Gitiles
Code Review Sign In
gerrit-public.fairphone.software / platform / external / clang / bdb10186618ebb73ebf09b44037faa6829dce0f4 / . / docs / UsersManual.html
blob: 91799f39b3dce1bcf27e841a0b2231068a0c039b [file] [log] [blame]
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]1<html>
2<head>
3<title>Clang Compiler User's Manual</title>
4<link type="text/css" rel="stylesheet" href="../menu.css" />
5<link type="text/css" rel="stylesheet" href="../content.css" />
6<style type="text/css">
7td {
8 vertical-align: top;
9}
10</style>
11</head>
12<body>
13
14<!--#include virtual="../menu.html.incl"-->
15
16<div id="content">
17
18<h1>Clang Compiler User's Manual</h1>
19
20<ul>
21<li><a href="#intro">Introduction</a>
22 <ul>
23 <li><a href="#terminology">Terminology</a></li>
24 <li><a href="#basicusage">Basic Usage</a></li>
25 </ul>
26</li>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]27<li><a href="#commandline">Command Line Options</a>
28 <ul>
29 <li><a href="#cl_diagnostics">Options to Control Error and Warning
30 Messages</a></li>
31 </ul>
32</li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]33<li><a href="#general_features">Language and Target-Independent Features</a>
34 <ul>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]35 <li><a href="#diagnostics">Controlling Errors and Warnings</a>
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]36 <ul>
37 <li><a href="#diagnostics_display">Controlling How Clang Displays Diagnostics</a></li>
38 <li><a href="#diagnostics_mappings">Diagnostic Mappings</a></li>
Chris Lattner67db8cd2010-05-30 23:42:51 +0000[diff] [blame]39 <li><a href="#diagnostics_categories">Diagnostic Categories</a></li>
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]40 <li><a href="#diagnostics_commandline">Controlling Diagnostics via Command Line Flags</a></li>
41 <li><a href="#diagnostics_pragmas">Controlling Diagnostics via Pragmas</a></li>
Ted Kremenekf391fa72011-08-18 01:17:05 +0000[diff] [blame]42 <li><a href="#diagnostics_enable_everything">Enabling All Warnings</a></li>
Ted Kremenek2fb11eb2010-08-24 18:12:35 +0000[diff] [blame]43 <li><a href="#analyzer_diagnositics">Controlling Static Analyzer Diagnostics</a></li>
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]44 </ul>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]45 </li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]46 <li><a href="#precompiledheaders">Precompiled Headers</a></li>
Mike Stump53664ca2009-12-14 23:53:10 +0000[diff] [blame]47 <li><a href="#codegen">Controlling Code Generation</a></li>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]48 </ul>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]49</li>
50<li><a href="#c">C Language Features</a>
51 <ul>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]52 <li><a href="#c_ext">Extensions supported by clang</a></li>
53 <li><a href="#c_modes">Differences between various standard modes</a></li>
54 <li><a href="#c_unimpl_gcc">GCC extensions not implemented yet</a></li>
55 <li><a href="#c_unsupp_gcc">Intentionally unsupported GCC extensions</a></li>
56 <li><a href="#c_ms">Microsoft extensions</a></li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]57 </ul>
58</li>
Richard Smithbdb10182011-11-21 20:54:59 +0000[diff] [blame^]59<li><a href="#cxx">C++ Language Features</a>
60 <ul>
61 <li><a href="#cxx_implimits">Controlling implementation limits</a></li>
62 </ul>
63</li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]64<li><a href="#target_features">Target-Specific Features and Limitations</a>
65 <ul>
66 <li><a href="#target_arch">CPU Architectures Features and Limitations</a>
67 <ul>
68 <li><a href="#target_arch_x86">X86</a></li>
Eli Friedman3b658d32009-06-08 05:12:39 +0000[diff] [blame]69 <li><a href="#target_arch_arm">ARM</a></li>
70 <li><a href="#target_arch_other">Other platforms</a></li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]71 </ul>
72 </li>
73 <li><a href="#target_os">Operating System Features and Limitations</a>
74 <ul>
75 <li><a href="#target_os_darwin">Darwin (Mac OS/X)</a></li>
76 <li>Linux, etc.</li>
NAKAMURA Takumia75fdb22011-04-04 15:02:41 +0000[diff] [blame]77 <li><a href="#target_os_win32">Windows</a></li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]78 </ul>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]79 </li>
80 </ul>
81</li>
82</ul>
83
84
85<!-- ======================================================================= -->
86<h2 id="intro">Introduction</h2>
87<!-- ======================================================================= -->
88
89<p>The Clang Compiler is an open-source compiler for the C family of programming
90languages, aiming to be the best in class implementation of these languages.
91Clang builds on the LLVM optimizer and code generator, allowing it to provide
92high-quality optimization and code generation support for many targets. For
93more general information, please see the <a href="http://clang.llvm.org">Clang
94Web Site</a> or the <a href="http://llvm.org">LLVM Web Site</a>.</p>
95
96<p>This document describes important notes about using Clang as a compiler for
97an end-user, documenting the supported features, command line options, etc. If
98you are interested in using Clang to build a tool that processes code, please
99see <a href="InternalsManual.html">the Clang Internals Manual</a>. If you are
100interested in the <a href="http://clang.llvm.org/StaticAnalysis.html">Clang
101Static Analyzer</a>, please see its web page.</p>
102
103<p>Clang is designed to support the C family of programming languages, which
104includes <a href="#c">C</a>, <a href="#objc">Objective-C</a>, <a
105href="#cxx">C++</a>, and <a href="#objcxx">Objective-C++</a> as well as many
106dialects of those. For language-specific information, please see the
107corresponding language specific section:</p>
108
109<ul>
110<li><a href="#c">C Language</a>: K&amp;R C, ANSI C89, ISO C90, ISO C94
111 (C89+AMD1), ISO C99 (+TC1, TC2, TC3). </li>
112<li><a href="#objc">Objective-C Language</a>: ObjC 1, ObjC 2, ObjC 2.1, plus
113 variants depending on base language.</li>
Richard Smithbdb10182011-11-21 20:54:59 +0000[diff] [blame^]114<li><a href="#cxx">C++ Language</a></li>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]115<li><a href="#objcxx">Objective C++ Language</a></li>
116</ul>
117
118<p>In addition to these base languages and their dialects, Clang supports a
119broad variety of language extensions, which are documented in the corresponding
120language section. These extensions are provided to be compatible with the GCC,
121Microsoft, and other popular compilers as well as to improve functionality
122through Clang-specific features. The Clang driver and language features are
123intentionally designed to be as compatible with the GNU GCC compiler as
124reasonably possible, easing migration from GCC to Clang. In most cases, code
125"just works".</p>
126
127<p>In addition to language specific features, Clang has a variety of features
128that depend on what CPU architecture or operating system is being compiled for.
