Gitiles
Code Review Sign In
gerrit-public.fairphone.software / platform / external / clang / e74dc199e95cc523c8806e0de41d8814e8bc5c42 / . / docs / LanguageExtensions.html
blob: 8c0e5b7ffcb0f1532874c1a683be12b83b760041 [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>
Daniel Dunbar85ff9692012-04-05 17:10:06 +0000[diff] [blame]33<li><a href="#user_specified_system_framework">'User-Specified' System Frameworks</a></li>
Douglas Gregor93a70672012-03-11 04:53:21 +0000[diff] [blame]34<li><a href="#availability">Availability attribute</a></li>
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]35<li><a href="#checking_language_features">Checks for Standard Language Features</a>
Ted Kremenek22c34102009-12-03 02:05:57 +0000[diff] [blame]36 <ul>
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]37 <li><a href="#cxx98">C++98</a>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]38 <ul>
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]39 <li><a href="#cxx_exceptions">C++ exceptions</a></li>
40 <li><a href="#cxx_rtti">C++ RTTI</a></li>
41 </ul></li>
42 <li><a href="#cxx11">C++11</a>
43 <ul>
44 <li><a href="#cxx_access_control_sfinae">C++11 SFINAE includes access control</a></li>
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]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>
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]71 <li><a href="#cxx_strong_enums">C++11 strongly-typed enumerations</a></li>
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]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>
Richard Smithfafbf062012-04-11 17:55:32 +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>
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]81 <li><a href="#c_atomic">C11 atomic operations</a></li>
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]82 <li><a href="#c_generic_selections">C11 generic selections</a></li>
83 <li><a href="#c_static_assert">C11 <tt>_Static_assert()</tt></a></li>
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]84 </ul></li>
85</ul></li>
Douglas Gregorafdf1372011-02-03 21:57:35 +0000[diff] [blame]86<li><a href="#checking_type_traits">Checks for Type Traits</a></li>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]87<li><a href="#blocks">Blocks</a></li>
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]88<li><a href="#objc_features">Objective-C Features</a>
89 <ul>
90 <li><a href="#objc_instancetype">Related result types</a></li>
John McCallf85e1932011-06-15 23:02:42 +0000[diff] [blame]91 <li><a href="#objc_arc">Automatic reference counting</a></li>
Douglas Gregor5471bc82011-09-08 17:18:35 +0000[diff] [blame]92 <li><a href="#objc_fixed_enum">Enumerations with a fixed underlying type</a></li>
Douglas Gregor8a4e1822012-03-09 23:24:48 +0000[diff] [blame]93 <li><a href="#objc_lambdas">Interoperability with C++11 lambdas</a></li>
Patrick Beardeb382ec2012-04-19 00:25:12 +0000[diff] [blame]94 <li><a href="#objc_object_literals_subscripting">Object Literals and Subscripting</a></li>
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]95 </ul>
96</li>
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]97<li><a href="#overloading-in-c">Function Overloading in C</a></li>
Eli Friedman0c706c22011-09-19 23:17:44 +0000[diff] [blame]98<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]99<li><a href="#builtins">Builtin Functions</a>
100 <ul>
Hal Finkela841c192012-08-05 22:03:08 +0000[diff] [blame]101 <li><a href="#__builtin_readcyclecounter">__builtin_readcyclecounter</a></li>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]102 <li><a href="#__builtin_shufflevector">__builtin_shufflevector</a></li>
Chris Lattner21190d52009-09-21 03:09:59 +0000[diff] [blame]103 <li><a href="#__builtin_unreachable">__builtin_unreachable</a></li>
Chris Lattner23aa9c82011-04-09 03:57:26 +0000[diff] [blame]104 <li><a href="#__sync_swap">__sync_swap</a></li>
Douglas Gregorafdf1372011-02-03 21:57:35 +0000[diff] [blame]105 </ul>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]106</li>
Richard Smithe0d3b4c2012-05-03 18:27:39 +0000[diff] [blame]107<li><a href="#non-standard-attributes">Non-standard C++11 Attributes</a>
108<ul>
109 <li><a href="#clang__fallthrough">The <tt>clang::fallthrough</tt> attribute</a></li>
110</ul>
111</li>
Chris Lattner1177f912009-04-09 19:58:15 +0000[diff] [blame]112<li><a href="#targetspecific">Target-Specific Extensions</a>
113 <ul>
114 <li><a href="#x86-specific">X86/X86-64 Language Extensions</a></li>
115 </ul>
116</li>
John McCall87494012011-03-18 03:51:49 +0000[diff] [blame]117<li><a href="#analyzerspecific">Static Analysis-Specific Extensions</a></li>
Benjamin Kramer665a8dc2012-01-15 15:26:07 +0000[diff] [blame]118<li><a href="#dynamicanalyzerspecific">Dynamic Analysis-Specific Extensions</a>
Kostya Serebryanyce98c9b2011-11-28 20:51:02 +0000[diff] [blame]119 <ul>
120 <li><a href="#address_sanitizer">AddressSanitizer</a></li>
121 </ul>
Benjamin Kramer665a8dc2012-01-15 15:26:07 +0000[diff] [blame]122</li>
123<li><a href="#threadsafety">Thread Safety Annotation Checking</a>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]124 <ul>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]125 <li><a href="#ts_noanal"><tt>no_thread_safety_analysis</tt></a></li>
126 <li><a href="#ts_lockable"><tt>lockable</tt></a></li>
127 <li><a href="#ts_scopedlockable"><tt>scoped_lockable</tt></a></li>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]128 <li><a href="#ts_guardedvar"><tt>guarded_var</tt></a></li>
129 <li><a href="#ts_ptguardedvar"><tt>pt_guarded_var</tt></a></li>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]130 <li><a href="#ts_guardedby"><tt>guarded_by(l)</tt></a></li>
131 <li><a href="#ts_ptguardedby"><tt>pt_guarded_by(l)</tt></a></li>
132 <li><a href="#ts_acquiredbefore"><tt>acquired_before(...)</tt></a></li>
133 <li><a href="#ts_acquiredafter"><tt>acquired_after(...)</tt></a></li>
134 <li><a href="#ts_elf"><tt>exclusive_lock_function(...)</tt></a></li>
135 <li><a href="#ts_slf"><tt>shared_lock_function(...)</tt></a></li>
136 <li><a href="#ts_etf"><tt>exclusive_trylock_function(...)</tt></a></li>
137 <li><a href="#ts_stf"><tt>shared_trylock_function(...)</tt></a></li>
138 <li><a href="#ts_uf"><tt>unlock_function(...)</tt></a></li>
139 <li><a href="#ts_lr"><tt>lock_returned(l)</tt></a></li>
140 <li><a href="#ts_le"><tt>locks_excluded(...)</tt></a></li>
141 <li><a href="#ts_elr"><tt>exclusive_locks_required(...)</tt></a></li>
142 <li><a href="#ts_slr"><tt>shared_locks_required(...)</tt></a></li>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]143 </ul>
Benjamin Kramer665a8dc2012-01-15 15:26:07 +0000[diff] [blame]144</li>
Dmitri Gribenko0d5a0692012-08-17 00:08:38 +0000[diff] [blame]145<li><a href="#type_safety">Type Safety Checking</a>
146 <ul>
147 <li><a href="#argument_with_type_tag"><tt>argument_with_type_tag(...)</tt></a></li>
148 <li><a href="#pointer_with_type_tag"><tt>pointer_with_type_tag(...)</tt></a></li>
149 <li><a href="#type_tag_for_datatype"><tt>type_tag_for_datatype(...)</tt></a></li>
150 </ul>
151</li>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]152</ul>
153
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]154<!-- ======================================================================= -->
155<h2 id="intro">Introduction</h2>
156<!-- ======================================================================= -->
157
158<p>This document describes the language extensions provided by Clang. In
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]159addition to the language extensions listed here, Clang aims to support a broad
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]160range of GCC extensions. Please see the <a
161href="http://gcc.gnu.org/onlinedocs/gcc/C-Extensions.html">GCC manual</a> for
162more information on these extensions.</p>
163
164<!-- ======================================================================= -->
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]165<h2 id="feature_check">Feature Checking Macros</h2>
166<!-- ======================================================================= -->
167
168<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]169on them. In order to allow fine-grain features checks, we support three builtin
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]170function-like macros. This allows you to directly test for a feature in your
171code without having to resort to something like autoconf or fragile "compiler
172version checks".</p>
173
174<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]175<h3><a name="__has_builtin">__has_builtin</a></h3>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]176<!-- ======================================================================= -->
177
178<p>This function-like macro takes a single identifier argument that is the name
179of a builtin function. It evaluates to 1 if the builtin is supported or 0 if
180not. It can be used like this:</p>
181
182<blockquote>
183<pre>
184#ifndef __has_builtin // Optional of course.
185 #define __has_builtin(x) 0 // Compatibility with non-clang compilers.
186#endif
187
188...
189#if __has_builtin(__builtin_trap)
190 __builtin_trap();
191#else
192 abort();
193#endif
194...
195</pre>
196</blockquote>
197
198
199<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]200<h3><a name="__has_feature_extension"> __has_feature and __has_extension</a></h3>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]201<!-- ======================================================================= -->
202
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]203<p>These function-like macros take a single identifier argument that is the
204name of a feature. <code>__has_feature</code> evaluates to 1 if the feature
205is both supported by Clang and standardized in the current language standard
206or 0 if not (but see <a href="#has_feature_back_compat">below</a>), while
207<code>__has_extension</code> evaluates to 1 if the feature is supported by
208Clang in the current language (either as a language extension or a standard
209language feature) or 0 if not. They can be used like this:</p>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]210
211<blockquote>
212<pre>
213#ifndef __has_feature // Optional of course.
214 #define __has_feature(x) 0 // Compatibility with non-clang compilers.
215#endif
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]216#ifndef __has_extension
217 #define __has_extension __has_feature // Compatibility with pre-3.0 compilers.
218#endif
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]219
220...
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]221#if __has_feature(cxx_rvalue_references)
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]222// This code will only be compiled with the -std=c++11 and -std=gnu++11
223// options, because rvalue references are only standardized in C++11.
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]224#endif
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]225
226#if __has_extension(cxx_rvalue_references)
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]227// 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]228// and -std=gnu++98 options, because rvalue references are supported as a
229// language extension in C++98.
230#endif
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]231</pre>
232</blockquote>
233
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]234<p id="has_feature_back_compat">For backwards compatibility reasons,
235<code>__has_feature</code> can also be used to test for support for
236non-standardized features, i.e. features not prefixed <code>c_</code>,
237<code>cxx_</code> or <code>objc_</code>.</p>
238
Kostya Serebryanyce98c9b2011-11-28 20:51:02 +0000[diff] [blame]239<p id="has_feature_for_non_language_features">
240Another use of <code>__has_feature</code> is to check for compiler features
241not related to the language standard, such as e.g.
242<a href="AddressSanitizer.html">AddressSanitizer</a>.
243
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]244<p>If the <code>-pedantic-errors</code> option is given,
245<code>__has_extension</code> is equivalent to <code>__has_feature</code>.</p>
246
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]247<p>The feature tag is described along with the language feature below.</p>
248
Richard Smith5297d712012-02-25 10:41:10 +0000[diff] [blame]249<p>The feature name or extension name can also be specified with a preceding and
250following <code>__</code> (double underscore) to avoid interference from a macro
Richard Smith1d9f4c12012-03-01 02:12:07 +0000[diff] [blame]251with the same name. For instance, <code>__cxx_rvalue_references__</code> can be
252used instead of <code>cxx_rvalue_references</code>.</p>
Richard Smith5297d712012-02-25 10:41:10 +0000[diff] [blame]253
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]254<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]255<h3><a name="__has_attribute">__has_attribute</a></h3>
Anders Carlssoncae50952010-10-20 02:31:43 +0000[diff] [blame]256<!-- ======================================================================= -->
257
258<p>This function-like macro takes a single identifier argument that is the name
259of an attribute. It evaluates to 1 if the attribute is supported or 0 if not. It
260can be used like this:</p>
261
262<blockquote>
263<pre>
264#ifndef __has_attribute // Optional of course.
265 #define __has_attribute(x) 0 // Compatibility with non-clang compilers.
266#endif
267
268...
Anders Carlsson961003d2011-01-24 03:54:51 +0000[diff] [blame]269#if __has_attribute(always_inline)
270#define ALWAYS_INLINE __attribute__((always_inline))
Anders Carlssoncae50952010-10-20 02:31:43 +0000[diff] [blame]271#else
Anders Carlsson961003d2011-01-24 03:54:51 +0000[diff] [blame]272#define ALWAYS_INLINE
Anders Carlssoncae50952010-10-20 02:31:43 +0000[diff] [blame]273#endif
274...
