configure.html revision 1.19 1 <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
2 <html>
3 <!-- Copyright (C) 1988-2024 Free Software Foundation, Inc.
4
5 Permission is granted to copy, distribute and/or modify this document
6 under the terms of the GNU Free Documentation License, Version 1.3 or
7 any later version published by the Free Software Foundation; with no
8 Invariant Sections, the Front-Cover texts being (a) (see below), and
9 with the Back-Cover Texts being (b) (see below). A copy of the
10 license is included in the section entitled "GNU
11 Free Documentation License".
12
13 (a) The FSF's Front-Cover Text is:
14
15 A GNU Manual
16
17 (b) The FSF's Back-Cover Text is:
18
19 You have freedom to copy and modify this GNU Manual, like GNU
20 software. Copies published by the Free Software Foundation raise
21 funds for GNU development. -->
22 <!-- Created by GNU Texinfo 6.5, http://www.gnu.org/software/texinfo/ -->
23 <head>
24 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
25 <title>Installing GCC: Configuration</title>
26
27 <meta name="description" content="Installing GCC: Configuration">
28 <meta name="keywords" content="Installing GCC: Configuration">
29 <meta name="resource-type" content="document">
30 <meta name="distribution" content="global">
31 <meta name="Generator" content="makeinfo">
32 <style type="text/css">
33 <!--
34 a.summary-letter {text-decoration: none}
35 blockquote.indentedblock {margin-right: 0em}
36 blockquote.smallindentedblock {margin-right: 0em; font-size: smaller}
37 blockquote.smallquotation {font-size: smaller}
38 div.display {margin-left: 3.2em}
39 div.example {margin-left: 3.2em}
40 div.lisp {margin-left: 3.2em}
41 div.smalldisplay {margin-left: 3.2em}
42 div.smallexample {margin-left: 3.2em}
43 div.smalllisp {margin-left: 3.2em}
44 kbd {font-style: oblique}
45 pre.display {font-family: inherit}
46 pre.format {font-family: inherit}
47 pre.menu-comment {font-family: serif}
48 pre.menu-preformatted {font-family: serif}
49 pre.smalldisplay {font-family: inherit; font-size: smaller}
50 pre.smallexample {font-size: smaller}
51 pre.smallformat {font-family: inherit; font-size: smaller}
52 pre.smalllisp {font-size: smaller}
53 span.nolinebreak {white-space: nowrap}
54 span.roman {font-family: initial; font-weight: normal}
55 span.sansserif {font-family: sans-serif; font-weight: normal}
56 ul.no-bullet {list-style: none}
57 -->
58 </style>
59
60
61 </head>
62
63 <body lang="en">
64 <h1 class="settitle" align="center">Installing GCC: Configuration</h1>
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87 <a name="index-Configuration"></a>
88 <a name="index-Installing-GCC_003a-Configuration"></a>
89
90 <p>Like most GNU software, GCC must be configured before it can be built.
91 This document describes the recommended configuration procedure
92 for both native and cross targets.
93 </p>
94 <p>We use <var>srcdir</var> to refer to the toplevel source directory for
95 GCC; we use <var>objdir</var> to refer to the toplevel build/object directory.
96 </p>
97 <p>If you obtained the sources by cloning the repository, <var>srcdir</var>
98 must refer to the top <samp>gcc</samp> directory, the one where the
99 <samp>MAINTAINERS</samp> file can be found, and not its <samp>gcc</samp>
100 subdirectory, otherwise the build will fail.
101 </p>
102 <p>If either <var>srcdir</var> or <var>objdir</var> is located on an automounted NFS
103 file system, the shell’s built-in <code>pwd</code> command will return
104 temporary pathnames. Using these can lead to various sorts of build
105 problems. To avoid this issue, set the <code>PWDCMD</code> environment
106 variable to an automounter-aware <code>pwd</code> command, e.g.,
107 <code>pawd</code> or ‘<samp>amq -w</samp>’, during the configuration and build
108 phases.
109 </p>
110 <p>First, we <strong>highly</strong> recommend that GCC be built into a
111 separate directory from the sources which does <strong>not</strong> reside
112 within the source tree. This is how we generally build GCC; building
113 where <var>objdir</var> is a subdirectory of <var>srcdir</var> should work as well;
114 building where <var>objdir</var> == <var>srcdir</var> is unsupported.
115 </p>
116 <p>If you have previously built GCC in the same directory for a
117 different target machine, do ‘<samp>make distclean</samp>’ to delete all files
118 that might be invalid. One of the files this deletes is <samp>Makefile</samp>;
119 if ‘<samp>make distclean</samp>’ complains that <samp>Makefile</samp> does not exist
120 or issues a message like “don’t know how to make distclean” it probably
121 means that the directory is already suitably clean. However, with the
122 recommended method of building in a separate <var>objdir</var>, you should
123 simply use a different <var>objdir</var> for each target.
124 </p>
125 <p>Second, when configuring a native system, either <code>cc</code> or
126 <code>gcc</code> must be in your path or you must set <code>CC</code> in
127 your environment before running configure. Otherwise the configuration
128 scripts may fail.
129 </p>
130
131 <p>To configure GCC:
132 </p>
133 <div class="smallexample">
134 <pre class="smallexample">% mkdir <var>objdir</var>
135 % cd <var>objdir</var>
136 % <var>srcdir</var>/configure [<var>options</var>] [<var>target</var>]
137 </pre></div>
138
139 <a name="Distributor-options"></a>
140 <h3 class="heading">Distributor options</h3>
141
142 <p>If you will be distributing binary versions of GCC, with modifications
143 to the source code, you should use the options described in this
144 section to make clear that your version contains modifications.
145 </p>
146 <dl compact="compact">
147 <dt><code>--with-pkgversion=<var>version</var></code></dt>
148 <dd><p>Specify a string that identifies your package. You may wish
149 to include a build number or build date. This version string will be
150 included in the output of <code>gcc --version</code>. This suffix does
151 not replace the default version string, only the ‘<samp>GCC</samp>’ part.
152 </p>
153 <p>The default value is ‘<samp>GCC</samp>’.
154 </p>
155 </dd>
156 <dt><code>--with-bugurl=<var>url</var></code></dt>
157 <dd><p>Specify the URL that users should visit if they wish to report a bug.
158 You are of course welcome to forward bugs reported to you to the FSF,
159 if you determine that they are not bugs in your modifications.
160 </p>
161 <p>The default value refers to the FSF’s GCC bug tracker.
162 </p>
163 </dd>
164 <dt><code>--with-documentation-root-url=<var>url</var></code></dt>
165 <dd><p>Specify the URL root that contains GCC option documentation. The <var>url</var>
166 should end with a <code>/</code> character.
167 </p>
168 <p>The default value is <a href="https://gcc.gnu.org/onlinedocs/">https://gcc.gnu.org/onlinedocs/</a>
169 on the GCC main development trunk. On release branches, the default
170 is <code>https://gcc.gnu.org/onlinedocs/gcc-<var>major</var>.<var>minor</var>.0/</code>.
171 </p>
172 </dd>
173 <dt><code>--with-changes-root-url=<var>url</var></code></dt>
174 <dd><p>Specify the URL root that contains information about changes in GCC
175 releases like <code>gcc-<var>version</var>/changes.html</code>.
176 The <var>url</var> should end with a <code>/</code> character.
177 </p>
178 <p>The default value is <a href="https://gcc.gnu.org/">https://gcc.gnu.org/</a>.
179 </p>
180 </dd>
181 </dl>
182
183 <a name="Host_002c-Build-and-Target-specification"></a>
184 <h3 class="heading">Host, Build and Target specification</h3>
185
186 <p>Specify the host, build and target machine configurations. You do this
187 when you run the <samp>configure</samp> script.
188 </p>
189 <p>The <em>build</em> machine is the system which you are using, the
190 <em>host</em> machine is the system where you want to run the resulting
191 compiler (normally the build machine), and the <em>target</em> machine is
192 the system for which you want the compiler to generate code.
193 </p>
194 <p>If you are building a compiler to produce code for the machine it runs
195 on (a native compiler), you normally do not need to specify any operands
196 to <samp>configure</samp>; it will try to guess the type of machine you are on
197 and use that as the build, host and target machines. So you don’t need
198 to specify a configuration when building a native compiler unless
199 <samp>configure</samp> cannot figure out what your configuration is or guesses
200 wrong.
201 </p>
202 <p>In those cases, specify the build machine’s <em>configuration name</em>
203 with the <samp>--host</samp> option; the host and target will default to be
204 the same as the host machine.
205 </p>
206 <p>Here is an example:
207 </p>
208 <div class="smallexample">
209 <pre class="smallexample">./configure --host=x86_64-pc-linux-gnu
210 </pre></div>
211
212 <p>A configuration name may be canonical or it may be more or less
213 abbreviated (<samp>config.sub</samp> script produces canonical versions).
214 </p>
215 <p>A canonical configuration name has three parts, separated by dashes.
216 It looks like this: ‘<samp><var>cpu</var>-<var>company</var>-<var>system</var></samp>’.
217 </p>
218 <p>Here are the possible CPU types:
219 </p>
220 <blockquote>
221 <p>aarch64, aarch64_be, alpha, alpha64, amdgcn, arc, arceb, arm, armeb, avr, bfin,
222 bpf, cris, csky, epiphany, fido, fr30, frv, ft32, h8300, hppa, hppa2.0,
223 hppa64, i486, i686, ia64, iq2000, lm32, loongarch64, m32c, m32r, m32rle, m68k,
224 mcore, microblaze, microblazeel, mips, mips64, mips64el, mips64octeon,
225 mips64orion, mips64vr, mipsel, mipsisa32, mipsisa32r2, mipsisa64, mipsisa64r2,
226 mipsisa64r2el, mipsisa64sb1, mipsisa64sr71k, mipstx39, mmix, mn10300, moxie,
227 msp430, nds32be, nds32le, nios2, nvptx, or1k, pdp11, powerpc, powerpc64,
228 powerpc64le, powerpcle, pru, riscv32, riscv32be, riscv64, riscv64be, rl78, rx,
229 s390, s390x, sh, shle, sparc, sparc64, tic6x, v850,
230 v850e, v850e1, vax, visium, x86_64, xstormy16, xtensa
231 </p></blockquote>
232
233 <p>Here is a list of system types:
234 </p>
235 <blockquote>
236 <p>aix<var>version</var>, amdhsa, aout, cygwin, darwin<var>version</var>,
237 eabi, eabialtivec, eabisim, eabisimaltivec, elf, elf32,
238 elfbare, elfoabi, freebsd<var>version</var>, gnu, hpux, hpux<var>version</var>,
239 kfreebsd-gnu, kopensolaris-gnu, linux-androideabi, linux-gnu,
240 linux-gnu_altivec, linux-musl, linux-uclibc, lynxos, mingw32, mingw32crt,
241 mmixware, msdosdjgpp, netbsd, netbsdelf<var>version</var>, nto-qnx, openbsd,
242 rtems, solaris<var>version</var>, symbianelf, tpf, uclinux, uclinux_eabi, vms,
243 vxworks, vxworksae, vxworksmils
244 </p></blockquote>
245
246 <a name="Options-specification"></a>
247 <h3 class="heading">Options specification</h3>
248
249 <p>Use <var>options</var> to override several configure time options for
250 GCC. A list of supported <var>options</var> follows; ‘<samp>configure
251 --help</samp>’ may list other options, but those not listed below may not
252 work and should not normally be used.
253 </p>
254 <p>Note that each <samp>--enable</samp> option has a corresponding
255 <samp>--disable</samp> option and that each <samp>--with</samp> option has a
256 corresponding <samp>--without</samp> option.
257 </p>
258 <dl compact="compact">
259 <dt><code>--prefix=<var>dirname</var></code></dt>
260 <dd><p>Specify the toplevel installation
261 directory. This is the recommended way to install the tools into a directory
262 other than the default. The toplevel installation directory defaults to
263 <samp>/usr/local</samp>.
264 </p>
265 <p>We <strong>highly</strong> recommend against <var>dirname</var> being the same or a
266 subdirectory of <var>objdir</var> or vice versa. If specifying a directory
267 beneath a user’s home directory tree, some shells will not expand
268 <var>dirname</var> correctly if it contains the ‘<samp>~</samp>’ metacharacter; use
269 <code>$HOME</code> instead.
270 </p>
271 <p>The following standard <code>autoconf</code> options are supported. Normally you
272 should not need to use these options.
273 </p><dl compact="compact">
274 <dt><code>--exec-prefix=<var>dirname</var></code></dt>
275 <dd><p>Specify the toplevel installation directory for architecture-dependent
276 files. The default is <samp><var>prefix</var></samp>.
277 </p>
278 </dd>
279 <dt><code>--bindir=<var>dirname</var></code></dt>
280 <dd><p>Specify the installation directory for the executables called by users
281 (such as <code>gcc</code> and <code>g++</code>). The default is
282 <samp><var>exec-prefix</var>/bin</samp>.
283 </p>
284 </dd>
285 <dt><code>--libdir=<var>dirname</var></code></dt>
286 <dd><p>Specify the installation directory for object code libraries and
287 internal data files of GCC. The default is <samp><var>exec-prefix</var>/lib</samp>.
288 </p>
289 </dd>
290 <dt><code>--libexecdir=<var>dirname</var></code></dt>
291 <dd><p>Specify the installation directory for internal executables of GCC.
292 The default is <samp><var>exec-prefix</var>/libexec</samp>.
293 </p>
294 </dd>
295 <dt><code>--with-slibdir=<var>dirname</var></code></dt>
296 <dd><p>Specify the installation directory for the shared libgcc library. The
297 default is <samp><var>libdir</var></samp>.
298 </p>
299 </dd>
300 <dt><code>--datarootdir=<var>dirname</var></code></dt>
301 <dd><p>Specify the root of the directory tree for read-only architecture-independent
302 data files referenced by GCC. The default is <samp><var>prefix</var>/share</samp>.
303 </p>
304 </dd>
305 <dt><code>--infodir=<var>dirname</var></code></dt>
306 <dd><p>Specify the installation directory for documentation in info format.
307 The default is <samp><var>datarootdir</var>/info</samp>.
308 </p>
309 </dd>
310 <dt><code>--datadir=<var>dirname</var></code></dt>
311 <dd><p>Specify the installation directory for some architecture-independent
312 data files referenced by GCC. The default is <samp><var>datarootdir</var></samp>.
313 </p>
314 </dd>
315 <dt><code>--docdir=<var>dirname</var></code></dt>
316 <dd><p>Specify the installation directory for documentation files (other
317 than Info) for GCC. The default is <samp><var>datarootdir</var>/doc</samp>.
318 </p>
319 </dd>
320 <dt><code>--htmldir=<var>dirname</var></code></dt>
321 <dd><p>Specify the installation directory for HTML documentation files.
322 The default is <samp><var>docdir</var></samp>.
323 </p>
324 </dd>
325 <dt><code>--pdfdir=<var>dirname</var></code></dt>
326 <dd><p>Specify the installation directory for PDF documentation files.
327 The default is <samp><var>docdir</var></samp>.
328 </p>
329 </dd>
330 <dt><code>--mandir=<var>dirname</var></code></dt>
331 <dd><p>Specify the installation directory for manual pages. The default is
332 <samp><var>datarootdir</var>/man</samp>. (Note that the manual pages are only extracts
333 from the full GCC manuals, which are provided in Texinfo format. The manpages
334 are derived by an automatic conversion process from parts of the full
335 manual.)
336 </p>
337 </dd>
338 <dt><code>--with-gxx-include-dir=<var>dirname</var></code></dt>
339 <dd><p>Specify
340 the installation directory for G++ header files. The default depends
341 on other configuration options, and differs between cross and native
342 configurations.
343 </p>
344 </dd>
345 <dt><code>--with-specs=<var>specs</var></code></dt>
346 <dd><p>Specify additional command line driver SPECS.
347 This can be useful if you need to turn on a non-standard feature by
348 default without modifying the compiler’s source code, for instance
349 <samp>--with-specs=%{!fcommon:%{!fno-common:-fno-common}}</samp>.
350 See “Spec Files” in the main manual
351 </p>
352 </dd>
353 </dl>
354
355 </dd>
356 <dt><code>--program-prefix=<var>prefix</var></code></dt>
357 <dd><p>GCC supports some transformations of the names of its programs when
358 installing them. This option prepends <var>prefix</var> to the names of
359 programs to install in <var>bindir</var> (see above). For example, specifying
360 <samp>--program-prefix=foo-</samp> would result in ‘<samp>gcc</samp>’
361 being installed as <samp>/usr/local/bin/foo-gcc</samp>.
