Home | History | Annotate | Line # | Download | only in INSTALL
configure.html revision 1.13
      1 <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
      2 <html>
      3 <!-- Copyright (C) 1988-2020 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 <a name="index-Configuration"></a>
     85 <a name="index-Installing-GCC_003a-Configuration"></a>
     86 
     87 <p>Like most GNU software, GCC must be configured before it can be built.
     88 This document describes the recommended configuration procedure
     89 for both native and cross targets.
     90 </p>
     91 <p>We use <var>srcdir</var> to refer to the toplevel source directory for
     92 GCC; we use <var>objdir</var> to refer to the toplevel build/object directory.
     93 </p>
     94 <p>If you obtained the sources by cloning the repository, <var>srcdir</var>
     95 must refer to the top <samp>gcc</samp> directory, the one where the
     96 <samp>MAINTAINERS</samp> file can be found, and not its <samp>gcc</samp>
     97 subdirectory, otherwise the build will fail.
     98 </p>
     99 <p>If either <var>srcdir</var> or <var>objdir</var> is located on an automounted NFS
    100 file system, the shell&rsquo;s built-in <code>pwd</code> command will return
    101 temporary pathnames.  Using these can lead to various sorts of build
    102 problems.  To avoid this issue, set the <code>PWDCMD</code> environment
    103 variable to an automounter-aware <code>pwd</code> command, e.g.,
    104 <code>pawd</code> or &lsquo;<samp>amq -w</samp>&rsquo;, during the configuration and build
    105 phases.
    106 </p>
    107 <p>First, we <strong>highly</strong> recommend that GCC be built into a
    108 separate directory from the sources which does <strong>not</strong> reside
    109 within the source tree.  This is how we generally build GCC; building
    110 where <var>srcdir</var> == <var>objdir</var> should still work, but doesn&rsquo;t
    111 get extensive testing; building where <var>objdir</var> is a subdirectory
    112 of <var>srcdir</var> is unsupported.
    113 </p>
    114 <p>If you have previously built GCC in the same directory for a
    115 different target machine, do &lsquo;<samp>make distclean</samp>&rsquo; to delete all files
    116 that might be invalid.  One of the files this deletes is <samp>Makefile</samp>;
    117 if &lsquo;<samp>make distclean</samp>&rsquo; complains that <samp>Makefile</samp> does not exist
    118 or issues a message like &ldquo;don&rsquo;t know how to make distclean&rdquo; it probably
    119 means that the directory is already suitably clean.  However, with the
    120 recommended method of building in a separate <var>objdir</var>, you should
    121 simply use a different <var>objdir</var> for each target.
    122 </p>
    123 <p>Second, when configuring a native system, either <code>cc</code> or
    124 <code>gcc</code> must be in your path or you must set <code>CC</code> in
    125 your environment before running configure.  Otherwise the configuration
    126 scripts may fail.
    127 </p>
    128 
    129 <p>To configure GCC:
    130 </p>
    131 <div class="smallexample">
    132 <pre class="smallexample">% mkdir <var>objdir</var>
    133 % cd <var>objdir</var>
    134 % <var>srcdir</var>/configure [<var>options</var>] [<var>target</var>]
    135 </pre></div>
    136 
    137 <a name="Distributor-options"></a>
    138 <h3 class="heading">Distributor options</h3>
    139 
    140 <p>If you will be distributing binary versions of GCC, with modifications
    141 to the source code, you should use the options described in this
    142 section to make clear that your version contains modifications.
    143 </p>
    144 <dl compact="compact">
    145 <dt><code>--with-pkgversion=<var>version</var></code></dt>
    146 <dd><p>Specify a string that identifies your package.  You may wish
    147 to include a build number or build date.  This version string will be
    148 included in the output of <code>gcc --version</code>.  This suffix does
    149 not replace the default version string, only the &lsquo;<samp>GCC</samp>&rsquo; part.
    150 </p>
    151 <p>The default value is &lsquo;<samp>GCC</samp>&rsquo;.
    152 </p>
    153 </dd>
    154 <dt><code>--with-bugurl=<var>url</var></code></dt>
    155 <dd><p>Specify the URL that users should visit if they wish to report a bug.
    156 You are of course welcome to forward bugs reported to you to the FSF,
    157 if you determine that they are not bugs in your modifications.
    158 </p>
    159 <p>The default value refers to the FSF&rsquo;s GCC bug tracker.
    160 </p>
    161 </dd>
    162 <dt><code>--with-documentation-root-url=<var>url</var></code></dt>
    163 <dd><p>Specify the URL root that contains GCC option documentation.  The <var>url</var>
    164 should end with a <code>/</code> character.
    165 </p>
    166 <p>The default value is <a href="https://gcc.gnu.org/onlinedocs/">https://gcc.gnu.org/onlinedocs/</a>.
    167 </p>
    168 </dd>
    169 <dt><code>--with-changes-root-url=<var>url</var></code></dt>
    170 <dd><p>Specify the URL root that contains information about changes in GCC
    171 releases like <code>gcc-<var>version</var>/changes.html</code>.
    172 The <var>url</var> should end with a <code>/</code> character.
    173 </p>
    174 <p>The default value is <a href="https://gcc.gnu.org/">https://gcc.gnu.org/</a>.
    175 </p>
    176 </dd>
    177 </dl>
    178 
    179 <a name="Target-specification"></a>
    180 <h3 class="heading">Target specification</h3>
    181 <ul>
    182 <li> GCC has code to correctly determine the correct value for <var>target</var>
    183 for nearly all native systems.  Therefore, we highly recommend you do
    184 not provide a configure target when configuring a native compiler.
    185 
    186 </li><li> <var>target</var> must be specified as <samp>--target=<var>target</var></samp>
    187 when configuring a cross compiler; examples of valid targets would be
    188 m68k-elf, sh-elf, etc.
    189 
    190 </li><li> Specifying just <var>target</var> instead of <samp>--target=<var>target</var></samp>
    191 implies that the host defaults to <var>target</var>.
    192 </li></ul>
    193 
    194 
    195 <a name="Options-specification"></a>
    196 <h3 class="heading">Options specification</h3>
    197 
    198 <p>Use <var>options</var> to override several configure time options for
    199 GCC.  A list of supported <var>options</var> follows; &lsquo;<samp>configure
    200 --help</samp>&rsquo; may list other options, but those not listed below may not
    201 work and should not normally be used.
    202 </p>
    203 <p>Note that each <samp>--enable</samp> option has a corresponding
    204 <samp>--disable</samp> option and that each <samp>--with</samp> option has a
    205 corresponding <samp>--without</samp> option.
    206 </p>
    207 <dl compact="compact">
    208 <dt><code>--prefix=<var>dirname</var></code></dt>
    209 <dd><p>Specify the toplevel installation
    210 directory.  This is the recommended way to install the tools into a directory
    211 other than the default.  The toplevel installation directory defaults to
    212 <samp>/usr/local</samp>.
    213 </p>
    214 <p>We <strong>highly</strong> recommend against <var>dirname</var> being the same or a
    215 subdirectory of <var>objdir</var> or vice versa.  If specifying a directory
    216 beneath a user&rsquo;s home directory tree, some shells will not expand
    217 <var>dirname</var> correctly if it contains the &lsquo;<samp>~</samp>&rsquo; metacharacter; use
    218 <code>$HOME</code> instead.
    219 </p>
    220 <p>The following standard <code>autoconf</code> options are supported.  Normally you
    221 should not need to use these options.
    222 </p><dl compact="compact">
    223 <dt><code>--exec-prefix=<var>dirname</var></code></dt>
    224 <dd><p>Specify the toplevel installation directory for architecture-dependent
    225 files.  The default is <samp><var>prefix</var></samp>.
    226 </p>
    227 </dd>
    228 <dt><code>--bindir=<var>dirname</var></code></dt>
    229 <dd><p>Specify the installation directory for the executables called by users
    230 (such as <code>gcc</code> and <code>g++</code>).  The default is
    231 <samp><var>exec-prefix</var>/bin</samp>.
    232 </p>
    233 </dd>
    234 <dt><code>--libdir=<var>dirname</var></code></dt>
    235 <dd><p>Specify the installation directory for object code libraries and
    236 internal data files of GCC.  The default is <samp><var>exec-prefix</var>/lib</samp>.
    237 </p>
    238 </dd>
    239 <dt><code>--libexecdir=<var>dirname</var></code></dt>
    240 <dd><p>Specify the installation directory for internal executables of GCC.
    241 The default is <samp><var>exec-prefix</var>/libexec</samp>.
    242 </p>
    243 </dd>
    244 <dt><code>--with-slibdir=<var>dirname</var></code></dt>
    245 <dd><p>Specify the installation directory for the shared libgcc library.  The
    246 default is <samp><var>libdir</var></samp>.
    247 </p>
    248 </dd>
    249 <dt><code>--datarootdir=<var>dirname</var></code></dt>
    250 <dd><p>Specify the root of the directory tree for read-only architecture-independent
    251 data files referenced by GCC.  The default is <samp><var>prefix</var>/share</samp>.
    252 </p>
    253 </dd>
    254 <dt><code>--infodir=<var>dirname</var></code></dt>
    255 <dd><p>Specify the installation directory for documentation in info format.
    256 The default is <samp><var>datarootdir</var>/info</samp>.
    257 </p>
    258 </dd>
    259 <dt><code>--datadir=<var>dirname</var></code></dt>
    260 <dd><p>Specify the installation directory for some architecture-independent
    261 data files referenced by GCC.  The default is <samp><var>datarootdir</var></samp>.
    262 </p>
    263 </dd>
    264 <dt><code>--docdir=<var>dirname</var></code></dt>
    265 <dd><p>Specify the installation directory for documentation files (other
    266 than Info) for GCC.  The default is <samp><var>datarootdir</var>/doc</samp>.
    267 </p>
    268 </dd>
    269 <dt><code>--htmldir=<var>dirname</var></code></dt>
    270 <dd><p>Specify the installation directory for HTML documentation files.
    271 The default is <samp><var>docdir</var></samp>.
    272 </p>
    273 </dd>
    274 <dt><code>--pdfdir=<var>dirname</var></code></dt>
    275 <dd><p>Specify the installation directory for PDF documentation files.
    276 The default is <samp><var>docdir</var></samp>.
    277 </p>
    278 </dd>
    279 <dt><code>--mandir=<var>dirname</var></code></dt>
    280 <dd><p>Specify the installation directory for manual pages.  The default is
    281 <samp><var>datarootdir</var>/man</samp>.  (Note that the manual pages are only extracts
    282 from the full GCC manuals, which are provided in Texinfo format.  The manpages
    283 are derived by an automatic conversion process from parts of the full
    284 manual.)
    285 </p>
    286 </dd>
    287 <dt><code>--with-gxx-include-dir=<var>dirname</var></code></dt>
    288 <dd><p>Specify
    289 the installation directory for G++ header files.  The default depends
    290 on other configuration options, and differs between cross and native
    291 configurations.
    292 </p>
    293 </dd>
    294 <dt><code>--with-specs=<var>specs</var></code></dt>
    295 <dd><p>Specify additional command line driver SPECS.
    296 This can be useful if you need to turn on a non-standard feature by
    297 default without modifying the compiler&rsquo;s source code, for instance
    298 <samp>--with-specs=%{!fcommon:%{!fno-common:-fno-common}}</samp>.
    299 See &ldquo;Spec Files&rdquo; in the main manual
    300 </p>
    301 </dd>
    302 </dl>
    303 
    304 </dd>
    305 <dt><code>--program-prefix=<var>prefix</var></code></dt>
    306 <dd><p>GCC supports some transformations of the names of its programs when
    307 installing them.  This option prepends <var>prefix</var> to the names of
    308 programs to install in <var>bindir</var> (see above).  For example, specifying
    309 <samp>--program-prefix=foo-</samp> would result in &lsquo;<samp>gcc</samp>&rsquo;
    310 being installed as <samp>/usr/local/bin/foo-gcc</samp>.
    311 </p>
    312 </dd>
    313 <dt><code>--program-suffix=<var>suffix</var></code></dt>
    314 <dd><p>Appends <var>suffix</var> to the names of programs to install in <var>bindir</var>
    315 (see above).  For example, specifying <samp>--program-suffix=-3.1</samp>
    316 would result in &lsquo;<samp>gcc</samp>&rsquo; being installed as
    317 <samp>/usr/local/bin/gcc-3.1</samp>.
    318 </p>
    319 </dd>
    320 <dt><code>--program-transform-name=<var>pattern</var></code></dt>
    321 <dd><p>Applies the &lsquo;<samp>sed</samp>&rsquo; script <var>pattern</var> to be applied to the names
    322 of programs to install in <var>bindir</var> (see above).  <var>pattern</var> has to
    323 consist of one or more basic &lsquo;<samp>sed</samp>&rsquo; editing commands, separated by
    324 semicolons.  For example, if you want the &lsquo;<samp>gcc</samp>&rsquo; program name to be
    325 transformed to the installed program <samp>/usr/local/bin/myowngcc</samp> and
    326 the &lsquo;<samp>g++</samp>&rsquo; program name to be transformed to
    327 <samp>/usr/local/bin/gspecial++</samp> without changing other program names,
    328 you could use the pattern
    329 <samp>--program-transform-name='s/^gcc$/myowngcc/; s/^g++$/gspecial++/'</samp>
    330 to achieve this effect.
    331 </p>
    332 <p>All three options can be combined and used together, resulting in more
    333 complex conversion patterns.  As a basic rule, <var>prefix</var> (and
    334 <var>suffix</var>) are prepended (appended) before further transformations
    335 can happen with a special transformation script <var>pattern</var>.
    336 </p>
    337 <p>As currently implemented, this option only takes effect for native
    338 builds; cross compiler binaries&rsquo; names are not transformed even when a
    339 transformation is explicitly asked for by one of these options.
    340 </p>
    341 <p>For native builds, some of the installed programs are also installed
    342 with the target alias in front of their name, as in
    343 &lsquo;<samp>i686-pc-linux-gnu-gcc</samp>&rsquo;.  All of the above transformations happen
    344 before the target alias is prepended to the name&mdash;so, specifying
    345 <samp>--program-prefix=foo-</samp> and <samp>program-suffix=-3.1</samp>, the
    346 resulting binary would be installed as
    347 <samp>/usr/local/bin/i686-pc-linux-gnu-foo-gcc-3.1</samp>.
    348 </p>
    349 <p>As a last shortcoming, none of the installed Ada programs are
    350 transformed yet, which will be fixed in some time.
    351 </p>
    352 </dd>
    353 <dt><code>--with-local-prefix=<var>dirname</var></code></dt>
    354 <dd><p>Specify the
    355 installation directory for local include files.  The default is
    356 <samp>/usr/local</samp>.  Specify this option if you want the compiler to
    357 search directory <samp><var>dirname</var>/include</samp> for locally installed
    358 header files <em>instead</em> of <samp>/usr/local/include</samp>.
    359 </p>
    360 <p>You should specify <samp>--with-local-prefix</samp> <strong>only</strong> if your
    361 site has a different convention (not <samp>/usr/local</samp>) for where to put
    362 site-specific files.
    363 </p>
    364 <p>The default value for <samp>--with-local-prefix</samp> is <samp>/usr/local</samp>
    365 regardless of the value of <samp>--prefix</samp>.  Specifying
    366 <samp>--prefix</samp> has no effect on which directory GCC searches for
    367 local header files.  This may seem counterintuitive, but actually it is
    368 logical.
