Gitiles
Code Review Sign In
gerrit-public.fairphone.software / platform / external / clang / 630b7aefc887c19e1f7a102098b4fa7907ae6b75 / . / docs / LanguageExtensions.html
blob: 78c3cd50bf37e1615b9f921d86ccd9f3bc93c971 [file] [log] [blame]
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1<html>
2<head>
3<title>Clang Language Extensions</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 Language Extensions</h1>
19
20<ul>
21<li><a href="#intro">Introduction</a></li>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]22<li><a href="#feature_check">Feature Checking Macros</a></li>
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]23<li><a href="#has_include">Include File Checking Macros</a></li>
Chris Lattner81edc9f2009-04-13 02:45:46 +0000[diff] [blame]24<li><a href="#builtinmacros">Builtin Macros</a></li>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]25<li><a href="#vectors">Vectors and Extended Vectors</a></li>
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]26<li><a href="#deprecated">Messages on <tt>deprecated</tt> and <tt>unavailable</tt> attributes</a></li>
27<li><a href="#attributes-on-enumerators">Attributes on enumerators</a></li>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]28<li><a href="#checking_language_features">Checks for Standard Language Features</a></li>
Ted Kremenek22c34102009-12-03 02:05:57 +0000[diff] [blame]29 <ul>
30 <li><a href="#cxx_exceptions">C++ exceptions</a></li>
31 <li><a href="#cxx_rtti">C++ RTTI</a></li>
32 </ul>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]33<li><a href="#checking_upcoming_features">Checks for Upcoming Standard Language Features</a></li>
34 <ul>
35 <li><a href="#cxx_attributes">C++0x attributes</a></li>
36 <li><a href="#cxx_decltype">C++0x <tt>decltype()</tt></a></li>
37 <li><a href="#cxx_deleted_functions">C++0x deleted functions</a></li>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]38 <li><a href="#cxx_lambdas">C++0x lambdas</a></li>
39 <li><a href="#cxx_nullptr">C++0x nullptr</a></li>
40 <li><a href="#cxx_rvalue_references">C++0x rvalue references</a></li>
41 <li><a href="#cxx_static_assert">C++0x <tt>static_assert()</tt></a></li>
42 <li><a href="#cxx_auto_type">C++0x type inference</a></li>
43 <li><a href="#cxx_variadic_templates">C++0x variadic templates</a></li>
Sebastian Redlf6c09772010-08-31 23:28:47 +0000[diff] [blame]44 <li><a href="#cxx_inline_namespaces">C++0x inline namespaces</a></li>
Douglas Gregor1274ccd2010-10-08 23:50:27 +0000[diff] [blame]45 <li><a href="#cxx_strong_enums">C++0x strongly-typed enumerations</a></li>
Douglas Gregordab60ad2010-10-01 18:44:50 +0000[diff] [blame]46 <li><a href="#cxx_trailing_return">C++0x trailing return type</a></li>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]47 </ul>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]48<li><a href="#blocks">Blocks</a></li>
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]49<li><a href="#overloading-in-c">Function Overloading in C</a></li>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]50<li><a href="#builtins">Builtin Functions</a>
51 <ul>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]52 <li><a href="#__builtin_shufflevector">__builtin_shufflevector</a></li>
Chris Lattner21190d52009-09-21 03:09:59 +0000[diff] [blame]53 <li><a href="#__builtin_unreachable">__builtin_unreachable</a></li>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]54 </ul>
55</li>
Chris Lattner1177f912009-04-09 19:58:15 +0000[diff] [blame]56<li><a href="#targetspecific">Target-Specific Extensions</a>
57 <ul>
58 <li><a href="#x86-specific">X86/X86-64 Language Extensions</a></li>
59 </ul>
60</li>
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]61<li><a href="#analyzerspecific">Static Analysis-Specific Extensions</a>
62 <ul>
63 <li><a href="#analyzerattributes">Analyzer Attributes</a></li>
64 </ul>
65</li>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]66</ul>
67
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]68<!-- ======================================================================= -->
69<h2 id="intro">Introduction</h2>
70<!-- ======================================================================= -->
71
72<p>This document describes the language extensions provided by Clang. In
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]73addition to the language extensions listed here, Clang aims to support a broad
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]74range of GCC extensions. Please see the <a
75href="http://gcc.gnu.org/onlinedocs/gcc/C-Extensions.html">GCC manual</a> for
76more information on these extensions.</p>
77
78<!-- ======================================================================= -->
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]79<h2 id="feature_check">Feature Checking Macros</h2>
80<!-- ======================================================================= -->
81
82<p>Language extensions can be very useful, but only if you know you can depend
83on them. In order to allow fine-grain features checks, we support two builtin
84function-like macros. This allows you to directly test for a feature in your
85code without having to resort to something like autoconf or fragile "compiler
86version checks".</p>
87
88<!-- ======================================================================= -->
89<h3 id="__has_builtin">__has_builtin</h3>
90<!-- ======================================================================= -->
91
92<p>This function-like macro takes a single identifier argument that is the name
93of a builtin function. It evaluates to 1 if the builtin is supported or 0 if
94not. It can be used like this:</p>
95
96<blockquote>
97<pre>
98#ifndef __has_builtin // Optional of course.