362 </p>
363 </dd>
364 <dt><code>--program-suffix=<var>suffix</var></code></dt>
365 <dd><p>Appends <var>suffix</var> to the names of programs to install in <var>bindir</var>
366 (see above). For example, specifying <samp>--program-suffix=-3.1</samp>
367 would result in ‘<samp>gcc</samp>’ being installed as
368 <samp>/usr/local/bin/gcc-3.1</samp>.
369 </p>
370 </dd>
371 <dt><code>--program-transform-name=<var>pattern</var></code></dt>
372 <dd><p>Applies the ‘<samp>sed</samp>’ script <var>pattern</var> to be applied to the names
373 of programs to install in <var>bindir</var> (see above). <var>pattern</var> has to
374 consist of one or more basic ‘<samp>sed</samp>’ editing commands, separated by
375 semicolons. For example, if you want the ‘<samp>gcc</samp>’ program name to be
376 transformed to the installed program <samp>/usr/local/bin/myowngcc</samp> and
377 the ‘<samp>g++</samp>’ program name to be transformed to
378 <samp>/usr/local/bin/gspecial++</samp> without changing other program names,
379 you could use the pattern
380 <samp>--program-transform-name='s/^gcc$/myowngcc/; s/^g++$/gspecial++/'</samp>
381 to achieve this effect.
382 </p>
383 <p>All three options can be combined and used together, resulting in more
384 complex conversion patterns. As a basic rule, <var>prefix</var> (and
385 <var>suffix</var>) are prepended (appended) before further transformations
386 can happen with a special transformation script <var>pattern</var>.
387 </p>
388 <p>As currently implemented, this option only takes effect for native
389 builds; cross compiler binaries’ names are not transformed even when a
390 transformation is explicitly asked for by one of these options.
391 </p>
392 <p>For native builds, some of the installed programs are also installed
393 with the target alias in front of their name, as in
394 ‘<samp>i686-pc-linux-gnu-gcc</samp>’. All of the above transformations happen
395 before the target alias is prepended to the name—so, specifying
396 <samp>--program-prefix=foo-</samp> and <samp>program-suffix=-3.1</samp>, the
397 resulting binary would be installed as
398 <samp>/usr/local/bin/i686-pc-linux-gnu-foo-gcc-3.1</samp>.
399 </p>
400 <p>As a last shortcoming, none of the installed Ada programs are
401 transformed yet, which will be fixed in some time.
402 </p>
403 </dd>
404 <dt><code>--with-local-prefix=<var>dirname</var></code></dt>
405 <dd><p>Specify the
406 installation directory for local include files. The default is
407 <samp>/usr/local</samp>. Specify this option if you want the compiler to
408 search directory <samp><var>dirname</var>/include</samp> for locally installed
409 header files <em>instead</em> of <samp>/usr/local/include</samp>.
410 </p>
411 <p>You should specify <samp>--with-local-prefix</samp> <strong>only</strong> if your
412 site has a different convention (not <samp>/usr/local</samp>) for where to put
413 site-specific files.
414 </p>
415 <p>The default value for <samp>--with-local-prefix</samp> is <samp>/usr/local</samp>
416 regardless of the value of <samp>--prefix</samp>. Specifying
417 <samp>--prefix</samp> has no effect on which directory GCC searches for
418 local header files. This may seem counterintuitive, but actually it is
419 logical.
420 </p>
421 <p>The purpose of <samp>--prefix</samp> is to specify where to <em>install
422 GCC</em>. The local header files in <samp>/usr/local/include</samp>—if you put
423 any in that directory—are not part of GCC. They are part of other
424 programs—perhaps many others. (GCC installs its own header files in
425 another directory which is based on the <samp>--prefix</samp> value.)
426 </p>
427 <p>Both the local-prefix include directory and the GCC-prefix include
428 directory are part of GCC’s “system include” directories. Although these
429 two directories are not fixed, they need to be searched in the proper
430 order for the correct processing of the include_next directive. The
431 local-prefix include directory is searched before the GCC-prefix
432 include directory. Another characteristic of system include directories
433 is that pedantic warnings are turned off for headers in these directories.
434 </p>
435 <p>Some autoconf macros add <samp>-I <var>directory</var></samp> options to the
436 compiler command line, to ensure that directories containing installed
437 packages’ headers are searched. When <var>directory</var> is one of GCC’s
438 system include directories, GCC will ignore the option so that system
439 directories continue to be processed in the correct order. This
440 may result in a search order different from what was specified but the
441 directory will still be searched.
442 </p>
443 <p>GCC automatically searches for ordinary libraries using
444 <code>GCC_EXEC_PREFIX</code>. Thus, when the same installation prefix is
445 used for both GCC and packages, GCC will automatically search for
446 both headers and libraries. This provides a configuration that is
447 easy to use. GCC behaves in a manner similar to that when it is
448 installed as a system compiler in <samp>/usr</samp>.
449 </p>
450 <p>Sites that need to install multiple versions of GCC may not want to
451 use the above simple configuration. It is possible to use the
452 <samp>--program-prefix</samp>, <samp>--program-suffix</samp> and
453 <samp>--program-transform-name</samp> options to install multiple versions
454 into a single directory, but it may be simpler to use different prefixes
455 and the <samp>--with-local-prefix</samp> option to specify the location of the
456 site-specific files for each version. It will then be necessary for
457 users to specify explicitly the location of local site libraries
458 (e.g., with <code>LIBRARY_PATH</code>).
459 </p>
460 <p>The same value can be used for both <samp>--with-local-prefix</samp> and
461 <samp>--prefix</samp> provided it is not <samp>/usr</samp>. This can be used
462 to avoid the default search of <samp>/usr/local/include</samp>.
463 </p>
464 <p><strong>Do not</strong> specify <samp>/usr</samp> as the <samp>--with-local-prefix</samp>!
465 The directory you use for <samp>--with-local-prefix</samp> <strong>must not</strong>
466 contain any of the system’s standard header files. If it did contain
467 them, certain programs would be miscompiled (including GNU Emacs, on
468 certain targets), because this would override and nullify the header
469 file corrections made by the <code>fixincludes</code> script.
470 </p>
471 <p>Indications are that people who use this option use it based on mistaken
472 ideas of what it is for. People use it as if it specified where to
473 install part of GCC. Perhaps they make this assumption because
474 installing GCC creates the directory.
475 </p>
476 </dd>
477 <dt><code>--with-gcc-major-version-only</code></dt>
478 <dd><p>Specifies that GCC should use only the major number rather than
479 <var>major</var>.<var>minor</var>.<var>patchlevel</var> in filesystem paths.
480 </p>
481 </dd>
482 <dt><code>--with-native-system-header-dir=<var>dirname</var></code></dt>
483 <dd><p>Specifies that <var>dirname</var> is the directory that contains native system
484 header files, rather than <samp>/usr/include</samp>. This option is most useful
485 if you are creating a compiler that should be isolated from the system
486 as much as possible. It is most commonly used with the
487 <samp>--with-sysroot</samp> option and will cause GCC to search
488 <var>dirname</var> inside the system root specified by that option.
489 </p>
490 </dd>
491 <dt><code>--enable-shared[=<var>package</var>[,…]]</code></dt>
492 <dd><p>Build shared versions of libraries, if shared libraries are supported on
493 the target platform. Unlike GCC 2.95.x and earlier, shared libraries
494 are enabled by default on all platforms that support shared libraries.
495 </p>
496 <p>If a list of packages is given as an argument, build shared libraries
497 only for the listed packages. For other packages, only static libraries
498 will be built. Package names currently recognized in the GCC tree are
499 ‘<samp>libgcc</samp>’ (also known as ‘<samp>gcc</samp>’), ‘<samp>libstdc++</samp>’ (not
500 ‘<samp>libstdc++-v3</samp>’), ‘<samp>libffi</samp>’, ‘<samp>zlib</samp>’, ‘<samp>boehm-gc</samp>’,
501 ‘<samp>ada</samp>’, ‘<samp>libada</samp>’, ‘<samp>libgo</samp>’, ‘<samp>libobjc</samp>’, and ‘<samp>libphobos</samp>’.
502 Note ‘<samp>libiberty</samp>’ does not support shared libraries at all.
503 </p>
504 <p>Use <samp>--disable-shared</samp> to build only static libraries. Note that
505 <samp>--disable-shared</samp> does not accept a list of package names as
506 argument, only <samp>--enable-shared</samp> does.
507 </p>
508 <p>Contrast with <samp>--enable-host-shared</samp>, which affects <em>host</em>
509 code.
510 </p>
511 </dd>
512 <dt><code>--enable-host-shared</code></dt>
513 <dd><p>Specify that the <em>host</em> code should be built into position-independent
514 machine code (with <samp>-fPIC</samp>), allowing it to be used within shared
515 libraries, but yielding a slightly slower compiler.
516 </p>
517 <p>This option is required when building the libgccjit.so library.
518 </p>
519 <p>Contrast with <samp>--enable-shared</samp>, which affects <em>target</em>
520 libraries.
521 </p>
522 </dd>
523 <dt><code>--enable-host-pie</code></dt>
524 <dd><p>Specify that the <em>host</em> executables should be built into
525 position-independent executables (with <samp>-fPIE</samp> and <samp>-pie</samp>),
526 yielding a slightly slower compiler (but faster than
527 <samp>--enable-host-shared</samp>). Position-independent executables are loaded
528 at random addresses each time they are executed, therefore provide additional
529 protection against Return Oriented Programming (ROP) attacks.
530 </p>
531 <p><samp>--enable-host-pie</samp> may be used with <samp>--enable-host-shared</samp>,
532 in which case <samp>-fPIC</samp> is used when compiling, and <samp>-pie</samp> when
533 linking.
534 </p>
535 </dd>
536 <dt><code>--enable-host-bind-now</code></dt>
537 <dd><p>Specify that the <em>host</em> executables should be linked with the option
538 <samp>-Wl,-z,now</samp>, which means that the dynamic linker will resolve all
539 symbols when the executables are started, and that in turn allows RELRO to
540 mark the GOT read-only, resulting in better security.
541 </p>
542 </dd>
543 <dt><code><a name="with-gnu-as"></a>--with-gnu-as</code></dt>
544 <dd><p>Specify that the compiler should assume that the
545 assembler it finds is the GNU assembler. However, this does not modify
546 the rules to find an assembler and will result in confusion if the
547 assembler found is not actually the GNU assembler. (Confusion may also
548 result if the compiler finds the GNU assembler but has not been
549 configured with <samp>--with-gnu-as</samp>.) If you have more than one
550 assembler installed on your system, you may want to use this option in
551 connection with <samp>--with-as=<var>pathname</var></samp> or
552 <samp>--with-build-time-tools=<var>pathname</var></samp>.
553 </p>
554 <p>The following systems are the only ones where it makes a difference
555 whether you use the GNU assembler. On any other system,
556 <samp>--with-gnu-as</samp> has no effect.
557 </p>
558 <ul>
559 <li> ‘<samp>hppa1.0-<var>any</var>-<var>any</var></samp>’
560 </li><li> ‘<samp>hppa1.1-<var>any</var>-<var>any</var></samp>’
561 </li><li> ‘<samp>*-*-solaris2.11</samp>’
562 </li></ul>
563
564 </dd>
565 <dt><code><a name="with-as"></a>--with-as=<var>pathname</var></code></dt>
566 <dd><p>Specify that the compiler should use the assembler pointed to by
567 <var>pathname</var>, rather than the one found by the standard rules to find
568 an assembler, which are:
569 </p><ul>
570 <li> Unless GCC is being built with a cross compiler, check the
571 <samp><var>libexec</var>/gcc/<var>target</var>/<var>version</var></samp> directory.
572 <var>libexec</var> defaults to <samp><var>exec-prefix</var>/libexec</samp>;
573 <var>exec-prefix</var> defaults to <var>prefix</var>, which
574 defaults to <samp>/usr/local</samp> unless overridden by the
575 <samp>--prefix=<var>pathname</var></samp> switch described above. <var>target</var>
576 is the target system triple, such as ‘<samp>sparc-sun-solaris2.11</samp>’, and
577 <var>version</var> denotes the GCC version, such as 3.0.
578
579 </li><li> If the target system is the same that you are building on, check
580 operating system specific directories.
581
582 </li><li> Check in the <code>PATH</code> for a tool whose name is prefixed by the
583 target system triple.
584
585 </li><li> Check in the <code>PATH</code> for a tool whose name is not prefixed by the
586 target system triple, if the host and target system triple are
587 the same (in other words, we use a host tool if it can be used for
588 the target as well).
589 </li></ul>
590
591 <p>You may want to use <samp>--with-as</samp> if no assembler
592 is installed in the directories listed above, or if you have multiple
593 assemblers installed and want to choose one that is not found by the
594 above rules.
595 </p>
596 </dd>
597 <dt><code><a name="with-gnu-ld"></a>--with-gnu-ld</code></dt>
598 <dd><p>Same as <a href="#with-gnu-as"><samp>--with-gnu-as</samp></a>
599 but for the linker.
600 </p>
601 </dd>
602 <dt><code>--with-ld=<var>pathname</var></code></dt>
603 <dd><p>Same as <a href="#with-as"><samp>--with-as</samp></a>
604 but for the linker.
605 </p>
606 </dd>
607 <dt><code>--with-dsymutil=<var>pathname</var></code></dt>
608 <dd><p>Same as <a href="#with-as"><samp>--with-as</samp></a>
609 but for the debug linker (only used on Darwin platforms so far).
610 </p>
611 </dd>
612 <dt><code>--with-tls=<var>dialect</var></code></dt>
613 <dd><p>Specify the default TLS dialect, for systems were there is a choice.
614 For ARM targets, possible values for <var>dialect</var> are <code>gnu</code> or
615 <code>gnu2</code>, which select between the original GNU dialect and the GNU TLS
616 descriptor-based dialect.
617 For RISC-V targets, possible values for <var>dialect</var> are <code>trad</code> or
618 <code>desc</code>, which select between the traditional GNU dialect and the GNU TLS
619 descriptor-based dialect.
620 </p>
621 </dd>
622 <dt><code>--enable-multiarch</code></dt>
623 <dd><p>Specify whether to enable or disable multiarch support. The default is
624 to check for glibc start files in a multiarch location, and enable it
625 if the files are found. The auto detection is enabled for native builds,
626 and for cross builds configured with <samp>--with-sysroot</samp>, and without
627 <samp>--with-native-system-header-dir</samp>.
628 More documentation about multiarch can be found at
629 <a href="https://wiki.debian.org/Multiarch">https://wiki.debian.org/Multiarch</a>.
630 </p>
631 </dd>
632 <dt><code>--enable-sjlj-exceptions</code></dt>
633 <dd><p>Force use of the <code>setjmp</code>/<code>longjmp</code>-based scheme for exceptions.
634 ‘<samp>configure</samp>’ ordinarily picks the correct value based on the platform.
635 Only use this option if you are sure you need a different setting.
636 </p>
637 </dd>
638 <dt><code>--enable-vtable-verify</code></dt>
639 <dd><p>Specify whether to enable or disable the vtable verification feature.
640 Enabling this feature causes libstdc++ to be built with its virtual calls
641 in verifiable mode. This means that, when linked with libvtv, every
642 virtual call in libstdc++ will verify the vtable pointer through which the
643 call will be made before actually making the call. If not linked with libvtv,
644 the verifier will call stub functions (in libstdc++ itself) and do nothing.
645 If vtable verification is disabled, then libstdc++ is not built with its
646 virtual calls in verifiable mode at all. However the libvtv library will
647 still be built (see <samp>--disable-libvtv</samp> to turn off building libvtv).
648 <samp>--disable-vtable-verify</samp> is the default.
649 </p>
650 </dd>
651 <dt><code>--disable-gcov</code></dt>
652 <dd><p>Specify that the run-time library used for coverage analysis
653 and associated host tools should not be built.
654 </p>
655 </dd>
656 <dt><code>--disable-multilib</code></dt>
657 <dd><p>Specify that multiple target
658 libraries to support different target variants, calling
659 conventions, etc. should not be built. The default is to build a
660 predefined set of them.
661 </p>
662 <p>Some targets provide finer-grained control over which multilibs are built
663 (e.g., <samp>--disable-softfloat</samp>):
664 </p><dl compact="compact">
665 <dt><code>arm-*-*</code></dt>
666 <dd><p>fpu, 26bit, underscore, interwork, biendian, nofmult.
667 </p>
668 </dd>
669 <dt><code>m68*-*-*</code></dt>
670 <dd><p>softfloat, m68881, m68000, m68020.
671 </p>
672 </dd>
673 <dt><code>mips*-*-*</code></dt>
674 <dd><p>single-float, biendian, softfloat.
675 </p>
676 </dd>
677 <dt><code>msp430-*-*</code></dt>
678 <dd><p>no-exceptions
679 </p>
680 </dd>
681 <dt><code>powerpc*-*-*, rs6000*-*-*</code></dt>
682 <dd><p>aix64, pthread, softfloat, powercpu, powerpccpu, powerpcos, biendian,
683 sysv, aix.
