{--with-arch=cf} and 68020 code otherwise.
3518 The @samp{m68k-*-linux-gnu}, @samp{m68k-*-netbsd} and
3519 @samp{m68k-*-openbsd} targets also support the @option{--with-arch}
3520 option. They will generate ColdFire CFV4e code when configured with
3521 @option{--with-arch=cf} and 68020 code otherwise.
3523 You can override the default processors listed above by configuring
3524 with @option{--with-cpu=@var{target}}. This @var{target} can either
3525 be a @option{-mcpu} argument or one of the following values:
3526 @samp{m68000}, @samp{m68010}, @samp{m68020}, @samp{m68030},
3527 @samp{m68040}, @samp{m68060}, @samp{m68020-40} and @samp{m68020-60}.
3532 @heading @anchor{m68k-hp-hpux}m68k-hp-hpux
3533 HP 9000 series 300 or 400 running HP-UX@. HP-UX version 8.0 has a bug in
3534 the assembler that prevents compilation of GCC@. This
3535 bug manifests itself during the first stage of compilation, while
3536 building @file{libgcc2.a}:
3540 cc1: warning: `-g' option not supported on this version of GCC
3541 cc1: warning: `-g1' option not supported on this version of GCC
3542 ./xgcc: Internal compiler error: program as got fatal signal 11
3545 A patched version of the assembler is available as the file
3546 @uref{ftp://altdorf.ai.mit.edu/archive/cph/hpux-8.0-assembler}. If you
3547 have HP software support, the patch can also be obtained directly from
3548 HP, as described in the following note:
3551 This is the patched assembler, to patch SR#1653-010439, where the
3552 assembler aborts on floating point constants.
3554 The bug is not really in the assembler, but in the shared library
3555 version of the function ``cvtnum(3c)''. The bug on ``cvtnum(3c)'' is
3556 SR#4701-078451. Anyway, the attached assembler uses the archive
3557 library version of ``cvtnum(3c)'' and thus does not exhibit the bug.
3560 This patch is also known as PHCO_4484.
3562 In addition gdb does not understand that native HP-UX format, so
3563 you must use gas if you wish to use gdb.
3565 On HP-UX version 8.05, but not on 8.07 or more recent versions, the
3566 @command{fixproto} shell script triggers a bug in the system shell. If you
3567 encounter this problem, upgrade your operating system or use BASH (the
3568 GNU shell) to run @command{fixproto}. This bug will cause the fixproto
3569 program to report an error of the form:
3572 ./fixproto: sh internal 1K buffer overflow
3575 To fix this, you can also change the first line of the fixproto script
3585 @heading @anchor{m68k-x-uclinux}m68k-*-uclinux
3586 GCC 4.3 changed the uClinux configuration so that it uses the
3587 @samp{m68k-linux-gnu} ABI rather than the @samp{m68k-elf} ABI.
3588 It also added improved support for C++ and flat shared libraries,
3589 both of which were ABI changes. However, you can still use the
3590 original ABI by configuring for @samp{m68k-uclinuxoldabi} or
3591 @samp{m68k-@var{vendor}-uclinuxoldabi}.
3596 @heading @anchor{mips-x-x}mips-*-*
3597 If on a MIPS system you get an error message saying ``does not have gp
3598 sections for all it's [sic] sectons [sic]'', don't worry about it. This
3599 happens whenever you use GAS with the MIPS linker, but there is not
3600 really anything wrong, and it is okay to use the output file. You can
3601 stop such warnings by installing the GNU linker.
3603 It would be nice to extend GAS to produce the gp tables, but they are
3604 optional, and there should not be a warning about their absence.
3606 The libstdc++ atomic locking routines for MIPS targets requires MIPS II
3607 and later. A patch went in just after the GCC 3.3 release to
3608 make @samp{mips*-*-*} use the generic implementation instead. You can also
3609 configure for @samp{mipsel-elf} as a workaround. The
3610 @samp{mips*-*-linux*} target continues to use the MIPS II routines. More
3611 work on this is expected in future releases.