275</pre>
276</blockquote>
277
Jean-Daniel Dupas8a5e7fd2012-03-01 14:53:16 +0000[diff] [blame]278<p>The attribute name can also be specified with a preceding and
279following <code>__</code> (double underscore) to avoid interference from a macro
280with the same name. For instance, <code>__always_inline__</code> can be used
281instead of <code>always_inline</code>.</p>
282
Anders Carlssoncae50952010-10-20 02:31:43 +0000[diff] [blame]283<!-- ======================================================================= -->
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]284<h2 id="has_include">Include File Checking Macros</h2>
285<!-- ======================================================================= -->
286
287<p>Not all developments systems have the same include files.
288The <a href="#__has_include">__has_include</a> and
289<a href="#__has_include_next">__has_include_next</a> macros allow you to
290check for the existence of an include file before doing
291a possibly failing #include directive.</p>
292
293<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]294<h3><a name="__has_include">__has_include</a></h3>
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]295<!-- ======================================================================= -->
296
297<p>This function-like macro takes a single file name string argument that
298is the name of an include file. It evaluates to 1 if the file can
299be found using the include paths, or 0 otherwise:</p>
300
301<blockquote>
302<pre>
303// Note the two possible file name string formats.
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]304#if __has_include("myinclude.h") &amp;&amp; __has_include(&lt;stdint.h&gt;)
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]305# include "myinclude.h"
306#endif
307
308// To avoid problem with non-clang compilers not having this macro.
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]309#if defined(__has_include) &amp;&amp; __has_include("myinclude.h")
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]310# include "myinclude.h"
311#endif
312</pre>
313</blockquote>
314
315<p>To test for this feature, use #if defined(__has_include).</p>
316
317<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]318<h3><a name="__has_include_next">__has_include_next</a></h3>
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]319<!-- ======================================================================= -->
320
321<p>This function-like macro takes a single file name string argument that
322is the name of an include file. It is like __has_include except that it
323looks for the second instance of the given file found in the include
324paths. It evaluates to 1 if the second instance of the file can
325be found using the include paths, or 0 otherwise:</p>
326
327<blockquote>
328<pre>
329// Note the two possible file name string formats.
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]330#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]331# include_next "myinclude.h"
332#endif
333
334// To avoid problem with non-clang compilers not having this macro.
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]335#if defined(__has_include_next) &amp;&amp; __has_include_next("myinclude.h")
John Thompson92bd8c72009-11-02 22:28:12 +0000[diff] [blame]336# include_next "myinclude.h"
337#endif
338</pre>
339</blockquote>
340
341<p>Note that __has_include_next, like the GNU extension
342#include_next directive, is intended for use in headers only,
343and will issue a warning if used in the top-level compilation
344file. A warning will also be issued if an absolute path
345is used in the file argument.</p>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]346
Ted Kremenekd7681502011-10-12 19:46:30 +0000[diff] [blame]347
348<!-- ======================================================================= -->
349<h3><a name="__has_warning">__has_warning</a></h3>
350<!-- ======================================================================= -->
351
352<p>This function-like macro takes a string literal that represents a command
353 line option for a warning and returns true if that is a valid warning
354 option.</p>
355
356<blockquote>
357<pre>
358#if __has_warning("-Wformat")
359...
360#endif
361</pre>
362</blockquote>
363
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]364<!-- ======================================================================= -->
Chris Lattner81edc9f2009-04-13 02:45:46 +0000[diff] [blame]365<h2 id="builtinmacros">Builtin Macros</h2>
366<!-- ======================================================================= -->
367
Douglas Gregor4290fbd2010-04-30 02:51:06 +0000[diff] [blame]368<dl>
369 <dt><code>__BASE_FILE__</code></dt>
370 <dd>Defined to a string that contains the name of the main input
371 file passed to Clang.</dd>
372
373 <dt><code>__COUNTER__</code></dt>
374 <dd>Defined to an integer value that starts at zero and is
375 incremented each time the <code>__COUNTER__</code> macro is
376 expanded.</dd>
377
378 <dt><code>__INCLUDE_LEVEL__</code></dt>
379 <dd>Defined to an integral value that is the include depth of the
380 file currently being translated. For the main file, this value is
381 zero.</dd>
382
383 <dt><code>__TIMESTAMP__</code></dt>
384 <dd>Defined to the date and time of the last modification of the
385 current source file.</dd>
386
387 <dt><code>__clang__</code></dt>
388 <dd>Defined when compiling with Clang</dd>
389
390 <dt><code>__clang_major__</code></dt>
Chris Lattnerd4b66b92011-12-15 19:06:36 +0000[diff] [blame]391 <dd>Defined to the major marketing version number of Clang (e.g., the
392 2 in 2.0.1). Note that marketing version numbers should not be used to
393 check for language features, as different vendors use different numbering
394 schemes. Instead, use the <a href="#feature_check">feature checking
395 macros</a>.</dd>
Douglas Gregor4290fbd2010-04-30 02:51:06 +0000[diff] [blame]396
397 <dt><code>__clang_minor__</code></dt>
398 <dd>Defined to the minor version number of Clang (e.g., the 0 in
Chris Lattnerd4b66b92011-12-15 19:06:36 +0000[diff] [blame]399 2.0.1). Note that marketing version numbers should not be used to
400 check for language features, as different vendors use different numbering
401 schemes. Instead, use the <a href="#feature_check">feature checking
402 macros</a>.</dd>
Douglas Gregor4290fbd2010-04-30 02:51:06 +0000[diff] [blame]403
404 <dt><code>__clang_patchlevel__</code></dt>
Chris Lattnerd4b66b92011-12-15 19:06:36 +0000[diff] [blame]405 <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]406
407 <dt><code>__clang_version__</code></dt>
Chris Lattnerd4b66b92011-12-15 19:06:36 +0000[diff] [blame]408 <dd>Defined to a string that captures the Clang marketing version, including
409 the Subversion tag or revision number, e.g., "1.5 (trunk 102332)".</dd>
Douglas Gregor4290fbd2010-04-30 02:51:06 +0000[diff] [blame]410</dl>
Chris Lattner81edc9f2009-04-13 02:45:46 +0000[diff] [blame]411
412<!-- ======================================================================= -->
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]413<h2 id="vectors">Vectors and Extended Vectors</h2>
414<!-- ======================================================================= -->
415
Anton Yartsevda90c772012-01-15 16:22:24 +0000[diff] [blame]416<p>Supports the GCC, OpenCL, AltiVec and NEON vector extensions.</p>
Owen Andersond2bf0cd2010-01-27 01:22:36 +0000[diff] [blame]417
Benjamin Kramer3419d7c2012-01-15 16:42:14 +0000[diff] [blame]418<p>OpenCL vector types are created using <tt>ext_vector_type</tt> attribute. It
419support for <tt>V.xyzw</tt> syntax and other tidbits as seen in OpenCL. An
420example is:</p>
Owen Andersond2bf0cd2010-01-27 01:22:36 +0000[diff] [blame]421
422<blockquote>
423<pre>
424typedef float float4 <b>__attribute__((ext_vector_type(4)))</b>;
425typedef float float2 <b>__attribute__((ext_vector_type(2)))</b>;
426
427float4 foo(float2 a, float2 b) {
428 float4 c;
429 c.xz = a;
430 c.yw = b;
431 return c;
432}
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]433</pre>
Owen Andersond2bf0cd2010-01-27 01:22:36 +0000[diff] [blame]434</blockquote>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]435
Benjamin Kramer3419d7c2012-01-15 16:42:14 +0000[diff] [blame]436<p>Query for this feature with
437<tt>__has_extension(attribute_ext_vector_type)</tt>.</p>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]438
Benjamin Kramer3419d7c2012-01-15 16:42:14 +0000[diff] [blame]439<p>Giving <tt>-faltivec</tt> option to clang enables support for AltiVec vector
440syntax and functions. For example:</p>
Anton Yartsevda90c772012-01-15 16:22:24 +0000[diff] [blame]441
442<blockquote>
443<pre>
444vector float foo(vector int a) {
445 vector int b;
446 b = vec_add(a, a) + a;
447 return (vector float)b;
448}
449</pre>
450</blockquote>
451
452<p>NEON vector types are created using <tt>neon_vector_type</tt> and
453<tt>neon_polyvector_type</tt> attributes. For example:</p>
454
455<blockquote>
456<pre>
457typedef <b>__attribute__((neon_vector_type(8)))</b> int8_t int8x8_t;
458typedef <b>__attribute__((neon_polyvector_type(16)))</b> poly8_t poly8x16_t;
459
460int8x8_t foo(int8x8_t a) {
461 int8x8_t v;
462 v = a;
463 return v;
464}
465</pre>
466</blockquote>
467
468<!-- ======================================================================= -->
469<h3><a name="vector_literals">Vector Literals</a></h3>
470<!-- ======================================================================= -->
471
472<p>Vector literals can be used to create vectors from a set of scalars, or
473vectors. Either parentheses or braces form can be used. In the parentheses form
474the number of literal values specified must be one, i.e. referring to a scalar
475value, or must match the size of the vector type being created. If a single
476scalar literal value is specified, the scalar literal value will be replicated
477to all the components of the vector type. In the brackets form any number of
478literals can be specified. For example:</p>
479
480<blockquote>
481<pre>
482typedef int v4si __attribute__((__vector_size__(16)));
483typedef float float4 __attribute__((ext_vector_type(4)));
484typedef float float2 __attribute__((ext_vector_type(2)));
485
486v4si vsi = (v4si){1, 2, 3, 4};
487float4 vf = (float4)(1.0f, 2.0f, 3.0f, 4.0f);
488vector int vi1 = (vector int)(1); // vi1 will be (1, 1, 1, 1).
489vector int vi2 = (vector int){1}; // vi2 will be (1, 0, 0, 0).
490vector int vi3 = (vector int)(1, 2); // error
491vector int vi4 = (vector int){1, 2}; // vi4 will be (1, 2, 0, 0).
492vector int vi5 = (vector int)(1, 2, 3, 4);
493float4 vf = (float4)((float2)(1.0f, 2.0f), (float2)(3.0f, 4.0f));
494</pre>
495</blockquote>
496
497<!-- ======================================================================= -->
498<h3><a name="vector_operations">Vector Operations</a></h3>
499<!-- ======================================================================= -->
500
501<p>The table below shows the support for each operation by vector extension.
502A dash indicates that an operation is not accepted according to a corresponding
503specification.</p>
504
505<table width="500" border="1" cellspacing="0">
506 <tr>
Benjamin Kramer3419d7c2012-01-15 16:42:14 +0000[diff] [blame]507 <th>Operator</th>
508 <th>OpenCL</th>
509 <th>AltiVec</th>
510 <th>GCC</th>
511 <th>NEON</th>
Anton Yartsevda90c772012-01-15 16:22:24 +0000[diff] [blame]512 </tr>
513 <tr>
514 <td>[]</td>
515 <td align="center">yes</td>
516 <td align="center">yes</td>
517 <td align="center">yes</td>
518 <td align="center">-</td>
519 </tr>
520 <tr>
521 <td>unary operators +, -</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>++, --</td>
529 <td align="center">yes</td>
530 <td align="center">yes</td>
531 <td align="center">-</td>
532 <td align="center">-</td>
533 </tr>
534 <tr>
535 <td>+, -, *, /, %</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>bitwise operators &, |, ^, ~</td>
543 <td align="center">yes</td>
544 <td align="center">yes</td>
545 <td align="center">yes</td>
546 <td align="center">-</td>
547 </tr>
548 <tr>
549 <td>&gt&gt, &lt&lt</td>
550 <td align="center">yes</td>
551 <td align="center">yes</td>
552 <td align="center">yes</td>
553 <td align="center">-</td>
554 </tr>
555 <tr>
556 <td>!, &&,||</td>
557 <td align="center">no</td>
558 <td align="center">-</td>
559 <td align="center">-</td>
560 <td align="center">-</td>
561 </tr>
562 <tr>
563 <td>==,!=, >, <, >=, <=</td>
564 <td align="center">yes</td>
565 <td align="center">yes</td>
566 <td align="center">-</td>
567 <td align="center">-</td>
568 </tr>
569 <tr>
570 <td>=</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 <tr>
577 <td>:?</td>
578 <td align="center">yes</td>
579 <td align="center">-</td>
580 <td align="center">-</td>
581 <td align="center">-</td>
582 </tr>
583 <tr>
584 <td>sizeof</td>
585 <td align="center">yes</td>
586 <td align="center">yes</td>
587 <td align="center">yes</td>
588 <td align="center">yes</td>
589 </tr>
590</table>
591
Owen Andersond2bf0cd2010-01-27 01:22:36 +0000[diff] [blame]592<p>See also <a href="#__builtin_shufflevector">__builtin_shufflevector</a>.</p>
593
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]594<!-- ======================================================================= -->
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]595<h2 id="deprecated">Messages on <tt>deprecated</tt> and <tt>unavailable</tt> Attributes</h2>
Fariborz Jahanianc784dc12010-10-06 23:12:32 +0000[diff] [blame]596<!-- ======================================================================= -->
597
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]598<p>An optional string message can be added to the <tt>deprecated</tt>
599and <tt>unavailable</tt> attributes. For example:</p>
Fariborz Jahanianc784dc12010-10-06 23:12:32 +0000[diff] [blame]600
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]601<blockquote>
Chris Lattner4836d6a2010-11-09 19:43:35 +0000[diff] [blame]602<pre>void explode(void) __attribute__((deprecated("extremely unsafe, use 'combust' instead!!!")));</pre>
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]603</blockquote>
604
605<p>If the deprecated or unavailable declaration is used, the message
606will be incorporated into the appropriate diagnostic:</p>
607
608<blockquote>
Benjamin Kramerb4556862012-03-19 19:12:30 +0000[diff] [blame]609<pre>harmless.c:4:3: warning: 'explode' is deprecated: extremely unsafe, use 'combust' instead!!!
