xref: /netbsd-src/external/gpl3/gcc.old/dist/INSTALL/specific.html (revision 5dd36a3bc8bf2a9dec29ceb6349550414570c447)
1<html lang="en">
2<head>
3<title>Host/Target specific installation notes for GCC</title>
4<meta http-equiv="Content-Type" content="text/html">
5<meta name="description" content="Host/Target specific installation notes for GCC">
6<meta name="generator" content="makeinfo 4.13">
7<link title="Top" rel="top" href="#Top">
8<link href="http://www.gnu.org/software/texinfo/" rel="generator-home" title="Texinfo Homepage">
9<!--
10Copyright (C) 1988-2017 Free Software Foundation, Inc.
11
12   Permission is granted to copy, distribute and/or modify this document
13under the terms of the GNU Free Documentation License, Version 1.3 or
14any later version published by the Free Software Foundation; with no
15Invariant Sections, the Front-Cover texts being (a) (see below), and
16with the Back-Cover Texts being (b) (see below).  A copy of the
17license is included in the section entitled "GNU Free Documentation License".
18
19(a) The FSF's Front-Cover Text is:
20
21     A GNU Manual
22
23(b) The FSF's Back-Cover Text is:
24
25     You have freedom to copy and modify this GNU Manual, like GNU
26     software.  Copies published by the Free Software Foundation raise
27     funds for GNU development.-->
28<meta http-equiv="Content-Style-Type" content="text/css">
29<style type="text/css"><!--
30  pre.display { font-family:inherit }
31  pre.format  { font-family:inherit }
32  pre.smalldisplay { font-family:inherit; font-size:smaller }
33  pre.smallformat  { font-family:inherit; font-size:smaller }
34  pre.smallexample { font-size:smaller }
35  pre.smalllisp    { font-size:smaller }
36  span.sc    { font-variant:small-caps }
37  span.roman { font-family:serif; font-weight:normal; }
38  span.sansserif { font-family:sans-serif; font-weight:normal; }
39--></style>
40</head>
41<body>
42<h1 class="settitle">Host/Target specific installation notes for GCC</h1>
43<a name="index-Specific-1"></a><a name="index-Specific-installation-notes-2"></a><a name="index-Target-specific-installation-3"></a><a name="index-Host-specific-installation-4"></a><a name="index-Target-specific-installation-notes-5"></a>
44Please read this document carefully <em>before</em> installing the
45GNU Compiler Collection on your machine.
46
47   <p>Note that this list of install notes is <em>not</em> a list of supported
48hosts or targets.  Not all supported hosts and targets are listed
49here, only the ones that require host-specific or target-specific
50information have to.
51
52     <ul>
53<li><a href="#aarch64-x-x">aarch64*-*-*</a>
54<li><a href="#alpha-x-x">alpha*-*-*</a>
55<li><a href="#alpha-dec-osf51">alpha*-dec-osf5.1</a>
56<li><a href="#amd64-x-solaris210">amd64-*-solaris2.10</a>
57<li><a href="#arm-x-eabi">arm-*-eabi</a>
58<li><a href="#avr">avr</a>
59<li><a href="#bfin">Blackfin</a>
60<li><a href="#dos">DOS</a>
61<li><a href="#x-x-freebsd">*-*-freebsd*</a>
62<li><a href="#h8300-hms">h8300-hms</a>
63<li><a href="#hppa-hp-hpux">hppa*-hp-hpux*</a>
64<li><a href="#hppa-hp-hpux10">hppa*-hp-hpux10</a>
65<li><a href="#hppa-hp-hpux11">hppa*-hp-hpux11</a>
66<li><a href="#x-x-linux-gnu">*-*-linux-gnu</a>
67<li><a href="#ix86-x-linux">i?86-*-linux*</a>
68<li><a href="#ix86-x-solaris210">i?86-*-solaris2.10</a>
69<li><a href="#ia64-x-linux">ia64-*-linux</a>
70<li><a href="#ia64-x-hpux">ia64-*-hpux*</a>
71<li><a href="#x-ibm-aix">*-ibm-aix*</a>
72<li><a href="#iq2000-x-elf">iq2000-*-elf</a>
73<li><a href="#lm32-x-elf">lm32-*-elf</a>
74<li><a href="#lm32-x-uclinux">lm32-*-uclinux</a>
75<li><a href="#m32c-x-elf">m32c-*-elf</a>
76<li><a href="#m32r-x-elf">m32r-*-elf</a>
77<li><a href="#m68k-x-x">m68k-*-*</a>
78<li><a href="#m68k-uclinux">m68k-uclinux</a>
79<li><a href="#microblaze-x-elf">microblaze-*-elf</a>
80<li><a href="#mips-x-x">mips-*-*</a>
81<li><a href="#mips-sgi-irix5">mips-sgi-irix5</a>
82<li><a href="#mips-sgi-irix6">mips-sgi-irix6</a>
83<li><a href="#nds32le-x-elf">nds32le-*-elf</a>
84<li><a href="#nds32be-x-elf">nds32be-*-elf</a>
85<li><a href="#nvptx-x-none">nvptx-*-none</a>
86<li><a href="#powerpc-x-x">powerpc*-*-*</a>
87<li><a href="#powerpc-x-darwin">powerpc-*-darwin*</a>
88<li><a href="#powerpc-x-elf">powerpc-*-elf</a>
89<li><a href="#powerpc-x-linux-gnu">powerpc*-*-linux-gnu*</a>
90<li><a href="#powerpc-x-netbsd">powerpc-*-netbsd*</a>
91<li><a href="#powerpc-x-eabisim">powerpc-*-eabisim</a>
92<li><a href="#powerpc-x-eabi">powerpc-*-eabi</a>
93<li><a href="#powerpcle-x-elf">powerpcle-*-elf</a>
94<li><a href="#powerpcle-x-eabisim">powerpcle-*-eabisim</a>
95<li><a href="#powerpcle-x-eabi">powerpcle-*-eabi</a>
96<li><a href="#riscv32-x-elf">riscv32-*-elf</a>
97<li><a href="#riscv32-x-linux">riscv32-*-linux</a>
98<li><a href="#riscv64-x-elf">riscv64-*-elf</a>
99<li><a href="#riscv64-x-linux">riscv64-*-linux</a>
100<li><a href="#s390-x-linux">s390-*-linux*</a>
101<li><a href="#s390x-x-linux">s390x-*-linux*</a>
102<li><a href="#s390x-ibm-tpf">s390x-ibm-tpf*</a>
103<li><a href="#x-x-solaris2">*-*-solaris2*</a>
104<li><a href="#sparc-x-x">sparc*-*-*</a>
105<li><a href="#sparc-sun-solaris2">sparc-sun-solaris2*</a>
106<li><a href="#sparc-sun-solaris210">sparc-sun-solaris2.10</a>
107<li><a href="#sparc-x-linux">sparc-*-linux*</a>
108<li><a href="#sparc64-x-solaris2">sparc64-*-solaris2*</a>
109<li><a href="#sparcv9-x-solaris2">sparcv9-*-solaris2*</a>
110<li><a href="#c6x-x-x">c6x-*-*</a>
111<li><a href="#tilegx-x-linux">tilegx-*-linux*</a>
112<li><a href="#tilegxbe-x-linux">tilegxbe-*-linux*</a>
113<li><a href="#tilepro-x-linux">tilepro-*-linux*</a>
114<li><a href="#visium-x-elf">visium-*-elf</a>
115<li><a href="#x-x-vxworks">*-*-vxworks*</a>
116<li><a href="#x86-64-x-x">x86_64-*-*</a> amd64-*-*
117<li><a href="#x86-64-x-solaris210">x86_64-*-solaris2.1[0-9]*</a>
118<li><a href="#xtensa-x-elf">xtensa*-*-elf</a>
119<li><a href="#xtensa-x-linux">xtensa*-*-linux*</a>
120<li><a href="#windows">Microsoft Windows</a>
121<li><a href="#x-x-cygwin">*-*-cygwin</a>
122<li><a href="#x-x-mingw32">*-*-mingw32</a>
123<li><a href="#os2">OS/2</a>
124<li><a href="#older">Older systems</a>
125</ul>
126
127     <ul>
128<li><a href="#elf">all ELF targets</a> (SVR4, Solaris 2, etc.)
129</ul>
130
131   <p><!-- -------- host/target specific issues start here ---------------- -->
132<hr /><a name="aarch64-x-x"></a>
133
134<h3 class="heading"><a name="TOC0"></a>aarch64*-*-*</h3>
135
136<p>Binutils pre 2.24 does not have support for selecting <samp><span class="option">-mabi</span></samp> and
137does not support ILP32.  If it is used to build GCC 4.9 or later, GCC will
138not support option <samp><span class="option">-mabi=ilp32</span></samp>.
139
140   <p>To enable a workaround for the Cortex-A53 erratum number 835769 by default
141(for all CPUs regardless of -mcpu option given) at configure time use the
142<samp><span class="option">--enable-fix-cortex-a53-835769</span></samp> option.  This will enable the fix by
143default and can be explicitly disabled during compilation by passing the
144<samp><span class="option">-mno-fix-cortex-a53-835769</span></samp> option.  Conversely,
145<samp><span class="option">--disable-fix-cortex-a53-835769</span></samp> will disable the workaround by
146default.  The workaround is disabled by default if neither of
147<samp><span class="option">--enable-fix-cortex-a53-835769</span></samp> or
148<samp><span class="option">--disable-fix-cortex-a53-835769</span></samp> is given at configure time.
149
150   <p>To enable a workaround for the Cortex-A53 erratum number 843419 by default
151(for all CPUs regardless of -mcpu option given) at configure time use the
152<samp><span class="option">--enable-fix-cortex-a53-843419</span></samp> option.  This workaround is applied at
153link time.  Enabling the workaround will cause GCC to pass the relevant option
154to the linker.  It can be explicitly disabled during compilation by passing the
155<samp><span class="option">-mno-fix-cortex-a53-843419</span></samp> option.  Conversely,
156<samp><span class="option">--disable-fix-cortex-a53-843419</span></samp> will disable the workaround by default.
157The workaround is disabled by default if neither of
158<samp><span class="option">--enable-fix-cortex-a53-843419</span></samp> or
159<samp><span class="option">--disable-fix-cortex-a53-843419</span></samp> is given at configure time.
160
161   <p><hr /><a name="alpha-x-x"></a>
162
163<h3 class="heading"><a name="TOC1"></a>alpha*-*-*</h3>
164
165<p>This section contains general configuration information for all
166alpha-based platforms using ELF (in particular, ignore this section for
167DEC OSF/1, Digital UNIX and Tru64 UNIX).  In addition to reading this
168section, please read all other sections that match your target.
169
170   <p>We require binutils 2.11.2 or newer.
171Previous binutils releases had a number of problems with DWARF 2
172debugging information, not the least of which is incorrect linking of
173shared libraries.
