/usr/share/doc/clang-3.5-doc/html/LanguageExtensions.html is in clang-3.5-doc 1:3.5.2-3ubuntu1.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 | <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<title>Clang Language Extensions — Clang 3.5 documentation</title>
<link rel="stylesheet" href="_static/haiku.css" type="text/css" />
<link rel="stylesheet" href="_static/pygments.css" type="text/css" />
<script type="text/javascript">
var DOCUMENTATION_OPTIONS = {
URL_ROOT: './',
VERSION: '3.5',
COLLAPSE_INDEX: false,
FILE_SUFFIX: '.html',
HAS_SOURCE: true
};
</script>
<script type="text/javascript" src="_static/jquery.js"></script>
<script type="text/javascript" src="_static/underscore.js"></script>
<script type="text/javascript" src="_static/doctools.js"></script>
<script type="text/javascript" src="https://cdn.mathjax.org/mathjax/latest/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
<link rel="top" title="Clang 3.5 documentation" href="index.html" />
<link rel="next" title="Objective-C Literals" href="ObjectiveCLiterals.html" />
<link rel="prev" title="Clang Compiler User’s Manual" href="UsersManual.html" />
</head>
<body role="document">
<div class="header" role="banner"><h1 class="heading"><a href="index.html">
<span>Clang 3.5 documentation</span></a></h1>
<h2 class="heading"><span>Clang Language Extensions</span></h2>
</div>
<div class="topnav" role="navigation" aria-label="top navigation">
<p>
«  <a href="UsersManual.html">Clang Compiler User’s Manual</a>
  ::  
<a class="uplink" href="index.html">Contents</a>
  ::  
<a href="ObjectiveCLiterals.html">Objective-C Literals</a>  »
</p>
</div>
<div class="content">
<div class="section" id="clang-language-extensions">
<h1>Clang Language Extensions<a class="headerlink" href="#clang-language-extensions" title="Permalink to this headline">¶</a></h1>
<div class="contents local topic" id="contents">
<ul class="simple">
<li><a class="reference internal" href="#introduction" id="id1">Introduction</a></li>
<li><a class="reference internal" href="#feature-checking-macros" id="id2">Feature Checking Macros</a></li>
<li><a class="reference internal" href="#include-file-checking-macros" id="id3">Include File Checking Macros</a></li>
<li><a class="reference internal" href="#builtin-macros" id="id4">Builtin Macros</a></li>
<li><a class="reference internal" href="#vectors-and-extended-vectors" id="id5">Vectors and Extended Vectors</a></li>
<li><a class="reference internal" href="#messages-on-deprecated-and-unavailable-attributes" id="id6">Messages on <code class="docutils literal"><span class="pre">deprecated</span></code> and <code class="docutils literal"><span class="pre">unavailable</span></code> Attributes</a></li>
<li><a class="reference internal" href="#attributes-on-enumerators" id="id7">Attributes on Enumerators</a></li>
<li><a class="reference internal" href="#user-specified-system-frameworks" id="id8">‘User-Specified’ System Frameworks</a></li>
<li><a class="reference internal" href="#checks-for-standard-language-features" id="id9">Checks for Standard Language Features</a></li>
<li><a class="reference internal" href="#checks-for-type-trait-primitives" id="id10">Checks for Type Trait Primitives</a></li>
<li><a class="reference internal" href="#blocks" id="id11">Blocks</a></li>
<li><a class="reference internal" href="#objective-c-features" id="id12">Objective-C Features</a></li>
<li><a class="reference internal" href="#initializer-lists-for-complex-numbers-in-c" id="id13">Initializer lists for complex numbers in C</a></li>
<li><a class="reference internal" href="#builtin-functions" id="id14">Builtin Functions</a></li>
<li><a class="reference internal" href="#non-standard-c-11-attributes" id="id15">Non-standard C++11 Attributes</a></li>
<li><a class="reference internal" href="#target-specific-extensions" id="id16">Target-Specific Extensions</a></li>
<li><a class="reference internal" href="#extensions-for-static-analysis" id="id17">Extensions for Static Analysis</a></li>
<li><a class="reference internal" href="#extensions-for-dynamic-analysis" id="id18">Extensions for Dynamic Analysis</a></li>
<li><a class="reference internal" href="#extensions-for-selectively-disabling-optimization" id="id19">Extensions for selectively disabling optimization</a></li>
<li><a class="reference internal" href="#extensions-for-loop-hint-optimizations" id="id20">Extensions for loop hint optimizations</a></li>
</ul>
</div>
<div class="toctree-wrapper compound">
</div>
<div class="section" id="introduction">
<h2><a class="toc-backref" href="#id1">Introduction</a><a class="headerlink" href="#introduction" title="Permalink to this headline">¶</a></h2>
<p>This document describes the language extensions provided by Clang. In addition
to the language extensions listed here, Clang aims to support a broad range of
GCC extensions. Please see the <a class="reference external" href="http://gcc.gnu.org/onlinedocs/gcc/C-Extensions.html">GCC manual</a> for more information on
these extensions.</p>
</div>
<div class="section" id="feature-checking-macros">
<span id="langext-feature-check"></span><h2><a class="toc-backref" href="#id2">Feature Checking Macros</a><a class="headerlink" href="#feature-checking-macros" title="Permalink to this headline">¶</a></h2>
<p>Language extensions can be very useful, but only if you know you can depend on
them. In order to allow fine-grain features checks, we support three builtin
function-like macros. This allows you to directly test for a feature in your
code without having to resort to something like autoconf or fragile “compiler
version checks”.</p>
<div class="section" id="has-builtin">
<h3><code class="docutils literal"><span class="pre">__has_builtin</span></code><a class="headerlink" href="#has-builtin" title="Permalink to this headline">¶</a></h3>
<p>This function-like macro takes a single identifier argument that is the name of
a builtin function. It evaluates to 1 if the builtin is supported or 0 if not.
It can be used like this:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#ifndef __has_builtin </span><span class="c1">// Optional of course.</span>
<span class="cp">#define __has_builtin(x) 0 </span><span class="c1">// Compatibility with non-clang compilers.</span>
<span class="cp">#endif</span>
<span class="p">...</span>
<span class="cp">#if __has_builtin(__builtin_trap)</span>
<span class="n">__builtin_trap</span><span class="p">();</span>
<span class="cp">#else</span>
<span class="n">abort</span><span class="p">();</span>
<span class="cp">#endif</span>
<span class="p">...</span>
</pre></div>
</div>
</div>
<div class="section" id="has-feature-and-has-extension">
<span id="langext-has-feature-has-extension"></span><h3><code class="docutils literal"><span class="pre">__has_feature</span></code> and <code class="docutils literal"><span class="pre">__has_extension</span></code><a class="headerlink" href="#has-feature-and-has-extension" title="Permalink to this headline">¶</a></h3>
<p>These function-like macros take a single identifier argument that is the name
of a feature. <code class="docutils literal"><span class="pre">__has_feature</span></code> evaluates to 1 if the feature is both
supported by Clang and standardized in the current language standard or 0 if
not (but see <a class="reference internal" href="#langext-has-feature-back-compat"><span>below</span></a>), while
<code class="docutils literal"><span class="pre">__has_extension</span></code> evaluates to 1 if the feature is supported by Clang in the
current language (either as a language extension or a standard language
feature) or 0 if not. They can be used like this:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#ifndef __has_feature </span><span class="c1">// Optional of course.</span>
<span class="cp">#define __has_feature(x) 0 </span><span class="c1">// Compatibility with non-clang compilers.</span>
<span class="cp">#endif</span>
<span class="cp">#ifndef __has_extension</span>
<span class="cp">#define __has_extension __has_feature </span><span class="c1">// Compatibility with pre-3.0 compilers.</span>
<span class="cp">#endif</span>
<span class="p">...</span>
<span class="cp">#if __has_feature(cxx_rvalue_references)</span>
<span class="c1">// This code will only be compiled with the -std=c++11 and -std=gnu++11</span>
<span class="c1">// options, because rvalue references are only standardized in C++11.</span>
<span class="cp">#endif</span>
<span class="cp">#if __has_extension(cxx_rvalue_references)</span>
<span class="c1">// This code will be compiled with the -std=c++11, -std=gnu++11, -std=c++98</span>
<span class="c1">// and -std=gnu++98 options, because rvalue references are supported as a</span>
<span class="c1">// language extension in C++98.</span>
<span class="cp">#endif</span>
</pre></div>
</div>
<p id="langext-has-feature-back-compat">For backward compatibility, <code class="docutils literal"><span class="pre">__has_feature</span></code> can also be used to test
for support for non-standardized features, i.e. features not prefixed <code class="docutils literal"><span class="pre">c_</span></code>,
<code class="docutils literal"><span class="pre">cxx_</span></code> or <code class="docutils literal"><span class="pre">objc_</span></code>.</p>
<p>Another use of <code class="docutils literal"><span class="pre">__has_feature</span></code> is to check for compiler features not related
to the language standard, such as e.g. <a class="reference internal" href="AddressSanitizer.html"><em>AddressSanitizer</em></a>.</p>
<p>If the <code class="docutils literal"><span class="pre">-pedantic-errors</span></code> option is given, <code class="docutils literal"><span class="pre">__has_extension</span></code> is equivalent
to <code class="docutils literal"><span class="pre">__has_feature</span></code>.</p>
<p>The feature tag is described along with the language feature below.</p>
<p>The feature name or extension name can also be specified with a preceding and
following <code class="docutils literal"><span class="pre">__</span></code> (double underscore) to avoid interference from a macro with
the same name. For instance, <code class="docutils literal"><span class="pre">__cxx_rvalue_references__</span></code> can be used instead
of <code class="docutils literal"><span class="pre">cxx_rvalue_references</span></code>.</p>
</div>
<div class="section" id="has-attribute">
<h3><code class="docutils literal"><span class="pre">__has_attribute</span></code><a class="headerlink" href="#has-attribute" title="Permalink to this headline">¶</a></h3>
<p>This function-like macro takes a single identifier argument that is the name of
an attribute. It evaluates to 1 if the attribute is supported by the current
compilation target, or 0 if not. It can be used like this:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#ifndef __has_attribute </span><span class="c1">// Optional of course.</span>
<span class="cp">#define __has_attribute(x) 0 </span><span class="c1">// Compatibility with non-clang compilers.</span>
<span class="cp">#endif</span>
<span class="p">...</span>
<span class="cp">#if __has_attribute(always_inline)</span>
<span class="cp">#define ALWAYS_INLINE __attribute__((always_inline))</span>
<span class="cp">#else</span>
<span class="cp">#define ALWAYS_INLINE</span>
<span class="cp">#endif</span>
<span class="p">...</span>
</pre></div>
</div>
<p>The attribute name can also be specified with a preceding and following <code class="docutils literal"><span class="pre">__</span></code>
(double underscore) to avoid interference from a macro with the same name. For
instance, <code class="docutils literal"><span class="pre">__always_inline__</span></code> can be used instead of <code class="docutils literal"><span class="pre">always_inline</span></code>.</p>
</div>
<div class="section" id="is-identifier">
<h3><code class="docutils literal"><span class="pre">__is_identifier</span></code><a class="headerlink" href="#is-identifier" title="Permalink to this headline">¶</a></h3>
<p>This function-like macro takes a single identifier argument that might be either
a reserved word or a regular identifier. It evaluates to 1 if the argument is just
a regular identifier and not a reserved word, in the sense that it can then be
used as the name of a user-defined function or variable. Otherwise it evaluates
to 0. It can be used like this:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="p">...</span>
<span class="cp">#ifdef __is_identifier </span><span class="c1">// Compatibility with non-clang compilers.</span>
<span class="cp">#if __is_identifier(__wchar_t)</span>
<span class="k">typedef</span> <span class="kt">wchar_t</span> <span class="kr">__wchar_t</span><span class="p">;</span>
<span class="cp">#endif</span>
<span class="cp">#endif</span>
<span class="kr">__wchar_t</span> <span class="n">WideCharacter</span><span class="p">;</span>
<span class="p">...</span>
</pre></div>
</div>
</div>
</div>
<div class="section" id="include-file-checking-macros">
<h2><a class="toc-backref" href="#id3">Include File Checking Macros</a><a class="headerlink" href="#include-file-checking-macros" title="Permalink to this headline">¶</a></h2>
<p>Not all developments systems have the same include files. The
<a class="reference internal" href="#langext-has-include"><span>__has_include</span></a> and <a class="reference internal" href="#langext-has-include-next"><span>__has_include_next</span></a> macros allow
you to check for the existence of an include file before doing a possibly
failing <code class="docutils literal"><span class="pre">#include</span></code> directive. Include file checking macros must be used
as expressions in <code class="docutils literal"><span class="pre">#if</span></code> or <code class="docutils literal"><span class="pre">#elif</span></code> preprocessing directives.</p>
<div class="section" id="has-include">
<span id="langext-has-include"></span><h3><code class="docutils literal"><span class="pre">__has_include</span></code><a class="headerlink" href="#has-include" title="Permalink to this headline">¶</a></h3>
<p>This function-like macro takes a single file name string argument that is the
name of an include file. It evaluates to 1 if the file can be found using the
include paths, or 0 otherwise:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="c1">// Note the two possible file name string formats.</span>
<span class="cp">#if __has_include("myinclude.h") && __has_include(<stdint.h>)</span>
<span class="cp"># include "myinclude.h"</span>
<span class="cp">#endif</span>
</pre></div>
</div>
<p>To test for this feature, use <code class="docutils literal"><span class="pre">#if</span> <span class="pre">defined(__has_include)</span></code>:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="c1">// To avoid problem with non-clang compilers not having this macro.</span>
<span class="cp">#if defined(__has_include)</span>
<span class="cp">#if __has_include("myinclude.h")</span>
<span class="cp"># include "myinclude.h"</span>
<span class="cp">#endif</span>
<span class="cp">#endif</span>
</pre></div>
</div>
</div>
<div class="section" id="has-include-next">
<span id="langext-has-include-next"></span><h3><code class="docutils literal"><span class="pre">__has_include_next</span></code><a class="headerlink" href="#has-include-next" title="Permalink to this headline">¶</a></h3>
<p>This function-like macro takes a single file name string argument that is the
name of an include file. It is like <code class="docutils literal"><span class="pre">__has_include</span></code> except that it looks for
the second instance of the given file found in the include paths. It evaluates
to 1 if the second instance of the file can be found using the include paths,
or 0 otherwise:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="c1">// Note the two possible file name string formats.</span>
<span class="cp">#if __has_include_next("myinclude.h") && __has_include_next(<stdint.h>)</span>
<span class="cp"># include_next "myinclude.h"</span>
<span class="cp">#endif</span>
<span class="c1">// To avoid problem with non-clang compilers not having this macro.</span>
<span class="cp">#if defined(__has_include_next)</span>
<span class="cp">#if __has_include_next("myinclude.h")</span>
<span class="cp"># include_next "myinclude.h"</span>
<span class="cp">#endif</span>
<span class="cp">#endif</span>
</pre></div>
</div>
<p>Note that <code class="docutils literal"><span class="pre">__has_include_next</span></code>, like the GNU extension <code class="docutils literal"><span class="pre">#include_next</span></code>
directive, is intended for use in headers only, and will issue a warning if
used in the top-level compilation file. A warning will also be issued if an
absolute path is used in the file argument.</p>
</div>
<div class="section" id="has-warning">
<h3><code class="docutils literal"><span class="pre">__has_warning</span></code><a class="headerlink" href="#has-warning" title="Permalink to this headline">¶</a></h3>
<p>This function-like macro takes a string literal that represents a command line
option for a warning and returns true if that is a valid warning option.</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#if __has_warning("-Wformat")</span>
<span class="p">...</span>
<span class="cp">#endif</span>
</pre></div>
</div>
</div>
</div>
<div class="section" id="builtin-macros">
<h2><a class="toc-backref" href="#id4">Builtin Macros</a><a class="headerlink" href="#builtin-macros" title="Permalink to this headline">¶</a></h2>
<dl class="docutils">
<dt><code class="docutils literal"><span class="pre">__BASE_FILE__</span></code></dt>
<dd>Defined to a string that contains the name of the main input file passed to
Clang.</dd>
<dt><code class="docutils literal"><span class="pre">__COUNTER__</span></code></dt>
<dd>Defined to an integer value that starts at zero and is incremented each time
the <code class="docutils literal"><span class="pre">__COUNTER__</span></code> macro is expanded.</dd>
<dt><code class="docutils literal"><span class="pre">__INCLUDE_LEVEL__</span></code></dt>
<dd>Defined to an integral value that is the include depth of the file currently
being translated. For the main file, this value is zero.</dd>
<dt><code class="docutils literal"><span class="pre">__TIMESTAMP__</span></code></dt>
<dd>Defined to the date and time of the last modification of the current source
file.</dd>
<dt><code class="docutils literal"><span class="pre">__clang__</span></code></dt>
<dd>Defined when compiling with Clang</dd>
<dt><code class="docutils literal"><span class="pre">__clang_major__</span></code></dt>
<dd>Defined to the major marketing version number of Clang (e.g., the 2 in
2.0.1). Note that marketing version numbers should not be used to check for
language features, as different vendors use different numbering schemes.