684 </p>
685 </dd>
686 </dl>
687
688 </dd>
689 <dt><code>--with-multilib-list=<var>list</var></code></dt>
690 <dt><code>--without-multilib-list</code></dt>
691 <dd><p>Specify what multilibs to build. <var>list</var> is a comma separated list of
692 values, possibly consisting of a single value. Currently only implemented
693 for aarch64*-*-*, amdgcn*-*-*, arm*-*-*, loongarch*-*-*, riscv*-*-*, sh*-*-*
694 and x86-64-*-linux*. The accepted values and meaning for each target is given
695 below.
696 </p>
697 <dl compact="compact">
698 <dt><code>aarch64*-*-*</code></dt>
699 <dd><p><var>list</var> is a comma separated list of <code>ilp32</code>, and <code>lp64</code>
700 to enable ILP32 and LP64 run-time libraries, respectively. If
701 <var>list</var> is empty, then there will be no multilibs and only the
702 default run-time library will be built. If <var>list</var> is
703 <code>default</code> or –with-multilib-list= is not specified, then the
704 default set of libraries is selected based on the value of
705 <samp>--target</samp>.
706 </p>
707 </dd>
708 <dt><code>amdgcn*-*-*</code></dt>
709 <dd><p><var>list</var> is a comma separated list of ISA names (allowed values: <code>fiji</code>,
710 <code>gfx900</code>, <code>gfx906</code>, <code>gfx908</code>, <code>gfx90a</code>, <code>gfx90c</code>,
711 <code>gfx1030</code>, <code>gfx1036</code>, <code>gfx1100</code>, <code>gfx1103</code>).
712 It ought not include the name of the default
713 ISA, specified via <samp>--with-arch</samp>. If <var>list</var> is empty, then there
714 will be no multilibs and only the default run-time library will be built. If
715 <var>list</var> is <code>default</code> or <samp>--with-multilib-list=</samp> is not
716 specified, then the default set of libraries is selected.
717 </p>
718 </dd>
719 <dt><code>arm*-*-*</code></dt>
720 <dd><p><var>list</var> is a comma separated list of <code>aprofile</code> and
721 <code>rmprofile</code> to build multilibs for A or R and M architecture
722 profiles respectively. Note that, due to some limitation of the current
723 multilib framework, using the combined <code>aprofile,rmprofile</code>
724 multilibs selects in some cases a less optimal multilib than when using
725 the multilib profile for the architecture targetted. The special value
726 <code>default</code> is also accepted and is equivalent to omitting the
727 option, i.e., only the default run-time library will be enabled.
728 </p>
729 <p><var>list</var> may instead contain <code>@name</code>, to use the multilib
730 configuration Makefile fragment <samp>name</samp> in <samp>gcc/config/arm</samp> in
731 the source tree (it is part of the corresponding sources, after all).
732 It is recommended, but not required, that files used for this purpose to
733 be named starting with <samp>t-ml-</samp>, to make their intended purpose
734 self-evident, in line with GCC conventions. Such files enable custom,
735 user-chosen multilib lists to be configured. Whether multiple such
736 files can be used together depends on the contents of the supplied
737 files. See <samp>gcc/config/arm/t-multilib</samp> and its supplementary
738 <samp>gcc/config/arm/t-*profile</samp> files for an example of what such
739 Makefile fragments might look like for this version of GCC. The macros
740 expected to be defined in these fragments are not stable across GCC
741 releases, so make sure they define the <code>MULTILIB</code>-related macros
742 expected by the version of GCC you are building.
743 See “Target Makefile Fragments” in the internals manual.
744 </p>
745 <p>The table below gives the combination of ISAs, architectures, FPUs and
746 floating-point ABIs for which multilibs are built for each predefined
747 profile. The union of these options is considered when specifying both
748 <code>aprofile</code> and <code>rmprofile</code>.
749 </p>
750 <table>
751 <tr><td width="15%">Option</td><td width="28%">aprofile</td><td width="30%">rmprofile</td></tr>
752 <tr><td width="15%">ISAs</td><td width="28%"><code>-marm</code> and <code>-mthumb</code></td><td width="30%"><code>-mthumb</code></td></tr>
753 <tr><td width="15%">Architectures<br><br><br><br><br><br></td><td width="28%">default architecture<br>
754 <code>-march=armv7-a</code><br>
755 <code>-march=armv7ve</code><br>
756 <code>-march=armv8-a</code><br><br><br></td><td width="30%">default architecture<br>
757 <code>-march=armv6s-m</code><br>
758 <code>-march=armv7-m</code><br>
759 <code>-march=armv7e-m</code><br>
760 <code>-march=armv8-m.base</code><br>
761 <code>-march=armv8-m.main</code><br>
762 <code>-march=armv7</code></td></tr>
763 <tr><td width="15%">FPUs<br><br><br><br><br></td><td width="28%">none<br>
764 <code>-mfpu=vfpv3-d16</code><br>
765 <code>-mfpu=neon</code><br>
766 <code>-mfpu=vfpv4-d16</code><br>
767 <code>-mfpu=neon-vfpv4</code><br>
768 <code>-mfpu=neon-fp-armv8</code></td><td width="30%">none<br>
769 <code>-mfpu=vfpv3-d16</code><br>
770 <code>-mfpu=fpv4-sp-d16</code><br>
771 <code>-mfpu=fpv5-sp-d16</code><br>
772 <code>-mfpu=fpv5-d16</code><br></td></tr>
773 <tr><td width="15%">floating-point ABIs<br><br></td><td width="28%"><code>-mfloat-abi=soft</code><br>
774 <code>-mfloat-abi=softfp</code><br>
775 <code>-mfloat-abi=hard</code></td><td width="30%"><code>-mfloat-abi=soft</code><br>
776 <code>-mfloat-abi=softfp</code><br>
777 <code>-mfloat-abi=hard</code></td></tr>
778 </table>
779
780 </dd>
781 <dt><code>loongarch*-*-*</code></dt>
782 <dd><p><var>list</var> is a comma-separated list, with each of the element starting with
783 the following ABI identifiers: <code>lp64d[/base]</code> <code>lp64f[/base]</code>
784 <code>lp64d[/base]</code> (the <code>/base</code> suffix may be omitted)
785 to enable their respective run-time libraries.
786 </p>
787 <p>A suffix <code>[/<var>arch</var>][/<var>option</var>/…]</code> may follow immediately
788 after the ABI identifier to customize the compiler options for building the
789 given set of libraries. <var>arch</var> denotes the architecture name recognized
790 by the <samp>-march=<var>arch</var></samp> compiler option, which acts as a basic target
791 ISA configuration that can be adjusted using the subsequent <var>option</var>
792 suffixes, where each <var>option</var> is a compiler option without a leading dash
793 (’-’).
794 </p>
795 <p>If no such suffix is present for a given multilib variant, the
796 configured value of <samp>--with-multilib-default</samp> is appended as a default
797 suffix. If <samp>--with-multilib-default</samp> is not given, the default build
798 option <samp>-march=abi-default</samp> is applied when building the variants
799 without a suffix.
800 </p>
801 <p>As a special case, <code>fixed</code> may be used in the position of <var>arch</var>,
802 which means using the architecture configured with
803 <samp>--with-arch=<var>arch</var></samp>, or its default value (e.g. <code>loongarch64</code>
804 for <code>loongarch64-*</code> targets).
805 </p>
806 <p>If <var>list</var> is empty or <code>default</code>, or if <samp>--with-multilib-list</samp>
807 is not specified, then only the default variant of the libraries are built,
808 where the default ABI is implied by the configured target triplet.
809 </p>
810 </dd>
811 <dt><code>riscv*-*-*</code></dt>
812 <dd><p><var>list</var> is a single ABI name. The target architecture must be either
813 <code>rv32gc</code> or <code>rv64gc</code>. This will build a single multilib for the
814 specified architecture and ABI pair. If <code>--with-multilib-list</code> is not
815 given, then a default set of multilibs is selected based on the value of
816 <samp>--target</samp>. This is usually a large set of multilibs.
817 </p>
818 </dd>
819 <dt><code>sh*-*-*</code></dt>
820 <dd><p><var>list</var> is a comma separated list of CPU names. These must be of the
821 form <code>sh*</code> or <code>m*</code> (in which case they match the compiler option
822 for that processor). The list should not contain any endian options -
823 these are handled by <samp>--with-endian</samp>.
824 </p>
825 <p>If <var>list</var> is empty, then there will be no multilibs for extra
826 processors. The multilib for the secondary endian remains enabled.
827 </p>
828 <p>As a special case, if an entry in the list starts with a <code>!</code>
829 (exclamation point), then it is added to the list of excluded multilibs.
830 Entries of this sort should be compatible with ‘<samp>MULTILIB_EXCLUDES</samp>’
831 (once the leading <code>!</code> has been stripped).
832 </p>
833 <p>If <samp>--with-multilib-list</samp> is not given, then a default set of
834 multilibs is selected based on the value of <samp>--target</samp>. This is
835 usually the complete set of libraries, but some targets imply a more
836 specialized subset.
837 </p>
838 <p>Example 1: to configure a compiler for SH4A only, but supporting both
839 endians, with little endian being the default:
840 </p><div class="smallexample">
841 <pre class="smallexample">--with-cpu=sh4a --with-endian=little,big --with-multilib-list=
842 </pre></div>
843
844 <p>Example 2: to configure a compiler for both SH4A and SH4AL-DSP, but with
845 only little endian SH4AL:
846 </p><div class="smallexample">
847 <pre class="smallexample">--with-cpu=sh4a --with-endian=little,big \
848 --with-multilib-list=sh4al,!mb/m4al
849 </pre></div>
850
851 </dd>
852 <dt><code>x86-64-*-linux*</code></dt>
853 <dd><p><var>list</var> is a comma separated list of <code>m32</code>, <code>m64</code> and
854 <code>mx32</code> to enable 32-bit, 64-bit and x32 run-time libraries,
855 respectively. If <var>list</var> is empty, then there will be no multilibs
856 and only the default run-time library will be enabled.
857 </p>
858 <p>If <samp>--with-multilib-list</samp> is not given, then only 32-bit and
859 64-bit run-time libraries will be enabled.
860 </p></dd>
861 </dl>
862
863 </dd>
864 <dt><code>--with-multilib-default</code></dt>
865 <dd><p>On LoongArch targets, set the default build options for enabled multilibs
866 without build options appended to their corresponding
867 <samp>--with-multilib-list</samp> items. The format of this value is
868 <code>[/<var>arch</var>][/<var>option</var>/…]</code>, where <var>arch</var> is an
869 architecture name recognized by <samp>-march=<var>arch</var></samp> compiler option,
870 and subsequent <var>option</var> suffixes are compiler options minus a leading
871 dash (’-’).
872 </p>
873 <p>Multiple <var>option</var>s may appear consecutively while <var>arch</var> may only
874 appear in the beginning or be omitted (which means <samp>-march=abi-default</samp>
875 is applied when building the libraries).
876 </p>
877 </dd>
878 <dt><code>--with-strict-align-lib</code></dt>
879 <dd><p>On LoongArch targets, build all enabled multilibs with <samp>-mstrict-align</samp>
880 (Not enabled by default).
881 </p>
882 </dd>
883 <dt><code>--with-multilib-generator=<var>config</var></code></dt>
884 <dd><p>Specify what multilibs to build. <var>config</var> is a semicolon separated list of
885 values, possibly consisting of a single value. Currently only implemented
886 for riscv*-*-elf*. The accepted values and meanings are given below.
887 </p>
888
889 <p>Every config is constructed with four components: architecture string, ABI,
890 reuse rule with architecture string and reuse rule with sub-extension.
891 </p>
892 <p>Example 1: Add multi-lib suppport for rv32i with ilp32.
893 </p><div class="smallexample">
894 <pre class="smallexample">rv32i-ilp32--
895 </pre></div>
896
897 <p>Example 2: Add multi-lib suppport for rv32i with ilp32 and rv32imafd with ilp32.
898 </p><div class="smallexample">
899 <pre class="smallexample">rv32i-ilp32--;rv32imafd-ilp32--
900 </pre></div>
901
902 <p>Example 3: Add multi-lib suppport for rv32i with ilp32; rv32im with ilp32 and
903 rv32ic with ilp32 will reuse this multi-lib set.
904 </p><div class="smallexample">
905 <pre class="smallexample">rv32i-ilp32-rv32im-c
906 </pre></div>
907
908 <p>Example 4: Add multi-lib suppport for rv64ima with lp64; rv64imaf with lp64,
909 rv64imac with lp64 and rv64imafc with lp64 will reuse this multi-lib set.
910 </p><div class="smallexample">
911 <pre class="smallexample">rv64ima-lp64--f,c,fc
912 </pre></div>
913
914 <p><samp>--with-multilib-generator</samp> have an optional configuration argument
915 <samp>--cmodel=val</samp> for code model, this option will expand with other
916 config options, <var>val</var> is a comma separated list of possible code model,
917 currently we support medlow and medany.
918 </p>
919 <p>Example 5: Add multi-lib suppport for rv64ima with lp64; rv64ima with lp64 and
920 medlow code model
921 </p><div class="smallexample">
922 <pre class="smallexample">rv64ima-lp64--;--cmodel=medlow
923 </pre></div>
924
925 <p>Example 6: Add multi-lib suppport for rv64ima with lp64; rv64ima with lp64 and
926 medlow code model; rv64ima with lp64 and medany code model
927 </p><div class="smallexample">
928 <pre class="smallexample">rv64ima-lp64--;--cmodel=medlow,medany
929 </pre></div>
930
931 </dd>
932 <dt><code>--with-endian=<var>endians</var></code></dt>
933 <dd><p>Specify what endians to use.
934 Currently only implemented for sh*-*-*.
935 </p>
936 <p><var>endians</var> may be one of the following:
937 </p><dl compact="compact">
938 <dt><code>big</code></dt>
939 <dd><p>Use big endian exclusively.
940 </p></dd>
941 <dt><code>little</code></dt>
942 <dd><p>Use little endian exclusively.
943 </p></dd>
944 <dt><code>big,little</code></dt>
945 <dd><p>Use big endian by default. Provide a multilib for little endian.
946 </p></dd>
947 <dt><code>little,big</code></dt>
948 <dd><p>Use little endian by default. Provide a multilib for big endian.
949 </p></dd>
950 </dl>
951
952 </dd>
953 <dt><code>--enable-threads</code></dt>
954 <dd><p>Specify that the target
955 supports threads. This affects the Objective-C compiler and runtime
956 library, and exception handling for other languages like C++.
957 On some systems, this is the default.
958 </p>
959 <p>In general, the best (and, in many cases, the only known) threading
960 model available will be configured for use. Beware that on some
961 systems, GCC has not been taught what threading models are generally
962 available for the system. In this case, <samp>--enable-threads</samp> is an
963 alias for <samp>--enable-threads=single</samp>.
964 </p>
965 </dd>
966 <dt><code>--disable-threads</code></dt>
967 <dd><p>Specify that threading support should be disabled for the system.
968 This is an alias for <samp>--enable-threads=single</samp>.
969 </p>
970 </dd>
971 <dt><code>--enable-threads=<var>lib</var></code></dt>
972 <dd><p>Specify that
973 <var>lib</var> is the thread support library. This affects the Objective-C
974 compiler and runtime library, and exception handling for other languages
975 like C++. The possibilities for <var>lib</var> are:
976 </p>
977 <dl compact="compact">
978 <dt><code>aix</code></dt>
979 <dd><p>AIX thread support.
980 </p></dd>
981 <dt><code>dce</code></dt>
982 <dd><p>DCE thread support.
983 </p></dd>
984 <dt><code>lynx</code></dt>
985 <dd><p>LynxOS thread support.
986 </p></dd>
987 <dt><code>mipssde</code></dt>
988 <dd><p>MIPS SDE thread support.
989 </p></dd>
990 <dt><code>no</code></dt>
991 <dd><p>This is an alias for ‘<samp>single</samp>’.
992 </p></dd>
993 <dt><code>posix</code></dt>
994 <dd><p>Generic POSIX/Unix98 thread support.
995 </p></dd>
996 <dt><code>rtems</code></dt>
997 <dd><p>RTEMS thread support.
998 </p></dd>
999 <dt><code>single</code></dt>
1000 <dd><p>Disable thread support, should work for all platforms.
1001 </p></dd>
1002 <dt><code>tpf</code></dt>
1003 <dd><p>TPF thread support.
1004 </p></dd>
1005 <dt><code>vxworks</code></dt>
1006 <dd><p>VxWorks thread support.