174
175   <p><hr /><a name="alpha-dec-osf51"></a>
176
177<h3 class="heading"><a name="TOC2"></a>alpha*-dec-osf5.1</h3>
178
179<p>Systems using processors that implement the DEC Alpha architecture and
180are running the DEC/Compaq/HP Unix (DEC OSF/1, Digital UNIX, or Compaq/HP
181Tru64 UNIX) operating system, for example the DEC Alpha AXP systems.
182
183   <p>Support for Tru64 UNIX V5.1 has been removed in GCC 4.8.  As of GCC 4.6,
184support for Tru64 UNIX V4.0 and V5.0 has been removed.  As of GCC 3.2,
185versions before <code>alpha*-dec-osf4</code> are no longer supported.  (These
186are the versions which identify themselves as DEC OSF/1.)
187
188   <p><hr /><a name="amd64-x-solaris210"></a>
189
190<h3 class="heading"><a name="TOC3"></a>amd64-*-solaris2.1[0-9]*</h3>
191
192<p>This is a synonym for &lsquo;<samp><span class="samp">x86_64-*-solaris2.1[0-9]*</span></samp>&rsquo;.
193
194   <p><hr /><a name="arc-x-elf32"></a>
195
196<h3 class="heading"><a name="TOC4"></a>arc-*-elf32</h3>
197
198<p>Use &lsquo;<samp><span class="samp">configure --target=arc-elf32 --with-cpu=</span><var>cpu</var><span class="samp"> --enable-languages="c,c++"</span></samp>&rsquo;
199to configure GCC, with <var>cpu</var> being one of &lsquo;<samp><span class="samp">arc600</span></samp>&rsquo;, &lsquo;<samp><span class="samp">arc601</span></samp>&rsquo;,
200or &lsquo;<samp><span class="samp">arc700</span></samp>&rsquo;.
201
202   <p><hr /><a name="arc-linux-uclibc"></a>
203
204<h3 class="heading"><a name="TOC5"></a>arc-linux-uclibc</h3>
205
206<p>Use &lsquo;<samp><span class="samp">configure --target=arc-linux-uclibc --with-cpu=arc700 --enable-languages="c,c++"</span></samp>&rsquo; to configure GCC.
207
208   <p><hr /><a name="arm-x-eabi"></a>
209
210<h3 class="heading"><a name="TOC6"></a>arm-*-eabi</h3>
211
212<p>ARM-family processors.
213
214   <p>Building the Ada frontend commonly fails (an infinite loop executing
215<code>xsinfo</code>) if the host compiler is GNAT 4.8.  Host compilers built from the
216GNAT 4.6, 4.9 or 5 release branches are known to succeed.
217
218   <p><hr /><a name="avr"></a>
219
220<h3 class="heading"><a name="TOC7"></a>avr</h3>
221
222<p>ATMEL AVR-family micro controllers.  These are used in embedded
223applications.  There are no standard Unix configurations.
224See &ldquo;AVR Options&rdquo; in the main manual
225for the list of supported MCU types.
226
227   <p>Use &lsquo;<samp><span class="samp">configure --target=avr --enable-languages="c"</span></samp>&rsquo; to configure GCC.
228
229   <p>Further installation notes and other useful information about AVR tools
230can also be obtained from:
231
232     <ul>
233<li><a href="http://www.nongnu.org/avr/">http://www.nongnu.org/avr/</a>
234<li><a href="http://www.amelek.gda.pl/avr/">http://www.amelek.gda.pl/avr/</a>
235</ul>
236
237   <p>The following error:
238<pre class="smallexample">     Error: register required
239</pre>
240   <p>indicates that you should upgrade to a newer version of the binutils.
241
242   <p><hr /><a name="bfin"></a>
243
244<h3 class="heading"><a name="TOC8"></a>Blackfin</h3>
245
246<p>The Blackfin processor, an Analog Devices DSP.
247See &ldquo;Blackfin Options&rdquo; in the main manual
248
249   <p>More information, and a version of binutils with support for this processor,
250is available at <a href="https://blackfin.uclinux.org">https://blackfin.uclinux.org</a>
251
252   <p><hr /><a name="cr16"></a>
253
254<h3 class="heading"><a name="TOC9"></a>CR16</h3>
255
256<p>The CR16 CompactRISC architecture is a 16-bit architecture. This
257architecture is used in embedded applications.
258
259   <p>See &ldquo;CR16 Options&rdquo; in the main manual for a list of CR16-specific options.
260
261   <p>Use &lsquo;<samp><span class="samp">configure --target=cr16-elf --enable-languages=c,c++</span></samp>&rsquo; to configure
262GCC for building a CR16 elf cross-compiler.
263
264   <p>Use &lsquo;<samp><span class="samp">configure --target=cr16-uclinux --enable-languages=c,c++</span></samp>&rsquo; to
265configure GCC for building a CR16 uclinux cross-compiler.
266
267   <p><hr /><a name="cris"></a>
268
269<h3 class="heading"><a name="TOC10"></a>CRIS</h3>
270
271<p>CRIS is the CPU architecture in Axis Communications ETRAX system-on-a-chip
272series.  These are used in embedded applications.
273
274   <p>See &ldquo;CRIS Options&rdquo; in the main manual
275for a list of CRIS-specific options.
276
277   <p>There are a few different CRIS targets:
278     <dl>
279<dt><code>cris-axis-elf</code><dd>Mainly for monolithic embedded systems.  Includes a multilib for the
280&lsquo;<samp><span class="samp">v10</span></samp>&rsquo; core used in &lsquo;<samp><span class="samp">ETRAX 100 LX</span></samp>&rsquo;.
281<br><dt><code>cris-axis-linux-gnu</code><dd>A GNU/Linux port for the CRIS architecture, currently targeting
282&lsquo;<samp><span class="samp">ETRAX 100 LX</span></samp>&rsquo; by default.
283</dl>
284
285   <p>Pre-packaged tools can be obtained from
286<a href="ftp://ftp.axis.com/pub/axis/tools/cris/compiler-kit/">ftp://ftp.axis.com/pub/axis/tools/cris/compiler-kit/</a>.  More
287information about this platform is available at
288<a href="http://developer.axis.com/">http://developer.axis.com/</a>.
289
290   <p><hr /><a name="dos"></a>
291
292<h3 class="heading"><a name="TOC11"></a>DOS</h3>
293
294<p>Please have a look at the <a href="binaries.html">binaries page</a>.
295
296   <p>You cannot install GCC by itself on MSDOS; it will not compile under
297any MSDOS compiler except itself.  You need to get the complete
298compilation package DJGPP, which includes binaries as well as sources,
299and includes all the necessary compilation tools and libraries.
300
301   <p><hr /><a name="epiphany-x-elf"></a>
302
303<h3 class="heading"><a name="TOC12"></a>epiphany-*-elf</h3>
304
305<p>Adapteva Epiphany.
306This configuration is intended for embedded systems.
307
308   <p><hr /><a name="x-x-freebsd"></a>
309
310<h3 class="heading"><a name="TOC13"></a>*-*-freebsd*</h3>
311
312<p>Support for FreeBSD 1 was discontinued in GCC 3.2.  Support for
313FreeBSD 2 (and any mutant a.out variants of FreeBSD 3) was
314discontinued in GCC 4.0.
315
316   <p>In order to better utilize FreeBSD base system functionality and match
317the configuration of the system compiler, GCC 4.5 and above as well as
318GCC 4.4 past 2010-06-20 leverage SSP support in libc (which is present
319on FreeBSD 7 or later) and the use of <code>__cxa_atexit</code> by default
320(on FreeBSD 6 or later).  The use of <code>dl_iterate_phdr</code> inside
321<samp><span class="file">libgcc_s.so.1</span></samp> and boehm-gc (on FreeBSD 7 or later) is enabled
322by GCC 4.5 and above.
323
324   <p>We support FreeBSD using the ELF file format with DWARF 2 debugging
325for all CPU architectures.  You may use <samp><span class="option">-gstabs</span></samp> instead of
326<samp><span class="option">-g</span></samp>, if you really want the old debugging format.  There are
327no known issues with mixing object files and libraries with different
328debugging formats.  Otherwise, this release of GCC should now match
329more of the configuration used in the stock FreeBSD configuration of
330GCC.  In particular, <samp><span class="option">--enable-threads</span></samp> is now configured by
331default.  However, as a general user, do not attempt to replace the
332system compiler with this release.  Known to bootstrap and check with
333good results on FreeBSD 7.2-STABLE.  In the past, known to bootstrap
334and check with good results on FreeBSD 3.0, 3.4, 4.0, 4.2, 4.3, 4.4,
3354.5, 4.8, 4.9 and 5-CURRENT.
336
337   <p>The version of binutils installed in <samp><span class="file">/usr/bin</span></samp> probably works
338with this release of GCC.  Bootstrapping against the latest GNU
339binutils and/or the version found in <samp><span class="file">/usr/ports/devel/binutils</span></samp> has
340been known to enable additional features and improve overall testsuite
341results.  However, it is currently known that boehm-gc may not configure
342properly on FreeBSD prior to the FreeBSD 7.0 release with GNU binutils
343after 2.16.1.
344
345   <p><hr /><a name="ft32-x-elf"></a>
346
347<h3 class="heading"><a name="TOC14"></a>ft32-*-elf</h3>
348
349<p>The FT32 processor.
350This configuration is intended for embedded systems.
351
352   <p><hr /><a name="h8300-hms"></a>
353
354<h3 class="heading"><a name="TOC15"></a>h8300-hms</h3>
355
356<p>Renesas H8/300 series of processors.
357
358   <p>Please have a look at the <a href="binaries.html">binaries page</a>.
359
360   <p>The calling convention and structure layout has changed in release 2.6.
361All code must be recompiled.  The calling convention now passes the
362first three arguments in function calls in registers.  Structures are no
363longer a multiple of 2 bytes.
364
365   <p><hr /><a name="hppa-hp-hpux"></a>
366
367<h3 class="heading"><a name="TOC16"></a>hppa*-hp-hpux*</h3>
368
369<p>Support for HP-UX version 9 and older was discontinued in GCC 3.4.
370
371   <p>We require using gas/binutils on all hppa platforms.  Version 2.19 or
372later is recommended.
373
374   <p>It may be helpful to configure GCC with the
375<a href="./configure.html#with-gnu-as"><samp><span class="option">--with-gnu-as</span></samp></a> and
376<samp><span class="option">--with-as=...</span></samp> options to ensure that GCC can find GAS.
377
378   <p>The HP assembler should not be used with GCC.  It is rarely tested and may
379not work.  It shouldn't be used with any languages other than C due to its
380many limitations.
