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 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 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}.
261 Necessary in some circumstances, optional in others. See the
262 host/target specific instructions for your platform for the exact
265 @item gzip version 1.2.4 (or later) or
266 @itemx bzip2 version 1.0.2 (or later)
268 Necessary to uncompress GCC @command{tar} files when source code is
269 obtained via FTP mirror sites.
271 @item GNU tar version 1.12 (or later)
273 Necessary (only on some platforms) to untar the source code. Many
274 systems' @command{tar} programs will also work, only try GNU
275 @command{tar} if you have problems.
280 @heading Tools/packages necessary for modifying GCC
283 @item autoconf version 2.13 (NO earlier or later versions) and
284 @itemx GNU m4 version 1.4 (or later)
286 Necessary when modifying @file{configure.in}, @file{aclocal.m4}, etc.@:
287 to regenerate @file{configure} and @file{config.in} files
289 @item automake version ???
291 Necessary when modifying a @file{Makefile.am} file to regenerate its
292 associated @file{Makefile.in}
294 @item gperf version 2.7.2 (or later)
296 Necessary when modifying @command{gperf} input files, e.g.@:
297 @file{gcc/cp/cfns.gperf} to regenerate its associated header file, e.g.@:
298 @file{gcc/cp/cfns.h}.
300 @item expect version ???
301 @itemx tcl version ???
302 @itemx dejagnu version ???
304 Necessary to run the GCC testsuite.
306 @item autogen version 5.5.4 (or later) and
307 @itemx guile version 1.4.1 (or later)
309 Necessary to regenerate @file{fixinc/fixincl.x} from
310 @file{fixinc/inclhack.def} and @file{fixinc/*.tpl}.
312 Necessary to run the @file{fixinc} @command{make check}.
314 Necessary to regenerate the top level @file{Makefile.am} files from
315 @file{Makefile.tpl} and @file{Makefile.def}.
317 @item GNU Bison version 1.28 (or later)
318 Berkeley @command{yacc} (@command{byacc}) is also reported to work other
321 Necessary when modifying @file{*.y} files.
323 Necessary to build GCC during development because the generated output
324 files are not included in the CVS repository. They are included in
327 @item Flex version 2.5.4 (or later)
329 Necessary when modifying @file{*.l} files.
331 Necessary to build GCC during development because the generated output
332 files are not included in the CVS repository. They are included in
335 @item Texinfo version 4.2 (or later)
337 Necessary for running @command{makeinfo} when modifying @file{*.texi}
338 files to test your changes.
340 Necessary to build GCC documentation during development because the
341 generated output files are not included in the CVS repository. They are
342 included in releases.
344 @item @TeX{} (any working version)
346 Necessary for running @command{texi2dvi}, used when running
347 @command{make dvi} to create DVI files.
349 @item cvs version 1.10 (or later)
350 @itemx ssh (any version)
352 Necessary to access the CVS repository. Public releases and weekly
353 snapshots of the development sources are also available via FTP.
355 @item perl version 5.6.1 (or later)
357 Necessary when regenerating @file{Makefile} dependencies in libiberty.
358 Necessary when regenerating something with intl??? (pod2man???)
361 @item GNU diffutils version 2.7 (or later)
363 Necessary when creating changes to GCC source code to submit for review.
365 @item patch version 2.5.4 (or later)
367 Necessary when applying patches, created with @command{diff}, to one's
377 @uref{./index.html,,Return to the GCC Installation page}
381 @c ***Downloading the source**************************************************
383 @comment node-name, next, previous, up
384 @node Downloading the source, Configuration, Prerequisites, Installing GCC
388 @chapter Downloading GCC
390 @cindex Downloading GCC
391 @cindex Downloading the Source
393 GCC is distributed via @uref{http://gcc.gnu.org/cvs.html,,CVS} and FTP
394 tarballs compressed with @command{gzip} or
395 @command{bzip2}. It is possible to download a full distribution or specific
398 Please refer to our @uref{http://gcc.gnu.org/releases.html,,releases web page}
399 for information on how to obtain GCC@.
401 The full distribution includes the C, C++, Objective-C, Fortran, Java,
402 and Ada (in case of GCC 3.1 and later) compilers. The full distribution
403 also includes runtime libraries for C++, Objective-C, Fortran, and Java.
404 In GCC 3.0 and later versions, GNU compiler testsuites are also included
405 in the full distribution.
407 If you choose to download specific components, you must download the core
408 GCC distribution plus any language specific distributions you wish to
409 use. The core distribution includes the C language front end as well as the
410 shared components. Each language has a tarball which includes the language
411 front end as well as the language runtime (when appropriate).
413 Unpack the core distribution as well as any language specific
414 distributions in the same directory.
416 If you also intend to build binutils (either to upgrade an existing
417 installation or for use in place of the corresponding tools of your
418 OS), unpack the binutils distribution either in the same directory or
419 a separate one. In the latter case, add symbolic links to any
420 components of the binutils you intend to build alongside the compiler
421 (@file{bfd}, @file{binutils}, @file{gas}, @file{gprof}, @file{ld},
422 @file{opcodes}, @dots{}) to the directory containing the GCC sources.
429 @uref{./index.html,,Return to the GCC Installation page}
433 @c ***Configuration***********************************************************
435 @comment node-name, next, previous, up
436 @node Configuration, Building, Downloading the source, Installing GCC
440 @chapter Installing GCC: Configuration
442 @cindex Configuration
443 @cindex Installing GCC: Configuration
445 Like most GNU software, GCC must be configured before it can be built.
446 This document describes the recommended configuration procedure
447 for both native and cross targets.
449 We use @var{srcdir} to refer to the toplevel source directory for
450 GCC; we use @var{objdir} to refer to the toplevel build/object directory.
452 If you obtained the sources via CVS, @var{srcdir} must refer to the top
453 @file{gcc} directory, the one where the @file{MAINTAINERS} can be found,
454 and not its @file{gcc} subdirectory, otherwise the build will fail.
456 If either @var{srcdir} or @var{objdir} is located on an automounted NFS
457 file system, the shell's built-in @command{pwd} command will return
458 temporary pathnames. Using these can lead to various sorts of build
459 problems. To avoid this issue, set the @env{PWDCMD} environment
460 variable to an automounter-aware @command{pwd} command, e.g.,
461 @command{pawd} or @samp{amq -w}, during the configuration and build
464 First, we @strong{highly} recommend that GCC be built into a
465 separate directory than the sources which does @strong{not} reside
466 within the source tree. This is how we generally build GCC; building
467 where @var{srcdir} == @var{objdir} should still work, but doesn't
468 get extensive testing; building where @var{objdir} is a subdirectory
469 of @var{srcdir} is unsupported.
471 If you have previously built GCC in the same directory for a
472 different target machine, do @samp{make distclean} to delete all files
473 that might be invalid. One of the files this deletes is @file{Makefile};
474 if @samp{make distclean} complains that @file{Makefile} does not exist
475 or issues a message like ``don't know how to make distclean'' it probably
476 means that the directory is already suitably clean. However, with the
477 recommended method of building in a separate @var{objdir}, you should
478 simply use a different @var{objdir} for each target.
480 Second, when configuring a native system, either @command{cc} or
481 @command{gcc} must be in your path or you must set @env{CC} in
482 your environment before running configure. Otherwise the configuration
485 Note that the bootstrap compiler and the resulting GCC must be link
486 compatible, else the bootstrap will fail with linker errors about
487 incompatible object file formats. Several multilibed targets are
488 affected by this requirement, see
490 @ref{Specific, host/target specific installation notes}.
493 @uref{specific.html,,host/target specific installation notes}.
501 % @var{srcdir}/configure [@var{options}] [@var{target}]
505 @heading Target specification
508 GCC has code to correctly determine the correct value for @var{target}
509 for nearly all native systems. Therefore, we highly recommend you not
510 provide a configure target when configuring a native compiler.
513 @var{target} must be specified as @option{--target=@var{target}}
514 when configuring a cross compiler; examples of valid targets would be
515 i960-rtems, m68k-coff, sh-elf, etc.
518 Specifying just @var{target} instead of @option{--target=@var{target}}
519 implies that the host defaults to @var{target}.
523 @heading Options specification
525 Use @var{options} to override several configure time options for
526 GCC@. A list of supported @var{options} follows; @samp{configure
527 --help} may list other options, but those not listed below may not
528 work and should not normally be used.
531 @item --prefix=@var{dirname}
532 Specify the toplevel installation
533 directory. This is the recommended way to install the tools into a directory
534 other than the default. The toplevel installation directory defaults to
537 We @strong{highly} recommend against @var{dirname} being the same or a
538 subdirectory of @var{objdir} or vice versa. If specifying a directory
539 beneath a user's home directory tree, some shells will not expand
540 @var{dirname} correctly if it contains the @samp{~} metacharacter; use
543 These additional options control where certain parts of the distribution
544 are installed. Normally you should not need to use these options.
546 @item --exec-prefix=@var{dirname}
547 Specify the toplevel installation directory for architecture-dependent
548 files. The default is @file{@var{prefix}}.
550 @item --bindir=@var{dirname}
551 Specify the installation directory for the executables called by users
552 (such as @command{gcc} and @command{g++}). The default is
553 @file{@var{exec-prefix}/bin}.
555 @item --libdir=@var{dirname}
556 Specify the installation directory for object code libraries and
557 internal parts of GCC@. The default is @file{@var{exec-prefix}/lib}.
559 @item --with-slibdir=@var{dirname}
560 Specify the installation directory for the shared libgcc library. The
561 default is @file{@var{libdir}}.
563 @item --infodir=@var{dirname}
564 Specify the installation directory for documentation in info format.
565 The default is @file{@var{prefix}/info}.
567 @item --datadir=@var{dirname}
568 Specify the installation directory for some architecture-independent
569 data files referenced by GCC@. The default is @file{@var{prefix}/share}.
571 @item --mandir=@var{dirname}
572 Specify the installation directory for manual pages. The default is
573 @file{@var{prefix}/man}. (Note that the manual pages are only extracts from
574 the full GCC manuals, which are provided in Texinfo format. The manpages
575 are derived by an automatic conversion process from parts of the full
578 @item --with-gxx-include-dir=@var{dirname}
580 the installation directory for G++ header files. The default is
581 @file{@var{prefix}/include/g++-v3}.
585 @item --program-prefix=@var{prefix}
586 GCC supports some transformations of the names of its programs when
587 installing them. This option prepends @var{prefix} to the names of
588 programs to install in @var{bindir} (see above). For example, specifying
589 @option{--program-prefix=foo-} would result in @samp{gcc}
590 being installed as @file{/usr/local/bin/foo-gcc}.
592 @item --program-suffix=@var{suffix}
593 Appends @var{suffix} to the names of programs to install in @var{bindir}
594 (see above). For example, specifying @option{--program-suffix=-3.1}
595 would result in @samp{gcc} being installed as
596 @file{/usr/local/bin/gcc-3.1}.
598 @item --program-transform-name=@var{pattern}
599 Applies the @samp{sed} script @var{pattern} to be applied to the names
600 of programs to install in @var{bindir} (see above). @var{pattern} has to
601 consist of one or more basic @samp{sed} editing commands, separated by
602 semicolons. For example, if you want the @samp{gcc} program name to be
603 transformed to the installed program @file{/usr/local/bin/myowngcc} and
604 the @samp{g++} program name to be transformed to
605 @file{/usr/local/bin/gspecial++} without changing other program names,
606 you could use the pattern
607 @option{--program-transform-name='s/^gcc$/myowngcc/; s/^g++$/gspecial++/'}
608 to achieve this effect.
610 All three options can be combined and used together, resulting in more
611 complex conversion patterns. As a basic rule, @var{prefix} (and
612 @var{suffix}) are prepended (appended) before further transformations
613 can happen with a special transformation script @var{pattern}.
615 As currently implemented, this option only takes effect for native
616 builds; cross compiler binaries' names are not transformed even when a
617 transformation is explicitly asked for by one of these options.
619 For native builds, some of the installed programs are also installed
620 with the target alias in front of their name, as in
621 @samp{i686-pc-linux-gnu-gcc}. All of the above transformations happen
622 before the target alias is prepended to the name - so, specifying
623 @option{--program-prefix=foo-} and @option{program-suffix=-3.1}, the
624 resulting binary would be installed as
625 @file{/usr/local/bin/i686-pc-linux-gnu-foo-gcc-3.1}.
627 As a last shortcoming, none of the installed Ada programs are
628 transformed yet, which will be fixed in some time.
630 @item --with-local-prefix=@var{dirname}
632 installation directory for local include files. The default is
633 @file{/usr/local}. Specify this option if you want the compiler to
634 search directory @file{@var{dirname}/include} for locally installed
635 header files @emph{instead} of @file{/usr/local/include}.
637 You should specify @option{--with-local-prefix} @strong{only} if your
638 site has a different convention (not @file{/usr/local}) for where to put
641 The default value for @option{--with-local-prefix} is @file{/usr/local}
642 regardless of the value of @option{--prefix}. Specifying
643 @option{--prefix} has no effect on which directory GCC searches for
644 local header files. This may seem counterintuitive, but actually it is
647 The purpose of @option{--prefix} is to specify where to @emph{install
648 GCC}. The local header files in @file{/usr/local/include}---if you put
649 any in that directory---are not part of GCC@. They are part of other
650 programs---perhaps many others. (GCC installs its own header files in
651 another directory which is based on the @option{--prefix} value.)
653 Both the local-prefix include directory and the GCC-prefix include
654 directory are part of GCC's "system include" directories. Although these
655 two directories are not fixed, they need to be searched in the proper
656 order for the correct processing of the include_next directive. The
657 local-prefix include directory is searched before the GCC-prefix
658 include directory. Another characteristic of system include directories
659 is that pedantic warnings are turned off for headers in these directories.
661 Some autoconf macros add @option{-I @var{directory}} options to the
662 compiler command line, to ensure that directories containing installed
663 packages' headers are searched. When @var{directory} is one of GCC's
664 system include directories, GCC will ignore the option so that system
665 directories continue to be processed in the correct order. This
666 may result in a search order different from what was specified but the
667 directory will still be searched.
669 GCC automatically searches for ordinary libraries using
670 @env{GCC_EXEC_PREFIX}. Thus, when the same installation prefix is
671 used for both GCC and packages, GCC will automatically search for
672 both headers and libraries. This provides a configuration that is
673 easy to use. GCC behaves in a manner similar to that when it is
674 installed as a system compiler in @file{/usr}.
676 Sites that need to install multiple versions of GCC may not want to
677 use the above simple configuration. It is possible to use the
678 @option{--program-prefix}, @option{--program-suffix} and
679 @option{--program-transform-name} options to install multiple versions
680 into a single directory, but it may be simpler to use different prefixes
681 and the @option{--with-local-prefix} option to specify the location of the
682 site-specific files for each version. It will then be necessary for
683 users to specify explicitly the location of local site libraries
684 (e.g., with @env{LIBRARY_PATH}).
686 The same value can be used for both @option{--with-local-prefix} and
687 @option{--prefix} provided it is not @file{/usr}. This can be used
688 to avoid the default search of @file{/usr/local/include}.
690 @strong{Do not} specify @file{/usr} as the @option{--with-local-prefix}!
691 The directory you use for @option{--with-local-prefix} @strong{must not}
692 contain any of the system's standard header files. If it did contain
693 them, certain programs would be miscompiled (including GNU Emacs, on
694 certain targets), because this would override and nullify the header
695 file corrections made by the @command{fixincludes} script.
