Gitiles
Code Review Sign In
gerrit-public.fairphone.software / platform / external / clang / 7b156ddae822a546aa9e8b23d683ac76da8e8ac8 / . / docs / LanguageExtensions.html
blob: d6053b90fbff97dd8fbc2a8161b83781e3c94396 [file] [log] [blame]
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
3<!-- Material used from: HTML 4.01 specs: http://www.w3.org/TR/html401/ -->
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]4<html>
5<head>
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]6 <META http-equiv="Content-Type" content="text/html; charset=ISO-8859-1">
Eli Friedman0c706c22011-09-19 23:17:44 +0000[diff] [blame]7 <title>Clang Language Extensions</title>
Sean Hunt64f857b2011-06-23 01:22:53 +0000[diff] [blame]8 <link type="text/css" rel="stylesheet" href="../menu.css">
9 <link type="text/css" rel="stylesheet" href="../content.css">
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]10 <style type="text/css">
11 td {
12 vertical-align: top;
13 }
Benjamin Kramer3419d7c2012-01-15 16:42:14 +0000[diff] [blame]14 th { background-color: #ffddaa; }
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]15 </style>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]16</head>
17<body>
18
19<!--#include virtual="../menu.html.incl"-->
20
21<div id="content">
22
23<h1>Clang Language Extensions</h1>
24
25<ul>
26<li><a href="#intro">Introduction</a></li>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]27<li><a href="#feature_check">Feature Checking Macros</a></li>
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]28<li><a href="#has_include">Include File Checking Macros</a></li>
Chris Lattner81edc9f2009-04-13 02:45:46 +0000[diff] [blame]29<li><a href="#builtinmacros">Builtin Macros</a></li>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]30<li><a href="#vectors">Vectors and Extended Vectors</a></li>
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]31<li><a href="#deprecated">Messages on <tt>deprecated</tt> and <tt>unavailable</tt> attributes</a></li>
32<li><a href="#attributes-on-enumerators">Attributes on enumerators</a></li>
Douglas Gregor93a70672012-03-11 04:53:21 +0000[diff] [blame]33<li><a href="#availability">Availability attribute</a></li>
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]34<li><a href="#checking_language_features">Checks for Standard Language Features</a>
Ted Kremenek22c34102009-12-03 02:05:57 +0000[diff] [blame]35 <ul>
36 <li><a href="#cxx_exceptions">C++ exceptions</a></li>
37 <li><a href="#cxx_rtti">C++ RTTI</a></li>
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]38 </ul></li>
39<li><a href="#checking_upcoming_features">Checks for Upcoming Standard Language Features</a>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]40 <ul>
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]41 <li><a href="#cxx0x">C++11</a>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]42 <ul>
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]43 <li><a href="#cxx_access_control_sfinae">C++11 SFINAE includes
44 access control</a></li>
45 <li><a href="#cxx_alias_templates">C++11 alias templates</a></li>
46 <li><a href="#cxx_alignas">C++11 alignment specifiers</a></li>
47 <li><a href="#cxx_attributes">C++11 attributes</a></li>
48 <li><a href="#cxx_constexpr">C++11 generalized constant expressions</a></li>
49 <li><a href="#cxx_decltype">C++11 <tt>decltype()</tt></a></li>
50 <li><a href="#cxx_default_function_template_args">C++11 default template arguments in function templates</a></li>
Douglas Gregorf695a692011-11-01 01:19:34 +0000[diff] [blame]51 <li><a href="#cxx_defaulted_functions">C++11 defaulted functions</a></li>
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]52 <li><a href="#cxx_delegating_constructor">C++11 delegating constructors</a></li>
53 <li><a href="#cxx_deleted_functions">C++11 deleted functions</a></li>
54 <li><a href="#cxx_explicit_conversions">C++11 explicit conversion functions</a></li>
55 <li><a href="#cxx_generalized_initializers">C++11 generalized initializers</a></li>
56 <li><a href="#cxx_implicit_moves">C++11 implicit move constructors/assignment operators</a></li>
57 <li><a href="#cxx_inheriting_constructors">C++11 inheriting constructors</a></li>
58 <li><a href="#cxx_inline_namespaces">C++11 inline namespaces</a></li>
59 <li><a href="#cxx_lambdas">C++11 lambdas</a></li>
Douglas Gregor7b156dd2012-04-04 00:48:39 +0000[diff] [blame^]60 <li><a href="#cxx_local_type_template_args">C++11 local and unnamed types as template arguments</a></li>
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]61 <li><a href="#cxx_noexcept">C++11 noexcept specification</a></li>
62 <li><a href="#cxx_nonstatic_member_init">C++11 in-class non-static data member initialization</a></li>
63 <li><a href="#cxx_nullptr">C++11 nullptr</a></li>
64 <li><a href="#cxx_override_control">C++11 override control</a></li>
65 <li><a href="#cxx_range_for">C++11 range-based for loop</a></li>
66 <li><a href="#cxx_raw_string_literals">C++11 raw string literals</a></li>
67 <li><a href="#cxx_rvalue_references">C++11 rvalue references</a></li>
68 <li><a href="#cxx_reference_qualified_functions">C++11 reference-qualified functions</a></li>
69 <li><a href="#cxx_static_assert">C++11 <tt>static_assert()</tt></a></li>
70 <li><a href="#cxx_auto_type">C++11 type inference</a></li>
71 <li><a href="#cxx_strong_enums">C++11 strongly-typed enumerations</a></li>
72 <li><a href="#cxx_trailing_return">C++11 trailing return type</a></li>
73 <li><a href="#cxx_unicode_literals">C++11 Unicode string literals</a></li>
74 <li><a href="#cxx_unrestricted_unions">C++11 unrestricted unions</a></li>
75 <li><a href="#cxx_user_literals">C++11 user-defined literals</a></li>
76 <li><a href="#cxx_variadic_templates">C++11 variadic templates</a></li>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]77 </ul></li>
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]78 <li><a href="#c11">C11</a>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]79 <ul>
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]80 <li><a href="#c_alignas">C11 alignment specifiers</a></li>
81 <li><a href="#c_generic_selections">C11 generic selections</a></li>
82 <li><a href="#c_static_assert">C11 <tt>_Static_assert()</tt></a></li>
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]83 </ul></li>
84 </ul> </li>
Douglas Gregorafdf1372011-02-03 21:57:35 +0000[diff] [blame]85<li><a href="#checking_type_traits">Checks for Type Traits</a></li>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]86<li><a href="#blocks">Blocks</a></li>
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]87<li><a href="#objc_features">Objective-C Features</a>
88 <ul>
89 <li><a href="#objc_instancetype">Related result types</a></li>
John McCallf85e1932011-06-15 23:02:42 +0000[diff] [blame]90 <li><a href="#objc_arc">Automatic reference counting</a></li>
Douglas Gregor5471bc82011-09-08 17:18:35 +0000[diff] [blame]91 <li><a href="#objc_fixed_enum">Enumerations with a fixed underlying type</a></li>
Douglas Gregor8a4e1822012-03-09 23:24:48 +0000[diff] [blame]92 <li><a href="#objc_lambdas">Interoperability with C++11 lambdas</a></li>
Patrick Beard62f12342012-03-20 21:51:03 +0000[diff] [blame]93 <li><a href="#object-literals-subscripting">Object Literals and Subscripting</a></li>
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]94 </ul>
95</li>
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]96<li><a href="#overloading-in-c">Function Overloading in C</a></li>
Eli Friedman0c706c22011-09-19 23:17:44 +0000[diff] [blame]97<li><a href="#complex-list-init">Initializer lists for complex numbers in C</a></li>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]98<li><a href="#builtins">Builtin Functions</a>
99 <ul>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]100 <li><a href="#__builtin_shufflevector">__builtin_shufflevector</a></li>
Chris Lattner21190d52009-09-21 03:09:59 +0000[diff] [blame]101 <li><a href="#__builtin_unreachable">__builtin_unreachable</a></li>
Chris Lattner23aa9c82011-04-09 03:57:26 +0000[diff] [blame]102 <li><a href="#__sync_swap">__sync_swap</a></li>
Douglas Gregorafdf1372011-02-03 21:57:35 +0000[diff] [blame]103 </ul>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]104</li>
Chris Lattner1177f912009-04-09 19:58:15 +0000[diff] [blame]105<li><a href="#targetspecific">Target-Specific Extensions</a>
106 <ul>
107 <li><a href="#x86-specific">X86/X86-64 Language Extensions</a></li>
108 </ul>
109</li>
John McCall87494012011-03-18 03:51:49 +0000[diff] [blame]110<li><a href="#analyzerspecific">Static Analysis-Specific Extensions</a></li>
Benjamin Kramer665a8dc2012-01-15 15:26:07 +0000[diff] [blame]111<li><a href="#dynamicanalyzerspecific">Dynamic Analysis-Specific Extensions</a>
Kostya Serebryanyce98c9b2011-11-28 20:51:02 +0000[diff] [blame]112 <ul>
113 <li><a href="#address_sanitizer">AddressSanitizer</a></li>
114 </ul>
Benjamin Kramer665a8dc2012-01-15 15:26:07 +0000[diff] [blame]115</li>
116<li><a href="#threadsafety">Thread Safety Annotation Checking</a>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]117 <ul>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]118 <li><a href="#ts_noanal"><tt>no_thread_safety_analysis</tt></a></li>
119 <li><a href="#ts_lockable"><tt>lockable</tt></a></li>
120 <li><a href="#ts_scopedlockable"><tt>scoped_lockable</tt></a></li>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]121 <li><a href="#ts_guardedvar"><tt>guarded_var</tt></a></li>
122 <li><a href="#ts_ptguardedvar"><tt>pt_guarded_var</tt></a></li>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]123 <li><a href="#ts_guardedby"><tt>guarded_by(l)</tt></a></li>
124 <li><a href="#ts_ptguardedby"><tt>pt_guarded_by(l)</tt></a></li>
125 <li><a href="#ts_acquiredbefore"><tt>acquired_before(...)</tt></a></li>
126 <li><a href="#ts_acquiredafter"><tt>acquired_after(...)</tt></a></li>
127 <li><a href="#ts_elf"><tt>exclusive_lock_function(...)</tt></a></li>
128 <li><a href="#ts_slf"><tt>shared_lock_function(...)</tt></a></li>
129 <li><a href="#ts_etf"><tt>exclusive_trylock_function(...)</tt></a></li>
130 <li><a href="#ts_stf"><tt>shared_trylock_function(...)</tt></a></li>
131 <li><a href="#ts_uf"><tt>unlock_function(...)</tt></a></li>
132 <li><a href="#ts_lr"><tt>lock_returned(l)</tt></a></li>
133 <li><a href="#ts_le"><tt>locks_excluded(...)</tt></a></li>
134 <li><a href="#ts_elr"><tt>exclusive_locks_required(...)</tt></a></li>
135 <li><a href="#ts_slr"><tt>shared_locks_required(...)</tt></a></li>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]136 </ul>
Benjamin Kramer665a8dc2012-01-15 15:26:07 +0000[diff] [blame]137</li>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]138</ul>
139
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]140<!-- ======================================================================= -->
141<h2 id="intro">Introduction</h2>
142<!-- ======================================================================= -->
143
144<p>This document describes the language extensions provided by Clang. In
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]145addition to the language extensions listed here, Clang aims to support a broad
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]146range of GCC extensions. Please see the <a
147href="http://gcc.gnu.org/onlinedocs/gcc/C-Extensions.html">GCC manual</a> for
148more information on these extensions.</p>
149
150<!-- ======================================================================= -->
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]151<h2 id="feature_check">Feature Checking Macros</h2>
152<!-- ======================================================================= -->
153
154<p>Language extensions can be very useful, but only if you know you can depend
Chris Lattnerc70e1932011-03-21 16:25:11 +0000[diff] [blame]155on them. In order to allow fine-grain features checks, we support three builtin
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]156function-like macros. This allows you to directly test for a feature in your
157code without having to resort to something like autoconf or fragile "compiler
158version checks".</p>
159
160<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]161<h3><a name="__has_builtin">__has_builtin</a></h3>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]162<!-- ======================================================================= -->
163
164<p>This function-like macro takes a single identifier argument that is the name
165of a builtin function. It evaluates to 1 if the builtin is supported or 0 if
166not. It can be used like this:</p>
167
168<blockquote>
169<pre>
170#ifndef __has_builtin // Optional of course.
