Training courses

Kernel and Embedded Linux

Bootlin training courses

Embedded Linux, kernel,
Yocto Project, Buildroot, real-time,
graphics, boot time, debugging...

Bootlin logo

Elixir Cross Referencer

   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
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
/*===---- smmintrin.h - SSE4 intrinsics ------------------------------------===
 *
 * Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 * See https://llvm.org/LICENSE.txt for license information.
 * SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 *
 *===-----------------------------------------------------------------------===
 */

#ifndef __SMMINTRIN_H
#define __SMMINTRIN_H

#include <tmmintrin.h>

/* Define the default attributes for the functions in this file. */
#define __DEFAULT_FN_ATTRS __attribute__((__always_inline__, __nodebug__, __target__("sse4.1"), __min_vector_width__(128)))

/* SSE4 Rounding macros. */
#define _MM_FROUND_TO_NEAREST_INT    0x00
#define _MM_FROUND_TO_NEG_INF        0x01
#define _MM_FROUND_TO_POS_INF        0x02
#define _MM_FROUND_TO_ZERO           0x03
#define _MM_FROUND_CUR_DIRECTION     0x04

#define _MM_FROUND_RAISE_EXC         0x00
#define _MM_FROUND_NO_EXC            0x08

#define _MM_FROUND_NINT      (_MM_FROUND_RAISE_EXC | _MM_FROUND_TO_NEAREST_INT)
#define _MM_FROUND_FLOOR     (_MM_FROUND_RAISE_EXC | _MM_FROUND_TO_NEG_INF)
#define _MM_FROUND_CEIL      (_MM_FROUND_RAISE_EXC | _MM_FROUND_TO_POS_INF)
#define _MM_FROUND_TRUNC     (_MM_FROUND_RAISE_EXC | _MM_FROUND_TO_ZERO)
#define _MM_FROUND_RINT      (_MM_FROUND_RAISE_EXC | _MM_FROUND_CUR_DIRECTION)
#define _MM_FROUND_NEARBYINT (_MM_FROUND_NO_EXC | _MM_FROUND_CUR_DIRECTION)

/// Rounds up each element of the 128-bit vector of [4 x float] to an
///    integer and returns the rounded values in a 128-bit vector of
///    [4 x float].
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128 _mm_ceil_ps(__m128 X);
/// \endcode
///
/// This intrinsic corresponds to the <c> VROUNDPS / ROUNDPS </c> instruction.
///
/// \param X
///    A 128-bit vector of [4 x float] values to be rounded up.
/// \returns A 128-bit vector of [4 x float] containing the rounded values.
#define _mm_ceil_ps(X)       _mm_round_ps((X), _MM_FROUND_CEIL)

/// Rounds up each element of the 128-bit vector of [2 x double] to an
///    integer and returns the rounded values in a 128-bit vector of
///    [2 x double].
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128d _mm_ceil_pd(__m128d X);
/// \endcode
///
/// This intrinsic corresponds to the <c> VROUNDPD / ROUNDPD </c> instruction.
///
/// \param X
///    A 128-bit vector of [2 x double] values to be rounded up.
/// \returns A 128-bit vector of [2 x double] containing the rounded values.
#define _mm_ceil_pd(X)       _mm_round_pd((X), _MM_FROUND_CEIL)

/// Copies three upper elements of the first 128-bit vector operand to
///    the corresponding three upper elements of the 128-bit result vector of
///    [4 x float]. Rounds up the lowest element of the second 128-bit vector
///    operand to an integer and copies it to the lowest element of the 128-bit
///    result vector of [4 x float].
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128 _mm_ceil_ss(__m128 X, __m128 Y);
/// \endcode
///
/// This intrinsic corresponds to the <c> VROUNDSS / ROUNDSS </c> instruction.
///
/// \param X
///    A 128-bit vector of [4 x float]. The values stored in bits [127:32] are
///    copied to the corresponding bits of the result.
/// \param Y
///    A 128-bit vector of [4 x float]. The value stored in bits [31:0] is
///    rounded up to the nearest integer and copied to the corresponding bits
///    of the result.
/// \returns A 128-bit vector of [4 x float] containing the copied and rounded
///    values.
#define _mm_ceil_ss(X, Y)    _mm_round_ss((X), (Y), _MM_FROUND_CEIL)