1007 </p></dd>
1008 <dt><code>win32</code></dt>
1009 <dd><p>Microsoft Win32 API thread support.
1010 </p></dd>
1011 </dl>
1012
1013 </dd>
1014 <dt><code>--enable-tls</code></dt>
1015 <dd><p>Specify that the target supports TLS (Thread Local Storage). Usually
1016 configure can correctly determine if TLS is supported. In cases where
1017 it guesses incorrectly, TLS can be explicitly enabled or disabled with
1018 <samp>--enable-tls</samp> or <samp>--disable-tls</samp>. This can happen if
1019 the assembler supports TLS but the C library does not, or if the
1020 assumptions made by the configure test are incorrect.
1021 </p>
1022 </dd>
1023 <dt><code>--disable-tls</code></dt>
1024 <dd><p>Specify that the target does not support TLS.
1025 This is an alias for <samp>--enable-tls=no</samp>.
1026 </p>
1027 </dd>
1028 <dt><code>--disable-tm-clone-registry</code></dt>
1029 <dd><p>Disable TM clone registry in libgcc. It is enabled in libgcc by default.
1030 This option helps to reduce code size for embedded targets which do
1031 not use transactional memory.
1032 </p>
1033 </dd>
1034 <dt><code>--with-cpu=<var>cpu</var></code></dt>
1035 <dt><code>--with-cpu-32=<var>cpu</var></code></dt>
1036 <dt><code>--with-cpu-64=<var>cpu</var></code></dt>
1037 <dd><p>Specify which cpu variant the compiler should generate code for by default.
1038 <var>cpu</var> will be used as the default value of the <samp>-mcpu=</samp> switch.
1039 This option is only supported on some targets, including ARC, ARM, i386, M68k,
1040 PowerPC, and SPARC. It is mandatory for ARC. The <samp>--with-cpu-32</samp> and
1041 <samp>--with-cpu-64</samp> options specify separate default CPUs for
1042 32-bit and 64-bit modes; these options are only supported for aarch64, i386,
1043 x86-64, PowerPC, and SPARC.
1044 </p>
1045 </dd>
1046 <dt><code>--with-schedule=<var>cpu</var></code></dt>
1047 <dt><code>--with-arch=<var>cpu</var></code></dt>
1048 <dt><code>--with-arch-32=<var>cpu</var></code></dt>
1049 <dt><code>--with-arch-64=<var>cpu</var></code></dt>
1050 <dt><code>--with-tune=<var>cpu</var></code></dt>
1051 <dt><code>--with-tune-32=<var>cpu</var></code></dt>
1052 <dt><code>--with-tune-64=<var>cpu</var></code></dt>
1053 <dt><code>--with-abi=<var>abi</var></code></dt>
1054 <dt><code>--with-fpu=<var>type</var></code></dt>
1055 <dt><code>--with-float=<var>type</var></code></dt>
1056 <dt><code>--with-simd=<var>type</var></code></dt>
1057 <dd><p>These configure options provide default values for the <samp>-mschedule=</samp>,
1058 <samp>-march=</samp>, <samp>-mtune=</samp>, <samp>-mabi=</samp>, and <samp>-mfpu=</samp>
1059 options and for <samp>-mhard-float</samp> or <samp>-msoft-float</samp>. As with
1060 <samp>--with-cpu</samp>, which switches will be accepted and acceptable values
1061 of the arguments depend on the target.
1062 </p>
1063 </dd>
1064 <dt><code>--with-mode=<var>mode</var></code></dt>
1065 <dd><p>Specify if the compiler should default to <samp>-marm</samp> or <samp>-mthumb</samp>.
1066 This option is only supported on ARM targets.
1067 </p>
1068 </dd>
1069 <dt><code>--with-stack-offset=<var>num</var></code></dt>
1070 <dd><p>This option sets the default for the -mstack-offset=<var>num</var> option,
1071 and will thus generally also control the setting of this option for
1072 libraries. This option is only supported on Epiphany targets.
1073 </p>
1074 </dd>
1075 <dt><code>--with-fpmath=<var>isa</var></code></dt>
1076 <dd><p>This options sets <samp>-mfpmath=sse</samp> by default and specifies the default
1077 ISA for floating-point arithmetics. You can select either ‘<samp>sse</samp>’ which
1078 enables <samp>-msse2</samp> or ‘<samp>avx</samp>’ which enables <samp>-mavx</samp> by default.
1079 This option is only supported on i386 and x86-64 targets.
1080 </p>
1081 </dd>
1082 <dt><code>--with-fp-32=<var>mode</var></code></dt>
1083 <dd><p>On MIPS targets, set the default value for the <samp>-mfp</samp> option when using
1084 the o32 ABI. The possibilities for <var>mode</var> are:
1085 </p><dl compact="compact">
1086 <dt><code>32</code></dt>
1087 <dd><p>Use the o32 FP32 ABI extension, as with the <samp>-mfp32</samp> command-line
1088 option.
1089 </p></dd>
1090 <dt><code>xx</code></dt>
1091 <dd><p>Use the o32 FPXX ABI extension, as with the <samp>-mfpxx</samp> command-line
1092 option.
1093 </p></dd>
1094 <dt><code>64</code></dt>
1095 <dd><p>Use the o32 FP64 ABI extension, as with the <samp>-mfp64</samp> command-line
1096 option.
1097 </p></dd>
1098 </dl>
1099 <p>In the absence of this configuration option the default is to use the o32
1100 FP32 ABI extension.
1101 </p>
1102 </dd>
1103 <dt><code>--with-odd-spreg-32</code></dt>
1104 <dd><p>On MIPS targets, set the <samp>-modd-spreg</samp> option by default when using
1105 the o32 ABI.
1106 </p>
1107 </dd>
1108 <dt><code>--without-odd-spreg-32</code></dt>
1109 <dd><p>On MIPS targets, set the <samp>-mno-odd-spreg</samp> option by default when using
1110 the o32 ABI. This is normally used in conjunction with
1111 <samp>--with-fp-32=64</samp> in order to target the o32 FP64A ABI extension.
1112 </p>
1113 </dd>
1114 <dt><code>--with-nan=<var>encoding</var></code></dt>
1115 <dd><p>On MIPS targets, set the default encoding convention to use for the
1116 special not-a-number (NaN) IEEE 754 floating-point data. The
1117 possibilities for <var>encoding</var> are:
1118 </p><dl compact="compact">
1119 <dt><code>legacy</code></dt>
1120 <dd><p>Use the legacy encoding, as with the <samp>-mnan=legacy</samp> command-line
1121 option.
1122 </p></dd>
1123 <dt><code>2008</code></dt>
1124 <dd><p>Use the 754-2008 encoding, as with the <samp>-mnan=2008</samp> command-line
1125 option.
1126 </p></dd>
1127 </dl>
1128 <p>To use this configuration option you must have an assembler version
1129 installed that supports the <samp>-mnan=</samp> command-line option too.
1130 In the absence of this configuration option the default convention is
1131 the legacy encoding, as when neither of the <samp>-mnan=2008</samp> and
1132 <samp>-mnan=legacy</samp> command-line options has been used.
1133 </p>
1134 </dd>
1135 <dt><code>--with-divide=<var>type</var></code></dt>
1136 <dd><p>Specify how the compiler should generate code for checking for
1137 division by zero. This option is only supported on the MIPS target.
1138 The possibilities for <var>type</var> are:
1139 </p><dl compact="compact">
1140 <dt><code>traps</code></dt>
1141 <dd><p>Division by zero checks use conditional traps (this is the default on
1142 systems that support conditional traps).
1143 </p></dd>
1144 <dt><code>breaks</code></dt>
1145 <dd><p>Division by zero checks use the break instruction.
1146 </p></dd>
1147 </dl>
1148
1149 </dd>
1150 <dt><code>--with-compact-branches=<var>policy</var></code></dt>
1151 <dd><p>Specify how the compiler should generate branch instructions.
1152 This option is only supported on the MIPS target.
1153 The possibilities for <var>type</var> are:
1154 </p><dl compact="compact">
1155 <dt><code>optimal</code></dt>
1156 <dd><p>Cause a delay slot branch to be used if one is available in the
1157 current ISA and the delay slot is successfully filled. If the delay slot
1158 is not filled, a compact branch will be chosen if one is available.
1159 </p></dd>
1160 <dt><code>never</code></dt>
1161 <dd><p>Ensures that compact branch instructions will never be generated.
1162 </p></dd>
1163 <dt><code>always</code></dt>
1164 <dd><p>Ensures that a compact branch instruction will be generated if available.
1165 If a compact branch instruction is not available,
1166 a delay slot form of the branch will be used instead.
1167 This option is supported from MIPS Release 6 onwards.
1168 For pre-R6/microMIPS/MIPS16, this option is just same as never/optimal.
1169 </p></dd>
1170 </dl>
1171
1172
1173 </dd>
1174 <dt><code>--with-llsc</code></dt>
1175 <dd><p>On MIPS targets, make <samp>-mllsc</samp> the default when no
1176 <samp>-mno-llsc</samp> option is passed. This is the default for
1177 Linux-based targets, as the kernel will emulate them if the ISA does
1178 not provide them.
1179 </p>
1180 </dd>
1181 <dt><code>--without-llsc</code></dt>
1182 <dd><p>On MIPS targets, make <samp>-mno-llsc</samp> the default when no
1183 <samp>-mllsc</samp> option is passed.
1184 </p>
1185 </dd>
1186 <dt><code>--with-synci</code></dt>
1187 <dd><p>On MIPS targets, make <samp>-msynci</samp> the default when no
1188 <samp>-mno-synci</samp> option is passed.
1189 </p>
1190 </dd>
1191 <dt><code>--without-synci</code></dt>
1192 <dd><p>On MIPS targets, make <samp>-mno-synci</samp> the default when no
1193 <samp>-msynci</samp> option is passed. This is the default.
1194 </p>
1195 </dd>
1196 <dt><code>--with-lxc1-sxc1</code></dt>
1197 <dd><p>On MIPS targets, make <samp>-mlxc1-sxc1</samp> the default when no
1198 <samp>-mno-lxc1-sxc1</samp> option is passed. This is the default.
1199 </p>
1200 </dd>
1201 <dt><code>--without-lxc1-sxc1</code></dt>
1202 <dd><p>On MIPS targets, make <samp>-mno-lxc1-sxc1</samp> the default when no
1203 <samp>-mlxc1-sxc1</samp> option is passed. The indexed load/store
1204 instructions are not directly a problem but can lead to unexpected
1205 behaviour when deployed in an application intended for a 32-bit address
1206 space but run on a 64-bit processor. The issue is seen because all
1207 known MIPS 64-bit Linux kernels execute o32 and n32 applications
1208 with 64-bit addressing enabled which affects the overflow behaviour
1209 of the indexed addressing mode. GCC will assume that ordinary
1210 32-bit arithmetic overflow behaviour is the same whether performed
1211 as an <code>addu</code> instruction or as part of the address calculation
1212 in <code>lwxc1</code> type instructions. This assumption holds true in a
1213 pure 32-bit environment and can hold true in a 64-bit environment if
1214 the address space is accurately set to be 32-bit for o32 and n32.
1215 </p>
1216 </dd>
1217 <dt><code>--with-madd4</code></dt>
1218 <dd><p>On MIPS targets, make <samp>-mmadd4</samp> the default when no
1219 <samp>-mno-madd4</samp> option is passed. This is the default.
1220 </p>
1221 </dd>
1222 <dt><code>--without-madd4</code></dt>
1223 <dd><p>On MIPS targets, make <samp>-mno-madd4</samp> the default when no
1224 <samp>-mmadd4</samp> option is passed. The <code>madd4</code> instruction
1225 family can be problematic when targeting a combination of cores that
1226 implement these instructions differently. There are two known cores
1227 that implement these as fused operations instead of unfused (where
1228 unfused is normally expected). Disabling these instructions is the
1229 only way to ensure compatible code is generated; this will incur
1230 a performance penalty.
1231 </p>
1232 </dd>
1233 <dt><code>--with-msa</code></dt>
1234 <dd><p>On MIPS targets, make <samp>-mmsa</samp> the default when no
1235 <samp>-mno-msa</samp> option is passed.
1236 </p>
1237 </dd>
1238 <dt><code>--without-msa</code></dt>
1239 <dd><p>On MIPS targets, make <samp>-mno-msa</samp> the default when no
1240 <samp>-mmsa</samp> option is passed. This is the default.
1241 </p>
1242 </dd>
1243 <dt><code>--with-mips-plt</code></dt>
1244 <dd><p>On MIPS targets, make use of copy relocations and PLTs.
1245 These features are extensions to the traditional
1246 SVR4-based MIPS ABIs and require support from GNU binutils
1247 and the runtime C library.
1248 </p>
1249 </dd>
1250 <dt><code>--with-stack-clash-protection-guard-size=<var>size</var></code></dt>
1251 <dd><p>On certain targets this option sets the default stack clash protection guard
1252 size as a power of two in bytes. On AArch64 <var>size</var> is required to be either
1253 12 (4KB) or 16 (64KB).
1254 </p>
1255 </dd>
1256 <dt><code>--with-isa-spec=<var>ISA-spec-string</var></code></dt>
1257 <dd><p>On RISC-V targets specify the default version of the RISC-V Unprivileged
1258 (formerly User-Level) ISA specification to produce code conforming to.
1259 The possibilities for <var>ISA-spec-string</var> are:
1260 </p><dl compact="compact">
1261 <dt><code>2.2</code></dt>
1262 <dd><p>Produce code conforming to version 2.2.
1263 </p></dd>
1264 <dt><code>20190608</code></dt>
1265 <dd><p>Produce code conforming to version 20190608.
1266 </p></dd>
1267 <dt><code>20191213</code></dt>
1268 <dd><p>Produce code conforming to version 20191213.
1269 </p></dd>
1270 </dl>
1271 <p>In the absence of this configuration option the default version is 20191213.
1272 </p>
1273 </dd>
1274 <dt><code>--enable-__cxa_atexit</code></dt>
1275 <dd><p>Define if you want to use __cxa_atexit, rather than atexit, to
1276 register C++ destructors for local statics and global objects.
1277 This is essential for fully standards-compliant handling of
1278 destructors, but requires __cxa_atexit in libc. This option is currently
1279 only available on systems with GNU libc. When enabled, this will cause
1280 <samp>-fuse-cxa-atexit</samp> to be passed by default.
1281 </p>
1282 </dd>
1283 <dt><code>--enable-gnu-indirect-function</code></dt>
1284 <dd><p>Define if you want to enable the <code>ifunc</code> attribute. This option is
1285 currently only available on systems with GNU libc on certain targets.
1286 </p>
1287 </dd>
1288 <dt><code>--enable-target-optspace</code></dt>
1289 <dd><p>Specify that target
1290 libraries should be optimized for code space instead of code speed.
1291 This is the default for the m32r platform.
1292 </p>
1293 </dd>
1294 <dt><code>--with-cpp-install-dir=<var>dirname</var></code></dt>
1295 <dd><p>Specify that the user visible <code>cpp</code> program should be installed
1296 in <samp><var>prefix</var>/<var>dirname</var>/cpp</samp>, in addition to <var>bindir</var>.
1297 </p>
1298 </dd>
1299 <dt><code>--enable-comdat</code></dt>
1300 <dd><p>Enable COMDAT group support. This is primarily used to override the
1301 automatically detected value.
1302 </p>
1303 </dd>
1304 <dt><code>--enable-initfini-array</code></dt>
1305 <dd><p>Force the use of sections <code>.init_array</code> and <code>.fini_array</code>
1306 (instead of <code>.init</code> and <code>.fini</code>) for constructors and
1307 destructors. Option <samp>--disable-initfini-array</samp> has the
1308 opposite effect. If neither option is specified, the configure script
1309 will try to guess whether the <code>.init_array</code> and
1310 <code>.fini_array</code> sections are supported and, if they are, use them.
1311 </p>
1312 </dd>
1313 <dt><code>--enable-link-mutex</code></dt>
1314 <dd><p>When building GCC, use a mutex to avoid linking the compilers for
1315 multiple languages at the same time, to avoid thrashing on build
1316 systems with limited free memory. The default is not to use such a mutex.
1317 </p>
1318 </dd>
1319 <dt><code>--enable-link-serialization</code></dt>
1320 <dd><p>When building GCC, use make dependencies to serialize linking the compilers for
1321 multiple languages, to avoid thrashing on build
1322 systems with limited free memory. The default is not to add such
1323 dependencies and thus with parallel make potentially link different
1324 compilers concurrently. If the argument is a positive integer, allow
1325 that number of concurrent link processes for the large binaries.
