1 \input texinfo.tex @c -*-texinfo-*-
4 @setfilename gccinstall.info
5 @settitle Installing GCC
10 @c Specify title for specific html page
12 @settitle Installing GCC
15 @settitle Host/Target specific installation notes for GCC
17 @ifset prerequisiteshtml
18 @settitle Prerequisites for GCC
21 @settitle Downloading GCC
24 @settitle Installing GCC: Configuration
27 @settitle Installing GCC: Building
30 @settitle Installing GCC: Testing
32 @ifset finalinstallhtml
33 @settitle Installing GCC: Final installation
36 @settitle Installing GCC: Binaries
39 @settitle Installing GCC: Old documentation
42 @settitle Installing GCC: GNU Free Documentation License
45 @c Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997, 1998,
46 @c 1999, 2000, 2001, 2002, 2003, 2004 Free Software Foundation, Inc.
47 @c *** Converted to texinfo by Dean Wakerley, dean@wakerley.com
49 @c Include everything if we're not making html
53 @set prerequisiteshtml
64 @c Part 2 Summary Description and Copyright
66 Copyright @copyright{} 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997, 1998,
67 1999, 2000, 2001, 2002, 2003, 2004 Free Software Foundation, Inc.
69 Permission is granted to copy, distribute and/or modify this document
70 under the terms of the GNU Free Documentation License, Version 1.2 or
71 any later version published by the Free Software Foundation; with no
72 Invariant Sections, the Front-Cover texts being (a) (see below), and
73 with the Back-Cover Texts being (b) (see below). A copy of the
74 license is included in the section entitled ``@uref{./gfdl.html,,GNU
75 Free Documentation License}''.
77 (a) The FSF's Front-Cover Text is:
81 (b) The FSF's Back-Cover Text is:
83 You have freedom to copy and modify this GNU Manual, like GNU
84 software. Copies published by the Free Software Foundation raise
85 funds for GNU development.
91 @c Part 3 Titlepage and Copyright
94 @comment The title is printed in a large font.
95 @center @titlefont{Installing GCC}
97 @c The following two commands start the copyright page.
99 @vskip 0pt plus 1filll
103 @c Part 4 Top node and Master Menu
106 @comment node-name, next, Previous, up
109 * Installing GCC:: This document describes the generic installation
110 procedure for GCC as well as detailing some target
111 specific installation instructions.
113 * Specific:: Host/target specific installation notes for GCC.
114 * Binaries:: Where to get pre-compiled binaries.
116 * Old:: Old installation documentation.
118 * GNU Free Documentation License:: How you can copy and share this manual.
119 * Concept Index:: This index has two entries.
123 @c Part 5 The Body of the Document
124 @c ***Installing GCC**********************************************************
126 @comment node-name, next, previous, up
127 @node Installing GCC, Binaries, , Top
131 @chapter Installing GCC
134 The latest version of this document is always available at
135 @uref{http://gcc.gnu.org/install/,,http://gcc.gnu.org/install/}.
137 This document describes the generic installation procedure for GCC as well
138 as detailing some target specific installation instructions.
140 GCC includes several components that previously were separate distributions
141 with their own installation instructions. This document supersedes all
142 package specific installation instructions.
144 @emph{Before} starting the build/install procedure please check the
146 @ref{Specific, host/target specific installation notes}.
149 @uref{specific.html,,host/target specific installation notes}.
151 We recommend you browse the entire generic installation instructions before
154 Lists of successful builds for released versions of GCC are
155 available at @uref{http://gcc.gnu.org/buildstat.html}.
156 These lists are updated as new information becomes available.
158 The installation procedure itself is broken into five steps.
163 * Downloading the source::
166 * Testing:: (optional)
173 @uref{prerequisites.html,,Prerequisites}
175 @uref{download.html,,Downloading the source}
177 @uref{configure.html,,Configuration}
179 @uref{build.html,,Building}
181 @uref{test.html,,Testing} (optional)
183 @uref{finalinstall.html,,Final install}
187 Please note that GCC does not support @samp{make uninstall} and probably
188 won't do so in the near future as this would open a can of worms. Instead,
189 we suggest that you install GCC into a directory of its own and simply
190 remove that directory when you do not need that specific version of GCC
191 any longer, and, if shared libraries are installed there as well, no
192 more binaries exist that use them.
195 There are also some @uref{old.html,,old installation instructions},
196 which are mostly obsolete but still contain some information which has
197 not yet been merged into the main part of this manual.
205 @uref{./index.html,,Return to the GCC Installation page}
211 @c ***Prerequisites**************************************************
213 @comment node-name, next, previous, up
214 @node Prerequisites, Downloading the source, , Installing GCC
216 @ifset prerequisiteshtml
218 @chapter Prerequisites
220 @cindex Prerequisites
222 GCC requires that various tools and packages be available for use in the
223 build procedure. Modifying GCC sources requires additional tools
226 @heading Tools/packages necessary for building GCC
228 @item ISO C90 compiler
229 Necessary to bootstrap the GCC package, although versions of GCC prior
230 to 3.4 also allow bootstrapping with a traditional (K&R) C compiler.
232 To make all languages in a cross-compiler or other configuration where
233 3-stage bootstrap is not performed, you need to start with an existing
234 GCC binary (version 2.95 or later) because source code for language
235 frontends other than C might use GCC extensions.
239 In order to build the Ada compiler (GNAT) you must already have GNAT
240 installed because portions of the Ada frontend are written in Ada (with
241 GNAT extensions.) Refer to the Ada installation instructions for more
242 specific information.
244 @item A ``working'' POSIX compatible shell, or GNU bash
246 Necessary when running @command{configure} because some
247 @command{/bin/sh} shells have bugs and may crash when configuring the
248 target libraries. In other cases, @command{/bin/sh} or even some
249 @command{ksh} have disastrous corner-case performance problems. This
250 can cause target @command{configure} runs to literally take days to
251 complete in some cases.
253 So on some platforms @command{/bin/ksh} is sufficient, on others it
254 isn't. See the host/target specific instructions for your platform, or
255 use @command{bash} to be sure. Then set @env{CONFIG_SHELL} in your
256 environment to your ``good'' shell prior to running
257 @command{configure}/@command{make}.
259 @command{zsh} is not a fully compliant POSIX shell and will not
260 work when configuring GCC.
264 Necessary in some circumstances, optional in others. See the
265 host/target specific instructions for your platform for the exact
268 @item gzip version 1.2.4 (or later) or
269 @itemx bzip2 version 1.0.2 (or later)
271 Necessary to uncompress GCC @command{tar} files when source code is
272 obtained via FTP mirror sites.
274 @item GNU make version 3.79.1 (or later)
276 You must have GNU make installed to build GCC.
278 @item GNU tar version 1.12 (or later)
280 Necessary (only on some platforms) to untar the source code. Many
281 systems' @command{tar} programs will also work, only try GNU
282 @command{tar} if you have problems.
287 @heading Tools/packages necessary for modifying GCC
290 @item autoconf versions 2.13 and 2.57
291 @itemx GNU m4 version 1.4 (or later)
293 Necessary when modifying @file{configure.in}, @file{aclocal.m4}, etc.@:
294 to regenerate @file{configure} and @file{config.in} files. Most
295 directories require autoconf 2.13 (exactly), but @file{libiberty},
296 @file{fastjar}, @file{libstdc++-v3}, and @file{gcc} require
297 autoconf 2.57 (exactly).
299 @item automake versions 1.4-p? and 1.7.?
301 Necessary when modifying a @file{Makefile.am} file to regenerate its
302 associated @file{Makefile.in}. Most directories require a 1.4 series
303 automake; @file{libstdc++-v3} and @file{fastjar} requires a 1.7 series
306 @item gettext version 0.12 (or later)
308 Needed to regenerate @file{gcc.pot}.
310 @item gperf version 2.7.2 (or later)
312 Necessary when modifying @command{gperf} input files, e.g.@:
313 @file{gcc/cp/cfns.gperf} to regenerate its associated header file, e.g.@:
314 @file{gcc/cp/cfns.h}.
316 @item expect version ???
317 @itemx tcl version ???
318 @itemx dejagnu version ???
320 Necessary to run the GCC testsuite.
322 @item autogen version 5.5.4 (or later) and
323 @itemx guile version 1.4.1 (or later)
325 Necessary to regenerate @file{fixinc/fixincl.x} from
326 @file{fixinc/inclhack.def} and @file{fixinc/*.tpl}.
328 Necessary to run the @file{fixinc} @command{make check}.
330 Necessary to regenerate the top level @file{Makefile.am} files from
331 @file{Makefile.tpl} and @file{Makefile.def}.
333 @item GNU Bison version 1.28 (or later)
334 Berkeley @command{yacc} (@command{byacc}) is also reported to work other
337 Necessary when modifying @file{*.y} files.
339 Necessary to build GCC during development because the generated output
340 files are not included in the CVS repository. They are included in
343 @item Flex version 2.5.4 (or later)
345 Necessary when modifying @file{*.l} files.
347 Necessary to build GCC during development because the generated output
348 files are not included in the CVS repository. They are included in
351 @item Texinfo version 4.2 (or later)
353 Necessary for running @command{makeinfo} when modifying @file{*.texi}
354 files to test your changes.
356 Necessary to build GCC documentation during development because the
357 generated output files are not included in the CVS repository. They are
358 included in releases.
360 @item @TeX{} (any working version)
362 Necessary for running @command{texi2dvi}, used when running
363 @command{make dvi} to create DVI files.
365 @item cvs version 1.10 (or later)
366 @itemx ssh (any version)
368 Necessary to access the CVS repository. Public releases and weekly
369 snapshots of the development sources are also available via FTP.
371 @item perl version 5.6.1 (or later)
373 Necessary when regenerating @file{Makefile} dependencies in libiberty.
374 Necessary when regenerating @file{libiberty/functions.texi}.
375 Necessary when generating manpages from Texinfo manuals.
376 Used by various scripts to generate some files included in CVS (mainly
377 Unicode-related and rarely changing) from source tables.
379 @item GNU diffutils version 2.7 (or later)
381 Necessary when creating changes to GCC source code to submit for review.
383 @item patch version 2.5.4 (or later)
385 Necessary when applying patches, created with @command{diff}, to one's
395 @uref{./index.html,,Return to the GCC Installation page}
399 @c ***Downloading the source**************************************************
401 @comment node-name, next, previous, up
402 @node Downloading the source, Configuration, Prerequisites, Installing GCC
406 @chapter Downloading GCC
408 @cindex Downloading GCC
409 @cindex Downloading the Source
411 GCC is distributed via @uref{http://gcc.gnu.org/cvs.html,,CVS} and FTP
412 tarballs compressed with @command{gzip} or
413 @command{bzip2}. It is possible to download a full distribution or specific
416 Please refer to our @uref{http://gcc.gnu.org/releases.html,,releases web page}
417 for information on how to obtain GCC@.
419 The full distribution includes the C, C++, Objective-C, Fortran, Java,
420 and Ada (in case of GCC 3.1 and later) compilers. The full distribution
421 also includes runtime libraries for C++, Objective-C, Fortran, and Java.
422 In GCC 3.0 and later versions, GNU compiler testsuites are also included
423 in the full distribution.
425 If you choose to download specific components, you must download the core
426 GCC distribution plus any language specific distributions you wish to
427 use. The core distribution includes the C language front end as well as the
428 shared components. Each language has a tarball which includes the language
429 front end as well as the language runtime (when appropriate).
431 Unpack the core distribution as well as any language specific
432 distributions in the same directory.
434 If you also intend to build binutils (either to upgrade an existing
435 installation or for use in place of the corresponding tools of your
436 OS), unpack the binutils distribution either in the same directory or
437 a separate one. In the latter case, add symbolic links to any
438 components of the binutils you intend to build alongside the compiler
439 (@file{bfd}, @file{binutils}, @file{gas}, @file{gprof}, @file{ld},
440 @file{opcodes}, @dots{}) to the directory containing the GCC sources.
447 @uref{./index.html,,Return to the GCC Installation page}
451 @c ***Configuration***********************************************************
453 @comment node-name, next, previous, up
454 @node Configuration, Building, Downloading the source, Installing GCC
458 @chapter Installing GCC: Configuration
460 @cindex Configuration
461 @cindex Installing GCC: Configuration
463 Like most GNU software, GCC must be configured before it can be built.
464 This document describes the recommended configuration procedure
465 for both native and cross targets.
467 We use @var{srcdir} to refer to the toplevel source directory for
468 GCC; we use @var{objdir} to refer to the toplevel build/object directory.
470 If you obtained the sources via CVS, @var{srcdir} must refer to the top
471 @file{gcc} directory, the one where the @file{MAINTAINERS} can be found,
472 and not its @file{gcc} subdirectory, otherwise the build will fail.
474 If either @var{srcdir} or @var{objdir} is located on an automounted NFS
475 file system, the shell's built-in @command{pwd} command will return
476 temporary pathnames. Using these can lead to various sorts of build
477 problems. To avoid this issue, set the @env{PWDCMD} environment
478 variable to an automounter-aware @command{pwd} command, e.g.,
479 @command{pawd} or @samp{amq -w}, during the configuration and build
482 First, we @strong{highly} recommend that GCC be built into a
483 separate directory than the sources which does @strong{not} reside
484 within the source tree. This is how we generally build GCC; building
485 where @var{srcdir} == @var{objdir} should still work, but doesn't
486 get extensive testing; building where @var{objdir} is a subdirectory
487 of @var{srcdir} is unsupported.
489 If you have previously built GCC in the same directory for a
490 different target machine, do @samp{make distclean} to delete all files
491 that might be invalid. One of the files this deletes is @file{Makefile};
492 if @samp{make distclean} complains that @file{Makefile} does not exist
493 or issues a message like ``don't know how to make distclean'' it probably
494 means that the directory is already suitably clean. However, with the
495 recommended method of building in a separate @var{objdir}, you should
496 simply use a different @var{objdir} for each target.
498 Second, when configuring a native system, either @command{cc} or
499 @command{gcc} must be in your path or you must set @env{CC} in
500 your environment before running configure. Otherwise the configuration
503 Note that the bootstrap compiler and the resulting GCC must be link
504 compatible, else the bootstrap will fail with linker errors about
505 incompatible object file formats. Several multilibed targets are
506 affected by this requirement, see
508 @ref{Specific, host/target specific installation notes}.
511 @uref{specific.html,,host/target specific installation notes}.
519 % @var{srcdir}/configure [@var{options}] [@var{target}]
523 @heading Target specification
526 GCC has code to correctly determine the correct value for @var{target}
527 for nearly all native systems. Therefore, we highly recommend you not
528 provide a configure target when configuring a native compiler.
531 @var{target} must be specified as @option{--target=@var{target}}
532 when configuring a cross compiler; examples of valid targets would be
533 i960-rtems, m68k-coff, sh-elf, etc.
536 Specifying just @var{target} instead of @option{--target=@var{target}}
537 implies that the host defaults to @var{target}.
541 @heading Options specification
543 Use @var{options} to override several configure time options for
544 GCC@. A list of supported @var{options} follows; @samp{configure
545 --help} may list other options, but those not listed below may not
546 work and should not normally be used.
549 @item --prefix=@var{dirname}
550 Specify the toplevel installation
551 directory. This is the recommended way to install the tools into a directory
552 other than the default. The toplevel installation directory defaults to
555 We @strong{highly} recommend against @var{dirname} being the same or a
556 subdirectory of @var{objdir} or vice versa. If specifying a directory
557 beneath a user's home directory tree, some shells will not expand
558 @var{dirname} correctly if it contains the @samp{~} metacharacter; use
561 The following standard @command{autoconf} options are supported. Normally you
562 should not need to use these options.
564 @item --exec-prefix=@var{dirname}
565 Specify the toplevel installation directory for architecture-dependent
566 files. The default is @file{@var{prefix}}.
568 @item --bindir=@var{dirname}
569 Specify the installation directory for the executables called by users
570 (such as @command{gcc} and @command{g++}). The default is
571 @file{@var{exec-prefix}/bin}.
573 @item --libdir=@var{dirname}
574 Specify the installation directory for object code libraries and
575 internal data files of GCC@. The default is @file{@var{exec-prefix}/lib}.
577 @item --libexecdir=@var{dirname}
578 Specify the installation directory for internal executables of GCC@.
579 The default is @file{@var{exec-prefix}/libexec}.
581 @item --with-slibdir=@var{dirname}
582 Specify the installation directory for the shared libgcc library. The
583 default is @file{@var{libdir}}.
585 @item --infodir=@var{dirname}
586 Specify the installation directory for documentation in info format.
587 The default is @file{@var{prefix}/info}.
589 @item --datadir=@var{dirname}
590 Specify the installation directory for some architecture-independent
591 data files referenced by GCC@. The default is @file{@var{prefix}/share}.
