1 \input texinfo.tex @c -*-texinfo-*-
4 @setfilename install.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
18 @settitle Downloading GCC
21 @settitle Installing GCC: Configuration
24 @settitle Installing GCC: Building
27 @settitle Installing GCC: Testing
29 @ifset finalinstallhtml
30 @settitle Installing GCC: Final installation
33 @settitle Installing GCC: Binaries
36 @settitle Installing GCC: Old documentation
39 @settitle Installing GCC: GNU Free Documentation License
42 @c Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997, 1998,
43 @c 1999, 2000, 2001, 2002 Free Software Foundation, Inc.
44 @c *** Converted to texinfo by Dean Wakerley, dean@wakerley.com
46 @c Include everything if we're not making html
60 @c Part 2 Summary Description and Copyright
61 @macro copyrightnotice
62 Copyright @copyright{} 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997, 1998,
63 1999, 2000, 2001, 2002 Free Software Foundation, Inc.
65 Permission is granted to copy, distribute and/or modify this document
66 under the terms of the GNU Free Documentation License, Version 1.1 or
67 any later version published by the Free Software Foundation; with no
68 Invariant Sections, the Front-Cover texts being (a) (see below), and
69 with the Back-Cover Texts being (b) (see below). A copy of the
70 license is included in the section entitled ``@uref{./gfdl.html,,GNU
71 Free Documentation License}''.
73 (a) The FSF's Front-Cover Text is:
77 (b) The FSF's Back-Cover Text is:
79 You have freedom to copy and modify this GNU Manual, like GNU
80 software. Copies published by the Free Software Foundation raise
81 funds for GNU development.
87 @c Part 3 Titlepage and Copyright
90 @comment The title is printed in a large font.
91 @center @titlefont{Installing GCC}
93 @c The following two commands start the copyright page.
95 @vskip 0pt plus 1filll
99 @c Part 4 Top node and Master Menu
102 @comment node-name, next, Previous, up
105 * Installing GCC:: This document describes the generic installation
106 procedure for GCC as well as detailing some target
107 specific installation instructions.
109 * Specific:: Host/target specific installation notes for GCC.
110 * Binaries:: Where to get pre-compiled binaries.
112 * Old:: Old installation documentation.
114 * GNU Free Documentation License:: How you can copy and share this manual.
115 * Concept Index:: This index has two entries.
119 @c Part 5 The Body of the Document
120 @c ***Installing GCC**********************************************************
122 @comment node-name, next, previous, up
123 @node Installing GCC, Binaries, , Top
127 <h1 align="center">Installing GCC</h1>
130 @chapter Installing GCC
133 The latest version of this document is always available at
134 @uref{http://gcc.gnu.org/install/,,http://gcc.gnu.org/install/}.
136 This document describes the generic installation procedure for GCC as well
137 as detailing some target specific installation instructions.
139 GCC includes several components that previously were separate distributions
140 with their own installation instructions. This document supersedes all
141 package specific installation instructions.
143 @emph{Before} starting the build/install procedure please check the
145 @ref{Specific, host/target specific installation notes}.
148 @uref{specific.html,,host/target specific installation notes}.
150 We recommend you browse the entire generic installation instructions before
153 Lists of successful builds for released versions of GCC are
154 available at our web pages for
155 @uref{http://gcc.gnu.org/gcc-3.1/buildstat.html,,3.1},
156 @uref{http://gcc.gnu.org/gcc-3.0/buildstat.html,,3.0}
158 @uref{http://gcc.gnu.org/gcc-2.95/buildstat.html,,2.95}.
159 These lists are updated as new information becomes available.
161 The installation procedure itself is broken into five steps.
165 * Downloading the source::
168 * Testing:: (optional)
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 ***Downloading the source**************************************************
213 @comment node-name, next, previous, up
214 @node Downloading the source, Configuration, , Installing GCC
218 <h1 align="center">Downloading GCC</h1>
221 @chapter Downloading GCC
223 @cindex Downloading GCC
224 @cindex Downloading the Source
226 GCC is distributed via @uref{http://gcc.gnu.org/cvs.html,,CVS} and FTP
227 tarballs compressed with @command{gzip} or
228 @command{bzip2}. It is possible to download a full distribution or specific
231 Please refer to our @uref{http://gcc.gnu.org/releases.html,,releases web page}
232 for information on how to obtain GCC@.
234 The full distribution includes the C, C++, Objective-C, Fortran, Java,
235 and Ada (in case of GCC 3.1 and later) compilers. The full distribution
236 also includes runtime libraries for C++, Objective-C, Fortran, and Java.
237 In GCC 3.0 and later versions, GNU compiler testsuites are also included
238 in the full distribution.
240 If you choose to download specific components, you must download the core
241 GCC distribution plus any language specific distributions you wish to
242 use. The core distribution includes the C language front end as well as the
243 shared components. Each language has a tarball which includes the language
244 front end as well as the language runtime (when appropriate).
246 Unpack the core distribution as well as any language specific
247 distributions in the same directory.
249 If you also intend to build binutils (either to upgrade an existing
250 installation or for use in place of the corresponding tools of your
251 OS), unpack the binutils distribution either in the same directory or
252 a separate one. In the latter case, add symbolic links to any
253 components of the binutils you intend to build alongside the compiler
254 (@file{bfd}, @file{binutils}, @file{gas}, @file{gprof}, @file{ld},
255 @file{opcodes}, @dots{}) to the directory containing the GCC sources.
262 @uref{./index.html,,Return to the GCC Installation page}
266 @c ***Configuration***********************************************************
268 @comment node-name, next, previous, up
269 @node Configuration, Building, Downloading the source, Installing GCC
273 <h1 align="center">Installing GCC: Configuration</h1>
276 @chapter Installing GCC: Configuration
278 @cindex Configuration
279 @cindex Installing GCC: Configuration
281 Like most GNU software, GCC must be configured before it can be built.
282 This document describes the recommended configuration procedure
283 for both native and cross targets.
285 We use @var{srcdir} to refer to the toplevel source directory for
286 GCC; we use @var{objdir} to refer to the toplevel build/object directory.
288 If you obtained the sources via CVS, @var{srcdir} must refer to the top
289 @file{gcc} directory, the one where the @file{MAINTAINERS} can be found,
290 and not its @file{gcc} subdirectory, otherwise the build will fail.
292 First, we @strong{highly} recommend that GCC be built into a
293 separate directory than the sources which does @strong{not} reside
294 within the source tree. This is how we generally build GCC; building
295 where @var{srcdir} == @var{objdir} should still work, but doesn't
296 get extensive testing; building where @var{objdir} is a subdirectory
297 of @var{srcdir} is unsupported.
299 If you have previously built GCC in the same directory for a
300 different target machine, do @samp{make distclean} to delete all files
301 that might be invalid. One of the files this deletes is
302 @file{Makefile}; if @samp{make distclean} complains that @file{Makefile}
303 does not exist, it probably means that the directory is already suitably
304 clean. However, with the recommended method of building in a separate
305 @var{objdir}, you should simply use a different @var{objdir} for each
308 Second, when configuring a native system, either @command{cc} or
309 @command{gcc} must be in your path or you must set @env{CC} in
310 your environment before running configure. Otherwise the configuration
313 Note that the bootstrap compiler and the resulting GCC must be link
314 compatible, else the bootstrap will fail with linker errors about
315 incompatible object file formats. Several multilibed targets are
316 affected by this requirement, see
318 @ref{Specific, host/target specific installation notes}.
321 @uref{specific.html,,host/target specific installation notes}.
329 % @var{srcdir}/configure [@var{options}] [@var{target}]
333 @heading Target specification
336 GCC has code to correctly determine the correct value for @var{target}
337 for nearly all native systems. Therefore, we highly recommend you not
338 provide a configure target when configuring a native compiler.
341 @var{target} must be specified as @option{--target=@var{target}}
342 when configuring a cross compiler; examples of valid targets would be
343 i960-rtems, m68k-coff, sh-elf, etc.
346 Specifying just @var{target} instead of @option{--target=@var{target}}
347 implies that the host defaults to @var{target}.
351 @heading Options specification
353 Use @var{options} to override several configure time options for
354 GCC@. A list of supported @var{options} follows; @command{configure
355 --help} may list other options, but those not listed below may not
356 work and should not normally be used.
359 @item --prefix=@var{dirname}
360 Specify the toplevel installation
361 directory. This is the recommended way to install the tools into a directory
362 other than the default. The toplevel installation directory defaults to
365 We @strong{highly} recommend against @var{dirname} being the same or a
366 subdirectory of @var{objdir} or vice versa.
368 These additional options control where certain parts of the distribution
369 are installed. Normally you should not need to use these options.
372 @item --exec-prefix=@var{dirname}
373 Specify the toplevel installation directory for architecture-dependent
374 files. The default is @file{@var{prefix}}.
376 @item --bindir=@var{dirname}
377 Specify the installation directory for the executables called by users
378 (such as @command{gcc} and @command{g++}). The default is
379 @file{@var{exec-prefix}/bin}.
381 @item --libdir=@var{dirname}
382 Specify the installation directory for object code libraries and
383 internal parts of GCC@. The default is @file{@var{exec-prefix}/lib}.
385 @item --with-slibdir=@var{dirname}
386 Specify the installation directory for the shared libgcc library. The
387 default is @file{@var{libdir}}.
389 @item --infodir=@var{dirname}
390 Specify the installation directory for documentation in info format.
391 The default is @file{@var{prefix}/info}.
393 @item --mandir=@var{dirname}
394 Specify the installation directory for manual pages. The default is
395 @file{@var{prefix}/man}. (Note that the manual pages are only extracts from
396 the full GCC manuals, which are provided in Texinfo format. The
397 @command{g77} manpage is unmaintained and may be out of date; the others
398 are derived by an automatic conversion process from parts of the full
401 @item --with-gxx-include-dir=@var{dirname}
403 the installation directory for G++ header files. The default is
404 @file{@var{prefix}/include/g++-v3}.
408 @item --program-prefix=@var{prefix}
409 GCC supports some transformations of the names of its programs when
410 installing them. This option prepends @var{prefix} to the names of
411 programs to install in @var{bindir} (see above). For example, specifying
412 @option{--program-prefix=foo-} would result in @samp{gcc}
413 being installed as @file{/usr/local/bin/foo-gcc}.
415 @item --program-suffix=@var{suffix}
416 Appends @var{suffix} to the names of programs to install in @var{bindir}
417 (see above). For example, specifying @option{--program-suffix=-3.1}
418 would result in @samp{gcc} being installed as
419 @file{/usr/local/bin/gcc-3.1}.
421 @item --program-transform-name=@var{pattern}
422 Applies the @samp{sed} script @var{pattern} to be applied to the names
423 of programs to install in @var{bindir} (see above). @var{pattern} has to
424 consist of one or more basic @samp{sed} editing commands, separated by
425 semicolons. For example, if you want the @samp{gcc} program name to be
426 transformed to the installed program @file{/usr/local/bin/myowngcc} and
427 the @samp{g++} program name to be transformed to
428 @file{/usr/local/bin/gspecial++} without changing other program names,
429 you could use the pattern
430 @option{--program-transform-name='s/^gcc$/myowngcc/; s/^g++$/gspecial++/'}
431 to achieve this effect.
433 All three options can be combined and used together, resulting in more
434 complex conversion patterns. As a basic rule, @var{prefix} (and
435 @var{suffix}) are prepended (appended) before further transformations
436 can happen with a special transformation script @var{pattern}.
438 As currently implemented, this options only take effect for native
439 builds; cross compiler binaries' names are not transformed even when a
440 transformation is explicitly asked for by one of this options.
442 For native builds, some of the installed programs are also installed
443 with the target alias in front of their name, as in
444 @samp{i686-pc-linux-gnu-gcc}. All of the above transformations happen
445 before the target alias is prepended to the name - so, specifying
446 @option{--program-prefix=foo-} and @option{program-suffix=-3.1}, the
447 resulting binary would be installed as
448 @file{/usr/local/bin/i686-pc-linux-gnu-foo-gcc-3.1}.
450 As a last shortcoming, none of the installed Ada programs are
451 transformed yet, which will be fixed in some time.
453 @item --with-local-prefix=@var{dirname}
455 installation directory for local include files. The default is
456 @file{/usr/local}. Specify this option if you want the compiler to
457 search directory @file{@var{dirname}/include} for locally installed
458 header files @emph{instead} of @file{/usr/local/include}.
460 You should specify @option{--with-local-prefix} @strong{only} if your
461 site has a different convention (not @file{/usr/local}) for where to put
464 The default value for @option{--with-local-prefix} is @file{/usr/local}
465 regardless of the value of @option{--prefix}. Specifying
466 @option{--prefix} has no effect on which directory GCC searches for
467 local header files. This may seem counterintuitive, but actually it is
470 The purpose of @option{--prefix} is to specify where to @emph{install
471 GCC}. The local header files in @file{/usr/local/include}---if you put
472 any in that directory---are not part of GCC@. They are part of other
473 programs---perhaps many others. (GCC installs its own header files in
474 another directory which is based on the @option{--prefix} value.)
476 @strong{Do not} specify @file{/usr} as the @option{--with-local-prefix}!
477 The directory you use for @option{--with-local-prefix} @strong{must not}
478 contain any of the system's standard header files. If it did contain
479 them, certain programs would be miscompiled (including GNU Emacs, on
480 certain targets), because this would override and nullify the header
481 file corrections made by the @code{fixincludes} script.
483 Indications are that people who use this option use it based on mistaken
484 ideas of what it is for. People use it as if it specified where to
485 install part of GCC@. Perhaps they make this assumption because
486 installing GCC creates the directory.
488 @item --enable-shared[=@var{package}[,@dots{}]]
489 Build shared versions of libraries, if shared libraries are supported on
490 the target platform. Unlike GCC 2.95.x and earlier, shared libraries
491 are enabled by default on all platforms that support shared libraries,
492 except for @samp{libobjc} which is built as a static library only by
495 If a list of packages is given as an argument, build shared libraries
496 only for the listed packages. For other packages, only static libraries
497 will be built. Package names currently recognized in the GCC tree are
498 @samp{libgcc} (also known as @samp{gcc}), @samp{libstdc++} (not
499 @samp{libstdc++-v3}), @samp{libffi}, @samp{zlib}, @samp{boehm-gc} and
500 @samp{libjava}. Note that @samp{libobjc} does not recognize itself by
501 any name, so, if you list package names in @option{--enable-shared},
502 you will only get static Objective-C libraries. @samp{libf2c} and
503 @samp{libiberty} do not support shared libraries at all.
505 Use @option{--disable-shared} to build only static libraries. Note that
506 @option{--disable-shared} does not accept a list of package names as
507 argument, only @option{--enable-shared} does.
509 @item @anchor{with-gnu-as}--with-gnu-as
510 Specify that the compiler should assume that the
511 assembler it finds is the GNU assembler. However, this does not modify
512 the rules to find an assembler and will result in confusion if found
513 assembler is not actually the GNU assembler. (Confusion will also
514 result if the compiler finds the GNU assembler but has not been
515 configured with @option{--with-gnu-as}.) If you have more than one
516 assembler installed on your system, you may want to use this option in
517 connection with @option{--with-as=@var{pathname}}.
519 The systems where it makes a difference whether you use the GNU assembler are
520 @samp{hppa1.0-@var{any}-@var{any}}, @samp{hppa1.1-@var{any}-@var{any}},
521 @samp{i386-@var{any}-sysv}, @samp{i386-@var{any}-isc},
522 @samp{i860-@var{any}-bsd}, @samp{m68k-bull-sysv},
523 @samp{m68k-hp-hpux}, @samp{m68k-sony-bsd},
524 @samp{m68k-altos-sysv}, @samp{m68000-hp-hpux},
525 @samp{m68000-att-sysv}, @samp{@var{any}-lynx-lynxos},
526 and @samp{mips-@var{any}}.
527 On any other system, @option{--with-gnu-as} has no effect.
529 On the systems listed above (except for the HP-PA, for ISC on the
530 386, and for @samp{mips-sgi-irix5.*}), if you use the GNU assembler,
531 you should also use the GNU linker (and specify @option{--with-gnu-ld}).
533 @item --with-as=@var{pathname}
535 compiler should use the assembler pointed to by @var{pathname}, rather
536 than the one found by the standard rules to find an assembler, which
541 @file{@var{exec_prefix}/lib/gcc-lib/@var{target}/@var{version}}
542 directory, where @var{exec_prefix} defaults to @var{prefix} which
543 defaults to @file{/usr/local} unless overridden by the
544 @option{--prefix=@var{pathname}} switch described above. @var{target} is the
545 target system triple, such as @samp{sparc-sun-solaris2.7}, and
546 @var{version} denotes the GCC version, such as 3.0.