697 Indications are that people who use this option use it based on mistaken
698 ideas of what it is for. People use it as if it specified where to
699 install part of GCC@. Perhaps they make this assumption because
700 installing GCC creates the directory.
702 @item --enable-shared[=@var{package}[,@dots{}]]
703 Build shared versions of libraries, if shared libraries are supported on
704 the target platform. Unlike GCC 2.95.x and earlier, shared libraries
705 are enabled by default on all platforms that support shared libraries,
706 except for @samp{libobjc} which is built as a static library only by
709 If a list of packages is given as an argument, build shared libraries
710 only for the listed packages. For other packages, only static libraries
711 will be built. Package names currently recognized in the GCC tree are
712 @samp{libgcc} (also known as @samp{gcc}), @samp{libstdc++} (not
713 @samp{libstdc++-v3}), @samp{libffi}, @samp{zlib}, @samp{boehm-gc} and
714 @samp{libjava}. Note that @samp{libobjc} does not recognize itself by
715 any name, so, if you list package names in @option{--enable-shared},
716 you will only get static Objective-C libraries. @samp{libf2c} and
717 @samp{libiberty} do not support shared libraries at all.
719 Use @option{--disable-shared} to build only static libraries. Note that
720 @option{--disable-shared} does not accept a list of package names as
721 argument, only @option{--enable-shared} does.
723 @item @anchor{with-gnu-as}--with-gnu-as
724 Specify that the compiler should assume that the
725 assembler it finds is the GNU assembler. However, this does not modify
726 the rules to find an assembler and will result in confusion if the
727 assembler found is not actually the GNU assembler. (Confusion may also
728 result if the compiler finds the GNU assembler but has not been
729 configured with @option{--with-gnu-as}.) If you have more than one
730 assembler installed on your system, you may want to use this option in
731 connection with @option{--with-as=@var{pathname}}.
733 The following systems are the only ones where it makes a difference
734 whether you use the GNU assembler. On any other system,
735 @option{--with-gnu-as} has no effect.
738 @item @samp{hppa1.0-@var{any}-@var{any}}
739 @item @samp{hppa1.1-@var{any}-@var{any}}
740 @item @samp{i386-@var{any}-sysv}
741 @item @samp{m68k-bull-sysv}
742 @item @samp{m68k-hp-hpux}
743 @item @samp{m68000-hp-hpux}
744 @item @samp{m68000-att-sysv}
745 @item @samp{@var{any}-lynx-lynxos}
746 @item @samp{mips-@var{any}}
747 @item @samp{sparc-sun-solaris2.@var{any}}
748 @item @samp{sparc64-@var{any}-solaris2.@var{any}}
751 On the systems listed above (except for the HP-PA, the SPARC, for ISC on
752 the 386, and for @samp{mips-sgi-irix5.*}), if you use the GNU assembler,
753 you should also use the GNU linker (and specify @option{--with-gnu-ld}).
755 @item @anchor{with-as}--with-as=@var{pathname}
757 compiler should use the assembler pointed to by @var{pathname}, rather
758 than the one found by the standard rules to find an assembler, which
763 @file{@var{exec_prefix}/lib/gcc-lib/@var{target}/@var{version}}
764 directory, where @var{exec_prefix} defaults to @var{prefix} which
765 defaults to @file{/usr/local} unless overridden by the
766 @option{--prefix=@var{pathname}} switch described above. @var{target} is the
767 target system triple, such as @samp{sparc-sun-solaris2.7}, and
768 @var{version} denotes the GCC version, such as 3.0.
770 Check operating system specific directories (e.g.@: @file{/usr/ccs/bin} on
773 Note that these rules do not check for the value of @env{PATH}. You may
774 want to use @option{--with-as} if no assembler is installed in the
775 directories listed above, or if you have multiple assemblers installed
776 and want to choose one that is not found by the above rules.
778 @item @anchor{with-gnu-ld}--with-gnu-ld
779 Same as @uref{#with-gnu-as,,@option{--with-gnu-as}}
782 @item --with-ld=@var{pathname}
783 Same as @uref{#with-as,,@option{--with-as}}
787 Specify that stabs debugging
788 information should be used instead of whatever format the host normally
789 uses. Normally GCC uses the same debug format as the host system.
791 On MIPS based systems and on Alphas, you must specify whether you want
792 GCC to create the normal ECOFF debugging format, or to use BSD-style
793 stabs passed through the ECOFF symbol table. The normal ECOFF debug
794 format cannot fully handle languages other than C@. BSD stabs format can
795 handle other languages, but it only works with the GNU debugger GDB@.
797 Normally, GCC uses the ECOFF debugging format by default; if you
798 prefer BSD stabs, specify @option{--with-stabs} when you configure GCC@.
800 No matter which default you choose when you configure GCC, the user
801 can use the @option{-gcoff} and @option{-gstabs+} options to specify explicitly
802 the debug format for a particular compilation.
804 @option{--with-stabs} is meaningful on the ISC system on the 386, also, if
805 @option{--with-gas} is used. It selects use of stabs debugging
806 information embedded in COFF output. This kind of debugging information
807 supports C++ well; ordinary COFF debugging information does not.
809 @option{--with-stabs} is also meaningful on 386 systems running SVR4. It
810 selects use of stabs debugging information embedded in ELF output. The
811 C++ compiler currently (2.6.0) does not support the DWARF debugging
812 information normally used on 386 SVR4 platforms; stabs provide a
813 workable alternative. This requires gas and gdb, as the normal SVR4
814 tools can not generate or interpret stabs.
816 @item --disable-multilib
817 Specify that multiple target
818 libraries to support different target variants, calling
819 conventions, etc should not be built. The default is to build a
820 predefined set of them.
822 Some targets provide finer-grained control over which multilibs are built
823 (e.g., @option{--disable-softfloat}):
829 fpu, 26bit, underscore, interwork, biendian, nofmult.
832 softfloat, m68881, m68000, m68020.
835 single-float, biendian, softfloat.
837 @item powerpc*-*-*, rs6000*-*-*
838 aix64, pthread, softfloat, powercpu, powerpccpu, powerpcos, biendian,
843 @item --enable-threads
844 Specify that the target
845 supports threads. This affects the Objective-C compiler and runtime
846 library, and exception handling for other languages like C++ and Java.
847 On some systems, this is the default.
849 In general, the best (and, in many cases, the only known) threading
850 model available will be configured for use. Beware that on some
851 systems, gcc has not been taught what threading models are generally
852 available for the system. In this case, @option{--enable-threads} is an
853 alias for @option{--enable-threads=single}.
855 @item --disable-threads
856 Specify that threading support should be disabled for the system.
857 This is an alias for @option{--enable-threads=single}.
859 @item --enable-threads=@var{lib}
861 @var{lib} is the thread support library. This affects the Objective-C
862 compiler and runtime library, and exception handling for other languages
863 like C++ and Java. The possibilities for @var{lib} are:
871 Ada tasking support. For non-Ada programs, this setting is equivalent
872 to @samp{single}. When used in conjunction with the Ada run time, it
873 causes GCC to use the same thread primitives as Ada uses. This option
874 is necessary when using both Ada and the back end exception handling,
875 which is the default for most Ada targets.
877 Generic MACH thread support, known to work on NeXTSTEP@. (Please note
878 that the file needed to support this configuration, @file{gthr-mach.h}, is
879 missing and thus this setting will cause a known bootstrap failure.)
881 This is an alias for @samp{single}.
883 Generic POSIX thread support.
885 Same as @samp{posix} on arm*-*-linux*, *-*-chorusos* and *-*-freebsd*
886 only. A future release of gcc might remove this alias or extend it
889 RTEMS thread support.
891 Disable thread support, should work for all platforms.
893 Sun Solaris 2 thread support.
895 VxWorks thread support.
897 Microsoft Win32 API thread support.
900 @item --with-cpu=@var{cpu}
901 Specify which cpu variant the compiler should generate code for by default.
902 @var{cpu} will be used as the default value of the @option{-mcpu=} switch.
903 This option is only supported on some targets, including ARM, i386, PowerPC,
906 @item --with-schedule=@var{cpu}
907 @itemx --with-arch=@var{cpu}
908 @itemx --with-tune=@var{cpu}
909 @itemx --with-abi=@var{abi}
910 @itemx --with-float=@var{type}
911 These configure options provide default values for the @option{-mschedule=},
912 @option{-march=}, @option{-mtune=}, and @option{-mabi=} options and for
913 @option{-mhard-float} or @option{-msoft-float}. As with @option{--with-cpu},
914 which switches will be accepted and acceptable values of the arguments depend
917 @item --enable-altivec
918 Specify that the target supports AltiVec vector enhancements. This
919 option will adjust the ABI for AltiVec enhancements, as well as generate
920 AltiVec code when appropriate. This option is only available for
923 @item --enable-target-optspace
925 libraries should be optimized for code space instead of code speed.
926 This is the default for the m32r platform.
929 Specify that a user visible @command{cpp} program should not be installed.
931 @item --with-cpp-install-dir=@var{dirname}
932 Specify that the user visible @command{cpp} program should be installed
933 in @file{@var{prefix}/@var{dirname}/cpp}, in addition to @var{bindir}.
935 @item --enable-initfini-array
936 Force the use of sections @code{.init_array} and @code{.fini_array}
937 (instead of @code{.init} and @code{.fini}) for constructors and
938 destructors. Option @option{--disable-initfini-array} has the
939 opposite effect. If neither option is specified, the configure script
940 will try to guess whether the @code{.init_array} and
941 @code{.fini_array} sections are supported and, if they are, use them.
943 @item --enable-maintainer-mode
945 regenerate the GCC master message catalog @file{gcc.pot} are normally
946 disabled. This is because it can only be rebuilt if the complete source
947 tree is present. If you have changed the sources and want to rebuild the
948 catalog, configuring with @option{--enable-maintainer-mode} will enable
949 this. Note that you need a recent version of the @code{gettext} tools
952 @item --enable-version-specific-runtime-libs
954 that runtime libraries should be installed in the compiler specific
955 subdirectory (@file{@var{libsubdir}}) rather than the usual places. In
956 addition, @samp{libstdc++}'s include files will be installed in
957 @file{@var{libsubdir}/include/g++} unless you overruled it by using
958 @option{--with-gxx-include-dir=@var{dirname}}. Using this option is
959 particularly useful if you intend to use several versions of GCC in
960 parallel. This is currently supported by @samp{libf2c} and
961 @samp{libstdc++}, and is the default for @samp{libobjc} which cannot be
962 changed in this case.
964 @item --enable-languages=@var{lang1},@var{lang2},@dots{}
965 Specify that only a particular subset of compilers and
966 their runtime libraries should be built. For a list of valid values for
967 @var{langN} you can issue the following command in the
968 @file{gcc} directory of your GCC source tree:@*
970 grep language= */config-lang.in
972 Currently, you can use any of the following:
973 @code{ada}, @code{c}, @code{c++}, @code{f77}, @code{java}, @code{objc}.
974 Building the Ada compiler has special requirements, see below.@*
975 If you do not pass this flag, all languages available in the @file{gcc}
976 sub-tree will be configured. Re-defining @code{LANGUAGES} when calling
977 @samp{make bootstrap} @strong{does not} work anymore, as those
978 language sub-directories might not have been configured!
980 @item --disable-libgcj
981 Specify that the run-time libraries
982 used by GCJ should not be built. This is useful in case you intend
983 to use GCJ with some other run-time, or you're going to install it
984 separately, or it just happens not to build on your particular
985 machine. In general, if the Java front end is enabled, the GCJ
986 libraries will be enabled too, unless they're known to not work on
987 the target platform. If GCJ is enabled but @samp{libgcj} isn't built, you
988 may need to port it; in this case, before modifying the top-level
989 @file{configure.in} so that @samp{libgcj} is enabled by default on this platform,
990 you may use @option{--enable-libgcj} to override the default.
993 Specify that the compiler should
994 use DWARF 2 debugging information as the default.
996 @item --enable-win32-registry
997 @itemx --enable-win32-registry=@var{key}
998 @itemx --disable-win32-registry
999 The @option{--enable-win32-registry} option enables Windows-hosted GCC
1000 to look up installations paths in the registry using the following key:
1003 @code{HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\@var{key}}
1006 @var{key} defaults to GCC version number, and can be overridden by the
1007 @option{--enable-win32-registry=@var{key}} option. Vendors and distributors
1008 who use custom installers are encouraged to provide a different key,
1009 perhaps one comprised of vendor name and GCC version number, to
1010 avoid conflict with existing installations. This feature is enabled
1011 by default, and can be disabled by @option{--disable-win32-registry}
1012 option. This option has no effect on the other hosts.
1015 Specify that the machine does not have a floating point unit. This
1016 option only applies to @samp{m68k-sun-sunos@var{n}}. On any other
1017 system, @option{--nfp} has no effect.
1019 @item --enable-werror
1020 @itemx --disable-werror
1021 @itemx --enable-werror=yes
1022 @itemx --enable-werror=no
1023 When you specify this option, it controls whether certain files in the
1024 compiler are built with @option{-Werror} in bootstrap stage2 and later.
1025 If you don't specify it, @option{-Werror} is turned on for the main
1026 development trunk. However it defaults to off for release branches and
1027 final releases. The specific files which get @option{-Werror} are
1028 controlled by the Makefiles.
1030 @item --enable-checking
1031 @itemx --enable-checking=@var{list}
1032 When you specify this option, the compiler is built to perform checking
1033 of tree node types when referencing fields of that node, and some other
1034 internal consistency checks. This does not change the generated code,
1035 but adds error checking within the compiler. This will slow down the
1036 compiler and may only work properly if you are building the compiler
1037 with GCC@. This is on by default when building from CVS or snapshots,
1038 but off for releases. More control over the checks may be had by
1039 specifying @var{list}; the categories of checks available are
1040 @samp{misc}, @samp{tree}, @samp{gc}, @samp{rtl}, @samp{rtlflag},
1041 @samp{fold}, @samp{gcac} and @samp{valgrind}. The check @samp{valgrind}
1042 requires the external @command{valgrind} simulator, available from
1043 @uref{http://developer.kde.org/~sewardj/}. The default when @var{list} is
1044 not specified is @samp{misc,tree,gc,rtlflag}; the checks @samp{rtl},
1045 @samp{gcac} and @samp{valgrind} are very expensive.
1047 @item --enable-coverage
1048 @item --enable-coverage=@var{level}
1049 With this option, the compiler is built to collect self coverage
1050 information, every time it is run. This is for internal development
1051 purposes, and only works when the compiler is being built with gcc. The
1052 @var{level} argument controls whether the compiler is built optimized or
1053 not, values are @samp{opt} and @samp{noopt}. For coverage analysis you
1054 want to disable optimization, for performance analysis you want to
1055 enable optimization. When coverage is enabled, the default level is
1056 without optimization.
1059 @itemx --disable-nls
1060 The @option{--enable-nls} option enables Native Language Support (NLS),
1061 which lets GCC output diagnostics in languages other than American
1062 English. Native Language Support is enabled by default if not doing a
1063 canadian cross build. The @option{--disable-nls} option disables NLS@.
1065 @item --with-included-gettext
1066 If NLS is enabled, the @option{--with-included-gettext} option causes the build
1067 procedure to prefer its copy of GNU @command{gettext}.