593 @item --mandir=@var{dirname}
594 Specify the installation directory for manual pages. The default is
595 @file{@var{prefix}/man}. (Note that the manual pages are only extracts from
596 the full GCC manuals, which are provided in Texinfo format. The manpages
597 are derived by an automatic conversion process from parts of the full
600 @item --with-gxx-include-dir=@var{dirname}
602 the installation directory for G++ header files. The default is
603 @file{@var{prefix}/include/g++-v3}.
607 @item --program-prefix=@var{prefix}
608 GCC supports some transformations of the names of its programs when
609 installing them. This option prepends @var{prefix} to the names of
610 programs to install in @var{bindir} (see above). For example, specifying
611 @option{--program-prefix=foo-} would result in @samp{gcc}
612 being installed as @file{/usr/local/bin/foo-gcc}.
614 @item --program-suffix=@var{suffix}
615 Appends @var{suffix} to the names of programs to install in @var{bindir}
616 (see above). For example, specifying @option{--program-suffix=-3.1}
617 would result in @samp{gcc} being installed as
618 @file{/usr/local/bin/gcc-3.1}.
620 @item --program-transform-name=@var{pattern}
621 Applies the @samp{sed} script @var{pattern} to be applied to the names
622 of programs to install in @var{bindir} (see above). @var{pattern} has to
623 consist of one or more basic @samp{sed} editing commands, separated by
624 semicolons. For example, if you want the @samp{gcc} program name to be
625 transformed to the installed program @file{/usr/local/bin/myowngcc} and
626 the @samp{g++} program name to be transformed to
627 @file{/usr/local/bin/gspecial++} without changing other program names,
628 you could use the pattern
629 @option{--program-transform-name='s/^gcc$/myowngcc/; s/^g++$/gspecial++/'}
630 to achieve this effect.
632 All three options can be combined and used together, resulting in more
633 complex conversion patterns. As a basic rule, @var{prefix} (and
634 @var{suffix}) are prepended (appended) before further transformations
635 can happen with a special transformation script @var{pattern}.
637 As currently implemented, this option only takes effect for native
638 builds; cross compiler binaries' names are not transformed even when a
639 transformation is explicitly asked for by one of these options.
641 For native builds, some of the installed programs are also installed
642 with the target alias in front of their name, as in
643 @samp{i686-pc-linux-gnu-gcc}. All of the above transformations happen
644 before the target alias is prepended to the name - so, specifying
645 @option{--program-prefix=foo-} and @option{program-suffix=-3.1}, the
646 resulting binary would be installed as
647 @file{/usr/local/bin/i686-pc-linux-gnu-foo-gcc-3.1}.
649 As a last shortcoming, none of the installed Ada programs are
650 transformed yet, which will be fixed in some time.
652 @item --with-local-prefix=@var{dirname}
654 installation directory for local include files. The default is
655 @file{/usr/local}. Specify this option if you want the compiler to
656 search directory @file{@var{dirname}/include} for locally installed
657 header files @emph{instead} of @file{/usr/local/include}.
659 You should specify @option{--with-local-prefix} @strong{only} if your
660 site has a different convention (not @file{/usr/local}) for where to put
663 The default value for @option{--with-local-prefix} is @file{/usr/local}
664 regardless of the value of @option{--prefix}. Specifying
665 @option{--prefix} has no effect on which directory GCC searches for
666 local header files. This may seem counterintuitive, but actually it is
669 The purpose of @option{--prefix} is to specify where to @emph{install
670 GCC}. The local header files in @file{/usr/local/include}---if you put
671 any in that directory---are not part of GCC@. They are part of other
672 programs---perhaps many others. (GCC installs its own header files in
673 another directory which is based on the @option{--prefix} value.)
675 Both the local-prefix include directory and the GCC-prefix include
676 directory are part of GCC's "system include" directories. Although these
677 two directories are not fixed, they need to be searched in the proper
678 order for the correct processing of the include_next directive. The
679 local-prefix include directory is searched before the GCC-prefix
680 include directory. Another characteristic of system include directories
681 is that pedantic warnings are turned off for headers in these directories.
683 Some autoconf macros add @option{-I @var{directory}} options to the
684 compiler command line, to ensure that directories containing installed
685 packages' headers are searched. When @var{directory} is one of GCC's
686 system include directories, GCC will ignore the option so that system
687 directories continue to be processed in the correct order. This
688 may result in a search order different from what was specified but the
689 directory will still be searched.
691 GCC automatically searches for ordinary libraries using
692 @env{GCC_EXEC_PREFIX}. Thus, when the same installation prefix is
693 used for both GCC and packages, GCC will automatically search for
694 both headers and libraries. This provides a configuration that is
695 easy to use. GCC behaves in a manner similar to that when it is
696 installed as a system compiler in @file{/usr}.
698 Sites that need to install multiple versions of GCC may not want to
699 use the above simple configuration. It is possible to use the
700 @option{--program-prefix}, @option{--program-suffix} and
701 @option{--program-transform-name} options to install multiple versions
702 into a single directory, but it may be simpler to use different prefixes
703 and the @option{--with-local-prefix} option to specify the location of the
704 site-specific files for each version. It will then be necessary for
705 users to specify explicitly the location of local site libraries
706 (e.g., with @env{LIBRARY_PATH}).
708 The same value can be used for both @option{--with-local-prefix} and
709 @option{--prefix} provided it is not @file{/usr}. This can be used
710 to avoid the default search of @file{/usr/local/include}.
712 @strong{Do not} specify @file{/usr} as the @option{--with-local-prefix}!
713 The directory you use for @option{--with-local-prefix} @strong{must not}
714 contain any of the system's standard header files. If it did contain
715 them, certain programs would be miscompiled (including GNU Emacs, on
716 certain targets), because this would override and nullify the header
717 file corrections made by the @command{fixincludes} script.
719 Indications are that people who use this option use it based on mistaken
720 ideas of what it is for. People use it as if it specified where to
721 install part of GCC@. Perhaps they make this assumption because
722 installing GCC creates the directory.
724 @item --enable-shared[=@var{package}[,@dots{}]]
725 Build shared versions of libraries, if shared libraries are supported on
726 the target platform. Unlike GCC 2.95.x and earlier, shared libraries
727 are enabled by default on all platforms that support shared libraries,
728 except for @samp{libobjc} which is built as a static library only by
731 If a list of packages is given as an argument, build shared libraries
732 only for the listed packages. For other packages, only static libraries
733 will be built. Package names currently recognized in the GCC tree are
734 @samp{libgcc} (also known as @samp{gcc}), @samp{libstdc++} (not
735 @samp{libstdc++-v3}), @samp{libffi}, @samp{zlib}, @samp{boehm-gc} and
736 @samp{libjava}. Note that @samp{libobjc} does not recognize itself by
737 any name, so, if you list package names in @option{--enable-shared},
738 you will only get static Objective-C libraries. @samp{libf2c} and
739 @samp{libiberty} do not support shared libraries at all.
741 Use @option{--disable-shared} to build only static libraries. Note that
742 @option{--disable-shared} does not accept a list of package names as
743 argument, only @option{--enable-shared} does.
745 @item @anchor{with-gnu-as}--with-gnu-as
746 Specify that the compiler should assume that the
747 assembler it finds is the GNU assembler. However, this does not modify
748 the rules to find an assembler and will result in confusion if the
749 assembler found is not actually the GNU assembler. (Confusion may also
750 result if the compiler finds the GNU assembler but has not been
751 configured with @option{--with-gnu-as}.) If you have more than one
752 assembler installed on your system, you may want to use this option in
753 connection with @option{--with-as=@var{pathname}}.
755 The following systems are the only ones where it makes a difference
756 whether you use the GNU assembler. On any other system,
757 @option{--with-gnu-as} has no effect.
760 @item @samp{hppa1.0-@var{any}-@var{any}}
761 @item @samp{hppa1.1-@var{any}-@var{any}}
762 @item @samp{i386-@var{any}-sysv}
763 @item @samp{m68k-bull-sysv}
764 @item @samp{m68k-hp-hpux}
765 @item @samp{m68000-hp-hpux}
766 @item @samp{m68000-att-sysv}
767 @item @samp{@var{any}-lynx-lynxos}
768 @item @samp{mips-@var{any}}
769 @item @samp{sparc-sun-solaris2.@var{any}}
770 @item @samp{sparc64-@var{any}-solaris2.@var{any}}
773 On the systems listed above (except for the HP-PA, the SPARC, for ISC on
774 the 386, and for @samp{mips-sgi-irix5.*}), if you use the GNU assembler,
775 you should also use the GNU linker (and specify @option{--with-gnu-ld}).
777 @item @anchor{with-as}--with-as=@var{pathname}
779 compiler should use the assembler pointed to by @var{pathname}, rather
780 than the one found by the standard rules to find an assembler, which
784 Check the @file{@var{libexec}/gcc/@var{target}/@var{version}}
785 directory, where @var{libexec} defaults to
786 @file{@var{exec-prefix}/libexec} and @var{exec-prefix} defaults to
787 @var{prefix} which defaults to @file{/usr/local} unless overridden by
788 the @option{--prefix=@var{pathname}} switch described
789 above. @var{target} is the target system triple, such as
790 @samp{sparc-sun-solaris2.7}, and @var{version} denotes the GCC
791 version, such as 3.0.
793 Check operating system specific directories (e.g.@: @file{/usr/ccs/bin} on
796 Note that these rules do not check for the value of @env{PATH}. You may
797 want to use @option{--with-as} if no assembler is installed in the
798 directories listed above, or if you have multiple assemblers installed
799 and want to choose one that is not found by the above rules.
801 @item @anchor{with-gnu-ld}--with-gnu-ld
802 Same as @uref{#with-gnu-as,,@option{--with-gnu-as}}
805 @item --with-ld=@var{pathname}
806 Same as @uref{#with-as,,@option{--with-as}}
810 Specify that stabs debugging
811 information should be used instead of whatever format the host normally
812 uses. Normally GCC uses the same debug format as the host system.
814 On MIPS based systems and on Alphas, you must specify whether you want
815 GCC to create the normal ECOFF debugging format, or to use BSD-style
816 stabs passed through the ECOFF symbol table. The normal ECOFF debug
817 format cannot fully handle languages other than C@. BSD stabs format can
818 handle other languages, but it only works with the GNU debugger GDB@.
820 Normally, GCC uses the ECOFF debugging format by default; if you
821 prefer BSD stabs, specify @option{--with-stabs} when you configure GCC@.
823 No matter which default you choose when you configure GCC, the user
824 can use the @option{-gcoff} and @option{-gstabs+} options to specify explicitly
825 the debug format for a particular compilation.
827 @option{--with-stabs} is meaningful on the ISC system on the 386, also, if
828 @option{--with-gas} is used. It selects use of stabs debugging
829 information embedded in COFF output. This kind of debugging information
830 supports C++ well; ordinary COFF debugging information does not.
832 @option{--with-stabs} is also meaningful on 386 systems running SVR4. It
833 selects use of stabs debugging information embedded in ELF output. The
834 C++ compiler currently (2.6.0) does not support the DWARF debugging
835 information normally used on 386 SVR4 platforms; stabs provide a
836 workable alternative. This requires gas and gdb, as the normal SVR4
837 tools can not generate or interpret stabs.
839 @item --disable-multilib
840 Specify that multiple target
841 libraries to support different target variants, calling
842 conventions, etc should not be built. The default is to build a
843 predefined set of them.
845 Some targets provide finer-grained control over which multilibs are built
846 (e.g., @option{--disable-softfloat}):
852 fpu, 26bit, underscore, interwork, biendian, nofmult.
855 softfloat, m68881, m68000, m68020.
858 single-float, biendian, softfloat.
860 @item powerpc*-*-*, rs6000*-*-*
861 aix64, pthread, softfloat, powercpu, powerpccpu, powerpcos, biendian,
866 @item --enable-threads
867 Specify that the target
868 supports threads. This affects the Objective-C compiler and runtime
869 library, and exception handling for other languages like C++ and Java.
870 On some systems, this is the default.
872 In general, the best (and, in many cases, the only known) threading
873 model available will be configured for use. Beware that on some
874 systems, gcc has not been taught what threading models are generally
875 available for the system. In this case, @option{--enable-threads} is an
876 alias for @option{--enable-threads=single}.
878 @item --disable-threads
879 Specify that threading support should be disabled for the system.
880 This is an alias for @option{--enable-threads=single}.
882 @item --enable-threads=@var{lib}
884 @var{lib} is the thread support library. This affects the Objective-C
885 compiler and runtime library, and exception handling for other languages
886 like C++ and Java. The possibilities for @var{lib} are:
894 Ada tasking support. For non-Ada programs, this setting is equivalent
895 to @samp{single}. When used in conjunction with the Ada run time, it
896 causes GCC to use the same thread primitives as Ada uses. This option
897 is necessary when using both Ada and the back end exception handling,
898 which is the default for most Ada targets.
900 Generic MACH thread support, known to work on NeXTSTEP@. (Please note
901 that the file needed to support this configuration, @file{gthr-mach.h}, is
902 missing and thus this setting will cause a known bootstrap failure.)
904 This is an alias for @samp{single}.
906 Generic POSIX thread support.
908 RTEMS thread support.
910 Disable thread support, should work for all platforms.
912 Sun Solaris 2 thread support.
914 VxWorks thread support.
916 Microsoft Win32 API thread support.
919 @item --with-cpu=@var{cpu}
920 Specify which cpu variant the compiler should generate code for by default.
921 @var{cpu} will be used as the default value of the @option{-mcpu=} switch.
922 This option is only supported on some targets, including ARM, i386, PowerPC,
925 @item --with-schedule=@var{cpu}
926 @itemx --with-arch=@var{cpu}
927 @itemx --with-tune=@var{cpu}
928 @itemx --with-abi=@var{abi}
929 @itemx --with-float=@var{type}
930 These configure options provide default values for the @option{-mschedule=},
931 @option{-march=}, @option{-mtune=}, and @option{-mabi=} options and for
932 @option{-mhard-float} or @option{-msoft-float}. As with @option{--with-cpu},
933 which switches will be accepted and acceptable values of the arguments depend
936 @item --enable-altivec
937 Specify that the target supports AltiVec vector enhancements. This
938 option will adjust the ABI for AltiVec enhancements, as well as generate
939 AltiVec code when appropriate. This option is only available for
942 @item --enable-target-optspace
944 libraries should be optimized for code space instead of code speed.
945 This is the default for the m32r platform.
948 Specify that a user visible @command{cpp} program should not be installed.
950 @item --with-cpp-install-dir=@var{dirname}
951 Specify that the user visible @command{cpp} program should be installed
952 in @file{@var{prefix}/@var{dirname}/cpp}, in addition to @var{bindir}.
954 @item --enable-initfini-array
955 Force the use of sections @code{.init_array} and @code{.fini_array}
956 (instead of @code{.init} and @code{.fini}) for constructors and
957 destructors. Option @option{--disable-initfini-array} has the
958 opposite effect. If neither option is specified, the configure script
959 will try to guess whether the @code{.init_array} and
960 @code{.fini_array} sections are supported and, if they are, use them.
962 @item --enable-maintainer-mode
964 regenerate the GCC master message catalog @file{gcc.pot} are normally
965 disabled. This is because it can only be rebuilt if the complete source
966 tree is present. If you have changed the sources and want to rebuild the
967 catalog, configuring with @option{--enable-maintainer-mode} will enable
968 this. Note that you need a recent version of the @code{gettext} tools
971 @item --enable-generated-files-in-srcdir
972 Neither the .c and .h files that are generated from bison and flex nor the
973 info manuals and man pages that are built from the .texi files are present
974 in the CVS development tree. When building GCC from that development tree,
975 or from a snapshot which are created from CVS, then those generated files
976 are placed in your build directory, which allows for the source to be in a
979 If you configure with @option{--enable-generated-files-in-srcdir} then those
980 generated files will go into the source directory. This is mainly intended
981 for generating release or prerelease tarballs of the GCC sources, since it
982 is not a requirement that the users of source releases to have flex, bison, or
985 @item --enable-version-specific-runtime-libs
987 that runtime libraries should be installed in the compiler specific
988 subdirectory (@file{@var{libdir}/gcc}) rather than the usual places. In
989 addition, @samp{libstdc++}'s include files will be installed into
990 @file{@var{libdir}} unless you overruled it by using
991 @option{--with-gxx-include-dir=@var{dirname}}. Using this option is
992 particularly useful if you intend to use several versions of GCC in
993 parallel. This is currently supported by @samp{libf2c} and
994 @samp{libstdc++}, and is the default for @samp{libobjc} which cannot be
995 changed in this case.
997 @item --enable-languages=@var{lang1},@var{lang2},@dots{}
998 Specify that only a particular subset of compilers and
999 their runtime libraries should be built. For a list of valid values for
1000 @var{langN} you can issue the following command in the
1001 @file{gcc} directory of your GCC source tree:@*
1003 grep language= */config-lang.in
1005 Currently, you can use any of the following:
1006 @code{ada}, @code{c}, @code{c++}, @code{f77}, @code{java}, @code{objc}.
