1<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd"> 2<html> 3<!-- Created by GNU Texinfo 6.8, https://www.gnu.org/software/texinfo/ --> 4<head> 5<meta http-equiv="Content-Type" content="text/html; charset=utf-8"> 6<!-- Copyright (C) 1988-2022 Free Software Foundation, Inc. 7 8Permission is granted to copy, distribute and/or modify this document 9under the terms of the GNU Free Documentation License, Version 1.3 or 10any later version published by the Free Software Foundation; with no 11Invariant Sections, the Front-Cover texts being (a) (see below), and 12with the Back-Cover Texts being (b) (see below). A copy of the 13license is included in the section entitled "GNU 14Free Documentation License". 15 16(a) The FSF's Front-Cover Text is: 17 18A GNU Manual 19 20(b) The FSF's Back-Cover Text is: 21 22You have freedom to copy and modify this GNU Manual, like GNU 23 software. Copies published by the Free Software Foundation raise 24 funds for GNU development. --> 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<meta name="viewport" content="width=device-width,initial-scale=1"> 33 34<style type="text/css"> 35<!-- 36a.copiable-anchor {visibility: hidden; text-decoration: none; line-height: 0em} 37a.summary-letter {text-decoration: none} 38blockquote.indentedblock {margin-right: 0em} 39div.display {margin-left: 3.2em} 40div.example {margin-left: 3.2em} 41kbd {font-style: oblique} 42pre.display {font-family: inherit} 43pre.format {font-family: inherit} 44pre.menu-comment {font-family: serif} 45pre.menu-preformatted {font-family: serif} 46span.nolinebreak {white-space: nowrap} 47span.roman {font-family: initial; font-weight: normal} 48span.sansserif {font-family: sans-serif; font-weight: normal} 49span:hover a.copiable-anchor {visibility: visible} 50ul.no-bullet {list-style: none} 51--> 52</style> 53 54 55</head> 56 57<body lang="en"> 58<h1 class="settitle" align="center">Installing GCC: Configuration</h1> 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78<span id="index-Configuration"></span> 79<span id="index-Installing-GCC_003a-Configuration"></span> 80 81<p>Like most GNU software, GCC must be configured before it can be built. 82This document describes the recommended configuration procedure 83for both native and cross targets. 84</p> 85<p>We use <var>srcdir</var> to refer to the toplevel source directory for 86GCC; we use <var>objdir</var> to refer to the toplevel build/object directory. 87</p> 88<p>If you obtained the sources by cloning the repository, <var>srcdir</var> 89must refer to the top <samp>gcc</samp> directory, the one where the 90<samp>MAINTAINERS</samp> file can be found, and not its <samp>gcc</samp> 91subdirectory, otherwise the build will fail. 92</p> 93<p>If either <var>srcdir</var> or <var>objdir</var> is located on an automounted NFS 94file system, the shell’s built-in <code>pwd</code> command will return 95temporary pathnames. Using these can lead to various sorts of build 96problems. To avoid this issue, set the <code>PWDCMD</code> environment 97variable to an automounter-aware <code>pwd</code> command, e.g., 98<code>pawd</code> or ‘<samp>amq -w</samp>’, during the configuration and build 99phases. 100</p> 101<p>First, we <strong>highly</strong> recommend that GCC be built into a 102separate directory from the sources which does <strong>not</strong> reside 103within the source tree. This is how we generally build GCC; building 104where <var>srcdir</var> == <var>objdir</var> should still work, but doesn’t 105get extensive testing; building where <var>objdir</var> is a subdirectory 106of <var>srcdir</var> is unsupported. 107</p> 108<p>If you have previously built GCC in the same directory for a 109different target machine, do ‘<samp>make distclean</samp>’ to delete all files 110that might be invalid. One of the files this deletes is <samp>Makefile</samp>; 111if ‘<samp>make distclean</samp>’ complains that <samp>Makefile</samp> does not exist 112or issues a message like “don’t know how to make distclean” it probably 113means that the directory is already suitably clean. However, with the 114recommended method of building in a separate <var>objdir</var>, you should 115simply use a different <var>objdir</var> for each target. 116</p> 117<p>Second, when configuring a native system, either <code>cc</code> or 118<code>gcc</code> must be in your path or you must set <code>CC</code> in 119your environment before running configure. Otherwise the configuration 120scripts may fail. 121</p> 122 123<p>To configure GCC: 124</p> 125<div class="example"> 126<pre class="example">% mkdir <var>objdir</var> 127% cd <var>objdir</var> 128% <var>srcdir</var>/configure [<var>options</var>] [<var>target</var>] 129</pre></div> 130 131<span id="Distributor-options"></span><h3 class="heading">Distributor options</h3> 132 133<p>If you will be distributing binary versions of GCC, with modifications 134to the source code, you should use the options described in this 135section to make clear that your version contains modifications. 136</p> 137<dl compact="compact"> 138<dt><span><code>--with-pkgversion=<var>version</var></code></span></dt> 139<dd><p>Specify a string that identifies your package. You may wish 140to include a build number or build date. This version string will be 141included in the output of <code>gcc --version</code>. This suffix does 142not replace the default version string, only the ‘<samp>GCC</samp>’ part. 143</p> 144<p>The default value is ‘<samp>GCC</samp>’. 145</p> 146</dd> 147<dt><span><code>--with-bugurl=<var>url</var></code></span></dt> 148<dd><p>Specify the URL that users should visit if they wish to report a bug. 149You are of course welcome to forward bugs reported to you to the FSF, 150if you determine that they are not bugs in your modifications. 151</p> 152<p>The default value refers to the FSF’s GCC bug tracker. 153</p> 154</dd> 155<dt><span><code>--with-documentation-root-url=<var>url</var></code></span></dt> 156<dd><p>Specify the URL root that contains GCC option documentation. The <var>url</var> 157should end with a <code>/</code> character. 158</p> 159<p>The default value is <a href="https://gcc.gnu.org/onlinedocs/">https://gcc.gnu.org/onlinedocs/</a>. 160</p> 161</dd> 162<dt><span><code>--with-changes-root-url=<var>url</var></code></span></dt> 163<dd><p>Specify the URL root that contains information about changes in GCC 164releases like <code>gcc-<var>version</var>/changes.html</code>. 165The <var>url</var> should end with a <code>/</code> character. 166</p> 167<p>The default value is <a href="https://gcc.gnu.org/">https://gcc.gnu.org/</a>. 168</p> 169</dd> 170</dl> 171 172<span id="Host_002c-Build-and-Target-specification"></span><h3 class="heading">Host, Build and Target specification</h3> 173 174<p>Specify the host, build and target machine configurations. You do this 175when you run the <samp>configure</samp> script. 176</p> 177<p>The <em>build</em> machine is the system which you are using, the 178<em>host</em> machine is the system where you want to run the resulting 179compiler (normally the build machine), and the <em>target</em> machine is 180the system for which you want the compiler to generate code. 181</p> 182<p>If you are building a compiler to produce code for the machine it runs 183on (a native compiler), you normally do not need to specify any operands 184to <samp>configure</samp>; it will try to guess the type of machine you are on 185and use that as the build, host and target machines. So you don’t need 186to specify a configuration when building a native compiler unless 187<samp>configure</samp> cannot figure out what your configuration is or guesses 188wrong. 189</p> 190<p>In those cases, specify the build machine’s <em>configuration name</em> 191with the <samp>--host</samp> option; the host and target will default to be 192the same as the host machine. 193</p> 194<p>Here is an example: 195</p> 196<div class="example"> 197<pre class="example">./configure --host=x86_64-pc-linux-gnu 198</pre></div> 199 200<p>A configuration name may be canonical or it may be more or less 201abbreviated (<samp>config.sub</samp> script produces canonical versions). 202</p> 203<p>A canonical configuration name has three parts, separated by dashes. 204It looks like this: ‘<samp><var>cpu</var>-<var>company</var>-<var>system</var></samp>’. 205</p> 206<p>Here are the possible CPU types: 207</p> 208<blockquote> 209<p>aarch64, aarch64_be, alpha, alpha64, amdgcn, arc, arceb, arm, armeb, avr, bfin, 210bpf, cr16, cris, csky, epiphany, fido, fr30, frv, ft32, h8300, hppa, hppa2.0, 211hppa64, i486, i686, ia64, iq2000, lm32, loongarch64, m32c, m32r, m32rle, m68k, 212mcore, microblaze, microblazeel, mips, mips64, mips64el, mips64octeon, 213mips64orion, mips64vr, mipsel, mipsisa32, mipsisa32r2, mipsisa64, mipsisa64r2, 214mipsisa64r2el, mipsisa64sb1, mipsisa64sr71k, mipstx39, mmix, mn10300, moxie, 215msp430, nds32be, nds32le, nios2, nvptx, or1k, pdp11, powerpc, powerpc64, 216powerpc64le, powerpcle, pru, riscv32, riscv32be, riscv64, riscv64be, rl78, rx, 217s390, s390x, sh, shle, sparc, sparc64, tic6x, tilegx, tilegxbe, tilepro, v850, 218v850e, v850e1, vax, visium, x86_64, xstormy16, xtensa 219</p></blockquote> 220 221<p>Here is a list of system types: 222</p> 223<blockquote> 224<p>aix<var>version</var>, amdhsa, aout, cygwin, darwin<var>version</var>, 225eabi, eabialtivec, eabisim, eabisimaltivec, elf, elf32, 226elfbare, elfoabi, freebsd<var>version</var>, gnu, hpux, hpux<var>version</var>, 227kfreebsd-gnu, kopensolaris-gnu, linux-androideabi, linux-gnu, 228linux-gnu_altivec, linux-musl, linux-uclibc, lynxos, mingw32, mingw32crt, 229mmixware, msdosdjgpp, netbsd, netbsdelf<var>version</var>, nto-qnx, openbsd, 230rtems, solaris<var>version</var>, symbianelf, tpf, uclinux, uclinux_eabi, vms, 231vxworks, vxworksae, vxworksmils 232</p></blockquote> 233 234<span id="Options-specification"></span><h3 class="heading">Options specification</h3> 235 236<p>Use <var>options</var> to override several configure time options for 237GCC. A list of supported <var>options</var> follows; ‘<samp>configure 238--help</samp>’ may list other options, but those not listed below may not 239work and should not normally be used. 240</p> 241<p>Note that each <samp>--enable</samp> option has a corresponding 242<samp>--disable</samp> option and that each <samp>--with</samp> option has a 243corresponding <samp>--without</samp> option. 244</p> 245<dl compact="compact"> 246<dt><span><code>--prefix=<var>dirname</var></code></span></dt> 247<dd><p>Specify the toplevel installation 248directory. This is the recommended way to install the tools into a directory 249other than the default. The toplevel installation directory defaults to 250<samp>/usr/local</samp>. 251</p> 252<p>We <strong>highly</strong> recommend against <var>dirname</var> being the same or a 253subdirectory of <var>objdir</var> or vice versa. If specifying a directory 254beneath a user’s home directory tree, some shells will not expand 255<var>dirname</var> correctly if it contains the ‘<samp>~</samp>’ metacharacter; use 256<code>$HOME</code> instead. 257</p> 258<p>The following standard <code>autoconf</code> options are supported. Normally you 259should not need to use these options. 260</p><dl compact="compact"> 261<dt><span><code>--exec-prefix=<var>dirname</var></code></span></dt> 262<dd><p>Specify the toplevel installation directory for architecture-dependent 263files. The default is <samp><var>prefix</var></samp>. 264</p> 265</dd> 266<dt><span><code>--bindir=<var>dirname</var></code></span></dt> 267<dd><p>Specify the installation directory for the executables called by users 268(such as <code>gcc</code> and <code>g++</code>). The default is 269<samp><var>exec-prefix</var>/bin</samp>. 270</p> 271</dd> 272<dt><span><code>--libdir=<var>dirname</var></code></span></dt> 273<dd><p>Specify the installation directory for object code libraries and 274internal data files of GCC. The default is <samp><var>exec-prefix</var>/lib</samp>. 275</p> 276</dd> 277<dt><span><code>--libexecdir=<var>dirname</var></code></span></dt> 278<dd><p>Specify the installation directory for internal executables of GCC. 279The default is <samp><var>exec-prefix</var>/libexec</samp>. 280</p> 281</dd> 282<dt><span><code>--with-slibdir=<var>dirname</var></code></span></dt> 283<dd><p>Specify the installation directory for the shared libgcc library. The 284default is <samp><var>libdir</var></samp>. 285</p> 286</dd> 287<dt><span><code>--datarootdir=<var>dirname</var></code></span></dt> 288<dd><p>Specify the root of the directory tree for read-only architecture-independent 289data files referenced by GCC. The default is <samp><var>prefix</var>/share</samp>. 290</p> 291</dd> 292<dt><span><code>--infodir=<var>dirname</var></code></span></dt> 293<dd><p>Specify the installation directory for documentation in info format. 294The default is <samp><var>datarootdir</var>/info</samp>. 295</p> 296</dd> 297<dt><span><code>--datadir=<var>dirname</var></code></span></dt> 298<dd><p>Specify the installation directory for some architecture-independent 299data files referenced by GCC. The default is <samp><var>datarootdir</var></samp>. 300</p> 301</dd> 302<dt><span><code>--docdir=<var>dirname</var></code></span></dt> 303<dd><p>Specify the installation directory for documentation files (other 304than Info) for GCC. The default is <samp><var>datarootdir</var>/doc</samp>. 305</p> 306</dd> 307<dt><span><code>--htmldir=<var>dirname</var></code></span></dt> 308<dd><p>Specify the installation directory for HTML documentation files. 309The default is <samp><var>docdir</var></samp>. 310</p> 311</dd> 312<dt><span><code>--pdfdir=<var>dirname</var></code></span></dt> 313<dd><p>Specify the installation directory for PDF documentation files. 314The default is <samp><var>docdir</var></samp>. 315</p> 316</dd> 317<dt><span><code>--mandir=<var>dirname</var></code></span></dt> 318<dd><p>Specify the installation directory for manual pages. The default is 319<samp><var>datarootdir</var>/man</samp>. (Note that the manual pages are only extracts 320from the full GCC manuals, which are provided in Texinfo format. The manpages 321are derived by an automatic conversion process from parts of the full 322manual.) 323</p> 324</dd> 325<dt><span><code>--with-gxx-include-dir=<var>dirname</var></code></span></dt> 326<dd><p>Specify 327the installation directory for G++ header files. The default depends 328on other configuration options, and differs between cross and native 329configurations. 330</p> 331</dd> 332<dt><span><code>--with-specs=<var>specs</var></code></span></dt> 333<dd><p>Specify additional command line driver SPECS. 334This can be useful if you need to turn on a non-standard feature by 335default without modifying the compiler’s source code, for instance 336<samp>--with-specs=%{!fcommon:%{!fno-common:-fno-common}}</samp>. 