Instead, use the <a class="reference internal" href="#langext-feature-check"><span>Feature Checking Macros</span></a>.</dd>
<dt><code class="docutils literal"><span class="pre">__clang_minor__</span></code></dt>
<dd>Defined to the minor version number of Clang (e.g., the 0 in 2.0.1). Note
that marketing version numbers should not be used to check for language
features, as different vendors use different numbering schemes. Instead, use
the <a class="reference internal" href="#langext-feature-check"><span>Feature Checking Macros</span></a>.</dd>
<dt><code class="docutils literal"><span class="pre">__clang_patchlevel__</span></code></dt>
<dd>Defined to the marketing patch level of Clang (e.g., the 1 in 2.0.1).</dd>
<dt><code class="docutils literal"><span class="pre">__clang_version__</span></code></dt>
<dd>Defined to a string that captures the Clang marketing version, including the
Subversion tag or revision number, e.g., “<code class="docutils literal"><span class="pre">1.5</span> <span class="pre">(trunk</span> <span class="pre">102332)</span></code>”.</dd>
</dl>
</div>
<div class="section" id="vectors-and-extended-vectors">
<span id="langext-vectors"></span><h2><a class="toc-backref" href="#id5">Vectors and Extended Vectors</a><a class="headerlink" href="#vectors-and-extended-vectors" title="Permalink to this headline">¶</a></h2>
<p>Supports the GCC, OpenCL, AltiVec and NEON vector extensions.</p>
<p>OpenCL vector types are created using <code class="docutils literal"><span class="pre">ext_vector_type</span></code> attribute. It
support for <code class="docutils literal"><span class="pre">V.xyzw</span></code> syntax and other tidbits as seen in OpenCL. An example
is:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="k">typedef</span> <span class="kt">float</span> <span class="n">float4</span> <span class="nf">__attribute__</span><span class="p">((</span><span class="n">ext_vector_type</span><span class="p">(</span><span class="mi">4</span><span class="p">)));</span>
<span class="k">typedef</span> <span class="kt">float</span> <span class="n">float2</span> <span class="nf">__attribute__</span><span class="p">((</span><span class="n">ext_vector_type</span><span class="p">(</span><span class="mi">2</span><span class="p">)));</span>
<span class="n">float4</span> <span class="nf">foo</span><span class="p">(</span><span class="n">float2</span> <span class="n">a</span><span class="p">,</span> <span class="n">float2</span> <span class="n">b</span><span class="p">)</span> <span class="p">{</span>
<span class="n">float4</span> <span class="n">c</span><span class="p">;</span>
<span class="n">c</span><span class="p">.</span><span class="n">xz</span> <span class="o">=</span> <span class="n">a</span><span class="p">;</span>
<span class="n">c</span><span class="p">.</span><span class="n">yw</span> <span class="o">=</span> <span class="n">b</span><span class="p">;</span>
<span class="k">return</span> <span class="n">c</span><span class="p">;</span>
<span class="p">}</span>
</pre></div>
</div>
<p>Query for this feature with <code class="docutils literal"><span class="pre">__has_extension(attribute_ext_vector_type)</span></code>.</p>
<p>Giving <code class="docutils literal"><span class="pre">-faltivec</span></code> option to clang enables support for AltiVec vector syntax
and functions. For example:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="n">vector</span> <span class="kt">float</span> <span class="nf">foo</span><span class="p">(</span><span class="n">vector</span> <span class="kt">int</span> <span class="n">a</span><span class="p">)</span> <span class="p">{</span>
<span class="n">vector</span> <span class="kt">int</span> <span class="n">b</span><span class="p">;</span>
<span class="n">b</span> <span class="o">=</span> <span class="n">vec_add</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">a</span><span class="p">)</span> <span class="o">+</span> <span class="n">a</span><span class="p">;</span>
<span class="k">return</span> <span class="p">(</span><span class="n">vector</span> <span class="kt">float</span><span class="p">)</span><span class="n">b</span><span class="p">;</span>
<span class="p">}</span>
</pre></div>
</div>
<p>NEON vector types are created using <code class="docutils literal"><span class="pre">neon_vector_type</span></code> and
<code class="docutils literal"><span class="pre">neon_polyvector_type</span></code> attributes. For example:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="k">typedef</span> <span class="nf">__attribute__</span><span class="p">((</span><span class="n">neon_vector_type</span><span class="p">(</span><span class="mi">8</span><span class="p">)))</span> <span class="kt">int8_t</span> <span class="n">int8x8_t</span><span class="p">;</span>
<span class="k">typedef</span> <span class="nf">__attribute__</span><span class="p">((</span><span class="n">neon_polyvector_type</span><span class="p">(</span><span class="mi">16</span><span class="p">)))</span> <span class="n">poly8_t</span> <span class="n">poly8x16_t</span><span class="p">;</span>
<span class="n">int8x8_t</span> <span class="nf">foo</span><span class="p">(</span><span class="n">int8x8_t</span> <span class="n">a</span><span class="p">)</span> <span class="p">{</span>
<span class="n">int8x8_t</span> <span class="n">v</span><span class="p">;</span>
<span class="n">v</span> <span class="o">=</span> <span class="n">a</span><span class="p">;</span>
<span class="k">return</span> <span class="n">v</span><span class="p">;</span>
<span class="p">}</span>
</pre></div>
</div>
<div class="section" id="vector-literals">
<h3>Vector Literals<a class="headerlink" href="#vector-literals" title="Permalink to this headline">¶</a></h3>
<p>Vector literals can be used to create vectors from a set of scalars, or
vectors. Either parentheses or braces form can be used. In the parentheses
form the number of literal values specified must be one, i.e. referring to a
scalar value, or must match the size of the vector type being created. If a
single scalar literal value is specified, the scalar literal value will be
replicated to all the components of the vector type. In the brackets form any
number of literals can be specified. For example:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="k">typedef</span> <span class="kt">int</span> <span class="n">v4si</span> <span class="nf">__attribute__</span><span class="p">((</span><span class="n">__vector_size__</span><span class="p">(</span><span class="mi">16</span><span class="p">)));</span>
<span class="k">typedef</span> <span class="kt">float</span> <span class="n">float4</span> <span class="nf">__attribute__</span><span class="p">((</span><span class="n">ext_vector_type</span><span class="p">(</span><span class="mi">4</span><span class="p">)));</span>
<span class="k">typedef</span> <span class="kt">float</span> <span class="n">float2</span> <span class="nf">__attribute__</span><span class="p">((</span><span class="n">ext_vector_type</span><span class="p">(</span><span class="mi">2</span><span class="p">)));</span>
<span class="n">v4si</span> <span class="n">vsi</span> <span class="o">=</span> <span class="p">(</span><span class="n">v4si</span><span class="p">){</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">4</span><span class="p">};</span>
<span class="n">float4</span> <span class="n">vf</span> <span class="o">=</span> <span class="p">(</span><span class="n">float4</span><span class="p">)(</span><span class="mf">1.0f</span><span class="p">,</span> <span class="mf">2.0f</span><span class="p">,</span> <span class="mf">3.0f</span><span class="p">,</span> <span class="mf">4.0f</span><span class="p">);</span>
<span class="n">vector</span> <span class="kt">int</span> <span class="n">vi1</span> <span class="o">=</span> <span class="p">(</span><span class="n">vector</span> <span class="kt">int</span><span class="p">)(</span><span class="mi">1</span><span class="p">);</span> <span class="c1">// vi1 will be (1, 1, 1, 1).</span>
<span class="n">vector</span> <span class="kt">int</span> <span class="n">vi2</span> <span class="o">=</span> <span class="p">(</span><span class="n">vector</span> <span class="kt">int</span><span class="p">){</span><span class="mi">1</span><span class="p">};</span> <span class="c1">// vi2 will be (1, 0, 0, 0).</span>
<span class="n">vector</span> <span class="kt">int</span> <span class="n">vi3</span> <span class="o">=</span> <span class="p">(</span><span class="n">vector</span> <span class="kt">int</span><span class="p">)(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">);</span> <span class="c1">// error</span>
<span class="n">vector</span> <span class="kt">int</span> <span class="n">vi4</span> <span class="o">=</span> <span class="p">(</span><span class="n">vector</span> <span class="kt">int</span><span class="p">){</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">};</span> <span class="c1">// vi4 will be (1, 2, 0, 0).</span>
<span class="n">vector</span> <span class="kt">int</span> <span class="n">vi5</span> <span class="o">=</span> <span class="p">(</span><span class="n">vector</span> <span class="kt">int</span><span class="p">)(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">4</span><span class="p">);</span>
<span class="n">float4</span> <span class="n">vf</span> <span class="o">=</span> <span class="p">(</span><span class="n">float4</span><span class="p">)((</span><span class="n">float2</span><span class="p">)(</span><span class="mf">1.0f</span><span class="p">,</span> <span class="mf">2.0f</span><span class="p">),</span> <span class="p">(</span><span class="n">float2</span><span class="p">)(</span><span class="mf">3.0f</span><span class="p">,</span> <span class="mf">4.0f</span><span class="p">));</span>
</pre></div>
</div>
</div>
<div class="section" id="vector-operations">
<h3>Vector Operations<a class="headerlink" href="#vector-operations" title="Permalink to this headline">¶</a></h3>
<p>The table below shows the support for each operation by vector extension. A
dash indicates that an operation is not accepted according to a corresponding
specification.</p>
<table border="1" class="docutils">
<colgroup>
<col width="60%" />
<col width="12%" />
<col width="14%" />
<col width="6%" />
<col width="8%" />
</colgroup>
<thead valign="bottom">
<tr class="row-odd"><th class="head">Opeator</th>
<th class="head">OpenCL</th>
<th class="head">AltiVec</th>
<th class="head">GCC</th>
<th class="head">NEON</th>
</tr>
</thead>
<tbody valign="top">
<tr class="row-even"><td>[]</td>
<td>yes</td>
<td>yes</td>
<td>yes</td>
<td>–</td>
</tr>
<tr class="row-odd"><td>unary operators +, –</td>
<td>yes</td>
<td>yes</td>
<td>yes</td>
<td>–</td>
</tr>
<tr class="row-even"><td>++, – –</td>
<td>yes</td>
<td>yes</td>
<td>yes</td>
<td>–</td>
</tr>
<tr class="row-odd"><td>+,–,*,/,%</td>
<td>yes</td>
<td>yes</td>
<td>yes</td>
<td>–</td>
</tr>
<tr class="row-even"><td>bitwise operators &,|,^,~</td>
<td>yes</td>
<td>yes</td>
<td>yes</td>
<td>–</td>
</tr>
<tr class="row-odd"><td>>>,<<</td>
<td>yes</td>
<td>yes</td>
<td>yes</td>
<td>–</td>
</tr>
<tr class="row-even"><td>!, &&, ||</td>
<td>no</td>
<td>–</td>
<td>–</td>
<td>–</td>
</tr>
<tr class="row-odd"><td>==, !=, >, <, >=, <=</td>
<td>yes</td>
<td>yes</td>
<td>–</td>
<td>–</td>
</tr>
<tr class="row-even"><td>=</td>
<td>yes</td>
<td>yes</td>
<td>yes</td>
<td>yes</td>
</tr>
<tr class="row-odd"><td>:?</td>
<td>yes</td>
<td>–</td>
<td>–</td>
<td>–</td>
</tr>
<tr class="row-even"><td>sizeof</td>
<td>yes</td>
<td>yes</td>
<td>yes</td>
<td>yes</td>
</tr>
</tbody>
</table>
<p>See also <a class="reference internal" href="#langext-builtin-shufflevector"><span>__builtin_shufflevector</span></a>.</p>
</div>
</div>
<div class="section" id="messages-on-deprecated-and-unavailable-attributes">
<h2><a class="toc-backref" href="#id6">Messages on <code class="docutils literal"><span class="pre">deprecated</span></code> and <code class="docutils literal"><span class="pre">unavailable</span></code> Attributes</a><a class="headerlink" href="#messages-on-deprecated-and-unavailable-attributes" title="Permalink to this headline">¶</a></h2>
<p>An optional string message can be added to the <code class="docutils literal"><span class="pre">deprecated</span></code> and
<code class="docutils literal"><span class="pre">unavailable</span></code> attributes. For example:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="kt">void</span> <span class="nf">explode</span><span class="p">(</span><span class="kt">void</span><span class="p">)</span> <span class="n">__attribute__</span><span class="p">((</span><span class="n">deprecated</span><span class="p">(</span><span class="s">"extremely unsafe, use 'combust' instead!!!"</span><span class="p">)));</span>
</pre></div>
</div>
<p>If the deprecated or unavailable declaration is used, the message will be
incorporated into the appropriate diagnostic:</p>
<div class="highlight-c++"><div class="highlight"><pre>harmless.c:4:3: warning: 'explode' is deprecated: extremely unsafe, use 'combust' instead!!!
[-Wdeprecated-declarations]
explode();
^
</pre></div>
</div>
<p>Query for this feature with
<code class="docutils literal"><span class="pre">__has_extension(attribute_deprecated_with_message)</span></code> and
<code class="docutils literal"><span class="pre">__has_extension(attribute_unavailable_with_message)</span></code>.</p>
</div>
<div class="section" id="attributes-on-enumerators">
<h2><a class="toc-backref" href="#id7">Attributes on Enumerators</a><a class="headerlink" href="#attributes-on-enumerators" title="Permalink to this headline">¶</a></h2>
<p>Clang allows attributes to be written on individual enumerators. This allows
enumerators to be deprecated, made unavailable, etc. The attribute must appear
after the enumerator name and before any initializer, like so:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="k">enum</span> <span class="n">OperationMode</span> <span class="p">{</span>
<span class="n">OM_Invalid</span><span class="p">,</span>
<span class="n">OM_Normal</span><span class="p">,</span>
<span class="n">OM_Terrified</span> <span class="n">__attribute__</span><span class="p">((</span><span class="n">deprecated</span><span class="p">)),</span>
<span class="n">OM_AbortOnError</span> <span class="n">__attribute__</span><span class="p">((</span><span class="n">deprecated</span><span class="p">))</span> <span class="o">=</span> <span class="mi">4</span>
<span class="p">};</span>
</pre></div>
</div>
<p>Attributes on the <code class="docutils literal"><span class="pre">enum</span></code> declaration do not apply to individual enumerators.</p>
<p>Query for this feature with <code class="docutils literal"><span class="pre">__has_extension(enumerator_attributes)</span></code>.</p>
</div>
<div class="section" id="user-specified-system-frameworks">
<h2><a class="toc-backref" href="#id8">‘User-Specified’ System Frameworks</a><a class="headerlink" href="#user-specified-system-frameworks" title="Permalink to this headline">¶</a></h2>
<p>Clang provides a mechanism by which frameworks can be built in such a way that
they will always be treated as being “system frameworks”, even if they are not
present in a system framework directory. This can be useful to system
framework developers who want to be able to test building other applications
with development builds of their framework, including the manner in which the
compiler changes warning behavior for system headers.</p>
<p>Framework developers can opt-in to this mechanism by creating a
“<code class="docutils literal"><span class="pre">.system_framework</span></code>” file at the top-level of their framework. That is, the
framework should have contents like:</p>
<div class="highlight-none"><div class="highlight"><pre>.../TestFramework.framework
.../TestFramework.framework/.system_framework
.../TestFramework.framework/Headers
.../TestFramework.framework/Headers/TestFramework.h
...