171 #define __has_builtin(x) 0 // Compatibility with non-clang compilers.
172#endif
173
174...
175#if __has_builtin(__builtin_trap)
176 __builtin_trap();
177#else
178 abort();
179#endif
180...
181</pre>
182</blockquote>
183
184
185<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]186<h3><a name="__has_feature_extension"> __has_feature and __has_extension</a></h3>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]187<!-- ======================================================================= -->
188
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]189<p>These function-like macros take a single identifier argument that is the
190name of a feature. <code>__has_feature</code> evaluates to 1 if the feature
191is both supported by Clang and standardized in the current language standard
192or 0 if not (but see <a href="#has_feature_back_compat">below</a>), while
193<code>__has_extension</code> evaluates to 1 if the feature is supported by
194Clang in the current language (either as a language extension or a standard
195language feature) or 0 if not. They can be used like this:</p>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]196
197<blockquote>
198<pre>
199#ifndef __has_feature // Optional of course.
200 #define __has_feature(x) 0 // Compatibility with non-clang compilers.
201#endif
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]202#ifndef __has_extension
203 #define __has_extension __has_feature // Compatibility with pre-3.0 compilers.
204#endif
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]205
206...
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]207#if __has_feature(cxx_rvalue_references)
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]208// This code will only be compiled with the -std=c++11 and -std=gnu++11
209// options, because rvalue references are only standardized in C++11.
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]210#endif
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]211
212#if __has_extension(cxx_rvalue_references)
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]213// This code will be compiled with the -std=c++11, -std=gnu++11, -std=c++98
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]214// and -std=gnu++98 options, because rvalue references are supported as a
215// language extension in C++98.
216#endif
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]217</pre>
218</blockquote>
219
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]220<p id="has_feature_back_compat">For backwards compatibility reasons,
221<code>__has_feature</code> can also be used to test for support for
222non-standardized features, i.e. features not prefixed <code>c_</code>,
223<code>cxx_</code> or <code>objc_</code>.</p>
224
Kostya Serebryanyce98c9b2011-11-28 20:51:02 +0000[diff] [blame]225<p id="has_feature_for_non_language_features">
226Another use of <code>__has_feature</code> is to check for compiler features
227not related to the language standard, such as e.g.
228<a href="AddressSanitizer.html">AddressSanitizer</a>.
229
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]230<p>If the <code>-pedantic-errors</code> option is given,
231<code>__has_extension</code> is equivalent to <code>__has_feature</code>.</p>
232
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]233<p>The feature tag is described along with the language feature below.</p>
234
Richard Smith5297d712012-02-25 10:41:10 +0000[diff] [blame]235<p>The feature name or extension name can also be specified with a preceding and
236following <code>__</code> (double underscore) to avoid interference from a macro
Richard Smith1d9f4c12012-03-01 02:12:07 +0000[diff] [blame]237with the same name. For instance, <code>__cxx_rvalue_references__</code> can be
238used instead of <code>cxx_rvalue_references</code>.</p>
Richard Smith5297d712012-02-25 10:41:10 +0000[diff] [blame]239
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]240<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]241<h3><a name="__has_attribute">__has_attribute</a></h3>
Anders Carlssoncae50952010-10-20 02:31:43 +0000[diff] [blame]242<!-- ======================================================================= -->
243
244<p>This function-like macro takes a single identifier argument that is the name
245of an attribute. It evaluates to 1 if the attribute is supported or 0 if not. It
246can be used like this:</p>
247
248<blockquote>
249<pre>
250#ifndef __has_attribute // Optional of course.
251 #define __has_attribute(x) 0 // Compatibility with non-clang compilers.
252#endif
253
254...
Anders Carlsson961003d2011-01-24 03:54:51 +0000[diff] [blame]255#if __has_attribute(always_inline)
256#define ALWAYS_INLINE __attribute__((always_inline))
Anders Carlssoncae50952010-10-20 02:31:43 +0000[diff] [blame]257#else
Anders Carlsson961003d2011-01-24 03:54:51 +0000[diff] [blame]258#define ALWAYS_INLINE
Anders Carlssoncae50952010-10-20 02:31:43 +0000[diff] [blame]259#endif
260...
261</pre>
262</blockquote>
263
Jean-Daniel Dupas8a5e7fd2012-03-01 14:53:16 +0000[diff] [blame]264<p>The attribute name can also be specified with a preceding and
265following <code>__</code> (double underscore) to avoid interference from a macro
266with the same name. For instance, <code>__always_inline__</code> can be used
267instead of <code>always_inline</code>.</p>
268
Anders Carlssoncae50952010-10-20 02:31:43 +0000[diff] [blame]269<!-- ======================================================================= -->
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]270<h2 id="has_include">Include File Checking Macros</h2>
271<!-- ======================================================================= -->
272
273<p>Not all developments systems have the same include files.
274The <a href="#__has_include">__has_include</a> and
275<a href="#__has_include_next">__has_include_next</a> macros allow you to
276check for the existence of an include file before doing
277a possibly failing #include directive.</p>
278
279<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]280<h3><a name="__has_include">__has_include</a></h3>
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]281<!-- ======================================================================= -->
282
283<p>This function-like macro takes a single file name string argument that
284is the name of an include file. It evaluates to 1 if the file can
285be found using the include paths, or 0 otherwise:</p>
286
287<blockquote>
288<pre>
289// Note the two possible file name string formats.
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]290#if __has_include("myinclude.h") &amp;&amp; __has_include(&lt;stdint.h&gt;)
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]291# include "myinclude.h"
292#endif
293
294// To avoid problem with non-clang compilers not having this macro.
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]295#if defined(__has_include) &amp;&amp; __has_include("myinclude.h")
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]296# include "myinclude.h"
297#endif
298</pre>
299</blockquote>
300
301<p>To test for this feature, use #if defined(__has_include).</p>
302
303<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]304<h3><a name="__has_include_next">__has_include_next</a></h3>
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]305<!-- ======================================================================= -->
306
307<p>This function-like macro takes a single file name string argument that
308is the name of an include file. It is like __has_include except that it
309looks for the second instance of the given file found in the include
310paths. It evaluates to 1 if the second instance of the file can
311be found using the include paths, or 0 otherwise:</p>
312
313<blockquote>
314<pre>
315// Note the two possible file name string formats.
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]316#if __has_include_next("myinclude.h") &amp;&amp; __has_include_next(&lt;stdint.h&gt;)
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]317# include_next "myinclude.h"
318#endif
319
320// To avoid problem with non-clang compilers not having this macro.
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]321#if defined(__has_include_next) &amp;&amp; __has_include_next("myinclude.h")
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]322# include_next "myinclude.h"
323#endif
324</pre>
325</blockquote>
326
327<p>Note that __has_include_next, like the GNU extension
328#include_next directive, is intended for use in headers only,
329and will issue a warning if used in the top-level compilation
330file. A warning will also be issued if an absolute path
331is used in the file argument.</p>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]332
Ted Kremenekd7681502011-10-12 19:46:30 +0000[diff] [blame]333
334<!-- ======================================================================= -->
335<h3><a name="__has_warning">__has_warning</a></h3>
336<!-- ======================================================================= -->
337
338<p>This function-like macro takes a string literal that represents a command
339 line option for a warning and returns true if that is a valid warning
340 option.</p>
341
342<blockquote>
343<pre>
344#if __has_warning("-Wformat")
345...
346#endif
347</pre>
348</blockquote>
349
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]350<!-- ======================================================================= -->
Chris Lattner81edc9f2009-04-13 02:45:46 +0000[diff] [blame]351<h2 id="builtinmacros">Builtin Macros</h2>
352<!-- ======================================================================= -->
353
Douglas Gregor4290fbd2010-04-30 02:51:06 +0000[diff] [blame]354<dl>
355 <dt><code>__BASE_FILE__</code></dt>
356 <dd>Defined to a string that contains the name of the main input
357 file passed to Clang.</dd>
358
359 <dt><code>__COUNTER__</code></dt>
360 <dd>Defined to an integer value that starts at zero and is
361 incremented each time the <code>__COUNTER__</code> macro is
362 expanded.</dd>
363
364 <dt><code>__INCLUDE_LEVEL__</code></dt>
365 <dd>Defined to an integral value that is the include depth of the
366 file currently being translated. For the main file, this value is
367 zero.</dd>
368
369 <dt><code>__TIMESTAMP__</code></dt>
370 <dd>Defined to the date and time of the last modification of the
371 current source file.</dd>
372
373 <dt><code>__clang__</code></dt>
374 <dd>Defined when compiling with Clang</dd>
375
376 <dt><code>__clang_major__</code></dt>
Chris Lattnerd4b66b92011-12-15 19:06:36 +0000[diff] [blame]377 <dd>Defined to the major marketing version number of Clang (e.g., the
378 2 in 2.0.1). Note that marketing version numbers should not be used to
379 check for language features, as different vendors use different numbering
380 schemes. Instead, use the <a href="#feature_check">feature checking
381 macros</a>.</dd>
Douglas Gregor4290fbd2010-04-30 02:51:06 +0000[diff] [blame]382
383 <dt><code>__clang_minor__</code></dt>
384 <dd>Defined to the minor version number of Clang (e.g., the 0 in
Chris Lattnerd4b66b92011-12-15 19:06:36 +0000[diff] [blame]385 2.0.1). Note that marketing version numbers should not be used to
386 check for language features, as different vendors use different numbering
387 schemes. Instead, use the <a href="#feature_check">feature checking
388 macros</a>.</dd>
Douglas Gregor4290fbd2010-04-30 02:51:06 +0000[diff] [blame]389
390 <dt><code>__clang_patchlevel__</code></dt>
Chris Lattnerd4b66b92011-12-15 19:06:36 +0000[diff] [blame]391 <dd>Defined to the marketing patch level of Clang (e.g., the 1 in 2.0.1).</dd>
Douglas Gregor4290fbd2010-04-30 02:51:06 +0000[diff] [blame]392
393 <dt><code>__clang_version__</code></dt>
Chris Lattnerd4b66b92011-12-15 19:06:36 +0000[diff] [blame]394 <dd>Defined to a string that captures the Clang marketing version, including
395 the Subversion tag or revision number, e.g., "1.5 (trunk 102332)".</dd>
Douglas Gregor4290fbd2010-04-30 02:51:06 +0000[diff] [blame]396</dl>
Chris Lattner81edc9f2009-04-13 02:45:46 +0000[diff] [blame]397
398<!-- ======================================================================= -->
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]399<h2 id="vectors">Vectors and Extended Vectors</h2>
400<!-- ======================================================================= -->
401
Anton Yartsevda90c772012-01-15 16:22:24 +0000[diff] [blame]402<p>Supports the GCC, OpenCL, AltiVec and NEON vector extensions.</p>
Owen Andersond2bf0cd2010-01-27 01:22:36 +0000[diff] [blame]403
Benjamin Kramer3419d7c2012-01-15 16:42:14 +0000[diff] [blame]404<p>OpenCL vector types are created using <tt>ext_vector_type</tt> attribute. It
405support for <tt>V.xyzw</tt> syntax and other tidbits as seen in OpenCL. An
406example is:</p>
Owen Andersond2bf0cd2010-01-27 01:22:36 +0000[diff] [blame]407
408<blockquote>
409<pre>
410typedef float float4 <b>__attribute__((ext_vector_type(4)))</b>;
411typedef float float2 <b>__attribute__((ext_vector_type(2)))</b>;
412
413float4 foo(float2 a, float2 b) {
414 float4 c;
415 c.xz = a;
416 c.yw = b;
417 return c;
418}
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]419</pre>
Owen Andersond2bf0cd2010-01-27 01:22:36 +0000[diff] [blame]420</blockquote>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]421
Benjamin Kramer3419d7c2012-01-15 16:42:14 +0000[diff] [blame]422<p>Query for this feature with
423<tt>__has_extension(attribute_ext_vector_type)</tt>.</p>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]424
Benjamin Kramer3419d7c2012-01-15 16:42:14 +0000[diff] [blame]425<p>Giving <tt>-faltivec</tt> option to clang enables support for AltiVec vector
426syntax and functions. For example:</p>
Anton Yartsevda90c772012-01-15 16:22:24 +0000[diff] [blame]427
428<blockquote>
429<pre>
430vector float foo(vector int a) {
431 vector int b;
432 b = vec_add(a, a) + a;
433 return (vector float)b;
434}
435</pre>
436</blockquote>
437
438<p>NEON vector types are created using <tt>neon_vector_type</tt> and
439<tt>neon_polyvector_type</tt> attributes. For example:</p>
440
441<blockquote>
442<pre>
443typedef <b>__attribute__((neon_vector_type(8)))</b> int8_t int8x8_t;
444typedef <b>__attribute__((neon_polyvector_type(16)))</b> poly8_t poly8x16_t;
445
446int8x8_t foo(int8x8_t a) {
447 int8x8_t v;
448 v = a;
449 return v;
450}
451</pre>
452</blockquote>
453
454<!-- ======================================================================= -->
455<h3><a name="vector_literals">Vector Literals</a></h3>
456<!-- ======================================================================= -->
457
458<p>Vector literals can be used to create vectors from a set of scalars, or
459vectors. Either parentheses or braces form can be used. In the parentheses form
460the number of literal values specified must be one, i.e. referring to a scalar
461value, or must match the size of the vector type being created. If a single
462scalar literal value is specified, the scalar literal value will be replicated
463to all the components of the vector type. In the brackets form any number of
464literals can be specified. For example:</p>
465
466<blockquote>
467<pre>
468typedef int v4si __attribute__((__vector_size__(16)));
469typedef float float4 __attribute__((ext_vector_type(4)));
470typedef float float2 __attribute__((ext_vector_type(2)));
471
472v4si vsi = (v4si){1, 2, 3, 4};
473float4 vf = (float4)(1.0f, 2.0f, 3.0f, 4.0f);
474vector int vi1 = (vector int)(1); // vi1 will be (1, 1, 1, 1).