99 #define __has_builtin(x) 0 // Compatibility with non-clang compilers.
100#endif
101
102...
103#if __has_builtin(__builtin_trap)
104 __builtin_trap();
105#else
106 abort();
107#endif
108...
109</pre>
110</blockquote>
111
112
113<!-- ======================================================================= -->
114<h3 id="__has_feature">__has_feature</h3>
115<!-- ======================================================================= -->
116
117<p>This function-like macro takes a single identifier argument that is the name
118of a feature. It evaluates to 1 if the feature is supported or 0 if not. It
119can be used like this:</p>
120
121<blockquote>
122<pre>
123#ifndef __has_feature // Optional of course.
124 #define __has_feature(x) 0 // Compatibility with non-clang compilers.
125#endif
126
127...
128#if __has_feature(attribute_overloadable) || \
129 __has_feature(blocks)
130...
131#endif
132...
133</pre>
134</blockquote>
135
136<p>The feature tag is described along with the language feature below.</p>
137
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]138<!-- ======================================================================= -->
Anders Carlssoncae50952010-10-20 02:31:43 +0000[diff] [blame]139<h3 id="__has_attribute">__has_attribute</h3>
140<!-- ======================================================================= -->
141
142<p>This function-like macro takes a single identifier argument that is the name
143of an attribute. It evaluates to 1 if the attribute is supported or 0 if not. It
144can be used like this:</p>
145
146<blockquote>
147<pre>
148#ifndef __has_attribute // Optional of course.
149 #define __has_attribute(x) 0 // Compatibility with non-clang compilers.
150#endif
151
152...
Anders Carlsson961003d2011-01-24 03:54:51 +0000[diff] [blame]153#if __has_attribute(always_inline)
154#define ALWAYS_INLINE __attribute__((always_inline))
Anders Carlssoncae50952010-10-20 02:31:43 +0000[diff] [blame]155#else
Anders Carlsson961003d2011-01-24 03:54:51 +0000[diff] [blame]156#define ALWAYS_INLINE
Anders Carlssoncae50952010-10-20 02:31:43 +0000[diff] [blame]157#endif
158...
159</pre>
160</blockquote>
161
162<!-- ======================================================================= -->
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]163<h2 id="has_include">Include File Checking Macros</h2>
164<!-- ======================================================================= -->
165
166<p>Not all developments systems have the same include files.
167The <a href="#__has_include">__has_include</a> and
168<a href="#__has_include_next">__has_include_next</a> macros allow you to
169check for the existence of an include file before doing
170a possibly failing #include directive.</p>
171
172<!-- ======================================================================= -->
173<h3 id="__has_include">__has_include</h3>
174<!-- ======================================================================= -->
175
176<p>This function-like macro takes a single file name string argument that
177is the name of an include file. It evaluates to 1 if the file can
178be found using the include paths, or 0 otherwise:</p>
179
180<blockquote>
181<pre>
182// Note the two possible file name string formats.
183#if __has_include("myinclude.h") && __has_include(&lt;stdint.h&gt;)
184# include "myinclude.h"
185#endif
186
187// To avoid problem with non-clang compilers not having this macro.
188#if defined(__has_include) && __has_include("myinclude.h")
189# include "myinclude.h"
190#endif
191</pre>
192</blockquote>
193
194<p>To test for this feature, use #if defined(__has_include).</p>
195
196<!-- ======================================================================= -->
197<h3 id="__has_include_next">__has_include_next</h3>
198<!-- ======================================================================= -->
199
200<p>This function-like macro takes a single file name string argument that
201is the name of an include file. It is like __has_include except that it
202looks for the second instance of the given file found in the include
203paths. It evaluates to 1 if the second instance of the file can
204be found using the include paths, or 0 otherwise:</p>
205
206<blockquote>
207<pre>
208// Note the two possible file name string formats.
209#if __has_include_next("myinclude.h") && __has_include_next(&lt;stdint.h&gt;)
210# include_next "myinclude.h"
211#endif
212
213// To avoid problem with non-clang compilers not having this macro.
