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 @c Copyright (C) 2001 Free Software Foundation, Inc.
37 @c *** Converted to texinfo by Dean Wakerley, dean@wakerley.com
39 @c Include everything if we're not making html
51 @c Part 2 Summary Description and Copyright
54 Copyright @copyright{} 2001 Free Software Foundation, Inc.
57 @c Part 3 Titlepage and Copyright
60 @comment The title is printed in a large font.
61 @center @titlefont{Installing GCC}
63 @c The following two commands start the copyright page.
65 @vskip 0pt plus 1filll
66 Copyright @copyright{} 2001 Free Software Foundation, Inc.
69 @c Part 4 Top node and Master Menu
72 @comment node-name, next, Previous, up
75 * Installing GCC:: This document describes the generic installation
76 procedure for GCC as well as detailing some target
77 specific installation instructions.
79 * Specific:: Host/target specific installation notes for GCC.
80 * Binaries:: Where to get pre-compiled binaries.
82 * Concept Index:: This index has two entries.
86 @c Part 5 The Body of the Document
87 @c ***Installing GCC**********************************************************
89 @comment node-name, next, previous, up
90 @node Installing GCC, Binaries, , Top
94 <h1 align="center">Installing GCC</h1>
97 @chapter Installing GCC
100 The latest version of this document is always available at
101 @uref{http://gcc.gnu.org/install/,,http://gcc.gnu.org/install/}.
103 This document describes the generic installation procedure for GCC as well
104 as detailing some target specific installation instructions.
106 GCC includes several components that previously were separate distributions
107 with their own installation instructions. This document supersedes all
108 package specific installation instructions. We provide the component
109 specific installation information in the source distribution for historical
110 reference purposes only.
112 @emph{Before} starting the build/install procedure please check the
114 @xref{Specific, host/target specific installation notes}.
117 @uref{specific.html,,host/target specific installation notes}.
119 We recommend you browse the entire generic installation instructions before
122 The installation procedure itself is broken into five steps.
126 * Downloading the source::
129 * Testing:: (optional)
136 @uref{download.html,,Downloading the source}
138 @uref{configure.html,,Configuration}
140 @uref{build.html,,Building}
142 @uref{test.html,,Testing} (optional)
144 @uref{finalinstall.html,,Final install}
148 Please note that GCC does not support @samp{make uninstall} and probably
149 won't do so in the near future as this would open a can of worms. Instead,
150 we suggest that you install GCC into a directory of its own and simply
151 remove that directory when you do not need that specific version of GCC
159 @uref{./index.html,,Return to the GCC Installation page}
163 @c ***Downloading the source**************************************************
165 @comment node-name, next, previous, up
166 @node Downloading the source, Configuration, , Installing GCC
170 <h1 align="center">Downloading GCC</h1>
173 @chapter Downloading GCC
175 @cindex Downloading GCC
176 @cindex Downloading the Source
178 GCC is distributed via CVS and FTP tarballs compressed with @command{gzip} or
179 @command{bzip2}. It is possible to download a full distribution or specific
182 Please refer to our @uref{http://gcc.gnu.org/releases.html,,releases web page}
183 for information on how to obtain GCC@.
185 The full distribution includes the C, C++, Objective-C, Fortran, Java,
186 and Chill compilers. The full distribution also includes runtime libraries
187 for C++, Objective-C, Fortran, Java and Chill. (GCC 3.0 does not
188 include Chill. Releases before 3.0 do not include the Java runtime
189 library.) In GCC 3.0 and later versions, GNU compiler testsuites
190 are also included in the full distribution.
192 If you choose to download specific components, you must download the core
193 gcc distribution plus any language specific distributions you wish to
194 use. The core distribution includes the C language front end as well as the
195 shared components. Each language has a tarball which includes the language
196 front end as well as the language runtime (when appropriate).
198 Unpack the core distribution as well as any language specific
199 distributions in the same directory.
201 If you also intend to build binutils (either to upgrade an existing
202 installation or for use in place of the corresponding tools of your
203 OS), unpack the binutils distribution either in the same directory or
204 a separate one. In the latter case, add symbolic links to any
205 components of the binutils you intend to build alongside the compiler
206 (@file{bfd}, @file{binutils}, @file{gas}, @file{gprof}, @file{ld},
207 @file{opcodes}, @dots{}) to the directory containing the GCC sources.
214 @uref{./index.html,,Return to the GCC Installation page}
218 @c ***Configuration***********************************************************
220 @comment node-name, next, previous, up
221 @node Configuration, Building, Downloading the source, Installing GCC
225 <h1 align="center">Installing GCC: Configuration</h1>
228 @chapter Installing GCC: Configuration
230 @cindex Configuration
231 @cindex Installing GCC: Configuration
233 Like most GNU software, GCC must be configured before it can be built.
234 This document describes the recommended configuration procedure
235 for both native and cross targets.
237 We use @var{srcdir} to refer to the toplevel source directory for
238 GCC; we use @var{objdir} to refer to the toplevel build/object directory.
240 If you obtained the sources via CVS, @var{srcdir} must refer to the top
241 @file{gcc} directory, the one where the @file{MAINTAINERS} can be found,
242 and not its @file{gcc} subdirectory, otherwise the build will fail.
244 First, we @strong{highly} recommend that GCC be built into a
245 separate directory than the sources which does @strong{not} reside
246 within the source tree. This is how we generally build GCC; building
247 where @var{srcdir} == @var{objdir} should still work, but doesn't
248 get extensive testing; building where @var{objdir} is a subdirectory
249 of @var{srcdir} is unsupported.
251 If you have built GCC previously in the same directory for a
252 different target machine, do @samp{make distclean} to delete all files
253 that might be invalid. One of the files this deletes is
254 @file{Makefile}; if @samp{make distclean} complains that @file{Makefile}
255 does not exist, it probably means that the directory is already suitably
256 clean. However, with the recommended method of building in a separate
257 @var{objdir}, you should simply use a different @var{objdir} for each
260 Second, when configuring a native system, either @command{cc} or
261 @command{gcc} must be in your path or you must set @env{CC} in
262 your environment before running configure. Otherwise the configuration
270 % @var{srcdir}/configure [@var{target}] [@var{options}]
274 @heading Target specification
277 GCC has code to correctly determine the correct value for @var{target}
278 for nearly all native systems. Therefore, we highly recommend you not
279 provide a configure target when configuring a native compiler.
282 @var{target} must be specified as @option{--target=@var{target}}
283 when configuring a cross compiler; examples of valid targets would be
284 i960-rtems, m68k-coff, sh-elf, etc.
287 Specifying just @var{target} instead of @option{--target=@var{target}}
288 implies that the host defaults to @var{target}.
292 @heading Options specification
294 Use @var{options} to override several configure time options for
295 GCC@. A partial list of supported @var{options}:
298 @item --prefix=@var{dirname}
299 Specify the toplevel installation
300 directory. This is the recommended way to install the tools into a directory
301 other than the default. The toplevel installation directory defaults to
304 We @strong{highly} recommend against @var{dirname} being the same or a
305 subdirectory of @var{objdir} or vice versa.
307 These additional options control where certain parts of the distribution
308 are installed. Normally you should not need to use these options.
311 @item --exec-prefix=@var{dirname}
312 Specify the toplevel installation directory for architecture-dependent
313 files. The default is @file{@var{prefix}}.
315 @item --bindir=@var{dirname}
316 Specify the installation directory for the executables called by users
317 (such as @command{gcc} and @command{g++}). The default is
318 @file{@var{exec-prefix}/bin}.
320 @item --libdir=@var{dirname}
321 Specify the installation directory for object code libraries and
322 internal parts of GCC@. The default is @file{@var{exec-prefix}/lib}.
324 @item --with-slibdir=@var{dirname}
325 Specify the installation directory for the shared libgcc library. The
326 default is @file{@var{libdir}}.
328 @item --infodir=@var{dirname}
329 Specify the installation directory for documentation in info format.
330 The default is @file{@var{prefix}/info}.
332 @item --mandir=@var{dirname}
333 Specify the installation directory for manual pages. The default is
334 @file{@var{prefix}/man}. (Note that the manual pages are only extracts from
335 the full GCC manuals, which are provided in Texinfo format. The
336 @command{g77} manpage is unmaintained and may be out of date; the others
337 are derived by an automatic conversion process from parts of the full
340 @item --with-gxx-include-dir=@var{dirname}
342 the installation directory for g++ header files. The default is
343 @file{@var{prefix}/include/g++-v3}.
347 @item --with-local-prefix=@var{dirname}
349 installation directory for local include files. The default is
350 @file{/usr/local}. Specify this option if you want the compiler to
351 search directory @file{@var{dirname}/include} for locally installed
352 header files @emph{instead} of @file{/usr/local/include}.
354 You should specify @option{--with-local-prefix} @strong{only} if your
355 site has a different convention (not @file{/usr/local}) for where to put
358 The default value for @option{--with-local-prefix} is @file{/usr/local}
359 regardless of the value of @option{--prefix}. Specifying
360 @option{--prefix} has no effect on which directory GCC searches for
361 local header files. This may seem counterintuitive, but actually it is
364 The purpose of @option{--prefix} is to specify where to @emph{install
365 GCC}. The local header files in @file{/usr/local/include}---if you put
366 any in that directory---are not part of GCC@. They are part of other
367 programs---perhaps many others. (GCC installs its own header files in
368 another directory which is based on the @option{--prefix} value.)
370 @strong{Do not} specify @file{/usr} as the @option{--with-local-prefix}!
371 The directory you use for @option{--with-local-prefix} @strong{must not}
372 contain any of the system's standard header files. If it did contain
373 them, certain programs would be miscompiled (including GNU Emacs, on
374 certain targets), because this would override and nullify the header
375 file corrections made by the @code{fixincludes} script.
377 Indications are that people who use this option use it based on mistaken
378 ideas of what it is for. People use it as if it specified where to
379 install part of GCC@. Perhaps they make this assumption because
380 installing GCC creates the directory.
382 @item --enable-shared[=@var{package}[,@dots{}]]
383 Build shared versions of libraries, if shared libraries are supported on
384 the target platform. Unlike GCC 2.95.x and earlier, shared libraries
385 are enabled by default on all platforms that support shared libraries.
387 If a list of packages is given as an argument, build shared libraries
388 only for the listed packages. For other packages, only static libraries
389 will be built. Package names currently recognized in the GCC tree are
390 @samp{libgcc} (also known as @samp{gcc}), @samp{libstdc++} (not
391 @samp{libstdc++-v3}), @samp{libffi}, @samp{zlib}, @samp{boehm-gc} and
392 @samp{libjava}. Note that @samp{libobjc} does not recognize itself by
393 any name, so, if you list package names in @option{--enable-shared},
394 you'll only get static Objective-C libraries. @samp{libf2c} and
395 @samp{libiberty} do not support shared libraries at all.
397 Use @option{--disable-shared} to build only static libraries. Note that
398 @option{--disable-shared} does not accept a list of package names as
399 argument, only @option{--enable-shared} does.
401 @item @anchor{with-gnu-as}--with-gnu-as
402 Specify that the compiler should assume that the
403 assembler it finds is the GNU assembler. However, this does not modify
404 the rules to find an assembler and will result in confusion if found
405 assembler is not actually the GNU assembler. (Confusion will also
406 result if the compiler finds the GNU assembler but has not been
407 configured with @option{--with-gnu-as}.) If you have more than one
408 assembler installed on your system, you may want to use this option in
409 connection with @option{--with-as=@file{/path/to/gas}}.
411 The systems where it makes a difference whether you use the GNU assembler are
412 @samp{hppa1.0-@var{any}-@var{any}}, @samp{hppa1.1-@var{any}-@var{any}},
413 @samp{i386-@var{any}-sysv}, @samp{i386-@var{any}-isc},
414 @samp{i860-@var{any}-bsd}, @samp{m68k-bull-sysv},
415 @samp{m68k-hp-hpux}, @samp{m68k-sony-bsd},
416 @samp{m68k-altos-sysv}, @samp{m68000-hp-hpux},
417 @samp{m68000-att-sysv}, @samp{@var{any}-lynx-lynxos},
418 and @samp{mips-@var{any}}.
419 On any other system, @option{--with-gnu-as} has no effect.
421 On the systems listed above (except for the HP-PA, for ISC on the
422 386, and for @samp{mips-sgi-irix5.*}), if you use the GNU assembler,
423 you should also use the GNU linker (and specify @option{--with-gnu-ld}).
425 @item --with-as=@file{/path/to/as}
427 compiler should use the assembler pointed to by @var{pathname}, rather
428 than the one found by the standard rules to find an assembler, which
433 @file{@var{exec_prefix}/lib/gcc-lib/@var{target}/@var{version}}
434 directory, where @var{exec_prefix} defaults to @var{prefix} which
435 defaults to @file{/usr/local} unless overridden by the
436 @option{--prefix=/pathname} switch described above. @var{target} is the
437 target system triple, such as @var{sparc-sun-solaris2.7}, and
438 @var{version} denotes the GCC version, such as 2.95.2.
440 Check operating system specific directories (e.g.@: @file{/usr/ccs/bin} on
443 Note that these rules do not check for the value of @env{PATH}. You may
444 want to use @option{--with-as} if no assembler is installed in the
445 directories listed above, or if you have multiple assemblers installed
446 and want to choose one that is not found by the above rules.
448 @item @anchor{with-gnu-ld}--with-gnu-ld
449 Same as @uref{#with-gnu-as,,@option{--with-gnu-as}}
453 @item --with-ld=@file{/path/to/ld}
455 @option{--with-as}, but for the linker.
458 Specify that stabs debugging
459 information should be used instead of whatever format the host normally
460 uses. Normally GCC uses the same debug format as the host system.
462 On MIPS based systems and on Alphas, you must specify whether you want
463 GCC to create the normal ECOFF debugging format, or to use BSD-style
464 stabs passed through the ECOFF symbol table. The normal ECOFF debug
465 format cannot fully handle languages other than C@. BSD stabs format can
466 handle other languages, but it only works with the GNU debugger GDB@.
468 Normally, GCC uses the ECOFF debugging format by default; if you
469 prefer BSD stabs, specify @option{--with-stabs} when you configure GCC@.
471 No matter which default you choose when you configure GCC, the user
472 can use the @option{-gcoff} and @option{-gstabs+} options to specify explicitly
473 the debug format for a particular compilation.