610 [-Wdeprecated-declarations]
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]611 explode();
612 ^</pre>
613</blockquote>
614
615<p>Query for this feature
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]616with <tt>__has_extension(attribute_deprecated_with_message)</tt>
617and <tt>__has_extension(attribute_unavailable_with_message)</tt>.</p>
John McCall48209082010-11-08 19:48:17 +0000[diff] [blame]618
619<!-- ======================================================================= -->
620<h2 id="attributes-on-enumerators">Attributes on Enumerators</h2>
621<!-- ======================================================================= -->
622
623<p>Clang allows attributes to be written on individual enumerators.
624This allows enumerators to be deprecated, made unavailable, etc. The
625attribute must appear after the enumerator name and before any
626initializer, like so:</p>
627
628<blockquote>
629<pre>enum OperationMode {
630 OM_Invalid,
631 OM_Normal,
632 OM_Terrified __attribute__((deprecated)),
633 OM_AbortOnError __attribute__((deprecated)) = 4
634};</pre>
635</blockquote>
636
637<p>Attributes on the <tt>enum</tt> declaration do not apply to
638individual enumerators.</p>
639
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]640<p>Query for this feature with <tt>__has_extension(enumerator_attributes)</tt>.</p>
Fariborz Jahanianc784dc12010-10-06 23:12:32 +0000[diff] [blame]641
642<!-- ======================================================================= -->
Daniel Dunbar85ff9692012-04-05 17:10:06 +0000[diff] [blame]643<h2 id="user_specified_system_framework">'User-Specified' System Frameworks</h2>
644<!-- ======================================================================= -->
645
646<p>Clang provides a mechanism by which frameworks can be built in such a way
647that they will always be treated as being 'system frameworks', even if they are
648not present in a system framework directory. This can be useful to system
649framework developers who want to be able to test building other applications
650with development builds of their framework, including the manner in which the
651compiler changes warning behavior for system headers.</p>
652
653<p>Framework developers can opt-in to this mechanism by creating a
654'.system_framework' file at the top-level of their framework. That is, the
655framework should have contents like:</p>
656
657<pre>
658 .../TestFramework.framework
659 .../TestFramework.framework/.system_framework
660 .../TestFramework.framework/Headers
661 .../TestFramework.framework/Headers/TestFramework.h
662 ...
663</pre>
664
665<p>Clang will treat the presence of this file as an indicator that the framework
666should be treated as a system framework, regardless of how it was found in the
667framework search path. For consistency, we recommend that such files never be
668included in installed versions of the framework.</p>
669
670<!-- ======================================================================= -->
Dmitri Gribenkoccc4edf2012-05-27 14:08:44 +0000[diff] [blame]671<h2 id="availability">Availability attribute</h2>
Douglas Gregor93a70672012-03-11 04:53:21 +0000[diff] [blame]672<!-- ======================================================================= -->
673
674<p>Clang introduces the <code>availability</code> attribute, which can
675be placed on declarations to describe the lifecycle of that
676declaration relative to operating system versions. Consider the function declaration for a hypothetical function <code>f</code>:</p>
677
678<pre>
679void f(void) __attribute__((availability(macosx,introduced=10.4,deprecated=10.6,obsoleted=10.7)));
680</pre>
681
682<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>
683
684<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>
685
686<dl>
687 <dt>introduced=<i>version</i></dt>
688 <dd>The first version in which this declaration was introduced.</dd>
689
690 <dt>deprecated=<i>version</i></dt>
691 <dd>The first version in which this declaration was deprecated, meaning that users should migrate away from this API.</dd>
692
693 <dt>obsoleted=<i>version</i></dt>
694 <dd>The first version in which this declaration was obsoleted, meaning that it was removed completely and can no longer be used.</dd>
695
696 <dt>unavailable</dt>
697 <dd>This declaration is never available on this platform.</dd>
698
699 <dt>message=<i>string-literal</i></dt>
700 <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>
701</dl>
702
703<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>
704
705<dl>
706 <dt>ios</dt>
707 <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>
708
709 <dt>macosx</dt>
710 <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>
711</dl>
712
Douglas Gregor594f8412012-03-11 17:21:03 +0000[diff] [blame]713<p>A declaration can be used even when deploying back to a platform
714version prior to when the declaration was introduced. When this
715happens, the declaration is <a
716 href="https://developer.apple.com/library/mac/#documentation/MacOSX/Conceptual/BPFrameworks/Concepts/WeakLinking.html">weakly
717linked</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>
718
Douglas Gregor93a70672012-03-11 04:53:21 +0000[diff] [blame]719<!-- ======================================================================= -->
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]720<h2 id="checking_language_features">Checks for Standard Language Features</h2>
721<!-- ======================================================================= -->
722
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]723<p>The <tt>__has_feature</tt> macro can be used to query if certain standard
724language features are enabled. The <tt>__has_extension</tt> macro can be used
725to query if language features are available as an extension when compiling for
726a standard which does not provide them. The features which can be tested are
727listed here.</p>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]728
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]729<h3 id="cxx98">C++98</h3>
730
731<p>The features listed below are part of the C++98 standard. These features are
732enabled by default when compiling C++ code.</p>
733
734<h4 id="cxx_exceptions">C++ exceptions</h4>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]735
Ted Kremenek22c34102009-12-03 02:05:57 +0000[diff] [blame]736<p>Use <tt>__has_feature(cxx_exceptions)</tt> to determine if C++ exceptions have been enabled. For
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]737example, compiling code with <tt>-fno-exceptions</tt> disables C++ exceptions.</p>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]738
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]739<h4 id="cxx_rtti">C++ RTTI</h4>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]740
Ted Kremenek0eb95602009-12-03 02:06:43 +0000[diff] [blame]741<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]742compiling code with <tt>-fno-rtti</tt> disables the use of RTTI.</p>
Ted Kremenek87774fd2009-12-03 02:04:01 +0000[diff] [blame]743
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]744<h3 id="cxx11">C++11</h3>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]745
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]746<p>The features listed below are part of the C++11 standard. As a result, all
747these features are enabled with the <tt>-std=c++11</tt> or <tt>-std=gnu++11</tt>
748option when compiling C++ code.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]749
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]750<h4 id="cxx_access_control_sfinae">C++11 SFINAE includes access control</h4>
Douglas Gregor7822ee32011-05-11 23:45:11 +0000[diff] [blame]751
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]752<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]753
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]754<h4 id="cxx_alias_templates">C++11 alias templates</h4>
Richard Smith3e4c6c42011-05-05 21:57:07 +0000[diff] [blame]755
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]756<p>Use <tt>__has_feature(cxx_alias_templates)</tt> or
757<tt>__has_extension(cxx_alias_templates)</tt> to determine if support for
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]758C++11's alias declarations and alias templates is enabled.</p>
Richard Smith3e4c6c42011-05-05 21:57:07 +0000[diff] [blame]759
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]760<h4 id="cxx_alignas">C++11 alignment specifiers</h4>
Peter Collingbournefd5f6862011-10-14 23:44:46 +0000[diff] [blame]761
762<p>Use <tt>__has_feature(cxx_alignas)</tt> or
763<tt>__has_extension(cxx_alignas)</tt> to determine if support for alignment
764specifiers using <tt>alignas</tt> is enabled.</p>
765
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]766<h4 id="cxx_attributes">C++11 attributes</h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]767
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]768<p>Use <tt>__has_feature(cxx_attributes)</tt> or
769<tt>__has_extension(cxx_attributes)</tt> to determine if support for attribute
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]770parsing with C++11's square bracket notation is enabled.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]771
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]772<h4 id="cxx_constexpr">C++11 generalized constant expressions</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]773
774<p>Use <tt>__has_feature(cxx_constexpr)</tt> to determine if support
775for generalized constant expressions (e.g., <tt>constexpr</tt>) is
Richard Smithb5216aa2012-02-14 22:56:17 +0000[diff] [blame]776enabled.</p>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]777
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]778<h4 id="cxx_decltype">C++11 <tt>decltype()</tt></h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]779
780<p>Use <tt>__has_feature(cxx_decltype)</tt> or
781<tt>__has_extension(cxx_decltype)</tt> to determine if support for the
Douglas Gregor316551f2012-04-10 20:00:33 +0000[diff] [blame]782<tt>decltype()</tt> specifier is enabled. C++11's <tt>decltype</tt>
783does not require type-completeness of a function call expression.
