Home | History | Annotate | Line # | Download | only in INSTALL
specific.html revision 1.8
      1 <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
      2 <html>
      3 <!-- Copyright (C) 1988-2017 Free Software Foundation, Inc.
      4 
      5 Permission is granted to copy, distribute and/or modify this document
      6 under the terms of the GNU Free Documentation License, Version 1.3 or
      7 any later version published by the Free Software Foundation; with no
      8 Invariant Sections, the Front-Cover texts being (a) (see below), and
      9 with the Back-Cover Texts being (b) (see below).  A copy of the
     10 license is included in the section entitled "GNU
     11 Free Documentation License".
     12 
     13 (a) The FSF's Front-Cover Text is:
     14 
     15 A GNU Manual
     16 
     17 (b) The FSF's Back-Cover Text is:
     18 
     19 You have freedom to copy and modify this GNU Manual, like GNU
     20      software.  Copies published by the Free Software Foundation raise
     21      funds for GNU development. -->
     22 <!-- Created by GNU Texinfo 6.4, http://www.gnu.org/software/texinfo/ -->
     23 <head>
     24 <title>Installing GCC</title>
     25 
     26 <meta name="description" content="Installing GCC">
     27 <meta name="keywords" content="Installing GCC">
     28 <meta name="resource-type" content="document">
     29 <meta name="distribution" content="global">
     30 <meta name="Generator" content="makeinfo">
     31 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
     32 <style type="text/css">
     33 <!--
     34 a.summary-letter {text-decoration: none}
     35 blockquote.indentedblock {margin-right: 0em}
     36 blockquote.smallindentedblock {margin-right: 0em; font-size: smaller}
     37 blockquote.smallquotation {font-size: smaller}
     38 div.display {margin-left: 3.2em}
     39 div.example {margin-left: 3.2em}
     40 div.lisp {margin-left: 3.2em}
     41 div.smalldisplay {margin-left: 3.2em}
     42 div.smallexample {margin-left: 3.2em}
     43 div.smalllisp {margin-left: 3.2em}
     44 kbd {font-style: oblique}
     45 pre.display {font-family: inherit}
     46 pre.format {font-family: inherit}
     47 pre.menu-comment {font-family: serif}
     48 pre.menu-preformatted {font-family: serif}
     49 pre.smalldisplay {font-family: inherit; font-size: smaller}
     50 pre.smallexample {font-size: smaller}
     51 pre.smallformat {font-family: inherit; font-size: smaller}
     52 pre.smalllisp {font-size: smaller}
     53 span.nolinebreak {white-space: nowrap}
     54 span.roman {font-family: initial; font-weight: normal}
     55 span.sansserif {font-family: sans-serif; font-weight: normal}
     56 ul.no-bullet {list-style: none}
     57 -->
     58 </style>
     59 
     60 
     61 </head>
     62 
     63 <body lang="en">
     64 <h1 class="settitle" align="center">Installing GCC</h1>
     65 
     66 
     67 
     68 
     69 
     70 
     71 
     72 
     73 
     74 
     75 
     76 
     77 
     78 
     79 
     80 
     81 
     82 
     83 
     84 
     85 
     86 
     87 
     88 
     89 <a name="index-Specific"></a>
     90 <a name="index-Specific-installation-notes"></a>
     91 <a name="index-Target-specific-installation"></a>
     92 <a name="index-Host-specific-installation"></a>
     93 <a name="index-Target-specific-installation-notes"></a>
     94 
     95 <p>Please read this document carefully <em>before</em> installing the
     96 GNU Compiler Collection on your machine.
     97 </p>
     98 <p>Note that this list of install notes is <em>not</em> a list of supported
     99 hosts or targets.  Not all supported hosts and targets are listed
    100 here, only the ones that require host-specific or target-specific
    101 information have to. 
    102 </p>
    103 <ul>
    104 <li> <a href="#aarch64-x-x">aarch64*-*-*</a>
    105 </li><li> <a href="#alpha-x-x">alpha*-*-*</a>
    106 </li><li> <a href="#alpha-dec-osf51">alpha*-dec-osf5.1</a>
    107 </li><li> <a href="#amd64-x-solaris210">amd64-*-solaris2.10</a>
    108 </li><li> <a href="#arm-x-eabi">arm-*-eabi</a>
    109 </li><li> <a href="#avr">avr</a>
    110 </li><li> <a href="#bfin">Blackfin</a>
    111 </li><li> <a href="#dos">DOS</a>
    112 </li><li> <a href="#x-x-freebsd">*-*-freebsd*</a>
    113 </li><li> <a href="#h8300-hms">h8300-hms</a>
    114 </li><li> <a href="#hppa-hp-hpux">hppa*-hp-hpux*</a>
    115 </li><li> <a href="#hppa-hp-hpux10">hppa*-hp-hpux10</a>
    116 </li><li> <a href="#hppa-hp-hpux11">hppa*-hp-hpux11</a>
    117 </li><li> <a href="#x-x-linux-gnu">*-*-linux-gnu</a>
    118 </li><li> <a href="#ix86-x-linux">i?86-*-linux*</a>
    119 </li><li> <a href="#ix86-x-solaris210">i?86-*-solaris2.10</a>
    120 </li><li> <a href="#ia64-x-linux">ia64-*-linux</a>
    121 </li><li> <a href="#ia64-x-hpux">ia64-*-hpux*</a>
    122 </li><li> <a href="#x-ibm-aix">*-ibm-aix*</a>
    123 </li><li> <a href="#iq2000-x-elf">iq2000-*-elf</a>
    124 </li><li> <a href="#lm32-x-elf">lm32-*-elf</a>
    125 </li><li> <a href="#lm32-x-uclinux">lm32-*-uclinux</a>
    126 </li><li> <a href="#m32c-x-elf">m32c-*-elf</a>
    127 </li><li> <a href="#m32r-x-elf">m32r-*-elf</a>
    128 </li><li> <a href="#m68k-x-x">m68k-*-*</a>
    129 </li><li> <a href="#m68k-uclinux">m68k-uclinux</a>
    130 </li><li> <a href="#mep-x-elf">mep-*-elf</a>
    131 </li><li> <a href="#microblaze-x-elf">microblaze-*-elf</a>
    132 </li><li> <a href="#mips-x-x">mips-*-*</a>
    133 </li><li> <a href="#mips-sgi-irix5">mips-sgi-irix5</a>
    134 </li><li> <a href="#mips-sgi-irix6">mips-sgi-irix6</a>
    135 </li><li> <a href="#nds32le-x-elf">nds32le-*-elf</a>
    136 </li><li> <a href="#nds32be-x-elf">nds32be-*-elf</a>
    137 </li><li> <a href="#nvptx-x-none">nvptx-*-none</a>
    138 </li><li> <a href="#powerpc-x-x">powerpc*-*-*</a>
    139 </li><li> <a href="#powerpc-x-darwin">powerpc-*-darwin*</a>
    140 </li><li> <a href="#powerpc-x-elf">powerpc-*-elf</a>
    141 </li><li> <a href="#powerpc-x-linux-gnu">powerpc*-*-linux-gnu*</a>
    142 </li><li> <a href="#powerpc-x-netbsd">powerpc-*-netbsd*</a>
    143 </li><li> <a href="#powerpc-x-eabisim">powerpc-*-eabisim</a>
    144 </li><li> <a href="#powerpc-x-eabi">powerpc-*-eabi</a>
    145 </li><li> <a href="#powerpcle-x-elf">powerpcle-*-elf</a>
    146 </li><li> <a href="#powerpcle-x-eabisim">powerpcle-*-eabisim</a>
    147 </li><li> <a href="#powerpcle-x-eabi">powerpcle-*-eabi</a>
    148 </li><li> <a href="#s390-x-linux">s390-*-linux*</a>
    149 </li><li> <a href="#s390x-x-linux">s390x-*-linux*</a>
    150 </li><li> <a href="#s390x-ibm-tpf">s390x-ibm-tpf*</a>
    151 </li><li> <a href="#x-x-solaris2">*-*-solaris2*</a>
    152 </li><li> <a href="#sparc-x-x">sparc*-*-*</a>
    153 </li><li> <a href="#sparc-sun-solaris2">sparc-sun-solaris2*</a>
    154 </li><li> <a href="#sparc-sun-solaris210">sparc-sun-solaris2.10</a>
    155 </li><li> <a href="#sparc-x-linux">sparc-*-linux*</a>
    156 </li><li> <a href="#sparc64-x-solaris2">sparc64-*-solaris2*</a>
    157 </li><li> <a href="#sparcv9-x-solaris2">sparcv9-*-solaris2*</a>
    158 </li><li> <a href="#c6x-x-x">c6x-*-*</a>
    159 </li><li> <a href="#tilegx-x-linux">tilegx-*-linux*</a>
    160 </li><li> <a href="#tilegxbe-x-linux">tilegxbe-*-linux*</a>
    161 </li><li> <a href="#tilepro-x-linux">tilepro-*-linux*</a>
    162 </li><li> <a href="#visium-x-elf">visium-*-elf</a>
    163 </li><li> <a href="#x-x-vxworks">*-*-vxworks*</a>
    164 </li><li> <a href="#x86-64-x-x">x86_64-*-*, amd64-*-*</a>
    165 </li><li> <a href="#x86-64-x-solaris210">x86_64-*-solaris2.1[0-9]*</a>
    166 </li><li> <a href="#xtensa-x-elf">xtensa*-*-elf</a>
    167 </li><li> <a href="#xtensa-x-linux">xtensa*-*-linux*</a>
    168 </li><li> <a href="#windows">Microsoft Windows</a>
    169 </li><li> <a href="#x-x-cygwin">*-*-cygwin</a>
    170 </li><li> <a href="#x-x-interix">*-*-interix</a>
    171 </li><li> <a href="#x-x-mingw32">*-*-mingw32</a>
    172 </li><li> <a href="#os2">OS/2</a>
    173 </li><li> <a href="#older">Older systems</a>
    174 </li></ul>
    175 
    176 <ul>
    177 <li> <a href="#elf">all ELF targets</a> (SVR4, Solaris 2, etc.)
    178 </li></ul>
    179 
    180 
    181 <!-- -------- host/target specific issues start here ---------------- -->
    182 <hr />
    183 <a name="aarch64_002dx_002dx"></a><a name="aarch64_002a_002d_002a_002d_002a"></a>
    184 <h3 class="heading">aarch64*-*-*</h3>
    185 <p>Binutils pre 2.24 does not have support for selecting <samp>-mabi</samp> and
    186 does not support ILP32.  If it is used to build GCC 4.9 or later, GCC will
    187 not support option <samp>-mabi=ilp32</samp>.
    188 </p>
    189 <p>To enable a workaround for the Cortex-A53 erratum number 835769 by default
    190 (for all CPUs regardless of -mcpu option given) at configure time use the
    191 <samp>--enable-fix-cortex-a53-835769</samp> option.  This will enable the fix by
    192 default and can be explicitly disabled during compilation by passing the
    193 <samp>-mno-fix-cortex-a53-835769</samp> option.  Conversely,
    194 <samp>--disable-fix-cortex-a53-835769</samp> will disable the workaround by
    195 default.  The workaround is disabled by default if neither of
    196 <samp>--enable-fix-cortex-a53-835769</samp> or
    197 <samp>--disable-fix-cortex-a53-835769</samp> is given at configure time.
    198 </p>
    199 <p>To enable a workaround for the Cortex-A53 erratum number 843419 by default
    200 (for all CPUs regardless of -mcpu option given) at configure time use the
    201 <samp>--enable-fix-cortex-a53-843419</samp> option.  This workaround is applied at
    202 link time.  Enabling the workaround will cause GCC to pass the relevant option
    203 to the linker.  It can be explicitly disabled during compilation by passing the
    204 <samp>-mno-fix-cortex-a53-843419</samp> option.  Conversely,
    205 <samp>--disable-fix-cortex-a53-843419</samp> will disable the workaround by default.