1069 @item --with-catgets
1070 If NLS is enabled, and if the host lacks @code{gettext} but has the
1071 inferior @code{catgets} interface, the GCC build procedure normally
1072 ignores @code{catgets} and instead uses GCC's copy of the GNU
1073 @code{gettext} library. The @option{--with-catgets} option causes the
1074 build procedure to use the host's @code{catgets} in this situation.
1076 @item --with-libiconv-prefix=@var{dir}
1077 Search for libiconv header files in @file{@var{dir}/include} and
1078 libiconv library files in @file{@var{dir}/lib}.
1080 @item --with-system-zlib
1081 Use installed zlib rather than that included with GCC@. This option
1082 only applies if the Java front end is being built.
1084 @item --enable-obsolete
1085 Enable configuration for an obsoleted system. If you attempt to
1086 configure GCC for a system (build, host, or target) which has been
1087 obsoleted, and you do not specify this flag, configure will halt with an
1090 All support for systems which have been obsoleted in one release of GCC
1091 is removed entirely in the next major release, unless someone steps
1092 forward to maintain the port.
1095 Some options which only apply to building cross compilers:
1097 @item --with-sysroot
1098 @itemx --with-sysroot=@var{dir}
1099 Tells GCC to consider @var{dir} as the root of a tree that contains a
1100 (subset of) the root filesystem of the target operating system.
1101 Target system headers, libraries and run-time object files will be
1102 searched in there. The specified directory is not copied into the
1103 install tree, unlike the options @option{--with-headers} and
1104 @option{--with-libs} that this option obsoletes. The default value,
1105 in case @option{--with-sysroot} is not given an argument, is
1106 @option{$@{gcc_tooldir@}/sys-root}. If the specified directory is a
1107 subdirectory of @option{$@{exec_prefix@}}, then it will be found relative to
1108 the GCC binaries if the installation tree is moved.
1110 @item --with-headers
1111 @itemx --with-headers=@var{dir}
1112 Deprecated in favor of @option{--with-sysroot}.
1113 Specifies that target headers are available when building a cross compiler.
1114 The @var{dir} argument specifies a directory which has the target include
1115 files. These include files will be copied into the @file{gcc} install
1116 directory. @emph{This option with the @var{dir} argument is required} when
1117 building a cross compiler, if @file{@var{prefix}/@var{target}/sys-include}
1118 doesn't pre-exist. If @file{@var{prefix}/@var{target}/sys-include} does
1119 pre-exist, the @var{dir} argument may be omitted. @command{fixincludes}
1120 will be run on these files to make them compatible with GCC.
1122 @item --without-headers
1123 Tells GCC not use any target headers from a libc when building a cross
1124 compiler. When crossing to GNU/Linux, you need the headers so gcc
1125 can build the exception handling for libgcc.
1126 See @uref{http://www.objsw.com/CrossGCC/,,CrossGCC} for more information
1130 @itemx --with-libs=``@var{dir1} @var{dir2} @dots{} @var{dirN}''
1131 Deprecated in favor of @option{--with-sysroot}.
1132 Specifies a list of directories which contain the target runtime
1133 libraries. These libraries will be copied into the @file{gcc} install
1134 directory. If the directory list is omitted, this option has no
1137 Specifies that @samp{newlib} is
1138 being used as the target C library. This causes @code{__eprintf} to be
1139 omitted from @file{libgcc.a} on the assumption that it will be provided by
1143 Note that each @option{--enable} option has a corresponding
1144 @option{--disable} option and that each @option{--with} option has a
1145 corresponding @option{--without} option.
1152 @uref{./index.html,,Return to the GCC Installation page}
1156 @c ***Building****************************************************************
1158 @comment node-name, next, previous, up
1159 @node Building, Testing, Configuration, Installing GCC
1165 @cindex Installing GCC: Building
1167 Now that GCC is configured, you are ready to build the compiler and
1170 We @strong{highly} recommend that GCC be built using GNU make;
1171 other versions may work, then again they might not.
1172 GNU make is required for compiling GNAT (the Ada compiler) and the Java
1175 (For example, many broken versions of make will fail if you use the
1176 recommended setup where @var{objdir} is different from @var{srcdir}.
1177 Other broken versions may recompile parts of the compiler when
1178 installing the compiler.)
1180 Some commands executed when making the compiler may fail (return a
1181 nonzero status) and be ignored by @command{make}. These failures, which
1182 are often due to files that were not found, are expected, and can safely
1185 It is normal to have compiler warnings when compiling certain files.
1186 Unless you are a GCC developer, you can generally ignore these warnings
1187 unless they cause compilation to fail. Developers should attempt to fix
1188 any warnings encountered, however they can temporarily continue past
1189 warnings-as-errors by specifying the configure flag
1190 @option{--disable-werror}.
1192 On certain old systems, defining certain environment variables such as
1193 @env{CC} can interfere with the functioning of @command{make}.
1195 If you encounter seemingly strange errors when trying to build the
1196 compiler in a directory other than the source directory, it could be
1197 because you have previously configured the compiler in the source
1198 directory. Make sure you have done all the necessary preparations.
1200 If you build GCC on a BSD system using a directory stored in an old System
1201 V file system, problems may occur in running @command{fixincludes} if the
1202 System V file system doesn't support symbolic links. These problems
1203 result in a failure to fix the declaration of @code{size_t} in
1204 @file{sys/types.h}. If you find that @code{size_t} is a signed type and
1205 that type mismatches occur, this could be the cause.
1207 The solution is not to use such a directory for building GCC@.
1209 When building from CVS or snapshots, or if you modify parser sources,
1210 you need the Bison parser generator installed. Any version 1.25 or
1211 later should work; older versions may also work. If you do not modify
1212 parser sources, releases contain the Bison-generated files and you do
1213 not need Bison installed to build them.
1215 When building from CVS or snapshots, or if you modify Texinfo
1216 documentation, you need version 4.2 or later of Texinfo installed if you
1217 want Info documentation to be regenerated. Releases contain Info
1218 documentation pre-built for the unmodified documentation in the release.
1220 @section Building a native compiler
1222 For a native build issue the command @samp{make bootstrap}. This
1223 will build the entire GCC system, which includes the following steps:
1227 Build host tools necessary to build the compiler such as texinfo, bison,
1231 Build target tools for use by the compiler such as binutils (bfd,
1232 binutils, gas, gprof, ld, and opcodes)
1233 if they have been individually linked
1234 or moved into the top level GCC source tree before configuring.
1237 Perform a 3-stage bootstrap of the compiler.
1240 Perform a comparison test of the stage2 and stage3 compilers.
1243 Build runtime libraries using the stage3 compiler from the previous step.
1247 If you are short on disk space you might consider @samp{make
1248 bootstrap-lean} instead. This is identical to @samp{make
1249 bootstrap} except that object files from the stage1 and
1250 stage2 of the 3-stage bootstrap of the compiler are deleted as
1251 soon as they are no longer needed.
1253 If you want to save additional space during the bootstrap and in
1254 the final installation as well, you can build the compiler binaries
1255 without debugging information as in the following example. This will save
1256 roughly 40% of disk space both for the bootstrap and the final installation.
1257 (Libraries will still contain debugging information.)
1260 make CFLAGS='-O' LIBCFLAGS='-g -O2' \
1261 LIBCXXFLAGS='-g -O2 -fno-implicit-templates' bootstrap
1264 If you wish to use non-default GCC flags when compiling the stage2 and
1265 stage3 compilers, set @code{BOOT_CFLAGS} on the command line when doing
1266 @samp{make bootstrap}. Non-default optimization flags are less well
1267 tested here than the default of @samp{-g -O2}, but should still work.
1268 In a few cases, you may find that you need to specify special flags such
1269 as @option{-msoft-float} here to complete the bootstrap; or, if the
1270 native compiler miscompiles the stage1 compiler, you may need to work
1271 around this, by choosing @code{BOOT_CFLAGS} to avoid the parts of the
1272 stage1 compiler that were miscompiled, or by using @samp{make
1273 bootstrap4} to increase the number of stages of bootstrap.
1275 If you used the flag @option{--enable-languages=@dots{}} to restrict
1276 the compilers to be built, only those you've actually enabled will be
1277 built. This will of course only build those runtime libraries, for
1278 which the particular compiler has been built. Please note,
1279 that re-defining @env{LANGUAGES} when calling @samp{make bootstrap}
1280 @strong{does not} work anymore!
1282 If the comparison of stage2 and stage3 fails, this normally indicates
1283 that the stage2 compiler has compiled GCC incorrectly, and is therefore
1284 a potentially serious bug which you should investigate and report. (On
1285 a few systems, meaningful comparison of object files is impossible; they
1286 always appear ``different''. If you encounter this problem, you will
1287 need to disable comparison in the @file{Makefile}.)
1289 @section Building a cross compiler
1291 We recommend reading the
1292 @uref{http://www.objsw.com/CrossGCC/,,crossgcc FAQ}
1293 for information about building cross compilers.
1295 When building a cross compiler, it is not generally possible to do a
1296 3-stage bootstrap of the compiler. This makes for an interesting problem
1297 as parts of GCC can only be built with GCC@.
1299 To build a cross compiler, we first recommend building and installing a
1300 native compiler. You can then use the native GCC compiler to build the
1301 cross compiler. The installed native compiler needs to be GCC version
1304 Assuming you have already installed a native copy of GCC and configured
1305 your cross compiler, issue the command @command{make}, which performs the
1310 Build host tools necessary to build the compiler such as texinfo, bison,
1314 Build target tools for use by the compiler such as binutils (bfd,
1315 binutils, gas, gprof, ld, and opcodes)
1316 if they have been individually linked or moved into the top level GCC source
1317 tree before configuring.
1320 Build the compiler (single stage only).
1323 Build runtime libraries using the compiler from the previous step.
1326 Note that if an error occurs in any step the make process will exit.
1328 @section Building in parallel
1330 You can use @samp{make bootstrap MAKE="make -j 2" -j 2}, or just
1331 @samp{make -j 2 bootstrap} for GNU Make 3.79 and above, instead of
1332 @samp{make bootstrap} to build GCC in parallel.
1333 You can also specify a bigger number, and in most cases using a value
1334 greater than the number of processors in your machine will result in
1335 fewer and shorter I/O latency hits, thus improving overall throughput;
1336 this is especially true for slow drives and network filesystems.
1338 @section Building the Ada compiler
1340 In order to build GNAT, the Ada compiler, you need a working GNAT
1341 compiler (GNAT version 3.14 or later, or GCC version 3.1 or later),
1342 since the Ada front end is written in Ada (with some
1343 GNAT-specific extensions), and GNU make.
1345 However, you do not need a full installation of GNAT, just the GNAT
1346 binary @file{gnat1}, a copy of @file{gnatbind}, and a compiler driver
1347 which can deal with Ada input (by invoking the @file{gnat1} binary).
1348 You can specify this compiler driver by setting the @env{ADAC}
1349 environment variable at the configure step. @command{configure} can
1350 detect the driver automatically if it has got a common name such as
1351 @command{gcc} or @command{gnatgcc}. Of course, you still need a working
1352 C compiler (the compiler driver can be different or not).
1353 @command{configure} does not test whether the GNAT installation works
1354 and has a sufficiently recent version; if too old a GNAT version is
1355 installed, the build will fail unless @option{--enable-languages} is
1356 used to disable building the Ada front end.
1358 Additional build tools (such as @command{gnatmake}) or a working GNAT
1359 run-time library installation are usually @emph{not} required. However,
1360 if you want to bootstrap the compiler using a minimal version of GNAT,
1361 you have to issue the following commands before invoking @samp{make
1362 bootstrap} (this assumes that you start with an unmodified and consistent
1363 source distribution):
1366 cd @var{srcdir}/gcc/ada
1367 touch treeprs.ads [es]info.h nmake.ad[bs]
1370 At the moment, the GNAT library and several tools for GNAT are not built
1371 by @samp{make bootstrap}. You have to invoke
1372 @samp{make gnatlib_and_tools} in the @file{@var{objdir}/gcc}
1373 subdirectory before proceeding with the next steps.
1375 For example, you can build a native Ada compiler by issuing the
1376 following commands (assuming @command{make} is GNU make):
1380 @var{srcdir}/configure --enable-languages=c,ada
1381 cd @var{srcdir}/gcc/ada
1382 touch treeprs.ads [es]info.h nmake.ad[bs]
1386 make gnatlib_and_tools
1390 Currently, when compiling the Ada front end, you cannot use the parallel
1391 build feature described in the previous section.
1393 @section Building with profile feedback
1395 It is possible to use profile feedback to optimize the compiler itself. This
1396 should result in a faster compiler binary. Experiments done on x86 using gcc
1397 3.3 showed approximately 7 percent speedup on compiling C programs. To
1398 bootstrap compiler with profile feedback, use @code{make profiledbootstrap}.
1400 When @samp{make profiledbootstrap} is run, it will first build a @code{stage1}
1401 compiler. This compiler is used to build a @code{stageprofile} compiler
1402 instrumented to collect execution counts of instruction and branch
1403 probabilities. Then runtime libraries are compiled with profile collected.
1404 Finally a @code{stagefeedback} compiler is built using the information collected.
1406 Unlike @samp{make bootstrap} several additional restrictions apply. The
1407 compiler used to build @code{stage1} needs to support a 64-bit integral type.
1408 It is recommended to only use GCC for this. Also parallel make is currently
1409 not supported since collisions in profile collecting may occur.
1416 @uref{./index.html,,Return to the GCC Installation page}
1420 @c ***Testing*****************************************************************
1422 @comment node-name, next, previous, up
1423 @node Testing, Final install, Building, Installing GCC
1427 @chapter Installing GCC: Testing
1430 @cindex Installing GCC: Testing
1433 Before you install GCC, we encourage you to run the testsuites and to
1434 compare your results with results from a similar configuration that have
1435 been submitted to the
1436 @uref{http://gcc.gnu.org/ml/gcc-testresults/,,gcc-testresults mailing list}.
1437 Some of these archived results are linked from the build status lists
1438 at @uref{http://gcc.gnu.org/buildstat.html}, although not everyone who
1439 reports a successful build runs the testsuites and submits the results.
1440 This step is optional and may require you to download additional software,
1441 but it can give you confidence in your new GCC installation or point out
1442 problems before you install and start using your new GCC.
1444 First, you must have @uref{download.html,,downloaded the testsuites}.
1445 These are part of the full distribution, but if you downloaded the
1446 ``core'' compiler plus any front ends, you must download the testsuites
1449 Second, you must have the testing tools installed. This includes
1450 @uref{http://www.gnu.org/software/dejagnu/,,DejaGnu} 1.4.1 or 1.4.3
1451 and later, Tcl, and Expect; the DejaGnu site has links to these.
1453 If the directories where @command{runtest} and @command{expect} were
1454 installed are not in the @env{PATH}, you may need to set the following
1455 environment variables appropriately, as in the following example (which
1456 assumes that DejaGnu has been installed under @file{/usr/local}):
1459 TCL_LIBRARY = /usr/local/share/tcl8.0
1460 DEJAGNULIBS = /usr/local/share/dejagnu
1463 (On systems such as Cygwin, these paths are required to be actual
1464 paths, not mounts or links; presumably this is due to some lack of
1465 portability in the DejaGnu code.)
1468 Finally, you can run the testsuite (which may take a long time):
1470 cd @var{objdir}; make -k check
1473 This will test various components of GCC, such as compiler
1474 front ends and runtime libraries. While running the testsuite, DejaGnu
1475 might emit some harmless messages resembling
1476 @samp{WARNING: Couldn't find the global config file.} or
1477 @samp{WARNING: Couldn't find tool init file} that can be ignored.