    369 </p>
    370 <p>The purpose of <samp>--prefix</samp> is to specify where to <em>install
    371 GCC</em>.  The local header files in <samp>/usr/local/include</samp>&mdash;if you put
    372 any in that directory&mdash;are not part of GCC.  They are part of other
    373 programs&mdash;perhaps many others.  (GCC installs its own header files in
    374 another directory which is based on the <samp>--prefix</samp> value.)
    375 </p>
    376 <p>Both the local-prefix include directory and the GCC-prefix include
    377 directory are part of GCC&rsquo;s &ldquo;system include&rdquo; directories.  Although these
    378 two directories are not fixed, they need to be searched in the proper
    379 order for the correct processing of the include_next directive.  The
    380 local-prefix include directory is searched before the GCC-prefix
    381 include directory.  Another characteristic of system include directories
    382 is that pedantic warnings are turned off for headers in these directories.
    383 </p>
    384 <p>Some autoconf macros add <samp>-I <var>directory</var></samp> options to the
    385 compiler command line, to ensure that directories containing installed
    386 packages&rsquo; headers are searched.  When <var>directory</var> is one of GCC&rsquo;s
    387 system include directories, GCC will ignore the option so that system
    388 directories continue to be processed in the correct order.  This
    389 may result in a search order different from what was specified but the
    390 directory will still be searched.
    391 </p>
    392 <p>GCC automatically searches for ordinary libraries using
    393 <code>GCC_EXEC_PREFIX</code>.  Thus, when the same installation prefix is
    394 used for both GCC and packages, GCC will automatically search for
    395 both headers and libraries.  This provides a configuration that is
    396 easy to use.  GCC behaves in a manner similar to that when it is
    397 installed as a system compiler in <samp>/usr</samp>.
    398 </p>
    399 <p>Sites that need to install multiple versions of GCC may not want to
    400 use the above simple configuration.  It is possible to use the
    401 <samp>--program-prefix</samp>, <samp>--program-suffix</samp> and
    402 <samp>--program-transform-name</samp> options to install multiple versions
    403 into a single directory, but it may be simpler to use different prefixes
    404 and the <samp>--with-local-prefix</samp> option to specify the location of the
    405 site-specific files for each version.  It will then be necessary for
    406 users to specify explicitly the location of local site libraries
    407 (e.g., with <code>LIBRARY_PATH</code>).
    408 </p>
    409 <p>The same value can be used for both <samp>--with-local-prefix</samp> and
    410 <samp>--prefix</samp> provided it is not <samp>/usr</samp>.  This can be used
    411 to avoid the default search of <samp>/usr/local/include</samp>.
    412 </p>
    413 <p><strong>Do not</strong> specify <samp>/usr</samp> as the <samp>--with-local-prefix</samp>!
    414 The directory you use for <samp>--with-local-prefix</samp> <strong>must not</strong>
    415 contain any of the system&rsquo;s standard header files.  If it did contain
    416 them, certain programs would be miscompiled (including GNU Emacs, on
    417 certain targets), because this would override and nullify the header
    418 file corrections made by the <code>fixincludes</code> script.
    419 </p>
    420 <p>Indications are that people who use this option use it based on mistaken
    421 ideas of what it is for.  People use it as if it specified where to
    422 install part of GCC.  Perhaps they make this assumption because
    423 installing GCC creates the directory.
    424 </p>
    425 </dd>
    426 <dt><code>--with-gcc-major-version-only</code></dt>
    427 <dd><p>Specifies that GCC should use only the major number rather than
    428 <var>major</var>.<var>minor</var>.<var>patchlevel</var> in filesystem paths.
    429 </p>
    430 </dd>
    431 <dt><code>--with-native-system-header-dir=<var>dirname</var></code></dt>
    432 <dd><p>Specifies that <var>dirname</var> is the directory that contains native system
    433 header files, rather than <samp>/usr/include</samp>.  This option is most useful
    434 if you are creating a compiler that should be isolated from the system
    435 as much as possible.  It is most commonly used with the
    436 <samp>--with-sysroot</samp> option and will cause GCC to search
    437 <var>dirname</var> inside the system root specified by that option.
    438 </p>
    439 </dd>
    440 <dt><code>--enable-shared[=<var>package</var>[,&hellip;]]</code></dt>
    441 <dd><p>Build shared versions of libraries, if shared libraries are supported on
    442 the target platform.  Unlike GCC 2.95.x and earlier, shared libraries
    443 are enabled by default on all platforms that support shared libraries.
    444 </p>
    445 <p>If a list of packages is given as an argument, build shared libraries
    446 only for the listed packages.  For other packages, only static libraries
    447 will be built.  Package names currently recognized in the GCC tree are
    448 &lsquo;<samp>libgcc</samp>&rsquo; (also known as &lsquo;<samp>gcc</samp>&rsquo;), &lsquo;<samp>libstdc++</samp>&rsquo; (not
    449 &lsquo;<samp>libstdc++-v3</samp>&rsquo;), &lsquo;<samp>libffi</samp>&rsquo;, &lsquo;<samp>zlib</samp>&rsquo;, &lsquo;<samp>boehm-gc</samp>&rsquo;,
    450 &lsquo;<samp>ada</samp>&rsquo;, &lsquo;<samp>libada</samp>&rsquo;, &lsquo;<samp>libgo</samp>&rsquo;, &lsquo;<samp>libobjc</samp>&rsquo;, and &lsquo;<samp>libphobos</samp>&rsquo;.
    451 Note &lsquo;<samp>libiberty</samp>&rsquo; does not support shared libraries at all.
    452 </p>
    453 <p>Use <samp>--disable-shared</samp> to build only static libraries.  Note that
    454 <samp>--disable-shared</samp> does not accept a list of package names as
    455 argument, only <samp>--enable-shared</samp> does.
    456 </p>
    457 <p>Contrast with <samp>--enable-host-shared</samp>, which affects <em>host</em>
    458 code.
    459 </p>
    460 </dd>
    461 <dt><code>--enable-host-shared</code></dt>
    462 <dd><p>Specify that the <em>host</em> code should be built into position-independent
    463 machine code (with -fPIC), allowing it to be used within shared libraries,
    464 but yielding a slightly slower compiler.
    465 </p>
    466 <p>This option is required when building the libgccjit.so library.
    467 </p>
    468 <p>Contrast with <samp>--enable-shared</samp>, which affects <em>target</em>
    469 libraries.
    470 </p>
    471 </dd>
    472 <dt><code><a name="with-gnu-as"></a>--with-gnu-as</code></dt>
    473 <dd><p>Specify that the compiler should assume that the
    474 assembler it finds is the GNU assembler.  However, this does not modify
    475 the rules to find an assembler and will result in confusion if the
    476 assembler found is not actually the GNU assembler.  (Confusion may also
    477 result if the compiler finds the GNU assembler but has not been
    478 configured with <samp>--with-gnu-as</samp>.)  If you have more than one
    479 assembler installed on your system, you may want to use this option in
    480 connection with <samp>--with-as=<var>pathname</var></samp> or
    481 <samp>--with-build-time-tools=<var>pathname</var></samp>.
    482 </p>
    483 <p>The following systems are the only ones where it makes a difference
    484 whether you use the GNU assembler.  On any other system,
    485 <samp>--with-gnu-as</samp> has no effect.
    486 </p>
    487 <ul>
    488 <li> &lsquo;<samp>hppa1.0-<var>any</var>-<var>any</var></samp>&rsquo;
    489 </li><li> &lsquo;<samp>hppa1.1-<var>any</var>-<var>any</var></samp>&rsquo;
    490 </li><li> &lsquo;<samp>sparc-sun-solaris2.<var>any</var></samp>&rsquo;
    491 </li><li> &lsquo;<samp>sparc64-<var>any</var>-solaris2.<var>any</var></samp>&rsquo;
    492 </li></ul>
    493 
    494 </dd>
    495 <dt><code><a name="with-as"></a>--with-as=<var>pathname</var></code></dt>
    496 <dd><p>Specify that the compiler should use the assembler pointed to by
    497 <var>pathname</var>, rather than the one found by the standard rules to find
    498 an assembler, which are:
    499 </p><ul>
    500 <li> Unless GCC is being built with a cross compiler, check the
    501 <samp><var>libexec</var>/gcc/<var>target</var>/<var>version</var></samp> directory.
    502 <var>libexec</var> defaults to <samp><var>exec-prefix</var>/libexec</samp>;
    503 <var>exec-prefix</var> defaults to <var>prefix</var>, which
    504 defaults to <samp>/usr/local</samp> unless overridden by the
    505 <samp>--prefix=<var>pathname</var></samp> switch described above.  <var>target</var>
    506 is the target system triple, such as &lsquo;<samp>sparc-sun-solaris2.7</samp>&rsquo;, and
    507 <var>version</var> denotes the GCC version, such as 3.0.
    508 
    509 </li><li> If the target system is the same that you are building on, check
    510 operating system specific directories (e.g. <samp>/usr/ccs/bin</samp> on
    511 Solaris 2).
    512 
    513 </li><li> Check in the <code>PATH</code> for a tool whose name is prefixed by the
    514 target system triple.
    515 
    516 </li><li> Check in the <code>PATH</code> for a tool whose name is not prefixed by the
    517 target system triple, if the host and target system triple are
    518 the same (in other words, we use a host tool if it can be used for
    519 the target as well).
    520 </li></ul>
    521 
    522 <p>You may want to use <samp>--with-as</samp> if no assembler
    523 is installed in the directories listed above, or if you have multiple
    524 assemblers installed and want to choose one that is not found by the
    525 above rules.
    526 </p>
    527 </dd>
    528 <dt><code><a name="with-gnu-ld"></a>--with-gnu-ld</code></dt>
    529 <dd><p>Same as <a href="#with-gnu-as"><samp>--with-gnu-as</samp></a>
    530 but for the linker.
    531 </p>
    532 </dd>
    533 <dt><code>--with-ld=<var>pathname</var></code></dt>
    534 <dd><p>Same as <a href="#with-as"><samp>--with-as</samp></a>
    535 but for the linker.
    536 </p>
    537 </dd>
    538 <dt><code>--with-stabs</code></dt>
    539 <dd><p>Specify that stabs debugging
    540 information should be used instead of whatever format the host normally
    541 uses.  Normally GCC uses the same debug format as the host system.
    542 </p>
    543 </dd>
    544 <dt><code>--with-tls=<var>dialect</var></code></dt>
    545 <dd><p>Specify the default TLS dialect, for systems were there is a choice.
    546 For ARM targets, possible values for <var>dialect</var> are <code>gnu</code> or
    547 <code>gnu2</code>, which select between the original GNU dialect and the GNU TLS
    548 descriptor-based dialect.
    549 </p>
    550 </dd>
    551 <dt><code>--enable-multiarch</code></dt>
    552 <dd><p>Specify whether to enable or disable multiarch support.  The default is
    553 to check for glibc start files in a multiarch location, and enable it
    554 if the files are found.  The auto detection is enabled for native builds,
    555 and for cross builds configured with <samp>--with-sysroot</samp>, and without
    556 <samp>--with-native-system-header-dir</samp>.
    557 More documentation about multiarch can be found at
    558 <a href="https://wiki.debian.org/Multiarch">https://wiki.debian.org/Multiarch</a>.
    559 </p>
    560 </dd>
    561 <dt><code>--enable-sjlj-exceptions</code></dt>
    562 <dd><p>Force use of the <code>setjmp</code>/<code>longjmp</code>-based scheme for exceptions.
    563 &lsquo;<samp>configure</samp>&rsquo; ordinarily picks the correct value based on the platform.
    564 Only use this option if you are sure you need a different setting.
    565 </p>
    566 </dd>
    567 <dt><code>--enable-vtable-verify</code></dt>
    568 <dd><p>Specify whether to enable or disable the vtable verification feature.
    569 Enabling this feature causes libstdc++ to be built with its virtual calls
    570 in verifiable mode.  This means that, when linked with libvtv, every
    571 virtual call in libstdc++ will verify the vtable pointer through which the
    572 call will be made before actually making the call.  If not linked with libvtv,
    573 the verifier will call stub functions (in libstdc++ itself) and do nothing.
    574 If vtable verification is disabled, then libstdc++ is not built with its
    575 virtual calls in verifiable mode at all.  However the libvtv library will
    576 still be built (see <samp>--disable-libvtv</samp> to turn off building libvtv).
    577 <samp>--disable-vtable-verify</samp> is the default.
    578 </p>
    579 </dd>
    580 <dt><code>--disable-gcov</code></dt>
    581 <dd><p>Specify that the run-time library used for coverage analysis
    582 and associated host tools should not be built.
    583 </p>
    584 </dd>
    585 <dt><code>--disable-multilib</code></dt>
    586 <dd><p>Specify that multiple target
    587 libraries to support different target variants, calling
    588 conventions, etc. should not be built.  The default is to build a
    589 predefined set of them.
    590 </p>
    591 <p>Some targets provide finer-grained control over which multilibs are built
    592 (e.g., <samp>--disable-softfloat</samp>):
    593 </p><dl compact="compact">
    594 <dt><code>arm-*-*</code></dt>
    595 <dd><p>fpu, 26bit, underscore, interwork, biendian, nofmult.
    596 </p>
    597 </dd>
    598 <dt><code>m68*-*-*</code></dt>
    599 <dd><p>softfloat, m68881, m68000, m68020.
    600 </p>
    601 </dd>
    602 <dt><code>mips*-*-*</code></dt>
    603 <dd><p>single-float, biendian, softfloat.
    604 </p>
    605 </dd>
    606 <dt><code>msp430-*-*</code></dt>
    607 <dd><p>no-exceptions
    608 </p>
    609 </dd>
    610 <dt><code>powerpc*-*-*, rs6000*-*-*</code></dt>
    611 <dd><p>aix64, pthread, softfloat, powercpu, powerpccpu, powerpcos, biendian,
    612 sysv, aix.
    613 </p>
    614 </dd>
    615 </dl>
    616 
    617 </dd>
    618 <dt><code>--with-multilib-list=<var>list</var></code></dt>
    619 <dt><code>--without-multilib-list</code></dt>
    620 <dd><p>Specify what multilibs to build.  <var>list</var> is a comma separated list of
    621 values, possibly consisting of a single value.  Currently only implemented
    622 for aarch64*-*-*, arm*-*-*, riscv*-*-*, sh*-*-* and x86-64-*-linux*.  The
    623 accepted values and meaning for each target is given below.
    624 </p>
    625 <dl compact="compact">
    626 <dt><code>aarch64*-*-*</code></dt>
    627 <dd><p><var>list</var> is a comma separated list of <code>ilp32</code>, and <code>lp64</code>
    628 to enable ILP32 and LP64 run-time libraries, respectively.  If
    629 <var>list</var> is empty, then there will be no multilibs and only the
    630 default run-time library will be built.  If <var>list</var> is
    631 <code>default</code> or &ndash;with-multilib-list= is not specified, then the
    632 default set of libraries is selected based on the value of
    633 <samp>--target</samp>.
    634 </p>
    635 </dd>
    636 <dt><code>arm*-*-*</code></dt>
    637 <dd><p><var>list</var> is a comma separated list of <code>aprofile</code> and
    638 <code>rmprofile</code> to build multilibs for A or R and M architecture
    639 profiles respectively.  Note that, due to some limitation of the current
    640 multilib framework, using the combined <code>aprofile,rmprofile</code>
    641 multilibs selects in some cases a less optimal multilib than when using
    642 the multilib profile for the architecture targetted.  The special value
    643 <code>default</code> is also accepted and is equivalent to omitting the
    644 option, i.e., only the default run-time library will be enabled.