381
382   <p>Specifically, <samp><span class="option">-g</span></samp> does not work (HP-UX uses a peculiar debugging
383format which GCC does not know about).  It also inserts timestamps
384into each object file it creates, causing the 3-stage comparison test to
385fail during a bootstrap.  You should be able to continue by saying
386&lsquo;<samp><span class="samp">make all-host all-target</span></samp>&rsquo; after getting the failure from &lsquo;<samp><span class="samp">make</span></samp>&rsquo;.
387
388   <p>Various GCC features are not supported.  For example, it does not support weak
389symbols or alias definitions.  As a result, explicit template instantiations
390are required when using C++.  This makes it difficult if not impossible to
391build many C++ applications.
392
393   <p>There are two default scheduling models for instructions.  These are
394PROCESSOR_7100LC and PROCESSOR_8000.  They are selected from the pa-risc
395architecture specified for the target machine when configuring.
396PROCESSOR_8000 is the default.  PROCESSOR_7100LC is selected when
397the target is a &lsquo;<samp><span class="samp">hppa1*</span></samp>&rsquo; machine.
398
399   <p>The PROCESSOR_8000 model is not well suited to older processors.  Thus,
400it is important to completely specify the machine architecture when
401configuring if you want a model other than PROCESSOR_8000.  The macro
402TARGET_SCHED_DEFAULT can be defined in BOOT_CFLAGS if a different
403default scheduling model is desired.
404
405   <p>As of GCC 4.0, GCC uses the UNIX 95 namespace for HP-UX 10.10
406through 11.00, and the UNIX 98 namespace for HP-UX 11.11 and later.
407This namespace change might cause problems when bootstrapping with
408an earlier version of GCC or the HP compiler as essentially the same
409namespace is required for an entire build.  This problem can be avoided
410in a number of ways.  With HP cc, <samp><span class="env">UNIX_STD</span></samp> can be set to &lsquo;<samp><span class="samp">95</span></samp>&rsquo;
411or &lsquo;<samp><span class="samp">98</span></samp>&rsquo;.  Another way is to add an appropriate set of predefines
412to <samp><span class="env">CC</span></samp>.  The description for the <samp><span class="option">munix=</span></samp> option contains
413a list of the predefines used with each standard.
414
415   <p>More specific information to &lsquo;<samp><span class="samp">hppa*-hp-hpux*</span></samp>&rsquo; targets follows.
416
417   <p><hr /><a name="hppa-hp-hpux10"></a>
418
419<h3 class="heading"><a name="TOC17"></a>hppa*-hp-hpux10</h3>
420
421<p>For hpux10.20, we <em>highly</em> recommend you pick up the latest sed patch
422<code>PHCO_19798</code> from HP.
423
424   <p>The C++ ABI has changed incompatibly in GCC 4.0.  COMDAT subspaces are
425used for one-only code and data.  This resolves many of the previous
426problems in using C++ on this target.  However, the ABI is not compatible
427with the one implemented under HP-UX 11 using secondary definitions.
428
429   <p><hr /><a name="hppa-hp-hpux11"></a>
430
431<h3 class="heading"><a name="TOC18"></a>hppa*-hp-hpux11</h3>
432
433<p>GCC 3.0 and up support HP-UX 11.  GCC 2.95.x is not supported and cannot
434be used to compile GCC 3.0 and up.
435
436   <p>The libffi library haven't been ported to 64-bit HP-UX and doesn't build.
437
438   <p>Refer to <a href="binaries.html">binaries</a> for information about obtaining
439precompiled GCC binaries for HP-UX.  Precompiled binaries must be obtained
440to build the Ada language as it cannot be bootstrapped using C.  Ada is
441only available for the 32-bit PA-RISC runtime.
442
443   <p>Starting with GCC 3.4 an ISO C compiler is required to bootstrap.  The
444bundled compiler supports only traditional C; you will need either HP's
445unbundled compiler, or a binary distribution of GCC.
446
447   <p>It is possible to build GCC 3.3 starting with the bundled HP compiler,
448but the process requires several steps.  GCC 3.3 can then be used to
449build later versions.
450
451   <p>There are several possible approaches to building the distribution.
452Binutils can be built first using the HP tools.  Then, the GCC
453distribution can be built.  The second approach is to build GCC
454first using the HP tools, then build binutils, then rebuild GCC.
455There have been problems with various binary distributions, so it
456is best not to start from a binary distribution.
457
458   <p>On 64-bit capable systems, there are two distinct targets.  Different
459installation prefixes must be used if both are to be installed on
460the same system.  The &lsquo;<samp><span class="samp">hppa[1-2]*-hp-hpux11*</span></samp>&rsquo; target generates code
461for the 32-bit PA-RISC runtime architecture and uses the HP linker.
462The &lsquo;<samp><span class="samp">hppa64-hp-hpux11*</span></samp>&rsquo; target generates 64-bit code for the
463PA-RISC 2.0 architecture.
464
465   <p>The script config.guess now selects the target type based on the compiler
466detected during configuration.  You must define <samp><span class="env">PATH</span></samp> or <samp><span class="env">CC</span></samp> so
467that configure finds an appropriate compiler for the initial bootstrap.
468When <samp><span class="env">CC</span></samp> is used, the definition should contain the options that are
469needed whenever <samp><span class="env">CC</span></samp> is used.
470
471   <p>Specifically, options that determine the runtime architecture must be
472in <samp><span class="env">CC</span></samp> to correctly select the target for the build.  It is also
473convenient to place many other compiler options in <samp><span class="env">CC</span></samp>.  For example,
474<samp><span class="env">CC="cc -Ac +DA2.0W -Wp,-H16376 -D_CLASSIC_TYPES -D_HPUX_SOURCE"</span></samp>
475can be used to bootstrap the GCC 3.3 branch with the HP compiler in
47664-bit K&amp;R/bundled mode.  The <samp><span class="option">+DA2.0W</span></samp> option will result in
477the automatic selection of the &lsquo;<samp><span class="samp">hppa64-hp-hpux11*</span></samp>&rsquo; target.  The
478macro definition table of cpp needs to be increased for a successful
479build with the HP compiler.  _CLASSIC_TYPES and _HPUX_SOURCE need to
480be defined when building with the bundled compiler, or when using the
481<samp><span class="option">-Ac</span></samp> option.  These defines aren't necessary with <samp><span class="option">-Ae</span></samp>.
482
483   <p>It is best to explicitly configure the &lsquo;<samp><span class="samp">hppa64-hp-hpux11*</span></samp>&rsquo; target
484with the <samp><span class="option">--with-ld=...</span></samp> option.  This overrides the standard
485search for ld.  The two linkers supported on this target require different
486commands.  The default linker is determined during configuration.  As a
487result, it's not possible to switch linkers in the middle of a GCC build.
488This has been reported to sometimes occur in unified builds of binutils
489and GCC.
490
491   <p>A recent linker patch must be installed for the correct operation of
492GCC 3.3 and later.  <code>PHSS_26559</code> and <code>PHSS_24304</code> are the
493oldest linker patches that are known to work.  They are for HP-UX
49411.00 and 11.11, respectively.  <code>PHSS_24303</code>, the companion to
495<code>PHSS_24304</code>, might be usable but it hasn't been tested.  These
496patches have been superseded.  Consult the HP patch database to obtain
497the currently recommended linker patch for your system.
498
499   <p>The patches are necessary for the support of weak symbols on the
50032-bit port, and for the running of initializers and finalizers.  Weak
501symbols are implemented using SOM secondary definition symbols.  Prior
502to HP-UX 11, there are bugs in the linker support for secondary symbols.
503The patches correct a problem of linker core dumps creating shared
504libraries containing secondary symbols, as well as various other
505linking issues involving secondary symbols.
506
507   <p>GCC 3.3 uses the ELF DT_INIT_ARRAY and DT_FINI_ARRAY capabilities to
508run initializers and finalizers on the 64-bit port.  The 32-bit port
509uses the linker <samp><span class="option">+init</span></samp> and <samp><span class="option">+fini</span></samp> options for the same
510purpose.  The patches correct various problems with the +init/+fini
511options, including program core dumps.  Binutils 2.14 corrects a
512problem on the 64-bit port resulting from HP's non-standard use of
513the .init and .fini sections for array initializers and finalizers.
514
515   <p>Although the HP and GNU linkers are both supported for the
516&lsquo;<samp><span class="samp">hppa64-hp-hpux11*</span></samp>&rsquo; target, it is strongly recommended that the
517HP linker be used for link editing on this target.
518
519   <p>At this time, the GNU linker does not support the creation of long
520branch stubs.  As a result, it cannot successfully link binaries
521containing branch offsets larger than 8 megabytes.  In addition,
522there are problems linking shared libraries, linking executables
523with <samp><span class="option">-static</span></samp>, and with dwarf2 unwind and exception support.
524It also doesn't provide stubs for internal calls to global functions
525in shared libraries, so these calls cannot be overloaded.
526
527   <p>The HP dynamic loader does not support GNU symbol versioning, so symbol
528versioning is not supported.  It may be necessary to disable symbol
529versioning with <samp><span class="option">--disable-symvers</span></samp> when using GNU ld.
530
531   <p>POSIX threads are the default.  The optional DCE thread library is not
532supported, so <samp><span class="option">--enable-threads=dce</span></samp> does not work.
533
534   <p><hr /><a name="x-x-linux-gnu"></a>
535
536<h3 class="heading"><a name="TOC19"></a>*-*-linux-gnu</h3>
537
538<p>Versions of libstdc++-v3 starting with 3.2.1 require bug fixes present
539in glibc 2.2.5 and later.  More information is available in the
540libstdc++-v3 documentation.
541
542   <p><hr /><a name="ix86-x-linux"></a>
543
544<h3 class="heading"><a name="TOC20"></a>i?86-*-linux*</h3>
545
546<p>As of GCC 3.3, binutils 2.13.1 or later is required for this platform.
547See <a href="http://gcc.gnu.org/PR10877">bug 10877</a> for more information.
548
549   <p>If you receive Signal 11 errors when building on GNU/Linux, then it is
550possible you have a hardware problem.  Further information on this can be
551found on <a href="http://www.bitwizard.nl/sig11/">www.bitwizard.nl</a>.
552
553   <p><hr /><a name="ix86-x-solaris210"></a>
554
555<h3 class="heading"><a name="TOC21"></a>i?86-*-solaris2.10</h3>
556
557<p>Use this for Solaris 10 or later on x86 and x86-64 systems.  Starting
558with GCC 4.7, there is also a 64-bit &lsquo;<samp><span class="samp">amd64-*-solaris2.1[0-9]*</span></samp>&rsquo; or
559&lsquo;<samp><span class="samp">x86_64-*-solaris2.1[0-9]*</span></samp>&rsquo; configuration that corresponds to
560&lsquo;<samp><span class="samp">sparcv9-sun-solaris2*</span></samp>&rsquo;.