    206 The workaround is disabled by default if neither of
    207 <samp>--enable-fix-cortex-a53-843419</samp> or
    208 <samp>--disable-fix-cortex-a53-843419</samp> is given at configure time.
    209 </p>
    210 <hr />
    211 <a name="alpha_002dx_002dx"></a><a name="alpha_002a_002d_002a_002d_002a"></a>
    212 <h3 class="heading">alpha*-*-*</h3>
    213 <p>This section contains general configuration information for all
    214 alpha-based platforms using ELF (in particular, ignore this section for
    215 DEC OSF/1, Digital UNIX and Tru64 UNIX).  In addition to reading this
    216 section, please read all other sections that match your target.
    217 </p>
    218 <p>We require binutils 2.11.2 or newer.
    219 Previous binutils releases had a number of problems with DWARF 2
    220 debugging information, not the least of which is incorrect linking of
    221 shared libraries.
    222 </p>
    223 <hr />
    224 <a name="alpha_002ddec_002dosf51"></a><a name="alpha_002a_002ddec_002dosf5_002e1"></a>
    225 <h3 class="heading">alpha*-dec-osf5.1</h3>
    226 <p>Systems using processors that implement the DEC Alpha architecture and
    227 are running the DEC/Compaq/HP Unix (DEC OSF/1, Digital UNIX, or Compaq/HP
    228 Tru64 UNIX) operating system, for example the DEC Alpha AXP systems.
    229 </p>
    230 <p>Support for Tru64 UNIX V5.1 has been removed in GCC 4.8.  As of GCC 4.6,
    231 support for Tru64 UNIX V4.0 and V5.0 has been removed.  As of GCC 3.2,
    232 versions before <code>alpha*-dec-osf4</code> are no longer supported.  (These
    233 are the versions which identify themselves as DEC OSF/1.)
    234 </p>
    235 <hr />
    236 <a name="amd64_002dx_002dsolaris210"></a><a name="amd64_002d_002a_002dsolaris2_002e1_005b0_002d9_005d_002a"></a>
    237 <h3 class="heading">amd64-*-solaris2.1[0-9]*</h3>
    238 <p>This is a synonym for &lsquo;<samp>x86_64-*-solaris2.1[0-9]*</samp>&rsquo;.
    239 </p>
    240 <hr />
    241 <a name="arc_002dx_002delf32"></a><a name="arc_002d_002a_002delf32"></a>
    242 <h3 class="heading">arc-*-elf32</h3>
    243 
    244 <p>Use &lsquo;<samp>configure --target=arc-elf32 --with-cpu=<var>cpu</var> --enable-languages=&quot;c,c++&quot;</samp>&rsquo;
    245 to configure GCC, with <var>cpu</var> being one of &lsquo;<samp>arc600</samp>&rsquo;, &lsquo;<samp>arc601</samp>&rsquo;,
    246 or &lsquo;<samp>arc700</samp>&rsquo;.
    247 </p>
    248 <hr />
    249 <a name="arc_002dlinux_002duclibc"></a><a name="arc_002dlinux_002duclibc-1"></a>
    250 <h3 class="heading">arc-linux-uclibc</h3>
    251 
    252 <p>Use &lsquo;<samp>configure --target=arc-linux-uclibc --with-cpu=arc700 --enable-languages=&quot;c,c++&quot;</samp>&rsquo; to configure GCC.
    253 </p>
    254 <hr />
    255 <a name="arm_002dx_002deabi"></a><a name="arm_002d_002a_002deabi"></a>
    256 <h3 class="heading">arm-*-eabi</h3>
    257 <p>ARM-family processors.  Subtargets that use the ELF object format
    258 require GNU binutils 2.13 or newer.  Such subtargets include:
    259 <code>arm-*-netbsdelf</code>, <code>arm-*-*linux-*</code>
    260 and <code>arm-*-rtemseabi</code>.
    261 </p>
    262 <p>Building the Ada frontend commonly fails (an infinite loop executing
    263 <code>xsinfo</code>) if the host compiler is GNAT 4.8.  Host compilers built from the
    264 GNAT 4.6, 4.9 or 5 release branches are known to succeed.
    265 </p>
    266 <hr />
    267 <a name="avr"></a><a name="avr-1"></a>
    268 <h3 class="heading">avr</h3>
    269 <p>ATMEL AVR-family micro controllers.  These are used in embedded
    270 applications.  There are no standard Unix configurations.
    271 See &ldquo;AVR Options&rdquo; in the main manual
    272 for the list of supported MCU types.
    273 </p>
    274 <p>Use &lsquo;<samp>configure --target=avr --enable-languages=&quot;c&quot;</samp>&rsquo; to configure GCC.
    275 </p>
    276 <p>Further installation notes and other useful information about AVR tools
    277 can also be obtained from:
    278 </p>
    279 <ul>
    280 <li> <a href="http://www.nongnu.org/avr/">http://www.nongnu.org/avr/</a>
    281 </li><li> <a href="http://www.amelek.gda.pl/avr/">http://www.amelek.gda.pl/avr/</a>
    282 </li></ul>
    283 
    284 <p>We <em>strongly</em> recommend using binutils 2.13 or newer.
    285 </p>
    286 <p>The following error:
    287 </p><div class="smallexample">
    288 <pre class="smallexample">Error: register required
    289 </pre></div>
    290 
    291 <p>indicates that you should upgrade to a newer version of the binutils.
    292 </p>
    293 <hr />
    294 <a name="bfin"></a><a name="Blackfin"></a>
    295 <h3 class="heading">Blackfin</h3>
    296 <p>The Blackfin processor, an Analog Devices DSP.
    297 See &ldquo;Blackfin Options&rdquo; in the main manual
    298 </p>
    299 <p>More information, and a version of binutils with support for this processor,
    300 is available at <a href="http://blackfin.uclinux.org">http://blackfin.uclinux.org</a>
    301 </p>
    302 <hr />
    303 <a name="cr16"></a><a name="CR16"></a>
    304 <h3 class="heading">CR16</h3>
    305 <p>The CR16 CompactRISC architecture is a 16-bit architecture. This
    306 architecture is used in embedded applications.
    307 </p>
    308 
    309 <p>See &ldquo;CR16 Options&rdquo; in the main manual for a list of CR16-specific options.
    310 </p>
    311 <p>Use &lsquo;<samp>configure --target=cr16-elf --enable-languages=c,c++</samp>&rsquo; to configure
    312 GCC&nbsp;for building a CR16 elf cross-compiler.
    313 </p>
    314 <p>Use &lsquo;<samp>configure --target=cr16-uclinux --enable-languages=c,c++</samp>&rsquo; to
    315 configure GCC&nbsp;for building a CR16 uclinux cross-compiler.
    316 </p>
    317 <hr />
    318 <a name="cris"></a><a name="CRIS"></a>
    319 <h3 class="heading">CRIS</h3>
    320 <p>CRIS is the CPU architecture in Axis Communications ETRAX system-on-a-chip
    321 series.  These are used in embedded applications.
    322 </p>
    323 <p>See &ldquo;CRIS Options&rdquo; in the main manual
    324 for a list of CRIS-specific options.
    325 </p>
    326 <p>There are a few different CRIS targets:
    327 </p><dl compact="compact">
    328 <dt><code>cris-axis-elf</code></dt>
    329 <dd><p>Mainly for monolithic embedded systems.  Includes a multilib for the
    330 &lsquo;<samp>v10</samp>&rsquo; core used in &lsquo;<samp>ETRAX 100 LX</samp>&rsquo;.
    331 </p></dd>
    332 <dt><code>cris-axis-linux-gnu</code></dt>
    333 <dd><p>A GNU/Linux port for the CRIS architecture, currently targeting
    334 &lsquo;<samp>ETRAX 100 LX</samp>&rsquo; by default.
    335 </p></dd>
    336 </dl>
    337 
    338 <p>For <code>cris-axis-elf</code> you need binutils 2.11
    339 or newer.  For <code>cris-axis-linux-gnu</code> you need binutils 2.12 or newer.
    340 </p>
    341 <p>Pre-packaged tools can be obtained from
    342 <a href="ftp://ftp.axis.com/pub/axis/tools/cris/compiler-kit/">ftp://ftp.axis.com/pub/axis/tools/cris/compiler-kit/</a>.  More
    343 information about this platform is available at
    344 <a href="http://developer.axis.com/">http://developer.axis.com/</a>.
    345 </p>
    346 <hr />
    347 <a name="dos"></a><a name="DOS"></a>
    348 <h3 class="heading">DOS</h3>
    349 <p>Please have a look at the <a href="binaries.html">binaries page</a>.
    350 </p>
    351 <p>You cannot install GCC by itself on MSDOS; it will not compile under
    352 any MSDOS compiler except itself.  You need to get the complete
    353 compilation package DJGPP, which includes binaries as well as sources,
    354 and includes all the necessary compilation tools and libraries.
    355 </p>
    356 <hr />
    357 <a name="epiphany_002dx_002delf"></a><a name="epiphany_002d_002a_002delf"></a>
    358 <h3 class="heading">epiphany-*-elf</h3>
    359 <p>Adapteva Epiphany.
    360 This configuration is intended for embedded systems.
    361 </p>
    362 <hr />
    363 <a name="x_002dx_002dfreebsd"></a><a name="g_t_002a_002d_002a_002dfreebsd_002a"></a>
    364 <h3 class="heading">*-*-freebsd*</h3>
    365 <p>Support for FreeBSD 1 was discontinued in GCC 3.2.  Support for
    366 FreeBSD 2 (and any mutant a.out variants of FreeBSD 3) was
    367 discontinued in GCC 4.0.
    368 </p>
    369 <p>In order to better utilize FreeBSD base system functionality and match
    370 the configuration of the system compiler, GCC 4.5 and above as well as
    371 GCC 4.4 past 2010-06-20 leverage SSP support in libc (which is present
    372 on FreeBSD 7 or later) and the use of <code>__cxa_atexit</code> by default
    373 (on FreeBSD 6 or later).  The use of <code>dl_iterate_phdr</code> inside
    374 <samp>libgcc_s.so.1</samp> and boehm-gc (on FreeBSD 7 or later) is enabled
    375 by GCC 4.5 and above.
    376 </p>
    377 <p>We support FreeBSD using the ELF file format with DWARF 2 debugging
    378 for all CPU architectures.  You may use <samp>-gstabs</samp> instead of
    379 <samp>-g</samp>, if you really want the old debugging format.  There are
    380 no known issues with mixing object files and libraries with different
    381 debugging formats.  Otherwise, this release of GCC should now match
    382 more of the configuration used in the stock FreeBSD configuration of
    383 GCC.  In particular, <samp>--enable-threads</samp> is now configured by
    384 default.  However, as a general user, do not attempt to replace the
    385 system compiler with this release.  Known to bootstrap and check with
    386 good results on FreeBSD 7.2-STABLE.  In the past, known to bootstrap
    387 and check with good results on FreeBSD 3.0, 3.4, 4.0, 4.2, 4.3, 4.4,
    388 4.5, 4.8, 4.9 and 5-CURRENT.
    389 </p>
    390 <p>The version of binutils installed in <samp>/usr/bin</samp> probably works
    391 with this release of GCC.  Bootstrapping against the latest GNU
    392 binutils and/or the version found in <samp>/usr/ports/devel/binutils</samp> has
    393 been known to enable additional features and improve overall testsuite
    394 results.  However, it is currently known that boehm-gc (which itself
    395 is required for java) may not configure properly on FreeBSD prior to
    396 the FreeBSD 7.0 release with GNU binutils after 2.16.1.
    397 </p>
    398 <hr />
    399 <a name="ft32_002dx_002delf"></a><a name="ft32_002d_002a_002delf"></a>
    400 <h3 class="heading">ft32-*-elf</h3>
    401 <p>The FT32 processor.
    402 This configuration is intended for embedded systems.
    403 </p>
    404 <hr />
    405 <a name="h8300_002dhms"></a><a name="h8300_002dhms-1"></a>
    406 <h3 class="heading">h8300-hms</h3>
    407 <p>Renesas H8/300 series of processors.
