Gitiles
Code Review Sign In
gerrit-public.fairphone.software / platform / external / clang / 5ab5249cbf6e6cd97ae7cdcad4236bc8cbd2f064 / . / docs / UsersManual.html
blob: 2356b0314678e2cbf4cc5d3deb25ba247cbcdf8c [file] [log] [blame]
Benjamin Kramer665a8dc2012-01-15 15:26:07 +0000[diff] [blame]1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]3<html>
4<head>
5<title>Clang Compiler User's Manual</title>
Benjamin Kramer665a8dc2012-01-15 15:26:07 +0000[diff] [blame]6<link type="text/css" rel="stylesheet" href="../menu.css">
7<link type="text/css" rel="stylesheet" href="../content.css">
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]8<style type="text/css">
9td {
10 vertical-align: top;
11}
12</style>
13</head>
14<body>
15
16<!--#include virtual="../menu.html.incl"-->
17
18<div id="content">
19
20<h1>Clang Compiler User's Manual</h1>
21
22<ul>
23<li><a href="#intro">Introduction</a>
24 <ul>
25 <li><a href="#terminology">Terminology</a></li>
26 <li><a href="#basicusage">Basic Usage</a></li>
27 </ul>
28</li>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]29<li><a href="#commandline">Command Line Options</a>
30 <ul>
31 <li><a href="#cl_diagnostics">Options to Control Error and Warning
32 Messages</a></li>
Chad Rosierb211a4b2012-02-22 19:14:46 +0000[diff] [blame]33 <li><a href="#cl_crash_diagnostics">Options to Control Clang Crash
34 Diagnostics</a></li>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]35 </ul>
36</li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]37<li><a href="#general_features">Language and Target-Independent Features</a>
38 <ul>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]39 <li><a href="#diagnostics">Controlling Errors and Warnings</a>
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]40 <ul>
41 <li><a href="#diagnostics_display">Controlling How Clang Displays Diagnostics</a></li>
42 <li><a href="#diagnostics_mappings">Diagnostic Mappings</a></li>
Chris Lattner67db8cd2010-05-30 23:42:51 +0000[diff] [blame]43 <li><a href="#diagnostics_categories">Diagnostic Categories</a></li>
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]44 <li><a href="#diagnostics_commandline">Controlling Diagnostics via Command Line Flags</a></li>
45 <li><a href="#diagnostics_pragmas">Controlling Diagnostics via Pragmas</a></li>
Ted Kremenekf391fa72011-08-18 01:17:05 +0000[diff] [blame]46 <li><a href="#diagnostics_enable_everything">Enabling All Warnings</a></li>
Ted Kremenek2fb11eb2010-08-24 18:12:35 +0000[diff] [blame]47 <li><a href="#analyzer_diagnositics">Controlling Static Analyzer Diagnostics</a></li>
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]48 </ul>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]49 </li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]50 <li><a href="#precompiledheaders">Precompiled Headers</a></li>
Mike Stump53664ca2009-12-14 23:53:10 +0000[diff] [blame]51 <li><a href="#codegen">Controlling Code Generation</a></li>
Alexey Samsonov5ab52492012-05-30 06:55:10 +0000[diff] [blame^]52 <li><a href="#debuginfosize">Controlling Size of Debug Information</a></li>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]53 </ul>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]54</li>
55<li><a href="#c">C Language Features</a>
56 <ul>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]57 <li><a href="#c_ext">Extensions supported by clang</a></li>
58 <li><a href="#c_modes">Differences between various standard modes</a></li>
59 <li><a href="#c_unimpl_gcc">GCC extensions not implemented yet</a></li>
60 <li><a href="#c_unsupp_gcc">Intentionally unsupported GCC extensions</a></li>
61 <li><a href="#c_ms">Microsoft extensions</a></li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]62 </ul>
63</li>
Richard Smithbdb10182011-11-21 20:54:59 +0000[diff] [blame]64<li><a href="#cxx">C++ Language Features</a>
65 <ul>
66 <li><a href="#cxx_implimits">Controlling implementation limits</a></li>
67 </ul>
68</li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]69<li><a href="#target_features">Target-Specific Features and Limitations</a>
70 <ul>
71 <li><a href="#target_arch">CPU Architectures Features and Limitations</a>
72 <ul>
73 <li><a href="#target_arch_x86">X86</a></li>
Eli Friedman3b658d32009-06-08 05:12:39 +0000[diff] [blame]74 <li><a href="#target_arch_arm">ARM</a></li>
75 <li><a href="#target_arch_other">Other platforms</a></li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]76 </ul>
77 </li>
78 <li><a href="#target_os">Operating System Features and Limitations</a>
79 <ul>
80 <li><a href="#target_os_darwin">Darwin (Mac OS/X)</a></li>
81 <li>Linux, etc.</li>
NAKAMURA Takumia75fdb22011-04-04 15:02:41 +0000[diff] [blame]82 <li><a href="#target_os_win32">Windows</a></li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]83 </ul>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]84 </li>
85 </ul>
86</li>
87</ul>
88
89
90<!-- ======================================================================= -->
91<h2 id="intro">Introduction</h2>
92<!-- ======================================================================= -->
93
94<p>The Clang Compiler is an open-source compiler for the C family of programming
95languages, aiming to be the best in class implementation of these languages.
96Clang builds on the LLVM optimizer and code generator, allowing it to provide
97high-quality optimization and code generation support for many targets. For
98more general information, please see the <a href="http://clang.llvm.org">Clang
99Web Site</a> or the <a href="http://llvm.org">LLVM Web Site</a>.</p>
100
101<p>This document describes important notes about using Clang as a compiler for
102an end-user, documenting the supported features, command line options, etc. If
103you are interested in using Clang to build a tool that processes code, please
104see <a href="InternalsManual.html">the Clang Internals Manual</a>. If you are
Michael J. Spencer8275fc02012-02-12 09:11:39 +0000[diff] [blame]105interested in the <a href="http://clang-analyzer.llvm.org">Clang
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]106Static Analyzer</a>, please see its web page.</p>
107
108<p>Clang is designed to support the C family of programming languages, which
109includes <a href="#c">C</a>, <a href="#objc">Objective-C</a>, <a
110href="#cxx">C++</a>, and <a href="#objcxx">Objective-C++</a> as well as many
111dialects of those. For language-specific information, please see the
112corresponding language specific section:</p>
113
114<ul>
115<li><a href="#c">C Language</a>: K&amp;R C, ANSI C89, ISO C90, ISO C94
116 (C89+AMD1), ISO C99 (+TC1, TC2, TC3). </li>
117<li><a href="#objc">Objective-C Language</a>: ObjC 1, ObjC 2, ObjC 2.1, plus
118 variants depending on base language.</li>
Richard Smithbdb10182011-11-21 20:54:59 +0000[diff] [blame]119<li><a href="#cxx">C++ Language</a></li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]120<li><a href="#objcxx">Objective C++ Language</a></li>
121</ul>
122
123<p>In addition to these base languages and their dialects, Clang supports a
124broad variety of language extensions, which are documented in the corresponding
125language section. These extensions are provided to be compatible with the GCC,
126Microsoft, and other popular compilers as well as to improve functionality
127through Clang-specific features. The Clang driver and language features are
128intentionally designed to be as compatible with the GNU GCC compiler as
129reasonably possible, easing migration from GCC to Clang. In most cases, code
130"just works".</p>
131
132<p>In addition to language specific features, Clang has a variety of features
133that depend on what CPU architecture or operating system is being compiled for.