1479 @section How can I run the test suite on selected tests?
1481 In order to run sets of tests selectively, there are targets
1482 @samp{make check-gcc} and @samp{make check-g++}
1483 in the @file{gcc} subdirectory of the object directory. You can also
1484 just run @samp{make check} in a subdirectory of the object directory.
1487 A more selective way to just run all @command{gcc} execute tests in the
1491 make check-gcc RUNTESTFLAGS="execute.exp @var{other-options}"
1494 Likewise, in order to run only the @command{g++} ``old-deja'' tests in
1495 the testsuite with filenames matching @samp{9805*}, you would use
1498 make check-g++ RUNTESTFLAGS="old-deja.exp=9805* @var{other-options}"
1501 The @file{*.exp} files are located in the testsuite directories of the GCC
1502 source, the most important ones being @file{compile.exp},
1503 @file{execute.exp}, @file{dg.exp} and @file{old-deja.exp}.
1504 To get a list of the possible @file{*.exp} files, pipe the
1505 output of @samp{make check} into a file and look at the
1506 @samp{Running @dots{} .exp} lines.
1508 The Java runtime tests can be executed via @samp{make check}
1509 in the @file{@var{target}/libjava/testsuite} directory in
1512 @section Additional testing for Java Class Libraries
1514 The @uref{http://sources.redhat.com/mauve/,,Mauve Project} provides
1515 a suite of tests for the Java Class Libraries. This suite can be run
1516 as part of libgcj testing by placing the Mauve tree within the libjava
1517 testsuite at @file{libjava/testsuite/libjava.mauve/mauve}, or by
1518 specifying the location of that tree when invoking @samp{make}, as in
1519 @samp{make MAUVEDIR=~/mauve check}.
1521 @uref{http://www-124.ibm.com/developerworks/oss/cvs/jikes/~checkout~/jacks/jacks.html,,Jacks}
1522 is a free test suite that tests Java compiler front ends. This suite
1523 can be run as part of libgcj testing by placing the Jacks tree within
1524 the libjava testsuite at @file{libjava/testsuite/libjava.jacks/jacks}.
1526 @section How to interpret test results
1528 The result of running the testsuite are various @file{*.sum} and @file{*.log}
1529 files in the testsuite subdirectories. The @file{*.log} files contain a
1530 detailed log of the compiler invocations and the corresponding
1531 results, the @file{*.sum} files summarize the results. These summaries
1532 contain status codes for all tests:
1536 PASS: the test passed as expected
1538 XPASS: the test unexpectedly passed
1540 FAIL: the test unexpectedly failed
1542 XFAIL: the test failed as expected
1544 UNSUPPORTED: the test is not supported on this platform
1546 ERROR: the testsuite detected an error
1548 WARNING: the testsuite detected a possible problem
1551 It is normal for some tests to report unexpected failures. At the
1552 current time our testing harness does not allow fine grained control
1553 over whether or not a test is expected to fail. We expect to fix this
1554 problem in future releases.
1557 @section Submitting test results
1559 If you want to report the results to the GCC project, use the
1560 @file{contrib/test_summary} shell script. Start it in the @var{objdir} with
1563 @var{srcdir}/contrib/test_summary -p your_commentary.txt \
1564 -m gcc-testresults@@gcc.gnu.org |sh
1567 This script uses the @command{Mail} program to send the results, so
1568 make sure it is in your @env{PATH}. The file @file{your_commentary.txt} is
1569 prepended to the testsuite summary and should contain any special
1570 remarks you have on your results or your build environment. Please
1571 do not edit the testsuite result block or the subject line, as these
1572 messages may be automatically processed.
1579 @uref{./index.html,,Return to the GCC Installation page}
1583 @c ***Final install***********************************************************
1585 @comment node-name, next, previous, up
1586 @node Final install, , Testing, Installing GCC
1588 @ifset finalinstallhtml
1590 @chapter Installing GCC: Final installation
1593 Now that GCC has been built (and optionally tested), you can install it with
1595 cd @var{objdir}; make install
1598 We strongly recommend to install into a target directory where there is
1599 no previous version of GCC present.
1601 That step completes the installation of GCC; user level binaries can
1602 be found in @file{@var{prefix}/bin} where @var{prefix} is the value you
1603 specified with the @option{--prefix} to configure (or @file{/usr/local}
1604 by default). (If you specified @option{--bindir}, that directory will
1605 be used instead; otherwise, if you specified @option{--exec-prefix},
1606 @file{@var{exec-prefix}/bin} will be used.) Headers for the C++ and
1607 Java libraries are installed in @file{@var{prefix}/include}; libraries
1608 in @file{@var{libdir}} (normally @file{@var{prefix}/lib}); internal
1609 parts of the compiler in @file{@var{libdir}/gcc-lib}; documentation in
1610 info format in @file{@var{infodir}} (normally @file{@var{prefix}/info}).
1612 When installing cross-compilers, GCC's executables
1613 are not only installed into @file{@var{bindir}}, that
1614 is, @file{@var{exec-prefix}/bin}, but additionally into
1615 @file{@var{exec-prefix}/@var{target-alias}/bin}, if that directory
1616 exists. Typically, such @dfn{tooldirs} hold target-specific
1617 binutils, including assembler and linker.
1619 Installation into a temporary staging area or into a @command{chroot}
1620 jail can be achieved with the command
1623 make DESTDIR=@var{path-to-rootdir} install
1626 @noindent where @var{path-to-rootdir} is the absolute path of
1627 a directory relative to which all installation paths will be
1628 interpreted. Note that the directory specified by @code{DESTDIR}
1629 need not exist yet; it will be created if necessary.
1631 There is a subtle point with tooldirs and @code{DESTDIR}:
1632 If you relocate a cross-compiler installation with
1633 e.g.@: @samp{DESTDIR=@var{rootdir}}, then the directory
1634 @file{@var{rootdir}/@var{exec-prefix}/@var{target-alias}/bin} will
1635 be filled with duplicated GCC executables only if it already exists,
1636 it will not be created otherwise. This is regarded as a feature,
1637 not as a bug, because it gives slightly more control to the packagers
1638 using the @code{DESTDIR} feature.
1640 If you built a released version of GCC using @samp{make bootstrap} then please
1641 quickly review the build status page for your release, available from
1642 @uref{http://gcc.gnu.org/buildstat.html}.
1643 If your system is not listed for the version of GCC that you built,
1645 @email{gcc@@gcc.gnu.org} indicating
1646 that you successfully built and installed GCC.
1647 Include the following information:
1651 Output from running @file{@var{srcdir}/config.guess}. Do not send us
1652 that file itself, just the one-line output from running it.
1655 The output of @samp{gcc -v} for your newly installed gcc.
1656 This tells us which version of GCC you built and the options you passed to
1660 Whether you enabled all languages or a subset of them. If you used a
1661 full distribution then this information is part of the configure
1662 options in the output of @samp{gcc -v}, but if you downloaded the
1663 ``core'' compiler plus additional front ends then it isn't apparent
1664 which ones you built unless you tell us about it.
1667 If the build was for GNU/Linux, also include:
1670 The distribution name and version (e.g., Red Hat 7.1 or Debian 2.2.3);
1671 this information should be available from @file{/etc/issue}.
1674 The version of the Linux kernel, available from @samp{uname --version}
1678 The version of glibc you used; for RPM-based systems like Red Hat,
1679 Mandrake, and SuSE type @samp{rpm -q glibc} to get the glibc version,
1680 and on systems like Debian and Progeny use @samp{dpkg -l libc6}.
1682 For other systems, you can include similar information if you think it is
1686 Any other information that you think would be useful to people building
1687 GCC on the same configuration. The new entry in the build status list
1688 will include a link to the archived copy of your message.
1691 We'd also like to know if the
1693 @ref{Specific, host/target specific installation notes}
1696 @uref{specific.html,,host/target specific installation notes}
1698 didn't include your host/target information or if that information is
1699 incomplete or out of date. Send a note to
1700 @email{gcc@@gcc.gnu.org} telling us how the information should be changed.
1702 If you find a bug, please report it following our
1703 @uref{../bugs.html,,bug reporting guidelines}.
1705 If you want to print the GCC manuals, do @samp{cd @var{objdir}; make
1706 dvi}. You will need to have @command{texi2dvi} (version at least 4.2)
1707 and @TeX{} installed. This creates a number of @file{.dvi} files in
1708 subdirectories of @file{@var{objdir}}; these may be converted for
1709 printing with programs such as @command{dvips}. You can also
1710 @uref{http://www.gnu.org/order/order.html,,buy printed manuals from the
1711 Free Software Foundation}, though such manuals may not be for the most
1712 recent version of GCC@.
1719 @uref{./index.html,,Return to the GCC Installation page}
1723 @c ***Binaries****************************************************************
1725 @comment node-name, next, previous, up
1726 @node Binaries, Specific, Installing GCC, Top
1730 @chapter Installing GCC: Binaries
1733 @cindex Installing GCC: Binaries
1735 We are often asked about pre-compiled versions of GCC@. While we cannot
1736 provide these for all platforms, below you'll find links to binaries for
1737 various platforms where creating them by yourself is not easy due to various
1740 Please note that we did not create these binaries, nor do we
1741 support them. If you have any problems installing them, please
1742 contact their makers.
1749 @uref{http://www.bullfreeware.com,,Bull's Freeware and Shareware Archive for AIX};
1752 @uref{http://aixpdslib.seas.ucla.edu,,UCLA Software Library for AIX}.
1756 DOS---@uref{http://www.delorie.com/djgpp/,,DJGPP}.
1759 Renesas H8/300[HS]---@uref{http://h8300-hms.sourceforge.net/,,GNU
1760 Development Tools for the Renesas H8/300[HS] Series}.
1766 @uref{http://hpux.cae.wisc.edu/,,HP-UX Porting Center};
1769 @uref{ftp://sunsite.informatik.rwth-aachen.de/pub/packages/gcc_hpux/,,Binaries for HP-UX 11.00 at Aachen University of Technology}.
1773 Motorola 68HC11/68HC12---@uref{http://www.gnu-m68hc11.org,,GNU
1774 Development Tools for the Motorola 68HC11/68HC12}.
1777 @uref{http://www.sco.com/skunkware/devtools/index.html#gcc,,SCO
1778 OpenServer/Unixware}.
1781 Sinix/Reliant Unix---@uref{ftp://ftp.fujitsu-siemens.com/pub/pd/gnu/gcc/,,Siemens}.
1784 Solaris 2 (SPARC, Intel)---@uref{http://www.sunfreeware.com/,,Sunfreeware}.
1787 SGI---@uref{http://freeware.sgi.com/,,SGI Freeware}.
1793 The @uref{http://sources.redhat.com/cygwin/,,Cygwin} project;
1795 The @uref{http://www.mingw.org/,,MinGW} project.
1799 @uref{ftp://ftp.thewrittenword.com/packages/by-name/,,The
1800 Written Word} offers binaries for
1803 Digital UNIX 4.0D and 5.1,
1805 HP-UX 10.20, 11.00, and 11.11, and
1806 Solaris/SPARC 2.5.1, 2.6, 2.7, 8, and 9,
1809 In addition to those specific offerings, you can get a binary
1810 distribution CD-ROM from the
1811 @uref{http://www.fsf.org/order/order.html,,Free Software Foundation}.
1812 It contains binaries for a number of platforms, and
1813 includes not only GCC, but other stuff as well. The current CD does
1814 not contain the latest version of GCC, but it should allow
1815 bootstrapping the compiler. An updated version of that disk is in the
1823 @uref{./index.html,,Return to the GCC Installation page}
1827 @c ***Specific****************************************************************
1829 @comment node-name, next, previous, up
1830 @node Specific, Old, Binaries, Top
1834 @chapter Host/target specific installation notes for GCC
1837 @cindex Specific installation notes
1838 @cindex Target specific installation
1839 @cindex Host specific installation
1840 @cindex Target specific installation notes
1842 Please read this document carefully @emph{before} installing the
1843 GNU Compiler Collection on your machine.
1848 @uref{#alpha*-*-*,,alpha*-*-*}
1850 @uref{#alpha*-dec-osf*,,alpha*-dec-osf*}
1852 @uref{#alphaev5-cray-unicosmk*,,alphaev5-cray-unicosmk*}
1854 @uref{#arc-*-elf,,arc-*-elf}
1856 @uref{#arm-*-elf,,arm-*-elf}
1857 @uref{#arm-*-coff,,arm-*-coff}
1858 @uref{#arm-*-aout,,arm-*-aout}
1860 @uref{#xscale-*-*,,xscale-*-*}
1868 @uref{#dsp16xx,,dsp16xx}
1870 @uref{#*-*-freebsd*,,*-*-freebsd*}
1872 @uref{#h8300-hms,,h8300-hms}
1874 @uref{#hppa*-hp-hpux*,,hppa*-hp-hpux*}
1876 @uref{#hppa*-hp-hpux10,,hppa*-hp-hpux10}
1878 @uref{#hppa*-hp-hpux11,,hppa*-hp-hpux11}
1880 @uref{#i370-*-*,,i370-*-*}
1882 @uref{#*-*-linux-gnu,,*-*-linux-gnu}
1884 @uref{#ix86-*-linux*aout,,i?86-*-linux*aout}
1886 @uref{#ix86-*-linux*,,i?86-*-linux*}
1888 @uref{#ix86-*-sco3.2v5*,,i?86-*-sco3.2v5*}
1890 @uref{#ix86-*-udk,,i?86-*-udk}
1892 @uref{#ix86-*-esix,,i?86-*-esix}
1894 @uref{#ia64-*-linux,,ia64-*-linux}
1896 @uref{#ia64-*-hpux*,,ia64-*-hpux*}
1898 @uref{#*-ibm-aix*,,*-ibm-aix*}
1900 @uref{#ip2k-*-elf,,ip2k-*-elf}
1902 @uref{#m32r-*-elf,,m32r-*-elf}
1904 @uref{#m6811-elf,,m6811-elf}
1906 @uref{#m6812-elf,,m6812-elf}
1908 @uref{#m68k-hp-hpux,,m68k-hp-hpux}
1910 @uref{#mips-*-*,,mips-*-*}
1912 @uref{#mips-sgi-irix5,,mips-sgi-irix5}
1914 @uref{#mips-sgi-irix6,,mips-sgi-irix6}
1916 @uref{#powerpc*-*-*,,powerpc*-*-*, powerpc-*-sysv4}
1918 @uref{#powerpc-*-darwin*,,powerpc-*-darwin*}
1920 @uref{#powerpc-*-elf,,powerpc-*-elf, powerpc-*-sysv4}
1922 @uref{#powerpc-*-linux-gnu*,,powerpc-*-linux-gnu*}
1924 @uref{#powerpc-*-netbsd*,,powerpc-*-netbsd*}
1926 @uref{#powerpc-*-eabiaix,,powerpc-*-eabiaix}
1928 @uref{#powerpc-*-eabisim,,powerpc-*-eabisim}
1930 @uref{#powerpc-*-eabi,,powerpc-*-eabi}
1932 @uref{#powerpcle-*-elf,,powerpcle-*-elf, powerpcle-*-sysv4}
1934 @uref{#powerpcle-*-eabisim,,powerpcle-*-eabisim}
1936 @uref{#powerpcle-*-eabi,,powerpcle-*-eabi}
1938 @uref{#s390-*-linux*,,s390-*-linux*}
1940 @uref{#s390x-*-linux*,,s390x-*-linux*}
1942 @uref{#*-*-solaris2*,,*-*-solaris2*}
1944 @uref{#sparc-sun-solaris2*,,sparc-sun-solaris2*}
1946 @uref{#sparc-sun-solaris2.7,,sparc-sun-solaris2.7}
1948 @uref{#sparc-*-linux*,,sparc-*-linux*}
1950 @uref{#sparc64-*-solaris2*,,sparc64-*-solaris2*}
1952 @uref{#sparcv9-*-solaris2*,,sparcv9-*-solaris2*}
1954 @uref{#*-*-sysv*,,*-*-sysv*}
1956 @uref{#vax-dec-ultrix,,vax-dec-ultrix}
1958 @uref{#*-*-vxworks*,,*-*-vxworks*}
1960 @uref{#xtensa-*-elf,,xtensa-*-elf}
1962 @uref{#xtensa-*-linux*,,xtensa-*-linux*}
1964 @uref{#windows,,Microsoft Windows}
1968 @uref{#older,,Older systems}
1973 @uref{#elf_targets,,all ELF targets} (SVR4, Solaris 2, etc.)