1007 Building the Ada compiler has special requirements, see below.@*
1008 If you do not pass this flag, all languages available in the @file{gcc}
1009 sub-tree will be configured. Re-defining @code{LANGUAGES} when calling
1010 @samp{make bootstrap} @strong{does not} work anymore, as those
1011 language sub-directories might not have been configured!
1013 @item --disable-libgcj
1014 Specify that the run-time libraries
1015 used by GCJ should not be built. This is useful in case you intend
1016 to use GCJ with some other run-time, or you're going to install it
1017 separately, or it just happens not to build on your particular
1018 machine. In general, if the Java front end is enabled, the GCJ
1019 libraries will be enabled too, unless they're known to not work on
1020 the target platform. If GCJ is enabled but @samp{libgcj} isn't built, you
1021 may need to port it; in this case, before modifying the top-level
1022 @file{configure.in} so that @samp{libgcj} is enabled by default on this platform,
1023 you may use @option{--enable-libgcj} to override the default.
1026 Specify that the compiler should
1027 use DWARF 2 debugging information as the default.
1029 @item --enable-win32-registry
1030 @itemx --enable-win32-registry=@var{key}
1031 @itemx --disable-win32-registry
1032 The @option{--enable-win32-registry} option enables Microsoft Windows-hosted GCC
1033 to look up installations paths in the registry using the following key:
1036 @code{HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\@var{key}}
1039 @var{key} defaults to GCC version number, and can be overridden by the
1040 @option{--enable-win32-registry=@var{key}} option. Vendors and distributors
1041 who use custom installers are encouraged to provide a different key,
1042 perhaps one comprised of vendor name and GCC version number, to
1043 avoid conflict with existing installations. This feature is enabled
1044 by default, and can be disabled by @option{--disable-win32-registry}
1045 option. This option has no effect on the other hosts.
1048 Specify that the machine does not have a floating point unit. This
1049 option only applies to @samp{m68k-sun-sunos@var{n}}. On any other
1050 system, @option{--nfp} has no effect.
1052 @item --enable-werror
1053 @itemx --disable-werror
1054 @itemx --enable-werror=yes
1055 @itemx --enable-werror=no
1056 When you specify this option, it controls whether certain files in the
1057 compiler are built with @option{-Werror} in bootstrap stage2 and later.
1058 If you don't specify it, @option{-Werror} is turned on for the main
1059 development trunk. However it defaults to off for release branches and
1060 final releases. The specific files which get @option{-Werror} are
1061 controlled by the Makefiles.
1063 @item --enable-checking
1064 @itemx --enable-checking=@var{list}
1065 When you specify this option, the compiler is built to perform checking
1066 of tree node types when referencing fields of that node, and some other
1067 internal consistency checks. This does not change the generated code,
1068 but adds error checking within the compiler. This will slow down the
1069 compiler and may only work properly if you are building the compiler
1070 with GCC@. This is on by default when building from CVS or snapshots,
1071 but off for releases. More control over the checks may be had by
1072 specifying @var{list}; the categories of checks available are
1073 @samp{misc}, @samp{tree}, @samp{gc}, @samp{rtl}, @samp{rtlflag},
1074 @samp{fold}, @samp{gcac} and @samp{valgrind}. The check @samp{valgrind}
1075 requires the external @command{valgrind} simulator, available from
1076 @uref{http://valgrind.kde.org/}. The default when @var{list} is
1077 not specified is @samp{misc,tree,gc,rtlflag}; the checks @samp{rtl},
1078 @samp{gcac} and @samp{valgrind} are very expensive.
1080 @item --enable-coverage
1081 @itemx --enable-coverage=@var{level}
1082 With this option, the compiler is built to collect self coverage
1083 information, every time it is run. This is for internal development
1084 purposes, and only works when the compiler is being built with gcc. The
1085 @var{level} argument controls whether the compiler is built optimized or
1086 not, values are @samp{opt} and @samp{noopt}. For coverage analysis you
1087 want to disable optimization, for performance analysis you want to
1088 enable optimization. When coverage is enabled, the default level is
1089 without optimization.
1091 @item --enable-gather-detailed-mem-stats
1092 When this option is specified more detailed information on memory
1093 allocation is gathered. This information is printed when using
1094 @option{-fmem-report}.
1097 @itemx --disable-nls
1098 The @option{--enable-nls} option enables Native Language Support (NLS),
1099 which lets GCC output diagnostics in languages other than American
1100 English. Native Language Support is enabled by default if not doing a
1101 canadian cross build. The @option{--disable-nls} option disables NLS@.
1103 @item --with-included-gettext
1104 If NLS is enabled, the @option{--with-included-gettext} option causes the build
1105 procedure to prefer its copy of GNU @command{gettext}.
1107 @item --with-catgets
1108 If NLS is enabled, and if the host lacks @code{gettext} but has the
1109 inferior @code{catgets} interface, the GCC build procedure normally
1110 ignores @code{catgets} and instead uses GCC's copy of the GNU
1111 @code{gettext} library. The @option{--with-catgets} option causes the
1112 build procedure to use the host's @code{catgets} in this situation.
1114 @item --with-libiconv-prefix=@var{dir}
1115 Search for libiconv header files in @file{@var{dir}/include} and
1116 libiconv library files in @file{@var{dir}/lib}.
1118 @item --with-system-zlib
1119 Use installed zlib rather than that included with GCC@. This option
1120 only applies if the Java front end is being built.
1122 @item --enable-obsolete
1123 Enable configuration for an obsoleted system. If you attempt to
1124 configure GCC for a system (build, host, or target) which has been
1125 obsoleted, and you do not specify this flag, configure will halt with an
1128 All support for systems which have been obsoleted in one release of GCC
1129 is removed entirely in the next major release, unless someone steps
1130 forward to maintain the port.
1133 Some options which only apply to building cross compilers:
1135 @item --with-sysroot
1136 @itemx --with-sysroot=@var{dir}
1137 Tells GCC to consider @var{dir} as the root of a tree that contains a
1138 (subset of) the root filesystem of the target operating system.
1139 Target system headers, libraries and run-time object files will be
1140 searched in there. The specified directory is not copied into the
1141 install tree, unlike the options @option{--with-headers} and
1142 @option{--with-libs} that this option obsoletes. The default value,
1143 in case @option{--with-sysroot} is not given an argument, is
1144 @option{$@{gcc_tooldir@}/sys-root}. If the specified directory is a
1145 subdirectory of @option{$@{exec_prefix@}}, then it will be found relative to
1146 the GCC binaries if the installation tree is moved.
1148 @item --with-headers
1149 @itemx --with-headers=@var{dir}
1150 Deprecated in favor of @option{--with-sysroot}.
1151 Specifies that target headers are available when building a cross compiler.
1152 The @var{dir} argument specifies a directory which has the target include
1153 files. These include files will be copied into the @file{gcc} install
1154 directory. @emph{This option with the @var{dir} argument is required} when
1155 building a cross compiler, if @file{@var{prefix}/@var{target}/sys-include}
1156 doesn't pre-exist. If @file{@var{prefix}/@var{target}/sys-include} does
1157 pre-exist, the @var{dir} argument may be omitted. @command{fixincludes}
1158 will be run on these files to make them compatible with GCC.
1160 @item --without-headers
1161 Tells GCC not use any target headers from a libc when building a cross
1162 compiler. When crossing to GNU/Linux, you need the headers so gcc
1163 can build the exception handling for libgcc.
1164 See @uref{http://www.objsw.com/CrossGCC/,,CrossGCC} for more information
1168 @itemx --with-libs=``@var{dir1} @var{dir2} @dots{} @var{dirN}''
1169 Deprecated in favor of @option{--with-sysroot}.
1170 Specifies a list of directories which contain the target runtime
1171 libraries. These libraries will be copied into the @file{gcc} install
1172 directory. If the directory list is omitted, this option has no
1175 Specifies that @samp{newlib} is
1176 being used as the target C library. This causes @code{__eprintf} to be
1177 omitted from @file{libgcc.a} on the assumption that it will be provided by
1181 Note that each @option{--enable} option has a corresponding
1182 @option{--disable} option and that each @option{--with} option has a
1183 corresponding @option{--without} option.
1190 @uref{./index.html,,Return to the GCC Installation page}
1194 @c ***Building****************************************************************
1196 @comment node-name, next, previous, up
1197 @node Building, Testing, Configuration, Installing GCC
1203 @cindex Installing GCC: Building
1205 Now that GCC is configured, you are ready to build the compiler and
1208 We @strong{highly} recommend that GCC be built using GNU make;
1209 other versions may work, then again they might not.
1210 GNU make is required for compiling GNAT (the Ada compiler) and the Java
1213 (For example, many broken versions of make will fail if you use the
1214 recommended setup where @var{objdir} is different from @var{srcdir}.
1215 Other broken versions may recompile parts of the compiler when
1216 installing the compiler.)
1218 Some commands executed when making the compiler may fail (return a
1219 nonzero status) and be ignored by @command{make}. These failures, which
1220 are often due to files that were not found, are expected, and can safely
1223 It is normal to have compiler warnings when compiling certain files.
1224 Unless you are a GCC developer, you can generally ignore these warnings
1225 unless they cause compilation to fail. Developers should attempt to fix
1226 any warnings encountered, however they can temporarily continue past
1227 warnings-as-errors by specifying the configure flag
1228 @option{--disable-werror}.
1230 On certain old systems, defining certain environment variables such as
1231 @env{CC} can interfere with the functioning of @command{make}.
1233 If you encounter seemingly strange errors when trying to build the
1234 compiler in a directory other than the source directory, it could be
1235 because you have previously configured the compiler in the source
1236 directory. Make sure you have done all the necessary preparations.
1238 If you build GCC on a BSD system using a directory stored in an old System
1239 V file system, problems may occur in running @command{fixincludes} if the
1240 System V file system doesn't support symbolic links. These problems
1241 result in a failure to fix the declaration of @code{size_t} in
1242 @file{sys/types.h}. If you find that @code{size_t} is a signed type and
1243 that type mismatches occur, this could be the cause.
1245 The solution is not to use such a directory for building GCC@.
1247 When building from CVS or snapshots, or if you modify parser sources,
1248 you need the Bison parser generator installed. Any version 1.25 or
1249 later should work; older versions may also work. If you do not modify
1250 parser sources, releases contain the Bison-generated files and you do
1251 not need Bison installed to build them.
1253 When building from CVS or snapshots, or if you modify Texinfo
1254 documentation, you need version 4.2 or later of Texinfo installed if you
1255 want Info documentation to be regenerated. Releases contain Info
1256 documentation pre-built for the unmodified documentation in the release.
1258 @section Building a native compiler
1260 For a native build issue the command @samp{make bootstrap}. This
1261 will build the entire GCC system, which includes the following steps:
1265 Build host tools necessary to build the compiler such as texinfo, bison,
1269 Build target tools for use by the compiler such as binutils (bfd,
1270 binutils, gas, gprof, ld, and opcodes)
1271 if they have been individually linked
1272 or moved into the top level GCC source tree before configuring.
1275 Perform a 3-stage bootstrap of the compiler.
1278 Perform a comparison test of the stage2 and stage3 compilers.
1281 Build runtime libraries using the stage3 compiler from the previous step.
1285 If you are short on disk space you might consider @samp{make
1286 bootstrap-lean} instead. This is identical to @samp{make
1287 bootstrap} except that object files from the stage1 and
1288 stage2 of the 3-stage bootstrap of the compiler are deleted as
1289 soon as they are no longer needed.
1291 If you want to save additional space during the bootstrap and in
1292 the final installation as well, you can build the compiler binaries
1293 without debugging information as in the following example. This will save
1294 roughly 40% of disk space both for the bootstrap and the final installation.
1295 (Libraries will still contain debugging information.)
1298 make CFLAGS='-O' LIBCFLAGS='-g -O2' \
1299 LIBCXXFLAGS='-g -O2 -fno-implicit-templates' bootstrap
1302 If you wish to use non-default GCC flags when compiling the stage2 and
1303 stage3 compilers, set @code{BOOT_CFLAGS} on the command line when doing
1304 @samp{make bootstrap}. Non-default optimization flags are less well
1305 tested here than the default of @samp{-g -O2}, but should still work.
1306 In a few cases, you may find that you need to specify special flags such
1307 as @option{-msoft-float} here to complete the bootstrap; or, if the
1308 native compiler miscompiles the stage1 compiler, you may need to work
1309 around this, by choosing @code{BOOT_CFLAGS} to avoid the parts of the
1310 stage1 compiler that were miscompiled, or by using @samp{make
1311 bootstrap4} to increase the number of stages of bootstrap.
1313 If you used the flag @option{--enable-languages=@dots{}} to restrict
1314 the compilers to be built, only those you've actually enabled will be
1315 built. This will of course only build those runtime libraries, for
1316 which the particular compiler has been built. Please note,
1317 that re-defining @env{LANGUAGES} when calling @samp{make bootstrap}
1318 @strong{does not} work anymore!
1320 If the comparison of stage2 and stage3 fails, this normally indicates
1321 that the stage2 compiler has compiled GCC incorrectly, and is therefore
1322 a potentially serious bug which you should investigate and report. (On
1323 a few systems, meaningful comparison of object files is impossible; they
1324 always appear ``different''. If you encounter this problem, you will
1325 need to disable comparison in the @file{Makefile}.)
1327 @section Building a cross compiler
1329 We recommend reading the
1330 @uref{http://www.objsw.com/CrossGCC/,,crossgcc FAQ}
1331 for information about building cross compilers.
1333 When building a cross compiler, it is not generally possible to do a
1334 3-stage bootstrap of the compiler. This makes for an interesting problem
1335 as parts of GCC can only be built with GCC@.
1337 To build a cross compiler, we first recommend building and installing a
1338 native compiler. You can then use the native GCC compiler to build the
1339 cross compiler. The installed native compiler needs to be GCC version
1342 Assuming you have already installed a native copy of GCC and configured
1343 your cross compiler, issue the command @command{make}, which performs the
1348 Build host tools necessary to build the compiler such as texinfo, bison,
1352 Build target tools for use by the compiler such as binutils (bfd,
1353 binutils, gas, gprof, ld, and opcodes)
1354 if they have been individually linked or moved into the top level GCC source
1355 tree before configuring.
1358 Build the compiler (single stage only).
1361 Build runtime libraries using the compiler from the previous step.
1364 Note that if an error occurs in any step the make process will exit.
1366 If you are not building GNU binutils in the same source tree as GCC,
1367 you will need a cross-assembler and cross-linker installed before
1368 configuring GCC@. Put them in the directory
1369 @file{@var{prefix}/@var{target}/bin}. Here is a table of the tools
1370 you should put in this directory:
1374 This should be the cross-assembler.
1377 This should be the cross-linker.
1380 This should be the cross-archiver: a program which can manipulate
1381 archive files (linker libraries) in the target machine's format.
1384 This should be a program to construct a symbol table in an archive file.
1387 The installation of GCC will find these programs in that directory,
1388 and copy or link them to the proper place to for the cross-compiler to
1389 find them when run later.
1391 The easiest way to provide these files is to build the Binutils package.
1392 Configure it with the same @option{--host} and @option{--target}
1393 options that you use for configuring GCC, then build and install
1394 them. They install their executables automatically into the proper
1395 directory. Alas, they do not support all the targets that GCC
1398 If you are not building a C library in the same source tree as GCC,
1399 you should also provide the target libraries and headers before
1400 configuring GCC, specifying the directories with
1401 @option{--with-sysroot} or @option{--with-headers} and
1402 @option{--with-libs}. Many targets also require ``start files'' such
1403 as @file{crt0.o} and
1404 @file{crtn.o} which are linked into each executable. There may be several
1405 alternatives for @file{crt0.o}, for use with profiling or other
1406 compilation options. Check your target's definition of
1407 @code{STARTFILE_SPEC} to find out what start files it uses.
1409 @section Building in parallel
1411 You can use @samp{make bootstrap MAKE="make -j 2" -j 2}, or just
1412 @samp{make -j 2 bootstrap} for GNU Make 3.79 and above, instead of
1413 @samp{make bootstrap} to build GCC in parallel.
1414 You can also specify a bigger number, and in most cases using a value
1415 greater than the number of processors in your machine will result in
1416 fewer and shorter I/O latency hits, thus improving overall throughput;
1417 this is especially true for slow drives and network filesystems.
1419 @section Building the Ada compiler
1421 In order to build GNAT, the Ada compiler, you need a working GNAT
1422 compiler (GNAT version 3.14 or later, or GCC version 3.1 or later),
1423 including GNAT tools such as @command{gnatmake} and @command{gnatlink},
1424 since the Ada front end is written in Ada (with some
1425 GNAT-specific extensions), and GNU make.
1427 @command{configure} does not test whether the GNAT installation works
1428 and has a sufficiently recent version; if too old a GNAT version is
1429 installed, the build will fail unless @option{--enable-languages} is
1430 used to disable building the Ada front end.