475 @option{--with-stabs} is meaningful on the ISC system on the 386, also, if
476 @option{--with-gas} is used. It selects use of stabs debugging
477 information embedded in COFF output. This kind of debugging information
478 supports C++ well; ordinary COFF debugging information does not.
480 @option{--with-stabs} is also meaningful on 386 systems running SVR4. It
481 selects use of stabs debugging information embedded in ELF output. The
482 C++ compiler currently (2.6.0) does not support the DWARF debugging
483 information normally used on 386 SVR4 platforms; stabs provide a
484 workable alternative. This requires gas and gdb, as the normal SVR4
485 tools can not generate or interpret stabs.
487 @item --enable-multilib
488 Specify that multiple target
489 libraries should be built to support different target variants, calling
490 conventions, etc. This is the default.
492 @item --enable-threads
493 Specify that the target
494 supports threads. This affects the Objective-C compiler and runtime
495 library, and exception handling for other languages like C++ and Java.
496 On some systems, this is the default.
498 In general, the best (and, in many cases, the only known) threading
499 model available will be configured for use. Beware that on some
500 systems, gcc has not been taught what threading models are generally
501 available for the system. In this case, @option{--enable-threads} is an
502 alias for @option{--enable-threads=single}.
504 @item --disable-threads
505 Specify that threading support should be disabled for the system.
506 This is an alias for @option{--enable-threads=single}.
508 @item --enable-threads=@var{lib}
510 @var{lib} is the thread support library. This affects the Objective-C
511 compiler and runtime library, and exception handling for other languages
512 like C++ and Java. The possibilities for @var{lib} are:
520 Generic MACH thread support, known to work on NEXTSTEP@. (Please note
521 that the file needed to support this configuration, @file{gthr-mach.h}, is
522 missing and thus this setting will cause a known bootstrap failure.)
524 This is an alias for @samp{single}.
526 Generic POSIX thread support.
528 Same as @samp{posix} on arm*-*-linux*, *-*-chorusos* and *-*-freebsd*
529 only. A future release of gcc might remove this alias or extend it
532 RTEMS thread support.
534 Disable thread support, should work for all platforms.
536 SUN Solaris thread support.
538 VxWorks thread support.
540 Microsoft Win32 API thread support.
543 @item --with-cpu=@var{cpu}
544 Specify which cpu variant the
545 compiler should generate code for by default. This is currently
546 only supported on the some ports, specifically arm, powerpc, and
547 SPARC@. If configure does not recognize the model name (e.g.@: arm700,
548 603e, or ultrasparc) you provide, please check the configure script
549 for a complete list of supported models.
551 @item --enable-target-optspace
553 libraries should be optimized for code space instead of code speed.
554 This is the default for the m32r platform.
557 Specify that a user visible @command{cpp} program should not be installed.
559 @item --with-cpp-install-dir=@var{dirname}
560 Specify that the user visible @command{cpp} program should be installed
561 in @file{@var{prefix}/@var{dirname}/cpp}, in addition to @var{bindir}.
563 @item --enable-maintainer-mode
565 regenerate the GCC master message catalog @file{gcc.pot} are normally
566 disabled. This is because it can only be rebuilt if the complete source
567 tree is present. If you have changed the sources and want to rebuild the
568 catalog, configuring with @option{--enable-maintainer-mode} will enable
569 this. Note that you need a recent version of the @code{gettext} tools
572 @item --enable-version-specific-runtime-libs
574 that runtime libraries should be installed in the compiler specific
575 subdirectory (@file{@var{libsubdir}}) rather than the usual places. In
576 addition, libstdc++'s include files will be installed in
577 @file{@var{libsubdir}/include/g++} unless you overruled it by using
578 @option{--with-gxx-include-dir=@var{dirname}}. Using this option is
579 particularly useful if you intend to use several versions of GCC in
580 parallel. This is currently supported by @samp{libf2c} and
583 @item --enable-languages=@var{lang1},@var{lang2},@dots{}
584 Specify that only a particular subset of compilers and
585 their runtime libraries should be built. For a list of valid values for
586 @var{langN} you can issue the following command in the
587 @file{gcc} directory of your GCC source tree:@* @samp{grep language=
588 */config-lang.in}@* Currently, you can use any of the following:
589 @code{c}, @code{c++}, @code{f77}, @code{java} and @code{objc}.
590 @code{CHILL} is not currently maintained, and will almost
591 certainly fail to compile. Note that this switch does not work with
592 EGCS 1.1.2 or older versions of egcs. It is supported in GCC 2.95
593 and newer versions.@*
594 If you do not pass this flag, all languages available in the @file{gcc}
595 sub-tree will be configured. Re-defining @code{LANGUAGES} when calling
596 @samp{make bootstrap} @strong{does not} work anymore, as those
597 language sub-directories might not have been configured!
599 @item --disable-libgcj
600 Specify that the run-time libraries
601 used by GCJ should not be built. This is useful in case you intend
602 to use GCJ with some other run-time, or you're going to install it
603 separately, or it just happens not to build on your particular
604 machine. In general, if the Java front end is enabled, the GCJ
605 libraries will be enabled too, unless they're known to not work on
606 the target platform. If GCJ is enabled but libgcj isn't built, you
607 may need to port it; in this case, before modifying the top-level
608 @file{configure.in} so that libgcj is enabled by default on this platform,
609 you may use @option{--enable-libgcj} to override the default.
612 Specify that the compiler should
613 use DWARF2 debugging information as the default.
615 @item --enable-win32-registry
616 @itemx --enable-win32-registry=@var{KEY}
617 @itemx --disable-win32-registry
618 The @option{--enable-win32-registry} option enables Windows-hosted GCC
619 to look up installations paths in the registry using the following key:
622 @code{HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\@var{KEY}}
625 @var{KEY} defaults to GCC version number, and can be overridden by the
626 @option{--enable-win32-registry=@var{KEY}} option. Vendors and distributors
627 who use custom installers are encouraged to provide a different key,
628 perhaps one comprised of vendor name and GCC version number, to
629 avoid conflict with existing installations. This feature is enabled
630 by default, and can be disabled by @option{--disable-win32-registry}
631 option. This option has no effect on the other hosts.
634 Specify that the machine does not have a floating point unit. This
635 option only applies to @samp{m68k-sun-sunos@var{n}} and
636 @samp{m68k-isi-bsd}. On any other system, @option{--nfp} has no effect.
638 @item --enable-checking
639 @itemx --enable-checking=@var{list}
640 When you specify this option, the compiler is built to perform checking
641 of tree node types when referencing fields of that node, and some other
642 internal consistency checks. This does not change the generated code,
643 but adds error checking within the compiler. This will slow down the
644 compiler and may only work properly if you are building the compiler
645 with GCC@. This is on by default when building from CVS or snapshots,
646 but off for releases. More control over the checks may be had by
647 specifying @var{list}; the categories of checks available are
648 @samp{misc}, @samp{tree}, @samp{gc}, @samp{rtl} and @samp{gcac}. The
649 default when @var{list} is not specified is @samp{misc,tree,gc}; the
650 checks @samp{rtl} and @samp{gcac} are very expensive.
654 The @option{--enable-nls} option enables Native Language Support (NLS),
655 which lets GCC output diagnostics in languages other than American
656 English. Native Language Support is enabled by default if not doing a
657 canadian cross build. The @option{--disable-nls} option disables NLS@.
659 @item --with-included-gettext
660 If NLS is enabled, the @option{--with-included-gettext} option causes the build
661 procedure to prefer its copy of GNU @code{gettext}.
664 If NLS is enabled, and if the host lacks @code{gettext} but has the
665 inferior @code{catgets} interface, the GCC build procedure normally
666 ignores @code{catgets} and instead uses GCC's copy of the GNU
667 @code{gettext} library. The @option{--with-catgets} option causes the
668 build procedure to use the host's @code{catgets} in this situation.
671 Some options which only apply to building cross compilers:
673 @item --with-headers=@var{dir}
674 Specifies a directory
675 which has target include files.
676 @emph{This options is required} when building a cross
677 compiler, if @file{@var{prefix}/@var{target}/sys-include} doesn't pre-exist.
678 These include files will be copied into the @file{gcc} install directory.
679 Fixincludes will be run on these files to make them compatible with
681 @item --with-libs=``@var{dir1} @var{dir2} @dots{} @var{dirN}''
682 Specifies a list of directories which contain the target runtime
683 libraries. These libraries will be copied into the @file{gcc} install
686 Specifies that ``newlib'' is
687 being used as the target C library. This causes @code{__eprintf} to be
688 omitted from libgcc.a on the assumption that it will be provided by
692 Note that each @option{--enable} option has a corresponding
693 @option{--disable} option and that each @option{--with} option has a
694 corresponding @option{--without} option.
701 @uref{./index.html,,Return to the GCC Installation page}
705 @c ***Building****************************************************************
707 @comment node-name, next, previous, up
708 @node Building, Testing, Configuration, Installing GCC
712 <h1 align="center">Installing GCC: Building</h1>
717 @cindex Installing GCC: Building
719 Now that GCC is configured, you are ready to build the compiler and
722 We @strong{highly} recommend that GCC be built using GNU make;
723 other versions may work, then again they might not.
725 (For example, many broken versions of make will fail if you use the
726 recommended setup where @var{objdir} is different from @var{srcdir}.
727 Other broken versions may recompile parts of the compiler when
728 installing the compiler.)
730 Some commands executed when making the compiler may fail (return a
731 non-zero status) and be ignored by @code{make}. These failures, which
732 are often due to files that were not found, are expected, and can safely
735 It is normal to have compiler warnings when compiling certain files.
736 Unless you are a GCC developer, you can generally ignore these warnings
737 unless they cause compilation to fail.
739 On certain old systems, defining certain environment variables such as
740 @env{CC} can interfere with the functioning of @command{make}.
742 If you encounter seemingly strange errors when trying to build the
743 compiler in a directory other than the source directory, it could be
744 because you have previously configured the compiler in the source
745 directory. Make sure you have done all the necessary preparations.
747 If you build GCC on a BSD system using a directory stored in an old System
748 V file system, problems may occur in running @code{fixincludes} if the
749 System V file system doesn't support symbolic links. These problems
750 result in a failure to fix the declaration of @code{size_t} in
751 @file{sys/types.h}. If you find that @code{size_t} is a signed type and
752 that type mismatches occur, this could be the cause.
754 The solution is not to use such a directory for building GCC@.
756 When building from CVS or snapshots, or if you modify parser sources,
757 you need the Bison parser generator installed. Any version 1.25 or
758 later should work; older versions may also work. If you do not modify
759 parser sources, releases contain the Bison-generated files and you do
760 not need Bison installed to build them.
762 When building from CVS or snapshots, or if you modify Texinfo
763 documentation, you need version 4.0 or later of Texinfo installed if you
764 want Info documentation to be regenerated. Releases contain Info
765 documentation pre-built for the unmodified documentation in the release.
767 @section Building a native compiler
769 For a native build issue the command @samp{make bootstrap}. This
770 will build the entire GCC system, which includes the following steps:
774 Build host tools necessary to build the compiler such as texinfo, bison,
778 Build target tools for use by the compiler such as binutils (bfd,
779 binutils, gas, gprof, ld, and opcodes)@*
780 if they have been individually linked
781 or moved into the top level GCC source tree before configuring.
784 Perform a 3-stage bootstrap of the compiler.
787 Perform a comparison test of the stage2 and stage3 compilers.
790 Build runtime libraries using the stage3 compiler from the previous step.
794 If you are short on disk space you might consider @samp{make
795 bootstrap-lean} instead. This is identical to @samp{make
796 bootstrap} except that object files from the stage1 and
797 stage2 of the 3-stage bootstrap of the compiler are deleted as
798 soon as they are no longer needed.
801 If you want to save additional space during the bootstrap and in
802 the final installation as well, you can build the compiler binaries
803 without debugging information with @samp{make CFLAGS='-O' LIBCFLAGS='-g
804 -O2' LIBCXXFLAGS='-g -O2 -fno-implicit-templates' bootstrap}. This will save
805 roughly 40% of disk space both for the bootstrap and the final installation.
806 (Libraries will still contain debugging information.)
808 If you wish to use non-default flags when compiling the stage2 and
809 stage3 compile, set @code{BOOT_CFLAGS} on the command line when doing
810 @samp{make bootstrap}. Non-default optimization flags are less well
811 tested here than the default of @samp{-g -O2}, but should still work.
812 In a few cases, you may find that you need to specify special flags such
813 as @option{-msoft-float} here to complete the bootstrap; or, if the
814 native compiler miscompiles the stage1 compiler, you may need to work
815 around this, by choosing @code{BOOT_CFLAGS} to avoid the parts of the
816 stage1 compiler that were miscompiled, or by using @samp{make
817 bootstrap4} to increase the number of stages of bootstrap.
819 If you used the flag @option{--enable-languages=@dots{}} to restrict
820 the compilers to be built, only those you've actually enabled will be
821 built. This will of course only build those runtime libraries, for
822 which the particular compiler has been built. Please note,
823 that re-defining LANGUAGES when calling @samp{make bootstrap}
824 @strong{does not} work anymore!
826 If the comparison of stage2 and stage3 fails, this normally indicates
827 that the stage 2 compiler has compiled GCC incorrectly, and is therefore
828 a potentially serious bug which you should investigate and report. (On
829 a few systems, meaningful comparison of object files is impossible; they
830 always appear ``different''. If you encounter this problem, you will
831 need to disable comparison in the @file{Makefile}.)
833 @section Building a cross compiler
835 We recommend reading the
836 @uref{http://www.objsw.com/CrossGCC/,,crossgcc FAQ}
837 for information about building cross compilers.
839 When building a cross compiler, it is not generally possible to do a
840 3-stage bootstrap of the compiler. This makes for an interesting problem
841 as parts of GCC can only be built with GCC@.
843 To build a cross compiler, we first recommend building and installing a
844 native compiler. You can then use the native GCC compiler to build the
847 Assuming you have already installed a native copy of GCC and configured
848 your cross compiler, issue the command @command{make}, which performs the
853 Build host tools necessary to build the compiler such as texinfo, bison,
857 Build target tools for use by the compiler such as binutils (bfd,
858 binutils, gas, gprof, ld, and opcodes)
859 if they have been individually linked or moved into the top level GCC source
860 tree before configuring.