337See “Spec Files” in the main manual 338</p> 339</dd> 340</dl> 341 342</dd> 343<dt><span><code>--program-prefix=<var>prefix</var></code></span></dt> 344<dd><p>GCC supports some transformations of the names of its programs when 345installing them. This option prepends <var>prefix</var> to the names of 346programs to install in <var>bindir</var> (see above). For example, specifying 347<samp>--program-prefix=foo-</samp> would result in ‘<samp>gcc</samp>’ 348being installed as <samp>/usr/local/bin/foo-gcc</samp>. 349</p> 350</dd> 351<dt><span><code>--program-suffix=<var>suffix</var></code></span></dt> 352<dd><p>Appends <var>suffix</var> to the names of programs to install in <var>bindir</var> 353(see above). For example, specifying <samp>--program-suffix=-3.1</samp> 354would result in ‘<samp>gcc</samp>’ being installed as 355<samp>/usr/local/bin/gcc-3.1</samp>. 356</p> 357</dd> 358<dt><span><code>--program-transform-name=<var>pattern</var></code></span></dt> 359<dd><p>Applies the ‘<samp>sed</samp>’ script <var>pattern</var> to be applied to the names 360of programs to install in <var>bindir</var> (see above). <var>pattern</var> has to 361consist of one or more basic ‘<samp>sed</samp>’ editing commands, separated by 362semicolons. For example, if you want the ‘<samp>gcc</samp>’ program name to be 363transformed to the installed program <samp>/usr/local/bin/myowngcc</samp> and 364the ‘<samp>g++</samp>’ program name to be transformed to 365<samp>/usr/local/bin/gspecial++</samp> without changing other program names, 366you could use the pattern 367<samp>--program-transform-name='s/^gcc$/myowngcc/; s/^g++$/gspecial++/'</samp> 368to achieve this effect. 369</p> 370<p>All three options can be combined and used together, resulting in more 371complex conversion patterns. As a basic rule, <var>prefix</var> (and 372<var>suffix</var>) are prepended (appended) before further transformations 373can happen with a special transformation script <var>pattern</var>. 374</p> 375<p>As currently implemented, this option only takes effect for native 376builds; cross compiler binaries’ names are not transformed even when a 377transformation is explicitly asked for by one of these options. 378</p> 379<p>For native builds, some of the installed programs are also installed 380with the target alias in front of their name, as in 381‘<samp>i686-pc-linux-gnu-gcc</samp>’. All of the above transformations happen 382before the target alias is prepended to the name—so, specifying 383<samp>--program-prefix=foo-</samp> and <samp>program-suffix=-3.1</samp>, the 384resulting binary would be installed as 385<samp>/usr/local/bin/i686-pc-linux-gnu-foo-gcc-3.1</samp>. 386</p> 387<p>As a last shortcoming, none of the installed Ada programs are 388transformed yet, which will be fixed in some time. 389</p> 390</dd> 391<dt><span><code>--with-local-prefix=<var>dirname</var></code></span></dt> 392<dd><p>Specify the 393installation directory for local include files. The default is 394<samp>/usr/local</samp>. Specify this option if you want the compiler to 395search directory <samp><var>dirname</var>/include</samp> for locally installed 396header files <em>instead</em> of <samp>/usr/local/include</samp>. 397</p> 398<p>You should specify <samp>--with-local-prefix</samp> <strong>only</strong> if your 399site has a different convention (not <samp>/usr/local</samp>) for where to put 400site-specific files. 401</p> 402<p>The default value for <samp>--with-local-prefix</samp> is <samp>/usr/local</samp> 403regardless of the value of <samp>--prefix</samp>. Specifying 404<samp>--prefix</samp> has no effect on which directory GCC searches for 405local header files. This may seem counterintuitive, but actually it is 406logical. 407</p> 408<p>The purpose of <samp>--prefix</samp> is to specify where to <em>install 409GCC</em>. The local header files in <samp>/usr/local/include</samp>—if you put 410any in that directory—are not part of GCC. They are part of other 411programs—perhaps many others. (GCC installs its own header files in 412another directory which is based on the <samp>--prefix</samp> value.) 413</p> 414<p>Both the local-prefix include directory and the GCC-prefix include 415directory are part of GCC’s “system include” directories. Although these 416two directories are not fixed, they need to be searched in the proper 417order for the correct processing of the include_next directive. The 418local-prefix include directory is searched before the GCC-prefix 419include directory. Another characteristic of system include directories 420is that pedantic warnings are turned off for headers in these directories. 421</p> 422<p>Some autoconf macros add <samp>-I <var>directory</var></samp> options to the 423compiler command line, to ensure that directories containing installed 424packages’ headers are searched. When <var>directory</var> is one of GCC’s 425system include directories, GCC will ignore the option so that system 426directories continue to be processed in the correct order. This 427may result in a search order different from what was specified but the 428directory will still be searched. 429</p> 430<p>GCC automatically searches for ordinary libraries using 431<code>GCC_EXEC_PREFIX</code>. Thus, when the same installation prefix is 432used for both GCC and packages, GCC will automatically search for 433both headers and libraries. This provides a configuration that is 434easy to use. GCC behaves in a manner similar to that when it is 435installed as a system compiler in <samp>/usr</samp>. 436</p> 437<p>Sites that need to install multiple versions of GCC may not want to 438use the above simple configuration. It is possible to use the 439<samp>--program-prefix</samp>, <samp>--program-suffix</samp> and 440<samp>--program-transform-name</samp> options to install multiple versions 441into a single directory, but it may be simpler to use different prefixes 442and the <samp>--with-local-prefix</samp> option to specify the location of the 443site-specific files for each version. It will then be necessary for 444users to specify explicitly the location of local site libraries 445(e.g., with <code>LIBRARY_PATH</code>). 446</p> 447<p>The same value can be used for both <samp>--with-local-prefix</samp> and 448<samp>--prefix</samp> provided it is not <samp>/usr</samp>. This can be used 449to avoid the default search of <samp>/usr/local/include</samp>. 450</p> 451<p><strong>Do not</strong> specify <samp>/usr</samp> as the <samp>--with-local-prefix</samp>! 452The directory you use for <samp>--with-local-prefix</samp> <strong>must not</strong> 453contain any of the system’s standard header files. If it did contain 454them, certain programs would be miscompiled (including GNU Emacs, on 455certain targets), because this would override and nullify the header 456file corrections made by the <code>fixincludes</code> script. 457</p> 458<p>Indications are that people who use this option use it based on mistaken 459ideas of what it is for. People use it as if it specified where to 460install part of GCC. Perhaps they make this assumption because 461installing GCC creates the directory. 462</p> 463</dd> 464<dt><span><code>--with-gcc-major-version-only</code></span></dt> 465<dd><p>Specifies that GCC should use only the major number rather than 466<var>major</var>.<var>minor</var>.<var>patchlevel</var> in filesystem paths. 467</p> 468</dd> 469<dt><span><code>--with-native-system-header-dir=<var>dirname</var></code></span></dt> 470<dd><p>Specifies that <var>dirname</var> is the directory that contains native system 471header files, rather than <samp>/usr/include</samp>. This option is most useful 472if you are creating a compiler that should be isolated from the system 473as much as possible. It is most commonly used with the 474<samp>--with-sysroot</samp> option and will cause GCC to search 475<var>dirname</var> inside the system root specified by that option. 476</p> 477</dd> 478<dt><span><code>--enable-shared[=<var>package</var>[,…]]</code></span></dt> 479<dd><p>Build shared versions of libraries, if shared libraries are supported on 480the target platform. Unlike GCC 2.95.x and earlier, shared libraries 481are enabled by default on all platforms that support shared libraries. 482</p> 483<p>If a list of packages is given as an argument, build shared libraries 484only for the listed packages. For other packages, only static libraries 485will be built. Package names currently recognized in the GCC tree are 486‘<samp>libgcc</samp>’ (also known as ‘<samp>gcc</samp>’), ‘<samp>libstdc++</samp>’ (not 487‘<samp>libstdc++-v3</samp>’), ‘<samp>libffi</samp>’, ‘<samp>zlib</samp>’, ‘<samp>boehm-gc</samp>’, 488‘<samp>ada</samp>’, ‘<samp>libada</samp>’, ‘<samp>libgo</samp>’, ‘<samp>libobjc</samp>’, and ‘<samp>libphobos</samp>’. 489Note ‘<samp>libiberty</samp>’ does not support shared libraries at all. 490</p> 491<p>Use <samp>--disable-shared</samp> to build only static libraries. Note that 492<samp>--disable-shared</samp> does not accept a list of package names as 493argument, only <samp>--enable-shared</samp> does. 494</p> 495<p>Contrast with <samp>--enable-host-shared</samp>, which affects <em>host</em> 496code. 497</p> 498</dd> 499<dt><span><code>--enable-host-shared</code></span></dt> 500<dd><p>Specify that the <em>host</em> code should be built into position-independent 501machine code (with -fPIC), allowing it to be used within shared libraries, 502but yielding a slightly slower compiler. 503</p> 504<p>This option is required when building the libgccjit.so library. 505</p> 506<p>Contrast with <samp>--enable-shared</samp>, which affects <em>target</em> 507libraries. 508</p> 509</dd> 510<dt><span><code><span id="with-gnu-as"></span>--with-gnu-as</code></span></dt> 511<dd><p>Specify that the compiler should assume that the 512assembler it finds is the GNU assembler. However, this does not modify 513the rules to find an assembler and will result in confusion if the 514assembler found is not actually the GNU assembler. (Confusion may also 515result if the compiler finds the GNU assembler but has not been 516configured with <samp>--with-gnu-as</samp>.) If you have more than one 517assembler installed on your system, you may want to use this option in 518connection with <samp>--with-as=<var>pathname</var></samp> or 519<samp>--with-build-time-tools=<var>pathname</var></samp>. 520</p> 521<p>The following systems are the only ones where it makes a difference 522whether you use the GNU assembler. On any other system, 523<samp>--with-gnu-as</samp> has no effect. 524</p> 525<ul> 526<li> ‘<samp>hppa1.0-<var>any</var>-<var>any</var></samp>’ 527</li><li> ‘<samp>hppa1.1-<var>any</var>-<var>any</var></samp>’ 528</li><li> ‘<samp>sparc-sun-solaris2.<var>any</var></samp>’ 529</li><li> ‘<samp>sparc64-<var>any</var>-solaris2.<var>any</var></samp>’ 530</li></ul> 531 532</dd> 533<dt><span><code><span id="with-as"></span>--with-as=<var>pathname</var></code></span></dt> 534<dd><p>Specify that the compiler should use the assembler pointed to by 535<var>pathname</var>, rather than the one found by the standard rules to find 536an assembler, which are: 537</p><ul> 538<li> Unless GCC is being built with a cross compiler, check the 539<samp><var>libexec</var>/gcc/<var>target</var>/<var>version</var></samp> directory. 540<var>libexec</var> defaults to <samp><var>exec-prefix</var>/libexec</samp>; 541<var>exec-prefix</var> defaults to <var>prefix</var>, which 542defaults to <samp>/usr/local</samp> unless overridden by the 543<samp>--prefix=<var>pathname</var></samp> switch described above. <var>target</var> 544is the target system triple, such as ‘<samp>sparc-sun-solaris2.7</samp>’, and 545<var>version</var> denotes the GCC version, such as 3.0. 546 547</li><li> If the target system is the same that you are building on, check 548operating system specific directories (e.g. <samp>/usr/ccs/bin</samp> on 549Solaris 2). 550 551</li><li> Check in the <code>PATH</code> for a tool whose name is prefixed by the 552target system triple. 553 554</li><li> Check in the <code>PATH</code> for a tool whose name is not prefixed by the 555target system triple, if the host and target system triple are 556the same (in other words, we use a host tool if it can be used for 557the target as well). 558</li></ul> 559 560<p>You may want to use <samp>--with-as</samp> if no assembler 561is installed in the directories listed above, or if you have multiple 562assemblers installed and want to choose one that is not found by the 563above rules. 564</p> 565</dd> 566<dt><span><code><span id="with-gnu-ld"></span>--with-gnu-ld</code></span></dt> 567<dd><p>Same as <a href="#with-gnu-as"><samp>--with-gnu-as</samp></a> 568but for the linker. 569</p> 570</dd> 571<dt><span><code>--with-ld=<var>pathname</var></code></span></dt> 572<dd><p>Same as <a href="#with-as"><samp>--with-as</samp></a> 573but for the linker. 574</p> 575</dd> 576<dt><span><code>--with-dsymutil=<var>pathname</var></code></span></dt> 577<dd><p>Same as <a href="#with-as"><samp>--with-as</samp></a> 578but for the debug linker (only used on Darwin platforms so far). 579</p> 580</dd> 581<dt><span><code>--with-tls=<var>dialect</var></code></span></dt> 582<dd><p>Specify the default TLS dialect, for systems were there is a choice. 583For ARM targets, possible values for <var>dialect</var> are <code>gnu</code> or 584<code>gnu2</code>, which select between the original GNU dialect and the GNU TLS 585descriptor-based dialect. 586</p> 587</dd> 588<dt><span><code>--enable-multiarch</code></span></dt> 589<dd><p>Specify whether to enable or disable multiarch support. The default is 590to check for glibc start files in a multiarch location, and enable it 591if the files are found. The auto detection is enabled for native builds, 592and for cross builds configured with <samp>--with-sysroot</samp>, and without 593<samp>--with-native-system-header-dir</samp>. 594More documentation about multiarch can be found at 595<a href="https://wiki.debian.org/Multiarch">https://wiki.debian.org/Multiarch</a>. 596</p> 597</dd> 598<dt><span><code>--enable-sjlj-exceptions</code></span></dt> 599<dd><p>Force use of the <code>setjmp</code>/<code>longjmp</code>-based scheme for exceptions. 600‘<samp>configure</samp>’ ordinarily picks the correct value based on the platform. 601Only use this option if you are sure you need a different setting. 602</p> 603</dd> 604<dt><span><code>--enable-vtable-verify</code></span></dt> 605<dd><p>Specify whether to enable or disable the vtable verification feature. 