561
562   <p>It is recommended that you configure GCC to use the GNU assembler.  The
563versions included in Solaris 10, from GNU binutils 2.15 (in
564<samp><span class="file">/usr/sfw/bin/gas</span></samp>), and Solaris 11, from GNU binutils 2.19 or
565newer (also available as <samp><span class="file">/usr/bin/gas</span></samp> and
566<samp><span class="file">/usr/gnu/bin/as</span></samp>), work fine.  Please note that the current
567version, from GNU binutils 2.26, only works on Solaris 12 when using the
568Solaris linker.  On Solaris 10 and 11, you either have to wait for GNU
569binutils 2.26.1 or newer, or stay with GNU binutils 2.25.1.  Recent
570versions of the Solaris assembler in <samp><span class="file">/usr/ccs/bin/as</span></samp> work almost
571as well, though.
572<!-- FIXME: as patch requirements? -->
573
574   <p>For linking, the Solaris linker, is preferred.  If you want to use the GNU
575linker instead, note that due to a packaging bug the version in Solaris
57610, from GNU binutils 2.15 (in <samp><span class="file">/usr/sfw/bin/gld</span></samp>), cannot be used,
577while the version in Solaris 11, from GNU binutils 2.19 or newer (also
578in <samp><span class="file">/usr/gnu/bin/ld</span></samp> and <samp><span class="file">/usr/bin/gld</span></samp>), works, as does the
579latest version, from GNU binutils 2.26.
580
581   <p>To use GNU <samp><span class="command">as</span></samp>, configure with the options
582<samp><span class="option">--with-gnu-as --with-as=/usr/sfw/bin/gas</span></samp>.  It may be necessary
583to configure with <samp><span class="option">--without-gnu-ld --with-ld=/usr/ccs/bin/ld</span></samp> to
584guarantee use of Sun <samp><span class="command">ld</span></samp>.
585<!-- FIXME: why -without-gnu-ld -with-ld? -->
586
587   <p><hr /><a name="ia64-x-linux"></a>
588
589<h3 class="heading"><a name="TOC22"></a>ia64-*-linux</h3>
590
591<p>IA-64 processor (also known as IPF, or Itanium Processor Family)
592running GNU/Linux.
593
594   <p>If you are using the installed system libunwind library with
595<samp><span class="option">--with-system-libunwind</span></samp>, then you must use libunwind 0.98 or
596later.
597
598   <p>None of the following versions of GCC has an ABI that is compatible
599with any of the other versions in this list, with the exception that
600Red Hat 2.96 and Trillian 000171 are compatible with each other:
6013.1, 3.0.2, 3.0.1, 3.0, Red Hat 2.96, and Trillian 000717.
602This primarily affects C++ programs and programs that create shared libraries.
603GCC 3.1 or later is recommended for compiling linux, the kernel.
604As of version 3.1 GCC is believed to be fully ABI compliant, and hence no
605more major ABI changes are expected.
606
607   <p><hr /><a name="ia64-x-hpux"></a>
608
609<h3 class="heading"><a name="TOC23"></a>ia64-*-hpux*</h3>
610
611<p>Building GCC on this target requires the GNU Assembler.  The bundled HP
612assembler will not work.  To prevent GCC from using the wrong assembler,
613the option <samp><span class="option">--with-gnu-as</span></samp> may be necessary.
614
615   <p>The GCC libunwind library has not been ported to HPUX.  This means that for
616GCC versions 3.2.3 and earlier, <samp><span class="option">--enable-libunwind-exceptions</span></samp>
617is required to build GCC.  For GCC 3.3 and later, this is the default.
618For gcc 3.4.3 and later, <samp><span class="option">--enable-libunwind-exceptions</span></samp> is
619removed and the system libunwind library will always be used.
620
621   <p><hr />
622<!-- rs6000-ibm-aix*, powerpc-ibm-aix* --><a name="x-ibm-aix"></a>
623
624<h3 class="heading"><a name="TOC24"></a>*-ibm-aix*</h3>
625
626<p>Support for AIX version 3 and older was discontinued in GCC 3.4.
627Support for AIX version 4.2 and older was discontinued in GCC 4.5.
628
629   <p>&ldquo;out of memory&rdquo; bootstrap failures may indicate a problem with
630process resource limits (ulimit).  Hard limits are configured in the
631<samp><span class="file">/etc/security/limits</span></samp> system configuration file.
632
633   <p>GCC 4.9 and above require a C++ compiler for bootstrap.  IBM VAC++ / xlC
634cannot bootstrap GCC.  xlc can bootstrap an older version of GCC and
635G++ can bootstrap recent releases of GCC.
636
637   <p>GCC can bootstrap with recent versions of IBM XLC, but bootstrapping
638with an earlier release of GCC is recommended.  Bootstrapping with XLC
639requires a larger data segment, which can be enabled through the
640<var>LDR_CNTRL</var> environment variable, e.g.,
641
642<pre class="smallexample">     % LDR_CNTRL=MAXDATA=0x50000000
643     % export LDR_CNTRL
644</pre>
645   <p>One can start with a pre-compiled version of GCC to build from
646sources.  One may delete GCC's &ldquo;fixed&rdquo; header files when starting
647with a version of GCC built for an earlier release of AIX.
648
649   <p>To speed up the configuration phases of bootstrapping and installing GCC,
650one may use GNU Bash instead of AIX <samp><span class="command">/bin/sh</span></samp>, e.g.,
651
652<pre class="smallexample">     % CONFIG_SHELL=/opt/freeware/bin/bash
653     % export CONFIG_SHELL
654</pre>
655   <p>and then proceed as described in <a href="build.html">the build instructions</a>, where we strongly recommend specifying an absolute path
656to invoke <var>srcdir</var>/configure.
657
658   <p>Because GCC on AIX is built as a 32-bit executable by default,
659(although it can generate 64-bit programs) the GMP and MPFR libraries
660required by gfortran must be 32-bit libraries.  Building GMP and MPFR
661as static archive libraries works better than shared libraries.
662
663   <p>Errors involving <code>alloca</code> when building GCC generally are due
664to an incorrect definition of <code>CC</code> in the Makefile or mixing files
665compiled with the native C compiler and GCC.  During the stage1 phase of
666the build, the native AIX compiler <strong>must</strong> be invoked as <samp><span class="command">cc</span></samp>
667(not <samp><span class="command">xlc</span></samp>).  Once <samp><span class="command">configure</span></samp> has been informed of
668<samp><span class="command">xlc</span></samp>, one needs to use &lsquo;<samp><span class="samp">make distclean</span></samp>&rsquo; to remove the
669configure cache files and ensure that <samp><span class="env">CC</span></samp> environment variable
670does not provide a definition that will confuse <samp><span class="command">configure</span></samp>.
671If this error occurs during stage2 or later, then the problem most likely
672is the version of Make (see above).
673
674   <p>The native <samp><span class="command">as</span></samp> and <samp><span class="command">ld</span></samp> are recommended for
675bootstrapping on AIX.  The GNU Assembler, GNU Linker, and GNU
676Binutils version 2.20 is the minimum level that supports bootstrap on
677AIX 5.  The GNU Assembler has not been updated to support AIX 6 or
678AIX 7.  The native AIX tools do interoperate with GCC.
679
680   <p>AIX 7.1 added partial support for DWARF debugging, but full support
681requires AIX 7.1 TL03 SP7 that supports additional DWARF sections and
682fixes a bug in the assembler.  AIX 7.1 TL03 SP5 distributed a version
683of libm.a missing important symbols; a fix for IV77796 will be
684included in SP6.
685
686   <p>AIX 5.3 TL10, AIX 6.1 TL05 and AIX 7.1 TL00 introduced an AIX
687assembler change that sometimes produces corrupt assembly files
688causing AIX linker errors.  The bug breaks GCC bootstrap on AIX and
689can cause compilation failures with existing GCC installations.  An
690AIX iFix for AIX 5.3 is available (APAR IZ98385 for AIX 5.3 TL10, APAR
691IZ98477 for AIX 5.3 TL11 and IZ98134 for AIX 5.3 TL12). AIX 5.3 TL11 SP8,
692AIX 5.3 TL12 SP5, AIX 6.1 TL04 SP11, AIX 6.1 TL05 SP7, AIX 6.1 TL06 SP6,
693AIX 6.1 TL07 and AIX 7.1 TL01 should include the fix.
694
695   <p>Building <samp><span class="file">libstdc++.a</span></samp> requires a fix for an AIX Assembler bug
696APAR IY26685 (AIX 4.3) or APAR IY25528 (AIX 5.1).  It also requires a
697fix for another AIX Assembler bug and a co-dependent AIX Archiver fix
698referenced as APAR IY53606 (AIX 5.2) or as APAR IY54774 (AIX 5.1)
699
700   <p><a name="TransferAixShobj"></a>&lsquo;<samp><span class="samp">libstdc++</span></samp>&rsquo; in GCC 3.4 increments the major version number of the
701shared object and GCC installation places the <samp><span class="file">libstdc++.a</span></samp>
702shared library in a common location which will overwrite the and GCC
7033.3 version of the shared library.  Applications either need to be
704re-linked against the new shared library or the GCC 3.1 and GCC 3.3
705versions of the &lsquo;<samp><span class="samp">libstdc++</span></samp>&rsquo; shared object needs to be available
706to the AIX runtime loader.  The GCC 3.1 &lsquo;<samp><span class="samp">libstdc++.so.4</span></samp>&rsquo;, if
707present, and GCC 3.3 &lsquo;<samp><span class="samp">libstdc++.so.5</span></samp>&rsquo; shared objects can be
708installed for runtime dynamic loading using the following steps to set
709the &lsquo;<samp><span class="samp">F_LOADONLY</span></samp>&rsquo; flag in the shared object for <em>each</em>
710multilib <samp><span class="file">libstdc++.a</span></samp> installed:
711
712   <p>Extract the shared objects from the currently installed
713<samp><span class="file">libstdc++.a</span></samp> archive:
714<pre class="smallexample">     % ar -x libstdc++.a libstdc++.so.4 libstdc++.so.5
715</pre>
716   <p>Enable the &lsquo;<samp><span class="samp">F_LOADONLY</span></samp>&rsquo; flag so that the shared object will be
717available for runtime dynamic loading, but not linking:
718<pre class="smallexample">     % strip -e libstdc++.so.4 libstdc++.so.5
719</pre>
720   <p>Archive the runtime-only shared object in the GCC 3.4
721<samp><span class="file">libstdc++.a</span></samp> archive:
722<pre class="smallexample">     % ar -q libstdc++.a libstdc++.so.4 libstdc++.so.5
723</pre>
724   <p>Eventually, the
725<a href="./configure.html#WithAixSoname"><samp><span class="option">--with-aix-soname=svr4</span></samp></a>
726configure option may drop the need for this procedure for libraries that
727support it.
728
729   <p>Linking executables and shared libraries may produce warnings of
730duplicate symbols.  The assembly files generated by GCC for AIX always
731have included multiple symbol definitions for certain global variable
732and function declarations in the original program.  The warnings should
733not prevent the linker from producing a correct library or runnable
734executable.