Douglas Gregorcd5a5052009-11-09 15:15:41 +0000[diff] [blame]129Please see the <a href="#target_features">Target-Specific Features and
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]130Limitations</a> section for more details.</p>
131
132<p>The rest of the introduction introduces some basic <a
133href="#terminology">compiler terminology</a> that is used throughout this manual
134and contains a basic <a href="#basicusage">introduction to using Clang</a>
135as a command line compiler.</p>
136
137<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
138<h3 id="terminology">Terminology</h3>
139<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
140
141<p>Front end, parser, backend, preprocessor, undefined behavior, diagnostic,
142 optimizer</p>
143
144<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
145<h3 id="basicusage">Basic Usage</h3>
146<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
147
148<p>Intro to how to use a C compiler for newbies.</p>
149<p>
150compile + link
151
152compile then link
153
154debug info
155
156enabling optimizations
157
158picking a language to use, defaults to C99 by default. Autosenses based on
159extension.
160
161using a makefile
162</p>
163
164
165<!-- ======================================================================= -->
166<h2 id="commandline">Command Line Options</h2>
167<!-- ======================================================================= -->
168
169<p>
170This section is generally an index into other sections. It does not go into
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]171depth on the ones that are covered by other sections. However, the first part
172introduces the language selection and other high level options like -c, -g, etc.
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]173</p>
174
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]175
176<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
177<h3 id="cl_diagnostics">Options to Control Error and Warning Messages</h3>
178<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
179
180<p><b>-Werror</b>: Turn warnings into errors.</p>
181<p><b>-Werror=foo</b>: Turn warning "foo" into an error.</p>
182<p><b>-Wno-error=foo</b>: Turn warning "foo" into an warning even if -Werror is
183 specified.</p>
184<p><b>-Wfoo</b>: Enable warning foo</p>
185<p><b>-Wno-foo</b>: Disable warning foo</p>
186<p><b>-w</b>: Disable all warnings.</p>
187<p><b>-pedantic</b>: Warn on language extensions.</p>
188<p><b>-pedantic-errors</b>: Error on language extensions.</p>
189<p><b>-Wsystem-headers</b>: Enable warnings from system headers.</p>
190
Chris Lattner0f0c9632010-04-07 20:49:23 +0000[diff] [blame]191<p><b>-ferror-limit=123</b>: Stop emitting diagnostics after 123 errors have
192 been produced. The default is 20, and the error limit can be disabled with
193 -ferror-limit=0.</p>
194
Douglas Gregor575cf372010-04-20 07:18:24 +0000[diff] [blame]195<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>
196
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]197<!-- ================================================= -->
198<h4 id="cl_diag_formatting">Formatting of Diagnostics</h4>
199<!-- ================================================= -->
200
201<p>Clang aims to produce beautiful diagnostics by default, particularly for new
202users that first come to Clang. However, different people have different
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]203preferences, and sometimes Clang is driven by another program that wants to
204parse simple and consistent output, not a person. For these cases, Clang
205provides a wide range of options to control the exact output format of the
206diagnostics that it generates.</p>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]207
208<dl>
209
210<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
211<dt id="opt_fshow-column"><b>-f[no-]show-column</b>: Print column number in
212diagnostic.</dt>
213<dd>This option, which defaults to on, controls whether or not Clang prints the
214column number of a diagnostic. For example, when this is enabled, Clang will
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]215print something like:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]216
217<pre>
218 test.c:28:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
219 #endif bad
220 ^
221 //
222</pre>
223
224<p>When this is disabled, Clang will print "test.c:28: warning..." with no
225column number.</p>
226</dd>
227
228<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
229<dt id="opt_fshow-source-location"><b>-f[no-]show-source-location</b>: Print
230source file/line/column information in diagnostic.</dt>
231<dd>This option, which defaults to on, controls whether or not Clang prints the
232filename, line number and column number of a diagnostic. For example,
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]233when this is enabled, Clang will print something like:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]234
235<pre>
236 test.c:28:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
237 #endif bad
238 ^
239 //
240</pre>
241
242<p>When this is disabled, Clang will not print the "test.c:28:8: " part.</p>
243</dd>
244
245<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
246<dt id="opt_fcaret-diagnostics"><b>-f[no-]caret-diagnostics</b>: Print source
247line and ranges from source code in diagnostic.</dt>
248<dd>This option, which defaults to on, controls whether or not Clang prints the
249source line, source ranges, and caret when emitting a diagnostic. For example,
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]250when this is enabled, Clang will print something like:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]251
252<pre>
253 test.c:28:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
254 #endif bad
255 ^
256 //
257</pre>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]258</dd>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]259<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
Douglas Gregore8d44dd2010-07-09 16:31:58 +0000[diff] [blame]260<dt id="opt_fcolor_diagnostics"><b>-f[no-]color-diagnostics</b>: </dt>
261<dd>This option, which defaults to on when a color-capable terminal is
262 detected, controls whether or not Clang prints diagnostics in color.
263 When this option is enabled, Clang will use colors to highlight
264 specific parts of the diagnostic, e.g.,
Chris Lattner4e1c53d2011-01-24 03:47:34 +0000[diff] [blame]265 <pre>
266 <b><font color="black">test.c:28:8: <font color="magenta">warning</font>: extra tokens at end of #endif directive [-Wextra-tokens]</font></b>
267 #endif bad
268 <font color="green">^</font>
269 <font color="green">//</font>
Douglas Gregore8d44dd2010-07-09 16:31:58 +0000[diff] [blame]270</pre>
Chris Lattner4e1c53d2011-01-24 03:47:34 +0000[diff] [blame]271
272<p>When this is disabled, Clang will just print:</p>
273
274<pre>
275 test.c:2:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
276 #endif bad
277 ^
278 //
279</pre>
280</dd>
Douglas Gregore8d44dd2010-07-09 16:31:58 +0000[diff] [blame]281<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
Douglas Gregorc9471b02011-05-21 17:07:29 +0000[diff] [blame]282<dt id="opt_fdiagnostics-format"><b>-fdiagnostics-format=clang/msvc/vi</b>:
283Changes diagnostic output format to better match IDEs and command line tools.</dt>
284<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:
285
286 <dl>
287 <dt><b>clang</b> (default)</dt>
288 <dd>
289 <pre>t.c:3:11: warning: conversion specifies type 'char *' but the argument has type 'int'</pre>
290 </dd>
291
292 <dt><b>msvc</b></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>vi</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 </dl>
302</dd>
303
304<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
Argyrios Kyrtzidis477aab62011-05-25 05:05:01 +0000[diff] [blame]305<dt id="opt_fdiagnostics-show-name"><b>-f[no-]diagnostics-show-name</b>:
306Enable the display of the diagnostic name.</dt>
307<dd>This option, which defaults to off, controls whether or not
308Clang prints the associated name.</dd>
309<br>
310<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]311<dt id="opt_fdiagnostics-show-option"><b>-f[no-]diagnostics-show-option</b>:
312Enable <tt>[-Woption]</tt> information in diagnostic line.</dt>
313<dd>This option, which defaults to on,
314controls whether or not Clang prints the associated <A
315href="#cl_diag_warning_groups">warning group</a> option name when outputting
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]316a warning diagnostic. For example, in this output:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]317
318<pre>
319 test.c:28:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
320 #endif bad
321 ^
322 //
323</pre>
324
325<p>Passing <b>-fno-diagnostics-show-option</b> will prevent Clang from printing
326the [<a href="#opt_Wextra-tokens">-Wextra-tokens</a>] information in the
327diagnostic. This information tells you the flag needed to enable or disable the
328diagnostic, either from the command line or through <a
329href="#pragma_GCC_diagnostic">#pragma GCC diagnostic</a>.</dd>
330
Chris Lattner28a43a42010-05-05 01:35:28 +0000[diff] [blame]331<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
332<dt id="opt_fdiagnostics-show-category"><b>-fdiagnostics-show-category=none/id/name</b>:
333Enable printing category information in diagnostic line.</dt>
334<dd>This option, which defaults to "none",
335controls whether or not Clang prints the category associated with a diagnostic
336when emitting it. Each diagnostic may or many not have an associated category,
337if it has one, it is listed in the diagnostic categorization field of the
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]338diagnostic line (in the []'s).