</pre></div>
</div>
<p>Clang will treat the presence of this file as an indicator that the framework
should be treated as a system framework, regardless of how it was found in the
framework search path. For consistency, we recommend that such files never be
included in installed versions of the framework.</p>
</div>
<div class="section" id="checks-for-standard-language-features">
<h2><a class="toc-backref" href="#id9">Checks for Standard Language Features</a><a class="headerlink" href="#checks-for-standard-language-features" title="Permalink to this headline">¶</a></h2>
<p>The <code class="docutils literal"><span class="pre">__has_feature</span></code> macro can be used to query if certain standard language
features are enabled. The <code class="docutils literal"><span class="pre">__has_extension</span></code> macro can be used to query if
language features are available as an extension when compiling for a standard
which does not provide them. The features which can be tested are listed here.</p>
<div class="section" id="c-98">
<h3>C++98<a class="headerlink" href="#c-98" title="Permalink to this headline">¶</a></h3>
<p>The features listed below are part of the C++98 standard. These features are
enabled by default when compiling C++ code.</p>
<div class="section" id="c-exceptions">
<h4>C++ exceptions<a class="headerlink" href="#c-exceptions" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_exceptions)</span></code> to determine if C++ exceptions have been
enabled. For example, compiling code with <code class="docutils literal"><span class="pre">-fno-exceptions</span></code> disables C++
exceptions.</p>
</div>
<div class="section" id="c-rtti">
<h4>C++ RTTI<a class="headerlink" href="#c-rtti" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_rtti)</span></code> to determine if C++ RTTI has been enabled. For
example, compiling code with <code class="docutils literal"><span class="pre">-fno-rtti</span></code> disables the use of RTTI.</p>
</div>
</div>
<div class="section" id="c-11">
<h3>C++11<a class="headerlink" href="#c-11" title="Permalink to this headline">¶</a></h3>
<p>The features listed below are part of the C++11 standard. As a result, all
these features are enabled with the <code class="docutils literal"><span class="pre">-std=c++11</span></code> or <code class="docutils literal"><span class="pre">-std=gnu++11</span></code> option
when compiling C++ code.</p>
<div class="section" id="c-11-sfinae-includes-access-control">
<h4>C++11 SFINAE includes access control<a class="headerlink" href="#c-11-sfinae-includes-access-control" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_access_control_sfinae)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_access_control_sfinae)</span></code> 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 class="reference external" href="http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#1170">C++ DR1170</a>.</p>
</div>
<div class="section" id="c-11-alias-templates">
<h4>C++11 alias templates<a class="headerlink" href="#c-11-alias-templates" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_alias_templates)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_alias_templates)</span></code> to determine if support for C++11’s
alias declarations and alias templates is enabled.</p>
</div>
<div class="section" id="c-11-alignment-specifiers">
<h4>C++11 alignment specifiers<a class="headerlink" href="#c-11-alignment-specifiers" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_alignas)</span></code> or <code class="docutils literal"><span class="pre">__has_extension(cxx_alignas)</span></code> to
determine if support for alignment specifiers using <code class="docutils literal"><span class="pre">alignas</span></code> is enabled.</p>
</div>
<div class="section" id="c-11-attributes">
<h4>C++11 attributes<a class="headerlink" href="#c-11-attributes" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_attributes)</span></code> or <code class="docutils literal"><span class="pre">__has_extension(cxx_attributes)</span></code> to
determine if support for attribute parsing with C++11’s square bracket notation
is enabled.</p>
</div>
<div class="section" id="c-11-generalized-constant-expressions">
<h4>C++11 generalized constant expressions<a class="headerlink" href="#c-11-generalized-constant-expressions" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_constexpr)</span></code> to determine if support for generalized
constant expressions (e.g., <code class="docutils literal"><span class="pre">constexpr</span></code>) is enabled.</p>
</div>
<div class="section" id="c-11-decltype">
<h4>C++11 <code class="docutils literal"><span class="pre">decltype()</span></code><a class="headerlink" href="#c-11-decltype" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_decltype)</span></code> or <code class="docutils literal"><span class="pre">__has_extension(cxx_decltype)</span></code> to
determine if support for the <code class="docutils literal"><span class="pre">decltype()</span></code> specifier is enabled. C++11’s
<code class="docutils literal"><span class="pre">decltype</span></code> does not require type-completeness of a function call expression.
Use <code class="docutils literal"><span class="pre">__has_feature(cxx_decltype_incomplete_return_types)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_decltype_incomplete_return_types)</span></code> to determine if
support for this feature is enabled.</p>
</div>
<div class="section" id="c-11-default-template-arguments-in-function-templates">
<h4>C++11 default template arguments in function templates<a class="headerlink" href="#c-11-default-template-arguments-in-function-templates" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_default_function_template_args)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_default_function_template_args)</span></code> to determine if support
for default template arguments in function templates is enabled.</p>
</div>
<div class="section" id="c-11-defaulted-functions">
<h4>C++11 <code class="docutils literal"><span class="pre">default</span></code>ed functions<a class="headerlink" href="#c-11-defaulted-functions" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_defaulted_functions)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_defaulted_functions)</span></code> to determine if support for
defaulted function definitions (with <code class="docutils literal"><span class="pre">=</span> <span class="pre">default</span></code>) is enabled.</p>
</div>
<div class="section" id="c-11-delegating-constructors">
<h4>C++11 delegating constructors<a class="headerlink" href="#c-11-delegating-constructors" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_delegating_constructors)</span></code> to determine if support for
delegating constructors is enabled.</p>
</div>
<div class="section" id="c-11-deleted-functions">
<h4>C++11 <code class="docutils literal"><span class="pre">deleted</span></code> functions<a class="headerlink" href="#c-11-deleted-functions" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_deleted_functions)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_deleted_functions)</span></code> to determine if support for deleted
function definitions (with <code class="docutils literal"><span class="pre">=</span> <span class="pre">delete</span></code>) is enabled.</p>
</div>
<div class="section" id="c-11-explicit-conversion-functions">
<h4>C++11 explicit conversion functions<a class="headerlink" href="#c-11-explicit-conversion-functions" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_explicit_conversions)</span></code> to determine if support for
<code class="docutils literal"><span class="pre">explicit</span></code> conversion functions is enabled.</p>
</div>
<div class="section" id="c-11-generalized-initializers">
<h4>C++11 generalized initializers<a class="headerlink" href="#c-11-generalized-initializers" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_generalized_initializers)</span></code> to determine if support for
generalized initializers (using braced lists and <code class="docutils literal"><span class="pre">std::initializer_list</span></code>) is
enabled.</p>
</div>
<div class="section" id="c-11-implicit-move-constructors-assignment-operators">
<h4>C++11 implicit move constructors/assignment operators<a class="headerlink" href="#c-11-implicit-move-constructors-assignment-operators" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_implicit_moves)</span></code> to determine if Clang will implicitly
generate move constructors and move assignment operators where needed.</p>
</div>
<div class="section" id="c-11-inheriting-constructors">
<h4>C++11 inheriting constructors<a class="headerlink" href="#c-11-inheriting-constructors" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_inheriting_constructors)</span></code> to determine if support for
inheriting constructors is enabled.</p>
</div>
<div class="section" id="c-11-inline-namespaces">
<h4>C++11 inline namespaces<a class="headerlink" href="#c-11-inline-namespaces" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_inline_namespaces)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_inline_namespaces)</span></code> to determine if support for inline
namespaces is enabled.</p>
</div>
<div class="section" id="c-11-lambdas">
<h4>C++11 lambdas<a class="headerlink" href="#c-11-lambdas" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_lambdas)</span></code> or <code class="docutils literal"><span class="pre">__has_extension(cxx_lambdas)</span></code> to
determine if support for lambdas is enabled.</p>
</div>
<div class="section" id="c-11-local-and-unnamed-types-as-template-arguments">
<h4>C++11 local and unnamed types as template arguments<a class="headerlink" href="#c-11-local-and-unnamed-types-as-template-arguments" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_local_type_template_args)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_local_type_template_args)</span></code> to determine if support for
local and unnamed types as template arguments is enabled.</p>
</div>
<div class="section" id="c-11-noexcept">
<h4>C++11 noexcept<a class="headerlink" href="#c-11-noexcept" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_noexcept)</span></code> or <code class="docutils literal"><span class="pre">__has_extension(cxx_noexcept)</span></code> to
determine if support for noexcept exception specifications is enabled.</p>
</div>
<div class="section" id="c-11-in-class-non-static-data-member-initialization">
<h4>C++11 in-class non-static data member initialization<a class="headerlink" href="#c-11-in-class-non-static-data-member-initialization" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_nonstatic_member_init)</span></code> to determine whether in-class
initialization of non-static data members is enabled.</p>
</div>
<div class="section" id="c-11-nullptr">
<h4>C++11 <code class="docutils literal"><span class="pre">nullptr</span></code><a class="headerlink" href="#c-11-nullptr" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_nullptr)</span></code> or <code class="docutils literal"><span class="pre">__has_extension(cxx_nullptr)</span></code> to
determine if support for <code class="docutils literal"><span class="pre">nullptr</span></code> is enabled.</p>
</div>
<div class="section" id="c-11-override-control">
<h4>C++11 <code class="docutils literal"><span class="pre">override</span> <span class="pre">control</span></code><a class="headerlink" href="#c-11-override-control" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_override_control)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_override_control)</span></code> to determine if support for the
override control keywords is enabled.</p>
</div>
<div class="section" id="c-11-reference-qualified-functions">
<h4>C++11 reference-qualified functions<a class="headerlink" href="#c-11-reference-qualified-functions" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_reference_qualified_functions)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_reference_qualified_functions)</span></code> to determine if support
for reference-qualified functions (e.g., member functions with <code class="docutils literal"><span class="pre">&</span></code> or <code class="docutils literal"><span class="pre">&&</span></code>
applied to <code class="docutils literal"><span class="pre">*this</span></code>) is enabled.</p>
</div>
<div class="section" id="c-11-range-based-for-loop">
<h4>C++11 range-based <code class="docutils literal"><span class="pre">for</span></code> loop<a class="headerlink" href="#c-11-range-based-for-loop" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_range_for)</span></code> or <code class="docutils literal"><span class="pre">__has_extension(cxx_range_for)</span></code> to
determine if support for the range-based for loop is enabled.</p>
</div>
<div class="section" id="c-11-raw-string-literals">
<h4>C++11 raw string literals<a class="headerlink" href="#c-11-raw-string-literals" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_raw_string_literals)</span></code> to determine if support for raw
string literals (e.g., <code class="docutils literal"><span class="pre">R"x(foo\bar)x"</span></code>) is enabled.</p>
</div>
<div class="section" id="c-11-rvalue-references">
<h4>C++11 rvalue references<a class="headerlink" href="#c-11-rvalue-references" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_rvalue_references)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_rvalue_references)</span></code> to determine if support for rvalue
references is enabled.</p>
</div>
<div class="section" id="c-11-static-assert">
<h4>C++11 <code class="docutils literal"><span class="pre">static_assert()</span></code><a class="headerlink" href="#c-11-static-assert" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_static_assert)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_static_assert)</span></code> to determine if support for compile-time
assertions using <code class="docutils literal"><span class="pre">static_assert</span></code> is enabled.</p>
</div>
<div class="section" id="c-11-thread-local">
<h4>C++11 <code class="docutils literal"><span class="pre">thread_local</span></code><a class="headerlink" href="#c-11-thread-local" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_thread_local)</span></code> to determine if support for
<code class="docutils literal"><span class="pre">thread_local</span></code> variables is enabled.</p>
</div>
<div class="section" id="c-11-type-inference">
<h4>C++11 type inference<a class="headerlink" href="#c-11-type-inference" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_auto_type)</span></code> or <code class="docutils literal"><span class="pre">__has_extension(cxx_auto_type)</span></code> to
determine C++11 type inference is supported using the <code class="docutils literal"><span class="pre">auto</span></code> specifier. If
this is disabled, <code class="docutils literal"><span class="pre">auto</span></code> will instead be a storage class specifier, as in C
or C++98.</p>
</div>
<div class="section" id="c-11-strongly-typed-enumerations">
<h4>C++11 strongly typed enumerations<a class="headerlink" href="#c-11-strongly-typed-enumerations" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_strong_enums)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_strong_enums)</span></code> to determine if support for strongly
typed, scoped enumerations is enabled.</p>
</div>
<div class="section" id="c-11-trailing-return-type">
<h4>C++11 trailing return type<a class="headerlink" href="#c-11-trailing-return-type" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_trailing_return)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_trailing_return)</span></code> to determine if support for the
alternate function declaration syntax with trailing return type is enabled.</p>
</div>
<div class="section" id="c-11-unicode-string-literals">
<h4>C++11 Unicode string literals<a class="headerlink" href="#c-11-unicode-string-literals" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_unicode_literals)</span></code> to determine if support for Unicode
string literals is enabled.</p>
</div>
<div class="section" id="c-11-unrestricted-unions">
<h4>C++11 unrestricted unions<a class="headerlink" href="#c-11-unrestricted-unions" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_unrestricted_unions)</span></code> to determine if support for
unrestricted unions is enabled.</p>
</div>
<div class="section" id="c-11-user-defined-literals">
<h4>C++11 user-defined literals<a class="headerlink" href="#c-11-user-defined-literals" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_user_literals)</span></code> to determine if support for
user-defined literals is enabled.</p>
</div>
<div class="section" id="c-11-variadic-templates">
<h4>C++11 variadic templates<a class="headerlink" href="#c-11-variadic-templates" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_variadic_templates)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_variadic_templates)</span></code> to determine if support for
variadic templates is enabled.</p>
</div>
</div>
<div class="section" id="c-1y">
<h3>C++1y<a class="headerlink" href="#c-1y" title="Permalink to this headline">¶</a></h3>
<p>The features listed below are part of the committee draft for the C++1y
standard. As a result, all these features are enabled with the <code class="docutils literal"><span class="pre">-std=c++1y</span></code>
or <code class="docutils literal"><span class="pre">-std=gnu++1y</span></code> option when compiling C++ code.</p>
<div class="section" id="c-1y-binary-literals">
<h4>C++1y binary literals<a class="headerlink" href="#c-1y-binary-literals" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_binary_literals)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_binary_literals)</span></code> to determine whether
binary literals (for instance, <code class="docutils literal"><span class="pre">0b10010</span></code>) are recognized. Clang supports this
feature as an extension in all language modes.</p>
</div>
<div class="section" id="c-1y-contextual-conversions">
<h4>C++1y contextual conversions<a class="headerlink" href="#c-1y-contextual-conversions" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_contextual_conversions)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_contextual_conversions)</span></code> to determine if the C++1y rules
are used when performing an implicit conversion for an array bound in a
<em>new-expression</em>, the operand of a <em>delete-expression</em>, an integral constant
expression, or a condition in a <code class="docutils literal"><span class="pre">switch</span></code> statement.</p>
</div>
<div class="section" id="c-1y-decltype-auto">
<h4>C++1y decltype(auto)<a class="headerlink" href="#c-1y-decltype-auto" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_decltype_auto)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_decltype_auto)</span></code> to determine if support
for the <code class="docutils literal"><span class="pre">decltype(auto)</span></code> placeholder type is enabled.</p>
</div>
<div class="section" id="c-1y-default-initializers-for-aggregates">
<h4>C++1y default initializers for aggregates<a class="headerlink" href="#c-1y-default-initializers-for-aggregates" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_aggregate_nsdmi)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_aggregate_nsdmi)</span></code> to determine if support
for default initializers in aggregate members is enabled.</p>
</div>
<div class="section" id="c-1y-generalized-lambda-capture">
<h4>C++1y generalized lambda capture<a class="headerlink" href="#c-1y-generalized-lambda-capture" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_init_captures)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_init_captures)</span></code> to determine if support for
lambda captures with explicit initializers is enabled
(for instance, <code class="docutils literal"><span class="pre">[n(0)]</span> <span class="pre">{</span> <span class="pre">return</span> <span class="pre">++n;</span> <span class="pre">}</span></code>).</p>
</div>
<div class="section" id="c-1y-generic-lambdas">
<h4>C++1y generic lambdas<a class="headerlink" href="#c-1y-generic-lambdas" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_generic_lambdas)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_generic_lambdas)</span></code> to determine if support for generic
(polymorphic) lambdas is enabled
(for instance, <code class="docutils literal"><span class="pre">[]</span> <span class="pre">(auto</span> <span class="pre">x)</span> <span class="pre">{</span> <span class="pre">return</span> <span class="pre">x</span> <span class="pre">+</span> <span class="pre">1;</span> <span class="pre">}</span></code>).</p>
</div>
<div class="section" id="c-1y-relaxed-constexpr">
<h4>C++1y relaxed constexpr<a class="headerlink" href="#c-1y-relaxed-constexpr" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_relaxed_constexpr)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_relaxed_constexpr)</span></code> to determine if variable
declarations, local variable modification, and control flow constructs
are permitted in <code class="docutils literal"><span class="pre">constexpr</span></code> functions.</p>
</div>
<div class="section" id="c-1y-return-type-deduction">
<h4>C++1y return type deduction<a class="headerlink" href="#c-1y-return-type-deduction" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_return_type_deduction)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_return_type_deduction)</span></code> to determine if support
for return type deduction for functions (using <code class="docutils literal"><span class="pre">auto</span></code> as a return type)
is enabled.</p>
</div>
<div class="section" id="c-1y-runtime-sized-arrays">
<h4>C++1y runtime-sized arrays<a class="headerlink" href="#c-1y-runtime-sized-arrays" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_runtime_array)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_runtime_array)</span></code> to determine if support
for arrays of runtime bound (a restricted form of variable-length arrays)
is enabled.