/// Copies the upper element of the first 128-bit vector operand to the
///    corresponding upper element of the 128-bit result vector of [2 x double].
///    Rounds up the lower element of the second 128-bit vector operand to an
///    integer and copies it to the lower element of the 128-bit result vector
///    of [2 x double].
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128d _mm_ceil_sd(__m128d X, __m128d Y);
/// \endcode
///
/// This intrinsic corresponds to the <c> VROUNDSD / ROUNDSD </c> instruction.
///
/// \param X
///    A 128-bit vector of [2 x double]. The value stored in bits [127:64] is
///    copied to the corresponding bits of the result.
/// \param Y
///    A 128-bit vector of [2 x double]. The value stored in bits [63:0] is
///    rounded up to the nearest integer and copied to the corresponding bits
///    of the result.
/// \returns A 128-bit vector of [2 x double] containing the copied and rounded
///    values.
#define _mm_ceil_sd(X, Y)    _mm_round_sd((X), (Y), _MM_FROUND_CEIL)

/// Rounds down each element of the 128-bit vector of [4 x float] to an
///    an integer and returns the rounded values in a 128-bit vector of
///    [4 x float].
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128 _mm_floor_ps(__m128 X);
/// \endcode
///
/// This intrinsic corresponds to the <c> VROUNDPS / ROUNDPS </c> instruction.
///
/// \param X
///    A 128-bit vector of [4 x float] values to be rounded down.
/// \returns A 128-bit vector of [4 x float] containing the rounded values.
#define _mm_floor_ps(X)      _mm_round_ps((X), _MM_FROUND_FLOOR)

/// Rounds down each element of the 128-bit vector of [2 x double] to an
///    integer and returns the rounded values in a 128-bit vector of
///    [2 x double].
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128d _mm_floor_pd(__m128d X);
/// \endcode
///
/// This intrinsic corresponds to the <c> VROUNDPD / ROUNDPD </c> instruction.
///
/// \param X
///    A 128-bit vector of [2 x double].
/// \returns A 128-bit vector of [2 x double] containing the rounded values.
#define _mm_floor_pd(X)      _mm_round_pd((X), _MM_FROUND_FLOOR)

/// Copies three upper elements of the first 128-bit vector operand to
///    the corresponding three upper elements of the 128-bit result vector of
///    [4 x float]. Rounds down the lowest element of the second 128-bit vector
///    operand to an integer and copies it to the lowest element of the 128-bit
///    result vector of [4 x float].
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128 _mm_floor_ss(__m128 X, __m128 Y);
/// \endcode
///
/// This intrinsic corresponds to the <c> VROUNDSS / ROUNDSS </c> instruction.
///
/// \param X
///    A 128-bit vector of [4 x float]. The values stored in bits [127:32] are
///    copied to the corresponding bits of the result.
/// \param Y
///    A 128-bit vector of [4 x float]. The value stored in bits [31:0] is
///    rounded down to the nearest integer and copied to the corresponding bits
///    of the result.
/// \returns A 128-bit vector of [4 x float] containing the copied and rounded
///    values.
#define _mm_floor_ss(X, Y)   _mm_round_ss((X), (Y), _MM_FROUND_FLOOR)

/// Copies the upper element of the first 128-bit vector operand to the
///    corresponding upper element of the 128-bit result vector of [2 x double].
///    Rounds down the lower element of the second 128-bit vector operand to an
///    integer and copies it to the lower element of the 128-bit result vector
///    of [2 x double].
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128d _mm_floor_sd(__m128d X, __m128d Y);
/// \endcode
///
/// This intrinsic corresponds to the <c> VROUNDSD / ROUNDSD </c> instruction.
///
/// \param X
///    A 128-bit vector of [2 x double]. The value stored in bits [127:64] is
///    copied to the corresponding bits of the result.
/// \param Y
///    A 128-bit vector of [2 x double]. The value stored in bits [63:0] is
///    rounded down to the nearest integer and copied to the corresponding bits
///    of the result.
/// \returns A 128-bit vector of [2 x double] containing the copied and rounded
///    values.
#define _mm_floor_sd(X, Y)   _mm_round_sd((X), (Y), _MM_FROUND_FLOOR)