Chris Lattner28a43a42010-05-05 01:35:28 +0000[diff] [blame]339
340<p>For example, a format string warning will produce these three renditions
341based on the setting of this option:</p>
342
343<pre>
344 t.c:3:11: warning: conversion specifies type 'char *' but the argument has type 'int' [-Wformat]
345 t.c:3:11: warning: conversion specifies type 'char *' but the argument has type 'int' [-Wformat<b>,1</b>]
346 t.c:3:11: warning: conversion specifies type 'char *' but the argument has type 'int' [-Wformat<b>,Format String</b>]
347</pre>
348
349<p>This category can be used by clients that want to group diagnostics by
350category, so it should be a high level category. We want dozens of these, not
351hundreds or thousands of them.</p>
352</dd>
353
354
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]355
356<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
357<dt id="opt_fdiagnostics-fixit-info"><b>-f[no-]diagnostics-fixit-info</b>:
358Enable "FixIt" information in the diagnostics output.</dt>
359<dd>This option, which defaults to on, controls whether or not Clang prints the
360information on how to fix a specific diagnostic underneath it when it knows.
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]361For example, in this output:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]362
363<pre>
364 test.c:28:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
365 #endif bad
366 ^
367 //
368</pre>
369
370<p>Passing <b>-fno-diagnostics-fixit-info</b> will prevent Clang from printing
371the "//" line at the end of the message. This information is useful for users
372who may not understand what is wrong, but can be confusing for machine
373parsing.</p>
374</dd>
375
376<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
Chris Lattner2a9cc232009-04-21 05:35:32 +0000[diff] [blame]377<dt id="opt_fdiagnostics-print-source-range-info">
378<b>-f[no-]diagnostics-print-source-range-info</b>:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]379Print machine parsable information about source ranges.</dt>
380<dd>This option, which defaults to off, controls whether or not Clang prints
381information about source ranges in a machine parsable format after the
382file/line/column number information. The information is a simple sequence of
383brace enclosed ranges, where each range lists the start and end line/column
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]384locations. For example, in this output:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]385
386<pre>
387exprs.c:47:15:{47:8-47:14}{47:17-47:24}: error: invalid operands to binary expression ('int *' and '_Complex float')
388 P = (P-42) + Gamma*4;
389 ~~~~~~ ^ ~~~~~~~
390</pre>
391
Chris Lattner2a9cc232009-04-21 05:35:32 +0000[diff] [blame]392<p>The {}'s are generated by -fdiagnostics-print-source-range-info.</p>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]393</dd>
394
Douglas Gregor4786c152010-08-19 20:24:43 +0000[diff] [blame]395<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
396<dt id="opt_fdiagnostics-parseable-fixits">
397<b>-fdiagnostics-parseable-fixits</b>:
398Print Fix-Its in a machine parseable form.</dt>
399<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>
400
401<pre>
Douglas Gregorbf5e09d2010-08-20 03:17:33 +0000[diff] [blame]402 fix-it:"t.cpp":{7:25-7:29}:"Gamma"
Douglas Gregor4786c152010-08-19 20:24:43 +0000[diff] [blame]403</pre>
404
NAKAMURA Takumi44626362011-04-05 00:57:02 +0000[diff] [blame]405<p>The range printed is a half-open range, so in this example the characters at
406column 25 up to but not including column 29 on line 7 in t.cpp should be
407replaced with the string &quot;Gamma&quot;. Either the range or the replacement
408string may be empty (representing strict insertions and strict erasures,
409respectively). Both the file name and the insertion string escape backslash (as
410&quot;\\&quot;), tabs (as &quot;\t&quot;), newlines (as &quot;\n&quot;), double
411quotes(as &quot;\&quot;&quot;) and non-printable characters (as octal
412&quot;\xxx&quot;).</p>
Douglas Gregor4786c152010-08-19 20:24:43 +0000[diff] [blame]413</dd>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]414
415</dl>
416
417
418
419
420<!-- ===================================================== -->
421<h4 id="cl_diag_warning_groups">Individual Warning Groups</h4>
422<!-- ===================================================== -->
423
424<p>TODO: Generate this from tblgen. Define one anchor per warning group.</p>
425
426
427<dl>
428
429
430<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
431<dt id="opt_Wextra-tokens"><b>-Wextra-tokens</b>: Warn about excess tokens at
432 the end of a preprocessor directive.</dt>
433<dd>This option, which defaults to on, enables warnings about extra tokens at
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]434the end of preprocessor directives. For example:
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]435
436<pre>
437 test.c:28:8: warning: extra tokens at end of #endif directive [-Wextra-tokens]
438 #endif bad
439 ^
440</pre>
441
442<p>These extra tokens are not strictly conforming, and are usually best handled
443by commenting them out.</p>
444
445<p>This option is also enabled by <a href="">-Wfoo</a>, <a href="">-Wbar</a>,
446 and <a href="">-Wbaz</a>.</p>
447</dd>
448
Jeffrey Yasskin21d07e42010-06-05 01:39:57 +0000[diff] [blame]449<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
450<dt id="opt_Wambiguous-member-template"><b>-Wambiguous-member-template</b>:
451Warn about unqualified uses of a member template whose name resolves
452to another template at the location of the use.</dt>
453<dd>This option, which defaults to on, enables a warning in the
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]454following code:
Jeffrey Yasskin21d07e42010-06-05 01:39:57 +0000[diff] [blame]455
456<pre>
457template&lt;typename T> struct set{};
458template&lt;typename T> struct trait { typedef const T& type; };
459struct Value {
460 template&lt;typename T> void set(typename trait&lt;T>::type value) {}
461};
462void foo() {
463 Value v;
464 v.set&lt;double>(3.2);
465}
466</pre>
467
468<p>C++ [basic.lookup.classref] requires this to be an error, but,
469because it's hard to work around, Clang downgrades it to a warning as
470an extension.</p>
471</dd>
472
Jeffrey Yasskin57d12fd2010-06-07 15:58:05 +0000[diff] [blame]473<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
474<dt id="opt_Wbind-to-temporary-copy"><b>-Wbind-to-temporary-copy</b>: Warn about
475an unusable copy constructor when binding a reference to a temporary.</dt>
476<dd>This option, which defaults to on, enables warnings about binding a
477reference to a temporary when the temporary doesn't have a usable copy
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]478constructor. For example:
Jeffrey Yasskin57d12fd2010-06-07 15:58:05 +0000[diff] [blame]479
480<pre>
481 struct NonCopyable {
482 NonCopyable();
483 private:
484 NonCopyable(const NonCopyable&);
485 };
486 void foo(const NonCopyable&);
487 void bar() {
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]488 foo(NonCopyable()); // Disallowed in C++98; allowed in C++11.