Clang’s implementation of this feature is incomplete.</p>
</div>
<div class="section" id="c-1y-variable-templates">
<h4>C++1y variable templates<a class="headerlink" href="#c-1y-variable-templates" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(cxx_variable_templates)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(cxx_variable_templates)</span></code> to determine if support for
templated variable declarations is enabled.</p>
</div>
</div>
<div class="section" id="c11">
<h3>C11<a class="headerlink" href="#c11" title="Permalink to this headline">¶</a></h3>
<p>The features listed below are part of the C11 standard. As a result, all these
features are enabled with the <code class="docutils literal"><span class="pre">-std=c11</span></code> or <code class="docutils literal"><span class="pre">-std=gnu11</span></code> option when
compiling C code. Additionally, because these features are all
backward-compatible, they are available as extensions in all language modes.</p>
<div class="section" id="c11-alignment-specifiers">
<h4>C11 alignment specifiers<a class="headerlink" href="#c11-alignment-specifiers" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(c_alignas)</span></code> or <code class="docutils literal"><span class="pre">__has_extension(c_alignas)</span></code> to determine
if support for alignment specifiers using <code class="docutils literal"><span class="pre">_Alignas</span></code> is enabled.</p>
</div>
<div class="section" id="c11-atomic-operations">
<h4>C11 atomic operations<a class="headerlink" href="#c11-atomic-operations" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(c_atomic)</span></code> or <code class="docutils literal"><span class="pre">__has_extension(c_atomic)</span></code> to determine
if support for atomic types using <code class="docutils literal"><span class="pre">_Atomic</span></code> is enabled. Clang also provides
<a class="reference internal" href="#langext-c11-atomic"><span>a set of builtins</span></a> which can be used to implement
the <code class="docutils literal"><span class="pre"><stdatomic.h></span></code> operations on <code class="docutils literal"><span class="pre">_Atomic</span></code> types.</p>
</div>
<div class="section" id="c11-generic-selections">
<h4>C11 generic selections<a class="headerlink" href="#c11-generic-selections" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(c_generic_selections)</span></code> or
<code class="docutils literal"><span class="pre">__has_extension(c_generic_selections)</span></code> to determine if support for generic
selections is enabled.</p>
<p>As an extension, the C11 generic selection expression is available in all
languages supported by Clang. The syntax is the same as that given in the C11
standard.</p>
<p>In C, type compatibility is decided according to the rules given in the
appropriate standard, but in C++, which lacks the type compatibility rules used
in C, types are considered compatible only if they are equivalent.</p>
</div>
<div class="section" id="c11-static-assert">
<h4>C11 <code class="docutils literal"><span class="pre">_Static_assert()</span></code><a class="headerlink" href="#c11-static-assert" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(c_static_assert)</span></code> or <code class="docutils literal"><span class="pre">__has_extension(c_static_assert)</span></code>
to determine if support for compile-time assertions using <code class="docutils literal"><span class="pre">_Static_assert</span></code> is
enabled.</p>
</div>
<div class="section" id="c11-thread-local">
<h4>C11 <code class="docutils literal"><span class="pre">_Thread_local</span></code><a class="headerlink" href="#c11-thread-local" title="Permalink to this headline">¶</a></h4>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(c_thread_local)</span></code> or <code class="docutils literal"><span class="pre">__has_extension(c_thread_local)</span></code>
to determine if support for <code class="docutils literal"><span class="pre">_Thread_local</span></code> variables is enabled.</p>
</div>
</div>
</div>
<div class="section" id="checks-for-type-trait-primitives">
<h2><a class="toc-backref" href="#id10">Checks for Type Trait Primitives</a><a class="headerlink" href="#checks-for-type-trait-primitives" title="Permalink to this headline">¶</a></h2>
<p>Type trait primitives are special builtin constant expressions that can be used
by the standard C++ library to facilitate or simplify the implementation of
user-facing type traits in the <type_traits> header.</p>
<p>They are not intended to be used directly by user code because they are
implementation-defined and subject to change – as such they’re tied closely to
the supported set of system headers, currently:</p>
<ul class="simple">
<li>LLVM’s own libc++</li>
<li>GNU libstdc++</li>
<li>The Microsoft standard C++ library</li>
</ul>
<p>Clang supports the <a class="reference external" href="http://gcc.gnu.org/onlinedocs/gcc/Type-Traits.html">GNU C++ type traits</a> and a subset of the
<a class="reference external" href="http://msdn.microsoft.com/en-us/library/ms177194(v=VS.100).aspx">Microsoft Visual C++ Type traits</a>.</p>
<p>Feature detection is supported only for some of the primitives at present. User
code should not use these checks because they bear no direct relation to the
actual set of type traits supported by the C++ standard library.</p>
<p>For type trait <code class="docutils literal"><span class="pre">__X</span></code>, <code class="docutils literal"><span class="pre">__has_extension(X)</span></code> indicates the presence of the
type trait primitive in the compiler. A simplistic usage example as might be
seen in standard C++ headers follows:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#if __has_extension(is_convertible_to)</span>
<span class="k">template</span><span class="o"><</span><span class="k">typename</span> <span class="n">From</span><span class="p">,</span> <span class="k">typename</span> <span class="n">To</span><span class="o">></span>
<span class="k">struct</span> <span class="n">is_convertible_to</span> <span class="p">{</span>
<span class="k">static</span> <span class="k">const</span> <span class="kt">bool</span> <span class="n">value</span> <span class="o">=</span> <span class="n">__is_convertible_to</span><span class="p">(</span><span class="n">From</span><span class="p">,</span> <span class="n">To</span><span class="p">);</span>
<span class="p">};</span>
<span class="cp">#else</span>
<span class="c1">// Emulate type trait for compatibility with other compilers.</span>
<span class="cp">#endif</span>
</pre></div>
</div>
<p>The following type trait primitives are supported by Clang:</p>
<ul class="simple">
<li><code class="docutils literal"><span class="pre">__has_nothrow_assign</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__has_nothrow_copy</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__has_nothrow_constructor</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__has_trivial_assign</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__has_trivial_copy</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__has_trivial_constructor</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__has_trivial_destructor</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__has_virtual_destructor</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__is_abstract</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__is_base_of</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__is_class</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__is_convertible_to</span></code> (Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__is_empty</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__is_enum</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__is_interface_class</span></code> (Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__is_pod</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__is_polymorphic</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__is_union</span></code> (GNU, Microsoft)</li>
<li><code class="docutils literal"><span class="pre">__is_literal(type)</span></code>: Determines whether the given type is a literal type</li>
<li><code class="docutils literal"><span class="pre">__is_final</span></code>: Determines whether the given type is declared with a
<code class="docutils literal"><span class="pre">final</span></code> class-virt-specifier.</li>
<li><code class="docutils literal"><span class="pre">__underlying_type(type)</span></code>: Retrieves the underlying type for a given
<code class="docutils literal"><span class="pre">enum</span></code> type. This trait is required to implement the C++11 standard
library.</li>
<li><code class="docutils literal"><span class="pre">__is_trivially_assignable(totype,</span> <span class="pre">fromtype)</span></code>: Determines whether a value
of type <code class="docutils literal"><span class="pre">totype</span></code> can be assigned to from a value of type <code class="docutils literal"><span class="pre">fromtype</span></code> 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>
<li><code class="docutils literal"><span class="pre">__is_trivially_constructible(type,</span> <span class="pre">argtypes...)</span></code>: Determines whether a
value of type <code class="docutils literal"><span class="pre">type</span></code> can be direct-initialized with arguments of types
<code class="docutils literal"><span class="pre">argtypes...</span></code> 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>
<li><code class="docutils literal"><span class="pre">__is_destructible</span></code> (MSVC 2013): partially implemented</li>
<li><code class="docutils literal"><span class="pre">__is_nothrow_destructible</span></code> (MSVC 2013): partially implemented</li>
<li><code class="docutils literal"><span class="pre">__is_nothrow_assignable</span></code> (MSVC 2013, clang)</li>
<li><code class="docutils literal"><span class="pre">__is_constructible</span></code> (MSVC 2013, clang)</li>
<li><code class="docutils literal"><span class="pre">__is_nothrow_constructible</span></code> (MSVC 2013, clang)</li>
</ul>
</div>
<div class="section" id="blocks">
<h2><a class="toc-backref" href="#id11">Blocks</a><a class="headerlink" href="#blocks" title="Permalink to this headline">¶</a></h2>
<p>The syntax and high level language feature description is in
<a class="reference internal" href="BlockLanguageSpec.html"><em>BlockLanguageSpec</em></a>. Implementation and ABI details for
the clang implementation are in <a class="reference internal" href="Block-ABI-Apple.html"><em>Block-ABI-Apple</em></a>.</p>
<p>Query for this feature with <code class="docutils literal"><span class="pre">__has_extension(blocks)</span></code>.</p>
</div>
<div class="section" id="objective-c-features">
<h2><a class="toc-backref" href="#id12">Objective-C Features</a><a class="headerlink" href="#objective-c-features" title="Permalink to this headline">¶</a></h2>
<div class="section" id="related-result-types">
<h3>Related result types<a class="headerlink" href="#related-result-types" title="Permalink to this headline">¶</a></h3>
<p>According to Cocoa conventions, Objective-C methods with certain names
(“<code class="docutils literal"><span class="pre">init</span></code>”, “<code class="docutils literal"><span class="pre">alloc</span></code>”, 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>
<div class="highlight-objc"><div class="highlight"><pre><span class="k">@interface</span> <span class="bp">NSObject</span>
<span class="p">+</span> <span class="p">(</span><span class="kt">id</span><span class="p">)</span><span class="nf">alloc</span><span class="p">;</span>
<span class="p">-</span> <span class="p">(</span><span class="kt">id</span><span class="p">)</span><span class="nf">init</span><span class="p">;</span>
<span class="k">@end</span>
<span class="k">@interface</span> <span class="bp">NSArray</span> : <span class="bp">NSObject</span>
<span class="k">@end</span>
</pre></div>
</div>
<p>and this common initialization pattern</p>
<div class="highlight-objc"><div class="highlight"><pre><span class="bp">NSArray</span> <span class="o">*</span><span class="n">array</span> <span class="o">=</span> <span class="p">[[</span><span class="bp">NSArray</span> <span class="n">alloc</span><span class="p">]</span> <span class="n">init</span><span class="p">];</span>
</pre></div>
</div>
<p>the type of the expression <code class="docutils literal"><span class="pre">[NSArray</span> <span class="pre">alloc]</span></code> is <code class="docutils literal"><span class="pre">NSArray*</span></code> because
<code class="docutils literal"><span class="pre">alloc</span></code> implicitly has a related result type. Similarly, the type of the
expression <code class="docutils literal"><span class="pre">[[NSArray</span> <span class="pre">alloc]</span> <span class="pre">init]</span></code> is <code class="docutils literal"><span class="pre">NSArray*</span></code>, since <code class="docutils literal"><span class="pre">init</span></code> has a
related result type and its receiver is known to have the type <code class="docutils literal"><span class="pre">NSArray</span> <span class="pre">*</span></code>.