3613 MIPS systems check for division by zero (unless
3614 @option{-mno-check-zero-division} is passed to the compiler) by
3615 generating either a conditional trap or a break instruction. Using
3616 trap results in smaller code, but is only supported on MIPS II and
3617 later. Also, some versions of the Linux kernel have a bug that
3618 prevents trap from generating the proper signal (@code{SIGFPE}). To enable
3619 the use of break, use the @option{--with-divide=breaks}
3620 @command{configure} option when configuring GCC@. The default is to
3621 use traps on systems that support them.
3623 Cross-compilers for the MIPS as target using the MIPS assembler
3624 currently do not work, because the auxiliary programs
3625 @file{mips-tdump.c} and @file{mips-tfile.c} can't be compiled on
3626 anything but a MIPS. It does work to cross compile for a MIPS
3627 if you use the GNU assembler and linker.
3629 The assembler from GNU binutils 2.17 and earlier has a bug in the way
3630 it sorts relocations for REL targets (o32, o64, EABI). This can cause
3631 bad code to be generated for simple C++ programs. Also the linker
3632 from GNU binutils versions prior to 2.17 has a bug which causes the
3633 runtime linker stubs in very large programs, like @file{libgcj.so}, to
3634 be incorrectly generated. Binutils CVS snapshots and releases made
3635 after Nov. 9, 2006 are thought to be free from both of these problems.
3640 @heading @anchor{mips-sgi-irix5}mips-sgi-irix5
3642 In order to compile GCC on an SGI running IRIX 5, the @samp{compiler_dev.hdr}
3643 subsystem must be installed from the IDO CD-ROM supplied by SGI@.
3644 It is also available for download from
3645 @uref{ftp://ftp.sgi.com/sgi/IRIX5.3/iris-development-option-5.3.tardist}.
3647 If you use the MIPS C compiler to bootstrap, it may be necessary
3648 to increase its table size for switch statements with the
3649 @option{-Wf,-XNg1500} option. If you use the @option{-O2}
3650 optimization option, you also need to use @option{-Olimit 3000}.
3652 To enable debugging under IRIX 5, you must use GNU binutils 2.15 or
3653 later, and use the @option{--with-gnu-ld} @command{configure} option
3654 when configuring GCC@. You need to use GNU @command{ar} and @command{nm},
3655 also distributed with GNU binutils.
3657 Some users have reported that @command{/bin/sh} will hang during bootstrap.
3658 This problem can be avoided by running the commands:
3661 % CONFIG_SHELL=/bin/ksh
3662 % export CONFIG_SHELL
3665 before starting the build.
3670 @heading @anchor{mips-sgi-irix6}mips-sgi-irix6
3672 If you are using SGI's MIPSpro @command{cc} as your bootstrap compiler, you must
3673 ensure that the N32 ABI is in use. To test this, compile a simple C
3674 file with @command{cc} and then run @command{file} on the
3675 resulting object file. The output should look like:
3678 test.o: ELF N32 MSB @dots{}
3684 test.o: ELF 32-bit MSB @dots{}
3690 test.o: ELF 64-bit MSB @dots{}
3693 then your version of @command{cc} uses the O32 or N64 ABI by default. You
3694 should set the environment variable @env{CC} to @samp{cc -n32}
3695 before configuring GCC@.
3697 If you want the resulting @command{gcc} to run on old 32-bit systems
3698 with the MIPS R4400 CPU, you need to ensure that only code for the @samp{mips3}
3699 instruction set architecture (ISA) is generated. While GCC 3.x does
3700 this correctly, both GCC 2.95 and SGI's MIPSpro @command{cc} may change
3701 the ISA depending on the machine where GCC is built. Using one of them
3702 as the bootstrap compiler may result in @samp{mips4} code, which won't run at
3703 all on @samp{mips3}-only systems. For the test program above, you should see:
3706 test.o: ELF N32 MSB mips-3 @dots{}
3712 test.o: ELF N32 MSB mips-4 @dots{}
3715 instead, you should set the environment variable @env{CC} to @samp{cc
3716 -n32 -mips3} or @samp{gcc -mips3} respectively before configuring GCC@.
3718 MIPSpro C 7.4 may cause bootstrap failures, due to a bug when inlining
3719 @code{memcmp}. Either add @code{-U__INLINE_INTRINSICS} to the @env{CC}
3720 environment variable as a workaround or upgrade to MIPSpro C 7.4.1m.