Jeffrey Yasskin57d12fd2010-06-07 15:58:05 +0000[diff] [blame]489 }
490</pre>
491<pre>
492 struct NonCopyable2 {
493 NonCopyable2();
494 NonCopyable2(NonCopyable2&);
495 };
496 void foo(const NonCopyable2&);
497 void bar() {
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]498 foo(NonCopyable2()); // Disallowed in C++98; allowed in C++11.
Jeffrey Yasskin57d12fd2010-06-07 15:58:05 +0000[diff] [blame]499 }
500</pre>
501
502<p>Note that if <tt>NonCopyable2::NonCopyable2()</tt> has a default
503argument whose instantiation produces a compile error, that error will
504still be a hard error in C++98 mode even if this warning is turned
505off.</p>
506
507</dd>
508
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]509</dl>
510
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]511<!-- ======================================================================= -->
512<h2 id="general_features">Language and Target-Independent Features</h2>
513<!-- ======================================================================= -->
514
515
516<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
517<h3 id="diagnostics">Controlling Errors and Warnings</h3>
518<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
519
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]520<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]521emit errors and warning messages, and how they are displayed to the console.</p>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]522
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]523<h4 id="diagnostics_display">Controlling How Clang Displays Diagnostics</h4>
Chris Lattner65a795b2009-04-20 06:00:23 +0000[diff] [blame]524
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]525<p>When Clang emits a diagnostic, it includes rich information in the output,
526and gives you fine-grain control over which information is printed. Clang has
527the ability to print this information, and these are the options that control
528it:</p>
529
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]530<ol>
531<li>A file/line/column indicator that shows exactly where the diagnostic occurs
532 in your code [<a href="#opt_fshow-column">-fshow-column</a>, <a
533 href="#opt_fshow-source-location">-fshow-source-location</a>].</li>
534<li>A categorization of the diagnostic as a note, warning, error, or fatal
535 error.</li>
536<li>A text string that describes what the problem is.</li>
Argyrios Kyrtzidis477aab62011-05-25 05:05:01 +0000[diff] [blame]537<li>An option that indicates whether to print the diagnostic name [<a
538 href="#opt_fdiagnostics-show-name">-fdiagnostics-show-name</a>].</li>
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]539<li>An option that indicates how to control the diagnostic (for diagnostics that
540 support it) [<a
541 href="#opt_fdiagnostics-show-option">-fdiagnostics-show-option</a>].</li>
Chris Lattner3f145382010-05-24 21:35:18 +0000[diff] [blame]542<li>A <a href="#diagnostics_categories">high-level category</a> for the
543 diagnostic for clients that want to group diagnostics by class (for
544 diagnostics that support it) [<a
Chris Lattner28a43a42010-05-05 01:35:28 +0000[diff] [blame]545 href="#opt_fdiagnostics-show-category">-fdiagnostics-show-category</a>].</li>
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]546<li>The line of source code that the issue occurs on, along with a caret and
547 ranges that indicate the important locations [<a
548 href="opt_fcaret-diagnostics">-fcaret-diagnostics</a>].</li>
549<li>"FixIt" information, which is a concise explanation of how to fix the
550 problem (when Clang is certain it knows) [<a
551 href="opt_fdiagnostics-fixit-info">-fdiagnostics-fixit-info</a>].</li>
552<li>A machine-parsable representation of the ranges involved (off by
553 default) [<a
Chris Lattner2a9cc232009-04-21 05:35:32 +0000[diff] [blame]554 href="opt_fdiagnostics-print-source-range-info">-fdiagnostics-print-source-range-info</a>].</li>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]555</ol>
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]556
557<p>For more information please see <a href="#cl_diag_formatting">Formatting of
558Diagnostics</a>.</p>
559
Chris Lattner3f145382010-05-24 21:35:18 +0000[diff] [blame]560
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]561<h4 id="diagnostics_mappings">Diagnostic Mappings</h4>
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]562
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]563<p>All diagnostics are mapped into one of these 5 classes:</p>
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]564
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]565<ul>
566<li>Ignored</li>
567<li>Note</li>
568<li>Warning</li>
569<li>Error</li>
570<li>Fatal</li>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]571</ul>
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]572
Chris Lattner3f145382010-05-24 21:35:18 +0000[diff] [blame]573<h4 id="diagnostics_categories">Diagnostic Categories</h4>
574
575<p>Though not shown by default, diagnostics may each be associated with a
576 high-level category. This category is intended to make it possible to triage
577 builds that produce a large number of errors or warnings in a grouped way.
578</p>
579
580<p>Categories are not shown by default, but they can be turned on with the
581<a href="#opt_fdiagnostics-show-category">-fdiagnostics-show-category</a> option.
582When set to "<tt>name</tt>", the category is printed textually in the diagnostic
583output. When it is set to "<tt>id</tt>", a category number is printed. The
584mapping of category names to category id's can be obtained by running '<tt>clang
585 --print-diagnostic-categories</tt>'.
586</p>
587
588<h4 id="diagnostics_commandline">Controlling Diagnostics via Command Line
589 Flags</h4>
Chris Lattner8217f4e2009-04-20 06:26:18 +0000[diff] [blame]590
591<p>-W flags, -pedantic, etc</p>
592
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]593<h4 id="diagnostics_pragmas">Controlling Diagnostics via Pragmas</h4>
594
595<p>Clang can also control what diagnostics are enabled through the use of
596pragmas in the source code. This is useful for turning off specific warnings
597in a section of source code. Clang supports GCC's pragma for compatibility
598with existing source code, as well as several extensions. </p>
599
600<p>The pragma may control any warning that can be used from the command line.
601Warnings may be set to ignored, warning, error, or fatal. The following
602example code will tell Clang or GCC to ignore the -Wall warnings:</p>
603
604<pre>
605#pragma GCC diagnostic ignored "-Wall"
606</pre>
607
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]608<p>In addition to all of the functionality provided by GCC's pragma, Clang
Chris Lattner3401cf82009-07-12 21:22:10 +0000[diff] [blame]609also allows you to push and pop the current warning state. This is particularly
610useful when writing a header file that will be compiled by other people, because
611you don't know what warning flags they build with.</p>
612
613<p>In the below example
614-Wmultichar is ignored for only a single line of code, after which the
615diagnostics return to whatever state had previously existed.</p>
616
617<pre>
618#pragma clang diagnostic push
619#pragma clang diagnostic ignored "-Wmultichar"
620
621char b = 'df'; // no warning.