475vector int vi2 = (vector int){1}; // vi2 will be (1, 0, 0, 0).
476vector int vi3 = (vector int)(1, 2); // error
477vector int vi4 = (vector int){1, 2}; // vi4 will be (1, 2, 0, 0).
478vector int vi5 = (vector int)(1, 2, 3, 4);
479float4 vf = (float4)((float2)(1.0f, 2.0f), (float2)(3.0f, 4.0f));
480</pre>
481</blockquote>
482
483<!-- ======================================================================= -->
484<h3><a name="vector_operations">Vector Operations</a></h3>
485<!-- ======================================================================= -->
486
487<p>The table below shows the support for each operation by vector extension.
488A dash indicates that an operation is not accepted according to a corresponding
489specification.</p>
490
491<table width="500" border="1" cellspacing="0">
492 <tr>
Benjamin Kramer3419d7c2012-01-15 16:42:14 +0000[diff] [blame]493 <th>Operator</th>
494 <th>OpenCL</th>
495 <th>AltiVec</th>
496 <th>GCC</th>
497 <th>NEON</th>
Anton Yartsevda90c772012-01-15 16:22:24 +0000[diff] [blame]498 </tr>
499 <tr>
500 <td>[]</td>
501 <td align="center">yes</td>
502 <td align="center">yes</td>
503 <td align="center">yes</td>
504 <td align="center">-</td>
505 </tr>
506 <tr>
507 <td>unary operators +, -</td>
508 <td align="center">yes</td>
509 <td align="center">yes</td>
510 <td align="center">yes</td>
511 <td align="center">-</td>
512 </tr>
513 <tr>
514 <td>++, --</td>
515 <td align="center">yes</td>
516 <td align="center">yes</td>
517 <td align="center">-</td>
518 <td align="center">-</td>
519 </tr>
520 <tr>
521 <td>+, -, *, /, %</td>
522 <td align="center">yes</td>
523 <td align="center">yes</td>
524 <td align="center">yes</td>
525 <td align="center">-</td>
526 </tr>
527 <tr>
528 <td>bitwise operators &, |, ^, ~</td>
529 <td align="center">yes</td>
530 <td align="center">yes</td>
531 <td align="center">yes</td>
532 <td align="center">-</td>
533 </tr>
534 <tr>
535 <td>&gt&gt, &lt&lt</td>
536 <td align="center">yes</td>
537 <td align="center">yes</td>
538 <td align="center">yes</td>
539 <td align="center">-</td>
540 </tr>
541 <tr>
542 <td>!, &&,||</td>
543 <td align="center">no</td>
544 <td align="center">-</td>
545 <td align="center">-</td>
546 <td align="center">-</td>
547 </tr>
548 <tr>
549 <td>==,!=, >, <, >=, <=</td>
550 <td align="center">yes</td>
551 <td align="center">yes</td>
552 <td align="center">-</td>
553 <td align="center">-</td>
554 </tr>
555 <tr>
556 <td>=</td>
557 <td align="center">yes</td>
558 <td align="center">yes</td>
559 <td align="center">yes</td>
560 <td align="center">yes</td>
561 </tr>
562 <tr>
563 <td>:?</td>
564 <td align="center">yes</td>
565 <td align="center">-</td>
566 <td align="center">-</td>
567 <td align="center">-</td>
568 </tr>
569 <tr>
570 <td>sizeof</td>
571 <td align="center">yes</td>
572 <td align="center">yes</td>
573 <td align="center">yes</td>
574 <td align="center">yes</td>
575 </tr>
576</table>
577
Owen Andersond2bf0cd2010-01-27 01:22:36 +0000[diff] [blame]578<p>See also <a href="#__builtin_shufflevector">__builtin_shufflevector</a>.</p>
579
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]580<!-- ======================================================================= -->
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]581<h2 id="deprecated">Messages on <tt>deprecated</tt> and <tt>unavailable</tt> Attributes</h2>
Fariborz Jahanianc784dc12010-10-06 23:12:32 +0000[diff] [blame]582<!-- ======================================================================= -->
583
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]584<p>An optional string message can be added to the <tt>deprecated</tt>
585and <tt>unavailable</tt> attributes. For example:</p>
Fariborz Jahanianc784dc12010-10-06 23:12:32 +0000[diff] [blame]586
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]587<blockquote>
Chris Lattner4836d6a2010-11-09 19:43:35 +0000[diff] [blame]588<pre>void explode(void) __attribute__((deprecated("extremely unsafe, use 'combust' instead!!!")));</pre>
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]589</blockquote>
590
591<p>If the deprecated or unavailable declaration is used, the message
592will be incorporated into the appropriate diagnostic:</p>
593
594<blockquote>
Benjamin Kramerb4556862012-03-19 19:12:30 +0000[diff] [blame]595<pre>harmless.c:4:3: warning: 'explode' is deprecated: extremely unsafe, use 'combust' instead!!!
596 [-Wdeprecated-declarations]
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]597 explode();
598 ^</pre>
599</blockquote>
600
601<p>Query for this feature
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]602with <tt>__has_extension(attribute_deprecated_with_message)</tt>
603and <tt>__has_extension(attribute_unavailable_with_message)</tt>.</p>
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]604
605<!-- ======================================================================= -->
606<h2 id="attributes-on-enumerators">Attributes on Enumerators</h2>
607<!-- ======================================================================= -->
608
609<p>Clang allows attributes to be written on individual enumerators.
610This allows enumerators to be deprecated, made unavailable, etc. The
611attribute must appear after the enumerator name and before any
612initializer, like so:</p>
613
614<blockquote>
615<pre>enum OperationMode {
616 OM_Invalid,
617 OM_Normal,
618 OM_Terrified __attribute__((deprecated)),
619 OM_AbortOnError __attribute__((deprecated)) = 4
620};</pre>
621</blockquote>
622
623<p>Attributes on the <tt>enum</tt> declaration do not apply to
624individual enumerators.</p>
625
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]626<p>Query for this feature with <tt>__has_extension(enumerator_attributes)</tt>.</p>
Fariborz Jahanianc784dc12010-10-06 23:12:32 +0000[diff] [blame]627
628<!-- ======================================================================= -->
Douglas Gregor93a70672012-03-11 04:53:21 +0000[diff] [blame]629<h2 id="availability">Availability attribute</h2
630<!-- ======================================================================= -->
631
632<p>Clang introduces the <code>availability</code> attribute, which can
633be placed on declarations to describe the lifecycle of that
634declaration relative to operating system versions. Consider the function declaration for a hypothetical function <code>f</code>:</p>
635
636<pre>
637void f(void) __attribute__((availability(macosx,introduced=10.4,deprecated=10.6,obsoleted=10.7)));
638</pre>
639
640<p>The availability attribute states that <code>f</code> was introduced in Mac OS X 10.4, deprecated in Mac OS X 10.6, and obsoleted in Mac OS X 10.7. This information is used by Clang to determine when it is safe to use <code>f</code>: for example, if Clang is instructed to compile code for Mac OS X 10.5, a call to <code>f()</code> succeeds. If Clang is instructed to compile code for Mac OS X 10.6, the call succeeds but Clang emits a warning specifying that the function is deprecated. Finally, if Clang is instructed to compile code for Mac OS X 10.7, the call fails because <code>f()</code> is no longer available.</p>
641
642<p>The availablility attribute is a comma-separated list starting with the platform name and then including clauses specifying important milestones in the declaration's lifetime (in any order) along with additional information. Those clauses can be:</p>
643
644<dl>
645 <dt>introduced=<i>version</i></dt>
646 <dd>The first version in which this declaration was introduced.</dd>
647
648 <dt>deprecated=<i>version</i></dt>
649 <dd>The first version in which this declaration was deprecated, meaning that users should migrate away from this API.</dd>
650
651 <dt>obsoleted=<i>version</i></dt>
652 <dd>The first version in which this declaration was obsoleted, meaning that it was removed completely and can no longer be used.</dd>
653
654 <dt>unavailable</dt>
655 <dd>This declaration is never available on this platform.</dd>
656
657 <dt>message=<i>string-literal</i></dt>
658 <dd>Additional message text that Clang will provide when emitting a warning or error about use of a deprecated or obsoleted declaration. Useful to direct users to replacement APIs.</dd>
659</dl>
660
661<p>Multiple availability attributes can be placed on a declaration, which may correspond to different platforms. Only the availability attribute with the platform corresponding to the target platform will be used; any others will be ignored. If no availability attribute specifies availability for the current target platform, the availability attributes are ignored. Supported platforms are:</p>
662
663<dl>
664 <dt>ios</dt>
665 <dd>Apple's iOS operating system. The minimum deployment target is specified by the <code>-mios-version-min=<i>version</i></code> or <code>-miphoneos-version-min=<i>version</i></code> command-line arguments.</dd>
666
667 <dt>macosx</dt>
668 <dd>Apple's Mac OS X operating system. The minimum deployment target is specified by the <code>-mmacosx-version-min=<i>version</i></code> command-line argument.</dd>
669</dl>
670
Douglas Gregor594f8412012-03-11 17:21:03 +0000[diff] [blame]671<p>A declaration can be used even when deploying back to a platform
672version prior to when the declaration was introduced. When this
673happens, the declaration is <a
674 href="https://developer.apple.com/library/mac/#documentation/MacOSX/Conceptual/BPFrameworks/Concepts/WeakLinking.html">weakly
675linked</a>, as if the <code>weak_import</code> attribute were added to the declaration. A weakly-linked declaration may or may not be present a run-time, and a program can determine whether the declaration is present by checking whether the address of that declaration is non-NULL.</p>
676
Douglas Gregor93a70672012-03-11 04:53:21 +0000[diff] [blame]677<!-- ======================================================================= -->
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]678<h2 id="checking_language_features">Checks for Standard Language Features</h2>
679<!-- ======================================================================= -->
680
681<p>The <tt>__has_feature</tt> macro can be used to query if certain standard language features are
682enabled. Those features are listed here.</p>
683
Ted Kremenek22c34102009-12-03 02:05:57 +0000[diff] [blame]684<h3 id="cxx_exceptions">C++ exceptions</h3>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]685
Ted Kremenek22c34102009-12-03 02:05:57 +0000[diff] [blame]686<p>Use <tt>__has_feature(cxx_exceptions)</tt> to determine if C++ exceptions have been enabled. For
Sean Hunt647ba1b2011-06-23 00:42:53 +0000[diff] [blame]687example, compiling code with <tt>-fexceptions</tt> enables C++ exceptions.</p>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]688
Ted Kremenek22c34102009-12-03 02:05:57 +0000[diff] [blame]689<h3 id="cxx_rtti">C++ RTTI</h3>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]690
Ted Kremenek0eb95602009-12-03 02:06:43 +0000[diff] [blame]691<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]692compiling code with <tt>-fno-rtti</tt> disables the use of RTTI.</p>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]693
694<!-- ======================================================================= -->
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]695<h2 id="checking_upcoming_features">Checks for Upcoming Standard Language Features</h2>
696<!-- ======================================================================= -->
697
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]698<p>The <tt>__has_feature</tt> or <tt>__has_extension</tt> macros can be used
699to query if certain upcoming standard language features are enabled. Those
700features are listed here. Features that are not yet implemented will be
701noted.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]702
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]703<h3 id="cxx0x">C++11</h3>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]704
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]705<p>The features listed below are slated for inclusion in the upcoming