3722 GCC on IRIX 6 is usually built to support the N32, O32 and N64 ABIs. If
3723 you build GCC on a system that doesn't have the N64 libraries installed
3724 or cannot run 64-bit binaries,
3725 you need to configure with @option{--disable-multilib} so GCC doesn't
3726 try to use them. This will disable building the O32 libraries, too.
3727 Look for @file{/usr/lib64/libc.so.1} to see if you
3728 have the 64-bit libraries installed.
3730 To enable debugging for the O32 ABI, you must use GNU @command{as} from
3731 GNU binutils 2.15 or later. You may also use GNU @command{ld}, but
3732 this is not required and currently causes some problems with Ada.
3734 The @option{--enable-libgcj}
3735 option is disabled by default: IRIX 6 uses a very low default limit
3736 (20480) for the command line length. Although @command{libtool} contains a
3737 workaround for this problem, at least the N64 @samp{libgcj} is known not
3738 to build despite this, running into an internal error of the native
3739 @command{ld}. A sure fix is to increase this limit (@samp{ncargs}) to
3740 its maximum of 262144 bytes. If you have root access, you can use the
3741 @command{systune} command to do this.
3743 @code{wchar_t} support in @samp{libstdc++} is not available for old
3744 IRIX 6.5.x releases, @math{x < 19}. The problem cannot be autodetected
3745 and in order to build GCC for such targets you need to configure with
3746 @option{--disable-wchar_t}.
3748 See @uref{http://freeware.sgi.com/} for more
3749 information about using GCC on IRIX platforms.
3754 @heading @anchor{powerpc-x-x}powerpc-*-*
3756 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
3757 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
3762 @heading @anchor{powerpc-x-darwin}powerpc-*-darwin*
3763 PowerPC running Darwin (Mac OS X kernel).
3765 Pre-installed versions of Mac OS X may not include any developer tools,
3766 meaning that you will not be able to build GCC from source. Tool
3767 binaries are available at
3768 @uref{http://developer.apple.com/darwin/projects/compiler/} (free
3769 registration required).
3771 This version of GCC requires at least cctools-590.36. The
3772 cctools-590.36 package referenced from
3773 @uref{http://gcc.gnu.org/ml/gcc/2006-03/msg00507.html} will not work
3774 on systems older than 10.3.9 (aka darwin7.9.0).
3779 @heading @anchor{powerpc-x-elf}powerpc-*-elf, powerpc-*-sysv4
3780 PowerPC system in big endian mode, running System V.4.
3785 @heading @anchor{powerpc-x-linux-gnu}powerpc*-*-linux-gnu*
3788 @uref{ftp://ftp.kernel.org/pub/linux/devel/binutils,,binutils 2.15}
3789 or newer for a working GCC@.
3794 @heading @anchor{powerpc-x-netbsd}powerpc-*-netbsd*
3795 PowerPC system in big endian mode running NetBSD@. To build the
3796 documentation you will need Texinfo version 4.4 (NetBSD 1.5.1 included
3797 Texinfo version 3.12).
3802 @heading @anchor{powerpc-x-eabisim}powerpc-*-eabisim
3803 Embedded PowerPC system in big endian mode for use in running under the
3809 @heading @anchor{powerpc-x-eabi}powerpc-*-eabi
3810 Embedded PowerPC system in big endian mode.
3815 @heading @anchor{powerpcle-x-elf}powerpcle-*-elf, powerpcle-*-sysv4
3816 PowerPC system in little endian mode, running System V.4.
3821 @heading @anchor{powerpcle-x-eabisim}powerpcle-*-eabisim
3822 Embedded PowerPC system in little endian mode for use in running under
3828 @heading @anchor{powerpcle-x-eabi}powerpcle-*-eabi
3829 Embedded PowerPC system in little endian mode.
3834 @heading @anchor{s390-x-linux}s390-*-linux*
3835 S/390 system running GNU/Linux for S/390@.
3840 @heading @anchor{s390x-x-linux}s390x-*-linux*
3841 zSeries system (64-bit) running GNU/Linux for zSeries@.