1326 </p>
1327 </dd>
1328 <dt><code>--enable-maintainer-mode</code></dt>
1329 <dd><p>The build rules that regenerate the Autoconf and Automake output files as
1330 well as the GCC master message catalog <samp>gcc.pot</samp> are normally
1331 disabled. This is because it can only be rebuilt if the complete source
1332 tree is present. If you have changed the sources and want to rebuild the
1333 catalog, configuring with <samp>--enable-maintainer-mode</samp> will enable
1334 this. Note that you need a recent version of the <code>gettext</code> tools
1335 to do so.
1336 </p>
1337 </dd>
1338 <dt><code>--disable-bootstrap</code></dt>
1339 <dd><p>For a native build, the default configuration is to perform
1340 a 3-stage bootstrap of the compiler when ‘<samp>make</samp>’ is invoked,
1341 testing that GCC can compile itself correctly. If you want to disable
1342 this process, you can configure with <samp>--disable-bootstrap</samp>.
1343 </p>
1344 </dd>
1345 <dt><code>--enable-bootstrap</code></dt>
1346 <dd><p>In special cases, you may want to perform a 3-stage build
1347 even if the target and host triplets are different.
1348 This is possible when the host can run code compiled for
1349 the target (e.g. host is i686-linux, target is i486-linux).
1350 Starting from GCC 4.2, to do this you have to configure explicitly
1351 with <samp>--enable-bootstrap</samp>.
1352 </p>
1353 </dd>
1354 <dt><code>--enable-generated-files-in-srcdir</code></dt>
1355 <dd><p>Neither the .c and .h files that are generated from Bison and flex nor the
1356 info manuals and man pages that are built from the .texi files are present
1357 in the repository development tree. When building GCC from that development tree,
1358 or from one of our snapshots, those generated files are placed in your
1359 build directory, which allows for the source to be in a readonly
1360 directory.
1361 </p>
1362 <p>If you configure with <samp>--enable-generated-files-in-srcdir</samp> then those
1363 generated files will go into the source directory. This is mainly intended
1364 for generating release or prerelease tarballs of the GCC sources, since it
1365 is not a requirement that the users of source releases to have flex, Bison,
1366 or makeinfo.
1367 </p>
1368 </dd>
1369 <dt><code>--enable-version-specific-runtime-libs</code></dt>
1370 <dd><p>Specify
1371 that runtime libraries should be installed in the compiler specific
1372 subdirectory (<samp><var>libdir</var>/gcc</samp>) rather than the usual places. In
1373 addition, ‘<samp>libstdc++</samp>’’s include files will be installed into
1374 <samp><var>libdir</var></samp> unless you overruled it by using
1375 <samp>--with-gxx-include-dir=<var>dirname</var></samp>. Using this option is
1376 particularly useful if you intend to use several versions of GCC in
1377 parallel. The default is ‘<samp>yes</samp>’ for ‘<samp>libada</samp>’, and ‘<samp>no</samp>’ for
1378 the remaining libraries.
1379 </p>
1380 </dd>
1381 <dt><code>--with-darwin-extra-rpath</code></dt>
1382 <dd><p>This is provided to allow distributions to add a single additional
1383 runpath on Darwin / macOS systems. This allows for cases where the
1384 installed GCC library directories are then symlinked to a common
1385 directory outside of the GCC installation.
1386 </p>
1387 </dd>
1388 <dt><code><a name="WithAixSoname"></a>--with-aix-soname=‘<samp>aix</samp>’, ‘<samp>svr4</samp>’ or ‘<samp>both</samp>’</code></dt>
1389 <dd><p>Traditional AIX shared library versioning (versioned <code>Shared Object</code>
1390 files as members of unversioned <code>Archive Library</code> files named
1391 ‘<samp>lib.a</samp>’) causes numerous headaches for package managers. However,
1392 <code>Import Files</code> as members of <code>Archive Library</code> files allow for
1393 <strong>filename-based versioning</strong> of shared libraries as seen on Linux/SVR4,
1394 where this is called the "SONAME". But as they prevent static linking,
1395 <code>Import Files</code> may be used with <code>Runtime Linking</code> only, where the
1396 linker does search for ‘<samp>libNAME.so</samp>’ before ‘<samp>libNAME.a</samp>’ library
1397 filenames with the ‘<samp>-lNAME</samp>’ linker flag.
1398 </p>
1399 <a name="AixLdCommand"></a><p>For detailed information please refer to the AIX
1400 <a href="https://www.ibm.com/support/knowledgecenter/search/%22the%20ld%20command%2C%20also%20called%20the%20linkage%20editor%20or%20binder%22">ld
1401 Command</a> reference.
1402 </p>
1403 <p>As long as shared library creation is enabled, upon:
1404 </p><dl compact="compact">
1405 <dt><code>--with-aix-soname=aix</code></dt>
1406 <dt><code>--with-aix-soname=both</code></dt>
1407 <dd><p>A (traditional AIX) <code>Shared Archive Library</code> file is created:
1408 </p><ul>
1409 <li> using the ‘<samp>libNAME.a</samp>’ filename scheme
1410 </li><li> with the <code>Shared Object</code> file as archive member named
1411 ‘<samp>libNAME.so.V</samp>’ (except for ‘<samp>libgcc_s</samp>’, where the <code>Shared
1412 Object</code> file is named ‘<samp>shr.o</samp>’ for backwards compatibility), which
1413 <ul class="no-bullet">
1414 <li>- is used for runtime loading from inside the ‘<samp>libNAME.a</samp>’ file
1415 </li><li>- is used for dynamic loading via
1416 <code>dlopen("libNAME.a(libNAME.so.V)", RTLD_MEMBER)</code>
1417 </li><li>- is used for shared linking
1418 </li><li>- is used for static linking, so no separate <code>Static Archive
1419 Library</code> file is needed
1420 </li></ul>
1421 </li></ul>
1422 </dd>
1423 <dt><code>--with-aix-soname=both</code></dt>
1424 <dt><code>--with-aix-soname=svr4</code></dt>
1425 <dd><p>A (second) <code>Shared Archive Library</code> file is created:
1426 </p><ul>
1427 <li> using the ‘<samp>libNAME.so.V</samp>’ filename scheme
1428 </li><li> with the <code>Shared Object</code> file as archive member named
1429 ‘<samp>shr.o</samp>’, which
1430 <ul class="no-bullet">
1431 <li>- is created with the <code>-G linker flag</code>
1432 </li><li>- has the <code>F_LOADONLY</code> flag set
1433 </li><li>- is used for runtime loading from inside the ‘<samp>libNAME.so.V</samp>’ file
1434 </li><li>- is used for dynamic loading via <code>dlopen("libNAME.so.V(shr.o)",
1435 RTLD_MEMBER)</code>
1436 </li></ul>
1437 </li><li> with the <code>Import File</code> as archive member named ‘<samp>shr.imp</samp>’,
1438 which
1439 <ul class="no-bullet">
1440 <li>- refers to ‘<samp>libNAME.so.V(shr.o)</samp>’ as the "SONAME", to be recorded
1441 in the <code>Loader Section</code> of subsequent binaries
1442 </li><li>- indicates whether ‘<samp>libNAME.so.V(shr.o)</samp>’ is 32 or 64 bit
1443 </li><li>- lists all the public symbols exported by ‘<samp>lib.so.V(shr.o)</samp>’,
1444 eventually decorated with the <code>‘<samp>weak</samp>’ Keyword</code>
1445 </li><li>- is necessary for shared linking against ‘<samp>lib.so.V(shr.o)</samp>’
1446 </li></ul>
1447 </li></ul>
1448 <p>A symbolic link using the ‘<samp>libNAME.so</samp>’ filename scheme is created:
1449 </p><ul>
1450 <li> pointing to the ‘<samp>libNAME.so.V</samp>’ <code>Shared Archive Library</code> file
1451 </li><li> to permit the <code>ld Command</code> to find ‘<samp>lib.so.V(shr.imp)</samp>’ via
1452 the ‘<samp>-lNAME</samp>’ argument (requires <code>Runtime Linking</code> to be enabled)
1453 </li><li> to permit dynamic loading of ‘<samp>lib.so.V(shr.o)</samp>’ without the need
1454 to specify the version number via <code>dlopen("libNAME.so(shr.o)",
1455 RTLD_MEMBER)</code>
1456 </li></ul>
1457 </dd>
1458 </dl>
1459
1460 <p>As long as static library creation is enabled, upon:
1461 </p><dl compact="compact">
1462 <dt><code>--with-aix-soname=svr4</code></dt>
1463 <dd><p>A <code>Static Archive Library</code> is created:
1464 </p><ul>
1465 <li> using the ‘<samp>libNAME.a</samp>’ filename scheme
1466 </li><li> with all the <code>Static Object</code> files as archive members, which
1467 <ul class="no-bullet">
1468 <li>- are used for static linking
1469 </li></ul>
1470 </li></ul>
1471 </dd>
1472 </dl>
1473
1474 <p>While the aix-soname=‘<samp>svr4</samp>’ option does not create <code>Shared Object</code>
1475 files as members of unversioned <code>Archive Library</code> files any more, package
1476 managers still are responsible to
1477 <a href="./specific.html#TransferAixShobj">transfer</a> <code>Shared Object</code> files
1478 found as member of a previously installed unversioned <code>Archive Library</code>
1479 file into the newly installed <code>Archive Library</code> file with the same
1480 filename.
1481 </p>
1482 <p><em>WARNING:</em> Creating <code>Shared Object</code> files with <code>Runtime Linking</code>
1483 enabled may bloat the TOC, eventually leading to <code>TOC overflow</code> errors,
1484 requiring the use of either the <samp>-Wl,-bbigtoc</samp> linker flag (seen to
1485 break with the <code>GDB</code> debugger) or some of the TOC-related compiler flags,
1486 see “RS/6000 and PowerPC Options” in the main manual.
1487 </p>
1488 <p><samp>--with-aix-soname</samp> is currently supported by ‘<samp>libgcc_s</samp>’ only, so
1489 this option is still experimental and not for normal use yet.
1490 </p>
1491 <p>Default is the traditional behavior <samp>--with-aix-soname=‘<samp>aix</samp>’</samp>.
1492 </p>
1493 </dd>
1494 <dt><code>--enable-languages=<var>lang1</var>,<var>lang2</var>,…</code></dt>
1495 <dd><p>Specify that only a particular subset of compilers and
1496 their runtime libraries should be built. For a list of valid values for
1497 <var>langN</var> you can issue the following command in the
1498 <samp>gcc</samp> directory of your GCC source tree:<br>
1499 </p><div class="smallexample">
1500 <pre class="smallexample">grep ^language= */config-lang.in
1501 </pre></div>
1502 <p>Currently, you can use any of the following:
1503 <code>all</code>, <code>default</code>, <code>ada</code>, <code>c</code>, <code>c++</code>, <code>d</code>,
1504 <code>fortran</code>, <code>go</code>, <code>jit</code>, <code>lto</code>, <code>m2</code>,
1505 <code>objc</code>, <code>obj-c++</code>.
1506 Building the Ada compiler has special requirements, see below.
1507 If you do not pass this flag, or specify the option <code>default</code>, then the
1508 default languages available in the <samp>gcc</samp> sub-tree will be configured.
1509 Ada, D, Go, Jit, Objective-C++ and Modula-2 are not default languages.
1510 LTO is not a
1511 default language, but is built by default because <samp>--enable-lto</samp> is
1512 enabled by default. The other languages are default languages. If
1513 <code>all</code> is specified, then all available languages are built. An
1514 exception is <code>jit</code> language, which requires
1515 <samp>--enable-host-shared</samp> to be included with <code>all</code>.
1516 </p>
1517 </dd>
1518 <dt><code>--enable-stage1-languages=<var>lang1</var>,<var>lang2</var>,…</code></dt>
1519 <dd><p>Specify that a particular subset of compilers and their runtime
1520 libraries should be built with the system C compiler during stage 1 of
1521 the bootstrap process, rather than only in later stages with the
1522 bootstrapped C compiler. The list of valid values is the same as for
1523 <samp>--enable-languages</samp>, and the option <code>all</code> will select all
1524 of the languages enabled by <samp>--enable-languages</samp>. This option is
1525 primarily useful for GCC development; for instance, when a development
1526 version of the compiler cannot bootstrap due to compiler bugs, or when
1527 one is debugging front ends other than the C front end. When this
1528 option is used, one can then build the target libraries for the
1529 specified languages with the stage-1 compiler by using <code>make
1530 stage1-bubble all-target</code>, or run the testsuite on the stage-1 compiler
1531 for the specified languages using <code>make stage1-start check-gcc</code>.
1532 </p>
1533 </dd>
1534 <dt><code>--disable-libada</code></dt>
1535 <dd><p>Specify that the run-time libraries and tools used by GNAT should not
1536 be built. This can be useful for debugging, or for compatibility with
1537 previous Ada build procedures, when it was required to explicitly
1538 do a ‘<samp>make -C gcc gnatlib_and_tools</samp>’.
1539 </p>
1540 </dd>
1541 <dt><code>--disable-libgm2</code></dt>
1542 <dd><p>Specify that the run-time libraries and tools used by Modula-2 should not
1543 be built. This can be useful for debugging.
1544 </p>
1545 </dd>
1546 <dt><code>--disable-libsanitizer</code></dt>
1547 <dd><p>Specify that the run-time libraries for the various sanitizers should
1548 not be built.
1549 </p>
1550 </dd>
1551 <dt><code>--disable-libssp</code></dt>
1552 <dd><p>Specify that the run-time libraries for stack smashing protection
1553 should not be built or linked against. On many targets library support
1554 is provided by the C library instead.
1555 </p>
1556 </dd>
1557 <dt><code>--disable-libquadmath</code></dt>
1558 <dd><p>Specify that the GCC quad-precision math library should not be built.
1559 On some systems, the library is required to be linkable when building
1560 the Fortran front end, unless <samp>--disable-libquadmath-support</samp>
1561 is used.
1562 </p>
1563 </dd>
1564 <dt><code>--disable-libquadmath-support</code></dt>
1565 <dd><p>Specify that the Fortran front end and <code>libgfortran</code> do not add
1566 support for <code>libquadmath</code> on systems supporting it.
1567 </p>
1568 </dd>
1569 <dt><code>--disable-libgomp</code></dt>
1570 <dd><p>Specify that the GNU Offloading and Multi Processing Runtime Library
1571 should not be built.
1572 </p>
1573 </dd>
1574 <dt><code>--disable-libvtv</code></dt>
1575 <dd><p>Specify that the run-time libraries used by vtable verification
1576 should not be built.
1577 </p>
1578 </dd>
1579 <dt><code>--with-dwarf2</code></dt>
1580 <dd><p>Specify that the compiler should
1581 use DWARF debugging information as the default; the exact
1582 DWARF version that is the default is target-specific.
1583 </p>
1584 </dd>
1585 <dt><code>--with-advance-toolchain=<var>at</var></code></dt>
1586 <dd><p>On 64-bit PowerPC Linux systems, configure the compiler to use the
1587 header files, library files, and the dynamic linker from the Advance
1588 Toolchain release <var>at</var> instead of the default versions that are
1589 provided by the Linux distribution. In general, this option is
1590 intended for the developers of GCC, and it is not intended for general
1591 use.
1592 </p>
1593 </dd>
1594 <dt><code>--enable-targets=all</code></dt>
1595 <dt><code>--enable-targets=<var>target_list</var></code></dt>
1596 <dd><p>Some GCC targets, e.g. powerpc64-linux, build bi-arch compilers.
1597 These are compilers that are able to generate either 64-bit or 32-bit
1598 code. Typically, the corresponding 32-bit target, e.g.
1599 powerpc-linux for powerpc64-linux, only generates 32-bit code. This
1600 option enables the 32-bit target to be a bi-arch compiler, which is
1601 useful when you want a bi-arch compiler that defaults to 32-bit, and
1602 you are building a bi-arch or multi-arch binutils in a combined tree.
1603 On mips-linux, this will build a tri-arch compiler (ABI o32/n32/64),
1604 defaulted to o32.
1605 Currently, this option only affects sparc-linux, powerpc-linux, x86-linux,
1606 mips-linux and s390-linux.
1607 </p>
1608 </dd>
1609 <dt><code>--enable-default-pie</code></dt>
1610 <dd><p>Turn on <samp>-fPIE</samp> and <samp>-pie</samp> by default.
1611 </p>
1612 </dd>
1613 <dt><code>--enable-secureplt</code></dt>
1614 <dd><p>This option enables <samp>-msecure-plt</samp> by default for powerpc-linux.
1615 See “RS/6000 and PowerPC Options” in the main manual
1616 </p>
1617 </dd>
1618 <dt><code>--enable-default-ssp</code></dt>
1619 <dd><p>Turn on <samp>-fstack-protector-strong</samp> by default.
1620 </p>
1621 </dd>
1622 <dt><code>--enable-cld</code></dt>
1623 <dd><p>This option enables <samp>-mcld</samp> by default for 32-bit x86 targets.