    408 </p>
    409 <p>Please have a look at the <a href="binaries.html">binaries page</a>.
    410 </p>
    411 <p>The calling convention and structure layout has changed in release 2.6.
    412 All code must be recompiled.  The calling convention now passes the
    413 first three arguments in function calls in registers.  Structures are no
    414 longer a multiple of 2 bytes.
    415 </p>
    416 <hr />
    417 <a name="hppa_002dhp_002dhpux"></a><a name="hppa_002a_002dhp_002dhpux_002a"></a>
    418 <h3 class="heading">hppa*-hp-hpux*</h3>
    419 <p>Support for HP-UX version 9 and older was discontinued in GCC 3.4.
    420 </p>
    421 <p>We require using gas/binutils on all hppa platforms.  Version 2.19 or
    422 later is recommended.
    423 </p>
    424 <p>It may be helpful to configure GCC with the
    425 <a href="./configure.html#with-gnu-as"><samp>--with-gnu-as</samp></a> and
    426 <samp>--with-as=&hellip;</samp> options to ensure that GCC can find GAS.
    427 </p>
    428 <p>The HP assembler should not be used with GCC.  It is rarely tested and may
    429 not work.  It shouldn&rsquo;t be used with any languages other than C due to its
    430 many limitations.
    431 </p>
    432 <p>Specifically, <samp>-g</samp> does not work (HP-UX uses a peculiar debugging
    433 format which GCC does not know about).  It also inserts timestamps
    434 into each object file it creates, causing the 3-stage comparison test to
    435 fail during a bootstrap.  You should be able to continue by saying
    436 &lsquo;<samp>make all-host all-target</samp>&rsquo; after getting the failure from &lsquo;<samp>make</samp>&rsquo;.
    437 </p>
    438 <p>Various GCC features are not supported.  For example, it does not support weak
    439 symbols or alias definitions.  As a result, explicit template instantiations
    440 are required when using C++.  This makes it difficult if not impossible to
    441 build many C++ applications.
    442 </p>
    443 <p>There are two default scheduling models for instructions.  These are
    444 PROCESSOR_7100LC and PROCESSOR_8000.  They are selected from the pa-risc
    445 architecture specified for the target machine when configuring.
    446 PROCESSOR_8000 is the default.  PROCESSOR_7100LC is selected when
    447 the target is a &lsquo;<samp>hppa1*</samp>&rsquo; machine.
    448 </p>
    449 <p>The PROCESSOR_8000 model is not well suited to older processors.  Thus,
    450 it is important to completely specify the machine architecture when
    451 configuring if you want a model other than PROCESSOR_8000.  The macro
    452 TARGET_SCHED_DEFAULT can be defined in BOOT_CFLAGS if a different
    453 default scheduling model is desired.
    454 </p>
    455 <p>As of GCC 4.0, GCC uses the UNIX 95 namespace for HP-UX 10.10
    456 through 11.00, and the UNIX 98 namespace for HP-UX 11.11 and later.
    457 This namespace change might cause problems when bootstrapping with
    458 an earlier version of GCC or the HP compiler as essentially the same
    459 namespace is required for an entire build.  This problem can be avoided
    460 in a number of ways.  With HP cc, <code>UNIX_STD</code> can be set to &lsquo;<samp>95</samp>&rsquo;
    461 or &lsquo;<samp>98</samp>&rsquo;.  Another way is to add an appropriate set of predefines
    462 to <code>CC</code>.  The description for the <samp>munix=</samp> option contains
    463 a list of the predefines used with each standard.
    464 </p>
    465 <p>More specific information to &lsquo;<samp>hppa*-hp-hpux*</samp>&rsquo; targets follows.
    466 </p>
    467 <hr />
    468 <a name="hppa_002dhp_002dhpux10"></a><a name="hppa_002a_002dhp_002dhpux10"></a>
    469 <h3 class="heading">hppa*-hp-hpux10</h3>
    470 <p>For hpux10.20, we <em>highly</em> recommend you pick up the latest sed patch
    471 <code>PHCO_19798</code> from HP.
    472 </p>
    473 <p>The C++ ABI has changed incompatibly in GCC 4.0.  COMDAT subspaces are
    474 used for one-only code and data.  This resolves many of the previous
    475 problems in using C++ on this target.  However, the ABI is not compatible
    476 with the one implemented under HP-UX 11 using secondary definitions.
    477 </p>
    478 <hr />
    479 <a name="hppa_002dhp_002dhpux11"></a><a name="hppa_002a_002dhp_002dhpux11"></a>
    480 <h3 class="heading">hppa*-hp-hpux11</h3>
    481 <p>GCC 3.0 and up support HP-UX 11.  GCC 2.95.x is not supported and cannot
    482 be used to compile GCC 3.0 and up.
    483 </p>
    484 <p>The libffi and libjava libraries haven&rsquo;t been ported to 64-bit HP-UX&nbsp;and don&rsquo;t build.
    485 </p>
    486 <p>Refer to <a href="binaries.html">binaries</a> for information about obtaining
    487 precompiled GCC binaries for HP-UX.  Precompiled binaries must be obtained
    488 to build the Ada language as it can&rsquo;t be bootstrapped using C.  Ada is
    489 only available for the 32-bit PA-RISC runtime.
    490 </p>
    491 <p>Starting with GCC 3.4 an ISO C compiler is required to bootstrap.  The
    492 bundled compiler supports only traditional C; you will need either HP&rsquo;s
    493 unbundled compiler, or a binary distribution of GCC.
    494 </p>
    495 <p>It is possible to build GCC 3.3 starting with the bundled HP compiler,
    496 but the process requires several steps.  GCC 3.3 can then be used to
    497 build later versions.  The fastjar program contains ISO C code and
    498 can&rsquo;t be built with the HP bundled compiler.  This problem can be
    499 avoided by not building the Java language.  For example, use the
    500 <samp>--enable-languages=&quot;c,c++,f77,objc&quot;</samp> option in your configure
    501 command.
    502 </p>
    503 <p>There are several possible approaches to building the distribution.
    504 Binutils can be built first using the HP tools.  Then, the GCC
    505 distribution can be built.  The second approach is to build GCC
    506 first using the HP tools, then build binutils, then rebuild GCC.
    507 There have been problems with various binary distributions, so it
    508 is best not to start from a binary distribution.
    509 </p>
    510 <p>On 64-bit capable systems, there are two distinct targets.  Different
    511 installation prefixes must be used if both are to be installed on
    512 the same system.  The &lsquo;<samp>hppa[1-2]*-hp-hpux11*</samp>&rsquo; target generates code
    513 for the 32-bit PA-RISC runtime architecture and uses the HP linker.
    514 The &lsquo;<samp>hppa64-hp-hpux11*</samp>&rsquo; target generates 64-bit code for the
    515 PA-RISC 2.0 architecture.
    516 </p>
    517 <p>The script config.guess now selects the target type based on the compiler
    518 detected during configuration.  You must define <code>PATH</code> or <code>CC</code> so
    519 that configure finds an appropriate compiler for the initial bootstrap.
    520 When <code>CC</code> is used, the definition should contain the options that are
    521 needed whenever <code>CC</code> is used.
    522 </p>
    523 <p>Specifically, options that determine the runtime architecture must be
    524 in <code>CC</code> to correctly select the target for the build.  It is also
    525 convenient to place many other compiler options in <code>CC</code>.  For example,
    526 <code>CC=&quot;cc -Ac +DA2.0W -Wp,-H16376 -D_CLASSIC_TYPES -D_HPUX_SOURCE&quot;</code>
    527 can be used to bootstrap the GCC 3.3 branch with the HP compiler in
    528 64-bit K&amp;R/bundled mode.  The <samp>+DA2.0W</samp> option will result in
    529 the automatic selection of the &lsquo;<samp>hppa64-hp-hpux11*</samp>&rsquo; target.  The
    530 macro definition table of cpp needs to be increased for a successful
    531 build with the HP compiler.  _CLASSIC_TYPES and _HPUX_SOURCE need to
    532 be defined when building with the bundled compiler, or when using the
    533 <samp>-Ac</samp> option.  These defines aren&rsquo;t necessary with <samp>-Ae</samp>.
    534 </p>
    535 <p>It is best to explicitly configure the &lsquo;<samp>hppa64-hp-hpux11*</samp>&rsquo; target
    536 with the <samp>--with-ld=&hellip;</samp> option.  This overrides the standard
    537 search for ld.  The two linkers supported on this target require different
    538 commands.  The default linker is determined during configuration.  As a
    539 result, it&rsquo;s not possible to switch linkers in the middle of a GCC build.
    540 This has been reported to sometimes occur in unified builds of binutils
    541 and GCC.
    542 </p>
    543 <p>A recent linker patch must be installed for the correct operation of
    544 GCC 3.3 and later.  <code>PHSS_26559</code> and <code>PHSS_24304</code> are the
    545 oldest linker patches that are known to work.  They are for HP-UX
    546 11.00 and 11.11, respectively.  <code>PHSS_24303</code>, the companion to
    547 <code>PHSS_24304</code>, might be usable but it hasn&rsquo;t been tested.  These
    548 patches have been superseded.  Consult the HP patch database to obtain
    549 the currently recommended linker patch for your system.
    550 </p>
    551 <p>The patches are necessary for the support of weak symbols on the
    552 32-bit port, and for the running of initializers and finalizers.  Weak
    553 symbols are implemented using SOM secondary definition symbols.  Prior
    554 to HP-UX 11, there are bugs in the linker support for secondary symbols.
    555 The patches correct a problem of linker core dumps creating shared
    556 libraries containing secondary symbols, as well as various other
    557 linking issues involving secondary symbols.
    558 </p>
    559 <p>GCC 3.3 uses the ELF DT_INIT_ARRAY and DT_FINI_ARRAY capabilities to
    560 run initializers and finalizers on the 64-bit port.  The 32-bit port
    561 uses the linker <samp>+init</samp> and <samp>+fini</samp> options for the same
    562 purpose.  The patches correct various problems with the +init/+fini
    563 options, including program core dumps.  Binutils 2.14 corrects a
    564 problem on the 64-bit port resulting from HP&rsquo;s non-standard use of
    565 the .init and .fini sections for array initializers and finalizers.
    566 </p>
    567 <p>Although the HP and GNU linkers are both supported for the
    568 &lsquo;<samp>hppa64-hp-hpux11*</samp>&rsquo; target, it is strongly recommended that the
    569 HP linker be used for link editing on this target.
    570 </p>
    571 <p>At this time, the GNU linker does not support the creation of long
    572 branch stubs.  As a result, it can&rsquo;t successfully link binaries
    573 containing branch offsets larger than 8 megabytes.  In addition,
    574 there are problems linking shared libraries, linking executables
    575 with <samp>-static</samp>, and with dwarf2 unwind and exception support.
    576 It also doesn&rsquo;t provide stubs for internal calls to global functions
    577 in shared libraries, so these calls can&rsquo;t be overloaded.
    578 </p>
    579 <p>The HP dynamic loader does not support GNU symbol versioning, so symbol
    580 versioning is not supported.  It may be necessary to disable symbol
    581 versioning with <samp>--disable-symvers</samp> when using GNU ld.
    582 </p>
    583 <p>POSIX threads are the default.  The optional DCE thread library is not
    584 supported, so <samp>--enable-threads=dce</samp> does not work.
    585 </p>
    586 <hr />
    587 <a name="x_002dx_002dlinux_002dgnu"></a><a name="g_t_002a_002d_002a_002dlinux_002dgnu"></a>
    588 <h3 class="heading">*-*-linux-gnu</h3>
    589 <p>Versions of libstdc++-v3 starting with 3.2.1 require bug fixes present
    590 in glibc 2.2.5 and later.  More information is available in the
    591 libstdc++-v3 documentation.
    592 </p>
    593 <hr />
    594 <a name="ix86_002dx_002dlinux"></a><a name="i_003f86_002d_002a_002dlinux_002a"></a>
    595 <h3 class="heading">i?86-*-linux*</h3>
    596 <p>As of GCC 3.3, binutils 2.13.1 or later is required for this platform.