863 Build the compiler (single stage only).
866 Build runtime libraries using the compiler from the previous step.
869 Note that if an error occurs in any step the make process will exit.
871 @section Building in parallel
873 If you have a multiprocessor system you can use @samp{make bootstrap
874 MAKE="make -j 2" -j 2} or just @samp{make -j 2 bootstrap}
875 for GNU Make 3.79 and above instead of just @samp{make bootstrap}
876 when building GCC@. You can use a bigger number instead of two if
877 you like. In most cases, it won't help to use a number bigger than
878 the number of processors in your machine.
885 @uref{./index.html,,Return to the GCC Installation page}
889 @c ***Testing*****************************************************************
891 @comment node-name, next, previous, up
892 @node Testing, Final install, Building, Installing GCC
896 <h1 align="center">Installing GCC: Testing</h1>
899 @chapter Installing GCC: Testing
902 @cindex Installing GCC: Testing
905 @strong{Please note that this is only applicable
906 to current development versions of GCC and GCC 3.0 or later.
907 GCC 2.95.x does not come with a testsuite.}
909 Before you install GCC, you might wish to run the testsuite. This
910 step is optional and may require you to download additional software.
912 First, you must have @uref{download.html,,downloaded the testsuites}.
913 The full distribution contains testsuites; only if you downloaded the
914 ``core'' compiler plus any front ends, you do not have the testsuites.
916 Second, you must have a @uref{http://www.gnu.org/software/dejagnu/,,current version of DejaGnu} installed;
917 dejagnu 1.3 is not sufficient.
919 Now you may need specific preparations:
923 In order to run the libio tests in GCC 2.95 and earlier versions of GCC
924 on targets which do not fully
925 support Unix/POSIX commands (e.g.@: Cygwin), the references to the @file{dbz}
926 directory have to be deleted from @file{libio/configure.in}.
929 The following environment variables may need to be set appropriately, as in
930 the following example (which assumes that DejaGnu has been installed
931 under @file{/usr/local}):
934 TCL_LIBRARY = /usr/local/share/tcl8.0
935 DEJAGNULIBS = /usr/local/share/dejagnu
938 On systems such as Cygwin, these paths are required to be actual
939 paths, not mounts or links; presumably this is due to some lack of
940 portability in the DejaGnu code.
942 If the directories where @command{runtest} and @command{expect} were
943 installed are in the @env{PATH}, it should not be necessary to set these
944 environment variables.
948 Finally, you can run the testsuite (which may take a long time):
950 cd @var{objdir}; make -k check
953 The testing process will try to test as many components in the GCC
954 distribution as possible, including the C, C++, Objective-C and Fortran
955 compilers as well as the C++ and Java runtime libraries.
957 @section How can I run the test suite on selected tests?
959 As a first possibility to cut down the number of tests that are run it is
960 possible to use @samp{make check-gcc} or @samp{make check-g++}
961 in the gcc subdirectory of the object directory. To further cut down the
962 tests the following is possible:
965 make check-gcc RUNTESTFLAGS="execute.exp @var{other-options}"
968 This will run all gcc execute tests in the testsuite.
971 make check-g++ RUNTESTFLAGS="old-deja.exp=9805* @var{other-options}"
974 This will run the g++ ``old-deja'' tests in the testsuite where the filename
975 matches @samp{9805*}.
977 The @file{*.exp} files are located in the testsuite directories of the GCC
978 source, the most important ones being @file{compile.exp},
979 @file{execute.exp}, @file{dg.exp} and @file{old-deja.exp}.
980 To get a list of the possible @file{*.exp} files, pipe the
981 output of @samp{make check} into a file and look at the
982 @samp{Running @dots{} .exp} lines.
984 @section How to interpret test results
986 After the testsuite has run you'll find various @file{*.sum} and @file{*.log}
987 files in the testsuite subdirectories. The @file{*.log} files contain a
988 detailed log of the compiler invocations and the corresponding
989 results, the @file{*.sum} files summarize the results. These summaries list
990 all the tests that have been run with a corresponding status code:
994 PASS: the test passed as expected
996 XPASS: the test unexpectedly passed
998 FAIL: the test unexpectedly failed
1000 XFAIL: the test failed as expected
1002 UNSUPPORTED: the test is not supported on this platform
1004 ERROR: the testsuite detected an error
1006 WARNING: the testsuite detected a possible problem
1009 It is normal for some tests to report unexpected failures. At the
1010 current time our testing harness does not allow fine grained control
1011 over whether or not a test is expected to fail. We expect to fix this
1012 problem in future releases.
1015 @section Submitting test results
1017 If you want to report the results to the GCC project, use the
1018 @file{contrib/test_summary} shell script. Start it in the @var{objdir} with
1021 @var{srcdir}/contrib/test_summary -p your_commentary.txt \
1022 -m gcc-testresults@@gcc.gnu.org |sh
1025 This script uses the @command{Mail} program to send the results, so
1026 make sure it is in your @env{PATH}. The file @file{your_commentary.txt} is
1027 prepended to the testsuite summary and should contain any special
1028 remarks you have on your results or your build environment. Please
1029 do not edit the testsuite result block or the subject line, as these
1030 messages are automatically parsed and presented at the
1031 @uref{http://gcc.gnu.org/testresults/,,GCC testresults} web
1032 page. Here you can also gather information on how specific tests
1033 behave on different platforms and compare them with your results. A
1034 few failing testcases are possible even on released versions and you
1035 should look here first if you think your results are unreasonable.
1039 @c ***Final install***********************************************************
1041 @comment node-name, next, previous, up
1042 @node Final install, , Testing, Installing GCC
1044 @ifset finalinstallhtml
1046 <h1 align="center">Installing GCC: Final installation</h1>
1049 @chapter Installing GCC: Final installation
1052 Now that GCC has been built and tested, you can install it with
1053 @samp{cd @var{objdir}; make install}.
1055 That step completes the installation of GCC; user level binaries can
1056 be found in @file{@var{prefix}/bin} where @var{prefix} is the value you
1057 specified with the @option{--prefix} to configure (or @file{/usr/local}
1058 by default). (If you specified @option{--bindir}, that directory will
1059 be used instead; otherwise, if you specified @option{--exec-prefix},
1060 @file{@var{exec-prefix}/bin} will be used.) Headers for the C++ and
1061 Java libraries are installed in @file{@var{prefix}/include}; libraries
1062 in @file{@var{libdir}} (normally @file{@var{prefix}/lib}); internal
1063 parts of the compiler in @file{@var{libdir}/gcc-lib}; documentation in
1064 info format in @file{@var{infodir}} (normally @file{@var{prefix}/info}).
1066 If you don't mind, please quickly review the
1067 @uref{http://gcc.gnu.org/gcc-3.0/buildstat.html,,build status page}.
1068 If your system is not listed, send a note to
1069 @uref{mailto:gcc@@gcc.gnu.org,,gcc@@gcc.gnu.org} indicating
1070 that you successfully built and installed GCC@.
1072 Include the output from running @file{@var{srcdir}/config.guess}. (Do
1073 not send us the config.guess file itself, just the one-line output from
1076 If you find a bug, please report it following our
1077 @uref{../bugs.html,,bug reporting guidelines}.
1079 If you want to print the GCC manuals, do @samp{cd @var{objdir}; make
1080 dvi}. You will need to have @command{texi2dvi} (version at least 4.0)
1081 and @TeX{} installed. This creates a number of @file{.dvi} files in
1082 subdirectories of @file{@var{objdir}}; these may be converted for
1083 printing with programs such as @command{dvips}. You can also
1084 @uref{http://www.gnu.org/order/order.html,,buy printed manuals from the
1085 Free Software Foundation}, though such manuals may not be for the most
1086 recent version of GCC@.
1093 @uref{./index.html,,Return to the GCC Installation page}
1097 @c ***Binaries****************************************************************
1099 @comment node-name, next, previous, up
1100 @node Binaries, Specific, Installing GCC, Top
1104 <h1 align="center">Installing GCC: Binaries</h1>
1107 @chapter Installing GCC: Binaries
1110 @cindex Installing GCC: Binaries
1112 We are often asked about pre-compiled versions of GCC@. While we cannot
1113 provide these for all platforms, below you'll find links to binaries for
1114 various platforms where creating them by yourself is not easy due to various
1117 Please note that we did not create these binaries, nor do we
1118 support them. If you have any problems installing them, please
1119 contact their makers.
1126 @uref{http://www-frec.bull.com/docs/download.htm,,Bull's Freeware and
1127 Shareware Archive for AIX};
1130 @uref{http://aixpdlib.seas.ucla.edu,,UCLA Software Library for AIX};
1134 DOS---@uref{http://www.delorie.com/djgpp/,,DJGPP};
1137 @uref{http://hpux.cae.wisc.edu/,,HP-UX Porting Center};
1140 @uref{http://www.sco.com/skunkware/devtools/index.html#gcc,,SCO
1141 OpenServer/Unixware};
1144 Solaris (SPARC, Intel)---@uref{http://www.sunfreeware.com/,,Sunfreeware};
1147 SGI---@uref{http://freeware.sgi.com/,,SGI Freeware};
1150 Windows 95, 98, and NT:
1153 The @uref{http://sources.redhat.com/cygwin/,,Cygwin} project;
1155 @uref{http://www.xraylith.wisc.edu/~khan/software/gnu-win32/,,GNU Win32}
1156 related projects by Mumit Khan.
1160 @uref{ftp://ftp.thewrittenword.com/packages/free/by-name/gcc-2.95.2/,,The
1161 Written Word} offers binaries for Solaris 2.5.1, 2.6, 2.7/SPARC, 2.7/Intel,
1162 IRIX 6.2, 6.5, Digital UNIX 4.0D, HP-UX 10.20, and HP-UX 11.00.
1165 Hitachi H8/300[HS]---@uref{http://h8300-hms.sourceforge.net/,,GNU
1166 Development Tools for the Hitachi H8/300[HS] Series}
1170 In addition to those specific offerings, you can get a binary
1171 distribution CD-ROM from the
1172 @uref{http://www.fsf.org/order/order.html,,Free Software Foundation}.
1173 It contains binaries for a number of platforms, and
1174 includes not only GCC, but other stuff as well. The current CD does
1175 not contain the latest version of GCC, but it should allow
1176 bootstrapping the compiler. An updated version of that disk is in the
1184 @uref{./index.html,,Return to the GCC Installation page}
1188 @c ***Specific****************************************************************
1190 @comment node-name, next, previous, up
1191 @node Specific, Concept Index, Binaries, Top
1195 <h1 align="center">Host/target specific installation notes for GCC</h1>
1198 @chapter Host/target specific installation notes for GCC
1201 @cindex Specific installation notes
1202 @cindex Target specific installation
1203 @cindex Host specific installation
1204 @cindex Target specific installation notes
1206 Please read this document carefully @emph{before} installing the
1207 GNU Compiler Collection on your machine.
1212 @uref{#1750a-*-*,,1750a-*-*}
1216 @uref{#a29k-*-bsd,,a29k-*-bsd}
1218 @uref{#alpha*-*-*,,alpha*-*-*}
1220 @uref{#alpha*-dec-osf*,,alpha*-dec-osf*}
1222 @uref{#arc-*-elf,,arc-*-elf}
1224 @uref{#arm-*-aout,,arm-*-aout}
1226 @uref{#arm-*-elf,,arm-*-elf}
1228 @uref{#arm*-*-linux-gnu,,arm*-*-linux-gnu}
1230 @uref{#arm-*-riscix,,arm-*-riscix}
1236 @uref{#decstation-*,,decstation-*}
1240 @uref{#dsp16xx,,dsp16xx}
1242 @uref{#elxsi-elxsi-bsd,,elxsi-elxsi-bsd}
1244 @uref{#h8300-hms,,h8300-hms}
1246 @uref{#hppa*-hp-hpux*,,hppa*-hp-hpux*}
1248 @uref{#hppa*-hp-hpux9,,hppa*-hp-hpux9}
1250 @uref{#hppa*-hp-hpux10,,hppa*-hp-hpux10}
1252 @uref{#hppa*-hp-hpux11,,hppa*-hp-hpux11}
1254 @uref{#*-*-freebsd*,,*-*-freebsd*}
1256 @uref{#i370-*-*,,i370-*-*}
1258 @uref{#*-*-linux-gnu,,*-*-linux-gnu}
1260 @uref{#ix86-*-linux*oldld,,i?86-*-linux*oldld}
1262 @uref{#ix86-*-linux*aout,,i?86-*-linux*aout}
1264 @uref{#ix86-*-linux*,,i?86-*-linux*}
1266 @uref{#ix86-*-sco,,i?86-*-sco}
1268 @uref{#ix86-*-sco3.2v4,,i?86-*-sco3.2v4}
1270 @uref{#ix86-*-sco3.2v5*,,i?86-*-sco3.2v5*}
1272 @uref{#ix86-*-udk,,i?86-*-udk}
1274 @uref{#ix86-*-isc,,i?86-*-isc}
1276 @uref{#ix86-*-esix,,i?86-*-esix}
1278 @uref{#ix86-ibm-aix,,i?86-ibm-aix}
1280 @uref{#ix86-sequent-bsd,,i?86-sequent-bsd}
1282 @uref{#ix86-sequent-ptx1*,,i?86-sequent-ptx1*, i?86-sequent-ptx2*}
1284 @uref{#ix86-*-sysv3*,,i?86-*-sysv3*}
1286 @uref{#i860-intel-osf*,,i860-intel-osf*}
1288 @uref{#*-lynx-lynxos,,*-lynx-lynxos}
1290 @uref{#*-ibm-aix*,,*-ibm-aix*}
1292 @uref{#m32r-*-elf,,m32r-*-elf}
1294 @uref{#m68000-hp-bsd,,m68000-hp-bsd}
1296 @uref{#m6811-elf,,m6811-elf}
1298 @uref{#m6812-elf,,m6812-elf}
1300 @uref{#m68k-altos,,m68k-altos}
1302 @uref{#m68k-apple-aux,,m68k-apple-aux}
1304 @uref{#m68k-att-sysv,,m68k-att-sysv}
1306 @uref{#m68k-bull-sysv,,m68k-bull-sysv}
1308 @uref{#m68k-crds-unox,,m68k-crds-unox}
1310 @uref{#m68k-hp-hpux,,m68k-hp-hpux}
1312 @uref{#m68k-*-nextstep*,,m68k-*-nextstep*}
1314 @uref{#m68k-ncr-*,,m68k-ncr-*}
1316 @uref{#m68k-sun,,m68k-sun}
1318 @uref{#m68k-sun-sunos4.1.1,,m68k-sun-sunos4.1.1}
1320 @uref{#m88k-*-svr3,,m88k-*-svr3}
1322 @uref{#m88k-*-dgux,,m88k-*-dgux}
1324 @uref{#m88k-tektronix-sysv3,,m88k-tektronix-sysv3}
1326 @uref{#mips-*-*,,mips-*-*}
1328 @uref{#mips-mips-bsd,,mips-mips-bsd}
1330 @uref{#mips-mips-riscos*,,mips-mips-riscos*}
1332 @uref{#mips*-sgi-irix4,,mips*-sgi-irix4}
1334 @uref{#mips*-sgi-irix5,,mips*-sgi-irix5}
1336 @uref{#mips*-sgi-irix6,,mips*-sgi-irix6}
1338 @uref{#mips-sony-sysv,,mips-sony-sysv}
1340 @uref{#ns32k-encore,,ns32k-encore}
1342 @uref{#ns32k-*-genix,,ns32k-*-genix}
1344 @uref{#ns32k-sequent,,ns32k-sequent}
1346 @uref{#ns32k-utek,,ns32k-utek}
1348 @uref{#powerpc-*-elf,,powerpc-*-elf, powerpc-*-sysv4}
1350 @uref{#powerpc-*-linux-gnu*,,powerpc-*-linux-gnu*}
1352 @uref{#powerpc-*-eabiaix,,powerpc-*-eabiaix}
1354 @uref{#powerpc-*-eabisim,,powerpc-*-eabisim}
1356 @uref{#powerpc-*-eabi,,powerpc-*-eabi}
1358 @uref{#powerpcle-*-elf,,powerpcle-*-elf, powerpcle-*-sysv4}
1360 @uref{#powerpcle-*-eabisim,,powerpcle-*-eabisim}
1362 @uref{#powerpcle-*-eabi,,powerpcle-*-eabi}
1364 @uref{#powerpcle-*-winnt,,powerpcle-*-winnt, powerpcle-*-pe}
1366 @uref{#romp-*-aos,,romp-*-aos, romp-*-mach}
1368 @uref{#*-*-solaris*,,*-*-solaris*}
1370 @uref{#sparc-sun-*,,sparc-sun-*}
1372 @uref{#sparc-sun-solaris*,,sparc-sun-solaris*}
1374 @uref{#sparc-sun-solaris2.7,,sparc-sun-solaris2.7}
1376 @uref{#*-sun-solaris2.8,,*-sun-solaris2.8}
1378 @uref{#sunv5,,Sun V5.0 Compiler Bugs}
1380 @uref{#sparc-sun-sunos*,,sparc-sun-sunos*}
1382 @uref{#sparc-unknown-linux-gnulibc1,,sparc-unknown-linux-gnulibc1}
1384 @uref{#sparc64-*-*,,sparc64-*-*}
1386 @uref{#*-*-sysv*,,*-*-sysv*}
1388 @uref{#vax-dec-ultrix,,vax-dec-ultrix}
1390 @uref{#we32k-*-*,,we32k-*-*}
1392 @uref{#windows,,Microsoft Windows}
1396 @uref{#older,,Older systems}
1401 @uref{#elf_targets,,all ELF targets} (SVR4, Solaris, etc.)