    645 </p>
    646 <p><var>list</var> may instead contain <code>@name</code>, to use the multilib
    647 configuration Makefile fragment <samp>name</samp> in <samp>gcc/config/arm</samp> in
    648 the source tree (it is part of the corresponding sources, after all).
    649 It is recommended, but not required, that files used for this purpose to
    650 be named starting with <samp>t-ml-</samp>, to make their intended purpose
    651 self-evident, in line with GCC conventions.  Such files enable custom,
    652 user-chosen multilib lists to be configured.  Whether multiple such
    653 files can be used together depends on the contents of the supplied
    654 files.  See <samp>gcc/config/arm/t-multilib</samp> and its supplementary
    655 <samp>gcc/config/arm/t-*profile</samp> files for an example of what such
    656 Makefile fragments might look like for this version of GCC.  The macros
    657 expected to be defined in these fragments are not stable across GCC
    658 releases, so make sure they define the <code>MULTILIB</code>-related macros
    659 expected by the version of GCC you are building.
    660 See &ldquo;Target Makefile Fragments&rdquo; in the internals manual.
    661 </p>
    662 <p>The table below gives the combination of ISAs, architectures, FPUs and
    663 floating-point ABIs for which multilibs are built for each predefined
    664 profile.  The union of these options is considered when specifying both
    665 <code>aprofile</code> and <code>rmprofile</code>.
    666 </p>
    667 <table>
    668 <tr><td width="15%">Option</td><td width="28%">aprofile</td><td width="30%">rmprofile</td></tr>
    669 <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>
    670 <tr><td width="15%">Architectures<br><br><br><br><br><br></td><td width="28%">default architecture<br>
    671 <code>-march=armv7-a</code><br>
    672 <code>-march=armv7ve</code><br>
    673 <code>-march=armv8-a</code><br><br><br></td><td width="30%">default architecture<br>
    674 <code>-march=armv6s-m</code><br>
    675 <code>-march=armv7-m</code><br>
    676 <code>-march=armv7e-m</code><br>
    677 <code>-march=armv8-m.base</code><br>
    678 <code>-march=armv8-m.main</code><br>
    679 <code>-march=armv7</code></td></tr>
    680 <tr><td width="15%">FPUs<br><br><br><br><br></td><td width="28%">none<br>
    681 <code>-mfpu=vfpv3-d16</code><br>
    682 <code>-mfpu=neon</code><br>
    683 <code>-mfpu=vfpv4-d16</code><br>
    684 <code>-mfpu=neon-vfpv4</code><br>
    685 <code>-mfpu=neon-fp-armv8</code></td><td width="30%">none<br>
    686 <code>-mfpu=vfpv3-d16</code><br>
    687 <code>-mfpu=fpv4-sp-d16</code><br>
    688 <code>-mfpu=fpv5-sp-d16</code><br>
    689 <code>-mfpu=fpv5-d16</code><br></td></tr>
    690 <tr><td width="15%">floating-point ABIs<br><br></td><td width="28%"><code>-mfloat-abi=soft</code><br>
    691 <code>-mfloat-abi=softfp</code><br>
    692 <code>-mfloat-abi=hard</code></td><td width="30%"><code>-mfloat-abi=soft</code><br>
    693 <code>-mfloat-abi=softfp</code><br>
    694 <code>-mfloat-abi=hard</code></td></tr>
    695 </table>
    696 
    697 </dd>
    698 <dt><code>riscv*-*-*</code></dt>
    699 <dd><p><var>list</var> is a single ABI name.  The target architecture must be either
    700 <code>rv32gc</code> or <code>rv64gc</code>.  This will build a single multilib for the
    701 specified architecture and ABI pair.  If <code>--with-multilib-list</code> is not
    702 given, then a default set of multilibs is selected based on the value of
    703 <samp>--target</samp>.  This is usually a large set of multilibs.
    704 </p>
    705 </dd>
    706 <dt><code>sh*-*-*</code></dt>
    707 <dd><p><var>list</var> is a comma separated list of CPU names.  These must be of the
    708 form <code>sh*</code> or <code>m*</code> (in which case they match the compiler option
    709 for that processor).  The list should not contain any endian options -
    710 these are handled by <samp>--with-endian</samp>.
    711 </p>
    712 <p>If <var>list</var> is empty, then there will be no multilibs for extra
    713 processors.  The multilib for the secondary endian remains enabled.
    714 </p>
    715 <p>As a special case, if an entry in the list starts with a <code>!</code>
    716 (exclamation point), then it is added to the list of excluded multilibs.
    717 Entries of this sort should be compatible with &lsquo;<samp>MULTILIB_EXCLUDES</samp>&rsquo;
    718 (once the leading <code>!</code> has been stripped).
    719 </p>
    720 <p>If <samp>--with-multilib-list</samp> is not given, then a default set of
    721 multilibs is selected based on the value of <samp>--target</samp>.  This is
    722 usually the complete set of libraries, but some targets imply a more
    723 specialized subset.
    724 </p>
    725 <p>Example 1: to configure a compiler for SH4A only, but supporting both
    726 endians, with little endian being the default:
    727 </p><div class="smallexample">
    728 <pre class="smallexample">--with-cpu=sh4a --with-endian=little,big --with-multilib-list=
    729 </pre></div>
    730 
    731 <p>Example 2: to configure a compiler for both SH4A and SH4AL-DSP, but with
    732 only little endian SH4AL:
    733 </p><div class="smallexample">
    734 <pre class="smallexample">--with-cpu=sh4a --with-endian=little,big \
    735 --with-multilib-list=sh4al,!mb/m4al
    736 </pre></div>
    737 
    738 </dd>
    739 <dt><code>x86-64-*-linux*</code></dt>
    740 <dd><p><var>list</var> is a comma separated list of <code>m32</code>, <code>m64</code> and
    741 <code>mx32</code> to enable 32-bit, 64-bit and x32 run-time libraries,
    742 respectively.  If <var>list</var> is empty, then there will be no multilibs
    743 and only the default run-time library will be enabled.
    744 </p>
    745 <p>If <samp>--with-multilib-list</samp> is not given, then only 32-bit and
    746 64-bit run-time libraries will be enabled.
    747 </p></dd>
    748 </dl>
    749 
    750 </dd>
    751 <dt><code>--with-endian=<var>endians</var></code></dt>
    752 <dd><p>Specify what endians to use.
    753 Currently only implemented for sh*-*-*.
    754 </p>
    755 <p><var>endians</var> may be one of the following:
    756 </p><dl compact="compact">
    757 <dt><code>big</code></dt>
    758 <dd><p>Use big endian exclusively.
    759 </p></dd>
    760 <dt><code>little</code></dt>
    761 <dd><p>Use little endian exclusively.
    762 </p></dd>
    763 <dt><code>big,little</code></dt>
    764 <dd><p>Use big endian by default.  Provide a multilib for little endian.
    765 </p></dd>
    766 <dt><code>little,big</code></dt>
    767 <dd><p>Use little endian by default.  Provide a multilib for big endian.
    768 </p></dd>
    769 </dl>
    770 
    771 </dd>
    772 <dt><code>--enable-threads</code></dt>
    773 <dd><p>Specify that the target
    774 supports threads.  This affects the Objective-C compiler and runtime
    775 library, and exception handling for other languages like C++.
    776 On some systems, this is the default.
    777 </p>
    778 <p>In general, the best (and, in many cases, the only known) threading
    779 model available will be configured for use.  Beware that on some
    780 systems, GCC has not been taught what threading models are generally
    781 available for the system.  In this case, <samp>--enable-threads</samp> is an
    782 alias for <samp>--enable-threads=single</samp>.
    783 </p>
    784 </dd>
    785 <dt><code>--disable-threads</code></dt>
    786 <dd><p>Specify that threading support should be disabled for the system.
    787 This is an alias for <samp>--enable-threads=single</samp>.
    788 </p>
    789 </dd>
    790 <dt><code>--enable-threads=<var>lib</var></code></dt>
    791 <dd><p>Specify that
    792 <var>lib</var> is the thread support library.  This affects the Objective-C
    793 compiler and runtime library, and exception handling for other languages
    794 like C++.  The possibilities for <var>lib</var> are:
    795 </p>
    796 <dl compact="compact">
    797 <dt><code>aix</code></dt>
    798 <dd><p>AIX thread support.
    799 </p></dd>
    800 <dt><code>dce</code></dt>
    801 <dd><p>DCE thread support.
    802 </p></dd>
    803 <dt><code>lynx</code></dt>
    804 <dd><p>LynxOS thread support.
    805 </p></dd>
    806 <dt><code>mipssde</code></dt>
    807 <dd><p>MIPS SDE thread support.
    808 </p></dd>
    809 <dt><code>no</code></dt>
    810 <dd><p>This is an alias for &lsquo;<samp>single</samp>&rsquo;.
    811 </p></dd>
    812 <dt><code>posix</code></dt>
    813 <dd><p>Generic POSIX/Unix98 thread support.
    814 </p></dd>
    815 <dt><code>rtems</code></dt>
    816 <dd><p>RTEMS thread support.
    817 </p></dd>
    818 <dt><code>single</code></dt>
    819 <dd><p>Disable thread support, should work for all platforms.
    820 </p></dd>
    821 <dt><code>tpf</code></dt>
    822 <dd><p>TPF thread support.
    823 </p></dd>
    824 <dt><code>vxworks</code></dt>
    825 <dd><p>VxWorks thread support.
    826 </p></dd>
    827 <dt><code>win32</code></dt>
    828 <dd><p>Microsoft Win32 API thread support.
    829 </p></dd>
    830 </dl>
    831 
    832 </dd>
    833 <dt><code>--enable-tls</code></dt>
    834 <dd><p>Specify that the target supports TLS (Thread Local Storage).  Usually
    835 configure can correctly determine if TLS is supported.  In cases where
    836 it guesses incorrectly, TLS can be explicitly enabled or disabled with
    837 <samp>--enable-tls</samp> or <samp>--disable-tls</samp>.  This can happen if
    838 the assembler supports TLS but the C library does not, or if the
    839 assumptions made by the configure test are incorrect.
    840 </p>
    841 </dd>
    842 <dt><code>--disable-tls</code></dt>
    843 <dd><p>Specify that the target does not support TLS.
    844 This is an alias for <samp>--enable-tls=no</samp>.
    845 </p>
    846 </dd>
    847 <dt><code>--disable-tm-clone-registry</code></dt>
    848 <dd><p>Disable TM clone registry in libgcc. It is enabled in libgcc by default.
    849 This option helps to reduce code size for embedded targets which do
    850 not use transactional memory.
    851 </p>
    852 </dd>
    853 <dt><code>--with-cpu=<var>cpu</var></code></dt>
    854 <dt><code>--with-cpu-32=<var>cpu</var></code></dt>
    855 <dt><code>--with-cpu-64=<var>cpu</var></code></dt>
    856 <dd><p>Specify which cpu variant the compiler should generate code for by default.
    857 <var>cpu</var> will be used as the default value of the <samp>-mcpu=</samp> switch.
    858 This option is only supported on some targets, including ARC, ARM, i386, M68k,
    859 PowerPC, and SPARC.  It is mandatory for ARC.  The <samp>--with-cpu-32</samp> and
    860 <samp>--with-cpu-64</samp> options specify separate default CPUs for
    861 32-bit and 64-bit modes; these options are only supported for aarch64, i386,
    862 x86-64, PowerPC, and SPARC.
    863 </p>
    864 </dd>
    865 <dt><code>--with-schedule=<var>cpu</var></code></dt>
    866 <dt><code>--with-arch=<var>cpu</var></code></dt>
    867 <dt><code>--with-arch-32=<var>cpu</var></code></dt>
    868 <dt><code>--with-arch-64=<var>cpu</var></code></dt>
    869 <dt><code>--with-tune=<var>cpu</var></code></dt>
    870 <dt><code>--with-tune-32=<var>cpu</var></code></dt>
    871 <dt><code>--with-tune-64=<var>cpu</var></code></dt>
    872 <dt><code>--with-abi=<var>abi</var></code></dt>
    873 <dt><code>--with-fpu=<var>type</var></code></dt>
    874 <dt><code>--with-float=<var>type</var></code></dt>
    875 <dd><p>These configure options provide default values for the <samp>-mschedule=</samp>,
    876 <samp>-march=</samp>, <samp>-mtune=</samp>, <samp>-mabi=</samp>, and <samp>-mfpu=</samp>
    877 options and for <samp>-mhard-float</samp> or <samp>-msoft-float</samp>.  As with
    878 <samp>--with-cpu</samp>, which switches will be accepted and acceptable values
    879 of the arguments depend on the target.
    880 </p>
    881 </dd>
    882 <dt><code>--with-mode=<var>mode</var></code></dt>
    883 <dd><p>Specify if the compiler should default to <samp>-marm</samp> or <samp>-mthumb</samp>.
    884 This option is only supported on ARM targets.
    885 </p>
    886 </dd>
    887 <dt><code>--with-stack-offset=<var>num</var></code></dt>
    888 <dd><p>This option sets the default for the -mstack-offset=<var>num</var> option,
    889 and will thus generally also control the setting of this option for
    890 libraries.  This option is only supported on Epiphany targets.
    891 </p>
    892 </dd>
    893 <dt><code>--with-fpmath=<var>isa</var></code></dt>
    894 <dd><p>This options sets <samp>-mfpmath=sse</samp> by default and specifies the default
    895 ISA for floating-point arithmetics.  You can select either &lsquo;<samp>sse</samp>&rsquo; which
    896 enables <samp>-msse2</samp> or &lsquo;<samp>avx</samp>&rsquo; which enables <samp>-mavx</samp> by default.
    897 This option is only supported on i386 and x86-64 targets.
    898 </p>
    899 </dd>
    900 <dt><code>--with-fp-32=<var>mode</var></code></dt>
    901 <dd><p>On MIPS targets, set the default value for the <samp>-mfp</samp> option when using
    902 the o32 ABI.  The possibilities for <var>mode</var> are:
    903 </p><dl compact="compact">
    904 <dt><code>32</code></dt>
    905 <dd><p>Use the o32 FP32 ABI extension, as with the <samp>-mfp32</samp> command-line
    906 option.
    907 </p></dd>
    908 <dt><code>xx</code></dt>
    909 <dd><p>Use the o32 FPXX ABI extension, as with the <samp>-mfpxx</samp> command-line
    910 option.
    911 </p></dd>
    912 <dt><code>64</code></dt>
    913 <dd><p>Use the o32 FP64 ABI extension, as with the <samp>-mfp64</samp> command-line
    914 option.
    915 </p></dd>
    916 </dl>
    917 <p>In the absence of this configuration option the default is to use the o32
    918 FP32 ABI extension.
    919 </p>
    920 </dd>
    921 <dt><code>--with-odd-spreg-32</code></dt>
    922 <dd><p>On MIPS targets, set the <samp>-modd-spreg</samp> option by default when using
    923 the o32 ABI.
    924 </p>
    925 </dd>
    926 <dt><code>--without-odd-spreg-32</code></dt>
    927 <dd><p>On MIPS targets, set the <samp>-mno-odd-spreg</samp> option by default when using
    928 the o32 ABI.  This is normally used in conjunction with
    929 <samp>--with-fp-32=64</samp> in order to target the o32 FP64A ABI extension.