Douglas Gregorcd5a5052009-11-09 15:15:41 +0000[diff] [blame]134Please see the <a href="#target_features">Target-Specific Features and
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]135Limitations</a> section for more details.</p>
136
137<p>The rest of the introduction introduces some basic <a
138href="#terminology">compiler terminology</a> that is used throughout this manual
139and contains a basic <a href="#basicusage">introduction to using Clang</a>
140as a command line compiler.</p>
141
142<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
143<h3 id="terminology">Terminology</h3>
144<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
145
146<p>Front end, parser, backend, preprocessor, undefined behavior, diagnostic,
147 optimizer</p>
148
149<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
150<h3 id="basicusage">Basic Usage</h3>
151<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
152
153<p>Intro to how to use a C compiler for newbies.</p>
154<p>
155compile + link
156
157compile then link
158
159debug info
160
161enabling optimizations
162
163picking a language to use, defaults to C99 by default. Autosenses based on
164extension.
165
166using a makefile
167</p>
168
169
170<!-- ======================================================================= -->
171<h2 id="commandline">Command Line Options</h2>
172<!-- ======================================================================= -->
173
174<p>
175This section is generally an index into other sections. It does not go into
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]176depth on the ones that are covered by other sections. However, the first part
177introduces the language selection and other high level options like -c, -g, etc.
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]178</p>
179
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]180
181<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
182<h3 id="cl_diagnostics">Options to Control Error and Warning Messages</h3>
183<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
184
185<p><b>-Werror</b>: Turn warnings into errors.</p>
186<p><b>-Werror=foo</b>: Turn warning "foo" into an error.</p>
187<p><b>-Wno-error=foo</b>: Turn warning "foo" into an warning even if -Werror is
188 specified.</p>
189<p><b>-Wfoo</b>: Enable warning foo</p>
190<p><b>-Wno-foo</b>: Disable warning foo</p>
191<p><b>-w</b>: Disable all warnings.</p>
192<p><b>-pedantic</b>: Warn on language extensions.</p>
193<p><b>-pedantic-errors</b>: Error on language extensions.</p>
194<p><b>-Wsystem-headers</b>: Enable warnings from system headers.</p>
195
Chris Lattner0f0c9632010-04-07 20:49:23 +0000[diff] [blame]196<p><b>-ferror-limit=123</b>: Stop emitting diagnostics after 123 errors have
197 been produced. The default is 20, and the error limit can be disabled with
198 -ferror-limit=0.</p>
199
Douglas Gregor575cf372010-04-20 07:18:24 +0000[diff] [blame]200<p><b>-ftemplate-backtrace-limit=123</b>: Only emit up to 123 template instantiation notes within the template instantiation backtrace for a single warning or error. The default is 10, and the limit can be disabled with -ftemplate-backtrace-limit=0.</p>
201
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]202<!-- ================================================= -->
203<h4 id="cl_diag_formatting">Formatting of Diagnostics</h4>
204<!-- ================================================= -->
205
206<p>Clang aims to produce beautiful diagnostics by default, particularly for new
207users that first come to Clang. However, different people have different
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]208preferences, and sometimes Clang is driven by another program that wants to
209parse simple and consistent output, not a person. For these cases, Clang
210provides a wide range of options to control the exact output format of the
211diagnostics that it generates.</p>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]212
213<dl>
214
215<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
216<dt id="opt_fshow-column"><b>-f[no-]show-column</b>: Print column number in
217diagnostic.</dt>
218<dd>This option, which defaults to on, controls whether or not Clang prints the
219column number of a diagnostic. For example, when this is enabled, Clang will
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]220print something like:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]221
222<pre>
223 test.c:28:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
224 #endif bad
225 ^
226 //
227</pre>
228
229<p>When this is disabled, Clang will print "test.c:28: warning..." with no
230column number.</p>
231</dd>
232
233<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
234<dt id="opt_fshow-source-location"><b>-f[no-]show-source-location</b>: Print
235source file/line/column information in diagnostic.</dt>
236<dd>This option, which defaults to on, controls whether or not Clang prints the
237filename, line number and column number of a diagnostic. For example,
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]238when this is enabled, Clang will print something like:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]239
240<pre>
241 test.c:28:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
242 #endif bad
243 ^
244 //
245</pre>
246
247<p>When this is disabled, Clang will not print the "test.c:28:8: " part.</p>
248</dd>
249
250<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
251<dt id="opt_fcaret-diagnostics"><b>-f[no-]caret-diagnostics</b>: Print source
252line and ranges from source code in diagnostic.</dt>
253<dd>This option, which defaults to on, controls whether or not Clang prints the
254source line, source ranges, and caret when emitting a diagnostic. For example,
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]255when this is enabled, Clang will print something like:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]256
257<pre>
258 test.c:28:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
259 #endif bad
260 ^
261 //
262</pre>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]263</dd>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]264<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
Douglas Gregore8d44dd2010-07-09 16:31:58 +0000[diff] [blame]265<dt id="opt_fcolor_diagnostics"><b>-f[no-]color-diagnostics</b>: </dt>
266<dd>This option, which defaults to on when a color-capable terminal is
267 detected, controls whether or not Clang prints diagnostics in color.
268 When this option is enabled, Clang will use colors to highlight
269 specific parts of the diagnostic, e.g.,
Chris Lattner4e1c53d2011-01-24 03:47:34 +0000[diff] [blame]270 <pre>
Benjamin Kramer665a8dc2012-01-15 15:26:07 +0000[diff] [blame]271 <b><span style="color:black">test.c:28:8: <span style="color:magenta">warning</span>: extra tokens at end of #endif directive [-Wextra-tokens]</span></b>
Chris Lattner4e1c53d2011-01-24 03:47:34 +0000[diff] [blame]272 #endif bad
Benjamin Kramer665a8dc2012-01-15 15:26:07 +0000[diff] [blame]273 <span style="color:green">^</span>
274 <span style="color:green">//</span>
Douglas Gregore8d44dd2010-07-09 16:31:58 +0000[diff] [blame]275</pre>
Chris Lattner4e1c53d2011-01-24 03:47:34 +0000[diff] [blame]276
277<p>When this is disabled, Clang will just print:</p>
278
279<pre>
280 test.c:2:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
281 #endif bad
282 ^
283 //
284</pre>
285</dd>
Douglas Gregore8d44dd2010-07-09 16:31:58 +0000[diff] [blame]286<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
Douglas Gregorc9471b02011-05-21 17:07:29 +0000[diff] [blame]287<dt id="opt_fdiagnostics-format"><b>-fdiagnostics-format=clang/msvc/vi</b>:
288Changes diagnostic output format to better match IDEs and command line tools.</dt>
289<dd>This option controls the output format of the filename, line number, and column printed in diagnostic messages. The options, and their affect on formatting a simple conversion diagnostic, follow:
290
291 <dl>
292 <dt><b>clang</b> (default)</dt>
293 <dd>
294 <pre>t.c:3:11: warning: conversion specifies type 'char *' but the argument has type 'int'</pre>
295 </dd>
296
297 <dt><b>msvc</b></dt>
298 <dd>
299 <pre>t.c(3,11) : warning: conversion specifies type 'char *' but the argument has type 'int'</pre>
300 </dd>
301
302 <dt><b>vi</b></dt>
303 <dd>
304 <pre>t.c +3:11: warning: conversion specifies type 'char *' but the argument has type 'int'</pre>
305 </dd>
306 </dl>
307</dd>
308
309<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
Argyrios Kyrtzidis477aab62011-05-25 05:05:01 +0000[diff] [blame]310<dt id="opt_fdiagnostics-show-name"><b>-f[no-]diagnostics-show-name</b>:
311Enable the display of the diagnostic name.</dt>
312<dd>This option, which defaults to off, controls whether or not
Benjamin Kramer665a8dc2012-01-15 15:26:07 +0000[diff] [blame]313Clang prints the associated name.<p></p></dd>
Argyrios Kyrtzidis477aab62011-05-25 05:05:01 +0000[diff] [blame]314<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]315<dt id="opt_fdiagnostics-show-option"><b>-f[no-]diagnostics-show-option</b>:
316Enable <tt>[-Woption]</tt> information in diagnostic line.</dt>
317<dd>This option, which defaults to on,
318controls whether or not Clang prints the associated <A
319href="#cl_diag_warning_groups">warning group</a> option name when outputting
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]320a warning diagnostic. For example, in this output:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]321
322<pre>
323 test.c:28:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
324 #endif bad
325 ^
326 //
327</pre>
328
329<p>Passing <b>-fno-diagnostics-show-option</b> will prevent Clang from printing
330the [<a href="#opt_Wextra-tokens">-Wextra-tokens</a>] information in the
331diagnostic. This information tells you the flag needed to enable or disable the
332diagnostic, either from the command line or through <a
333href="#pragma_GCC_diagnostic">#pragma GCC diagnostic</a>.</dd>
334
Chris Lattner28a43a42010-05-05 01:35:28 +0000[diff] [blame]335<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
336<dt id="opt_fdiagnostics-show-category"><b>-fdiagnostics-show-category=none/id/name</b>:
337Enable printing category information in diagnostic line.</dt>
338<dd>This option, which defaults to "none",
339controls whether or not Clang prints the category associated with a diagnostic
340when emitting it. Each diagnostic may or many not have an associated category,
341if it has one, it is listed in the diagnostic categorization field of the
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]342diagnostic line (in the []'s).