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]706C++11 standard. As a result, all these features are enabled
707with the <tt>-std=c++11</tt> option when compiling C++ code.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]708
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]709<h4 id="cxx_access_control_sfinae">C++11 SFINAE includes access control</h4>
Douglas Gregor7822ee32011-05-11 23:45:11 +0000[diff] [blame]710
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]711<p>Use <tt>__has_feature(cxx_access_control_sfinae)</tt> or <tt>__has_extension(cxx_access_control_sfinae)</tt> to determine whether access-control errors (e.g., calling a private constructor) are considered to be template argument deduction errors (aka SFINAE errors), per <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#1170">C++ DR1170</a>.</p>
Douglas Gregor7822ee32011-05-11 23:45:11 +0000[diff] [blame]712
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]713<h4 id="cxx_alias_templates">C++11 alias templates</h4>
Richard Smith3e4c6c42011-05-05 21:57:07 +0000[diff] [blame]714
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]715<p>Use <tt>__has_feature(cxx_alias_templates)</tt> or
716<tt>__has_extension(cxx_alias_templates)</tt> to determine if support for
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]717C++11's alias declarations and alias templates is enabled.</p>
Richard Smith3e4c6c42011-05-05 21:57:07 +0000[diff] [blame]718
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]719<h4 id="cxx_alignas">C++11 alignment specifiers</h4>
Peter Collingbournefd5f6862011-10-14 23:44:46 +0000[diff] [blame]720
721<p>Use <tt>__has_feature(cxx_alignas)</tt> or
722<tt>__has_extension(cxx_alignas)</tt> to determine if support for alignment
723specifiers using <tt>alignas</tt> is enabled.</p>
724
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]725<h4 id="cxx_attributes">C++11 attributes</h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]726
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]727<p>Use <tt>__has_feature(cxx_attributes)</tt> or
728<tt>__has_extension(cxx_attributes)</tt> to determine if support for attribute
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]729parsing with C++11's square bracket notation is enabled.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]730
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]731<h4 id="cxx_constexpr">C++11 generalized constant expressions</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]732
733<p>Use <tt>__has_feature(cxx_constexpr)</tt> to determine if support
734for generalized constant expressions (e.g., <tt>constexpr</tt>) is
Richard Smithb5216aa2012-02-14 22:56:17 +0000[diff] [blame]735enabled.</p>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]736
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]737<h4 id="cxx_decltype">C++11 <tt>decltype()</tt></h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]738
739<p>Use <tt>__has_feature(cxx_decltype)</tt> or
740<tt>__has_extension(cxx_decltype)</tt> to determine if support for the
741<tt>decltype()</tt> specifier is enabled.</p>
742
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]743<h4 id="cxx_default_function_template_args">C++11 default template arguments in function templates</h4>
Douglas Gregor07508002011-02-05 20:35:30 +0000[diff] [blame]744
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]745<p>Use <tt>__has_feature(cxx_default_function_template_args)</tt> or
746<tt>__has_extension(cxx_default_function_template_args)</tt> to determine
747if support for default template arguments in function templates is enabled.</p>
Douglas Gregor07508002011-02-05 20:35:30 +0000[diff] [blame]748
Douglas Gregorf695a692011-11-01 01:19:34 +0000[diff] [blame]749<h4 id="cxx_defaulted_functions">C++11 <tt>default</tt>ed functions</h4>
750
751<p>Use <tt>__has_feature(cxx_defaulted_functions)</tt> or
752<tt>__has_extension(cxx_defaulted_functions)</tt> to determine if support for
753defaulted function definitions (with <tt>= default</tt>) is enabled.</p>
754
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]755<h4 id="cxx_delegating_constructors">C++11 delegating constructors</h4>
Sean Huntd9624992011-06-23 06:11:37 +0000[diff] [blame]756
757<p>Use <tt>__has_feature(cxx_delegating_constructors)</tt> to determine if
758support for delegating constructors is enabled.</p>
759
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]760<h4 id="cxx_deleted_functions">C++11 <tt>delete</tt>d functions</h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]761
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]762<p>Use <tt>__has_feature(cxx_deleted_functions)</tt> or
763<tt>__has_extension(cxx_deleted_functions)</tt> to determine if support for
Sebastian Redlf6c09772010-08-31 23:28:47 +0000[diff] [blame]764deleted function definitions (with <tt>= delete</tt>) is enabled.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]765
Benjamin Kramer665a8dc2012-01-15 15:26:07 +0000[diff] [blame]766<h4 id="cxx_explicit_conversions">C++11 explicit conversion functions</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]767<p>Use <tt>__has_feature(cxx_explicit_conversions)</tt> to determine if support for <tt>explicit</tt> conversion functions is enabled.</p>
768
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]769<h4 id="cxx_generalized_initializers">C++11 generalized initializers</h4>
Sean Hunte1f6dea2011-08-07 00:34:32 +0000[diff] [blame]770
771<p>Use <tt>__has_feature(cxx_generalized_initializers)</tt> to determine if
772support for generalized initializers (using braced lists and
Richard Smith88189552012-02-26 07:09:21 +0000[diff] [blame]773<tt>std::initializer_list</tt>) is enabled.</p>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]774
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]775<h4 id="cxx_implicit_moves">C++11 implicit move constructors/assignment operators</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]776
Sebastian Redl72a81d22011-10-10 18:10:00 +0000[diff] [blame]777<p>Use <tt>__has_feature(cxx_implicit_moves)</tt> to determine if Clang will
778implicitly generate move constructors and move assignment operators where needed.</p>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]779
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]780<h4 id="cxx_inheriting_constructors">C++11 inheriting constructors</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]781
782<p>Use <tt>__has_feature(cxx_inheriting_constructors)</tt> to determine if support for inheriting constructors is enabled. Clang does not currently implement this feature.</p>
783
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]784<h4 id="cxx_inline_namespaces">C++11 inline namespaces</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]785
786<p>Use <tt>__has_feature(cxx_inline_namespaces)</tt> or
787<tt>__has_extension(cxx_inline_namespaces)</tt> to determine if support for
788inline namespaces is enabled.</p>
Sean Hunte1f6dea2011-08-07 00:34:32 +0000[diff] [blame]789
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]790<h4 id="cxx_lambdas">C++11 lambdas</h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]791
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]792<p>Use <tt>__has_feature(cxx_lambdas)</tt> or
793<tt>__has_extension(cxx_lambdas)</tt> to determine if support for lambdas
Douglas Gregor46e021e2012-02-23 05:44:09 +0000[diff] [blame]794is enabled. </p>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]795
Douglas Gregor7b156dd2012-04-04 00:48:39 +0000[diff] [blame^]796<h4 id="cxx_local_type_template_args">C++11 local and unnamed types as template arguments</h4>
797
798<p>Use <tt>__has_feature(cxx_local_type_template_args)</tt> or
799<tt>__has_extension(cxx_local_type_template_args)</tt> to determine if
800support for local and unnamed types as template arguments is enabled.</p>
801
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]802<h4 id="cxx_noexcept">C++11 noexcept</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]803
804<p>Use <tt>__has_feature(cxx_noexcept)</tt> or
805<tt>__has_extension(cxx_noexcept)</tt> to determine if support for noexcept
806exception specifications is enabled.</p>
807
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]808<h4 id="cxx_nonstatic_member_init">C++11 in-class non-static data member initialization</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]809
810<p>Use <tt>__has_feature(cxx_nonstatic_member_init)</tt> to determine whether in-class initialization of non-static data members is enabled.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]811
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]812<h4 id="cxx_nullptr">C++11 <tt>nullptr</tt></h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]813
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]814<p>Use <tt>__has_feature(cxx_nullptr)</tt> or
815<tt>__has_extension(cxx_nullptr)</tt> to determine if support for
Douglas Gregor84ee2ee2011-05-21 23:15:46 +0000[diff] [blame]816<tt>nullptr</tt> is enabled.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]817
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]818<h4 id="cxx_override_control">C++11 <tt>override control</tt></h4>
Anders Carlssonc8b9f792011-03-25 15:04:23 +0000[diff] [blame]819
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]820<p>Use <tt>__has_feature(cxx_override_control)</tt> or
821<tt>__has_extension(cxx_override_control)</tt> to determine if support for
Anders Carlssonc8b9f792011-03-25 15:04:23 +0000[diff] [blame]822the override control keywords is enabled.</p>
823
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]824<h4 id="cxx_reference_qualified_functions">C++11 reference-qualified functions</h4>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]825<p>Use <tt>__has_feature(cxx_reference_qualified_functions)</tt> or
826<tt>__has_extension(cxx_reference_qualified_functions)</tt> to determine
827if support for reference-qualified functions (e.g., member functions with
828<code>&amp;</code> or <code>&amp;&amp;</code> applied to <code>*this</code>)
829is enabled.</p>
Douglas Gregor56209ff2011-01-26 21:25:54 +0000[diff] [blame]830
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]831<h4 id="cxx_range_for">C++11 range-based <tt>for</tt> loop</h4>
Richard Smitha391a462011-04-15 15:14:40 +0000[diff] [blame]832
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]833<p>Use <tt>__has_feature(cxx_range_for)</tt> or
834<tt>__has_extension(cxx_range_for)</tt> to determine if support for the
835range-based for loop is enabled. </p>
Richard Smitha391a462011-04-15 15:14:40 +0000[diff] [blame]836
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]837<h4 id="cxx_raw_string_literals">C++11 raw string literals</h4>
Richard Smith80134582012-03-07 08:57:31 +0000[diff] [blame]838<p>Use <tt>__has_feature(cxx_raw_string_literals)</tt> to determine if support
839for raw string literals (e.g., <tt>R"x(foo\bar)x"</tt>) is enabled.</p>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]840
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]841<h4 id="cxx_rvalue_references">C++11 rvalue references</h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]842
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]843<p>Use <tt>__has_feature(cxx_rvalue_references)</tt> or
844<tt>__has_extension(cxx_rvalue_references)</tt> to determine if support for
Douglas Gregor56209ff2011-01-26 21:25:54 +0000[diff] [blame]845rvalue references is enabled. </p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]846
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]847<h4 id="cxx_static_assert">C++11 <tt>static_assert()</tt></h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]848
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]849<p>Use <tt>__has_feature(cxx_static_assert)</tt> or
850<tt>__has_extension(cxx_static_assert)</tt> to determine if support for
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]851compile-time assertions using <tt>static_assert</tt> is enabled.</p>
852