784Use <tt>__has_feature(cxx_decltype_incomplete_return_types)</tt>
785or <tt>__has_extension(cxx_decltype_incomplete_return_types)</tt>
786to determine if support for this feature is enabled.</p>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]787
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]788<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]789
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]790<p>Use <tt>__has_feature(cxx_default_function_template_args)</tt> or
791<tt>__has_extension(cxx_default_function_template_args)</tt> to determine
792if support for default template arguments in function templates is enabled.</p>
Douglas Gregor07508002011-02-05 20:35:30 +0000[diff] [blame]793
Douglas Gregorf695a692011-11-01 01:19:34 +0000[diff] [blame]794<h4 id="cxx_defaulted_functions">C++11 <tt>default</tt>ed functions</h4>
795
796<p>Use <tt>__has_feature(cxx_defaulted_functions)</tt> or
797<tt>__has_extension(cxx_defaulted_functions)</tt> to determine if support for
798defaulted function definitions (with <tt>= default</tt>) is enabled.</p>
799
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]800<h4 id="cxx_delegating_constructors">C++11 delegating constructors</h4>
Sean Huntd9624992011-06-23 06:11:37 +0000[diff] [blame]801
802<p>Use <tt>__has_feature(cxx_delegating_constructors)</tt> to determine if
803support for delegating constructors is enabled.</p>
804
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]805<h4 id="cxx_deleted_functions">C++11 <tt>delete</tt>d functions</h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]806
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]807<p>Use <tt>__has_feature(cxx_deleted_functions)</tt> or
808<tt>__has_extension(cxx_deleted_functions)</tt> to determine if support for
Sebastian Redlf6c09772010-08-31 23:28:47 +0000[diff] [blame]809deleted function definitions (with <tt>= delete</tt>) is enabled.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]810
Benjamin Kramer665a8dc2012-01-15 15:26:07 +0000[diff] [blame]811<h4 id="cxx_explicit_conversions">C++11 explicit conversion functions</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]812<p>Use <tt>__has_feature(cxx_explicit_conversions)</tt> to determine if support for <tt>explicit</tt> conversion functions is enabled.</p>
813
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]814<h4 id="cxx_generalized_initializers">C++11 generalized initializers</h4>
Sean Hunte1f6dea2011-08-07 00:34:32 +0000[diff] [blame]815
816<p>Use <tt>__has_feature(cxx_generalized_initializers)</tt> to determine if
817support for generalized initializers (using braced lists and
Richard Smith88189552012-02-26 07:09:21 +0000[diff] [blame]818<tt>std::initializer_list</tt>) is enabled.</p>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]819
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]820<h4 id="cxx_implicit_moves">C++11 implicit move constructors/assignment operators</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]821
Sebastian Redl72a81d22011-10-10 18:10:00 +0000[diff] [blame]822<p>Use <tt>__has_feature(cxx_implicit_moves)</tt> to determine if Clang will
823implicitly generate move constructors and move assignment operators where needed.</p>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]824
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]825<h4 id="cxx_inheriting_constructors">C++11 inheriting constructors</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]826
827<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>
828
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]829<h4 id="cxx_inline_namespaces">C++11 inline namespaces</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]830
831<p>Use <tt>__has_feature(cxx_inline_namespaces)</tt> or
832<tt>__has_extension(cxx_inline_namespaces)</tt> to determine if support for
833inline namespaces is enabled.</p>
Sean Hunte1f6dea2011-08-07 00:34:32 +0000[diff] [blame]834
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]835<h4 id="cxx_lambdas">C++11 lambdas</h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]836
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]837<p>Use <tt>__has_feature(cxx_lambdas)</tt> or
838<tt>__has_extension(cxx_lambdas)</tt> to determine if support for lambdas
Douglas Gregor46e021e2012-02-23 05:44:09 +0000[diff] [blame]839is enabled. </p>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]840
Douglas Gregor7b156dd2012-04-04 00:48:39 +0000[diff] [blame]841<h4 id="cxx_local_type_template_args">C++11 local and unnamed types as template arguments</h4>
842
843<p>Use <tt>__has_feature(cxx_local_type_template_args)</tt> or
844<tt>__has_extension(cxx_local_type_template_args)</tt> to determine if
845support for local and unnamed types as template arguments is enabled.</p>
846
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]847<h4 id="cxx_noexcept">C++11 noexcept</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]848
849<p>Use <tt>__has_feature(cxx_noexcept)</tt> or
850<tt>__has_extension(cxx_noexcept)</tt> to determine if support for noexcept
851exception specifications is enabled.</p>
852
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]853<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]854
855<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]856
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]857<h4 id="cxx_nullptr">C++11 <tt>nullptr</tt></h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]858
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]859<p>Use <tt>__has_feature(cxx_nullptr)</tt> or
860<tt>__has_extension(cxx_nullptr)</tt> to determine if support for
Douglas Gregor84ee2ee2011-05-21 23:15:46 +0000[diff] [blame]861<tt>nullptr</tt> is enabled.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]862
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]863<h4 id="cxx_override_control">C++11 <tt>override control</tt></h4>
Anders Carlssonc8b9f792011-03-25 15:04:23 +0000[diff] [blame]864
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]865<p>Use <tt>__has_feature(cxx_override_control)</tt> or
866<tt>__has_extension(cxx_override_control)</tt> to determine if support for
Anders Carlssonc8b9f792011-03-25 15:04:23 +0000[diff] [blame]867the override control keywords is enabled.</p>
868
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]869<h4 id="cxx_reference_qualified_functions">C++11 reference-qualified functions</h4>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]870<p>Use <tt>__has_feature(cxx_reference_qualified_functions)</tt> or
871<tt>__has_extension(cxx_reference_qualified_functions)</tt> to determine
872if support for reference-qualified functions (e.g., member functions with
873<code>&amp;</code> or <code>&amp;&amp;</code> applied to <code>*this</code>)
874is enabled.</p>
Douglas Gregor56209ff2011-01-26 21:25:54 +0000[diff] [blame]875
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]876<h4 id="cxx_range_for">C++11 range-based <tt>for</tt> loop</h4>
Richard Smitha391a462011-04-15 15:14:40 +0000[diff] [blame]877
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]878<p>Use <tt>__has_feature(cxx_range_for)</tt> or
879<tt>__has_extension(cxx_range_for)</tt> to determine if support for the
880range-based for loop is enabled. </p>
Richard Smitha391a462011-04-15 15:14:40 +0000[diff] [blame]881
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]882<h4 id="cxx_raw_string_literals">C++11 raw string literals</h4>
Richard Smith80134582012-03-07 08:57:31 +0000[diff] [blame]883<p>Use <tt>__has_feature(cxx_raw_string_literals)</tt> to determine if support
884for 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]885
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]886<h4 id="cxx_rvalue_references">C++11 rvalue references</h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]887
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]888<p>Use <tt>__has_feature(cxx_rvalue_references)</tt> or
889<tt>__has_extension(cxx_rvalue_references)</tt> to determine if support for
Douglas Gregor56209ff2011-01-26 21:25:54 +0000[diff] [blame]890rvalue references is enabled. </p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]891
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]892<h4 id="cxx_static_assert">C++11 <tt>static_assert()</tt></h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]893
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]894<p>Use <tt>__has_feature(cxx_static_assert)</tt> or
895<tt>__has_extension(cxx_static_assert)</tt> to determine if support for
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]896compile-time assertions using <tt>static_assert</tt> is enabled.</p>
897
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]898<h4 id="cxx_auto_type">C++11 type inference</h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]899
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]900<p>Use <tt>__has_feature(cxx_auto_type)</tt> or
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]901<tt>__has_extension(cxx_auto_type)</tt> to determine C++11 type inference is
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]902supported using the <tt>auto</tt> specifier. If this is disabled, <tt>auto</tt>
903will instead be a storage class specifier, as in C or C++98.</p>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]904
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]905<h4 id="cxx_strong_enums">C++11 strongly typed enumerations</h4>
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]906
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]907<p>Use <tt>__has_feature(cxx_strong_enums)</tt> or
908<tt>__has_extension(cxx_strong_enums)</tt> to determine if support for
909strongly typed, scoped enumerations is enabled.</p>
Sebastian Redlf6c09772010-08-31 23:28:47 +0000[diff] [blame]910
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]911<h4 id="cxx_trailing_return">C++11 trailing return type</h4>
Douglas Gregordab60ad2010-10-01 18:44:50 +0000[diff] [blame]912
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]913<p>Use <tt>__has_feature(cxx_trailing_return)</tt> or
914<tt>__has_extension(cxx_trailing_return)</tt> to determine if support for the
915alternate function declaration syntax with trailing return type is enabled.</p>
Douglas Gregordab60ad2010-10-01 18:44:50 +0000[diff] [blame]916
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]917<h4 id="cxx_unicode_literals">C++11 Unicode string literals</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]918<p>Use <tt>__has_feature(cxx_unicode_literals)</tt> to determine if
919support for Unicode string literals is enabled.</p>
Sebastian Redl4561ecd2011-03-15 21:17:12 +0000[diff] [blame]920
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]921<h4 id="cxx_unrestricted_unions">C++11 unrestricted unions</h4>
Sebastian Redl4561ecd2011-03-15 21:17:12 +0000[diff] [blame]922
Richard Smithec92bc72012-03-03 23:51:05 +0000[diff] [blame]923<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]924
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]925<h4 id="cxx_user_literals">C++11 user-defined literals</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]926
Richard Smith9c1dda72012-03-09 08:41:27 +0000[diff] [blame]927<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]928
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]929<h4 id="cxx_variadic_templates">C++11 variadic templates</h4>
Douglas Gregorece38942011-08-29 17:28:38 +0000[diff] [blame]930
931<p>Use <tt>__has_feature(cxx_variadic_templates)</tt> or
932<tt>__has_extension(cxx_variadic_templates)</tt> to determine if support
933for variadic templates is enabled.</p>
Douglas Gregor1274ccd2010-10-08 23:50:27 +0000[diff] [blame]934
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]935<h3 id="c11">C11</h3>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]936
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]937<p>The features listed below are part of the C11 standard. As a result, all
938these features are enabled with the <tt>-std=c11</tt> or <tt>-std=gnu11</tt>
939option when compiling C code. Additionally, because these features are all
940backward-compatible, they are available as extensions in all language modes.</p>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]941
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]942<h4 id="c_alignas">C11 alignment specifiers</h4>
Peter Collingbournefd5f6862011-10-14 23:44:46 +0000[diff] [blame]943
944<p>Use <tt>__has_feature(c_alignas)</tt> or <tt>__has_extension(c_alignas)</tt>
945to determine if support for alignment specifiers using <tt>_Alignas</tt>
946is enabled.</p>
947
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]948<h4 id="c_atomic">C11 atomic operations</h4>
949
950<p>Use <tt>__has_feature(c_atomic)</tt> or <tt>__has_extension(c_atomic)</tt>
951to determine if support for atomic types using <tt>_Atomic</tt> is enabled.
952Clang also provides <a href="#__c11_atomic">a set of builtins</a> which can be
Richard Smithc495e602012-04-19 17:46:52 +0000[diff] [blame]953used to implement the <tt>&lt;stdatomic.h&gt;</tt> operations on
954<tt>_Atomic</tt> types.</p>
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]955
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]956<h4 id="c_generic_selections">C11 generic selections</h4>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]957
958<p>Use <tt>__has_feature(c_generic_selections)</tt> or
959<tt>__has_extension(c_generic_selections)</tt> to determine if support for
960generic selections is enabled.</p>
961
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]962<p>As an extension, the C11 generic selection expression is available in all
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]963languages supported by Clang. The syntax is the same as that given in the
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]964C11 standard.</p>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]965
966<p>In C, type compatibility is decided according to the rules given in the
967appropriate standard, but in C++, which lacks the type compatibility rules
968used in C, types are considered compatible only if they are equivalent.</p>
969
Benjamin Kramerffbe9b92011-12-23 17:00:35 +0000[diff] [blame]970<h4 id="c_static_assert">C11 <tt>_Static_assert()</tt></h4>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]971
972<p>Use <tt>__has_feature(c_static_assert)</tt> or
973<tt>__has_extension(c_static_assert)</tt> to determine if support for
974compile-time assertions using <tt>_Static_assert</tt> is enabled.</p>
975
Sean Hunt4ef4c6b2010-01-13 08:31:49 +0000[diff] [blame]976<!-- ======================================================================= -->
Douglas Gregorafdf1372011-02-03 21:57:35 +0000[diff] [blame]977<h2 id="checking_type_traits">Checks for Type Traits</h2>
978<!-- ======================================================================= -->
979
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]980<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]981<blockquote>
982<pre>
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]983#if __has_extension(is_convertible_to)
Douglas Gregorafdf1372011-02-03 21:57:35 +0000[diff] [blame]984template&lt;typename From, typename To&gt;
985struct is_convertible_to {
986 static const bool value = __is_convertible_to(From, To);
987};
988#else
989// Emulate type trait
990#endif
991</pre>
992</blockquote>
993
994<p>The following type traits are supported by Clang:</p>
995<ul>
996 <li><code>__has_nothrow_assign</code> (GNU, Microsoft)</li>
997 <li><code>__has_nothrow_copy</code> (GNU, Microsoft)</li>
998 <li><code>__has_nothrow_constructor</code> (GNU, Microsoft)</li>
999 <li><code>__has_trivial_assign</code> (GNU, Microsoft)</li>
1000 <li><code>__has_trivial_copy</code> (GNU, Microsoft)</li>
1001 <li><code>__has_trivial_constructor</code> (GNU, Microsoft)</li>
1002 <li><code>__has_trivial_destructor</code> (GNU, Microsoft)</li>
1003 <li><code>__has_virtual_destructor</code> (GNU, Microsoft)</li>
1004 <li><code>__is_abstract</code> (GNU, Microsoft)</li>
1005 <li><code>__is_base_of</code> (GNU, Microsoft)</li>
1006 <li><code>__is_class</code> (GNU, Microsoft)</li>
1007 <li><code>__is_convertible_to</code> (Microsoft)</li>
1008 <li><code>__is_empty</code> (GNU, Microsoft)</li>
1009 <li><code>__is_enum</code> (GNU, Microsoft)</li>
John McCallea30e2f2012-09-25 07:32:49 +0000[diff] [blame]1010 <li><code>__is_interface_class</code> (Microsoft)</li>
Douglas Gregorafdf1372011-02-03 21:57:35 +0000[diff] [blame]1011 <li><code>__is_pod</code> (GNU, Microsoft)</li>
1012 <li><code>__is_polymorphic</code> (GNU, Microsoft)</li>
1013 <li><code>__is_union</code> (GNU, Microsoft)</li>
1014 <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]1015 <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]1016 <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]1017 <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>
1018 <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]1019</ul>
1020
1021<!-- ======================================================================= -->
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1022<h2 id="blocks">Blocks</h2>
1023<!-- ======================================================================= -->
1024
Chris Lattnera7dbdf52009-03-09 07:03:22 +0000[diff] [blame]1025<p>The syntax and high level language feature description is in <a
1026href="BlockLanguageSpec.txt">BlockLanguageSpec.txt</a>. Implementation and ABI
1027details for the clang implementation are in <a
Chris Lattner5d7650b2010-03-16 21:43:03 +0000[diff] [blame]1028href="Block-ABI-Apple.txt">Block-ABI-Apple.txt</a>.</p>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1029
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]1030
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]1031<p>Query for this feature with __has_extension(blocks).</p>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]1032
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1033<!-- ======================================================================= -->
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]1034<h2 id="objc_features">Objective-C Features</h2>
1035<!-- ======================================================================= -->
1036
1037<h3 id="objc_instancetype">Related result types</h3>
1038
1039<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>
1040
1041<blockquote>
1042<pre>
1043@interface NSObject
1044+ (id)alloc;
1045- (id)init;
1046@end
1047
1048@interface NSArray : NSObject
1049@end
1050</pre>
1051</blockquote>
1052
1053<p>and this common initialization pattern</p>
1054
1055<blockquote>
1056<pre>
1057NSArray *array = [[NSArray alloc] init];
1058</pre>
1059</blockquote>
1060
1061<p>the type of the expression <code>[NSArray alloc]</code> is
1062<code>NSArray*</code> because <code>alloc</code> implicitly has a
1063related result type. Similarly, the type of the expression
1064<code>[[NSArray alloc] init]</code> is <code>NSArray*</code>, since
1065<code>init</code> has a related result type and its receiver is known
1066to 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>
1067
Douglas Gregore97179c2011-09-08 01:46:34 +0000[diff] [blame]1068<p>A method with a related result type can be declared by using the
1069type <tt>instancetype</tt> as its result type. <tt>instancetype</tt>
1070is a contextual keyword that is only permitted in the result type of
1071an Objective-C method, e.g.</p>
1072
1073<pre>
1074@interface A
1075+ (<b>instancetype</b>)constructAnA;
1076@end
1077</pre>
1078
1079<p>The related result type can also be inferred for some methods.