    597 See <a href="http://gcc.gnu.org/PR10877">bug 10877</a> for more information.
    598 </p>
    599 <p>If you receive Signal 11 errors when building on GNU/Linux, then it is
    600 possible you have a hardware problem.  Further information on this can be
    601 found on <a href="http://www.bitwizard.nl/sig11/">www.bitwizard.nl</a>.
    602 </p>
    603 <hr />
    604 <a name="ix86_002dx_002dsolaris210"></a><a name="i_003f86_002d_002a_002dsolaris2_002e10"></a>
    605 <h3 class="heading">i?86-*-solaris2.10</h3>
    606 <p>Use this for Solaris 10 or later on x86 and x86-64 systems.  Starting
    607 with GCC 4.7, there is also a 64-bit &lsquo;<samp>amd64-*-solaris2.1[0-9]*</samp>&rsquo; or
    608 &lsquo;<samp>x86_64-*-solaris2.1[0-9]*</samp>&rsquo; configuration that corresponds to
    609 &lsquo;<samp>sparcv9-sun-solaris2*</samp>&rsquo;.
    610 </p>
    611 <p>It is recommended that you configure GCC to use the GNU assembler.  The
    612 versions included in Solaris 10, from GNU binutils 2.15 (in
    613 <samp>/usr/sfw/bin/gas</samp>), and Solaris 11, from GNU binutils 2.19 or
    614 newer (also available as <samp>/usr/bin/gas</samp> and
    615 <samp>/usr/gnu/bin/as</samp>), work fine.  Please note that the current
    616 version, from GNU binutils 2.26, only works on Solaris 12 when using the
    617 Solaris linker.  On Solaris 10 and 11, you either have to wait for GNU
    618 binutils 2.26.1 or newer, or stay with GNU binutils 2.25.1.  Recent
    619 versions of the Solaris assembler in <samp>/usr/ccs/bin/as</samp> work almost
    620 as well, though.
    621 </p>
    622 <p>For linking, the Solaris linker, is preferred.  If you want to use the GNU
    623 linker instead, note that due to a packaging bug the version in Solaris
    624 10, from GNU binutils 2.15 (in <samp>/usr/sfw/bin/gld</samp>), cannot be used,
    625 while the version in Solaris 11, from GNU binutils 2.19 or newer (also
    626 in <samp>/usr/gnu/bin/ld</samp> and <samp>/usr/bin/gld</samp>), works, as does the
    627 latest version, from GNU binutils 2.26.
    628 </p>
    629 <p>To use GNU <code>as</code>, configure with the options
    630 <samp>--with-gnu-as --with-as=/usr/sfw/bin/gas</samp>.  It may be necessary
    631 to configure with <samp>--without-gnu-ld --with-ld=/usr/ccs/bin/ld</samp> to
    632 guarantee use of Sun <code>ld</code>.
    633 </p>
    634 <hr />
    635 <a name="ia64_002dx_002dlinux"></a><a name="ia64_002d_002a_002dlinux"></a>
    636 <h3 class="heading">ia64-*-linux</h3>
    637 <p>IA-64 processor (also known as IPF, or Itanium Processor Family)
    638 running GNU/Linux.
    639 </p>
    640 <p>If you are using the installed system libunwind library with
    641 <samp>--with-system-libunwind</samp>, then you must use libunwind 0.98 or
    642 later.
    643 </p>
    644 <p>None of the following versions of GCC has an ABI that is compatible
    645 with any of the other versions in this list, with the exception that
    646 Red Hat 2.96 and Trillian 000171 are compatible with each other:
    647 3.1, 3.0.2, 3.0.1, 3.0, Red Hat 2.96, and Trillian 000717.
    648 This primarily affects C++ programs and programs that create shared libraries.
    649 GCC 3.1 or later is recommended for compiling linux, the kernel.
    650 As of version 3.1 GCC is believed to be fully ABI compliant, and hence no
    651 more major ABI changes are expected.
    652 </p>
    653 <hr />
    654 <a name="ia64_002dx_002dhpux"></a><a name="ia64_002d_002a_002dhpux_002a"></a>
    655 <h3 class="heading">ia64-*-hpux*</h3>
    656 <p>Building GCC on this target requires the GNU Assembler.  The bundled HP
    657 assembler will not work.  To prevent GCC from using the wrong assembler,
    658 the option <samp>--with-gnu-as</samp> may be necessary.
    659 </p>
    660 <p>The GCC libunwind library has not been ported to HPUX.  This means that for
    661 GCC versions 3.2.3 and earlier, <samp>--enable-libunwind-exceptions</samp>
    662 is required to build GCC.  For GCC 3.3 and later, this is the default.
    663 For gcc 3.4.3 and later, <samp>--enable-libunwind-exceptions</samp> is
    664 removed and the system libunwind library will always be used.
    665 </p>
    666 <hr />
    667 <!-- rs6000-ibm-aix*, powerpc-ibm-aix* -->
    668 <a name="x_002dibm_002daix"></a><a name="g_t_002a_002dibm_002daix_002a"></a>
    669 <h3 class="heading">*-ibm-aix*</h3>
    670 <p>Support for AIX version 3 and older was discontinued in GCC 3.4.
    671 Support for AIX version 4.2 and older was discontinued in GCC 4.5.
    672 </p>
    673 <p>&ldquo;out of memory&rdquo; bootstrap failures may indicate a problem with
    674 process resource limits (ulimit).  Hard limits are configured in the
    675 <samp>/etc/security/limits</samp> system configuration file.
    676 </p>
    677 <p>GCC 4.9 and above require a C++ compiler for bootstrap.  IBM VAC++ / xlC
    678 cannot bootstrap GCC.  xlc can bootstrap an older version of GCC and
    679 G++ can bootstrap recent releases of GCC.
    680 </p>
    681 <p>GCC can bootstrap with recent versions of IBM XLC, but bootstrapping
    682 with an earlier release of GCC is recommended.  Bootstrapping with XLC
    683 requires a larger data segment, which can be enabled through the
    684 <var>LDR_CNTRL</var> environment variable, e.g.,
    685 </p>
    686 <div class="smallexample">
    687 <pre class="smallexample">% LDR_CNTRL=MAXDATA=0x50000000
    688 % export LDR_CNTRL
    689 </pre></div>
    690 
    691 <p>One can start with a pre-compiled version of GCC to build from
    692 sources.  One may delete GCC&rsquo;s &ldquo;fixed&rdquo; header files when starting
    693 with a version of GCC built for an earlier release of AIX.
    694 </p>
    695 <p>To speed up the configuration phases of bootstrapping and installing GCC,
    696 one may use GNU Bash instead of AIX <code>/bin/sh</code>, e.g.,
    697 </p>
    698 <div class="smallexample">
    699 <pre class="smallexample">% CONFIG_SHELL=/opt/freeware/bin/bash
    700 % export CONFIG_SHELL
    701 </pre></div>
    702 
    703 <p>and then proceed as described in <a href="build.html">the build
    704 instructions</a>, where we strongly recommend specifying an absolute path
    705 to invoke <var>srcdir</var>/configure.
    706 </p>
    707 <p>Because GCC on AIX is built as a 32-bit executable by default,
    708 (although it can generate 64-bit programs) the GMP and MPFR libraries
    709 required by gfortran must be 32-bit libraries.  Building GMP and MPFR
    710 as static archive libraries works better than shared libraries.
    711 </p>
    712 <p>Errors involving <code>alloca</code> when building GCC generally are due
    713 to an incorrect definition of <code>CC</code> in the Makefile or mixing files
    714 compiled with the native C compiler and GCC.  During the stage1 phase of
    715 the build, the native AIX compiler <strong>must</strong> be invoked as <code>cc</code>
    716 (not <code>xlc</code>).  Once <code>configure</code> has been informed of
    717 <code>xlc</code>, one needs to use &lsquo;<samp>make distclean</samp>&rsquo; to remove the
    718 configure cache files and ensure that <code>CC</code> environment variable
    719 does not provide a definition that will confuse <code>configure</code>.
    720 If this error occurs during stage2 or later, then the problem most likely
    721 is the version of Make (see above).
    722 </p>
    723 <p>The native <code>as</code> and <code>ld</code> are recommended for
    724 bootstrapping on AIX.  The GNU Assembler, GNU Linker, and GNU
    725 Binutils version 2.20 is the minimum level that supports bootstrap on
    726 AIX 5.  The GNU Assembler has not been updated to support AIX 6&nbsp;or
    727 AIX 7.  The native AIX tools do interoperate with GCC.
    728 </p>
    729 <p>AIX 7.1 added partial support for DWARF debugging, but full support
    730 requires AIX 7.1 TL03 SP7 that supports additional DWARF sections and
    731 fixes a bug in the assembler.  AIX 7.1 TL03 SP5 distributed a version
    732 of libm.a missing important symbols; a fix for IV77796 will be
    733 included in SP6.
    734 </p>
    735 <p>AIX 5.3 TL10, AIX 6.1 TL05 and AIX 7.1 TL00 introduced an AIX
    736 assembler change that sometimes produces corrupt assembly files
    737 causing AIX linker errors.  The bug breaks GCC bootstrap on AIX and
    738 can cause compilation failures with existing GCC installations.  An
    739 AIX iFix for AIX 5.3 is available (APAR IZ98385 for AIX 5.3 TL10, APAR
    740 IZ98477 for AIX 5.3 TL11 and IZ98134 for AIX 5.3 TL12). AIX 5.3 TL11 SP8,
    741 AIX 5.3 TL12 SP5, AIX 6.1 TL04 SP11, AIX 6.1 TL05 SP7, AIX 6.1 TL06 SP6,
    742 AIX 6.1 TL07 and AIX 7.1 TL01 should include the fix.
    743 </p>
    744 <p>Building <samp>libstdc++.a</samp> requires a fix for an AIX Assembler bug
    745 APAR IY26685 (AIX 4.3) or APAR IY25528 (AIX 5.1).  It also requires a
    746 fix for another AIX Assembler bug and a co-dependent AIX Archiver fix
    747 referenced as APAR IY53606 (AIX 5.2) or as APAR IY54774 (AIX 5.1)
    748 </p>
    749 <a name="TransferAixShobj"></a><p>&lsquo;<samp>libstdc++</samp>&rsquo; in GCC 3.4 increments the major version number of the
    750 shared object and GCC installation places the <samp>libstdc++.a</samp>
    751 shared library in a common location which will overwrite the and GCC
    752 3.3 version of the shared library.  Applications either need to be
    753 re-linked against the new shared library or the GCC 3.1 and GCC 3.3
    754 versions of the &lsquo;<samp>libstdc++</samp>&rsquo; shared object needs to be available
    755 to the AIX runtime loader.  The GCC 3.1 &lsquo;<samp>libstdc++.so.4</samp>&rsquo;, if
    756 present, and GCC 3.3 &lsquo;<samp>libstdc++.so.5</samp>&rsquo; shared objects can be
    757 installed for runtime dynamic loading using the following steps to set
    758 the &lsquo;<samp>F_LOADONLY</samp>&rsquo; flag in the shared object for <em>each</em>
    759 multilib <samp>libstdc++.a</samp> installed:
    760 </p>
    761 <p>Extract the shared objects from the currently installed
    762 <samp>libstdc++.a</samp> archive:
    763 </p><div class="smallexample">
    764 <pre class="smallexample">% ar -x libstdc++.a libstdc++.so.4 libstdc++.so.5
    765 </pre></div>
    766 
    767 <p>Enable the &lsquo;<samp>F_LOADONLY</samp>&rsquo; flag so that the shared object will be
    768 available for runtime dynamic loading, but not linking:
    769 </p><div class="smallexample">
    770 <pre class="smallexample">% strip -e libstdc++.so.4 libstdc++.so.5
    771 </pre></div>
    772 
    773 <p>Archive the runtime-only shared object in the GCC 3.4
    774 <samp>libstdc++.a</samp> archive:
    775 </p><div class="smallexample">
    776 <pre class="smallexample">% ar -q libstdc++.a libstdc++.so.4 libstdc++.so.5
    777 </pre></div>
    778 
    779 <p>Eventually, the
    780 <a href="./configure.html#WithAixSoname"><samp>--with-aix-soname=svr4</samp></a>
    781 configure option may drop the need for this procedure for libraries that
    782 support it.