214#if defined(__has_include_next) && __has_include_next("myinclude.h")
215# include_next "myinclude.h"
216#endif
217</pre>
218</blockquote>
219
220<p>Note that __has_include_next, like the GNU extension
221#include_next directive, is intended for use in headers only,
222and will issue a warning if used in the top-level compilation
223file. A warning will also be issued if an absolute path
224is used in the file argument.</p>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]225
226<!-- ======================================================================= -->
Chris Lattner81edc9f2009-04-13 02:45:46 +0000[diff] [blame]227<h2 id="builtinmacros">Builtin Macros</h2>
228<!-- ======================================================================= -->
229
Douglas Gregor4290fbd2010-04-30 02:51:06 +0000[diff] [blame]230<dl>
231 <dt><code>__BASE_FILE__</code></dt>
232 <dd>Defined to a string that contains the name of the main input
233 file passed to Clang.</dd>
234
235 <dt><code>__COUNTER__</code></dt>
236 <dd>Defined to an integer value that starts at zero and is
237 incremented each time the <code>__COUNTER__</code> macro is
238 expanded.</dd>
239
240 <dt><code>__INCLUDE_LEVEL__</code></dt>
241 <dd>Defined to an integral value that is the include depth of the
242 file currently being translated. For the main file, this value is
243 zero.</dd>
244
245 <dt><code>__TIMESTAMP__</code></dt>
246 <dd>Defined to the date and time of the last modification of the
247 current source file.</dd>
248
249 <dt><code>__clang__</code></dt>
250 <dd>Defined when compiling with Clang</dd>
251
252 <dt><code>__clang_major__</code></dt>
253 <dd>Defined to the major version number of Clang (e.g., the 2 in
254 2.0.1).</dd>
255
256 <dt><code>__clang_minor__</code></dt>
257 <dd>Defined to the minor version number of Clang (e.g., the 0 in
258 2.0.1).</dd>
259
260 <dt><code>__clang_patchlevel__</code></dt>
261 <dd>Defined to the patch level of Clang (e.g., the 1 in 2.0.1).</dd>
262
263 <dt><code>__clang_version__</code></dt>
264 <dd>Defined to a string that captures the Clang version, including
265 the Subversion tag or revision number, e.g., "1.5 (trunk
266 102332)".</dd>
267</dl>
Chris Lattner81edc9f2009-04-13 02:45:46 +0000[diff] [blame]268
269<!-- ======================================================================= -->
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]270<h2 id="vectors">Vectors and Extended Vectors</h2>
271<!-- ======================================================================= -->
272
Owen Andersond2bf0cd2010-01-27 01:22:36 +0000[diff] [blame]273<p>Supports the GCC vector extensions, plus some stuff like V[1].</p>
274
275<p>Also supports <tt>ext_vector</tt>, which additionally support for V.xyzw
276syntax and other tidbits as seen in OpenCL. An example is:</p>
277
278<blockquote>
279<pre>
280typedef float float4 <b>__attribute__((ext_vector_type(4)))</b>;
281typedef float float2 <b>__attribute__((ext_vector_type(2)))</b>;
282
283float4 foo(float2 a, float2 b) {
284 float4 c;
285 c.xz = a;
286 c.yw = b;
287 return c;
288}
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]289</pre>
Owen Andersond2bf0cd2010-01-27 01:22:36 +0000[diff] [blame]290</blockquote>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]291
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]292<p>Query for this feature with __has_feature(attribute_ext_vector_type).</p>
293
Owen Andersond2bf0cd2010-01-27 01:22:36 +0000[diff] [blame]294<p>See also <a href="#__builtin_shufflevector">__builtin_shufflevector</a>.</p>
295
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]296<!-- ======================================================================= -->
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]297<h2 id="deprecated">Messages on <tt>deprecated</tt> and <tt>unavailable</tt> Attributes</h2>
Fariborz Jahanianc784dc12010-10-06 23:12:32 +0000[diff] [blame]298<!-- ======================================================================= -->
299
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]300<p>An optional string message can be added to the <tt>deprecated</tt>
301and <tt>unavailable</tt> attributes. For example:</p>
Fariborz Jahanianc784dc12010-10-06 23:12:32 +0000[diff] [blame]302
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]303<blockquote>
Chris Lattner4836d6a2010-11-09 19:43:35 +0000[diff] [blame]304<pre>void explode(void) __attribute__((deprecated("extremely unsafe, use 'combust' instead!!!")));</pre>
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]305</blockquote>
306
307<p>If the deprecated or unavailable declaration is used, the message
308will be incorporated into the appropriate diagnostic:</p>
309
310<blockquote>
Chris Lattner4836d6a2010-11-09 19:43:35 +0000[diff] [blame]311<pre>harmless.c:4:3: warning: 'explode' is deprecated: extremely unsafe, use 'combust' instead!!! [-Wdeprecated-declarations]
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]312 explode();
313 ^</pre>
314</blockquote>
315
316<p>Query for this feature
317with <tt>__has_feature(attribute_deprecated_with_message)</tt>
318and <tt>__has_feature(attribute_unavailable_with_message)</tt>.</p>
319
320<!-- ======================================================================= -->
321<h2 id="attributes-on-enumerators">Attributes on Enumerators</h2>
322<!-- ======================================================================= -->
323
324<p>Clang allows attributes to be written on individual enumerators.