548 Check operating system specific directories (e.g.@: @file{/usr/ccs/bin} on
551 Note that these rules do not check for the value of @env{PATH}. You may
552 want to use @option{--with-as} if no assembler is installed in the
553 directories listed above, or if you have multiple assemblers installed
554 and want to choose one that is not found by the above rules.
556 @item @anchor{with-gnu-ld}--with-gnu-ld
557 Same as @uref{#with-gnu-as,,@option{--with-gnu-as}}
561 @item --with-ld=@var{pathname}
563 @option{--with-as}, but for the linker.
566 Specify that stabs debugging
567 information should be used instead of whatever format the host normally
568 uses. Normally GCC uses the same debug format as the host system.
570 On MIPS based systems and on Alphas, you must specify whether you want
571 GCC to create the normal ECOFF debugging format, or to use BSD-style
572 stabs passed through the ECOFF symbol table. The normal ECOFF debug
573 format cannot fully handle languages other than C@. BSD stabs format can
574 handle other languages, but it only works with the GNU debugger GDB@.
576 Normally, GCC uses the ECOFF debugging format by default; if you
577 prefer BSD stabs, specify @option{--with-stabs} when you configure GCC@.
579 No matter which default you choose when you configure GCC, the user
580 can use the @option{-gcoff} and @option{-gstabs+} options to specify explicitly
581 the debug format for a particular compilation.
583 @option{--with-stabs} is meaningful on the ISC system on the 386, also, if
584 @option{--with-gas} is used. It selects use of stabs debugging
585 information embedded in COFF output. This kind of debugging information
586 supports C++ well; ordinary COFF debugging information does not.
588 @option{--with-stabs} is also meaningful on 386 systems running SVR4. It
589 selects use of stabs debugging information embedded in ELF output. The
590 C++ compiler currently (2.6.0) does not support the DWARF debugging
591 information normally used on 386 SVR4 platforms; stabs provide a
592 workable alternative. This requires gas and gdb, as the normal SVR4
593 tools can not generate or interpret stabs.
595 @item --disable-multilib
596 Specify that multiple target
597 libraries to support different target variants, calling
598 conventions, etc should not be built. The default is to build a
599 predefined set of them.
601 Some targets provide finer-grained control over which multilibs are built
602 (e.g., @option{--disable-softfloat}):
609 fpu, 26bit, underscore, interwork, biendian, nofmult.
612 softfloat, m68881, m68000, m68020.
615 single-float, biendian, softfloat.
617 @item powerpc*-*-*, rs6000*-*-*
618 aix64, pthread, softfloat, powercpu, powerpccpu, powerpcos, biendian,
623 @item --enable-threads
624 Specify that the target
625 supports threads. This affects the Objective-C compiler and runtime
626 library, and exception handling for other languages like C++ and Java.
627 On some systems, this is the default.
629 In general, the best (and, in many cases, the only known) threading
630 model available will be configured for use. Beware that on some
631 systems, gcc has not been taught what threading models are generally
632 available for the system. In this case, @option{--enable-threads} is an
633 alias for @option{--enable-threads=single}.
635 @item --disable-threads
636 Specify that threading support should be disabled for the system.
637 This is an alias for @option{--enable-threads=single}.
639 @item --enable-threads=@var{lib}
641 @var{lib} is the thread support library. This affects the Objective-C
642 compiler and runtime library, and exception handling for other languages
643 like C++ and Java. The possibilities for @var{lib} are:
651 Generic MACH thread support, known to work on NeXTSTEP@. (Please note
652 that the file needed to support this configuration, @file{gthr-mach.h}, is
653 missing and thus this setting will cause a known bootstrap failure.)
655 This is an alias for @samp{single}.
657 Generic POSIX thread support.
659 Same as @samp{posix} on arm*-*-linux*, *-*-chorusos* and *-*-freebsd*
660 only. A future release of gcc might remove this alias or extend it
663 RTEMS thread support.
665 Disable thread support, should work for all platforms.
667 Sun Solaris 2 thread support.
669 VxWorks thread support.
671 Microsoft Win32 API thread support.
674 @item --with-cpu=@var{cpu}
675 Specify which cpu variant the
676 compiler should generate code for by default. This is currently
677 only supported on the some ports, specifically arm, powerpc, and
678 SPARC@. If configure does not recognize the model name (e.g.@: arm700,
679 603e, or ultrasparc) you provide, please check the configure script
680 for a complete list of supported models.
682 @item --enable-altivec
683 Specify that the target supports AltiVec vector enhancements. This
684 option will adjust the ABI for AltiVec enhancements, as well as generate
685 AltiVec code when appropriate. This option is only available for
688 @item --enable-target-optspace
690 libraries should be optimized for code space instead of code speed.
691 This is the default for the m32r platform.
694 Specify that a user visible @command{cpp} program should not be installed.
696 @item --with-cpp-install-dir=@var{dirname}
697 Specify that the user visible @command{cpp} program should be installed
698 in @file{@var{prefix}/@var{dirname}/cpp}, in addition to @var{bindir}.
700 @item --enable-maintainer-mode
702 regenerate the GCC master message catalog @file{gcc.pot} are normally
703 disabled. This is because it can only be rebuilt if the complete source
704 tree is present. If you have changed the sources and want to rebuild the
705 catalog, configuring with @option{--enable-maintainer-mode} will enable
706 this. Note that you need a recent version of the @code{gettext} tools
709 @item --enable-version-specific-runtime-libs
711 that runtime libraries should be installed in the compiler specific
712 subdirectory (@file{@var{libsubdir}}) rather than the usual places. In
713 addition, @samp{libstdc++}'s include files will be installed in
714 @file{@var{libsubdir}/include/g++} unless you overruled it by using
715 @option{--with-gxx-include-dir=@var{dirname}}. Using this option is
716 particularly useful if you intend to use several versions of GCC in
717 parallel. This is currently supported by @samp{libf2c} and
718 @samp{libstdc++}, and is the default for @samp{libobjc} which cannot be
719 changed in this case.
721 @item --enable-languages=@var{lang1},@var{lang2},@dots{}
722 Specify that only a particular subset of compilers and
723 their runtime libraries should be built. For a list of valid values for
724 @var{langN} you can issue the following command in the
725 @file{gcc} directory of your GCC source tree:@*
727 grep language= */config-lang.in
729 Currently, you can use any of the following:
730 @code{ada}, @code{c}, @code{c++}, @code{f77}, @code{java}, @code{objc}.
731 Building the Ada compiler has special requirements, see below.@*
732 If you do not pass this flag, all languages available in the @file{gcc}
733 sub-tree will be configured. Re-defining @code{LANGUAGES} when calling
734 @samp{make bootstrap} @strong{does not} work anymore, as those
735 language sub-directories might not have been configured!
737 @item --disable-libgcj
738 Specify that the run-time libraries
739 used by GCJ should not be built. This is useful in case you intend
740 to use GCJ with some other run-time, or you're going to install it
741 separately, or it just happens not to build on your particular
742 machine. In general, if the Java front end is enabled, the GCJ
743 libraries will be enabled too, unless they're known to not work on
744 the target platform. If GCJ is enabled but @samp{libgcj} isn't built, you
745 may need to port it; in this case, before modifying the top-level
746 @file{configure.in} so that @samp{libgcj} is enabled by default on this platform,
747 you may use @option{--enable-libgcj} to override the default.
750 Specify that the compiler should
751 use DWARF 2 debugging information as the default.
753 @item --enable-win32-registry
754 @itemx --enable-win32-registry=@var{key}
755 @itemx --disable-win32-registry
756 The @option{--enable-win32-registry} option enables Windows-hosted GCC
757 to look up installations paths in the registry using the following key:
760 @code{HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\@var{key}}
763 @var{key} defaults to GCC version number, and can be overridden by the
764 @option{--enable-win32-registry=@var{key}} option. Vendors and distributors
765 who use custom installers are encouraged to provide a different key,
766 perhaps one comprised of vendor name and GCC version number, to
767 avoid conflict with existing installations. This feature is enabled
768 by default, and can be disabled by @option{--disable-win32-registry}
769 option. This option has no effect on the other hosts.
772 Specify that the machine does not have a floating point unit. This
773 option only applies to @samp{m68k-sun-sunos@var{n}} and
774 @samp{m68k-isi-bsd}. On any other system, @option{--nfp} has no effect.
776 @item --enable-checking
777 @itemx --enable-checking=@var{list}
778 When you specify this option, the compiler is built to perform checking
779 of tree node types when referencing fields of that node, and some other
780 internal consistency checks. This does not change the generated code,
781 but adds error checking within the compiler. This will slow down the
782 compiler and may only work properly if you are building the compiler
783 with GCC@. This is on by default when building from CVS or snapshots,
784 but off for releases. More control over the checks may be had by
785 specifying @var{list}; the categories of checks available are
786 @samp{misc}, @samp{tree}, @samp{gc}, @samp{rtl} and @samp{gcac}. The
787 default when @var{list} is not specified is @samp{misc,tree,gc}; the
788 checks @samp{rtl} and @samp{gcac} are very expensive.
792 The @option{--enable-nls} option enables Native Language Support (NLS),
793 which lets GCC output diagnostics in languages other than American
794 English. Native Language Support is enabled by default if not doing a
795 canadian cross build. The @option{--disable-nls} option disables NLS@.
797 @item --with-included-gettext
798 If NLS is enabled, the @option{--with-included-gettext} option causes the build
799 procedure to prefer its copy of GNU @command{gettext}.
802 If NLS is enabled, and if the host lacks @code{gettext} but has the
803 inferior @code{catgets} interface, the GCC build procedure normally
804 ignores @code{catgets} and instead uses GCC's copy of the GNU
805 @code{gettext} library. The @option{--with-catgets} option causes the
806 build procedure to use the host's @code{catgets} in this situation.
808 @item --with-libiconv-prefix=@var{dir}
809 Search for libiconv header files in @file{@var{dir}/include} and
810 libiconv library files in @file{@var{dir}/lib}.
812 @item --with-system-zlib
813 Use installed zlib rather than that included with GCC@. This option
814 only applies if the Java front end is being built.
816 @item --enable-obsolete
817 Enable configuration for an obsoleted system. If you attempt to
818 configure GCC for a system (build, host, or target) which has been
819 obsoleted, and you do not specify this flag, configure will halt with an
822 All support for systems which have been obsoleted in one release of GCC
823 is removed entirely in the next major release, unless someone steps
824 forward to maintain the port.
827 Some options which only apply to building cross compilers:
829 @item --with-headers=@var{dir}
830 Specifies a directory
831 which has target include files.
832 @emph{This options is required} when building a cross
833 compiler, if @file{@var{prefix}/@var{target}/sys-include} doesn't pre-exist.
834 These include files will be copied into the @file{gcc} install directory.
835 Fixincludes will be run on these files to make them compatible with
837 @item --with-libs=``@var{dir1} @var{dir2} @dots{} @var{dirN}''
838 Specifies a list of directories which contain the target runtime
839 libraries. These libraries will be copied into the @file{gcc} install
842 Specifies that @samp{newlib} is
843 being used as the target C library. This causes @code{__eprintf} to be
844 omitted from @file{libgcc.a} on the assumption that it will be provided by
848 Note that each @option{--enable} option has a corresponding
849 @option{--disable} option and that each @option{--with} option has a
850 corresponding @option{--without} option.
857 @uref{./index.html,,Return to the GCC Installation page}
861 @c ***Building****************************************************************
863 @comment node-name, next, previous, up
864 @node Building, Testing, Configuration, Installing GCC
868 <h1 align="center">Installing GCC: Building</h1>
873 @cindex Installing GCC: Building
875 Now that GCC is configured, you are ready to build the compiler and
878 We @strong{highly} recommend that GCC be built using GNU make;
879 other versions may work, then again they might not.
880 GNU make is required for compiling GNAT (the Ada compiler) and the Java
883 (For example, many broken versions of make will fail if you use the
884 recommended setup where @var{objdir} is different from @var{srcdir}.
885 Other broken versions may recompile parts of the compiler when
886 installing the compiler.)
888 Some commands executed when making the compiler may fail (return a
889 nonzero status) and be ignored by @code{make}. These failures, which
890 are often due to files that were not found, are expected, and can safely
893 It is normal to have compiler warnings when compiling certain files.
894 Unless you are a GCC developer, you can generally ignore these warnings
895 unless they cause compilation to fail.
897 On certain old systems, defining certain environment variables such as
898 @env{CC} can interfere with the functioning of @command{make}.
900 If you encounter seemingly strange errors when trying to build the
901 compiler in a directory other than the source directory, it could be
902 because you have previously configured the compiler in the source
903 directory. Make sure you have done all the necessary preparations.
905 If you build GCC on a BSD system using a directory stored in an old System
906 V file system, problems may occur in running @code{fixincludes} if the
907 System V file system doesn't support symbolic links. These problems
908 result in a failure to fix the declaration of @code{size_t} in
909 @file{sys/types.h}. If you find that @code{size_t} is a signed type and
910 that type mismatches occur, this could be the cause.
912 The solution is not to use such a directory for building GCC@.
914 When building from CVS or snapshots, or if you modify parser sources,
915 you need the Bison parser generator installed. Any version 1.25 or
916 later should work; older versions may also work. If you do not modify
917 parser sources, releases contain the Bison-generated files and you do
918 not need Bison installed to build them.
920 When building from CVS or snapshots, or if you modify Texinfo
921 documentation, you need version 4.1 or later of Texinfo installed if you
922 want Info documentation to be regenerated. Releases contain Info
923 documentation pre-built for the unmodified documentation in the release.
925 @section Building a native compiler
927 For a native build issue the command @samp{make bootstrap}. This
928 will build the entire GCC system, which includes the following steps:
932 Build host tools necessary to build the compiler such as texinfo, bison,
936 Build target tools for use by the compiler such as binutils (bfd,
937 binutils, gas, gprof, ld, and opcodes)
938 if they have been individually linked
939 or moved into the top level GCC source tree before configuring.
942 Perform a 3-stage bootstrap of the compiler.
945 Perform a comparison test of the stage2 and stage3 compilers.
948 Build runtime libraries using the stage3 compiler from the previous step.
952 If you are short on disk space you might consider @samp{make
953 bootstrap-lean} instead. This is identical to @samp{make
954 bootstrap} except that object files from the stage1 and
955 stage2 of the 3-stage bootstrap of the compiler are deleted as
956 soon as they are no longer needed.
959 If you want to save additional space during the bootstrap and in
960 the final installation as well, you can build the compiler binaries
961 without debugging information with @samp{make CFLAGS='-O' LIBCFLAGS='-g
962 -O2' LIBCXXFLAGS='-g -O2 -fno-implicit-templates' bootstrap}. This will save
963 roughly 40% of disk space both for the bootstrap and the final installation.
964 (Libraries will still contain debugging information.)
966 If you wish to use non-default GCC flags when compiling the stage2 and
967 stage3 compilers, set @code{BOOT_CFLAGS} on the command line when doing
968 @samp{make bootstrap}. Non-default optimization flags are less well
969 tested here than the default of @samp{-g -O2}, but should still work.
970 In a few cases, you may find that you need to specify special flags such
971 as @option{-msoft-float} here to complete the bootstrap; or, if the
972 native compiler miscompiles the stage1 compiler, you may need to work
973 around this, by choosing @code{BOOT_CFLAGS} to avoid the parts of the
974 stage1 compiler that were miscompiled, or by using @samp{make
975 bootstrap4} to increase the number of stages of bootstrap.
977 If you used the flag @option{--enable-languages=@dots{}} to restrict
978 the compilers to be built, only those you've actually enabled will be
979 built. This will of course only build those runtime libraries, for
980 which the particular compiler has been built. Please note,
981 that re-defining @env{LANGUAGES} when calling @samp{make bootstrap}
982 @strong{does not} work anymore!
984 If the comparison of stage2 and stage3 fails, this normally indicates
985 that the stage2 compiler has compiled GCC incorrectly, and is therefore
986 a potentially serious bug which you should investigate and report. (On
987 a few systems, meaningful comparison of object files is impossible; they
988 always appear ``different''. If you encounter this problem, you will
989 need to disable comparison in the @file{Makefile}.)
991 @section Building a cross compiler
993 We recommend reading the
994 @uref{http://www.objsw.com/CrossGCC/,,crossgcc FAQ}
995 for information about building cross compilers.
997 When building a cross compiler, it is not generally possible to do a
998 3-stage bootstrap of the compiler. This makes for an interesting problem
999 as parts of GCC can only be built with GCC@.
1001 To build a cross compiler, we first recommend building and installing a
1002 native compiler. You can then use the native GCC compiler to build the
1003 cross compiler. The installed native compiler needs to be GCC version
1006 Assuming you have already installed a native copy of GCC and configured
1007 your cross compiler, issue the command @command{make}, which performs the
1012 Build host tools necessary to build the compiler such as texinfo, bison,
1016 Build target tools for use by the compiler such as binutils (bfd,
1017 binutils, gas, gprof, ld, and opcodes)
1018 if they have been individually linked or moved into the top level GCC source
1019 tree before configuring.