If neither <code class="docutils literal"><span class="pre">alloc</span></code> nor <code class="docutils literal"><span class="pre">init</span></code> had a related result type, the expressions
would have had type <code class="docutils literal"><span class="pre">id</span></code>, as declared in the method signature.</p>
<p>A method with a related result type can be declared by using the type
<code class="docutils literal"><span class="pre">instancetype</span></code> as its result type. <code class="docutils literal"><span class="pre">instancetype</span></code> is a contextual keyword
that is only permitted in the result type of an Objective-C method, e.g.</p>
<div class="highlight-objc"><div class="highlight"><pre><span class="k">@interface</span> <span class="nc">A</span>
<span class="p">+</span> <span class="p">(</span><span class="kt">instancetype</span><span class="p">)</span><span class="nf">constructAnA</span><span class="p">;</span>
<span class="k">@end</span>
</pre></div>
</div>
<p>The related result type can also be inferred for some methods. To determine
whether a method has an inferred related result type, the first word in the
camel-case selector (e.g., “<code class="docutils literal"><span class="pre">init</span></code>” in “<code class="docutils literal"><span class="pre">initWithObjects</span></code>”) is considered,
and the method will have a related result type if its return type is compatible
with the type of its class and if:</p>
<ul class="simple">
<li>the first word is “<code class="docutils literal"><span class="pre">alloc</span></code>” or “<code class="docutils literal"><span class="pre">new</span></code>”, and the method is a class method,
or</li>
<li>the first word is “<code class="docutils literal"><span class="pre">autorelease</span></code>”, “<code class="docutils literal"><span class="pre">init</span></code>”, “<code class="docutils literal"><span class="pre">retain</span></code>”, or “<code class="docutils literal"><span class="pre">self</span></code>”,
and the method is an instance method.</li>
</ul>
<p>If a method with a related result type is overridden by a subclass method, the
subclass method must also return a type that is compatible with the subclass
type. For example:</p>
<div class="highlight-objc"><div class="highlight"><pre><span class="k">@interface</span> <span class="bp">NSString</span> : <span class="bp">NSObject</span>
<span class="p">-</span> <span class="p">(</span><span class="n">NSUnrelated</span> <span class="o">*</span><span class="p">)</span><span class="nf">init</span><span class="p">;</span> <span class="c1">// incorrect usage: NSUnrelated is not NSString or a superclass of NSString</span>
<span class="k">@end</span>
</pre></div>
</div>
<p>Related result types only affect the type of a message send or property access
via the given method. In all other respects, a method with a related result
type is treated the same way as method that returns <code class="docutils literal"><span class="pre">id</span></code>.</p>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(objc_instancetype)</span></code> to determine whether the
<code class="docutils literal"><span class="pre">instancetype</span></code> contextual keyword is available.</p>
</div>
<div class="section" id="automatic-reference-counting">
<h3>Automatic reference counting<a class="headerlink" href="#automatic-reference-counting" title="Permalink to this headline">¶</a></h3>
<p>Clang provides support for <a class="reference internal" href="AutomaticReferenceCounting.html"><em>automated reference counting</em></a> in Objective-C, which eliminates the need
for manual <code class="docutils literal"><span class="pre">retain</span></code>/<code class="docutils literal"><span class="pre">release</span></code>/<code class="docutils literal"><span class="pre">autorelease</span></code> message sends. There are two
feature macros associated with automatic reference counting:
<code class="docutils literal"><span class="pre">__has_feature(objc_arc)</span></code> indicates the availability of automated reference
counting in general, while <code class="docutils literal"><span class="pre">__has_feature(objc_arc_weak)</span></code> indicates that
automated reference counting also includes support for <code class="docutils literal"><span class="pre">__weak</span></code> pointers to
Objective-C objects.</p>
</div>
<div class="section" id="enumerations-with-a-fixed-underlying-type">
<span id="objc-fixed-enum"></span><h3>Enumerations with a fixed underlying type<a class="headerlink" href="#enumerations-with-a-fixed-underlying-type" title="Permalink to this headline">¶</a></h3>
<p>Clang provides support for C++11 enumerations with a fixed underlying type
within Objective-C. For example, one can write an enumeration type as:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="k">typedef</span> <span class="k">enum</span> <span class="o">:</span> <span class="kt">unsigned</span> <span class="kt">char</span> <span class="p">{</span> <span class="n">Red</span><span class="p">,</span> <span class="n">Green</span><span class="p">,</span> <span class="n">Blue</span> <span class="p">}</span> <span class="n">Color</span><span class="p">;</span>
</pre></div>
</div>
<p>This specifies that the underlying type, which is used to store the enumeration
value, is <code class="docutils literal"><span class="pre">unsigned</span> <span class="pre">char</span></code>.</p>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(objc_fixed_enum)</span></code> to determine whether support for fixed
underlying types is available in Objective-C.</p>
</div>
<div class="section" id="interoperability-with-c-11-lambdas">
<h3>Interoperability with C++11 lambdas<a class="headerlink" href="#interoperability-with-c-11-lambdas" title="Permalink to this headline">¶</a></h3>
<p>Clang provides interoperability between C++11 lambdas and blocks-based APIs, by
permitting a lambda to be implicitly converted to a block pointer with the
corresponding signature. For example, consider an API such as <code class="docutils literal"><span class="pre">NSArray</span></code>‘s
array-sorting method:</p>
<div class="highlight-objc"><div class="highlight"><pre><span class="p">-</span> <span class="p">(</span><span class="bp">NSArray</span> <span class="o">*</span><span class="p">)</span><span class="nf">sortedArrayUsingComparator:</span><span class="p">(</span><span class="n">NSComparator</span><span class="p">)</span><span class="nv">cmptr</span><span class="p">;</span>
</pre></div>
</div>
<p><code class="docutils literal"><span class="pre">NSComparator</span></code> is simply a typedef for the block pointer <code class="docutils literal"><span class="pre">NSComparisonResult</span>
<span class="pre">(^)(id,</span> <span class="pre">id)</span></code>, and parameters of this type are generally provided with block
literals as arguments. However, one can also use a C++11 lambda so long as it
provides the same signature (in this case, accepting two parameters of type
<code class="docutils literal"><span class="pre">id</span></code> and returning an <code class="docutils literal"><span class="pre">NSComparisonResult</span></code>):</p>
<div class="highlight-objc"><div class="highlight"><pre><span class="bp">NSArray</span> <span class="o">*</span><span class="n">array</span> <span class="o">=</span> <span class="l">@[</span><span class="s">@"string 1"</span><span class="p">,</span> <span class="s">@"string 21"</span><span class="p">,</span> <span class="s">@"string 12"</span><span class="p">,</span> <span class="s">@"String 11"</span><span class="p">,</span>
<span class="s">@"String 02"</span><span class="l">]</span><span class="p">;</span>
<span class="k">const</span> <span class="n">NSStringCompareOptions</span> <span class="n">comparisonOptions</span>
<span class="o">=</span> <span class="n">NSCaseInsensitiveSearch</span> <span class="o">|</span> <span class="n">NSNumericSearch</span> <span class="o">|</span>
<span class="n">NSWidthInsensitiveSearch</span> <span class="o">|</span> <span class="n">NSForcedOrderingSearch</span><span class="p">;</span>
<span class="bp">NSLocale</span> <span class="o">*</span><span class="n">currentLocale</span> <span class="o">=</span> <span class="p">[</span><span class="bp">NSLocale</span> <span class="n">currentLocale</span><span class="p">];</span>
<span class="bp">NSArray</span> <span class="o">*</span><span class="n">sorted</span>
<span class="o">=</span> <span class="p">[</span><span class="n">array</span> <span class="nl">sortedArrayUsingComparator</span><span class="p">:[</span><span class="o">=</span><span class="p">](</span><span class="kt">id</span> <span class="n">s1</span><span class="p">,</span> <span class="kt">id</span> <span class="n">s2</span><span class="p">)</span> <span class="o">-></span> <span class="n">NSComparisonResult</span> <span class="p">{</span>
<span class="n">NSRange</span> <span class="n">string1Range</span> <span class="o">=</span> <span class="n">NSMakeRange</span><span class="p">(</span><span class="mi">0</span><span class="p">,</span> <span class="p">[</span><span class="n">s1</span> <span class="n">length</span><span class="p">]);</span>
<span class="k">return</span> <span class="p">[</span><span class="n">s1</span> <span class="nl">compare</span><span class="p">:</span><span class="n">s2</span> <span class="nl">options</span><span class="p">:</span><span class="n">comparisonOptions</span>
<span class="nl">range</span><span class="p">:</span><span class="n">string1Range</span> <span class="nl">locale</span><span class="p">:</span><span class="n">currentLocale</span><span class="p">];</span>
<span class="p">}];</span>
<span class="n">NSLog</span><span class="p">(</span><span class="s">@"sorted: %@"</span><span class="p">,</span> <span class="n">sorted</span><span class="p">);</span>
</pre></div>
</div>
<p>This code relies on an implicit conversion from the type of the lambda
expression (an unnamed, local class type called the <em>closure type</em>) to the
corresponding block pointer type. The conversion itself is expressed by a
conversion operator in that closure type that produces a block pointer with the
same signature as the lambda itself, e.g.,</p>
<div class="highlight-objc"><div class="highlight"><pre><span class="n">operator</span> <span class="nf">NSComparisonResult</span> <span class="p">(</span><span class="o">^</span><span class="p">)(</span><span class="kt">id</span><span class="p">,</span> <span class="kt">id</span><span class="p">)()</span> <span class="k">const</span><span class="p">;</span>
</pre></div>
</div>
<p>This conversion function returns a new block that simply forwards the two
parameters to the lambda object (which it captures by copy), then returns the
result. The returned block is first copied (with <code class="docutils literal"><span class="pre">Block_copy</span></code>) and then
autoreleased. As an optimization, if a lambda expression is immediately
converted to a block pointer (as in the first example, above), then the block
is not copied and autoreleased: rather, it is given the same lifetime as a
block literal written at that point in the program, which avoids the overhead
of copying a block to the heap in the common case.</p>
<p>The conversion from a lambda to a block pointer is only available in
Objective-C++, and not in C++ with blocks, due to its use of Objective-C memory
management (autorelease).</p>
</div>
<div class="section" id="object-literals-and-subscripting">
<h3>Object Literals and Subscripting<a class="headerlink" href="#object-literals-and-subscripting" title="Permalink to this headline">¶</a></h3>
<p>Clang provides support for <a class="reference internal" href="ObjectiveCLiterals.html"><em>Object Literals and Subscripting</em></a> in Objective-C, which simplifies common Objective-C
programming patterns, makes programs more concise, and improves the safety of
container creation. There are several feature macros associated with object
literals and subscripting: <code class="docutils literal"><span class="pre">__has_feature(objc_array_literals)</span></code> tests the
availability of array literals; <code class="docutils literal"><span class="pre">__has_feature(objc_dictionary_literals)</span></code>
tests the availability of dictionary literals;
<code class="docutils literal"><span class="pre">__has_feature(objc_subscripting)</span></code> tests the availability of object
subscripting.</p>
</div>
<div class="section" id="objective-c-autosynthesis-of-properties">
<h3>Objective-C Autosynthesis of Properties<a class="headerlink" href="#objective-c-autosynthesis-of-properties" title="Permalink to this headline">¶</a></h3>
<p>Clang provides support for autosynthesis of declared properties. Using this
feature, clang provides default synthesis of those properties not declared
@dynamic and not having user provided backing getter and setter methods.
<code class="docutils literal"><span class="pre">__has_feature(objc_default_synthesize_properties)</span></code> checks for availability
of this feature in version of clang being used.</p>
</div>
<div class="section" id="objective-c-retaining-behavior-attributes">
<span id="langext-objc-retain-release"></span><h3>Objective-C retaining behavior attributes<a class="headerlink" href="#objective-c-retaining-behavior-attributes" title="Permalink to this headline">¶</a></h3>
<p>In Objective-C, functions and methods are generally assumed to follow the
<a class="reference external" href="http://developer.apple.com/library/mac/#documentation/Cocoa/Conceptual/MemoryMgmt/Articles/mmRules.html">Cocoa Memory Management</a>
conventions for ownership of object arguments and
return values. However, there are exceptions, and so Clang provides attributes
to allow these exceptions to be documented. This are used by ARC and the
<a class="reference external" href="http://clang-analyzer.llvm.org">static analyzer</a> Some exceptions may be
better described using the <code class="docutils literal"><span class="pre">objc_method_family</span></code> attribute instead.</p>
<p><strong>Usage</strong>: The <code class="docutils literal"><span class="pre">ns_returns_retained</span></code>, <code class="docutils literal"><span class="pre">ns_returns_not_retained</span></code>,
<code class="docutils literal"><span class="pre">ns_returns_autoreleased</span></code>, <code class="docutils literal"><span class="pre">cf_returns_retained</span></code>, and
<code class="docutils literal"><span class="pre">cf_returns_not_retained</span></code> attributes can be placed on methods and functions
that return Objective-C or CoreFoundation objects. They are commonly placed at
the end of a function prototype or method declaration:</p>
<div class="highlight-objc"><div class="highlight"><pre><span class="kt">id</span> <span class="nf">foo</span><span class="p">()</span> <span class="n">__attribute__</span><span class="p">((</span><span class="n">ns_returns_retained</span><span class="p">));</span>
<span class="p">-</span> <span class="p">(</span><span class="bp">NSString</span> <span class="o">*</span><span class="p">)</span><span class="nf">bar:</span><span class="p">(</span><span class="kt">int</span><span class="p">)</span><span class="nv">x</span> <span class="n">__attribute__</span><span class="p">((</span><span class="n">ns_returns_retained</span><span class="p">));</span>
</pre></div>
</div>
<p>The <code class="docutils literal"><span class="pre">*_returns_retained</span></code> attributes specify that the returned object has a +1
retain count. The <code class="docutils literal"><span class="pre">*_returns_not_retained</span></code> attributes specify that the return
object has a +0 retain count, even if the normal convention for its selector
would be +1. <code class="docutils literal"><span class="pre">ns_returns_autoreleased</span></code> specifies that the returned object is
+0, but is guaranteed to live at least as long as the next flush of an
autorelease pool.</p>
<p><strong>Usage</strong>: The <code class="docutils literal"><span class="pre">ns_consumed</span></code> and <code class="docutils literal"><span class="pre">cf_consumed</span></code> attributes can be placed on
an parameter declaration; they specify that the argument is expected to have a
+1 retain count, which will be balanced in some way by the function or method.
The <code class="docutils literal"><span class="pre">ns_consumes_self</span></code> attribute can only be placed on an Objective-C
method; it specifies that the method expects its <code class="docutils literal"><span class="pre">self</span></code> parameter to have a
+1 retain count, which it will balance in some way.</p>
<div class="highlight-objc"><div class="highlight"><pre><span class="kt">void</span> <span class="nf">foo</span><span class="p">(</span><span class="n">__attribute__</span><span class="p">((</span><span class="n">ns_consumed</span><span class="p">))</span> <span class="bp">NSString</span> <span class="o">*</span><span class="n">string</span><span class="p">);</span>
<span class="p">-</span> <span class="p">(</span><span class="kt">void</span><span class="p">)</span> <span class="nf">bar</span> <span class="n">__attribute__</span><span class="p">((</span><span class="n">ns_consumes_self</span><span class="p">));</span>
<span class="p">-</span> <span class="p">(</span><span class="kt">void</span><span class="p">)</span> <span class="nf">baz:</span><span class="p">(</span><span class="kt">id</span><span class="p">)</span> <span class="nv">__attribute__</span><span class="p">((</span><span class="n">ns_consumed</span><span class="p">))</span> <span class="nv">x</span><span class="p">;</span>
</pre></div>
</div>
<p>Further examples of these attributes are available in the static analyzer’s <a class="reference external" href="http://clang-analyzer.llvm.org/annotations.html#cocoa_mem">list of annotations for analysis</a>.</p>
<p>Query for these features with <code class="docutils literal"><span class="pre">__has_attribute(ns_consumed)</span></code>,
<code class="docutils literal"><span class="pre">__has_attribute(ns_returns_retained)</span></code>, etc.</p>
</div>
<div class="section" id="objective-c-abi-protocol-qualifier-mangling-of-parameters">
<h3>Objective-C++ ABI: protocol-qualifier mangling of parameters<a class="headerlink" href="#objective-c-abi-protocol-qualifier-mangling-of-parameters" title="Permalink to this headline">¶</a></h3>
<p>Starting with LLVM 3.4, Clang produces a new mangling for parameters whose
type is a qualified-<code class="docutils literal"><span class="pre">id</span></code> (e.g., <code class="docutils literal"><span class="pre">id<Foo></span></code>). This mangling allows such
parameters to be differentiated from those with the regular unqualified <code class="docutils literal"><span class="pre">id</span></code>
type.</p>
<p>This was a non-backward compatible mangling change to the ABI. This change
allows proper overloading, and also prevents mangling conflicts with template
parameters of protocol-qualified type.</p>
<p>Query the presence of this new mangling with
<code class="docutils literal"><span class="pre">__has_feature(objc_protocol_qualifier_mangling)</span></code>.</p>
</div>
</div>
<div class="section" id="initializer-lists-for-complex-numbers-in-c">
<span id="langext-overloading"></span><h2><a class="toc-backref" href="#id13">Initializer lists for complex numbers in C</a><a class="headerlink" href="#initializer-lists-for-complex-numbers-in-c" title="Permalink to this headline">¶</a></h2>
<p>clang supports an extension which allows the following in C:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#include</span> <span class="cpf"><math.h></span><span class="cp"></span>
<span class="cp">#include</span> <span class="cpf"><complex.h></span><span class="cp"></span>
<span class="n">complex</span> <span class="kt">float</span> <span class="n">x</span> <span class="o">=</span> <span class="p">{</span> <span class="mf">1.0f</span><span class="p">,</span> <span class="n">INFINITY</span> <span class="p">};</span> <span class="c1">// Init to (1, Inf)</span>
</pre></div>
</div>
<p>This construct is useful because there is no way to separately initialize the
real and imaginary parts of a complex variable in standard C, given that clang
does not support <code class="docutils literal"><span class="pre">_Imaginary</span></code>. (Clang also supports the <code class="docutils literal"><span class="pre">__real__</span></code> and
<code class="docutils literal"><span class="pre">__imag__</span></code> extensions from gcc, which help in some cases, but are not usable
in static initializers.)</p>
<p>Note that this extension does not allow eliding the braces; the meaning of the
following two lines is different:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="n">complex</span> <span class="kt">float</span> <span class="n">x</span><span class="p">[]</span> <span class="o">=</span> <span class="p">{</span> <span class="p">{</span> <span class="mf">1.0f</span><span class="p">,</span> <span class="mf">1.0f</span> <span class="p">}</span> <span class="p">};</span> <span class="c1">// [0] = (1, 1)</span>
<span class="n">complex</span> <span class="kt">float</span> <span class="n">x</span><span class="p">[]</span> <span class="o">=</span> <span class="p">{</span> <span class="mf">1.0f</span><span class="p">,</span> <span class="mf">1.0f</span> <span class="p">};</span> <span class="c1">// [0] = (1, 0), [1] = (1, 0)</span>
</pre></div>
</div>
<p>This extension also works in C++ mode, as far as that goes, but does not apply
to the C++ <code class="docutils literal"><span class="pre">std::complex</span></code>. (In C++11, list initialization allows the same
syntax to be used with <code class="docutils literal"><span class="pre">std::complex</span></code> with the same meaning.)</p>
</div>
<div class="section" id="builtin-functions">
<h2><a class="toc-backref" href="#id14">Builtin Functions</a><a class="headerlink" href="#builtin-functions" title="Permalink to this headline">¶</a></h2>
<p>Clang supports a number of builtin library functions with the same syntax as
GCC, including things like <code class="docutils literal"><span class="pre">__builtin_nan</span></code>, <code class="docutils literal"><span class="pre">__builtin_constant_p</span></code>,
<code class="docutils literal"><span class="pre">__builtin_choose_expr</span></code>, <code class="docutils literal"><span class="pre">__builtin_types_compatible_p</span></code>,
<code class="docutils literal"><span class="pre">__sync_fetch_and_add</span></code>, etc. In addition to the GCC builtins, Clang supports
a number of builtins that GCC does not, which are listed here.</p>
<p>Please note that Clang does not and will not support all of the GCC builtins
for vector operations. Instead of using builtins, you should use the functions
defined in target-specific header files like <code class="docutils literal"><span class="pre"><xmmintrin.h></span></code>, which define
portable wrappers for these. Many of the Clang versions of these functions are
implemented directly in terms of <a class="reference internal" href="#langext-vectors"><span>extended vector support</span></a> instead of builtins, in order to reduce the number of
builtins that we need to implement.</p>
<div class="section" id="builtin-readcyclecounter">
<h3><code class="docutils literal"><span class="pre">__builtin_readcyclecounter</span></code><a class="headerlink" href="#builtin-readcyclecounter" title="Permalink to this headline">¶</a></h3>
<p><code class="docutils literal"><span class="pre">__builtin_readcyclecounter</span></code> is used to access the cycle counter register (or
a similar low-latency, high-accuracy clock) on those targets that support it.</p>
<p><strong>Syntax</strong>:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="n">__builtin_readcyclecounter</span><span class="p">()</span>
</pre></div>
</div>
<p><strong>Example of Use</strong>:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">t0</span> <span class="o">=</span> <span class="n">__builtin_readcyclecounter</span><span class="p">();</span>
<span class="n">do_something</span><span class="p">();</span>
<span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">t1</span> <span class="o">=</span> <span class="n">__builtin_readcyclecounter</span><span class="p">();</span>
<span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">cycles_to_do_something</span> <span class="o">=</span> <span class="n">t1</span> <span class="o">-</span> <span class="n">t0</span><span class="p">;</span> <span class="c1">// assuming no overflow</span>
</pre></div>
</div>
<p><strong>Description</strong>:</p>
<p>The <code class="docutils literal"><span class="pre">__builtin_readcyclecounter()</span></code> builtin returns the cycle counter value,
which may be either global or process/thread-specific depending on the target.