1432 At the moment, the GNAT library and several tools for GNAT are not built
1433 by @samp{make bootstrap}. For a native build, you have to invoke
1434 @samp{make gnatlib_and_tools} in the @file{@var{objdir}/gcc}
1435 subdirectory before proceeding with the next steps.
1436 For a cross build, you need to invoke
1437 @samp{make gnatlib cross-gnattools ada.all.cross}. For a canadian
1438 cross you only need to invoke @samp{make cross-gnattools}; the GNAT
1439 library would be the same as the one built for the cross compiler.
1441 For example, you can build a native Ada compiler by issuing the
1442 following commands (assuming @command{make} is GNU make):
1446 @var{srcdir}/configure --enable-languages=c,ada
1450 make gnatlib_and_tools
1454 Currently, when compiling the Ada front end, you cannot use the parallel
1455 build feature described in the previous section.
1457 @section Building with profile feedback
1459 It is possible to use profile feedback to optimize the compiler itself. This
1460 should result in a faster compiler binary. Experiments done on x86 using gcc
1461 3.3 showed approximately 7 percent speedup on compiling C programs. To
1462 bootstrap compiler with profile feedback, use @code{make profiledbootstrap}.
1464 When @samp{make profiledbootstrap} is run, it will first build a @code{stage1}
1465 compiler. This compiler is used to build a @code{stageprofile} compiler
1466 instrumented to collect execution counts of instruction and branch
1467 probabilities. Then runtime libraries are compiled with profile collected.
1468 Finally a @code{stagefeedback} compiler is built using the information collected.
1470 Unlike @samp{make bootstrap} several additional restrictions apply. The
1471 compiler used to build @code{stage1} needs to support a 64-bit integral type.
1472 It is recommended to only use GCC for this. Also parallel make is currently
1473 not supported since collisions in profile collecting may occur.
1480 @uref{./index.html,,Return to the GCC Installation page}
1484 @c ***Testing*****************************************************************
1486 @comment node-name, next, previous, up
1487 @node Testing, Final install, Building, Installing GCC
1491 @chapter Installing GCC: Testing
1494 @cindex Installing GCC: Testing
1497 Before you install GCC, we encourage you to run the testsuites and to
1498 compare your results with results from a similar configuration that have
1499 been submitted to the
1500 @uref{http://gcc.gnu.org/ml/gcc-testresults/,,gcc-testresults mailing list}.
1501 Some of these archived results are linked from the build status lists
1502 at @uref{http://gcc.gnu.org/buildstat.html}, although not everyone who
1503 reports a successful build runs the testsuites and submits the results.
1504 This step is optional and may require you to download additional software,
1505 but it can give you confidence in your new GCC installation or point out
1506 problems before you install and start using your new GCC.
1508 First, you must have @uref{download.html,,downloaded the testsuites}.
1509 These are part of the full distribution, but if you downloaded the
1510 ``core'' compiler plus any front ends, you must download the testsuites
1513 Second, you must have the testing tools installed. This includes
1514 @uref{http://www.gnu.org/software/dejagnu/,,DejaGnu} 1.4.1 or 1.4.3
1515 and later, Tcl, and Expect; the DejaGnu site has links to these.
1517 If the directories where @command{runtest} and @command{expect} were
1518 installed are not in the @env{PATH}, you may need to set the following
1519 environment variables appropriately, as in the following example (which
1520 assumes that DejaGnu has been installed under @file{/usr/local}):
1523 TCL_LIBRARY = /usr/local/share/tcl8.0
1524 DEJAGNULIBS = /usr/local/share/dejagnu
1527 (On systems such as Cygwin, these paths are required to be actual
1528 paths, not mounts or links; presumably this is due to some lack of
1529 portability in the DejaGnu code.)
1532 Finally, you can run the testsuite (which may take a long time):
1534 cd @var{objdir}; make -k check
1537 This will test various components of GCC, such as compiler
1538 front ends and runtime libraries. While running the testsuite, DejaGnu
1539 might emit some harmless messages resembling
1540 @samp{WARNING: Couldn't find the global config file.} or
1541 @samp{WARNING: Couldn't find tool init file} that can be ignored.
1543 @section How can I run the test suite on selected tests?
1545 In order to run sets of tests selectively, there are targets
1546 @samp{make check-gcc} and @samp{make check-g++}
1547 in the @file{gcc} subdirectory of the object directory. You can also
1548 just run @samp{make check} in a subdirectory of the object directory.
1551 A more selective way to just run all @command{gcc} execute tests in the
1555 make check-gcc RUNTESTFLAGS="execute.exp @var{other-options}"
1558 Likewise, in order to run only the @command{g++} ``old-deja'' tests in
1559 the testsuite with filenames matching @samp{9805*}, you would use
1562 make check-g++ RUNTESTFLAGS="old-deja.exp=9805* @var{other-options}"
1565 The @file{*.exp} files are located in the testsuite directories of the GCC
1566 source, the most important ones being @file{compile.exp},
1567 @file{execute.exp}, @file{dg.exp} and @file{old-deja.exp}.
1568 To get a list of the possible @file{*.exp} files, pipe the
1569 output of @samp{make check} into a file and look at the
1570 @samp{Running @dots{} .exp} lines.
1572 @section Passing options and running multiple testsuites
1574 You can pass multiple options to the testsuite using the
1575 @samp{--target_board} option of DejaGNU, either passed as part of
1576 @samp{RUNTESTFLAGS}, or directly to @command{runtest} if you prefer to
1577 work outside the makefiles. For example,
1580 make check-g++ RUNTESTFLAGS="--target_board=unix/-O3/-fno-strength-reduce"
1583 will run the standard @command{g++} testsuites (``unix'' is the target name
1584 for a standard native testsuite situation), passing
1585 @samp{-O3 -fno-strength-reduce} to the compiler on every test, i.e.,
1586 slashes separate options.
1588 You can run the testsuites multiple times using combinations of options
1589 with a syntax similar to the brace expansion of popular shells:
1592 @dots{}"--target_board=arm-sim@{-mhard-float,-msoft-float@}@{-O1,-O2,-O3,@}"
1595 (Note the empty option caused by the trailing comma in the final group.)
1596 The following will run each testsuite eight times using the @samp{arm-sim}
1597 target, as if you had specified all possible combinations yourself:
1600 --target_board=arm-sim/-mhard-float/-O1
1601 --target_board=arm-sim/-mhard-float/-O2
1602 --target_board=arm-sim/-mhard-float/-O3
1603 --target_board=arm-sim/-mhard-float
1604 --target_board=arm-sim/-msoft-float/-O1
1605 --target_board=arm-sim/-msoft-float/-O2
1606 --target_board=arm-sim/-msoft-float/-O3
1607 --target_board=arm-sim/-msoft-float
1610 They can be combined as many times as you wish, in arbitrary ways. This
1614 @dots{}"--target_board=unix/-Wextra@{-O3,-fno-strength-reduce@}@{-fomit-frame-pointer,@}"
1617 will generate four combinations, all involving @samp{-Wextra}.
1619 The disadvantage to this method is that the testsuites are run in serial,
1620 which is a waste on multiprocessor systems. For users with GNU Make and
1621 a shell which performs brace expansion, you can run the testsuites in
1622 parallel by having the shell perform the combinations and @command{make}
1623 do the parallel runs. Instead of using @samp{--target_board}, use a
1624 special makefile target:
1627 make -j@var{N} check-@var{testsuite}//@var{test-target}/@var{option1}/@var{option2}/@dots{}
1633 make -j3 check-gcc//sh-hms-sim/@{-m1,-m2,-m3,-m3e,-m4@}/@{,-nofpu@}
1636 will run three concurrent ``make-gcc'' testsuites, eventually testing all
1637 ten combinations as described above. Note that this is currently only
1638 supported in the @file{gcc} subdirectory. (To see how this works, try
1639 typing @command{echo} before the example given here.)
1642 @section Additional testing for Java Class Libraries
1644 The Java runtime tests can be executed via @samp{make check}
1645 in the @file{@var{target}/libjava/testsuite} directory in
1648 The @uref{http://sources.redhat.com/mauve/,,Mauve Project} provides
1649 a suite of tests for the Java Class Libraries. This suite can be run
1650 as part of libgcj testing by placing the Mauve tree within the libjava
1651 testsuite at @file{libjava/testsuite/libjava.mauve/mauve}, or by
1652 specifying the location of that tree when invoking @samp{make}, as in
1653 @samp{make MAUVEDIR=~/mauve check}.
1655 @uref{http://www-124.ibm.com/developerworks/oss/cvs/jikes/~checkout~/jacks/jacks.html,,Jacks}
1656 is a free test suite that tests Java compiler front ends. This suite
1657 can be run as part of libgcj testing by placing the Jacks tree within
1658 the libjava testsuite at @file{libjava/testsuite/libjava.jacks/jacks}.
1660 @section How to interpret test results
1662 The result of running the testsuite are various @file{*.sum} and @file{*.log}
1663 files in the testsuite subdirectories. The @file{*.log} files contain a
1664 detailed log of the compiler invocations and the corresponding
1665 results, the @file{*.sum} files summarize the results. These summaries
1666 contain status codes for all tests:
1670 PASS: the test passed as expected
1672 XPASS: the test unexpectedly passed
1674 FAIL: the test unexpectedly failed
1676 XFAIL: the test failed as expected
1678 UNSUPPORTED: the test is not supported on this platform
1680 ERROR: the testsuite detected an error
1682 WARNING: the testsuite detected a possible problem
1685 It is normal for some tests to report unexpected failures. At the
1686 current time our testing harness does not allow fine grained control
1687 over whether or not a test is expected to fail. We expect to fix this
1688 problem in future releases.
1691 @section Submitting test results
1693 If you want to report the results to the GCC project, use the
1694 @file{contrib/test_summary} shell script. Start it in the @var{objdir} with
1697 @var{srcdir}/contrib/test_summary -p your_commentary.txt \
1698 -m gcc-testresults@@gcc.gnu.org |sh
1701 This script uses the @command{Mail} program to send the results, so
1702 make sure it is in your @env{PATH}. The file @file{your_commentary.txt} is
1703 prepended to the testsuite summary and should contain any special
1704 remarks you have on your results or your build environment. Please
1705 do not edit the testsuite result block or the subject line, as these
1706 messages may be automatically processed.
1713 @uref{./index.html,,Return to the GCC Installation page}
1717 @c ***Final install***********************************************************
1719 @comment node-name, next, previous, up
1720 @node Final install, , Testing, Installing GCC
1722 @ifset finalinstallhtml
1724 @chapter Installing GCC: Final installation
1727 Now that GCC has been built (and optionally tested), you can install it with
1729 cd @var{objdir}; make install
1732 We strongly recommend to install into a target directory where there is
1733 no previous version of GCC present.
1735 That step completes the installation of GCC; user level binaries can
1736 be found in @file{@var{prefix}/bin} where @var{prefix} is the value
1737 you specified with the @option{--prefix} to configure (or
1738 @file{/usr/local} by default). (If you specified @option{--bindir},
1739 that directory will be used instead; otherwise, if you specified
1740 @option{--exec-prefix}, @file{@var{exec-prefix}/bin} will be used.)
1741 Headers for the C++ and Java libraries are installed in
1742 @file{@var{prefix}/include}; libraries in @file{@var{libdir}}
1743 (normally @file{@var{prefix}/lib}); internal parts of the compiler in
1744 @file{@var{libdir}/gcc} and @file{@var{libexecdir}/gcc}; documentation
1745 in info format in @file{@var{infodir}} (normally
1746 @file{@var{prefix}/info}).
1748 When installing cross-compilers, GCC's executables
1749 are not only installed into @file{@var{bindir}}, that
1750 is, @file{@var{exec-prefix}/bin}, but additionally into
1751 @file{@var{exec-prefix}/@var{target-alias}/bin}, if that directory
1752 exists. Typically, such @dfn{tooldirs} hold target-specific
1753 binutils, including assembler and linker.
1755 Installation into a temporary staging area or into a @command{chroot}
1756 jail can be achieved with the command
1759 make DESTDIR=@var{path-to-rootdir} install
1762 @noindent where @var{path-to-rootdir} is the absolute path of
1763 a directory relative to which all installation paths will be
1764 interpreted. Note that the directory specified by @code{DESTDIR}
1765 need not exist yet; it will be created if necessary.
1767 There is a subtle point with tooldirs and @code{DESTDIR}:
1768 If you relocate a cross-compiler installation with
1769 e.g.@: @samp{DESTDIR=@var{rootdir}}, then the directory
1770 @file{@var{rootdir}/@var{exec-prefix}/@var{target-alias}/bin} will
1771 be filled with duplicated GCC executables only if it already exists,
1772 it will not be created otherwise. This is regarded as a feature,
1773 not as a bug, because it gives slightly more control to the packagers
1774 using the @code{DESTDIR} feature.
1776 If you built a released version of GCC using @samp{make bootstrap} then please
1777 quickly review the build status page for your release, available from
1778 @uref{http://gcc.gnu.org/buildstat.html}.
1779 If your system is not listed for the version of GCC that you built,
1781 @email{gcc@@gcc.gnu.org} indicating
1782 that you successfully built and installed GCC.
1783 Include the following information:
1787 Output from running @file{@var{srcdir}/config.guess}. Do not send us
1788 that file itself, just the one-line output from running it.
1791 The output of @samp{gcc -v} for your newly installed gcc.
1792 This tells us which version of GCC you built and the options you passed to
1796 Whether you enabled all languages or a subset of them. If you used a
1797 full distribution then this information is part of the configure
1798 options in the output of @samp{gcc -v}, but if you downloaded the
1799 ``core'' compiler plus additional front ends then it isn't apparent
1800 which ones you built unless you tell us about it.
1803 If the build was for GNU/Linux, also include:
1806 The distribution name and version (e.g., Red Hat 7.1 or Debian 2.2.3);
1807 this information should be available from @file{/etc/issue}.
1810 The version of the Linux kernel, available from @samp{uname --version}
1814 The version of glibc you used; for RPM-based systems like Red Hat,
1815 Mandrake, and SuSE type @samp{rpm -q glibc} to get the glibc version,
1816 and on systems like Debian and Progeny use @samp{dpkg -l libc6}.
1818 For other systems, you can include similar information if you think it is
1822 Any other information that you think would be useful to people building
1823 GCC on the same configuration. The new entry in the build status list
1824 will include a link to the archived copy of your message.
1827 We'd also like to know if the
1829 @ref{Specific, host/target specific installation notes}
1832 @uref{specific.html,,host/target specific installation notes}
1834 didn't include your host/target information or if that information is
1835 incomplete or out of date. Send a note to
1836 @email{gcc@@gcc.gnu.org} telling us how the information should be changed.
1838 If you find a bug, please report it following our
1839 @uref{../bugs.html,,bug reporting guidelines}.
1841 If you want to print the GCC manuals, do @samp{cd @var{objdir}; make
1842 dvi}. You will need to have @command{texi2dvi} (version at least 4.2)
1843 and @TeX{} installed. This creates a number of @file{.dvi} files in
1844 subdirectories of @file{@var{objdir}}; these may be converted for
1845 printing with programs such as @command{dvips}. You can also
1846 @uref{http://www.gnu.org/order/order.html,,buy printed manuals from the
1847 Free Software Foundation}, though such manuals may not be for the most
1848 recent version of GCC@.
1855 @uref{./index.html,,Return to the GCC Installation page}
1859 @c ***Binaries****************************************************************
1861 @comment node-name, next, previous, up
1862 @node Binaries, Specific, Installing GCC, Top
1866 @chapter Installing GCC: Binaries
1869 @cindex Installing GCC: Binaries
1871 We are often asked about pre-compiled versions of GCC@. While we cannot
1872 provide these for all platforms, below you'll find links to binaries for
1873 various platforms where creating them by yourself is not easy due to various
1876 Please note that we did not create these binaries, nor do we
1877 support them. If you have any problems installing them, please
1878 contact their makers.
1885 @uref{http://www.bullfreeware.com,,Bull's Freeware and Shareware Archive for AIX};
1888 @uref{http://aixpdslib.seas.ucla.edu,,UCLA Software Library for AIX}.
1892 DOS---@uref{http://www.delorie.com/djgpp/,,DJGPP}.
1895 Renesas H8/300[HS]---@uref{http://h8300-hms.sourceforge.net/,,GNU
1896 Development Tools for the Renesas H8/300[HS] Series}.
1902 @uref{http://hpux.cae.wisc.edu/,,HP-UX Porting Center};
1905 @uref{ftp://sunsite.informatik.rwth-aachen.de/pub/packages/gcc_hpux/,,Binaries for HP-UX 11.00 at Aachen University of Technology}.