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]853<h4 id="cxx_auto_type">C++11 type inference</h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]854
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]855<p>Use <tt>__has_feature(cxx_auto_type)</tt> or
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]856<tt>__has_extension(cxx_auto_type)</tt> to determine C++11 type inference is
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]857supported using the <tt>auto</tt> specifier. If this is disabled, <tt>auto</tt>
858will instead be a storage class specifier, as in C or C++98.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]859
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]860<h4 id="cxx_strong_enums">C++11 strongly typed enumerations</h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]861
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]862<p>Use <tt>__has_feature(cxx_strong_enums)</tt> or
863<tt>__has_extension(cxx_strong_enums)</tt> to determine if support for
864strongly typed, scoped enumerations is enabled.</p>
Sebastian Redlf6c09772010-08-31 23:28:47 +0000[diff] [blame]865
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]866<h4 id="cxx_trailing_return">C++11 trailing return type</h4>
Douglas Gregordab60ad2010-10-01 18:44:50 +0000[diff] [blame]867
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]868<p>Use <tt>__has_feature(cxx_trailing_return)</tt> or
869<tt>__has_extension(cxx_trailing_return)</tt> to determine if support for the
870alternate function declaration syntax with trailing return type is enabled.</p>
Douglas Gregordab60ad2010-10-01 18:44:50 +0000[diff] [blame]871
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]872<h4 id="cxx_unicode_literals">C++11 Unicode string literals</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]873<p>Use <tt>__has_feature(cxx_unicode_literals)</tt> to determine if
874support for Unicode string literals is enabled.</p>
Sebastian Redl4561ecd2011-03-15 21:17:12 +0000[diff] [blame]875
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]876<h4 id="cxx_unrestricted_unions">C++11 unrestricted unions</h4>
Sebastian Redl4561ecd2011-03-15 21:17:12 +0000[diff] [blame]877
Richard Smithec92bc72012-03-03 23:51:05 +0000[diff] [blame]878<p>Use <tt>__has_feature(cxx_unrestricted_unions)</tt> to determine if support for unrestricted unions is enabled.</p>
Douglas Gregor1274ccd2010-10-08 23:50:27 +0000[diff] [blame]879
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]880<h4 id="cxx_user_literals">C++11 user-defined literals</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]881
Richard Smith9c1dda72012-03-09 08:41:27 +0000[diff] [blame]882<p>Use <tt>__has_feature(cxx_user_literals)</tt> to determine if support for user-defined literals is enabled.</p>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]883
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]884<h4 id="cxx_variadic_templates">C++11 variadic templates</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]885
886<p>Use <tt>__has_feature(cxx_variadic_templates)</tt> or
887<tt>__has_extension(cxx_variadic_templates)</tt> to determine if support
888for variadic templates is enabled.</p>
Douglas Gregor1274ccd2010-10-08 23:50:27 +0000[diff] [blame]889
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]890<h3 id="c11">C11</h3>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]891
892<p>The features listed below are slated for inclusion in the upcoming
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]893C11 standard. As a result, all these features are enabled
894with the <tt>-std=c11</tt> option when compiling C code.</p>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]895
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]896<h4 id="c_alignas">C11 alignment specifiers</h4>
Peter Collingbournefd5f6862011-10-14 23:44:46 +0000[diff] [blame]897
898<p>Use <tt>__has_feature(c_alignas)</tt> or <tt>__has_extension(c_alignas)</tt>
899to determine if support for alignment specifiers using <tt>_Alignas</tt>
900is enabled.</p>
901
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]902<h4 id="c_generic_selections">C11 generic selections</h4>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]903
904<p>Use <tt>__has_feature(c_generic_selections)</tt> or
905<tt>__has_extension(c_generic_selections)</tt> to determine if support for
906generic selections is enabled.</p>
907
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]908<p>As an extension, the C11 generic selection expression is available in all
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]909languages supported by Clang. The syntax is the same as that given in the
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]910C11 standard.</p>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]911
912<p>In C, type compatibility is decided according to the rules given in the
913appropriate standard, but in C++, which lacks the type compatibility rules
914used in C, types are considered compatible only if they are equivalent.</p>
915
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]916<h4 id="c_static_assert">C11 <tt>_Static_assert()</tt></h4>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]917
918<p>Use <tt>__has_feature(c_static_assert)</tt> or
919<tt>__has_extension(c_static_assert)</tt> to determine if support for
920compile-time assertions using <tt>_Static_assert</tt> is enabled.</p>
921
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]922<!-- ======================================================================= -->
Douglas Gregorafdf1372011-02-03 21:57:35 +0000[diff] [blame]923<h2 id="checking_type_traits">Checks for Type Traits</h2>
924<!-- ======================================================================= -->
925
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]926<p>Clang supports the <a href="http://gcc.gnu.org/onlinedocs/gcc/Type-Traits.html">GNU C++ type traits</a> and a subset of the <a href="http://msdn.microsoft.com/en-us/library/ms177194(v=VS.100).aspx">Microsoft Visual C++ Type traits</a>. For each supported type trait <code>__X</code>, <code>__has_extension(X)</code> indicates the presence of the type trait. For example:
Douglas Gregorafdf1372011-02-03 21:57:35 +0000[diff] [blame]927<blockquote>
928<pre>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]929#if __has_extension(is_convertible_to)
Douglas Gregorafdf1372011-02-03 21:57:35 +0000[diff] [blame]930template&lt;typename From, typename To&gt;
931struct is_convertible_to {
932 static const bool value = __is_convertible_to(From, To);
933};
934#else
935// Emulate type trait
936#endif
937</pre>
938</blockquote>
939
940<p>The following type traits are supported by Clang:</p>
941<ul>
942 <li><code>__has_nothrow_assign</code> (GNU, Microsoft)</li>
943 <li><code>__has_nothrow_copy</code> (GNU, Microsoft)</li>
944 <li><code>__has_nothrow_constructor</code> (GNU, Microsoft)</li>
945 <li><code>__has_trivial_assign</code> (GNU, Microsoft)</li>
946 <li><code>__has_trivial_copy</code> (GNU, Microsoft)</li>
947 <li><code>__has_trivial_constructor</code> (GNU, Microsoft)</li>
948 <li><code>__has_trivial_destructor</code> (GNU, Microsoft)</li>
949 <li><code>__has_virtual_destructor</code> (GNU, Microsoft)</li>
950 <li><code>__is_abstract</code> (GNU, Microsoft)</li>
951 <li><code>__is_base_of</code> (GNU, Microsoft)</li>
952 <li><code>__is_class</code> (GNU, Microsoft)</li>
953 <li><code>__is_convertible_to</code> (Microsoft)</li>
954 <li><code>__is_empty</code> (GNU, Microsoft)</li>
955 <li><code>__is_enum</code> (GNU, Microsoft)</li>
956 <li><code>__is_pod</code> (GNU, Microsoft)</li>
957 <li><code>__is_polymorphic</code> (GNU, Microsoft)</li>
958 <li><code>__is_union</code> (GNU, Microsoft)</li>
959 <li><code>__is_literal(type)</code>: Determines whether the given type is a literal type</li>
Douglas Gregor5e9392b2011-12-03 18:14:24 +0000[diff] [blame]960 <li><code>__is_final</code>: Determines whether the given type is declared with a <code>final</code> class-virt-specifier.</li>
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]961 <li><code>__underlying_type(type)</code>: Retrieves the underlying type for a given <code>enum</code> type. This trait is required to implement the C++11 standard library.</li>
Douglas Gregor4ca8ac22012-02-24 07:38:34 +0000[diff] [blame]962 <li><code>__is_trivially_assignable(totype, fromtype)</code>: Determines whether a value of type <tt>totype</tt> can be assigned to from a value of type <tt>fromtype</tt> such that no non-trivial functions are called as part of that assignment. This trait is required to implement the C++11 standard library.</li>
963 <li><code>__is_trivially_constructible(type, argtypes...)</code>: Determines whether a value of type <tt>type</tt> can be direct-initialized with arguments of types <tt>argtypes...</tt> such that no non-trivial functions are called as part of that initialization. This trait is required to implement the C++11 standard library.</li>
Douglas Gregorafdf1372011-02-03 21:57:35 +0000[diff] [blame]964</ul>
965
966<!-- ======================================================================= -->
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]967<h2 id="blocks">Blocks</h2>
968<!-- ======================================================================= -->
969
Chris Lattnera7dbdf52009-03-09 07:03:22 +0000[diff] [blame]970<p>The syntax and high level language feature description is in <a
971href="BlockLanguageSpec.txt">BlockLanguageSpec.txt</a>. Implementation and ABI
972details for the clang implementation are in <a
Chris Lattner5d7650b2010-03-16 21:43:03 +0000[diff] [blame]973href="Block-ABI-Apple.txt">Block-ABI-Apple.txt</a>.</p>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]974
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]975
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]976<p>Query for this feature with __has_extension(blocks).</p>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]977
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]978<!-- ======================================================================= -->
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]979<h2 id="objc_features">Objective-C Features</h2>
980<!-- ======================================================================= -->
981
982<h3 id="objc_instancetype">Related result types</h3>
983
984<p>According to Cocoa conventions, Objective-C methods with certain names ("init", "alloc", etc.) always return objects that are an instance of the receiving class's type. Such methods are said to have a "related result type", meaning that a message send to one of these methods will have the same static type as an instance of the receiver class. For example, given the following classes:</p>
985
986<blockquote>
987<pre>
988@interface NSObject
989+ (id)alloc;
990- (id)init;
991@end
992
993@interface NSArray : NSObject
994@end
995</pre>
996</blockquote>
997
998<p>and this common initialization pattern</p>
999
1000<blockquote>
1001<pre>
1002NSArray *array = [[NSArray alloc] init];
1003</pre>
1004</blockquote>
1005
1006<p>the type of the expression <code>[NSArray alloc]</code> is
1007<code>NSArray*</code> because <code>alloc</code> implicitly has a
1008related result type. Similarly, the type of the expression
1009<code>[[NSArray alloc] init]</code> is <code>NSArray*</code>, since
1010<code>init</code> has a related result type and its receiver is known
1011to have the type <code>NSArray *</code>. If neither <code>alloc</code> nor <code>init</code> had a related result type, the expressions would have had type <code>id</code>, as declared in the method signature.</p>
1012
Douglas Gregore97179c2011-09-08 01:46:34 +0000[diff] [blame]1013<p>A method with a related result type can be declared by using the
1014type <tt>instancetype</tt> as its result type. <tt>instancetype</tt>
1015is a contextual keyword that is only permitted in the result type of
1016an Objective-C method, e.g.</p>
1017
1018<pre>
1019@interface A
1020+ (<b>instancetype</b>)constructAnA;
1021@end
1022</pre>
1023
1024<p>The related result type can also be inferred for some methods.