    783 </p>
    784 <p>Linking executables and shared libraries may produce warnings of
    785 duplicate symbols.  The assembly files generated by GCC for AIX always
    786 have included multiple symbol definitions for certain global variable
    787 and function declarations in the original program.  The warnings should
    788 not prevent the linker from producing a correct library or runnable
    789 executable.
    790 </p>
    791 <p>AIX 4.3 utilizes a &ldquo;large format&rdquo; archive to support both 32-bit and
    792 64-bit object modules.  The routines provided in AIX 4.3.0 and AIX 4.3.1
    793 to parse archive libraries did not handle the new format correctly.
    794 These routines are used by GCC and result in error messages during
    795 linking such as &ldquo;not a COFF file&rdquo;.  The version of the routines shipped
    796 with AIX 4.3.1 should work for a 32-bit environment.  The <samp>-g</samp>
    797 option of the archive command may be used to create archives of 32-bit
    798 objects using the original &ldquo;small format&rdquo;.  A correct version of the
    799 routines is shipped with AIX 4.3.2 and above.
    800 </p>
    801 <p>Some versions of the AIX binder (linker) can fail with a relocation
    802 overflow severe error when the <samp>-bbigtoc</samp> option is used to link
    803 GCC-produced object files into an executable that overflows the TOC.  A fix
    804 for APAR IX75823 (OVERFLOW DURING LINK WHEN USING GCC AND -BBIGTOC) is
    805 available from IBM Customer Support and from its
    806 <a href="http://techsupport.services.ibm.com/">techsupport.services.ibm.com</a>
    807 website as PTF U455193.
    808 </p>
    809 <p>The AIX 4.3.2.1 linker (bos.rte.bind_cmds Level 4.3.2.1) will dump core
    810 with a segmentation fault when invoked by any version of GCC.  A fix for
    811 APAR IX87327 is available from IBM Customer Support and from its
    812 <a href="http://techsupport.services.ibm.com/">techsupport.services.ibm.com</a>
    813 website as PTF U461879.  This fix is incorporated in AIX 4.3.3 and above.
    814 </p>
    815 <p>The initial assembler shipped with AIX 4.3.0 generates incorrect object
    816 files.  A fix for APAR IX74254 (64BIT DISASSEMBLED OUTPUT FROM COMPILER FAILS
    817 TO ASSEMBLE/BIND) is available from IBM Customer Support and from its
    818 <a href="http://techsupport.services.ibm.com/">techsupport.services.ibm.com</a>
    819 website as PTF U453956.  This fix is incorporated in AIX 4.3.1 and above.
    820 </p>
    821 <p>AIX provides National Language Support (NLS).  Compilers and assemblers
    822 use NLS to support locale-specific representations of various data
    823 formats including floating-point numbers (e.g., &lsquo;<samp>.</samp>&rsquo;  vs &lsquo;<samp>,</samp>&rsquo; for
    824 separating decimal fractions).  There have been problems reported where
    825 GCC does not produce the same floating-point formats that the assembler
    826 expects.  If one encounters this problem, set the <code>LANG</code>
    827 environment variable to &lsquo;<samp>C</samp>&rsquo; or &lsquo;<samp>En_US</samp>&rsquo;.
    828 </p>
    829 <p>A default can be specified with the <samp>-mcpu=<var>cpu_type</var></samp>
    830 switch and using the configure option <samp>--with-cpu-<var>cpu_type</var></samp>.
    831 </p>
    832 <hr />
    833 <a name="iq2000_002dx_002delf"></a><a name="iq2000_002d_002a_002delf"></a>
    834 <h3 class="heading">iq2000-*-elf</h3>
    835 <p>Vitesse IQ2000 processors.  These are used in embedded
    836 applications.  There are no standard Unix configurations.
    837 </p>
    838 <hr />
    839 <a name="lm32_002dx_002delf"></a><a name="lm32_002d_002a_002delf"></a>
    840 <h3 class="heading">lm32-*-elf</h3>
    841 <p>Lattice Mico32 processor.
    842 This configuration is intended for embedded systems.
    843 </p>
    844 <hr />
    845 <a name="lm32_002dx_002duclinux"></a><a name="lm32_002d_002a_002duclinux"></a>
    846 <h3 class="heading">lm32-*-uclinux</h3>
    847 <p>Lattice Mico32 processor.
    848 This configuration is intended for embedded systems running uClinux.
    849 </p>
    850 <hr />
    851 <a name="m32c_002dx_002delf"></a><a name="m32c_002d_002a_002delf"></a>
    852 <h3 class="heading">m32c-*-elf</h3>
    853 <p>Renesas M32C processor.
    854 This configuration is intended for embedded systems.
    855 </p>
    856 <hr />
    857 <a name="m32r_002dx_002delf"></a><a name="m32r_002d_002a_002delf"></a>
    858 <h3 class="heading">m32r-*-elf</h3>
    859 <p>Renesas M32R processor.
    860 This configuration is intended for embedded systems.
    861 </p>
    862 <hr />
    863 <a name="m68k_002dx_002dx"></a><a name="m68k_002d_002a_002d_002a"></a>
    864 <h3 class="heading">m68k-*-*</h3>
    865 <p>By default,
    866 &lsquo;<samp>m68k-*-elf*</samp>&rsquo;, &lsquo;<samp>m68k-*-rtems</samp>&rsquo;,  &lsquo;<samp>m68k-*-uclinux</samp>&rsquo; and
    867 &lsquo;<samp>m68k-*-linux</samp>&rsquo;
    868 build libraries for both M680x0 and ColdFire processors.  If you only
    869 need the M680x0 libraries, you can omit the ColdFire ones by passing
    870 <samp>--with-arch=m68k</samp> to <code>configure</code>.  Alternatively, you
    871 can omit the M680x0 libraries by passing <samp>--with-arch=cf</samp> to
    872 <code>configure</code>.  These targets default to 5206 or 5475 code as
    873 appropriate for the target system when
    874 configured with <samp>--with-arch=cf</samp> and 68020 code otherwise.
    875 </p>
    876 <p>The &lsquo;<samp>m68k-*-netbsd</samp>&rsquo; and
    877 &lsquo;<samp>m68k-*-openbsd</samp>&rsquo; targets also support the <samp>--with-arch</samp>
    878 option.  They will generate ColdFire CFV4e code when configured with
    879 <samp>--with-arch=cf</samp> and 68020 code otherwise.
    880 </p>
    881 <p>You can override the default processors listed above by configuring
    882 with <samp>--with-cpu=<var>target</var></samp>.  This <var>target</var> can either
    883 be a <samp>-mcpu</samp> argument or one of the following values:
    884 &lsquo;<samp>m68000</samp>&rsquo;, &lsquo;<samp>m68010</samp>&rsquo;, &lsquo;<samp>m68020</samp>&rsquo;, &lsquo;<samp>m68030</samp>&rsquo;,
    885 &lsquo;<samp>m68040</samp>&rsquo;, &lsquo;<samp>m68060</samp>&rsquo;, &lsquo;<samp>m68020-40</samp>&rsquo; and &lsquo;<samp>m68020-60</samp>&rsquo;.
    886 </p>
    887 <p>GCC requires at least binutils version 2.17 on these targets.
    888 </p>
    889 <hr />
    890 <a name="m68k_002dx_002duclinux"></a><a name="m68k_002d_002a_002duclinux"></a>
    891 <h3 class="heading">m68k-*-uclinux</h3>
    892 <p>GCC 4.3 changed the uClinux configuration so that it uses the
    893 &lsquo;<samp>m68k-linux-gnu</samp>&rsquo; ABI rather than the &lsquo;<samp>m68k-elf</samp>&rsquo; ABI.
    894 It also added improved support for C++ and flat shared libraries,
    895 both of which were ABI changes.
    896 </p>
    897 <hr />
    898 <a name="mep_002dx_002delf"></a><a name="mep_002d_002a_002delf"></a>
    899 <h3 class="heading">mep-*-elf</h3>
    900 <p>Toshiba Media embedded Processor.
    901 This configuration is intended for embedded systems.
    902 </p>
    903 <hr />
    904 <a name="microblaze_002dx_002delf"></a><a name="microblaze_002d_002a_002delf"></a>
    905 <h3 class="heading">microblaze-*-elf</h3>
    906 <p>Xilinx MicroBlaze processor.
    907 This configuration is intended for embedded systems.
    908 </p>
    909 <hr />
    910 <a name="mips_002dx_002dx"></a><a name="mips_002d_002a_002d_002a"></a>
    911 <h3 class="heading">mips-*-*</h3>
    912 <p>If on a MIPS system you get an error message saying &ldquo;does not have gp
    913 sections for all it&rsquo;s [sic] sectons [sic]&rdquo;, don&rsquo;t worry about it.  This
    914 happens whenever you use GAS with the MIPS linker, but there is not
    915 really anything wrong, and it is okay to use the output file.  You can
    916 stop such warnings by installing the GNU linker.
    917 </p>
    918 <p>It would be nice to extend GAS to produce the gp tables, but they are
    919 optional, and there should not be a warning about their absence.
    920 </p>
    921 <p>The libstdc++ atomic locking routines for MIPS targets requires MIPS II
    922 and later.  A patch went in just after the GCC 3.3 release to
    923 make &lsquo;<samp>mips*-*-*</samp>&rsquo; use the generic implementation instead.  You can also
    924 configure for &lsquo;<samp>mipsel-elf</samp>&rsquo; as a workaround.  The
    925 &lsquo;<samp>mips*-*-linux*</samp>&rsquo; target continues to use the MIPS II routines.  More
    926 work on this is expected in future releases.
    927 </p>
    928 
    929 <p>The built-in <code>__sync_*</code> functions are available on MIPS II and
    930 later systems and others that support the &lsquo;<samp>ll</samp>&rsquo;, &lsquo;<samp>sc</samp>&rsquo; and
    931 &lsquo;<samp>sync</samp>&rsquo; instructions.  This can be overridden by passing
    932 <samp>--with-llsc</samp> or <samp>--without-llsc</samp> when configuring GCC.
    933 Since the Linux kernel emulates these instructions if they are
    934 missing, the default for &lsquo;<samp>mips*-*-linux*</samp>&rsquo; targets is
    935 <samp>--with-llsc</samp>.  The <samp>--with-llsc</samp> and
    936 <samp>--without-llsc</samp> configure options may be overridden at compile
    937 time by passing the <samp>-mllsc</samp> or <samp>-mno-llsc</samp> options to
    938 the compiler.
    939 </p>
    940 <p>MIPS systems check for division by zero (unless
    941 <samp>-mno-check-zero-division</samp> is passed to the compiler) by
    942 generating either a conditional trap or a break instruction.  Using
    943 trap results in smaller code, but is only supported on MIPS II and
    944 later.  Also, some versions of the Linux kernel have a bug that
    945 prevents trap from generating the proper signal (<code>SIGFPE</code>).  To enable
    946 the use of break, use the <samp>--with-divide=breaks</samp>
    947 <code>configure</code> option when configuring GCC.  The default is to
    948 use traps on systems that support them.
    949 </p>
    950 <p>The assembler from GNU binutils 2.17 and earlier has a bug in the way
    951 it sorts relocations for REL targets (o32, o64, EABI).  This can cause
    952 bad code to be generated for simple C++ programs.  Also the linker
    953 from GNU binutils versions prior to 2.17 has a bug which causes the
    954 runtime linker stubs in very large programs, like <samp>libgcj.so</samp>, to
    955 be incorrectly generated.  GNU Binutils 2.18 and later (and snapshots
    956 made after Nov. 9, 2006) should be free from both of these problems.