735
736   <p>AIX 4.3 utilizes a &ldquo;large format&rdquo; archive to support both 32-bit and
73764-bit object modules.  The routines provided in AIX 4.3.0 and AIX 4.3.1
738to parse archive libraries did not handle the new format correctly.
739These routines are used by GCC and result in error messages during
740linking such as &ldquo;not a COFF file&rdquo;.  The version of the routines shipped
741with AIX 4.3.1 should work for a 32-bit environment.  The <samp><span class="option">-g</span></samp>
742option of the archive command may be used to create archives of 32-bit
743objects using the original &ldquo;small format&rdquo;.  A correct version of the
744routines is shipped with AIX 4.3.2 and above.
745
746   <p>Some versions of the AIX binder (linker) can fail with a relocation
747overflow severe error when the <samp><span class="option">-bbigtoc</span></samp> option is used to link
748GCC-produced object files into an executable that overflows the TOC.  A fix
749for APAR IX75823 (OVERFLOW DURING LINK WHEN USING GCC AND -BBIGTOC) is
750available from IBM Customer Support and from its
751<a href="http://techsupport.services.ibm.com/">techsupport.services.ibm.com</a>
752website as PTF U455193.
753
754   <p>The AIX 4.3.2.1 linker (bos.rte.bind_cmds Level 4.3.2.1) will dump core
755with a segmentation fault when invoked by any version of GCC.  A fix for
756APAR IX87327 is available from IBM Customer Support and from its
757<a href="http://techsupport.services.ibm.com/">techsupport.services.ibm.com</a>
758website as PTF U461879.  This fix is incorporated in AIX 4.3.3 and above.
759
760   <p>The initial assembler shipped with AIX 4.3.0 generates incorrect object
761files.  A fix for APAR IX74254 (64BIT DISASSEMBLED OUTPUT FROM COMPILER FAILS
762TO ASSEMBLE/BIND) is available from IBM Customer Support and from its
763<a href="http://techsupport.services.ibm.com/">techsupport.services.ibm.com</a>
764website as PTF U453956.  This fix is incorporated in AIX 4.3.1 and above.
765
766   <p>AIX provides National Language Support (NLS).  Compilers and assemblers
767use NLS to support locale-specific representations of various data
768formats including floating-point numbers (e.g., &lsquo;<samp><span class="samp">.</span></samp>&rsquo;  vs &lsquo;<samp><span class="samp">,</span></samp>&rsquo; for
769separating decimal fractions).  There have been problems reported where
770GCC does not produce the same floating-point formats that the assembler
771expects.  If one encounters this problem, set the <samp><span class="env">LANG</span></samp>
772environment variable to &lsquo;<samp><span class="samp">C</span></samp>&rsquo; or &lsquo;<samp><span class="samp">En_US</span></samp>&rsquo;.
773
774   <p>A default can be specified with the <samp><span class="option">-mcpu=</span><var>cpu_type</var></samp>
775switch and using the configure option <samp><span class="option">--with-cpu-</span><var>cpu_type</var></samp>.
776
777   <p><hr /><a name="iq2000-x-elf"></a>
778
779<h3 class="heading"><a name="TOC25"></a>iq2000-*-elf</h3>
780
781<p>Vitesse IQ2000 processors.  These are used in embedded
782applications.  There are no standard Unix configurations.
783
784   <p><hr /><a name="lm32-x-elf"></a>
785
786<h3 class="heading"><a name="TOC26"></a>lm32-*-elf</h3>
787
788<p>Lattice Mico32 processor.
789This configuration is intended for embedded systems.
790
791   <p><hr /><a name="lm32-x-uclinux"></a>
792
793<h3 class="heading"><a name="TOC27"></a>lm32-*-uclinux</h3>
794
795<p>Lattice Mico32 processor.
796This configuration is intended for embedded systems running uClinux.
797
798   <p><hr /><a name="m32c-x-elf"></a>
799
800<h3 class="heading"><a name="TOC28"></a>m32c-*-elf</h3>
801
802<p>Renesas M32C processor.
803This configuration is intended for embedded systems.
804
805   <p><hr /><a name="m32r-x-elf"></a>
806
807<h3 class="heading"><a name="TOC29"></a>m32r-*-elf</h3>
808
809<p>Renesas M32R processor.
810This configuration is intended for embedded systems.
811
812   <p><hr /><a name="m68k-x-x"></a>
813
814<h3 class="heading"><a name="TOC30"></a>m68k-*-*</h3>
815
816<p>By default,
817&lsquo;<samp><span class="samp">m68k-*-elf*</span></samp>&rsquo;, &lsquo;<samp><span class="samp">m68k-*-rtems</span></samp>&rsquo;,  &lsquo;<samp><span class="samp">m68k-*-uclinux</span></samp>&rsquo; and
818&lsquo;<samp><span class="samp">m68k-*-linux</span></samp>&rsquo;
819build libraries for both M680x0 and ColdFire processors.  If you only
820need the M680x0 libraries, you can omit the ColdFire ones by passing
821<samp><span class="option">--with-arch=m68k</span></samp> to <samp><span class="command">configure</span></samp>.  Alternatively, you
822can omit the M680x0 libraries by passing <samp><span class="option">--with-arch=cf</span></samp> to
823<samp><span class="command">configure</span></samp>.  These targets default to 5206 or 5475 code as
824appropriate for the target system when
825configured with <samp><span class="option">--with-arch=cf</span></samp> and 68020 code otherwise.
826
827   <p>The &lsquo;<samp><span class="samp">m68k-*-netbsd</span></samp>&rsquo; and
828&lsquo;<samp><span class="samp">m68k-*-openbsd</span></samp>&rsquo; targets also support the <samp><span class="option">--with-arch</span></samp>
829option.  They will generate ColdFire CFV4e code when configured with
830<samp><span class="option">--with-arch=cf</span></samp> and 68020 code otherwise.
831
832   <p>You can override the default processors listed above by configuring
833with <samp><span class="option">--with-cpu=</span><var>target</var></samp>.  This <var>target</var> can either
834be a <samp><span class="option">-mcpu</span></samp> argument or one of the following values:
835&lsquo;<samp><span class="samp">m68000</span></samp>&rsquo;, &lsquo;<samp><span class="samp">m68010</span></samp>&rsquo;, &lsquo;<samp><span class="samp">m68020</span></samp>&rsquo;, &lsquo;<samp><span class="samp">m68030</span></samp>&rsquo;,
836&lsquo;<samp><span class="samp">m68040</span></samp>&rsquo;, &lsquo;<samp><span class="samp">m68060</span></samp>&rsquo;, &lsquo;<samp><span class="samp">m68020-40</span></samp>&rsquo; and &lsquo;<samp><span class="samp">m68020-60</span></samp>&rsquo;.
837
838   <p>GCC requires at least binutils version 2.17 on these targets.
839
840   <p><hr /><a name="m68k-x-uclinux"></a>
841
842<h3 class="heading"><a name="TOC31"></a>m68k-*-uclinux</h3>
843
844<p>GCC 4.3 changed the uClinux configuration so that it uses the
845&lsquo;<samp><span class="samp">m68k-linux-gnu</span></samp>&rsquo; ABI rather than the &lsquo;<samp><span class="samp">m68k-elf</span></samp>&rsquo; ABI.
846It also added improved support for C++ and flat shared libraries,
847both of which were ABI changes.
848
849   <p><hr /><a name="microblaze-x-elf"></a>
850
851<h3 class="heading"><a name="TOC32"></a>microblaze-*-elf</h3>
852
853<p>Xilinx MicroBlaze processor.
854This configuration is intended for embedded systems.
855
856   <p><hr /><a name="mips-x-x"></a>
857
858<h3 class="heading"><a name="TOC33"></a>mips-*-*</h3>
859
860<p>If on a MIPS system you get an error message saying &ldquo;does not have gp
861sections for all it's [sic] sectons [sic]&rdquo;, don't worry about it.  This
862happens whenever you use GAS with the MIPS linker, but there is not
863really anything wrong, and it is okay to use the output file.  You can
864stop such warnings by installing the GNU linker.
865
866   <p>It would be nice to extend GAS to produce the gp tables, but they are
867optional, and there should not be a warning about their absence.
868
869   <p>The libstdc++ atomic locking routines for MIPS targets requires MIPS II
870and later.  A patch went in just after the GCC 3.3 release to
871make &lsquo;<samp><span class="samp">mips*-*-*</span></samp>&rsquo; use the generic implementation instead.  You can also
872configure for &lsquo;<samp><span class="samp">mipsel-elf</span></samp>&rsquo; as a workaround.  The
873&lsquo;<samp><span class="samp">mips*-*-linux*</span></samp>&rsquo; target continues to use the MIPS II routines.  More
874work on this is expected in future releases.
875
876<!-- If you make -with-llsc the default for another target, please also -->
877<!-- update the description of the -with-llsc option. -->
878   <p>The built-in <code>__sync_*</code> functions are available on MIPS II and
879later systems and others that support the &lsquo;<samp><span class="samp">ll</span></samp>&rsquo;, &lsquo;<samp><span class="samp">sc</span></samp>&rsquo; and
880&lsquo;<samp><span class="samp">sync</span></samp>&rsquo; instructions.  This can be overridden by passing
881<samp><span class="option">--with-llsc</span></samp> or <samp><span class="option">--without-llsc</span></samp> when configuring GCC.
882Since the Linux kernel emulates these instructions if they are
883missing, the default for &lsquo;<samp><span class="samp">mips*-*-linux*</span></samp>&rsquo; targets is
884<samp><span class="option">--with-llsc</span></samp>.  The <samp><span class="option">--with-llsc</span></samp> and
885<samp><span class="option">--without-llsc</span></samp> configure options may be overridden at compile
886time by passing the <samp><span class="option">-mllsc</span></samp> or <samp><span class="option">-mno-llsc</span></samp> options to
887the compiler.
888
889   <p>MIPS systems check for division by zero (unless
890<samp><span class="option">-mno-check-zero-division</span></samp> is passed to the compiler) by
891generating either a conditional trap or a break instruction.  Using
892trap results in smaller code, but is only supported on MIPS II and
893later.  Also, some versions of the Linux kernel have a bug that
894prevents trap from generating the proper signal (<code>SIGFPE</code>).  To enable
895the use of break, use the <samp><span class="option">--with-divide=breaks</span></samp>
896<samp><span class="command">configure</span></samp> option when configuring GCC.  The default is to
897use traps on systems that support them.
898
899   <p><hr /><a name="mips-sgi-irix5"></a>
900
901<h3 class="heading"><a name="TOC34"></a>mips-sgi-irix5</h3>
902
903<p>Support for IRIX 5 has been removed in GCC 4.6.