    930 </p>
    931 </dd>
    932 <dt><code>--with-nan=<var>encoding</var></code></dt>
    933 <dd><p>On MIPS targets, set the default encoding convention to use for the
    934 special not-a-number (NaN) IEEE 754 floating-point data.  The
    935 possibilities for <var>encoding</var> are:
    936 </p><dl compact="compact">
    937 <dt><code>legacy</code></dt>
    938 <dd><p>Use the legacy encoding, as with the <samp>-mnan=legacy</samp> command-line
    939 option.
    940 </p></dd>
    941 <dt><code>2008</code></dt>
    942 <dd><p>Use the 754-2008 encoding, as with the <samp>-mnan=2008</samp> command-line
    943 option.
    944 </p></dd>
    945 </dl>
    946 <p>To use this configuration option you must have an assembler version
    947 installed that supports the <samp>-mnan=</samp> command-line option too.
    948 In the absence of this configuration option the default convention is
    949 the legacy encoding, as when neither of the <samp>-mnan=2008</samp> and
    950 <samp>-mnan=legacy</samp> command-line options has been used.
    951 </p>
    952 </dd>
    953 <dt><code>--with-divide=<var>type</var></code></dt>
    954 <dd><p>Specify how the compiler should generate code for checking for
    955 division by zero.  This option is only supported on the MIPS target.
    956 The possibilities for <var>type</var> are:
    957 </p><dl compact="compact">
    958 <dt><code>traps</code></dt>
    959 <dd><p>Division by zero checks use conditional traps (this is the default on
    960 systems that support conditional traps).
    961 </p></dd>
    962 <dt><code>breaks</code></dt>
    963 <dd><p>Division by zero checks use the break instruction.
    964 </p></dd>
    965 </dl>
    966 
    967 
    968 </dd>
    969 <dt><code>--with-llsc</code></dt>
    970 <dd><p>On MIPS targets, make <samp>-mllsc</samp> the default when no
    971 <samp>-mno-llsc</samp> option is passed.  This is the default for
    972 Linux-based targets, as the kernel will emulate them if the ISA does
    973 not provide them.
    974 </p>
    975 </dd>
    976 <dt><code>--without-llsc</code></dt>
    977 <dd><p>On MIPS targets, make <samp>-mno-llsc</samp> the default when no
    978 <samp>-mllsc</samp> option is passed.
    979 </p>
    980 </dd>
    981 <dt><code>--with-synci</code></dt>
    982 <dd><p>On MIPS targets, make <samp>-msynci</samp> the default when no
    983 <samp>-mno-synci</samp> option is passed.
    984 </p>
    985 </dd>
    986 <dt><code>--without-synci</code></dt>
    987 <dd><p>On MIPS targets, make <samp>-mno-synci</samp> the default when no
    988 <samp>-msynci</samp> option is passed.  This is the default.
    989 </p>
    990 </dd>
    991 <dt><code>--with-lxc1-sxc1</code></dt>
    992 <dd><p>On MIPS targets, make <samp>-mlxc1-sxc1</samp> the default when no
    993 <samp>-mno-lxc1-sxc1</samp> option is passed.  This is the default.
    994 </p>
    995 </dd>
    996 <dt><code>--without-lxc1-sxc1</code></dt>
    997 <dd><p>On MIPS targets, make <samp>-mno-lxc1-sxc1</samp> the default when no
    998 <samp>-mlxc1-sxc1</samp> option is passed.  The indexed load/store
    999 instructions are not directly a problem but can lead to unexpected
   1000 behaviour when deployed in an application intended for a 32-bit address
   1001 space but run on a 64-bit processor.  The issue is seen because all
   1002 known MIPS 64-bit Linux kernels execute o32 and n32 applications
   1003 with 64-bit addressing enabled which affects the overflow behaviour
   1004 of the indexed addressing mode.  GCC will assume that ordinary
   1005 32-bit arithmetic overflow behaviour is the same whether performed
   1006 as an <code>addu</code> instruction or as part of the address calculation
   1007 in <code>lwxc1</code> type instructions.  This assumption holds true in a
   1008 pure 32-bit environment and can hold true in a 64-bit environment if
   1009 the address space is accurately set to be 32-bit for o32 and n32.
   1010 </p>
   1011 </dd>
   1012 <dt><code>--with-madd4</code></dt>
   1013 <dd><p>On MIPS targets, make <samp>-mmadd4</samp> the default when no
   1014 <samp>-mno-madd4</samp> option is passed.  This is the default.
   1015 </p>
   1016 </dd>
   1017 <dt><code>--without-madd4</code></dt>
   1018 <dd><p>On MIPS targets, make <samp>-mno-madd4</samp> the default when no
   1019 <samp>-mmadd4</samp> option is passed.  The <code>madd4</code> instruction
   1020 family can be problematic when targeting a combination of cores that
   1021 implement these instructions differently.  There are two known cores
   1022 that implement these as fused operations instead of unfused (where
   1023 unfused is normally expected).  Disabling these instructions is the
   1024 only way to ensure compatible code is generated; this will incur
   1025 a performance penalty.
   1026 </p>
   1027 </dd>
   1028 <dt><code>--with-mips-plt</code></dt>
   1029 <dd><p>On MIPS targets, make use of copy relocations and PLTs.
   1030 These features are extensions to the traditional
   1031 SVR4-based MIPS ABIs and require support from GNU binutils
   1032 and the runtime C library.
   1033 </p>
   1034 </dd>
   1035 <dt><code>--with-stack-clash-protection-guard-size=<var>size</var></code></dt>
   1036 <dd><p>On certain targets this option sets the default stack clash protection guard
   1037 size as a power of two in bytes.  On AArch64 <var>size</var> is required to be either
   1038 12 (4KB) or 16 (64KB).
   1039 </p>
   1040 </dd>
   1041 <dt><code>--enable-__cxa_atexit</code></dt>
   1042 <dd><p>Define if you want to use __cxa_atexit, rather than atexit, to
   1043 register C++ destructors for local statics and global objects.
   1044 This is essential for fully standards-compliant handling of
   1045 destructors, but requires __cxa_atexit in libc.  This option is currently
   1046 only available on systems with GNU libc.  When enabled, this will cause
   1047 <samp>-fuse-cxa-atexit</samp> to be passed by default.
   1048 </p>
   1049 </dd>
   1050 <dt><code>--enable-gnu-indirect-function</code></dt>
   1051 <dd><p>Define if you want to enable the <code>ifunc</code> attribute.  This option is
   1052 currently only available on systems with GNU libc on certain targets.
   1053 </p>
   1054 </dd>
   1055 <dt><code>--enable-target-optspace</code></dt>
   1056 <dd><p>Specify that target
   1057 libraries should be optimized for code space instead of code speed.
   1058 This is the default for the m32r platform.
   1059 </p>
   1060 </dd>
   1061 <dt><code>--with-cpp-install-dir=<var>dirname</var></code></dt>
   1062 <dd><p>Specify that the user visible <code>cpp</code> program should be installed
   1063 in <samp><var>prefix</var>/<var>dirname</var>/cpp</samp>, in addition to <var>bindir</var>.
   1064 </p>
   1065 </dd>
   1066 <dt><code>--enable-comdat</code></dt>
   1067 <dd><p>Enable COMDAT group support.  This is primarily used to override the
   1068 automatically detected value.
   1069 </p>
   1070 </dd>
   1071 <dt><code>--enable-initfini-array</code></dt>
   1072 <dd><p>Force the use of sections <code>.init_array</code> and <code>.fini_array</code>
   1073 (instead of <code>.init</code> and <code>.fini</code>) for constructors and
   1074 destructors.  Option <samp>--disable-initfini-array</samp> has the
   1075 opposite effect.  If neither option is specified, the configure script
   1076 will try to guess whether the <code>.init_array</code> and
   1077 <code>.fini_array</code> sections are supported and, if they are, use them.
   1078 </p>
   1079 </dd>
   1080 <dt><code>--enable-link-mutex</code></dt>
   1081 <dd><p>When building GCC, use a mutex to avoid linking the compilers for
   1082 multiple languages at the same time, to avoid thrashing on build
   1083 systems with limited free memory.  The default is not to use such a mutex.
   1084 </p>
   1085 </dd>
   1086 <dt><code>--enable-maintainer-mode</code></dt>
   1087 <dd><p>The build rules that regenerate the Autoconf and Automake output files as
   1088 well as the GCC master message catalog <samp>gcc.pot</samp> are normally
   1089 disabled.  This is because it can only be rebuilt if the complete source
   1090 tree is present.  If you have changed the sources and want to rebuild the
   1091 catalog, configuring with <samp>--enable-maintainer-mode</samp> will enable
   1092 this.  Note that you need a recent version of the <code>gettext</code> tools
   1093 to do so.
   1094 </p>
   1095 </dd>
   1096 <dt><code>--disable-bootstrap</code></dt>
   1097 <dd><p>For a native build, the default configuration is to perform
   1098 a 3-stage bootstrap of the compiler when &lsquo;<samp>make</samp>&rsquo; is invoked,
   1099 testing that GCC can compile itself correctly.  If you want to disable
   1100 this process, you can configure with <samp>--disable-bootstrap</samp>.
   1101 </p>
   1102 </dd>
   1103 <dt><code>--enable-bootstrap</code></dt>
   1104 <dd><p>In special cases, you may want to perform a 3-stage build
   1105 even if the target and host triplets are different.
   1106 This is possible when the host can run code compiled for
   1107 the target (e.g. host is i686-linux, target is i486-linux).
   1108 Starting from GCC 4.2, to do this you have to configure explicitly
   1109 with <samp>--enable-bootstrap</samp>.
   1110 </p>
   1111 </dd>
   1112 <dt><code>--enable-generated-files-in-srcdir</code></dt>
   1113 <dd><p>Neither the .c and .h files that are generated from Bison and flex nor the
   1114 info manuals and man pages that are built from the .texi files are present
   1115 in the repository development tree.  When building GCC from that development tree,
   1116 or from one of our snapshots, those generated files are placed in your
   1117 build directory, which allows for the source to be in a readonly
   1118 directory.
   1119 </p>
   1120 <p>If you configure with <samp>--enable-generated-files-in-srcdir</samp> then those
   1121 generated files will go into the source directory.  This is mainly intended
   1122 for generating release or prerelease tarballs of the GCC sources, since it
   1123 is not a requirement that the users of source releases to have flex, Bison,
   1124 or makeinfo.
   1125 </p>
   1126 </dd>
   1127 <dt><code>--enable-version-specific-runtime-libs</code></dt>
   1128 <dd><p>Specify
   1129 that runtime libraries should be installed in the compiler specific
   1130 subdirectory (<samp><var>libdir</var>/gcc</samp>) rather than the usual places.  In
   1131 addition, &lsquo;<samp>libstdc++</samp>&rsquo;&rsquo;s include files will be installed into
   1132 <samp><var>libdir</var></samp> unless you overruled it by using
   1133 <samp>--with-gxx-include-dir=<var>dirname</var></samp>.  Using this option is
   1134 particularly useful if you intend to use several versions of GCC in
   1135 parallel.  The default is &lsquo;<samp>yes</samp>&rsquo; for &lsquo;<samp>libada</samp>&rsquo;, and &lsquo;<samp>no</samp>&rsquo; for
   1136 the remaining libraries.
   1137 </p>
   1138 </dd>
   1139 <dt><code><a name="WithAixSoname"></a>--with-aix-soname=&lsquo;<samp>aix</samp>&rsquo;, &lsquo;<samp>svr4</samp>&rsquo; or &lsquo;<samp>both</samp>&rsquo;</code></dt>
   1140 <dd><p>Traditional AIX shared library versioning (versioned <code>Shared Object</code>
   1141 files as members of unversioned <code>Archive Library</code> files named
   1142 &lsquo;<samp>lib.a</samp>&rsquo;) causes numerous headaches for package managers. However,
   1143 <code>Import Files</code> as members of <code>Archive Library</code> files allow for
   1144 <strong>filename-based versioning</strong> of shared libraries as seen on Linux/SVR4,
   1145 where this is called the &quot;SONAME&quot;. But as they prevent static linking,
   1146 <code>Import Files</code> may be used with <code>Runtime Linking</code> only, where the
   1147 linker does search for &lsquo;<samp>libNAME.so</samp>&rsquo; before &lsquo;<samp>libNAME.a</samp>&rsquo; library
   1148 filenames with the &lsquo;<samp>-lNAME</samp>&rsquo; linker flag.
   1149 </p>
   1150 <a name="AixLdCommand"></a><p>For detailed information please refer to the AIX
   1151 <a href="https://www.ibm.com/support/knowledgecenter/search/%22the%20ld%20command%2C%20also%20called%20the%20linkage%20editor%20or%20binder%22">ld