    957 </p>
    958 <hr />
    959 <a name="mips_002dsgi_002dirix5"></a><a name="mips_002dsgi_002dirix5-1"></a>
    960 <h3 class="heading">mips-sgi-irix5</h3>
    961 <p>Support for IRIX 5 has been removed in GCC 4.6.
    962 </p>
    963 <hr />
    964 <a name="mips_002dsgi_002dirix6"></a><a name="mips_002dsgi_002dirix6-1"></a>
    965 <h3 class="heading">mips-sgi-irix6</h3>
    966 <p>Support for IRIX 6.5 has been removed in GCC 4.8.  Support for IRIX 6
    967 releases before 6.5 has been removed in GCC 4.6, as well as support for
    968 the O32 ABI.
    969 </p>
    970 <hr />
    971 <a name="moxie_002dx_002delf"></a><a name="moxie_002d_002a_002delf"></a>
    972 <h3 class="heading">moxie-*-elf</h3>
    973 <p>The moxie processor.
    974 </p>
    975 <hr />
    976 <a name="msp430_002dx_002delf"></a><a name="msp430_002d_002a_002delf"></a>
    977 <h3 class="heading">msp430-*-elf</h3>
    978 <p>TI MSP430 processor.
    979 This configuration is intended for embedded systems.
    980 </p>
    981 <hr />
    982 <a name="nds32le_002dx_002delf"></a><a name="nds32le_002d_002a_002delf"></a>
    983 <h3 class="heading">nds32le-*-elf</h3>
    984 <p>Andes NDS32 target in little endian mode.
    985 </p>
    986 <hr />
    987 <a name="nds32be_002dx_002delf"></a><a name="nds32be_002d_002a_002delf"></a>
    988 <h3 class="heading">nds32be-*-elf</h3>
    989 <p>Andes NDS32 target in big endian mode.
    990 </p>
    991 <hr />
    992 <a name="nvptx_002dx_002dnone"></a><a name="nvptx_002d_002a_002dnone"></a>
    993 <h3 class="heading">nvptx-*-none</h3>
    994 <p>Nvidia PTX target.
    995 </p>
    996 <p>Instead of GNU binutils, you will need to install
    997 <a href="https://github.com/MentorEmbedded/nvptx-tools/">nvptx-tools</a>.
    998 Tell GCC where to find it:
    999 <samp>--with-build-time-tools=[install-nvptx-tools]/nvptx-none/bin</samp>.
   1000 </p>
   1001 <p>A nvptx port of newlib is available at
   1002 <a href="https://github.com/MentorEmbedded/nvptx-newlib/">nvptx-newlib</a>.
   1003 It can be automatically built together with GCC.  For this, add a
   1004 symbolic link to nvptx-newlib&rsquo;s <samp>newlib</samp> directory to the
   1005 directory containing the GCC sources.
   1006 </p>
   1007 <p>Use the <samp>--disable-sjlj-exceptions</samp> and
   1008 <samp>--enable-newlib-io-long-long</samp> options when configuring.
   1009 </p>
   1010 <hr />
   1011 <a name="powerpc_002dx_002dx"></a><a name="powerpc_002d_002a_002d_002a"></a>
   1012 <h3 class="heading">powerpc-*-*</h3>
   1013 <p>You can specify a default version for the <samp>-mcpu=<var>cpu_type</var></samp>
   1014 switch by using the configure option <samp>--with-cpu-<var>cpu_type</var></samp>.
   1015 </p>
   1016 <p>You will need
   1017 <a href="ftp://ftp.kernel.org/pub/linux/devel/binutils">binutils 2.15</a>
   1018 or newer for a working GCC.
   1019 </p>
   1020 <hr />
   1021 <a name="powerpc_002dx_002ddarwin"></a><a name="powerpc_002d_002a_002ddarwin_002a"></a>
   1022 <h3 class="heading">powerpc-*-darwin*</h3>
   1023 <p>PowerPC running Darwin (Mac OS X kernel).
   1024 </p>
   1025 <p>Pre-installed versions of Mac OS X may not include any developer tools,
   1026 meaning that you will not be able to build GCC from source.  Tool
   1027 binaries are available at
   1028 <a href="http://opensource.apple.com/">http://opensource.apple.com/</a>.
   1029 </p>
   1030 <p>This version of GCC requires at least cctools-590.36.  The
   1031 cctools-590.36 package referenced from
   1032 <a href="http://gcc.gnu.org/ml/gcc/2006-03/msg00507.html">http://gcc.gnu.org/ml/gcc/2006-03/msg00507.html</a> will not work
   1033 on systems older than 10.3.9 (aka darwin7.9.0).
   1034 </p>
   1035 <hr />
   1036 <a name="powerpc_002dx_002delf"></a><a name="powerpc_002d_002a_002delf"></a>
   1037 <h3 class="heading">powerpc-*-elf</h3>
   1038 <p>PowerPC system in big endian mode, running System V.4.
   1039 </p>
   1040 <hr />
   1041 <a name="powerpc_002dx_002dlinux_002dgnu"></a><a name="powerpc_002a_002d_002a_002dlinux_002dgnu_002a"></a>
   1042 <h3 class="heading">powerpc*-*-linux-gnu*</h3>
   1043 <p>PowerPC system in big endian mode running Linux.
   1044 </p>
   1045 <hr />
   1046 <a name="powerpc_002dx_002dnetbsd"></a><a name="powerpc_002d_002a_002dnetbsd_002a"></a>
   1047 <h3 class="heading">powerpc-*-netbsd*</h3>
   1048 <p>PowerPC system in big endian mode running NetBSD.
   1049 </p>
   1050 <hr />
   1051 <a name="powerpc_002dx_002deabisim"></a><a name="powerpc_002d_002a_002deabisim"></a>
   1052 <h3 class="heading">powerpc-*-eabisim</h3>
   1053 <p>Embedded PowerPC system in big endian mode for use in running under the
   1054 PSIM simulator.
   1055 </p>
   1056 <hr />
   1057 <a name="powerpc_002dx_002deabi"></a><a name="powerpc_002d_002a_002deabi"></a>
   1058 <h3 class="heading">powerpc-*-eabi</h3>
   1059 <p>Embedded PowerPC system in big endian mode.
   1060 </p>
   1061 <hr />
   1062 <a name="powerpcle_002dx_002delf"></a><a name="powerpcle_002d_002a_002delf"></a>
   1063 <h3 class="heading">powerpcle-*-elf</h3>
   1064 <p>PowerPC system in little endian mode, running System V.4.
   1065 </p>
   1066 <hr />
   1067 <a name="powerpcle_002dx_002deabisim"></a><a name="powerpcle_002d_002a_002deabisim"></a>
   1068 <h3 class="heading">powerpcle-*-eabisim</h3>
   1069 <p>Embedded PowerPC system in little endian mode for use in running under
   1070 the PSIM simulator.
   1071 </p>
   1072 <hr />
   1073 <a name="powerpcle_002dx_002deabi"></a><a name="powerpcle_002d_002a_002deabi"></a>
   1074 <h3 class="heading">powerpcle-*-eabi</h3>
   1075 <p>Embedded PowerPC system in little endian mode.
   1076 </p>
   1077 <hr />
   1078 <a name="rl78_002dx_002delf"></a><a name="rl78_002d_002a_002delf"></a>
   1079 <h3 class="heading">rl78-*-elf</h3>
   1080 <p>The Renesas RL78 processor.
   1081 This configuration is intended for embedded systems.
   1082 </p>
   1083 <hr />
   1084 <a name="rx_002dx_002delf"></a><a name="rx_002d_002a_002delf"></a>
   1085 <h3 class="heading">rx-*-elf</h3>
   1086 <p>The Renesas RX processor.  See
   1087 <a href="http://eu.renesas.com/fmwk.jsp?cnt=rx600_series_landing.jsp&fp=/products/mpumcu/rx_family/rx600_series">http://eu.renesas.com/fmwk.jsp?cnt=rx600_series_landing.jsp&fp=/products/mpumcu/rx_family/rx600_series</a>
   1088 for more information about this processor.
   1089 </p>
   1090 <hr />
   1091 <a name="s390_002dx_002dlinux"></a><a name="s390_002d_002a_002dlinux_002a"></a>
   1092 <h3 class="heading">s390-*-linux*</h3>
   1093 <p>S/390 system running GNU/Linux for S/390.
   1094 </p>
   1095 <hr />
   1096 <a name="s390x_002dx_002dlinux"></a><a name="s390x_002d_002a_002dlinux_002a"></a>
   1097 <h3 class="heading">s390x-*-linux*</h3>
   1098 <p>zSeries system (64-bit) running GNU/Linux for zSeries.
   1099 </p>
   1100 <hr />
   1101 <a name="s390x_002dibm_002dtpf"></a><a name="s390x_002dibm_002dtpf_002a"></a>
   1102 <h3 class="heading">s390x-ibm-tpf*</h3>
   1103 <p>zSeries system (64-bit) running TPF.  This platform is
   1104 supported as cross-compilation target only.
   1105 </p>
   1106 <hr />
   1107 <a name="x_002dx_002dsolaris2"></a><a name="g_t_002a_002d_002a_002dsolaris2_002a"></a>
   1108 <h3 class="heading">*-*-solaris2*</h3>
   1109 <p>Support for Solaris 9 has been removed in GCC 5.  Support for Solaris
   1110 8 has been removed in GCC 4.8.  Support for Solaris 7 has been removed
   1111 in GCC 4.6.
   1112 </p>
   1113 <p>Sun does not ship a C compiler with Solaris 2 before Solaris 10, though
   1114 you can download the Sun Studio compilers for free.  In Solaris 10 and
   1115 11, GCC 3.4.3 is available as <code>/usr/sfw/bin/gcc</code>.  Solaris 11
   1116 also provides GCC 4.5.2, 4.7.3, and 4.8.2 as
   1117 <code>/usr/gcc/4.5/bin/gcc</code> or similar.  Alternatively,
   1118 you can install a pre-built GCC to bootstrap and install GCC.  See the
   1119 <a href="binaries.html">binaries page</a> for details.
   1120 </p>
   1121 <p>The Solaris 2 <code>/bin/sh</code> will often fail to configure
   1122 &lsquo;<samp>libstdc++-v3</samp>&rsquo;, &lsquo;<samp>boehm-gc</samp>&rsquo; or &lsquo;<samp>libjava</samp>&rsquo;.  We therefore
   1123 recommend using the following initial sequence of commands
   1124 </p>
   1125 <div class="smallexample">
   1126 <pre class="smallexample">% CONFIG_SHELL=/bin/ksh
   1127 % export CONFIG_SHELL
   1128 </pre></div>
   1129 
   1130 <p>and proceed as described in <a href="configure.html">the configure instructions</a>.
   1131 In addition we strongly recommend specifying an absolute path to invoke
   1132 <code><var>srcdir</var>/configure</code>.
   1133 </p>
   1134 <p>Solaris 10 comes with a number of optional OS packages.  Some of these
   1135 are needed to use GCC fully, namely <code>SUNWarc</code>,
   1136 <code>SUNWbtool</code>, <code>SUNWesu</code>, <code>SUNWhea</code>, <code>SUNWlibm</code>,
   1137 <code>SUNWsprot</code>, and <code>SUNWtoo</code>.  If you did not install all
   1138 optional packages when installing Solaris 10, you will need to verify that
   1139 the packages that GCC needs are installed.
   1140 To check whether an optional package is installed, use
   1141 the <code>pkginfo</code> command.  To add an optional package, use the
   1142 <code>pkgadd</code> command.  For further details, see the Solaris 10
   1143 documentation.
   1144 </p>
   1145 <p>Starting with Solaris 11, the package management has changed, so you
   1146 need to check for <code>system/header</code>, <code>system/linker</code>, and
   1147 <code>developer/assembler</code> packages.  Checking for and installing
   1148 packages is done with the <code>pkg</code> command now.
   1149 </p>
   1150 <p>Trying to use the linker and other tools in
   1151 <samp>/usr/ucb</samp> to install GCC has been observed to cause trouble.