1080To determine whether a method has an inferred related result type, the first
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]1081word in the camel-case selector (e.g., "init" in "initWithObjects") is
Douglas Gregor8a0ace62011-11-03 18:33:01 +0000[diff] [blame]1082considered, and the method will have a related result type if its return
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]1083type is compatible with the type of its class and if</p>
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]1084
1085<ul>
1086
1087 <li>the first word is "alloc" or "new", and the method is a class
1088 method, or</li>
1089
1090 <li>the first word is "autorelease", "init", "retain", or "self",
1091 and the method is an instance method.</li>
1092
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]1093</ul>
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]1094
1095<p>If a method with a related result type is overridden by a subclass
1096method, the subclass method must also return a type that is compatible
1097with the subclass type. For example:</p>
1098
1099<blockquote>
1100<pre>
1101@interface NSString : NSObject
1102- (NSUnrelated *)init; // incorrect usage: NSUnrelated is not NSString or a superclass of NSString
1103@end
1104</pre>
1105</blockquote>
1106
1107<p>Related result types only affect the type of a message send or
1108property access via the given method. In all other respects, a method
Douglas Gregore97179c2011-09-08 01:46:34 +0000[diff] [blame]1109with a related result type is treated the same way as method that
1110returns <tt>id</tt>.</p>
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]1111
Douglas Gregoraebb6532011-09-08 17:19:31 +0000[diff] [blame]1112<p>Use <tt>__has_feature(objc_instancetype)</tt> to determine whether
1113the <tt>instancetype</tt> contextual keyword is available.</p>
1114
Douglas Gregor926df6c2011-06-11 01:09:30 +0000[diff] [blame]1115<!-- ======================================================================= -->
John McCallf85e1932011-06-15 23:02:42 +0000[diff] [blame]1116<h2 id="objc_arc">Automatic reference counting </h2>
1117<!-- ======================================================================= -->
1118
1119<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>
1120
1121<!-- ======================================================================= -->
Douglas Gregor5471bc82011-09-08 17:18:35 +0000[diff] [blame]1122<h2 id="objc_fixed_enum">Enumerations with a fixed underlying type</h2>
1123<!-- ======================================================================= -->
1124
David Blaikie5090e9f2011-10-18 05:49:30 +0000[diff] [blame]1125<p>Clang provides support for C++11 enumerations with a fixed
Douglas Gregor5471bc82011-09-08 17:18:35 +0000[diff] [blame]1126underlying type within Objective-C. For example, one can write an
1127enumeration type as:</p>
1128
1129<pre>
1130typedef enum : unsigned char { Red, Green, Blue } Color;
1131</pre>
1132
1133<p>This specifies that the underlying type, which is used to store the
1134enumeration value, is <tt>unsigned char</tt>.</p>
1135
1136<p>Use <tt>__has_feature(objc_fixed_enum)</tt> to determine whether
1137support for fixed underlying types is available in Objective-C.</p>
1138
1139<!-- ======================================================================= -->
Douglas Gregor8a4e1822012-03-09 23:24:48 +0000[diff] [blame]1140<h2 id="objc_lambdas">Interoperability with C++11 lambdas</h2>
1141<!-- ======================================================================= -->
1142
1143<p>Clang provides interoperability between C++11 lambdas and
1144blocks-based APIs, by permitting a lambda to be implicitly converted
1145to a block pointer with the corresponding signature. For example,
1146consider an API such as <code>NSArray</code>'s array-sorting
1147method:</p>
1148
1149<pre> - (NSArray *)sortedArrayUsingComparator:(NSComparator)cmptr; </pre>
1150
1151<p><code>NSComparator</code> is simply a typedef for the block pointer
1152<code>NSComparisonResult (^)(id, id)</code>, and parameters of this
1153type are generally provided with block literals as arguments. However,
1154one can also use a C++11 lambda so long as it provides the same
1155signature (in this case, accepting two parameters of type
1156<code>id</code> and returning an <code>NSComparisonResult</code>):</p>
1157
1158<pre>
1159 NSArray *array = @[@"string 1", @"string 21", @"string 12", @"String 11",
1160 @"String 02"];
1161 const NSStringCompareOptions comparisonOptions
1162 = NSCaseInsensitiveSearch | NSNumericSearch |
1163 NSWidthInsensitiveSearch | NSForcedOrderingSearch;
1164 NSLocale *currentLocale = [NSLocale currentLocale];
1165 NSArray *sorted
1166 = [array sortedArrayUsingComparator:<b>[=](id s1, id s2) -&gt; NSComparisonResult {
1167 NSRange string1Range = NSMakeRange(0, [s1 length]);
1168 return [s1 compare:s2 options:comparisonOptions
1169 range:string1Range locale:currentLocale];
1170 }</b>];
1171 NSLog(@"sorted: %@", sorted);
1172</pre>
1173
1174<p>This code relies on an implicit conversion from the type of the
1175lambda expression (an unnamed, local class type called the <i>closure
1176type</i>) to the corresponding block pointer type. The conversion
1177itself is expressed by a conversion operator in that closure type
1178that produces a block pointer with the same signature as the lambda
1179itself, e.g.,</p>
1180
1181<pre>
1182 operator NSComparisonResult (^)(id, id)() const;
1183</pre>
1184
1185<p>This conversion function returns a new block that simply forwards
1186the two parameters to the lambda object (which it captures by copy),
1187then returns the result. The returned block is first copied (with
1188<tt>Block_copy</tt>) and then autoreleased. As an optimization, if a
1189lambda expression is immediately converted to a block pointer (as in
1190the first example, above), then the block is not copied and
1191autoreleased: rather, it is given the same lifetime as a block literal
1192written at that point in the program, which avoids the overhead of
1193copying a block to the heap in the common case.</p>
1194
Douglas Gregorbccda482012-03-10 22:20:11 +0000[diff] [blame]1195<p>The conversion from a lambda to a block pointer is only available
1196in Objective-C++, and not in C++ with blocks, due to its use of
1197Objective-C memory management (autorelease).</p>
1198
Douglas Gregor8a4e1822012-03-09 23:24:48 +0000[diff] [blame]1199<!-- ======================================================================= -->
Patrick Beardeb382ec2012-04-19 00:25:12 +0000[diff] [blame]1200<h2 id="objc_object_literals_subscripting">Object Literals and Subscripting</h2>
Patrick Beard62f12342012-03-20 21:51:03 +0000[diff] [blame]1201<!-- ======================================================================= -->
1202
Patrick Beardeb382ec2012-04-19 00:25:12 +0000[diff] [blame]1203<p>Clang provides support for <a href="ObjectiveCLiterals.html">Object Literals
1204and Subscripting</a> in Objective-C, which simplifies common Objective-C
1205programming patterns, makes programs more concise, and improves the safety of
1206container creation. There are several feature macros associated with object
1207literals and subscripting: <code>__has_feature(objc_array_literals)</code>
1208tests the availability of array literals;
1209<code>__has_feature(objc_dictionary_literals)</code> tests the availability of
1210dictionary literals; <code>__has_feature(objc_subscripting)</code> tests the
1211availability of object subscripting.</p>
Patrick Beard62f12342012-03-20 21:51:03 +0000[diff] [blame]1212
1213<!-- ======================================================================= -->
Fariborz Jahanian790880b2012-06-18 17:13:17 +0000[diff] [blame]1214<h2 id="objc_default_synthesize_properties">Objective-C Autosynthesis of Properties</h2>
1215<!-- ======================================================================= -->
1216
1217<p> Clang provides support for autosynthesis of declared properties. Using this
1218feature, clang provides default synthesis of those properties not declared @dynamic
1219and not having user provided backing getter and setter methods.
1220<code>__has_feature(objc_default_synthesize_properties)</code> checks for availability
1221of this feature in version of clang being used.</p>
1222
1223<!-- ======================================================================= -->
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]1224<h2 id="overloading-in-c">Function Overloading in C</h2>
1225<!-- ======================================================================= -->
1226
Chris Lattnerf161d412009-02-13 21:51:45 +0000[diff] [blame]1227<p>Clang provides support for C++ function overloading in C. Function
1228overloading in C is introduced using the <tt>overloadable</tt> attribute. For
1229example, one might provide several overloaded versions of a <tt>tgsin</tt>
1230function that invokes the appropriate standard function computing the sine of a
1231value with <tt>float</tt>, <tt>double</tt>, or <tt>long double</tt>
1232precision:</p>
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]1233
1234<blockquote>
1235<pre>
1236#include &lt;math.h&gt;
1237float <b>__attribute__((overloadable))</b> tgsin(float x) { return sinf(x); }
1238double <b>__attribute__((overloadable))</b> tgsin(double x) { return sin(x); }
1239long double <b>__attribute__((overloadable))</b> tgsin(long double x) { return sinl(x); }
1240</pre>
1241</blockquote>
1242
1243<p>Given these declarations, one can call <tt>tgsin</tt> with a
1244<tt>float</tt> value to receive a <tt>float</tt> result, with a
1245<tt>double</tt> to receive a <tt>double</tt> result, etc. Function
1246overloading in C follows the rules of C++ function overloading to pick
1247the best overload given the call arguments, with a few C-specific
1248semantics:</p>
1249<ul>
1250 <li>Conversion from <tt>float</tt> or <tt>double</tt> to <tt>long
1251 double</tt> is ranked as a floating-point promotion (per C99) rather
1252 than as a floating-point conversion (as in C++).</li>
1253
1254 <li>A conversion from a pointer of type <tt>T*</tt> to a pointer of type
1255 <tt>U*</tt> is considered a pointer conversion (with conversion
1256 rank) if <tt>T</tt> and <tt>U</tt> are compatible types.</li>
1257
1258 <li>A conversion from type <tt>T</tt> to a value of type <tt>U</tt>
1259 is permitted if <tt>T</tt> and <tt>U</tt> are compatible types. This
1260 conversion is given "conversion" rank.</li>
1261</ul>
1262
1263<p>The declaration of <tt>overloadable</tt> functions is restricted to
1264function declarations and definitions. Most importantly, if any
1265function with a given name is given the <tt>overloadable</tt>
1266attribute, then all function declarations and definitions with that
1267name (and in that scope) must have the <tt>overloadable</tt>
Chris Lattnerf161d412009-02-13 21:51:45 +0000[diff] [blame]1268attribute. This rule even applies to redeclarations of functions whose original
1269declaration had the <tt>overloadable</tt> attribute, e.g.,</p>
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]1270
1271<blockquote>
1272<pre>
1273int f(int) __attribute__((overloadable));
1274float f(float); <i>// error: declaration of "f" must have the "overloadable" attribute</i>
1275
1276int g(int) __attribute__((overloadable));
1277int g(int) { } <i>// error: redeclaration of "g" must also have the "overloadable" attribute</i>
1278</pre>
1279</blockquote>
1280
Douglas Gregor965acbb2009-02-18 07:07:28 +0000[diff] [blame]1281<p>Functions marked <tt>overloadable</tt> must have
1282prototypes. Therefore, the following code is ill-formed:</p>
1283
1284<blockquote>
1285<pre>
1286int h() __attribute__((overloadable)); <i>// error: h does not have a prototype</i>
1287</pre>
1288</blockquote>
1289
1290<p>However, <tt>overloadable</tt> functions are allowed to use a
1291ellipsis 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>
1292
1293<blockquote>
1294<pre>
Chris Lattner02246802009-02-18 22:27:46 +0000[diff] [blame]1295void honeypot(...) __attribute__((overloadable, unavailable)); <i>// calling me is an error</i>
Douglas Gregor965acbb2009-02-18 07:07:28 +0000[diff] [blame]1296</pre>
1297</blockquote>
1298
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]1299<p>Functions declared with the <tt>overloadable</tt> attribute have
1300their names mangled according to the same rules as C++ function
1301names. For example, the three <tt>tgsin</tt> functions in our
1302motivating example get the mangled names <tt>_Z5tgsinf</tt>,
Chris Lattner71b48d62010-11-28 18:19:13 +0000[diff] [blame]1303<tt>_Z5tgsind</tt>, and <tt>_Z5tgsine</tt>, respectively. There are two
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]1304caveats to this use of name mangling:</p>
1305
1306<ul>
1307
1308 <li>Future versions of Clang may change the name mangling of
1309 functions overloaded in C, so you should not depend on an specific
1310 mangling. To be completely safe, we strongly urge the use of
1311 <tt>static inline</tt> with <tt>overloadable</tt> functions.</li>
1312
1313 <li>The <tt>overloadable</tt> attribute has almost no meaning when
1314 used in C++, because names will already be mangled and functions are
1315 already overloadable. However, when an <tt>overloadable</tt>
1316 function occurs within an <tt>extern "C"</tt> linkage specification,
1317 it's name <i>will</i> be mangled in the same way as it would in
1318 C.</li>
1319</ul>
1320
Peter Collingbournec1b5fa42011-05-13 20:54:45 +0000[diff] [blame]1321<p>Query for this feature with __has_extension(attribute_overloadable).</p>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]1322
Eli Friedman0c706c22011-09-19 23:17:44 +0000[diff] [blame]1323<!-- ======================================================================= -->
1324<h2 id="complex-list-init">Initializer lists for complex numbers in C</h2>
1325<!-- ======================================================================= -->
1326
1327<p>clang supports an extension which allows the following in C:</p>
1328
1329<blockquote>
1330<pre>
1331#include &lt;math.h&gt;
1332#include &lt;complex.h&gt;
1333complex float x = { 1.0f, INFINITY }; // Init to (1, Inf)
1334</pre>
1335</blockquote>
1336
1337<p>This construct is useful because there is no way to separately
1338initialize the real and imaginary parts of a complex variable in
1339standard C, given that clang does not support <code>_Imaginary</code>.