1022 Build the compiler (single stage only).
1025 Build runtime libraries using the compiler from the previous step.
1028 Note that if an error occurs in any step the make process will exit.
1030 @section Building in parallel
1032 If you have a multiprocessor system you can use @samp{make bootstrap
1033 MAKE="make -j 2" -j 2} or just @samp{make -j 2 bootstrap}
1034 for GNU Make 3.79 and above instead of just @samp{make bootstrap}
1035 when building GCC@. You can use a bigger number instead of two if
1036 you like. In most cases, it won't help to use a number bigger than
1037 the number of processors in your machine.
1039 @section Building the Ada compiler
1041 In order to build GNAT, the Ada compiler, you need a working GNAT
1042 compiler (GNAT version 3.13 or later, or GCC version 3.1 or later),
1043 since the Ada front end is written in Ada (with some
1044 GNAT-specific extensions), and GNU make.
1046 However, you do not need a full installation of GNAT, just the GNAT
1047 binary @file{gnat1}, a copy of @file{gnatbind}, and a compiler driver
1048 which can deal with Ada input (by invoking the @file{gnat1} binary).
1049 You can specify this compiler driver by setting the @env{ADAC}
1050 environment variable at the configure step. @command{configure} can
1051 detect the driver automatically if it has got a common name such as
1052 @command{gcc} or @command{gnatgcc}. Of course, you still need a working
1053 C compiler (the compiler driver can be different or not).
1054 @command{configure} does not test whether the GNAT installation works
1055 and has a sufficiently recent version; if too old a GNAT version is
1056 installed, the build will fail unless @option{--enable-languages} is
1057 used to disable building the Ada front end.
1059 Additional build tools (such as @command{gnatmake}) or a working GNAT
1060 run-time library installation are usually @emph{not} required. However,
1061 if you want to boostrap the compiler using a minimal version of GNAT,
1062 you have to issue the following commands before invoking @samp{make
1063 boostrap} (this assumes that you start with an unmodified and consistent
1064 source distribution):
1067 cd @var{srcdir}/gcc/ada
1068 touch treeprs.ads [es]info.h nmake.ad[bs]
1071 At the moment, the GNAT library and several tools for GNAT are not built
1072 by @samp{make bootstrap}. You have to invoke
1073 @samp{make gnatlib_and_tools} in the @file{@var{objdir}/gcc}
1074 subdirectory before proceeding with the next steps.
1076 For example, you can build a native Ada compiler by issuing the
1077 following commands (assuming @command{make} is GNU make):
1081 @var{srcdir}/configure --enable-languages=c,ada
1082 cd @var{srcdir}/gcc/ada
1083 touch treeprs.ads [es]info.h nmake.ad[bs]
1087 make gnatlib_and_tools
1091 Currently, when compiling the Ada front end, you cannot use the parallel
1092 build feature described in the previous section.
1099 @uref{./index.html,,Return to the GCC Installation page}
1103 @c ***Testing*****************************************************************
1105 @comment node-name, next, previous, up
1106 @node Testing, Final install, Building, Installing GCC
1110 <h1 align="center">Installing GCC: Testing</h1>
1113 @chapter Installing GCC: Testing
1116 @cindex Installing GCC: Testing
1119 Before you install GCC, we encourage you to run the testsuites and to
1120 compare your results with results from a similar configuration that have
1121 been submitted to the
1122 @uref{http://gcc.gnu.org/ml/gcc-testresults/,,gcc-testresults mailing list}.
1123 This step is optional and may require you to download additional software,
1124 but it can give you confidence in your new GCC installation or point out
1125 problems before you install and start using your new GCC.
1127 First, you must have @uref{download.html,,downloaded the testsuites}.
1128 These are part of the full distribution, but if you downloaded the
1129 ``core'' compiler plus any front ends, you must download the testsuites
1132 Second, you must have the testing tools installed. This includes
1133 a @uref{http://www.gnu.org/software/dejagnu/,,current version of DejaGnu};
1134 dejagnu 1.3 is not sufficient.
1135 It also includes Tcl and Expect; the DejaGnu site has links to these.
1137 Now you may need specific preparations:
1142 The following environment variables may need to be set appropriately, as in
1143 the following example (which assumes that DejaGnu has been installed
1144 under @file{/usr/local}):
1147 TCL_LIBRARY = /usr/local/share/tcl8.0
1148 DEJAGNULIBS = /usr/local/share/dejagnu
1151 On systems such as Cygwin, these paths are required to be actual
1152 paths, not mounts or links; presumably this is due to some lack of
1153 portability in the DejaGnu code.
1155 If the directories where @command{runtest} and @command{expect} were
1156 installed are in the @env{PATH}, it should not be necessary to set these
1157 environment variables.
1161 Finally, you can run the testsuite (which may take a long time):
1163 cd @var{objdir}; make -k check
1166 The testing process will try to test as many components in the GCC
1167 distribution as possible, including the C, C++, Objective-C and Fortran
1168 compilers as well as the C++ and Java runtime libraries.
1170 @section How can I run the test suite on selected tests?
1172 As a first possibility to cut down the number of tests that are run it is
1173 possible to use @samp{make check-gcc} or @samp{make check-g++}
1174 in the @file{gcc} subdirectory of the object directory. To further cut down the
1175 tests the following is possible:
1178 make check-gcc RUNTESTFLAGS="execute.exp @var{other-options}"
1181 This will run all @command{gcc} execute tests in the testsuite.
1184 make check-g++ RUNTESTFLAGS="old-deja.exp=9805* @var{other-options}"
1187 This will run the @command{g++} ``old-deja'' tests in the testsuite where the filename
1188 matches @samp{9805*}.
1190 The @file{*.exp} files are located in the testsuite directories of the GCC
1191 source, the most important ones being @file{compile.exp},
1192 @file{execute.exp}, @file{dg.exp} and @file{old-deja.exp}.
1193 To get a list of the possible @file{*.exp} files, pipe the
1194 output of @samp{make check} into a file and look at the
1195 @samp{Running @dots{} .exp} lines.
1197 @section How to interpret test results
1199 After the testsuite has run you'll find various @file{*.sum} and @file{*.log}
1200 files in the testsuite subdirectories. The @file{*.log} files contain a
1201 detailed log of the compiler invocations and the corresponding
1202 results, the @file{*.sum} files summarize the results. These summaries list
1203 all the tests that have been run with a corresponding status code:
1207 PASS: the test passed as expected
1209 XPASS: the test unexpectedly passed
1211 FAIL: the test unexpectedly failed
1213 XFAIL: the test failed as expected
1215 UNSUPPORTED: the test is not supported on this platform
1217 ERROR: the testsuite detected an error
1219 WARNING: the testsuite detected a possible problem
1222 It is normal for some tests to report unexpected failures. At the
1223 current time our testing harness does not allow fine grained control
1224 over whether or not a test is expected to fail. We expect to fix this
1225 problem in future releases.
1228 @section Submitting test results
1230 If you want to report the results to the GCC project, use the
1231 @file{contrib/test_summary} shell script. Start it in the @var{objdir} with
1234 @var{srcdir}/contrib/test_summary -p your_commentary.txt \
1235 -m gcc-testresults@@gcc.gnu.org |sh
1238 This script uses the @command{Mail} program to send the results, so
1239 make sure it is in your @env{PATH}. The file @file{your_commentary.txt} is
1240 prepended to the testsuite summary and should contain any special
1241 remarks you have on your results or your build environment. Please
1242 do not edit the testsuite result block or the subject line, as these
1243 messages are automatically parsed and presented at the
1244 @uref{http://gcc.gnu.org/testresults/,,GCC testresults} web
1245 page. Here you can also gather information on how specific tests
1246 behave on different platforms and compare them with your results. A
1247 few failing testcases are possible even on released versions and you
1248 should look here first if you think your results are unreasonable.
1255 @uref{./index.html,,Return to the GCC Installation page}
1259 @c ***Final install***********************************************************
1261 @comment node-name, next, previous, up
1262 @node Final install, , Testing, Installing GCC
1264 @ifset finalinstallhtml
1266 <h1 align="center">Installing GCC: Final installation</h1>
1269 @chapter Installing GCC: Final installation
1272 Now that GCC has been built (and optionally tested), you can install it with
1274 cd @var{objdir}; make install
1277 That step completes the installation of GCC; user level binaries can
1278 be found in @file{@var{prefix}/bin} where @var{prefix} is the value you
1279 specified with the @option{--prefix} to configure (or @file{/usr/local}
1280 by default). (If you specified @option{--bindir}, that directory will
1281 be used instead; otherwise, if you specified @option{--exec-prefix},
1282 @file{@var{exec-prefix}/bin} will be used.) Headers for the C++ and
1283 Java libraries are installed in @file{@var{prefix}/include}; libraries
1284 in @file{@var{libdir}} (normally @file{@var{prefix}/lib}); internal
1285 parts of the compiler in @file{@var{libdir}/gcc-lib}; documentation in
1286 info format in @file{@var{infodir}} (normally @file{@var{prefix}/info}).
1288 If you built a released version of GCC using @samp{make bootstrap} then please
1289 quickly review the build status page for
1290 @uref{http://gcc.gnu.org/gcc-3.1/buildstat.html,,3.1},
1291 @uref{http://gcc.gnu.org/gcc-3.0/buildstat.html,,3.0}, or
1292 @uref{http://gcc.gnu.org/gcc-2.95/buildstat.html,,2.95}.
1293 If your system is not listed for the version of GCC that you built,
1295 @email{gcc@@gcc.gnu.org} indicating
1296 that you successfully built and installed GCC.
1297 Include the following information:
1301 Output from running @file{@var{srcdir}/config.guess}. Do not send us
1302 that file itself, just the one-line output from running it.
1305 The output of @samp{gcc -v} for your newly installed gcc.
1306 This tells us which version of GCC you built and the options you passed to
1310 Whether you enabled all languages or a subset of them. If you used a
1311 full distribution then this information is part of the configure
1312 options in the output of @samp{gcc -v}, but if you downloaded the
1313 ``core'' compiler plus additional front ends then it isn't apparent
1314 which ones you built unless you tell us about it.
1317 If the build was for GNU/Linux, also include:
1320 The distribution name and version (e.g., Red Hat 7.1 or Debian 2.2.3);
1321 this information should be available from @file{/etc/issue}.
1324 The version of the Linux kernel, available from @samp{uname --version}
1328 The version of glibc you used; for RPM-based systems like Red Hat,
1329 Mandrake, and SuSE type @samp{rpm -q glibc} to get the glibc version,
1330 and on systems like Debian and Progeny use @samp{dpkg -l libc6}.
1332 For other systems, you can include similar information if you think it is
1336 Any other information that you think would be useful to people building
1337 GCC on the same configuration. The new entry in the build status list
1338 will include a link to the archived copy of your message.
1341 We'd also like to know if the
1343 @ref{Specific, host/target specific installation notes}
1346 @uref{specific.html,,host/target specific installation notes}
1348 didn't include your host/target information or if that information is
1349 incomplete or out of date. Send a note to
1350 @email{gcc@@gcc.gnu.org} telling us how the information should be changed.
1352 If you find a bug, please report it following our
1353 @uref{../bugs.html,,bug reporting guidelines}.
1355 If you want to print the GCC manuals, do @samp{cd @var{objdir}; make
1356 dvi}. You will need to have @command{texi2dvi} (version at least 4.1)
1357 and @TeX{} installed. This creates a number of @file{.dvi} files in
1358 subdirectories of @file{@var{objdir}}; these may be converted for
1359 printing with programs such as @command{dvips}. You can also
1360 @uref{http://www.gnu.org/order/order.html,,buy printed manuals from the
1361 Free Software Foundation}, though such manuals may not be for the most
1362 recent version of GCC@.
1369 @uref{./index.html,,Return to the GCC Installation page}
1373 @c ***Binaries****************************************************************
1375 @comment node-name, next, previous, up
1376 @node Binaries, Specific, Installing GCC, Top
1380 <h1 align="center">Installing GCC: Binaries</h1>
1383 @chapter Installing GCC: Binaries
1386 @cindex Installing GCC: Binaries
1388 We are often asked about pre-compiled versions of GCC@. While we cannot
1389 provide these for all platforms, below you'll find links to binaries for
1390 various platforms where creating them by yourself is not easy due to various
1393 Please note that we did not create these binaries, nor do we
1394 support them. If you have any problems installing them, please
1395 contact their makers.
1402 @uref{http://freeware.bull.net,,Bull's Freeware and Shareware Archive for AIX};
1405 @uref{http://aixpdslib.seas.ucla.edu,,UCLA Software Library for AIX};
1409 DOS---@uref{http://www.delorie.com/djgpp/,,DJGPP};
1415 @uref{http://hpux.cae.wisc.edu/,,HP-UX Porting Center};
1418 @uref{ftp://sunsite.informatik.rwth-aachen.de/pub/packages/gcc_hpux/,,Binaries for HP-UX 11.00 at Aachen University of Technology}.
1422 @uref{http://www.sco.com/skunkware/devtools/index.html#gcc,,SCO
1423 OpenServer/Unixware};
1426 Solaris 2 (SPARC, Intel)---@uref{http://www.sunfreeware.com/,,Sunfreeware};
1429 SGI---@uref{http://freeware.sgi.com/,,SGI Freeware};
1432 Windows 95, 98, and NT:
1435 The @uref{http://sources.redhat.com/cygwin/,,Cygwin} project;
1437 @uref{http://www.xraylith.wisc.edu/~khan/software/gnu-win32/,,GNU Win32}
1438 related projects by Mumit Khan.
1442 @uref{ftp://ftp.thewrittenword.com/packages/free/by-name/,,The
1443 Written Word} offers binaries for Solaris 2.5.1, 2.6, 2.7/SPARC, 2.7/Intel,
1444 IRIX 6.2, 6.5, Digital UNIX 4.0D, HP-UX 10.20, and HP-UX 11.00.
1447 Hitachi H8/300[HS]---@uref{http://h8300-hms.sourceforge.net/,,GNU
1448 Development Tools for the Hitachi H8/300[HS] Series}
1452 In addition to those specific offerings, you can get a binary
1453 distribution CD-ROM from the
1454 @uref{http://www.fsf.org/order/order.html,,Free Software Foundation}.