Chris Lattner28a43a42010-05-05 01:35:28 +0000[diff] [blame]343
344<p>For example, a format string warning will produce these three renditions
345based on the setting of this option:</p>
346
347<pre>
348 t.c:3:11: warning: conversion specifies type 'char *' but the argument has type 'int' [-Wformat]
349 t.c:3:11: warning: conversion specifies type 'char *' but the argument has type 'int' [-Wformat<b>,1</b>]
350 t.c:3:11: warning: conversion specifies type 'char *' but the argument has type 'int' [-Wformat<b>,Format String</b>]
351</pre>
352
353<p>This category can be used by clients that want to group diagnostics by
354category, so it should be a high level category. We want dozens of these, not
355hundreds or thousands of them.</p>
356</dd>
357
358
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]359
360<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
361<dt id="opt_fdiagnostics-fixit-info"><b>-f[no-]diagnostics-fixit-info</b>:
362Enable "FixIt" information in the diagnostics output.</dt>
363<dd>This option, which defaults to on, controls whether or not Clang prints the
364information on how to fix a specific diagnostic underneath it when it knows.
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]365For example, in this output:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]366
367<pre>
368 test.c:28:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
369 #endif bad
370 ^
371 //
372</pre>
373
374<p>Passing <b>-fno-diagnostics-fixit-info</b> will prevent Clang from printing
375the "//" line at the end of the message. This information is useful for users
376who may not understand what is wrong, but can be confusing for machine
377parsing.</p>
378</dd>
379
380<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
Chris Lattner2a9cc232009-04-21 05:35:32 +0000[diff] [blame]381<dt id="opt_fdiagnostics-print-source-range-info">
382<b>-f[no-]diagnostics-print-source-range-info</b>:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]383Print machine parsable information about source ranges.</dt>
384<dd>This option, which defaults to off, controls whether or not Clang prints
385information about source ranges in a machine parsable format after the
386file/line/column number information. The information is a simple sequence of
387brace enclosed ranges, where each range lists the start and end line/column
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]388locations. For example, in this output:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]389
390<pre>
391exprs.c:47:15:{47:8-47:14}{47:17-47:24}: error: invalid operands to binary expression ('int *' and '_Complex float')
392 P = (P-42) + Gamma*4;
393 ~~~~~~ ^ ~~~~~~~
394</pre>
395
Chris Lattner2a9cc232009-04-21 05:35:32 +0000[diff] [blame]396<p>The {}'s are generated by -fdiagnostics-print-source-range-info.</p>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]397</dd>
398
Douglas Gregor4786c152010-08-19 20:24:43 +0000[diff] [blame]399<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
400<dt id="opt_fdiagnostics-parseable-fixits">
401<b>-fdiagnostics-parseable-fixits</b>:
402Print Fix-Its in a machine parseable form.</dt>
403<dd><p>This option makes Clang print available Fix-Its in a machine parseable format at the end of diagnostics. The following example illustrates the format:</p>
404
405<pre>
Douglas Gregorbf5e09d2010-08-20 03:17:33 +0000[diff] [blame]406 fix-it:"t.cpp":{7:25-7:29}:"Gamma"
Douglas Gregor4786c152010-08-19 20:24:43 +0000[diff] [blame]407</pre>
408
NAKAMURA Takumi44626362011-04-05 00:57:02 +0000[diff] [blame]409<p>The range printed is a half-open range, so in this example the characters at
410column 25 up to but not including column 29 on line 7 in t.cpp should be
411replaced with the string &quot;Gamma&quot;. Either the range or the replacement
412string may be empty (representing strict insertions and strict erasures,
413respectively). Both the file name and the insertion string escape backslash (as
414&quot;\\&quot;), tabs (as &quot;\t&quot;), newlines (as &quot;\n&quot;), double
415quotes(as &quot;\&quot;&quot;) and non-printable characters (as octal
416&quot;\xxx&quot;).</p>
Douglas Gregor4786c152010-08-19 20:24:43 +0000[diff] [blame]417</dd>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]418
419</dl>
420
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]421
422
423<!-- ===================================================== -->
424<h4 id="cl_diag_warning_groups">Individual Warning Groups</h4>
425<!-- ===================================================== -->
426
427<p>TODO: Generate this from tblgen. Define one anchor per warning group.</p>
428
429
430<dl>
431
432
433<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
434<dt id="opt_Wextra-tokens"><b>-Wextra-tokens</b>: Warn about excess tokens at
435 the end of a preprocessor directive.</dt>
436<dd>This option, which defaults to on, enables warnings about extra tokens at
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]437the end of preprocessor directives. For example:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]438
439<pre>
440 test.c:28:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
441 #endif bad
442 ^
443</pre>
444
445<p>These extra tokens are not strictly conforming, and are usually best handled
446by commenting them out.</p>
447
448<p>This option is also enabled by <a href="">-Wfoo</a>, <a href="">-Wbar</a>,
449 and <a href="">-Wbaz</a>.</p>
450</dd>
451
Jeffrey Yasskin21d07e42010-06-05 01:39:57 +0000[diff] [blame]452<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
453<dt id="opt_Wambiguous-member-template"><b>-Wambiguous-member-template</b>:
454Warn about unqualified uses of a member template whose name resolves
455to another template at the location of the use.</dt>
456<dd>This option, which defaults to on, enables a warning in the
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]457following code:
Jeffrey Yasskin21d07e42010-06-05 01:39:57 +0000[diff] [blame]458
459<pre>
460template&lt;typename T> struct set{};
461template&lt;typename T> struct trait { typedef const T& type; };
462struct Value {
463 template&lt;typename T> void set(typename trait&lt;T>::type value) {}
464};
465void foo() {
466 Value v;
467 v.set&lt;double>(3.2);
468}
469</pre>
470
471<p>C++ [basic.lookup.classref] requires this to be an error, but,
472because it's hard to work around, Clang downgrades it to a warning as
473an extension.</p>
474</dd>
475
Jeffrey Yasskin57d12fd2010-06-07 15:58:05 +0000[diff] [blame]476<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
477<dt id="opt_Wbind-to-temporary-copy"><b>-Wbind-to-temporary-copy</b>: Warn about
478an unusable copy constructor when binding a reference to a temporary.</dt>
479<dd>This option, which defaults to on, enables warnings about binding a
480reference to a temporary when the temporary doesn't have a usable copy
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]481constructor. For example:
Jeffrey Yasskin57d12fd2010-06-07 15:58:05 +0000[diff] [blame]482
483<pre>
484 struct NonCopyable {
485 NonCopyable();
486 private:
487 NonCopyable(const NonCopyable&);
488 };
489 void foo(const NonCopyable&);
490 void bar() {
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]491 foo(NonCopyable()); // Disallowed in C++98; allowed in C++11.