1407 <!-- -------- host/target specific issues start here ---------------- -->
1410 @heading @anchor{1750a-*-*}1750a-*-*
1411 MIL-STD-1750A processors.
1413 The MIL-STD-1750A cross configuration produces output for
1414 @code{as1750}, an assembler/linker available under the GNU Public
1415 License for the 1750A@. @code{as1750} can be obtained at
1416 @uref{ftp://ftp.fta-berlin.de/pub/crossgcc/1750gals/}.
1417 A similarly licensed simulator for
1418 the 1750A is available from same address.
1420 You should ignore a fatal error during the building of libgcc (libgcc is
1421 not yet implemented for the 1750A@.)
1423 The @code{as1750} assembler requires the file @file{ms1750.inc}, which is
1424 found in the directory @file{config/1750a}.
1426 GCC produced the same sections as the Fairchild F9450 C Compiler,
1431 The program code section.
1434 The read/write (RAM) data section.
1437 The read-only (ROM) constants section.
1440 Initialization section (code to copy KREL to SREL)@.
1443 The smallest addressable unit is 16 bits (BITS_PER_UNIT is 16). This
1444 means that type @code{char} is represented with a 16-bit word per character.
1445 The 1750A's ``Load/Store Upper/Lower Byte'' instructions are not used by
1452 @heading @anchor{a29k}a29k
1453 AMD Am29k-family processors. These are normally used in embedded
1454 applications. There are no standard Unix configurations.
1456 corresponds to AMD's standard calling sequence and binary interface
1457 and is compatible with other 29k tools.
1459 You may need to make a variant of the file @file{a29k.h} for your
1460 particular configuration.
1466 @heading @anchor{a29k-*-bsd}a29k-*-bsd
1467 AMD Am29050 used in a system running a variant of BSD Unix.
1473 @heading @anchor{alpha*-*-*}alpha*-*-*
1475 This section contains general configuration information for all
1476 alpha-based platforms using ELF (in particular, ignore this section for
1477 DEC OSF/1, Digital UNIX and Tru64 UNIX)@. In addition to reading this
1478 section, please read all other sections that match your target.
1480 We require binutils 2.11.1 (as of yet unreleased), binutils with
1481 @samp{binutils-2_11-branch} tag after May 31, 2001 (as taken below), or newer.
1482 Previous binutils releases had a number of problems with DWARF2
1483 debugging information, not the least of which is incorrect linking of
1486 Until binutils 2.11.1 is released, these sample commands may be useful:
1489 mkdir binutils-2.11.X; cd binutils-2.11.X
1490 cvs -d :pserver:anoncvs@@anoncvs.cygnus.com:/cvs/src \
1491 co -rbinutils-2_11-branch -P binutils
1493 ../src/configure --prefix=@emph{an-absolute-path}
1494 make all check install
1497 When configuring gcc, provide explicit @option{--with-gnu-as}
1498 @option{--with-as=@emph{an-absolute-path/bin/as}} and
1499 @option{--with-gnu-ld} @option{--with-ld=@emph{an-absolute-path/bin/ld}}
1500 options to point into the prefix used above.
1506 @heading @anchor{alpha*-dec-osf*}alpha*-dec-osf*
1507 Systems using processors that implement the DEC Alpha architecture and
1508 are running the DEC/Compaq Unix (DEC OSF/1, Digital UNIX, or Compaq
1509 Tru64 UNIX) operating system, for example the DEC Alpha AXP systems.
1511 In Tru64 UNIX V5.1, Compaq introduced a new assembler that does not
1512 currently (2001-06-13) work with @command{mips-tfile}. As a workaround,
1513 we need to use the old assembler, invoked via the barely documented
1514 @option{-oldas} option. To bootstrap GCC, you either need to use the
1518 % CC=cc @var{srcdir}/configure [@var{target}] [@var{options}]
1521 or you can use a copy of GCC 2.95.3 or higher built on Tru64 UNIX V4.0:
1524 % CC=gcc -Wa,-oldas @var{srcdir}/configure [@var{target}] [@var{options}]
1527 GCC writes a @samp{.verstamp} directive to the assembler output file
1528 unless it is built as a cross-compiler. It gets the version to use from
1529 the system header file @file{/usr/include/stamp.h}. If you install a
1530 new version of DEC Unix, you should rebuild GCC to pick up the new version
1533 Note that since the Alpha is a 64-bit architecture, cross-compilers from
1534 32-bit machines will not generate code as efficient as that generated
1535 when the compiler is running on a 64-bit machine because many
1536 optimizations that depend on being able to represent a word on the
1537 target in an integral value on the host cannot be performed. Building
1538 cross-compilers on the Alpha for 32-bit machines has only been tested in
1539 a few cases and may not work properly.
1541 @code{make compare} may fail on old versions of DEC Unix unless you add
1542 @option{-save-temps} to @code{CFLAGS}. On these systems, the name of the
1543 assembler input file is stored in the object file, and that makes
1544 comparison fail if it differs between the @code{stage1} and
1545 @code{stage2} compilations. The option @option{-save-temps} forces a
1546 fixed name to be used for the assembler input file, instead of a
1547 randomly chosen name in @file{/tmp}. Do not add @option{-save-temps}
1548 unless the comparisons fail without that option. If you add
1549 @option{-save-temps}, you will have to manually delete the @samp{.i} and
1550 @samp{.s} files after each series of compilations.
1552 GCC now supports both the native (ECOFF) debugging format used by DBX
1553 and GDB and an encapsulated STABS format for use only with GDB@. See the
1554 discussion of the @option{--with-stabs} option of @file{configure} above
1555 for more information on these formats and how to select them.
1557 There is a bug in DEC's assembler that produces incorrect line numbers
1558 for ECOFF format when the @samp{.align} directive is used. To work
1559 around this problem, GCC will not emit such alignment directives
1560 while writing ECOFF format debugging information even if optimization is
1561 being performed. Unfortunately, this has the very undesirable
1562 side-effect that code addresses when @option{-O} is specified are
1563 different depending on whether or not @option{-g} is also specified.
1565 To avoid this behavior, specify @option{-gstabs+} and use GDB instead of
1566 DBX@. DEC is now aware of this problem with the assembler and hopes to
1567 provide a fix shortly.
1573 @heading @anchor{arc-*-elf}arc-*-elf
1574 Argonaut ARC processor.
1575 This configuration is intended for embedded systems.
1581 @heading @anchor{arm-*-aout}arm-*-aout
1582 Advanced RISC Machines ARM-family processors. These are often used in
1583 embedded applications. There are no standard Unix configurations.
1584 This configuration corresponds to the basic instruction sequences and will
1585 produce @file{a.out} format object modules.
1587 You may need to make a variant of the file @file{arm.h} for your particular
1594 @heading @anchor{arm-*-elf}arm-*-elf
1595 This configuration is intended for embedded systems.
1601 @heading @anchor{arm*-*-linux-gnu}arm*-*-linux-gnu
1603 We require GNU binutils 2.10 or newer.
1609 @heading @anchor{arm-*-riscix}arm-*-riscix
1610 The ARM2 or ARM3 processor running RISC iX, Acorn's port of BSD Unix.
1611 If you are running a version of RISC iX prior to 1.2 then you must
1612 specify the version number during configuration. Note that the
1613 assembler shipped with RISC iX does not support stabs debugging
1614 information; a new version of the assembler, with stabs support
1615 included, is now available from Acorn and via ftp
1616 @uref{ftp://ftp.acorn.com/pub/riscix/as+xterm.tar.Z}. To enable stabs
1617 debugging, pass @option{--with-gnu-as} to configure.
1619 You will need to install GNU @command{sed} before you can run configure.
1625 @heading @anchor{avr}avr
1627 ATMEL AVR-family micro controllers. These are used in embedded
1628 applications. There are no standard Unix configurations. @xref{AVR
1629 Options,, AVR Options, gcc, Using and Porting the GNU Compiler
1630 Collection (GCC)}, for the list of supported MCU types.
1632 Use @samp{configure --target=avr --enable-languages="c"} to configure GCC@.
1634 Further installation notes and other useful information about AVR tools
1635 can also be obtained from:
1639 @uref{http://home.overta.ru/users/denisc,,http://home.overta.ru/users/denisc}
1641 @uref{http://www.itnet.pl/amelektr/avr,,http://www.itnet.pl/amelektr/avr}
1644 We @emph{strongly} recommend using binutils 2.11 or newer.
1646 The following error:
1648 Error: register required
1651 indicates that you should upgrade to a newer version of the binutils.
1657 @heading @anchor{c4x}c4x
1659 Texas Instruments TMS320C3x and TMS320C4x Floating Point Digital Signal
1660 Processors. These are used in embedded applications. There are no
1661 standard Unix configurations. @xref{C4x Options,, C4x Options, gcc,
1662 Using and Porting the GNU Compiler Collection (GCC)}, for the list of
1663 supported MCU types.
1665 GCC can be configured as a cross compiler for both the C3x and C4x
1666 architectures on the same system. Use @samp{configure --target=c4x
1667 --enable-languages="c,c++"} to configure.
1670 Further installation notes and other useful information about C4x tools
1671 can also be obtained from:
1675 @uref{http://www.elec.canterbury.ac.nz/c4x,,http://www.elec.canterbury.ac.nz/c4x}
1682 @heading @anchor{decstation-*}decstation-*
1683 MIPS-based DECstations can support three different personalities:
1684 Ultrix, DEC OSF/1, and OSF/rose. (Alpha-based DECstation products have
1685 a configuration name beginning with @samp{alpha-dec}.) To configure GCC
1686 for these platforms use the following configurations:
1689 @item decstation-ultrix
1690 Ultrix configuration.
1692 @item decstation-osf1
1693 Dec's version of OSF/1.
1695 @item decstation-osfrose
1696 Open Software Foundation reference port of OSF/1 which uses the
1697 OSF/rose object file format instead of ECOFF@. Normally, you
1698 would not select this configuration.
1701 The MIPS C compiler needs to be told to increase its table size
1702 for switch statements with the @option{-Wf,-XNg1500} option in
1703 order to compile @file{cp/parse.c}. If you use the @option{-O2}
1704 optimization option, you also need to use @option{-Olimit 3000}.
1705 Both of these options are automatically generated in the
1706 @file{Makefile} that the shell script @file{configure} builds.
1707 If you override the @code{CC} make variable and use the MIPS
1708 compilers, you may need to add @option{-Wf,-XNg1500 -Olimit 3000}.
1714 @heading @anchor{dos}DOS
1716 Please have a look at our @uref{binaries.html,,binaries page}.
1718 You cannot install GCC by itself on MSDOS; it will not compile under
1719 any MSDOS compiler except itself. You need to get the complete
1720 compilation package DJGPP, which includes binaries as well as sources,
1721 and includes all the necessary compilation tools and libraries.
1727 @heading @anchor{dsp16xx}dsp16xx
1728 A port to the AT&T DSP1610 family of processors.
1734 @heading @anchor{elxsi-elxsi-bsd}elxsi-elxsi-bsd
1735 The Elxsi's C compiler has known limitations that prevent it from
1736 compiling GCC@. Please contact @email{mrs@@cygnus.com} for more details.