325This allows enumerators to be deprecated, made unavailable, etc. The
326attribute must appear after the enumerator name and before any
327initializer, like so:</p>
328
329<blockquote>
330<pre>enum OperationMode {
331 OM_Invalid,
332 OM_Normal,
333 OM_Terrified __attribute__((deprecated)),
334 OM_AbortOnError __attribute__((deprecated)) = 4
335};</pre>
336</blockquote>
337
338<p>Attributes on the <tt>enum</tt> declaration do not apply to
339individual enumerators.</p>
340
341<p>Query for this feature with <tt>__has_feature(enumerator_attributes)</tt>.</p>
Fariborz Jahanianc784dc12010-10-06 23:12:32 +0000[diff] [blame]342
343<!-- ======================================================================= -->
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]344<h2 id="checking_language_features">Checks for Standard Language Features</h2>
345<!-- ======================================================================= -->
346
347<p>The <tt>__has_feature</tt> macro can be used to query if certain standard language features are
348enabled. Those features are listed here.</p>
349
Ted Kremenek22c34102009-12-03 02:05:57 +0000[diff] [blame]350<h3 id="cxx_exceptions">C++ exceptions</h3>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]351
Ted Kremenek22c34102009-12-03 02:05:57 +0000[diff] [blame]352<p>Use <tt>__has_feature(cxx_exceptions)</tt> to determine if C++ exceptions have been enabled. For
353example, compiling code with <tt>-fexceptions</tt> enables C++ exceptions.</p>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]354
Ted Kremenek22c34102009-12-03 02:05:57 +0000[diff] [blame]355<h3 id="cxx_rtti">C++ RTTI</h3>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]356
Ted Kremenek0eb95602009-12-03 02:06:43 +0000[diff] [blame]357<p>Use <tt>__has_feature(cxx_rtti)</tt> to determine if C++ RTTI has been enabled. For example,
Ted Kremenek22c34102009-12-03 02:05:57 +0000[diff] [blame]358compiling code with <tt>-fno-rtti</tt> disables the use of RTTI.</p>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]359
360<!-- ======================================================================= -->
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]361<h2 id="checking_upcoming_features">Checks for Upcoming Standard Language Features</h2>
362<!-- ======================================================================= -->
363
364<p>The <tt>__has_feature</tt> macro can be used to query if certain upcoming
365standard language features are enabled. Those features are listed here.</p>
366
367<p>Currently, all features listed here are slated for inclusion in the upcoming
368C++0x standard. As a result, all the features that clang supports are enabled
369with the <tt>-std=c++0x</tt> option when compiling C++ code. Features that are
370not yet implemented will be noted.</p>
371
372<h3 id="cxx_decltype">C++0x <tt>decltype()</tt></h3>
373
374<p>Use <tt>__has_feature(cxx_decltype)</tt> to determine if support for the
375<tt>decltype()</tt> specifier is enabled.</p>
376
377<h3 id="cxx_attributes">C++0x attributes</h3>
378
379<p>Use <tt>__has_feature(cxx_attributes)</tt> to determine if support for
Sebastian Redlf6c09772010-08-31 23:28:47 +0000[diff] [blame]380attribute parsing with C++0x's square bracket notation is enabled.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]381
382<h3 id="cxx_deleted_functions">C++0x deleted functions</tt></h3>
383
384<p>Use <tt>__has_feature(cxx_deleted_functions)</tt> to determine if support for
Sebastian Redlf6c09772010-08-31 23:28:47 +0000[diff] [blame]385deleted function definitions (with <tt>= delete</tt>) is enabled.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]386
Douglas Gregor9cc90a32010-01-13 16:27:49 +0000[diff] [blame]387<h3 id="cxx_lambdas">C++0x lambdas</h3>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]388
389<p>Use <tt>__has_feature(cxx_lambdas)</tt> to determine if support for
Sebastian Redlf6c09772010-08-31 23:28:47 +0000[diff] [blame]390lambdas is enabled. clang does not currently implement this feature.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]391
392<h3 id="cxx_nullptr">C++0x <tt>nullptr</tt></h3>
393
394<p>Use <tt>__has_feature(cxx_nullptr)</tt> to determine if support for
Sebastian Redlf6c09772010-08-31 23:28:47 +0000[diff] [blame]395<tt>nullptr</tt> is enabled. clang does not yet fully implement this
396feature.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]397
398<h3 id="cxx_rvalue_references">C++0x rvalue references</tt></h3>
399
400<p>Use <tt>__has_feature(cxx_rvalue_references)</tt> to determine if support for
Sebastian Redlf6c09772010-08-31 23:28:47 +0000[diff] [blame]401rvalue references is enabled. clang does not yet fully implement this
402feature.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]403
404<h3 id="cxx_static_assert">C++0x <tt>static_assert()</tt></h3>
405
406<p>Use <tt>__has_feature(cxx_static_assert)</tt> to determine if support for