622
623#pragma clang diagnostic pop
624</pre>
625
626<p>The push and pop pragmas will save and restore the full diagnostic state of
627the compiler, regardless of how it was set. That means that it is possible to
628use push and pop around GCC compatible diagnostics and Clang will push and pop
629them appropriately, while GCC will ignore the pushes and pops as unknown
630pragmas. It should be noted that while Clang supports the GCC pragma, Clang and
631GCC do not support the exact same set of warnings, so even when using GCC
632compatible #pragmas there is no guarantee that they will have identical behaviour
633on both compilers. </p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]634
Ted Kremenekf391fa72011-08-18 01:17:05 +0000[diff] [blame]635<h4 id="diagnostics_enable_everything">Enabling All Warnings</h4>
636
637<p>In addition to the traditional <tt>-W</tt> flags, one can enable <b>all</b>
638 warnings by passing <tt>-Weverything</tt>.
639 This works as expected with <tt>-Werror</tt>,
640 and also includes the warnings from <tt>-pedantic</tt>.</p>
641
642<p>Note that when combined with <tt>-w</tt> (which disables all warnings), that
643 flag wins.</p>
644
Ted Kremenek2fb11eb2010-08-24 18:12:35 +0000[diff] [blame]645<h4 id="analyzer_diagnositics">Controlling Static Analyzer Diagnostics</h4>
646
647<p>While not strictly part of the compiler, the diagnostics from Clang's <a
648href="http://clang-analyzer.llvm.org">static analyzer</a> can also be influenced
649by the user via changes to the source code. This can be done in two ways:
650
651<ul>
652
653<li id="analyzer_annotations"><b>Annotations</b>: The static analyzer recognizes various GCC-style
654attributes (e.g., <tt>__attribute__((nonnull)))</tt>) that can either suppress
655static analyzer warnings or teach the analyzer about code invariants which
656enable it to find more bugs. While many of these attributes are standard GCC
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]657attributes, additional ones have been added to Clang to specifically support the
Ted Kremenek2fb11eb2010-08-24 18:12:35 +0000[diff] [blame]658static analyzer. Detailed information on these annotations can be found in the
659<a href="http://clang-analyzer.llvm.org/annotations.html">analyzer's
660documentation</a>.</li>
661
662<li><b><tt>__clang_analyzer__</tt></b>: When the static analyzer is using Clang
663to parse source files, it implicitly defines the preprocessor macro
664<tt>__clang_analyzer__</tt>. While discouraged, code can use this macro to
665selectively exclude code the analyzer examines. Here is an example:
666
667<pre>
668#ifndef __clang_analyzer__
669// Code not to be analyzed
670#endif
671</pre>
672
673In general, this usage is discouraged. Instead, we prefer that users file bugs
674against the analyzer when it flags false positives. There is also active
675discussion of allowing users in the future to selectively silence specific
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]676analyzer warnings (some of which can already be done using <a
Ted Kremenek2fb11eb2010-08-24 18:12:35 +0000[diff] [blame]677href="analyzer_annotations">annotations</a>).</li>
678
679</ul>
680
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]681<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
682<h3 id="precompiledheaders">Precompiled Headers</h3>
683<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
684
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]685<p><a href="http://en.wikipedia.org/wiki/Precompiled_header">Precompiled
686headers</a> are a general approach employed by many compilers to reduce
687compilation time. The underlying motivation of the approach is that it is
688common for the same (and often large) header files to be included by
689multiple source files. Consequently, compile times can often be greatly improved
690by caching some of the (redundant) work done by a compiler to process headers.
691Precompiled header files, which represent one of many ways to implement
692this optimization, are literally files that represent an on-disk cache that
693contains the vital information necessary to reduce some of the work
694needed to process a corresponding header file. While details of precompiled
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]695headers vary between compilers, precompiled headers have been shown to be
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]696highly effective at speeding up program compilation on systems with very large
697system headers (e.g., Mac OS/X).</p>
698
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]699<h4>Generating a PCH File</h4>
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]700
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]701<p>To generate a PCH file using Clang, one invokes Clang with
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]702the <b><tt>-x <i>&lt;language&gt;</i>-header</tt></b> option. This mirrors the
703interface in GCC for generating PCH files:</p>
704
705<pre>
706 $ gcc -x c-header test.h -o test.h.gch
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]707 $ clang -x c-header test.h -o test.h.pch
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]708</pre>
709
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]710<h4>Using a PCH File</h4>
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]711
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]712<p>A PCH file can then be used as a prefix header when a
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]713<b><tt>-include</tt></b> option is passed to <tt>clang</tt>:</p>
714
715<pre>
716 $ clang -include test.h test.c -o test
717</pre>
718
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]719<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]720is 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]721will be processed from the PCH file. Otherwise, Clang falls back to
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]722directly processing the content of <tt>test.h</tt>. This mirrors the behavior of
723GCC.</p>
724
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]725<p><b>NOTE:</b> Clang does <em>not</em> automatically use PCH files
Chris Lattner5c3074f2009-04-20 04:37:38 +0000[diff] [blame]726for headers that are directly included within a source file. For example:</p>
727
728<pre>
Chris Lattnere42ec542009-06-13 20:35:58 +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 $ cat test.c
731 #include "test.h"
732 $ clang test.c -o test
733</pre>
734
Douglas Gregorf4d59532009-06-03 22:37:00 +0000[diff] [blame]735<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]736<tt>test.h</tt> since <tt>test.h</tt> was included directly in the source file
737and not specified on the command line using <tt>-include</tt>.</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]738
Douglas Gregore650c8c2009-07-07 00:12:59 +0000[diff] [blame]739<h4>Relocatable PCH Files</h4>
740<p>It is sometimes necessary to build a precompiled header from headers that
741are not yet in their final, installed locations. For example, one might build a
742precompiled header within the build tree that is then meant to be installed
743alongside the headers. Clang permits the creation of "relocatable" precompiled
744headers, which are built with a given path (into the build directory) and can
745later be used from an installed location.</p>
746
747<p>To build a relocatable precompiled header, place your headers into a
748subdirectory whose structure mimics the installed location. For example, if you
749want to build a precompiled header for the header <code>mylib.h</code> that
750will be installed into <code>/usr/include</code>, create a subdirectory
751<code>build/usr/include</code> and place the header <code>mylib.h</code> into
752that subdirectory. If <code>mylib.h</code> depends on other headers, then
753they can be stored within <code>build/usr/include</code> in a way that mimics
754the installed location.</p>
755
756<p>Building a relocatable precompiled header requires two additional arguments.