1742 @heading @anchor{h8300-hms}h8300-hms
1743 Hitachi H8/300 series of processors.
1745 Please have a look at our @uref{binaries.html,,binaries page}.
1747 The calling convention and structure layout has changed in release 2.6.
1748 All code must be recompiled. The calling convention now passes the
1749 first three arguments in function calls in registers. Structures are no
1750 longer a multiple of 2 bytes.
1756 @heading @anchor{hppa*-hp-hpux*}hppa*-hp-hpux*
1758 We @emph{highly} recommend using gas/binutils-2.8 or newer on all hppa
1759 platforms; you may encounter a variety of problems when using the HP
1762 Specifically, @option{-g} does not work on HP-UX (since that system
1763 uses a peculiar debugging format which GCC does not know about), unless you
1764 use GAS and GDB and configure GCC with the
1765 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}} and
1766 @option{--with-as=@dots{}} options.
1768 If you wish to use pa-risc 2.0 architecture support, you must use either
1769 the HP assembler, gas/binutils-2.11 or a recent
1770 @uref{ftp://sources.redhat.com/pub/binutils/snapshots,,snapshot of gas}.
1772 More specific information to hppa*-hp-hpux* targets follows.
1778 @heading @anchor{hppa*-hp-hpux9}hppa*-hp-hpux9
1780 The HP assembler has major problems on this platform. We've tried to work
1781 around the worst of the problems. However, those workarounds may be causing
1782 linker crashes in some circumstances; the workarounds also probably prevent
1783 shared libraries from working. Use the GNU assembler to avoid these problems.
1786 The configuration scripts for GCC will also trigger a bug in the hpux9
1787 shell. To avoid this problem set @env{CONFIG_SHELL} to @file{/bin/ksh}
1788 and @env{SHELL} to @file{/bin/ksh} in your environment.
1795 @heading @anchor{hppa*-hp-hpux10}hppa*-hp-hpux10
1797 For hpux10.20, we @emph{highly} recommend you pick up the latest sed patch
1798 @code{PHCO_19798} from HP@. HP has two sites which provide patches free of
1804 <a href="http://us-support.external.hp.com">US, Canada, Asia-Pacific, and
1808 @uref{http://us-support.external.hp.com,,}US, Canada, Asia-Pacific, and
1812 @uref{http://europe-support.external.hp.com,,Europe}
1815 The HP assembler on these systems is much better than the hpux9 assembler,
1816 but still has some problems. Most notably the assembler inserts timestamps
1817 into each object file it creates, causing the 3-stage comparison test to fail
1818 during a @samp{make bootstrap}. You should be able to continue by
1819 saying @samp{make all} after getting the failure from @samp{make
1827 @heading @anchor{hppa*-hp-hpux11}hppa*-hp-hpux11
1829 GCC 3.0 supports HP-UX 11. You must use GNU binutils 2.11 or above on
1836 @heading @anchor{*-*-freebsd*}*-*-freebsd*
1838 The version of binutils installed in @file{/usr/bin} is known to work unless
1839 otherwise specified in any per-architecture notes. However, binutils
1840 2.11 is known to improve overall testsuite results.
1842 For FreeBSD 1, FreeBSD 2 or any mutant a.out versions of FreeBSD 3: All
1843 configuration support and files as shipped with gcc 2.95 are still in
1844 place. FreeBSD 2.2.7 has been known to bootstrap completely; however,
1845 it is unknown which version of binutils was used (it is assumed that it
1846 was the system copy in @file{/usr/bin}) and C++ EH failures were noted.
1848 For FreeBSD using the ELF file format: DWARF2 debugging is now the
1849 default for all CPU architectures. It had been the default on
1850 FreeBSD/alpha since its inception. You may use @option{-gstabs} instead
1851 of @option{-g}, if you really want the old debugging format. There are
1852 no known issues with mixing object files and libraries with different
1853 debugging formats. Otherwise, this release of gcc should now match more
1854 of the configuration used in the stock FreeBSD configuration of gcc. In
1855 particular, @option{--enable-threads} is now configured by default.
1856 However, as a general user, do not attempt to replace the system
1857 compiler with this release. Known to bootstrap and check with good
1858 results on FreeBSD 3.0, 3.4, 4.0, 4.2, 4.3 and 5-CURRENT@.
1860 At this time, @option{--enable-threads} is not compatible with
1861 @option{--enable-libgcj} on FreeBSD@.
1867 @heading @anchor{i370-*-*}i370-*-*
1868 This port is very preliminary and has many known bugs. We hope to
1869 have a higher-quality port for this machine soon.
1875 @heading @anchor{*-*-linux-gnu}*-*-linux-gnu
1877 If you use glibc 2.2 (or 2.1.9x), GCC 2.95.2 won't install
1878 out-of-the-box. You'll get compile errors while building libstdc++.
1879 The patch @uref{glibc-2.2.patch,,glibc-2.2.patch}, that is to be
1880 applied in the GCC source tree, fixes the compatibility problems.
1890 Currently Glibc 2.2.3 (and older releases) and GCC 3.0 are out of sync
1891 since the latest exception handling changes for GCC@. Compiling glibc
1892 with GCC 3.0 will give a binary incompatible glibc and therefore cause
1893 lots of problems and might make your system completly unusable. This
1894 will definitly need fixes in glibc but might also need fixes in GCC@. We
1895 strongly advise to wait for glibc 2.2.4 and to read the release notes of
1896 glibc 2.2.4 whether patches for GCC 3.0 are needed. You can use glibc
1897 2.2.3 with GCC 3.0, just do not try to recompile it.
1903 @heading @anchor{ix86-*-linux*oldld}i?86-*-linux*oldld
1904 Use this configuration to generate @file{a.out} binaries on Linux-based
1905 GNU systems if you do not have gas/binutils version 2.5.2 or later
1906 installed. This is an obsolete configuration.
1912 @heading @anchor{ix86-*-linux*aout}i?86-*-linux*aout
1913 Use this configuration to generate @file{a.out} binaries on Linux-based
1914 GNU systems. This configuration is being superseded. You must use
1915 gas/binutils version 2.5.2 or later.
1921 @heading @anchor{ix86-*-linux*}i?86-*-linux*
1923 You will need binutils-2.9.1.0.15 or newer for exception handling to work.
1925 If you receive Signal 11 errors when building on GNU/Linux, then it is
1926 possible you have a hardware problem. Further information on this can be
1927 found on @uref{http://www.bitwizard.nl/sig11/,,www.bitwizard.nl}.
1933 @heading @anchor{ix86-*-sco}i?86-*-sco
1934 Compilation with RCC is recommended. Also, it may be a good idea to
1935 link with GNU malloc instead of the malloc that comes with the system.
1941 @heading @anchor{ix86-*-sco3.2v4}i?86-*-sco3.2v4
1942 Use this configuration for SCO release 3.2 version 4.
1948 @heading @anchor{ix86-*-sco3.2v5*}i?86-*-sco3.2v5*
1949 Use this for the SCO OpenServer Release 5 family of operating systems.
1951 Unlike earlier versions of GCC, the ability to generate COFF with this
1952 target is no longer provided.
1954 Earlier versions of GCC emitted Dwarf-1 when generating ELF to allow
1955 the system debugger to be used. That support was too burdensome to
1956 maintain. GCC now emits only dwarf-2 for this target. This means you
1957 may use either the UDK debugger or GDB to debug programs built by this
1960 If you are building languages other than C, you must follow the instructions
1961 about invoking @samp{make bootstrap} because the native OpenServer
1962 compiler will build a @command{cc1plus} that will not correctly parse many
1963 valid C++ programs including those in @file{libgcc.a}.
1964 @strong{You must do a @samp{make bootstrap} if you are building with the
1967 Use of the @option{-march-pentiumpro} flag can result in
1968 unrecognized opcodes when using the native assembler on OS versions before
1969 5.0.6. (Support for P6 opcodes was added to the native ELF assembler in
1970 that version.) While it's rather rare to see these emitted by GCC yet,
1971 errors of the basic form:
1974 /usr/tmp/ccaNlqBc.s:22:unknown instruction: fcomip
1975 /usr/tmp/ccaNlqBc.s:50:unknown instruction: fucomip
1978 are symptoms of this problem. You may work around this by not
1979 building affected files with that flag, by using the GNU assembler, or
1980 by using the assembler provided with the current version of the OS@.
1981 Users of GNU assembler should see the note below for hazards on doing
1984 The native SCO assembler that is provided with the OS at no
1985 charge is normally required. If, however, you must be able to use
1986 the GNU assembler (perhaps you're compiling code with asms that
1987 require GAS syntax) you may configure this package using the flags
1988 @uref{./configure.html#with-gnu-as,,@option{--with-gnu-as}}. You must
1989 use a recent version of GNU binutils; versions past 2.9.1 seem to work
1992 In general, the @option{--with-gnu-as} option isn't as well tested
1993 as the native assembler.
1995 Look in @file{gcc/config/i386/sco5.h} (search for ``messy'') for
1996 additional OpenServer-specific flags.
1998 Systems based on OpenServer before 5.0.4 (@samp{uname -X}
1999 will tell you what you're running) require TLS597 from ftp.sco.com/TLS
2000 for C++ constructors and destructors to work right.
2002 The system linker in (at least) 5.0.4 and 5.0.5 will sometimes
2003 do the wrong thing for a construct that GCC will emit for PIC
2004 code. This can be seen as execution testsuite failures when using
2005 @option{-fPIC} on @file{921215-1.c}, @file{931002-1.c}, @file{nestfunc-1.c}, and @file{gcov-1.c}.
2006 For 5.0.5, an updated linker that will cure this problem is
2007 available. You must install both
2008 @uref{ftp://ftp.sco.com/Supplements/rs505a/,,ftp://ftp.sco.com/Supplements/rs505a/}
2009 and @uref{ftp://ftp.sco.com/SLS/,,OSS499A}.
2011 The dynamic linker in OpenServer 5.0.5 (earlier versions may show
2012 the same problem) aborts on certain g77-compiled programs. It's particularly
2013 likely to be triggered by building Fortran code with the @option{-fPIC} flag.
2014 Although it's conceivable that the error could be triggered by other
2015 code, only G77-compiled code has been observed to cause this abort.
2016 If you are getting core dumps immediately upon execution of your
2017 g77 program---and especially if it's compiled with @option{-fPIC}---try applying
2018 @uref{sco_osr5_g77.patch,,@file{sco_osr5_g77.patch}} to your libf2c and
2020 Affected faults, when analyzed in a debugger, will show a stack
2021 backtrace with a fault occurring in @code{rtld()} and the program
2022 running as @file{/usr/lib/ld.so.1}. This problem has been reported to SCO
2023 engineering and will hopefully be addressed in later releases.
2030 @heading @anchor{ix86-*-udk}i?86-*-udk
2032 This target emulates the SCO Universal Development Kit and requires that
2033 package be installed. (If it is installed, you will have a
2034 @file{/udk/usr/ccs/bin/cc} file present.) It's very much like the
2035 @code{i?86-*-unixware7*} target
2036 but is meant to be used when hosting on a system where UDK isn't the
2037 default compiler such as OpenServer 5 or Unixware 2. This target will
2038 generate binaries that will run on OpenServer, Unixware 2, or Unixware 7,
2039 with the same warnings and caveats as the SCO UDK@.
2041 You can stage1 with either your native compiler or with UDK@. If you
2042 don't do a full bootstrap when initially building with your native compiler
2043 you will have an utterly unusable pile of bits as your reward.
2045 This target is a little tricky to build because we have to distinguish
2046 it from the native tools (so it gets headers, startups, and libraries
2047 from the right place) while making the tools not think we're actually
2048 building a cross compiler. The easiest way to do this is with a configure
2051 @samp{CC=/udk/usr/ccs/bin/cc @var{/your/path/to}/gcc/configure
2052 --host=i686-pc-udk --target=i686-pc-udk --program-prefix=udk-}
2054 @emph{You should substitute @samp{i686} in the above command with the appropriate
2055 processor for your host.}
2057 You should follow this with a @samp{make bootstrap} then
2058 @samp{make install}. You can then access the UDK-targeted GCC
2059 tools by adding @command{udk-} before the commonly known name. For
2060 example, to invoke the C compiler, you would use @command{udk-gcc}.
2061 They will coexist peacefully with any native-target GCC tools you may
2069 @heading @anchor{ix86-*-isc}i?86-*-isc
2070 It may be a good idea to link with GNU malloc instead of the malloc that
2071 comes with the system.
2073 In ISC version 4.1, @command{sed} core dumps when building
2074 @file{deduced.h}. Use the version of @command{sed} from version 4.0.
2080 @heading @anchor{ix86-*-esix}i?86-*-esix
2081 It may be good idea to link with GNU malloc instead of the malloc that
2082 comes with the system.
2088 @heading @anchor{ix86-ibm-aix}i?86-ibm-aix
2089 You need to use GAS version 2.1 or later, and LD from
2090 GNU binutils version 2.2 or later.
2096 @heading @anchor{ix86-sequent-bsd}i?86-sequent-bsd
2097 Go to the Berkeley universe before compiling.
2103 @heading @anchor{ix86-sequent-ptx1*}i?86-sequent-ptx1*, i?86-sequent-ptx2*
2104 You must install GNU @file{sed} before running @file{configure}.
2110 @heading @anchor{#ix86-*-sysv3*}i?86-*-sysv3*
2111 The @code{fixproto} shell script may trigger a bug in the system shell.
2112 If you encounter this problem, upgrade your operating system or
2113 use BASH (the GNU shell) to run @code{fixproto}.
2120 @heading @anchor{i860-intel-osf*}i860-intel-osf*
2121 On the Intel Paragon (an i860 machine), if you are using operating
2122 system version 1.0, you will get warnings or errors about redefinition
2123 of @code{va_arg} when you build GCC@.
2125 If this happens, then you need to link most programs with the library
2126 @file{iclib.a}. You must also modify @file{stdio.h} as follows: before
2130 #if defined(__i860__) && !defined(_VA_LIST)
2131 #include <va_list.h>
2145 extern int vprintf(const char *, va_list );
2146 extern int vsprintf(char *, const char *, va_list );
2154 #endif /* __PGC__ */
2157 These problems don't exist in operating system version 1.1.
2163 @heading @anchor{*-lynx-lynxos}*-lynx-lynxos
2164 LynxOS 2.2 and earlier comes with GCC 1.x already installed as
2165 @file{/bin/gcc}. You should compile with this instead of @file{/bin/cc}.