1025To determine whether a method has an inferred related result type, the first
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]1026word in the camel-case selector (e.g., "init" in "initWithObjects") is
Douglas Gregor8a0ace62011-11-03 18:33:01 +0000[diff] [blame]1027considered, and the method will have a related result type if its return
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]1028type is compatible with the type of its class and if</p>
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]1029
1030<ul>
1031
1032 <li>the first word is "alloc" or "new", and the method is a class
1033 method, or</li>
1034
1035 <li>the first word is "autorelease", "init", "retain", or "self",
1036 and the method is an instance method.</li>
1037
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]1038</ul>
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]1039
1040<p>If a method with a related result type is overridden by a subclass
1041method, the subclass method must also return a type that is compatible
1042with the subclass type. For example:</p>
1043
1044<blockquote>
1045<pre>
1046@interface NSString : NSObject
1047- (NSUnrelated *)init; // incorrect usage: NSUnrelated is not NSString or a superclass of NSString
1048@end
1049</pre>
1050</blockquote>
1051
1052<p>Related result types only affect the type of a message send or
1053property access via the given method. In all other respects, a method
Douglas Gregore97179c2011-09-08 01:46:34 +0000[diff] [blame]1054with a related result type is treated the same way as method that
1055returns <tt>id</tt>.</p>
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]1056
Douglas Gregoraebb6532011-09-08 17:19:31 +0000[diff] [blame]1057<p>Use <tt>__has_feature(objc_instancetype)</tt> to determine whether
1058the <tt>instancetype</tt> contextual keyword is available.</p>
1059
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]1060<!-- ======================================================================= -->
John McCallf85e1932011-06-15 23:02:42 +0000[diff] [blame]1061<h2 id="objc_arc">Automatic reference counting </h2>
1062<!-- ======================================================================= -->
1063
1064<p>Clang provides support for <a href="AutomaticReferenceCounting.html">automated reference counting</a> in Objective-C, which eliminates the need for manual retain/release/autorelease message sends. There are two feature macros associated with automatic reference counting: <code>__has_feature(objc_arc)</code> indicates the availability of automated reference counting in general, while <code>__has_feature(objc_arc_weak)</code> indicates that automated reference counting also includes support for <code>__weak</code> pointers to Objective-C objects.</p>
1065
1066<!-- ======================================================================= -->
Douglas Gregor5471bc82011-09-08 17:18:35 +0000[diff] [blame]1067<h2 id="objc_fixed_enum">Enumerations with a fixed underlying type</h2>
1068<!-- ======================================================================= -->
1069
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]1070<p>Clang provides support for C++11 enumerations with a fixed
Douglas Gregor5471bc82011-09-08 17:18:35 +0000[diff] [blame]1071underlying type within Objective-C. For example, one can write an
1072enumeration type as:</p>
1073
1074<pre>
1075typedef enum : unsigned char { Red, Green, Blue } Color;
1076</pre>
1077
1078<p>This specifies that the underlying type, which is used to store the
1079enumeration value, is <tt>unsigned char</tt>.</p>
1080
1081<p>Use <tt>__has_feature(objc_fixed_enum)</tt> to determine whether
1082support for fixed underlying types is available in Objective-C.</p>
1083
1084<!-- ======================================================================= -->
Douglas Gregor8a4e1822012-03-09 23:24:48 +0000[diff] [blame]1085<h2 id="objc_lambdas">Interoperability with C++11 lambdas</h2>
1086<!-- ======================================================================= -->
1087
1088<p>Clang provides interoperability between C++11 lambdas and
1089blocks-based APIs, by permitting a lambda to be implicitly converted
1090to a block pointer with the corresponding signature. For example,
1091consider an API such as <code>NSArray</code>'s array-sorting
1092method:</p>
1093
1094<pre> - (NSArray *)sortedArrayUsingComparator:(NSComparator)cmptr; </pre>
1095
1096<p><code>NSComparator</code> is simply a typedef for the block pointer
1097<code>NSComparisonResult (^)(id, id)</code>, and parameters of this
1098type are generally provided with block literals as arguments. However,
1099one can also use a C++11 lambda so long as it provides the same
1100signature (in this case, accepting two parameters of type
1101<code>id</code> and returning an <code>NSComparisonResult</code>):</p>
1102
1103<pre>
1104 NSArray *array = @[@"string 1", @"string 21", @"string 12", @"String 11",
1105 @"String 02"];
1106 const NSStringCompareOptions comparisonOptions
1107 = NSCaseInsensitiveSearch | NSNumericSearch |
1108 NSWidthInsensitiveSearch | NSForcedOrderingSearch;
1109 NSLocale *currentLocale = [NSLocale currentLocale];
1110 NSArray *sorted
1111 = [array sortedArrayUsingComparator:<b>[=](id s1, id s2) -&gt; NSComparisonResult {
1112 NSRange string1Range = NSMakeRange(0, [s1 length]);
1113 return [s1 compare:s2 options:comparisonOptions
1114 range:string1Range locale:currentLocale];
1115 }</b>];
1116 NSLog(@"sorted: %@", sorted);
1117</pre>
1118
1119<p>This code relies on an implicit conversion from the type of the
1120lambda expression (an unnamed, local class type called the <i>closure
1121type</i>) to the corresponding block pointer type. The conversion
1122itself is expressed by a conversion operator in that closure type
1123that produces a block pointer with the same signature as the lambda
1124itself, e.g.,</p>
1125
1126<pre>
1127 operator NSComparisonResult (^)(id, id)() const;
1128</pre>
1129
1130<p>This conversion function returns a new block that simply forwards
1131the two parameters to the lambda object (which it captures by copy),
1132then returns the result. The returned block is first copied (with
1133<tt>Block_copy</tt>) and then autoreleased. As an optimization, if a
1134lambda expression is immediately converted to a block pointer (as in
1135the first example, above), then the block is not copied and
1136autoreleased: rather, it is given the same lifetime as a block literal
1137written at that point in the program, which avoids the overhead of
1138copying a block to the heap in the common case.</p>
1139
Douglas Gregorbccda482012-03-10 22:20:11 +0000[diff] [blame]1140<p>The conversion from a lambda to a block pointer is only available
1141in Objective-C++, and not in C++ with blocks, due to its use of
1142Objective-C memory management (autorelease).</p>
1143
Douglas Gregor8a4e1822012-03-09 23:24:48 +0000[diff] [blame]1144<!-- ======================================================================= -->
Patrick Beard62f12342012-03-20 21:51:03 +0000[diff] [blame]1145<h2 id="object-literals-subscripting">Object Literals and Subscripting</h2>
1146<!-- ======================================================================= -->
1147
Patrick Beard1526f542012-03-21 16:22:53 +0000[diff] [blame]1148<p>Clang provides support for <a href="ObjectiveCLiterals.html">Object Literals and Subscripting</a> in Objective-C, which simplifies common Objective-C programming patterns, makes programs more concise, and improves the safety of container creation. There are several feature macros associated with object literals and subscripting: <code>__has_feature(objc_array_literals)</code> tests the availability of array literals; <code>__has_feature(objc_dictionary_literals)</code> tests the availability of dictionary literals; <code>__has_feature(objc_subscripting)</code> tests the availability of object subscripting.</p>
Patrick Beard62f12342012-03-20 21:51:03 +0000[diff] [blame]1149
1150<!-- ======================================================================= -->
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]1151<h2 id="overloading-in-c">Function Overloading in C</h2>
1152<!-- ======================================================================= -->
1153
Chris Lattnerf161d412009-02-13 21:51:45 +0000[diff] [blame]1154<p>Clang provides support for C++ function overloading in C. Function
1155overloading in C is introduced using the <tt>overloadable</tt> attribute. For
1156example, one might provide several overloaded versions of a <tt>tgsin</tt>
1157function that invokes the appropriate standard function computing the sine of a
1158value with <tt>float</tt>, <tt>double</tt>, or <tt>long double</tt>
1159precision:</p>
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]1160
1161<blockquote>
1162<pre>
1163#include &lt;math.h&gt;
1164float <b>__attribute__((overloadable))</b> tgsin(float x) { return sinf(x); }
1165double <b>__attribute__((overloadable))</b> tgsin(double x) { return sin(x); }
1166long double <b>__attribute__((overloadable))</b> tgsin(long double x) { return sinl(x); }
1167</pre>
1168</blockquote>
1169
1170<p>Given these declarations, one can call <tt>tgsin</tt> with a
1171<tt>float</tt> value to receive a <tt>float</tt> result, with a
1172<tt>double</tt> to receive a <tt>double</tt> result, etc. Function
1173overloading in C follows the rules of C++ function overloading to pick
1174the best overload given the call arguments, with a few C-specific
1175semantics:</p>
1176<ul>
1177 <li>Conversion from <tt>float</tt> or <tt>double</tt> to <tt>long
1178 double</tt> is ranked as a floating-point promotion (per C99) rather
1179 than as a floating-point conversion (as in C++).</li>
1180
1181 <li>A conversion from a pointer of type <tt>T*</tt> to a pointer of type
1182 <tt>U*</tt> is considered a pointer conversion (with conversion
1183 rank) if <tt>T</tt> and <tt>U</tt> are compatible types.</li>
1184
1185 <li>A conversion from type <tt>T</tt> to a value of type <tt>U</tt>
1186 is permitted if <tt>T</tt> and <tt>U</tt> are compatible types. This
1187 conversion is given "conversion" rank.</li>
1188</ul>
1189
1190<p>The declaration of <tt>overloadable</tt> functions is restricted to
1191function declarations and definitions. Most importantly, if any
1192function with a given name is given the <tt>overloadable</tt>
1193attribute, then all function declarations and definitions with that
1194name (and in that scope) must have the <tt>overloadable</tt>
Chris Lattnerf161d412009-02-13 21:51:45 +0000[diff] [blame]1195attribute. This rule even applies to redeclarations of functions whose original
1196declaration had the <tt>overloadable</tt> attribute, e.g.,</p>
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]1197
1198<blockquote>
1199<pre>
1200int f(int) __attribute__((overloadable));
1201float f(float); <i>// error: declaration of "f" must have the "overloadable" attribute</i>
1202
1203int g(int) __attribute__((overloadable));
1204int g(int) { } <i>// error: redeclaration of "g" must also have the "overloadable" attribute</i>
1205</pre>
1206</blockquote>
1207
Douglas Gregor965acbb2009-02-18 07:07:28 +0000[diff] [blame]1208<p>Functions marked <tt>overloadable</tt> must have
1209prototypes. Therefore, the following code is ill-formed:</p>
1210
1211<blockquote>
1212<pre>
1213int h() __attribute__((overloadable)); <i>// error: h does not have a prototype</i>
1214</pre>
1215</blockquote>
1216
1217<p>However, <tt>overloadable</tt> functions are allowed to use a
1218ellipsis 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>
1219
1220<blockquote>
1221<pre>
Chris Lattner02246802009-02-18 22:27:46 +0000[diff] [blame]1222void honeypot(...) __attribute__((overloadable, unavailable)); <i>// calling me is an error</i>
Douglas Gregor965acbb2009-02-18 07:07:28 +0000[diff] [blame]1223</pre>
1224</blockquote>
1225
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]1226<p>Functions declared with the <tt>overloadable</tt> attribute have
1227their names mangled according to the same rules as C++ function
1228names. For example, the three <tt>tgsin</tt> functions in our
1229motivating example get the mangled names <tt>_Z5tgsinf</tt>,
Chris Lattner71b48d62010-11-28 18:19:13 +0000[diff] [blame]1230<tt>_Z5tgsind</tt>, and <tt>_Z5tgsine</tt>, respectively. There are two
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]1231caveats to this use of name mangling:</p>
1232
1233<ul>
1234
1235 <li>Future versions of Clang may change the name mangling of
1236 functions overloaded in C, so you should not depend on an specific
1237 mangling. To be completely safe, we strongly urge the use of
1238 <tt>static inline</tt> with <tt>overloadable</tt> functions.</li>
1239
1240 <li>The <tt>overloadable</tt> attribute has almost no meaning when
1241 used in C++, because names will already be mangled and functions are
1242 already overloadable. However, when an <tt>overloadable</tt>
1243 function occurs within an <tt>extern "C"</tt> linkage specification,
1244 it's name <i>will</i> be mangled in the same way as it would in
1245 C.</li>
1246</ul>
1247
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]1248<p>Query for this feature with __has_extension(attribute_overloadable).</p>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]1249
Eli Friedman0c706c22011-09-19 23:17:44 +0000[diff] [blame]1250<!-- ======================================================================= -->
1251<h2 id="complex-list-init">Initializer lists for complex numbers in C</h2>
1252<!-- ======================================================================= -->
1253
1254<p>clang supports an extension which allows the following in C:</p>
1255
1256<blockquote>
1257<pre>
1258#include &lt;math.h&gt;
1259#include &lt;complex.h&gt;
1260complex float x = { 1.0f, INFINITY }; // Init to (1, Inf)
1261</pre>
1262</blockquote>
1263
1264<p>This construct is useful because there is no way to separately
1265initialize the real and imaginary parts of a complex variable in
1266standard C, given that clang does not support <code>_Imaginary</code>.