606Enabling this feature causes libstdc++ to be built with its virtual calls 607in verifiable mode. This means that, when linked with libvtv, every 608virtual call in libstdc++ will verify the vtable pointer through which the 609call will be made before actually making the call. If not linked with libvtv, 610the verifier will call stub functions (in libstdc++ itself) and do nothing. 611If vtable verification is disabled, then libstdc++ is not built with its 612virtual calls in verifiable mode at all. However the libvtv library will 613still be built (see <samp>--disable-libvtv</samp> to turn off building libvtv). 614<samp>--disable-vtable-verify</samp> is the default. 615</p> 616</dd> 617<dt><span><code>--disable-gcov</code></span></dt> 618<dd><p>Specify that the run-time library used for coverage analysis 619and associated host tools should not be built. 620</p> 621</dd> 622<dt><span><code>--disable-multilib</code></span></dt> 623<dd><p>Specify that multiple target 624libraries to support different target variants, calling 625conventions, etc. should not be built. The default is to build a 626predefined set of them. 627</p> 628<p>Some targets provide finer-grained control over which multilibs are built 629(e.g., <samp>--disable-softfloat</samp>): 630</p><dl compact="compact"> 631<dt><span><code>arm-*-*</code></span></dt> 632<dd><p>fpu, 26bit, underscore, interwork, biendian, nofmult. 633</p> 634</dd> 635<dt><span><code>m68*-*-*</code></span></dt> 636<dd><p>softfloat, m68881, m68000, m68020. 637</p> 638</dd> 639<dt><span><code>mips*-*-*</code></span></dt> 640<dd><p>single-float, biendian, softfloat. 641</p> 642</dd> 643<dt><span><code>msp430-*-*</code></span></dt> 644<dd><p>no-exceptions 645</p> 646</dd> 647<dt><span><code>powerpc*-*-*, rs6000*-*-*</code></span></dt> 648<dd><p>aix64, pthread, softfloat, powercpu, powerpccpu, powerpcos, biendian, 649sysv, aix. 650</p> 651</dd> 652</dl> 653 654</dd> 655<dt><span><code>--with-multilib-list=<var>list</var></code></span></dt> 656<dt><span><code>--without-multilib-list</code></span></dt> 657<dd><p>Specify what multilibs to build. <var>list</var> is a comma separated list of 658values, possibly consisting of a single value. Currently only implemented 659for aarch64*-*-*, arm*-*-*, loongarch64-*-*, riscv*-*-*, sh*-*-* and 660x86-64-*-linux*. The accepted values and meaning for each target is given 661below. 662</p> 663<dl compact="compact"> 664<dt><span><code>aarch64*-*-*</code></span></dt> 665<dd><p><var>list</var> is a comma separated list of <code>ilp32</code>, and <code>lp64</code> 666to enable ILP32 and LP64 run-time libraries, respectively. If 667<var>list</var> is empty, then there will be no multilibs and only the 668default run-time library will be built. If <var>list</var> is 669<code>default</code> or –with-multilib-list= is not specified, then the 670default set of libraries is selected based on the value of 671<samp>--target</samp>. 672</p> 673</dd> 674<dt><span><code>arm*-*-*</code></span></dt> 675<dd><p><var>list</var> is a comma separated list of <code>aprofile</code> and 676<code>rmprofile</code> to build multilibs for A or R and M architecture 677profiles respectively. Note that, due to some limitation of the current 678multilib framework, using the combined <code>aprofile,rmprofile</code> 679multilibs selects in some cases a less optimal multilib than when using 680the multilib profile for the architecture targetted. The special value 681<code>default</code> is also accepted and is equivalent to omitting the 682option, i.e., only the default run-time library will be enabled. 683</p> 684<p><var>list</var> may instead contain <code>@name</code>, to use the multilib 685configuration Makefile fragment <samp>name</samp> in <samp>gcc/config/arm</samp> in 686the source tree (it is part of the corresponding sources, after all). 687It is recommended, but not required, that files used for this purpose to 688be named starting with <samp>t-ml-</samp>, to make their intended purpose 689self-evident, in line with GCC conventions. Such files enable custom, 690user-chosen multilib lists to be configured. Whether multiple such 691files can be used together depends on the contents of the supplied 692files. See <samp>gcc/config/arm/t-multilib</samp> and its supplementary 693<samp>gcc/config/arm/t-*profile</samp> files for an example of what such 694Makefile fragments might look like for this version of GCC. The macros 695expected to be defined in these fragments are not stable across GCC 696releases, so make sure they define the <code>MULTILIB</code>-related macros 697expected by the version of GCC you are building. 698See “Target Makefile Fragments” in the internals manual. 699</p> 700<p>The table below gives the combination of ISAs, architectures, FPUs and 701floating-point ABIs for which multilibs are built for each predefined 702profile. The union of these options is considered when specifying both 703<code>aprofile</code> and <code>rmprofile</code>. 704</p> 705<table> 706<tr><td width="15%">Option</td><td width="28%">aprofile</td><td width="30%">rmprofile</td></tr> 707<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> 708<tr><td width="15%">Architectures<br><br><br><br><br><br></td><td width="28%">default architecture<br> 709<code>-march=armv7-a</code><br> 710<code>-march=armv7ve</code><br> 711<code>-march=armv8-a</code><br><br><br></td><td width="30%">default architecture<br> 712<code>-march=armv6s-m</code><br> 713<code>-march=armv7-m</code><br> 714<code>-march=armv7e-m</code><br> 715<code>-march=armv8-m.base</code><br> 716<code>-march=armv8-m.main</code><br> 717<code>-march=armv7</code></td></tr> 718<tr><td width="15%">FPUs<br><br><br><br><br></td><td width="28%">none<br> 719<code>-mfpu=vfpv3-d16</code><br> 720<code>-mfpu=neon</code><br> 721<code>-mfpu=vfpv4-d16</code><br> 722<code>-mfpu=neon-vfpv4</code><br> 723<code>-mfpu=neon-fp-armv8</code></td><td width="30%">none<br> 724<code>-mfpu=vfpv3-d16</code><br> 725<code>-mfpu=fpv4-sp-d16</code><br> 726<code>-mfpu=fpv5-sp-d16</code><br> 727<code>-mfpu=fpv5-d16</code><br></td></tr> 728<tr><td width="15%">floating-point ABIs<br><br></td><td width="28%"><code>-mfloat-abi=soft</code><br> 729<code>-mfloat-abi=softfp</code><br> 730<code>-mfloat-abi=hard</code></td><td width="30%"><code>-mfloat-abi=soft</code><br> 731<code>-mfloat-abi=softfp</code><br> 732<code>-mfloat-abi=hard</code></td></tr> 733</table> 734 735</dd> 736<dt><span><code>loongarch*-*-*</code></span></dt> 737<dd><p><var>list</var> is a comma-separated list of the following ABI identifiers: 738<code>lp64d[/base]</code> <code>lp64f[/base]</code> <code>lp64d[/base]</code>, where the 739<code>/base</code> suffix may be omitted, to enable their respective run-time 740libraries. If <var>list</var> is empty or <code>default</code>, 741or if <samp>--with-multilib-list</samp> is not specified, then the default ABI 742as specified by <samp>--with-abi</samp> or implied by <samp>--target</samp> is selected. 743</p> 744</dd> 745<dt><span><code>riscv*-*-*</code></span></dt> 746<dd><p><var>list</var> is a single ABI name. The target architecture must be either 747<code>rv32gc</code> or <code>rv64gc</code>. This will build a single multilib for the 748specified architecture and ABI pair. If <code>--with-multilib-list</code> is not 749given, then a default set of multilibs is selected based on the value of 750<samp>--target</samp>. This is usually a large set of multilibs. 751</p> 752</dd> 753<dt><span><code>sh*-*-*</code></span></dt> 754<dd><p><var>list</var> is a comma separated list of CPU names. These must be of the 755form <code>sh*</code> or <code>m*</code> (in which case they match the compiler option 756for that processor). The list should not contain any endian options - 757these are handled by <samp>--with-endian</samp>. 758</p> 759<p>If <var>list</var> is empty, then there will be no multilibs for extra 760processors. The multilib for the secondary endian remains enabled. 761</p> 762<p>As a special case, if an entry in the list starts with a <code>!</code> 763(exclamation point), then it is added to the list of excluded multilibs. 764Entries of this sort should be compatible with ‘<samp>MULTILIB_EXCLUDES</samp>’ 765(once the leading <code>!</code> has been stripped). 766</p> 767<p>If <samp>--with-multilib-list</samp> is not given, then a default set of 768multilibs is selected based on the value of <samp>--target</samp>. This is 769usually the complete set of libraries, but some targets imply a more 770specialized subset. 771</p> 772<p>Example 1: to configure a compiler for SH4A only, but supporting both 773endians, with little endian being the default: 774</p><div class="example"> 775<pre class="example">--with-cpu=sh4a --with-endian=little,big --with-multilib-list= 776</pre></div> 777 778<p>Example 2: to configure a compiler for both SH4A and SH4AL-DSP, but with 779only little endian SH4AL: 780</p><div class="example"> 781<pre class="example">--with-cpu=sh4a --with-endian=little,big \ 782--with-multilib-list=sh4al,!mb/m4al 783</pre></div> 784 785</dd> 786<dt><span><code>x86-64-*-linux*</code></span></dt> 787<dd><p><var>list</var> is a comma separated list of <code>m32</code>, <code>m64</code> and 788<code>mx32</code> to enable 32-bit, 64-bit and x32 run-time libraries, 789respectively. If <var>list</var> is empty, then there will be no multilibs 790and only the default run-time library will be enabled. 791</p> 792<p>If <samp>--with-multilib-list</samp> is not given, then only 32-bit and 79364-bit run-time libraries will be enabled. 794</p></dd> 795</dl> 796 797</dd> 798<dt><span><code>--with-multilib-generator=<var>config</var></code></span></dt> 799<dd><p>Specify what multilibs to build. <var>config</var> is a semicolon separated list of 800values, possibly consisting of a single value. Currently only implemented 801for riscv*-*-elf*. The accepted values and meanings are given below. 802</p> 803 804<p>Every config is constructed with four components: architecture string, ABI, 805reuse rule with architecture string and reuse rule with sub-extension. 806</p> 807<p>Example 1: Add multi-lib suppport for rv32i with ilp32. 808</p><div class="example"> 809<pre class="example">rv32i-ilp32-- 810</pre></div> 811 812<p>Example 2: Add multi-lib suppport for rv32i with ilp32 and rv32imafd with ilp32. 813</p><div class="example"> 814<pre class="example">rv32i-ilp32--;rv32imafd-ilp32-- 815</pre></div> 816 817<p>Example 3: Add multi-lib suppport for rv32i with ilp32; rv32im with ilp32 and 818rv32ic with ilp32 will reuse this multi-lib set. 819</p><div class="example"> 820<pre class="example">rv32i-ilp32-rv32im-c 821</pre></div> 822 823<p>Example 4: Add multi-lib suppport for rv64ima with lp64; rv64imaf with lp64, 824rv64imac with lp64 and rv64imafc with lp64 will reuse this multi-lib set. 825</p><div class="example"> 826<pre class="example">rv64ima-lp64--f,c,fc 827</pre></div> 828 829<p><samp>--with-multilib-generator</samp> have an optional configuration argument 830<samp>--cmodel=val</samp> for code model, this option will expand with other 831config options, <var>val</var> is a comma separated list of possible code model, 832currently we support medlow and medany. 833</p> 834<p>Example 5: Add multi-lib suppport for rv64ima with lp64; rv64ima with lp64 and 835medlow code model 836</p><div class="example"> 837<pre class="example">rv64ima-lp64--;--cmodel=medlow 838</pre></div> 839 840<p>Example 6: Add multi-lib suppport for rv64ima with lp64; rv64ima with lp64 and 841medlow code model; rv64ima with lp64 and medany code model 842</p><div class="example"> 843<pre class="example">rv64ima-lp64--;--cmodel=medlow,medany 844</pre></div> 845 846</dd> 847<dt><span><code>--with-endian=<var>endians</var></code></span></dt> 848<dd><p>Specify what endians to use. 849Currently only implemented for sh*-*-*. 850</p> 851<p><var>endians</var> may be one of the following: 852</p><dl compact="compact"> 853<dt><span><code>big</code></span></dt> 854<dd><p>Use big endian exclusively. 855</p></dd> 856<dt><span><code>little</code></span></dt> 857<dd><p>Use little endian exclusively. 858</p></dd> 859<dt><span><code>big,little</code></span></dt> 860<dd><p>Use big endian by default. Provide a multilib for little endian. 861</p></dd> 862<dt><span><code>little,big</code></span></dt> 863<dd><p>Use little endian by default. Provide a multilib for big endian. 864</p></dd> 865</dl> 866 867</dd> 868<dt><span><code>--enable-threads</code></span></dt> 869<dd><p>Specify that the target 870supports threads. This affects the Objective-C compiler and runtime 871library, and exception handling for other languages like C++. 872On some systems, this is the default. 873</p> 874<p>In general, the best (and, in many cases, the only known) threading 875model available will be configured for use. Beware that on some 876systems, GCC has not been taught what threading models are generally 877available for the system. In this case, <samp>--enable-threads</samp> is an 878alias for <samp>--enable-threads=single</samp>. 879</p> 880</dd> 881<dt><span><code>--disable-threads</code></span></dt> 882<dd><p>Specify that threading support should be disabled for the system. 883This is an alias for <samp>--enable-threads=single</samp>. 884</p> 885</dd> 886<dt><span><code>--enable-threads=<var>lib</var></code></span></dt> 887<dd><p>Specify that 888<var>lib</var> is the thread support library. This affects the Objective-C 889compiler and runtime library, and exception handling for other languages 890like C++. The possibilities for <var>lib</var> are: 891</p> 892<dl compact="compact"> 893<dt><span><code>aix</code></span></dt> 894<dd><p>AIX thread support. 895</p></dd> 896<dt><span><code>dce</code></span></dt> 897<dd><p>DCE thread support. 898</p></dd> 899<dt><span><code>lynx</code></span></dt> 900<dd><p>LynxOS thread support. 901</p></dd> 902<dt><span><code>mipssde</code></span></dt> 903<dd><p>MIPS SDE thread support. 904</p></dd> 905<dt><span><code>no</code></span></dt> 906<dd><p>This is an alias for ‘<samp>single</samp>’. 907</p></dd> 908<dt><span><code>posix</code></span></dt> 909<dd><p>Generic POSIX/Unix98 thread support. 910</p></dd> 911<dt><span><code>rtems</code></span></dt> 912<dd><p>RTEMS thread support. 913</p></dd> 914<dt><span><code>single</code></span></dt> 915<dd><p>Disable thread support, should work for all platforms. 916</p></dd> 917<dt><span><code>tpf</code></span></dt> 918<dd><p>TPF thread support. 919</p></dd> 920<dt><span><code>vxworks</code></span></dt> 921<dd><p>VxWorks thread support. 922</p></dd> 923<dt><span><code>win32</code></span></dt> 924<dd><p>Microsoft Win32 API thread support. 925</p></dd> 926</dl> 927 928</dd> 929<dt><span><code>--enable-tls</code></span></dt> 930<dd><p>Specify that the target supports TLS (Thread Local Storage). Usually 931configure can correctly determine if TLS is supported. In cases where 932it guesses incorrectly, TLS can be explicitly enabled or disabled with 933<samp>--enable-tls</samp> or <samp>--disable-tls</samp>. This can happen if 934the assembler supports TLS but the C library does not, or if the 935assumptions made by the configure test are incorrect. 