1624 See “i386 and x86-64 Options” in the main manual
1625 </p>
1626 </dd>
1627 <dt><code>--enable-large-address-aware</code></dt>
1628 <dd><p>The <samp>--enable-large-address-aware</samp> option arranges for MinGW
1629 executables to be linked using the <samp>--large-address-aware</samp>
1630 option, that enables the use of more than 2GB of memory. If GCC is
1631 configured with this option, its effects can be reversed by passing the
1632 <samp>-Wl,--disable-large-address-aware</samp> option to the so-configured
1633 compiler driver.
1634 </p>
1635 </dd>
1636 <dt><code>--enable-win32-registry</code></dt>
1637 <dt><code>--enable-win32-registry=<var>key</var></code></dt>
1638 <dt><code>--disable-win32-registry</code></dt>
1639 <dd><p>The <samp>--enable-win32-registry</samp> option enables Microsoft Windows-hosted GCC
1640 to look up installations paths in the registry using the following key:
1641 </p>
1642 <div class="smallexample">
1643 <pre class="smallexample"><code>HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\<var>key</var></code>
1644 </pre></div>
1645
1646 <p><var>key</var> defaults to GCC version number, and can be overridden by the
1647 <samp>--enable-win32-registry=<var>key</var></samp> option. Vendors and distributors
1648 who use custom installers are encouraged to provide a different key,
1649 perhaps one comprised of vendor name and GCC version number, to
1650 avoid conflict with existing installations. This feature is enabled
1651 by default, and can be disabled by <samp>--disable-win32-registry</samp>
1652 option. This option has no effect on the other hosts.
1653 </p>
1654 </dd>
1655 <dt><code>--nfp</code></dt>
1656 <dd><p>Specify that the machine does not have a floating point unit. This
1657 option only applies to ‘<samp>m68k-sun-sunos<var>n</var></samp>’. On any other
1658 system, <samp>--nfp</samp> has no effect.
1659 </p>
1660 </dd>
1661 <dt><code>--enable-werror</code></dt>
1662 <dt><code>--disable-werror</code></dt>
1663 <dt><code>--enable-werror=yes</code></dt>
1664 <dt><code>--enable-werror=no</code></dt>
1665 <dd><p>When you specify this option, it controls whether certain files in the
1666 compiler are built with <samp>-Werror</samp> in bootstrap stage2 and later.
1667 If you don’t specify it, <samp>-Werror</samp> is turned on for the main
1668 development trunk. However it defaults to off for release branches and
1669 final releases. The specific files which get <samp>-Werror</samp> are
1670 controlled by the Makefiles.
1671 </p>
1672 </dd>
1673 <dt><code>--enable-checking</code></dt>
1674 <dt><code>--disable-checking</code></dt>
1675 <dt><code>--enable-checking=<var>list</var></code></dt>
1676 <dd><p>This option controls performing internal consistency checks in the compiler.
1677 It does not change the generated code, but adds error checking of the
1678 requested complexity. This slows down the compiler and may only work
1679 properly if you are building the compiler with GCC.
1680 </p>
1681 <p>When the option is not specified, the active set of checks depends on context.
1682 Namely, bootstrap stage 1 defaults to ‘<samp>--enable-checking=yes</samp>’, builds
1683 from release branches or release archives default to
1684 ‘<samp>--enable-checking=release</samp>’, and otherwise
1685 ‘<samp>--enable-checking=yes,extra</samp>’ is used. When the option is
1686 specified without a <var>list</var>, the result is the same as
1687 ‘<samp>--enable-checking=yes</samp>’. Likewise, ‘<samp>--disable-checking</samp>’ is
1688 equivalent to ‘<samp>--enable-checking=no</samp>’.
1689 </p>
1690 <p>The categories of checks available in <var>list</var> are ‘<samp>yes</samp>’ (most common
1691 checks ‘<samp>assert,misc,gc,gimple,rtlflag,runtime,tree,types</samp>’), ‘<samp>no</samp>’
1692 (no checks at all), ‘<samp>all</samp>’ (all but ‘<samp>valgrind</samp>’), ‘<samp>release</samp>’
1693 (cheapest checks ‘<samp>assert,runtime</samp>’) or ‘<samp>none</samp>’ (same as ‘<samp>no</samp>’).
1694 ‘<samp>release</samp>’ checks are always on and to disable them
1695 ‘<samp>--disable-checking</samp>’ or ‘<samp>--enable-checking=no[,<other checks>]</samp>’
1696 must be explicitly requested. Disabling assertions makes the compiler and
1697 runtime slightly faster but increases the risk of undetected internal errors
1698 causing wrong code to be generated.
1699 </p>
1700 <p>Individual checks can be enabled with these flags: ‘<samp>assert</samp>’, ‘<samp>df</samp>’,
1701 ‘<samp>extra</samp>’, ‘<samp>fold</samp>’, ‘<samp>gc</samp>’, ‘<samp>gcac</samp>’, ‘<samp>gimple</samp>’,
1702 ‘<samp>misc</samp>’, ‘<samp>rtl</samp>’, ‘<samp>rtlflag</samp>’, ‘<samp>runtime</samp>’, ‘<samp>tree</samp>’,
1703 ‘<samp>types</samp>’ and ‘<samp>valgrind</samp>’. ‘<samp>extra</samp>’ extends ‘<samp>misc</samp>’
1704 checking with extra checks that might affect code generation and should
1705 therefore not differ between stage1 and later stages in bootstrap.
1706 </p>
1707 <p>The ‘<samp>valgrind</samp>’ check requires the external <code>valgrind</code> simulator,
1708 available from <a href="https://valgrind.org">https://valgrind.org</a>. The ‘<samp>rtl</samp>’ checks are
1709 expensive and the ‘<samp>df</samp>’, ‘<samp>gcac</samp>’ and ‘<samp>valgrind</samp>’ checks are very
1710 expensive.
1711 </p>
1712 </dd>
1713 <dt><code>--disable-stage1-checking</code></dt>
1714 <dt><code>--enable-stage1-checking</code></dt>
1715 <dt><code>--enable-stage1-checking=<var>list</var></code></dt>
1716 <dd><p>This option affects only bootstrap build. If no <samp>--enable-checking</samp>
1717 option is specified the stage1 compiler is built with ‘<samp>yes</samp>’ checking
1718 enabled, otherwise the stage1 checking flags are the same as specified by
1719 <samp>--enable-checking</samp>. To build the stage1 compiler with
1720 different checking options use <samp>--enable-stage1-checking</samp>.
1721 The list of checking options is the same as for <samp>--enable-checking</samp>.
1722 If your system is too slow or too small to bootstrap a released compiler
1723 with checking for stage1 enabled, you can use ‘<samp>--disable-stage1-checking</samp>’
1724 to disable checking for the stage1 compiler.
1725 </p>
1726 </dd>
1727 <dt><code>--enable-coverage</code></dt>
1728 <dt><code>--enable-coverage=<var>level</var></code></dt>
1729 <dd><p>With this option, the compiler is built to collect self coverage
1730 information, every time it is run. This is for internal development
1731 purposes, and only works when the compiler is being built with gcc. The
1732 <var>level</var> argument controls whether the compiler is built optimized or
1733 not, values are ‘<samp>opt</samp>’ and ‘<samp>noopt</samp>’. For coverage analysis you
1734 want to disable optimization, for performance analysis you want to
1735 enable optimization. When coverage is enabled, the default level is
1736 without optimization.
1737 </p>
1738 </dd>
1739 <dt><code>--enable-gather-detailed-mem-stats</code></dt>
1740 <dd><p>When this option is specified more detailed information on memory
1741 allocation is gathered. This information is printed when using
1742 <samp>-fmem-report</samp>.
1743 </p>
1744 </dd>
1745 <dt><code>--enable-valgrind-annotations</code></dt>
1746 <dd><p>Mark selected memory related operations in the compiler when run under
1747 valgrind to suppress false positives.
1748 </p>
1749 </dd>
1750 <dt><code>--enable-nls</code></dt>
1751 <dt><code>--disable-nls</code></dt>
1752 <dd><p>The <samp>--enable-nls</samp> option enables Native Language Support (NLS),
1753 which lets GCC output diagnostics in languages other than American
1754 English. Native Language Support is enabled by default if not doing a
1755 canadian cross build. The <samp>--disable-nls</samp> option disables NLS.
1756 </p>
1757 <p>Note that this functionality requires either libintl (provided by GNU
1758 gettext) or C standard library that contains support for gettext (such
1759 as the GNU C Library).
1760 See <a href="#with-included-gettext">–with-included-gettext</a> for more
1761 information on the conditions required to get gettext support.
1762 </p>
1763 </dd>
1764 <dt><code>--with-libintl-prefix=<var>dir</var></code></dt>
1765 <dt><code>--without-libintl-prefix</code></dt>
1766 <dd><p>Searches for libintl in <samp><var>dir</var>/include</samp> and
1767 <samp><var>dir</var>/lib</samp>, or disables manual searching for it, letting the
1768 linker handle it.
1769 </p>
1770 </dd>
1771 <dt><code>--with-libintl-type=<var>type</var></code></dt>
1772 <dd><p>Specifies the type of library to search for when looking for libintl.
1773 <var>type</var> can be one of <code>auto</code>, <code>static</code> or <code>shared</code>.
1774 </p>
1775 <a name="with-included-gettext"></a></dd>
1776 <dt><code>--with-included-gettext</code></dt>
1777 <dd><p>Only available if <samp>gettext</samp> is present in the source tree.
1778 </p>
1779 <p>Forces the gettext tree to be configured to build and use a new static
1780 libintl, overriding the system libintl. Results in GCC being built
1781 against the newly built libintl rather than the system libintl.
1782 </p>
1783 <p>The build system makes a somewhat complicated choice when picking where
1784 to get gettext routines from. The following table is a summary of the
1785 possible options:
1786 </p>
1787 <table>
1788 <thead><tr><th width="12%">GNU gettext present in sources</th><th width="12%">libintl installed on the system</th><th width="12%"><code>gettext</code> present in libc</th><th width="20%"><code>--with-included-gettext</code></th><th width="44%">Effects on localization</th></tr></thead>
1789 <tr><td width="12%">No</td><td width="12%">No</td><td width="12%">No</td><td width="20%">(ignored)</td><td width="44%">No localization</td></tr>
1790 <tr><td width="12%">No</td><td width="12%">No</td><td width="12%">Yes</td><td width="20%">(ignored)</td><td width="44%">Localized, libc gettext</td></tr>
1791 <tr><td width="12%">No</td><td width="12%">Yes</td><td width="12%">No</td><td width="20%">(ignored)</td><td width="44%">Localized, libintl</td></tr>
1792 <tr><td width="12%">No</td><td width="12%">Yes</td><td width="12%">Yes</td><td width="20%">(ignored)</td><td width="44%">Localized, libintl</td></tr>
1793 <tr><td width="12%">Yes</td><td width="12%">No</td><td width="12%">No</td><td width="20%">No</td><td width="44%">Localized, new, static libintl</td></tr>
1794 <tr><td width="12%">Yes</td><td width="12%">No</td><td width="12%">No</td><td width="20%">Yes</td><td width="44%">Localized, new, static libintl</td></tr>
1795 <tr><td width="12%">Yes</td><td width="12%">No</td><td width="12%">Yes</td><td width="20%">No</td><td width="44%">Localized, libc gettext</td></tr>
1796 <tr><td width="12%">Yes</td><td width="12%">No</td><td width="12%">Yes</td><td width="20%">Yes</td><td width="44%">Localized, new, static libintl</td></tr>
1797 <tr><td width="12%">Yes</td><td width="12%">Yes</td><td width="12%">No</td><td width="20%">No</td><td width="44%">Localized, libintl</td></tr>
1798 <tr><td width="12%">Yes</td><td width="12%">Yes</td><td width="12%">No</td><td width="20%">Yes</td><td width="44%">Localized, new, static libintl</td></tr>
1799 <tr><td width="12%">Yes</td><td width="12%">Yes</td><td width="12%">Yes</td><td width="20%">No</td><td width="44%">Localized, libintl</td></tr>
1800 <tr><td width="12%">Yes</td><td width="12%">Yes</td><td width="12%">Yes</td><td width="20%">Yes</td><td width="44%">Localized, new, static libintl</td></tr>
1801 </table>
1802
1803 </dd>
1804 <dt><code>--with-catgets</code></dt>
1805 <dd><p>If NLS is enabled, and if the host lacks <code>gettext</code> but has the
1806 inferior <code>catgets</code> interface, the GCC build procedure normally
1807 ignores <code>catgets</code> and instead uses GCC’s copy of the GNU
1808 <code>gettext</code> library. The <samp>--with-catgets</samp> option causes the
1809 build procedure to use the host’s <code>catgets</code> in this situation.
1810 </p>
1811 </dd>
1812 <dt><code>--with-libiconv-prefix=<var>dir</var></code></dt>
1813 <dd><p>Search for libiconv header files in <samp><var>dir</var>/include</samp> and
1814 libiconv library files in <samp><var>dir</var>/lib</samp>.
1815 </p>
1816 </dd>
1817 <dt><code>--enable-obsolete</code></dt>
1818 <dd><p>Enable configuration for an obsoleted system. If you attempt to
1819 configure GCC for a system (build, host, or target) which has been
1820 obsoleted, and you do not specify this flag, configure will halt with an
1821 error message.
1822 </p>
1823 <p>All support for systems which have been obsoleted in one release of GCC
1824 is removed entirely in the next major release, unless someone steps
1825 forward to maintain the port.
1826 </p>
1827 </dd>
1828 <dt><code>--enable-decimal-float</code></dt>
1829 <dt><code>--enable-decimal-float=yes</code></dt>
1830 <dt><code>--enable-decimal-float=no</code></dt>
1831 <dt><code>--enable-decimal-float=bid</code></dt>
1832 <dt><code>--enable-decimal-float=dpd</code></dt>
1833 <dt><code>--disable-decimal-float</code></dt>
1834 <dd><p>Enable (or disable) support for the C decimal floating point extension
1835 that is in the IEEE 754-2008 standard. This is enabled by default
1836 only on AArch64, PowerPC, i386, and x86_64 GNU/Linux systems. Other
1837 systems may also support it, but require the user to specifically
1838 enable it. You can optionally control which decimal floating point
1839 format is used (either ‘<samp>bid</samp>’ or ‘<samp>dpd</samp>’). The ‘<samp>bid</samp>’
1840 (binary integer decimal) format is default on AArch64, i386 and x86_64
1841 systems, and the ‘<samp>dpd</samp>’ (densely packed decimal) format is default
1842 on PowerPC systems.
1843 </p>
1844 </dd>
1845 <dt><code>--enable-fixed-point</code></dt>
1846 <dt><code>--disable-fixed-point</code></dt>
1847 <dd><p>Enable (or disable) support for C fixed-point arithmetic.
1848 This option is enabled by default for some targets (such as MIPS) which
1849 have hardware-support for fixed-point operations. On other targets, you
1850 may enable this option manually.
1851 </p>
1852 </dd>
1853 <dt><code>--with-long-double-128</code></dt>
1854 <dd><p>Specify if <code>long double</code> type should be 128-bit by default on selected
1855 GNU/Linux architectures. If using <code>--without-long-double-128</code>,
1856 <code>long double</code> will be by default 64-bit, the same as <code>double</code> type.
1857 When neither of these configure options are used, the default will be
1858 128-bit <code>long double</code> when built against GNU C Library 2.4 and later,
1859 64-bit <code>long double</code> otherwise.
1860 </p>
1861 </dd>
1862 <dt><code>--with-long-double-format=ibm</code></dt>
1863 <dt><code>--with-long-double-format=ieee</code></dt>
1864 <dd><p>Specify whether <code>long double</code> uses the IBM extended double format
1865 or the IEEE 128-bit floating point format on PowerPC Linux systems.
1866 This configuration switch will only work on little endian PowerPC
1867 Linux systems and on big endian 64-bit systems where the default cpu
1868 is at least power7 (i.e. <samp>--with-cpu=power7</samp>,
1869 <samp>--with-cpu=power8</samp>, or <samp>--with-cpu=power9</samp> is used).
1870 </p>
1871 <p>If you use the <samp>--with-long-double-64</samp> configuration option,
1872 the <samp>--with-long-double-format=ibm</samp> and
1873 <samp>--with-long-double-format=ieee</samp> options are ignored.
1874 </p>
1875 <p>The default <code>long double</code> format is to use IBM extended double.
1876 Until all of the libraries are converted to use IEEE 128-bit floating
1877 point, it is not recommended to use
1878 <samp>--with-long-double-format=ieee</samp>.
1879 </p>
1880 </dd>
1881 <dt><code>--enable-fdpic</code></dt>
1882 <dd><p>On SH Linux systems, generate ELF FDPIC code.