1979 <!-- -------- host/target specific issues start here ---------------- -->
1982 @heading @anchor{alpha*-*-*}alpha*-*-*
1984 This section contains general configuration information for all
1985 alpha-based platforms using ELF (in particular, ignore this section for
1986 DEC OSF/1, Digital UNIX and Tru64 UNIX)@. In addition to reading this
1987 section, please read all other sections that match your target.
1989 We require binutils 2.11.2 or newer.
1990 Previous binutils releases had a number of problems with DWARF 2
1991 debugging information, not the least of which is incorrect linking of
1997 @heading @anchor{alpha*-dec-osf*}alpha*-dec-osf*
1998 Systems using processors that implement the DEC Alpha architecture and
1999 are running the DEC/Compaq Unix (DEC OSF/1, Digital UNIX, or Compaq
2000 Tru64 UNIX) operating system, for example the DEC Alpha AXP systems.
2002 As of GCC 3.2, versions before @code{alpha*-dec-osf4} are no longer
2003 supported. (These are the versions which identify themselves as DEC
2006 In Digital Unix V4.0, virtual memory exhausted bootstrap failures
2007 may be fixed by configuring with @option{--with-gc=simple},
2008 reconfiguring Kernel Virtual Memory and Swap parameters
2009 per the @command{/usr/sbin/sys_check} Tuning Suggestions,
2010 or applying the patch in
2011 @uref{http://gcc.gnu.org/ml/gcc/2002-08/msg00822.html}.
2013 In Tru64 UNIX V5.1, Compaq introduced a new assembler that does not
2014 currently (2001-06-13) work with @command{mips-tfile}. As a workaround,
2015 we need to use the old assembler, invoked via the barely documented
2016 @option{-oldas} option. To bootstrap GCC, you either need to use the
2020 % CC=cc @var{srcdir}/configure [@var{options}] [@var{target}]
2023 or you can use a copy of GCC 2.95.3 or higher built on Tru64 UNIX V4.0:
2026 % CC=gcc -Wa,-oldas @var{srcdir}/configure [@var{options}] [@var{target}]
2029 As of GNU binutils 2.11.2, neither GNU @command{as} nor GNU @command{ld}
2030 are supported on Tru64 UNIX, so you must not configure GCC with
2031 @option{--with-gnu-as} or @option{--with-gnu-ld}.
2033 GCC writes a @samp{.verstamp} directive to the assembler output file
2034 unless it is built as a cross-compiler. It gets the version to use from
2035 the system header file @file{/usr/include/stamp.h}. If you install a
2036 new version of DEC Unix, you should rebuild GCC to pick up the new version
2039 Note that since the Alpha is a 64-bit architecture, cross-compilers from
2040 32-bit machines will not generate code as efficient as that generated
2041 when the compiler is running on a 64-bit machine because many
2042 optimizations that depend on being able to represent a word on the
2043 target in an integral value on the host cannot be performed. Building
2044 cross-compilers on the Alpha for 32-bit machines has only been tested in
2045 a few cases and may not work properly.
2047 @samp{make compare} may fail on old versions of DEC Unix unless you add
2048 @option{-save-temps} to @code{CFLAGS}. On these systems, the name of the
2049 assembler input file is stored in the object file, and that makes
2050 comparison fail if it differs between the @code{stage1} and
2051 @code{stage2} compilations. The option @option{-save-temps} forces a
2052 fixed name to be used for the assembler input file, instead of a
2053 randomly chosen name in @file{/tmp}. Do not add @option{-save-temps}
2054 unless the comparisons fail without that option. If you add
2055 @option{-save-temps}, you will have to manually delete the @samp{.i} and
2056 @samp{.s} files after each series of compilations.
2058 GCC now supports both the native (ECOFF) debugging format used by DBX
2059 and GDB and an encapsulated STABS format for use only with GDB@. See the
2060 discussion of the @option{--with-stabs} option of @file{configure} above
2061 for more information on these formats and how to select them.
2063 There is a bug in DEC's assembler that produces incorrect line numbers
2064 for ECOFF format when the @samp{.align} directive is used. To work
2065 around this problem, GCC will not emit such alignment directives
2066 while writing ECOFF format debugging information even if optimization is
2067 being performed. Unfortunately, this has the very undesirable
2068 side-effect that code addresses when @option{-O} is specified are
2069 different depending on whether or not @option{-g} is also specified.
2071 To avoid this behavior, specify @option{-gstabs+} and use GDB instead of
2072 DBX@. DEC is now aware of this problem with the assembler and hopes to
2073 provide a fix shortly.
2078 @heading @anchor{alphaev5-cray-unicosmk*}alphaev5-cray-unicosmk*
2079 Cray T3E systems running Unicos/Mk.
2081 This port is incomplete and has many known bugs. We hope to improve the
2082 support for this target soon. Currently, only the C front end is supported,
2083 and it is not possible to build parallel applications. Cray modules are not
2084 supported; in particular, Craylibs are assumed to be in
2085 @file{/opt/ctl/craylibs/craylibs}.
2087 You absolutely @strong{must} use GNU make on this platform. Also, you
2088 need to tell GCC where to find the assembler and the linker. The
2089 simplest way to do so is by providing @option{--with-as} and
2090 @option{--with-ld} to @file{configure}, e.g.@:
2093 configure --with-as=/opt/ctl/bin/cam --with-ld=/opt/ctl/bin/cld \
2094 --enable-languages=c
2097 The comparison test during @samp{make bootstrap} fails on Unicos/Mk
2098 because the assembler inserts timestamps into object files. You should
2099 be able to work around this by doing @samp{make all} after getting this
2105 @heading @anchor{arc-*-elf}arc-*-elf
2106 Argonaut ARC processor.
2107 This configuration is intended for embedded systems.
2112 @heading @anchor{arm-*-elf}arm-*-elf
2113 @heading @anchor{xscale-*-*}xscale-*-*
2114 ARM-family processors. Subtargets that use the ELF object format
2115 require GNU binutils 2.13 or newer. Such subtargets include:
2116 @code{arm-*-freebsd}, @code{arm-*-netbsdelf}, @code{arm-*-*linux},
2117 @code{arm-*-rtems} and @code{arm-*-kaos}.
2122 @heading @anchor{arm-*-coff}arm-*-coff
2123 ARM-family processors. Note that there are two different varieties
2124 of PE format subtarget supported: @code{arm-wince-pe} and
2125 @code{arm-pe} as well as a standard COFF target @code{arm-*-coff}.
2130 @heading @anchor{arm-*-aout}arm-*-aout
2131 ARM-family processors. These targets support the AOUT file format:
2132 @code{arm-*-aout}, @code{arm-*-netbsd}.
2137 @heading @anchor{avr}avr
2139 ATMEL AVR-family micro controllers. These are used in embedded
2140 applications. There are no standard Unix configurations.
2142 @xref{AVR Options,, AVR Options, gcc, Using and Porting the GNU Compiler
2146 See ``AVR Options'' in the main manual
2148 for the list of supported MCU types.
2150 Use @samp{configure --target=avr --enable-languages="c"} to configure GCC@.
2152 Further installation notes and other useful information about AVR tools
2153 can also be obtained from:
2157 @uref{http://www.openavr.org,,http://www.openavr.org}
2159 @uref{http://home.overta.ru/users/denisc/,,http://home.overta.ru/users/denisc/}
2161 @uref{http://www.amelek.gda.pl/avr/,,http://www.amelek.gda.pl/avr/}
2164 We @emph{strongly} recommend using binutils 2.13 or newer.
2166 The following error:
2168 Error: register required
2171 indicates that you should upgrade to a newer version of the binutils.
2176 @heading @anchor{c4x}c4x
2178 Texas Instruments TMS320C3x and TMS320C4x Floating Point Digital Signal
2179 Processors. These are used in embedded applications. There are no
2180 standard Unix configurations.
2182 @xref{TMS320C3x/C4x Options,, TMS320C3x/C4x Options, gcc, Using and
2183 Porting the GNU Compiler Collection (GCC)},
2186 See ``TMS320C3x/C4x Options'' in the main manual
2188 for the list of supported MCU types.
2190 GCC can be configured as a cross compiler for both the C3x and C4x
2191 architectures on the same system. Use @samp{configure --target=c4x
2192 --enable-languages="c,c++"} to configure.
2195 Further installation notes and other useful information about C4x tools
2196 can also be obtained from:
2200 @uref{http://www.elec.canterbury.ac.nz/c4x/,,http://www.elec.canterbury.ac.nz/c4x/}
2206 @heading @anchor{cris}CRIS
2208 CRIS is the CPU architecture in Axis Communications ETRAX system-on-a-chip
2209 series. These are used in embedded applications.
2212 @xref{CRIS Options,, CRIS Options, gcc, Using and Porting the GNU Compiler
2216 See ``CRIS Options'' in the main manual
2218 for a list of CRIS-specific options.
2220 There are a few different CRIS targets:
2222 @item cris-axis-aout
2223 Old target. Includes a multilib for the @samp{elinux} a.out-based
2224 target. No multilibs for newer architecture variants.
2226 Mainly for monolithic embedded systems. Includes a multilib for the
2227 @samp{v10} core used in @samp{ETRAX 100 LX}.
2228 @item cris-axis-linux-gnu
2229 A GNU/Linux port for the CRIS architecture, currently targeting
2230 @samp{ETRAX 100 LX} by default.
2233 For @code{cris-axis-aout} and @code{cris-axis-elf} you need binutils 2.11
2234 or newer. For @code{cris-axis-linux-gnu} you need binutils 2.12 or newer.
2236 Pre-packaged tools can be obtained from
2237 @uref{ftp://ftp.axis.com/pub/axis/tools/cris/compiler-kit/}. More
2238 information about this platform is available at
2239 @uref{http://developer.axis.com/}.
2244 @heading @anchor{dos}DOS
2246 Please have a look at our @uref{binaries.html,,binaries page}.
2248 You cannot install GCC by itself on MSDOS; it will not compile under
2249 any MSDOS compiler except itself. You need to get the complete
2250 compilation package DJGPP, which includes binaries as well as sources,
2251 and includes all the necessary compilation tools and libraries.
2256 @heading @anchor{dsp16xx}dsp16xx
2257 A port to the AT&T DSP1610 family of processors.
2262 @heading @anchor{*-*-freebsd*}*-*-freebsd*
2264 The version of binutils installed in @file{/usr/bin} is known to work unless
2265 otherwise specified in any per-architecture notes. However, binutils
2266 2.12.1 or greater is known to improve overall testsuite results.
2268 FreeBSD 1 is no longer supported.
2270 For FreeBSD 2 or any mutant a.out versions of FreeBSD 3: All
2271 configuration support and files as shipped with GCC 2.95 are still in
2272 place. FreeBSD 2.2.7 has been known to bootstrap completely; however,
2273 it is unknown which version of binutils was used (it is assumed that it
2274 was the system copy in @file{/usr/bin}) and C++ EH failures were noted.
2276 For FreeBSD using the ELF file format: DWARF 2 debugging is now the
2277 default for all CPU architectures. It had been the default on
2278 FreeBSD/alpha since its inception. You may use @option{-gstabs} instead
2279 of @option{-g}, if you really want the old debugging format. There are
2280 no known issues with mixing object files and libraries with different
2281 debugging formats. Otherwise, this release of GCC should now match more
2282 of the configuration used in the stock FreeBSD configuration of GCC. In
2283 particular, @option{--enable-threads} is now configured by default.
2284 However, as a general user, do not attempt to replace the system
2285 compiler with this release. Known to bootstrap and check with good
2286 results on FreeBSD 4.8-STABLE and 5-CURRENT@. In the past, known to
2287 bootstrap and check with good results on FreeBSD 3.0, 3.4, 4.0, 4.2,
2288 4.3, 4.4, 4.5-STABLE@.
2290 In principle, @option{--enable-threads} is now compatible with
2291 @option{--enable-libgcj} on FreeBSD@. However, it has only been built
2292 and tested on @samp{i386-*-freebsd[45]} and @samp{alpha-*-freebsd[45]}.
2294 library may be incorrectly built (symbols are missing at link time).
2295 There is a rare timing-based startup hang (probably involves an
2296 assumption about the thread library). Multi-threaded boehm-gc (required for
2297 libjava) exposes severe threaded signal-handling bugs on FreeBSD before
2298 4.5-RELEASE@. Other CPU architectures
2299 supported by FreeBSD will require additional configuration tuning in, at
2300 the very least, both boehm-gc and libffi.
2302 Shared @file{libgcc_s.so} is now built and installed by default.
2307 @heading @anchor{h8300-hms}h8300-hms
2308 Renesas H8/300 series of processors.
2310 Please have a look at our @uref{binaries.html,,binaries page}.
2312 The calling convention and structure layout has changed in release 2.6.
2313 All code must be recompiled. The calling convention now passes the
2314 first three arguments in function calls in registers. Structures are no
2315 longer a multiple of 2 bytes.
2320 @heading @anchor{hppa*-hp-hpux*}hppa*-hp-hpux*
2321 HP-UX version 9 or older is no longer supported.
2323 We @emph{highly} recommend using gas/binutils 2.8 or newer on all hppa
2324 platforms; you may encounter a variety of problems when using the HP
2327 Specifically, @option{-g} does not work on HP-UX (since that system
2328 uses a peculiar debugging format which GCC does not know about), unless you
2329 use GAS and GDB and configure GCC with the
2330 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}} and
2331 @option{--with-as=@dots{}} options.
2333 If you wish to use the pa-risc 2.0 architecture support with a 32-bit
2334 runtime, you must use either the HP assembler, gas/binutils 2.11 or newer,
2336 @uref{ftp://sources.redhat.com/pub/binutils/snapshots,,snapshot of gas}.
2338 There are two default scheduling models for instructions. These are
2339 PROCESSOR_7100LC and PROCESSOR_8000. They are selected from the pa-risc
2340 architecture specified for the target machine when configuring.
2341 PROCESSOR_8000 is the default. PROCESSOR_7100LC is selected when
2342 the target is a @samp{hppa1*} machine.
2344 The PROCESSOR_8000 model is not well suited to older processors. Thus,
2345 it is important to completely specify the machine architecture when
2346 configuring if you want a model other than PROCESSOR_8000. The macro
2347 TARGET_SCHED_DEFAULT can be defined in BOOT_CFLAGS if a different
2348 default scheduling model is desired.