1909 Motorola 68HC11/68HC12---@uref{http://www.gnu-m68hc11.org,,GNU
1910 Development Tools for the Motorola 68HC11/68HC12}.
1913 @uref{http://www.sco.com/skunkware/devtools/index.html#gcc,,SCO
1914 OpenServer/Unixware}.
1917 Sinix/Reliant Unix---@uref{ftp://ftp.fujitsu-siemens.com/pub/pd/gnu/gcc/,,Siemens}.
1920 Solaris 2 (SPARC, Intel)---@uref{http://www.sunfreeware.com/,,Sunfreeware}.
1923 SGI---@uref{http://freeware.sgi.com/,,SGI Freeware}.
1929 The @uref{http://sources.redhat.com/cygwin/,,Cygwin} project;
1931 The @uref{http://www.mingw.org/,,MinGW} project.
1935 @uref{ftp://ftp.thewrittenword.com/packages/by-name/,,The
1936 Written Word} offers binaries for
1939 Digital UNIX 4.0D and 5.1,
1941 HP-UX 10.20, 11.00, and 11.11, and
1942 Solaris/SPARC 2.5.1, 2.6, 2.7, 8, and 9,
1945 In addition to those specific offerings, you can get a binary
1946 distribution CD-ROM from the
1947 @uref{http://www.fsf.org/order/order.html,,Free Software Foundation}.
1948 It contains binaries for a number of platforms, and
1949 includes not only GCC, but other stuff as well. The current CD does
1950 not contain the latest version of GCC, but it should allow
1951 bootstrapping the compiler. An updated version of that disk is in the
1959 @uref{./index.html,,Return to the GCC Installation page}
1963 @c ***Specific****************************************************************
1965 @comment node-name, next, previous, up
1966 @node Specific, Old, Binaries, Top
1970 @chapter Host/target specific installation notes for GCC
1973 @cindex Specific installation notes
1974 @cindex Target specific installation
1975 @cindex Host specific installation
1976 @cindex Target specific installation notes
1978 Please read this document carefully @emph{before} installing the
1979 GNU Compiler Collection on your machine.
1984 @uref{#alpha*-*-*,,alpha*-*-*}
1986 @uref{#alpha*-dec-osf*,,alpha*-dec-osf*}
1988 @uref{#alphaev5-cray-unicosmk*,,alphaev5-cray-unicosmk*}
1990 @uref{#arc-*-elf,,arc-*-elf}
1992 @uref{#arm-*-elf,,arm-*-elf}
1993 @uref{#arm-*-coff,,arm-*-coff}
1994 @uref{#arm-*-aout,,arm-*-aout}
1996 @uref{#xscale-*-*,,xscale-*-*}
2004 @uref{#dsp16xx,,dsp16xx}
2006 @uref{#*-*-freebsd*,,*-*-freebsd*}
2008 @uref{#h8300-hms,,h8300-hms}
2010 @uref{#hppa*-hp-hpux*,,hppa*-hp-hpux*}
2012 @uref{#hppa*-hp-hpux10,,hppa*-hp-hpux10}
2014 @uref{#hppa*-hp-hpux11,,hppa*-hp-hpux11}
2016 @uref{#i370-*-*,,i370-*-*}
2018 @uref{#*-*-linux-gnu,,*-*-linux-gnu}
2020 @uref{#ix86-*-linux*aout,,i?86-*-linux*aout}
2022 @uref{#ix86-*-linux*,,i?86-*-linux*}
2024 @uref{#ix86-*-sco3.2v5*,,i?86-*-sco3.2v5*}
2026 @uref{#ix86-*-udk,,i?86-*-udk}
2028 @uref{#ix86-*-esix,,i?86-*-esix}
2030 @uref{#ia64-*-linux,,ia64-*-linux}
2032 @uref{#ia64-*-hpux*,,ia64-*-hpux*}
2034 @uref{#*-ibm-aix*,,*-ibm-aix*}
2036 @uref{#ip2k-*-elf,,ip2k-*-elf}
2038 @uref{#iq2000-*-elf,,iq2000-*-elf}
2040 @uref{#m32r-*-elf,,m32r-*-elf}
2042 @uref{#m6811-elf,,m6811-elf}
2044 @uref{#m6812-elf,,m6812-elf}
2046 @uref{#m68k-hp-hpux,,m68k-hp-hpux}
2048 @uref{#mips-*-*,,mips-*-*}
2050 @uref{#mips-sgi-irix5,,mips-sgi-irix5}
2052 @uref{#mips-sgi-irix6,,mips-sgi-irix6}
2054 @uref{#powerpc*-*-*,,powerpc*-*-*, powerpc-*-sysv4}
2056 @uref{#powerpc-*-darwin*,,powerpc-*-darwin*}
2058 @uref{#powerpc-*-elf,,powerpc-*-elf, powerpc-*-sysv4}
2060 @uref{#powerpc-*-linux-gnu*,,powerpc-*-linux-gnu*}
2062 @uref{#powerpc-*-netbsd*,,powerpc-*-netbsd*}
2064 @uref{#powerpc-*-eabiaix,,powerpc-*-eabiaix}
2066 @uref{#powerpc-*-eabisim,,powerpc-*-eabisim}
2068 @uref{#powerpc-*-eabi,,powerpc-*-eabi}
2070 @uref{#powerpcle-*-elf,,powerpcle-*-elf, powerpcle-*-sysv4}
2072 @uref{#powerpcle-*-eabisim,,powerpcle-*-eabisim}
2074 @uref{#powerpcle-*-eabi,,powerpcle-*-eabi}
2076 @uref{#s390-*-linux*,,s390-*-linux*}
2078 @uref{#s390x-*-linux*,,s390x-*-linux*}
2080 @uref{#s390x-ibm-tpf*,,s390x-ibm-tpf*}
2082 @uref{#*-*-solaris2*,,*-*-solaris2*}
2084 @uref{#sparc-sun-solaris2*,,sparc-sun-solaris2*}
2086 @uref{#sparc-sun-solaris2.7,,sparc-sun-solaris2.7}
2088 @uref{#sparc-*-linux*,,sparc-*-linux*}
2090 @uref{#sparc64-*-solaris2*,,sparc64-*-solaris2*}
2092 @uref{#sparcv9-*-solaris2*,,sparcv9-*-solaris2*}
2094 @uref{#*-*-sysv*,,*-*-sysv*}
2096 @uref{#vax-dec-ultrix,,vax-dec-ultrix}
2098 @uref{#*-*-vxworks*,,*-*-vxworks*}
2100 @uref{#x86_64-*-*,,x86_64-*-*, amd64-*-*}
2102 @uref{#xtensa-*-elf,,xtensa-*-elf}
2104 @uref{#xtensa-*-linux*,,xtensa-*-linux*}
2106 @uref{#windows,,Microsoft Windows}
2110 @uref{#older,,Older systems}
2115 @uref{#elf_targets,,all ELF targets} (SVR4, Solaris 2, etc.)
2121 <!-- -------- host/target specific issues start here ---------------- -->
2124 @heading @anchor{alpha*-*-*}alpha*-*-*
2126 This section contains general configuration information for all
2127 alpha-based platforms using ELF (in particular, ignore this section for
2128 DEC OSF/1, Digital UNIX and Tru64 UNIX)@. In addition to reading this
2129 section, please read all other sections that match your target.
2131 We require binutils 2.11.2 or newer.
2132 Previous binutils releases had a number of problems with DWARF 2
2133 debugging information, not the least of which is incorrect linking of
2139 @heading @anchor{alpha*-dec-osf*}alpha*-dec-osf*
2140 Systems using processors that implement the DEC Alpha architecture and
2141 are running the DEC/Compaq Unix (DEC OSF/1, Digital UNIX, or Compaq
2142 Tru64 UNIX) operating system, for example the DEC Alpha AXP systems.
2144 As of GCC 3.2, versions before @code{alpha*-dec-osf4} are no longer
2145 supported. (These are the versions which identify themselves as DEC
2148 In Digital Unix V4.0, virtual memory exhausted bootstrap failures
2149 may be fixed by configuring with @option{--with-gc=simple},
2150 reconfiguring Kernel Virtual Memory and Swap parameters
2151 per the @command{/usr/sbin/sys_check} Tuning Suggestions,
2152 or applying the patch in
2153 @uref{http://gcc.gnu.org/ml/gcc/2002-08/msg00822.html}.
2155 In Tru64 UNIX V5.1, Compaq introduced a new assembler that does not
2156 currently (2001-06-13) work with @command{mips-tfile}. As a workaround,
2157 we need to use the old assembler, invoked via the barely documented
2158 @option{-oldas} option. To bootstrap GCC, you either need to use the
2162 % CC=cc @var{srcdir}/configure [@var{options}] [@var{target}]
2165 or you can use a copy of GCC 2.95.3 or higher built on Tru64 UNIX V4.0:
2168 % CC=gcc -Wa,-oldas @var{srcdir}/configure [@var{options}] [@var{target}]
2171 As of GNU binutils 2.11.2, neither GNU @command{as} nor GNU @command{ld}
2172 are supported on Tru64 UNIX, so you must not configure GCC with
2173 @option{--with-gnu-as} or @option{--with-gnu-ld}.
2175 GCC writes a @samp{.verstamp} directive to the assembler output file
2176 unless it is built as a cross-compiler. It gets the version to use from
2177 the system header file @file{/usr/include/stamp.h}. If you install a
2178 new version of DEC Unix, you should rebuild GCC to pick up the new version
2181 Note that since the Alpha is a 64-bit architecture, cross-compilers from
2182 32-bit machines will not generate code as efficient as that generated
2183 when the compiler is running on a 64-bit machine because many
2184 optimizations that depend on being able to represent a word on the
2185 target in an integral value on the host cannot be performed. Building
2186 cross-compilers on the Alpha for 32-bit machines has only been tested in
2187 a few cases and may not work properly.
2189 @samp{make compare} may fail on old versions of DEC Unix unless you add
2190 @option{-save-temps} to @code{CFLAGS}. On these systems, the name of the
2191 assembler input file is stored in the object file, and that makes
2192 comparison fail if it differs between the @code{stage1} and
2193 @code{stage2} compilations. The option @option{-save-temps} forces a
2194 fixed name to be used for the assembler input file, instead of a
2195 randomly chosen name in @file{/tmp}. Do not add @option{-save-temps}
2196 unless the comparisons fail without that option. If you add
2197 @option{-save-temps}, you will have to manually delete the @samp{.i} and
2198 @samp{.s} files after each series of compilations.
2200 GCC now supports both the native (ECOFF) debugging format used by DBX
2201 and GDB and an encapsulated STABS format for use only with GDB@. See the
2202 discussion of the @option{--with-stabs} option of @file{configure} above
2203 for more information on these formats and how to select them.
2205 There is a bug in DEC's assembler that produces incorrect line numbers
2206 for ECOFF format when the @samp{.align} directive is used. To work
2207 around this problem, GCC will not emit such alignment directives
2208 while writing ECOFF format debugging information even if optimization is
2209 being performed. Unfortunately, this has the very undesirable
2210 side-effect that code addresses when @option{-O} is specified are
2211 different depending on whether or not @option{-g} is also specified.
2213 To avoid this behavior, specify @option{-gstabs+} and use GDB instead of
2214 DBX@. DEC is now aware of this problem with the assembler and hopes to
2215 provide a fix shortly.
2220 @heading @anchor{alphaev5-cray-unicosmk*}alphaev5-cray-unicosmk*
2221 Cray T3E systems running Unicos/Mk.
2223 This port is incomplete and has many known bugs. We hope to improve the
2224 support for this target soon. Currently, only the C front end is supported,
2225 and it is not possible to build parallel applications. Cray modules are not
2226 supported; in particular, Craylibs are assumed to be in
2227 @file{/opt/ctl/craylibs/craylibs}.
2229 You absolutely @strong{must} use GNU make on this platform. Also, you
2230 need to tell GCC where to find the assembler and the linker. The
2231 simplest way to do so is by providing @option{--with-as} and
2232 @option{--with-ld} to @file{configure}, e.g.@:
2235 configure --with-as=/opt/ctl/bin/cam --with-ld=/opt/ctl/bin/cld \
2236 --enable-languages=c
2239 The comparison test during @samp{make bootstrap} fails on Unicos/Mk
2240 because the assembler inserts timestamps into object files. You should
2241 be able to work around this by doing @samp{make all} after getting this
2247 @heading @anchor{arc-*-elf}arc-*-elf
2248 Argonaut ARC processor.
2249 This configuration is intended for embedded systems.
2254 @heading @anchor{arm-*-elf}arm-*-elf
2255 @heading @anchor{xscale-*-*}xscale-*-*
2256 ARM-family processors. Subtargets that use the ELF object format
2257 require GNU binutils 2.13 or newer. Such subtargets include:
2258 @code{arm-*-freebsd}, @code{arm-*-netbsdelf}, @code{arm-*-*linux},
2259 @code{arm-*-rtems} and @code{arm-*-kaos}.
2264 @heading @anchor{arm-*-coff}arm-*-coff
2265 ARM-family processors. Note that there are two different varieties
2266 of PE format subtarget supported: @code{arm-wince-pe} and
2267 @code{arm-pe} as well as a standard COFF target @code{arm-*-coff}.
2272 @heading @anchor{arm-*-aout}arm-*-aout
2273 ARM-family processors. These targets support the AOUT file format:
2274 @code{arm-*-aout}, @code{arm-*-netbsd}.
2279 @heading @anchor{avr}avr
2281 ATMEL AVR-family micro controllers. These are used in embedded
2282 applications. There are no standard Unix configurations.
2284 @xref{AVR Options,, AVR Options, gcc, Using and Porting the GNU Compiler
2288 See ``AVR Options'' in the main manual
2290 for the list of supported MCU types.
2292 Use @samp{configure --target=avr --enable-languages="c"} to configure GCC@.
2294 Further installation notes and other useful information about AVR tools
2295 can also be obtained from:
2299 @uref{http://www.openavr.org,,http://www.openavr.org}
2301 @uref{http://home.overta.ru/users/denisc/,,http://home.overta.ru/users/denisc/}
2303 @uref{http://www.amelek.gda.pl/avr/,,http://www.amelek.gda.pl/avr/}
2306 We @emph{strongly} recommend using binutils 2.13 or newer.
2308 The following error:
2310 Error: register required
2313 indicates that you should upgrade to a newer version of the binutils.
2318 @heading @anchor{c4x}c4x
2320 Texas Instruments TMS320C3x and TMS320C4x Floating Point Digital Signal
2321 Processors. These are used in embedded applications. There are no
2322 standard Unix configurations.
2324 @xref{TMS320C3x/C4x Options,, TMS320C3x/C4x Options, gcc, Using and
2325 Porting the GNU Compiler Collection (GCC)},
2328 See ``TMS320C3x/C4x Options'' in the main manual
2330 for the list of supported MCU types.
2332 GCC can be configured as a cross compiler for both the C3x and C4x
2333 architectures on the same system. Use @samp{configure --target=c4x
2334 --enable-languages="c,c++"} to configure.
2337 Further installation notes and other useful information about C4x tools
2338 can also be obtained from:
2342 @uref{http://www.elec.canterbury.ac.nz/c4x/,,http://www.elec.canterbury.ac.nz/c4x/}
2348 @heading @anchor{cris}CRIS
2350 CRIS is the CPU architecture in Axis Communications ETRAX system-on-a-chip
2351 series. These are used in embedded applications.
2354 @xref{CRIS Options,, CRIS Options, gcc, Using and Porting the GNU Compiler
2358 See ``CRIS Options'' in the main manual
2360 for a list of CRIS-specific options.
2362 There are a few different CRIS targets:
2364 @item cris-axis-aout
2365 Old target. Includes a multilib for the @samp{elinux} a.out-based
2366 target. No multilibs for newer architecture variants.
2368 Mainly for monolithic embedded systems. Includes a multilib for the
2369 @samp{v10} core used in @samp{ETRAX 100 LX}.
2370 @item cris-axis-linux-gnu
2371 A GNU/Linux port for the CRIS architecture, currently targeting
2372 @samp{ETRAX 100 LX} by default.
2375 For @code{cris-axis-aout} and @code{cris-axis-elf} you need binutils 2.11
2376 or newer. For @code{cris-axis-linux-gnu} you need binutils 2.12 or newer.
2378 Pre-packaged tools can be obtained from
2379 @uref{ftp://ftp.axis.com/pub/axis/tools/cris/compiler-kit/}. More
2380 information about this platform is available at
2381 @uref{http://developer.axis.com/}.
2386 @heading @anchor{dos}DOS
2388 Please have a look at our @uref{binaries.html,,binaries page}.
2390 You cannot install GCC by itself on MSDOS; it will not compile under
2391 any MSDOS compiler except itself. You need to get the complete
2392 compilation package DJGPP, which includes binaries as well as sources,
2393 and includes all the necessary compilation tools and libraries.
2398 @heading @anchor{dsp16xx}dsp16xx
2399 A port to the AT&T DSP1610 family of processors.
2404 @heading @anchor{*-*-freebsd*}*-*-freebsd*
2406 The version of binutils installed in @file{/usr/bin} is known to work unless
2407 otherwise specified in any per-architecture notes. However, binutils
2408 2.12.1 or greater is known to improve overall testsuite results.