1883 </p>
1884 </dd>
1885 <dt><code>--with-gmp=<var>pathname</var></code></dt>
1886 <dt><code>--with-gmp-include=<var>pathname</var></code></dt>
1887 <dt><code>--with-gmp-lib=<var>pathname</var></code></dt>
1888 <dt><code>--with-mpfr=<var>pathname</var></code></dt>
1889 <dt><code>--with-mpfr-include=<var>pathname</var></code></dt>
1890 <dt><code>--with-mpfr-lib=<var>pathname</var></code></dt>
1891 <dt><code>--with-mpc=<var>pathname</var></code></dt>
1892 <dt><code>--with-mpc-include=<var>pathname</var></code></dt>
1893 <dt><code>--with-mpc-lib=<var>pathname</var></code></dt>
1894 <dd><p>If you want to build GCC but do not have the GMP library, the MPFR
1895 library and/or the MPC library installed in a standard location and
1896 do not have their sources present in the GCC source tree then you
1897 can explicitly specify the directory where they are installed
1898 (‘<samp>--with-gmp=<var>gmpinstalldir</var></samp>’,
1899 ‘<samp>--with-mpfr=<var>mpfrinstalldir</var></samp>’,
1900 ‘<samp>--with-mpc=<var>mpcinstalldir</var></samp>’). The
1901 <samp>--with-gmp=<var>gmpinstalldir</var></samp> option is shorthand for
1902 <samp>--with-gmp-lib=<var>gmpinstalldir</var>/lib</samp> and
1903 <samp>--with-gmp-include=<var>gmpinstalldir</var>/include</samp>. Likewise the
1904 <samp>--with-mpfr=<var>mpfrinstalldir</var></samp> option is shorthand for
1905 <samp>--with-mpfr-lib=<var>mpfrinstalldir</var>/lib</samp> and
1906 <samp>--with-mpfr-include=<var>mpfrinstalldir</var>/include</samp>, also the
1907 <samp>--with-mpc=<var>mpcinstalldir</var></samp> option is shorthand for
1908 <samp>--with-mpc-lib=<var>mpcinstalldir</var>/lib</samp> and
1909 <samp>--with-mpc-include=<var>mpcinstalldir</var>/include</samp>. If these
1910 shorthand assumptions are not correct, you can use the explicit
1911 include and lib options directly. You might also need to ensure the
1912 shared libraries can be found by the dynamic linker when building and
1913 using GCC, for example by setting the runtime shared library path
1914 variable (<code>LD_LIBRARY_PATH</code> on GNU/Linux and Solaris systems).
1915 </p>
1916 <p>These flags are applicable to the host platform only. When building
1917 a cross compiler, they will not be used to configure target libraries.
1918 </p>
1919 </dd>
1920 <dt><code>--with-isl=<var>pathname</var></code></dt>
1921 <dt><code>--with-isl-include=<var>pathname</var></code></dt>
1922 <dt><code>--with-isl-lib=<var>pathname</var></code></dt>
1923 <dd><p>If you do not have the isl library installed in a standard location and you
1924 want to build GCC, you can explicitly specify the directory where it is
1925 installed (‘<samp>--with-isl=<var>islinstalldir</var></samp>’). The
1926 <samp>--with-isl=<var>islinstalldir</var></samp> option is shorthand for
1927 <samp>--with-isl-lib=<var>islinstalldir</var>/lib</samp> and
1928 <samp>--with-isl-include=<var>islinstalldir</var>/include</samp>. If this
1929 shorthand assumption is not correct, you can use the explicit
1930 include and lib options directly.
1931 </p>
1932 <p>These flags are applicable to the host platform only. When building
1933 a cross compiler, they will not be used to configure target libraries.
1934 </p>
1935 </dd>
1936 <dt><code>--with-stage1-ldflags=<var>flags</var></code></dt>
1937 <dd><p>This option may be used to set linker flags to be used when linking
1938 stage 1 of GCC. These are also used when linking GCC if configured with
1939 <samp>--disable-bootstrap</samp>. If <samp>--with-stage1-libs</samp> is not set to a
1940 value, then the default is ‘<samp>-static-libstdc++ -static-libgcc</samp>’, if
1941 supported.
1942 </p>
1943 </dd>
1944 <dt><code>--with-stage1-libs=<var>libs</var></code></dt>
1945 <dd><p>This option may be used to set libraries to be used when linking stage 1
1946 of GCC. These are also used when linking GCC if configured with
1947 <samp>--disable-bootstrap</samp>.
1948 </p>
1949 </dd>
1950 <dt><code>--with-boot-ldflags=<var>flags</var></code></dt>
1951 <dd><p>This option may be used to set linker flags to be used when linking
1952 stage 2 and later when bootstrapping GCC. If –with-boot-libs
1953 is not is set to a value, then the default is
1954 ‘<samp>-static-libstdc++ -static-libgcc</samp>’.
1955 </p>
1956 </dd>
1957 <dt><code>--with-boot-libs=<var>libs</var></code></dt>
1958 <dd><p>This option may be used to set libraries to be used when linking stage 2
1959 and later when bootstrapping GCC.
1960 </p>
1961 </dd>
1962 <dt><code>--with-debug-prefix-map=<var>map</var></code></dt>
1963 <dd><p>Convert source directory names using <samp>-fdebug-prefix-map</samp> when
1964 building runtime libraries. ‘<samp><var>map</var></samp>’ is a space-separated
1965 list of maps of the form ‘<samp><var>old</var>=<var>new</var></samp>’.
1966 </p>
1967 </dd>
1968 <dt><code>--enable-linker-build-id</code></dt>
1969 <dd><p>Tells GCC to pass <samp>--build-id</samp> option to the linker for all final
1970 links (links performed without the <samp>-r</samp> or <samp>--relocatable</samp>
1971 option), if the linker supports it. If you specify
1972 <samp>--enable-linker-build-id</samp>, but your linker does not
1973 support <samp>--build-id</samp> option, a warning is issued and the
1974 <samp>--enable-linker-build-id</samp> option is ignored. The default is off.
1975 </p>
1976 </dd>
1977 <dt><code>--with-linker-hash-style=<var>choice</var></code></dt>
1978 <dd><p>Tells GCC to pass <samp>--hash-style=<var>choice</var></samp> option to the
1979 linker for all final links. <var>choice</var> can be one of
1980 ‘<samp>sysv</samp>’, ‘<samp>gnu</samp>’, and ‘<samp>both</samp>’ where ‘<samp>sysv</samp>’ is the default.
1981 </p>
1982 </dd>
1983 <dt><code>--enable-gnu-unique-object</code></dt>
1984 <dt><code>--disable-gnu-unique-object</code></dt>
1985 <dd><p>Tells GCC to use the gnu_unique_object relocation for C++ template
1986 static data members and inline function local statics. Enabled by
1987 default for a toolchain with an assembler that accepts it and
1988 GLIBC 2.11 or above, otherwise disabled.
1989 </p>
1990 </dd>
1991 <dt><code>--with-diagnostics-color=<var>choice</var></code></dt>
1992 <dd><p>Tells GCC to use <var>choice</var> as the default for <samp>-fdiagnostics-color=</samp>
1993 option (if not used explicitly on the command line). <var>choice</var>
1994 can be one of ‘<samp>never</samp>’, ‘<samp>auto</samp>’, ‘<samp>always</samp>’, and ‘<samp>auto-if-env</samp>’
1995 where ‘<samp>auto</samp>’ is the default. ‘<samp>auto-if-env</samp>’ makes
1996 <samp>-fdiagnostics-color=auto</samp> the default if <code>GCC_COLORS</code>
1997 is present and non-empty in the environment of the compiler, and
1998 <samp>-fdiagnostics-color=never</samp> otherwise.
1999 </p>
2000 </dd>
2001 <dt><code>--with-diagnostics-urls=<var>choice</var></code></dt>
2002 <dd><p>Tells GCC to use <var>choice</var> as the default for <samp>-fdiagnostics-urls=</samp>
2003 option (if not used explicitly on the command line). <var>choice</var>
2004 can be one of ‘<samp>never</samp>’, ‘<samp>auto</samp>’, ‘<samp>always</samp>’, and ‘<samp>auto-if-env</samp>’
2005 where ‘<samp>auto</samp>’ is the default. ‘<samp>auto-if-env</samp>’ makes
2006 <samp>-fdiagnostics-urls=auto</samp> the default if <code>GCC_URLS</code>
2007 or <code>TERM_URLS</code> is present and non-empty in the environment of the
2008 compiler, and <samp>-fdiagnostics-urls=never</samp> otherwise.
2009 </p>
2010 </dd>
2011 <dt><code>--enable-lto</code></dt>
2012 <dt><code>--disable-lto</code></dt>
2013 <dd><p>Enable support for link-time optimization (LTO). This is enabled by
2014 default, and may be disabled using <samp>--disable-lto</samp>.
2015 </p>
2016 </dd>
2017 <dt><code>--enable-linker-plugin-configure-flags=FLAGS</code></dt>
2018 <dt><code>--enable-linker-plugin-flags=FLAGS</code></dt>
2019 <dd><p>By default, linker plugins (such as the LTO plugin) are built for the
2020 host system architecture. For the case that the linker has a
2021 different (but run-time compatible) architecture, these flags can be
2022 specified to build plugins that are compatible to the linker. For
2023 example, if you are building GCC for a 64-bit x86_64
2024 (‘<samp>x86_64-pc-linux-gnu</samp>’) host system, but have a 32-bit x86
2025 GNU/Linux (‘<samp>i686-pc-linux-gnu</samp>’) linker executable (which is
2026 executable on the former system), you can configure GCC as follows for
2027 getting compatible linker plugins:
2028 </p>
2029 <div class="smallexample">
2030 <pre class="smallexample">% <var>srcdir</var>/configure \
2031 --host=x86_64-pc-linux-gnu \
2032 --enable-linker-plugin-configure-flags=--host=i686-pc-linux-gnu \
2033 --enable-linker-plugin-flags='CC=gcc\ -m32\ -Wl,-rpath,[...]/i686-pc-linux-gnu/lib'
2034 </pre></div>
2035
2036 </dd>
2037 <dt><code>--with-plugin-ld=<var>pathname</var></code></dt>
2038 <dd><p>Enable an alternate linker to be used at link-time optimization (LTO)
2039 link time when <samp>-fuse-linker-plugin</samp> is enabled.
2040 This linker should have plugin support such as gold starting with
2041 version 2.20 or GNU ld starting with version 2.21.
2042 See <samp>-fuse-linker-plugin</samp> for details.
2043 </p>
2044 </dd>
2045 <dt><code>--enable-canonical-system-headers</code></dt>
2046 <dt><code>--disable-canonical-system-headers</code></dt>
2047 <dd><p>Enable system header path canonicalization for <samp>libcpp</samp>. This can
2048 produce shorter header file paths in diagnostics and dependency output
2049 files, but these changed header paths may conflict with some compilation
2050 environments. Enabled by default, and may be disabled using
2051 <samp>--disable-canonical-system-headers</samp>.
2052 </p>
2053 </dd>
2054 <dt><code>--with-glibc-version=<var>major</var>.<var>minor</var></code></dt>
2055 <dd><p>Tell GCC that when the GNU C Library (glibc) is used on the target it
2056 will be version <var>major</var>.<var>minor</var> or later. Normally this can
2057 be detected from the C library’s header files, but this option may be
2058 needed when bootstrapping a cross toolchain without the header files
2059 available for building the initial bootstrap compiler.
2060 </p>
2061 <p>If GCC is configured with some multilibs that use glibc and some that
2062 do not, this option applies only to the multilibs that use glibc.
2063 However, such configurations may not work well as not all the relevant
2064 configuration in GCC is on a per-multilib basis.
2065 </p>
2066 </dd>
2067 <dt><code>--enable-as-accelerator-for=<var>target</var></code></dt>
2068 <dd><p>Build as offload target compiler. Specify offload host triple by <var>target</var>.
2069 </p>
2070 </dd>
2071 <dt><code>--enable-offload-targets=<var>target1</var>[=<var>path1</var>],…,<var>targetN</var>[=<var>pathN</var>]</code></dt>
2072 <dd><p>Enable offloading to targets <var>target1</var>, …, <var>targetN</var>.
2073 Offload compilers are expected to be already installed. Default search
2074 path for them is <samp><var>exec-prefix</var></samp>, but it can be changed by
2075 specifying paths <var>path1</var>, …, <var>pathN</var>.
2076 </p>
2077 <div class="smallexample">
2078 <pre class="smallexample">% <var>srcdir</var>/configure \
2079 --enable-offload-targets=amdgcn-amdhsa,nvptx-none
2080 </pre></div>
2081
2082 </dd>
2083 <dt><code>--enable-offload-defaulted</code></dt>
2084 <dd>
2085 <p>Tell GCC that configured but not installed offload compilers and libgomp
2086 plugins are silently ignored. Useful for distribution compilers where
2087 those are in separate optional packages and where the presence or absence
2088 of those optional packages should determine the actual supported offloading
2089 target set rather than the GCC configure-time selection.
2090 </p>
2091 </dd>
2092 <dt><code>--enable-cet</code></dt>
2093 <dt><code>--disable-cet</code></dt>
2094 <dd><p>Enable building target run-time libraries with control-flow
2095 instrumentation, see <samp>-fcf-protection</samp> option. When
2096 <code>--enable-cet</code> is specified target libraries are configured
2097 to add <samp>-fcf-protection</samp> and, if needed, other target
2098 specific options to a set of building options.
2099 </p>
2100 <p><code>--enable-cet=auto</code> is default. CET is enabled on Linux/x86 if
2101 target binutils supports <code>Intel CET</code> instructions and disabled
2102 otherwise. In this case, the target libraries are configured to get
2103 additional <samp>-fcf-protection</samp> option.
2104 </p>
2105 </dd>
2106 <dt><code>--with-riscv-attribute=‘<samp>yes</samp>’, ‘<samp>no</samp>’ or ‘<samp>default</samp>’</code></dt>
2107 <dd><p>Generate RISC-V attribute by default, in order to record extra build
2108 information in object.
2109 </p>
2110 <p>The option is disabled by default. It is enabled on RISC-V/ELF (bare-metal)
2111 target if target binutils supported.
2112 </p>
2113 </dd>
2114 <dt><code>--enable-s390-excess-float-precision</code></dt>
2115 <dt><code>--disable-s390-excess-float-precision</code></dt>
2116 <dd><p>On s390(x) targets, enable treatment of float expressions with double precision
2117 when in standards-compliant mode (e.g., when <code>--std=c99</code> or
2118 <code>-fexcess-precision=standard</code> are given).
2119 </p>
2120 <p>For a native build and cross compiles that have target headers, the option’s
2121 default is derived from glibc’s behavior. When glibc clamps float_t to double,
2122 GCC follows and enables the option. For other cross compiles, the default is
2123 disabled.
2124 </p>
2125 </dd>
2126 <dt><code>--with-zstd=<var>pathname</var></code></dt>
2127 <dt><code>--with-zstd-include=<var>pathname</var></code></dt>
2128 <dt><code>--with-zstd-lib=<var>pathname</var></code></dt>
2129 <dd><p>If you do not have the <code>zstd</code> library installed in a standard
2130 location and you want to build GCC, you can explicitly specify the
2131 directory where it is installed (‘<samp>--with-zstd=<var>zstdinstalldir</var></samp>’).
2132 The <samp>--with-zstd=<var>zstdinstalldir</var></samp> option is shorthand for
2133 <samp>--with-zstd-lib=<var>zstdinstalldir</var>/lib</samp> and
2134 <samp>--with-zstd-include=<var>zstdinstalldir</var>/include</samp>. If this
2135 shorthand assumption is not correct, you can use the explicit
2136 include and lib options directly.
2137 </p>
2138 <p>These flags are applicable to the host platform only. When building
2139 a cross compiler, they will not be used to configure target libraries.
2140 </p></dd>
2141 </dl>
2142
2143 <a name="Cross-Compiler-Specific-Options"></a>
2144 <h4 class="subheading">Cross-Compiler-Specific Options</h4>
2145 <p>The following options only apply to building cross compilers.
2146 </p>
2147 <dl compact="compact">
2148 <dt><code>--with-toolexeclibdir=<var>dir</var></code></dt>
2149 <dd><p>Specify the installation directory for libraries built with a cross compiler.
2150 The default is <samp>${gcc_tooldir}/lib</samp>.
2151 </p>
2152 </dd>
2153 <dt><code>--with-sysroot</code></dt>
2154 <dt><code>--with-sysroot=<var>dir</var></code></dt>
2155 <dd><p>Tells GCC to consider <var>dir</var> as the root of a tree that contains
2156 (a subset of) the root filesystem of the target operating system.