407compile-time assertions using <tt>static_assert</tt> is enabled.</p>
408
409<h3 id="cxx_auto_type">C++0x type inference</h3>
410
411<p>Use <tt>__has_feature(cxx_auto_type)</tt> to determine C++0x type inference
412is supported using the <tt>auto</tt> specifier. If this is disabled,
413<tt>auto</tt> will instead be a storage class specifier, as in C or C++98.</p>
414
Sebastian Redlf6c09772010-08-31 23:28:47 +0000[diff] [blame]415<h3 id="cxx_variadic_templates">C++0x variadic templates</h3>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]416
417<p>Use <tt>__has_feature(cxx_variadic_templates)</tt> to determine if support
Douglas Gregor83d77812011-01-19 23:15:20 +0000[diff] [blame]418for variadic templates is enabled.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]419
Sebastian Redlf6c09772010-08-31 23:28:47 +0000[diff] [blame]420<h3 id="cxx_inline_namespaces">C++0x inline namespaces</h3>
421
422<p>Use <tt>__has_feature(cxx_inline_namespaces)</tt> to determine if support for
423inline namespaces is enabled.</p>
424
Douglas Gregordab60ad2010-10-01 18:44:50 +0000[diff] [blame]425<h3 id="cxx_trailing_return">C++0x trailing return type</h3>
426
427<p>Use <tt>__has_feature(cxx_trailing_return)</tt> to determine if support for
428the alternate function declaration syntax with trailing return type is enabled.</p>
429
Douglas Gregor1274ccd2010-10-08 23:50:27 +0000[diff] [blame]430<h3 id="cxx_strong_enums">C++0x strongly typed enumerations</h3>
431
432<p>Use <tt>__has_feature(cxx_strong_enums)</tt> to determine if support for
433strongly typed, scoped enumerations is enabled.</p>
434
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]435<!-- ======================================================================= -->
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]436<h2 id="blocks">Blocks</h2>
437<!-- ======================================================================= -->
438
Chris Lattnera7dbdf52009-03-09 07:03:22 +0000[diff] [blame]439<p>The syntax and high level language feature description is in <a
440href="BlockLanguageSpec.txt">BlockLanguageSpec.txt</a>. Implementation and ABI
441details for the clang implementation are in <a
Chris Lattner5d7650b2010-03-16 21:43:03 +0000[diff] [blame]442href="Block-ABI-Apple.txt">Block-ABI-Apple.txt</a>.</p>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]443
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]444
445<p>Query for this feature with __has_feature(blocks).</p>
446
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]447<!-- ======================================================================= -->
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]448<h2 id="overloading-in-c">Function Overloading in C</h2>
449<!-- ======================================================================= -->
450
Chris Lattnerf161d412009-02-13 21:51:45 +0000[diff] [blame]451<p>Clang provides support for C++ function overloading in C. Function
452overloading in C is introduced using the <tt>overloadable</tt> attribute. For
453example, one might provide several overloaded versions of a <tt>tgsin</tt>
454function that invokes the appropriate standard function computing the sine of a
455value with <tt>float</tt>, <tt>double</tt>, or <tt>long double</tt>
456precision:</p>
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]457
458<blockquote>
459<pre>
460#include &lt;math.h&gt;
461float <b>__attribute__((overloadable))</b> tgsin(float x) { return sinf(x); }
462double <b>__attribute__((overloadable))</b> tgsin(double x) { return sin(x); }
463long double <b>__attribute__((overloadable))</b> tgsin(long double x) { return sinl(x); }
464</pre>
465</blockquote>
466
467<p>Given these declarations, one can call <tt>tgsin</tt> with a
468<tt>float</tt> value to receive a <tt>float</tt> result, with a
469<tt>double</tt> to receive a <tt>double</tt> result, etc. Function
470overloading in C follows the rules of C++ function overloading to pick
471the best overload given the call arguments, with a few C-specific
472semantics:</p>
473<ul>
474 <li>Conversion from <tt>float</tt> or <tt>double</tt> to <tt>long
475 double</tt> is ranked as a floating-point promotion (per C99) rather
476 than as a floating-point conversion (as in C++).</li>
477
478 <li>A conversion from a pointer of type <tt>T*</tt> to a pointer of type
479 <tt>U*</tt> is considered a pointer conversion (with conversion
480 rank) if <tt>T</tt> and <tt>U</tt> are compatible types.</li>
481
482 <li>A conversion from type <tt>T</tt> to a value of type <tt>U</tt>
483 is permitted if <tt>T</tt> and <tt>U</tt> are compatible types. This
484 conversion is given "conversion" rank.</li>
485</ul>
486
487<p>The declaration of <tt>overloadable</tt> functions is restricted to
488function declarations and definitions. Most importantly, if any
489function with a given name is given the <tt>overloadable</tt>
490attribute, then all function declarations and definitions with that