1455 It contains binaries for a number of platforms, and
1456 includes not only GCC, but other stuff as well. The current CD does
1457 not contain the latest version of GCC, but it should allow
1458 bootstrapping the compiler. An updated version of that disk is in the
1466 @uref{./index.html,,Return to the GCC Installation page}
1470 @c ***Specific****************************************************************
1472 @comment node-name, next, previous, up
1473 @node Specific, Old, Binaries, Top
1477 <h1 align="center">Host/target specific installation notes for GCC</h1>
1480 @chapter Host/target specific installation notes for GCC
1483 @cindex Specific installation notes
1484 @cindex Target specific installation
1485 @cindex Host specific installation
1486 @cindex Target specific installation notes
1488 Please read this document carefully @emph{before} installing the
1489 GNU Compiler Collection on your machine.
1494 @uref{#1750a-*-*,,1750a-*-*}
1498 @uref{#a29k-*-bsd,,a29k-*-bsd}
1500 @uref{#alpha*-*-*,,alpha*-*-*}
1502 @uref{#alpha*-dec-osf*,,alpha*-dec-osf*}
1504 @uref{#alphaev5-cray-unicosmk*,,alphaev5-cray-unicosmk*}
1506 @uref{#arc-*-elf,,arc-*-elf}
1508 @uref{#arm-*-aout,,arm-*-aout}
1510 @uref{#arm-*-elf,,arm-*-elf}
1512 @uref{#arm*-*-linux-gnu,,arm*-*-linux-gnu}
1514 @uref{#arm-*-riscix,,arm-*-riscix}
1522 @uref{#dsp16xx,,dsp16xx}
1524 @uref{#elxsi-elxsi-bsd,,elxsi-elxsi-bsd}
1526 @uref{#*-*-freebsd*,,*-*-freebsd*}
1528 @uref{#h8300-hms,,h8300-hms}
1530 @uref{#hppa*-hp-hpux*,,hppa*-hp-hpux*}
1532 @uref{#hppa*-hp-hpux9,,hppa*-hp-hpux9}
1534 @uref{#hppa*-hp-hpux10,,hppa*-hp-hpux10}
1536 @uref{#hppa*-hp-hpux11,,hppa*-hp-hpux11}
1538 @uref{#i370-*-*,,i370-*-*}
1540 @uref{#*-*-linux-gnu,,*-*-linux-gnu}
1542 @uref{#ix86-*-linux*oldld,,i?86-*-linux*oldld}
1544 @uref{#ix86-*-linux*aout,,i?86-*-linux*aout}
1546 @uref{#ix86-*-linux*,,i?86-*-linux*}
1548 @uref{#ix86-*-sco,,i?86-*-sco}
1550 @uref{#ix86-*-sco3.2v4,,i?86-*-sco3.2v4}
1552 @uref{#ix86-*-sco3.2v5*,,i?86-*-sco3.2v5*}
1554 @uref{#ix86-*-udk,,i?86-*-udk}
1556 @uref{#ix86-*-isc,,i?86-*-isc}
1558 @uref{#ix86-*-esix,,i?86-*-esix}
1560 @uref{#ix86-ibm-aix,,i?86-ibm-aix}
1562 @uref{#ix86-sequent-bsd,,i?86-sequent-bsd}
1564 @uref{#ix86-sequent-ptx1*,,i?86-sequent-ptx1*, i?86-sequent-ptx2*, i?86-sequent-sysv3*}
1566 @uref{#i860-intel-osf*,,i860-intel-osf*}
1568 @uref{#ia64-*-linux,,ia64-*-linux}
1570 @uref{#*-lynx-lynxos,,*-lynx-lynxos}
1572 @uref{#*-ibm-aix*,,*-ibm-aix*}
1574 @uref{#m32r-*-elf,,m32r-*-elf}
1576 @uref{#m68000-hp-bsd,,m68000-hp-bsd}
1578 @uref{#m6811-elf,,m6811-elf}
1580 @uref{#m6812-elf,,m6812-elf}
1582 @uref{#m68k-altos,,m68k-altos}
1584 @uref{#m68k-apple-aux,,m68k-apple-aux}
1586 @uref{#m68k-att-sysv,,m68k-att-sysv}
1588 @uref{#m68k-bull-sysv,,m68k-bull-sysv}
1590 @uref{#m68k-crds-unos,,m68k-crds-unos}
1592 @uref{#m68k-hp-hpux,,m68k-hp-hpux}
1594 @uref{#m68k-*-nextstep*,,m68k-*-nextstep*}
1596 @uref{#m68k-ncr-*,,m68k-ncr-*}
1598 @uref{#m68k-sun,,m68k-sun}
1600 @uref{#m68k-sun-sunos4.1.1,,m68k-sun-sunos4.1.1}
1602 @uref{#m88k-*-svr3,,m88k-*-svr3}
1604 @uref{#m88k-*-dgux,,m88k-*-dgux}
1606 @uref{#m88k-tektronix-sysv3,,m88k-tektronix-sysv3}
1608 @uref{#mips-*-*,,mips-*-*}
1610 @uref{#mips-dec-*,,mips-dec-*}
1612 @uref{#mips-mips-bsd,,mips-mips-bsd}
1614 @uref{#mips-mips-riscos*,,mips-mips-riscos*}
1616 @uref{#mips-sgi-irix4,,mips-sgi-irix4}
1618 @uref{#mips-sgi-irix5,,mips-sgi-irix5}
1620 @uref{#mips-sgi-irix6,,mips-sgi-irix6}
1622 @uref{#mips-sony-sysv,,mips-sony-sysv}
1624 @uref{#ns32k-encore,,ns32k-encore}
1626 @uref{#ns32k-*-genix,,ns32k-*-genix}
1628 @uref{#ns32k-sequent,,ns32k-sequent}
1630 @uref{#ns32k-utek,,ns32k-utek}
1632 @uref{#powerpc*-*-*,,powerpc*-*-*, powerpc-*-sysv4}
1634 @uref{#powerpc-*-darwin*,,powerpc-*-darwin*}
1636 @uref{#powerpc-*-elf,,powerpc-*-elf, powerpc-*-sysv4}
1638 @uref{#powerpc-*-linux-gnu*,,powerpc-*-linux-gnu*}
1640 @uref{#powerpc-*-netbsd*,,powerpc-*-netbsd*}
1642 @uref{#powerpc-*-eabiaix,,powerpc-*-eabiaix}
1644 @uref{#powerpc-*-eabisim,,powerpc-*-eabisim}
1646 @uref{#powerpc-*-eabi,,powerpc-*-eabi}
1648 @uref{#powerpcle-*-elf,,powerpcle-*-elf, powerpcle-*-sysv4}
1650 @uref{#powerpcle-*-eabisim,,powerpcle-*-eabisim}
1652 @uref{#powerpcle-*-eabi,,powerpcle-*-eabi}
1654 @uref{#powerpcle-*-winnt,,powerpcle-*-winnt, powerpcle-*-pe}
1656 @uref{#romp-*-aos,,romp-*-aos, romp-*-mach}
1658 @uref{#s390-*-linux*}
1660 @uref{#s390x-*-linux*}
1662 @uref{#*-*-solaris2*,,*-*-solaris2*}
1664 @uref{#sparc-sun-solaris2*,,sparc-sun-solaris2*}
1666 @uref{#sparc-sun-solaris2.7,,sparc-sun-solaris2.7}
1668 @uref{#sparc-sun-sunos4*,,sparc-sun-sunos4*}
1670 @uref{#sparc-unknown-linux-gnulibc1,,sparc-unknown-linux-gnulibc1}
1672 @uref{#sparc-*-linux*,,sparc-*-linux*}
1674 @uref{#sparc64-*-*,,sparc64-*-*}
1676 @uref{#sparcv9-*-solaris2*,,sparcv9-*-solaris2*}
1678 @uref{#*-*-sysv*,,*-*-sysv*}
1680 @uref{#vax-dec-ultrix,,vax-dec-ultrix}
1682 @uref{#we32k-*-*,,we32k-*-*}
1684 @uref{#xtensa-*-elf,,xtensa-*-elf}
1686 @uref{#xtensa-*-linux*,,xtensa-*-linux*}
1688 @uref{#windows,,Microsoft Windows}
1692 @uref{#older,,Older systems}
1697 @uref{#elf_targets,,all ELF targets} (SVR4, Solaris 2, etc.)
1703 <!-- -------- host/target specific issues start here ---------------- -->
1706 @heading @anchor{1750a-*-*}1750a-*-*
1707 MIL-STD-1750A processors. This target is obsoleted in GCC 3.1.
1709 The MIL-STD-1750A cross configuration produces output for
1710 @code{as1750}, an assembler/linker available under the GNU General Public
1711 License for the 1750A@. @code{as1750} can be obtained at
1712 @uref{ftp://ftp.fta-berlin.de/pub/crossgcc/1750gals/}.
1713 A similarly licensed simulator for
1714 the 1750A is available from same address.
1716 You should ignore a fatal error during the building of @samp{libgcc}
1717 (@samp{libgcc} is not yet implemented for the 1750A@.)
1719 The @code{as1750} assembler requires the file @file{ms1750.inc}, which is
1720 found in the directory @file{gcc/config/1750a}.
1722 GCC produced the same sections as the Fairchild F9450 C Compiler,
1727 The program code section.
1730 The read/write (RAM) data section.
1733 The read-only (ROM) constants section.
1736 Initialization section (code to copy KREL to SREL)@.
1739 The smallest addressable unit is 16 bits (@code{BITS_PER_UNIT} is 16). This
1740 means that type @code{char} is represented with a 16-bit word per character.
1741 The 1750A's ``Load/Store Upper/Lower Byte'' instructions are not used by
1748 @heading @anchor{a29k}a29k
1749 AMD Am29k-family processors. These are normally used in embedded
1750 applications. This configuration corresponds to AMD's standard calling
1751 sequence and binary interface and is compatible with other 29k tools.
1753 AMD has abandoned this processor. All existing a29k targets are obsoleted
1756 You may need to make a variant of the file @file{a29k.h} for your
1757 particular configuration.
1763 @heading @anchor{a29k-*-bsd}a29k-*-bsd
1764 AMD Am29050 used in a system running a variant of BSD Unix.
1770 @heading @anchor{alpha*-*-*}alpha*-*-*
1772 This section contains general configuration information for all
1773 alpha-based platforms using ELF (in particular, ignore this section for
1774 DEC OSF/1, Digital UNIX and Tru64 UNIX)@. In addition to reading this
1775 section, please read all other sections that match your target.
1777 We require binutils 2.11.2 or newer.
1778 Previous binutils releases had a number of problems with DWARF 2
1779 debugging information, not the least of which is incorrect linking of
1786 @heading @anchor{alpha*-dec-osf*}alpha*-dec-osf*
1787 Systems using processors that implement the DEC Alpha architecture and
1788 are running the DEC/Compaq Unix (DEC OSF/1, Digital UNIX, or Compaq
1789 Tru64 UNIX) operating system, for example the DEC Alpha AXP systems.
1791 Support for versions before @code{alpha*-dec-osf4} is obsoleted in GCC
1792 3.1. (These are the versions which identify themselves as DEC OSF/1.)
1794 In Tru64 UNIX V5.1, Compaq introduced a new assembler that does not
1795 currently (2001-06-13) work with @command{mips-tfile}. As a workaround,
1796 we need to use the old assembler, invoked via the barely documented
1797 @option{-oldas} option. To bootstrap GCC, you either need to use the
1801 % CC=cc @var{srcdir}/configure [@var{options}] [@var{target}]
1804 or you can use a copy of GCC 2.95.3 or higher built on Tru64 UNIX V4.0:
1807 % CC=gcc -Wa,-oldas @var{srcdir}/configure [@var{options}] [@var{target}]
1810 As of GNU binutils 2.11.2, neither GNU @command{as} nor GNU @command{ld}
1811 are supported on Tru64 UNIX, so you must not configure GCC with
1812 @option{--with-gnu-as} or @option{--with-gnu-ld}.
1814 The @option{--enable-threads} options isn't supported yet. A patch is
1815 in preparation for a future release.
1817 GCC writes a @samp{.verstamp} directive to the assembler output file
1818 unless it is built as a cross-compiler. It gets the version to use from
1819 the system header file @file{/usr/include/stamp.h}. If you install a
1820 new version of DEC Unix, you should rebuild GCC to pick up the new version
1823 Note that since the Alpha is a 64-bit architecture, cross-compilers from
1824 32-bit machines will not generate code as efficient as that generated
1825 when the compiler is running on a 64-bit machine because many
1826 optimizations that depend on being able to represent a word on the
1827 target in an integral value on the host cannot be performed. Building
1828 cross-compilers on the Alpha for 32-bit machines has only been tested in
1829 a few cases and may not work properly.
1831 @code{make compare} may fail on old versions of DEC Unix unless you add
1832 @option{-save-temps} to @code{CFLAGS}. On these systems, the name of the
1833 assembler input file is stored in the object file, and that makes
1834 comparison fail if it differs between the @code{stage1} and
1835 @code{stage2} compilations. The option @option{-save-temps} forces a
1836 fixed name to be used for the assembler input file, instead of a
1837 randomly chosen name in @file{/tmp}. Do not add @option{-save-temps}
1838 unless the comparisons fail without that option. If you add
1839 @option{-save-temps}, you will have to manually delete the @samp{.i} and
1840 @samp{.s} files after each series of compilations.
1842 GCC now supports both the native (ECOFF) debugging format used by DBX
1843 and GDB and an encapsulated STABS format for use only with GDB@. See the
1844 discussion of the @option{--with-stabs} option of @file{configure} above
1845 for more information on these formats and how to select them.
1847 There is a bug in DEC's assembler that produces incorrect line numbers
1848 for ECOFF format when the @samp{.align} directive is used. To work
1849 around this problem, GCC will not emit such alignment directives
1850 while writing ECOFF format debugging information even if optimization is
1851 being performed. Unfortunately, this has the very undesirable
1852 side-effect that code addresses when @option{-O} is specified are
1853 different depending on whether or not @option{-g} is also specified.
1855 To avoid this behavior, specify @option{-gstabs+} and use GDB instead of
1856 DBX@. DEC is now aware of this problem with the assembler and hopes to
1857 provide a fix shortly.
1863 @heading @anchor{alphaev5-cray-unicosmk*}alphaev5-cray-unicosmk*
1864 Cray T3E systems running Unicos/Mk.
1866 This port is incomplete and has many known bugs. We hope to improve the
1867 support for this target soon. Currently, only the C front end is supported,
1868 and it is not possible to build parallel applications. Cray modules are not
1869 supported; in particular, Craylibs are assumed to be in
1870 @file{/opt/ctl/craylibs/craylibs}.
1872 You absolutely @strong{must} use GNU make on this platform. Also, you
1873 need to tell GCC where to find the assembler and the linker. The
1874 simplest way to do so is by providing @option{--with-as} and
1875 @option{--with-ld} to @file{configure}, e.g.@:
1877 @samp{configure --with-as=/opt/ctl/bin/cam --with-ld=/opt/ctl/bin/cld
1878 --enable-languages=c}
1880 The comparison test during @samp{make bootstrap} fails on Unicos/Mk
1881 because the assembler inserts timestamps into object files. You should
1882 be able to work around this by doing @samp{make all} after getting this
1889 @heading @anchor{arc-*-elf}arc-*-elf
1890 Argonaut ARC processor.
1891 This configuration is intended for embedded systems.
1897 @heading @anchor{arm-*-aout}arm-*-aout
1898 Advanced RISC Machines ARM-family processors. These are often used in
1899 embedded applications. There are no standard Unix configurations.
1900 This configuration corresponds to the basic instruction sequences and will
1901 produce @file{a.out} format object modules.
1903 You may need to make a variant of the file @file{arm.h} for your particular
1910 @heading @anchor{arm-*-elf}arm-*-elf
1911 This configuration is intended for embedded systems.
1917 @heading @anchor{arm*-*-linux-gnu}arm*-*-linux-gnu
1919 We require GNU binutils 2.10 or newer.
1925 @heading @anchor{arm-*-riscix}arm-*-riscix
1926 The ARM2 or ARM3 processor running RISC iX, Acorn's port of BSD Unix.
1927 This configuration is obsoleted in GCC 3.1.
1929 If you are running a version of RISC iX prior to 1.2 then you must
1930 specify the version number during configuration. Note that the
1931 assembler shipped with RISC iX does not support stabs debugging
1932 information; a new version of the assembler, with stabs support
1933 included, is now available from Acorn and via ftp
1934 @uref{ftp://ftp.acorn.com/pub/riscix/as+xterm.tar.Z}. To enable stabs
1935 debugging, pass @option{--with-gnu-as} to configure.
1937 You will need to install GNU @command{sed} before you can run configure.
1943 @heading @anchor{avr}avr
1945 ATMEL AVR-family micro controllers. These are used in embedded
1946 applications. There are no standard Unix configurations.
1948 @xref{AVR Options,, AVR Options, gcc, Using and Porting the GNU Compiler
1952 See ``AVR Options'' in the main manual
1954 for the list of supported MCU types.
1956 Use @samp{configure --target=avr --enable-languages="c"} to configure GCC@.
1958 Further installation notes and other useful information about AVR tools
1959 can also be obtained from:
1963 @uref{http://home.overta.ru/users/denisc,,http://home.overta.ru/users/denisc}
1965 @uref{http://www.amelek.gda.pl/avr,,http://www.amelek.gda.pl/avr}
1968 We @emph{strongly} recommend using binutils 2.11 or newer.
1970 The following error:
1972 Error: register required
1975 indicates that you should upgrade to a newer version of the binutils.
1981 @heading @anchor{c4x}c4x
1983 Texas Instruments TMS320C3x and TMS320C4x Floating Point Digital Signal
1984 Processors. These are used in embedded applications. There are no
1985 standard Unix configurations.
1987 @xref{TMS320C3x/C4x Options,, TMS320C3x/C4x Options, gcc, Using and
1988 Porting the GNU Compiler Collection (GCC)},
1991 See ``TMS320C3x/C4x Options'' in the main manual
1993 for the list of supported MCU types.
1995 GCC can be configured as a cross compiler for both the C3x and C4x
1996 architectures on the same system. Use @samp{configure --target=c4x
1997 --enable-languages="c,c++"} to configure.
2000 Further installation notes and other useful information about C4x tools
2001 can also be obtained from:
2005 @uref{http://www.elec.canterbury.ac.nz/c4x/,,http://www.elec.canterbury.ac.nz/c4x/}
2012 @heading @anchor{cris}CRIS
2014 CRIS is the CPU architecture in Axis Communications ETRAX system-on-a-chip
2015 series. These are used in embedded applications.
2018 @xref{CRIS Options,, CRIS Options, gcc, Using and Porting the GNU Compiler
2022 See ``CRIS Options'' in the main manual
2024 for a list of CRIS-specific options.
2026 There are a few different CRIS targets:
2028 @item cris-axis-aout
2029 Old target. Includes a multilib for the @samp{elinux} a.out-based
2030 target. No multilibs for newer architecture variants.
2032 Mainly for monolithic embedded systems. Includes a multilib for the
2033 @samp{v10} core used in @samp{ETRAX 100 LX}.