1340(clang also supports the <code>__real__</code> and <code>__imag__</code>
1341extensions from gcc, which help in some cases, but are not usable in
1342static initializers.)
1343
1344<p>Note that this extension does not allow eliding the braces; the
1345meaning of the following two lines is different:</p>
1346
1347<blockquote>
1348<pre>
1349complex float x[] = { { 1.0f, 1.0f } }; // [0] = (1, 1)
1350complex float x[] = { 1.0f, 1.0f }; // [0] = (1, 0), [1] = (1, 0)
1351</pre>
1352</blockquote>
1353
1354<p>This extension also works in C++ mode, as far as that goes, but does not
1355 apply to the C++ <code>std::complex</code>. (In C++11, list
1356 initialization allows the same syntax to be used with
1357 <code>std::complex</code> with the same meaning.)
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]1358
Douglas Gregorcb54d432009-02-13 00:57:04 +0000[diff] [blame]1359<!-- ======================================================================= -->
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1360<h2 id="builtins">Builtin Functions</h2>
1361<!-- ======================================================================= -->
1362
1363<p>Clang supports a number of builtin library functions with the same syntax as
1364GCC, including things like <tt>__builtin_nan</tt>,
1365<tt>__builtin_constant_p</tt>, <tt>__builtin_choose_expr</tt>,
1366<tt>__builtin_types_compatible_p</tt>, <tt>__sync_fetch_and_add</tt>, etc. In
1367addition to the GCC builtins, Clang supports a number of builtins that GCC does
1368not, which are listed here.</p>
1369
1370<p>Please note that Clang does not and will not support all of the GCC builtins
1371for vector operations. Instead of using builtins, you should use the functions
1372defined in target-specific header files like <tt>&lt;xmmintrin.h&gt;</tt>, which
1373define portable wrappers for these. Many of the Clang versions of these
1374functions are implemented directly in terms of <a href="#vectors">extended
1375vector support</a> instead of builtins, in order to reduce the number of
1376builtins that we need to implement.</p>
1377
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1378<!-- ======================================================================= -->
Hal Finkela841c192012-08-05 22:03:08 +0000[diff] [blame]1379<h3><a name="__builtin_readcyclecounter">__builtin_readcyclecounter</a></h3>
1380<!-- ======================================================================= -->
1381
1382<p><tt>__builtin_readcyclecounter</tt> is used to access the cycle counter
1383register (or a similar low-latency, high-accuracy clock) on those targets that
1384support it.
1385</p>
1386
1387<p><b>Syntax:</b></p>
1388
1389<pre>
1390__builtin_readcyclecounter()
1391</pre>
1392
1393<p><b>Example of Use:</b></p>
1394
1395<pre>
1396unsigned long long t0 = __builtin_readcyclecounter();
1397do_something();
1398unsigned long long t1 = __builtin_readcyclecounter();
1399unsigned long long cycles_to_do_something = t1 - t0; // assuming no overflow
1400</pre>
1401
1402<p><b>Description:</b></p>
1403
1404<p>The __builtin_readcyclecounter() builtin returns the cycle counter value,
1405which may be either global or process/thread-specific depending on the target.
1406As the backing counters often overflow quickly (on the order of
1407seconds) this should only be used for timing small intervals. When not
1408supported by the target, the return value is always zero. This builtin
1409takes no arguments and produces an unsigned long long result.
1410</p>
1411
1412<p>Query for this feature with __has_builtin(__builtin_readcyclecounter).</p>
1413
1414<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]1415<h3><a name="__builtin_shufflevector">__builtin_shufflevector</a></h3>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1416<!-- ======================================================================= -->
1417
Chris Lattneraad826b2009-09-16 18:56:12 +0000[diff] [blame]1418<p><tt>__builtin_shufflevector</tt> is used to express generic vector
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]1419permutation/shuffle/swizzle operations. This builtin is also very important for
1420the implementation of various target-specific header files like
1421<tt>&lt;xmmintrin.h&gt;</tt>.
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1422</p>
1423
1424<p><b>Syntax:</b></p>
1425
1426<pre>
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]1427__builtin_shufflevector(vec1, vec2, index1, index2, ...)
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1428</pre>
1429
1430<p><b>Examples:</b></p>
1431
1432<pre>
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]1433 // Identity operation - return 4-element vector V1.
1434 __builtin_shufflevector(V1, V1, 0, 1, 2, 3)
1435
1436 // "Splat" element 0 of V1 into a 4-element result.
1437 __builtin_shufflevector(V1, V1, 0, 0, 0, 0)
1438
1439 // Reverse 4-element vector V1.
1440 __builtin_shufflevector(V1, V1, 3, 2, 1, 0)
1441
1442 // Concatenate every other element of 4-element vectors V1 and V2.
1443 __builtin_shufflevector(V1, V2, 0, 2, 4, 6)
1444
1445 // Concatenate every other element of 8-element vectors V1 and V2.
1446 __builtin_shufflevector(V1, V2, 0, 2, 4, 6, 8, 10, 12, 14)
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1447</pre>
1448
1449<p><b>Description:</b></p>
1450
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]1451<p>The first two arguments to __builtin_shufflevector are vectors that have the
1452same element type. The remaining arguments are a list of integers that specify
1453the elements indices of the first two vectors that should be extracted and
1454returned in a new vector. These element indices are numbered sequentially
1455starting with the first vector, continuing into the second vector. Thus, if
1456vec1 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]1457</p>
1458
Chris Lattner6f72da52009-02-13 20:00:20 +0000[diff] [blame]1459<p>The result of __builtin_shufflevector is a vector
1460with the same element type as vec1/vec2 but that has an element count equal to
1461the number of indices specified.
1462</p>
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]1463
Chris Lattner21190d52009-09-21 03:09:59 +0000[diff] [blame]1464<p>Query for this feature with __has_builtin(__builtin_shufflevector).</p>
1465
1466<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]1467<h3><a name="__builtin_unreachable">__builtin_unreachable</a></h3>
Chris Lattner21190d52009-09-21 03:09:59 +0000[diff] [blame]1468<!-- ======================================================================= -->
1469
1470<p><tt>__builtin_unreachable</tt> is used to indicate that a specific point in
1471the program cannot be reached, even if the compiler might otherwise think it
1472can. This is useful to improve optimization and eliminates certain warnings.
1473For example, without the <tt>__builtin_unreachable</tt> in the example below,
1474the compiler assumes that the inline asm can fall through and prints a "function
1475declared 'noreturn' should not return" warning.
1476</p>
1477
1478<p><b>Syntax:</b></p>
1479
1480<pre>
1481__builtin_unreachable()
1482</pre>
1483
1484<p><b>Example of Use:</b></p>
1485
1486<pre>
1487void myabort(void) __attribute__((noreturn));
1488void myabort(void) {
1489 asm("int3");
1490 __builtin_unreachable();
1491}
1492</pre>
1493
1494<p><b>Description:</b></p>
1495
1496<p>The __builtin_unreachable() builtin has completely undefined behavior. Since
1497it has undefined behavior, it is a statement that it is never reached and the
1498optimizer can take advantage of this to produce better code. This builtin takes
1499no arguments and produces a void result.
1500</p>
1501
1502<p>Query for this feature with __has_builtin(__builtin_unreachable).</p>
1503
Chris Lattner23aa9c82011-04-09 03:57:26 +0000[diff] [blame]1504<!-- ======================================================================= -->
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]1505<h3><a name="__sync_swap">__sync_swap</a></h3>
Chris Lattner23aa9c82011-04-09 03:57:26 +0000[diff] [blame]1506<!-- ======================================================================= -->
1507
1508<p><tt>__sync_swap</tt> is used to atomically swap integers or pointers in
1509memory.
1510</p>
1511
1512<p><b>Syntax:</b></p>
1513
1514<pre>
1515<i>type</i> __sync_swap(<i>type</i> *ptr, <i>type</i> value, ...)
1516</pre>
1517
1518<p><b>Example of Use:</b></p>
1519
1520<pre>
Sean Hunt7e98b472011-06-23 01:21:01 +0000[diff] [blame]1521int old_value = __sync_swap(&amp;value, new_value);
Chris Lattner23aa9c82011-04-09 03:57:26 +0000[diff] [blame]1522</pre>
1523
1524<p><b>Description:</b></p>
1525
1526<p>The __sync_swap() builtin extends the existing __sync_*() family of atomic
1527intrinsics to allow code to atomically swap the current value with the new
1528value. More importantly, it helps developers write more efficient and correct
1529code by avoiding expensive loops around __sync_bool_compare_and_swap() or
1530relying on the platform specific implementation details of
1531__sync_lock_test_and_set(). The __sync_swap() builtin is a full barrier.