2350 More specific information to @samp{hppa*-hp-hpux*} targets follows.
2355 @heading @anchor{hppa*-hp-hpux10}hppa*-hp-hpux10
2357 For hpux10.20, we @emph{highly} recommend you pick up the latest sed patch
2358 @code{PHCO_19798} from HP@. HP has two sites which provide patches free of
2364 <a href="http://us.itrc.hp.com/service/home/home.do">US, Canada, Asia-Pacific, and
2368 @uref{http://us.itrc.hp.com/service/home/home.do,,} US, Canada, Asia-Pacific,
2372 @uref{http://europe.itrc.hp.com/service/home/home.do,,} Europe.
2375 The HP assembler on these systems has some problems. Most notably the
2376 assembler inserts timestamps into each object file it creates, causing
2377 the 3-stage comparison test to fail during a @samp{make bootstrap}.
2378 You should be able to continue by saying @samp{make all} after getting
2379 the failure from @samp{make bootstrap}.
2385 @heading @anchor{hppa*-hp-hpux11}hppa*-hp-hpux11
2387 GCC 3.0 and up support HP-UX 11. On 64-bit capable systems, there
2388 are two distinct ports. The @samp{hppa2.0w-hp-hpux11*} port generates
2389 code for the 32-bit pa-risc runtime architecture. It uses the HP
2390 linker. The @samp{hppa64-hp-hpux11*} port generates 64-bit code for the
2391 pa-risc 2.0 architecture. The script config.guess now selects the port
2392 type based on the type compiler detected during configuration. You must
2393 set your @env{PATH} or define @env{CC} so that configure finds an appropriate
2394 compiler for the initial bootstrap. Different prefixes must be used if
2395 both ports are to be installed on the same system.
2397 It is best to explicitly configure the @samp{hppa64-hp-hpux11*} target
2398 with the @option{--with-ld=@dots{}} option. We support both the HP
2399 and GNU linkers for this target. The two linkers require different
2400 link commands. Thus, it's not possible to switch linkers during a
2401 GCC build. This has been been reported to occur in a unified build
2402 of binutils and GCC.
2404 GCC 2.95.x is not supported under HP-UX 11 and cannot be used to
2405 compile GCC 3.0 and up. Refer to @uref{binaries.html,,binaries} for
2406 information about obtaining precompiled GCC binaries for HP-UX.
2408 You must use GNU binutils 2.11 or above with the 32-bit port. Thread
2409 support is not currently implemented, so @option{--enable-threads} does
2413 @item @uref{http://gcc.gnu.org/ml/gcc-prs/2002-01/msg00551.html}
2414 @item @uref{http://gcc.gnu.org/ml/gcc-bugs/2002-01/msg00663.html}
2417 GCC 3.3 and later support weak symbols on the 32-bit port using SOM
2418 secondary definition symbols. This feature is not enabled for earlier
2419 versions of HP-UX since there have been bugs in the linker support for
2420 secondary symbols. The HP linker patches @code{PHSS_26559} and
2421 @code{PHSS_24304} for HP-UX 11.00 and 11.11, respectively, correct the
2422 problem of linker core dumps creating C++ libraries. Earlier patches
2423 may work but they have not been tested.
2425 GCC 3.3 nows uses the ELF DT_INIT_ARRAY and DT_FINI_ARRAY capability
2426 to run initializers and finalizers on the 64-bit port. The feature
2427 requires CVS binutils as of January 2, 2003, or a subsequent release
2428 to correct a problem arising from HP's non-standard use of the .init
2429 and .fini sections. The 32-bit port uses the linker @option{+init}
2430 and @option{+fini} options. As with the support for secondary symbols,
2431 there have been bugs in the order in which these options are executed
2432 by the HP linker. So, again a recent linker patch is recommended.
2434 The HP assembler has many limitations and is not recommended for either
2435 the 32 or 64-bit ports. For example, it does not support weak symbols
2436 or alias definitions. As a result, explicit template instantiations
2437 are required when using C++. This will make it difficult if not
2438 impossible to build many C++ applications. You also can't generate
2439 debugging information when using the HP assembler with GCC.
2441 There are a number of issues to consider in selecting which linker to
2442 use with the 64-bit port. The GNU 64-bit linker can only create dynamic
2443 binaries. The @option{-static} option causes linking with archive
2444 libraries but doesn't produce a truly static binary. Dynamic binaries
2445 still require final binding by the dynamic loader to resolve a set of
2446 dynamic-loader-defined symbols. The default behavior of the HP linker
2447 is the same as the GNU linker. However, it can generate true 64-bit
2448 static binaries using the @option{+compat} option.
2450 The HP 64-bit linker doesn't support linkonce semantics. As a
2451 result, C++ programs have many more sections than they should.
2453 The GNU 64-bit linker has some issues with shared library support
2454 and exceptions. As a result, we only support libgcc in archive
2455 format. For similar reasons, dwarf2 unwind and exception support
2456 are disabled. The GNU linker also has problems creating binaries
2457 with @option{-static}. It doesn't provide stubs for internal
2458 calls to global functions in shared libraries, so these calls
2459 can't be overloaded.
2461 There are several possible approaches to building the distribution.
2462 Binutils can be built first using the HP tools. Then, the GCC
2463 distribution can be built. The second approach is to build GCC
2464 first using the HP tools, then build binutils, then rebuild GCC.
2465 There have been problems with various binary distributions, so
2466 it is best not to start from a binary distribution.
2468 Starting with GCC 3.4 an ISO C compiler is required to bootstrap.
2469 The bundled compiler supports only traditional C; you will need
2470 either HP's unbundled compiler, or a binary distribution of GCC@.
2472 This port still is undergoing significant development.
2477 @heading @anchor{i370-*-*}i370-*-*
2478 This port is very preliminary and has many known bugs. We hope to
2479 have a higher-quality port for this machine soon.
2484 @heading @anchor{*-*-linux-gnu}*-*-linux-gnu
2486 Versions of libstdc++-v3 starting with 3.2.1 require bugfixes present
2487 in glibc 2.2.5 and later. More information is available in the
2488 libstdc++-v3 documentation.
2490 If you use glibc 2.2 (or 2.1.9x), GCC 2.95.2 won't install
2491 out-of-the-box. You'll get compile errors while building @samp{libstdc++}.
2492 The patch @uref{glibc-2.2.patch,,glibc-2.2.patch}, that is to be
2493 applied in the GCC source tree, fixes the compatibility problems.
2495 Currently Glibc 2.2.3 (and older releases) and GCC 3.0 are out of sync
2496 since the latest exception handling changes for GCC@. Compiling glibc
2497 with GCC 3.0 will give a binary incompatible glibc and therefore cause
2498 lots of problems and might make your system completely unusable. This
2499 will definitely need fixes in glibc but might also need fixes in GCC@. We
2500 strongly advise to wait for glibc 2.2.4 and to read the release notes of
2501 glibc 2.2.4 whether patches for GCC 3.0 are needed. You can use glibc
2502 2.2.3 with GCC 3.0, just do not try to recompile it.
2507 @heading @anchor{ix86-*-linux*aout}i?86-*-linux*aout
2508 Use this configuration to generate @file{a.out} binaries on Linux-based
2509 GNU systems. This configuration is being superseded.
2514 @heading @anchor{ix86-*-linux*}i?86-*-linux*
2516 As of GCC 3.3, binutils 2.13.1 or later is required for this platform.
2517 See @uref{http://gcc.gnu.org/PR10877,,bug 10877} for more information.
2519 If you receive Signal 11 errors when building on GNU/Linux, then it is
2520 possible you have a hardware problem. Further information on this can be
2521 found on @uref{http://www.bitwizard.nl/sig11/,,www.bitwizard.nl}.
2526 @heading @anchor{ix86-*-sco3.2v5*}i?86-*-sco3.2v5*
2527 Use this for the SCO OpenServer Release 5 family of operating systems.
2529 Unlike earlier versions of GCC, the ability to generate COFF with this
2530 target is no longer provided.
2532 Earlier versions of GCC emitted DWARF 1 when generating ELF to allow
2533 the system debugger to be used. That support was too burdensome to
2534 maintain. GCC now emits only DWARF 2 for this target. This means you
2535 may use either the UDK debugger or GDB to debug programs built by this
2538 GCC is now only supported on releases 5.0.4 and later, and requires that
2539 you install Support Level Supplement OSS646B or later, and Support Level
2540 Supplement OSS631C or later. If you are using release 5.0.7 of
2541 OpenServer, you must have at least the first maintenance pack installed
2542 (this includes the relevant portions of OSS646). OSS646, also known as
2543 the "Execution Environment Update", provides updated link editors and
2544 assemblers, as well as updated standard C and math libraries. The C
2545 startup modules are also updated to support the System V gABI draft, and
2546 GCC relies on that behavior. OSS631 provides a collection of commonly
2547 used open source libraries, some of which GCC depends on (such as GNU
2548 gettext and zlib). SCO OpenServer Release 5.0.7 has all of this built
2549 in by default, but OSS631C and later also apply to that release. Please
2551 @uref{ftp://ftp.sco.com/pub/openserver5,,ftp://ftp.sco.com/pub/openserver5}
2552 for the latest versions of these (and other potentially useful)
2555 Although there is support for using the native assembler, it is
2556 recommended that you configure GCC to use the GNU assembler. You do
2557 this by using the flags
2558 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}}. You should
2559 use a modern version of GNU binutils. Version 2.13.2.1 was used for all
2560 testing. In general, only the @option{--with-gnu-as} option is tested.
2561 A modern bintuils (as well as a plethora of other development related
2562 GNU utilities) can be found in Support Level Supplement OSS658A, the
2563 "GNU Development Tools" package. See the SCO web and ftp sites for details.
2564 That package also contains the currently "officially supported" version of
2565 GCC, version 2.95.3. It is useful for bootstrapping this version.
2570 @heading @anchor{ix86-*-udk}i?86-*-udk
2572 This target emulates the SCO Universal Development Kit and requires that
2573 package be installed. (If it is installed, you will have a
2574 @file{/udk/usr/ccs/bin/cc} file present.) It's very much like the
2575 @samp{i?86-*-unixware7*} target
2576 but is meant to be used when hosting on a system where UDK isn't the
2577 default compiler such as OpenServer 5 or Unixware 2. This target will
2578 generate binaries that will run on OpenServer, Unixware 2, or Unixware 7,
2579 with the same warnings and caveats as the SCO UDK@.
2581 This target is a little tricky to build because we have to distinguish
2582 it from the native tools (so it gets headers, startups, and libraries
2583 from the right place) while making the tools not think we're actually
2584 building a cross compiler. The easiest way to do this is with a configure
2588 CC=/udk/usr/ccs/bin/cc @var{/your/path/to}/gcc/configure \
2589 --host=i686-pc-udk --target=i686-pc-udk --program-prefix=udk-
2592 @emph{You should substitute @samp{i686} in the above command with the appropriate
2593 processor for your host.}
2595 After the usual @samp{make bootstrap} and
2596 @samp{make install}, you can then access the UDK-targeted GCC
2597 tools by adding @command{udk-} before the commonly known name. For
2598 example, to invoke the C compiler, you would use @command{udk-gcc}.
2599 They will coexist peacefully with any native-target GCC tools you may
2606 @heading @anchor{ia64-*-linux}ia64-*-linux
2607 IA-64 processor (also known as IPF, or Itanium Processor Family)
2610 None of the following versions of GCC has an ABI that is compatible
2611 with any of the other versions in this list, with the exception that
2612 Red Hat 2.96 and Trillian 000171 are compatible with each other:
2613 3.1, 3.0.2, 3.0.1, 3.0, Red Hat 2.96, and Trillian 000717.
2614 This primarily affects C++ programs and programs that create shared libraries.
2615 GCC 3.1 or later is recommended for compiling linux, the kernel.
2616 As of version 3.1 GCC is believed to be fully ABI compliant, and hence no
2617 more major ABI changes are expected.
2622 @heading @anchor{ia64-*-hpux*}ia64-*-hpux*
2623 Building GCC on this target requires the GNU Assembler. The bundled HP
2624 assembler will not work. To prevent GCC from using the wrong assembler,
2625 the option @option{--with-gnu-as} may be necessary.
2627 The GCC libunwind library has not been ported to HPUX. This means that for
2628 GCC versions 3.2.3 and earlier, @option{--enable-libunwind-exceptions}
2629 is required to build GCC. For GCC 3.3 and later, this is the default.
2633 <!-- rs6000-ibm-aix*, powerpc-ibm-aix* -->
2635 @heading @anchor{*-ibm-aix*}*-ibm-aix*
2636 AIX version 3 or older is no longer supported.
2638 AIX Make frequently has problems with GCC makefiles. GNU Make 3.79.1 or
2639 newer is recommended to build on this platform.
2641 Errors involving @code{alloca} when building GCC generally are due
2642 to an incorrect definition of @code{CC} in the Makefile or mixing files
2643 compiled with the native C compiler and GCC@. During the stage1 phase of
2644 the build, the native AIX compiler @strong{must} be invoked as @command{cc}
2645 (not @command{xlc}). Once @command{configure} has been informed of
2646 @command{xlc}, one needs to use @samp{make distclean} to remove the
2647 configure cache files and ensure that @env{CC} environment variable
2648 does not provide a definition that will confuse @command{configure}.
2649 If this error occurs during stage2 or later, then the problem most likely
2650 is the version of Make (see above).
2652 The native @command{as} and @command{ld} are recommended for bootstrapping
2653 on AIX 4 and required for bootstrapping on AIX 5L. The GNU Assembler
2654 reports that it supports WEAK symbols on AIX 4, which causes GCC to try to
2655 utilize weak symbol functionality although it is not supported. The GNU
2656 Assembler and Linker do not support AIX 5L sufficiently to bootstrap GCC.
2657 The native AIX tools do interoperate with GCC@.
2659 Building @file{libstdc++.a} requires a fix for an AIX Assembler bug
2660 APAR IY26685 (AIX 4.3) or APAR IY25528 (AIX 5.1).
2662 @samp{libstdc++} in GCC 3.2 increments the major version number of the
2663 shared object and GCC installation places the @file{libstdc++.a}
2664 shared library in a common location which will overwrite the GCC 3.1
2665 version of the shared library. Applications either need to be
2666 re-linked against the new shared library or the GCC 3.1 version of the
2667 @samp{libstdc++} shared object needs to be available to the AIX
2668 runtime loader. The GCC 3.1 @samp{libstdc++.so.4} shared object can
2669 be installed for runtime dynamic loading using the following steps to
2670 set the @samp{F_LOADONLY} flag in the shared object for @emph{each}
2671 multilib @file{libstdc++.a} installed:
2673 Extract the shared object from each the GCC 3.1 @file{libstdc++.a}
2676 % ar -x libstdc++.a libstdc++.so.4
2679 Enable the @samp{F_LOADONLY} flag so that the shared object will be
2680 available for runtime dynamic loading, but not linking:
2682 % strip -e libstdc++.so.4
2685 Archive the runtime-only shared object in the GCC 3.2
2686 @file{libstdc++.a} archive:
2688 % ar -q libstdc++.a libstdc++.so.4
2691 Linking executables and shared libraries may produce warnings of
2692 duplicate symbols. The assembly files generated by GCC for AIX always
2693 have included multiple symbol definitions for certain global variable
2694 and function declarations in the original program. The warnings should
2695 not prevent the linker from producing a correct library or runnable
2698 AIX 4.3 utilizes a ``large format'' archive to support both 32-bit and
2699 64-bit object modules. The routines provided in AIX 4.3.0 and AIX 4.3.1
2700 to parse archive libraries did not handle the new format correctly.