904
905   <p><hr /><a name="mips-sgi-irix6"></a>
906
907<h3 class="heading"><a name="TOC35"></a>mips-sgi-irix6</h3>
908
909<p>Support for IRIX 6.5 has been removed in GCC 4.8.  Support for IRIX 6
910releases before 6.5 has been removed in GCC 4.6, as well as support for
911the O32 ABI.
912
913   <p><hr /><a name="moxie-x-elf"></a>
914
915<h3 class="heading"><a name="TOC36"></a>moxie-*-elf</h3>
916
917<p>The moxie processor.
918
919   <p><hr /><a name="msp430-x-elf"></a>
920
921<h3 class="heading"><a name="TOC37"></a>msp430-*-elf</h3>
922
923<p>TI MSP430 processor.
924This configuration is intended for embedded systems.
925
926   <p><hr /><a name="nds32le-x-elf"></a>
927
928<h3 class="heading"><a name="TOC38"></a>nds32le-*-elf</h3>
929
930<p>Andes NDS32 target in little endian mode.
931
932   <p><hr /><a name="nds32be-x-elf"></a>
933
934<h3 class="heading"><a name="TOC39"></a>nds32be-*-elf</h3>
935
936<p>Andes NDS32 target in big endian mode.
937
938   <p><hr /><a name="nvptx-x-none"></a>
939
940<h3 class="heading"><a name="TOC40"></a>nvptx-*-none</h3>
941
942<p>Nvidia PTX target.
943
944   <p>Instead of GNU binutils, you will need to install
945<a href="https://github.com/MentorEmbedded/nvptx-tools/">nvptx-tools</a>.
946Tell GCC where to find it:
947<samp><span class="option">--with-build-time-tools=[install-nvptx-tools]/nvptx-none/bin</span></samp>.
948
949   <p>A nvptx port of newlib is available at
950<a href="https://github.com/MentorEmbedded/nvptx-newlib/">nvptx-newlib</a>.
951It can be automatically built together with GCC.  For this, add a
952symbolic link to nvptx-newlib's <samp><span class="file">newlib</span></samp> directory to the
953directory containing the GCC sources.
954
955   <p>Use the <samp><span class="option">--disable-sjlj-exceptions</span></samp> and
956<samp><span class="option">--enable-newlib-io-long-long</span></samp> options when configuring.
957
958   <p><hr /><a name="powerpc-x-x"></a>
959
960<h3 class="heading"><a name="TOC41"></a>powerpc-*-*</h3>
961
962<p>You can specify a default version for the <samp><span class="option">-mcpu=</span><var>cpu_type</var></samp>
963switch by using the configure option <samp><span class="option">--with-cpu-</span><var>cpu_type</var></samp>.
964
965   <p>You will need GNU binutils 2.15 or newer.
966
967   <p><hr /><a name="powerpc-x-darwin"></a>
968
969<h3 class="heading"><a name="TOC42"></a>powerpc-*-darwin*</h3>
970
971<p>PowerPC running Darwin (Mac OS X kernel).
972
973   <p>Pre-installed versions of Mac OS X may not include any developer tools,
974meaning that you will not be able to build GCC from source.  Tool
975binaries are available at
976<a href="https://opensource.apple.com">https://opensource.apple.com</a>.
977
978   <p>This version of GCC requires at least cctools-590.36.  The
979cctools-590.36 package referenced from
980<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
981on systems older than 10.3.9 (aka darwin7.9.0).
982
983   <p><hr /><a name="powerpc-x-elf"></a>
984
985<h3 class="heading"><a name="TOC43"></a>powerpc-*-elf</h3>
986
987<p>PowerPC system in big endian mode, running System V.4.
988
989   <p><hr /><a name="powerpc-x-linux-gnu"></a>
990
991<h3 class="heading"><a name="TOC44"></a>powerpc*-*-linux-gnu*</h3>
992
993<p>PowerPC system in big endian mode running Linux.
994
995   <p><hr /><a name="powerpc-x-netbsd"></a>
996
997<h3 class="heading"><a name="TOC45"></a>powerpc-*-netbsd*</h3>
998
999<p>PowerPC system in big endian mode running NetBSD.
1000
1001   <p><hr /><a name="powerpc-x-eabisim"></a>
1002
1003<h3 class="heading"><a name="TOC46"></a>powerpc-*-eabisim</h3>
1004
1005<p>Embedded PowerPC system in big endian mode for use in running under the
1006PSIM simulator.
1007
1008   <p><hr /><a name="powerpc-x-eabi"></a>
1009
1010<h3 class="heading"><a name="TOC47"></a>powerpc-*-eabi</h3>
1011
1012<p>Embedded PowerPC system in big endian mode.
1013
1014   <p><hr /><a name="powerpcle-x-elf"></a>
1015
1016<h3 class="heading"><a name="TOC48"></a>powerpcle-*-elf</h3>
1017
1018<p>PowerPC system in little endian mode, running System V.4.
1019
1020   <p><hr /><a name="powerpcle-x-eabisim"></a>
1021
1022<h3 class="heading"><a name="TOC49"></a>powerpcle-*-eabisim</h3>
1023
1024<p>Embedded PowerPC system in little endian mode for use in running under
1025the PSIM simulator.
1026
1027   <p><hr /><a name="powerpcle-x-eabi"></a>
1028
1029<h3 class="heading"><a name="TOC50"></a>powerpcle-*-eabi</h3>
1030
1031<p>Embedded PowerPC system in little endian mode.
1032
1033   <p><hr /><a name="rl78-x-elf"></a>
1034
1035<h3 class="heading"><a name="TOC51"></a>rl78-*-elf</h3>
1036
1037<p>The Renesas RL78 processor.
1038This configuration is intended for embedded systems.
1039
1040   <p><hr /><a name="riscv32-x-elf"></a>
1041
1042<h3 class="heading"><a name="TOC52"></a>riscv32-*-elf</h3>
1043
1044<p>The RISC-V RV32 instruction set.
1045This configuration is intended for embedded systems.
1046This (and all other RISC-V) targets are supported upstream as of the
1047binutils 2.28 release.
1048
1049   <p><hr /><a name="riscv32-x-linux"></a>
1050
1051<h3 class="heading"><a name="TOC53"></a>riscv32-*-linux</h3>
1052
1053<p>The RISC-V RV32 instruction set running GNU/Linux.
1054This (and all other RISC-V) targets are supported upstream as of the
1055binutils 2.28 release.
1056
1057   <p><hr /><a name="riscv64-x-elf"></a>
1058
1059<h3 class="heading"><a name="TOC54"></a>riscv64-*-elf</h3>
1060
1061<p>The RISC-V RV64 instruction set.
1062This configuration is intended for embedded systems.
1063This (and all other RISC-V) targets are supported upstream as of the
1064binutils 2.28 release.
1065
1066   <p><hr /><a name="riscv64-x-linux"></a>
1067
1068<h3 class="heading"><a name="TOC55"></a>riscv64-*-linux</h3>
1069
1070<p>The RISC-V RV64 instruction set running GNU/Linux.
1071This (and all other RISC-V) targets are supported upstream as of the
1072binutils 2.28 release.
1073
1074   <p><hr /><a name="rx-x-elf"></a>
1075
1076<h3 class="heading"><a name="TOC56"></a>rx-*-elf</h3>
1077
1078<p>The Renesas RX processor.
1079
1080   <p><hr /><a name="s390-x-linux"></a>
1081
1082<h3 class="heading"><a name="TOC57"></a>s390-*-linux*</h3>
1083
1084<p>S/390 system running GNU/Linux for S/390.
1085
1086   <p><hr /><a name="s390x-x-linux"></a>
1087
1088<h3 class="heading"><a name="TOC58"></a>s390x-*-linux*</h3>
1089
1090<p>zSeries system (64-bit) running GNU/Linux for zSeries.
1091
1092   <p><hr /><a name="s390x-ibm-tpf"></a>
1093
1094<h3 class="heading"><a name="TOC59"></a>s390x-ibm-tpf*</h3>
1095
1096<p>zSeries system (64-bit) running TPF.  This platform is
1097supported as cross-compilation target only.
1098
1099   <p><hr /><!-- Please use Solaris 2 to refer to all release of Solaris, starting -->
1100<!-- with 2.0 until 2.6, 7, 8, etc.  Solaris 1 was a marketing name for -->
1101<!-- SunOS 4 releases which we don't use to avoid confusion.  Solaris -->
1102<!-- alone is too unspecific and must be avoided. -->
1103<a name="x-x-solaris2"></a>
1104
1105<h3 class="heading"><a name="TOC60"></a>*-*-solaris2*</h3>
1106
1107<p>Support for Solaris 9 has been removed in GCC 5.  Support for Solaris
11088 has been removed in GCC 4.8.  Support for Solaris 7 has been removed
1109in GCC 4.6.
1110
1111   <p>Sun does not ship a C compiler with Solaris 2 before Solaris 10, though
1112you can download the Sun Studio compilers for free.  In Solaris 10 and
111311, GCC 3.4.3 is available as <samp><span class="command">/usr/sfw/bin/gcc</span></samp>.  Solaris 11
1114also provides GCC 4.5.2, 4.7.3, and 4.8.2 as
1115<samp><span class="command">/usr/gcc/4.5/bin/gcc</span></samp> or similar.  Alternatively,
1116you can install a pre-built GCC to bootstrap and install GCC.  See the
1117<a href="binaries.html">binaries page</a> for details.
1118
1119   <p>The Solaris 2 <samp><span class="command">/bin/sh</span></samp> will often fail to configure
1120&lsquo;<samp><span class="samp">libstdc++-v3</span></samp>&rsquo;or &lsquo;<samp><span class="samp">boehm-gc</span></samp>&rsquo;.  We therefore recommend using the
1121following initial sequence of commands
1122
1123<pre class="smallexample">     % CONFIG_SHELL=/bin/ksh
1124     % export CONFIG_SHELL
1125</pre>
1126   <p class="noindent">and proceed as described in <a href="configure.html">the configure instructions</a>.
1127In addition we strongly recommend specifying an absolute path to invoke
1128<samp><var>srcdir</var><span class="command">/configure</span></samp>.
1129
1130   <p>Solaris 10 comes with a number of optional OS packages.  Some of these
1131are needed to use GCC fully, namely <code>SUNWarc</code>,
1132<code>SUNWbtool</code>, <code>SUNWesu</code>, <code>SUNWhea</code>, <code>SUNWlibm</code>,
1133<code>SUNWsprot</code>, and <code>SUNWtoo</code>.  If you did not install all
1134optional packages when installing Solaris 10, you will need to verify that
1135the packages that GCC needs are installed.
1136To check whether an optional package is installed, use
1137the <samp><span class="command">pkginfo</span></samp> command.  To add an optional package, use the
1138<samp><span class="command">pkgadd</span></samp> command.  For further details, see the Solaris 10
1139documentation.