As the backing counters often overflow quickly (on the order of seconds) this
should only be used for timing small intervals. When not supported by the
target, the return value is always zero. This builtin takes no arguments and
produces an unsigned long long result.</p>
<p>Query for this feature with <code class="docutils literal"><span class="pre">__has_builtin(__builtin_readcyclecounter)</span></code>. Note
that even if present, its use may depend on run-time privilege or other OS
controlled state.</p>
</div>
<div class="section" id="builtin-shufflevector">
<span id="langext-builtin-shufflevector"></span><h3><code class="docutils literal"><span class="pre">__builtin_shufflevector</span></code><a class="headerlink" href="#builtin-shufflevector" title="Permalink to this headline">¶</a></h3>
<p><code class="docutils literal"><span class="pre">__builtin_shufflevector</span></code> is used to express generic vector
permutation/shuffle/swizzle operations. This builtin is also very important
for the implementation of various target-specific header files like
<code class="docutils literal"><span class="pre"><xmmintrin.h></span></code>.</p>
<p><strong>Syntax</strong>:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="n">__builtin_shufflevector</span><span class="p">(</span><span class="n">vec1</span><span class="p">,</span> <span class="n">vec2</span><span class="p">,</span> <span class="n">index1</span><span class="p">,</span> <span class="n">index2</span><span class="p">,</span> <span class="p">...)</span>
</pre></div>
</div>
<p><strong>Examples</strong>:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="c1">// identity operation - return 4-element vector v1.</span>
<span class="n">__builtin_shufflevector</span><span class="p">(</span><span class="n">v1</span><span class="p">,</span> <span class="n">v1</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">)</span>
<span class="c1">// "Splat" element 0 of V1 into a 4-element result.</span>
<span class="n">__builtin_shufflevector</span><span class="p">(</span><span class="n">V1</span><span class="p">,</span> <span class="n">V1</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">0</span><span class="p">)</span>
<span class="c1">// Reverse 4-element vector V1.</span>
<span class="n">__builtin_shufflevector</span><span class="p">(</span><span class="n">V1</span><span class="p">,</span> <span class="n">V1</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">0</span><span class="p">)</span>
<span class="c1">// Concatenate every other element of 4-element vectors V1 and V2.</span>
<span class="n">__builtin_shufflevector</span><span class="p">(</span><span class="n">V1</span><span class="p">,</span> <span class="n">V2</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">6</span><span class="p">)</span>
<span class="c1">// Concatenate every other element of 8-element vectors V1 and V2.</span>
<span class="n">__builtin_shufflevector</span><span class="p">(</span><span class="n">V1</span><span class="p">,</span> <span class="n">V2</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">6</span><span class="p">,</span> <span class="mi">8</span><span class="p">,</span> <span class="mi">10</span><span class="p">,</span> <span class="mi">12</span><span class="p">,</span> <span class="mi">14</span><span class="p">)</span>
<span class="c1">// Shuffle v1 with some elements being undefined</span>
<span class="n">__builtin_shufflevector</span><span class="p">(</span><span class="n">v1</span><span class="p">,</span> <span class="n">v1</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">)</span>
</pre></div>
</div>
<p><strong>Description</strong>:</p>
<p>The first two arguments to <code class="docutils literal"><span class="pre">__builtin_shufflevector</span></code> are vectors that have
the same element type. The remaining arguments are a list of integers that
specify the elements indices of the first two vectors that should be extracted
and returned in a new vector. These element indices are numbered sequentially
starting with the first vector, continuing into the second vector. Thus, if
<code class="docutils literal"><span class="pre">vec1</span></code> is a 4-element vector, index 5 would refer to the second element of
<code class="docutils literal"><span class="pre">vec2</span></code>. An index of -1 can be used to indicate that the corresponding element
in the returned vector is a don’t care and can be optimized by the backend.</p>
<p>The result of <code class="docutils literal"><span class="pre">__builtin_shufflevector</span></code> is a vector with the same element
type as <code class="docutils literal"><span class="pre">vec1</span></code>/<code class="docutils literal"><span class="pre">vec2</span></code> but that has an element count equal to the number of
indices specified.</p>
<p>Query for this feature with <code class="docutils literal"><span class="pre">__has_builtin(__builtin_shufflevector)</span></code>.</p>
</div>
<div class="section" id="builtin-convertvector">
<h3><code class="docutils literal"><span class="pre">__builtin_convertvector</span></code><a class="headerlink" href="#builtin-convertvector" title="Permalink to this headline">¶</a></h3>
<p><code class="docutils literal"><span class="pre">__builtin_convertvector</span></code> is used to express generic vector
type-conversion operations. The input vector and the output vector
type must have the same number of elements.</p>
<p><strong>Syntax</strong>:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="n">__builtin_convertvector</span><span class="p">(</span><span class="n">src_vec</span><span class="p">,</span> <span class="n">dst_vec_type</span><span class="p">)</span>
</pre></div>
</div>
<p><strong>Examples</strong>:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="k">typedef</span> <span class="kt">double</span> <span class="n">vector4double</span> <span class="nf">__attribute__</span><span class="p">((</span><span class="n">__vector_size__</span><span class="p">(</span><span class="mi">32</span><span class="p">)));</span>
<span class="k">typedef</span> <span class="kt">float</span> <span class="n">vector4float</span> <span class="nf">__attribute__</span><span class="p">((</span><span class="n">__vector_size__</span><span class="p">(</span><span class="mi">16</span><span class="p">)));</span>
<span class="k">typedef</span> <span class="kt">short</span> <span class="n">vector4short</span> <span class="nf">__attribute__</span><span class="p">((</span><span class="n">__vector_size__</span><span class="p">(</span><span class="mi">8</span><span class="p">)));</span>
<span class="n">vector4float</span> <span class="n">vf</span><span class="p">;</span> <span class="n">vector4short</span> <span class="n">vs</span><span class="p">;</span>
<span class="c1">// convert from a vector of 4 floats to a vector of 4 doubles.</span>
<span class="n">__builtin_convertvector</span><span class="p">(</span><span class="n">vf</span><span class="p">,</span> <span class="n">vector4double</span><span class="p">)</span>
<span class="c1">// equivalent to:</span>
<span class="p">(</span><span class="n">vector4double</span><span class="p">)</span> <span class="p">{</span> <span class="p">(</span><span class="kt">double</span><span class="p">)</span> <span class="n">vf</span><span class="p">[</span><span class="mi">0</span><span class="p">],</span> <span class="p">(</span><span class="kt">double</span><span class="p">)</span> <span class="n">vf</span><span class="p">[</span><span class="mi">1</span><span class="p">],</span> <span class="p">(</span><span class="kt">double</span><span class="p">)</span> <span class="n">vf</span><span class="p">[</span><span class="mi">2</span><span class="p">],</span> <span class="p">(</span><span class="kt">double</span><span class="p">)</span> <span class="n">vf</span><span class="p">[</span><span class="mi">3</span><span class="p">]</span> <span class="p">}</span>
<span class="c1">// convert from a vector of 4 shorts to a vector of 4 floats.</span>
<span class="n">__builtin_convertvector</span><span class="p">(</span><span class="n">vs</span><span class="p">,</span> <span class="n">vector4float</span><span class="p">)</span>
<span class="c1">// equivalent to:</span>
<span class="p">(</span><span class="n">vector4float</span><span class="p">)</span> <span class="p">{</span> <span class="p">(</span><span class="kt">float</span><span class="p">)</span> <span class="n">vf</span><span class="p">[</span><span class="mi">0</span><span class="p">],</span> <span class="p">(</span><span class="kt">float</span><span class="p">)</span> <span class="n">vf</span><span class="p">[</span><span class="mi">1</span><span class="p">],</span> <span class="p">(</span><span class="kt">float</span><span class="p">)</span> <span class="n">vf</span><span class="p">[</span><span class="mi">2</span><span class="p">],</span> <span class="p">(</span><span class="kt">float</span><span class="p">)</span> <span class="n">vf</span><span class="p">[</span><span class="mi">3</span><span class="p">]</span> <span class="p">}</span>
</pre></div>
</div>
<p><strong>Description</strong>:</p>
<p>The first argument to <code class="docutils literal"><span class="pre">__builtin_convertvector</span></code> is a vector, and the second
argument is a vector type with the same number of elements as the first
argument.</p>
<p>The result of <code class="docutils literal"><span class="pre">__builtin_convertvector</span></code> is a vector with the same element
type as the second argument, with a value defined in terms of the action of a
C-style cast applied to each element of the first argument.</p>
<p>Query for this feature with <code class="docutils literal"><span class="pre">__has_builtin(__builtin_convertvector)</span></code>.</p>
</div>
<div class="section" id="builtin-unreachable">
<h3><code class="docutils literal"><span class="pre">__builtin_unreachable</span></code><a class="headerlink" href="#builtin-unreachable" title="Permalink to this headline">¶</a></h3>
<p><code class="docutils literal"><span class="pre">__builtin_unreachable</span></code> is used to indicate that a specific point in the
program cannot be reached, even if the compiler might otherwise think it can.
This is useful to improve optimization and eliminates certain warnings. For
example, without the <code class="docutils literal"><span class="pre">__builtin_unreachable</span></code> in the example below, the
compiler assumes that the inline asm can fall through and prints a “function
declared ‘<code class="docutils literal"><span class="pre">noreturn</span></code>‘ should not return” warning.</p>
<p><strong>Syntax</strong>:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="n">__builtin_unreachable</span><span class="p">()</span>
</pre></div>
</div>
<p><strong>Example of use</strong>:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="kt">void</span> <span class="nf">myabort</span><span class="p">(</span><span class="kt">void</span><span class="p">)</span> <span class="n">__attribute__</span><span class="p">((</span><span class="n">noreturn</span><span class="p">));</span>
<span class="kt">void</span> <span class="nf">myabort</span><span class="p">(</span><span class="kt">void</span><span class="p">)</span> <span class="p">{</span>
<span class="k">asm</span><span class="p">(</span><span class="s">"int3"</span><span class="p">);</span>
<span class="n">__builtin_unreachable</span><span class="p">();</span>
<span class="p">}</span>
</pre></div>
</div>
<p><strong>Description</strong>:</p>
<p>The <code class="docutils literal"><span class="pre">__builtin_unreachable()</span></code> builtin has completely undefined behavior.