2701 These routines are used by GCC and result in error messages during
2702 linking such as ``not a COFF file''. The version of the routines shipped
2703 with AIX 4.3.1 should work for a 32-bit environment. The @option{-g}
2704 option of the archive command may be used to create archives of 32-bit
2705 objects using the original ``small format''. A correct version of the
2706 routines is shipped with AIX 4.3.2 and above.
2708 Some versions of the AIX binder (linker) can fail with a relocation
2709 overflow severe error when the @option{-bbigtoc} option is used to link
2710 GCC-produced object files into an executable that overflows the TOC@. A fix
2711 for APAR IX75823 (OVERFLOW DURING LINK WHEN USING GCC AND -BBIGTOC) is
2712 available from IBM Customer Support and from its
2713 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
2714 website as PTF U455193.
2716 The AIX 4.3.2.1 linker (bos.rte.bind_cmds Level 4.3.2.1) will dump core
2717 with a segmentation fault when invoked by any version of GCC@. A fix for
2718 APAR IX87327 is available from IBM Customer Support and from its
2719 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
2720 website as PTF U461879. This fix is incorporated in AIX 4.3.3 and above.
2722 The initial assembler shipped with AIX 4.3.0 generates incorrect object
2723 files. A fix for APAR IX74254 (64BIT DISASSEMBLED OUTPUT FROM COMPILER FAILS
2724 TO ASSEMBLE/BIND) is available from IBM Customer Support and from its
2725 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
2726 website as PTF U453956. This fix is incorporated in AIX 4.3.1 and above.
2728 AIX provides National Language Support (NLS)@. Compilers and assemblers
2729 use NLS to support locale-specific representations of various data
2730 formats including floating-point numbers (e.g., @samp{.} vs @samp{,} for
2731 separating decimal fractions). There have been problems reported where
2732 GCC does not produce the same floating-point formats that the assembler
2733 expects. If one encounters this problem, set the @env{LANG}
2734 environment variable to @samp{C} or @samp{En_US}.
2736 By default, GCC for AIX 4.1 and above produces code that can be used on
2737 both Power or PowerPC processors.
2739 A default can be specified with the @option{-mcpu=@var{cpu_type}}
2740 switch and using the configure option @option{--with-cpu-@var{cpu_type}}.
2745 @heading @anchor{ip2k-*-elf}ip2k-*-elf
2746 Ubicom IP2022 micro controller.
2747 This configuration is intended for embedded systems.
2748 There are no standard Unix configurations.
2750 Use @samp{configure --target=ip2k-elf --enable-languages=c} to configure GCC@.
2755 @heading @anchor{m32r-*-elf}m32r-*-elf
2756 Renesas M32R processor.
2757 This configuration is intended for embedded systems.
2762 @heading @anchor{m6811-elf}m6811-elf
2763 Motorola 68HC11 family micro controllers. These are used in embedded
2764 applications. There are no standard Unix configurations.
2769 @heading @anchor{m6812-elf}m6812-elf
2770 Motorola 68HC12 family micro controllers. These are used in embedded
2771 applications. There are no standard Unix configurations.
2776 @heading @anchor{m68k-hp-hpux}m68k-hp-hpux
2777 HP 9000 series 300 or 400 running HP-UX@. HP-UX version 8.0 has a bug in
2778 the assembler that prevents compilation of GCC@. This
2779 bug manifests itself during the first stage of compilation, while
2780 building @file{libgcc2.a}:
2784 cc1: warning: `-g' option not supported on this version of GCC
2785 cc1: warning: `-g1' option not supported on this version of GCC
2786 ./xgcc: Internal compiler error: program as got fatal signal 11
2789 A patched version of the assembler is available as the file
2790 @uref{ftp://altdorf.ai.mit.edu/archive/cph/hpux-8.0-assembler}. If you
2791 have HP software support, the patch can also be obtained directly from
2792 HP, as described in the following note:
2795 This is the patched assembler, to patch SR#1653-010439, where the
2796 assembler aborts on floating point constants.
2798 The bug is not really in the assembler, but in the shared library
2799 version of the function ``cvtnum(3c)''. The bug on ``cvtnum(3c)'' is
2800 SR#4701-078451. Anyway, the attached assembler uses the archive
2801 library version of ``cvtnum(3c)'' and thus does not exhibit the bug.
2804 This patch is also known as PHCO_4484.
2806 In addition gdb does not understand that native HP-UX format, so
2807 you must use gas if you wish to use gdb.
2809 On HP-UX version 8.05, but not on 8.07 or more recent versions, the
2810 @command{fixproto} shell script triggers a bug in the system shell. If you
2811 encounter this problem, upgrade your operating system or use BASH (the
2812 GNU shell) to run @command{fixproto}. This bug will cause the fixproto
2813 program to report an error of the form:
2816 ./fixproto: sh internal 1K buffer overflow
2819 To fix this, you can also change the first line of the fixproto script
2829 @heading @anchor{mips-*-*}mips-*-*
2830 If on a MIPS system you get an error message saying ``does not have gp
2831 sections for all it's [sic] sectons [sic]'', don't worry about it. This
2832 happens whenever you use GAS with the MIPS linker, but there is not
2833 really anything wrong, and it is okay to use the output file. You can
2834 stop such warnings by installing the GNU linker.
2836 It would be nice to extend GAS to produce the gp tables, but they are
2837 optional, and there should not be a warning about their absence.
2839 The libstdc++ atomic locking routines for MIPS targets requires MIPS II
2840 and later. A patch went in just after the GCC 3.3 release to
2841 make @samp{mips*-*-*} use the generic implementation instead. You can also
2842 configure for @samp{mipsel-elf} as a workaround. The
2843 @samp{mips*-*-linux*} target continues to use the MIPS II routines. More
2844 work on this is expected in future releases.
2849 @heading @anchor{mips-sgi-irix5}mips-sgi-irix5
2851 This configuration has considerable problems, which will be fixed in a
2854 In order to compile GCC on an SGI running IRIX 5, the ``compiler_dev.hdr''
2855 subsystem must be installed from the IDO CD-ROM supplied by Silicon
2856 Graphics. It is also available for download from
2857 @uref{http://www.sgi.com/developers/devtools/apis/ido.html,,http://www.sgi.com/developers/devtools/apis/ido.html}.
2859 @samp{make compare} may fail on version 5 of IRIX unless you add
2860 @option{-save-temps} to @code{CFLAGS}. On these systems, the name of the
2861 assembler input file is stored in the object file, and that makes
2862 comparison fail if it differs between the @code{stage1} and
2863 @code{stage2} compilations. The option @option{-save-temps} forces a
2864 fixed name to be used for the assembler input file, instead of a
2865 randomly chosen name in @file{/tmp}. Do not add @option{-save-temps}
2866 unless the comparisons fail without that option. If you do you
2867 @option{-save-temps}, you will have to manually delete the @samp{.i} and
2868 @samp{.s} files after each series of compilations.
2870 If you use the MIPS C compiler to bootstrap, it may be necessary
2871 to increase its table size for switch statements with the
2872 @option{-Wf,-XNg1500} option. If you use the @option{-O2}
2873 optimization option, you also need to use @option{-Olimit 3000}.
2875 To enable debugging under IRIX 5, you must use GNU @command{as} 2.11.2
2877 and use the @option{--with-gnu-as} configure option when configuring GCC.
2878 GNU @command{as} is distributed as part of the binutils package.
2879 When using release 2.11.2, you need to apply a patch
2880 @uref{http://sources.redhat.com/ml/binutils/2001-07/msg00352.html,,http://sources.redhat.com/ml/binutils/2001-07/msg00352.html}
2881 which will be included in the next release of binutils.
2883 When building GCC, the build process loops rebuilding @command{cc1} over
2884 and over again. This happens on @samp{mips-sgi-irix5.2}, and possibly
2885 other platforms. It has been reported that this is a known bug in the
2886 @command{make} shipped with IRIX 5.2. We recommend you use GNU
2887 @command{make} instead of the vendor supplied @command{make} program;
2888 however, you may have success with @command{smake} on IRIX 5.2 if you do
2889 not have GNU @command{make} available.
2894 @heading @anchor{mips-sgi-irix6}mips-sgi-irix6
2896 If you are using IRIX @command{cc} as your bootstrap compiler, you must
2897 ensure that the N32 ABI is in use. To test this, compile a simple C
2898 file with @command{cc} and then run @command{file} on the
2899 resulting object file. The output should look like:
2902 test.o: ELF N32 MSB @dots{}
2908 test.o: ELF 32-bit MSB @dots{}
2914 test.o: ELF 64-bit MSB @dots{}
2917 then your version of @command{cc} uses the O32 or N64 ABI by default. You
2918 should set the environment variable @env{CC} to @samp{cc -n32}
2919 before configuring GCC@.
2921 If you want the resulting @command{gcc} to run on old 32-bit systems
2922 with the MIPS R4400 CPU, you need to ensure that only code for the mips3
2923 instruction set architecture (ISA) is generated. While GCC 3.x does
2924 this correctly, both GCC 2.95 and SGI's MIPSpro @command{cc} may change
2925 the ISA depending on the machine where GCC is built. Using one of them
2926 as the bootstrap compiler may result in mips4 code, which won't run at
2927 all on mips3-only systems. For the test program above, you should see:
2930 test.o: ELF N32 MSB mips-3 @dots{}
2936 test.o: ELF N32 MSB mips-4 @dots{}
2939 instead, you should set the environment variable @env{CC} to @samp{cc
2940 -n32 -mips3} or @samp{gcc -mips3} respectively before configuring GCC@.
2942 GCC on IRIX 6 is usually built to support both the N32 and N64 ABIs. If
2943 you build GCC on a system that doesn't have the N64 libraries installed,
2944 you need to configure with @option{--disable-multilib} so GCC doesn't
2945 try to use them. Look for @file{/usr/lib64/libc.so.1} to see if you
2946 have the 64-bit libraries installed.
2948 You must @emph{not} use GNU @command{as} (which isn't built anyway as of
2949 binutils 2.11.2) on IRIX 6 platforms; doing so will only cause problems.
2951 GCC does not currently support generating O32 ABI binaries in the
2952 @samp{mips-sgi-irix6} configurations. It is possible to create a GCC
2953 with O32 ABI only support by configuring it for the @samp{mips-sgi-irix5}
2954 target and using a patched GNU @command{as} 2.11.2 as documented in the
2955 @uref{#mips-sgi-irix5,,@samp{mips-sgi-irix5}} section above. Using the
2956 native assembler requires patches to GCC which will be included in a
2957 future release. It is
2958 expected that O32 ABI support will be available again in a future release.
2960 The @option{--enable-threads} option doesn't currently work, a patch is
2961 in preparation for a future release. The @option{--enable-libgcj}
2962 option is disabled by default: IRIX 6 uses a very low default limit
2963 (20480) for the command line length. Although libtool contains a
2964 workaround for this problem, at least the N64 @samp{libgcj} is known not
2965 to build despite this, running into an internal error of the native
2966 @command{ld}. A sure fix is to increase this limit (@samp{ncargs}) to
2967 its maximum of 262144 bytes. If you have root access, you can use the
2968 @command{systune} command to do this.
2970 GCC does not correctly pass/return structures which are
2971 smaller than 16 bytes and which are not 8 bytes. The problem is very
2972 involved and difficult to fix. It affects a number of other targets also,
2973 but IRIX 6 is affected the most, because it is a 64-bit target, and 4 byte
2974 structures are common. The exact problem is that structures are being padded
2975 at the wrong end, e.g.@: a 4 byte structure is loaded into the lower 4 bytes
2976 of the register when it should be loaded into the upper 4 bytes of the
2979 GCC is consistent with itself, but not consistent with the SGI C compiler
2980 (and the SGI supplied runtime libraries), so the only failures that can
2981 happen are when there are library functions that take/return such
2982 structures. There are very few such library functions. Currently this
2983 is known to affect @code{inet_ntoa}, @code{inet_lnaof},
2984 @code{inet_netof}, @code{inet_makeaddr}, and @code{semctl}. Until the
2985 bug is fixed, GCC contains workarounds for the known affected functions.
2987 See @uref{http://freeware.sgi.com/,,http://freeware.sgi.com/} for more
2988 information about using GCC on IRIX platforms.
2993 @heading @anchor{powerpc*-*-*}powerpc-*-*
2995 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
2996 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
3001 @heading @anchor{powerpc-*-darwin*}powerpc-*-darwin*
3002 PowerPC running Darwin (Mac OS X kernel).
3004 Pre-installed versions of Mac OS X may not include any developer tools,
3005 meaning that you will not be able to build GCC from source. Tool
3006 binaries are available at
3007 @uref{http://developer.apple.com/tools/compilers.html} (free
3008 registration required).
3010 The default stack limit of 512K is too small, which may cause compiles
3011 to fail with 'Bus error'. Set the stack larger, for instance
3012 by doing @samp{limit stack 800}. It's a good idea to use the GNU
3013 preprocessor instead of Apple's @file{cpp-precomp} during the first stage of
3014 bootstrapping; this is automatic when doing @samp{make bootstrap}, but
3015 to do it from the toplevel objdir you will need to say @samp{make
3016 CC='cc -no-cpp-precomp' bootstrap}.
3018 The version of GCC shipped by Apple typically includes a number of
3019 extensions not available in a standard GCC release. These extensions
3020 are generally specific to Mac programming.
3025 @heading @anchor{powerpc-*-elf}powerpc-*-elf, powerpc-*-sysv4
3026 PowerPC system in big endian mode, running System V.4.
3031 @heading @anchor{powerpc-*-linux-gnu*}powerpc-*-linux-gnu*
3034 @uref{ftp://ftp.kernel.org/pub/linux/devel/binutils,,binutils 2.13.90.0.10}
3035 or newer for a working GCC@.
3040 @heading @anchor{powerpc-*-netbsd*}powerpc-*-netbsd*
3041 PowerPC system in big endian mode running NetBSD@. To build the
3042 documentation you will need Texinfo version 4.2 (NetBSD 1.5.1 included
3043 Texinfo version 3.12).
3048 @heading @anchor{powerpc-*-eabisim}powerpc-*-eabisim
3049 Embedded PowerPC system in big endian mode for use in running under the
3055 @heading @anchor{powerpc-*-eabi}powerpc-*-eabi
3056 Embedded PowerPC system in big endian mode.
3061 @heading @anchor{powerpcle-*-elf}powerpcle-*-elf, powerpcle-*-sysv4
3062 PowerPC system in little endian mode, running System V.4.
3067 @heading @anchor{powerpcle-*-eabisim}powerpcle-*-eabisim
3068 Embedded PowerPC system in little endian mode for use in running under
3074 @heading @anchor{powerpcle-*-eabi}powerpcle-*-eabi
3075 Embedded PowerPC system in little endian mode.
3080 @heading @anchor{s390-*-linux*}s390-*-linux*
3081 S/390 system running Linux for S/390@.
3086 @heading @anchor{s390x-*-linux*}s390x-*-linux*
3087 zSeries system (64-bit) running Linux for zSeries@.
3092 @c Please use Solaris 2 to refer to all release of Solaris, starting
3093 @c with 2.0 until 2.6, 7, and 8. Solaris 1 was a marketing name for
3094 @c SunOS 4 releases which we don't use to avoid confusion. Solaris
3095 @c alone is too unspecific and must be avoided.