Jeffrey Yasskin57d12fd2010-06-07 15:58:05 +0000[diff] [blame]492 }
493</pre>
494<pre>
495 struct NonCopyable2 {
496 NonCopyable2();
497 NonCopyable2(NonCopyable2&);
498 };
499 void foo(const NonCopyable2&);
500 void bar() {
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]501 foo(NonCopyable2()); // Disallowed in C++98; allowed in C++11.
Jeffrey Yasskin57d12fd2010-06-07 15:58:05 +0000[diff] [blame]502 }
503</pre>
504
505<p>Note that if <tt>NonCopyable2::NonCopyable2()</tt> has a default
506argument whose instantiation produces a compile error, that error will
507still be a hard error in C++98 mode even if this warning is turned
508off.</p>
509
510</dd>
511
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]512</dl>
513
Chad Rosiercd376a12012-02-22 20:26:10 +0000[diff] [blame]514<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
515<h3 id="cl_crash_diagnostics">Options to Control Clang Crash Diagnostics</h3>
516<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
517
518<p>As unbelievable as it may sound, Clang does crash from time to time.
519Generally, this only occurs to those living on the
520<a href="http://llvm.org/releases/download.html#svn">bleeding edge</a>. Clang
521goes to great lengths to assist you in filing a bug report. Specifically, Clang
522generates preprocessed source file(s) and associated run script(s) upon a
523crash. These files should be attached to a bug report to ease reproducibility
524of the failure. Below are the command line options to control the crash
525diagnostics.
526</p>
527
528<p><b>-fno-crash-diagnostics</b>: Disable auto-generation of preprocessed
529source files during a clang crash.</p>
530
531<p>The -fno-crash-diagnostics flag can be helpful for speeding the process of
532generating a delta reduced test case.</p>
533
534
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]535<!-- ======================================================================= -->
536<h2 id="general_features">Language and Target-Independent Features</h2>
537<!-- ======================================================================= -->
538
539
540<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
541<h3 id="diagnostics">Controlling Errors and Warnings</h3>
542<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
543
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]544<p>Clang provides a number of ways to control which code constructs cause it to
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]545emit errors and warning messages, and how they are displayed to the console.</p>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]546
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]547<h4 id="diagnostics_display">Controlling How Clang Displays Diagnostics</h4>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]548
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]549<p>When Clang emits a diagnostic, it includes rich information in the output,
550and gives you fine-grain control over which information is printed. Clang has
551the ability to print this information, and these are the options that control
552it:</p>
553
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]554<ol>
555<li>A file/line/column indicator that shows exactly where the diagnostic occurs
556 in your code [<a href="#opt_fshow-column">-fshow-column</a>, <a
557 href="#opt_fshow-source-location">-fshow-source-location</a>].</li>
558<li>A categorization of the diagnostic as a note, warning, error, or fatal
559 error.</li>
560<li>A text string that describes what the problem is.</li>
561<li>An option that indicates how to control the diagnostic (for diagnostics that
562 support it) [<a
563 href="#opt_fdiagnostics-show-option">-fdiagnostics-show-option</a>].</li>
Chris Lattner3f145382010-05-24 21:35:18 +0000[diff] [blame]564<li>A <a href="#diagnostics_categories">high-level category</a> for the
565 diagnostic for clients that want to group diagnostics by class (for
566 diagnostics that support it) [<a
Chris Lattner28a43a42010-05-05 01:35:28 +0000[diff] [blame]567 href="#opt_fdiagnostics-show-category">-fdiagnostics-show-category</a>].</li>
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]568<li>The line of source code that the issue occurs on, along with a caret and
569 ranges that indicate the important locations [<a
570 href="opt_fcaret-diagnostics">-fcaret-diagnostics</a>].</li>
571<li>"FixIt" information, which is a concise explanation of how to fix the
572 problem (when Clang is certain it knows) [<a
573 href="opt_fdiagnostics-fixit-info">-fdiagnostics-fixit-info</a>].</li>
574<li>A machine-parsable representation of the ranges involved (off by
575 default) [<a
Chris Lattner2a9cc232009-04-21 05:35:32 +0000[diff] [blame]576 href="opt_fdiagnostics-print-source-range-info">-fdiagnostics-print-source-range-info</a>].</li>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]577</ol>
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]578
579<p>For more information please see <a href="#cl_diag_formatting">Formatting of
580Diagnostics</a>.</p>
581
Chris Lattner3f145382010-05-24 21:35:18 +0000[diff] [blame]582
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]583<h4 id="diagnostics_mappings">Diagnostic Mappings</h4>
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]584
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]585<p>All diagnostics are mapped into one of these 5 classes:</p>
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]586
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]587<ul>
588<li>Ignored</li>
589<li>Note</li>
590<li>Warning</li>
591<li>Error</li>
592<li>Fatal</li>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]593</ul>
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]594
Chris Lattner3f145382010-05-24 21:35:18 +0000[diff] [blame]595<h4 id="diagnostics_categories">Diagnostic Categories</h4>
596
597<p>Though not shown by default, diagnostics may each be associated with a
598 high-level category. This category is intended to make it possible to triage
599 builds that produce a large number of errors or warnings in a grouped way.
600</p>
601
602<p>Categories are not shown by default, but they can be turned on with the
603<a href="#opt_fdiagnostics-show-category">-fdiagnostics-show-category</a> option.
604When set to "<tt>name</tt>", the category is printed textually in the diagnostic
605output. When it is set to "<tt>id</tt>", a category number is printed. The
606mapping of category names to category id's can be obtained by running '<tt>clang
607 --print-diagnostic-categories</tt>'.
608</p>
609
610<h4 id="diagnostics_commandline">Controlling Diagnostics via Command Line
611 Flags</h4>
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]612
613<p>-W flags, -pedantic, etc</p>
614
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]615<h4 id="diagnostics_pragmas">Controlling Diagnostics via Pragmas</h4>
616
617<p>Clang can also control what diagnostics are enabled through the use of
618pragmas in the source code. This is useful for turning off specific warnings
619in a section of source code. Clang supports GCC's pragma for compatibility
620with existing source code, as well as several extensions. </p>
621
622<p>The pragma may control any warning that can be used from the command line.
623Warnings may be set to ignored, warning, error, or fatal. The following
624example code will tell Clang or GCC to ignore the -Wall warnings:</p>
625
626<pre>
627#pragma GCC diagnostic ignored "-Wall"
628</pre>
629
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]630<p>In addition to all of the functionality provided by GCC's pragma, Clang
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]631also allows you to push and pop the current warning state. This is particularly
632useful when writing a header file that will be compiled by other people, because
633you don't know what warning flags they build with.</p>
634
635<p>In the below example
636-Wmultichar is ignored for only a single line of code, after which the
637diagnostics return to whatever state had previously existed.</p>
638
639<pre>
640#pragma clang diagnostic push
641#pragma clang diagnostic ignored "-Wmultichar"
642
643char b = 'df'; // no warning.