   1152 Command</a> reference.
   1153 </p>
   1154 <p>As long as shared library creation is enabled, upon:
   1155 </p><dl compact="compact">
   1156 <dt><code>--with-aix-soname=aix</code></dt>
   1157 <dt><code>--with-aix-soname=both</code></dt>
   1158 <dd><p>A (traditional AIX) <code>Shared Archive Library</code> file is created:
   1159  </p><ul>
   1160 <li> using the &lsquo;<samp>libNAME.a</samp>&rsquo; filename scheme
   1161   </li><li> with the <code>Shared Object</code> file as archive member named
   1162   &lsquo;<samp>libNAME.so.V</samp>&rsquo; (except for &lsquo;<samp>libgcc_s</samp>&rsquo;, where the <code>Shared
   1163   Object</code> file is named &lsquo;<samp>shr.o</samp>&rsquo; for backwards compatibility), which
   1164   <ul class="no-bullet">
   1165 <li>- is used for runtime loading from inside the &lsquo;<samp>libNAME.a</samp>&rsquo; file
   1166    </li><li>- is used for dynamic loading via
   1167    <code>dlopen(&quot;libNAME.a(libNAME.so.V)&quot;, RTLD_MEMBER)</code>
   1168    </li><li>- is used for shared linking
   1169    </li><li>- is used for static linking, so no separate <code>Static Archive
   1170    Library</code> file is needed
   1171   </li></ul>
   1172 </li></ul>
   1173 </dd>
   1174 <dt><code>--with-aix-soname=both</code></dt>
   1175 <dt><code>--with-aix-soname=svr4</code></dt>
   1176 <dd><p>A (second) <code>Shared Archive Library</code> file is created:
   1177  </p><ul>
   1178 <li> using the &lsquo;<samp>libNAME.so.V</samp>&rsquo; filename scheme
   1179  </li><li> with the <code>Shared Object</code> file as archive member named
   1180  &lsquo;<samp>shr.o</samp>&rsquo;, which
   1181   <ul class="no-bullet">
   1182 <li>- is created with the <code>-G linker flag</code>
   1183    </li><li>- has the <code>F_LOADONLY</code> flag set
   1184    </li><li>- is used for runtime loading from inside the &lsquo;<samp>libNAME.so.V</samp>&rsquo; file
   1185    </li><li>- is used for dynamic loading via <code>dlopen(&quot;libNAME.so.V(shr.o)&quot;,
   1186    RTLD_MEMBER)</code>
   1187   </li></ul>
   1188 </li><li> with the <code>Import File</code> as archive member named &lsquo;<samp>shr.imp</samp>&rsquo;,
   1189  which
   1190   <ul class="no-bullet">
   1191 <li>- refers to &lsquo;<samp>libNAME.so.V(shr.o)</samp>&rsquo; as the &quot;SONAME&quot;, to be recorded
   1192    in the <code>Loader Section</code> of subsequent binaries
   1193    </li><li>- indicates whether &lsquo;<samp>libNAME.so.V(shr.o)</samp>&rsquo; is 32 or 64 bit
   1194    </li><li>- lists all the public symbols exported by &lsquo;<samp>lib.so.V(shr.o)</samp>&rsquo;,
   1195    eventually decorated with the <code>&lsquo;<samp>weak</samp>&rsquo; Keyword</code>
   1196    </li><li>- is necessary for shared linking against &lsquo;<samp>lib.so.V(shr.o)</samp>&rsquo;
   1197    </li></ul>
   1198 </li></ul>
   1199 <p>A symbolic link using the &lsquo;<samp>libNAME.so</samp>&rsquo; filename scheme is created:
   1200   </p><ul>
   1201 <li> pointing to the &lsquo;<samp>libNAME.so.V</samp>&rsquo; <code>Shared Archive Library</code> file
   1202   </li><li> to permit the <code>ld Command</code> to find &lsquo;<samp>lib.so.V(shr.imp)</samp>&rsquo; via
   1203   the &lsquo;<samp>-lNAME</samp>&rsquo; argument (requires <code>Runtime Linking</code> to be enabled)
   1204   </li><li> to permit dynamic loading of &lsquo;<samp>lib.so.V(shr.o)</samp>&rsquo; without the need
   1205   to specify the version number via <code>dlopen(&quot;libNAME.so(shr.o)&quot;,
   1206   RTLD_MEMBER)</code>
   1207   </li></ul>
   1208 </dd>
   1209 </dl>
   1210 
   1211 <p>As long as static library creation is enabled, upon:
   1212 </p><dl compact="compact">
   1213 <dt><code>--with-aix-soname=svr4</code></dt>
   1214 <dd><p>A <code>Static Archive Library</code> is created:
   1215  </p><ul>
   1216 <li> using the &lsquo;<samp>libNAME.a</samp>&rsquo; filename scheme
   1217  </li><li> with all the <code>Static Object</code> files as archive members, which
   1218   <ul class="no-bullet">
   1219 <li>- are used for static linking
   1220   </li></ul>
   1221 </li></ul>
   1222 </dd>
   1223 </dl>
   1224 
   1225 <p>While the aix-soname=&lsquo;<samp>svr4</samp>&rsquo; option does not create <code>Shared Object</code>
   1226 files as members of unversioned <code>Archive Library</code> files any more, package
   1227 managers still are responsible to
   1228 <a href="./specific.html#TransferAixShobj">transfer</a> <code>Shared Object</code> files
   1229 found as member of a previously installed unversioned <code>Archive Library</code>
   1230 file into the newly installed <code>Archive Library</code> file with the same
   1231 filename.
   1232 </p>
   1233 <p><em>WARNING:</em> Creating <code>Shared Object</code> files with <code>Runtime Linking</code>
   1234 enabled may bloat the TOC, eventually leading to <code>TOC overflow</code> errors,
   1235 requiring the use of either the <samp>-Wl,-bbigtoc</samp> linker flag (seen to
   1236 break with the <code>GDB</code> debugger) or some of the TOC-related compiler flags,
   1237 see &ldquo;RS/6000 and PowerPC Options&rdquo; in the main manual.
   1238 </p>
   1239 <p><samp>--with-aix-soname</samp> is currently supported by &lsquo;<samp>libgcc_s</samp>&rsquo; only, so
   1240 this option is still experimental and not for normal use yet.
   1241 </p>
   1242 <p>Default is the traditional behavior <samp>--with-aix-soname=&lsquo;<samp>aix</samp>&rsquo;</samp>.
   1243 </p>
   1244 </dd>
   1245 <dt><code>--enable-languages=<var>lang1</var>,<var>lang2</var>,&hellip;</code></dt>
   1246 <dd><p>Specify that only a particular subset of compilers and
   1247 their runtime libraries should be built.  For a list of valid values for
   1248 <var>langN</var> you can issue the following command in the
   1249 <samp>gcc</samp> directory of your GCC source tree:<br>
   1250 </p><div class="smallexample">
   1251 <pre class="smallexample">grep ^language= */config-lang.in
   1252 </pre></div>
   1253 <p>Currently, you can use any of the following:
   1254 <code>all</code>, <code>default</code>, <code>ada</code>, <code>c</code>, <code>c++</code>, <code>d</code>,
   1255 <code>fortran</code>, <code>go</code>, <code>jit</code>, <code>lto</code>, <code>objc</code>, <code>obj-c++</code>.
   1256 Building the Ada compiler has special requirements, see below.
   1257 If you do not pass this flag, or specify the option <code>default</code>, then the
   1258 default languages available in the <samp>gcc</samp> sub-tree will be configured.
   1259 Ada, D, Go, Jit, and Objective-C++ are not default languages.  LTO is not a
   1260 default language, but is built by default because <samp>--enable-lto</samp> is
   1261 enabled by default.  The other languages are default languages.  If
   1262 <code>all</code> is specified, then all available languages are built.  An
   1263 exception is <code>jit</code> language, which requires
   1264 <samp>--enable-host-shared</samp> to be included with <code>all</code>.
   1265 </p>
   1266 </dd>
   1267 <dt><code>--enable-stage1-languages=<var>lang1</var>,<var>lang2</var>,&hellip;</code></dt>
   1268 <dd><p>Specify that a particular subset of compilers and their runtime
   1269 libraries should be built with the system C compiler during stage 1 of
   1270 the bootstrap process, rather than only in later stages with the
   1271 bootstrapped C compiler.  The list of valid values is the same as for
   1272 <samp>--enable-languages</samp>, and the option <code>all</code> will select all
   1273 of the languages enabled by <samp>--enable-languages</samp>.  This option is
   1274 primarily useful for GCC development; for instance, when a development
   1275 version of the compiler cannot bootstrap due to compiler bugs, or when
   1276 one is debugging front ends other than the C front end.  When this
   1277 option is used, one can then build the target libraries for the
   1278 specified languages with the stage-1 compiler by using <code>make
   1279 stage1-bubble all-target</code>, or run the testsuite on the stage-1 compiler
   1280 for the specified languages using <code>make stage1-start check-gcc</code>.
   1281 </p>
   1282 </dd>
   1283 <dt><code>--disable-libada</code></dt>
   1284 <dd><p>Specify that the run-time libraries and tools used by GNAT should not
   1285 be built.  This can be useful for debugging, or for compatibility with
   1286 previous Ada build procedures, when it was required to explicitly
   1287 do a &lsquo;<samp>make -C gcc gnatlib_and_tools</samp>&rsquo;.
   1288 </p>
   1289 </dd>
   1290 <dt><code>--disable-libsanitizer</code></dt>
   1291 <dd><p>Specify that the run-time libraries for the various sanitizers should
   1292 not be built.
   1293 </p>
   1294 </dd>
   1295 <dt><code>--disable-libssp</code></dt>
   1296 <dd><p>Specify that the run-time libraries for stack smashing protection
   1297 should not be built or linked against.  On many targets library support
   1298 is provided by the C library instead.
   1299 </p>
   1300 </dd>
   1301 <dt><code>--disable-libquadmath</code></dt>
   1302 <dd><p>Specify that the GCC quad-precision math library should not be built.
   1303 On some systems, the library is required to be linkable when building
   1304 the Fortran front end, unless <samp>--disable-libquadmath-support</samp>
   1305 is used.
   1306 </p>
   1307 </dd>
   1308 <dt><code>--disable-libquadmath-support</code></dt>
   1309 <dd><p>Specify that the Fortran front end and <code>libgfortran</code> do not add
   1310 support for <code>libquadmath</code> on systems supporting it.
   1311 </p>
   1312 </dd>
   1313 <dt><code>--disable-libgomp</code></dt>
   1314 <dd><p>Specify that the GNU Offloading and Multi Processing Runtime Library
   1315 should not be built.
   1316 </p>
   1317 </dd>
   1318 <dt><code>--disable-libvtv</code></dt>
   1319 <dd><p>Specify that the run-time libraries used by vtable verification
   1320 should not be built.
   1321 </p>
   1322 </dd>
   1323 <dt><code>--with-dwarf2</code></dt>
   1324 <dd><p>Specify that the compiler should
   1325 use DWARF 2 debugging information as the default.
   1326 </p>
   1327 </dd>
   1328 <dt><code>--with-advance-toolchain=<var>at</var></code></dt>
   1329 <dd><p>On 64-bit PowerPC Linux systems, configure the compiler to use the
   1330 header files, library files, and the dynamic linker from the Advance
   1331 Toolchain release <var>at</var> instead of the default versions that are
   1332 provided by the Linux distribution.  In general, this option is
   1333 intended for the developers of GCC, and it is not intended for general
   1334 use.
   1335 </p>
   1336 </dd>
   1337 <dt><code>--enable-targets=all</code></dt>
   1338 <dt><code>--enable-targets=<var>target_list</var></code></dt>
   1339 <dd><p>Some GCC targets, e.g. powerpc64-linux, build bi-arch compilers.
   1340 These are compilers that are able to generate either 64-bit or 32-bit
   1341 code.  Typically, the corresponding 32-bit target, e.g.
   1342 powerpc-linux for powerpc64-linux, only generates 32-bit code.  This
   1343 option enables the 32-bit target to be a bi-arch compiler, which is
   1344 useful when you want a bi-arch compiler that defaults to 32-bit, and
   1345 you are building a bi-arch or multi-arch binutils in a combined tree.
   1346 On mips-linux, this will build a tri-arch compiler (ABI o32/n32/64),
   1347 defaulted to o32.
   1348 Currently, this option only affects sparc-linux, powerpc-linux, x86-linux,
   1349 mips-linux and s390-linux.
   1350 </p>
   1351 </dd>
   1352 <dt><code>--enable-default-pie</code></dt>
   1353 <dd><p>Turn on <samp>-fPIE</samp> and <samp>-pie</samp> by default.
   1354 </p>
   1355 </dd>
   1356 <dt><code>--enable-secureplt</code></dt>
   1357 <dd><p>This option enables <samp>-msecure-plt</samp> by default for powerpc-linux.
   1358 See &ldquo;RS/6000 and PowerPC Options&rdquo; in the main manual
   1359 </p>
   1360 </dd>
   1361 <dt><code>--enable-default-ssp</code></dt>
   1362 <dd><p>Turn on <samp>-fstack-protector-strong</samp> by default.
   1363 </p>
   1364 </dd>
   1365 <dt><code>--enable-cld</code></dt>
   1366 <dd><p>This option enables <samp>-mcld</samp> by default for 32-bit x86 targets.
   1367 See &ldquo;i386 and x86-64 Options&rdquo; in the main manual
   1368 </p>
   1369 </dd>
   1370 <dt><code>--enable-large-address-aware</code></dt>
   1371 <dd><p>The <samp>--enable-large-address-aware</samp> option arranges for MinGW
   1372 executables to be linked using the <samp>--large-address-aware</samp>
   1373 option, that enables the use of more than 2GB of memory.  If GCC is
   1374 configured with this option, its effects can be reversed by passing the
   1375 <samp>-Wl,--disable-large-address-aware</samp> option to the so-configured
   1376 compiler driver.
   1377 </p>
   1378 </dd>
   1379 <dt><code>--enable-win32-registry</code></dt>
   1380 <dt><code>--enable-win32-registry=<var>key</var></code></dt>
   1381 <dt><code>--disable-win32-registry</code></dt>
   1382 <dd><p>The <samp>--enable-win32-registry</samp> option enables Microsoft Windows-hosted GCC
   1383 to look up installations paths in the registry using the following key:
   1384 </p>
   1385 <div class="smallexample">
   1386 <pre class="smallexample"><code>HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\<var>key</var></code>
   1387 </pre></div>
   1388 
   1389 <p><var>key</var> defaults to GCC version number, and can be overridden by the
   1390 <samp>--enable-win32-registry=<var>key</var></samp> option.  Vendors and distributors
   1391 who use custom installers are encouraged to provide a different key,
   1392 perhaps one comprised of vendor name and GCC version number, to
   1393 avoid conflict with existing installations.  This feature is enabled
   1394 by default, and can be disabled by <samp>--disable-win32-registry</samp>
   1395 option.  This option has no effect on the other hosts.
   1396 </p>
   1397 </dd>
   1398 <dt><code>--nfp</code></dt>
   1399 <dd><p>Specify that the machine does not have a floating point unit.  This
   1400 option only applies to &lsquo;<samp>m68k-sun-sunos<var>n</var></samp>&rsquo;.  On any other
   1401 system, <samp>--nfp</samp> has no effect.
   1402 </p>
   1403 </dd>
   1404 <dt><code>--enable-werror</code></dt>
   1405 <dt><code>--disable-werror</code></dt>
   1406 <dt><code>--enable-werror=yes</code></dt>
   1407 <dt><code>--enable-werror=no</code></dt>
   1408 <dd><p>When you specify this option, it controls whether certain files in the
   1409 compiler are built with <samp>-Werror</samp> in bootstrap stage2 and later.
   1410 If you don&rsquo;t specify it, <samp>-Werror</samp> is turned on for the main
   1411 development trunk.  However it defaults to off for release branches and
   1412 final releases.  The specific files which get <samp>-Werror</samp> are
   1413 controlled by the Makefiles.
   1414 </p>
   1415 </dd>
   1416 <dt><code>--enable-checking</code></dt>
   1417 <dt><code>--disable-checking</code></dt>
   1418 <dt><code>--enable-checking=<var>list</var></code></dt>
   1419 <dd><p>This option controls performing internal consistency checks in the compiler.
   1420 It does not change the generated code, but adds error checking of the
   1421 requested complexity.  This slows down the compiler and may only work
   1422 properly if you are building the compiler with GCC.
   1423 </p>
   1424 <p>When the option is not specified, the active set of checks depends on context.
   1425 Namely, bootstrap stage 1 defaults to &lsquo;<samp>--enable-checking=yes</samp>&rsquo;, builds
   1426 from release branches or release archives default to
   1427 &lsquo;<samp>--enable-checking=release</samp>&rsquo;, and otherwise
   1428 &lsquo;<samp>--enable-checking=yes,extra</samp>&rsquo; is used.  When the option is
   1429 specified without a <var>list</var>, the result is the same as
   1430 &lsquo;<samp>--enable-checking=yes</samp>&rsquo;.  Likewise, &lsquo;<samp>--disable-checking</samp>&rsquo; is
   1431 equivalent to &lsquo;<samp>--enable-checking=no</samp>&rsquo;.
   1432 </p>
   1433 <p>The categories of checks available in <var>list</var> are &lsquo;<samp>yes</samp>&rsquo; (most common
   1434 checks &lsquo;<samp>assert,misc,gc,gimple,rtlflag,runtime,tree,types</samp>&rsquo;), &lsquo;<samp>no</samp>&rsquo;
   1435 (no checks at all), &lsquo;<samp>all</samp>&rsquo; (all but &lsquo;<samp>valgrind</samp>&rsquo;), &lsquo;<samp>release</samp>&rsquo;
   1436 (cheapest checks &lsquo;<samp>assert,runtime</samp>&rsquo;) or &lsquo;<samp>none</samp>&rsquo; (same as &lsquo;<samp>no</samp>&rsquo;).