3096 @heading @anchor{*-*-solaris2*}*-*-solaris2*
3098 Sun does not ship a C compiler with Solaris 2. To bootstrap and install
3099 GCC you first have to install a pre-built compiler, see our
3100 @uref{binaries.html,,binaries page} for details.
3102 The Solaris 2 @command{/bin/sh} will often fail to configure
3103 @file{libstdc++-v3}, @file{boehm-gc} or @file{libjava}. We therefore
3104 recommend to use the following sequence of commands to bootstrap and
3108 % CONFIG_SHELL=/bin/ksh
3109 % export CONFIG_SHELL
3110 % @var{srcdir}/configure [@var{options}] [@var{target}]
3115 As explained in the @uref{build.html,,build} instructions, we recommend
3116 to use GNU make, which we call @command{gmake} here to distinguish it
3119 Solaris 2 comes with a number of optional OS packages. Some of these
3120 are needed to use GCC fully, namely @code{SUNWarc},
3121 @code{SUNWbtool}, @code{SUNWesu}, @code{SUNWhea}, @code{SUNWlibm},
3122 @code{SUNWsprot}, and @code{SUNWtoo}. If you did not install all
3123 optional packages when installing Solaris 2, you will need to verify that
3124 the packages that GCC needs are installed.
3126 To check whether an optional package is installed, use
3127 the @command{pkginfo} command. To add an optional package, use the
3128 @command{pkgadd} command. For further details, see the Solaris 2
3131 Trying to use the linker and other tools in
3132 @file{/usr/ucb} to install GCC has been observed to cause trouble.
3133 For example, the linker may hang indefinitely. The fix is to remove
3134 @file{/usr/ucb} from your @env{PATH}.
3136 The build process works more smoothly with the legacy Sun tools so, if you
3137 have @file{/usr/xpg4/bin} in your @env{PATH}, we recommend that you place
3138 @file{/usr/bin} before @file{/usr/xpg4/bin} for the duration of the build.
3140 All releases of GNU binutils prior to 2.11.2 have known bugs on this
3141 platform. We recommend the use of GNU binutils 2.11.2 or the vendor
3142 tools (Sun @command{as}, Sun @command{ld}).
3144 Sun bug 4296832 turns up when compiling X11 headers with GCC 2.95 or
3145 newer: @command{g++} will complain that types are missing. These headers assume
3146 that omitting the type means @code{int}; this assumption worked for C89 but
3147 is wrong for C++, and is now wrong for C99 also.
3149 @command{g++} accepts such (invalid) constructs with the option
3150 @option{-fpermissive}; it
3151 will assume that any missing type is @code{int} (as defined by C89).
3153 There are patches for Solaris 2.6 (105633-56 or newer for SPARC,
3154 106248-42 or newer for Intel), Solaris 7 (108376-21 or newer for SPARC,
3155 108377-20 for Intel), and Solaris 8 (108652-24 or newer for SPARC,
3156 108653-22 for Intel) that fix this bug.
3161 @heading @anchor{sparc-sun-solaris2*}sparc-sun-solaris2*
3163 When GCC is configured to use binutils 2.11.2 or later the binaries
3164 produced are smaller than the ones produced using Sun's native tools;
3165 this difference is quite significant for binaries containing debugging
3168 Sun @command{as} 4.x is broken in that it cannot cope with long symbol names.
3169 A typical error message might look similar to the following:
3172 /usr/ccs/bin/as: "/var/tmp/ccMsw135.s", line 11041: error:
3173 can't compute value of an expression involving an external symbol.
3176 This is Sun bug 4237974. This is fixed with patch 108908-02 for Solaris
3177 2.6 and has been fixed in later (5.x) versions of the assembler,
3178 starting with Solaris 7.
3180 Starting with Solaris 7, the operating system is capable of executing
3181 64-bit SPARC V9 binaries. GCC 3.1 and later properly supports
3182 this; the @option{-m64} option enables 64-bit code generation.
3183 However, if all you want is code tuned for the UltraSPARC CPU, you
3184 should try the @option{-mtune=ultrasparc} option instead, which produces
3185 code that, unlike full 64-bit code, can still run on non-UltraSPARC
3188 When configuring on a Solaris 7 or later system that is running a kernel
3189 that supports only 32-bit binaries, one must configure with
3190 @option{--disable-multilib}, since we will not be able to build the
3191 64-bit target libraries.
3196 @heading @anchor{sparc-sun-solaris2.7}sparc-sun-solaris2.7
3198 Sun patch 107058-01 (1999-01-13) for Solaris 7/SPARC triggers a bug in
3199 the dynamic linker. This problem (Sun bug 4210064) affects GCC 2.8
3200 and later, including all EGCS releases. Sun formerly recommended
3201 107058-01 for all Solaris 7 users, but around 1999-09-01 it started to
3202 recommend it only for people who use Sun's compilers.
3204 Here are some workarounds to this problem:
3207 Do not install Sun patch 107058-01 until after Sun releases a
3208 complete patch for bug 4210064. This is the simplest course to take,
3209 unless you must also use Sun's C compiler. Unfortunately 107058-01
3210 is preinstalled on some new Solaris 7-based hosts, so you may have to
3214 Copy the original, unpatched Solaris 7
3215 @command{/usr/ccs/bin/as} into
3216 @command{/usr/local/lib/gcc-lib/sparc-sun-solaris2.7/3.1/as},
3217 adjusting the latter name to fit your local conventions and software
3221 Install Sun patch 106950-03 (1999-05-25) or later. Nobody with
3222 both 107058-01 and 106950-03 installed has reported the bug with GCC
3223 and Sun's dynamic linker. This last course of action is riskiest,
3224 for two reasons. First, you must install 106950 on all hosts that
3225 run code generated by GCC; it doesn't suffice to install it only on
3226 the hosts that run GCC itself. Second, Sun says that 106950-03 is
3227 only a partial fix for bug 4210064, but Sun doesn't know whether the
3228 partial fix is adequate for GCC@. Revision -08 or later should fix
3229 the bug. The current (as of 2001-09-24) revision is -14, and is included in
3230 the Solaris 7 Recommended Patch Cluster.
3233 GCC 3.3 triggers a bug in version 5.0 Alpha 03/27/98 of the Sun assembler,
3234 which causes a bootstrap failure when linking the 64-bit shared version of
3235 libgcc. A typical error message is:
3238 ld: fatal: relocation error: R_SPARC_32: file libgcc/sparcv9/_muldi3.o:
3239 symbol <unknown>: offset 0xffffffff7ec133e7 is non-aligned.
3242 This bug has been fixed in the final 5.0 version of the assembler.
3247 @heading @anchor{sparc-*-linux*}sparc-*-linux*
3249 GCC versions 3.0 and higher require binutils 2.11.2 and glibc 2.2.4
3250 or newer on this platform. All earlier binutils and glibc
3251 releases mishandled unaligned relocations on @code{sparc-*-*} targets.
3257 @heading @anchor{sparc64-*-solaris2*}sparc64-*-solaris2*
3259 The following compiler flags must be specified in the configure
3260 step in order to bootstrap this target with the Sun compiler:
3263 % CC="cc -xildoff -xarch=v9" @var{srcdir}/configure [@var{options}] [@var{target}]
3266 @option{-xildoff} turns off the incremental linker, and @option{-xarch=v9}
3267 specifies the SPARC-V9 architecture to the Sun linker and assembler.
3272 @heading @anchor{sparcv9-*-solaris2*}sparcv9-*-solaris2*
3274 This is a synonym for sparc64-*-solaris2*.
3279 @heading @anchor{#*-*-sysv*}*-*-sysv*
3280 On System V release 3, you may get this error message
3284 ld fatal: failed to write symbol name @var{something}
3285 in strings table for file @var{whatever}
3288 This probably indicates that the disk is full or your ulimit won't allow
3289 the file to be as large as it needs to be.
3291 This problem can also result because the kernel parameter @code{MAXUMEM}
3292 is too small. If so, you must regenerate the kernel and make the value
3293 much larger. The default value is reported to be 1024; a value of 32768
3294 is said to work. Smaller values may also work.
3296 On System V, if you get an error like this,
3299 /usr/local/lib/bison.simple: In function `yyparse':
3300 /usr/local/lib/bison.simple:625: virtual memory exhausted
3304 that too indicates a problem with disk space, ulimit, or @code{MAXUMEM}.
3306 On a System V release 4 system, make sure @file{/usr/bin} precedes
3307 @file{/usr/ucb} in @code{PATH}. The @command{cc} command in
3308 @file{/usr/ucb} uses libraries which have bugs.
3313 @heading @anchor{vax-dec-ultrix}vax-dec-ultrix
3314 Don't try compiling with VAX C (@command{vcc}). It produces incorrect code
3315 in some cases (for example, when @code{alloca} is used).
3320 @heading @anchor{*-*-vxworks*}*-*-vxworks*
3321 Support for VxWorks is in flux. At present GCC supports @emph{only} the
3322 very recent VxWorks 5.5 (aka Tornado 2.2) release, and only on PowerPC.
3323 We welcome patches for other architectures supported by VxWorks 5.5.
3324 Support for VxWorks AE would also be welcome; we believe this is merely
3325 a matter of writing an appropriate ``configlette'' (see below). We are
3326 not interested in supporting older, a.out or COFF-based, versions of
3329 VxWorks comes with an older version of GCC installed in
3330 @file{@var{$WIND_BASE}/host}; we recommend you do not overwrite it.
3331 Choose an installation @var{prefix} entirely outside @var{$WIND_BASE}.
3332 Before running @command{configure}, create the directories @file{@var{prefix}}
3333 and @file{@var{prefix}/bin}. Link or copy the appropriate assembler,
3334 linker, etc. into @file{@var{prefix}/bin}, and set your @var{PATH} to
3335 include that directory while running both @command{configure} and
3338 You must give @command{configure} the
3339 @option{--with-headers=@var{$WIND_BASE}/target/h} switch so that it can
3340 find the VxWorks system headers. Since VxWorks is a cross compilation
3341 target only, you must also specify @option{--target=@var{target}}.
3342 @command{configure} will attempt to create the directory
3343 @file{@var{prefix}/@var{target}/sys-include} and copy files into it;
3344 make sure the user running @command{configure} has sufficient privilege
3347 GCC's exception handling runtime requires a special ``configlette''
3348 module, @file{contrib/gthr_supp_vxw_5x.c}. Follow the instructions in
3349 that file to add the module to your kernel build. (Future versions of
3350 VxWorks will incorporate this module.)
3355 @heading @anchor{xtensa-*-elf}xtensa-*-elf
3357 This target is intended for embedded Xtensa systems using the
3358 @samp{newlib} C library. It uses ELF but does not support shared
3359 objects. Designed-defined instructions specified via the
3360 Tensilica Instruction Extension (TIE) language are only supported
3361 through inline assembly.
3363 The Xtensa configuration information must be specified prior to
3364 building GCC@. The @file{include/xtensa-config.h} header
3365 file contains the configuration information. If you created your
3366 own Xtensa configuration with the Xtensa Processor Generator, the
3367 downloaded files include a customized copy of this header file,
3368 which you can use to replace the default header file.
3373 @heading @anchor{xtensa-*-linux*}xtensa-*-linux*
3375 This target is for Xtensa systems running GNU/Linux. It supports ELF
3376 shared objects and the GNU C library (glibc). It also generates
3377 position-independent code (PIC) regardless of whether the
3378 @option{-fpic} or @option{-fPIC} options are used. In other
3379 respects, this target is the same as the
3380 @uref{#xtensa-*-elf,,@samp{xtensa-*-elf}} target.
3385 @heading @anchor{windows}Microsoft Windows (32-bit)
3387 A port of GCC 2.95.2 and 3.x is included with the
3388 @uref{http://www.cygwin.com/,,Cygwin environment}.
3390 Current (as of early 2001) snapshots of GCC will build under Cygwin
3391 without modification.
3393 GCC does not currently build with Microsoft's C++ compiler and there
3394 are no plans to make it do so.
3399 @heading @anchor{os2}OS/2
3401 GCC does not currently support OS/2. However, Andrew Zabolotny has been
3402 working on a generic OS/2 port with pgcc. The current code can be found
3403 at @uref{http://www.goof.com/pcg/os2/,,http://www.goof.com/pcg/os2/}.
3405 An older copy of GCC 2.8.1 is included with the EMX tools available at
3406 @uref{ftp://ftp.leo.org/pub/comp/os/os2/leo/devtools/emx+gcc/,,
3407 ftp://ftp.leo.org/pub/comp/os/os2/leo/devtools/emx+gcc/}.
3412 @heading @anchor{older}Older systems
3414 GCC contains support files for many older (1980s and early
3415 1990s) Unix variants. For the most part, support for these systems
3416 has not been deliberately removed, but it has not been maintained for
3417 several years and may suffer from bitrot.
3419 Starting with GCC 3.1, each release has a list of ``obsoleted'' systems.
3420 Support for these systems is still present in that release, but
3421 @command{configure} will fail unless the @option{--enable-obsolete}
3422 option is given. Unless a maintainer steps forward, support for these
3423 systems will be removed from the next release of GCC@.
3425 Support for old systems as hosts for GCC can cause problems if the
3426 workarounds for compiler, library and operating system bugs affect the
3427 cleanliness or maintainability of the rest of GCC@. In some cases, to
3428 bring GCC up on such a system, if still possible with current GCC, may
3429 require first installing an old version of GCC which did work on that
3430 system, and using it to compile a more recent GCC, to avoid bugs in the
3431 vendor compiler. Old releases of GCC 1 and GCC 2 are available in the
3432 @file{old-releases} directory on the @uref{../mirrors.html,,GCC mirror
3433 sites}. Header bugs may generally be avoided using
3434 @command{fixincludes}, but bugs or deficiencies in libraries and the
3435 operating system may still cause problems.
3437 Support for older systems as targets for cross-compilation is less
3438 problematic than support for them as hosts for GCC; if an enthusiast
3439 wishes to make such a target work again (including resurrecting any of
3440 the targets that never worked with GCC 2, starting from the last CVS
3441 version before they were removed), patches
3442 @uref{../contribute.html,,following the usual requirements} would be
3443 likely to be accepted, since they should not affect the support for more
3446 For some systems, old versions of GNU binutils may also be useful,
3447 and are available from @file{pub/binutils/old-releases} on
3448 @uref{http://sources.redhat.com/mirrors.html,,sources.redhat.com mirror sites}.
3450 Some of the information on specific systems above relates to
3451 such older systems, but much of the information
3452 about GCC on such systems (which may no longer be applicable to
3453 current GCC) is to be found in the GCC texinfo manual.
3458 @heading @anchor{elf_targets}all ELF targets (SVR4, Solaris 2, etc.)
3460 C++ support is significantly better on ELF targets if you use the
3461 @uref{./configure.html#with-gnu-ld,,GNU linker}; duplicate copies of
3462 inlines, vtables and template instantiations will be discarded
3471 @uref{./index.html,,Return to the GCC Installation page}
3475 @c ***Old documentation******************************************************
3477 @include install-old.texi
3483 @uref{./index.html,,Return to the GCC Installation page}
3487 @c ***GFDL********************************************************************
3495 @uref{./index.html,,Return to the GCC Installation page}
3499 @c ***************************************************************************
3500 @c Part 6 The End of the Document
3502 @comment node-name, next, previous, up
3503 @node Concept Index, , GNU Free Documentation License, Top
3507 @unnumbered Concept Index