644
645#pragma clang diagnostic pop
646</pre>
647
648<p>The push and pop pragmas will save and restore the full diagnostic state of
649the compiler, regardless of how it was set. That means that it is possible to
650use push and pop around GCC compatible diagnostics and Clang will push and pop
651them appropriately, while GCC will ignore the pushes and pops as unknown
652pragmas. It should be noted that while Clang supports the GCC pragma, Clang and
653GCC do not support the exact same set of warnings, so even when using GCC
654compatible #pragmas there is no guarantee that they will have identical behaviour
655on both compilers. </p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]656
Ted Kremenekf391fa72011-08-18 01:17:05 +0000[diff] [blame]657<h4 id="diagnostics_enable_everything">Enabling All Warnings</h4>
658
659<p>In addition to the traditional <tt>-W</tt> flags, one can enable <b>all</b>
660 warnings by passing <tt>-Weverything</tt>.
661 This works as expected with <tt>-Werror</tt>,
662 and also includes the warnings from <tt>-pedantic</tt>.</p>
663
664<p>Note that when combined with <tt>-w</tt> (which disables all warnings), that
665 flag wins.</p>
666
Ted Kremenek2fb11eb2010-08-24 18:12:35 +0000[diff] [blame]667<h4 id="analyzer_diagnositics">Controlling Static Analyzer Diagnostics</h4>
668
669<p>While not strictly part of the compiler, the diagnostics from Clang's <a
670href="http://clang-analyzer.llvm.org">static analyzer</a> can also be influenced
671by the user via changes to the source code. This can be done in two ways:
672
673<ul>
674
675<li id="analyzer_annotations"><b>Annotations</b>: The static analyzer recognizes various GCC-style
676attributes (e.g., <tt>__attribute__((nonnull)))</tt>) that can either suppress
677static analyzer warnings or teach the analyzer about code invariants which
678enable it to find more bugs. While many of these attributes are standard GCC
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]679attributes, additional ones have been added to Clang to specifically support the
Ted Kremenek2fb11eb2010-08-24 18:12:35 +0000[diff] [blame]680static analyzer. Detailed information on these annotations can be found in the
681<a href="http://clang-analyzer.llvm.org/annotations.html">analyzer's
682documentation</a>.</li>
683
684<li><b><tt>__clang_analyzer__</tt></b>: When the static analyzer is using Clang
685to parse source files, it implicitly defines the preprocessor macro
686<tt>__clang_analyzer__</tt>. While discouraged, code can use this macro to
687selectively exclude code the analyzer examines. Here is an example:
688
689<pre>
690#ifndef __clang_analyzer__
691// Code not to be analyzed
692#endif
693</pre>
694
695In general, this usage is discouraged. Instead, we prefer that users file bugs
696against the analyzer when it flags false positives. There is also active
697discussion of allowing users in the future to selectively silence specific
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]698analyzer warnings (some of which can already be done using <a
Chris Lattner2c2a3852011-12-19 07:49:23 +0000[diff] [blame]699href="#analyzer_annotations">annotations</a>).</li>
Ted Kremenek2fb11eb2010-08-24 18:12:35 +0000[diff] [blame]700
701</ul>
702
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]703<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
704<h3 id="precompiledheaders">Precompiled Headers</h3>
705<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
706
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]707<p><a href="http://en.wikipedia.org/wiki/Precompiled_header">Precompiled
708headers</a> are a general approach employed by many compilers to reduce
709compilation time. The underlying motivation of the approach is that it is
710common for the same (and often large) header files to be included by
711multiple source files. Consequently, compile times can often be greatly improved
712by caching some of the (redundant) work done by a compiler to process headers.
713Precompiled header files, which represent one of many ways to implement
714this optimization, are literally files that represent an on-disk cache that
715contains the vital information necessary to reduce some of the work
716needed to process a corresponding header file. While details of precompiled
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]717headers vary between compilers, precompiled headers have been shown to be
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]718highly effective at speeding up program compilation on systems with very large
719system headers (e.g., Mac OS/X).</p>
720
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]721<h4>Generating a PCH File</h4>
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]722
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]723<p>To generate a PCH file using Clang, one invokes Clang with
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]724the <b><tt>-x <i>&lt;language&gt;</i>-header</tt></b> option. This mirrors the
725interface in GCC for generating PCH files:</p>
726
727<pre>
728 $ gcc -x c-header test.h -o test.h.gch
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]729 $ clang -x c-header test.h -o test.h.pch
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]730</pre>
731
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]732<h4>Using a PCH File</h4>
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]733
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]734<p>A PCH file can then be used as a prefix header when a
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]735<b><tt>-include</tt></b> option is passed to <tt>clang</tt>:</p>
736
737<pre>
738 $ clang -include test.h test.c -o test
739</pre>
740
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]741<p>The <tt>clang</tt> driver will first check if a PCH file for <tt>test.h</tt>
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]742is available; if so, the contents of <tt>test.h</tt> (and the files it includes)
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]743will be processed from the PCH file. Otherwise, Clang falls back to
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]744directly processing the content of <tt>test.h</tt>. This mirrors the behavior of
745GCC.</p>
746
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]747<p><b>NOTE:</b> Clang does <em>not</em> automatically use PCH files
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]748for headers that are directly included within a source file. For example:</p>
749
750<pre>
Chris Lattnere42ec542009-06-13 20:35:58 +0000[diff] [blame]751 $ clang -x c-header test.h -o test.h.pch
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]752 $ cat test.c
753 #include "test.h"
754 $ clang test.c -o test
755</pre>
756
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]757<p>In this example, <tt>clang</tt> will not automatically use the PCH file for
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]758<tt>test.h</tt> since <tt>test.h</tt> was included directly in the source file
759and not specified on the command line using <tt>-include</tt>.</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]760
Douglas Gregore650c8c2009-07-07 00:12:59 +0000[diff] [blame]761<h4>Relocatable PCH Files</h4>
762<p>It is sometimes necessary to build a precompiled header from headers that
763are not yet in their final, installed locations. For example, one might build a
764precompiled header within the build tree that is then meant to be installed
765alongside the headers. Clang permits the creation of "relocatable" precompiled
766headers, which are built with a given path (into the build directory) and can
767later be used from an installed location.</p>
768
769<p>To build a relocatable precompiled header, place your headers into a
770subdirectory whose structure mimics the installed location. For example, if you
771want to build a precompiled header for the header <code>mylib.h</code> that
772will be installed into <code>/usr/include</code>, create a subdirectory
773<code>build/usr/include</code> and place the header <code>mylib.h</code> into
774that subdirectory. If <code>mylib.h</code> depends on other headers, then
775they can be stored within <code>build/usr/include</code> in a way that mimics
776the installed location.</p>
777
778<p>Building a relocatable precompiled header requires two additional arguments.
779First, pass the <code>--relocatable-pch</code> flag to indicate that the
780resulting PCH file should be relocatable. Second, pass
781<code>-isysroot /path/to/build</code>, which makes all includes for your
782library relative to the build directory. For example:</p>
783
784<pre>
785 # clang -x c-header --relocatable-pch -isysroot /path/to/build /path/to/build/mylib.h mylib.h.pch
786</pre>
787
788<p>When loading the relocatable PCH file, the various headers used in the PCH
789file are found from the system header root. For example, <code>mylib.h</code>
790can be found in <code>/usr/include/mylib.h</code>. If the headers are installed
791in some other system root, the <code>-isysroot</code> option can be used provide
792a different system root from which the headers will be based. For example,
793<code>-isysroot /Developer/SDKs/MacOSX10.4u.sdk</code> will look for
794<code>mylib.h</code> in
795<code>/Developer/SDKs/MacOSX10.4u.sdk/usr/include/mylib.h</code>.</p>
796
797<p>Relocatable precompiled headers are intended to be used in a limited number
798of cases where the compilation environment is tightly controlled and the
799precompiled header cannot be generated after headers have been installed.