491name (and in that scope) must have the <tt>overloadable</tt>
Chris Lattnerf161d412009-02-13 21:51:45 +0000[diff] [blame]492attribute. This rule even applies to redeclarations of functions whose original
493declaration had the <tt>overloadable</tt> attribute, e.g.,</p>
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]494
495<blockquote>
496<pre>
497int f(int) __attribute__((overloadable));
498float f(float); <i>// error: declaration of "f" must have the "overloadable" attribute</i>
499
500int g(int) __attribute__((overloadable));
501int g(int) { } <i>// error: redeclaration of "g" must also have the "overloadable" attribute</i>
502</pre>
503</blockquote>
504
Douglas Gregor965acbb2009-02-18 07:07:28 +0000[diff] [blame]505<p>Functions marked <tt>overloadable</tt> must have
506prototypes. Therefore, the following code is ill-formed:</p>
507
508<blockquote>
509<pre>
510int h() __attribute__((overloadable)); <i>// error: h does not have a prototype</i>
511</pre>
512</blockquote>
513
514<p>However, <tt>overloadable</tt> functions are allowed to use a
515ellipsis even if there are no named parameters (as is permitted in C++). This feature is particularly useful when combined with the <tt>unavailable</tt> attribute:</p>
516
517<blockquote>
518<pre>
Chris Lattner02246802009-02-18 22:27:46 +0000[diff] [blame]519void honeypot(...) __attribute__((overloadable, unavailable)); <i>// calling me is an error</i>
Douglas Gregor965acbb2009-02-18 07:07:28 +0000[diff] [blame]520</pre>
521</blockquote>
522
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]523<p>Functions declared with the <tt>overloadable</tt> attribute have
524their names mangled according to the same rules as C++ function
525names. For example, the three <tt>tgsin</tt> functions in our
526motivating example get the mangled names <tt>_Z5tgsinf</tt>,
Chris Lattner71b48d62010-11-28 18:19:13 +0000[diff] [blame]527<tt>_Z5tgsind</tt>, and <tt>_Z5tgsine</tt>, respectively. There are two
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]528caveats to this use of name mangling:</p>
529
530<ul>
531
532 <li>Future versions of Clang may change the name mangling of
533 functions overloaded in C, so you should not depend on an specific
534 mangling. To be completely safe, we strongly urge the use of
535 <tt>static inline</tt> with <tt>overloadable</tt> functions.</li>
536
537 <li>The <tt>overloadable</tt> attribute has almost no meaning when
538 used in C++, because names will already be mangled and functions are
539 already overloadable. However, when an <tt>overloadable</tt>
540 function occurs within an <tt>extern "C"</tt> linkage specification,
541 it's name <i>will</i> be mangled in the same way as it would in
542 C.</li>
543</ul>
544
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]545<p>Query for this feature with __has_feature(attribute_overloadable).</p>
546
547
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]548<!-- ======================================================================= -->
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]549<h2 id="builtins">Builtin Functions</h2>
550<!-- ======================================================================= -->
551
552<p>Clang supports a number of builtin library functions with the same syntax as
553GCC, including things like <tt>__builtin_nan</tt>,
554<tt>__builtin_constant_p</tt>, <tt>__builtin_choose_expr</tt>,
555<tt>__builtin_types_compatible_p</tt>, <tt>__sync_fetch_and_add</tt>, etc. In
556addition to the GCC builtins, Clang supports a number of builtins that GCC does
557not, which are listed here.</p>
558
559<p>Please note that Clang does not and will not support all of the GCC builtins
560for vector operations. Instead of using builtins, you should use the functions
561defined in target-specific header files like <tt>&lt;xmmintrin.h&gt;</tt>, which
562define portable wrappers for these. Many of the Clang versions of these
563functions are implemented directly in terms of <a href="#vectors">extended
564vector support</a> instead of builtins, in order to reduce the number of
565builtins that we need to implement.</p>
566
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]567<!-- ======================================================================= -->
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]568<h3 id="__builtin_shufflevector">__builtin_shufflevector</h3>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]569<!-- ======================================================================= -->
570
Chris Lattneraad826b2009-09-16 18:56:12 +0000[diff] [blame]571<p><tt>__builtin_shufflevector</tt> is used to express generic vector
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]572permutation/shuffle/swizzle operations. This builtin is also very important for
573the implementation of various target-specific header files like
574<tt>&lt;xmmintrin.h&gt;</tt>.