Since it has undefined behavior, it is a statement that it is never reached and
the optimizer can take advantage of this to produce better code. This builtin
takes no arguments and produces a void result.</p>
<p>Query for this feature with <code class="docutils literal"><span class="pre">__has_builtin(__builtin_unreachable)</span></code>.</p>
</div>
<div class="section" id="sync-swap">
<h3><code class="docutils literal"><span class="pre">__sync_swap</span></code><a class="headerlink" href="#sync-swap" title="Permalink to this headline">¶</a></h3>
<p><code class="docutils literal"><span class="pre">__sync_swap</span></code> is used to atomically swap integers or pointers in memory.</p>
<p><strong>Syntax</strong>:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="n">type</span> <span class="n">__sync_swap</span><span class="p">(</span><span class="n">type</span> <span class="o">*</span><span class="n">ptr</span><span class="p">,</span> <span class="n">type</span> <span class="n">value</span><span class="p">,</span> <span class="p">...)</span>
</pre></div>
</div>
<p><strong>Example of Use</strong>:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="kt">int</span> <span class="n">old_value</span> <span class="o">=</span> <span class="n">__sync_swap</span><span class="p">(</span><span class="o">&</span><span class="n">value</span><span class="p">,</span> <span class="n">new_value</span><span class="p">);</span>
</pre></div>
</div>
<p><strong>Description</strong>:</p>
<p>The <code class="docutils literal"><span class="pre">__sync_swap()</span></code> builtin extends the existing <code class="docutils literal"><span class="pre">__sync_*()</span></code> family of
atomic intrinsics to allow code to atomically swap the current value with the
new value. More importantly, it helps developers write more efficient and
correct code by avoiding expensive loops around
<code class="docutils literal"><span class="pre">__sync_bool_compare_and_swap()</span></code> or relying on the platform specific
implementation details of <code class="docutils literal"><span class="pre">__sync_lock_test_and_set()</span></code>. The
<code class="docutils literal"><span class="pre">__sync_swap()</span></code> builtin is a full barrier.</p>
</div>
<div class="section" id="builtin-addressof">
<h3><code class="docutils literal"><span class="pre">__builtin_addressof</span></code><a class="headerlink" href="#builtin-addressof" title="Permalink to this headline">¶</a></h3>
<p><code class="docutils literal"><span class="pre">__builtin_addressof</span></code> performs the functionality of the built-in <code class="docutils literal"><span class="pre">&</span></code>
operator, ignoring any <code class="docutils literal"><span class="pre">operator&</span></code> overload. This is useful in constant
expressions in C++11, where there is no other way to take the address of an
object that overloads <code class="docutils literal"><span class="pre">operator&</span></code>.</p>
<p><strong>Example of use</strong>:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="k">template</span><span class="o"><</span><span class="k">typename</span> <span class="n">T</span><span class="o">></span> <span class="k">constexpr</span> <span class="n">T</span> <span class="o">*</span><span class="n">addressof</span><span class="p">(</span><span class="n">T</span> <span class="o">&</span><span class="n">value</span><span class="p">)</span> <span class="p">{</span>
<span class="k">return</span> <span class="n">__builtin_addressof</span><span class="p">(</span><span class="n">value</span><span class="p">);</span>
<span class="p">}</span>
</pre></div>
</div>
</div>
<div class="section" id="builtin-operator-new-and-builtin-operator-delete">
<h3><code class="docutils literal"><span class="pre">__builtin_operator_new</span></code> and <code class="docutils literal"><span class="pre">__builtin_operator_delete</span></code><a class="headerlink" href="#builtin-operator-new-and-builtin-operator-delete" title="Permalink to this headline">¶</a></h3>
<p><code class="docutils literal"><span class="pre">__builtin_operator_new</span></code> allocates memory just like a non-placement non-class
<em>new-expression</em>. This is exactly like directly calling the normal
non-placement <code class="docutils literal"><span class="pre">::operator</span> <span class="pre">new</span></code>, except that it allows certain optimizations
that the C++ standard does not permit for a direct function call to
<code class="docutils literal"><span class="pre">::operator</span> <span class="pre">new</span></code> (in particular, removing <code class="docutils literal"><span class="pre">new</span></code> / <code class="docutils literal"><span class="pre">delete</span></code> pairs and
merging allocations).</p>
<p>Likewise, <code class="docutils literal"><span class="pre">__builtin_operator_delete</span></code> deallocates memory just like a
non-class <em>delete-expression</em>, and is exactly like directly calling the normal
<code class="docutils literal"><span class="pre">::operator</span> <span class="pre">delete</span></code>, except that it permits optimizations. Only the unsized
form of <code class="docutils literal"><span class="pre">__builtin_operator_delete</span></code> is currently available.</p>
<p>These builtins are intended for use in the implementation of <code class="docutils literal"><span class="pre">std::allocator</span></code>
and other similar allocation libraries, and are only available in C++.</p>
</div>
<div class="section" id="multiprecision-arithmetic-builtins">
<h3>Multiprecision Arithmetic Builtins<a class="headerlink" href="#multiprecision-arithmetic-builtins" title="Permalink to this headline">¶</a></h3>
<p>Clang provides a set of builtins which expose multiprecision arithmetic in a
manner amenable to C. They all have the following form:</p>
<div class="highlight-c"><div class="highlight"><pre><span class="kt">unsigned</span> <span class="n">x</span> <span class="o">=</span> <span class="p">...,</span> <span class="n">y</span> <span class="o">=</span> <span class="p">...,</span> <span class="n">carryin</span> <span class="o">=</span> <span class="p">...,</span> <span class="n">carryout</span><span class="p">;</span>
<span class="kt">unsigned</span> <span class="n">sum</span> <span class="o">=</span> <span class="n">__builtin_addc</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="n">y</span><span class="p">,</span> <span class="n">carryin</span><span class="p">,</span> <span class="o">&</span><span class="n">carryout</span><span class="p">);</span>
</pre></div>
</div>
<p>Thus one can form a multiprecision addition chain in the following manner:</p>
<div class="highlight-c"><div class="highlight"><pre><span class="kt">unsigned</span> <span class="o">*</span><span class="n">x</span><span class="p">,</span> <span class="o">*</span><span class="n">y</span><span class="p">,</span> <span class="o">*</span><span class="n">z</span><span class="p">,</span> <span class="n">carryin</span><span class="o">=</span><span class="mi">0</span><span class="p">,</span> <span class="n">carryout</span><span class="p">;</span>
<span class="n">z</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span> <span class="o">=</span> <span class="n">__builtin_addc</span><span class="p">(</span><span class="n">x</span><span class="p">[</span><span class="mi">0</span><span class="p">],</span> <span class="n">y</span><span class="p">[</span><span class="mi">0</span><span class="p">],</span> <span class="n">carryin</span><span class="p">,</span> <span class="o">&</span><span class="n">carryout</span><span class="p">);</span>
<span class="n">carryin</span> <span class="o">=</span> <span class="n">carryout</span><span class="p">;</span>
<span class="n">z</span><span class="p">[</span><span class="mi">1</span><span class="p">]</span> <span class="o">=</span> <span class="n">__builtin_addc</span><span class="p">(</span><span class="n">x</span><span class="p">[</span><span class="mi">1</span><span class="p">],</span> <span class="n">y</span><span class="p">[</span><span class="mi">1</span><span class="p">],</span> <span class="n">carryin</span><span class="p">,</span> <span class="o">&</span><span class="n">carryout</span><span class="p">);</span>
<span class="n">carryin</span> <span class="o">=</span> <span class="n">carryout</span><span class="p">;</span>
<span class="n">z</span><span class="p">[</span><span class="mi">2</span><span class="p">]</span> <span class="o">=</span> <span class="n">__builtin_addc</span><span class="p">(</span><span class="n">x</span><span class="p">[</span><span class="mi">2</span><span class="p">],</span> <span class="n">y</span><span class="p">[</span><span class="mi">2</span><span class="p">],</span> <span class="n">carryin</span><span class="p">,</span> <span class="o">&</span><span class="n">carryout</span><span class="p">);</span>
<span class="n">carryin</span> <span class="o">=</span> <span class="n">carryout</span><span class="p">;</span>
<span class="n">z</span><span class="p">[</span><span class="mi">3</span><span class="p">]</span> <span class="o">=</span> <span class="n">__builtin_addc</span><span class="p">(</span><span class="n">x</span><span class="p">[</span><span class="mi">3</span><span class="p">],</span> <span class="n">y</span><span class="p">[</span><span class="mi">3</span><span class="p">],</span> <span class="n">carryin</span><span class="p">,</span> <span class="o">&</span><span class="n">carryout</span><span class="p">);</span>
</pre></div>
</div>
<p>The complete list of builtins are:</p>
<div class="highlight-c"><div class="highlight"><pre><span class="kt">unsigned</span> <span class="kt">char</span> <span class="nf">__builtin_addcb</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="kt">char</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">char</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">char</span> <span class="n">carryin</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">char</span> <span class="o">*</span><span class="n">carryout</span><span class="p">);</span>
<span class="kt">unsigned</span> <span class="kt">short</span> <span class="nf">__builtin_addcs</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="kt">short</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">short</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">short</span> <span class="n">carryin</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">short</span> <span class="o">*</span><span class="n">carryout</span><span class="p">);</span>
<span class="kt">unsigned</span> <span class="nf">__builtin_addc</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="n">carryin</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="o">*</span><span class="n">carryout</span><span class="p">);</span>
<span class="kt">unsigned</span> <span class="kt">long</span> <span class="nf">__builtin_addcl</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="n">carryin</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="o">*</span><span class="n">carryout</span><span class="p">);</span>
<span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="nf">__builtin_addcll</span><span class="p">(</span><span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">carryin</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="o">*</span><span class="n">carryout</span><span class="p">);</span>
<span class="kt">unsigned</span> <span class="kt">char</span> <span class="nf">__builtin_subcb</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="kt">char</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">char</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">char</span> <span class="n">carryin</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">char</span> <span class="o">*</span><span class="n">carryout</span><span class="p">);</span>
<span class="kt">unsigned</span> <span class="kt">short</span> <span class="nf">__builtin_subcs</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="kt">short</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">short</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">short</span> <span class="n">carryin</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">short</span> <span class="o">*</span><span class="n">carryout</span><span class="p">);</span>
<span class="kt">unsigned</span> <span class="nf">__builtin_subc</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="n">carryin</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="o">*</span><span class="n">carryout</span><span class="p">);</span>
<span class="kt">unsigned</span> <span class="kt">long</span> <span class="nf">__builtin_subcl</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="n">carryin</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="o">*</span><span class="n">carryout</span><span class="p">);</span>
<span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="nf">__builtin_subcll</span><span class="p">(</span><span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">carryin</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="o">*</span><span class="n">carryout</span><span class="p">);</span>
</pre></div>
</div>
</div>
<div class="section" id="checked-arithmetic-builtins">
<h3>Checked Arithmetic Builtins<a class="headerlink" href="#checked-arithmetic-builtins" title="Permalink to this headline">¶</a></h3>
<p>Clang provides a set of builtins that implement checked arithmetic for security
critical applications in a manner that is fast and easily expressable in C. As
an example of their usage:</p>
<div class="highlight-c"><div class="highlight"><pre><span class="n">errorcode_t</span> <span class="nf">security_critical_application</span><span class="p">(...)</span> <span class="p">{</span>
<span class="kt">unsigned</span> <span class="n">x</span><span class="p">,</span> <span class="n">y</span><span class="p">,</span> <span class="n">result</span><span class="p">;</span>
<span class="p">...</span>
<span class="k">if</span> <span class="p">(</span><span class="n">__builtin_umul_overflow</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="n">y</span><span class="p">,</span> <span class="o">&</span><span class="n">result</span><span class="p">))</span>
<span class="k">return</span> <span class="n">kErrorCodeHackers</span><span class="p">;</span>
<span class="p">...</span>
<span class="n">use_multiply</span><span class="p">(</span><span class="n">result</span><span class="p">);</span>
<span class="p">...</span>
<span class="p">}</span>
</pre></div>
</div>
<p>A complete enumeration of the builtins are:</p>
<div class="highlight-c"><div class="highlight"><pre><span class="kt">bool</span> <span class="nf">__builtin_uadd_overflow</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="o">*</span><span class="n">sum</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_uaddl_overflow</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="o">*</span><span class="n">sum</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_uaddll_overflow</span><span class="p">(</span><span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="o">*</span><span class="n">sum</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_usub_overflow</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="o">*</span><span class="n">diff</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_usubl_overflow</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="o">*</span><span class="n">diff</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_usubll_overflow</span><span class="p">(</span><span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="o">*</span><span class="n">diff</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_umul_overflow</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="o">*</span><span class="n">prod</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_umull_overflow</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="o">*</span><span class="n">prod</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_umulll_overflow</span><span class="p">(</span><span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="kt">long</span> <span class="kt">long</span> <span class="o">*</span><span class="n">prod</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_sadd_overflow</span> <span class="p">(</span><span class="kt">int</span> <span class="n">x</span><span class="p">,</span> <span class="kt">int</span> <span class="n">y</span><span class="p">,</span> <span class="kt">int</span> <span class="o">*</span><span class="n">sum</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_saddl_overflow</span> <span class="p">(</span><span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">long</span> <span class="o">*</span><span class="n">sum</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_saddll_overflow</span><span class="p">(</span><span class="kt">long</span> <span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">long</span> <span class="kt">long</span> <span class="o">*</span><span class="n">sum</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_ssub_overflow</span> <span class="p">(</span><span class="kt">int</span> <span class="n">x</span><span class="p">,</span> <span class="kt">int</span> <span class="n">y</span><span class="p">,</span> <span class="kt">int</span> <span class="o">*</span><span class="n">diff</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_ssubl_overflow</span> <span class="p">(</span><span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">long</span> <span class="o">*</span><span class="n">diff</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_ssubll_overflow</span><span class="p">(</span><span class="kt">long</span> <span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">long</span> <span class="kt">long</span> <span class="o">*</span><span class="n">diff</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_smul_overflow</span> <span class="p">(</span><span class="kt">int</span> <span class="n">x</span><span class="p">,</span> <span class="kt">int</span> <span class="n">y</span><span class="p">,</span> <span class="kt">int</span> <span class="o">*</span><span class="n">prod</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_smull_overflow</span> <span class="p">(</span><span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">long</span> <span class="o">*</span><span class="n">prod</span><span class="p">);</span>
<span class="kt">bool</span> <span class="nf">__builtin_smulll_overflow</span><span class="p">(</span><span class="kt">long</span> <span class="kt">long</span> <span class="n">x</span><span class="p">,</span> <span class="kt">long</span> <span class="kt">long</span> <span class="n">y</span><span class="p">,</span> <span class="kt">long</span> <span class="kt">long</span> <span class="o">*</span><span class="n">prod</span><span class="p">);</span>
</pre></div>
</div>
</div>
<div class="section" id="c11-atomic-builtins">
<span id="langext-c11-atomic"></span><h3>__c11_atomic builtins<a class="headerlink" href="#c11-atomic-builtins" title="Permalink to this headline">¶</a></h3>
<p>Clang provides a set of builtins which are intended to be used to implement
C11’s <code class="docutils literal"><span class="pre"><stdatomic.h></span></code> header. These builtins provide the semantics of the
<code class="docutils literal"><span class="pre">_explicit</span></code> form of the corresponding C11 operation, and are named with a
<code class="docutils literal"><span class="pre">__c11_</span></code> prefix. The supported operations are:</p>
<ul class="simple">
<li><code class="docutils literal"><span class="pre">__c11_atomic_init</span></code></li>
<li><code class="docutils literal"><span class="pre">__c11_atomic_thread_fence</span></code></li>
<li><code class="docutils literal"><span class="pre">__c11_atomic_signal_fence</span></code></li>
<li><code class="docutils literal"><span class="pre">__c11_atomic_is_lock_free</span></code></li>
<li><code class="docutils literal"><span class="pre">__c11_atomic_store</span></code></li>
<li><code class="docutils literal"><span class="pre">__c11_atomic_load</span></code></li>
<li><code class="docutils literal"><span class="pre">__c11_atomic_exchange</span></code></li>
<li><code class="docutils literal"><span class="pre">__c11_atomic_compare_exchange_strong</span></code></li>
<li><code class="docutils literal"><span class="pre">__c11_atomic_compare_exchange_weak</span></code></li>
<li><code class="docutils literal"><span class="pre">__c11_atomic_fetch_add</span></code></li>
<li><code class="docutils literal"><span class="pre">__c11_atomic_fetch_sub</span></code></li>
<li><code class="docutils literal"><span class="pre">__c11_atomic_fetch_and</span></code></li>
<li><code class="docutils literal"><span class="pre">__c11_atomic_fetch_or</span></code></li>
<li><code class="docutils literal"><span class="pre">__c11_atomic_fetch_xor</span></code></li>
</ul>
</div>
<div class="section" id="low-level-arm-exclusive-memory-builtins">
<h3>Low-level ARM exclusive memory builtins<a class="headerlink" href="#low-level-arm-exclusive-memory-builtins" title="Permalink to this headline">¶</a></h3>
<p>Clang provides overloaded builtins giving direct access to the three key ARM
instructions for implementing atomic operations.</p>
<div class="highlight-c"><div class="highlight"><pre><span class="n">T</span> <span class="nf">__builtin_arm_ldrex</span><span class="p">(</span><span class="k">const</span> <span class="k">volatile</span> <span class="n">T</span> <span class="o">*</span><span class="n">addr</span><span class="p">);</span>
<span class="n">T</span> <span class="nf">__builtin_arm_ldaex</span><span class="p">(</span><span class="k">const</span> <span class="k">volatile</span> <span class="n">T</span> <span class="o">*</span><span class="n">addr</span><span class="p">);</span>
<span class="kt">int</span> <span class="nf">__builtin_arm_strex</span><span class="p">(</span><span class="n">T</span> <span class="n">val</span><span class="p">,</span> <span class="k">volatile</span> <span class="n">T</span> <span class="o">*</span><span class="n">addr</span><span class="p">);</span>
<span class="kt">int</span> <span class="nf">__builtin_arm_stlex</span><span class="p">(</span><span class="n">T</span> <span class="n">val</span><span class="p">,</span> <span class="k">volatile</span> <span class="n">T</span> <span class="o">*</span><span class="n">addr</span><span class="p">);</span>
<span class="kt">void</span> <span class="nf">__builtin_arm_clrex</span><span class="p">(</span><span class="kt">void</span><span class="p">);</span>
</pre></div>
</div>
<p>The types <code class="docutils literal"><span class="pre">T</span></code> currently supported are:
* Integer types with width at most 64 bits (or 128 bits on AArch64).