936</p> 937</dd> 938<dt><span><code>--disable-tls</code></span></dt> 939<dd><p>Specify that the target does not support TLS. 940This is an alias for <samp>--enable-tls=no</samp>. 941</p> 942</dd> 943<dt><span><code>--disable-tm-clone-registry</code></span></dt> 944<dd><p>Disable TM clone registry in libgcc. It is enabled in libgcc by default. 945This option helps to reduce code size for embedded targets which do 946not use transactional memory. 947</p> 948</dd> 949<dt><span><code>--with-cpu=<var>cpu</var></code></span></dt> 950<dt><span><code>--with-cpu-32=<var>cpu</var></code></span></dt> 951<dt><span><code>--with-cpu-64=<var>cpu</var></code></span></dt> 952<dd><p>Specify which cpu variant the compiler should generate code for by default. 953<var>cpu</var> will be used as the default value of the <samp>-mcpu=</samp> switch. 954This option is only supported on some targets, including ARC, ARM, i386, M68k, 955PowerPC, and SPARC. It is mandatory for ARC. The <samp>--with-cpu-32</samp> and 956<samp>--with-cpu-64</samp> options specify separate default CPUs for 95732-bit and 64-bit modes; these options are only supported for aarch64, i386, 958x86-64, PowerPC, and SPARC. 959</p> 960</dd> 961<dt><span><code>--with-schedule=<var>cpu</var></code></span></dt> 962<dt><span><code>--with-arch=<var>cpu</var></code></span></dt> 963<dt><span><code>--with-arch-32=<var>cpu</var></code></span></dt> 964<dt><span><code>--with-arch-64=<var>cpu</var></code></span></dt> 965<dt><span><code>--with-tune=<var>cpu</var></code></span></dt> 966<dt><span><code>--with-tune-32=<var>cpu</var></code></span></dt> 967<dt><span><code>--with-tune-64=<var>cpu</var></code></span></dt> 968<dt><span><code>--with-abi=<var>abi</var></code></span></dt> 969<dt><span><code>--with-fpu=<var>type</var></code></span></dt> 970<dt><span><code>--with-float=<var>type</var></code></span></dt> 971<dd><p>These configure options provide default values for the <samp>-mschedule=</samp>, 972<samp>-march=</samp>, <samp>-mtune=</samp>, <samp>-mabi=</samp>, and <samp>-mfpu=</samp> 973options and for <samp>-mhard-float</samp> or <samp>-msoft-float</samp>. As with 974<samp>--with-cpu</samp>, which switches will be accepted and acceptable values 975of the arguments depend on the target. 976</p> 977</dd> 978<dt><span><code>--with-mode=<var>mode</var></code></span></dt> 979<dd><p>Specify if the compiler should default to <samp>-marm</samp> or <samp>-mthumb</samp>. 980This option is only supported on ARM targets. 981</p> 982</dd> 983<dt><span><code>--with-stack-offset=<var>num</var></code></span></dt> 984<dd><p>This option sets the default for the -mstack-offset=<var>num</var> option, 985and will thus generally also control the setting of this option for 986libraries. This option is only supported on Epiphany targets. 987</p> 988</dd> 989<dt><span><code>--with-fpmath=<var>isa</var></code></span></dt> 990<dd><p>This options sets <samp>-mfpmath=sse</samp> by default and specifies the default 991ISA for floating-point arithmetics. You can select either ‘<samp>sse</samp>’ which 992enables <samp>-msse2</samp> or ‘<samp>avx</samp>’ which enables <samp>-mavx</samp> by default. 993This option is only supported on i386 and x86-64 targets. 994</p> 995</dd> 996<dt><span><code>--with-fp-32=<var>mode</var></code></span></dt> 997<dd><p>On MIPS targets, set the default value for the <samp>-mfp</samp> option when using 998the o32 ABI. The possibilities for <var>mode</var> are: 999</p><dl compact="compact"> 1000<dt><span><code>32</code></span></dt> 1001<dd><p>Use the o32 FP32 ABI extension, as with the <samp>-mfp32</samp> command-line 1002option. 1003</p></dd> 1004<dt><span><code>xx</code></span></dt> 1005<dd><p>Use the o32 FPXX ABI extension, as with the <samp>-mfpxx</samp> command-line 1006option. 1007</p></dd> 1008<dt><span><code>64</code></span></dt> 1009<dd><p>Use the o32 FP64 ABI extension, as with the <samp>-mfp64</samp> command-line 1010option. 1011</p></dd> 1012</dl> 1013<p>In the absence of this configuration option the default is to use the o32 1014FP32 ABI extension. 1015</p> 1016</dd> 1017<dt><span><code>--with-odd-spreg-32</code></span></dt> 1018<dd><p>On MIPS targets, set the <samp>-modd-spreg</samp> option by default when using 1019the o32 ABI. 1020</p> 1021</dd> 1022<dt><span><code>--without-odd-spreg-32</code></span></dt> 1023<dd><p>On MIPS targets, set the <samp>-mno-odd-spreg</samp> option by default when using 1024the o32 ABI. This is normally used in conjunction with 1025<samp>--with-fp-32=64</samp> in order to target the o32 FP64A ABI extension. 1026</p> 1027</dd> 1028<dt><span><code>--with-nan=<var>encoding</var></code></span></dt> 1029<dd><p>On MIPS targets, set the default encoding convention to use for the 1030special not-a-number (NaN) IEEE 754 floating-point data. The 1031possibilities for <var>encoding</var> are: 1032</p><dl compact="compact"> 1033<dt><span><code>legacy</code></span></dt> 1034<dd><p>Use the legacy encoding, as with the <samp>-mnan=legacy</samp> command-line 1035option. 1036</p></dd> 1037<dt><span><code>2008</code></span></dt> 1038<dd><p>Use the 754-2008 encoding, as with the <samp>-mnan=2008</samp> command-line 1039option. 1040</p></dd> 1041</dl> 1042<p>To use this configuration option you must have an assembler version 1043installed that supports the <samp>-mnan=</samp> command-line option too. 1044In the absence of this configuration option the default convention is 1045the legacy encoding, as when neither of the <samp>-mnan=2008</samp> and 1046<samp>-mnan=legacy</samp> command-line options has been used. 1047</p> 1048</dd> 1049<dt><span><code>--with-divide=<var>type</var></code></span></dt> 1050<dd><p>Specify how the compiler should generate code for checking for 1051division by zero. This option is only supported on the MIPS target. 1052The possibilities for <var>type</var> are: 1053</p><dl compact="compact"> 1054<dt><span><code>traps</code></span></dt> 1055<dd><p>Division by zero checks use conditional traps (this is the default on 1056systems that support conditional traps). 1057</p></dd> 1058<dt><span><code>breaks</code></span></dt> 1059<dd><p>Division by zero checks use the break instruction. 1060</p></dd> 1061</dl> 1062 1063 1064</dd> 1065<dt><span><code>--with-llsc</code></span></dt> 1066<dd><p>On MIPS targets, make <samp>-mllsc</samp> the default when no 1067<samp>-mno-llsc</samp> option is passed. This is the default for 1068Linux-based targets, as the kernel will emulate them if the ISA does 1069not provide them. 1070</p> 1071</dd> 1072<dt><span><code>--without-llsc</code></span></dt> 1073<dd><p>On MIPS targets, make <samp>-mno-llsc</samp> the default when no 1074<samp>-mllsc</samp> option is passed. 1075</p> 1076</dd> 1077<dt><span><code>--with-synci</code></span></dt> 1078<dd><p>On MIPS targets, make <samp>-msynci</samp> the default when no 1079<samp>-mno-synci</samp> option is passed. 1080</p> 1081</dd> 1082<dt><span><code>--without-synci</code></span></dt> 1083<dd><p>On MIPS targets, make <samp>-mno-synci</samp> the default when no 1084<samp>-msynci</samp> option is passed. This is the default. 1085</p> 1086</dd> 1087<dt><span><code>--with-lxc1-sxc1</code></span></dt> 1088<dd><p>On MIPS targets, make <samp>-mlxc1-sxc1</samp> the default when no 1089<samp>-mno-lxc1-sxc1</samp> option is passed. This is the default. 1090</p> 1091</dd> 1092<dt><span><code>--without-lxc1-sxc1</code></span></dt> 1093<dd><p>On MIPS targets, make <samp>-mno-lxc1-sxc1</samp> the default when no 1094<samp>-mlxc1-sxc1</samp> option is passed. The indexed load/store 1095instructions are not directly a problem but can lead to unexpected 1096behaviour when deployed in an application intended for a 32-bit address 1097space but run on a 64-bit processor. The issue is seen because all 1098known MIPS 64-bit Linux kernels execute o32 and n32 applications 1099with 64-bit addressing enabled which affects the overflow behaviour 1100of the indexed addressing mode. GCC will assume that ordinary 110132-bit arithmetic overflow behaviour is the same whether performed 1102as an <code>addu</code> instruction or as part of the address calculation 1103in <code>lwxc1</code> type instructions. This assumption holds true in a 1104pure 32-bit environment and can hold true in a 64-bit environment if 1105the address space is accurately set to be 32-bit for o32 and n32. 1106</p> 1107</dd> 1108<dt><span><code>--with-madd4</code></span></dt> 1109<dd><p>On MIPS targets, make <samp>-mmadd4</samp> the default when no 1110<samp>-mno-madd4</samp> option is passed. This is the default. 1111</p> 1112</dd> 1113<dt><span><code>--without-madd4</code></span></dt> 1114<dd><p>On MIPS targets, make <samp>-mno-madd4</samp> the default when no 1115<samp>-mmadd4</samp> option is passed. The <code>madd4</code> instruction 1116family can be problematic when targeting a combination of cores that 1117implement these instructions differently. There are two known cores 1118that implement these as fused operations instead of unfused (where 1119unfused is normally expected). Disabling these instructions is the 1120only way to ensure compatible code is generated; this will incur 1121a performance penalty. 1122</p> 1123</dd> 1124<dt><span><code>--with-mips-plt</code></span></dt> 1125<dd><p>On MIPS targets, make use of copy relocations and PLTs. 1126These features are extensions to the traditional 1127SVR4-based MIPS ABIs and require support from GNU binutils 1128and the runtime C library. 1129</p> 1130</dd> 1131<dt><span><code>--with-stack-clash-protection-guard-size=<var>size</var></code></span></dt> 1132<dd><p>On certain targets this option sets the default stack clash protection guard 1133size as a power of two in bytes. On AArch64 <var>size</var> is required to be either 113412 (4KB) or 16 (64KB). 1135</p> 1136</dd> 1137<dt><span><code>--with-isa-spec=<var>ISA-spec-string</var></code></span></dt> 1138<dd><p>On RISC-V targets specify the default version of the RISC-V Unprivileged 1139(formerly User-Level) ISA specification to produce code conforming to. 1140The possibilities for <var>ISA-spec-string</var> are: 1141</p><dl compact="compact"> 1142<dt><span><code>2.2</code></span></dt> 1143<dd><p>Produce code conforming to version 2.2. 1144</p></dd> 1145<dt><span><code>20190608</code></span></dt> 1146<dd><p>Produce code conforming to version 20190608. 1147</p></dd> 1148<dt><span><code>20191213</code></span></dt> 1149<dd><p>Produce code conforming to version 20191213. 1150</p></dd> 1151</dl> 1152<p>In the absence of this configuration option the default version is 20191213. 1153</p> 1154</dd> 1155<dt><span><code>--enable-__cxa_atexit</code></span></dt> 1156<dd><p>Define if you want to use __cxa_atexit, rather than atexit, to 1157register C++ destructors for local statics and global objects. 1158This is essential for fully standards-compliant handling of 1159destructors, but requires __cxa_atexit in libc. This option is currently 1160only available on systems with GNU libc. When enabled, this will cause 1161<samp>-fuse-cxa-atexit</samp> to be passed by default. 1162</p> 1163</dd> 1164<dt><span><code>--enable-gnu-indirect-function</code></span></dt> 1165<dd><p>Define if you want to enable the <code>ifunc</code> attribute. This option is 1166currently only available on systems with GNU libc on certain targets. 1167</p> 1168</dd> 1169<dt><span><code>--enable-target-optspace</code></span></dt> 1170<dd><p>Specify that target 1171libraries should be optimized for code space instead of code speed. 1172This is the default for the m32r platform. 1173</p> 1174</dd> 1175<dt><span><code>--with-cpp-install-dir=<var>dirname</var></code></span></dt> 1176<dd><p>Specify that the user visible <code>cpp</code> program should be installed 1177in <samp><var>prefix</var>/<var>dirname</var>/cpp</samp>, in addition to <var>bindir</var>. 1178</p> 1179</dd> 1180<dt><span><code>--enable-comdat</code></span></dt> 1181<dd><p>Enable COMDAT group support. This is primarily used to override the 1182automatically detected value. 1183</p> 1184</dd> 1185<dt><span><code>--enable-initfini-array</code></span></dt> 1186<dd><p>Force the use of sections <code>.init_array</code> and <code>.fini_array</code> 1187(instead of <code>.init</code> and <code>.fini</code>) for constructors and 1188destructors. Option <samp>--disable-initfini-array</samp> has the 1189opposite effect. If neither option is specified, the configure script 1190will try to guess whether the <code>.init_array</code> and 1191<code>.fini_array</code> sections are supported and, if they are, use them. 1192</p> 1193</dd> 1194<dt><span><code>--enable-link-mutex</code></span></dt> 1195<dd><p>When building GCC, use a mutex to avoid linking the compilers for 1196multiple languages at the same time, to avoid thrashing on build 1197systems with limited free memory. The default is not to use such a mutex. 1198</p> 1199</dd> 1200<dt><span><code>--enable-link-serialization</code></span></dt> 1201<dd><p>When building GCC, use make dependencies to serialize linking the compilers for 1202multiple languages, to avoid thrashing on build 1203systems with limited free memory. The default is not to add such 1204dependencies and thus with parallel make potentially link different 1205compilers concurrently. If the argument is a positive integer, allow 1206that number of concurrent link processes for the large binaries. 1207</p> 1208</dd> 1209<dt><span><code>--enable-maintainer-mode</code></span></dt> 1210<dd><p>The build rules that regenerate the Autoconf and Automake output files as 1211well as the GCC master message catalog <samp>gcc.pot</samp> are normally 1212disabled. This is because it can only be rebuilt if the complete source 1213tree is present. If you have changed the sources and want to rebuild the 1214catalog, configuring with <samp>--enable-maintainer-mode</samp> will enable 1215this. Note that you need a recent version of the <code>gettext</code> tools 1216to do so. 1217</p> 1218</dd> 1219<dt><span><code>--disable-bootstrap</code></span></dt> 1220<dd><p>For a native build, the default configuration is to perform 1221a 3-stage bootstrap of the compiler when ‘<samp>make</samp>’ is invoked, 1222testing that GCC can compile itself correctly. If you want to disable 1223this process, you can configure with <samp>--disable-bootstrap</samp>. 1224</p> 1225</dd> 1226<dt><span><code>--enable-bootstrap</code></span></dt> 1227<dd><p>In special cases, you may want to perform a 3-stage build 1228even if the target and host triplets are different. 