3846 @heading @anchor{s390x-ibm-tpf}s390x-ibm-tpf*
3847 zSeries system (64-bit) running TPF@. This platform is
3848 supported as cross-compilation target only.
3853 @c Please use Solaris 2 to refer to all release of Solaris, starting
3854 @c with 2.0 until 2.6, 7, 8, etc. Solaris 1 was a marketing name for
3855 @c SunOS 4 releases which we don't use to avoid confusion. Solaris
3856 @c alone is too unspecific and must be avoided.
3857 @heading @anchor{x-x-solaris2}*-*-solaris2*
3859 Sun does not ship a C compiler with Solaris 2. To bootstrap and install
3860 GCC you first have to install a pre-built compiler, see the
3861 @uref{binaries.html,,binaries page} for details.
3863 The Solaris 2 @command{/bin/sh} will often fail to configure
3864 @file{libstdc++-v3}, @file{boehm-gc} or @file{libjava}. We therefore
3865 recommend using the following initial sequence of commands
3868 % CONFIG_SHELL=/bin/ksh
3869 % export CONFIG_SHELL
3872 and proceed as described in @uref{configure.html,,the configure instructions}.
3873 In addition we strongly recommend specifying an absolute path to invoke
3874 @var{srcdir}/configure.
3876 Solaris 2 comes with a number of optional OS packages. Some of these
3877 are needed to use GCC fully, namely @code{SUNWarc},
3878 @code{SUNWbtool}, @code{SUNWesu}, @code{SUNWhea}, @code{SUNWlibm},
3879 @code{SUNWsprot}, and @code{SUNWtoo}. If you did not install all
3880 optional packages when installing Solaris 2, you will need to verify that
3881 the packages that GCC needs are installed.
3883 To check whether an optional package is installed, use
3884 the @command{pkginfo} command. To add an optional package, use the
3885 @command{pkgadd} command. For further details, see the Solaris 2
3888 Trying to use the linker and other tools in
3889 @file{/usr/ucb} to install GCC has been observed to cause trouble.
3890 For example, the linker may hang indefinitely. The fix is to remove
3891 @file{/usr/ucb} from your @env{PATH}.
3893 The build process works more smoothly with the legacy Sun tools so, if you
3894 have @file{/usr/xpg4/bin} in your @env{PATH}, we recommend that you place
3895 @file{/usr/bin} before @file{/usr/xpg4/bin} for the duration of the build.
3897 All releases of GNU binutils prior to 2.11.2 have known bugs on this
3898 platform. We recommend the use of GNU binutils 2.11.2 or later, or the
3899 vendor tools (Sun @command{as}, Sun @command{ld}). Note that your mileage
3900 may vary if you use a combination of the GNU tools and the Sun tools: while
3901 the combination GNU @command{as} + Sun @command{ld} should reasonably work,
3902 the reverse combination Sun @command{as} + GNU @command{ld} is known to
3903 cause memory corruption at runtime in some cases for C++ programs.
3905 The stock GNU binutils 2.15 release is broken on this platform because of a
3906 single bug. It has been fixed on the 2.15 branch in the CVS repository.
3907 You can obtain a working version by checking out the binutils-2_15-branch
3908 from the CVS repository or applying the patch
3909 @uref{http://sourceware.org/ml/binutils-cvs/2004-09/msg00036.html} to the
3912 We recommend using GNU binutils 2.16 or later in conjunction with GCC 4.x,
3913 or the vendor tools (Sun @command{as}, Sun @command{ld}). However, for
3914 Solaris 10 and above, an additional patch is required in order for the GNU
3915 linker to be able to cope with a new flavor of shared libraries. You
3916 can obtain a working version by checking out the binutils-2_16-branch from
3917 the CVS repository or applying the patch
3918 @uref{http://sourceware.org/ml/binutils-cvs/2005-07/msg00122.html} to the
3921 Sun bug 4296832 turns up when compiling X11 headers with GCC 2.95 or
3922 newer: @command{g++} will complain that types are missing. These headers assume
3923 that omitting the type means @code{int}; this assumption worked for C89 but
3924 is wrong for C++, and is now wrong for C99 also.