800Relocatable precompiled headers also have some performance impact, because
801the difference in location between the header locations at PCH build time vs.
802at the time of PCH use requires one of the PCH optimizations,
803<code>stat()</code> caching, to be disabled. However, this change is only
804likely to affect PCH files that reference a large number of headers.</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]805
Mike Stump53664ca2009-12-14 23:53:10 +0000[diff] [blame]806<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
807<h3 id="codegen">Controlling Code Generation</h3>
808<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
809
810<p>Clang provides a number of ways to control code generation. The options are listed below.</p>
811
812<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]813<dl>
Mike Stump53664ca2009-12-14 23:53:10 +0000[diff] [blame]814<dt id="opt_fcatch-undefined-behavior"><b>-fcatch-undefined-behavior</b>: Turn
815on runtime code generation to check for undefined behavior.</dt>
816
817<dd>This option, which defaults to off, controls whether or not Clang
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]818adds runtime checks for undefined runtime behavior. If a check fails,
Mike Stump53664ca2009-12-14 23:53:10 +0000[diff] [blame]819<tt>__builtin_trap()</tt> is used to indicate failure.
820The checks are:
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]821<ul>
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]822<li>Subscripting where the static type of one operand is a variable
Mike Stump88b2a172009-12-16 03:25:12 +0000[diff] [blame]823 which is decayed from an array type and the other operand is
824 greater than the size of the array or less than zero.</li>
Mike Stump53664ca2009-12-14 23:53:10 +0000[diff] [blame]825<li>Shift operators where the amount shifted is greater or equal to the
826 promoted bit-width of the left-hand-side or less than zero.</li>
Mike Stump8f6a3ed2009-12-16 03:18:14 +0000[diff] [blame]827<li>If control flow reaches __builtin_unreachable.
828<li>When llvm implements more __builtin_object_size support, reads and
829 writes for objects that __builtin_object_size indicates we aren't
830 accessing valid memory. Bit-fields and vectors are not yet checked.
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]831</ul>
Mike Stump53664ca2009-12-14 23:53:10 +0000[diff] [blame]832</dd>
833
Kostya Serebryany48a8cd92012-03-01 23:14:19 +0000[diff] [blame]834<dt id="opt_faddress-sanitizer"><b>-f[no-]address-sanitizer</b>:
835Turn on <a href="AddressSanitizer.html">AddressSanitizer</a>,
836a memory error detector.
837
838<dt id="opt_fthread-sanitizer"><b>-f[no-]thread-sanitizer</b>:
839Turn on ThreadSanitizer, an <em>experimental</em> data race detector.
840Not ready for widespread use.
841
Nuno Lopesaa526242009-12-17 10:00:52 +0000[diff] [blame]842<dt id="opt_fno-assume-sane-operator-new"><b>-fno-assume-sane-operator-new</b>:
843Don't assume that the C++'s new operator is sane.</dt>
Nuno Lopesb23f20d2009-12-17 10:15:49 +0000[diff] [blame]844<dd>This option tells the compiler to do not assume that C++'s global new
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]845operator will always return a pointer that does not
Nuno Lopesaa526242009-12-17 10:00:52 +0000[diff] [blame]846alias any other pointer when the function returns.</dd>
Evan Chengfda026b2011-04-08 22:18:01 +0000[diff] [blame]847
Evan Cheng4cf4b592011-04-08 22:34:21 +0000[diff] [blame]848<dt id="opt_ftrap-function"><b>-ftrap-function=[name]</b>: Instruct code
Evan Chengfda026b2011-04-08 22:18:01 +0000[diff] [blame]849generator to emit a function call to the specified function name for
850<tt>__builtin_trap()</tt>.</dt>
851
852<dd>LLVM code generator translates <tt>__builtin_trap()</tt> to a trap
853instruction if it is supported by the target ISA. Otherwise, the builtin is
854translated into a call to <tt>abort</tt>. If this option is set, then the code
855generator will always lower the builtin to a call to the specified function
856regardless of whether the target ISA has a trap instruction. This option is
Chris Lattner02e266b2011-04-08 22:58:43 +0000[diff] [blame]857useful for environments (e.g. deeply embedded) where a trap cannot be properly
858handled, or when some custom behavior is desired.</dd>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]859</dl>
Nuno Lopesaa526242009-12-17 10:00:52 +0000[diff] [blame]860
Alexey Samsonov5ab52492012-05-30 06:55:10 +0000[diff] [blame^]861<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
862<h3 id="debuginfosize">Controlling Size of Debug Information</h3>
863<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
864
865<p>Debug info kind generated by Clang can be set by one of the flags listed
866below. If multiple flags are present, the last one is used.</p>
867
868<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
869<dl>
870<dt id="opt_g0"><b>-g0</b>: Don't generate any debug info (default).
871
872<dt id="opt_gline-tables-only"><b>-gline-tables-only</b>:
873Generate line number tables only.
874<dd>
875This kind of debug info allows to obtain stack traces with function
876names, file names and line numbers (by such tools as
877gdb or addr2line). It doesn't contain any other data (e.g.
878description of local variables or function parameters).
879</dd>
880
881<dt id="opt_g"><b>-g</b>: Generate complete debug info.
882</dl>
883
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]884<!-- ======================================================================= -->
885<h2 id="c">C Language Features</h2>
886<!-- ======================================================================= -->
887
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]888<p>The support for standard C in clang is feature-complete except for the C99
889floating-point pragmas.</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]890
891<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]892<h3 id="c_ext">Extensions supported by clang</h3>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]893<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
894
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]895<p>See <a href="LanguageExtensions.html">clang language extensions</a>.</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]896
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]897<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
898<h3 id="c_modes">Differences between various standard modes</h3>
899<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]900
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]901<p>clang supports the -std option, which changes what language mode clang uses.
902The supported modes for C are c89, gnu89, c94, c99, gnu99 and various aliases
903for those modes. If no -std option is specified, clang defaults to gnu99 mode.