1229This is possible when the host can run code compiled for 1230the target (e.g. host is i686-linux, target is i486-linux). 1231Starting from GCC 4.2, to do this you have to configure explicitly 1232with <samp>--enable-bootstrap</samp>. 1233</p> 1234</dd> 1235<dt><span><code>--enable-generated-files-in-srcdir</code></span></dt> 1236<dd><p>Neither the .c and .h files that are generated from Bison and flex nor the 1237info manuals and man pages that are built from the .texi files are present 1238in the repository development tree. When building GCC from that development tree, 1239or from one of our snapshots, those generated files are placed in your 1240build directory, which allows for the source to be in a readonly 1241directory. 1242</p> 1243<p>If you configure with <samp>--enable-generated-files-in-srcdir</samp> then those 1244generated files will go into the source directory. This is mainly intended 1245for generating release or prerelease tarballs of the GCC sources, since it 1246is not a requirement that the users of source releases to have flex, Bison, 1247or makeinfo. 1248</p> 1249</dd> 1250<dt><span><code>--enable-version-specific-runtime-libs</code></span></dt> 1251<dd><p>Specify 1252that runtime libraries should be installed in the compiler specific 1253subdirectory (<samp><var>libdir</var>/gcc</samp>) rather than the usual places. In 1254addition, ‘<samp>libstdc++</samp>’’s include files will be installed into 1255<samp><var>libdir</var></samp> unless you overruled it by using 1256<samp>--with-gxx-include-dir=<var>dirname</var></samp>. Using this option is 1257particularly useful if you intend to use several versions of GCC in 1258parallel. The default is ‘<samp>yes</samp>’ for ‘<samp>libada</samp>’, and ‘<samp>no</samp>’ for 1259the remaining libraries. 1260</p> 1261</dd> 1262<dt><span><code><span id="WithAixSoname"></span>--with-aix-soname=‘<samp>aix</samp>’, ‘<samp>svr4</samp>’ or ‘<samp>both</samp>’</code></span></dt> 1263<dd><p>Traditional AIX shared library versioning (versioned <code>Shared Object</code> 1264files as members of unversioned <code>Archive Library</code> files named 1265‘<samp>lib.a</samp>’) causes numerous headaches for package managers. However, 1266<code>Import Files</code> as members of <code>Archive Library</code> files allow for 1267<strong>filename-based versioning</strong> of shared libraries as seen on Linux/SVR4, 1268where this is called the "SONAME". But as they prevent static linking, 1269<code>Import Files</code> may be used with <code>Runtime Linking</code> only, where the 1270linker does search for ‘<samp>libNAME.so</samp>’ before ‘<samp>libNAME.a</samp>’ library 1271filenames with the ‘<samp>-lNAME</samp>’ linker flag. 1272</p> 1273<span id="AixLdCommand"></span><p>For detailed information please refer to the AIX 1274<a href="https://www.ibm.com/support/knowledgecenter/search/%22the%20ld%20command%2C%20also%20called%20the%20linkage%20editor%20or%20binder%22">ld 1275Command</a> reference. 1276</p> 1277<p>As long as shared library creation is enabled, upon: 1278</p><dl compact="compact"> 1279<dt><span><code>--with-aix-soname=aix</code></span></dt> 1280<dt><span><code>--with-aix-soname=both</code></span></dt> 1281<dd><p>A (traditional AIX) <code>Shared Archive Library</code> file is created: 1282 </p><ul> 1283<li> using the ‘<samp>libNAME.a</samp>’ filename scheme 1284 </li><li> with the <code>Shared Object</code> file as archive member named 1285 ‘<samp>libNAME.so.V</samp>’ (except for ‘<samp>libgcc_s</samp>’, where the <code>Shared 1286 Object</code> file is named ‘<samp>shr.o</samp>’ for backwards compatibility), which 1287 <ul class="no-bullet"> 1288<li>- is used for runtime loading from inside the ‘<samp>libNAME.a</samp>’ file 1289 </li><li>- is used for dynamic loading via 1290 <code>dlopen("libNAME.a(libNAME.so.V)", RTLD_MEMBER)</code> 1291 </li><li>- is used for shared linking 1292 </li><li>- is used for static linking, so no separate <code>Static Archive 1293 Library</code> file is needed 1294 </li></ul> 1295</li></ul> 1296</dd> 1297<dt><span><code>--with-aix-soname=both</code></span></dt> 1298<dt><span><code>--with-aix-soname=svr4</code></span></dt> 1299<dd><p>A (second) <code>Shared Archive Library</code> file is created: 1300 </p><ul> 1301<li> using the ‘<samp>libNAME.so.V</samp>’ filename scheme 1302 </li><li> with the <code>Shared Object</code> file as archive member named 1303 ‘<samp>shr.o</samp>’, which 1304 <ul class="no-bullet"> 1305<li>- is created with the <code>-G linker flag</code> 1306 </li><li>- has the <code>F_LOADONLY</code> flag set 1307 </li><li>- is used for runtime loading from inside the ‘<samp>libNAME.so.V</samp>’ file 1308 </li><li>- is used for dynamic loading via <code>dlopen("libNAME.so.V(shr.o)", 1309 RTLD_MEMBER)</code> 1310 </li></ul> 1311</li><li> with the <code>Import File</code> as archive member named ‘<samp>shr.imp</samp>’, 1312 which 1313 <ul class="no-bullet"> 1314<li>- refers to ‘<samp>libNAME.so.V(shr.o)</samp>’ as the "SONAME", to be recorded 1315 in the <code>Loader Section</code> of subsequent binaries 1316 </li><li>- indicates whether ‘<samp>libNAME.so.V(shr.o)</samp>’ is 32 or 64 bit 1317 </li><li>- lists all the public symbols exported by ‘<samp>lib.so.V(shr.o)</samp>’, 1318 eventually decorated with the <code>‘<samp>weak</samp>’ Keyword</code> 1319 </li><li>- is necessary for shared linking against ‘<samp>lib.so.V(shr.o)</samp>’ 1320 </li></ul> 1321</li></ul> 1322<p>A symbolic link using the ‘<samp>libNAME.so</samp>’ filename scheme is created: 1323 </p><ul> 1324<li> pointing to the ‘<samp>libNAME.so.V</samp>’ <code>Shared Archive Library</code> file 1325 </li><li> to permit the <code>ld Command</code> to find ‘<samp>lib.so.V(shr.imp)</samp>’ via 1326 the ‘<samp>-lNAME</samp>’ argument (requires <code>Runtime Linking</code> to be enabled) 1327 </li><li> to permit dynamic loading of ‘<samp>lib.so.V(shr.o)</samp>’ without the need 1328 to specify the version number via <code>dlopen("libNAME.so(shr.o)", 1329 RTLD_MEMBER)</code> 1330 </li></ul> 1331</dd> 1332</dl> 1333 1334<p>As long as static library creation is enabled, upon: 1335</p><dl compact="compact"> 1336<dt><span><code>--with-aix-soname=svr4</code></span></dt> 1337<dd><p>A <code>Static Archive Library</code> is created: 1338 </p><ul> 1339<li> using the ‘<samp>libNAME.a</samp>’ filename scheme 1340 </li><li> with all the <code>Static Object</code> files as archive members, which 1341 <ul class="no-bullet"> 1342<li>- are used for static linking 1343 </li></ul> 1344</li></ul> 1345</dd> 1346</dl> 1347 1348<p>While the aix-soname=‘<samp>svr4</samp>’ option does not create <code>Shared Object</code> 1349files as members of unversioned <code>Archive Library</code> files any more, package 1350managers still are responsible to 1351<a href="./specific.html#TransferAixShobj">transfer</a> <code>Shared Object</code> files 1352found as member of a previously installed unversioned <code>Archive Library</code> 1353file into the newly installed <code>Archive Library</code> file with the same 1354filename. 1355</p> 1356<p><em>WARNING:</em> Creating <code>Shared Object</code> files with <code>Runtime Linking</code> 1357enabled may bloat the TOC, eventually leading to <code>TOC overflow</code> errors, 1358requiring the use of either the <samp>-Wl,-bbigtoc</samp> linker flag (seen to 1359break with the <code>GDB</code> debugger) or some of the TOC-related compiler flags, 1360see “RS/6000 and PowerPC Options” in the main manual. 1361</p> 1362<p><samp>--with-aix-soname</samp> is currently supported by ‘<samp>libgcc_s</samp>’ only, so 1363this option is still experimental and not for normal use yet. 1364</p> 1365<p>Default is the traditional behavior <samp>--with-aix-soname=‘<samp>aix</samp>’</samp>. 1366</p> 1367</dd> 1368<dt><span><code>--enable-languages=<var>lang1</var>,<var>lang2</var>,…</code></span></dt> 1369<dd><p>Specify that only a particular subset of compilers and 1370their runtime libraries should be built. For a list of valid values for 1371<var>langN</var> you can issue the following command in the 1372<samp>gcc</samp> directory of your GCC source tree:<br> 1373</p><div class="example"> 1374<pre class="example">grep ^language= */config-lang.in 1375</pre></div> 1376<p>Currently, you can use any of the following: 1377<code>all</code>, <code>default</code>, <code>ada</code>, <code>c</code>, <code>c++</code>, <code>d</code>, 1378<code>fortran</code>, <code>go</code>, <code>jit</code>, <code>lto</code>, <code>objc</code>, <code>obj-c++</code>. 1379Building the Ada compiler has special requirements, see below. 1380If you do not pass this flag, or specify the option <code>default</code>, then the 1381default languages available in the <samp>gcc</samp> sub-tree will be configured. 1382Ada, D, Go, Jit, and Objective-C++ are not default languages. LTO is not a 1383default language, but is built by default because <samp>--enable-lto</samp> is 1384enabled by default. The other languages are default languages. If 1385<code>all</code> is specified, then all available languages are built. An 1386exception is <code>jit</code> language, which requires 1387<samp>--enable-host-shared</samp> to be included with <code>all</code>. 1388</p> 1389</dd> 1390<dt><span><code>--enable-stage1-languages=<var>lang1</var>,<var>lang2</var>,…</code></span></dt> 1391<dd><p>Specify that a particular subset of compilers and their runtime 1392libraries should be built with the system C compiler during stage 1 of 1393the bootstrap process, rather than only in later stages with the 1394bootstrapped C compiler. The list of valid values is the same as for 1395<samp>--enable-languages</samp>, and the option <code>all</code> will select all 1396of the languages enabled by <samp>--enable-languages</samp>. This option is 1397primarily useful for GCC development; for instance, when a development 1398version of the compiler cannot bootstrap due to compiler bugs, or when 1399one is debugging front ends other than the C front end. When this 1400option is used, one can then build the target libraries for the 1401specified languages with the stage-1 compiler by using <code>make 1402stage1-bubble all-target</code>, or run the testsuite on the stage-1 compiler 1403for the specified languages using <code>make stage1-start check-gcc</code>. 1404</p> 1405</dd> 1406<dt><span><code>--disable-libada</code></span></dt> 1407<dd><p>Specify that the run-time libraries and tools used by GNAT should not 1408be built. This can be useful for debugging, or for compatibility with 1409previous Ada build procedures, when it was required to explicitly 1410do a ‘<samp>make -C gcc gnatlib_and_tools</samp>’. 1411</p> 1412</dd> 1413<dt><span><code>--disable-libsanitizer</code></span></dt> 1414<dd><p>Specify that the run-time libraries for the various sanitizers should 1415not be built. 1416</p> 1417</dd> 1418<dt><span><code>--disable-libssp</code></span></dt> 1419<dd><p>Specify that the run-time libraries for stack smashing protection 1420should not be built or linked against. On many targets library support 1421is provided by the C library instead. 1422</p> 1423</dd> 1424<dt><span><code>--disable-libquadmath</code></span></dt> 1425<dd><p>Specify that the GCC quad-precision math library should not be built. 1426On some systems, the library is required to be linkable when building 1427the Fortran front end, unless <samp>--disable-libquadmath-support</samp> 1428is used. 1429</p> 1430</dd> 1431<dt><span><code>--disable-libquadmath-support</code></span></dt> 1432<dd><p>Specify that the Fortran front end and <code>libgfortran</code> do not add 1433support for <code>libquadmath</code> on systems supporting it. 1434</p> 1435</dd> 1436<dt><span><code>--disable-libgomp</code></span></dt> 1437<dd><p>Specify that the GNU Offloading and Multi Processing Runtime Library 1438should not be built. 1439</p> 1440</dd> 1441<dt><span><code>--disable-libvtv</code></span></dt> 1442<dd><p>Specify that the run-time libraries used by vtable verification 1443should not be built. 1444</p> 1445</dd> 1446<dt><span><code>--with-dwarf2</code></span></dt> 1447<dd><p>Specify that the compiler should 1448use DWARF 2 debugging information as the default. 1449</p> 1450</dd> 1451<dt><span><code>--with-advance-toolchain=<var>at</var></code></span></dt> 1452<dd><p>On 64-bit PowerPC Linux systems, configure the compiler to use the 1453header files, library files, and the dynamic linker from the Advance 1454Toolchain release <var>at</var> instead of the default versions that are 1455provided by the Linux distribution. In general, this option is 1456intended for the developers of GCC, and it is not intended for general 1457use. 1458</p> 1459</dd> 1460<dt><span><code>--enable-targets=all</code></span></dt> 1461<dt><span><code>--enable-targets=<var>target_list</var></code></span></dt> 1462<dd><p>Some GCC targets, e.g. powerpc64-linux, build bi-arch compilers. 1463These are compilers that are able to generate either 64-bit or 32-bit 1464code. Typically, the corresponding 32-bit target, e.g. 1465powerpc-linux for powerpc64-linux, only generates 32-bit code. This 1466option enables the 32-bit target to be a bi-arch compiler, which is 1467useful when you want a bi-arch compiler that defaults to 32-bit, and 1468you are building a bi-arch or multi-arch binutils in a combined tree. 1469On mips-linux, this will build a tri-arch compiler (ABI o32/n32/64), 1470defaulted to o32. 1471Currently, this option only affects sparc-linux, powerpc-linux, x86-linux, 1472mips-linux and s390-linux. 1473</p> 1474</dd> 1475<dt><span><code>--enable-default-pie</code></span></dt> 1476<dd><p>Turn on <samp>-fPIE</samp> and <samp>-pie</samp> by default. 1477</p> 1478</dd> 1479<dt><span><code>--enable-secureplt</code></span></dt> 1480<dd><p>This option enables <samp>-msecure-plt</samp> by default for powerpc-linux. 1481See “RS/6000 and PowerPC Options” in the main manual 1482</p> 1483</dd> 1484<dt><span><code>--enable-default-ssp</code></span></dt> 1485<dd><p>Turn on <samp>-fstack-protector-strong</samp> by default. 1486</p> 1487</dd> 1488<dt><span><code>--enable-cld</code></span></dt> 1489<dd><p>This option enables <samp>-mcld</samp> by default for 32-bit x86 targets. 1490See “i386 and x86-64 Options” in the main manual 1491</p> 1492</dd> 1493<dt><span><code>--enable-large-address-aware</code></span></dt> 1494<dd><p>The <samp>--enable-large-address-aware</samp> option arranges for MinGW 1495executables to be linked using the <samp>--large-address-aware</samp> 1496option, that enables the use of more than 2GB of memory. If GCC is 1497configured with this option, its effects can be reversed by passing the 1498<samp>-Wl,--disable-large-address-aware</samp> option to the so-configured 1499compiler driver. 