1532</p>
1533
Richard Smithfafbf062012-04-11 17:55:32 +0000[diff] [blame]1534<!-- ======================================================================= -->
1535<h3><a name="__c11_atomic">__c11_atomic builtins</a></h3>
1536<!-- ======================================================================= -->
1537
1538<p>Clang provides a set of builtins which are intended to be used to implement
1539C11's <tt>&lt;stdatomic.h&gt;</tt> header. These builtins provide the semantics
1540of the <tt>_explicit</tt> form of the corresponding C11 operation, and are named
1541with a <tt>__c11_</tt> prefix. The supported operations are:</p>
1542
1543<ul>
1544 <li><tt>__c11_atomic_init</tt></li>
1545 <li><tt>__c11_atomic_thread_fence</tt></li>
1546 <li><tt>__c11_atomic_signal_fence</tt></li>
1547 <li><tt>__c11_atomic_is_lock_free</tt></li>
1548 <li><tt>__c11_atomic_store</tt></li>
1549 <li><tt>__c11_atomic_load</tt></li>
1550 <li><tt>__c11_atomic_exchange</tt></li>
1551 <li><tt>__c11_atomic_compare_exchange_strong</tt></li>
1552 <li><tt>__c11_atomic_compare_exchange_weak</tt></li>
1553 <li><tt>__c11_atomic_fetch_add</tt></li>
1554 <li><tt>__c11_atomic_fetch_sub</tt></li>
1555 <li><tt>__c11_atomic_fetch_and</tt></li>
1556 <li><tt>__c11_atomic_fetch_or</tt></li>
1557 <li><tt>__c11_atomic_fetch_xor</tt></li>
1558</ul>
1559
Richard Smithe0d3b4c2012-05-03 18:27:39 +0000[diff] [blame]1560<!-- ======================================================================= -->
1561<h2 id="non-standard-attributes">Non-standard C++11 Attributes</h2>
1562<!-- ======================================================================= -->
1563
Richard Smith207653c2012-05-03 20:05:46 +0000[diff] [blame]1564<p>Clang supports one non-standard C++11 attribute. It resides in the
1565<tt>clang</tt> attribute namespace.</p>
Richard Smithe0d3b4c2012-05-03 18:27:39 +0000[diff] [blame]1566
1567<!-- ======================================================================= -->
1568<h3 id="clang__fallthrough">The <tt>clang::fallthrough</tt> attribute</h3>
1569<!-- ======================================================================= -->
1570
Richard Smith207653c2012-05-03 20:05:46 +0000[diff] [blame]1571<p>The <tt>clang::fallthrough</tt> attribute is used along with the
1572<tt>-Wimplicit-fallthrough</tt> argument to annotate intentional fall-through
1573between switch labels. It can only be applied to a null statement placed at a
Richard Smithe0d3b4c2012-05-03 18:27:39 +0000[diff] [blame]1574point of execution between any statement and the next switch label. It is common
1575to mark these places with a specific comment, but this attribute is meant to
1576replace comments with a more strict annotation, which can be checked by the
1577compiler. This attribute doesn't change semantics of the code and can be used
Richard Smith207653c2012-05-03 20:05:46 +0000[diff] [blame]1578wherever an intended fall-through occurs. It is designed to mimic
1579control-flow statements like <tt>break;</tt>, so it can be placed in most places
1580where <tt>break;</tt> can, but only if there are no statements on the execution
1581path between it and the next switch label.</p>
Richard Smithe0d3b4c2012-05-03 18:27:39 +0000[diff] [blame]1582<p>Here is an example:</p>
1583<pre>
1584// compile with -Wimplicit-fallthrough
1585switch (n) {
Alexander Kornienko9d8c9d32012-09-17 21:45:21 +0000[diff] [blame]1586case 22:
1587case 33: // no warning: no statements between case labels
Richard Smithe0d3b4c2012-05-03 18:27:39 +0000[diff] [blame]1588 f();
1589case 44: // warning: unannotated fall-through
1590 g();
1591 <b>[[clang::fallthrough]];</b>
1592case 55: // no warning
1593 if (x) {
1594 h();
1595 break;
1596 }
1597 else {
1598 i();
1599 <b>[[clang::fallthrough]];</b>
1600 }
1601case 66: // no warning
1602 p();
Richard Smithfff4a442012-05-03 20:10:41 +0000[diff] [blame]1603 <b>[[clang::fallthrough]];</b> // warning: fallthrough annotation does not directly precede case label
Richard Smithe0d3b4c2012-05-03 18:27:39 +0000[diff] [blame]1604 q();
1605case 77: // warning: unannotated fall-through
1606 r();
1607}
1608</pre>
Chris Lattner21190d52009-09-21 03:09:59 +0000[diff] [blame]1609
Chris Lattner1177f912009-04-09 19:58:15 +0000[diff] [blame]1610<!-- ======================================================================= -->
1611<h2 id="targetspecific">Target-Specific Extensions</h2>
1612<!-- ======================================================================= -->
1613
1614<p>Clang supports some language features conditionally on some targets.</p>
1615
1616<!-- ======================================================================= -->
1617<h3 id="x86-specific">X86/X86-64 Language Extensions</h3>
1618<!-- ======================================================================= -->
1619
1620<p>The X86 backend has these language extensions:</p>
1621
1622<!-- ======================================================================= -->
1623<h4 id="x86-gs-segment">Memory references off the GS segment</h4>
1624<!-- ======================================================================= -->
1625
1626<p>Annotating a pointer with address space #256 causes it to be code generated
Chris Lattnera021e7c2009-05-05 18:54:47 +0000[diff] [blame]1627relative to the X86 GS segment register, and address space #257 causes it to be
1628relative to the X86 FS segment. Note that this is a very very low-level
1629feature that should only be used if you know what you're doing (for example in
1630an OS kernel).</p>
Chris Lattner1177f912009-04-09 19:58:15 +0000[diff] [blame]1631
1632<p>Here is an example:</p>
1633
1634<pre>
1635#define GS_RELATIVE __attribute__((address_space(256)))
1636int foo(int GS_RELATIVE *P) {
1637 return *P;
1638}
1639</pre>
1640
1641<p>Which compiles to (on X86-32):</p>
1642
1643<pre>
1644_foo:
1645 movl 4(%esp), %eax
1646 movl %gs:(%eax), %eax
1647 ret
1648</pre>
1649
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]1650<!-- ======================================================================= -->
1651<h2 id="analyzerspecific">Static Analysis-Specific Extensions</h2>
1652<!-- ======================================================================= -->
1653
1654<p>Clang supports additional attributes that are useful for documenting program
1655invariants and rules for static analysis tools. The extensions documented here
1656are used by the <a
1657href="http://clang.llvm.org/StaticAnalysis.html">path-sensitive static analyzer
1658engine</a> that is part of Clang's Analysis library.</p>
1659
John McCall87494012011-03-18 03:51:49 +0000[diff] [blame]1660<h3 id="attr_analyzer_noreturn">The <tt>analyzer_noreturn</tt> attribute</h3>
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]1661
1662<p>Clang's static analysis engine understands the standard <tt>noreturn</tt>
Ted Kremenek4df21142009-04-10 05:04:22 +0000[diff] [blame]1663attribute. This attribute, which is typically affixed to a function prototype,
1664indicates that a call to a given function never returns. Function prototypes for
1665common functions like <tt>exit</tt> are typically annotated with this attribute,
1666as well as a variety of common assertion handlers. Users can educate the static
1667analyzer about their own custom assertion handles (thus cutting down on false
1668positives due to false paths) by marking their own &quot;panic&quot; functions
1669with this attribute.</p>
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]1670
1671<p>While useful, <tt>noreturn</tt> is not applicable in all cases. Sometimes
Nick Lewycky625b5862009-06-14 04:08:08 +0000[diff] [blame]1672there are special functions that for all intents and purposes should be
1673considered panic functions (i.e., they are only called when an internal program
1674error occurs) but may actually return so that the program can fail gracefully.
1675The <tt>analyzer_noreturn</tt> attribute allows one to annotate such functions
1676as being interpreted as &quot;no return&quot; functions by the analyzer (thus
Chris Lattner28935892009-04-10 05:54:56 +0000[diff] [blame]1677pruning bogus paths) but will not affect compilation (as in the case of
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]1678<tt>noreturn</tt>).</p>
1679
1680<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]1681same places where the <tt>noreturn</tt> attribute can be placed. It is commonly
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]1682placed at the end of function prototypes:</p>
1683
1684<pre>
1685 void foo() <b>__attribute__((analyzer_noreturn))</b>;
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]1686</pre>
1687
John McCall87494012011-03-18 03:51:49 +0000[diff] [blame]1688<p>Query for this feature with
1689<tt>__has_attribute(analyzer_noreturn)</tt>.</p>
Chris Lattner148772a2009-06-13 07:13:28 +0000[diff] [blame]1690
John McCall87494012011-03-18 03:51:49 +0000[diff] [blame]1691<h3 id="attr_method_family">The <tt>objc_method_family</tt> attribute</h3>
1692
1693<p>Many methods in Objective-C have conventional meanings determined
1694by their selectors. For the purposes of static analysis, it is
1695sometimes useful to be able to mark a method as having a particular
1696conventional meaning despite not having the right selector, or as not
1697having the conventional meaning that its selector would suggest.
1698For these use cases, we provide an attribute to specifically describe
1699the <q>method family</q> that a method belongs to.</p>
1700
1701<p><b>Usage</b>: <tt>__attribute__((objc_method_family(X)))</tt>,
1702where <tt>X</tt> is one of <tt>none</tt>, <tt>alloc</tt>, <tt>copy</tt>,
1703<tt>init</tt>, <tt>mutableCopy</tt>, or <tt>new</tt>. This attribute
1704can only be placed at the end of a method declaration:</p>
1705
1706<pre>
1707 - (NSString*) initMyStringValue <b>__attribute__((objc_method_family(none)))</b>;
1708</pre>
1709
1710<p>Users who do not wish to change the conventional meaning of a
1711method, and who merely want to document its non-standard retain and
1712release semantics, should use the
1713<a href="#attr_retain_release">retaining behavior attributes</a>
1714described below.</p>
1715
1716<p>Query for this feature with
1717<tt>__has_attribute(objc_method_family)</tt>.</p>
1718
1719<h3 id="attr_retain_release">Objective-C retaining behavior attributes</h3>
John McCall630b7ae2011-01-25 04:26:21 +0000[diff] [blame]1720
1721<p>In Objective-C, functions and methods are generally assumed to take
1722and return objects with +0 retain counts, with some exceptions for
1723special methods like <tt>+alloc</tt> and <tt>init</tt>. However,
1724there are exceptions, and so Clang provides attributes to allow these
1725exceptions to be documented, which helps the analyzer find leaks (and
John McCall87494012011-03-18 03:51:49 +0000[diff] [blame]1726ignore non-leaks). Some exceptions may be better described using
1727the <a href="#attr_method_family"><tt>objc_method_family</tt></a>
1728attribute instead.</p>
John McCall630b7ae2011-01-25 04:26:21 +0000[diff] [blame]1729
1730<p><b>Usage</b>: The <tt>ns_returns_retained</tt>, <tt>ns_returns_not_retained</tt>,
1731<tt>ns_returns_autoreleased</tt>, <tt>cf_returns_retained</tt>,
1732and <tt>cf_returns_not_retained</tt> attributes can be placed on
1733methods and functions that return Objective-C or CoreFoundation
1734objects. They are commonly placed at the end of a function prototype
1735or method declaration:</p>
1736
1737<pre>
1738 id foo() <b>__attribute__((ns_returns_retained))</b>;
1739
1740 - (NSString*) bar: (int) x <b>__attribute__((ns_returns_retained))</b>;
1741</pre>
1742
1743<p>The <tt>*_returns_retained</tt> attributes specify that the
1744returned object has a +1 retain count.
1745The <tt>*_returns_not_retained</tt> attributes specify that the return
1746object has a +0 retain count, even if the normal convention for its
1747selector would be +1. <tt>ns_returns_autoreleased</tt> specifies that the
1748returned object is +0, but is guaranteed to live at least as long as the
1749next flush of an autorelease pool.</p>
1750
1751<p><b>Usage</b>: The <tt>ns_consumed</tt> and <tt>cf_consumed</tt>
1752attributes can be placed on an parameter declaration; they specify
1753that the argument is expected to have a +1 retain count, which will be
1754balanced in some way by the function or method.
1755The <tt>ns_consumes_self</tt> attribute can only be placed on an
1756Objective-C method; it specifies that the method expects
1757its <tt>self</tt> parameter to have a +1 retain count, which it will
1758balance in some way.</p>
1759
1760<pre>
1761 void <b>foo(__attribute__((ns_consumed))</b> NSString *string);
1762
1763 - (void) bar <b>__attribute__((ns_consumes_self))</b>;
1764 - (void) baz: (id) <b>__attribute__((ns_consumed))</b> x;
1765</pre>
Ted Kremeneked869312009-04-10 05:03:33 +0000[diff] [blame]1766
John McCall87494012011-03-18 03:51:49 +0000[diff] [blame]1767<p>Query for these features with <tt>__has_attribute(ns_consumed)</tt>,
1768<tt>__has_attribute(ns_returns_retained)</tt>, etc.</p>
1769
Kostya Serebryanyce98c9b2011-11-28 20:51:02 +0000[diff] [blame]1770<!-- ======================================================================= -->
1771<h2 id="dynamicanalyzerspecific">Dynamic Analysis-Specific Extensions</h2>
1772<!-- ======================================================================= -->
1773<h3 id="address_sanitizer">AddressSanitizer</h3>
1774<p> Use <code>__has_feature(address_sanitizer)</code>
1775to check if the code is being built with <a
1776 href="AddressSanitizer.html">AddressSanitizer</a>.
1777</p>
Kostya Serebryany71efba02012-01-24 19:25:38 +0000[diff] [blame]1778<p>Use <tt>__attribute__((no_address_safety_analysis))</tt> on a function
1779declaration to specify that address safety instrumentation (e.g.
1780AddressSanitizer) should not be applied to that function.
1781</p>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]1782
1783<!-- ======================================================================= -->
Caitlin Sadowski73cbbc82011-07-28 18:38:36 +0000[diff] [blame]1784<h2 id="threadsafety">Thread-Safety Annotation Checking</h2>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]1785<!-- ======================================================================= -->
1786
1787<p>Clang supports additional attributes for checking basic locking policies in
1788multithreaded programs.
1789Clang currently parses the following list of attributes, although
1790<b>the implementation for these annotations is currently in development.</b>
1791For more details, see the
1792<a href="http://gcc.gnu.org/wiki/ThreadSafetyAnnotation">GCC implementation</a>.
1793</p>
1794
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1795<h4 id="ts_noanal">no_thread_safety_analysis</h4>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]1796
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1797<p>Use <tt>__attribute__((no_thread_safety_analysis))</tt> on a function
1798declaration to specify that the thread safety analysis should not be run on that
1799function. This attribute provides an escape hatch (e.g. for situations when it
1800is difficult to annotate the locking policy). </p>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]1801
1802<h4 id="ts_lockable">lockable</h4>
1803
1804<p>Use <tt>__attribute__((lockable))</tt> on a class definition to specify
1805that it has a lockable type (e.g. a Mutex class). This annotation is primarily
1806used to check consistency.</p>
1807
1808<h4 id="ts_scopedlockable">scoped_lockable</h4>
1809
1810<p>Use <tt>__attribute__((scoped_lockable))</tt> on a class definition to
1811specify that it has a "scoped" lockable type. Objects of this type will acquire
1812the lock upon construction and release it upon going out of scope.