2034 @item cris-axis-linux-gnu
2035 A GNU/Linux port for the CRIS architecture, currently targeting
2036 @samp{ETRAX 100 LX} by default.
2039 For @code{cris-axis-aout} and @code{cris-axis-elf} you need binutils 2.11
2040 or newer. For @code{cris-axis-linux-gnu} you need binutils 2.12 or newer.
2042 Pre-packaged tools can be obtained from
2043 @uref{ftp://ftp.axis.com/pub/axis/tools/cris/compiler-kit/}. More
2044 information about this platform is available at
2045 @uref{http://developer.axis.com/}.
2051 @heading @anchor{dos}DOS
2053 Please have a look at our @uref{binaries.html,,binaries page}.
2055 You cannot install GCC by itself on MSDOS; it will not compile under
2056 any MSDOS compiler except itself. You need to get the complete
2057 compilation package DJGPP, which includes binaries as well as sources,
2058 and includes all the necessary compilation tools and libraries.
2064 @heading @anchor{dsp16xx}dsp16xx
2065 A port to the AT&T DSP1610 family of processors.
2071 @heading @anchor{*-*-freebsd*}*-*-freebsd*
2073 The version of binutils installed in @file{/usr/bin} is known to work unless
2074 otherwise specified in any per-architecture notes. However, binutils
2075 2.12.1 or greater is known to improve overall testsuite results.
2077 For FreeBSD 1, FreeBSD 2 or any mutant a.out versions of FreeBSD 3: All
2078 configuration support and files as shipped with GCC 2.95 are still in
2079 place. FreeBSD 2.2.7 has been known to bootstrap completely; however,
2080 it is unknown which version of binutils was used (it is assumed that it
2081 was the system copy in @file{/usr/bin}) and C++ EH failures were noted.
2083 Support for FreeBSD 1 is obsoleted in GCC 3.1.
2085 For FreeBSD using the ELF file format: DWARF 2 debugging is now the
2086 default for all CPU architectures. It had been the default on
2087 FreeBSD/alpha since its inception. You may use @option{-gstabs} instead
2088 of @option{-g}, if you really want the old debugging format. There are
2089 no known issues with mixing object files and libraries with different
2090 debugging formats. Otherwise, this release of GCC should now match more
2091 of the configuration used in the stock FreeBSD configuration of GCC. In
2092 particular, @option{--enable-threads} is now configured by default.
2093 However, as a general user, do not attempt to replace the system
2094 compiler with this release. Known to bootstrap and check with good
2095 results on FreeBSD 3.0, 3.4, 4.0, 4.2, 4.3, 4.4, 4.5-STABLE and 5-CURRENT@.
2097 In principle, @option{--enable-threads} is now compatible with
2098 @option{--enable-libgcj} on FreeBSD@. However, it has only been built
2099 and tested on i386-*-freebsd4.5 and alpha-*-freebsd5.0. The static
2100 library may be incorrectly built (symbols are missing at link time).
2101 There is a rare timing-based startup hang (probably involves an
2102 assupmtion about the thread library). Multi-threaded boehm-gc (required for
2103 libjava) exposes severe threaded signal-handling bugs on FreeBSD before
2104 4.5-RELEASE. The alpha port may not fully bootstrap without some manual
2105 intervention: gcjh will crash with a floating-point exception while
2106 generating @file{java/lang/Double.h} (just copy the version built on
2107 i386-*-freebsd* and rerun the top-level gmake with no arguments and it
2108 should properly complete the bootstrap). Other CPU architectures
2109 supported by FreeBSD will require additional configuration tuning in, at
2110 the very least, both boehm-gc and libffi.
2112 Shared @file{libgcc_s.so} is now built and installed by default.
2118 @heading @anchor{elxsi-elxsi-bsd}elxsi-elxsi-bsd
2119 The Elxsi's C compiler has known limitations that prevent it from
2120 compiling GCC@. Please contact @email{mrs@@wrs.com} for more details.
2122 Support for this processor is obsoleted in GCC 3.1.
2128 @heading @anchor{h8300-hms}h8300-hms
2129 Hitachi H8/300 series of processors.
2131 Please have a look at our @uref{binaries.html,,binaries page}.
2133 The calling convention and structure layout has changed in release 2.6.
2134 All code must be recompiled. The calling convention now passes the
2135 first three arguments in function calls in registers. Structures are no
2136 longer a multiple of 2 bytes.
2142 @heading @anchor{hppa*-hp-hpux*}hppa*-hp-hpux*
2144 We @emph{highly} recommend using gas/binutils 2.8 or newer on all hppa
2145 platforms; you may encounter a variety of problems when using the HP
2148 Specifically, @option{-g} does not work on HP-UX (since that system
2149 uses a peculiar debugging format which GCC does not know about), unless you
2150 use GAS and GDB and configure GCC with the
2151 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}} and
2152 @option{--with-as=@dots{}} options.
2154 If you wish to use pa-risc 2.0 architecture support, you must use either
2155 the HP assembler, gas/binutils 2.11 or a recent
2156 @uref{ftp://sources.redhat.com/pub/binutils/snapshots,,snapshot of gas}.
2158 More specific information to @samp{hppa*-hp-hpux*} targets follows.
2164 @heading @anchor{hppa*-hp-hpux9}hppa*-hp-hpux9
2166 The HP assembler has major problems on this platform. We've tried to work
2167 around the worst of the problems. However, those workarounds may be causing
2168 linker crashes in some circumstances; the workarounds also probably prevent
2169 shared libraries from working. Use the GNU assembler to avoid these problems.
2172 The configuration scripts for GCC will also trigger a bug in the hpux9
2173 shell. To avoid this problem set @env{CONFIG_SHELL} to @file{/bin/ksh}
2174 and @env{SHELL} to @file{/bin/ksh} in your environment.
2181 @heading @anchor{hppa*-hp-hpux10}hppa*-hp-hpux10
2183 For hpux10.20, we @emph{highly} recommend you pick up the latest sed patch
2184 @code{PHCO_19798} from HP@. HP has two sites which provide patches free of
2190 <a href="http://us-support.external.hp.com">US, Canada, Asia-Pacific, and
2194 @uref{http://us-support.external.hp.com,,}US, Canada, Asia-Pacific, and
2198 @uref{http://europe-support.external.hp.com,,Europe}
2201 The HP assembler on these systems is much better than the hpux9 assembler,
2202 but still has some problems. Most notably the assembler inserts timestamps
2203 into each object file it creates, causing the 3-stage comparison test to fail
2204 during a @samp{make bootstrap}. You should be able to continue by
2205 saying @samp{make all} after getting the failure from @samp{make
2213 @heading @anchor{hppa*-hp-hpux11}hppa*-hp-hpux11
2215 GCC 3.0 and up support HP-UX 11. You must use GNU binutils 2.11 or above on
2216 this platform. Thread support is not currently implemented for this
2217 platform, so @option{--enable-threads} does not work.
2218 See @uref{http://gcc.gnu.org/ml/gcc-prs/2002-01/msg00551.html}
2219 and @uref{http://gcc.gnu.org/ml/gcc-bugs/2002-01/msg00663.html}.
2220 GCC 2.95.x is not supported under HP-UX 11 and cannot be used to
2221 compile GCC 3.0 and up. Refer to @uref{binaries.html,,binaries} for
2222 information about obtaining precompiled GCC binaries for HP-UX.
2228 @heading @anchor{i370-*-*}i370-*-*
2229 This port is very preliminary and has many known bugs. We hope to
2230 have a higher-quality port for this machine soon.
2236 @heading @anchor{*-*-linux-gnu}*-*-linux-gnu
2238 If you use glibc 2.2 (or 2.1.9x), GCC 2.95.2 won't install
2239 out-of-the-box. You'll get compile errors while building @samp{libstdc++}.
2240 The patch @uref{glibc-2.2.patch,,glibc-2.2.patch}, that is to be
2241 applied in the GCC source tree, fixes the compatibility problems.
2251 Currently Glibc 2.2.3 (and older releases) and GCC 3.0 are out of sync
2252 since the latest exception handling changes for GCC@. Compiling glibc
2253 with GCC 3.0 will give a binary incompatible glibc and therefore cause
2254 lots of problems and might make your system completly unusable. This
2255 will definitly need fixes in glibc but might also need fixes in GCC@. We
2256 strongly advise to wait for glibc 2.2.4 and to read the release notes of
2257 glibc 2.2.4 whether patches for GCC 3.0 are needed. You can use glibc
2258 2.2.3 with GCC 3.0, just do not try to recompile it.
2264 @heading @anchor{ix86-*-linux*oldld}i?86-*-linux*oldld
2265 Use this configuration to generate @file{a.out} binaries on Linux-based
2266 GNU systems if you do not have gas/binutils version 2.5.2 or later
2269 This configuration is obsoleted in GCC 3.1.
2275 @heading @anchor{ix86-*-linux*aout}i?86-*-linux*aout
2276 Use this configuration to generate @file{a.out} binaries on Linux-based
2277 GNU systems. This configuration is being superseded. You must use
2278 gas/binutils version 2.5.2 or later.
2284 @heading @anchor{ix86-*-linux*}i?86-*-linux*
2286 You will need binutils 2.9.1.0.15 or newer for exception handling to work.
2288 If you receive Signal 11 errors when building on GNU/Linux, then it is
2289 possible you have a hardware problem. Further information on this can be
2290 found on @uref{http://www.bitwizard.nl/sig11/,,www.bitwizard.nl}.
2296 @heading @anchor{ix86-*-sco}i?86-*-sco
2297 Compilation with RCC is recommended. Also, it may be a good idea to
2298 link with GNU malloc instead of the malloc that comes with the system.
2304 @heading @anchor{ix86-*-sco3.2v4}i?86-*-sco3.2v4
2305 Use this configuration for SCO release 3.2 version 4.
2311 @heading @anchor{ix86-*-sco3.2v5*}i?86-*-sco3.2v5*
2312 Use this for the SCO OpenServer Release 5 family of operating systems.
2314 Unlike earlier versions of GCC, the ability to generate COFF with this
2315 target is no longer provided.
2317 Earlier versions of GCC emitted DWARF 1 when generating ELF to allow
2318 the system debugger to be used. That support was too burdensome to
2319 maintain. GCC now emits only DWARF 2 for this target. This means you
2320 may use either the UDK debugger or GDB to debug programs built by this
2323 Use of the @option{-march=pentiumpro} flag can result in
2324 unrecognized opcodes when using the native assembler on OS versions before
2325 5.0.6. (Support for P6 opcodes was added to the native ELF assembler in
2326 that version.) While it's rather rare to see these emitted by GCC yet,
2327 errors of the basic form:
2330 /usr/tmp/ccaNlqBc.s:22:unknown instruction: fcomip
2331 /usr/tmp/ccaNlqBc.s:50:unknown instruction: fucomip
2334 are symptoms of this problem. You may work around this by not
2335 building affected files with that flag, by using the GNU assembler, or
2336 by using the assembler provided with the current version of the OS@.
2337 Users of GNU assembler should see the note below for hazards on doing
2340 The native SCO assembler that is provided with the OS at no
2341 charge is normally required. If, however, you must be able to use
2342 the GNU assembler (perhaps you're compiling code with asms that
2343 require GAS syntax) you may configure this package using the flags
2344 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}}. You must
2345 use a recent version of GNU binutils; versions past 2.9.1 seem to work
2348 In general, the @option{--with-gnu-as} option isn't as well tested
2349 as the native assembler.
2351 Look in @file{gcc/config/i386/sco5.h} (search for ``messy'') for
2352 additional OpenServer-specific flags.
2354 Systems based on OpenServer before 5.0.4 (@samp{uname -X}
2355 will tell you what you're running) require TLS597 from
2356 @uref{ftp://ftp.sco.com/TLS/,,ftp://ftp.sco.com/TLS/}
2357 for C++ constructors and destructors to work right.
2359 The system linker in (at least) 5.0.4 and 5.0.5 will sometimes
2360 do the wrong thing for a construct that GCC will emit for PIC
2361 code. This can be seen as execution testsuite failures when using
2362 @option{-fPIC} on @file{921215-1.c}, @file{931002-1.c}, @file{nestfunc-1.c}, and @file{gcov-1.c}.
2363 For 5.0.5, an updated linker that will cure this problem is
2364 available. You must install both
2365 @uref{ftp://ftp.sco.com/Supplements/rs505a/,,ftp://ftp.sco.com/Supplements/rs505a/}
2366 and @uref{ftp://ftp.sco.com/SLS/,,OSS499A}.
2368 The dynamic linker in OpenServer 5.0.5 (earlier versions may show
2369 the same problem) aborts on certain G77-compiled programs. It's particularly
2370 likely to be triggered by building Fortran code with the @option{-fPIC} flag.
2371 Although it's conceivable that the error could be triggered by other
2372 code, only G77-compiled code has been observed to cause this abort.
2373 If you are getting core dumps immediately upon execution of your
2374 G77 program---and especially if it's compiled with @option{-fPIC}---try applying
2375 @uref{sco_osr5_g77.patch,,@file{sco_osr5_g77.patch}} to your @samp{libf2c} and
2377 Affected faults, when analyzed in a debugger, will show a stack
2378 backtrace with a fault occurring in @code{rtld()} and the program
2379 running as @file{/usr/lib/ld.so.1}. This problem has been reported to SCO
2380 engineering and will hopefully be addressed in later releases.
2387 @heading @anchor{ix86-*-udk}i?86-*-udk
2389 This target emulates the SCO Universal Development Kit and requires that
2390 package be installed. (If it is installed, you will have a
2391 @file{/udk/usr/ccs/bin/cc} file present.) It's very much like the
2392 @samp{i?86-*-unixware7*} target
2393 but is meant to be used when hosting on a system where UDK isn't the
2394 default compiler such as OpenServer 5 or Unixware 2. This target will
2395 generate binaries that will run on OpenServer, Unixware 2, or Unixware 7,
2396 with the same warnings and caveats as the SCO UDK@.
2398 This target is a little tricky to build because we have to distinguish
2399 it from the native tools (so it gets headers, startups, and libraries
2400 from the right place) while making the tools not think we're actually
2401 building a cross compiler. The easiest way to do this is with a configure
2404 @samp{CC=/udk/usr/ccs/bin/cc @var{/your/path/to}/gcc/configure
2405 --host=i686-pc-udk --target=i686-pc-udk --program-prefix=udk-}
2407 @emph{You should substitute @samp{i686} in the above command with the appropriate
2408 processor for your host.}
2410 After the usual @samp{make bootstrap} and
2411 @samp{make install}, you can then access the UDK-targeted GCC
2412 tools by adding @command{udk-} before the commonly known name. For
2413 example, to invoke the C compiler, you would use @command{udk-gcc}.
2414 They will coexist peacefully with any native-target GCC tools you may
2422 @heading @anchor{ix86-*-isc}i?86-*-isc
2423 This configuration is obsoleted in GCC 3.1.
2425 It may be a good idea to link with GNU malloc instead of the malloc that
2426 comes with the system.
2428 In ISC version 4.1, @command{sed} core dumps when building
2429 @file{deduced.h}. Use the version of @command{sed} from version 4.0.
2435 @heading @anchor{ix86-ibm-aix}i?86-ibm-aix
2436 This configuration is obsoleted in GCC 3.1.
2438 You need to use GAS version 2.1 or later, and LD from
2439 GNU binutils version 2.2 or later.
2445 @heading @anchor{ix86-sequent-bsd}i?86-sequent-bsd
2446 This configuration is obsoleted in GCC 3.1.
2448 Go to the Berkeley universe before compiling.
2454 @heading @anchor{ix86-sequent-ptx1*}i?86-sequent-ptx1*, i?86-sequent-ptx2*, i?86-sequent-sysv3*
2455 This configuration is obsoleted in GCC 3.1.
2457 You must install GNU @command{sed} before running @command{configure}.
2459 The @code{fixproto} shell script may trigger a bug in the system shell.
2460 If you encounter this problem, upgrade your operating system or
2461 use @command{bash} (the GNU shell) to run @code{fixproto}.
2467 @heading @anchor{i860-intel-osf*}i860-intel-osf*
2468 All support for the i860 processor is obsoleted in GCC 3.1.
2470 On the Intel Paragon (an i860 machine), if you are using operating
2471 system version 1.0, you will get warnings or errors about redefinition
2472 of @code{va_arg} when you build GCC@.