* Floating-point types
* Pointer types.</p>
<p>Note that the compiler does not guarantee it will not insert stores which clear
the exclusive monitor in between an <code class="docutils literal"><span class="pre">ldrex</span></code> type operation and its paired
<code class="docutils literal"><span class="pre">strex</span></code>. In practice this is only usually a risk when the extra store is on
the same cache line as the variable being modified and Clang will only insert
stack stores on its own, so it is best not to use these operations on variables
with automatic storage duration.</p>
<p>Also, loads and stores may be implicit in code written between the <code class="docutils literal"><span class="pre">ldrex</span></code> and
<code class="docutils literal"><span class="pre">strex</span></code>. Clang will not necessarily mitigate the effects of these either, so
care should be exercised.</p>
<p>For these reasons the higher level atomic primitives should be preferred where
possible.</p>
</div>
</div>
<div class="section" id="non-standard-c-11-attributes">
<h2><a class="toc-backref" href="#id15">Non-standard C++11 Attributes</a><a class="headerlink" href="#non-standard-c-11-attributes" title="Permalink to this headline">¶</a></h2>
<p>Clang’s non-standard C++11 attributes live in the <code class="docutils literal"><span class="pre">clang</span></code> attribute
namespace.</p>
<p>Clang supports GCC’s <code class="docutils literal"><span class="pre">gnu</span></code> attribute namespace. All GCC attributes which
are accepted with the <code class="docutils literal"><span class="pre">__attribute__((foo))</span></code> syntax are also accepted as
<code class="docutils literal"><span class="pre">[[gnu::foo]]</span></code>. This only extends to attributes which are specified by GCC
(see the list of <a class="reference external" href="http://gcc.gnu.org/onlinedocs/gcc/Function-Attributes.html">GCC function attributes</a>, <a class="reference external" href="http://gcc.gnu.org/onlinedocs/gcc/Variable-Attributes.html">GCC variable
attributes</a>, and
<a class="reference external" href="http://gcc.gnu.org/onlinedocs/gcc/Type-Attributes.html">GCC type attributes</a>). As with the GCC
implementation, these attributes must appertain to the <em>declarator-id</em> in a
declaration, which means they must go either at the start of the declaration or
immediately after the name being declared.</p>
<p>For example, this applies the GNU <code class="docutils literal"><span class="pre">unused</span></code> attribute to <code class="docutils literal"><span class="pre">a</span></code> and <code class="docutils literal"><span class="pre">f</span></code>, and
also applies the GNU <code class="docutils literal"><span class="pre">noreturn</span></code> attribute to <code class="docutils literal"><span class="pre">f</span></code>.</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="p">[[</span><span class="n">gnu</span><span class="o">::</span><span class="n">unused</span><span class="p">]]</span> <span class="kt">int</span> <span class="n">a</span><span class="p">,</span> <span class="n">f</span> <span class="p">[[</span><span class="n">gnu</span><span class="o">::</span><span class="n">noreturn</span><span class="p">]]</span> <span class="p">();</span>
</pre></div>
</div>
</div>
<div class="section" id="target-specific-extensions">
<h2><a class="toc-backref" href="#id16">Target-Specific Extensions</a><a class="headerlink" href="#target-specific-extensions" title="Permalink to this headline">¶</a></h2>
<p>Clang supports some language features conditionally on some targets.</p>
<div class="section" id="arm-aarch64-language-extensions">
<h3>ARM/AArch64 Language Extensions<a class="headerlink" href="#arm-aarch64-language-extensions" title="Permalink to this headline">¶</a></h3>
<div class="section" id="memory-barrier-intrinsics">
<h4>Memory Barrier Intrinsics<a class="headerlink" href="#memory-barrier-intrinsics" title="Permalink to this headline">¶</a></h4>
<p>Clang implements the <code class="docutils literal"><span class="pre">__dmb</span></code>, <code class="docutils literal"><span class="pre">__dsb</span></code> and <code class="docutils literal"><span class="pre">__isb</span></code> intrinsics as defined
in the <a class="reference external" href="http://infocenter.arm.com/help/topic/com.arm.doc.ihi0053c/IHI0053C_acle_2_0.pdf">ARM C Language Extensions Release 2.0</a>.
Note that these intrinsics are implemented as motion barriers that block
reordering of memory accesses and side effect instructions. Other instructions
like simple arithmatic may be reordered around the intrinsic. If you expect to
have no reordering at all, use inline assembly instead.</p>
</div>
</div>
<div class="section" id="x86-x86-64-language-extensions">
<h3>X86/X86-64 Language Extensions<a class="headerlink" href="#x86-x86-64-language-extensions" title="Permalink to this headline">¶</a></h3>
<p>The X86 backend has these language extensions:</p>
<div class="section" id="memory-references-off-the-gs-segment">
<h4>Memory references off the GS segment<a class="headerlink" href="#memory-references-off-the-gs-segment" title="Permalink to this headline">¶</a></h4>
<p>Annotating a pointer with address space #256 causes it to be code generated
relative to the X86 GS segment register, and address space #257 causes it to be
relative to the X86 FS segment. Note that this is a very very low-level
feature that should only be used if you know what you’re doing (for example in
an OS kernel).</p>
<p>Here is an example:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#define GS_RELATIVE __attribute__((address_space(256)))</span>
<span class="kt">int</span> <span class="nf">foo</span><span class="p">(</span><span class="kt">int</span> <span class="n">GS_RELATIVE</span> <span class="o">*</span><span class="n">P</span><span class="p">)</span> <span class="p">{</span>
<span class="k">return</span> <span class="o">*</span><span class="n">P</span><span class="p">;</span>
<span class="p">}</span>
</pre></div>
</div>
<p>Which compiles to (on X86-32):</p>
<div class="highlight-gas"><div class="highlight"><pre><span class="nl">_foo:</span>
<span class="nf">movl</span> <span class="mi">4</span><span class="p">(</span><span class="nv">%esp</span><span class="p">),</span> <span class="nv">%eax</span>
<span class="nf">movl</span> <span class="nv">%gs</span><span class="p">:(</span><span class="nv">%eax</span><span class="p">),</span> <span class="nv">%eax</span>
<span class="nf">ret</span>
</pre></div>
</div>
</div>
</div>
</div>
<div class="section" id="extensions-for-static-analysis">
<h2><a class="toc-backref" href="#id17">Extensions for Static Analysis</a><a class="headerlink" href="#extensions-for-static-analysis" title="Permalink to this headline">¶</a></h2>
<p>Clang supports additional attributes that are useful for documenting program
invariants and rules for static analysis tools, such as the <a class="reference external" href="http://clang-analyzer.llvm.org/">Clang Static
Analyzer</a>. These attributes are documented
in the analyzer’s <a class="reference external" href="http://clang-analyzer.llvm.org/annotations.html">list of source-level annotations</a>.</p>
</div>
<div class="section" id="extensions-for-dynamic-analysis">
<h2><a class="toc-backref" href="#id18">Extensions for Dynamic Analysis</a><a class="headerlink" href="#extensions-for-dynamic-analysis" title="Permalink to this headline">¶</a></h2>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(address_sanitizer)</span></code> to check if the code is being built
with <a class="reference internal" href="AddressSanitizer.html"><em>AddressSanitizer</em></a>.</p>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(thread_sanitizer)</span></code> to check if the code is being built
with <a class="reference internal" href="ThreadSanitizer.html"><em>ThreadSanitizer</em></a>.</p>
<p>Use <code class="docutils literal"><span class="pre">__has_feature(memory_sanitizer)</span></code> to check if the code is being built
with <a class="reference internal" href="MemorySanitizer.html"><em>MemorySanitizer</em></a>.</p>
</div>
<div class="section" id="extensions-for-selectively-disabling-optimization">
<h2><a class="toc-backref" href="#id19">Extensions for selectively disabling optimization</a><a class="headerlink" href="#extensions-for-selectively-disabling-optimization" title="Permalink to this headline">¶</a></h2>
<p>Clang provides a mechanism for selectively disabling optimizations in functions
and methods.</p>
<p>To disable optimizations in a single function definition, the GNU-style or C++11
non-standard attribute <code class="docutils literal"><span class="pre">optnone</span></code> can be used.</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="c1">// The following functions will not be optimized.</span>
<span class="c1">// GNU-style attribute</span>
<span class="n">__attribute__</span><span class="p">((</span><span class="n">optnone</span><span class="p">))</span> <span class="kt">int</span> <span class="n">foo</span><span class="p">()</span> <span class="p">{</span>
<span class="c1">// ... code</span>
<span class="p">}</span>
<span class="c1">// C++11 attribute</span>
<span class="p">[[</span><span class="n">clang</span><span class="o">::</span><span class="n">optnone</span><span class="p">]]</span> <span class="kt">int</span> <span class="n">bar</span><span class="p">()</span> <span class="p">{</span>
<span class="c1">// ... code</span>
<span class="p">}</span>
</pre></div>
</div>
<p>To facilitate disabling optimization for a range of function definitions, a
range-based pragma is provided. Its syntax is <code class="docutils literal"><span class="pre">#pragma</span> <span class="pre">clang</span> <span class="pre">optimize</span></code>
followed by <code class="docutils literal"><span class="pre">off</span></code> or <code class="docutils literal"><span class="pre">on</span></code>.</p>
<p>All function definitions in the region between an <code class="docutils literal"><span class="pre">off</span></code> and the following
<code class="docutils literal"><span class="pre">on</span></code> will be decorated with the <code class="docutils literal"><span class="pre">optnone</span></code> attribute unless doing so would
conflict with explicit attributes already present on the function (e.g. the
ones that control inlining).</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#pragma clang optimize off</span>
<span class="c1">// This function will be decorated with optnone.</span>
<span class="kt">int</span> <span class="nf">foo</span><span class="p">()</span> <span class="p">{</span>
<span class="c1">// ... code</span>
<span class="p">}</span>
<span class="c1">// optnone conflicts with always_inline, so bar() will not be decorated.</span>
<span class="n">__attribute__</span><span class="p">((</span><span class="n">always_inline</span><span class="p">))</span> <span class="kt">int</span> <span class="n">bar</span><span class="p">()</span> <span class="p">{</span>
<span class="c1">// ... code</span>
<span class="p">}</span>
<span class="cp">#pragma clang optimize on</span>
</pre></div>
</div>
<p>If no <code class="docutils literal"><span class="pre">on</span></code> is found to close an <code class="docutils literal"><span class="pre">off</span></code> region, the end of the region is the
end of the compilation unit.</p>
<p>Note that a stray <code class="docutils literal"><span class="pre">#pragma</span> <span class="pre">clang</span> <span class="pre">optimize</span> <span class="pre">on</span></code> does not selectively enable
additional optimizations when compiling at low optimization levels. This feature
can only be used to selectively disable optimizations.</p>
<p>The pragma has an effect on functions only at the point of their definition; for
function templates, this means that the state of the pragma at the point of an
instantiation is not necessarily relevant. Consider the following example:</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="k">template</span><span class="o"><</span><span class="k">typename</span> <span class="n">T</span><span class="o">></span> <span class="n">T</span> <span class="n">twice</span><span class="p">(</span><span class="n">T</span> <span class="n">t</span><span class="p">)</span> <span class="p">{</span>
<span class="k">return</span> <span class="mi">2</span> <span class="o">*</span> <span class="n">t</span><span class="p">;</span>
<span class="p">}</span>
<span class="cp">#pragma clang optimize off</span>
<span class="k">template</span><span class="o"><</span><span class="k">typename</span> <span class="n">T</span><span class="o">></span> <span class="n">T</span> <span class="n">thrice</span><span class="p">(</span><span class="n">T</span> <span class="n">t</span><span class="p">)</span> <span class="p">{</span>
<span class="k">return</span> <span class="mi">3</span> <span class="o">*</span> <span class="n">t</span><span class="p">;</span>
<span class="p">}</span>
<span class="kt">int</span> <span class="n">container</span><span class="p">(</span><span class="kt">int</span> <span class="n">a</span><span class="p">,</span> <span class="kt">int</span> <span class="n">b</span><span class="p">)</span> <span class="p">{</span>
<span class="k">return</span> <span class="n">twice</span><span class="p">(</span><span class="n">a</span><span class="p">)</span> <span class="o">+</span> <span class="n">thrice</span><span class="p">(</span><span class="n">b</span><span class="p">);</span>
<span class="p">}</span>
<span class="cp">#pragma clang optimize on</span>
</pre></div>
</div>
<p>In this example, the definition of the template function <code class="docutils literal"><span class="pre">twice</span></code> is outside
the pragma region, whereas the definition of <code class="docutils literal"><span class="pre">thrice</span></code> is inside the region.
The <code class="docutils literal"><span class="pre">container</span></code> function is also in the region and will not be optimized, but
it causes the instantiation of <code class="docutils literal"><span class="pre">twice</span></code> and <code class="docutils literal"><span class="pre">thrice</span></code> with an <code class="docutils literal"><span class="pre">int</span></code> type; of
these two instantiations, <code class="docutils literal"><span class="pre">twice</span></code> will be optimized (because its definition
was outside the region) and <code class="docutils literal"><span class="pre">thrice</span></code> will not be optimized.</p>
</div>
<div class="section" id="extensions-for-loop-hint-optimizations">
<span id="langext-pragma-loop"></span><h2><a class="toc-backref" href="#id20">Extensions for loop hint optimizations</a><a class="headerlink" href="#extensions-for-loop-hint-optimizations" title="Permalink to this headline">¶</a></h2>
<p>The <code class="docutils literal"><span class="pre">#pragma</span> <span class="pre">clang</span> <span class="pre">loop</span></code> directive is used to specify hints for optimizing the
subsequent for, while, do-while, or c++11 range-based for loop. The directive
provides options for vectorization and interleaving. Loop hints can be specified
before any loop and will be ignored if the optimization is not safe to apply.</p>
<p>A vectorized loop performs multiple iterations of the original loop
in parallel using vector instructions. The instruction set of the target
processor determines which vector instructions are available and their vector
widths. This restricts the types of loops that can be vectorized. The vectorizer
automatically determines if the loop is safe and profitable to vectorize. A
vector instruction cost model is used to select the vector width.</p>
<p>Interleaving multiple loop iterations allows modern processors to further
improve instruction-level parallelism (ILP) using advanced hardware features,
such as multiple execution units and out-of-order execution. The vectorizer uses
a cost model that depends on the register pressure and generated code size to
select the interleaving count.</p>
<p>Vectorization is enabled by <code class="docutils literal"><span class="pre">vectorize(enable)</span></code> and interleaving is enabled
by <code class="docutils literal"><span class="pre">interleave(enable)</span></code>. This is useful when compiling with <code class="docutils literal"><span class="pre">-Os</span></code> to
manually enable vectorization or interleaving.</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#pragma clang loop vectorize(enable)</span>
<span class="cp">#pragma clang loop interleave(enable)</span>
<span class="k">for</span><span class="p">(...)</span> <span class="p">{</span>
<span class="p">...</span>
<span class="p">}</span>
</pre></div>
</div>
<p>The vector width is specified by <code class="docutils literal"><span class="pre">vectorize_width(_value_)</span></code> and the interleave
count is specified by <code class="docutils literal"><span class="pre">interleave_count(_value_)</span></code>, where
_value_ is a positive integer. This is useful for specifying the optimal
width/count of the set of target architectures supported by your application.</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#pragma clang loop vectorize_width(2)</span>
<span class="cp">#pragma clang loop interleave_count(2)</span>
<span class="k">for</span><span class="p">(...)</span> <span class="p">{</span>
<span class="p">...</span>
<span class="p">}</span>
</pre></div>
</div>
<p>Specifying a width/count of 1 disables the optimization, and is equivalent to
<code class="docutils literal"><span class="pre">vectorize(disable)</span></code> or <code class="docutils literal"><span class="pre">interleave(disable)</span></code>.</p>
<p>For convenience multiple loop hints can be specified on a single line.</p>
<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#pragma clang loop vectorize_width(4) interleave_count(8)</span>
<span class="k">for</span><span class="p">(...)</span> <span class="p">{</span>
<span class="p">...</span>
<span class="p">}</span>
</pre></div>
</div>
<p>If an optimization cannot be applied any hints that apply to it will be ignored.
For example, the hint <code class="docutils literal"><span class="pre">vectorize_width(4)</span></code> is ignored if the loop is not
proven safe to vectorize. To identify and diagnose optimization issues use
<cite>-Rpass</cite>, <cite>-Rpass-missed</cite>, and <cite>-Rpass-analysis</cite> command line options. See the
user guide for details.</p>
</div>
</div>
</div>
<div class="bottomnav" role="navigation" aria-label="bottom navigation">
<p>
«  <a href="UsersManual.html">Clang Compiler User’s Manual</a>
  ::  
<a class="uplink" href="index.html">Contents</a>
  ::  
<a href="ObjectiveCLiterals.html">Objective-C Literals</a>  »
</p>
</div>
<div class="footer" role="contentinfo">
© Copyright 2007-2014, The Clang Team.
Created using <a href="http://sphinx-doc.org/">Sphinx</a> 1.3.6.
</div>
</body>
</html>
|