1500</p> 1501</dd> 1502<dt><span><code>--enable-win32-registry</code></span></dt> 1503<dt><span><code>--enable-win32-registry=<var>key</var></code></span></dt> 1504<dt><span><code>--disable-win32-registry</code></span></dt> 1505<dd><p>The <samp>--enable-win32-registry</samp> option enables Microsoft Windows-hosted GCC 1506to look up installations paths in the registry using the following key: 1507</p> 1508<div class="example"> 1509<pre class="example"><code>HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\<var>key</var></code> 1510</pre></div> 1511 1512<p><var>key</var> defaults to GCC version number, and can be overridden by the 1513<samp>--enable-win32-registry=<var>key</var></samp> option. Vendors and distributors 1514who use custom installers are encouraged to provide a different key, 1515perhaps one comprised of vendor name and GCC version number, to 1516avoid conflict with existing installations. This feature is enabled 1517by default, and can be disabled by <samp>--disable-win32-registry</samp> 1518option. This option has no effect on the other hosts. 1519</p> 1520</dd> 1521<dt><span><code>--nfp</code></span></dt> 1522<dd><p>Specify that the machine does not have a floating point unit. This 1523option only applies to ‘<samp>m68k-sun-sunos<var>n</var></samp>’. On any other 1524system, <samp>--nfp</samp> has no effect. 1525</p> 1526</dd> 1527<dt><span><code>--enable-werror</code></span></dt> 1528<dt><span><code>--disable-werror</code></span></dt> 1529<dt><span><code>--enable-werror=yes</code></span></dt> 1530<dt><span><code>--enable-werror=no</code></span></dt> 1531<dd><p>When you specify this option, it controls whether certain files in the 1532compiler are built with <samp>-Werror</samp> in bootstrap stage2 and later. 1533If you don’t specify it, <samp>-Werror</samp> is turned on for the main 1534development trunk. However it defaults to off for release branches and 1535final releases. The specific files which get <samp>-Werror</samp> are 1536controlled by the Makefiles. 1537</p> 1538</dd> 1539<dt><span><code>--enable-checking</code></span></dt> 1540<dt><span><code>--disable-checking</code></span></dt> 1541<dt><span><code>--enable-checking=<var>list</var></code></span></dt> 1542<dd><p>This option controls performing internal consistency checks in the compiler. 1543It does not change the generated code, but adds error checking of the 1544requested complexity. This slows down the compiler and may only work 1545properly if you are building the compiler with GCC. 1546</p> 1547<p>When the option is not specified, the active set of checks depends on context. 1548Namely, bootstrap stage 1 defaults to ‘<samp>--enable-checking=yes</samp>’, builds 1549from release branches or release archives default to 1550‘<samp>--enable-checking=release</samp>’, and otherwise 1551‘<samp>--enable-checking=yes,extra</samp>’ is used. When the option is 1552specified without a <var>list</var>, the result is the same as 1553‘<samp>--enable-checking=yes</samp>’. Likewise, ‘<samp>--disable-checking</samp>’ is 1554equivalent to ‘<samp>--enable-checking=no</samp>’. 1555</p> 1556<p>The categories of checks available in <var>list</var> are ‘<samp>yes</samp>’ (most common 1557checks ‘<samp>assert,misc,gc,gimple,rtlflag,runtime,tree,types</samp>’), ‘<samp>no</samp>’ 1558(no checks at all), ‘<samp>all</samp>’ (all but ‘<samp>valgrind</samp>’), ‘<samp>release</samp>’ 1559(cheapest checks ‘<samp>assert,runtime</samp>’) or ‘<samp>none</samp>’ (same as ‘<samp>no</samp>’). 1560‘<samp>release</samp>’ checks are always on and to disable them 1561‘<samp>--disable-checking</samp>’ or ‘<samp>--enable-checking=no[,<other checks>]</samp>’ 1562must be explicitly requested. Disabling assertions makes the compiler and 1563runtime slightly faster but increases the risk of undetected internal errors 1564causing wrong code to be generated. 1565</p> 1566<p>Individual checks can be enabled with these flags: ‘<samp>assert</samp>’, ‘<samp>df</samp>’, 1567‘<samp>extra</samp>’, ‘<samp>fold</samp>’, ‘<samp>gc</samp>’, ‘<samp>gcac</samp>’, ‘<samp>gimple</samp>’, 1568‘<samp>misc</samp>’, ‘<samp>rtl</samp>’, ‘<samp>rtlflag</samp>’, ‘<samp>runtime</samp>’, ‘<samp>tree</samp>’, 1569‘<samp>types</samp>’ and ‘<samp>valgrind</samp>’. ‘<samp>extra</samp>’ extends ‘<samp>misc</samp>’ 1570checking with extra checks that might affect code generation and should 1571therefore not differ between stage1 and later stages in bootstrap. 1572</p> 1573<p>The ‘<samp>valgrind</samp>’ check requires the external <code>valgrind</code> simulator, 1574available from <a href="https://valgrind.org">https://valgrind.org</a>. The ‘<samp>rtl</samp>’ checks are 1575expensive and the ‘<samp>df</samp>’, ‘<samp>gcac</samp>’ and ‘<samp>valgrind</samp>’ checks are very 1576expensive. 1577</p> 1578</dd> 1579<dt><span><code>--disable-stage1-checking</code></span></dt> 1580<dt><span><code>--enable-stage1-checking</code></span></dt> 1581<dt><span><code>--enable-stage1-checking=<var>list</var></code></span></dt> 1582<dd><p>This option affects only bootstrap build. If no <samp>--enable-checking</samp> 1583option is specified the stage1 compiler is built with ‘<samp>yes</samp>’ checking 1584enabled, otherwise the stage1 checking flags are the same as specified by 1585<samp>--enable-checking</samp>. To build the stage1 compiler with 1586different checking options use <samp>--enable-stage1-checking</samp>. 1587The list of checking options is the same as for <samp>--enable-checking</samp>. 1588If your system is too slow or too small to bootstrap a released compiler 1589with checking for stage1 enabled, you can use ‘<samp>--disable-stage1-checking</samp>’ 1590to disable checking for the stage1 compiler. 1591</p> 1592</dd> 1593<dt><span><code>--enable-coverage</code></span></dt> 1594<dt><span><code>--enable-coverage=<var>level</var></code></span></dt> 1595<dd><p>With this option, the compiler is built to collect self coverage 1596information, every time it is run. This is for internal development 1597purposes, and only works when the compiler is being built with gcc. The 1598<var>level</var> argument controls whether the compiler is built optimized or 1599not, values are ‘<samp>opt</samp>’ and ‘<samp>noopt</samp>’. For coverage analysis you 1600want to disable optimization, for performance analysis you want to 1601enable optimization. When coverage is enabled, the default level is 1602without optimization. 1603</p> 1604</dd> 1605<dt><span><code>--enable-gather-detailed-mem-stats</code></span></dt> 1606<dd><p>When this option is specified more detailed information on memory 1607allocation is gathered. This information is printed when using 1608<samp>-fmem-report</samp>. 1609</p> 1610</dd> 1611<dt><span><code>--enable-valgrind-annotations</code></span></dt> 1612<dd><p>Mark selected memory related operations in the compiler when run under 1613valgrind to suppress false positives. 1614</p> 1615</dd> 1616<dt><span><code>--enable-nls</code></span></dt> 1617<dt><span><code>--disable-nls</code></span></dt> 1618<dd><p>The <samp>--enable-nls</samp> option enables Native Language Support (NLS), 1619which lets GCC output diagnostics in languages other than American 1620English. Native Language Support is enabled by default if not doing a 1621canadian cross build. The <samp>--disable-nls</samp> option disables NLS. 1622</p> 1623</dd> 1624<dt><span><code>--with-included-gettext</code></span></dt> 1625<dd><p>If NLS is enabled, the <samp>--with-included-gettext</samp> option causes the build 1626procedure to prefer its copy of GNU <code>gettext</code>. 1627</p> 1628</dd> 1629<dt><span><code>--with-catgets</code></span></dt> 1630<dd><p>If NLS is enabled, and if the host lacks <code>gettext</code> but has the 1631inferior <code>catgets</code> interface, the GCC build procedure normally 1632ignores <code>catgets</code> and instead uses GCC’s copy of the GNU 1633<code>gettext</code> library. The <samp>--with-catgets</samp> option causes the 1634build procedure to use the host’s <code>catgets</code> in this situation. 1635</p> 1636</dd> 1637<dt><span><code>--with-libiconv-prefix=<var>dir</var></code></span></dt> 1638<dd><p>Search for libiconv header files in <samp><var>dir</var>/include</samp> and 1639libiconv library files in <samp><var>dir</var>/lib</samp>. 1640</p> 1641</dd> 1642<dt><span><code>--enable-obsolete</code></span></dt> 1643<dd><p>Enable configuration for an obsoleted system. If you attempt to 1644configure GCC for a system (build, host, or target) which has been 1645obsoleted, and you do not specify this flag, configure will halt with an 1646error message. 1647</p> 1648<p>All support for systems which have been obsoleted in one release of GCC 1649is removed entirely in the next major release, unless someone steps 1650forward to maintain the port. 1651</p> 1652</dd> 1653<dt><span><code>--enable-decimal-float</code></span></dt> 1654<dt><span><code>--enable-decimal-float=yes</code></span></dt> 1655<dt><span><code>--enable-decimal-float=no</code></span></dt> 1656<dt><span><code>--enable-decimal-float=bid</code></span></dt> 1657<dt><span><code>--enable-decimal-float=dpd</code></span></dt> 1658<dt><span><code>--disable-decimal-float</code></span></dt> 1659<dd><p>Enable (or disable) support for the C decimal floating point extension 1660that is in the IEEE 754-2008 standard. This is enabled by default only 1661on PowerPC, i386, and x86_64 GNU/Linux systems. Other systems may also 1662support it, but require the user to specifically enable it. You can 1663optionally control which decimal floating point format is used (either 1664‘<samp>bid</samp>’ or ‘<samp>dpd</samp>’). The ‘<samp>bid</samp>’ (binary integer decimal) 1665format is default on i386 and x86_64 systems, and the ‘<samp>dpd</samp>’ 1666(densely packed decimal) format is default on PowerPC systems. 1667</p> 1668</dd> 1669<dt><span><code>--enable-fixed-point</code></span></dt> 1670<dt><span><code>--disable-fixed-point</code></span></dt> 1671<dd><p>Enable (or disable) support for C fixed-point arithmetic. 1672This option is enabled by default for some targets (such as MIPS) which 1673have hardware-support for fixed-point operations. On other targets, you 1674may enable this option manually. 1675</p> 1676</dd> 1677<dt><span><code>--with-long-double-128</code></span></dt> 1678<dd><p>Specify if <code>long double</code> type should be 128-bit by default on selected 1679GNU/Linux architectures. If using <code>--without-long-double-128</code>, 1680<code>long double</code> will be by default 64-bit, the same as <code>double</code> type. 1681When neither of these configure options are used, the default will be 1682128-bit <code>long double</code> when built against GNU C Library 2.4 and later, 168364-bit <code>long double</code> otherwise. 1684</p> 1685</dd> 1686<dt><span><code>--with-long-double-format=ibm</code></span></dt> 1687<dt><span><code>--with-long-double-format=ieee</code></span></dt> 1688<dd><p>Specify whether <code>long double</code> uses the IBM extended double format 1689or the IEEE 128-bit floating point format on PowerPC Linux systems. 1690This configuration switch will only work on little endian PowerPC 1691Linux systems and on big endian 64-bit systems where the default cpu 1692is at least power7 (i.e. <samp>--with-cpu=power7</samp>, 1693<samp>--with-cpu=power8</samp>, or <samp>--with-cpu=power9</samp> is used). 1694</p> 1695<p>If you use the <samp>--with-long-double-64</samp> configuration option, 1696the <samp>--with-long-double-format=ibm</samp> and 1697<samp>--with-long-double-format=ieee</samp> options are ignored. 1698</p> 1699<p>The default <code>long double</code> format is to use IBM extended double. 1700Until all of the libraries are converted to use IEEE 128-bit floating 1701point, it is not recommended to use 1702<samp>--with-long-double-format=ieee</samp>. 1703</p> 1704</dd> 1705<dt><span><code>--enable-fdpic</code></span></dt> 1706<dd><p>On SH Linux systems, generate ELF FDPIC code. 1707</p> 1708</dd> 1709<dt><span><code>--with-gmp=<var>pathname</var></code></span></dt> 1710<dt><span><code>--with-gmp-include=<var>pathname</var></code></span></dt> 1711<dt><span><code>--with-gmp-lib=<var>pathname</var></code></span></dt> 1712<dt><span><code>--with-mpfr=<var>pathname</var></code></span></dt> 1713<dt><span><code>--with-mpfr-include=<var>pathname</var></code></span></dt> 1714<dt><span><code>--with-mpfr-lib=<var>pathname</var></code></span></dt> 1715<dt><span><code>--with-mpc=<var>pathname</var></code></span></dt> 1716<dt><span><code>--with-mpc-include=<var>pathname</var></code></span></dt> 1717<dt><span><code>--with-mpc-lib=<var>pathname</var></code></span></dt> 1718<dd><p>If you want to build GCC but do not have the GMP library, the MPFR 1719library and/or the MPC library installed in a standard location and 1720do not have their sources present in the GCC source tree then you 1721can explicitly specify the directory where they are installed 1722(‘<samp>--with-gmp=<var>gmpinstalldir</var></samp>’, 1723‘<samp>--with-mpfr=<var>mpfrinstalldir</var></samp>’, 1724‘<samp>--with-mpc=<var>mpcinstalldir</var></samp>’). The 1725<samp>--with-gmp=<var>gmpinstalldir</var></samp> option is shorthand for 1726<samp>--with-gmp-lib=<var>gmpinstalldir</var>/lib</samp> and 1727<samp>--with-gmp-include=<var>gmpinstalldir</var>/include</samp>. Likewise the 1728<samp>--with-mpfr=<var>mpfrinstalldir</var></samp> option is shorthand for 1729<samp>--with-mpfr-lib=<var>mpfrinstalldir</var>/lib</samp> and 1730<samp>--with-mpfr-include=<var>mpfrinstalldir</var>/include</samp>, also the 1731<samp>--with-mpc=<var>mpcinstalldir</var></samp> option is shorthand for 1732<samp>--with-mpc-lib=<var>mpcinstalldir</var>/lib</samp> and 1733<samp>--with-mpc-include=<var>mpcinstalldir</var>/include</samp>. If these 1734shorthand assumptions are not correct, you can use the explicit 1735include and lib options directly. You might also need to ensure the 1736shared libraries can be found by the dynamic linker when building and 1737using GCC, for example by setting the runtime shared library path 1738variable (<code>LD_LIBRARY_PATH</code> on GNU/Linux and Solaris systems). 1739</p> 1740<p>These flags are applicable to the host platform only. When building 1741a cross compiler, they will not be used to configure target libraries. 1742</p> 1743</dd> 1744<dt><span><code>--with-isl=<var>pathname</var></code></span></dt> 1745<dt><span><code>--with-isl-include=<var>pathname</var></code></span></dt> 1746<dt><span><code>--with-isl-lib=<var>pathname</var></code></span></dt> 1747<dd><p>If you do not have the isl library installed in a standard location and you 1748want to build GCC, you can explicitly specify the directory where it is 1749installed (‘<samp>--with-isl=<var>islinstalldir</var></samp>’). The 1750<samp>--with-isl=<var>islinstalldir</var></samp> option is shorthand for 1751<samp>--with-isl-lib=<var>islinstalldir</var>/lib</samp> and 1752<samp>--with-isl-include=<var>islinstalldir</var>/include</samp>. If this 1753shorthand assumption is not correct, you can use the explicit 1754include and lib options directly. 