2166 You can tell GCC to use the GNU assembler and linker, by specifying
2167 @samp{--with-gnu-as --with-gnu-ld} when configuring. These will produce
2168 COFF format object files and executables; otherwise GCC will use the
2169 installed tools, which produce @file{a.out} format executables.
2174 <!-- rs6000-ibm-aix*, powerpc-ibm-aix* -->
2176 @heading @anchor{*-ibm-aix*}*-ibm-aix*
2178 AIX Make frequently has problems with GCC makefiles. GNU Make 3.76 or
2179 newer is recommended to build on this platform.
2181 Errors involving @code{alloca} when building GCC generally are due
2182 to an incorrect definition of @var{CC} in the Makefile or mixing files
2183 compiled with the native C compiler and GCC@. During the stage1 phase of
2184 the build, the native AIX compiler @strong{must} be invoked as @command{cc}
2185 (not @command{xlc}). Once @command{configure} has been informed of
2186 @command{xlc}, one needs to use @samp{make distclean} to remove the
2187 configure cache files and ensure that @env{CC} environment variable
2188 does not provide a definition that will confuse @command{configure}.
2189 If this error occurs during stage2 or later, then the problem most likely
2190 is the version of Make (see above).
2192 Binutils 2.10 does not support AIX 4.3. Binutils available from the
2193 @uref{http://www-1.ibm.com/servers/aix/products/aixos/linux/,,AIX
2194 Toolbox for Linux: GNU and Open Source tools for AIX};
2195 website does work. Binutils 2.11 is expected to include AIX 4.3
2196 support. The GNU Assembler is necessary for libstdc++ to build. The
2197 AIX native ld still is recommended. The native AIX tools do
2198 interoperate with GCC@.
2200 Linking executables and shared libraries may produce warnings of
2201 duplicate symbols. The assembly files generated by GCC for AIX always
2202 have included multiple symbol definitions for certain global variable
2203 and function declarations in the original program. The warnings should
2204 not prevent the linker from producing a correct library or runnable
2207 AIX 4.3 utilizes a ``large format'' archive to support both 32-bit and
2208 64-bit object modules. The routines provided in AIX 4.3.0 and AIX 4.3.1
2209 to parse archive libraries did not handle the new format correctly.
2210 These routines are used by GCC and result in error messages during
2211 linking such as ``not a COFF file''. The version of the routines shipped
2212 with AIX 4.3.1 should work for a 32-bit environment. The @option{-g}
2213 option of the archive command may be used to create archives of 32-bit
2214 objects using the original ``small format''. A correct version of the
2215 routines is shipped with AIX 4.3.2.
2217 Some versions of the AIX binder (linker) can fail with a relocation
2218 overflow severe error when the @option{-bbigtoc} option is used to link
2219 GCC-produced object files into an executable that overflows the TOC@. A fix
2220 for APAR IX75823 (OVERFLOW DURING LINK WHEN USING GCC AND -BBIGTOC) is
2221 available from IBM Customer Support and from its
2222 @uref{http://service.boulder.ibm.com/,,service.boulder.ibm.com}
2223 website as PTF U455193.
2225 The AIX 4.3.2.1 linker (bos.rte.bind_cmds Level 4.3.2.1) will dump core
2226 with a segmentation fault when invoked by any version of GCC@. A fix for
2227 APAR IX87327 is available from IBM Customer Support and from its
2228 @uref{http://service.boulder.ibm.com/,,service.boulder.ibm.com}
2229 website as PTF U461879. This fix is incorporated in AIX 4.3.3 and above.
2231 The initial assembler shipped with AIX 4.3.0 generates incorrect object
2232 files. A fix for APAR IX74254 (64BIT DISASSEMBLED OUTPUT FROM COMPILER FAILS
2233 TO ASSEMBLE/BIND) is available from IBM Customer Support and from its
2234 @uref{http://service.boulder.ibm.com/,,service.boulder.ibm.com}
2235 website as PTF U453956. This fix is incorporated in AIX 4.3.1 and above.
2237 AIX provides National Language Support (NLS)@. Compilers and assemblers
2238 use NLS to support locale-specific representations of various data
2239 formats including floating-point numbers (e.g., @samp{.} vs @samp{,} for
2240 separating decimal fractions). There have been problems reported where
2241 GCC does not produce the same floating-point formats that the assembler
2242 expects. If one encounters this problem, set the @env{LANG}
2243 environment variable to @samp{C} or @samp{En_US}.
2245 By default, GCC for AIX 4.1 and above produces code that can be used on
2246 both Power or PowerPC processors.
2248 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
2249 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
2255 @heading @anchor{m32r-*-elf}m32r-*-elf
2256 Mitsubishi M32R processor.
2257 This configuration is intended for embedded systems.
2263 @heading @anchor{m68000-hp-bsd}m68000-hp-bsd
2264 HP 9000 series 200 running BSD@. Note that the C compiler that comes
2265 with this system cannot compile GCC; contact @email{law@@cygnus.com}
2266 to get binaries of GCC for bootstrapping.
2272 @heading @anchor{m6811-elf}m6811-elf
2273 Motorola 68HC11 family micro controllers. These are used in embedded
2274 applications. There are no standard Unix configurations.
2280 @heading @anchor{m6812-elf}m6812-elf
2281 Motorola 68HC12 family micro controllers. These are used in embedded
2282 applications. There are no standard Unix configurations.
2288 @heading @anchor{m68k-altos}m68k-altos
2289 Altos 3068. You must use the GNU assembler, linker and debugger.
2290 Also, you must fix a kernel bug. Details in the file @file{README.ALTOS}.
2296 @heading @anchor{m68k-apple-aux}m68k-apple-aux
2297 Apple Macintosh running A/UX@.
2298 You may configure GCC to use either the system assembler and
2299 linker or the GNU assembler and linker. You should use the GNU configuration
2300 if you can, especially if you also want to use G++. You enabled
2301 that configuration with + the @option{--with-gnu-as} and @option{--with-gnu-ld}
2302 options to @code{configure}.
2304 Note the C compiler that comes
2305 with this system cannot compile GCC@. You can find binaries of GCC
2306 for bootstrapping on @code{jagubox.gsfc.nasa.gov}.
2307 You will also a patched version of @file{/bin/ld} there that
2308 raises some of the arbitrary limits found in the original.
2314 @heading @anchor{m68k-att-sysv}m68k-att-sysv
2315 AT&T 3b1, a.k.a.@: 7300 PC@. This version of GCC cannot
2316 be compiled with the system C compiler, which is too buggy.
2317 You will need to get a previous version of GCC and use it to
2318 bootstrap. Binaries are available from the OSU-CIS archive, at
2319 @uref{ftp://archive.cis.ohio-state.edu/pub/att7300/}.
2325 @heading @anchor{m68k-bull-sysv}m68k-bull-sysv
2326 Bull DPX/2 series 200 and 300 with BOS-2.00.45 up to BOS-2.01. GCC works
2327 either with native assembler or GNU assembler. You can use
2328 GNU assembler with native coff generation by providing @option{--with-gnu-as} to
2329 the configure script or use GNU assembler with dbx-in-coff encapsulation
2330 by providing @samp{--with-gnu-as --stabs}. For any problem with native
2331 assembler or for availability of the DPX/2 port of GAS, contact
2332 @email{F.Pierresteguy@@frcl.bull.fr}.
2338 @heading @anchor{m68k-crds-unox}m68k-crds-unox
2339 Use @samp{configure unos} for building on Unos.
2341 The Unos assembler is named @code{casm} instead of @code{as}. For some
2342 strange reason linking @file{/bin/as} to @file{/bin/casm} changes the
2343 behavior, and does not work. So, when installing GCC, you should
2344 install the following script as @file{as} in the subdirectory where
2345 the passes of GCC are installed:
2352 The default Unos library is named @file{libunos.a} instead of
2353 @file{libc.a}. To allow GCC to function, either change all
2354 references to @option{-lc} in @file{gcc.c} to @option{-lunos} or link
2355 @file{/lib/libc.a} to @file{/lib/libunos.a}.
2357 @cindex @code{alloca}, for Unos
2358 When compiling GCC with the standard compiler, to overcome bugs in
2359 the support of @code{alloca}, do not use @option{-O} when making stage 2.
2360 Then use the stage 2 compiler with @option{-O} to make the stage 3
2361 compiler. This compiler will have the same characteristics as the usual
2362 stage 2 compiler on other systems. Use it to make a stage 4 compiler
2363 and compare that with stage 3 to verify proper compilation.
2365 (Perhaps simply defining @code{ALLOCA} in @file{x-crds} as described in
2366 the comments there will make the above paragraph superfluous. Please
2367 inform us of whether this works.)
2369 Unos uses memory segmentation instead of demand paging, so you will need
2370 a lot of memory. 5 Mb is barely enough if no other tasks are running.
2371 If linking @file{cc1} fails, try putting the object files into a library
2372 and linking from that library.
2378 @heading @anchor{m68k-hp-hpux}m68k-hp-hpux
2379 HP 9000 series 300 or 400 running HP-UX@. HP-UX version 8.0 has a bug in
2380 the assembler that prevents compilation of GCC@. This
2381 bug manifests itself during the first stage of compilation, while
2382 building @file{libgcc2.a}:
2386 cc1: warning: `-g' option not supported on this version of GCC
2387 cc1: warning: `-g1' option not supported on this version of GCC
2388 ./xgcc: Internal compiler error: program as got fatal signal 11
2391 A patched version of the assembler is available as the file
2392 @uref{ftp://altdorf.ai.mit.edu/archive/cph/hpux-8.0-assembler}. If you
2393 have HP software support, the patch can also be obtained directly from
2394 HP, as described in the following note:
2397 This is the patched assembler, to patch SR#1653-010439, where the
2398 assembler aborts on floating point constants.
2400 The bug is not really in the assembler, but in the shared library
2401 version of the function ``cvtnum(3c)''. The bug on ``cvtnum(3c)'' is
2402 SR#4701-078451. Anyway, the attached assembler uses the archive
2403 library version of ``cvtnum(3c)'' and thus does not exhibit the bug.
2406 This patch is also known as PHCO_4484.
2408 In addition, if you wish to use gas @option{--with-gnu-as} you must use
2409 gas version 2.1 or later, and you must use the GNU linker version 2.1 or
2410 later. Earlier versions of gas relied upon a program which converted the
2411 gas output into the native HP-UX format, but that program has not been
2412 kept up to date. gdb does not understand that native HP-UX format, so
2413 you must use gas if you wish to use gdb.
2415 On HP-UX version 8.05, but not on 8.07 or more recent versions, the
2416 @code{fixproto} shell script triggers a bug in the system shell. If you
2417 encounter this problem, upgrade your operating system or use BASH (the
2418 GNU shell) to run @code{fixproto}. This bug will cause the fixproto
2419 program to report an error of the form:
2422 ./fixproto: sh internal 1K buffer overflow
2425 To fix this, you can also change the first line of the fixproto script
2437 @heading @anchor{m68k-*-nextstep*}m68k-*-nextstep*
2439 Current GCC versions probably do not work on version 2 of the NeXT
2442 On NeXTStep 3.0, the Objective-C compiler does not work, due,
2443 apparently, to a kernel bug that it happens to trigger. This problem
2444 does not happen on 3.1.
2446 You absolutely @strong{must} use GNU sed and GNU make on this platform.
2449 On NEXTSTEP 3.x where x < 3 the build of GCC will abort during
2450 stage1 with an error message like this:
2454 /usr/tmp/ccbbsZ0U.s:987:Unknown pseudo-op: .section
2455 /usr/tmp/ccbbsZ0U.s:987:Rest of line ignored. 1st junk character
2459 The reason for this is the fact that NeXT's assembler for these
2460 versions of the operating system does not support the @samp{.section}
2461 pseudo op that's needed for full C++ exception functionality.
2463 As NeXT's assembler is a derived work from GNU as, a free
2464 replacement that does can be obtained at
2465 @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}.
2467 If you try to build the integrated C++ & C++ runtime libraries on this system
2468 you will run into trouble with include files. The way to get around this is
2469 to use the following sequence. Note you must have write permission to
2470 the directory @var{prefix} you specified in the configuration process of GCC
2471 for this sequence to work.
2475 make all-texinfo all-bison all-byacc all-binutils all-gas all-ld
2478 make install-headers-tar
2487 @heading @anchor{m68k-ncr-*}m68k-ncr-*
2488 On the Tower models 4@var{n}0 and 6@var{n}0, by default a process is not
2489 allowed to have more than one megabyte of memory. GCC cannot compile
2490 itself (or many other programs) with @option{-O} in that much memory.
2492 To solve this problem, reconfigure the kernel adding the following line
2493 to the configuration file:
2504 @heading @anchor{m68k-sun}m68k-sun
2505 Sun 3. We do not provide a configuration file to use the Sun FPA by
2506 default, because programs that establish signal handlers for floating
2507 point traps inherently cannot work with the FPA@.
2513 @heading @anchor{m68k-sun-sunos4.1.1}m68k-sun-sunos4.1.1
2515 It is reported that you may need the GNU assembler on this platform.
2522 @heading @anchor{m88k-*-svr3}m88k-*-svr3
2523 Motorola m88k running the AT&T/Unisoft/Motorola V.3 reference port.
2524 These systems tend to use the Green Hills C, revision 1.8.5, as the
2525 standard C compiler. There are apparently bugs in this compiler that
2526 result in object files differences between stage 2 and stage 3. If this
2527 happens, make the stage 4 compiler and compare it to the stage 3
2528 compiler. If the stage 3 and stage 4 object files are identical, this
2529 suggests you encountered a problem with the standard C compiler; the
2530 stage 3 and 4 compilers may be usable.
2532 It is best, however, to use an older version of GCC for bootstrapping
2539 @heading @anchor{m88k-*-dgux}m88k-*-dgux
2540 Motorola m88k running DG/UX@. To build 88open BCS native or cross
2541 compilers on DG/UX, specify the configuration name as
2542 @samp{m88k-*-dguxbcs} and build in the 88open BCS software development
2543 environment. To build ELF native or cross compilers on DG/UX, specify
2544 @samp{m88k-*-dgux} and build in the DG/UX ELF development environment.