1267(clang also supports the <code>__real__</code> and <code>__imag__</code>
1268extensions from gcc, which help in some cases, but are not usable in
1269static initializers.)
1270
1271<p>Note that this extension does not allow eliding the braces; the
1272meaning of the following two lines is different:</p>
1273
1274<blockquote>
1275<pre>
1276complex float x[] = { { 1.0f, 1.0f } }; // [0] = (1, 1)
1277complex float x[] = { 1.0f, 1.0f }; // [0] = (1, 0), [1] = (1, 0)
1278</pre>
1279</blockquote>
1280
1281<p>This extension also works in C++ mode, as far as that goes, but does not
1282 apply to the C++ <code>std::complex</code>. (In C++11, list
1283 initialization allows the same syntax to be used with
1284 <code>std::complex</code> with the same meaning.)
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]1285
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]1286<!-- ======================================================================= -->
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1287<h2 id="builtins">Builtin Functions</h2>
1288<!-- ======================================================================= -->
1289
1290<p>Clang supports a number of builtin library functions with the same syntax as
1291GCC, including things like <tt>__builtin_nan</tt>,
1292<tt>__builtin_constant_p</tt>, <tt>__builtin_choose_expr</tt>,
1293<tt>__builtin_types_compatible_p</tt>, <tt>__sync_fetch_and_add</tt>, etc. In
1294addition to the GCC builtins, Clang supports a number of builtins that GCC does
1295not, which are listed here.</p>
1296
1297<p>Please note that Clang does not and will not support all of the GCC builtins
1298for vector operations. Instead of using builtins, you should use the functions
1299defined in target-specific header files like <tt>&lt;xmmintrin.h&gt;</tt>, which
1300define portable wrappers for these. Many of the Clang versions of these
1301functions are implemented directly in terms of <a href="#vectors">extended
1302vector support</a> instead of builtins, in order to reduce the number of
1303builtins that we need to implement.</p>
1304
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1305<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]1306<h3><a name="__builtin_shufflevector">__builtin_shufflevector</a></h3>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1307<!-- ======================================================================= -->
1308
Chris Lattneraad826b2009-09-16 18:56:12 +0000[diff] [blame]1309<p><tt>__builtin_shufflevector</tt> is used to express generic vector
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]1310permutation/shuffle/swizzle operations. This builtin is also very important for
1311the implementation of various target-specific header files like
1312<tt>&lt;xmmintrin.h&gt;</tt>.
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1313</p>
1314
1315<p><b>Syntax:</b></p>
1316
1317<pre>
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]1318__builtin_shufflevector(vec1, vec2, index1, index2, ...)
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1319</pre>
1320
1321<p><b>Examples:</b></p>
1322
1323<pre>
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]1324 // Identity operation - return 4-element vector V1.
1325 __builtin_shufflevector(V1, V1, 0, 1, 2, 3)
1326
1327 // "Splat" element 0 of V1 into a 4-element result.
1328 __builtin_shufflevector(V1, V1, 0, 0, 0, 0)
1329
1330 // Reverse 4-element vector V1.
1331 __builtin_shufflevector(V1, V1, 3, 2, 1, 0)
1332
1333 // Concatenate every other element of 4-element vectors V1 and V2.
1334 __builtin_shufflevector(V1, V2, 0, 2, 4, 6)
1335
1336 // Concatenate every other element of 8-element vectors V1 and V2.
1337 __builtin_shufflevector(V1, V2, 0, 2, 4, 6, 8, 10, 12, 14)
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1338</pre>
1339
1340<p><b>Description:</b></p>
1341
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]1342<p>The first two arguments to __builtin_shufflevector are vectors that have the
1343same element type. The remaining arguments are a list of integers that specify
1344the elements indices of the first two vectors that should be extracted and
1345returned in a new vector. These element indices are numbered sequentially
1346starting with the first vector, continuing into the second vector. Thus, if
1347vec1 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]1348</p>
1349
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]1350<p>The result of __builtin_shufflevector is a vector
1351with the same element type as vec1/vec2 but that has an element count equal to
1352the number of indices specified.
1353</p>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1354
Chris Lattner21190d52009-09-21 03:09:59 +0000[diff] [blame]1355<p>Query for this feature with __has_builtin(__builtin_shufflevector).</p>
1356
1357<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]1358<h3><a name="__builtin_unreachable">__builtin_unreachable</a></h3>
Chris Lattner21190d52009-09-21 03:09:59 +0000[diff] [blame]1359<!-- ======================================================================= -->
1360
1361<p><tt>__builtin_unreachable</tt> is used to indicate that a specific point in
1362the program cannot be reached, even if the compiler might otherwise think it
1363can. This is useful to improve optimization and eliminates certain warnings.
1364For example, without the <tt>__builtin_unreachable</tt> in the example below,
1365the compiler assumes that the inline asm can fall through and prints a "function
1366declared 'noreturn' should not return" warning.
1367</p>
1368
1369<p><b>Syntax:</b></p>
1370
1371<pre>
1372__builtin_unreachable()
1373</pre>
1374
1375<p><b>Example of Use:</b></p>
1376
1377<pre>
1378void myabort(void) __attribute__((noreturn));
1379void myabort(void) {
1380 asm("int3");
1381 __builtin_unreachable();
1382}
1383</pre>
1384
1385<p><b>Description:</b></p>
1386
1387<p>The __builtin_unreachable() builtin has completely undefined behavior. Since
1388it has undefined behavior, it is a statement that it is never reached and the
1389optimizer can take advantage of this to produce better code. This builtin takes
1390no arguments and produces a void result.
1391</p>
1392
1393<p>Query for this feature with __has_builtin(__builtin_unreachable).</p>
1394
Chris Lattner23aa9c82011-04-09 03:57:26 +0000[diff] [blame]1395<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]1396<h3><a name="__sync_swap">__sync_swap</a></h3>
Chris Lattner23aa9c82011-04-09 03:57:26 +0000[diff] [blame]1397<!-- ======================================================================= -->
1398
1399<p><tt>__sync_swap</tt> is used to atomically swap integers or pointers in
1400memory.
1401</p>
1402
1403<p><b>Syntax:</b></p>
1404
1405<pre>
1406<i>type</i> __sync_swap(<i>type</i> *ptr, <i>type</i> value, ...)
1407</pre>
1408
1409<p><b>Example of Use:</b></p>
1410
1411<pre>
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]1412int old_value = __sync_swap(&amp;value, new_value);
Chris Lattner23aa9c82011-04-09 03:57:26 +0000[diff] [blame]1413</pre>
1414
1415<p><b>Description:</b></p>
1416
1417<p>The __sync_swap() builtin extends the existing __sync_*() family of atomic
1418intrinsics to allow code to atomically swap the current value with the new
1419value. More importantly, it helps developers write more efficient and correct
1420code by avoiding expensive loops around __sync_bool_compare_and_swap() or
1421relying on the platform specific implementation details of
1422__sync_lock_test_and_set(). The __sync_swap() builtin is a full barrier.
1423</p>
1424
Chris Lattner21190d52009-09-21 03:09:59 +0000[diff] [blame]1425
Chris Lattner1177f912009-04-09 19:58:15 +0000[diff] [blame]1426<!-- ======================================================================= -->
1427<h2 id="targetspecific">Target-Specific Extensions</h2>
1428<!-- ======================================================================= -->
1429
1430<p>Clang supports some language features conditionally on some targets.</p>
1431
1432<!-- ======================================================================= -->
1433<h3 id="x86-specific">X86/X86-64 Language Extensions</h3>
1434<!-- ======================================================================= -->
1435
1436<p>The X86 backend has these language extensions:</p>
1437
1438<!-- ======================================================================= -->
1439<h4 id="x86-gs-segment">Memory references off the GS segment</h4>
1440<!-- ======================================================================= -->
1441
1442<p>Annotating a pointer with address space #256 causes it to be code generated
Chris Lattnera021e7c2009-05-05 18:54:47 +0000[diff] [blame]1443relative to the X86 GS segment register, and address space #257 causes it to be
1444relative to the X86 FS segment. Note that this is a very very low-level
1445feature that should only be used if you know what you're doing (for example in
1446an OS kernel).</p>
Chris Lattner1177f912009-04-09 19:58:15 +0000[diff] [blame]1447
1448<p>Here is an example:</p>
1449
1450<pre>
1451#define GS_RELATIVE __attribute__((address_space(256)))
1452int foo(int GS_RELATIVE *P) {
1453 return *P;
1454}
1455</pre>
1456
1457<p>Which compiles to (on X86-32):</p>
1458
1459<pre>
1460_foo:
1461 movl 4(%esp), %eax
1462 movl %gs:(%eax), %eax
1463 ret
1464</pre>
1465
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]1466<!-- ======================================================================= -->
1467<h2 id="analyzerspecific">Static Analysis-Specific Extensions</h2>
1468<!-- ======================================================================= -->
1469
1470<p>Clang supports additional attributes that are useful for documenting program
1471invariants and rules for static analysis tools. The extensions documented here
1472are used by the <a
1473href="http://clang.llvm.org/StaticAnalysis.html">path-sensitive static analyzer
1474engine</a> that is part of Clang's Analysis library.</p>
1475
John McCall87494012011-03-18 03:51:49 +0000[diff] [blame]1476<h3 id="attr_analyzer_noreturn">The <tt>analyzer_noreturn</tt> attribute</h3>
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]1477
1478<p>Clang's static analysis engine understands the standard <tt>noreturn</tt>
Ted Kremenek4df21142009-04-10 05:04:22 +0000[diff] [blame]1479attribute. This attribute, which is typically affixed to a function prototype,
1480indicates that a call to a given function never returns. Function prototypes for
1481common functions like <tt>exit</tt> are typically annotated with this attribute,
1482as well as a variety of common assertion handlers. Users can educate the static
1483analyzer about their own custom assertion handles (thus cutting down on false
1484positives due to false paths) by marking their own &quot;panic&quot; functions
1485with this attribute.</p>
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]1486
1487<p>While useful, <tt>noreturn</tt> is not applicable in all cases. Sometimes
Nick Lewycky625b5862009-06-14 04:08:08 +0000[diff] [blame]1488there are special functions that for all intents and purposes should be
1489considered panic functions (i.e., they are only called when an internal program
1490error occurs) but may actually return so that the program can fail gracefully.
1491The <tt>analyzer_noreturn</tt> attribute allows one to annotate such functions
1492as being interpreted as &quot;no return&quot; functions by the analyzer (thus
Chris Lattner28935892009-04-10 05:54:56 +0000[diff] [blame]1493pruning bogus paths) but will not affect compilation (as in the case of
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]1494<tt>noreturn</tt>).</p>
1495
1496<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]1497same places where the <tt>noreturn</tt> attribute can be placed. It is commonly
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]1498placed at the end of function prototypes:</p>
1499
1500<pre>
1501 void foo() <b>__attribute__((analyzer_noreturn))</b>;
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]1502</pre>
1503
John McCall87494012011-03-18 03:51:49 +0000[diff] [blame]1504<p>Query for this feature with
1505<tt>__has_attribute(analyzer_noreturn)</tt>.</p>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]1506
John McCall87494012011-03-18 03:51:49 +0000[diff] [blame]1507<h3 id="attr_method_family">The <tt>objc_method_family</tt> attribute</h3>
1508
1509<p>Many methods in Objective-C have conventional meanings determined
1510by their selectors. For the purposes of static analysis, it is
1511sometimes useful to be able to mark a method as having a particular
1512conventional meaning despite not having the right selector, or as not
1513having the conventional meaning that its selector would suggest.