   1437 &lsquo;<samp>release</samp>&rsquo; checks are always on and to disable them
   1438 &lsquo;<samp>--disable-checking</samp>&rsquo; or &lsquo;<samp>--enable-checking=no[,&lt;other checks&gt;]</samp>&rsquo;
   1439 must be explicitly requested.  Disabling assertions makes the compiler and
   1440 runtime slightly faster but increases the risk of undetected internal errors
   1441 causing wrong code to be generated.
   1442 </p>
   1443 <p>Individual checks can be enabled with these flags: &lsquo;<samp>assert</samp>&rsquo;, &lsquo;<samp>df</samp>&rsquo;,
   1444 &lsquo;<samp>extra</samp>&rsquo;, &lsquo;<samp>fold</samp>&rsquo;, &lsquo;<samp>gc</samp>&rsquo;, &lsquo;<samp>gcac</samp>&rsquo;, &lsquo;<samp>gimple</samp>&rsquo;,
   1445 &lsquo;<samp>misc</samp>&rsquo;, &lsquo;<samp>rtl</samp>&rsquo;, &lsquo;<samp>rtlflag</samp>&rsquo;, &lsquo;<samp>runtime</samp>&rsquo;, &lsquo;<samp>tree</samp>&rsquo;,
   1446 &lsquo;<samp>types</samp>&rsquo; and &lsquo;<samp>valgrind</samp>&rsquo;.  &lsquo;<samp>extra</samp>&rsquo; extends &lsquo;<samp>misc</samp>&rsquo;
   1447 checking with extra checks that might affect code generation and should
   1448 therefore not differ between stage1 and later stages in bootstrap.
   1449 </p>
   1450 <p>The &lsquo;<samp>valgrind</samp>&rsquo; check requires the external <code>valgrind</code> simulator,
   1451 available from <a href="http://valgrind.org/">http://valgrind.org/</a>.  The &lsquo;<samp>rtl</samp>&rsquo; checks are
   1452 expensive and the &lsquo;<samp>df</samp>&rsquo;, &lsquo;<samp>gcac</samp>&rsquo; and &lsquo;<samp>valgrind</samp>&rsquo; checks are very
   1453 expensive.
   1454 </p>
   1455 </dd>
   1456 <dt><code>--disable-stage1-checking</code></dt>
   1457 <dt><code>--enable-stage1-checking</code></dt>
   1458 <dt><code>--enable-stage1-checking=<var>list</var></code></dt>
   1459 <dd><p>This option affects only bootstrap build.  If no <samp>--enable-checking</samp>
   1460 option is specified the stage1 compiler is built with &lsquo;<samp>yes</samp>&rsquo; checking
   1461 enabled, otherwise the stage1 checking flags are the same as specified by
   1462 <samp>--enable-checking</samp>.  To build the stage1 compiler with
   1463 different checking options use <samp>--enable-stage1-checking</samp>.
   1464 The list of checking options is the same as for <samp>--enable-checking</samp>.
   1465 If your system is too slow or too small to bootstrap a released compiler
   1466 with checking for stage1 enabled, you can use &lsquo;<samp>--disable-stage1-checking</samp>&rsquo;
   1467 to disable checking for the stage1 compiler.
   1468 </p>
   1469 </dd>
   1470 <dt><code>--enable-coverage</code></dt>
   1471 <dt><code>--enable-coverage=<var>level</var></code></dt>
   1472 <dd><p>With this option, the compiler is built to collect self coverage
   1473 information, every time it is run.  This is for internal development
   1474 purposes, and only works when the compiler is being built with gcc.  The
   1475 <var>level</var> argument controls whether the compiler is built optimized or
   1476 not, values are &lsquo;<samp>opt</samp>&rsquo; and &lsquo;<samp>noopt</samp>&rsquo;.  For coverage analysis you
   1477 want to disable optimization, for performance analysis you want to
   1478 enable optimization.  When coverage is enabled, the default level is
   1479 without optimization.
   1480 </p>
   1481 </dd>
   1482 <dt><code>--enable-gather-detailed-mem-stats</code></dt>
   1483 <dd><p>When this option is specified more detailed information on memory
   1484 allocation is gathered.  This information is printed when using
   1485 <samp>-fmem-report</samp>.
   1486 </p>
   1487 </dd>
   1488 <dt><code>--enable-valgrind-annotations</code></dt>
   1489 <dd><p>Mark selected memory related operations in the compiler when run under
   1490 valgrind to suppress false positives.
   1491 </p>
   1492 </dd>
   1493 <dt><code>--enable-nls</code></dt>
   1494 <dt><code>--disable-nls</code></dt>
   1495 <dd><p>The <samp>--enable-nls</samp> option enables Native Language Support (NLS),
   1496 which lets GCC output diagnostics in languages other than American
   1497 English.  Native Language Support is enabled by default if not doing a
   1498 canadian cross build.  The <samp>--disable-nls</samp> option disables NLS.
   1499 </p>
   1500 </dd>
   1501 <dt><code>--with-included-gettext</code></dt>
   1502 <dd><p>If NLS is enabled, the <samp>--with-included-gettext</samp> option causes the build
   1503 procedure to prefer its copy of GNU <code>gettext</code>.
   1504 </p>
   1505 </dd>
   1506 <dt><code>--with-catgets</code></dt>
   1507 <dd><p>If NLS is enabled, and if the host lacks <code>gettext</code> but has the
   1508 inferior <code>catgets</code> interface, the GCC build procedure normally
   1509 ignores <code>catgets</code> and instead uses GCC&rsquo;s copy of the GNU
   1510 <code>gettext</code> library.  The <samp>--with-catgets</samp> option causes the
   1511 build procedure to use the host&rsquo;s <code>catgets</code> in this situation.
   1512 </p>
   1513 </dd>
   1514 <dt><code>--with-libiconv-prefix=<var>dir</var></code></dt>
   1515 <dd><p>Search for libiconv header files in <samp><var>dir</var>/include</samp> and
   1516 libiconv library files in <samp><var>dir</var>/lib</samp>.
   1517 </p>
   1518 </dd>
   1519 <dt><code>--enable-obsolete</code></dt>
   1520 <dd><p>Enable configuration for an obsoleted system.  If you attempt to
   1521 configure GCC for a system (build, host, or target) which has been
   1522 obsoleted, and you do not specify this flag, configure will halt with an
   1523 error message.
   1524 </p>
   1525 <p>All support for systems which have been obsoleted in one release of GCC
   1526 is removed entirely in the next major release, unless someone steps
   1527 forward to maintain the port.
   1528 </p>
   1529 </dd>
   1530 <dt><code>--enable-decimal-float</code></dt>
   1531 <dt><code>--enable-decimal-float=yes</code></dt>
   1532 <dt><code>--enable-decimal-float=no</code></dt>
   1533 <dt><code>--enable-decimal-float=bid</code></dt>
   1534 <dt><code>--enable-decimal-float=dpd</code></dt>
   1535 <dt><code>--disable-decimal-float</code></dt>
   1536 <dd><p>Enable (or disable) support for the C decimal floating point extension
   1537 that is in the IEEE 754-2008 standard.  This is enabled by default only
   1538 on PowerPC, i386, and x86_64 GNU/Linux systems.  Other systems may also
   1539 support it, but require the user to specifically enable it.  You can
   1540 optionally control which decimal floating point format is used (either
   1541 &lsquo;<samp>bid</samp>&rsquo; or &lsquo;<samp>dpd</samp>&rsquo;).  The &lsquo;<samp>bid</samp>&rsquo; (binary integer decimal)
   1542 format is default on i386 and x86_64 systems, and the &lsquo;<samp>dpd</samp>&rsquo;
   1543 (densely packed decimal) format is default on PowerPC systems.
   1544 </p>
   1545 </dd>
   1546 <dt><code>--enable-fixed-point</code></dt>
   1547 <dt><code>--disable-fixed-point</code></dt>
   1548 <dd><p>Enable (or disable) support for C fixed-point arithmetic.
   1549 This option is enabled by default for some targets (such as MIPS) which
   1550 have hardware-support for fixed-point operations.  On other targets, you
   1551 may enable this option manually.
   1552 </p>
   1553 </dd>
   1554 <dt><code>--with-long-double-128</code></dt>
   1555 <dd><p>Specify if <code>long double</code> type should be 128-bit by default on selected
   1556 GNU/Linux architectures.  If using <code>--without-long-double-128</code>,
   1557 <code>long double</code> will be by default 64-bit, the same as <code>double</code> type.
   1558 When neither of these configure options are used, the default will be
   1559 128-bit <code>long double</code> when built against GNU C Library 2.4 and later,
   1560 64-bit <code>long double</code> otherwise.
   1561 </p>
   1562 </dd>
   1563 <dt><code>--with-long-double-format=ibm</code></dt>
   1564 <dt><code>--with-long-double-format=ieee</code></dt>
   1565 <dd><p>Specify whether <code>long double</code> uses the IBM extended double format
   1566 or the IEEE 128-bit floating point format on PowerPC Linux systems.
   1567 This configuration switch will only work on little endian PowerPC
   1568 Linux systems and on big endian 64-bit systems where the default cpu
   1569 is at least power7 (i.e. <samp>--with-cpu=power7</samp>,
   1570 <samp>--with-cpu=power8</samp>, or <samp>--with-cpu=power9</samp> is used).
   1571 </p>
   1572 <p>If you use the <samp>--with-long-double-64</samp> configuration option,
   1573 the <samp>--with-long-double-format=ibm</samp> and
   1574 <samp>--with-long-double-format=ieee</samp> options are ignored.
   1575 </p>
   1576 <p>The default <code>long double</code> format is to use IBM extended double.
   1577 Until all of the libraries are converted to use IEEE 128-bit floating
   1578 point, it is not recommended to use
   1579 <samp>--with-long-double-format=ieee</samp>.
   1580 </p>
   1581 <p>On little endian PowerPC Linux systems, if you explicitly set the
   1582 <code>long double</code> type, it will build multilibs to allow you to
   1583 select either <code>long double</code> format, unless you disable multilibs
   1584 with the <code>--disable-multilib</code> option.  At present,
   1585 <code>long double</code> multilibs are not built on big endian PowerPC Linux
   1586 systems.  If you are building multilibs, you will need to configure
   1587 the compiler using the <samp>--with-system-zlib</samp> option.
   1588 </p>
   1589 <p>If you do not set the <code>long double</code> type explicitly, no multilibs
   1590 will be generated.
   1591 </p>
   1592 </dd>
   1593 <dt><code>--enable-fdpic</code></dt>
   1594 <dd><p>On SH Linux systems, generate ELF FDPIC code.
   1595 </p>
   1596 </dd>
   1597 <dt><code>--with-gmp=<var>pathname</var></code></dt>
   1598 <dt><code>--with-gmp-include=<var>pathname</var></code></dt>
   1599 <dt><code>--with-gmp-lib=<var>pathname</var></code></dt>
   1600 <dt><code>--with-mpfr=<var>pathname</var></code></dt>
   1601 <dt><code>--with-mpfr-include=<var>pathname</var></code></dt>
   1602 <dt><code>--with-mpfr-lib=<var>pathname</var></code></dt>
   1603 <dt><code>--with-mpc=<var>pathname</var></code></dt>
   1604 <dt><code>--with-mpc-include=<var>pathname</var></code></dt>
   1605 <dt><code>--with-mpc-lib=<var>pathname</var></code></dt>
   1606 <dd><p>If you want to build GCC but do not have the GMP library, the MPFR
   1607 library and/or the MPC library installed in a standard location and
   1608 do not have their sources present in the GCC source tree then you
   1609 can explicitly specify the directory where they are installed
   1610 (&lsquo;<samp>--with-gmp=<var>gmpinstalldir</var></samp>&rsquo;,
   1611 &lsquo;<samp>--with-mpfr=<var>mpfrinstalldir</var></samp>&rsquo;,
   1612 &lsquo;<samp>--with-mpc=<var>mpcinstalldir</var></samp>&rsquo;).  The
   1613 <samp>--with-gmp=<var>gmpinstalldir</var></samp> option is shorthand for
   1614 <samp>--with-gmp-lib=<var>gmpinstalldir</var>/lib</samp> and
   1615 <samp>--with-gmp-include=<var>gmpinstalldir</var>/include</samp>.  Likewise the
   1616 <samp>--with-mpfr=<var>mpfrinstalldir</var></samp> option is shorthand for
   1617 <samp>--with-mpfr-lib=<var>mpfrinstalldir</var>/lib</samp> and
   1618 <samp>--with-mpfr-include=<var>mpfrinstalldir</var>/include</samp>, also the
   1619 <samp>--with-mpc=<var>mpcinstalldir</var></samp> option is shorthand for
   1620 <samp>--with-mpc-lib=<var>mpcinstalldir</var>/lib</samp> and
   1621 <samp>--with-mpc-include=<var>mpcinstalldir</var>/include</samp>.  If these
   1622 shorthand assumptions are not correct, you can use the explicit
   1623 include and lib options directly.  You might also need to ensure the
   1624 shared libraries can be found by the dynamic linker when building and
   1625 using GCC, for example by setting the runtime shared library path
   1626 variable (<code>LD_LIBRARY_PATH</code> on GNU/Linux and Solaris systems).
   1627 </p>
   1628 <p>These flags are applicable to the host platform only.  When building
   1629 a cross compiler, they will not be used to configure target libraries.
   1630 </p>
   1631 </dd>
   1632 <dt><code>--with-isl=<var>pathname</var></code></dt>
   1633 <dt><code>--with-isl-include=<var>pathname</var></code></dt>
   1634 <dt><code>--with-isl-lib=<var>pathname</var></code></dt>
   1635 <dd><p>If you do not have the isl library installed in a standard location and you
   1636 want to build GCC, you can explicitly specify the directory where it is
   1637 installed (&lsquo;<samp>--with-isl=<var>islinstalldir</var></samp>&rsquo;). The
   1638 <samp>--with-isl=<var>islinstalldir</var></samp> option is shorthand for
   1639 <samp>--with-isl-lib=<var>islinstalldir</var>/lib</samp> and
   1640 <samp>--with-isl-include=<var>islinstalldir</var>/include</samp>. If this
   1641 shorthand assumption is not correct, you can use the explicit
   1642 include and lib options directly.
   1643 </p>
   1644 <p>These flags are applicable to the host platform only.  When building
   1645 a cross compiler, they will not be used to configure target libraries.
   1646 </p>
   1647 </dd>
   1648 <dt><code>--with-stage1-ldflags=<var>flags</var></code></dt>
   1649 <dd><p>This option may be used to set linker flags to be used when linking
   1650 stage 1 of GCC.  These are also used when linking GCC if configured with
   1651 <samp>--disable-bootstrap</samp>.  If <samp>--with-stage1-libs</samp> is not set to a
   1652 value, then the default is &lsquo;<samp>-static-libstdc++ -static-libgcc</samp>&rsquo;, if
   1653 supported.
   1654 </p>
   1655 </dd>
   1656 <dt><code>--with-stage1-libs=<var>libs</var></code></dt>
   1657 <dd><p>This option may be used to set libraries to be used when linking stage 1
   1658 of GCC.  These are also used when linking GCC if configured with
   1659 <samp>--disable-bootstrap</samp>.
   1660 </p>
   1661 </dd>
   1662 <dt><code>--with-boot-ldflags=<var>flags</var></code></dt>
   1663 <dd><p>This option may be used to set linker flags to be used when linking
   1664 stage 2 and later when bootstrapping GCC.  If &ndash;with-boot-libs
   1665 is not is set to a value, then the default is
   1666 &lsquo;<samp>-static-libstdc++ -static-libgcc</samp>&rsquo;.