2545 You set the software development environment by issuing
2546 @samp{sde-target} command and specifying either @samp{m88kbcs} or
2547 @samp{m88kdguxelf} as the operand.
2549 If you do not specify a configuration name, @file{configure} guesses the
2550 configuration based on the current software development environment.
2556 @heading @anchor{m88k-tektronix-sysv3}m88k-tektronix-sysv3
2557 Tektronix XD88 running UTekV 3.2e. Do not turn on
2558 optimization while building stage1 if you bootstrap with
2559 the buggy Green Hills compiler. Also, The bundled LAI
2560 System V NFS is buggy so if you build in an NFS mounted
2561 directory, start from a fresh reboot, or avoid NFS all together.
2562 Otherwise you may have trouble getting clean comparisons
2569 @heading @anchor{mips-*-*}mips-*-*
2570 If you use the 1.31 version of the MIPS assembler (such as was shipped
2571 with Ultrix 3.1), you will need to use the @option{-fno-delayed-branch} switch
2572 when optimizing floating point code. Otherwise, the assembler will
2573 complain when the GCC compiler fills a branch delay slot with a
2574 floating point instruction, such as @code{add.d}.
2576 If on a MIPS system you get an error message saying ``does not have gp
2577 sections for all it's [sic] sectons [sic]'', don't worry about it. This
2578 happens whenever you use GAS with the MIPS linker, but there is not
2579 really anything wrong, and it is okay to use the output file. You can
2580 stop such warnings by installing the GNU linker.
2582 It would be nice to extend GAS to produce the gp tables, but they are
2583 optional, and there should not be a warning about their absence.
2585 Users have reported some problems with version 2.0 of the MIPS
2586 compiler tools that were shipped with Ultrix 4.1. Version 2.10
2587 which came with Ultrix 4.2 seems to work fine.
2589 Users have also reported some problems with version 2.20 of the
2590 MIPS compiler tools that were shipped with RISC/os 4.x. The earlier
2591 version 2.11 seems to work fine.
2593 Some versions of the MIPS linker will issue an assertion failure
2594 when linking code that uses @code{alloca} against shared
2595 libraries on RISC-OS 5.0, and DEC's OSF/1 systems. This is a bug
2596 in the linker, that is supposed to be fixed in future revisions.
2597 To protect against this, GCC passes @option{-non_shared} to the
2598 linker unless you pass an explicit @option{-shared} or
2599 @option{-call_shared} switch.
2601 @heading @anchor{mips-mips-bsd}mips-mips-bsd
2602 MIPS machines running the MIPS operating system in BSD mode. It's
2603 possible that some old versions of the system lack the functions
2604 @code{memcpy}, @code{memmove}, @code{memcmp}, and @code{memset}. If your
2605 system lacks these, you must remove or undo the definition of
2606 @code{TARGET_MEM_FUNCTIONS} in @file{mips-bsd.h}.
2608 The MIPS C compiler needs to be told to increase its table size
2609 for switch statements with the @option{-Wf,-XNg1500} option in
2610 order to compile @file{cp/parse.c}. If you use the @option{-O2}
2611 optimization option, you also need to use @option{-Olimit 3000}.
2612 Both of these options are automatically generated in the
2613 @file{Makefile} that the shell script @file{configure} builds.
2614 If you override the @code{CC} make variable and use the MIPS
2615 compilers, you may need to add @option{-Wf,-XNg1500 -Olimit 3000}.
2621 @heading @anchor{mips-mips-riscos*}mips-mips-riscos*
2622 The MIPS C compiler needs to be told to increase its table size
2623 for switch statements with the @option{-Wf,-XNg1500} option in
2624 order to compile @file{cp/parse.c}. If you use the @option{-O2}
2625 optimization option, you also need to use @option{-Olimit 3000}.
2626 Both of these options are automatically generated in the
2627 @file{Makefile} that the shell script @file{configure} builds.
2628 If you override the @code{CC} make variable and use the MIPS
2629 compilers, you may need to add @samp{-Wf,-XNg1500 -Olimit 3000}.
2631 MIPS computers running RISC-OS can support four different
2632 personalities: default, BSD 4.3, System V.3, and System V.4
2633 (older versions of RISC-OS don't support V.4). To configure GCC
2634 for these platforms use the following configurations:
2637 @item mips-mips-riscos@code{rev}
2638 Default configuration for RISC-OS, revision @code{rev}.
2640 @item mips-mips-riscos@code{rev}bsd
2641 BSD 4.3 configuration for RISC-OS, revision @code{rev}.
2643 @item mips-mips-riscos@code{rev}sysv4
2644 System V.4 configuration for RISC-OS, revision @code{rev}.
2650 @item mips-mips-riscos@code{rev}sysv
2651 System V.3 configuration for RISC-OS, revision @code{rev}.
2654 The revision @code{rev} mentioned above is the revision of
2655 RISC-OS to use. You must reconfigure GCC when going from a
2656 RISC-OS revision 4 to RISC-OS revision 5. This has the effect of
2657 avoiding a linker bug.
2663 @heading @anchor{mips*-sgi-irix4*}mips*-sgi-irix4*
2665 In order to compile GCC on an SGI running IRIX 4, the ``c.hdr.lib''
2666 option must be installed from the CD-ROM supplied from Silicon Graphics.
2667 This is found on the 2nd CD in release 4.0.1.
2669 On IRIX version 4.0.5F, and perhaps on some other versions as well,
2670 there is an assembler bug that reorders instructions incorrectly. To
2671 work around it, specify the target configuration
2672 @samp{mips-sgi-irix4loser}. This configuration inhibits assembler
2675 In a compiler configured with target @samp{mips-sgi-irix4}, you can turn
2676 off assembler optimization by using the @option{-noasmopt} option. This
2677 compiler option passes the option @option{-O0} to the assembler, to
2680 The @option{-noasmopt} option can be useful for testing whether a problem
2681 is due to erroneous assembler reordering. Even if a problem does not go
2682 away with @option{-noasmopt}, it may still be due to assembler
2683 reordering---perhaps GCC itself was miscompiled as a result.
2685 You may get the following warning on IRIX 4 platforms, it can be safely
2688 warning: foo.o does not have gp tables for all its sections.
2695 @heading @anchor{mips*-sgi-irix5*}mips*-sgi-irix5*
2697 In order to compile GCC on an SGI running IRIX 5, the ``compiler_dev.hdr''
2698 subsystem must be installed from the IDO CD-ROM supplied by Silicon
2699 Graphics. It is also available for download from
2700 @uref{http://www.sgi.com/developers/devtools/apis/ido.html,,http://www.sgi.com/developers/devtools/apis/ido.html}.
2702 @code{make compare} may fail on version 5 of IRIX unless you add
2703 @option{-save-temps} to @code{CFLAGS}. On these systems, the name of the
2704 assembler input file is stored in the object file, and that makes
2705 comparison fail if it differs between the @code{stage1} and
2706 @code{stage2} compilations. The option @option{-save-temps} forces a
2707 fixed name to be used for the assembler input file, instead of a
2708 randomly chosen name in @file{/tmp}. Do not add @option{-save-temps}
2709 unless the comparisons fail without that option. If you do you
2710 @option{-save-temps}, you will have to manually delete the @samp{.i} and
2711 @samp{.s} files after each series of compilations.
2713 If you use the MIPS C compiler to bootstrap, it may be necessary
2714 to increase its table size for switch statements with the
2715 @option{-Wf,-XNg1500} option. If you use the @option{-O2}
2716 optimization option, you also need to use @option{-Olimit 3000}.
2718 To enable debugging under IRIX 5, you must use GNU @command{as} 2.5 or later,
2719 and use the @option{--with-gnu-as} configure option when configuring GCC.
2720 GNU @command{as} is distributed as part of the binutils package.
2722 You must use GNU @command{as} on these platforms, as the native
2723 assembler can not handle the code for exception handling support. Either
2724 of these messages indicates that you are using the MIPS assembler when
2725 instead you should be using GNU @command{as}:
2728 as0: Error: ./libgcc2.c, line 1:Badly delimited numeric literal
2729 .4byte $LECIE1-$LSCIE1
2730 as0: Error: ./libgcc2.c, line 1:malformed statement
2736 as0: Error: ./libgcc2.c, line 1:undefined symbol in expression
2737 .word $LECIE1-$LSCIE1
2740 When building GCC, the build process loops rebuilding @command{cc1} over
2741 and over again. This happens on @samp{mips-sgi-irix5.2}, and possibly
2742 other platforms. It has been reported that this is a known bug in the
2743 @command{make} shipped with IRIX 5.2. We recommend you use GNU
2744 @command{make} instead of the vendor supplied @command{make} program;
2745 however, you may have success with @command{smake} on IRIX 5.2 if you do
2746 not have GNU @command{make} available.
2752 @heading @anchor{mips*-sgi-irix6}mips*-sgi-irix6
2754 If you are using IRIX @command{cc} as your bootstrap compiler, you must
2755 ensure that the N32 ABI is in use. To test this, compile a simple C
2756 file with @command{cc} and then run @command{file} on the
2757 resulting object file. The output should look like:
2760 test.o: ELF N32 MSB @dots{}
2766 test.o: ELF 32-bit MSB @dots{}
2772 test.o: ELF 64-bit MSB @dots{}
2775 then your version of @command{cc} uses the O32 or N64 ABI by default. You
2776 should set the environment variable @env{CC} to @samp{cc -n32}
2777 before configuring GCC@.
2779 GCC on IRIX 6 is usually built to support both the N32 and N64 ABIs. If
2780 you build GCC on a system that doesn't have the N64 libraries installed,
2781 you need to configure with @option{--disable-multilib} so GCC doesn't
2782 try to use them. Look for @file{/usr/lib64/libc.so.1} to see if you
2783 have the 64-bit libraries installed.
2785 You must @emph{not} use GNU @command{as} (which isn't built anyway as of
2786 binutils 2.11.2) on IRIX 6 platforms; doing so will only cause problems.
2788 GCC does not currently support generating O32 ABI binaries in the
2789 @samp{mips-sgi-irix6} configurations. It used to be possible to create a GCC
2790 with O32 ABI only support by configuring it for the @samp{mips-sgi-irix5}
2791 target, which doesn't currently (2001-06-13) work itself. It is
2792 expected that O32 ABI support will be available again in a future release.
2794 GCC does not correctly pass/return structures which are
2795 smaller than 16 bytes and which are not 8 bytes. The problem is very
2796 involved and difficult to fix. It affects a number of other targets also,
2797 but IRIX 6 is affected the most, because it is a 64 bit target, and 4 byte
2798 structures are common. The exact problem is that structures are being padded
2799 at the wrong end, e.g.@: a 4 byte structure is loaded into the lower 4 bytes
2800 of the register when it should be loaded into the upper 4 bytes of the
2803 GCC is consistent with itself, but not consistent with the SGI C compiler
2804 (and the SGI supplied runtime libraries), so the only failures that can
2805 happen are when there are library functions that take/return such
2806 structures. There are very few such library functions. Currently this
2807 is known to affect @code{inet_ntoa}, @code{inet_lnaof},
2808 @code{inet_netof}, @code{inet_makeaddr}, and @code{semctl}.
2810 See @uref{http://freeware.sgi.com/,,http://freeware.sgi.com/} for more
2811 information about using GCC on IRIX platforms.
2817 @heading @anchor{mips-sony-sysv}mips-sony-sysv
2818 Sony MIPS NEWS@. This works in NEWSOS 5.0.1, but not in 5.0.2 (which
2819 uses ELF instead of COFF)@. Support for 5.0.2 will probably be provided
2820 soon by volunteers. In particular, the linker does not like the
2821 code generated by GCC when shared libraries are linked in.
2828 @heading @anchor{ns32k-encore}ns32k-encore
2829 Encore ns32000 system. Encore systems are supported only under BSD@.
2835 @heading @anchor{ns32k-*-genix}ns32k-*-genix
2836 National Semiconductor ns32000 system. Genix has bugs in @code{alloca}
2837 and @code{malloc}; you must get the compiled versions of these from GNU
2844 @heading @anchor{ns32k-sequent}ns32k-sequent
2845 Go to the Berkeley universe before compiling.
2851 @heading @anchor{ns32k-utek}ns32k-utek
2852 UTEK ns32000 system (``merlin''). The C compiler that comes with this
2853 system cannot compile GCC; contact @samp{tektronix!reed!mason} to get
2854 binaries of GCC for bootstrapping.
2861 @heading @anchor{powerpc-*-elf}powerpc-*-elf, powerpc-*-sysv4
2862 PowerPC system in big endian mode, running System V.4.
2864 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
2865 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
2871 @heading @anchor{powerpc-*-linux-gnu*}powerpc-*-linux-gnu*
2874 @uref{ftp://ftp.varesearch.com/pub/support/hjl/binutils,,binutils-2.9.4.0.8}
2875 or newer for a working GCC@. It is strongly recommended to recompile binutils
2876 if you initially built it with gcc-2.7.2.x.
2878 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
2879 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
2885 @heading @anchor{powerpc-*-eabiaix}powerpc-*-eabiaix
2886 Embedded PowerPC system in big endian mode with @option{-mcall-aix} selected as
2889 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
2890 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
2896 @heading @anchor{powerpc-*-eabisim}powerpc-*-eabisim
2897 Embedded PowerPC system in big endian mode for use in running under the
2900 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
2901 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
2907 @heading @anchor{powerpc-*-eabi}powerpc-*-eabi
2908 Embedded PowerPC system in big endian mode.
2910 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
2911 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
2917 @heading @anchor{powerpcle-*-elf}powerpcle-*-elf, powerpcle-*-sysv4
2918 PowerPC system in little endian mode, running System V.4.
2920 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
2921 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
2927 @heading @anchor{powerpcle-*-eabisim}powerpcle-*-eabisim
2928 Embedded PowerPC system in little endian mode for use in running under
2935 @heading @anchor{powerpcle-*-eabi}powerpcle-*-eabi
2936 Embedded PowerPC system in little endian mode.
2938 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
2939 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
2945 @heading @anchor{powerpcle-*-winnt}powerpcle-*-winnt, powerpcle-*-pe
2946 PowerPC system in little endian mode running Windows NT@.
2948 You can specify a default version for the @option{-mcpu=@var{cpu_type}}
2949 switch by using the configure option @option{--with-cpu-@var{cpu_type}}.