2410 Support for FreeBSD 1 was discontinued in GCC 3.2.
2412 For FreeBSD 2 or any mutant a.out versions of FreeBSD 3: All
2413 configuration support and files as shipped with GCC 2.95 are still in
2414 place. FreeBSD 2.2.7 has been known to bootstrap completely; however,
2415 it is unknown which version of binutils was used (it is assumed that it
2416 was the system copy in @file{/usr/bin}) and C++ EH failures were noted.
2418 For FreeBSD using the ELF file format: DWARF 2 debugging is now the
2419 default for all CPU architectures. It had been the default on
2420 FreeBSD/alpha since its inception. You may use @option{-gstabs} instead
2421 of @option{-g}, if you really want the old debugging format. There are
2422 no known issues with mixing object files and libraries with different
2423 debugging formats. Otherwise, this release of GCC should now match more
2424 of the configuration used in the stock FreeBSD configuration of GCC. In
2425 particular, @option{--enable-threads} is now configured by default.
2426 However, as a general user, do not attempt to replace the system
2427 compiler with this release. Known to bootstrap and check with good
2428 results on FreeBSD 4.8-STABLE and 5-CURRENT@. In the past, known to
2429 bootstrap and check with good results on FreeBSD 3.0, 3.4, 4.0, 4.2,
2430 4.3, 4.4, 4.5-STABLE@.
2432 In principle, @option{--enable-threads} is now compatible with
2433 @option{--enable-libgcj} on FreeBSD@. However, it has only been built
2434 and tested on @samp{i386-*-freebsd[45]} and @samp{alpha-*-freebsd[45]}.
2436 library may be incorrectly built (symbols are missing at link time).
2437 There is a rare timing-based startup hang (probably involves an
2438 assumption about the thread library). Multi-threaded boehm-gc (required for
2439 libjava) exposes severe threaded signal-handling bugs on FreeBSD before
2440 4.5-RELEASE@. Other CPU architectures
2441 supported by FreeBSD will require additional configuration tuning in, at
2442 the very least, both boehm-gc and libffi.
2444 Shared @file{libgcc_s.so} is now built and installed by default.
2449 @heading @anchor{h8300-hms}h8300-hms
2450 Renesas H8/300 series of processors.
2452 Please have a look at our @uref{binaries.html,,binaries page}.
2454 The calling convention and structure layout has changed in release 2.6.
2455 All code must be recompiled. The calling convention now passes the
2456 first three arguments in function calls in registers. Structures are no
2457 longer a multiple of 2 bytes.
2462 @heading @anchor{hppa*-hp-hpux*}hppa*-hp-hpux*
2463 Support for HP-UX version 9 and older was discontinued in GCC 3.4.
2465 We @emph{highly} recommend using gas/binutils 2.8 or newer on all hppa
2466 platforms; you may encounter a variety of problems when using the HP
2469 Specifically, @option{-g} does not work on HP-UX (since that system
2470 uses a peculiar debugging format which GCC does not know about), unless you
2471 use GAS and GDB and configure GCC with the
2472 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}} and
2473 @option{--with-as=@dots{}} options.
2475 If you wish to use the pa-risc 2.0 architecture support with a 32-bit
2476 runtime, you must use either the HP assembler, gas/binutils 2.11 or newer,
2478 @uref{ftp://sources.redhat.com/pub/binutils/snapshots,,snapshot of gas}.
2480 There are two default scheduling models for instructions. These are
2481 PROCESSOR_7100LC and PROCESSOR_8000. They are selected from the pa-risc
2482 architecture specified for the target machine when configuring.
2483 PROCESSOR_8000 is the default. PROCESSOR_7100LC is selected when
2484 the target is a @samp{hppa1*} machine.
2486 The PROCESSOR_8000 model is not well suited to older processors. Thus,
2487 it is important to completely specify the machine architecture when
2488 configuring if you want a model other than PROCESSOR_8000. The macro
2489 TARGET_SCHED_DEFAULT can be defined in BOOT_CFLAGS if a different
2490 default scheduling model is desired.
2492 More specific information to @samp{hppa*-hp-hpux*} targets follows.
2497 @heading @anchor{hppa*-hp-hpux10}hppa*-hp-hpux10
2499 For hpux10.20, we @emph{highly} recommend you pick up the latest sed patch
2500 @code{PHCO_19798} from HP@. HP has two sites which provide patches free of
2506 <a href="http://us.itrc.hp.com/service/home/home.do">US, Canada, Asia-Pacific, and
2510 @uref{http://us.itrc.hp.com/service/home/home.do,,} US, Canada, Asia-Pacific,
2514 @uref{http://europe.itrc.hp.com/service/home/home.do,,} Europe.
2517 The HP assembler on these systems has some problems. Most notably the
2518 assembler inserts timestamps into each object file it creates, causing
2519 the 3-stage comparison test to fail during a @samp{make bootstrap}.
2520 You should be able to continue by saying @samp{make all} after getting
2521 the failure from @samp{make bootstrap}.
2527 @heading @anchor{hppa*-hp-hpux11}hppa*-hp-hpux11
2529 GCC 3.0 and up support HP-UX 11. On 64-bit capable systems, there
2530 are two distinct ports. The @samp{hppa2.0w-hp-hpux11*} port generates
2531 code for the 32-bit pa-risc runtime architecture. It uses the HP
2532 linker. The @samp{hppa64-hp-hpux11*} port generates 64-bit code for the
2533 pa-risc 2.0 architecture. The script config.guess now selects the port
2534 type based on the type compiler detected during configuration. You must
2535 set your @env{PATH} or define @env{CC} so that configure finds an appropriate
2536 compiler for the initial bootstrap. Different prefixes must be used if
2537 both ports are to be installed on the same system.
2539 It is best to explicitly configure the @samp{hppa64-hp-hpux11*} target
2540 with the @option{--with-ld=@dots{}} option. We support both the HP
2541 and GNU linkers for this target. The two linkers require different
2542 link commands. Thus, it's not possible to switch linkers during a
2543 GCC build. This has been been reported to occur in a unified build
2544 of binutils and GCC.
2546 GCC 2.95.x is not supported under HP-UX 11 and cannot be used to
2547 compile GCC 3.0 and up. Refer to @uref{binaries.html,,binaries} for
2548 information about obtaining precompiled GCC binaries for HP-UX.
2550 You must use GNU binutils 2.11 or above with the 32-bit port. Thread
2551 support is not currently implemented, so @option{--enable-threads} does
2555 @item @uref{http://gcc.gnu.org/ml/gcc-prs/2002-01/msg00551.html}
2556 @item @uref{http://gcc.gnu.org/ml/gcc-bugs/2002-01/msg00663.html}
2559 GCC 3.3 and later support weak symbols on the 32-bit port using SOM
2560 secondary definition symbols. This feature is not enabled for earlier
2561 versions of HP-UX since there have been bugs in the linker support for
2562 secondary symbols. The HP linker patches @code{PHSS_26559} and
2563 @code{PHSS_24304} for HP-UX 11.00 and 11.11, respectively, correct the
2564 problem of linker core dumps creating C++ libraries. Earlier patches
2565 may work but they have not been tested.
2567 GCC 3.3 nows uses the ELF DT_INIT_ARRAY and DT_FINI_ARRAY capability
2568 to run initializers and finalizers on the 64-bit port. The feature
2569 requires CVS binutils as of January 2, 2003, or a subsequent release
2570 to correct a problem arising from HP's non-standard use of the .init
2571 and .fini sections. The 32-bit port uses the linker @option{+init}
2572 and @option{+fini} options. As with the support for secondary symbols,
2573 there have been bugs in the order in which these options are executed
2574 by the HP linker. So, again a recent linker patch is recommended.
2576 The HP assembler has many limitations and is not recommended for either
2577 the 32 or 64-bit ports. For example, it does not support weak symbols
2578 or alias definitions. As a result, explicit template instantiations
2579 are required when using C++. This will make it difficult if not
2580 impossible to build many C++ applications. You also can't generate
2581 debugging information when using the HP assembler with GCC.
2583 There are a number of issues to consider in selecting which linker to
2584 use with the 64-bit port. The GNU 64-bit linker can only create dynamic
2585 binaries. The @option{-static} option causes linking with archive
2586 libraries but doesn't produce a truly static binary. Dynamic binaries
2587 still require final binding by the dynamic loader to resolve a set of
2588 dynamic-loader-defined symbols. The default behavior of the HP linker
2589 is the same as the GNU linker. However, it can generate true 64-bit
2590 static binaries using the @option{+compat} option.
2592 The HP 64-bit linker doesn't support linkonce semantics. As a
2593 result, C++ programs have many more sections than they should.
2595 The GNU 64-bit linker has some issues with shared library support
2596 and exceptions. As a result, we only support libgcc in archive
2597 format. For similar reasons, dwarf2 unwind and exception support
2598 are disabled. The GNU linker also has problems creating binaries
2599 with @option{-static}. It doesn't provide stubs for internal
2600 calls to global functions in shared libraries, so these calls
2601 can't be overloaded.
2603 There are several possible approaches to building the distribution.
2604 Binutils can be built first using the HP tools. Then, the GCC
2605 distribution can be built. The second approach is to build GCC
2606 first using the HP tools, then build binutils, then rebuild GCC.
2607 There have been problems with various binary distributions, so
2608 it is best not to start from a binary distribution.
2610 Starting with GCC 3.4 an ISO C compiler is required to bootstrap.
2611 The bundled compiler supports only traditional C; you will need
2612 either HP's unbundled compiler, or a binary distribution of GCC@.
2614 This port still is undergoing significant development.
2619 @heading @anchor{i370-*-*}i370-*-*
2620 This port is very preliminary and has many known bugs. We hope to
2621 have a higher-quality port for this machine soon.
2626 @heading @anchor{*-*-linux-gnu}*-*-linux-gnu
2628 Versions of libstdc++-v3 starting with 3.2.1 require bugfixes present
2629 in glibc 2.2.5 and later. More information is available in the
2630 libstdc++-v3 documentation.
2635 @heading @anchor{ix86-*-linux*aout}i?86-*-linux*aout
2636 Use this configuration to generate @file{a.out} binaries on Linux-based
2637 GNU systems. This configuration is being superseded.
2642 @heading @anchor{ix86-*-linux*}i?86-*-linux*
2644 As of GCC 3.3, binutils 2.13.1 or later is required for this platform.
2645 See @uref{http://gcc.gnu.org/PR10877,,bug 10877} for more information.
2647 If you receive Signal 11 errors when building on GNU/Linux, then it is
2648 possible you have a hardware problem. Further information on this can be
2649 found on @uref{http://www.bitwizard.nl/sig11/,,www.bitwizard.nl}.
2654 @heading @anchor{ix86-*-sco3.2v5*}i?86-*-sco3.2v5*
2655 Use this for the SCO OpenServer Release 5 family of operating systems.
2657 Unlike earlier versions of GCC, the ability to generate COFF with this
2658 target is no longer provided.
2660 Earlier versions of GCC emitted DWARF 1 when generating ELF to allow
2661 the system debugger to be used. That support was too burdensome to
2662 maintain. GCC now emits only DWARF 2 for this target. This means you
2663 may use either the UDK debugger or GDB to debug programs built by this
2666 GCC is now only supported on releases 5.0.4 and later, and requires that
2667 you install Support Level Supplement OSS646B or later, and Support Level
2668 Supplement OSS631C or later. If you are using release 5.0.7 of
2669 OpenServer, you must have at least the first maintenance pack installed
2670 (this includes the relevant portions of OSS646). OSS646, also known as
2671 the "Execution Environment Update", provides updated link editors and
2672 assemblers, as well as updated standard C and math libraries. The C
2673 startup modules are also updated to support the System V gABI draft, and
2674 GCC relies on that behavior. OSS631 provides a collection of commonly
2675 used open source libraries, some of which GCC depends on (such as GNU
2676 gettext and zlib). SCO OpenServer Release 5.0.7 has all of this built
2677 in by default, but OSS631C and later also apply to that release. Please
2679 @uref{ftp://ftp.sco.com/pub/openserver5,,ftp://ftp.sco.com/pub/openserver5}
2680 for the latest versions of these (and other potentially useful)
2683 Although there is support for using the native assembler, it is
2684 recommended that you configure GCC to use the GNU assembler. You do
2685 this by using the flags
2686 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}}. You should
2687 use a modern version of GNU binutils. Version 2.13.2.1 was used for all
2688 testing. In general, only the @option{--with-gnu-as} option is tested.
2689 A modern bintuils (as well as a plethora of other development related
2690 GNU utilities) can be found in Support Level Supplement OSS658A, the
2691 "GNU Development Tools" package. See the SCO web and ftp sites for details.
2692 That package also contains the currently "officially supported" version of
2693 GCC, version 2.95.3. It is useful for bootstrapping this version.
2698 @heading @anchor{ix86-*-udk}i?86-*-udk
2700 This target emulates the SCO Universal Development Kit and requires that
2701 package be installed. (If it is installed, you will have a
2702 @file{/udk/usr/ccs/bin/cc} file present.) It's very much like the
2703 @samp{i?86-*-unixware7*} target
2704 but is meant to be used when hosting on a system where UDK isn't the
2705 default compiler such as OpenServer 5 or Unixware 2. This target will
2706 generate binaries that will run on OpenServer, Unixware 2, or Unixware 7,
2707 with the same warnings and caveats as the SCO UDK@.
2709 This target is a little tricky to build because we have to distinguish
2710 it from the native tools (so it gets headers, startups, and libraries
2711 from the right place) while making the tools not think we're actually
2712 building a cross compiler. The easiest way to do this is with a configure
2716 CC=/udk/usr/ccs/bin/cc @var{/your/path/to}/gcc/configure \
2717 --host=i686-pc-udk --target=i686-pc-udk --program-prefix=udk-
2720 @emph{You should substitute @samp{i686} in the above command with the appropriate
2721 processor for your host.}
2723 After the usual @samp{make bootstrap} and
2724 @samp{make install}, you can then access the UDK-targeted GCC
2725 tools by adding @command{udk-} before the commonly known name. For
2726 example, to invoke the C compiler, you would use @command{udk-gcc}.
2727 They will coexist peacefully with any native-target GCC tools you may
2734 @heading @anchor{ia64-*-linux}ia64-*-linux
2735 IA-64 processor (also known as IPF, or Itanium Processor Family)
2738 If you are using the optional libunwind library, then you must use
2739 libunwind 0.96 or later.
2741 None of the following versions of GCC has an ABI that is compatible
2742 with any of the other versions in this list, with the exception that
2743 Red Hat 2.96 and Trillian 000171 are compatible with each other:
2744 3.1, 3.0.2, 3.0.1, 3.0, Red Hat 2.96, and Trillian 000717.
2745 This primarily affects C++ programs and programs that create shared libraries.
2746 GCC 3.1 or later is recommended for compiling linux, the kernel.
2747 As of version 3.1 GCC is believed to be fully ABI compliant, and hence no
2748 more major ABI changes are expected.
2753 @heading @anchor{ia64-*-hpux*}ia64-*-hpux*
2754 Building GCC on this target requires the GNU Assembler. The bundled HP
2755 assembler will not work. To prevent GCC from using the wrong assembler,
2756 the option @option{--with-gnu-as} may be necessary.
2758 The GCC libunwind library has not been ported to HPUX. This means that for
2759 GCC versions 3.2.3 and earlier, @option{--enable-libunwind-exceptions}
2760 is required to build GCC. For GCC 3.3 and later, this is the default.
2764 <!-- rs6000-ibm-aix*, powerpc-ibm-aix* -->
2766 @heading @anchor{*-ibm-aix*}*-ibm-aix*
2767 Support for AIX version 3 and older was discontinued in GCC 3.4.
2769 AIX Make frequently has problems with GCC makefiles. GNU Make 3.79.1 or
2770 newer is recommended to build on this platform.
2772 Errors involving @code{alloca} when building GCC generally are due
2773 to an incorrect definition of @code{CC} in the Makefile or mixing files
2774 compiled with the native C compiler and GCC@. During the stage1 phase of
2775 the build, the native AIX compiler @strong{must} be invoked as @command{cc}
2776 (not @command{xlc}). Once @command{configure} has been informed of
2777 @command{xlc}, one needs to use @samp{make distclean} to remove the
2778 configure cache files and ensure that @env{CC} environment variable
2779 does not provide a definition that will confuse @command{configure}.
2780 If this error occurs during stage2 or later, then the problem most likely
2781 is the version of Make (see above).