   1667 </p>
   1668 </dd>
   1669 <dt><code>--with-boot-libs=<var>libs</var></code></dt>
   1670 <dd><p>This option may be used to set libraries to be used when linking stage 2
   1671 and later when bootstrapping GCC.
   1672 </p>
   1673 </dd>
   1674 <dt><code>--with-debug-prefix-map=<var>map</var></code></dt>
   1675 <dd><p>Convert source directory names using <samp>-fdebug-prefix-map</samp> when
   1676 building runtime libraries.  &lsquo;<samp><var>map</var></samp>&rsquo; is a space-separated
   1677 list of maps of the form &lsquo;<samp><var>old</var>=<var>new</var></samp>&rsquo;.
   1678 </p>
   1679 </dd>
   1680 <dt><code>--enable-linker-build-id</code></dt>
   1681 <dd><p>Tells GCC to pass <samp>--build-id</samp> option to the linker for all final
   1682 links (links performed without the <samp>-r</samp> or <samp>--relocatable</samp>
   1683 option), if the linker supports it.  If you specify
   1684 <samp>--enable-linker-build-id</samp>, but your linker does not
   1685 support <samp>--build-id</samp> option, a warning is issued and the
   1686 <samp>--enable-linker-build-id</samp> option is ignored.  The default is off.
   1687 </p>
   1688 </dd>
   1689 <dt><code>--with-linker-hash-style=<var>choice</var></code></dt>
   1690 <dd><p>Tells GCC to pass <samp>--hash-style=<var>choice</var></samp> option to the
   1691 linker for all final links. <var>choice</var> can be one of
   1692 &lsquo;<samp>sysv</samp>&rsquo;, &lsquo;<samp>gnu</samp>&rsquo;, and &lsquo;<samp>both</samp>&rsquo; where &lsquo;<samp>sysv</samp>&rsquo; is the default.
   1693 </p>
   1694 </dd>
   1695 <dt><code>--enable-gnu-unique-object</code></dt>
   1696 <dt><code>--disable-gnu-unique-object</code></dt>
   1697 <dd><p>Tells GCC to use the gnu_unique_object relocation for C++ template
   1698 static data members and inline function local statics.  Enabled by
   1699 default for a toolchain with an assembler that accepts it and
   1700 GLIBC 2.11 or above, otherwise disabled.
   1701 </p>
   1702 </dd>
   1703 <dt><code>--with-diagnostics-color=<var>choice</var></code></dt>
   1704 <dd><p>Tells GCC to use <var>choice</var> as the default for <samp>-fdiagnostics-color=</samp>
   1705 option (if not used explicitly on the command line).  <var>choice</var>
   1706 can be one of &lsquo;<samp>never</samp>&rsquo;, &lsquo;<samp>auto</samp>&rsquo;, &lsquo;<samp>always</samp>&rsquo;, and &lsquo;<samp>auto-if-env</samp>&rsquo;
   1707 where &lsquo;<samp>auto</samp>&rsquo; is the default.  &lsquo;<samp>auto-if-env</samp>&rsquo; makes
   1708 <samp>-fdiagnostics-color=auto</samp> the default if <code>GCC_COLORS</code>
   1709 is present and non-empty in the environment of the compiler, and
   1710 <samp>-fdiagnostics-color=never</samp> otherwise.
   1711 </p>
   1712 </dd>
   1713 <dt><code>--with-diagnostics-urls=<var>choice</var></code></dt>
   1714 <dd><p>Tells GCC to use <var>choice</var> as the default for <samp>-fdiagnostics-urls=</samp>
   1715 option (if not used explicitly on the command line).  <var>choice</var>
   1716 can be one of &lsquo;<samp>never</samp>&rsquo;, &lsquo;<samp>auto</samp>&rsquo;, &lsquo;<samp>always</samp>&rsquo;, and &lsquo;<samp>auto-if-env</samp>&rsquo;
   1717 where &lsquo;<samp>auto</samp>&rsquo; is the default.  &lsquo;<samp>auto-if-env</samp>&rsquo; makes
   1718 <samp>-fdiagnostics-urls=auto</samp> the default if <code>GCC_URLS</code>
   1719 or <code>TERM_URLS</code> is present and non-empty in the environment of the
   1720 compiler, and <samp>-fdiagnostics-urls=never</samp> otherwise.
   1721 </p>
   1722 </dd>
   1723 <dt><code>--enable-lto</code></dt>
   1724 <dt><code>--disable-lto</code></dt>
   1725 <dd><p>Enable support for link-time optimization (LTO).  This is enabled by
   1726 default, and may be disabled using <samp>--disable-lto</samp>.
   1727 </p>
   1728 </dd>
   1729 <dt><code>--enable-linker-plugin-configure-flags=FLAGS</code></dt>
   1730 <dt><code>--enable-linker-plugin-flags=FLAGS</code></dt>
   1731 <dd><p>By default, linker plugins (such as the LTO plugin) are built for the
   1732 host system architecture.  For the case that the linker has a
   1733 different (but run-time compatible) architecture, these flags can be
   1734 specified to build plugins that are compatible to the linker.  For
   1735 example, if you are building GCC for a 64-bit x86_64
   1736 (&lsquo;<samp>x86_64-pc-linux-gnu</samp>&rsquo;) host system, but have a 32-bit x86
   1737 GNU/Linux (&lsquo;<samp>i686-pc-linux-gnu</samp>&rsquo;) linker executable (which is
   1738 executable on the former system), you can configure GCC as follows for
   1739 getting compatible linker plugins:
   1740 </p>
   1741 <div class="smallexample">
   1742 <pre class="smallexample">% <var>srcdir</var>/configure \
   1743     --host=x86_64-pc-linux-gnu \
   1744     --enable-linker-plugin-configure-flags=--host=i686-pc-linux-gnu \
   1745     --enable-linker-plugin-flags='CC=gcc\ -m32\ -Wl,-rpath,[...]/i686-pc-linux-gnu/lib'
   1746 </pre></div>
   1747 
   1748 </dd>
   1749 <dt><code>--with-plugin-ld=<var>pathname</var></code></dt>
   1750 <dd><p>Enable an alternate linker to be used at link-time optimization (LTO)
   1751 link time when <samp>-fuse-linker-plugin</samp> is enabled.
   1752 This linker should have plugin support such as gold starting with
   1753 version 2.20 or GNU ld starting with version 2.21.
   1754 See <samp>-fuse-linker-plugin</samp> for details.
   1755 </p>
   1756 </dd>
   1757 <dt><code>--enable-canonical-system-headers</code></dt>
   1758 <dt><code>--disable-canonical-system-headers</code></dt>
   1759 <dd><p>Enable system header path canonicalization for <samp>libcpp</samp>.  This can
   1760 produce shorter header file paths in diagnostics and dependency output
   1761 files, but these changed header paths may conflict with some compilation
   1762 environments.  Enabled by default, and may be disabled using
   1763 <samp>--disable-canonical-system-headers</samp>.
   1764 </p>
   1765 </dd>
   1766 <dt><code>--with-glibc-version=<var>major</var>.<var>minor</var></code></dt>
   1767 <dd><p>Tell GCC that when the GNU C Library (glibc) is used on the target it
   1768 will be version <var>major</var>.<var>minor</var> or later.  Normally this can
   1769 be detected from the C library&rsquo;s header files, but this option may be
   1770 needed when bootstrapping a cross toolchain without the header files
   1771 available for building the initial bootstrap compiler.
   1772 </p>
   1773 <p>If GCC is configured with some multilibs that use glibc and some that
   1774 do not, this option applies only to the multilibs that use glibc.
   1775 However, such configurations may not work well as not all the relevant
   1776 configuration in GCC is on a per-multilib basis.
   1777 </p>
   1778 </dd>
   1779 <dt><code>--enable-as-accelerator-for=<var>target</var></code></dt>
   1780 <dd><p>Build as offload target compiler. Specify offload host triple by <var>target</var>.
   1781 </p>
   1782 </dd>
   1783 <dt><code>--enable-offload-targets=<var>target1</var>[=<var>path1</var>],&hellip;,<var>targetN</var>[=<var>pathN</var>]</code></dt>
   1784 <dd><p>Enable offloading to targets <var>target1</var>, &hellip;, <var>targetN</var>.
   1785 Offload compilers are expected to be already installed.  Default search
   1786 path for them is <samp><var>exec-prefix</var></samp>, but it can be changed by
   1787 specifying paths <var>path1</var>, &hellip;, <var>pathN</var>.
   1788 </p>
   1789 <div class="smallexample">
   1790 <pre class="smallexample">% <var>srcdir</var>/configure \
   1791     --enable-offload-targets=x86_64-intelmicemul-linux-gnu=/path/to/x86_64/compiler,nvptx-none,hsa
   1792 </pre></div>
   1793 
   1794 <p>If &lsquo;<samp>hsa</samp>&rsquo; is specified as one of the targets, the compiler will be
   1795 built with support for HSA GPU accelerators.  Because the same
   1796 compiler will emit the accelerator code, no path should be specified.
   1797 </p>
   1798 </dd>
   1799 <dt><code>--with-hsa-runtime=<var>pathname</var></code></dt>
   1800 <dt><code>--with-hsa-runtime-include=<var>pathname</var></code></dt>
   1801 <dt><code>--with-hsa-runtime-lib=<var>pathname</var></code></dt>
   1802 <dd>
   1803 <p>If you configure GCC with HSA offloading but do not have the HSA
   1804 run-time library installed in a standard location then you can
   1805 explicitly specify the directory where they are installed.  The
   1806 <samp>--with-hsa-runtime=<var>hsainstalldir</var></samp> option is a
   1807 shorthand for
   1808 <samp>--with-hsa-runtime-lib=<var>hsainstalldir</var>/lib</samp> and
   1809 <samp>--with-hsa-runtime-include=<var>hsainstalldir</var>/include</samp>.
   1810 </p>
   1811 </dd>
   1812 <dt><code>--enable-cet</code></dt>
   1813 <dt><code>--disable-cet</code></dt>
   1814 <dd><p>Enable building target run-time libraries with control-flow
   1815 instrumentation, see <samp>-fcf-protection</samp> option.  When
   1816 <code>--enable-cet</code> is specified target libraries are configured
   1817 to add <samp>-fcf-protection</samp> and, if needed, other target
   1818 specific options to a set of building options.
   1819 </p>
   1820 <p>The option is disabled by default.  When <code>--enable-cet=auto</code>
   1821 is used, it is enabled on Linux/x86 if target binutils
   1822 supports <code>Intel CET</code> instructions and disabled otherwise.
   1823 In this case the target libraries are configured to get additional
   1824 <samp>-fcf-protection</samp> option.
   1825 </p>
   1826 </dd>
   1827 <dt><code>--with-riscv-attribute=&lsquo;<samp>yes</samp>&rsquo;, &lsquo;<samp>no</samp>&rsquo; or &lsquo;<samp>default</samp>&rsquo;</code></dt>
   1828 <dd><p>Generate RISC-V attribute by default, in order to record extra build
   1829 information in object.
   1830 </p>
   1831 <p>The option is disabled by default. It is enabled on RISC-V/ELF (bare-metal)
   1832 target if target binutils supported.
   1833 </p></dd>
   1834 </dl>
   1835 
   1836 <a name="Cross-Compiler-Specific-Options"></a>
   1837 <h4 class="subheading">Cross-Compiler-Specific Options</h4>
   1838 <p>The following options only apply to building cross compilers.
   1839 </p>
   1840 <dl compact="compact">
   1841 <dt><code>--with-toolexeclibdir=<var>dir</var></code></dt>
   1842 <dd><p>Specify the installation directory for libraries built with a cross compiler.
   1843 The default is <samp>${gcc_tooldir}/lib</samp>.
   1844 </p>
   1845 </dd>
   1846 <dt><code>--with-sysroot</code></dt>
   1847 <dt><code>--with-sysroot=<var>dir</var></code></dt>
   1848 <dd><p>Tells GCC to consider <var>dir</var> as the root of a tree that contains
   1849 (a subset of) the root filesystem of the target operating system.
   1850 Target system headers, libraries and run-time object files will be
   1851 searched for in there.  More specifically, this acts as if
   1852 <samp>--sysroot=<var>dir</var></samp> was added to the default options of the built
   1853 compiler.  The specified directory is not copied into the
   1854 install tree, unlike the options <samp>--with-headers</samp> and
   1855 <samp>--with-libs</samp> that this option obsoletes.  The default value,
   1856 in case <samp>--with-sysroot</samp> is not given an argument, is
   1857 <samp>${gcc_tooldir}/sys-root</samp>.  If the specified directory is a
   1858 subdirectory of <samp>${exec_prefix}</samp>, then it will be found relative to
   1859 the GCC binaries if the installation tree is moved.
   1860 </p>
   1861 <p>This option affects the system root for the compiler used to build
   1862 target libraries (which runs on the build system) and the compiler newly
   1863 installed with <code>make install</code>; it does not affect the compiler which is
   1864 used to build GCC itself.
   1865 </p>
   1866 <p>If you specify the <samp>--with-native-system-header-dir=<var>dirname</var></samp>
   1867 option then the compiler will search that directory within <var>dirname</var> for
   1868 native system headers rather than the default <samp>/usr/include</samp>.
   1869 </p>
   1870 </dd>
   1871 <dt><code>--with-build-sysroot</code></dt>
   1872 <dt><code>--with-build-sysroot=<var>dir</var></code></dt>
   1873 <dd><p>Tells GCC to consider <var>dir</var> as the system root (see
   1874 <samp>--with-sysroot</samp>) while building target libraries, instead of
   1875 the directory specified with <samp>--with-sysroot</samp>.  This option is
   1876 only useful when you are already using <samp>--with-sysroot</samp>.  You
   1877 can use <samp>--with-build-sysroot</samp> when you are configuring with
   1878 <samp>--prefix</samp> set to a directory that is different from the one in
   1879 which you are installing GCC and your target libraries.
   1880 </p>
   1881 <p>This option affects the system root for the compiler used to build
   1882 target libraries (which runs on the build system); it does not affect
   1883 the compiler which is used to build GCC itself.
   1884 </p>
   1885 <p>If you specify the <samp>--with-native-system-header-dir=<var>dirname</var></samp>
   1886 option then the compiler will search that directory within <var>dirname</var> for
   1887 native system headers rather than the default <samp>/usr/include</samp>.
   1888 </p>
   1889 </dd>
   1890 <dt><code>--with-headers</code></dt>
   1891 <dt><code>--with-headers=<var>dir</var></code></dt>
   1892 <dd><p>Deprecated in favor of <samp>--with-sysroot</samp>.
   1893 Specifies that target headers are available when building a cross compiler.
   1894 The <var>dir</var> argument specifies a directory which has the target include
   1895 files.  These include files will be copied into the <samp>gcc</samp> install
   1896 directory.  <em>This option with the <var>dir</var> argument is required</em> when
   1897 building a cross compiler, if <samp><var>prefix</var>/<var>target</var>/sys-include</samp>
   1898 doesn&rsquo;t pre-exist.  If <samp><var>prefix</var>/<var>target</var>/sys-include</samp> does
   1899 pre-exist, the <var>dir</var> argument may be omitted.  <code>fixincludes</code>
   1900 will be run on these files to make them compatible with GCC.