757First, pass the <code>--relocatable-pch</code> flag to indicate that the
758resulting PCH file should be relocatable. Second, pass
759<code>-isysroot /path/to/build</code>, which makes all includes for your
760library relative to the build directory. For example:</p>
761
762<pre>
763 # clang -x c-header --relocatable-pch -isysroot /path/to/build /path/to/build/mylib.h mylib.h.pch
764</pre>
765
766<p>When loading the relocatable PCH file, the various headers used in the PCH
767file are found from the system header root. For example, <code>mylib.h</code>
768can be found in <code>/usr/include/mylib.h</code>. If the headers are installed
769in some other system root, the <code>-isysroot</code> option can be used provide
770a different system root from which the headers will be based. For example,
771<code>-isysroot /Developer/SDKs/MacOSX10.4u.sdk</code> will look for
772<code>mylib.h</code> in
773<code>/Developer/SDKs/MacOSX10.4u.sdk/usr/include/mylib.h</code>.</p>
774
775<p>Relocatable precompiled headers are intended to be used in a limited number
776of cases where the compilation environment is tightly controlled and the
777precompiled header cannot be generated after headers have been installed.
778Relocatable precompiled headers also have some performance impact, because
779the difference in location between the header locations at PCH build time vs.
780at the time of PCH use requires one of the PCH optimizations,
781<code>stat()</code> caching, to be disabled. However, this change is only
782likely to affect PCH files that reference a large number of headers.</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]783
Mike Stump53664ca2009-12-14 23:53:10 +0000[diff] [blame]784<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
785<h3 id="codegen">Controlling Code Generation</h3>
786<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
787
788<p>Clang provides a number of ways to control code generation. The options are listed below.</p>
789
790<!-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -->
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]791<dl>
Mike Stump53664ca2009-12-14 23:53:10 +0000[diff] [blame]792<dt id="opt_fcatch-undefined-behavior"><b>-fcatch-undefined-behavior</b>: Turn
793on runtime code generation to check for undefined behavior.</dt>
794
795<dd>This option, which defaults to off, controls whether or not Clang
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]796adds runtime checks for undefined runtime behavior. If a check fails,
Mike Stump53664ca2009-12-14 23:53:10 +0000[diff] [blame]797<tt>__builtin_trap()</tt> is used to indicate failure.
798The checks are:
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]799<ul>
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]800<li>Subscripting where the static type of one operand is a variable
Mike Stump88b2a172009-12-16 03:25:12 +0000[diff] [blame]801 which is decayed from an array type and the other operand is
802 greater than the size of the array or less than zero.</li>
Mike Stump53664ca2009-12-14 23:53:10 +0000[diff] [blame]803<li>Shift operators where the amount shifted is greater or equal to the
804 promoted bit-width of the left-hand-side or less than zero.</li>
Mike Stump8f6a3ed2009-12-16 03:18:14 +0000[diff] [blame]805<li>If control flow reaches __builtin_unreachable.
806<li>When llvm implements more __builtin_object_size support, reads and
807 writes for objects that __builtin_object_size indicates we aren't
808 accessing valid memory. Bit-fields and vectors are not yet checked.
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]809</ul>
Mike Stump53664ca2009-12-14 23:53:10 +0000[diff] [blame]810</dd>
811
Nuno Lopesaa526242009-12-17 10:00:52 +0000[diff] [blame]812<dt id="opt_fno-assume-sane-operator-new"><b>-fno-assume-sane-operator-new</b>:
813Don't assume that the C++'s new operator is sane.</dt>
Nuno Lopesb23f20d2009-12-17 10:15:49 +0000[diff] [blame]814<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]815operator will always return a pointer that does not
Nuno Lopesaa526242009-12-17 10:00:52 +0000[diff] [blame]816alias any other pointer when the function returns.</dd>
Evan Chengfda026b2011-04-08 22:18:01 +0000[diff] [blame]817
Evan Cheng4cf4b592011-04-08 22:34:21 +0000[diff] [blame]818<dt id="opt_ftrap-function"><b>-ftrap-function=[name]</b>: Instruct code
Evan Chengfda026b2011-04-08 22:18:01 +0000[diff] [blame]819generator to emit a function call to the specified function name for
820<tt>__builtin_trap()</tt>.</dt>
821
822<dd>LLVM code generator translates <tt>__builtin_trap()</tt> to a trap
823instruction if it is supported by the target ISA. Otherwise, the builtin is
824translated into a call to <tt>abort</tt>. If this option is set, then the code
825generator will always lower the builtin to a call to the specified function
826regardless of whether the target ISA has a trap instruction. This option is
Chris Lattner02e266b2011-04-08 22:58:43 +0000[diff] [blame]827useful for environments (e.g. deeply embedded) where a trap cannot be properly
828handled, or when some custom behavior is desired.</dd>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]829</dl>
Nuno Lopesaa526242009-12-17 10:00:52 +0000[diff] [blame]830
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]831<!-- ======================================================================= -->
832<h2 id="c">C Language Features</h2>
833<!-- ======================================================================= -->
834
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]835<p>The support for standard C in clang is feature-complete except for the C99
836floating-point pragmas.</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]837
838<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]839<h3 id="c_ext">Extensions supported by clang</h3>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]840<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
841
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]842<p>See <a href="LanguageExtensions.html">clang language extensions</a>.</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]843
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]844<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
845<h3 id="c_modes">Differences between various standard modes</h3>
846<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]847
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]848<p>clang supports the -std option, which changes what language mode clang uses.
849The supported modes for C are c89, gnu89, c94, c99, gnu99 and various aliases
850for those modes. If no -std option is specified, clang defaults to gnu99 mode.
851</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]852
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]853<p>Differences between all c* and gnu* modes:</p>
854<ul>
855<li>c* modes define "__STRICT_ANSI__".</li>
Eli Friedman26fa0ed2009-05-27 23:02:57 +0000[diff] [blame]856<li>Target-specific defines not prefixed by underscores, like "linux", are
857defined in gnu* modes.</li>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]858<li>Trigraphs default to being off in gnu* modes; they can be enabled by the
859-trigraphs option.</li>
860<li>The parser recognizes "asm" and "typeof" as keywords in gnu* modes; the
861variants "__asm__" and "__typeof__" are recognized in all modes.</li>
Eli Friedmanb0ac5452009-05-16 23:17:30 +0000[diff] [blame]862<li>The Apple "blocks" extension is recognized by default in gnu* modes
863on some platforms; it can be enabled in any mode with the "-fblocks"
864option.</li>
Chris Lattner741be6a2011-06-15 00:41:40 +0000[diff] [blame]865<li>Arrays that are VLA's according to the standard, but which can be constant
866 folded by the frontend are treated as fixed size arrays. This occurs for
867 things like "int X[(1, 2)];", which is technically a VLA. c* modes are
868 strictly compliant and treat these as VLAs.</li>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]869</ul>
870
871<p>Differences between *89 and *99 modes:</p>
872<ul>
873<li>The *99 modes default to implementing "inline" as specified in C99, while
874the *89 modes implement the GNU version. This can be overridden for individual
875functions with the __gnu_inline__ attribute.</li>
Eli Friedman26fa0ed2009-05-27 23:02:57 +0000[diff] [blame]876<li>Digraphs are not recognized in c89 mode.</li>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]877<li>The scope of names defined inside a "for", "if", "switch", "while", or "do"
878statement is different. (example: "if ((struct x {int x;}*)0) {}".)</li>
879<li>__STDC_VERSION__ is not defined in *89 modes.</li>
Eli Friedman26fa0ed2009-05-27 23:02:57 +0000[diff] [blame]880<li>"inline" is not recognized as a keyword in c89 mode.</li>
881<li>"restrict" is not recognized as a keyword in *89 modes.</li>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]882<li>Commas are allowed in integer constant expressions in *99 modes.</li>
883<li>Arrays which are not lvalues are not implicitly promoted to pointers in
884*89 modes.</li>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]885<li>Some warnings are different.</li>
886</ul>
887
888<p>c94 mode is identical to c89 mode except that digraphs are enabled in
889c94 mode (FIXME: And __STDC_VERSION__ should be defined!).</p>
890
891<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
892<h3 id="c_unimpl_gcc">GCC extensions not implemented yet</h3>
893<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
894
895<p>clang tries to be compatible with gcc as much as possible, but some gcc
896extensions are not implemented yet:</p>
897
898<ul>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]899
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]900<li>clang does not support #pragma weak
Eli Friedman4da92552009-06-02 08:21:31 +0000[diff] [blame]901(<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]902the uses described in the bug, this is likely to be implemented at some
903point, at least partially.</li>
904
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]905<li>clang does not support decimal floating point types (_Decimal32 and
906friends) or fixed-point types (_Fract and friends); nobody has expressed
907interest in these features yet, so it's hard to say when they will be
908implemented.</li>
909
910<li>clang does not support nested functions; this is a complex feature which
911is infrequently used, so it is unlikely to be implemented anytime soon.</li>
912
Daniel Dunbar05fa6292009-05-12 23:12:07 +0000[diff] [blame]913<li>clang does not support global register variables, this is unlikely
Eli Friedmanca130bd2009-06-12 20:04:25 +0000[diff] [blame]914to be implemented soon because it requires additional LLVM backend support.