2474 If this happens, then you need to link most programs with the library
2475 @file{iclib.a}. You must also modify @file{stdio.h} as follows: before
2479 #if defined(__i860__) && !defined(_VA_LIST)
2480 #include <va_list.h>
2494 extern int vprintf(const char *, va_list );
2495 extern int vsprintf(char *, const char *, va_list );
2503 #endif /* __PGC__ */
2506 These problems don't exist in operating system version 1.1.
2512 @heading @anchor{ia64-*-linux}ia64-*-linux
2513 IA-64 processor (also known as IPF, or Itanium Processor Family)
2516 The toolchain is not completely finished, so requirements will continue
2518 GCC 3.0.1 and later require glibc 2.2.4.
2519 GCC 3.0.2 requires binutils from 2001-09-05 or later.
2520 GCC 3.0.1 requires binutils 2.11.1 or later.
2522 None of the following versions of GCC has an ABI that is compatible
2523 with any of the other versions in this list, with the exception that
2524 Red Hat 2.96 and Trillian 000171 are compatible with each other:
2525 3.0.2, 3.0.1, 3.0, Red Hat 2.96, and Trillian 000717.
2526 This primarily affects C++ programs and programs that create shared libraries.
2527 Because of these ABI incompatibilities, GCC 3.0.2 is not recommended for
2528 user programs on GNU/Linux systems built using earlier compiler releases.
2529 GCC 3.0.2 is recommended for compiling linux, the kernel.
2530 GCC 3.0.2 is believed to be fully ABI compliant, and hence no more major
2531 ABI changes are expected.
2537 @heading @anchor{*-lynx-lynxos}*-lynx-lynxos
2538 LynxOS 2.2 and earlier comes with GCC 1.x already installed as
2539 @file{/bin/gcc}. You should compile with this instead of @file{/bin/cc}.
2540 You can tell GCC to use the GNU assembler and linker, by specifying
2541 @samp{--with-gnu-as --with-gnu-ld} when configuring. These will produce
2542 COFF format object files and executables; otherwise GCC will use the
2543 installed tools, which produce @file{a.out} format executables.
2548 <!-- rs6000-ibm-aix*, powerpc-ibm-aix* -->
2550 @heading @anchor{*-ibm-aix*}*-ibm-aix*
2552 AIX Make frequently has problems with GCC makefiles. GNU Make 3.76 or
2553 newer is recommended to build on this platform.
2555 Errors involving @code{alloca} when building GCC generally are due
2556 to an incorrect definition of @code{CC} in the Makefile or mixing files
2557 compiled with the native C compiler and GCC@. During the stage1 phase of
2558 the build, the native AIX compiler @strong{must} be invoked as @command{cc}
2559 (not @command{xlc}). Once @command{configure} has been informed of
2560 @command{xlc}, one needs to use @samp{make distclean} to remove the
2561 configure cache files and ensure that @env{CC} environment variable
2562 does not provide a definition that will confuse @command{configure}.
2563 If this error occurs during stage2 or later, then the problem most likely
2564 is the version of Make (see above).
2566 The GNU Assembler incorrectly reports that it supports WEAK symbols on
2567 AIX which causes GCC to try to utilize weak symbol functionality which
2568 is not really supported on the platform. The native @command{as} and
2569 @command{ld} still are recommended. The native AIX tools do
2570 interoperate with GCC@.
2572 Building @file{libstdc++.a} requires a fix for a AIX Assembler bug
2573 APAR IY26685 (AIX 4.3) or APAR IY25528 (AIX 5.1).
2575 Linking executables and shared libraries may produce warnings of
2576 duplicate symbols. The assembly files generated by GCC for AIX always
2577 have included multiple symbol definitions for certain global variable
2578 and function declarations in the original program. The warnings should
2579 not prevent the linker from producing a correct library or runnable
2582 AIX 4.3 utilizes a ``large format'' archive to support both 32-bit and
2583 64-bit object modules. The routines provided in AIX 4.3.0 and AIX 4.3.1
2584 to parse archive libraries did not handle the new format correctly.
2585 These routines are used by GCC and result in error messages during
2586 linking such as ``not a COFF file''. The version of the routines shipped
2587 with AIX 4.3.1 should work for a 32-bit environment. The @option{-g}
2588 option of the archive command may be used to create archives of 32-bit
2589 objects using the original ``small format''. A correct version of the
2590 routines is shipped with AIX 4.3.2 and above.
2592 Some versions of the AIX binder (linker) can fail with a relocation
2593 overflow severe error when the @option{-bbigtoc} option is used to link
2594 GCC-produced object files into an executable that overflows the TOC@. A fix
2595 for APAR IX75823 (OVERFLOW DURING LINK WHEN USING GCC AND -BBIGTOC) is
2596 available from IBM Customer Support and from its
2597 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
2598 website as PTF U455193.
2600 The AIX 4.3.2.1 linker (bos.rte.bind_cmds Level 4.3.2.1) will dump core
2601 with a segmentation fault when invoked by any version of GCC@. A fix for
2602 APAR IX87327 is available from IBM Customer Support and from its
2603 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
2604 website as PTF U461879. This fix is incorporated in AIX 4.3.3 and above.
2606 The initial assembler shipped with AIX 4.3.0 generates incorrect object
2607 files. A fix for APAR IX74254 (64BIT DISASSEMBLED OUTPUT FROM COMPILER FAILS
2608 TO ASSEMBLE/BIND) is available from IBM Customer Support and from its
2609 @uref{http://techsupport.services.ibm.com/,,techsupport.services.ibm.com}
2610 website as PTF U453956. This fix is incorporated in AIX 4.3.1 and above.
2612 AIX provides National Language Support (NLS)@. Compilers and assemblers
2613 use NLS to support locale-specific representations of various data
2614 formats including floating-point numbers (e.g., @samp{.} vs @samp{,} for
2615 separating decimal fractions). There have been problems reported where
2616 GCC does not produce the same floating-point formats that the assembler
2617 expects. If one encounters this problem, set the @env{LANG}
2618 environment variable to @samp{C} or @samp{En_US}.
2620 By default, GCC for AIX 4.1 and above produces code that can be used on
2621 both Power or PowerPC processors.
2623 A default can be specified with the @option{-mcpu=@var{cpu_type}}
2624 switch and using the configure option @option{--with-cpu-@var{cpu_type}}.
2630 @heading @anchor{m32r-*-elf}m32r-*-elf
2631 Mitsubishi M32R processor.
2632 This configuration is intended for embedded systems.
2638 @heading @anchor{m68000-hp-bsd}m68000-hp-bsd
2639 HP 9000 series 200 running BSD@. Note that the C compiler that comes
2640 with this system cannot compile GCC; contact @email{law@@cygnus.com}
2641 to get binaries of GCC for bootstrapping.
2647 @heading @anchor{m6811-elf}m6811-elf
2648 Motorola 68HC11 family micro controllers. These are used in embedded
2649 applications. There are no standard Unix configurations.
2655 @heading @anchor{m6812-elf}m6812-elf
2656 Motorola 68HC12 family micro controllers. These are used in embedded
2657 applications. There are no standard Unix configurations.
2663 @heading @anchor{m68k-altos}m68k-altos
2664 Altos 3068. This configuration is obsoleted in GCC 3.1.
2666 You must use the GNU assembler, linker and debugger.
2667 Also, you must fix a kernel bug.
2673 @heading @anchor{m68k-apple-aux}m68k-apple-aux
2674 Apple Macintosh running A/UX@.
2675 This configuration is obsoleted in GCC 3.1.
2677 You may configure GCC to use either the system assembler and
2678 linker or the GNU assembler and linker. You should use the GNU configuration
2679 if you can, especially if you also want to use G++. You enable
2680 that configuration with the @option{--with-gnu-as} and @option{--with-gnu-ld}
2681 options to @code{configure}.
2683 Note the C compiler that comes
2684 with this system cannot compile GCC@. You can find binaries of GCC
2685 for bootstrapping on @code{jagubox.gsfc.nasa.gov}.
2686 You will also a patched version of @file{/bin/ld} there that
2687 raises some of the arbitrary limits found in the original.
2693 @heading @anchor{m68k-att-sysv}m68k-att-sysv
2694 AT&T 3b1, a.k.a.@: 7300 PC@. This version of GCC cannot
2695 be compiled with the system C compiler, which is too buggy.
2696 You will need to get a previous version of GCC and use it to
2697 bootstrap. Binaries are available from the OSU-CIS archive, at
2698 @uref{ftp://archive.cis.ohio-state.edu/pub/att7300/}.
2704 @heading @anchor{m68k-bull-sysv}m68k-bull-sysv
2705 Bull DPX/2 series 200 and 300 with BOS-2.00.45 up to BOS-2.01.
2706 This configuration is obsoleted in GCC 3.1.
2709 either with native assembler or GNU assembler. You can use
2710 GNU assembler with native COFF generation by providing @option{--with-gnu-as} to
2711 the configure script or use GNU assembler with stabs-in-COFF encapsulation
2712 by providing @samp{--with-gnu-as --stabs}. For any problem with the native
2713 assembler or for availability of the DPX/2 port of GAS, contact
2714 @email{F.Pierresteguy@@frcl.bull.fr}.
2720 @heading @anchor{m68k-crds-unos}m68k-crds-unos
2721 Use @samp{configure unos} for building on Unos.
2723 The Unos assembler is named @code{casm} instead of @code{as}. For some
2724 strange reason linking @file{/bin/as} to @file{/bin/casm} changes the
2725 behavior, and does not work. So, when installing GCC, you should
2726 install the following script as @file{as} in the subdirectory where
2727 the passes of GCC are installed:
2734 The default Unos library is named @file{libunos.a} instead of
2735 @file{libc.a}. To allow GCC to function, either change all
2736 references to @option{-lc} in @file{gcc.c} to @option{-lunos} or link
2737 @file{/lib/libc.a} to @file{/lib/libunos.a}.
2739 @cindex @code{alloca}, for Unos
2740 When compiling GCC with the standard compiler, to overcome bugs in
2741 the support of @code{alloca}, do not use @option{-O} when making stage 2.
2742 Then use the stage 2 compiler with @option{-O} to make the stage 3
2743 compiler. This compiler will have the same characteristics as the usual
2744 stage 2 compiler on other systems. Use it to make a stage 4 compiler
2745 and compare that with stage 3 to verify proper compilation.
2747 (Perhaps simply defining @code{ALLOCA} in @file{x-crds} as described in
2748 the comments there will make the above paragraph superfluous. Please
2749 inform us of whether this works.)
2751 Unos uses memory segmentation instead of demand paging, so you will need
2752 a lot of memory. 5 Mb is barely enough if no other tasks are running.
2753 If linking @file{cc1} fails, try putting the object files into a library
2754 and linking from that library.
2760 @heading @anchor{m68k-hp-hpux}m68k-hp-hpux
2761 HP 9000 series 300 or 400 running HP-UX@. HP-UX version 8.0 has a bug in
2762 the assembler that prevents compilation of GCC@. This
2763 bug manifests itself during the first stage of compilation, while
2764 building @file{libgcc2.a}:
2768 cc1: warning: `-g' option not supported on this version of GCC
2769 cc1: warning: `-g1' option not supported on this version of GCC
2770 ./xgcc: Internal compiler error: program as got fatal signal 11
2773 A patched version of the assembler is available as the file
2774 @uref{ftp://altdorf.ai.mit.edu/archive/cph/hpux-8.0-assembler}. If you
2775 have HP software support, the patch can also be obtained directly from
2776 HP, as described in the following note:
2779 This is the patched assembler, to patch SR#1653-010439, where the
2780 assembler aborts on floating point constants.
2782 The bug is not really in the assembler, but in the shared library
2783 version of the function ``cvtnum(3c)''. The bug on ``cvtnum(3c)'' is
2784 SR#4701-078451. Anyway, the attached assembler uses the archive
2785 library version of ``cvtnum(3c)'' and thus does not exhibit the bug.
2788 This patch is also known as PHCO_4484.
2790 In addition, if you wish to use gas, you must use
2791 gas version 2.1 or later, and you must use the GNU linker version 2.1 or
2792 later. Earlier versions of gas relied upon a program which converted the
2793 gas output into the native HP-UX format, but that program has not been
2794 kept up to date. gdb does not understand that native HP-UX format, so
2795 you must use gas if you wish to use gdb.
2797 On HP-UX version 8.05, but not on 8.07 or more recent versions, the
2798 @code{fixproto} shell script triggers a bug in the system shell. If you
2799 encounter this problem, upgrade your operating system or use BASH (the
2800 GNU shell) to run @code{fixproto}. This bug will cause the fixproto
2801 program to report an error of the form:
2804 ./fixproto: sh internal 1K buffer overflow
2807 To fix this, you can also change the first line of the fixproto script
2819 @heading @anchor{m68k-*-nextstep*}m68k-*-nextstep*
2820 These configurations are obsoleted in GCC 3.1.
2822 Current GCC versions probably do not work on version 2 of the NeXT
2825 On NeXTStep 3.0, the Objective-C compiler does not work, due,
2826 apparently, to a kernel bug that it happens to trigger. This problem
2827 does not happen on 3.1.
2829 You absolutely @strong{must} use GNU sed and GNU make on this platform.
2832 On NeXTSTEP 3.x where x < 3 the build of GCC will abort during
2833 stage1 with an error message like this:
2837 /usr/tmp/ccbbsZ0U.s:987:Unknown pseudo-op: .section
2838 /usr/tmp/ccbbsZ0U.s:987:Rest of line ignored. 1st junk character
2842 The reason for this is the fact that NeXT's assembler for these
2843 versions of the operating system does not support the @samp{.section}
2844 pseudo op that's needed for full C++ exception functionality.
2846 As NeXT's assembler is a derived work from GNU as, a free
2847 replacement that does can be obtained at
2848 @uref{ftp://ftp.next.peak.org:/next-ftp/next/apps/devtools/as.3.3.NIHS.s.tar.gz,,ftp://ftp.next.peak.org:/next-ftp/next/apps/devtools/as.3.3.NIHS.s.tar.gz}.
2850 If you try to build the integrated C++ & C++ runtime libraries on this system
2851 you will run into trouble with include files. The way to get around this is
2852 to use the following sequence. Note you must have write permission to
2853 the directory @var{prefix} you specified in the configuration process of GCC
2854 for this sequence to work.
2858 make all-texinfo all-bison all-byacc all-binutils all-gas all-ld
2861 make install-headers-tar
2870 @heading @anchor{m68k-ncr-*}m68k-ncr-*
2871 On the Tower models 4@var{n}0 and 6@var{n}0, by default a process is not
2872 allowed to have more than one megabyte of memory. GCC cannot compile
2873 itself (or many other programs) with @option{-O} in that much memory.
2875 To solve this problem, reconfigure the kernel adding the following line
2876 to the configuration file:
2887 @heading @anchor{m68k-sun}m68k-sun
2888 Sun 3. We do not provide a configuration file to use the Sun FPA by
2889 default, because programs that establish signal handlers for floating
2890 point traps inherently cannot work with the FPA@.
2896 @heading @anchor{m68k-sun-sunos4.1.1}m68k-sun-sunos4.1.1
2898 It is reported that you may need the GNU assembler on this platform.
2905 @heading @anchor{m88k-*-svr3}m88k-*-svr3
2906 Motorola m88k running the AT&T/Unisoft/Motorola V.3 reference port.
2907 These configurations are obsoleted in GCC 3.1.
2909 These systems tend to use the Green Hills C, revision 1.8.5, as the
2910 standard C compiler. There are apparently bugs in this compiler that
2911 result in object files differences between stage 2 and stage 3. If this
2912 happens, make the stage 4 compiler and compare it to the stage 3
2913 compiler. If the stage 3 and stage 4 object files are identical, this
2914 suggests you encountered a problem with the standard C compiler; the
2915 stage 3 and 4 compilers may be usable.
2917 It is best, however, to use an older version of GCC for bootstrapping
2924 @heading @anchor{m88k-*-dgux}m88k-*-dgux
2925 Motorola m88k running DG/UX@.
2926 These configurations are obsoleted in GCC 3.1.
2928 To build 88open BCS native or cross
2929 compilers on DG/UX, specify the configuration name as
2930 @samp{m88k-*-dguxbcs} and build in the 88open BCS software development
2931 environment. To build ELF native or cross compilers on DG/UX, specify
2932 @samp{m88k-*-dgux} and build in the DG/UX ELF development environment.
2933 You set the software development environment by issuing
2934 @samp{sde-target} command and specifying either @samp{m88kbcs} or
2935 @samp{m88kdguxelf} as the operand.
2937 If you do not specify a configuration name, @file{configure} guesses the
2938 configuration based on the current software development environment.
2944 @heading @anchor{m88k-tektronix-sysv3}m88k-tektronix-sysv3
2945 Tektronix XD88 running UTekV 3.2e.
2946 These configurations are obsoleted in GCC 3.1.
2949 optimization while building stage1 if you bootstrap with
2950 the buggy Green Hills compiler. Also, the bundled LAI
2951 System V NFS is buggy so if you build in an NFS mounted
2952 directory, start from a fresh reboot, or avoid NFS all together.