904</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]905
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]906<p>Differences between all c* and gnu* modes:</p>
907<ul>
908<li>c* modes define "__STRICT_ANSI__".</li>
Eli Friedman26fa0ed2009-05-27 23:02:57 +0000[diff] [blame]909<li>Target-specific defines not prefixed by underscores, like "linux", are
910defined in gnu* modes.</li>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]911<li>Trigraphs default to being off in gnu* modes; they can be enabled by the
912-trigraphs option.</li>
913<li>The parser recognizes "asm" and "typeof" as keywords in gnu* modes; the
914variants "__asm__" and "__typeof__" are recognized in all modes.</li>
Eli Friedmanb0ac5452009-05-16 23:17:30 +0000[diff] [blame]915<li>The Apple "blocks" extension is recognized by default in gnu* modes
916on some platforms; it can be enabled in any mode with the "-fblocks"
917option.</li>
Chris Lattner741be6a2011-06-15 00:41:40 +0000[diff] [blame]918<li>Arrays that are VLA's according to the standard, but which can be constant
919 folded by the frontend are treated as fixed size arrays. This occurs for
920 things like "int X[(1, 2)];", which is technically a VLA. c* modes are
921 strictly compliant and treat these as VLAs.</li>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]922</ul>
923
924<p>Differences between *89 and *99 modes:</p>
925<ul>
926<li>The *99 modes default to implementing "inline" as specified in C99, while
927the *89 modes implement the GNU version. This can be overridden for individual
928functions with the __gnu_inline__ attribute.</li>
Eli Friedman26fa0ed2009-05-27 23:02:57 +0000[diff] [blame]929<li>Digraphs are not recognized in c89 mode.</li>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]930<li>The scope of names defined inside a "for", "if", "switch", "while", or "do"
931statement is different. (example: "if ((struct x {int x;}*)0) {}".)</li>
932<li>__STDC_VERSION__ is not defined in *89 modes.</li>
Eli Friedman26fa0ed2009-05-27 23:02:57 +0000[diff] [blame]933<li>"inline" is not recognized as a keyword in c89 mode.</li>
934<li>"restrict" is not recognized as a keyword in *89 modes.</li>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]935<li>Commas are allowed in integer constant expressions in *99 modes.</li>
936<li>Arrays which are not lvalues are not implicitly promoted to pointers in
937*89 modes.</li>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]938<li>Some warnings are different.</li>
939</ul>
940
941<p>c94 mode is identical to c89 mode except that digraphs are enabled in
942c94 mode (FIXME: And __STDC_VERSION__ should be defined!).</p>
943
944<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
945<h3 id="c_unimpl_gcc">GCC extensions not implemented yet</h3>
946<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
947
948<p>clang tries to be compatible with gcc as much as possible, but some gcc
949extensions are not implemented yet:</p>
950
951<ul>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]952
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]953<li>clang does not support #pragma weak
Eli Friedman4da92552009-06-02 08:21:31 +0000[diff] [blame]954(<a href="http://llvm.org/bugs/show_bug.cgi?id=3679">bug 3679</a>). Due to
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]955the uses described in the bug, this is likely to be implemented at some
956point, at least partially.</li>
957
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]958<li>clang does not support decimal floating point types (_Decimal32 and
959friends) or fixed-point types (_Fract and friends); nobody has expressed
960interest in these features yet, so it's hard to say when they will be
961implemented.</li>
962
963<li>clang does not support nested functions; this is a complex feature which
964is infrequently used, so it is unlikely to be implemented anytime soon.</li>
965
Daniel Dunbar05fa6292009-05-12 23:12:07 +0000[diff] [blame]966<li>clang does not support global register variables, this is unlikely
Eli Friedmanca130bd2009-06-12 20:04:25 +0000[diff] [blame]967to be implemented soon because it requires additional LLVM backend support.
968</li>
Daniel Dunbar05fa6292009-05-12 23:12:07 +0000[diff] [blame]969
970<li>clang does not support static initialization of flexible array
971members. This appears to be a rarely used extension, but could be
972implemented pending user demand.</li>
973
Eli Friedmanca130bd2009-06-12 20:04:25 +0000[diff] [blame]974<li>clang does not support __builtin_va_arg_pack/__builtin_va_arg_pack_len.
975This is used rarely, but in some potentially interesting places, like the
976glibc headers, so it may be implemented pending user demand. Note that
977because clang pretends to be like GCC 4.2, and this extension was introduced
978in 4.3, the glibc headers will not try to use this extension with clang at
979the moment.</li>
980
981<li>clang does not support the gcc extension for forward-declaring function
982parameters; this has not showed up in any real-world code yet, though, so it
983might never be implemented.</li>
984
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]985</ul>
986
987<p>This is not a complete list; if you find an unsupported extension
988missing from this list, please send an e-mail to cfe-dev. This list
989currently excludes C++; see <a href="#cxx">C++ Language Features</a>.
990Also, this list does not include bugs in mostly-implemented features; please
991see the <a href="http://llvm.org/bugs/buglist.cgi?quicksearch=product%3Aclang+component%3A-New%2BBugs%2CAST%2CBasic%2CDriver%2CHeaders%2CLLVM%2BCodeGen%2Cparser%2Cpreprocessor%2CSemantic%2BAnalyzer">
992bug tracker</a> for known existing bugs (FIXME: Is there a section for
993bug-reporting guidelines somewhere?).</p>
994
995<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
996<h3 id="c_unsupp_gcc">Intentionally unsupported GCC extensions</h3>
997<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
998
Eli Friedman0b326002009-06-12 20:11:05 +0000[diff] [blame]999<ul>
1000
Eli Friedman2f3fca02009-06-12 20:08:48 +0000[diff] [blame]1001<li>clang does not support the gcc extension that allows variable-length arrays
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]1002in structures. This is for a few reasons: one, it is tricky
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]1003to implement, two, the extension is completely undocumented, and three, the
Chris Lattnerb9d511c2010-10-28 02:20:32 +0000[diff] [blame]1004extension appears to be rarely used. Note that clang <em>does</em> support
1005flexible array members (arrays with a zero or unspecified size at the end of
1006a structure).</li>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]1007
Eli Friedmanca130bd2009-06-12 20:04:25 +0000[diff] [blame]1008<li>clang does not have an equivalent to gcc's "fold"; this means that
1009clang doesn't accept some constructs gcc might accept in contexts where a
Eli Friedman248f9732011-06-13 23:12:01 +0000[diff] [blame]1010constant expression is required, like "x-x" where x is a variable.</li>
Eli Friedmanca130bd2009-06-12 20:04:25 +0000[diff] [blame]1011
Eli Friedmanca130bd2009-06-12 20:04:25 +0000[diff] [blame]1012<li>clang does not support __builtin_apply and friends; this extension is
1013extremely obscure and difficult to implement reliably.</li>
1014
Eli Friedman0b326002009-06-12 20:11:05 +0000[diff] [blame]1015</ul>
1016
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]1017<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1018<h3 id="c_ms">Microsoft extensions</h3>
1019<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1020
1021<p>clang has some experimental support for extensions from
1022Microsoft Visual C++; to enable it, use the -fms-extensions command-line
Eli Friedmana1821b52009-06-08 06:21:03 +0000[diff] [blame]1023option. This is the default for Windows targets. Note that the
1024support is incomplete; enabling Microsoft extensions will silently drop
1025certain constructs (including __declspec and Microsoft-style asm statements).