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]575</p>
576
577<p><b>Syntax:</b></p>
578
579<pre>
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]580__builtin_shufflevector(vec1, vec2, index1, index2, ...)
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]581</pre>
582
583<p><b>Examples:</b></p>
584
585<pre>
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]586 // Identity operation - return 4-element vector V1.
587 __builtin_shufflevector(V1, V1, 0, 1, 2, 3)
588
589 // "Splat" element 0 of V1 into a 4-element result.
590 __builtin_shufflevector(V1, V1, 0, 0, 0, 0)
591
592 // Reverse 4-element vector V1.
593 __builtin_shufflevector(V1, V1, 3, 2, 1, 0)
594
595 // Concatenate every other element of 4-element vectors V1 and V2.
596 __builtin_shufflevector(V1, V2, 0, 2, 4, 6)
597
598 // Concatenate every other element of 8-element vectors V1 and V2.
599 __builtin_shufflevector(V1, V2, 0, 2, 4, 6, 8, 10, 12, 14)
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]600</pre>
601
602<p><b>Description:</b></p>
603
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]604<p>The first two arguments to __builtin_shufflevector are vectors that have the
605same element type. The remaining arguments are a list of integers that specify
606the elements indices of the first two vectors that should be extracted and
607returned in a new vector. These element indices are numbered sequentially
608starting with the first vector, continuing into the second vector. Thus, if
609vec1 is a 4-element vector, index 5 would refer to the second element of vec2.
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]610</p>
611
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]612<p>The result of __builtin_shufflevector is a vector
613with the same element type as vec1/vec2 but that has an element count equal to
614the number of indices specified.
615</p>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]616
Chris Lattner21190d52009-09-21 03:09:59 +0000[diff] [blame]617<p>Query for this feature with __has_builtin(__builtin_shufflevector).</p>
618
619<!-- ======================================================================= -->
620<h3 id="__builtin_unreachable">__builtin_unreachable</h3>
621<!-- ======================================================================= -->
622
623<p><tt>__builtin_unreachable</tt> is used to indicate that a specific point in
624the program cannot be reached, even if the compiler might otherwise think it
625can. This is useful to improve optimization and eliminates certain warnings.
626For example, without the <tt>__builtin_unreachable</tt> in the example below,
627the compiler assumes that the inline asm can fall through and prints a "function
628declared 'noreturn' should not return" warning.
629</p>
630
631<p><b>Syntax:</b></p>
632
633<pre>
634__builtin_unreachable()
635</pre>
636
637<p><b>Example of Use:</b></p>
638
639<pre>
640void myabort(void) __attribute__((noreturn));
641void myabort(void) {
642 asm("int3");
643 __builtin_unreachable();
644}
645</pre>
646
647<p><b>Description:</b></p>
648
649<p>The __builtin_unreachable() builtin has completely undefined behavior. Since
650it has undefined behavior, it is a statement that it is never reached and the
651optimizer can take advantage of this to produce better code. This builtin takes
652no arguments and produces a void result.
653</p>
654
655<p>Query for this feature with __has_builtin(__builtin_unreachable).</p>
656
657
Chris Lattner1177f912009-04-09 19:58:15 +0000[diff] [blame]658<!-- ======================================================================= -->
659<h2 id="targetspecific">Target-Specific Extensions</h2>
660<!-- ======================================================================= -->
661
662<p>Clang supports some language features conditionally on some targets.</p>
663
664<!-- ======================================================================= -->
665<h3 id="x86-specific">X86/X86-64 Language Extensions</h3>
666<!-- ======================================================================= -->
667
668<p>The X86 backend has these language extensions:</p>
669
670<!-- ======================================================================= -->
671<h4 id="x86-gs-segment">Memory references off the GS segment</h4>
672<!-- ======================================================================= -->
673
674<p>Annotating a pointer with address space #256 causes it to be code generated
Chris Lattnera021e7c2009-05-05 18:54:47 +0000[diff] [blame]675relative to the X86 GS segment register, and address space #257 causes it to be
676relative to the X86 FS segment. Note that this is a very very low-level
677feature that should only be used if you know what you're doing (for example in
678an OS kernel).</p>
Chris Lattner1177f912009-04-09 19:58:15 +0000[diff] [blame]679
680<p>Here is an example:</p>
681
682<pre>
683#define GS_RELATIVE __attribute__((address_space(256)))
684int foo(int GS_RELATIVE *P) {
685 return *P;
686}
687</pre>
688
689<p>Which compiles to (on X86-32):</p>
690
691<pre>
692_foo:
693 movl 4(%esp), %eax
694 movl %gs:(%eax), %eax
695 ret
696</pre>
697
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]698<!-- ======================================================================= -->