2953 Otherwise you may have trouble getting clean comparisons
2960 @heading @anchor{mips-*-*}mips-*-*
2961 If you use the 1.31 version of the MIPS assembler (such as was shipped
2962 with Ultrix 3.1), you will need to use the @option{-fno-delayed-branch} switch
2963 when optimizing floating point code. Otherwise, the assembler will
2964 complain when the GCC compiler fills a branch delay slot with a
2965 floating point instruction, such as @code{add.d}.
2967 If on a MIPS system you get an error message saying ``does not have gp
2968 sections for all it's [sic] sectons [sic]'', don't worry about it. This
2969 happens whenever you use GAS with the MIPS linker, but there is not
2970 really anything wrong, and it is okay to use the output file. You can
2971 stop such warnings by installing the GNU linker.
2973 It would be nice to extend GAS to produce the gp tables, but they are
2974 optional, and there should not be a warning about their absence.
2976 Users have reported some problems with version 2.0 of the MIPS
2977 compiler tools that were shipped with Ultrix 4.1. Version 2.10
2978 which came with Ultrix 4.2 seems to work fine.
2980 Users have also reported some problems with version 2.20 of the
2981 MIPS compiler tools that were shipped with RISC/os 4.x. The earlier
2982 version 2.11 seems to work fine.
2984 Some versions of the MIPS linker will issue an assertion failure
2985 when linking code that uses @code{alloca} against shared
2986 libraries on RISC-OS 5.0, and DEC's OSF/1 systems. This is a bug
2987 in the linker, that is supposed to be fixed in future revisions.
2988 To protect against this, GCC passes @option{-non_shared} to the
2989 linker unless you pass an explicit @option{-shared} or
2990 @option{-call_shared} switch.
2992 @heading @anchor{mips-mips-bsd}mips-mips-bsd
2993 MIPS machines running the MIPS operating system in BSD mode.
2994 These configurations are obsoleted in GCC 3.1.
2996 It's possible that some old versions of the system lack the functions
2997 @code{memcpy}, @code{memmove}, @code{memcmp}, and @code{memset}. If your
2998 system lacks these, you must remove or undo the definition of
2999 @code{TARGET_MEM_FUNCTIONS} in @file{mips-bsd.h}.
3001 If you use the MIPS C compiler to bootstrap, it may be necessary
3002 to increase its table size for switch statements with the
3003 @option{-Wf,-XNg1500} option. If you use the @option{-O2}
3004 optimization option, you also need to use @option{-Olimit 3000}.
3005 Both of these options are automatically generated in the
3006 @file{Makefile} that the shell script @file{configure} builds.
3007 If you override the @code{CC} make variable and use the MIPS
3008 compilers, you may need to add @option{-Wf,-XNg1500 -Olimit 3000}.
3014 @heading @anchor{mips-dec-*}mips-dec-*
3015 These configurations are obsoleted in GCC 3.1.
3017 MIPS-based DECstations can support three different personalities:
3018 Ultrix, DEC OSF/1, and OSF/rose. (Alpha-based DECstation products have
3019 a configuration name beginning with @samp{alpha*-dec}.) To configure GCC
3020 for these platforms use the following configurations:
3023 @item mips-dec-ultrix
3024 Ultrix configuration.
3027 DEC's version of OSF/1.
3029 @item mips-dec-osfrose
3030 Open Software Foundation reference port of OSF/1 which uses the
3031 OSF/rose object file format instead of ECOFF@. Normally, you
3032 would not select this configuration.
3035 If you use the MIPS C compiler to bootstrap, it may be necessary
3036 to increase its table size for switch statements with the
3037 @option{-Wf,-XNg1500} option. If you use the @option{-O2}
3038 optimization option, you also need to use @option{-Olimit 3000}.
3039 Both of these options are automatically generated in the
3040 @file{Makefile} that the shell script @file{configure} builds.
3041 If you override the @code{CC} make variable and use the MIPS
3042 compilers, you may need to add @option{-Wf,-XNg1500 -Olimit 3000}.
3048 @heading @anchor{mips-mips-riscos*}mips-mips-riscos*
3049 These configurations are obsoleted in GCC 3.1.
3051 If you use the MIPS C compiler to bootstrap, it may be necessary
3052 to increase its table size for switch statements with the
3053 @option{-Wf,-XNg1500} option. If you use the @option{-O2}
3054 optimization option, you also need to use @option{-Olimit 3000}.
3055 Both of these options are automatically generated in the
3056 @file{Makefile} that the shell script @file{configure} builds.
3057 If you override the @code{CC} make variable and use the MIPS
3058 compilers, you may need to add @samp{-Wf,-XNg1500 -Olimit 3000}.
3060 MIPS computers running RISC-OS can support four different
3061 personalities: default, BSD 4.3, System V.3, and System V.4
3062 (older versions of RISC-OS don't support V.4). To configure GCC
3063 for these platforms use the following configurations:
3066 @item mips-mips-riscos@var{rev}
3067 Default configuration for RISC-OS, revision @var{rev}.
3069 @item mips-mips-riscos@var{rev}bsd
3070 BSD 4.3 configuration for RISC-OS, revision @var{rev}.
3072 @item mips-mips-riscos@var{rev}sysv4
3073 System V.4 configuration for RISC-OS, revision @var{rev}.
3079 @item mips-mips-riscos@var{rev}sysv
3080 System V.3 configuration for RISC-OS, revision @var{rev}.
3083 The revision @code{rev} mentioned above is the revision of
3084 RISC-OS to use. You must reconfigure GCC when going from a
3085 RISC-OS revision 4 to RISC-OS revision 5. This has the effect of
3086 avoiding a linker bug.
3092 @heading @anchor{mips-sgi-irix4}mips-sgi-irix4
3093 This configuration is obsoleted in GCC 3.1.
3095 In order to compile GCC on an SGI running IRIX 4, the ``c.hdr.lib''
3096 option must be installed from the CD-ROM supplied from Silicon Graphics.
3097 This is found on the 2nd CD in release 4.0.1.
3099 On IRIX version 4.0.5F, and perhaps on some other versions as well,
3100 there is an assembler bug that reorders instructions incorrectly. To
3101 work around it, specify the target configuration
3102 @samp{mips-sgi-irix4loser}. This configuration inhibits assembler
3105 In a compiler configured with target @samp{mips-sgi-irix4}, you can turn
3106 off assembler optimization by using the @option{-noasmopt} option. This
3107 compiler option passes the option @option{-O0} to the assembler, to
3110 The @option{-noasmopt} option can be useful for testing whether a problem
3111 is due to erroneous assembler reordering. Even if a problem does not go
3112 away with @option{-noasmopt}, it may still be due to assembler
3113 reordering---perhaps GCC itself was miscompiled as a result.
3115 You may get the following warning on IRIX 4 platforms, it can be safely
3118 warning: foo.o does not have gp tables for all its sections.
3125 @heading @anchor{mips-sgi-irix5}mips-sgi-irix5
3127 This configuration has considerable problems, which will be fixed in a
3130 In order to compile GCC on an SGI running IRIX 5, the ``compiler_dev.hdr''
3131 subsystem must be installed from the IDO CD-ROM supplied by Silicon
3132 Graphics. It is also available for download from
3133 @uref{http://www.sgi.com/developers/devtools/apis/ido.html,,http://www.sgi.com/developers/devtools/apis/ido.html}.
3135 @code{make compare} may fail on version 5 of IRIX unless you add
3136 @option{-save-temps} to @code{CFLAGS}. On these systems, the name of the
3137 assembler input file is stored in the object file, and that makes
3138 comparison fail if it differs between the @code{stage1} and
3139 @code{stage2} compilations. The option @option{-save-temps} forces a
3140 fixed name to be used for the assembler input file, instead of a
3141 randomly chosen name in @file{/tmp}. Do not add @option{-save-temps}
3142 unless the comparisons fail without that option. If you do you
3143 @option{-save-temps}, you will have to manually delete the @samp{.i} and
3144 @samp{.s} files after each series of compilations.
3146 If you use the MIPS C compiler to bootstrap, it may be necessary
3147 to increase its table size for switch statements with the
3148 @option{-Wf,-XNg1500} option. If you use the @option{-O2}
3149 optimization option, you also need to use @option{-Olimit 3000}.
3151 To enable debugging under IRIX 5, you must use GNU @command{as} 2.11.2
3153 and use the @option{--with-gnu-as} configure option when configuring GCC.
3154 GNU @command{as} is distributed as part of the binutils package.
3155 When using release 2.11.2, you need to apply a patch
3156 @uref{http://sources.redhat.com/ml/binutils/2001-07/msg00352.html,,http://sources.redhat.com/ml/binutils/2001-07/msg00352.html}
3157 which will be included in the next release of binutils.
3159 When building GCC, the build process loops rebuilding @command{cc1} over
3160 and over again. This happens on @samp{mips-sgi-irix5.2}, and possibly
3161 other platforms. It has been reported that this is a known bug in the
3162 @command{make} shipped with IRIX 5.2. We recommend you use GNU
3163 @command{make} instead of the vendor supplied @command{make} program;
3164 however, you may have success with @command{smake} on IRIX 5.2 if you do
3165 not have GNU @command{make} available.
3171 @heading @anchor{mips-sgi-irix6}mips-sgi-irix6
3173 If you are using IRIX @command{cc} as your bootstrap compiler, you must
3174 ensure that the N32 ABI is in use. To test this, compile a simple C
3175 file with @command{cc} and then run @command{file} on the
3176 resulting object file. The output should look like:
3179 test.o: ELF N32 MSB @dots{}
3185 test.o: ELF 32-bit MSB @dots{}
3191 test.o: ELF 64-bit MSB @dots{}
3194 then your version of @command{cc} uses the O32 or N64 ABI by default. You
3195 should set the environment variable @env{CC} to @samp{cc -n32}
3196 before configuring GCC@.
3198 GCC on IRIX 6 is usually built to support both the N32 and N64 ABIs. If
3199 you build GCC on a system that doesn't have the N64 libraries installed,
3200 you need to configure with @option{--disable-multilib} so GCC doesn't
3201 try to use them. Look for @file{/usr/lib64/libc.so.1} to see if you
3202 have the 64-bit libraries installed.
3204 You must @emph{not} use GNU @command{as} (which isn't built anyway as of
3205 binutils 2.11.2) on IRIX 6 platforms; doing so will only cause problems.
3207 GCC does not currently support generating O32 ABI binaries in the
3208 @samp{mips-sgi-irix6} configurations. It is possible to create a GCC
3209 with O32 ABI only support by configuring it for the @samp{mips-sgi-irix5}
3210 target and using a patched GNU @command{as} 2.11.2 as documented in the
3211 @uref{#mips-sgi-irix5,,@samp{mips-sgi-irix5}} section above. Using the
3212 native assembler requires patches to GCC which will be included in a
3213 future release. It is
3214 expected that O32 ABI support will be available again in a future release.
3216 The @option{--enable-threads} option doesn't currently work, a patch is
3217 in preparation for a future release. The @option{--enable-libgcj}
3218 option is disabled by default: IRIX 6 uses a very low default limit
3219 (20480) for the command line length. Although libtool contains a
3220 workaround for this problem, at least the N64 @samp{libgcj} is known not
3221 to build despite this, running into an internal error of the native
3222 @command{ld}. A sure fix is to increase this limit (@samp{ncargs}) to
3223 its maximum of 262144 bytes. If you have root access, you can use the
3224 @command{systune} command to do this.
3226 GCC does not correctly pass/return structures which are
3227 smaller than 16 bytes and which are not 8 bytes. The problem is very
3228 involved and difficult to fix. It affects a number of other targets also,
3229 but IRIX 6 is affected the most, because it is a 64 bit target, and 4 byte
3230 structures are common. The exact problem is that structures are being padded
3231 at the wrong end, e.g.@: a 4 byte structure is loaded into the lower 4 bytes
3232 of the register when it should be loaded into the upper 4 bytes of the
3235 GCC is consistent with itself, but not consistent with the SGI C compiler
3236 (and the SGI supplied runtime libraries), so the only failures that can
3237 happen are when there are library functions that take/return such
3238 structures. There are very few such library functions. Currently this
3239 is known to affect @code{inet_ntoa}, @code{inet_lnaof},
3240 @code{inet_netof}, @code{inet_makeaddr}, and @code{semctl}. Until the
3241 bug is fixed, GCC contains workarounds for the known affected functions.
3243 See @uref{http://freeware.sgi.com/,,http://freeware.sgi.com/} for more
3244 information about using GCC on IRIX platforms.
3250 @heading @anchor{mips-sony-sysv}mips-sony-sysv
3251 Sony MIPS NEWS@. This configuration is obsoleted in GCC 3.1.
3253 This works in NEWSOS 5.0.1, but not in 5.0.2 (which uses ELF instead of
3254 COFF)@. In particular, the linker does not like the code generated by
3255 GCC when shared libraries are linked in.
3261 @heading @anchor{ns32k-encore}ns32k-encore
3262 This configuration is obsoleted in GCC 3.1.
3264 Encore ns32000 system. Encore systems are supported only under BSD@.
3270 @heading @anchor{ns32k-*-genix}ns32k-*-genix
3271 National Semiconductor ns32000 system. This configuration is obsoleted
3274 Genix has bugs in @code{alloca} and @code{malloc}; you must get the
3275 compiled versions of these from GNU Emacs.
3281 @heading @anchor{ns32k-sequent}ns32k-sequent
3282 This configuration is obsoleted in GCC 3.1.
3284 Go to the Berkeley universe before compiling.
3290 @heading @anchor{ns32k-utek}ns32k-utek
3291 UTEK ns32000 system (``merlin''). This configuration is obsoleted in
3294 The C compiler that comes with this system cannot compile GCC; contact
3295 @samp{tektronix!reed!mason} to get binaries of GCC for bootstrapping.
3302 @heading @anchor{powerpc*-*-*}powerpc-*-*
3304 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
3305 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
3311 @heading @anchor{powerpc-*-darwin*}powerpc-*-darwin*
3312 PowerPC running Darwin (Mac OS X kernel).
3314 GCC 3.0 does not support Darwin, but 3.1 and later releases will work.
3316 Pre-installed versions of Mac OS X may not include any developer tools,
3317 meaning that you will not be able to build GCC from source. Tool
3318 binaries are available at
3319 @uref{http://www.opensource.apple.com/projects/darwin} (free
3320 registration required).
3322 Versions of the assembler prior to ``cctools-364'' cannot handle the
3323 4-argument form of rlwinm and related mask-using instructions. Darwin
3324 1.3 (Mac OS X 10.0) uses cctools-353 for instance. To get cctools-364,
3325 check out @file{cctools} with tag @samp{Apple-364}, build it, and
3326 install the assembler as @file{usr/bin/as}. See
3327 @uref{http://www.opensource.apple.com/tools/cvs/docs.html} for details.
3329 Also, the default stack limit of 512K is too small, and a bootstrap will
3330 typically fail when self-compiling @file{expr.c}. Set the stack to 800K
3331 or more, for instance by doing @samp{limit stack 800}. It's also
3332 convenient to use the GNU preprocessor instead of Apple's during the
3333 first stage of bootstrapping; this is automatic when doing @samp{make
3334 bootstrap}, but to do it from the toplevel objdir you will need to say
3335 @samp{make CC='cc -no-cpp-precomp' bootstrap}.
3337 Note that the version of GCC shipped by Apple typically includes a
3338 number of extensions not available in a standard GCC release. These
3339 extensions are generally specific to Mac programming.
3345 @heading @anchor{powerpc-*-elf}powerpc-*-elf, powerpc-*-sysv4
3346 PowerPC system in big endian mode, running System V.4.
3352 @heading @anchor{powerpc-*-linux-gnu*}powerpc-*-linux-gnu*
3355 @uref{ftp://ftp.kernel.org/pub/linux/devel/binutils,,binutils 2.12.90.0.7}
3356 or newer for a working GCC@.
3362 @heading @anchor{powerpc-*-netbsd*}powerpc-*-netbsd*
3363 PowerPC system in big endian mode running NetBSD@. To build the
3364 documentation you will need Texinfo version 4.1 (NetBSD 1.5.1 included
3365 Texinfo version 3.12).
3371 @heading @anchor{powerpc-*-eabiaix}powerpc-*-eabiaix
3372 Embedded PowerPC system in big endian mode with @option{-mcall-aix} selected as
3379 @heading @anchor{powerpc-*-eabisim}powerpc-*-eabisim
3380 Embedded PowerPC system in big endian mode for use in running under the
3387 @heading @anchor{powerpc-*-eabi}powerpc-*-eabi
3388 Embedded PowerPC system in big endian mode.
3394 @heading @anchor{powerpcle-*-elf}powerpcle-*-elf, powerpcle-*-sysv4
3395 PowerPC system in little endian mode, running System V.4.
3401 @heading @anchor{powerpcle-*-eabisim}powerpcle-*-eabisim
3402 Embedded PowerPC system in little endian mode for use in running under
3409 @heading @anchor{powerpcle-*-eabi}powerpcle-*-eabi
3410 Embedded PowerPC system in little endian mode.
3416 @heading @anchor{powerpcle-*-winnt}powerpcle-*-winnt, powerpcle-*-pe
3417 PowerPC system in little endian mode running Windows NT@.
3423 @heading @anchor{romp-*-aos}romp-*-aos, romp-*-mach
3424 These configurations are obsoleted in GCC 3.1.
3426 We recommend you compile GCC with an earlier version of itself; if you
3427 compile GCC with @command{hc}, the Metaware compiler, it will work, but
3428 you will get mismatches between the stage 2 and stage 3 compilers in
3429 various files. These errors are minor differences in some
3430 floating-point constants and can be safely ignored; the stage 3 compiler
3437 @heading @anchor{s390-*-linux*}s390-*-linux*
3438 S/390 system running Linux for S/390@.
3444 @heading @anchor{s390x-*-linux*}s390x-*-linux*
3445 zSeries system (64 Bit) running Linux for zSeries@.