1026</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]1027
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]1028<ul>
Douglas Gregord3b227d2010-12-14 16:21:49 +0000[diff] [blame]1029<li>clang allows setting _MSC_VER with -fmsc-version=. It defaults to 1300 which
Michael J. Spencerdae4ac42010-10-21 05:21:48 +0000[diff] [blame]1030is the same as Visual C/C++ 2003. Any number is supported and can greatly affect
1031what Windows SDK and c++stdlib headers clang can compile. This option will be
1032removed when clang supports the full set of MS extensions required for these
1033headers.</li>
1034
Daniel Dunbar05fa6292009-05-12 23:12:07 +0000[diff] [blame]1035<li>clang does not support the Microsoft extension where anonymous
Daniel Dunbar869e0542009-05-13 00:23:53 +0000[diff] [blame]1036record members can be declared using user defined typedefs.</li>
Daniel Dunbar05fa6292009-05-12 23:12:07 +0000[diff] [blame]1037
Daniel Dunbar9375ed12009-05-13 21:40:49 +0000[diff] [blame]1038<li>clang supports the Microsoft "#pragma pack" feature for
1039controlling record layout. GCC also contains support for this feature,
1040however where MSVC and GCC are incompatible clang follows the MSVC
1041definition.</li>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]1042</ul>
Daniel Dunbar9375ed12009-05-13 21:40:49 +0000[diff] [blame]1043
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]1044<!-- ======================================================================= -->
Richard Smithbdb10182011-11-21 20:54:59 +0000[diff] [blame]1045<h2 id="cxx">C++ Language Features</h2>
1046<!-- ======================================================================= -->
1047
1048<p>clang fully implements all of standard C++98 except for exported templates
1049(which were removed in C++11), and
1050<a href="http://clang.llvm.org/cxx_status.html">many C++11 features</a> are also
1051implemented.</p>
1052
1053<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1054<h3 id="cxx_implimits">Controlling implementation limits</h3>
1055<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1056
1057<p><b>-fconstexpr-depth=N</b>: Sets the limit for recursive constexpr function
1058invocations to N. The default is 512.</p>
1059
1060<p><b>-ftemplate-depth=N</b>: Sets the limit for recursively nested template
1061instantiations to N. The default is 1024.</p>
1062
1063<!-- ======================================================================= -->
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]1064<h2 id="target_features">Target-Specific Features and Limitations</h2>
1065<!-- ======================================================================= -->
1066
1067
1068<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1069<h3 id="target_arch">CPU Architectures Features and Limitations</h3>
1070<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1071
1072<!-- ======================== -->
1073<h4 id="target_arch_x86">X86</h4>
1074<!-- ======================== -->
Daniel Dunbarbcaf7aa2010-09-19 19:26:59 +0000[diff] [blame]1075
1076<p>The support for X86 (both 32-bit and 64-bit) is considered stable on Darwin
1077(Mac OS/X), Linux, FreeBSD, and Dragonfly BSD: it has been tested to correctly
1078compile many large C, C++, Objective-C, and Objective-C++ codebases.</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]1079
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]1080<p>On x86_64-mingw32, passing i128(by value) is incompatible to Microsoft x64
1081calling conversion. You might need to tweak WinX86_64ABIInfo::classify()
1082in lib/CodeGen/TargetInfo.cpp.</p>
NAKAMURA Takumia75fdb22011-04-04 15:02:41 +0000[diff] [blame]1083
Eli Friedman3b658d32009-06-08 05:12:39 +0000[diff] [blame]1084<!-- ======================== -->
1085<h4 id="target_arch_arm">ARM</h4>
1086<!-- ======================== -->
Daniel Dunbarbcaf7aa2010-09-19 19:26:59 +0000[diff] [blame]1087
1088<p>The support for ARM (specifically ARMv6 and ARMv7) is considered stable on
1089Darwin (iOS): it has been tested to correctly compile many large C, C++,
Bob Wilson4ea8dfa2011-01-10 17:55:14 +0000[diff] [blame]1090Objective-C, and Objective-C++ codebases. Clang only supports a limited number
1091of ARM architectures. It does not yet fully support ARMv5, for example.</p>
Eli Friedman3b658d32009-06-08 05:12:39 +0000[diff] [blame]1092
1093<!-- ======================== -->
1094<h4 id="target_arch_other">Other platforms</h4>
1095<!-- ======================== -->
1096clang currently contains some support for PPC and Sparc; however, significant
1097pieces of code generation are still missing, and they haven't undergone
1098significant testing.
1099
Eli Friedman3b658d32009-06-08 05:12:39 +0000[diff] [blame]1100<p>clang contains limited support for the MSP430 embedded processor, but both
1101the clang support and the LLVM backend support are highly experimental.
1102
1103<p>Other platforms are completely unsupported at the moment. Adding the
1104minimal support needed for parsing and semantic analysis on a new platform
1105is quite easy; see lib/Basic/Targets.cpp in the clang source tree. This level
1106of support is also sufficient for conversion to LLVM IR for simple programs.
1107Proper support for conversion to LLVM IR requires adding code to
1108lib/CodeGen/CGCall.cpp at the moment; this is likely to change soon, though.
1109Generating assembly requires a suitable LLVM backend.
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]1110
1111<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1112<h3 id="target_os">Operating System Features and Limitations</h3>
1113<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1114
1115<!-- ======================================= -->
1116<h4 id="target_os_darwin">Darwin (Mac OS/X)</h4>
1117<!-- ======================================= -->
1118
Chris Lattner396663b2012-05-04 21:11:08 +0000[diff] [blame]1119<p>None</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]1120
NAKAMURA Takumia75fdb22011-04-04 15:02:41 +0000[diff] [blame]1121<!-- ======================================= -->
1122<h4 id="target_os_win32">Windows</h4>
1123<!-- ======================================= -->
1124
1125<p>Experimental supports are on Cygming.</p>
1126
1127<h5>Cygwin</h5>
1128
1129<p>Clang works on Cygwin-1.7.</p>
1130
1131<h5>MinGW32</h5>
1132
1133<p>Clang works on some mingw32 distributions.
1134Clang assumes directories as below;</p>
1135
1136<ul>
1137<li><tt>C:/mingw/include</tt></li>
1138<li><tt>C:/mingw/lib</tt></li>
1139<li><tt>C:/mingw/lib/gcc/mingw32/4.[3-5].0/include/c++</tt></li>
1140</ul>
1141
NAKAMURA Takumi0e53df62011-11-17 07:54:13 +0000[diff] [blame]1142<p>On MSYS, a few tests might fail.</p>
NAKAMURA Takumia75fdb22011-04-04 15:02:41 +0000[diff] [blame]1143
1144<h5>MinGW-w64</h5>
1145
Douglas Gregordad879a2011-08-30 00:40:12 +0000[diff] [blame]1146<p>For 32-bit (i686-w64-mingw32), and 64-bit (x86_64-w64-mingw32), Clang assumes as below;<p>
NAKAMURA Takumia75fdb22011-04-04 15:02:41 +0000[diff] [blame]1147
1148<ul>
Douglas Gregordad879a2011-08-30 00:40:12 +0000[diff] [blame]1149<li><tt>GCC versions 4.5.0 to 4.5.3, 4.6.0 to 4.6.2, or 4.7.0 (for the C++ header search path)</tt></li>
1150<li><tt>some_directory/bin/gcc.exe</tt></li>
1151<li><tt>some_directory/bin/clang.exe</tt></li>
1152<li><tt>some_directory/bin/clang++.exe</tt></li>
1153<li><tt>some_directory/bin/../include/c++/GCC_version</tt></li>
1154<li><tt>some_directory/bin/../include/c++/GCC_version/x86_64-w64-mingw32</tt></li>
1155<li><tt>some_directory/bin/../include/c++/GCC_version/i686-w64-mingw32</tt></li>
1156<li><tt>some_directory/bin/../include/c++/GCC_version/backward</tt></li>
1157<li><tt>some_directory/bin/../x86_64-w64-mingw32/include</tt></li>
1158<li><tt>some_directory/bin/../i686-w64-mingw32/include</tt></li>
1159<li><tt>some_directory/bin/../include</tt></li>
NAKAMURA Takumia75fdb22011-04-04 15:02:41 +0000[diff] [blame]1160</ul>
1161
David Blaikiea9639622012-04-19 17:21:46 +0000[diff] [blame]1162<p>This directory layout is standard for any toolchain you will find on the official <a href="http://mingw-w64.sourceforge.net">MinGW-w64 website</a>.
Douglas Gregordad879a2011-08-30 00:40:12 +0000[diff] [blame]1163
1164<p>Clang expects the GCC executable &quot;gcc.exe&quot; compiled for i686-w64-mingw32 (or x86_64-w64-mingw32) to be present on PATH.</p>
1165
1166<p><a href="http://llvm.org/bugs/show_bug.cgi?id=9072">Some tests might fail</a>
1167on x86_64-w64-mingw32.</p>
NAKAMURA Takumia75fdb22011-04-04 15:02:41 +0000[diff] [blame]1168
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]1169</div>
1170</body>
1171</html>