915</li>
Daniel Dunbar05fa6292009-05-12 23:12:07 +0000[diff] [blame]916
917<li>clang does not support static initialization of flexible array
918members. This appears to be a rarely used extension, but could be
919implemented pending user demand.</li>
920
Eli Friedmanca130bd2009-06-12 20:04:25 +0000[diff] [blame]921<li>clang does not support __builtin_va_arg_pack/__builtin_va_arg_pack_len.
922This is used rarely, but in some potentially interesting places, like the
923glibc headers, so it may be implemented pending user demand. Note that
924because clang pretends to be like GCC 4.2, and this extension was introduced
925in 4.3, the glibc headers will not try to use this extension with clang at
926the moment.</li>
927
928<li>clang does not support the gcc extension for forward-declaring function
929parameters; this has not showed up in any real-world code yet, though, so it
930might never be implemented.</li>
931
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]932</ul>
933
934<p>This is not a complete list; if you find an unsupported extension
935missing from this list, please send an e-mail to cfe-dev. This list
936currently excludes C++; see <a href="#cxx">C++ Language Features</a>.
937Also, this list does not include bugs in mostly-implemented features; please
938see 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">
939bug tracker</a> for known existing bugs (FIXME: Is there a section for
940bug-reporting guidelines somewhere?).</p>
941
942<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
943<h3 id="c_unsupp_gcc">Intentionally unsupported GCC extensions</h3>
944<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
945
Eli Friedman0b326002009-06-12 20:11:05 +0000[diff] [blame]946<ul>
947
Eli Friedman2f3fca02009-06-12 20:08:48 +0000[diff] [blame]948<li>clang does not support the gcc extension that allows variable-length arrays
Chris Lattnerd462b6a2011-01-24 03:47:59 +0000[diff] [blame]949in structures. This is for a few reasons: one, it is tricky
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]950to implement, two, the extension is completely undocumented, and three, the
Chris Lattnerb9d511c2010-10-28 02:20:32 +0000[diff] [blame]951extension appears to be rarely used. Note that clang <em>does</em> support
952flexible array members (arrays with a zero or unspecified size at the end of
953a structure).</li>
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]954
Eli Friedmanca130bd2009-06-12 20:04:25 +0000[diff] [blame]955<li>clang does not have an equivalent to gcc's "fold"; this means that
956clang doesn't accept some constructs gcc might accept in contexts where a
Eli Friedman248f9732011-06-13 23:12:01 +0000[diff] [blame]957constant expression is required, like "x-x" where x is a variable.</li>
Eli Friedmanca130bd2009-06-12 20:04:25 +0000[diff] [blame]958
Eli Friedman2f3fca02009-06-12 20:08:48 +0000[diff] [blame]959<li>clang does not support multiple alternative constraints in inline asm; this
Eli Friedmanca130bd2009-06-12 20:04:25 +0000[diff] [blame]960is an extremely obscure feature which would be complicated to implement
Eli Friedman2f3fca02009-06-12 20:08:48 +0000[diff] [blame]961correctly.</li>
Eli Friedmanca130bd2009-06-12 20:04:25 +0000[diff] [blame]962
963<li>clang does not support __builtin_apply and friends; this extension is
964extremely obscure and difficult to implement reliably.</li>
965
Eli Friedman0b326002009-06-12 20:11:05 +0000[diff] [blame]966</ul>
967
Eli Friedman660a5ac2009-04-28 18:48:34 +0000[diff] [blame]968<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
969<h3 id="c_ms">Microsoft extensions</h3>
970<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
971
972<p>clang has some experimental support for extensions from
973Microsoft Visual C++; to enable it, use the -fms-extensions command-line
Eli Friedmana1821b52009-06-08 06:21:03 +0000[diff] [blame]974option. This is the default for Windows targets. Note that the
975support is incomplete; enabling Microsoft extensions will silently drop
976certain constructs (including __declspec and Microsoft-style asm statements).