2955 @heading @anchor{romp-*-aos}romp-*-aos, romp-*-mach
2956 The only operating systems supported for the IBM RT PC are AOS and
2957 MACH@. GCC does not support AIX running on the RT@. We recommend you
2958 compile GCC with an earlier version of itself; if you compile GCC
2959 with @code{hc}, the Metaware compiler, it will work, but you will get
2960 mismatches between the stage 2 and stage 3 compilers in various files.
2961 These errors are minor differences in some floating-point constants and
2962 can be safely ignored; the stage 3 compiler is correct.
2968 @heading @anchor{*-*-solaris*}*-*-solaris*
2970 Starting with Solaris, Sun does not ship a C compiler any more. To
2971 bootstrap and install GCC you first have to install a pre-built
2972 compiler, see our @uref{binaries.html,,binaries page} for
2975 You must use GNU Make to build GCC on Solaris 2. If you don't have GNU
2976 Make installed, you can use the prebuilt compiler mentioned above to
2979 Sun as 4.X is broken in that it cannot cope with long symbol names.
2980 A typical error message might look similar to the following:
2982 @samp{/usr/ccs/bin/as: "/var/tmp/ccMsw135.s", line 11041:
2983 error: can't compute value of an expression involving an external symbol.}
2985 This is Sun bug 4237974. This is fixed with patch 108908-02 and has
2986 been fixed in later (5.x) versions of the assembler.
2988 Solaris' @file{/bin/sh} will often fail to configure libstdc++-v3, boehm-gc or
2989 libjava. If you encounter this problem, set @var{CONFIG_SHELL} to
2990 @file{/bin/ksh} in your environment and run @samp{make bootstrap} again.
2991 Another possibility that sometimes helps is to remove
2992 @file{*-*-solaris*/config.cache}.
2998 @heading @anchor{sparc-sun-*}sparc-sun-*
2999 Sometimes on a Sun 4 you may observe a crash in the program
3000 @code{genflags} or @code{genoutput} while building GCC@. This is said to
3001 be due to a bug in @code{sh}. You can probably get around it by running
3002 @code{genflags} or @code{genoutput} manually and then retrying the
3009 @heading @anchor{sparc-sun-solaris*}sparc-sun-solaris*
3010 On Solaris 2, executables of GCC version 2.0.2 are commonly
3011 available, but they have a bug that shows up when compiling current
3012 versions of GCC: undefined symbol errors occur during assembly if you
3015 The solution is to compile the current version of GCC without
3016 @option{-g}. That makes a working compiler which you can use to recompile
3019 Solaris 2 comes with a number of optional OS packages. Some of these
3020 packages are needed to use GCC fully. If you did not install all
3021 optional packages when installing Solaris, you will need to verify that
3022 the packages that GCC needs are installed.
3024 To check whether an optional package is installed, use
3025 the @code{pkginfo} command. To add an optional package, use the
3026 @code{pkgadd} command. For further details, see the Solaris
3029 For Solaris 2.0 and 2.1, GCC needs six packages: @samp{SUNWarc},
3030 @samp{SUNWbtool}, @samp{SUNWesu}, @samp{SUNWhea}, @samp{SUNWlibm}, and
3033 For Solaris 2.2, GCC needs an additional seventh package: @samp{SUNWsprot}.
3035 On Solaris 2, trying to use the linker and other tools in
3036 @file{/usr/ucb} to install GCC has been observed to cause trouble.
3037 For example, the linker may hang indefinitely. The fix is to remove
3038 @file{/usr/ucb} from your @code{PATH}.
3040 All releases of GNU binutils prior to 2.11.1 have known bugs on this
3041 platform. We recommend the use of GNU binutils 2.11.1 or the vendor
3042 tools (Sun as, Sun ld).
3044 Unfortunately, C++ shared libraries, including libstdc++, won't work
3045 properly if assembled with Sun as: the linker will complain about
3046 relocations in read-only sections, in the definition of virtual
3047 tables. Also, Sun as fails to process long symbols resulting from
3048 mangling template-heavy C++ function names.
3054 @heading @anchor{sparc-sun-solaris2.7}sparc-sun-solaris2.7
3056 Sun patch 107058-01 (1999-01-13) for SPARC Solaris 7 triggers a bug in
3057 the dynamic linker. This problem (Sun bug 4210064) affects GCC 2.8
3058 and later, including all EGCS releases. Sun formerly recommended
3059 107058-01 for all Solaris 7 users, but around 1999-09-01 it started to
3060 recommend it only for people who use Sun's compilers.
3062 Here are some workarounds to this problem:
3065 Do not install Sun patch 107058-01 until after Sun releases a
3066 complete patch for bug 4210064. This is the simplest course to take,
3067 unless you must also use Sun's C compiler. Unfortunately 107058-01
3068 is preinstalled on some new Solaris-based hosts, so you may have to
3072 Copy the original, unpatched Solaris 7
3073 @command{/usr/ccs/bin/as} into
3074 @command{/usr/local/lib/gcc-lib/sparc-sun-solaris2.7/2.95.1/as},
3075 adjusting the latter name to fit your local conventions and software
3079 Install Sun patch 106950-03 (1999-05-25) or later. Nobody with
3080 both 107058-01 and 106950-03 installed has reported the bug with GCC
3081 and Sun's dynamic linker. This last course of action is riskiest,
3082 for two reasons. First, you must install 106950 on all hosts that
3083 run code generated by GCC; it doesn't suffice to install it only on
3084 the hosts that run GCC itself. Second, Sun says that 106950-03 is
3085 only a partial fix for bug 4210064, but Sun doesn't know whether the
3086 partial fix is adequate for GCC@. Revision -08 or later should fix
3087 the bug, but (as of 1999-10-06) it is still being tested.
3094 <!-- ripped from the same FAQ that I answered -->
3096 @heading @anchor{*-sun-solaris2.8}*-sun-solaris2.8
3098 Sun bug 4296832 turns up when compiling X11 headers with GCC 2.95 or
3099 newer: @command{g++} will complain that types are missing. These headers assume
3100 that omitting the type means @code{int}; this assumption worked for C89 but
3101 is wrong for C++, and is now wrong for C99 also.
3103 @command{g++} accepts such (illegal) constructs with the option @option{-fpermissive}; it
3104 will assume that any missing type is @code{int} (as defined by C89).
3106 For Solaris 8, this is fixed by revision 24 or later of patch 108652
3107 (for SPARCs) or 108653 (for Intels).
3109 Solaris 8's linker fails to link some libjava programs if
3110 previously-installed GCC java libraries already exist in the configured
3111 prefix. For this reason, libgcj is disabled by default on Solaris 8.
3112 If you use GNU ld, or if you don't have a previously-installed libgcj in
3113 the same prefix, use @option{--enable-libgcj} to build and install the
3120 @heading @anchor{sunv5}Sun V5.0 Compiler Bugs
3122 The Sun V5.0 compilers are known to mis-compile GCC 2.95 and GCC 2.95.1,
3123 which in turn causes GCC to fail its bootstrap comparison test.
3124 GCC 2.95.2 has a workaround.
3131 @heading @anchor{sparc-sun-sunos*}sparc-sun-sunos*
3133 A bug in the SunOS4 linker will cause it to crash when linking
3134 @option{-fPIC} compiled objects (and will therefore not allow you to build
3137 To fix this problem you can either use the most recent version of
3138 binutils or get the latest SunOS4 linker patch (patch ID 100170-10)
3139 from Sun's patch site.
3146 @heading @anchor{sparc-unknown-linux-gnulibc1}sparc-unknown-linux-gnulibc1
3148 It has been reported that you might need
3149 @uref{ftp://ftp.yggdrasil.com/private/hjl,,binutils-2.8.1.0.23}
3150 for this platform, too.
3157 @heading @anchor{sparc64-*-*}sparc64-*-*
3159 GCC version 2.95 is not able to compile code correctly for
3160 @code{sparc64} targets. Users of the Linux kernel, at least,
3161 12~can use the @code{sparc32} program to start up a new shell
3162 invocation with an environment that causes @command{configure} to
3163 recognize (via @samp{uname -a}) the system as @var{sparc-*-*} instead.
3170 @heading @anchor{#*-*-sysv*}*-*-sysv*
3171 On System V release 3, you may get this error message
3175 ld fatal: failed to write symbol name @var{something}
3176 in strings table for file @var{whatever}
3179 This probably indicates that the disk is full or your ULIMIT won't allow
3180 the file to be as large as it needs to be.
3182 This problem can also result because the kernel parameter @code{MAXUMEM}
3183 is too small. If so, you must regenerate the kernel and make the value
3184 much larger. The default value is reported to be 1024; a value of 32768
3185 is said to work. Smaller values may also work.
3187 On System V, if you get an error like this,
3190 /usr/local/lib/bison.simple: In function `yyparse':
3191 /usr/local/lib/bison.simple:625: virtual memory exhausted
3195 that too indicates a problem with disk space, ULIMIT, or @code{MAXUMEM}.
3197 On a System V release 4 system, make sure @file{/usr/bin} precedes
3198 @file{/usr/ucb} in @code{PATH}. The @code{cc} command in
3199 @file{/usr/ucb} uses libraries which have bugs.
3205 @heading @anchor{vax-dec-ultrix}vax-dec-ultrix
3206 Don't try compiling with Vax C (@code{vcc}). It produces incorrect code
3207 in some cases (for example, when @code{alloca} is used).
3209 Meanwhile, compiling @file{cp/parse.c} with pcc does not work because of
3210 an internal table size limitation in that compiler. To avoid this
3211 problem, compile just the GNU C compiler first, and use it to recompile
3212 building all the languages that you want to run.
3218 @heading @anchor{we32k-*-*}we32k-*-*
3219 These computers are also known as the 3b2, 3b5, 3b20 and other similar
3220 names. (However, the 3b1 is actually a 68000.)
3222 Don't use @option{-g} when compiling with the system's compiler. The
3223 system's linker seems to be unable to handle such a large program with
3224 debugging information.
3226 The system's compiler runs out of capacity when compiling @file{stmt.c}
3227 in GCC@. You can work around this by building @file{cpp} in GCC
3228 first, then use that instead of the system's preprocessor with the
3229 system's C compiler to compile @file{stmt.c}. Here is how:
3232 mv /lib/cpp /lib/cpp.att
3234 echo '/lib/cpp.gnu -traditional $@{1+"$@@"@}' > /lib/cpp
3238 The system's compiler produces bad code for some of the GCC
3239 optimization files. So you must build the stage 2 compiler without
3240 optimization. Then build a stage 3 compiler with optimization.
3241 That executable should work. Here are the necessary commands:
3244 make LANGUAGES=c CC=stage1/xgcc CFLAGS="-Bstage1/ -g"
3246 make CC=stage2/xgcc CFLAGS="-Bstage2/ -g -O"
3249 You may need to raise the ULIMIT setting to build a C++ compiler,
3250 as the file @file{cc1plus} is larger than one megabyte.
3256 @heading @anchor{windows}Microsoft Windows (32 bit)
3258 A port of GCC 2.95.x is included with the
3259 @uref{http://www.cygwin.com/,,Cygwin environment}.
3261 Current (as of early 2001) snapshots of GCC will build under Cygwin
3262 without modification.
3268 @heading @anchor{os2}OS/2
3270 GCC does not currently support OS/2. However, Andrew Zabolotny has been
3271 working on a generic OS/2 port with pgcc. The current code code can be found
3272 at @uref{http://www.goof.com/pcg/os2/,,http://www.goof.com/pcg/os2/}.
3274 An older copy of GCC 2.8.1 is included with the EMX tools available at
3275 @uref{ftp://ftp.leo.org/pub/comp/os/os2/leo/devtools/emx+gcc/,,
3276 ftp://ftp.leo.org/pub/comp/os/os2/leo/devtools/emx+gcc/}.
3282 @heading @anchor{older}Older systems
3284 GCC contains support files for many older (1980s and early
3285 1990s) Unix variants. For the most part, support for these systems
3286 has not been deliberately removed, but it has not been maintained for
3287 several years and may suffer from bitrot. Support from some systems
3288 has been removed from GCC 3: fx80, ns32-ns-genix, pyramid, tahoe,
3289 gmicro, spur; most of these targets had not been updated since GCC
3292 Support for older systems as targets for cross-compilation is less
3293 problematic than support for them as hosts for GCC; if an enthusiast
3294 wishes to make such a target work again (including resurrecting any
3295 of the targets that never worked with GCC 2, starting from the last
3296 CVS version before they were removed), patches
3297 @uref{../contribute.html,,following the usual requirements}
3298 would be likely to be accepted, since they should not affect the
3299 support for more modern targets.
3301 Support for old systems as hosts for GCC can cause problems if the
3302 workarounds for compiler, library and operating system bugs affect the
3303 cleanliness or maintainability of the rest of GCC@. In some cases, to
3304 bring GCC up on such a system, if still possible with current GCC, may
3305 require first installing an old version of GCC which did work on that
3306 system, and using it to compile a more recent GCC, to avoid bugs in
3307 the vendor compiler. Old releases of GCC 1 and GCC 2 are available in
3308 the old-releases directory on the
3309 @uref{../mirrors.html,,GCC mirror sites}. Header bugs may generally
3310 be avoided using @command{fixincludes}, but bugs or deficiencies in
3311 libraries and the operating system may still cause problems.
3313 For some systems, old versions of GNU binutils may also be useful,
3314 and are available from pub/binutils/old-releases on
3315 @uref{http://sources.redhat.com/mirrors.html,,sources.redhat.com mirror sites}.
3317 Some of the information on specific systems above relates to
3318 such older systems, but much of the information
3319 about GCC on such systems (which may no longer be applicable to
3320 current GCC) is to be found in the GCC texinfo manual.
3326 @heading @anchor{elf_targets}all ELF targets (SVR4, Solaris, etc.)
3328 C++ support is significantly better on ELF targets if you use the
3329 @uref{./configure.html#with-gnu-ld,,GNU linker}; duplicate copies of
3330 inlines, vtables and template instantiations will be discarded
3340 @uref{./index.html,,Return to the GCC Installation page}
3344 @c ***************************************************************************
3345 @c Part 6 The End of the Document
3347 @comment node-name, next, previous, up
3348 @node Concept Index, , Specific, Top
3352 @unnumbered Concept Index