1813 This annotation is primarily used to check
1814consistency.</p>
1815
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1816<h4 id="ts_guardedvar">guarded_var</h4>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]1817
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1818<p>Use <tt>__attribute__((guarded_var))</tt> on a variable declaration to
1819specify that the variable must be accessed while holding some lock.</p>
1820
1821<h4 id="ts_ptguardedvar">pt_guarded_var</h4>
1822
1823<p>Use <tt>__attribute__((pt_guarded_var))</tt> on a pointer declaration to
1824specify that the pointer must be dereferenced while holding some lock.</p>
1825
1826<h4 id="ts_guardedby">guarded_by(l)</h4>
1827
1828<p>Use <tt>__attribute__((guarded_by(l)))</tt> on a variable declaration to
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1829specify that the variable must be accessed while holding lock <tt>l</tt>.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1830
1831<h4 id="ts_ptguardedby">pt_guarded_by(l)</h4>
1832
1833<p>Use <tt>__attribute__((pt_guarded_by(l)))</tt> on a pointer declaration to
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1834specify that the pointer must be dereferenced while holding lock <tt>l</tt>.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1835
1836<h4 id="ts_acquiredbefore">acquired_before(...)</h4>
1837
1838<p>Use <tt>__attribute__((acquired_before(...)))</tt> on a declaration
1839of a lockable variable to specify that the lock must be acquired before all
1840attribute arguments. Arguments must be lockable type, and there must be at
1841least one argument.</p>
1842
1843<h4 id="ts_acquiredafter">acquired_after(...)</h4>
1844
1845<p>Use <tt>__attribute__((acquired_after(...)))</tt> on a declaration
1846of a lockable variable to specify that the lock must be acquired after all
1847attribute arguments. Arguments must be lockable type, and there must be at
1848least one argument.</p>
1849
1850<h4 id="ts_elf">exclusive_lock_function(...)</h4>
1851
1852<p>Use <tt>__attribute__((exclusive_lock_function(...)))</tt> on a function
1853declaration to specify that the function acquires all listed locks
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1854exclusively. This attribute takes zero or more arguments: either of lockable
1855type or integers indexing into function parameters of lockable type. If no
1856arguments are given, the acquired lock is implicitly <tt>this</tt> of the
1857enclosing object.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1858
1859<h4 id="ts_slf">shared_lock_function(...)</h4>
1860
1861<p>Use <tt>__attribute__((shared_lock_function(...)))</tt> on a function
1862declaration to specify that the function acquires all listed locks, although
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1863 the locks may be shared (e.g. read locks). This attribute takes zero or more
1864arguments: either of lockable type or integers indexing into function
1865parameters of lockable type. If no arguments are given, the acquired lock is
1866implicitly <tt>this</tt> of the enclosing object.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1867
1868<h4 id="ts_etf">exclusive_trylock_function(...)</h4>
1869
1870<p>Use <tt>__attribute__((exclusive_lock_function(...)))</tt> on a function
1871declaration to specify that the function will try (without blocking) to acquire
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1872all listed locks exclusively. This attribute takes one or more arguments. The
1873first argument is an integer or boolean value specifying the return value of a
1874successful lock acquisition. The remaining arugments are either of lockable type
1875or integers indexing into function parameters of lockable type. If only one
1876argument is given, the acquired lock is implicitly <tt>this</tt> of the
1877enclosing object.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1878
1879<h4 id="ts_stf">shared_trylock_function(...)</h4>
1880
1881<p>Use <tt>__attribute__((shared_lock_function(...)))</tt> on a function
1882declaration to specify that the function will try (without blocking) to acquire
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1883all listed locks, although the locks may be shared (e.g. read locks). This
1884attribute takes one or more arguments. The first argument is an integer or
1885boolean value specifying the return value of a successful lock acquisition. The
1886remaining arugments are either of lockable type or integers indexing into
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1887function parameters of lockable type. If only one argument is given, the
1888acquired lock is implicitly <tt>this</tt> of the enclosing object.</p>
1889
1890<h4 id="ts_uf">unlock_function(...)</h4>
1891
1892<p>Use <tt>__attribute__((unlock_function(...)))</tt> on a function
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1893declaration to specify that the function release all listed locks. This
1894attribute takes zero or more arguments: either of lockable type or integers
1895indexing into function parameters of lockable type. If no arguments are given,
1896the acquired lock is implicitly <tt>this</tt> of the enclosing object.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1897
1898<h4 id="ts_lr">lock_returned(l)</h4>
1899
1900<p>Use <tt>__attribute__((lock_returned(l)))</tt> on a function
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1901declaration to specify that the function returns lock <tt>l</tt> (<tt>l</tt>
1902must be of lockable type). This annotation is used to aid in resolving lock
1903expressions.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1904
1905<h4 id="ts_le">locks_excluded(...)</h4>
1906
1907<p>Use <tt>__attribute__((locks_excluded(...)))</tt> on a function declaration
Caitlin Sadowskib51e0312011-08-09 17:59:31 +0000[diff] [blame]1908to specify that the function must not be called with the listed locks. Arguments
1909must be lockable type, and there must be at least one argument.</p>
Caitlin Sadowskidb33e142011-07-28 20:12:35 +0000[diff] [blame]1910
1911<h4 id="ts_elr">exclusive_locks_required(...)</h4>
1912
1913<p>Use <tt>__attribute__((exclusive_locks_required(...)))</tt> on a function
1914declaration to specify that the function must be called while holding the listed
1915exclusive locks. Arguments must be lockable type, and there must be at
1916least one argument.</p>
1917
1918<h4 id="ts_slr">shared_locks_required(...)</h4>
1919
1920<p>Use <tt>__attribute__((shared_locks_required(...)))</tt> on a function
1921declaration to specify that the function must be called while holding the listed
1922shared locks. Arguments must be lockable type, and there must be at
1923least one argument.</p>
Caitlin Sadowskifdde9e72011-07-28 17:21:07 +0000[diff] [blame]1924
Dmitri Gribenko0d5a0692012-08-17 00:08:38 +0000[diff] [blame]1925<!-- ======================================================================= -->
1926<h2 id="type_safety">Type Safety Checking</h2>
1927<!-- ======================================================================= -->
1928
1929<p>Clang supports additional attributes to enable checking type safety
1930properties that can't be enforced by C type system. Usecases include:</p>
1931<ul>
1932<li>MPI library implementations, where these attributes enable checking that
1933 buffer type matches the passed <tt>MPI_Datatype</tt>;</li>
1934<li>for HDF5 library there is a similar usecase as MPI;</li>
1935<li>checking types of variadic functions' arguments for functions like
1936 <tt>fcntl()</tt> and <tt>ioctl()</tt>.</li>
1937</ul>
1938
1939<p>You can detect support for these attributes with __has_attribute(). For
1940example:</p>
1941
1942<blockquote>
1943<pre>
1944#if defined(__has_attribute)
1945# if __has_attribute(argument_with_type_tag) &amp;&amp; \
1946 __has_attribute(pointer_with_type_tag) &amp;&amp; \
1947 __has_attribute(type_tag_for_datatype)
1948# define ATTR_MPI_PWT(buffer_idx, type_idx) __attribute__((pointer_with_type_tag(mpi,buffer_idx,type_idx)))
1949/* ... other macros ... */
1950# endif
1951#endif
1952
1953#if !defined(ATTR_MPI_PWT)
1954#define ATTR_MPI_PWT(buffer_idx, type_idx)
1955#endif
1956
1957int MPI_Send(void *buf, int count, MPI_Datatype datatype /*, other args omitted */)
1958 ATTR_MPI_PWT(1,3);
1959</pre>
1960</blockquote>
1961
1962<h3 id="argument_with_type_tag"><tt>argument_with_type_tag(...)</tt></h3>
1963
1964<p>Use <tt>__attribute__((argument_with_type_tag(arg_kind, arg_idx,
1965type_tag_idx)))</tt> on a function declaration to specify that the function
1966accepts a type tag that determines the type of some other argument.
1967<tt>arg_kind</tt> is an identifier that should be used when annotating all
1968applicable type tags.</p>
1969
1970<p>This attribute is primarily useful for checking arguments of variadic
1971functions (<tt>pointer_with_type_tag</tt> can be used in most of non-variadic
1972cases).</p>
1973
1974<p>For example:</p>
1975<blockquote>
1976<pre>
1977int fcntl(int fd, int cmd, ...)
1978 __attribute__(( argument_with_type_tag(fcntl,3,2) ));
1979</pre>
1980</blockquote>
1981
1982<h3 id="pointer_with_type_tag"><tt>pointer_with_type_tag(...)</tt></h3>
1983
1984<p>Use <tt>__attribute__((pointer_with_type_tag(ptr_kind, ptr_idx,
1985type_tag_idx)))</tt> on a function declaration to specify that the
Dmitri Gribenkodfccbd92012-08-31 03:19:55 +0000[diff] [blame]1986function accepts a type tag that determines the pointee type of some other
Dmitri Gribenkoa8ba40e2012-08-31 02:39:31 +0000[diff] [blame]1987pointer argument.</p>
Dmitri Gribenko0d5a0692012-08-17 00:08:38 +0000[diff] [blame]1988
1989<p>For example:</p>
1990<blockquote>
1991<pre>
1992int MPI_Send(void *buf, int count, MPI_Datatype datatype /*, other args omitted */)
1993 __attribute__(( pointer_with_type_tag(mpi,1,3) ));
1994</pre>
1995</blockquote>
1996
1997<h3 id="type_tag_for_datatype"><tt>type_tag_for_datatype(...)</tt></h3>
1998
1999<p>Clang supports annotating type tags of two forms.</p>
2000
2001<ul>
2002<li><b>Type tag that is an expression containing a reference to some declared
2003identifier.</b> Use <tt>__attribute__((type_tag_for_datatype(kind, type)))</tt>
2004on a declaration with that identifier:
2005
2006<blockquote>
2007<pre>
2008extern struct mpi_datatype mpi_datatype_int
2009 __attribute__(( type_tag_for_datatype(mpi,int) ));
2010#define MPI_INT ((MPI_Datatype) &amp;mpi_datatype_int)
2011</pre>
2012</blockquote></li>
2013
2014<li><b>Type tag that is an integral literal.</b> Introduce a <tt>static
2015const</tt> variable with a corresponding initializer value and attach
2016<tt>__attribute__((type_tag_for_datatype(kind, type)))</tt> on that
2017declaration, for example:
2018
2019<blockquote>
2020<pre>
2021#define MPI_INT ((MPI_Datatype) 42)
2022static const MPI_Datatype mpi_datatype_int
2023 __attribute__(( type_tag_for_datatype(mpi,int) )) = 42
2024</pre>
2025</blockquote></li>
2026</ul>
2027
2028<p>The attribute also accepts an optional third argument that determines how
2029the expression is compared to the type tag. There are two supported flags:</p>
2030
2031<ul><li><tt>layout_compatible</tt> will cause types to be compared according to
2032layout-compatibility rules (C++11 [class.mem] p&nbsp;17, 18). This is
2033implemented to support annotating types like <tt>MPI_DOUBLE_INT</tt>.
2034
2035<p>For example:</p>
2036<blockquote>
2037<pre>
2038/* In mpi.h */
2039struct internal_mpi_double_int { double d; int i; };
2040extern struct mpi_datatype mpi_datatype_double_int
2041 __attribute__(( type_tag_for_datatype(mpi, struct internal_mpi_double_int,
2042 layout_compatible) ));
2043
2044#define MPI_DOUBLE_INT ((MPI_Datatype) &amp;mpi_datatype_double_int)
2045
2046/* In user code */
2047struct my_pair { double a; int b; };
2048struct my_pair *buffer;
2049MPI_Send(buffer, 1, MPI_DOUBLE_INT /*, ... */); // no warning
2050
2051struct my_int_pair { int a; int b; }
2052struct my_int_pair *buffer2;
2053MPI_Send(buffer2, 1, MPI_DOUBLE_INT /*, ... */); // warning: actual buffer element
2054 // type 'struct my_int_pair'
2055 // doesn't match specified MPI_Datatype
2056</pre>
2057</blockquote>
2058</li>
2059
2060<li><tt>must_be_null</tt> specifies that the expression should be a null
2061pointer constant, for example:
2062
2063<blockquote>
2064<pre>
2065/* In mpi.h */
2066extern struct mpi_datatype mpi_datatype_null
2067 __attribute__(( type_tag_for_datatype(mpi, void, must_be_null) ));
2068
2069#define MPI_DATATYPE_NULL ((MPI_Datatype) &amp;mpi_datatype_null)
2070
2071/* In user code */
2072MPI_Send(buffer, 1, MPI_DATATYPE_NULL /*, ... */); // warning: MPI_DATATYPE_NULL
2073 // was specified but buffer
2074 // is not a null pointer
2075</pre>
2076</blockquote>
2077</li>
2078</ul>
2079
Chris Lattner5ce933f2009-02-09 08:46:11 +0000[diff] [blame]2080</div>
2081</body>
2082</html>