2783 The native @command{as} and @command{ld} are recommended for bootstrapping
2784 on AIX 4 and required for bootstrapping on AIX 5L. The GNU Assembler
2785 reports that it supports WEAK symbols on AIX 4, which causes GCC to try to
2786 utilize weak symbol functionality although it is not supported. The GNU
2787 Assembler and Linker do not support AIX 5L sufficiently to bootstrap GCC.
2788 The native AIX tools do interoperate with GCC@.
2790 Building @file{libstdc++.a} requires a fix for an AIX Assembler bug
2791 APAR IY26685 (AIX 4.3) or APAR IY25528 (AIX 5.1).
2793 @samp{libstdc++} in GCC 3.2 increments the major version number of the
2794 shared object and GCC installation places the @file{libstdc++.a}
2795 shared library in a common location which will overwrite the GCC 3.1
2796 version of the shared library. Applications either need to be
2797 re-linked against the new shared library or the GCC 3.1 version of the
2798 @samp{libstdc++} shared object needs to be available to the AIX
2799 runtime loader. The GCC 3.1 @samp{libstdc++.so.4} shared object can
2800 be installed for runtime dynamic loading using the following steps to
2801 set the @samp{F_LOADONLY} flag in the shared object for @emph{each}
2802 multilib @file{libstdc++.a} installed:
2804 Extract the shared object from each the GCC 3.1 @file{libstdc++.a}
2807 % ar -x libstdc++.a libstdc++.so.4
2810 Enable the @samp{F_LOADONLY} flag so that the shared object will be
2811 available for runtime dynamic loading, but not linking:
2813 % strip -e libstdc++.so.4
2816 Archive the runtime-only shared object in the GCC 3.2
2817 @file{libstdc++.a} archive:
2819 % ar -q libstdc++.a libstdc++.so.4
2822 Linking executables and shared libraries may produce warnings of
2823 duplicate symbols. The assembly files generated by GCC for AIX always
2824 have included multiple symbol definitions for certain global variable
2825 and function declarations in the original program. The warnings should
2826 not prevent the linker from producing a correct library or runnable
2829 AIX 4.3 utilizes a ``large format'' archive to support both 32-bit and
2830 64-bit object modules. The routines provided in AIX 4.3.0 and AIX 4.3.1
2831 to parse archive libraries did not handle the new format correctly.
2832 These routines are used by GCC and result in error messages during
2833 linking such as ``not a COFF file''. The version of the routines shipped
2834 with AIX 4.3.1 should work for a 32-bit environment. The @option{-g}
2835 option of the archive command may be used to create archives of 32-bit
2836 objects using the original ``small format''. A correct version of the
2837 routines is shipped with AIX 4.3.2 and above.
2839 Some versions of the AIX binder (linker) can fail with a relocation
2840 overflow severe error when the @option{-bbigtoc} option is used to link
2841 GCC-produced object files into an executable that overflows the TOC@. A fix
2842 for APAR IX75823 (OVERFLOW DURING LINK WHEN USING GCC AND -BBIGTOC) is
2843 available from IBM Customer Support and from its
2844 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
2845 website as PTF U455193.
2847 The AIX 4.3.2.1 linker (bos.rte.bind_cmds Level 4.3.2.1) will dump core
2848 with a segmentation fault when invoked by any version of GCC@. A fix for
2849 APAR IX87327 is available from IBM Customer Support and from its
2850 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
2851 website as PTF U461879. This fix is incorporated in AIX 4.3.3 and above.
2853 The initial assembler shipped with AIX 4.3.0 generates incorrect object
2854 files. A fix for APAR IX74254 (64BIT DISASSEMBLED OUTPUT FROM COMPILER FAILS
2855 TO ASSEMBLE/BIND) is available from IBM Customer Support and from its
2856 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
2857 website as PTF U453956. This fix is incorporated in AIX 4.3.1 and above.
2859 AIX provides National Language Support (NLS)@. Compilers and assemblers
2860 use NLS to support locale-specific representations of various data
2861 formats including floating-point numbers (e.g., @samp{.} vs @samp{,} for
2862 separating decimal fractions). There have been problems reported where
2863 GCC does not produce the same floating-point formats that the assembler
2864 expects. If one encounters this problem, set the @env{LANG}
2865 environment variable to @samp{C} or @samp{En_US}.
2867 By default, GCC for AIX 4.1 and above produces code that can be used on
2868 both Power or PowerPC processors.
2870 A default can be specified with the @option{-mcpu=@var{cpu_type}}
2871 switch and using the configure option @option{--with-cpu-@var{cpu_type}}.
2876 @heading @anchor{ip2k-*-elf}ip2k-*-elf
2877 Ubicom IP2022 micro controller.
2878 This configuration is intended for embedded systems.
2879 There are no standard Unix configurations.
2881 Use @samp{configure --target=ip2k-elf --enable-languages=c} to configure GCC@.
2886 @heading @anchor{iq2000-*-elf}iq2000-*-elf
2887 Vitesse IQ2000 processors. These are used in embedded
2888 applications. There are no standard Unix configurations.
2893 @heading @anchor{m32r-*-elf}m32r-*-elf
2894 Renesas M32R processor.
2895 This configuration is intended for embedded systems.
2900 @heading @anchor{m6811-elf}m6811-elf
2901 Motorola 68HC11 family micro controllers. These are used in embedded
2902 applications. There are no standard Unix configurations.
2907 @heading @anchor{m6812-elf}m6812-elf
2908 Motorola 68HC12 family micro controllers. These are used in embedded
2909 applications. There are no standard Unix configurations.
2914 @heading @anchor{m68k-hp-hpux}m68k-hp-hpux
2915 HP 9000 series 300 or 400 running HP-UX@. HP-UX version 8.0 has a bug in
2916 the assembler that prevents compilation of GCC@. This
2917 bug manifests itself during the first stage of compilation, while
2918 building @file{libgcc2.a}:
2922 cc1: warning: `-g' option not supported on this version of GCC
2923 cc1: warning: `-g1' option not supported on this version of GCC
2924 ./xgcc: Internal compiler error: program as got fatal signal 11
2927 A patched version of the assembler is available as the file
2928 @uref{ftp://altdorf.ai.mit.edu/archive/cph/hpux-8.0-assembler}. If you
2929 have HP software support, the patch can also be obtained directly from
2930 HP, as described in the following note:
2933 This is the patched assembler, to patch SR#1653-010439, where the
2934 assembler aborts on floating point constants.
2936 The bug is not really in the assembler, but in the shared library
2937 version of the function ``cvtnum(3c)''. The bug on ``cvtnum(3c)'' is
2938 SR#4701-078451. Anyway, the attached assembler uses the archive
2939 library version of ``cvtnum(3c)'' and thus does not exhibit the bug.
2942 This patch is also known as PHCO_4484.
2944 In addition gdb does not understand that native HP-UX format, so
2945 you must use gas if you wish to use gdb.
2947 On HP-UX version 8.05, but not on 8.07 or more recent versions, the
2948 @command{fixproto} shell script triggers a bug in the system shell. If you
2949 encounter this problem, upgrade your operating system or use BASH (the
2950 GNU shell) to run @command{fixproto}. This bug will cause the fixproto
2951 program to report an error of the form:
2954 ./fixproto: sh internal 1K buffer overflow
2957 To fix this, you can also change the first line of the fixproto script
2967 @heading @anchor{mips-*-*}mips-*-*
2968 If on a MIPS system you get an error message saying ``does not have gp
2969 sections for all it's [sic] sectons [sic]'', don't worry about it. This
2970 happens whenever you use GAS with the MIPS linker, but there is not
2971 really anything wrong, and it is okay to use the output file. You can
2972 stop such warnings by installing the GNU linker.
2974 It would be nice to extend GAS to produce the gp tables, but they are
2975 optional, and there should not be a warning about their absence.
2977 The libstdc++ atomic locking routines for MIPS targets requires MIPS II
2978 and later. A patch went in just after the GCC 3.3 release to
2979 make @samp{mips*-*-*} use the generic implementation instead. You can also
2980 configure for @samp{mipsel-elf} as a workaround. The
2981 @samp{mips*-*-linux*} target continues to use the MIPS II routines. More
2982 work on this is expected in future releases.
2984 Cross-compilers for the Mips as target using the Mips assembler
2985 currently do not work, because the auxiliary programs
2986 @file{mips-tdump.c} and @file{mips-tfile.c} can't be compiled on
2987 anything but a Mips. It does work to cross compile for a Mips
2988 if you use the GNU assembler and linker.
2993 @heading @anchor{mips-sgi-irix5}mips-sgi-irix5
2995 This configuration has considerable problems, which will be fixed in a
2998 In order to compile GCC on an SGI running IRIX 5, the ``compiler_dev.hdr''
2999 subsystem must be installed from the IDO CD-ROM supplied by Silicon
3000 Graphics. It is also available for download from
3001 @uref{http://www.sgi.com/developers/devtools/apis/ido.html,,http://www.sgi.com/developers/devtools/apis/ido.html}.
3003 @samp{make compare} may fail on version 5 of IRIX unless you add
3004 @option{-save-temps} to @code{CFLAGS}. On these systems, the name of the
3005 assembler input file is stored in the object file, and that makes
3006 comparison fail if it differs between the @code{stage1} and
3007 @code{stage2} compilations. The option @option{-save-temps} forces a
3008 fixed name to be used for the assembler input file, instead of a
3009 randomly chosen name in @file{/tmp}. Do not add @option{-save-temps}
3010 unless the comparisons fail without that option. If you do you
3011 @option{-save-temps}, you will have to manually delete the @samp{.i} and
3012 @samp{.s} files after each series of compilations.
3014 If you use the MIPS C compiler to bootstrap, it may be necessary
3015 to increase its table size for switch statements with the
3016 @option{-Wf,-XNg1500} option. If you use the @option{-O2}
3017 optimization option, you also need to use @option{-Olimit 3000}.
3019 To enable debugging under IRIX 5, you must use GNU @command{as} 2.11.2
3021 and use the @option{--with-gnu-as} configure option when configuring GCC.
3022 GNU @command{as} is distributed as part of the binutils package.
3023 When using release 2.11.2, you need to apply a patch
3024 @uref{http://sources.redhat.com/ml/binutils/2001-07/msg00352.html,,http://sources.redhat.com/ml/binutils/2001-07/msg00352.html}
3025 which will be included in the next release of binutils.
3027 When building GCC, the build process loops rebuilding @command{cc1} over
3028 and over again. This happens on @samp{mips-sgi-irix5.2}, and possibly
3029 other platforms. It has been reported that this is a known bug in the
3030 @command{make} shipped with IRIX 5.2. We recommend you use GNU
3031 @command{make} instead of the vendor supplied @command{make} program;
3032 however, you may have success with @command{smake} on IRIX 5.2 if you do
3033 not have GNU @command{make} available.
3038 @heading @anchor{mips-sgi-irix6}mips-sgi-irix6
3040 If you are using IRIX @command{cc} as your bootstrap compiler, you must
3041 ensure that the N32 ABI is in use. To test this, compile a simple C
3042 file with @command{cc} and then run @command{file} on the
3043 resulting object file. The output should look like:
3046 test.o: ELF N32 MSB @dots{}
3052 test.o: ELF 32-bit MSB @dots{}
3058 test.o: ELF 64-bit MSB @dots{}
3061 then your version of @command{cc} uses the O32 or N64 ABI by default. You
3062 should set the environment variable @env{CC} to @samp{cc -n32}
3063 before configuring GCC@.
3065 If you want the resulting @command{gcc} to run on old 32-bit systems
3066 with the MIPS R4400 CPU, you need to ensure that only code for the mips3
3067 instruction set architecture (ISA) is generated. While GCC 3.x does
3068 this correctly, both GCC 2.95 and SGI's MIPSpro @command{cc} may change
3069 the ISA depending on the machine where GCC is built. Using one of them
3070 as the bootstrap compiler may result in mips4 code, which won't run at
3071 all on mips3-only systems. For the test program above, you should see:
3074 test.o: ELF N32 MSB mips-3 @dots{}
3080 test.o: ELF N32 MSB mips-4 @dots{}
3083 instead, you should set the environment variable @env{CC} to @samp{cc
3084 -n32 -mips3} or @samp{gcc -mips3} respectively before configuring GCC@.
3086 GCC on IRIX 6 is usually built to support both the N32 and N64 ABIs. If
3087 you build GCC on a system that doesn't have the N64 libraries installed,
3088 you need to configure with @option{--disable-multilib} so GCC doesn't
3089 try to use them. Look for @file{/usr/lib64/libc.so.1} to see if you
3090 have the 64-bit libraries installed.
3092 You must @emph{not} use GNU @command{as} (which isn't built anyway as of
3093 binutils 2.11.2) on IRIX 6 platforms; doing so will only cause problems.
3095 GCC does not currently support generating O32 ABI binaries in the
3096 @samp{mips-sgi-irix6} configurations. It is possible to create a GCC
3097 with O32 ABI only support by configuring it for the @samp{mips-sgi-irix5}
3098 target and using a patched GNU @command{as} 2.11.2 as documented in the
3099 @uref{#mips-sgi-irix5,,@samp{mips-sgi-irix5}} section above. Using the
3100 native assembler requires patches to GCC which will be included in a
3101 future release. It is
3102 expected that O32 ABI support will be available again in a future release.
3104 The @option{--enable-threads} option doesn't currently work, a patch is
3105 in preparation for a future release. The @option{--enable-libgcj}
3106 option is disabled by default: IRIX 6 uses a very low default limit
3107 (20480) for the command line length. Although libtool contains a
3108 workaround for this problem, at least the N64 @samp{libgcj} is known not
3109 to build despite this, running into an internal error of the native
3110 @command{ld}. A sure fix is to increase this limit (@samp{ncargs}) to
3111 its maximum of 262144 bytes. If you have root access, you can use the
3112 @command{systune} command to do this.
3114 GCC does not correctly pass/return structures which are
3115 smaller than 16 bytes and which are not 8 bytes. The problem is very
3116 involved and difficult to fix. It affects a number of other targets also,
3117 but IRIX 6 is affected the most, because it is a 64-bit target, and 4 byte
3118 structures are common. The exact problem is that structures are being padded
3119 at the wrong end, e.g.@: a 4 byte structure is loaded into the lower 4 bytes
3120 of the register when it should be loaded into the upper 4 bytes of the
3123 GCC is consistent with itself, but not consistent with the SGI C compiler
3124 (and the SGI supplied runtime libraries), so the only failures that can
3125 happen are when there are library functions that take/return such
3126 structures. There are very few such library functions. Currently this
3127 is known to affect @code{inet_ntoa}, @code{inet_lnaof},
3128 @code{inet_netof}, @code{inet_makeaddr}, and @code{semctl}. Until the
3129 bug is fixed, GCC contains workarounds for the known affected functions.
3131 See @uref{http://freeware.sgi.com/,,http://freeware.sgi.com/} for more
3132 information about using GCC on IRIX platforms.
3137 @heading @anchor{powerpc*-*-*}powerpc-*-*
3139 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
3140 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
3145 @heading @anchor{powerpc-*-darwin*}powerpc-*-darwin*
3146 PowerPC running Darwin (Mac OS X kernel).
3148 Pre-installed versions of Mac OS X may not include any developer tools,
3149 meaning that you will not be able to build GCC from source. Tool
3150 binaries are available at
3151 @uref{http://developer.apple.com/tools/compilers.html} (free
3152 registration required).
3154 The default stack limit of 512K is too small, which may cause compiles
3155 to fail with 'Bus error'. Set the stack larger, for instance
3156 by doing @samp{limit stack 800}. It's a good idea to use the GNU
3157 preprocessor instead of Apple's @file{cpp-precomp} during the first stage of
3158 bootstrapping; this is automatic when doing @samp{make bootstrap}, but
3159 to do it from the toplevel objdir you will need to say @samp{make
3160 CC='cc -no-cpp-precomp' bootstrap}.
3162 The version of GCC shipped by Apple typically includes a number of
3163 extensions not available in a standard GCC release. These extensions
3164 are generally specific to Mac programming.
3169 @heading @anchor{powerpc-*-elf}powerpc-*-elf, powerpc-*-sysv4
3170 PowerPC system in big endian mode, running System V.4.
3175 @heading @anchor{powerpc-*-linux-gnu*}powerpc-*-linux-gnu*
3178 @uref{ftp://ftp.kernel.org/pub/linux/devel/binutils,,binutils 2.13.90.0.10}
3179 or newer for a working GCC@.
3184 @heading @anchor{powerpc-*-netbsd*}powerpc-*-netbsd*
3185 PowerPC system in big endian mode running NetBSD@. To build the
3186 documentation you will need Texinfo version 4.2 (NetBSD 1.5.1 included
3187 Texinfo version 3.12).
3192 @heading @anchor{powerpc-*-eabisim}powerpc-*-eabisim
3193 Embedded PowerPC system in big endian mode for use in running under the
3199 @heading @anchor{powerpc-*-eabi}powerpc-*-eabi
3200 Embedded PowerPC system in big endian mode.
3205 @heading @anchor{powerpcle-*-elf}powerpcle-*-elf, powerpcle-*-sysv4
3206 PowerPC system in little endian mode, running System V.4.
3211 @heading @anchor{powerpcle-*-eabisim}powerpcle-*-eabisim
3212 Embedded PowerPC system in little endian mode for use in running under
3218 @heading @anchor{powerpcle-*-eabi}powerpcle-*-eabi
3219 Embedded PowerPC system in little endian mode.
3224 @heading @anchor{s390-*-linux*}s390-*-linux*
3225 S/390 system running GNU/Linux for S/390@.
3230 @heading @anchor{s390x-*-linux*}s390x-*-linux*
3231 zSeries system (64-bit) running GNU/Linux for zSeries@.