1140
1141   <p>Starting with Solaris 11, the package management has changed, so you
1142need to check for <code>system/header</code>, <code>system/linker</code>, and
1143<code>developer/assembler</code> packages.  Checking for and installing
1144packages is done with the <samp><span class="command">pkg</span></samp> command now.
1145
1146   <p>Trying to use the linker and other tools in
1147<samp><span class="file">/usr/ucb</span></samp> to install GCC has been observed to cause trouble.
1148For example, the linker may hang indefinitely.  The fix is to remove
1149<samp><span class="file">/usr/ucb</span></samp> from your <samp><span class="env">PATH</span></samp>.
1150
1151   <p>The build process works more smoothly with the legacy Sun tools so, if you
1152have <samp><span class="file">/usr/xpg4/bin</span></samp> in your <samp><span class="env">PATH</span></samp>, we recommend that you place
1153<samp><span class="file">/usr/bin</span></samp> before <samp><span class="file">/usr/xpg4/bin</span></samp> for the duration of the build.
1154
1155   <p>We recommend the use of the Solaris assembler or the GNU assembler, in
1156conjunction with the Solaris linker.  The GNU <samp><span class="command">as</span></samp>
1157versions included in Solaris 10, from GNU binutils 2.15 (in
1158<samp><span class="file">/usr/sfw/bin/gas</span></samp>), and Solaris 11,
1159from GNU binutils 2.19 or newer (also in <samp><span class="file">/usr/bin/gas</span></samp> and
1160<samp><span class="file">/usr/gnu/bin/as</span></samp>), are known to work.
1161Current versions of GNU binutils (2.26)
1162are known to work as well, with the caveat mentioned in
1163<a href="#ix86-x-solaris210">i?86-*-solaris2.10</a> .  Note that your mileage may vary
1164if you use a combination of the GNU tools and the Solaris tools: while the
1165combination GNU <samp><span class="command">as</span></samp> + Sun <samp><span class="command">ld</span></samp> should reasonably work,
1166the reverse combination Sun <samp><span class="command">as</span></samp> + GNU <samp><span class="command">ld</span></samp> may fail to
1167build or cause memory corruption at runtime in some cases for C++ programs.
1168<!-- FIXME: still? -->
1169GNU <samp><span class="command">ld</span></samp> usually works as well, although the version included in
1170Solaris 10 cannot be used due to several bugs.  Again, the current
1171version (2.26) is known to work, but generally lacks platform specific
1172features, so better stay with Solaris <samp><span class="command">ld</span></samp>.  To use the LTO linker
1173plugin (<samp><span class="option">-fuse-linker-plugin</span></samp>) with GNU <samp><span class="command">ld</span></samp>, GNU
1174binutils <em>must</em> be configured with <samp><span class="option">--enable-largefile</span></samp>.
1175
1176   <p>To enable symbol versioning in &lsquo;<samp><span class="samp">libstdc++</span></samp>&rsquo; with the Solaris linker,
1177you need to have any version of GNU <samp><span class="command">c++filt</span></samp>, which is part of
1178GNU binutils.  &lsquo;<samp><span class="samp">libstdc++</span></samp>&rsquo; symbol versioning will be disabled if no
1179appropriate version is found.  Solaris <samp><span class="command">c++filt</span></samp> from the Solaris
1180Studio compilers does <em>not</em> work.
1181
1182   <p>Sun bug 4927647 sometimes causes random spurious testsuite failures
1183related to missing diagnostic output.  This bug doesn't affect GCC
1184itself, rather it is a kernel bug triggered by the <samp><span class="command">expect</span></samp>
1185program which is used only by the GCC testsuite driver.  When the bug
1186causes the <samp><span class="command">expect</span></samp> program to miss anticipated output, extra
1187testsuite failures appear.
1188
1189   <p><hr /><a name="sparc-x-x"></a>
1190
1191<h3 class="heading"><a name="TOC61"></a>sparc*-*-*</h3>
1192
1193<p>This section contains general configuration information for all
1194SPARC-based platforms.  In addition to reading this section, please
1195read all other sections that match your target.
1196
1197   <p>Newer versions of the GNU Multiple Precision Library (GMP), the MPFR
1198library and the MPC library are known to be miscompiled by earlier
1199versions of GCC on these platforms.  We therefore recommend the use
1200of the exact versions of these libraries listed as minimal versions
1201in <a href="prerequisites.html">the prerequisites</a>.
1202
1203   <p><hr /><a name="sparc-sun-solaris2"></a>
1204
1205<h3 class="heading"><a name="TOC62"></a>sparc-sun-solaris2*</h3>
1206
1207<p>When GCC is configured to use GNU binutils 2.14 or later, the binaries
1208produced are smaller than the ones produced using Sun's native tools;
1209this difference is quite significant for binaries containing debugging
1210information.
1211
1212   <p>Starting with Solaris 7, the operating system is capable of executing
121364-bit SPARC V9 binaries.  GCC 3.1 and later properly supports
1214this; the <samp><span class="option">-m64</span></samp> option enables 64-bit code generation.
1215However, if all you want is code tuned for the UltraSPARC CPU, you
1216should try the <samp><span class="option">-mtune=ultrasparc</span></samp> option instead, which produces
1217code that, unlike full 64-bit code, can still run on non-UltraSPARC
1218machines.
1219
1220   <p>When configuring the GNU Multiple Precision Library (GMP), the MPFR
1221library or the MPC library on a Solaris 7 or later system, the canonical
1222target triplet must be specified as the <samp><span class="command">build</span></samp> parameter on the
1223configure line.  This target triplet can be obtained by invoking <samp><span class="command">./config.guess</span></samp> in the toplevel source directory of GCC (and
1224not that of GMP or MPFR or MPC).  For example on a Solaris 9 system:
1225
1226<pre class="smallexample">     % ./configure --build=sparc-sun-solaris2.9 --prefix=xxx
1227</pre>
1228   <p><hr /><a name="sparc-sun-solaris210"></a>
1229
1230<h3 class="heading"><a name="TOC63"></a>sparc-sun-solaris2.10</h3>
1231
1232<p>There is a bug in older versions of the Sun assembler which breaks
1233thread-local storage (TLS).  A typical error message is
1234
1235<pre class="smallexample">     ld: fatal: relocation error: R_SPARC_TLS_LE_HIX22: file /var/tmp//ccamPA1v.o:
1236       symbol &lt;unknown&gt;: bad symbol type SECT: symbol type must be TLS
1237</pre>
1238   <p class="noindent">This bug is fixed in Sun patch 118683-03 or later.
1239
1240   <p><hr /><a name="sparc-x-linux"></a>
1241
1242<h3 class="heading"><a name="TOC64"></a>sparc-*-linux*</h3>
1243
1244<p><hr /><a name="sparc64-x-solaris2"></a>
1245
1246<h3 class="heading"><a name="TOC65"></a>sparc64-*-solaris2*</h3>
1247
1248<p>When configuring the GNU Multiple Precision Library (GMP), the MPFR
1249library or the MPC library, the canonical target triplet must be specified
1250as the <samp><span class="command">build</span></samp> parameter on the configure line.  For example
1251on a Solaris 9 system:
1252
1253<pre class="smallexample">     % ./configure --build=sparc64-sun-solaris2.9 --prefix=xxx
1254</pre>
1255   <p><hr /><a name="sparcv9-x-solaris2"></a>
1256
1257<h3 class="heading"><a name="TOC66"></a>sparcv9-*-solaris2*</h3>
1258
1259<p>This is a synonym for &lsquo;<samp><span class="samp">sparc64-*-solaris2*</span></samp>&rsquo;.
1260
1261   <p><hr /><a name="c6x-x-x"></a>
1262
1263<h3 class="heading"><a name="TOC67"></a>c6x-*-*</h3>
1264
1265<p>The C6X family of processors. This port requires binutils-2.22 or newer.
1266
1267   <p><hr /><a name="tilegx-*-linux"></a>
1268
1269<h3 class="heading"><a name="TOC68"></a>tilegx-*-linux*</h3>
1270
1271<p>The TILE-Gx processor in little endian mode, running GNU/Linux.  This
1272port requires binutils-2.22 or newer.
1273
1274   <p><hr /><a name="tilegxbe-*-linux"></a>
1275
1276<h3 class="heading"><a name="TOC69"></a>tilegxbe-*-linux*</h3>
1277
1278<p>The TILE-Gx processor in big endian mode, running GNU/Linux.  This
1279port requires binutils-2.23 or newer.
1280
1281   <p><hr /><a name="tilepro-*-linux"></a>
1282
1283<h3 class="heading"><a name="TOC70"></a>tilepro-*-linux*</h3>
1284
1285<p>The TILEPro processor running GNU/Linux.  This port requires
1286binutils-2.22 or newer.
1287
1288   <p><hr /><a name="visium-x-elf"></a>
1289
1290<h3 class="heading"><a name="TOC71"></a>visium-*-elf</h3>
1291
1292<p>CDS VISIUMcore processor.
1293This configuration is intended for embedded systems.
1294
1295   <p><hr /><a name="x-x-vxworks"></a>
1296
1297<h3 class="heading"><a name="TOC72"></a>*-*-vxworks*</h3>
1298
1299<p>Support for VxWorks is in flux.  At present GCC supports <em>only</em> the
1300very recent VxWorks 5.5 (aka Tornado 2.2) release, and only on PowerPC.
1301We welcome patches for other architectures supported by VxWorks 5.5.
1302Support for VxWorks AE would also be welcome; we believe this is merely
1303a matter of writing an appropriate &ldquo;configlette&rdquo; (see below).  We are
1304not interested in supporting older, a.out or COFF-based, versions of
1305VxWorks in GCC 3.
1306
1307   <p>VxWorks comes with an older version of GCC installed in
1308<samp><var>$WIND_BASE</var><span class="file">/host</span></samp>; we recommend you do not overwrite it.
1309Choose an installation <var>prefix</var> entirely outside <var>$WIND_BASE</var>.
1310Before running <samp><span class="command">configure</span></samp>, create the directories <samp><var>prefix</var></samp>
1311and <samp><var>prefix</var><span class="file">/bin</span></samp>.  Link or copy the appropriate assembler,
1312linker, etc. into <samp><var>prefix</var><span class="file">/bin</span></samp>, and set your <var>PATH</var> to
1313include that directory while running both <samp><span class="command">configure</span></samp> and
1314<samp><span class="command">make</span></samp>.
1315
1316   <p>You must give <samp><span class="command">configure</span></samp> the
1317<samp><span class="option">--with-headers=</span><var>$WIND_BASE</var><span class="option">/target/h</span></samp> switch so that it can
1318find the VxWorks system headers.  Since VxWorks is a cross compilation
1319target only, you must also specify <samp><span class="option">--target=</span><var>target</var></samp>.