3926 @command{g++} accepts such (invalid) constructs with the option
3927 @option{-fpermissive}; it
3928 will assume that any missing type is @code{int} (as defined by C89).
3930 There are patches for Solaris 2.6 (105633-56 or newer for SPARC,
3931 106248-42 or newer for Intel), Solaris 7 (108376-21 or newer for SPARC,
3932 108377-20 for Intel), and Solaris 8 (108652-24 or newer for SPARC,
3933 108653-22 for Intel) that fix this bug.
3935 Sun bug 4927647 sometimes causes random spurious testsuite failures
3936 related to missing diagnostic output. This bug doesn't affect GCC
3937 itself, rather it is a kernel bug triggered by the @command{expect}
3938 program which is used only by the GCC testsuite driver. When the bug
3939 causes the @command{expect} program to miss anticipated output, extra
3940 testsuite failures appear.
3942 There are patches for Solaris 8 (117350-12 or newer for SPARC,
3943 117351-12 or newer for Intel) and Solaris 9 (117171-11 or newer for
3944 SPARC, 117172-11 or newer for Intel) that address this problem.
3949 @heading @anchor{sparc-sun-solaris2}sparc-sun-solaris2*
3951 When GCC is configured to use binutils 2.11.2 or later the binaries
3952 produced are smaller than the ones produced using Sun's native tools;
3953 this difference is quite significant for binaries containing debugging
3956 Sun @command{as} 4.x is broken in that it cannot cope with long symbol names.
3957 A typical error message might look similar to the following:
3960 /usr/ccs/bin/as: "/var/tmp/ccMsw135.s", line 11041: error:
3961 can't compute value of an expression involving an external symbol.
3964 This is Sun bug 4237974. This is fixed with patch 108908-02 for Solaris
3965 2.6 and has been fixed in later (5.x) versions of the assembler,
3966 starting with Solaris 7.
3968 Starting with Solaris 7, the operating system is capable of executing
3969 64-bit SPARC V9 binaries. GCC 3.1 and later properly supports
3970 this; the @option{-m64} option enables 64-bit code generation.
3971 However, if all you want is code tuned for the UltraSPARC CPU, you
3972 should try the @option{-mtune=ultrasparc} option instead, which produces
3973 code that, unlike full 64-bit code, can still run on non-UltraSPARC
3976 When configuring on a Solaris 7 or later system that is running a kernel
3977 that supports only 32-bit binaries, one must configure with
3978 @option{--disable-multilib}, since we will not be able to build the
3979 64-bit target libraries.
3981 GCC 3.3 and GCC 3.4 trigger code generation bugs in earlier versions of
3982 the GNU compiler (especially GCC 3.0.x versions), which lead to the
3983 miscompilation of the stage1 compiler and the subsequent failure of the
3984 bootstrap process. A workaround is to use GCC 3.2.3 as an intermediary
3985 stage, i.e.@: to bootstrap that compiler with the base compiler and then
3986 use it to bootstrap the final compiler.
3988 GCC 3.4 triggers a code generation bug in versions 5.4 (Sun ONE Studio 7)
3989 and 5.5 (Sun ONE Studio 8) of the Sun compiler, which causes a bootstrap
3990 failure in form of a miscompilation of the stage1 compiler by the Sun
3991 compiler. This is Sun bug 4974440. This is fixed with patch 112760-07.
3993 GCC 3.4 changed the default debugging format from STABS to DWARF-2 for
3994 32-bit code on Solaris 7 and later. If you use the Sun assembler, this
3995 change apparently runs afoul of Sun bug 4910101 (which is referenced as
3996 a x86-only problem by Sun, probably because they do not use DWARF-2).
3997 A symptom of the problem is that you cannot compile C++ programs like
3998 @command{groff} 1.19.1 without getting messages similar to the following:
4001 ld: warning: relocation error: R_SPARC_UA32: @dots{}
4002 external symbolic relocation against non-allocatable section
4003 .debug_info cannot be processed at runtime: relocation ignored.
4006 To work around this problem, compile with @option{-gstabs+} instead of
4009 When configuring the GNU Multiple Precision Library (GMP) or the MPFR
4010 library on a Solaris 7 or later system, the canonical target triplet
4011 must be specified as the @command{build} parameter on the configure
4012 line. This triplet can be obtained by invoking ./config.guess in
4013 the toplevel source directory of GCC (and not that of GMP or MPFR).