   1152 For example, the linker may hang indefinitely.  The fix is to remove
   1153 <samp>/usr/ucb</samp> from your <code>PATH</code>.
   1154 </p>
   1155 <p>The build process works more smoothly with the legacy Sun tools so, if you
   1156 have <samp>/usr/xpg4/bin</samp> in your <code>PATH</code>, we recommend that you place
   1157 <samp>/usr/bin</samp> before <samp>/usr/xpg4/bin</samp> for the duration of the build.
   1158 </p>
   1159 <p>We recommend the use of the Solaris assembler or the GNU assembler, in
   1160 conjunction with the Solaris linker.  The GNU <code>as</code>
   1161 versions included in Solaris 10, from GNU binutils 2.15 (in
   1162 <samp>/usr/sfw/bin/gas</samp>), and Solaris 11,
   1163 from GNU binutils 2.19 or newer (also in <samp>/usr/bin/gas</samp> and
   1164 <samp>/usr/gnu/bin/as</samp>), are known to work.
   1165 Current versions of GNU binutils (2.26)
   1166 are known to work as well, with the caveat mentioned in
   1167 <a href="#ix86-x-solaris210">i?86-*-solaris2.10</a> .  Note that your mileage may vary
   1168 if you use a combination of the GNU tools and the Solaris tools: while the
   1169 combination GNU <code>as</code> + Sun <code>ld</code> should reasonably work,
   1170 the reverse combination Sun <code>as</code> + GNU <code>ld</code> may fail to
   1171 build or cause memory corruption at runtime in some cases for C++ programs.
   1172 GNU <code>ld</code> usually works as well, although the version included in
   1173 Solaris 10 cannot be used due to several bugs.  Again, the current
   1174 version (2.26) is known to work, but generally lacks platform specific
   1175 features, so better stay with Solaris <code>ld</code>.  To use the LTO linker
   1176 plugin (<samp>-fuse-linker-plugin</samp>) with GNU <code>ld</code>, GNU
   1177 binutils <em>must</em> be configured with <samp>--enable-largefile</samp>.
   1178 </p>
   1179 <p>To enable symbol versioning in &lsquo;<samp>libstdc++</samp>&rsquo; with the Solaris linker,
   1180 you need to have any version of GNU <code>c++filt</code>, which is part of
   1181 GNU binutils.  &lsquo;<samp>libstdc++</samp>&rsquo; symbol versioning will be disabled if no
   1182 appropriate version is found.  Solaris <code>c++filt</code> from the Solaris
   1183 Studio compilers does <em>not</em> work.
   1184 </p>
   1185 <p>GNU <code>make</code> version 3.81 or later is required to build libjava
   1186 with the Solaris linker.
   1187 </p>
   1188 <p>Sun bug 4927647 sometimes causes random spurious testsuite failures
   1189 related to missing diagnostic output.  This bug doesn&rsquo;t affect GCC
   1190 itself, rather it is a kernel bug triggered by the <code>expect</code>
   1191 program which is used only by the GCC testsuite driver.  When the bug
   1192 causes the <code>expect</code> program to miss anticipated output, extra
   1193 testsuite failures appear.
   1194 </p>
   1195 <hr />
   1196 <a name="sparc_002dx_002dx"></a><a name="sparc_002a_002d_002a_002d_002a"></a>
   1197 <h3 class="heading">sparc*-*-*</h3>
   1198 <p>This section contains general configuration information for all
   1199 SPARC-based platforms.  In addition to reading this section, please
   1200 read all other sections that match your target.
   1201 </p>
   1202 <p>Newer versions of the GNU Multiple Precision Library (GMP), the MPFR
   1203 library and the MPC library are known to be miscompiled by earlier
   1204 versions of GCC on these platforms.  We therefore recommend the use
   1205 of the exact versions of these libraries listed as minimal versions
   1206 in <a href="prerequisites.html">the prerequisites</a>.
   1207 </p>
   1208 <hr />
   1209 <a name="sparc_002dsun_002dsolaris2"></a><a name="sparc_002dsun_002dsolaris2_002a"></a>
   1210 <h3 class="heading">sparc-sun-solaris2*</h3>
   1211 <p>When GCC is configured to use GNU binutils 2.14 or later, the binaries
   1212 produced are smaller than the ones produced using Sun&rsquo;s native tools;
   1213 this difference is quite significant for binaries containing debugging
   1214 information.
   1215 </p>
   1216 <p>Starting with Solaris 7, the operating system is capable of executing
   1217 64-bit SPARC V9 binaries.  GCC 3.1 and later properly supports
   1218 this; the <samp>-m64</samp> option enables 64-bit code generation.
   1219 However, if all you want is code tuned for the UltraSPARC CPU, you
   1220 should try the <samp>-mtune=ultrasparc</samp> option instead, which produces
   1221 code that, unlike full 64-bit code, can still run on non-UltraSPARC
   1222 machines.
   1223 </p>
   1224 <p>When configuring the GNU Multiple Precision Library (GMP), the MPFR
   1225 library or the MPC library on a Solaris 7 or later system, the canonical
   1226 target triplet must be specified as the <code>build</code> parameter on the
   1227 configure line.  This target triplet can be obtained by invoking <code>./config.guess</code> in the toplevel source directory of GCC (and
   1228 not that of GMP or MPFR or MPC).  For example on a Solaris 9 system:
   1229 </p>
   1230 <div class="smallexample">
   1231 <pre class="smallexample">% ./configure --build=sparc-sun-solaris2.9 --prefix=xxx
   1232 </pre></div>
   1233 
   1234 <hr />
   1235 <a name="sparc_002dsun_002dsolaris210"></a><a name="sparc_002dsun_002dsolaris2_002e10"></a>
   1236 <h3 class="heading">sparc-sun-solaris2.10</h3>
   1237 <p>There is a bug in older versions of the Sun assembler which breaks
   1238 thread-local storage (TLS).  A typical error message is
   1239 </p>
   1240 <div class="smallexample">
   1241 <pre class="smallexample">ld: fatal: relocation error: R_SPARC_TLS_LE_HIX22: file /var/tmp//ccamPA1v.o:
   1242   symbol &lt;unknown&gt;: bad symbol type SECT: symbol type must be TLS
   1243 </pre></div>
   1244 
   1245 <p>This bug is fixed in Sun patch 118683-03 or later.
   1246 </p>
   1247 <hr />
   1248 <a name="sparc_002dx_002dlinux"></a><a name="sparc_002d_002a_002dlinux_002a"></a>
   1249 <h3 class="heading">sparc-*-linux*</h3>
   1250 
   1251 <p>GCC versions 3.0 and higher require binutils 2.11.2 and glibc 2.2.4
   1252 or newer on this platform.  All earlier binutils and glibc
   1253 releases mishandled unaligned relocations on <code>sparc-*-*</code> targets.
   1254 </p>
   1255 
   1256 <hr />
   1257 <a name="sparc64_002dx_002dsolaris2"></a><a name="sparc64_002d_002a_002dsolaris2_002a"></a>
   1258 <h3 class="heading">sparc64-*-solaris2*</h3>
   1259 <p>When configuring the GNU Multiple Precision Library (GMP), the MPFR
   1260 library or the MPC library, the canonical target triplet must be specified
   1261 as the <code>build</code> parameter on the configure line.  For example
   1262 on a Solaris 9 system:
   1263 </p>
   1264 <div class="smallexample">
   1265 <pre class="smallexample">% ./configure --build=sparc64-sun-solaris2.9 --prefix=xxx
   1266 </pre></div>
   1267 
   1268 <hr />
   1269 <a name="sparcv9_002dx_002dsolaris2"></a><a name="sparcv9_002d_002a_002dsolaris2_002a"></a>
   1270 <h3 class="heading">sparcv9-*-solaris2*</h3>
   1271 <p>This is a synonym for &lsquo;<samp>sparc64-*-solaris2*</samp>&rsquo;.
   1272 </p>
   1273 <hr />
   1274 <a name="c6x_002dx_002dx"></a><a name="c6x_002d_002a_002d_002a"></a>
   1275 <h3 class="heading">c6x-*-*</h3>
   1276 <p>The C6X family of processors. This port requires binutils-2.22 or newer.
   1277 </p>
   1278 <hr />
   1279 <a name="tilegx_002d_002a_002dlinux"></a><a name="tilegx_002d_002a_002dlinux_002a"></a>
   1280 <h3 class="heading">tilegx-*-linux*</h3>
   1281 <p>The TILE-Gx processor in little endian mode, running GNU/Linux.  This
   1282 port requires binutils-2.22 or newer.
   1283 </p>
   1284 <hr />
   1285 <a name="tilegxbe_002d_002a_002dlinux"></a><a name="tilegxbe_002d_002a_002dlinux_002a"></a>
   1286 <h3 class="heading">tilegxbe-*-linux*</h3>
   1287 <p>The TILE-Gx processor in big endian mode, running GNU/Linux.  This
   1288 port requires binutils-2.23 or newer.
   1289 </p>
   1290 <hr />
   1291 <a name="tilepro_002d_002a_002dlinux"></a><a name="tilepro_002d_002a_002dlinux_002a"></a>
   1292 <h3 class="heading">tilepro-*-linux*</h3>
   1293 <p>The TILEPro processor running GNU/Linux.  This port requires
   1294 binutils-2.22 or newer.
   1295 </p>
   1296 <hr />
   1297 <a name="visium_002dx_002delf"></a><a name="visium_002d_002a_002delf"></a>
   1298 <h3 class="heading">visium-*-elf</h3>
   1299 <p>CDS VISIUMcore processor.
   1300 This configuration is intended for embedded systems.
   1301 </p>
   1302 <hr />
   1303 <a name="x_002dx_002dvxworks"></a><a name="g_t_002a_002d_002a_002dvxworks_002a"></a>
   1304 <h3 class="heading">*-*-vxworks*</h3>
   1305 <p>Support for VxWorks is in flux.  At present GCC supports <em>only</em> the
   1306 very recent VxWorks 5.5 (aka Tornado 2.2) release, and only on PowerPC.
   1307 We welcome patches for other architectures supported by VxWorks 5.5.
   1308 Support for VxWorks AE would also be welcome; we believe this is merely
   1309 a matter of writing an appropriate &ldquo;configlette&rdquo; (see below).  We are
   1310 not interested in supporting older, a.out or COFF-based, versions of
   1311 VxWorks in GCC 3.
   1312 </p>
   1313 <p>VxWorks comes with an older version of GCC installed in
   1314 <samp><var>$WIND_BASE</var>/host</samp>; we recommend you do not overwrite it.
   1315 Choose an installation <var>prefix</var> entirely outside <var>$WIND_BASE</var>.
   1316 Before running <code>configure</code>, create the directories <samp><var>prefix</var></samp>
   1317 and <samp><var>prefix</var>/bin</samp>.  Link or copy the appropriate assembler,
   1318 linker, etc. into <samp><var>prefix</var>/bin</samp>, and set your <var>PATH</var> to
   1319 include that directory while running both <code>configure</code> and
   1320 <code>make</code>.
   1321 </p>
   1322 <p>You must give <code>configure</code> the
   1323 <samp>--with-headers=<var>$WIND_BASE</var>/target/h</samp> switch so that it can
   1324 find the VxWorks system headers.  Since VxWorks is a cross compilation
   1325 target only, you must also specify <samp>--target=<var>target</var></samp>.
   1326 <code>configure</code> will attempt to create the directory
   1327 <samp><var>prefix</var>/<var>target</var>/sys-include</samp> and copy files into it;
   1328 make sure the user running <code>configure</code> has sufficient privilege
   1329 to do so.
   1330 </p>
   1331 <p>GCC&rsquo;s exception handling runtime requires a special &ldquo;configlette&rdquo;
   1332 module, <samp>contrib/gthr_supp_vxw_5x.c</samp>.  Follow the instructions in
   1333 that file to add the module to your kernel build.  (Future versions of
   1334 VxWorks will incorporate this module.)
   1335 </p>
   1336 <hr />
   1337 <a name="x86_002d64_002dx_002dx"></a><a name="x86_005f64_002d_002a_002d_002a_002c-amd64_002d_002a_002d_002a"></a>
   1338 <h3 class="heading">x86_64-*-*, amd64-*-*</h3>
   1339 <p>GCC supports the x86-64 architecture implemented by the AMD64 processor
   1340 (amd64-*-* is an alias for x86_64-*-*) on GNU/Linux, FreeBSD and NetBSD.