1755</p> 1756<p>These flags are applicable to the host platform only. When building 1757a cross compiler, they will not be used to configure target libraries. 1758</p> 1759</dd> 1760<dt><span><code>--with-stage1-ldflags=<var>flags</var></code></span></dt> 1761<dd><p>This option may be used to set linker flags to be used when linking 1762stage 1 of GCC. These are also used when linking GCC if configured with 1763<samp>--disable-bootstrap</samp>. If <samp>--with-stage1-libs</samp> is not set to a 1764value, then the default is ‘<samp>-static-libstdc++ -static-libgcc</samp>’, if 1765supported. 1766</p> 1767</dd> 1768<dt><span><code>--with-stage1-libs=<var>libs</var></code></span></dt> 1769<dd><p>This option may be used to set libraries to be used when linking stage 1 1770of GCC. These are also used when linking GCC if configured with 1771<samp>--disable-bootstrap</samp>. 1772</p> 1773</dd> 1774<dt><span><code>--with-boot-ldflags=<var>flags</var></code></span></dt> 1775<dd><p>This option may be used to set linker flags to be used when linking 1776stage 2 and later when bootstrapping GCC. If –with-boot-libs 1777is not is set to a value, then the default is 1778‘<samp>-static-libstdc++ -static-libgcc</samp>’. 1779</p> 1780</dd> 1781<dt><span><code>--with-boot-libs=<var>libs</var></code></span></dt> 1782<dd><p>This option may be used to set libraries to be used when linking stage 2 1783and later when bootstrapping GCC. 1784</p> 1785</dd> 1786<dt><span><code>--with-debug-prefix-map=<var>map</var></code></span></dt> 1787<dd><p>Convert source directory names using <samp>-fdebug-prefix-map</samp> when 1788building runtime libraries. ‘<samp><var>map</var></samp>’ is a space-separated 1789list of maps of the form ‘<samp><var>old</var>=<var>new</var></samp>’. 1790</p> 1791</dd> 1792<dt><span><code>--enable-linker-build-id</code></span></dt> 1793<dd><p>Tells GCC to pass <samp>--build-id</samp> option to the linker for all final 1794links (links performed without the <samp>-r</samp> or <samp>--relocatable</samp> 1795option), if the linker supports it. If you specify 1796<samp>--enable-linker-build-id</samp>, but your linker does not 1797support <samp>--build-id</samp> option, a warning is issued and the 1798<samp>--enable-linker-build-id</samp> option is ignored. The default is off. 1799</p> 1800</dd> 1801<dt><span><code>--with-linker-hash-style=<var>choice</var></code></span></dt> 1802<dd><p>Tells GCC to pass <samp>--hash-style=<var>choice</var></samp> option to the 1803linker for all final links. <var>choice</var> can be one of 1804‘<samp>sysv</samp>’, ‘<samp>gnu</samp>’, and ‘<samp>both</samp>’ where ‘<samp>sysv</samp>’ is the default. 1805</p> 1806</dd> 1807<dt><span><code>--enable-gnu-unique-object</code></span></dt> 1808<dt><span><code>--disable-gnu-unique-object</code></span></dt> 1809<dd><p>Tells GCC to use the gnu_unique_object relocation for C++ template 1810static data members and inline function local statics. Enabled by 1811default for a toolchain with an assembler that accepts it and 1812GLIBC 2.11 or above, otherwise disabled. 1813</p> 1814</dd> 1815<dt><span><code>--with-diagnostics-color=<var>choice</var></code></span></dt> 1816<dd><p>Tells GCC to use <var>choice</var> as the default for <samp>-fdiagnostics-color=</samp> 1817option (if not used explicitly on the command line). <var>choice</var> 1818can be one of ‘<samp>never</samp>’, ‘<samp>auto</samp>’, ‘<samp>always</samp>’, and ‘<samp>auto-if-env</samp>’ 1819where ‘<samp>auto</samp>’ is the default. ‘<samp>auto-if-env</samp>’ makes 1820<samp>-fdiagnostics-color=auto</samp> the default if <code>GCC_COLORS</code> 1821is present and non-empty in the environment of the compiler, and 1822<samp>-fdiagnostics-color=never</samp> otherwise. 1823</p> 1824</dd> 1825<dt><span><code>--with-diagnostics-urls=<var>choice</var></code></span></dt> 1826<dd><p>Tells GCC to use <var>choice</var> as the default for <samp>-fdiagnostics-urls=</samp> 1827option (if not used explicitly on the command line). <var>choice</var> 1828can be one of ‘<samp>never</samp>’, ‘<samp>auto</samp>’, ‘<samp>always</samp>’, and ‘<samp>auto-if-env</samp>’ 1829where ‘<samp>auto</samp>’ is the default. ‘<samp>auto-if-env</samp>’ makes 1830<samp>-fdiagnostics-urls=auto</samp> the default if <code>GCC_URLS</code> 1831or <code>TERM_URLS</code> is present and non-empty in the environment of the 1832compiler, and <samp>-fdiagnostics-urls=never</samp> otherwise. 1833</p> 1834</dd> 1835<dt><span><code>--enable-lto</code></span></dt> 1836<dt><span><code>--disable-lto</code></span></dt> 1837<dd><p>Enable support for link-time optimization (LTO). This is enabled by 1838default, and may be disabled using <samp>--disable-lto</samp>. 1839</p> 1840</dd> 1841<dt><span><code>--enable-linker-plugin-configure-flags=FLAGS</code></span></dt> 1842<dt><span><code>--enable-linker-plugin-flags=FLAGS</code></span></dt> 1843<dd><p>By default, linker plugins (such as the LTO plugin) are built for the 1844host system architecture. For the case that the linker has a 1845different (but run-time compatible) architecture, these flags can be 1846specified to build plugins that are compatible to the linker. For 1847example, if you are building GCC for a 64-bit x86_64 1848(‘<samp>x86_64-pc-linux-gnu</samp>’) host system, but have a 32-bit x86 1849GNU/Linux (‘<samp>i686-pc-linux-gnu</samp>’) linker executable (which is 1850executable on the former system), you can configure GCC as follows for 1851getting compatible linker plugins: 1852</p> 1853<div class="example"> 1854<pre class="example">% <var>srcdir</var>/configure \ 1855 --host=x86_64-pc-linux-gnu \ 1856 --enable-linker-plugin-configure-flags=--host=i686-pc-linux-gnu \ 1857 --enable-linker-plugin-flags='CC=gcc\ -m32\ -Wl,-rpath,[...]/i686-pc-linux-gnu/lib' 1858</pre></div> 1859 1860</dd> 1861<dt><span><code>--with-plugin-ld=<var>pathname</var></code></span></dt> 1862<dd><p>Enable an alternate linker to be used at link-time optimization (LTO) 1863link time when <samp>-fuse-linker-plugin</samp> is enabled. 1864This linker should have plugin support such as gold starting with 1865version 2.20 or GNU ld starting with version 2.21. 1866See <samp>-fuse-linker-plugin</samp> for details. 1867</p> 1868</dd> 1869<dt><span><code>--enable-canonical-system-headers</code></span></dt> 1870<dt><span><code>--disable-canonical-system-headers</code></span></dt> 1871<dd><p>Enable system header path canonicalization for <samp>libcpp</samp>. This can 1872produce shorter header file paths in diagnostics and dependency output 1873files, but these changed header paths may conflict with some compilation 1874environments. Enabled by default, and may be disabled using 1875<samp>--disable-canonical-system-headers</samp>. 1876</p> 1877</dd> 1878<dt><span><code>--with-glibc-version=<var>major</var>.<var>minor</var></code></span></dt> 1879<dd><p>Tell GCC that when the GNU C Library (glibc) is used on the target it 1880will be version <var>major</var>.<var>minor</var> or later. Normally this can 1881be detected from the C library’s header files, but this option may be 1882needed when bootstrapping a cross toolchain without the header files 1883available for building the initial bootstrap compiler. 1884</p> 1885<p>If GCC is configured with some multilibs that use glibc and some that 1886do not, this option applies only to the multilibs that use glibc. 1887However, such configurations may not work well as not all the relevant 1888configuration in GCC is on a per-multilib basis. 1889</p> 1890</dd> 1891<dt><span><code>--enable-as-accelerator-for=<var>target</var></code></span></dt> 1892<dd><p>Build as offload target compiler. Specify offload host triple by <var>target</var>. 1893</p> 1894</dd> 1895<dt><span><code>--enable-offload-targets=<var>target1</var>[=<var>path1</var>],…,<var>targetN</var>[=<var>pathN</var>]</code></span></dt> 1896<dd><p>Enable offloading to targets <var>target1</var>, …, <var>targetN</var>. 1897Offload compilers are expected to be already installed. Default search 1898path for them is <samp><var>exec-prefix</var></samp>, but it can be changed by 1899specifying paths <var>path1</var>, …, <var>pathN</var>. 1900</p> 1901<div class="example"> 1902<pre class="example">% <var>srcdir</var>/configure \ 1903 --enable-offload-targets=x86_64-intelmicemul-linux-gnu=/path/to/x86_64/compiler,nvptx-none 1904</pre></div> 1905 1906</dd> 1907<dt><span><code>--enable-offload-defaulted</code></span></dt> 1908<dd> 1909<p>Tell GCC that configured but not installed offload compilers and libgomp 1910plugins are silently ignored. Useful for distribution compilers where 1911those are in separate optional packages and where the presence or absence 1912of those optional packages should determine the actual supported offloading 1913target set rather than the GCC configure-time selection. 1914</p> 1915</dd> 1916<dt><span><code>--with-hsa-runtime=<var>pathname</var></code></span></dt> 1917<dt><span><code>--with-hsa-runtime-include=<var>pathname</var></code></span></dt> 1918<dt><span><code>--with-hsa-runtime-lib=<var>pathname</var></code></span></dt> 1919<dd> 1920<p>If you configure GCC with offloading which uses an HSA run-time such as 1921AMDGCN but do not have the HSA run-time library installed in a standard 1922location then you can explicitly specify the directory where they are 1923installed. The <samp>--with-hsa-runtime=<var>hsainstalldir</var></samp> option 1924is a shorthand for 1925<samp>--with-hsa-runtime-lib=<var>hsainstalldir</var>/lib</samp> and 1926<samp>--with-hsa-runtime-include=<var>hsainstalldir</var>/include</samp>. 1927</p> 1928</dd> 1929<dt><span><code>--enable-cet</code></span></dt> 1930<dt><span><code>--disable-cet</code></span></dt> 1931<dd><p>Enable building target run-time libraries with control-flow 1932instrumentation, see <samp>-fcf-protection</samp> option. When 1933<code>--enable-cet</code> is specified target libraries are configured 1934to add <samp>-fcf-protection</samp> and, if needed, other target 1935specific options to a set of building options. 1936</p> 1937<p><code>--enable-cet=auto</code> is default. CET is enabled on Linux/x86 if 1938target binutils supports <code>Intel CET</code> instructions and disabled 1939otherwise. In this case, the target libraries are configured to get 1940additional <samp>-fcf-protection</samp> option. 1941</p> 1942</dd> 1943<dt><span><code>--with-riscv-attribute=‘<samp>yes</samp>’, ‘<samp>no</samp>’ or ‘<samp>default</samp>’</code></span></dt> 1944<dd><p>Generate RISC-V attribute by default, in order to record extra build 1945information in object. 1946</p> 1947<p>The option is disabled by default. It is enabled on RISC-V/ELF (bare-metal) 1948target if target binutils supported. 1949</p> 1950</dd> 1951<dt><span><code>--enable-s390-excess-float-precision</code></span></dt> 1952<dt><span><code>--disable-s390-excess-float-precision</code></span></dt> 1953<dd><p>On s390(x) targets, enable treatment of float expressions with double precision 1954when in standards-compliant mode (e.g., when <code>--std=c99</code> or 1955<code>-fexcess-precision=standard</code> are given). 1956</p> 1957<p>For a native build and cross compiles that have target headers, the option’s 1958default is derived from glibc’s behavior. When glibc clamps float_t to double, 1959GCC follows and enables the option. For other cross compiles, the default is 1960disabled. 1961</p> 1962</dd> 1963<dt><span><code>--with-zstd=<var>pathname</var></code></span></dt> 1964<dt><span><code>--with-zstd-include=<var>pathname</var></code></span></dt> 1965<dt><span><code>--with-zstd-lib=<var>pathname</var></code></span></dt> 1966<dd><p>If you do not have the <code>zstd</code> library installed in a standard 1967location and you want to build GCC, you can explicitly specify the 1968directory where it is installed (‘<samp>--with-zstd=<var>zstdinstalldir</var></samp>’). 1969The <samp>--with-zstd=<var>zstdinstalldir</var></samp> option is shorthand for 1970<samp>--with-zstd-lib=<var>zstdinstalldir</var>/lib</samp> and 1971<samp>--with-zstd-include=<var>zstdinstalldir</var>/include</samp>. If this 1972shorthand assumption is not correct, you can use the explicit 1973include and lib options directly. 1974</p> 1975<p>These flags are applicable to the host platform only. When building 1976a cross compiler, they will not be used to configure target libraries. 1977</p></dd> 1978</dl> 1979 1980<span id="Cross-Compiler-Specific-Options"></span><h4 class="subheading">Cross-Compiler-Specific Options</h4> 1981<p>The following options only apply to building cross compilers. 1982</p> 1983<dl compact="compact"> 1984<dt><span><code>--with-toolexeclibdir=<var>dir</var></code></span></dt> 1985<dd><p>Specify the installation directory for libraries built with a cross compiler. 1986The default is <samp>${gcc_tooldir}/lib</samp>. 1987</p> 1988</dd> 1989<dt><span><code>--with-sysroot</code></span></dt> 1990<dt><span><code>--with-sysroot=<var>dir</var></code></span></dt> 1991<dd><p>Tells GCC to consider <var>dir</var> as the root of a tree that contains 1992(a subset of) the root filesystem of the target operating system. 1993Target system headers, libraries and run-time object files will be 1994searched for in there. More specifically, this acts as if 1995<samp>--sysroot=<var>dir</var></samp> was added to the default options of the built 1996compiler. The specified directory is not copied into the 1997install tree, unlike the options <samp>--with-headers</samp> and 1998<samp>--with-libs</samp> that this option obsoletes. The default value, 1999in case <samp>--with-sysroot</samp> is not given an argument, is 2000<samp>${gcc_tooldir}/sys-root</samp>. If the specified directory is a 2001subdirectory of <samp>${exec_prefix}</samp>, then it will be found relative to 2002the GCC binaries if the installation tree is moved. 2003</p> 2004<p>This option affects the system root for the compiler used to build 2005target libraries (which runs on the build system) and the compiler newly 2006installed with <code>make install</code>; it does not affect the compiler which is 2007used to build GCC itself. 2008</p> 2009<p>If you specify the <samp>--with-native-system-header-dir=<var>dirname</var></samp> 2010option then the compiler will search that directory within <var>dirname</var> for 2011native system headers rather than the default <samp>/usr/include</samp>. 