4014 For example on a Solaris 7 system:
4017 % ./configure --build=sparc-sun-solaris2.7 --prefix=xxx
4023 @heading @anchor{sparc-sun-solaris27}sparc-sun-solaris2.7
4025 Sun patch 107058-01 (1999-01-13) for Solaris 7/SPARC triggers a bug in
4026 the dynamic linker. This problem (Sun bug 4210064) affects GCC 2.8
4027 and later, including all EGCS releases. Sun formerly recommended
4028 107058-01 for all Solaris 7 users, but around 1999-09-01 it started to
4029 recommend it only for people who use Sun's compilers.
4031 Here are some workarounds to this problem:
4034 Do not install Sun patch 107058-01 until after Sun releases a
4035 complete patch for bug 4210064. This is the simplest course to take,
4036 unless you must also use Sun's C compiler. Unfortunately 107058-01
4037 is preinstalled on some new Solaris 7-based hosts, so you may have to
4041 Copy the original, unpatched Solaris 7
4042 @command{/usr/ccs/bin/as} into
4043 @command{/usr/local/libexec/gcc/sparc-sun-solaris2.7/3.4/as},
4044 adjusting the latter name to fit your local conventions and software
4048 Install Sun patch 106950-03 (1999-05-25) or later. Nobody with
4049 both 107058-01 and 106950-03 installed has reported the bug with GCC
4050 and Sun's dynamic linker. This last course of action is riskiest,
4051 for two reasons. First, you must install 106950 on all hosts that
4052 run code generated by GCC; it doesn't suffice to install it only on
4053 the hosts that run GCC itself. Second, Sun says that 106950-03 is
4054 only a partial fix for bug 4210064, but Sun doesn't know whether the
4055 partial fix is adequate for GCC@. Revision -08 or later should fix
4056 the bug. The current (as of 2004-05-23) revision is -24, and is included in
4057 the Solaris 7 Recommended Patch Cluster.
4060 GCC 3.3 triggers a bug in version 5.0 Alpha 03/27/98 of the Sun assembler,
4061 which causes a bootstrap failure when linking the 64-bit shared version of
4062 libgcc. A typical error message is:
4065 ld: fatal: relocation error: R_SPARC_32: file libgcc/sparcv9/_muldi3.o:
4066 symbol <unknown>: offset 0xffffffff7ec133e7 is non-aligned.
4069 This bug has been fixed in the final 5.0 version of the assembler.
4071 A similar problem was reported for version Sun WorkShop 6 99/08/18 of the
4072 Sun assembler, which causes a bootstrap failure with GCC 4.0.0:
4075 ld: fatal: relocation error: R_SPARC_DISP32:
4076 file .libs/libstdc++.lax/libsupc++convenience.a/vterminate.o:
4077 symbol <unknown>: offset 0xfccd33ad is non-aligned
4080 This bug has been fixed in more recent revisions of the assembler.
4085 @heading @anchor{sparc-x-linux}sparc-*-linux*
4087 GCC versions 3.0 and higher require binutils 2.11.2 and glibc 2.2.4
4088 or newer on this platform. All earlier binutils and glibc
4089 releases mishandled unaligned relocations on @code{sparc-*-*} targets.
4095 @heading @anchor{sparc64-x-solaris2}sparc64-*-solaris2*
4097 When configuring the GNU Multiple Precision Library (GMP) or the
4098 MPFR library, the canonical target triplet must be specified as
4099 the @command{build} parameter on the configure line. For example
4100 on a Solaris 7 system:
4103 % ./configure --build=sparc64-sun-solaris2.7 --prefix=xxx
4106 The following compiler flags must be specified in the configure
4107 step in order to bootstrap this target with the Sun compiler:
4110 % CC="cc -xarch=v9 -xildoff" @var{srcdir}/configure [@var{options}] [@var{target}]
4113 @option{-xarch=v9} specifies the SPARC-V9 architecture to the Sun toolchain
4114 and @option{-xildoff} turns off the incremental linker.