977</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]978
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]979<ul>
Douglas Gregord3b227d2010-12-14 16:21:49 +0000[diff] [blame]980<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]981is the same as Visual C/C++ 2003. Any number is supported and can greatly affect
982what Windows SDK and c++stdlib headers clang can compile. This option will be
983removed when clang supports the full set of MS extensions required for these
984headers.</li>
985
Daniel Dunbar05fa6292009-05-12 23:12:07 +0000[diff] [blame]986<li>clang does not support the Microsoft extension where anonymous
Daniel Dunbar869e0542009-05-13 00:23:53 +0000[diff] [blame]987record members can be declared using user defined typedefs.</li>
Daniel Dunbar05fa6292009-05-12 23:12:07 +0000[diff] [blame]988
Daniel Dunbar9375ed12009-05-13 21:40:49 +0000[diff] [blame]989<li>clang supports the Microsoft "#pragma pack" feature for
990controlling record layout. GCC also contains support for this feature,
991however where MSVC and GCC are incompatible clang follows the MSVC
992definition.</li>
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]993</ul>
Daniel Dunbar9375ed12009-05-13 21:40:49 +0000[diff] [blame]994
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]995<!-- ======================================================================= -->
Richard Smithbdb10182011-11-21 20:54:59 +0000[diff] [blame^]996<h2 id="cxx">C++ Language Features</h2>
997<!-- ======================================================================= -->
998
999<p>clang fully implements all of standard C++98 except for exported templates
1000(which were removed in C++11), and
1001<a href="http://clang.llvm.org/cxx_status.html">many C++11 features</a> are also
1002implemented.</p>
1003
1004<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1005<h3 id="cxx_implimits">Controlling implementation limits</h3>
1006<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1007
1008<p><b>-fconstexpr-depth=N</b>: Sets the limit for recursive constexpr function
1009invocations to N. The default is 512.</p>
1010
1011<p><b>-ftemplate-depth=N</b>: Sets the limit for recursively nested template
1012instantiations to N. The default is 1024.</p>
1013
1014<!-- ======================================================================= -->
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]1015<h2 id="target_features">Target-Specific Features and Limitations</h2>
1016<!-- ======================================================================= -->
1017
1018
1019<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1020<h3 id="target_arch">CPU Architectures Features and Limitations</h3>
1021<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1022
1023<!-- ======================== -->
1024<h4 id="target_arch_x86">X86</h4>
1025<!-- ======================== -->
Daniel Dunbarbcaf7aa2010-09-19 19:26:59 +0000[diff] [blame]1026
1027<p>The support for X86 (both 32-bit and 64-bit) is considered stable on Darwin
1028(Mac OS/X), Linux, FreeBSD, and Dragonfly BSD: it has been tested to correctly
1029compile many large C, C++, Objective-C, and Objective-C++ codebases.</p>
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]1030
NAKAMURA Takumiddf68ff2011-04-05 04:31:31 +0000[diff] [blame]1031<p>On x86_64-mingw32, passing i128(by value) is incompatible to Microsoft x64
1032calling conversion. You might need to tweak WinX86_64ABIInfo::classify()
1033in lib/CodeGen/TargetInfo.cpp.</p>
NAKAMURA Takumia75fdb22011-04-04 15:02:41 +0000[diff] [blame]1034
Eli Friedman3b658d32009-06-08 05:12:39 +0000[diff] [blame]1035<!-- ======================== -->
1036<h4 id="target_arch_arm">ARM</h4>
1037<!-- ======================== -->
Daniel Dunbarbcaf7aa2010-09-19 19:26:59 +0000[diff] [blame]1038
1039<p>The support for ARM (specifically ARMv6 and ARMv7) is considered stable on
1040Darwin (iOS): it has been tested to correctly compile many large C, C++,
Bob Wilson4ea8dfa2011-01-10 17:55:14 +0000[diff] [blame]1041Objective-C, and Objective-C++ codebases. Clang only supports a limited number
1042of ARM architectures. It does not yet fully support ARMv5, for example.</p>
Eli Friedman3b658d32009-06-08 05:12:39 +0000[diff] [blame]1043
1044<!-- ======================== -->
1045<h4 id="target_arch_other">Other platforms</h4>
1046<!-- ======================== -->
1047clang currently contains some support for PPC and Sparc; however, significant
1048pieces of code generation are still missing, and they haven't undergone
1049significant testing.
1050
Eli Friedman3b658d32009-06-08 05:12:39 +0000[diff] [blame]1051<p>clang contains limited support for the MSP430 embedded processor, but both
1052the clang support and the LLVM backend support are highly experimental.
1053
1054<p>Other platforms are completely unsupported at the moment. Adding the
1055minimal support needed for parsing and semantic analysis on a new platform
1056is quite easy; see lib/Basic/Targets.cpp in the clang source tree. This level
1057of support is also sufficient for conversion to LLVM IR for simple programs.
1058Proper support for conversion to LLVM IR requires adding code to
1059lib/CodeGen/CGCall.cpp at the moment; this is likely to change soon, though.
1060Generating assembly requires a suitable LLVM backend.
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]1061
1062<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1063<h3 id="target_os">Operating System Features and Limitations</h3>
1064<!-- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = -->
1065
1066<!-- ======================================= -->
1067<h4 id="target_os_darwin">Darwin (Mac OS/X)</h4>
1068<!-- ======================================= -->
1069
1070<p>No __thread support, 64-bit ObjC support requires SL tools.</p>
1071
NAKAMURA Takumia75fdb22011-04-04 15:02:41 +0000[diff] [blame]1072<!-- ======================================= -->
1073<h4 id="target_os_win32">Windows</h4>
1074<!-- ======================================= -->
1075
1076<p>Experimental supports are on Cygming.</p>
1077
1078<h5>Cygwin</h5>
1079
1080<p>Clang works on Cygwin-1.7.</p>
1081
1082<h5>MinGW32</h5>
1083
1084<p>Clang works on some mingw32 distributions.
1085Clang assumes directories as below;</p>
1086
1087<ul>
1088<li><tt>C:/mingw/include</tt></li>
1089<li><tt>C:/mingw/lib</tt></li>
1090<li><tt>C:/mingw/lib/gcc/mingw32/4.[3-5].0/include/c++</tt></li>
1091</ul>
1092
NAKAMURA Takumi0e53df62011-11-17 07:54:13 +0000[diff] [blame]1093<p>On MSYS, a few tests might fail.</p>
NAKAMURA Takumia75fdb22011-04-04 15:02:41 +0000[diff] [blame]1094
1095<h5>MinGW-w64</h5>
1096
Douglas Gregordad879a2011-08-30 00:40:12 +0000[diff] [blame]1097<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]1098
1099<ul>
Douglas Gregordad879a2011-08-30 00:40:12 +0000[diff] [blame]1100<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>
1101<li><tt>some_directory/bin/gcc.exe</tt></li>
1102<li><tt>some_directory/bin/clang.exe</tt></li>
1103<li><tt>some_directory/bin/clang++.exe</tt></li>
1104<li><tt>some_directory/bin/../include/c++/GCC_version</tt></li>
1105<li><tt>some_directory/bin/../include/c++/GCC_version/x86_64-w64-mingw32</tt></li>
1106<li><tt>some_directory/bin/../include/c++/GCC_version/i686-w64-mingw32</tt></li>
1107<li><tt>some_directory/bin/../include/c++/GCC_version/backward</tt></li>
1108<li><tt>some_directory/bin/../x86_64-w64-mingw32/include</tt></li>
1109<li><tt>some_directory/bin/../i686-w64-mingw32/include</tt></li>
1110<li><tt>some_directory/bin/../include</tt></li>
NAKAMURA Takumia75fdb22011-04-04 15:02:41 +0000[diff] [blame]1111</ul>
1112
Douglas Gregordad879a2011-08-30 00:40:12 +0000[diff] [blame]1113<p>This directory layout is standard for any toolchain you will find on the official <a href="mingw-w64.sourceforge.net">MinGW-w64 website</a>.
1114
1115<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>
1116
1117<p><a href="http://llvm.org/bugs/show_bug.cgi?id=9072">Some tests might fail</a>
1118on x86_64-w64-mingw32.</p>
NAKAMURA Takumia75fdb22011-04-04 15:02:41 +0000[diff] [blame]1119
Chris Lattnercf17d9d2009-04-20 04:23:09 +0000[diff] [blame]1120</div>
1121</body>
1122</html>