2012</p> 2013</dd> 2014<dt><span><code>--with-build-sysroot</code></span></dt> 2015<dt><span><code>--with-build-sysroot=<var>dir</var></code></span></dt> 2016<dd><p>Tells GCC to consider <var>dir</var> as the system root (see 2017<samp>--with-sysroot</samp>) while building target libraries, instead of 2018the directory specified with <samp>--with-sysroot</samp>. This option is 2019only useful when you are already using <samp>--with-sysroot</samp>. You 2020can use <samp>--with-build-sysroot</samp> when you are configuring with 2021<samp>--prefix</samp> set to a directory that is different from the one in 2022which you are installing GCC and your target libraries. 2023</p> 2024<p>This option affects the system root for the compiler used to build 2025target libraries (which runs on the build system); it does not affect 2026the compiler which is used to build GCC itself. 2027</p> 2028<p>If you specify the <samp>--with-native-system-header-dir=<var>dirname</var></samp> 2029option then the compiler will search that directory within <var>dirname</var> for 2030native system headers rather than the default <samp>/usr/include</samp>. 2031</p> 2032</dd> 2033<dt><span><code>--with-headers</code></span></dt> 2034<dt><span><code>--with-headers=<var>dir</var></code></span></dt> 2035<dd><p>Deprecated in favor of <samp>--with-sysroot</samp>. 2036Specifies that target headers are available when building a cross compiler. 2037The <var>dir</var> argument specifies a directory which has the target include 2038files. These include files will be copied into the <samp>gcc</samp> install 2039directory. <em>This option with the <var>dir</var> argument is required</em> when 2040building a cross compiler, if <samp><var>prefix</var>/<var>target</var>/sys-include</samp> 2041doesn’t pre-exist. If <samp><var>prefix</var>/<var>target</var>/sys-include</samp> does 2042pre-exist, the <var>dir</var> argument may be omitted. <code>fixincludes</code> 2043will be run on these files to make them compatible with GCC. 2044</p> 2045</dd> 2046<dt><span><code>--without-headers</code></span></dt> 2047<dd><p>Tells GCC not use any target headers from a libc when building a cross 2048compiler. When crossing to GNU/Linux, you need the headers so GCC 2049can build the exception handling for libgcc. 2050</p> 2051</dd> 2052<dt><span><code>--with-libs</code></span></dt> 2053<dt><span><code>--with-libs="<var>dir1</var> <var>dir2</var> … <var>dirN</var>"</code></span></dt> 2054<dd><p>Deprecated in favor of <samp>--with-sysroot</samp>. 2055Specifies a list of directories which contain the target runtime 2056libraries. These libraries will be copied into the <samp>gcc</samp> install 2057directory. If the directory list is omitted, this option has no 2058effect. 2059</p> 2060</dd> 2061<dt><span><code>--with-newlib</code></span></dt> 2062<dd><p>Specifies that ‘<samp>newlib</samp>’ is 2063being used as the target C library. This causes <code>__eprintf</code> to be 2064omitted from <samp>libgcc.a</samp> on the assumption that it will be provided by 2065‘<samp>newlib</samp>’. 2066</p> 2067<a name="avr"></a> 2068</dd> 2069<dt><span><code>--with-avrlibc</code></span></dt> 2070<dd><p>Only supported for the AVR target. Specifies that ‘<samp>AVR-Libc</samp>’ is 2071being used as the target C library. This causes float support 2072functions like <code>__addsf3</code> to be omitted from <samp>libgcc.a</samp> on 2073the assumption that it will be provided by <samp>libm.a</samp>. For more 2074technical details, cf. <a href="https://gcc.gnu.org/PR54461">PR54461</a>. 2075It is not supported for 2076RTEMS configurations, which currently use newlib. The option is 2077supported since version 4.7.2 and is the default in 4.8.0 and newer. 2078</p> 2079</dd> 2080<dt><span><code>--with-double={32|64|32,64|64,32}</code></span></dt> 2081<dt><span><code>--with-long-double={32|64|32,64|64,32|double}</code></span></dt> 2082<dd><p>Only supported for the AVR target since version 10. 2083Specify the default layout available for the C/C++ ‘<samp>double</samp>’ 2084and ‘<samp>long double</samp>’ type, respectively. The following rules apply: 2085</p><ul> 2086<li> The first value after the ‘<samp>=</samp>’ specifies the default layout (in bits) 2087of the type and also the default for the <samp>-mdouble=</samp> resp. 2088<samp>-mlong-double=</samp> compiler option. 2089</li><li> If more than one value is specified, respective multilib variants are 2090available, and <samp>-mdouble=</samp> resp. <samp>-mlong-double=</samp> acts 2091as a multilib option. 2092</li><li> If <samp>--with-long-double=double</samp> is specified, ‘<samp>double</samp>’ and 2093‘<samp>long double</samp>’ will have the same layout. 2094</li><li> The defaults are <samp>--with-long-double=64,32</samp> and 2095<samp>--with-double=32,64</samp>. The default ‘<samp>double</samp>’ layout imposed by 2096the latter is compatible with older versions of the compiler that implement 2097‘<samp>double</samp>’ as a 32-bit type, which does not comply to the language standard. 2098</li></ul> 2099<p>Not all combinations of <samp>--with-double=</samp> and 2100<samp>--with-long-double=</samp> are valid. For example, the combination 2101<samp>--with-double=32,64</samp> <samp>--with-long-double=32</samp> will be 2102rejected because the first option specifies the availability of 2103multilibs for ‘<samp>double</samp>’, whereas the second option implies 2104that ‘<samp>long double</samp>’ — and hence also ‘<samp>double</samp>’ — is always 210532 bits wide. 2106</p> 2107</dd> 2108<dt><span><code>--with-double-comparison={tristate|bool|libf7}</code></span></dt> 2109<dd><p>Only supported for the AVR target since version 10. 2110Specify what result format is returned by library functions that 2111compare 64-bit floating point values (<code>DFmode</code>). 2112The GCC default is ‘<samp>tristate</samp>’. If the floating point 2113implementation returns a boolean instead, set it to ‘<samp>bool</samp>’. 2114</p> 2115</dd> 2116<dt><span><code>--with-libf7={libgcc|math|math-symbols|no}</code></span></dt> 2117<dd><p>Only supported for the AVR target since version 10. 2118Specify to which degree code from LibF7 is included in libgcc. 2119LibF7 is an ad-hoc, AVR-specific, 64-bit floating point emulation 2120written in C and (inline) assembly. ‘<samp>libgcc</samp>’ adds support 2121for functions that one would usually expect in libgcc like double addition, 2122double comparisons and double conversions. ‘<samp>math</samp>’ also adds routines 2123that one would expect in <samp>libm.a</samp>, but with <code>__</code> (two underscores) 2124prepended to the symbol names as specified by <samp>math.h</samp>. 2125‘<samp>math-symbols</samp>’ also defines weak aliases for the functions 2126declared in <samp>math.h</samp>. However, <code>--with-libf7</code> won’t 2127install no <samp>math.h</samp> header file whatsoever, this file must come 2128from elsewhere. This option sets <samp>--with-double-comparison</samp> 2129to ‘<samp>bool</samp>’. 2130</p> 2131</dd> 2132<dt><span><code>--with-nds32-lib=<var>library</var></code></span></dt> 2133<dd><p>Specifies that <var>library</var> setting is used for building <samp>libgcc.a</samp>. 2134Currently, the valid <var>library</var> is ‘<samp>newlib</samp>’ or ‘<samp>mculib</samp>’. 2135This option is only supported for the NDS32 target. 2136</p> 2137</dd> 2138<dt><span><code>--with-build-time-tools=<var>dir</var></code></span></dt> 2139<dd><p>Specifies where to find the set of target tools (assembler, linker, etc.) 2140that will be used while building GCC itself. This option can be useful 2141if the directory layouts are different between the system you are building 2142GCC on, and the system where you will deploy it. 2143</p> 2144<p>For example, on an ‘<samp>ia64-hp-hpux</samp>’ system, you may have the GNU 2145assembler and linker in <samp>/usr/bin</samp>, and the native tools in a 2146different path, and build a toolchain that expects to find the 2147native tools in <samp>/usr/bin</samp>. 2148</p> 2149<p>When you use this option, you should ensure that <var>dir</var> includes 2150<code>ar</code>, <code>as</code>, <code>ld</code>, <code>nm</code>, 2151<code>ranlib</code> and <code>strip</code> if necessary, and possibly 2152<code>objdump</code>. Otherwise, GCC may use an inconsistent set of 2153tools. 2154</p></dd> 2155</dl> 2156 2157<span id="Overriding-configure-test-results"></span><h4 class="subsubheading">Overriding <code>configure</code> test results</h4> 2158 2159<p>Sometimes, it might be necessary to override the result of some 2160<code>configure</code> test, for example in order to ease porting to a new 2161system or work around a bug in a test. The toplevel <code>configure</code> 2162script provides three variables for this: 2163</p> 2164<dl compact="compact"> 2165<dt id='index-build_005fconfigargs'><span><code>build_configargs</code><a href='#index-build_005fconfigargs' class='copiable-anchor'> ¶</a></span></dt> 2166<dd><p>The contents of this variable is passed to all build <code>configure</code> 2167scripts. 2168</p> 2169</dd> 2170<dt id='index-host_005fconfigargs'><span><code>host_configargs</code><a href='#index-host_005fconfigargs' class='copiable-anchor'> ¶</a></span></dt> 2171<dd><p>The contents of this variable is passed to all host <code>configure</code> 2172scripts. 2173</p> 2174</dd> 2175<dt id='index-target_005fconfigargs'><span><code>target_configargs</code><a href='#index-target_005fconfigargs' class='copiable-anchor'> ¶</a></span></dt> 2176<dd><p>The contents of this variable is passed to all target <code>configure</code> 2177scripts. 2178</p> 2179</dd> 2180</dl> 2181 2182<p>In order to avoid shell and <code>make</code> quoting issues for complex 2183overrides, you can pass a setting for <code>CONFIG_SITE</code> and set 2184variables in the site file. 2185</p> 2186<span id="Objective-C-Specific-Options"></span><h4 class="subheading">Objective-C-Specific Options</h4> 2187 2188<p>The following options apply to the build of the Objective-C runtime library. 2189</p> 2190<dl compact="compact"> 2191<dt><span><code>--enable-objc-gc</code></span></dt> 2192<dd><p>Specify that an additional variant of the GNU Objective-C runtime library 2193is built, using an external build of the Boehm-Demers-Weiser garbage 2194collector (<a href="https://www.hboehm.info/gc/">https://www.hboehm.info/gc/</a>). This library needs to be 2195available for each multilib variant, unless configured with 2196<samp>--enable-objc-gc=‘<samp>auto</samp>’</samp> in which case the build of the 2197additional runtime library is skipped when not available and the build 2198continues. 2199</p> 2200</dd> 2201<dt><span><code>--with-target-bdw-gc=<var>list</var></code></span></dt> 2202<dt><span><code>--with-target-bdw-gc-include=<var>list</var></code></span></dt> 2203<dt><span><code>--with-target-bdw-gc-lib=<var>list</var></code></span></dt> 2204<dd><p>Specify search directories for the garbage collector header files and 2205libraries. <var>list</var> is a comma separated list of key value pairs of the 2206form ‘<samp><var>multilibdir</var>=<var>path</var></samp>’, where the default multilib key 2207is named as ‘<samp>.</samp>’ (dot), or is omitted (e.g. 2208‘<samp>--with-target-bdw-gc=/opt/bdw-gc,32=/opt-bdw-gc32</samp>’). 2209</p> 2210<p>The options <samp>--with-target-bdw-gc-include</samp> and 2211<samp>--with-target-bdw-gc-lib</samp> must always be specified together 2212for each multilib variant and they take precedence over 2213<samp>--with-target-bdw-gc</samp>. If <samp>--with-target-bdw-gc-include</samp> 2214is missing values for a multilib, then the value for the default 2215multilib is used (e.g. ‘<samp>--with-target-bdw-gc-include=/opt/bdw-gc/include</samp>’ 2216‘<samp>--with-target-bdw-gc-lib=/opt/bdw-gc/lib64,32=/opt-bdw-gc/lib32</samp>’). 2217If none of these options are specified, the library is assumed in 2218default locations. 2219</p></dd> 2220</dl> 2221 2222<span id="D-Specific-Options"></span><h4 class="subheading">D-Specific Options</h4> 2223 2224<p>The following options apply to the build of the D runtime library. 2225</p> 2226<dl compact="compact"> 2227<dt><span><code>--enable-libphobos-checking</code></span></dt> 2228<dt><span><code>--disable-libphobos-checking</code></span></dt> 2229<dt><span><code>--enable-libphobos-checking=<var>list</var></code></span></dt> 2230<dd><p>This option controls whether run-time checks and contracts are compiled into 2231the D runtime library. When the option is not specified, the library is built 2232with ‘<samp>release</samp>’ checking. When the option is specified without a 2233<var>list</var>, the result is the same as ‘<samp>--enable-libphobos-checking=yes</samp>’. 2234Likewise, ‘<samp>--disable-libphobos-checking</samp>’ is equivalent to 2235‘<samp>--enable-libphobos-checking=no</samp>’. 2236</p> 2237<p>The categories of checks available in <var>list</var> are ‘<samp>yes</samp>’ (compiles 2238libphobos with <samp>-fno-release</samp>), ‘<samp>no</samp>’ (compiles libphobos with 2239<samp>-frelease</samp>), ‘<samp>all</samp>’ (same as ‘<samp>yes</samp>’), ‘<samp>none</samp>’ or 2240‘<samp>release</samp>’ (same as ‘<samp>no</samp>’). 2241</p> 2242<p>Individual checks available in <var>list</var> are ‘<samp>assert</samp>’ (compiles libphobos 2243with an extra option <samp>-fassert</samp>). 2244</p> 2245</dd> 2246<dt><span><code>--with-libphobos-druntime-only</code></span></dt> 2247<dt><span><code>--with-libphobos-druntime-only=<var>choice</var></code></span></dt> 2248<dd><p>Specify whether to build only the core D runtime library (druntime), or both 2249the core and standard library (phobos) into libphobos. This is useful for 2250targets that have full support in druntime, but no or incomplete support 2251in phobos. <var>choice</var> can be one of ‘<samp>auto</samp>’, ‘<samp>yes</samp>’, and ‘<samp>no</samp>’ 2252where ‘<samp>auto</samp>’ is the default. 2253</p> 2254<p>When the option is not specified, the default choice ‘<samp>auto</samp>’ means that it 2255is inferred whether the target has support for the phobos standard library. 2256When the option is specified without a <var>choice</var>, the result is the same as 2257‘<samp>--with-libphobos-druntime-only=yes</samp>’. 2258</p> 2259</dd> 2260<dt><span><code>--with-target-system-zlib</code></span></dt> 2261<dd><p>Use installed ‘<samp>zlib</samp>’ rather than that included with GCC. This needs 2262to be available for each multilib variant, unless configured with 2263<samp>--with-target-system-zlib=‘<samp>auto</samp>’</samp> in which case the GCC included 2264‘<samp>zlib</samp>’ is only used when the system installed library is not available. 2265</p></dd> 2266</dl> 2267 2268<hr /> 2269<p> 2270<p><a href="./index.html">Return to the GCC Installation page</a> 2271</p> 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282</body> 2283</html> 2284