4119 @heading @anchor{sparcv9-x-solaris2}sparcv9-*-solaris2*
4121 This is a synonym for sparc64-*-solaris2*.
4126 @heading @anchor{x-x-sysv}*-*-sysv*
4127 On System V release 3, you may get this error message
4131 ld fatal: failed to write symbol name @var{something}
4132 in strings table for file @var{whatever}
4135 This probably indicates that the disk is full or your ulimit won't allow
4136 the file to be as large as it needs to be.
4138 This problem can also result because the kernel parameter @code{MAXUMEM}
4139 is too small. If so, you must regenerate the kernel and make the value
4140 much larger. The default value is reported to be 1024; a value of 32768
4141 is said to work. Smaller values may also work.
4143 On System V, if you get an error like this,
4146 /usr/local/lib/bison.simple: In function `yyparse':
4147 /usr/local/lib/bison.simple:625: virtual memory exhausted
4151 that too indicates a problem with disk space, ulimit, or @code{MAXUMEM}.
4153 On a System V release 4 system, make sure @file{/usr/bin} precedes
4154 @file{/usr/ucb} in @code{PATH}. The @command{cc} command in
4155 @file{/usr/ucb} uses libraries which have bugs.
4160 @heading @anchor{vax-dec-ultrix}vax-dec-ultrix
4161 Don't try compiling with VAX C (@command{vcc}). It produces incorrect code
4162 in some cases (for example, when @code{alloca} is used).
4167 @heading @anchor{x-x-vxworks}*-*-vxworks*
4168 Support for VxWorks is in flux. At present GCC supports @emph{only} the
4169 very recent VxWorks 5.5 (aka Tornado 2.2) release, and only on PowerPC@.
4170 We welcome patches for other architectures supported by VxWorks 5.5.
4171 Support for VxWorks AE would also be welcome; we believe this is merely
4172 a matter of writing an appropriate ``configlette'' (see below). We are
4173 not interested in supporting older, a.out or COFF-based, versions of
4176 VxWorks comes with an older version of GCC installed in
4177 @file{@var{$WIND_BASE}/host}; we recommend you do not overwrite it.
4178 Choose an installation @var{prefix} entirely outside @var{$WIND_BASE}.
4179 Before running @command{configure}, create the directories @file{@var{prefix}}
4180 and @file{@var{prefix}/bin}. Link or copy the appropriate assembler,
4181 linker, etc.@: into @file{@var{prefix}/bin}, and set your @var{PATH} to
4182 include that directory while running both @command{configure} and
4185 You must give @command{configure} the
4186 @option{--with-headers=@var{$WIND_BASE}/target/h} switch so that it can
4187 find the VxWorks system headers. Since VxWorks is a cross compilation
4188 target only, you must also specify @option{--target=@var{target}}.
4189 @command{configure} will attempt to create the directory
4190 @file{@var{prefix}/@var{target}/sys-include} and copy files into it;
4191 make sure the user running @command{configure} has sufficient privilege
4194 GCC's exception handling runtime requires a special ``configlette''
4195 module, @file{contrib/gthr_supp_vxw_5x.c}. Follow the instructions in
4196 that file to add the module to your kernel build. (Future versions of
4197 VxWorks will incorporate this module.)
4202 @heading @anchor{x86-64-x-x}x86_64-*-*, amd64-*-*
4204 GCC supports the x86-64 architecture implemented by the AMD64 processor
4205 (amd64-*-* is an alias for x86_64-*-*) on GNU/Linux, FreeBSD and NetBSD@.
4206 On GNU/Linux the default is a bi-arch compiler which is able to generate
4207 both 64-bit x86-64 and 32-bit x86 code (via the @option{-m32} switch).
4212 @heading @anchor{xtensa-x-elf}xtensa-*-elf
4214 This target is intended for embedded Xtensa systems using the
4215 @samp{newlib} C library. It uses ELF but does not support shared
4216 objects. Designed-defined instructions specified via the
4217 Tensilica Instruction Extension (TIE) language are only supported
4218 through inline assembly.
4220 The Xtensa configuration information must be specified prior to
4221 building GCC@. The @file{include/xtensa-config.h} header
4222 file contains the configuration information. If you created your
4223 own Xtensa configuration with the Xtensa Processor Generator, the
4224 downloaded files include a customized copy of this header file,
4225 which you can use to replace the default header file.