3236 @heading @anchor{s390x-ibm-tpf*}s390x-ibm-tpf*
3237 zSeries system (64-bit) running TPF. This platform is
3238 supported as cross-compilation target only.
3243 @c Please use Solaris 2 to refer to all release of Solaris, starting
3244 @c with 2.0 until 2.6, 7, and 8. Solaris 1 was a marketing name for
3245 @c SunOS 4 releases which we don't use to avoid confusion. Solaris
3246 @c alone is too unspecific and must be avoided.
3247 @heading @anchor{*-*-solaris2*}*-*-solaris2*
3249 Sun does not ship a C compiler with Solaris 2. To bootstrap and install
3250 GCC you first have to install a pre-built compiler, see our
3251 @uref{binaries.html,,binaries page} for details.
3253 The Solaris 2 @command{/bin/sh} will often fail to configure
3254 @file{libstdc++-v3}, @file{boehm-gc} or @file{libjava}. We therefore
3255 recommend to use the following sequence of commands to bootstrap and
3259 % CONFIG_SHELL=/bin/ksh
3260 % export CONFIG_SHELL
3263 and then proceed as described in @uref{build.html,,the build instructions},
3264 where we strongly recommend using GNU make and specifying an absolute path
3265 to invoke @var{srcdir}/configure.
3267 Solaris 2 comes with a number of optional OS packages. Some of these
3268 are needed to use GCC fully, namely @code{SUNWarc},
3269 @code{SUNWbtool}, @code{SUNWesu}, @code{SUNWhea}, @code{SUNWlibm},
3270 @code{SUNWsprot}, and @code{SUNWtoo}. If you did not install all
3271 optional packages when installing Solaris 2, you will need to verify that
3272 the packages that GCC needs are installed.
3274 To check whether an optional package is installed, use
3275 the @command{pkginfo} command. To add an optional package, use the
3276 @command{pkgadd} command. For further details, see the Solaris 2
3279 Trying to use the linker and other tools in
3280 @file{/usr/ucb} to install GCC has been observed to cause trouble.
3281 For example, the linker may hang indefinitely. The fix is to remove
3282 @file{/usr/ucb} from your @env{PATH}.
3284 The build process works more smoothly with the legacy Sun tools so, if you
3285 have @file{/usr/xpg4/bin} in your @env{PATH}, we recommend that you place
3286 @file{/usr/bin} before @file{/usr/xpg4/bin} for the duration of the build.
3288 All releases of GNU binutils prior to 2.11.2 have known bugs on this
3289 platform. We recommend the use of GNU binutils 2.11.2 or the vendor
3290 tools (Sun @command{as}, Sun @command{ld}).
3292 Sun bug 4296832 turns up when compiling X11 headers with GCC 2.95 or
3293 newer: @command{g++} will complain that types are missing. These headers assume
3294 that omitting the type means @code{int}; this assumption worked for C89 but
3295 is wrong for C++, and is now wrong for C99 also.
3297 @command{g++} accepts such (invalid) constructs with the option
3298 @option{-fpermissive}; it
3299 will assume that any missing type is @code{int} (as defined by C89).
3301 There are patches for Solaris 2.6 (105633-56 or newer for SPARC,
3302 106248-42 or newer for Intel), Solaris 7 (108376-21 or newer for SPARC,
3303 108377-20 for Intel), and Solaris 8 (108652-24 or newer for SPARC,
3304 108653-22 for Intel) that fix this bug.
3309 @heading @anchor{sparc-sun-solaris2*}sparc-sun-solaris2*
3311 When GCC is configured to use binutils 2.11.2 or later the binaries
3312 produced are smaller than the ones produced using Sun's native tools;
3313 this difference is quite significant for binaries containing debugging
3316 Sun @command{as} 4.x is broken in that it cannot cope with long symbol names.
3317 A typical error message might look similar to the following:
3320 /usr/ccs/bin/as: "/var/tmp/ccMsw135.s", line 11041: error:
3321 can't compute value of an expression involving an external symbol.
3324 This is Sun bug 4237974. This is fixed with patch 108908-02 for Solaris
3325 2.6 and has been fixed in later (5.x) versions of the assembler,
3326 starting with Solaris 7.
3328 Starting with Solaris 7, the operating system is capable of executing
3329 64-bit SPARC V9 binaries. GCC 3.1 and later properly supports
3330 this; the @option{-m64} option enables 64-bit code generation.
3331 However, if all you want is code tuned for the UltraSPARC CPU, you
3332 should try the @option{-mtune=ultrasparc} option instead, which produces
3333 code that, unlike full 64-bit code, can still run on non-UltraSPARC
3336 When configuring on a Solaris 7 or later system that is running a kernel
3337 that supports only 32-bit binaries, one must configure with
3338 @option{--disable-multilib}, since we will not be able to build the
3339 64-bit target libraries.
3344 @heading @anchor{sparc-sun-solaris2.7}sparc-sun-solaris2.7
3346 Sun patch 107058-01 (1999-01-13) for Solaris 7/SPARC triggers a bug in
3347 the dynamic linker. This problem (Sun bug 4210064) affects GCC 2.8
3348 and later, including all EGCS releases. Sun formerly recommended
3349 107058-01 for all Solaris 7 users, but around 1999-09-01 it started to
3350 recommend it only for people who use Sun's compilers.
3352 Here are some workarounds to this problem:
3355 Do not install Sun patch 107058-01 until after Sun releases a
3356 complete patch for bug 4210064. This is the simplest course to take,
3357 unless you must also use Sun's C compiler. Unfortunately 107058-01
3358 is preinstalled on some new Solaris 7-based hosts, so you may have to
3362 Copy the original, unpatched Solaris 7
3363 @command{/usr/ccs/bin/as} into
3364 @command{/usr/local/libexec/gcc/sparc-sun-solaris2.7/3.4/as},
3365 adjusting the latter name to fit your local conventions and software
3369 Install Sun patch 106950-03 (1999-05-25) or later. Nobody with
3370 both 107058-01 and 106950-03 installed has reported the bug with GCC
3371 and Sun's dynamic linker. This last course of action is riskiest,
3372 for two reasons. First, you must install 106950 on all hosts that
3373 run code generated by GCC; it doesn't suffice to install it only on
3374 the hosts that run GCC itself. Second, Sun says that 106950-03 is
3375 only a partial fix for bug 4210064, but Sun doesn't know whether the
3376 partial fix is adequate for GCC@. Revision -08 or later should fix
3377 the bug. The current (as of 2001-09-24) revision is -14, and is included in
3378 the Solaris 7 Recommended Patch Cluster.
3381 GCC 3.3 triggers a bug in version 5.0 Alpha 03/27/98 of the Sun assembler,
3382 which causes a bootstrap failure when linking the 64-bit shared version of
3383 libgcc. A typical error message is:
3386 ld: fatal: relocation error: R_SPARC_32: file libgcc/sparcv9/_muldi3.o:
3387 symbol <unknown>: offset 0xffffffff7ec133e7 is non-aligned.
3390 This bug has been fixed in the final 5.0 version of the assembler.
3395 @heading @anchor{sparc-*-linux*}sparc-*-linux*
3397 GCC versions 3.0 and higher require binutils 2.11.2 and glibc 2.2.4
3398 or newer on this platform. All earlier binutils and glibc
3399 releases mishandled unaligned relocations on @code{sparc-*-*} targets.
3405 @heading @anchor{sparc64-*-solaris2*}sparc64-*-solaris2*
3407 The following compiler flags must be specified in the configure
3408 step in order to bootstrap this target with the Sun compiler:
3411 % CC="cc -xildoff -xarch=v9" @var{srcdir}/configure [@var{options}] [@var{target}]
3414 @option{-xildoff} turns off the incremental linker, and @option{-xarch=v9}
3415 specifies the SPARC-V9 architecture to the Sun linker and assembler.
3420 @heading @anchor{sparcv9-*-solaris2*}sparcv9-*-solaris2*
3422 This is a synonym for sparc64-*-solaris2*.
3427 @heading @anchor{#*-*-sysv*}*-*-sysv*
3428 On System V release 3, you may get this error message
3432 ld fatal: failed to write symbol name @var{something}
3433 in strings table for file @var{whatever}
3436 This probably indicates that the disk is full or your ulimit won't allow
3437 the file to be as large as it needs to be.
3439 This problem can also result because the kernel parameter @code{MAXUMEM}
3440 is too small. If so, you must regenerate the kernel and make the value
3441 much larger. The default value is reported to be 1024; a value of 32768
3442 is said to work. Smaller values may also work.
3444 On System V, if you get an error like this,
3447 /usr/local/lib/bison.simple: In function `yyparse':
3448 /usr/local/lib/bison.simple:625: virtual memory exhausted
3452 that too indicates a problem with disk space, ulimit, or @code{MAXUMEM}.
3454 On a System V release 4 system, make sure @file{/usr/bin} precedes
3455 @file{/usr/ucb} in @code{PATH}. The @command{cc} command in
3456 @file{/usr/ucb} uses libraries which have bugs.
3461 @heading @anchor{vax-dec-ultrix}vax-dec-ultrix
3462 Don't try compiling with VAX C (@command{vcc}). It produces incorrect code
3463 in some cases (for example, when @code{alloca} is used).
3468 @heading @anchor{*-*-vxworks*}*-*-vxworks*
3469 Support for VxWorks is in flux. At present GCC supports @emph{only} the
3470 very recent VxWorks 5.5 (aka Tornado 2.2) release, and only on PowerPC.
3471 We welcome patches for other architectures supported by VxWorks 5.5.
3472 Support for VxWorks AE would also be welcome; we believe this is merely
3473 a matter of writing an appropriate ``configlette'' (see below). We are
3474 not interested in supporting older, a.out or COFF-based, versions of
3477 VxWorks comes with an older version of GCC installed in
3478 @file{@var{$WIND_BASE}/host}; we recommend you do not overwrite it.
3479 Choose an installation @var{prefix} entirely outside @var{$WIND_BASE}.
3480 Before running @command{configure}, create the directories @file{@var{prefix}}
3481 and @file{@var{prefix}/bin}. Link or copy the appropriate assembler,
3482 linker, etc. into @file{@var{prefix}/bin}, and set your @var{PATH} to
3483 include that directory while running both @command{configure} and
3486 You must give @command{configure} the
3487 @option{--with-headers=@var{$WIND_BASE}/target/h} switch so that it can
3488 find the VxWorks system headers. Since VxWorks is a cross compilation
3489 target only, you must also specify @option{--target=@var{target}}.
3490 @command{configure} will attempt to create the directory
3491 @file{@var{prefix}/@var{target}/sys-include} and copy files into it;
3492 make sure the user running @command{configure} has sufficient privilege
3495 GCC's exception handling runtime requires a special ``configlette''
3496 module, @file{contrib/gthr_supp_vxw_5x.c}. Follow the instructions in
3497 that file to add the module to your kernel build. (Future versions of
3498 VxWorks will incorporate this module.)
3503 @heading @anchor{x86_64-*-*}x86_64-*-*, amd64-*-*
3505 GCC supports the x86-64 architecture implemented by the AMD64 processor
3506 (amd64-*-* is an alias for x86_64-*-*) on GNU/Linux, FreeBSD and NetBSD.
3507 On GNU/Linux the default is a bi-arch compiler which is able to generate
3508 both 64-bit x86-64 and 32-bit x86 code (via the @option{-m32} switch).
3513 @heading @anchor{xtensa-*-elf}xtensa-*-elf
3515 This target is intended for embedded Xtensa systems using the
3516 @samp{newlib} C library. It uses ELF but does not support shared
3517 objects. Designed-defined instructions specified via the
3518 Tensilica Instruction Extension (TIE) language are only supported
3519 through inline assembly.
3521 The Xtensa configuration information must be specified prior to
3522 building GCC@. The @file{include/xtensa-config.h} header
3523 file contains the configuration information. If you created your
3524 own Xtensa configuration with the Xtensa Processor Generator, the
3525 downloaded files include a customized copy of this header file,
3526 which you can use to replace the default header file.
3531 @heading @anchor{xtensa-*-linux*}xtensa-*-linux*
3533 This target is for Xtensa systems running GNU/Linux. It supports ELF
3534 shared objects and the GNU C library (glibc). It also generates
3535 position-independent code (PIC) regardless of whether the
3536 @option{-fpic} or @option{-fPIC} options are used. In other
3537 respects, this target is the same as the
3538 @uref{#xtensa-*-elf,,@samp{xtensa-*-elf}} target.
3543 @heading @anchor{windows}Microsoft Windows (32-bit)
3545 A port of GCC 2.95.2 and 3.x is included with the
3546 @uref{http://www.cygwin.com/,,Cygwin environment}.
3548 Current (as of early 2001) snapshots of GCC will build under Cygwin
3549 without modification.
3551 GCC does not currently build with Microsoft's C++ compiler and there
3552 are no plans to make it do so.
3557 @heading @anchor{os2}OS/2
3559 GCC does not currently support OS/2. However, Andrew Zabolotny has been
3560 working on a generic OS/2 port with pgcc. The current code can be found
3561 at @uref{http://www.goof.com/pcg/os2/,,http://www.goof.com/pcg/os2/}.
3563 An older copy of GCC 2.8.1 is included with the EMX tools available at
3564 @uref{ftp://ftp.leo.org/pub/comp/os/os2/leo/devtools/emx+gcc/,,
3565 ftp://ftp.leo.org/pub/comp/os/os2/leo/devtools/emx+gcc/}.
3570 @heading @anchor{older}Older systems
3572 GCC contains support files for many older (1980s and early
3573 1990s) Unix variants. For the most part, support for these systems
3574 has not been deliberately removed, but it has not been maintained for
3575 several years and may suffer from bitrot.
3577 Starting with GCC 3.1, each release has a list of ``obsoleted'' systems.
3578 Support for these systems is still present in that release, but
3579 @command{configure} will fail unless the @option{--enable-obsolete}
3580 option is given. Unless a maintainer steps forward, support for these
3581 systems will be removed from the next release of GCC@.
3583 Support for old systems as hosts for GCC can cause problems if the
3584 workarounds for compiler, library and operating system bugs affect the
3585 cleanliness or maintainability of the rest of GCC@. In some cases, to
3586 bring GCC up on such a system, if still possible with current GCC, may
3587 require first installing an old version of GCC which did work on that
3588 system, and using it to compile a more recent GCC, to avoid bugs in the
3589 vendor compiler. Old releases of GCC 1 and GCC 2 are available in the
3590 @file{old-releases} directory on the @uref{../mirrors.html,,GCC mirror
3591 sites}. Header bugs may generally be avoided using
3592 @command{fixincludes}, but bugs or deficiencies in libraries and the
3593 operating system may still cause problems.
3595 Support for older systems as targets for cross-compilation is less
3596 problematic than support for them as hosts for GCC; if an enthusiast
3597 wishes to make such a target work again (including resurrecting any of
3598 the targets that never worked with GCC 2, starting from the last CVS
3599 version before they were removed), patches
3600 @uref{../contribute.html,,following the usual requirements} would be
3601 likely to be accepted, since they should not affect the support for more
3604 For some systems, old versions of GNU binutils may also be useful,
3605 and are available from @file{pub/binutils/old-releases} on
3606 @uref{http://sources.redhat.com/mirrors.html,,sources.redhat.com mirror sites}.
3608 Some of the information on specific systems above relates to
3609 such older systems, but much of the information
3610 about GCC on such systems (which may no longer be applicable to
3611 current GCC) is to be found in the GCC texinfo manual.
3616 @heading @anchor{elf_targets}all ELF targets (SVR4, Solaris 2, etc.)
3618 C++ support is significantly better on ELF targets if you use the
3619 @uref{./configure.html#with-gnu-ld,,GNU linker}; duplicate copies of
3620 inlines, vtables and template instantiations will be discarded
3629 @uref{./index.html,,Return to the GCC Installation page}
3633 @c ***Old documentation******************************************************
3635 @include install-old.texi
3641 @uref{./index.html,,Return to the GCC Installation page}
3645 @c ***GFDL********************************************************************
3653 @uref{./index.html,,Return to the GCC Installation page}
3657 @c ***************************************************************************
3658 @c Part 6 The End of the Document
3660 @comment node-name, next, previous, up
3661 @node Concept Index, , GNU Free Documentation License, Top
3665 @unnumbered Concept Index