2157 Target system headers, libraries and run-time object files will be
2158 searched for in there. More specifically, this acts as if
2159 <samp>--sysroot=<var>dir</var></samp> was added to the default options of the built
2160 compiler. The specified directory is not copied into the
2161 install tree, unlike the options <samp>--with-headers</samp> and
2162 <samp>--with-libs</samp> that this option obsoletes. The default value,
2163 in case <samp>--with-sysroot</samp> is not given an argument, is
2164 <samp>${gcc_tooldir}/sys-root</samp>. If the specified directory is a
2165 subdirectory of <samp>${exec_prefix}</samp>, then it will be found relative to
2166 the GCC binaries if the installation tree is moved.
2167 </p>
2168 <p>This option affects the system root for the compiler used to build
2169 target libraries (which runs on the build system) and the compiler newly
2170 installed with <code>make install</code>; it does not affect the compiler which is
2171 used to build GCC itself.
2172 </p>
2173 <p>If you specify the <samp>--with-native-system-header-dir=<var>dirname</var></samp>
2174 option then the compiler will search that directory within <var>dirname</var> for
2175 native system headers rather than the default <samp>/usr/include</samp>.
2176 </p>
2177 </dd>
2178 <dt><code>--with-build-sysroot</code></dt>
2179 <dt><code>--with-build-sysroot=<var>dir</var></code></dt>
2180 <dd><p>Tells GCC to consider <var>dir</var> as the system root (see
2181 <samp>--with-sysroot</samp>) while building target libraries, instead of
2182 the directory specified with <samp>--with-sysroot</samp>. This option is
2183 only useful when you are already using <samp>--with-sysroot</samp>. You
2184 can use <samp>--with-build-sysroot</samp> when you are configuring with
2185 <samp>--prefix</samp> set to a directory that is different from the one in
2186 which you are installing GCC and your target libraries.
2187 </p>
2188 <p>This option affects the system root for the compiler used to build
2189 target libraries (which runs on the build system); it does not affect
2190 the compiler which is used to build GCC itself.
2191 </p>
2192 <p>If you specify the <samp>--with-native-system-header-dir=<var>dirname</var></samp>
2193 option then the compiler will search that directory within <var>dirname</var> for
2194 native system headers rather than the default <samp>/usr/include</samp>.
2195 </p>
2196 </dd>
2197 <dt><code>--with-headers</code></dt>
2198 <dt><code>--with-headers=<var>dir</var></code></dt>
2199 <dd><p>Deprecated in favor of <samp>--with-sysroot</samp>.
2200 Specifies that target headers are available when building a cross compiler.
2201 The <var>dir</var> argument specifies a directory which has the target include
2202 files. These include files will be copied into the <samp>gcc</samp> install
2203 directory. <em>This option with the <var>dir</var> argument is required</em> when
2204 building a cross compiler, if <samp><var>prefix</var>/<var>target</var>/sys-include</samp>
2205 doesn’t pre-exist. If <samp><var>prefix</var>/<var>target</var>/sys-include</samp> does
2206 pre-exist, the <var>dir</var> argument may be omitted. <code>fixincludes</code>
2207 will be run on these files to make them compatible with GCC.
2208 </p>
2209 </dd>
2210 <dt><code>--without-headers</code></dt>
2211 <dd><p>Tells GCC not use any target headers from a libc when building a cross
2212 compiler. When crossing to GNU/Linux, you need the headers so GCC
2213 can build the exception handling for libgcc.
2214 </p>
2215 </dd>
2216 <dt><code>--with-libs</code></dt>
2217 <dt><code>--with-libs="<var>dir1</var> <var>dir2</var> … <var>dirN</var>"</code></dt>
2218 <dd><p>Deprecated in favor of <samp>--with-sysroot</samp>.
2219 Specifies a list of directories which contain the target runtime
2220 libraries. These libraries will be copied into the <samp>gcc</samp> install
2221 directory. If the directory list is omitted, this option has no
2222 effect.
2223 </p>
2224 </dd>
2225 <dt><code>--with-newlib</code></dt>
2226 <dd><p>Specifies that ‘<samp>newlib</samp>’ is
2227 being used as the target C library. This causes <code>__eprintf</code> to be
2228 omitted from <samp>libgcc.a</samp> on the assumption that it will be provided by
2229 ‘<samp>newlib</samp>’.
2230 </p>
2231 <a name="avr"></a>
2232 </dd>
2233 <dt><code>--with-avrlibc</code></dt>
2234 <dd><p>Only supported for the AVR target. Specifies that ‘<samp>AVR-Libc</samp>’ is
2235 being used as the target C library. This causes float support
2236 functions like <code>__addsf3</code> to be omitted from <samp>libgcc.a</samp> on
2237 the assumption that it will be provided by <samp>libm.a</samp>. For more
2238 technical details, cf. <a href="https://gcc.gnu.org/PR54461">PR54461</a>.
2239 It is not supported for
2240 RTEMS configurations, which currently use newlib. The option is
2241 supported since version 4.7.2 and is the default in 4.8.0 and newer.
2242 </p>
2243 </dd>
2244 <dt><code>--with-double={32|64|32,64|64,32}</code></dt>
2245 <dt><code>--with-long-double={32|64|32,64|64,32|double}</code></dt>
2246 <dd><p>Only supported for the AVR target since version 10.
2247 Specify the default layout available for the C/C++ ‘<samp>double</samp>’
2248 and ‘<samp>long double</samp>’ type, respectively. The following rules apply:
2249 </p><ul>
2250 <li> The first value after the ‘<samp>=</samp>’ specifies the default layout (in bits)
2251 of the type and also the default for the <samp>-mdouble=</samp> resp.
2252 <samp>-mlong-double=</samp> compiler option.
2253 </li><li> If more than one value is specified, respective multilib variants are
2254 available, and <samp>-mdouble=</samp> resp. <samp>-mlong-double=</samp> acts
2255 as a multilib option.
2256 </li><li> If <samp>--with-long-double=double</samp> is specified, ‘<samp>double</samp>’ and
2257 ‘<samp>long double</samp>’ will have the same layout.
2258 </li><li> The defaults are <samp>--with-long-double=64,32</samp> and
2259 <samp>--with-double=32,64</samp>. The default ‘<samp>double</samp>’ layout imposed by
2260 the latter is compatible with older versions of the compiler that implement
2261 ‘<samp>double</samp>’ as a 32-bit type, which does not comply to the language standard.
2262 </li></ul>
2263 <p>Not all combinations of <samp>--with-double=</samp> and
2264 <samp>--with-long-double=</samp> are valid. For example, the combination
2265 <samp>--with-double=32,64</samp> <samp>--with-long-double=32</samp> will be
2266 rejected because the first option specifies the availability of
2267 multilibs for ‘<samp>double</samp>’, whereas the second option implies
2268 that ‘<samp>long double</samp>’ — and hence also ‘<samp>double</samp>’ — is always
2269 32 bits wide.
2270 </p>
2271 </dd>
2272 <dt><code>--with-double-comparison={tristate|bool|libf7}</code></dt>
2273 <dd><p>Only supported for the AVR target since version 10.
2274 Specify what result format is returned by library functions that
2275 compare 64-bit floating point values (<code>DFmode</code>).
2276 The GCC default is ‘<samp>tristate</samp>’. If the floating point
2277 implementation returns a boolean instead, set it to ‘<samp>bool</samp>’.
2278 </p>
2279 </dd>
2280 <dt><code>--with-libf7={libgcc|math|math-symbols|no}</code></dt>
2281 <dd><p>Only supported for the AVR target since version 10.
2282 Specify to which degree code from LibF7 is included in libgcc.
2283 LibF7 is an ad-hoc, AVR-specific, 64-bit floating point emulation
2284 written in C and (inline) assembly. ‘<samp>libgcc</samp>’ adds support
2285 for functions that one would usually expect in libgcc like double addition,
2286 double comparisons and double conversions. ‘<samp>math</samp>’ also adds routines
2287 that one would expect in <samp>libm.a</samp>, but with <code>__</code> (two underscores)
2288 prepended to the symbol names as specified by <samp>math.h</samp>.
2289 ‘<samp>math-symbols</samp>’ also defines weak aliases for the functions
2290 declared in <samp>math.h</samp>. However, <code>--with-libf7</code> won’t
2291 install no <samp>math.h</samp> header file whatsoever, this file must come
2292 from elsewhere. This option sets <samp>--with-double-comparison</samp>
2293 to ‘<samp>bool</samp>’.
2294 </p>
2295 </dd>
2296 <dt><code>--with-nds32-lib=<var>library</var></code></dt>
2297 <dd><p>Specifies that <var>library</var> setting is used for building <samp>libgcc.a</samp>.
2298 Currently, the valid <var>library</var> is ‘<samp>newlib</samp>’ or ‘<samp>mculib</samp>’.
2299 This option is only supported for the NDS32 target.
2300 </p>
2301 </dd>
2302 <dt><code>--with-build-time-tools=<var>dir</var></code></dt>
2303 <dd><p>Specifies where to find the set of target tools (assembler, linker, etc.)
2304 that will be used while building GCC itself. This option can be useful
2305 if the directory layouts are different between the system you are building
2306 GCC on, and the system where you will deploy it.
2307 </p>
2308 <p>For example, on an ‘<samp>ia64-hp-hpux</samp>’ system, you may have the GNU
2309 assembler and linker in <samp>/usr/bin</samp>, and the native tools in a
2310 different path, and build a toolchain that expects to find the
2311 native tools in <samp>/usr/bin</samp>.
2312 </p>
2313 <p>When you use this option, you should ensure that <var>dir</var> includes
2314 <code>ar</code>, <code>as</code>, <code>ld</code>, <code>nm</code>,
2315 <code>ranlib</code> and <code>strip</code> if necessary, and possibly
2316 <code>objdump</code>. Otherwise, GCC may use an inconsistent set of
2317 tools.
2318 </p></dd>
2319 </dl>
2320
2321 <a name="Overriding-configure-test-results"></a>
2322 <h4 class="subsubheading">Overriding <code>configure</code> test results</h4>
2323
2324 <p>Sometimes, it might be necessary to override the result of some
2325 <code>configure</code> test, for example in order to ease porting to a new
2326 system or work around a bug in a test. The toplevel <code>configure</code>
2327 script provides three variables for this:
2328 </p>
2329 <dl compact="compact">
2330 <dd>
2331 <a name="index-build_005fconfigargs"></a>
2332 </dd>
2333 <dt><code>build_configargs</code></dt>
2334 <dd><p>The contents of this variable is passed to all build <code>configure</code>
2335 scripts.
2336 </p>
2337 <a name="index-host_005fconfigargs"></a>
2338 </dd>
2339 <dt><code>host_configargs</code></dt>
2340 <dd><p>The contents of this variable is passed to all host <code>configure</code>
2341 scripts.
2342 </p>
2343 <a name="index-target_005fconfigargs"></a>
2344 </dd>
2345 <dt><code>target_configargs</code></dt>
2346 <dd><p>The contents of this variable is passed to all target <code>configure</code>
2347 scripts.
2348 </p>
2349 </dd>
2350 </dl>
2351
2352 <p>In order to avoid shell and <code>make</code> quoting issues for complex
2353 overrides, you can pass a setting for <code>CONFIG_SITE</code> and set
2354 variables in the site file.
2355 </p>
2356 <a name="Objective-C-Specific-Options"></a>
2357 <h4 class="subheading">Objective-C-Specific Options</h4>
2358
2359 <p>The following options apply to the build of the Objective-C runtime library.
2360 </p>
2361 <dl compact="compact">
2362 <dt><code>--enable-objc-gc</code></dt>
2363 <dd><p>Specify that an additional variant of the GNU Objective-C runtime library
2364 is built, using an external build of the Boehm-Demers-Weiser garbage
2365 collector (<a href="https://www.hboehm.info/gc/">https://www.hboehm.info/gc/</a>). This library needs to be
2366 available for each multilib variant, unless configured with
2367 <samp>--enable-objc-gc=‘<samp>auto</samp>’</samp> in which case the build of the
2368 additional runtime library is skipped when not available and the build
2369 continues.
2370 </p>
2371 </dd>
2372 <dt><code>--with-target-bdw-gc=<var>list</var></code></dt>
2373 <dt><code>--with-target-bdw-gc-include=<var>list</var></code></dt>
2374 <dt><code>--with-target-bdw-gc-lib=<var>list</var></code></dt>
2375 <dd><p>Specify search directories for the garbage collector header files and
2376 libraries. <var>list</var> is a comma separated list of key value pairs of the
2377 form ‘<samp><var>multilibdir</var>=<var>path</var></samp>’, where the default multilib key
2378 is named as ‘<samp>.</samp>’ (dot), or is omitted (e.g.
2379 ‘<samp>--with-target-bdw-gc=/opt/bdw-gc,32=/opt-bdw-gc32</samp>’).
2380 </p>
2381 <p>The options <samp>--with-target-bdw-gc-include</samp> and
2382 <samp>--with-target-bdw-gc-lib</samp> must always be specified together
2383 for each multilib variant and they take precedence over
2384 <samp>--with-target-bdw-gc</samp>. If <samp>--with-target-bdw-gc-include</samp>
2385 is missing values for a multilib, then the value for the default
2386 multilib is used (e.g. ‘<samp>--with-target-bdw-gc-include=/opt/bdw-gc/include</samp>’
2387 ‘<samp>--with-target-bdw-gc-lib=/opt/bdw-gc/lib64,32=/opt-bdw-gc/lib32</samp>’).
2388 If none of these options are specified, the library is assumed in
2389 default locations.
2390 </p></dd>
2391 </dl>
2392
2393 <a name="D-Specific-Options"></a>
2394 <h4 class="subheading">D-Specific Options</h4>
2395
2396 <p>The following options apply to the build of the D runtime library.
2397 </p>
2398 <dl compact="compact">
2399 <dt><code>--enable-libphobos-checking</code></dt>
2400 <dt><code>--disable-libphobos-checking</code></dt>
2401 <dt><code>--enable-libphobos-checking=<var>list</var></code></dt>
2402 <dd><p>This option controls whether run-time checks and contracts are compiled into
2403 the D runtime library. When the option is not specified, the library is built
2404 with ‘<samp>release</samp>’ checking. When the option is specified without a
2405 <var>list</var>, the result is the same as ‘<samp>--enable-libphobos-checking=yes</samp>’.
2406 Likewise, ‘<samp>--disable-libphobos-checking</samp>’ is equivalent to
2407 ‘<samp>--enable-libphobos-checking=no</samp>’.
2408 </p>
2409 <p>The categories of checks available in <var>list</var> are ‘<samp>yes</samp>’ (compiles
2410 libphobos with <samp>-fno-release</samp>), ‘<samp>no</samp>’ (compiles libphobos with
2411 <samp>-frelease</samp>), ‘<samp>all</samp>’ (same as ‘<samp>yes</samp>’), ‘<samp>none</samp>’ or
2412 ‘<samp>release</samp>’ (same as ‘<samp>no</samp>’).
2413 </p>
2414 <p>Individual checks available in <var>list</var> are ‘<samp>assert</samp>’ (compiles libphobos
2415 with an extra option <samp>-fassert</samp>).
2416 </p>
2417 </dd>
2418 <dt><code>--with-libphobos-druntime-only</code></dt>
2419 <dt><code>--with-libphobos-druntime-only=<var>choice</var></code></dt>
2420 <dd><p>Specify whether to build only the core D runtime library (druntime), or both
2421 the core and standard library (phobos) into libphobos. This is useful for
2422 targets that have full support in druntime, but no or incomplete support
2423 in phobos. <var>choice</var> can be one of ‘<samp>auto</samp>’, ‘<samp>yes</samp>’, and ‘<samp>no</samp>’
2424 where ‘<samp>auto</samp>’ is the default.
2425 </p>
2426 <p>When the option is not specified, the default choice ‘<samp>auto</samp>’ means that it
2427 is inferred whether the target has support for the phobos standard library.
2428 When the option is specified without a <var>choice</var>, the result is the same as
2429 ‘<samp>--with-libphobos-druntime-only=yes</samp>’.
2430 </p>
2431 </dd>
2432 <dt><code>--with-target-system-zlib</code></dt>
2433 <dd><p>Use installed ‘<samp>zlib</samp>’ rather than that included with GCC. This needs
2434 to be available for each multilib variant, unless configured with
2435 <samp>--with-target-system-zlib=‘<samp>auto</samp>’</samp> in which case the GCC included
2436 ‘<samp>zlib</samp>’ is only used when the system installed library is not available.
2437 </p></dd>
2438 </dl>
2439
2440 <hr />
2441 <p>
2442 <p><a href="./index.html">Return to the GCC Installation page</a>
2443 </p>
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454 </body>
2455 </html>
2456