1320<samp><span class="command">configure</span></samp> will attempt to create the directory
1321<samp><var>prefix</var><span class="file">/</span><var>target</var><span class="file">/sys-include</span></samp> and copy files into it;
1322make sure the user running <samp><span class="command">configure</span></samp> has sufficient privilege
1323to do so.
1324
1325   <p>GCC's exception handling runtime requires a special &ldquo;configlette&rdquo;
1326module, <samp><span class="file">contrib/gthr_supp_vxw_5x.c</span></samp>.  Follow the instructions in
1327that file to add the module to your kernel build.  (Future versions of
1328VxWorks will incorporate this module.)
1329
1330   <p><hr /><a name="x86-64-x-x"></a>
1331
1332<h3 class="heading"><a name="TOC73"></a>x86_64-*-*, amd64-*-*</h3>
1333
1334<p>GCC supports the x86-64 architecture implemented by the AMD64 processor
1335(amd64-*-* is an alias for x86_64-*-*) on GNU/Linux, FreeBSD and NetBSD.
1336On GNU/Linux the default is a bi-arch compiler which is able to generate
1337both 64-bit x86-64 and 32-bit x86 code (via the <samp><span class="option">-m32</span></samp> switch).
1338
1339   <p><hr /><a name="x86-64-x-solaris210"></a>
1340
1341<h3 class="heading"><a name="TOC74"></a>x86_64-*-solaris2.1[0-9]*</h3>
1342
1343<p>GCC also supports the x86-64 architecture implemented by the AMD64
1344processor (&lsquo;<samp><span class="samp">amd64-*-*</span></samp>&rsquo; is an alias for &lsquo;<samp><span class="samp">x86_64-*-*</span></samp>&rsquo;) on
1345Solaris 10 or later.  Unlike other systems, without special options a
1346bi-arch compiler is built which generates 32-bit code by default, but
1347can generate 64-bit x86-64 code with the <samp><span class="option">-m64</span></samp> switch.  Since
1348GCC 4.7, there is also a configuration that defaults to 64-bit code, but
1349can generate 32-bit code with <samp><span class="option">-m32</span></samp>.  To configure and build
1350this way, you have to provide all support libraries like <samp><span class="file">libgmp</span></samp>
1351as 64-bit code, configure with <samp><span class="option">--target=x86_64-pc-solaris2.1x</span></samp>
1352and &lsquo;<samp><span class="samp">CC=gcc -m64</span></samp>&rsquo;.
1353
1354   <p><hr /><a name="xtensa-x-elf"></a>
1355
1356<h3 class="heading"><a name="TOC75"></a>xtensa*-*-elf</h3>
1357
1358<p>This target is intended for embedded Xtensa systems using the
1359&lsquo;<samp><span class="samp">newlib</span></samp>&rsquo; C library.  It uses ELF but does not support shared
1360objects.  Designed-defined instructions specified via the
1361Tensilica Instruction Extension (TIE) language are only supported
1362through inline assembly.
1363
1364   <p>The Xtensa configuration information must be specified prior to
1365building GCC.  The <samp><span class="file">include/xtensa-config.h</span></samp> header
1366file contains the configuration information.  If you created your
1367own Xtensa configuration with the Xtensa Processor Generator, the
1368downloaded files include a customized copy of this header file,
1369which you can use to replace the default header file.
1370
1371   <p><hr /><a name="xtensa-x-linux"></a>
1372
1373<h3 class="heading"><a name="TOC76"></a>xtensa*-*-linux*</h3>
1374
1375<p>This target is for Xtensa systems running GNU/Linux.  It supports ELF
1376shared objects and the GNU C library (glibc).  It also generates
1377position-independent code (PIC) regardless of whether the
1378<samp><span class="option">-fpic</span></samp> or <samp><span class="option">-fPIC</span></samp> options are used.  In other
1379respects, this target is the same as the
1380<a href="#xtensa*-*-elf">&lsquo;<samp><span class="samp">xtensa*-*-elf</span></samp>&rsquo;</a> target.
1381
1382   <p><hr /><a name="windows"></a>
1383
1384<h3 class="heading"><a name="TOC77"></a>Microsoft Windows</h3>
1385
1386<h4 class="subheading"><a name="TOC78"></a>Intel 16-bit versions</h4>
1387
1388<p>The 16-bit versions of Microsoft Windows, such as Windows 3.1, are not
1389supported.
1390
1391   <p>However, the 32-bit port has limited support for Microsoft
1392Windows 3.11 in the Win32s environment, as a target only.  See below.
1393
1394<h4 class="subheading"><a name="TOC79"></a>Intel 32-bit versions</h4>
1395
1396<p>The 32-bit versions of Windows, including Windows 95, Windows NT, Windows
1397XP, and Windows Vista, are supported by several different target
1398platforms.  These targets differ in which Windows subsystem they target
1399and which C libraries are used.
1400
1401     <ul>
1402<li>Cygwin <a href="#x-x-cygwin">*-*-cygwin</a>: Cygwin provides a user-space
1403Linux API emulation layer in the Win32 subsystem.
1404<li>MinGW <a href="#x-x-mingw32">*-*-mingw32</a>: MinGW is a native GCC port for
1405the Win32 subsystem that provides a subset of POSIX.
1406<li>MKS i386-pc-mks: NuTCracker from MKS.  See
1407<a href="http://www.mkssoftware.com/">http://www.mkssoftware.com/</a> for more information.
1408</ul>
1409
1410<h4 class="subheading"><a name="TOC80"></a>Intel 64-bit versions</h4>
1411
1412<p>GCC contains support for x86-64 using the mingw-w64
1413runtime library, available from <a href="http://mingw-w64.org/doku.php">http://mingw-w64.org/doku.php</a>.
1414This library should be used with the target triple x86_64-pc-mingw32.
1415
1416   <p>Presently Windows for Itanium is not supported.
1417
1418<h4 class="subheading"><a name="TOC81"></a>Windows CE</h4>
1419
1420<p>Windows CE is supported as a target only on Hitachi
1421SuperH (sh-wince-pe), and MIPS (mips-wince-pe).
1422
1423<h4 class="subheading"><a name="TOC82"></a>Other Windows Platforms</h4>
1424
1425<p>GCC no longer supports Windows NT on the Alpha or PowerPC.
1426
1427   <p>GCC no longer supports the Windows POSIX subsystem.  However, it does
1428support the Interix subsystem.  See above.
1429
1430   <p>Old target names including *-*-winnt and *-*-windowsnt are no longer used.
1431
1432   <p>PW32 (i386-pc-pw32) support was never completed, and the project seems to
1433be inactive.  See <a href="http://pw32.sourceforge.net/">http://pw32.sourceforge.net/</a> for more information.
1434
1435   <p>UWIN support has been removed due to a lack of maintenance.
1436
1437   <p><hr /><a name="x-x-cygwin"></a>
1438
1439<h3 class="heading"><a name="TOC83"></a>*-*-cygwin</h3>
1440
1441<p>Ports of GCC are included with the
1442<a href="http://www.cygwin.com/">Cygwin environment</a>.
1443
1444   <p>GCC will build under Cygwin without modification; it does not build
1445with Microsoft's C++ compiler and there are no plans to make it do so.
1446
1447   <p>The Cygwin native compiler can be configured to target any 32-bit x86
1448cpu architecture desired; the default is i686-pc-cygwin.  It should be
1449used with as up-to-date a version of binutils as possible; use either
1450the latest official GNU binutils release in the Cygwin distribution,
1451or version 2.20 or above if building your own.
1452
1453   <p><hr /><a name="x-x-mingw32"></a>
1454
1455<h3 class="heading"><a name="TOC84"></a>*-*-mingw32</h3>
1456
1457<p>GCC will build with and support only MinGW runtime 3.12 and later.
1458Earlier versions of headers are incompatible with the new default semantics
1459of <code>extern inline</code> in <code>-std=c99</code> and <code>-std=gnu99</code> modes.
1460
1461   <p><hr /><a name="older"></a>
1462
1463<h3 class="heading"><a name="TOC85"></a>Older systems</h3>
1464
1465<p>GCC contains support files for many older (1980s and early
14661990s) Unix variants.  For the most part, support for these systems
1467has not been deliberately removed, but it has not been maintained for
1468several years and may suffer from bitrot.
1469
1470   <p>Starting with GCC 3.1, each release has a list of &ldquo;obsoleted&rdquo; systems.
1471Support for these systems is still present in that release, but
1472<samp><span class="command">configure</span></samp> will fail unless the <samp><span class="option">--enable-obsolete</span></samp>
1473option is given.  Unless a maintainer steps forward, support for these
1474systems will be removed from the next release of GCC.
1475
1476   <p>Support for old systems as hosts for GCC can cause problems if the
1477workarounds for compiler, library and operating system bugs affect the
1478cleanliness or maintainability of the rest of GCC.  In some cases, to
1479bring GCC up on such a system, if still possible with current GCC, may
1480require first installing an old version of GCC which did work on that
1481system, and using it to compile a more recent GCC, to avoid bugs in the
1482vendor compiler.  Old releases of GCC 1 and GCC 2 are available in the
1483<samp><span class="file">old-releases</span></samp> directory on the <a href="../mirrors.html">GCC mirror sites</a>.  Header bugs may generally be avoided using
1484<samp><span class="command">fixincludes</span></samp>, but bugs or deficiencies in libraries and the
1485operating system may still cause problems.
1486
1487   <p>Support for older systems as targets for cross-compilation is less
1488problematic than support for them as hosts for GCC; if an enthusiast
1489wishes to make such a target work again (including resurrecting any of
1490the targets that never worked with GCC 2, starting from the last
1491version before they were removed), patches
1492<a href="../contribute.html">following the usual requirements</a> would be
1493likely to be accepted, since they should not affect the support for more
1494modern targets.
1495
1496   <p>For some systems, old versions of GNU binutils may also be useful,
1497and are available from <samp><span class="file">pub/binutils/old-releases</span></samp> on
1498<a href="https://sourceware.org/mirrors.html">sourceware.org mirror sites</a>.
1499
1500   <p>Some of the information on specific systems above relates to
1501such older systems, but much of the information
1502about GCC on such systems (which may no longer be applicable to
1503current GCC) is to be found in the GCC texinfo manual.
1504
1505   <p><hr /><a name="elf"></a>
1506
1507<h3 class="heading"><a name="TOC86"></a>all ELF targets (SVR4, Solaris 2, etc.)</h3>
1508
1509<p>C++ support is significantly better on ELF targets if you use the
1510<a href="./configure.html#with-gnu-ld">GNU linker</a>; duplicate copies of
1511inlines, vtables and template instantiations will be discarded
1512automatically.
1513
1514   <p><hr />
1515<p><a href="./index.html">Return to the GCC Installation page</a>
1516
1517<!-- ***Old documentation****************************************************** -->
1518<!-- ***GFDL******************************************************************** -->
1519<!-- *************************************************************************** -->
1520<!-- Part 6 The End of the Document -->
1521</body></html>
1522
1523