   1341 On GNU/Linux the default is a bi-arch compiler which is able to generate
   1342 both 64-bit x86-64 and 32-bit x86 code (via the <samp>-m32</samp> switch).
   1343 </p>
   1344 <hr />
   1345 <a name="x86_002d64_002dx_002dsolaris210"></a><a name="x86_005f64_002d_002a_002dsolaris2_002e1_005b0_002d9_005d_002a"></a>
   1346 <h3 class="heading">x86_64-*-solaris2.1[0-9]*</h3>
   1347 <p>GCC also supports the x86-64 architecture implemented by the AMD64
   1348 processor (&lsquo;<samp>amd64-*-*</samp>&rsquo; is an alias for &lsquo;<samp>x86_64-*-*</samp>&rsquo;) on
   1349 Solaris 10 or later.  Unlike other systems, without special options a
   1350 bi-arch compiler is built which generates 32-bit code by default, but
   1351 can generate 64-bit x86-64 code with the <samp>-m64</samp> switch.  Since
   1352 GCC 4.7, there is also a configuration that defaults to 64-bit code, but
   1353 can generate 32-bit code with <samp>-m32</samp>.  To configure and build
   1354 this way, you have to provide all support libraries like <samp>libgmp</samp>
   1355 as 64-bit code, configure with <samp>--target=x86_64-pc-solaris2.1x</samp>
   1356 and &lsquo;<samp>CC=gcc -m64</samp>&rsquo;.
   1357 </p>
   1358 <hr />
   1359 <a name="xtensa_002dx_002delf"></a><a name="xtensa_002a_002d_002a_002delf"></a>
   1360 <h3 class="heading">xtensa*-*-elf</h3>
   1361 <p>This target is intended for embedded Xtensa systems using the
   1362 &lsquo;<samp>newlib</samp>&rsquo; C library.  It uses ELF but does not support shared
   1363 objects.  Designed-defined instructions specified via the
   1364 Tensilica Instruction Extension (TIE) language are only supported
   1365 through inline assembly.
   1366 </p>
   1367 <p>The Xtensa configuration information must be specified prior to
   1368 building GCC.  The <samp>include/xtensa-config.h</samp> header
   1369 file contains the configuration information.  If you created your
   1370 own Xtensa configuration with the Xtensa Processor Generator, the
   1371 downloaded files include a customized copy of this header file,
   1372 which you can use to replace the default header file.
   1373 </p>
   1374 <hr />
   1375 <a name="xtensa_002dx_002dlinux"></a><a name="xtensa_002a_002d_002a_002dlinux_002a"></a>
   1376 <h3 class="heading">xtensa*-*-linux*</h3>
   1377 <p>This target is for Xtensa systems running GNU/Linux.  It supports ELF
   1378 shared objects and the GNU C library (glibc).  It also generates
   1379 position-independent code (PIC) regardless of whether the
   1380 <samp>-fpic</samp> or <samp>-fPIC</samp> options are used.  In other
   1381 respects, this target is the same as the
   1382 <a href="#xtensa*-*-elf">&lsquo;<samp>xtensa*-*-elf</samp>&rsquo;</a> target.
   1383 </p>
   1384 <hr />
   1385 <a name="windows"></a><a name="Microsoft-Windows"></a>
   1386 <h3 class="heading">Microsoft Windows</h3>
   1387 
   1388 <a name="Intel-16_002dbit-versions"></a>
   1389 <h4 class="subheading">Intel 16-bit versions</h4>
   1390 <p>The 16-bit versions of Microsoft Windows, such as Windows 3.1, are not
   1391 supported.
   1392 </p>
   1393 <p>However, the 32-bit port has limited support for Microsoft
   1394 Windows 3.11 in the Win32s environment, as a target only.  See below.
   1395 </p>
   1396 <a name="Intel-32_002dbit-versions"></a>
   1397 <h4 class="subheading">Intel 32-bit versions</h4>
   1398 <p>The 32-bit versions of Windows, including Windows 95, Windows NT, Windows
   1399 XP, and Windows Vista, are supported by several different target
   1400 platforms.  These targets differ in which Windows subsystem they target
   1401 and which C libraries are used.
   1402 </p>
   1403 <ul>
   1404 <li> Cygwin <a href="#x-x-cygwin">*-*-cygwin</a>: Cygwin provides a user-space
   1405 Linux API emulation layer in the Win32 subsystem.
   1406 </li><li> Interix <a href="#x-x-interix">*-*-interix</a>: The Interix subsystem
   1407 provides native support for POSIX.
   1408 </li><li> MinGW <a href="#x-x-mingw32">*-*-mingw32</a>: MinGW is a native GCC port for
   1409 the Win32 subsystem that provides a subset of POSIX.
   1410 </li><li> MKS i386-pc-mks: NuTCracker from MKS.  See
   1411 <a href="http://www.mkssoftware.com/">http://www.mkssoftware.com/</a> for more information.
   1412 </li></ul>
   1413 
   1414 <a name="Intel-64_002dbit-versions"></a>
   1415 <h4 class="subheading">Intel 64-bit versions</h4>
   1416 <p>GCC contains support for x86-64 using the mingw-w64
   1417 runtime library, available from <a href="http://mingw-w64.sourceforge.net/">http://mingw-w64.sourceforge.net/</a>.
   1418 This library should be used with the target triple x86_64-pc-mingw32.
   1419 </p>
   1420 <p>Presently Windows for Itanium is not supported.
   1421 </p>
   1422 <a name="Windows-CE"></a>
   1423 <h4 class="subheading">Windows CE</h4>
   1424 <p>Windows CE is supported as a target only on Hitachi
   1425 SuperH (sh-wince-pe), and MIPS (mips-wince-pe).
   1426 </p>
   1427 <a name="Other-Windows-Platforms"></a>
   1428 <h4 class="subheading">Other Windows Platforms</h4>
   1429 <p>GCC no longer supports Windows NT on the Alpha or PowerPC.
   1430 </p>
   1431 <p>GCC no longer supports the Windows POSIX subsystem.  However, it does
   1432 support the Interix subsystem.  See above.
   1433 </p>
   1434 <p>Old target names including *-*-winnt and *-*-windowsnt are no longer used.
   1435 </p>
   1436 <p>PW32 (i386-pc-pw32) support was never completed, and the project seems to
   1437 be inactive.  See <a href="http://pw32.sourceforge.net/">http://pw32.sourceforge.net/</a> for more information.
   1438 </p>
   1439 <p>UWIN support has been removed due to a lack of maintenance.
   1440 </p>
   1441 <hr />
   1442 <a name="x_002dx_002dcygwin"></a><a name="g_t_002a_002d_002a_002dcygwin"></a>
   1443 <h3 class="heading">*-*-cygwin</h3>
   1444 <p>Ports of GCC are included with the
   1445 <a href="http://www.cygwin.com/">Cygwin environment</a>.
   1446 </p>
   1447 <p>GCC will build under Cygwin without modification; it does not build
   1448 with Microsoft&rsquo;s C++ compiler and there are no plans to make it do so.
   1449 </p>
   1450 <p>The Cygwin native compiler can be configured to target any 32-bit x86
   1451 cpu architecture desired; the default is i686-pc-cygwin.  It should be
   1452 used with as up-to-date a version of binutils as possible; use either
   1453 the latest official GNU binutils release in the Cygwin distribution,
   1454 or version 2.20 or above if building your own.
   1455 </p>
   1456 <hr />
   1457 <a name="x_002dx_002dinterix"></a><a name="g_t_002a_002d_002a_002dinterix"></a>
   1458 <h3 class="heading">*-*-interix</h3>
   1459 <p>The Interix target is used by OpenNT, Interix, Services For UNIX (SFU),
   1460 and Subsystem for UNIX-based Applications (SUA).  Applications compiled
   1461 with this target run in the Interix subsystem, which is separate from
   1462 the Win32 subsystem.  This target was last known to work in GCC 3.3.
   1463 </p>
   1464 <hr />
   1465 <a name="x_002dx_002dmingw32"></a><a name="g_t_002a_002d_002a_002dmingw32"></a>
   1466 <h3 class="heading">*-*-mingw32</h3>
   1467 <p>GCC will build with and support only MinGW runtime 3.12 and later.
   1468 Earlier versions of headers are incompatible with the new default semantics
   1469 of <code>extern inline</code> in <code>-std=c99</code> and <code>-std=gnu99</code> modes.
   1470 </p>
   1471 <hr />
   1472 <a name="older"></a><a name="Older-systems"></a>
   1473 <h3 class="heading">Older systems</h3>
   1474 <p>GCC contains support files for many older (1980s and early
   1475 1990s) Unix variants.  For the most part, support for these systems
   1476 has not been deliberately removed, but it has not been maintained for
   1477 several years and may suffer from bitrot.
   1478 </p>
   1479 <p>Starting with GCC 3.1, each release has a list of &ldquo;obsoleted&rdquo; systems.
   1480 Support for these systems is still present in that release, but
   1481 <code>configure</code> will fail unless the <samp>--enable-obsolete</samp>
   1482 option is given.  Unless a maintainer steps forward, support for these
   1483 systems will be removed from the next release of GCC.
   1484 </p>
   1485 <p>Support for old systems as hosts for GCC can cause problems if the
   1486 workarounds for compiler, library and operating system bugs affect the
   1487 cleanliness or maintainability of the rest of GCC.  In some cases, to
   1488 bring GCC up on such a system, if still possible with current GCC, may
   1489 require first installing an old version of GCC which did work on that
   1490 system, and using it to compile a more recent GCC, to avoid bugs in the
   1491 vendor compiler.  Old releases of GCC 1 and GCC 2 are available in the
   1492 <samp>old-releases</samp> directory on the <a href="../mirrors.html">GCC mirror
   1493 sites</a>.  Header bugs may generally be avoided using
   1494 <code>fixincludes</code>, but bugs or deficiencies in libraries and the
   1495 operating system may still cause problems.
   1496 </p>
   1497 <p>Support for older systems as targets for cross-compilation is less
   1498 problematic than support for them as hosts for GCC; if an enthusiast
   1499 wishes to make such a target work again (including resurrecting any of
   1500 the targets that never worked with GCC 2, starting from the last
   1501 version before they were removed), patches
   1502 <a href="../contribute.html">following the usual requirements</a> would be
   1503 likely to be accepted, since they should not affect the support for more
   1504 modern targets.
   1505 </p>
   1506 <p>For some systems, old versions of GNU binutils may also be useful,
   1507 and are available from <samp>pub/binutils/old-releases</samp> on
   1508 <a href="http://sourceware.org/mirrors.html">sourceware.org mirror sites</a>.
   1509 </p>
   1510 <p>Some of the information on specific systems above relates to
   1511 such older systems, but much of the information
   1512 about GCC on such systems (which may no longer be applicable to
   1513 current GCC) is to be found in the GCC texinfo manual.
   1514 </p>
   1515 <hr />
   1516 <a name="elf"></a><a name="all-ELF-targets-_0028SVR4_002c-Solaris-2_002c-etc_002e_0029"></a>
   1517 <h3 class="heading">all ELF targets (SVR4, Solaris 2, etc.)</h3>
   1518 <p>C++ support is significantly better on ELF targets if you use the
   1519 <a href="./configure.html#with-gnu-ld">GNU linker</a>; duplicate copies of
   1520 inlines, vtables and template instantiations will be discarded
   1521 automatically.
   1522 </p>
   1523 
   1524 <hr />
   1525 <p>
   1526 <p><a href="./index.html">Return to the GCC Installation page</a>
   1527 </p>
   1528 
   1529 
   1530 
   1531 
   1532 
   1533 
   1534 </body>
   1535 </html>
   1536