1514For these use cases, we provide an attribute to specifically describe
1515the <q>method family</q> that a method belongs to.</p>
1516
1517<p><b>Usage</b>: <tt>__attribute__((objc_method_family(X)))</tt>,
1518where <tt>X</tt> is one of <tt>none</tt>, <tt>alloc</tt>, <tt>copy</tt>,
1519<tt>init</tt>, <tt>mutableCopy</tt>, or <tt>new</tt>. This attribute
1520can only be placed at the end of a method declaration:</p>
1521
1522<pre>
1523 - (NSString*) initMyStringValue <b>__attribute__((objc_method_family(none)))</b>;
1524</pre>
1525
1526<p>Users who do not wish to change the conventional meaning of a
1527method, and who merely want to document its non-standard retain and
1528release semantics, should use the
1529<a href="#attr_retain_release">retaining behavior attributes</a>
1530described below.</p>
1531
1532<p>Query for this feature with
1533<tt>__has_attribute(objc_method_family)</tt>.</p>
1534
1535<h3 id="attr_retain_release">Objective-C retaining behavior attributes</h3>
John McCall630b7ae2011-01-25 04:26:21 +0000[diff] [blame]1536
1537<p>In Objective-C, functions and methods are generally assumed to take
1538and return objects with +0 retain counts, with some exceptions for
1539special methods like <tt>+alloc</tt> and <tt>init</tt>. However,
1540there are exceptions, and so Clang provides attributes to allow these
1541exceptions to be documented, which helps the analyzer find leaks (and
John McCall87494012011-03-18 03:51:49 +0000[diff] [blame]1542ignore non-leaks). Some exceptions may be better described using
1543the <a href="#attr_method_family"><tt>objc_method_family</tt></a>
1544attribute instead.</p>
John McCall630b7ae2011-01-25 04:26:21 +0000[diff] [blame]1545
1546<p><b>Usage</b>: The <tt>ns_returns_retained</tt>, <tt>ns_returns_not_retained</tt>,
1547<tt>ns_returns_autoreleased</tt>, <tt>cf_returns_retained</tt>,
1548and <tt>cf_returns_not_retained</tt> attributes can be placed on
1549methods and functions that return Objective-C or CoreFoundation
1550objects. They are commonly placed at the end of a function prototype
1551or method declaration:</p>
1552
1553<pre>
1554 id foo() <b>__attribute__((ns_returns_retained))</b>;
1555
1556 - (NSString*) bar: (int) x <b>__attribute__((ns_returns_retained))</b>;
1557</pre>
1558
1559<p>The <tt>*_returns_retained</tt> attributes specify that the
1560returned object has a +1 retain count.
1561The <tt>*_returns_not_retained</tt> attributes specify that the return
1562object has a +0 retain count, even if the normal convention for its
1563selector would be +1. <tt>ns_returns_autoreleased</tt> specifies that the
1564returned object is +0, but is guaranteed to live at least as long as the
1565next flush of an autorelease pool.</p>
1566
1567<p><b>Usage</b>: The <tt>ns_consumed</tt> and <tt>cf_consumed</tt>
1568attributes can be placed on an parameter declaration; they specify
1569that the argument is expected to have a +1 retain count, which will be
1570balanced in some way by the function or method.
1571The <tt>ns_consumes_self</tt> attribute can only be placed on an
1572Objective-C method; it specifies that the method expects
1573its <tt>self</tt> parameter to have a +1 retain count, which it will
1574balance in some way.</p>
1575
1576<pre>
1577 void <b>foo(__attribute__((ns_consumed))</b> NSString *string);
1578
1579 - (void) bar <b>__attribute__((ns_consumes_self))</b>;
1580 - (void) baz: (id) <b>__attribute__((ns_consumed))</b> x;
1581</pre>
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]1582
John McCall87494012011-03-18 03:51:49 +0000[diff] [blame]1583<p>Query for these features with <tt>__has_attribute(ns_consumed)</tt>,
1584<tt>__has_attribute(ns_returns_retained)</tt>, etc.</p>
1585
Kostya Serebryanyce98c9b2011-11-28 20:51:02 +0000[diff] [blame]1586<!-- ======================================================================= -->
1587<h2 id="dynamicanalyzerspecific">Dynamic Analysis-Specific Extensions</h2>
1588<!-- ======================================================================= -->
1589<h3 id="address_sanitizer">AddressSanitizer</h3>
1590<p> Use <code>__has_feature(address_sanitizer)</code>
1591to check if the code is being built with <a
1592 href="AddressSanitizer.html">AddressSanitizer</a>.
1593</p>
Kostya Serebryany71efba02012-01-24 19:25:38 +0000[diff] [blame]1594<p>Use <tt>__attribute__((no_address_safety_analysis))</tt> on a function
1595declaration to specify that address safety instrumentation (e.g.
1596AddressSanitizer) should not be applied to that function.
1597</p>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]1598
1599<!-- ======================================================================= -->
Caitlin Sadowski73cbbc82011-07-28 18:38:36 +0000[diff] [blame]1600<h2 id="threadsafety">Thread-Safety Annotation Checking</h2>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]1601<!-- ======================================================================= -->
1602
1603<p>Clang supports additional attributes for checking basic locking policies in
1604multithreaded programs.
1605Clang currently parses the following list of attributes, although
1606<b>the implementation for these annotations is currently in development.</b>
1607For more details, see the
1608<a href="http://gcc.gnu.org/wiki/ThreadSafetyAnnotation">GCC implementation</a>.
1609</p>
1610
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1611<h4 id="ts_noanal">no_thread_safety_analysis</h4>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]1612
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1613<p>Use <tt>__attribute__((no_thread_safety_analysis))</tt> on a function
1614declaration to specify that the thread safety analysis should not be run on that
1615function. This attribute provides an escape hatch (e.g. for situations when it
1616is difficult to annotate the locking policy). </p>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]1617
1618<h4 id="ts_lockable">lockable</h4>
1619
1620<p>Use <tt>__attribute__((lockable))</tt> on a class definition to specify
1621that it has a lockable type (e.g. a Mutex class). This annotation is primarily
1622used to check consistency.</p>
1623
1624<h4 id="ts_scopedlockable">scoped_lockable</h4>
1625
1626<p>Use <tt>__attribute__((scoped_lockable))</tt> on a class definition to
1627specify that it has a "scoped" lockable type. Objects of this type will acquire
1628the lock upon construction and release it upon going out of scope.
1629 This annotation is primarily used to check
1630consistency.</p>
1631
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1632<h4 id="ts_guardedvar">guarded_var</h4>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]1633
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1634<p>Use <tt>__attribute__((guarded_var))</tt> on a variable declaration to
1635specify that the variable must be accessed while holding some lock.</p>
1636
1637<h4 id="ts_ptguardedvar">pt_guarded_var</h4>
1638
1639<p>Use <tt>__attribute__((pt_guarded_var))</tt> on a pointer declaration to
1640specify that the pointer must be dereferenced while holding some lock.</p>
1641
1642<h4 id="ts_guardedby">guarded_by(l)</h4>
1643
1644<p>Use <tt>__attribute__((guarded_by(l)))</tt> on a variable declaration to
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1645specify that the variable must be accessed while holding lock <tt>l</tt>.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1646
1647<h4 id="ts_ptguardedby">pt_guarded_by(l)</h4>
1648
1649<p>Use <tt>__attribute__((pt_guarded_by(l)))</tt> on a pointer declaration to
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1650specify that the pointer must be dereferenced while holding lock <tt>l</tt>.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1651
1652<h4 id="ts_acquiredbefore">acquired_before(...)</h4>
1653
1654<p>Use <tt>__attribute__((acquired_before(...)))</tt> on a declaration
1655of a lockable variable to specify that the lock must be acquired before all
1656attribute arguments. Arguments must be lockable type, and there must be at
1657least one argument.</p>
1658
1659<h4 id="ts_acquiredafter">acquired_after(...)</h4>
1660
1661<p>Use <tt>__attribute__((acquired_after(...)))</tt> on a declaration
1662of a lockable variable to specify that the lock must be acquired after all
1663attribute arguments. Arguments must be lockable type, and there must be at
1664least one argument.</p>
1665
1666<h4 id="ts_elf">exclusive_lock_function(...)</h4>
1667
1668<p>Use <tt>__attribute__((exclusive_lock_function(...)))</tt> on a function
1669declaration to specify that the function acquires all listed locks
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1670exclusively. This attribute takes zero or more arguments: either of lockable
1671type or integers indexing into function parameters of lockable type. If no
1672arguments are given, the acquired lock is implicitly <tt>this</tt> of the
1673enclosing object.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1674
1675<h4 id="ts_slf">shared_lock_function(...)</h4>
1676
1677<p>Use <tt>__attribute__((shared_lock_function(...)))</tt> on a function
1678declaration to specify that the function acquires all listed locks, although
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1679 the locks may be shared (e.g. read locks). This attribute takes zero or more
1680arguments: either of lockable type or integers indexing into function
1681parameters of lockable type. If no arguments are given, the acquired lock is
1682implicitly <tt>this</tt> of the enclosing object.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1683
1684<h4 id="ts_etf">exclusive_trylock_function(...)</h4>
1685
1686<p>Use <tt>__attribute__((exclusive_lock_function(...)))</tt> on a function
1687declaration to specify that the function will try (without blocking) to acquire
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1688all listed locks exclusively. This attribute takes one or more arguments. The
1689first argument is an integer or boolean value specifying the return value of a
1690successful lock acquisition. The remaining arugments are either of lockable type
1691or integers indexing into function parameters of lockable type. If only one
1692argument is given, the acquired lock is implicitly <tt>this</tt> of the
1693enclosing object.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1694
1695<h4 id="ts_stf">shared_trylock_function(...)</h4>
1696
1697<p>Use <tt>__attribute__((shared_lock_function(...)))</tt> on a function
1698declaration to specify that the function will try (without blocking) to acquire
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1699all listed locks, although the locks may be shared (e.g. read locks). This
1700attribute takes one or more arguments. The first argument is an integer or
1701boolean value specifying the return value of a successful lock acquisition. The
1702remaining arugments are either of lockable type or integers indexing into
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1703function parameters of lockable type. If only one argument is given, the
1704acquired lock is implicitly <tt>this</tt> of the enclosing object.</p>
1705
1706<h4 id="ts_uf">unlock_function(...)</h4>
1707
1708<p>Use <tt>__attribute__((unlock_function(...)))</tt> on a function
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1709declaration to specify that the function release all listed locks. This
1710attribute takes zero or more arguments: either of lockable type or integers
1711indexing into function parameters of lockable type. If no arguments are given,
1712the acquired lock is implicitly <tt>this</tt> of the enclosing object.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1713
1714<h4 id="ts_lr">lock_returned(l)</h4>
1715
1716<p>Use <tt>__attribute__((lock_returned(l)))</tt> on a function
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1717declaration to specify that the function returns lock <tt>l</tt> (<tt>l</tt>
1718must be of lockable type). This annotation is used to aid in resolving lock
1719expressions.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1720
1721<h4 id="ts_le">locks_excluded(...)</h4>
1722
1723<p>Use <tt>__attribute__((locks_excluded(...)))</tt> on a function declaration
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1724to specify that the function must not be called with the listed locks. Arguments
1725must be lockable type, and there must be at least one argument.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1726
1727<h4 id="ts_elr">exclusive_locks_required(...)</h4>
1728
1729<p>Use <tt>__attribute__((exclusive_locks_required(...)))</tt> on a function
1730declaration to specify that the function must be called while holding the listed
1731exclusive locks. Arguments must be lockable type, and there must be at
1732least one argument.</p>
1733
1734<h4 id="ts_slr">shared_locks_required(...)</h4>
1735
1736<p>Use <tt>__attribute__((shared_locks_required(...)))</tt> on a function
1737declaration to specify that the function must be called while holding the listed
1738shared locks. Arguments must be lockable type, and there must be at
1739least one argument.</p>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]1740
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1741</div>
1742</body>
1743</html>