3451 @c Please use Solaris 2 to refer to all release of Solaris, starting
3452 @c with 2.0 until 2.6, 7, and 8. Solaris 1 was a marketing name for
3453 @c SunOS 4 releases which we don't use to avoid confusion. Solaris
3454 @c alone is too unspecific and must be avoided.
3455 @heading @anchor{*-*-solaris2*}*-*-solaris2*
3457 Sun does not ship a C compiler with Solaris 2. To bootstrap and install
3458 GCC you first have to install a pre-built compiler, see our
3459 @uref{binaries.html,,binaries page} for details.
3461 The Solaris 2 @command{/bin/sh} will often fail to configure
3462 @file{libstdc++-v3}, @file{boehm-gc} or
3463 @file{libjava}. If you encounter this problem, set @env{CONFIG_SHELL} to
3464 @command{/bin/ksh} in your environment and run @command{make bootstrap} again.
3465 Another possibility that sometimes helps is to remove
3466 @file{*-*-solaris2*/config.cache}.
3468 Solaris 2 comes with a number of optional OS packages. Some of these
3469 packages are needed to use GCC fully, namely @code{SUNWarc},
3470 @code{SUNWbtool}, @code{SUNWesu}, @code{SUNWhea}, @code{SUNWlibm},
3471 @code{SUNWsprot}, and @code{SUNWtoo}. If you did not install all
3472 optional packages when installing Solaris 2, you will need to verify that
3473 the packages that GCC needs are installed.
3475 To check whether an optional package is installed, use
3476 the @command{pkginfo} command. To add an optional package, use the
3477 @command{pkgadd} command. For further details, see the Solaris 2
3480 Trying to use the linker and other tools in
3481 @file{/usr/ucb} to install GCC has been observed to cause trouble.
3482 For example, the linker may hang indefinitely. The fix is to remove
3483 @file{/usr/ucb} from your @env{PATH}.
3485 All releases of GNU binutils prior to 2.11.2 have known bugs on this
3486 platform. We recommend the use of GNU binutils 2.11.2 or the vendor
3487 tools (Sun @command{as}, Sun @command{ld}).
3489 Sun bug 4296832 turns up when compiling X11 headers with GCC 2.95 or
3490 newer: @command{g++} will complain that types are missing. These headers assume
3491 that omitting the type means @code{int}; this assumption worked for C89 but
3492 is wrong for C++, and is now wrong for C99 also.
3494 @command{g++} accepts such (invalid) constructs with the option
3495 @option{-fpermissive}; it
3496 will assume that any missing type is @code{int} (as defined by C89).
3498 There are patches for Solaris 2.6 (105633-56 or newer for SPARC,
3499 106248-42 or newer for Intel), Solaris 7 (108376-21 or newer for SPARC,
3500 108377-20 for Intel), and Solaris 8 (108652-24 or newer for SPARC,
3501 108653-22 for Intel) that fix this bug.
3507 @heading @anchor{sparc-sun-solaris2*}sparc-sun-solaris2*
3509 When GCC is configured to use binutils 2.11.2 or later the binaries
3510 produced are smaller than the ones produced using Sun's native tools;
3511 this difference is quite significant for binaries containing debugging
3514 Sun @command{as} 4.x is broken in that it cannot cope with long symbol names.
3515 A typical error message might look similar to the following:
3517 @samp{/usr/ccs/bin/as: "/var/tmp/ccMsw135.s", line 11041:
3518 error: can't compute value of an expression involving an external symbol.}
3520 This is Sun bug 4237974. This is fixed with patch 108908-02 for Solaris
3521 2.6 and has been fixed in later (5.x) versions of the assembler,
3522 starting with Solaris 7.
3524 Starting with Solaris 7, the operating system is capable of executing
3525 64-bit SPARC V9 binaries. GCC 3.1 and later properly supports
3526 this; the @option{-m64} option enables 64-bit code generation.
3527 However, if all you want is code tuned for the UltraSPARC CPU, you
3528 should try the @option{-mtune=ultrasparc} option instead, which produces
3529 code that, unlike full 64-bit code, can still run on non-UltraSPARC
3532 When configuring on a Solaris 7 or later system that is running a kernel
3533 that supports only 32-bit binaries, one must configure with
3534 @option{--disable-multilib}, since we will not be able to build the
3535 64-bit target libraries.
3541 @heading @anchor{sparc-sun-solaris2.7}sparc-sun-solaris2.7
3543 Sun patch 107058-01 (1999-01-13) for Solaris 7/SPARC triggers a bug in
3544 the dynamic linker. This problem (Sun bug 4210064) affects GCC 2.8
3545 and later, including all EGCS releases. Sun formerly recommended
3546 107058-01 for all Solaris 7 users, but around 1999-09-01 it started to
3547 recommend it only for people who use Sun's compilers.
3549 Here are some workarounds to this problem:
3552 Do not install Sun patch 107058-01 until after Sun releases a
3553 complete patch for bug 4210064. This is the simplest course to take,
3554 unless you must also use Sun's C compiler. Unfortunately 107058-01
3555 is preinstalled on some new Solaris 7-based hosts, so you may have to
3559 Copy the original, unpatched Solaris 7
3560 @command{/usr/ccs/bin/as} into
3561 @command{/usr/local/lib/gcc-lib/sparc-sun-solaris2.7/3.1/as},
3562 adjusting the latter name to fit your local conventions and software
3566 Install Sun patch 106950-03 (1999-05-25) or later. Nobody with
3567 both 107058-01 and 106950-03 installed has reported the bug with GCC
3568 and Sun's dynamic linker. This last course of action is riskiest,
3569 for two reasons. First, you must install 106950 on all hosts that
3570 run code generated by GCC; it doesn't suffice to install it only on
3571 the hosts that run GCC itself. Second, Sun says that 106950-03 is
3572 only a partial fix for bug 4210064, but Sun doesn't know whether the
3573 partial fix is adequate for GCC@. Revision -08 or later should fix
3574 the bug. The current (as of 2001-09-24) revision is -14, and is included in
3575 the Solaris 7 Recommended Patch Cluster.
3583 @heading @anchor{sparc-sun-sunos4*}sparc-sun-sunos4*
3585 A bug in the SunOS 4 linker will cause it to crash when linking
3586 @option{-fPIC} compiled objects (and will therefore not allow you to build
3589 To fix this problem you can either use the most recent version of
3590 binutils or get the latest SunOS 4 linker patch (patch ID 100170-10)
3591 from Sun's patch site.
3593 Sometimes on a Sun 4 you may observe a crash in the program
3594 @command{genflags} or @command{genoutput} while building GCC. This is said to
3595 be due to a bug in @command{sh}. You can probably get around it by running
3596 @command{genflags} or @command{genoutput} manually and then retrying the
3603 @heading @anchor{sparc-unknown-linux-gnulibc1}sparc-unknown-linux-gnulibc1
3605 It has been reported that you might need
3606 @uref{ftp://ftp.yggdrasil.com/private/hjl,,binutils 2.8.1.0.23}
3607 for this platform, too.
3614 @heading @anchor{sparc-*-linux*}sparc-*-linux*
3616 GCC versions 3.0 and higher require binutils 2.11.2 and glibc 2.2.4
3617 or newer on this platform. All earlier binutils and glibc
3618 releases mishandled unaligned relocations on @code{sparc-*-*} targets.
3625 @heading @anchor{sparc64-*-*}sparc64-*-*
3627 GCC version 2.95 is not able to compile code correctly for
3628 @code{sparc64} targets. Users of the Linux kernel, at least,
3629 can use the @code{sparc32} program to start up a new shell
3630 invocation with an environment that causes @command{configure} to
3631 recognize (via @samp{uname -a}) the system as @samp{sparc-*-*} instead.
3637 @heading @anchor{sparcv9-*-solaris2*}sparcv9-*-solaris2*
3639 The following compiler flags must be specified in the configure
3640 step in order to bootstrap this target with the Sun compiler:
3643 % CC="cc -xildoff -xarch=v9" @var{srcdir}/configure [@var{options}] [@var{target}]
3646 @option{-xildoff} turns off the incremental linker, and @option{-xarch=v9}
3647 specifies the v9 architecture to the Sun linker and assembler.
3653 @heading @anchor{#*-*-sysv*}*-*-sysv*
3654 On System V release 3, you may get this error message
3658 ld fatal: failed to write symbol name @var{something}
3659 in strings table for file @var{whatever}
3662 This probably indicates that the disk is full or your ulimit won't allow
3663 the file to be as large as it needs to be.
3665 This problem can also result because the kernel parameter @code{MAXUMEM}
3666 is too small. If so, you must regenerate the kernel and make the value
3667 much larger. The default value is reported to be 1024; a value of 32768
3668 is said to work. Smaller values may also work.
3670 On System V, if you get an error like this,
3673 /usr/local/lib/bison.simple: In function `yyparse':
3674 /usr/local/lib/bison.simple:625: virtual memory exhausted
3678 that too indicates a problem with disk space, ulimit, or @code{MAXUMEM}.
3680 On a System V release 4 system, make sure @file{/usr/bin} precedes
3681 @file{/usr/ucb} in @code{PATH}. The @code{cc} command in
3682 @file{/usr/ucb} uses libraries which have bugs.
3688 @heading @anchor{vax-dec-ultrix}vax-dec-ultrix
3689 Don't try compiling with VAX C (@code{vcc}). It produces incorrect code
3690 in some cases (for example, when @code{alloca} is used).
3696 @heading @anchor{we32k-*-*}we32k-*-*
3697 These computers are also known as the 3b2, 3b5, 3b20 and other similar
3698 names. (However, the 3b1 is actually a 68000.)
3699 These configurations are obsoleted in GCC 3.1.
3701 Don't use @option{-g} when compiling with the system's compiler. The
3702 system's linker seems to be unable to handle such a large program with
3703 debugging information.
3705 The system's compiler runs out of capacity when compiling @file{stmt.c}
3706 in GCC@. You can work around this by building @file{cpp} in GCC
3707 first, then use that instead of the system's preprocessor with the
3708 system's C compiler to compile @file{stmt.c}. Here is how:
3711 mv /lib/cpp /lib/cpp.att
3713 echo '/lib/cpp.gnu -traditional $@{1+"$@@"@}' > /lib/cpp
3717 The system's compiler produces bad code for some of the GCC
3718 optimization files. So you must build the stage 2 compiler without
3719 optimization. Then build a stage 3 compiler with optimization.
3720 That executable should work. Here are the necessary commands:
3723 make LANGUAGES=c CC=stage1/xgcc CFLAGS="-Bstage1/ -g"
3725 make CC=stage2/xgcc CFLAGS="-Bstage2/ -g -O"
3728 You may need to raise the ULIMIT setting to build a C++ compiler,
3729 as the file @file{cc1plus} is larger than one megabyte.
3735 @heading @anchor{xtensa-*-elf}xtensa-*-elf
3737 This target is intended for embedded Xtensa systems using the
3738 @samp{newlib} C library. It uses ELF but does not support shared
3739 objects. Designed-defined instructions specified via the
3740 Tensilica Instruction Extension (TIE) language are only supported
3741 through inline assembly.
3743 The Xtensa configuration information must be specified prior to
3744 building GCC@. The @file{gcc/config/xtensa/xtensa-config.h} header
3745 file contains the configuration information. If you created your
3746 own Xtensa configuration with the Xtensa Processor Generator, the
3747 downloaded files include a customized copy of this header file,
3748 which you can use to replace the default header file.
3754 @heading @anchor{xtensa-*-linux*}xtensa-*-linux*
3756 This target is for Xtensa systems running GNU/Linux. It supports ELF
3757 shared objects and the GNU C library (glibc). It also generates
3758 position-independent code (PIC) regardless of whether the
3759 @option{-fpic} or @option{-fPIC} options are used. In other
3760 respects, this target is the same as the
3761 @uref{#xtensa-*-elf,,@samp{xtensa-*-elf}} target.
3767 @heading @anchor{windows}Microsoft Windows (32 bit)
3769 A port of GCC 2.95.x is included with the
3770 @uref{http://www.cygwin.com/,,Cygwin environment}.
3772 Current (as of early 2001) snapshots of GCC will build under Cygwin
3773 without modification.
3779 @heading @anchor{os2}OS/2
3781 GCC does not currently support OS/2. However, Andrew Zabolotny has been
3782 working on a generic OS/2 port with pgcc. The current code can be found
3783 at @uref{http://www.goof.com/pcg/os2/,,http://www.goof.com/pcg/os2/}.
3785 An older copy of GCC 2.8.1 is included with the EMX tools available at
3786 @uref{ftp://ftp.leo.org/pub/comp/os/os2/leo/devtools/emx+gcc/,,
3787 ftp://ftp.leo.org/pub/comp/os/os2/leo/devtools/emx+gcc/}.
3793 @heading @anchor{older}Older systems
3795 GCC contains support files for many older (1980s and early
3796 1990s) Unix variants. For the most part, support for these systems
3797 has not been deliberately removed, but it has not been maintained for
3798 several years and may suffer from bitrot. Support from some systems
3799 has been removed from GCC 3: fx80, ns32-ns-genix, pyramid, tahoe,
3800 gmicro, spur; most of these targets had not been updated since GCC
3803 We are planning to remove support for more older systems, starting in
3804 GCC 3.1. Each release will have a list of ``obsoleted'' systems.
3805 Support for these systems is still present in that release, but
3806 @command{configure} will fail unless the @option{--enable-obsolete}
3807 option is given. Unless a maintainer steps forward, support for
3808 these systems will be removed from the next release of GCC@.
3810 Support for older systems as targets for cross-compilation is less
3811 problematic than support for them as hosts for GCC; if an enthusiast
3812 wishes to make such a target work again (including resurrecting any
3813 of the targets that never worked with GCC 2, starting from the last
3814 CVS version before they were removed), patches
3815 @uref{../contribute.html,,following the usual requirements}
3816 would be likely to be accepted, since they should not affect the
3817 support for more modern targets.
3819 Support for old systems as hosts for GCC can cause problems if the
3820 workarounds for compiler, library and operating system bugs affect the
3821 cleanliness or maintainability of the rest of GCC@. In some cases, to
3822 bring GCC up on such a system, if still possible with current GCC, may
3823 require first installing an old version of GCC which did work on that
3824 system, and using it to compile a more recent GCC, to avoid bugs in
3825 the vendor compiler. Old releases of GCC 1 and GCC 2 are available in
3826 the @file{old-releases} directory on the
3827 @uref{../mirrors.html,,GCC mirror sites}. Header bugs may generally
3828 be avoided using @command{fixincludes}, but bugs or deficiencies in
3829 libraries and the operating system may still cause problems.
3831 For some systems, old versions of GNU binutils may also be useful,
3832 and are available from @file{pub/binutils/old-releases} on
3833 @uref{http://sources.redhat.com/mirrors.html,,sources.redhat.com mirror sites}.
3835 Some of the information on specific systems above relates to
3836 such older systems, but much of the information
3837 about GCC on such systems (which may no longer be applicable to
3838 current GCC) is to be found in the GCC texinfo manual.
3844 @heading @anchor{elf_targets}all ELF targets (SVR4, Solaris 2, etc.)
3846 C++ support is significantly better on ELF targets if you use the
3847 @uref{./configure.html#with-gnu-ld,,GNU linker}; duplicate copies of
3848 inlines, vtables and template instantiations will be discarded
3858 @uref{./index.html,,Return to the GCC Installation page}
3862 @c ***Old documentation******************************************************
3864 @include install-old.texi
3871 @uref{./index.html,,Return to the GCC Installation page}
3875 @c ***GFDL********************************************************************
3884 @uref{./index.html,,Return to the GCC Installation page}
3888 @c ***************************************************************************
3889 @c Part 6 The End of the Document
3891 @comment node-name, next, previous, up
3892 @node Concept Index, , GNU Free Documentation License, Top
3896 @unnumbered Concept Index