4230 @heading @anchor{xtensa-x-linux}xtensa-*-linux*
4232 This target is for Xtensa systems running GNU/Linux. It supports ELF
4233 shared objects and the GNU C library (glibc). It also generates
4234 position-independent code (PIC) regardless of whether the
4235 @option{-fpic} or @option{-fPIC} options are used. In other
4236 respects, this target is the same as the
4237 @uref{#xtensa-*-elf,,@samp{xtensa-*-elf}} target.
4242 @heading @anchor{windows}Microsoft Windows (32-bit)
4244 Ports of GCC are included with the
4245 @uref{http://www.cygwin.com/,,Cygwin environment}.
4247 GCC will build under Cygwin without modification; it does not build
4248 with Microsoft's C++ compiler and there are no plans to make it do so.
4253 @heading @anchor{os2}OS/2
4255 GCC does not currently support OS/2. However, Andrew Zabolotny has been
4256 working on a generic OS/2 port with pgcc. The current code can be found
4257 at @uref{http://www.goof.com/pcg/os2/,,http://www.goof.com/pcg/os2/}.
4262 @heading @anchor{older}Older systems
4264 GCC contains support files for many older (1980s and early
4265 1990s) Unix variants. For the most part, support for these systems
4266 has not been deliberately removed, but it has not been maintained for
4267 several years and may suffer from bitrot.
4269 Starting with GCC 3.1, each release has a list of ``obsoleted'' systems.
4270 Support for these systems is still present in that release, but
4271 @command{configure} will fail unless the @option{--enable-obsolete}
4272 option is given. Unless a maintainer steps forward, support for these
4273 systems will be removed from the next release of GCC@.
4275 Support for old systems as hosts for GCC can cause problems if the
4276 workarounds for compiler, library and operating system bugs affect the
4277 cleanliness or maintainability of the rest of GCC@. In some cases, to
4278 bring GCC up on such a system, if still possible with current GCC, may
4279 require first installing an old version of GCC which did work on that
4280 system, and using it to compile a more recent GCC, to avoid bugs in the
4281 vendor compiler. Old releases of GCC 1 and GCC 2 are available in the
4282 @file{old-releases} directory on the @uref{../mirrors.html,,GCC mirror
4283 sites}. Header bugs may generally be avoided using
4284 @command{fixincludes}, but bugs or deficiencies in libraries and the
4285 operating system may still cause problems.
4287 Support for older systems as targets for cross-compilation is less
4288 problematic than support for them as hosts for GCC; if an enthusiast
4289 wishes to make such a target work again (including resurrecting any of
4290 the targets that never worked with GCC 2, starting from the last
4291 version before they were removed), patches
4292 @uref{../contribute.html,,following the usual requirements} would be
4293 likely to be accepted, since they should not affect the support for more
4296 For some systems, old versions of GNU binutils may also be useful,
4297 and are available from @file{pub/binutils/old-releases} on
4298 @uref{http://sourceware.org/mirrors.html,,sourceware.org mirror sites}.
4300 Some of the information on specific systems above relates to
4301 such older systems, but much of the information
4302 about GCC on such systems (which may no longer be applicable to
4303 current GCC) is to be found in the GCC texinfo manual.
4308 @heading @anchor{elf}all ELF targets (SVR4, Solaris 2, etc.)
4310 C++ support is significantly better on ELF targets if you use the
4311 @uref{./configure.html#with-gnu-ld,,GNU linker}; duplicate copies of
4312 inlines, vtables and template instantiations will be discarded
4321 @uref{./index.html,,Return to the GCC Installation page}
4325 @c ***Old documentation******************************************************
4327 @include install-old.texi
4333 @uref{./index.html,,Return to the GCC Installation page}
4337 @c ***GFDL********************************************************************
4345 @uref{./index.html,,Return to the GCC Installation page}
4349 @c ***************************************************************************
4350 @c Part 6 The End of the Document
4352 @comment node-name, next, previous, up
4353 @node Concept Index, , GNU Free Documentation License, Top
4357 @unnumbered Concept Index