   1901 </p>
   1902 </dd>
   1903 <dt><code>--without-headers</code></dt>
   1904 <dd><p>Tells GCC not use any target headers from a libc when building a cross
   1905 compiler.  When crossing to GNU/Linux, you need the headers so GCC
   1906 can build the exception handling for libgcc.
   1907 </p>
   1908 </dd>
   1909 <dt><code>--with-libs</code></dt>
   1910 <dt><code>--with-libs=&quot;<var>dir1</var> <var>dir2</var> &hellip; <var>dirN</var>&quot;</code></dt>
   1911 <dd><p>Deprecated in favor of <samp>--with-sysroot</samp>.
   1912 Specifies a list of directories which contain the target runtime
   1913 libraries.  These libraries will be copied into the <samp>gcc</samp> install
   1914 directory.  If the directory list is omitted, this option has no
   1915 effect.
   1916 </p>
   1917 </dd>
   1918 <dt><code>--with-newlib</code></dt>
   1919 <dd><p>Specifies that &lsquo;<samp>newlib</samp>&rsquo; is
   1920 being used as the target C library.  This causes <code>__eprintf</code> to be
   1921 omitted from <samp>libgcc.a</samp> on the assumption that it will be provided by
   1922 &lsquo;<samp>newlib</samp>&rsquo;.
   1923 </p>
   1924 <a name="avr"></a>
   1925 </dd>
   1926 <dt><code>--with-avrlibc</code></dt>
   1927 <dd><p>Only supported for the AVR target. Specifies that &lsquo;<samp>AVR-Libc</samp>&rsquo; is
   1928 being used as the target C&nbsp; library.  This causes float support
   1929 functions like <code>__addsf3</code> to be omitted from <samp>libgcc.a</samp> on
   1930 the assumption that it will be provided by <samp>libm.a</samp>.  For more
   1931 technical details, cf. <a href="http://gcc.gnu.org/PR54461">PR54461</a>.
   1932 It is not supported for
   1933 RTEMS configurations, which currently use newlib.  The option is
   1934 supported since version 4.7.2 and is the default in 4.8.0 and newer.
   1935 </p>
   1936 </dd>
   1937 <dt><code>--with-double={32|64|32,64|64,32}</code></dt>
   1938 <dt><code>--with-long-double={32|64|32,64|64,32|double}</code></dt>
   1939 <dd><p>Only supported for the AVR target since version&nbsp;10.
   1940 Specify the default layout available for the C/C++ &lsquo;<samp>double</samp>&rsquo;
   1941 and &lsquo;<samp>long double</samp>&rsquo; type, respectively. The following rules apply:
   1942 </p><ul>
   1943 <li> The first value after the &lsquo;<samp>=</samp>&rsquo; specifies the default layout (in bits)
   1944 of the type and also the default for the <samp>-mdouble=</samp> resp.
   1945 <samp>-mlong-double=</samp> compiler option.
   1946 </li><li> If more than one value is specified, respective multilib variants are
   1947 available, and  <samp>-mdouble=</samp> resp. <samp>-mlong-double=</samp> acts
   1948 as a multilib option.
   1949 </li><li> If <samp>--with-long-double=double</samp> is specified, &lsquo;<samp>double</samp>&rsquo; and
   1950 &lsquo;<samp>long double</samp>&rsquo; will have the same layout.
   1951 </li><li> The defaults are <samp>--with-long-double=64,32</samp> and
   1952 <samp>--with-double=32,64</samp>.  The default &lsquo;<samp>double</samp>&rsquo; layout imposed by
   1953 the latter is compatible with older versions of the compiler that implement
   1954 &lsquo;<samp>double</samp>&rsquo; as a 32-bit type, which does not comply to the language standard.
   1955 </li></ul>
   1956 <p>Not all combinations of <samp>--with-double=</samp> and
   1957 <samp>--with-long-double=</samp> are valid.  For example, the combination
   1958 <samp>--with-double=32,64</samp> <samp>--with-long-double=32</samp> will be
   1959 rejected because the first option specifies the availability of
   1960 multilibs for &lsquo;<samp>double</samp>&rsquo;, whereas the second option implies
   1961 that &lsquo;<samp>long double</samp>&rsquo; &mdash; and hence also &lsquo;<samp>double</samp>&rsquo; &mdash; is always
   1962 32&nbsp;bits wide.
   1963 </p>
   1964 </dd>
   1965 <dt><code>--with-double-comparison={tristate|bool|libf7}</code></dt>
   1966 <dd><p>Only supported for the AVR target since version&nbsp;10.
   1967 Specify what result format is returned by library functions that
   1968 compare 64-bit floating point values (<code>DFmode</code>).
   1969 The GCC default is &lsquo;<samp>tristate</samp>&rsquo;.  If the floating point
   1970 implementation returns a boolean instead, set it to &lsquo;<samp>bool</samp>&rsquo;.
   1971 </p>
   1972 </dd>
   1973 <dt><code>--with-libf7={libgcc|math|math-symbols|no}</code></dt>
   1974 <dd><p>Only supported for the AVR target since version&nbsp;10.
   1975 Specify to which degree code from LibF7 is included in libgcc.
   1976 LibF7 is an ad-hoc, AVR-specific, 64-bit floating point emulation
   1977 written in C and (inline) assembly. &lsquo;<samp>libgcc</samp>&rsquo; adds support
   1978 for functions that one would usually expect in libgcc like double addition,
   1979 double comparisons and double conversions. &lsquo;<samp>math</samp>&rsquo; also adds routines
   1980 that one would expect in <samp>libm.a</samp>, but with <code>__</code> (two underscores)
   1981 prepended to the symbol names as specified by <samp>math.h</samp>.
   1982 &lsquo;<samp>math-symbols</samp>&rsquo; also defines weak aliases for the functions
   1983 declared in <samp>math.h</samp>.  However, <code>--with-libf7</code> won&rsquo;t
   1984 install no <samp>math.h</samp> header file whatsoever, this file must come
   1985 from elsewhere.  This option sets <samp>--with-double-comparison</samp>
   1986 to &lsquo;<samp>bool</samp>&rsquo;.
   1987 </p>
   1988 </dd>
   1989 <dt><code>--with-nds32-lib=<var>library</var></code></dt>
   1990 <dd><p>Specifies that <var>library</var> setting is used for building <samp>libgcc.a</samp>.
   1991 Currently, the valid <var>library</var> is &lsquo;<samp>newlib</samp>&rsquo; or &lsquo;<samp>mculib</samp>&rsquo;.
   1992 This option is only supported for the NDS32 target.
   1993 </p>
   1994 </dd>
   1995 <dt><code>--with-build-time-tools=<var>dir</var></code></dt>
   1996 <dd><p>Specifies where to find the set of target tools (assembler, linker, etc.)
   1997 that will be used while building GCC itself.  This option can be useful
   1998 if the directory layouts are different between the system you are building
   1999 GCC on, and the system where you will deploy it.
   2000 </p>
   2001 <p>For example, on an &lsquo;<samp>ia64-hp-hpux</samp>&rsquo; system, you may have the GNU
   2002 assembler and linker in <samp>/usr/bin</samp>, and the native tools in a
   2003 different path, and build a toolchain that expects to find the
   2004 native tools in <samp>/usr/bin</samp>.
   2005 </p>
   2006 <p>When you use this option, you should ensure that <var>dir</var> includes
   2007 <code>ar</code>, <code>as</code>, <code>ld</code>, <code>nm</code>,
   2008 <code>ranlib</code> and <code>strip</code> if necessary, and possibly
   2009 <code>objdump</code>.  Otherwise, GCC may use an inconsistent set of
   2010 tools.
   2011 </p></dd>
   2012 </dl>
   2013 
   2014 <a name="Overriding-configure-test-results"></a>
   2015 <h4 class="subsubheading">Overriding <code>configure</code> test results</h4>
   2016 
   2017 <p>Sometimes, it might be necessary to override the result of some
   2018 <code>configure</code> test, for example in order to ease porting to a new
   2019 system or work around a bug in a test.  The toplevel <code>configure</code>
   2020 script provides three variables for this:
   2021 </p>
   2022 <dl compact="compact">
   2023 <dt><code>build_configargs</code></dt>
   2024 <dd><a name="index-build_005fconfigargs"></a>
   2025 <p>The contents of this variable is passed to all build <code>configure</code>
   2026 scripts.
   2027 </p>
   2028 </dd>
   2029 <dt><code>host_configargs</code></dt>
   2030 <dd><a name="index-host_005fconfigargs"></a>
   2031 <p>The contents of this variable is passed to all host <code>configure</code>
   2032 scripts.
   2033 </p>
   2034 </dd>
   2035 <dt><code>target_configargs</code></dt>
   2036 <dd><a name="index-target_005fconfigargs"></a>
   2037 <p>The contents of this variable is passed to all target <code>configure</code>
   2038 scripts.
   2039 </p>
   2040 </dd>
   2041 </dl>
   2042 
   2043 <p>In order to avoid shell and <code>make</code> quoting issues for complex
   2044 overrides, you can pass a setting for <code>CONFIG_SITE</code> and set
   2045 variables in the site file.
   2046 </p>
   2047 <a name="Objective-C-Specific-Options"></a>
   2048 <h4 class="subheading">Objective-C-Specific Options</h4>
   2049 
   2050 <p>The following options apply to the build of the Objective-C runtime library.
   2051 </p>
   2052 <dl compact="compact">
   2053 <dt><code>--enable-objc-gc</code></dt>
   2054 <dd><p>Specify that an additional variant of the GNU Objective-C runtime library
   2055 is built, using an external build of the Boehm-Demers-Weiser garbage
   2056 collector (<a href="https://www.hboehm.info/gc/">https://www.hboehm.info/gc/</a>).  This library needs to be
   2057 available for each multilib variant, unless configured with
   2058 <samp>--enable-objc-gc=&lsquo;<samp>auto</samp>&rsquo;</samp> in which case the build of the
   2059 additional runtime library is skipped when not available and the build
   2060 continues.
   2061 </p>
   2062 </dd>
   2063 <dt><code>--with-target-bdw-gc=<var>list</var></code></dt>
   2064 <dt><code>--with-target-bdw-gc-include=<var>list</var></code></dt>
   2065 <dt><code>--with-target-bdw-gc-lib=<var>list</var></code></dt>
   2066 <dd><p>Specify search directories for the garbage collector header files and
   2067 libraries. <var>list</var> is a comma separated list of key value pairs of the
   2068 form &lsquo;<samp><var>multilibdir</var>=<var>path</var></samp>&rsquo;, where the default multilib key
   2069 is named as &lsquo;<samp>.</samp>&rsquo; (dot), or is omitted (e.g.
   2070 &lsquo;<samp>--with-target-bdw-gc=/opt/bdw-gc,32=/opt-bdw-gc32</samp>&rsquo;).
   2071 </p>
   2072 <p>The options <samp>--with-target-bdw-gc-include</samp> and
   2073 <samp>--with-target-bdw-gc-lib</samp> must always be specified together
   2074 for each multilib variant and they take precedence over
   2075 <samp>--with-target-bdw-gc</samp>.  If <samp>--with-target-bdw-gc-include</samp>
   2076 is missing values for a multilib, then the value for the default
   2077 multilib is used (e.g. &lsquo;<samp>--with-target-bdw-gc-include=/opt/bdw-gc/include</samp>&rsquo;
   2078 &lsquo;<samp>--with-target-bdw-gc-lib=/opt/bdw-gc/lib64,32=/opt-bdw-gc/lib32</samp>&rsquo;).
   2079 If none of these options are specified, the library is assumed in
   2080 default locations.
   2081 </p></dd>
   2082 </dl>
   2083 
   2084 <a name="D-Specific-Options"></a>
   2085 <h4 class="subheading">D-Specific Options</h4>
   2086 
   2087 <p>The following options apply to the build of the D runtime library.
   2088 </p>
   2089 <dl compact="compact">
   2090 <dt><code>--enable-libphobos-checking</code></dt>
   2091 <dt><code>--disable-libphobos-checking</code></dt>
   2092 <dt><code>--enable-libphobos-checking=<var>list</var></code></dt>
   2093 <dd><p>This option controls whether run-time checks and contracts are compiled into
   2094 the D runtime library.  When the option is not specified, the library is built
   2095 with &lsquo;<samp>release</samp>&rsquo; checking.  When the option is specified without a
   2096 <var>list</var>, the result is the same as &lsquo;<samp>--enable-libphobos-checking=yes</samp>&rsquo;.
   2097 Likewise, &lsquo;<samp>--disable-libphobos-checking</samp>&rsquo; is equivalent to
   2098 &lsquo;<samp>--enable-libphobos-checking=no</samp>&rsquo;.
   2099 </p>
   2100 <p>The categories of checks available in <var>list</var> are &lsquo;<samp>yes</samp>&rsquo; (compiles
   2101 libphobos with <samp>-fno-release</samp>), &lsquo;<samp>no</samp>&rsquo; (compiles libphobos with
   2102 <samp>-frelease</samp>), &lsquo;<samp>all</samp>&rsquo; (same as &lsquo;<samp>yes</samp>&rsquo;), &lsquo;<samp>none</samp>&rsquo; or
   2103 &lsquo;<samp>release</samp>&rsquo; (same as &lsquo;<samp>no</samp>&rsquo;).
   2104 </p>
   2105 <p>Individual checks available in <var>list</var> are &lsquo;<samp>assert</samp>&rsquo; (compiles libphobos
   2106 with an extra option <samp>-fassert</samp>).
   2107 </p>
   2108 </dd>
   2109 <dt><code>--with-libphobos-druntime-only</code></dt>
   2110 <dt><code>--with-libphobos-druntime-only=<var>choice</var></code></dt>
   2111 <dd><p>Specify whether to build only the core D runtime library (druntime), or both
   2112 the core and standard library (phobos) into libphobos.  This is useful for
   2113 targets that have full support in druntime, but no or incomplete support
   2114 in phobos.  <var>choice</var> can be one of &lsquo;<samp>auto</samp>&rsquo;, &lsquo;<samp>yes</samp>&rsquo;, and &lsquo;<samp>no</samp>&rsquo;
   2115 where &lsquo;<samp>auto</samp>&rsquo; is the default.
   2116 </p>
   2117 <p>When the option is not specified, the default choice &lsquo;<samp>auto</samp>&rsquo; means that it
   2118 is inferred whether the target has support for the phobos standard library.
   2119 When the option is specified without a <var>choice</var>,  the result is the same as
   2120 &lsquo;<samp>--with-libphobos-druntime-only=yes</samp>&rsquo;.
   2121 </p>
   2122 </dd>
   2123 <dt><code>--with-target-system-zlib</code></dt>
   2124 <dd><p>Use installed &lsquo;<samp>zlib</samp>&rsquo; rather than that included with GCC.  This needs
   2125 to be available for each multilib variant, unless configured with
   2126 <samp>--with-target-system-zlib=&lsquo;<samp>auto</samp>&rsquo;</samp> in which case the GCC&nbsp;included
   2127 &lsquo;<samp>zlib</samp>&rsquo; is only used when the system installed library is not available.
   2128 </p></dd>
   2129 </dl>
   2130 
   2131 <hr />
   2132 <p>
   2133 <p><a href="./index.html">Return to the GCC Installation page</a>
   2134 </p>
   2135 
   2136 
   2137 
   2138 
   2139 
   2140 
   2141 
   2142 
   2143 
   2144 
   2145 
   2146 </body>
   2147 </html>
   2148