699<h2 id="analyzerspecific">Static Analysis-Specific Extensions</h2>
700<!-- ======================================================================= -->
701
702<p>Clang supports additional attributes that are useful for documenting program
703invariants and rules for static analysis tools. The extensions documented here
704are used by the <a
705href="http://clang.llvm.org/StaticAnalysis.html">path-sensitive static analyzer
706engine</a> that is part of Clang's Analysis library.</p>
707
708<!-- ======================================================================= -->
709<h3 id="analyzerattributes">Analyzer Attributes</h3>
710<!-- ======================================================================= -->
711
712<h4 id="attr_analyzer_noreturn"><tt>analyzer_noreturn</tt></h4>
713
714<p>Clang's static analysis engine understands the standard <tt>noreturn</tt>
Ted Kremenek4df21142009-04-10 05:04:22 +0000[diff] [blame]715attribute. This attribute, which is typically affixed to a function prototype,
716indicates that a call to a given function never returns. Function prototypes for
717common functions like <tt>exit</tt> are typically annotated with this attribute,
718as well as a variety of common assertion handlers. Users can educate the static
719analyzer about their own custom assertion handles (thus cutting down on false
720positives due to false paths) by marking their own &quot;panic&quot; functions
721with this attribute.</p>
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]722
723<p>While useful, <tt>noreturn</tt> is not applicable in all cases. Sometimes
Nick Lewycky625b5862009-06-14 04:08:08 +0000[diff] [blame]724there are special functions that for all intents and purposes should be
725considered panic functions (i.e., they are only called when an internal program
726error occurs) but may actually return so that the program can fail gracefully.
727The <tt>analyzer_noreturn</tt> attribute allows one to annotate such functions
728as being interpreted as &quot;no return&quot; functions by the analyzer (thus
Chris Lattner28935892009-04-10 05:54:56 +0000[diff] [blame]729pruning bogus paths) but will not affect compilation (as in the case of
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]730<tt>noreturn</tt>).</p>
731
732<p><b>Usage</b>: The <tt>analyzer_noreturn</tt> attribute can be placed in the
Chris Lattner28935892009-04-10 05:54:56 +0000[diff] [blame]733same places where the <tt>noreturn</tt> attribute can be placed. It is commonly
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]734placed at the end of function prototypes:</p>
735
736<pre>
737 void foo() <b>__attribute__((analyzer_noreturn))</b>;
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]738</pre>
739
740<p>Query for this feature with __has_feature(attribute_analyzer_noreturn).</p>
741
John McCall630b7ae2011-01-25 04:26:21 +0000[diff] [blame^]742<h4 id="attr_retain_release">Objective-C retaining behavior attributes</h4>
743
744<p>In Objective-C, functions and methods are generally assumed to take
745and return objects with +0 retain counts, with some exceptions for
746special methods like <tt>+alloc</tt> and <tt>init</tt>. However,
747there are exceptions, and so Clang provides attributes to allow these
748exceptions to be documented, which helps the analyzer find leaks (and
749ignore non-leaks).</p>
750
751<p><b>Usage</b>: The <tt>ns_returns_retained</tt>, <tt>ns_returns_not_retained</tt>,
752<tt>ns_returns_autoreleased</tt>, <tt>cf_returns_retained</tt>,
753and <tt>cf_returns_not_retained</tt> attributes can be placed on
754methods and functions that return Objective-C or CoreFoundation
755objects. They are commonly placed at the end of a function prototype
756or method declaration:</p>
757
758<pre>
759 id foo() <b>__attribute__((ns_returns_retained))</b>;
760
761 - (NSString*) bar: (int) x <b>__attribute__((ns_returns_retained))</b>;
762</pre>
763
764<p>The <tt>*_returns_retained</tt> attributes specify that the
765returned object has a +1 retain count.
766The <tt>*_returns_not_retained</tt> attributes specify that the return
767object has a +0 retain count, even if the normal convention for its
768selector would be +1. <tt>ns_returns_autoreleased</tt> specifies that the
769returned object is +0, but is guaranteed to live at least as long as the
770next flush of an autorelease pool.</p>
771
772<p><b>Usage</b>: The <tt>ns_consumed</tt> and <tt>cf_consumed</tt>
773attributes can be placed on an parameter declaration; they specify
774that the argument is expected to have a +1 retain count, which will be
775balanced in some way by the function or method.
776The <tt>ns_consumes_self</tt> attribute can only be placed on an
777Objective-C method; it specifies that the method expects
778its <tt>self</tt> parameter to have a +1 retain count, which it will
779balance in some way.</p>
780
781<pre>
782 void <b>foo(__attribute__((ns_consumed))</b> NSString *string);
783
784 - (void) bar <b>__attribute__((ns_consumes_self))</b>;
785 - (void) baz: (id) <b>__attribute__((ns_consumed))</b> x;
786</pre>
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]787
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]788</div>
789</body>
790</html>