/// Rounds each element of the 128-bit vector of [4 x float] to an
///    integer value according to the rounding control specified by the second
///    argument and returns the rounded values in a 128-bit vector of
///    [4 x float].
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128 _mm_round_ps(__m128 X, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VROUNDPS / ROUNDPS </c> instruction.
///
/// \param X
///    A 128-bit vector of [4 x float].
/// \param M
///    An integer value that specifies the rounding operation. \n
///    Bits [7:4] are reserved. \n
///    Bit [3] is a precision exception value: \n
///      0: A normal PE exception is used \n
///      1: The PE field is not updated \n
///    Bit [2] is the rounding control source: \n
///      0: Use bits [1:0] of \a M \n
///      1: Use the current MXCSR setting \n
///    Bits [1:0] contain the rounding control definition: \n
///      00: Nearest \n
///      01: Downward (toward negative infinity) \n
///      10: Upward (toward positive infinity) \n
///      11: Truncated
/// \returns A 128-bit vector of [4 x float] containing the rounded values.
#define _mm_round_ps(X, M) \
  (__m128)__builtin_ia32_roundps((__v4sf)(__m128)(X), (M))

/// Copies three upper elements of the first 128-bit vector operand to
///    the corresponding three upper elements of the 128-bit result vector of
///    [4 x float]. Rounds the lowest element of the second 128-bit vector
///    operand to an integer value according to the rounding control specified
///    by the third argument and copies it to the lowest element of the 128-bit
///    result vector of [4 x float].
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128 _mm_round_ss(__m128 X, __m128 Y, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VROUNDSS / ROUNDSS </c> instruction.
///
/// \param X
///    A 128-bit vector of [4 x float]. The values stored in bits [127:32] are
///    copied to the corresponding bits of the result.
/// \param Y
///    A 128-bit vector of [4 x float]. The value stored in bits [31:0] is
///    rounded to the nearest integer using the specified rounding control and
///    copied to the corresponding bits of the result.
/// \param M
///    An integer value that specifies the rounding operation. \n
///    Bits [7:4] are reserved. \n
///    Bit [3] is a precision exception value: \n
///      0: A normal PE exception is used \n
///      1: The PE field is not updated \n
///    Bit [2] is the rounding control source: \n
///      0: Use bits [1:0] of \a M \n
///      1: Use the current MXCSR setting \n
///    Bits [1:0] contain the rounding control definition: \n
///      00: Nearest \n
///      01: Downward (toward negative infinity) \n
///      10: Upward (toward positive infinity) \n
///      11: Truncated
/// \returns A 128-bit vector of [4 x float] containing the copied and rounded
///    values.
#define _mm_round_ss(X, Y, M) \
  (__m128)__builtin_ia32_roundss((__v4sf)(__m128)(X), \
                                 (__v4sf)(__m128)(Y), (M))

/// Rounds each element of the 128-bit vector of [2 x double] to an
///    integer value according to the rounding control specified by the second
///    argument and returns the rounded values in a 128-bit vector of
///    [2 x double].
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128d _mm_round_pd(__m128d X, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VROUNDPD / ROUNDPD </c> instruction.
///
/// \param X
///    A 128-bit vector of [2 x double].
/// \param M
///    An integer value that specifies the rounding operation. \n
///    Bits [7:4] are reserved. \n
///    Bit [3] is a precision exception value: \n
///      0: A normal PE exception is used \n
///      1: The PE field is not updated \n
///    Bit [2] is the rounding control source: \n
///      0: Use bits [1:0] of \a M \n
///      1: Use the current MXCSR setting \n
///    Bits [1:0] contain the rounding control definition: \n
///      00: Nearest \n
///      01: Downward (toward negative infinity) \n
///      10: Upward (toward positive infinity) \n
///      11: Truncated
/// \returns A 128-bit vector of [2 x double] containing the rounded values.
#define _mm_round_pd(X, M) \
  (__m128d)__builtin_ia32_roundpd((__v2df)(__m128d)(X), (M))

/// Copies the upper element of the first 128-bit vector operand to the
///    corresponding upper element of the 128-bit result vector of [2 x double].
///    Rounds the lower element of the second 128-bit vector operand to an
///    integer value according to the rounding control specified by the third
///    argument and copies it to the lower element of the 128-bit result vector
///    of [2 x double].
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128d _mm_round_sd(__m128d X, __m128d Y, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VROUNDSD / ROUNDSD </c> instruction.
///
/// \param X
///    A 128-bit vector of [2 x double]. The value stored in bits [127:64] is
///    copied to the corresponding bits of the result.
/// \param Y
///    A 128-bit vector of [2 x double]. The value stored in bits [63:0] is
///    rounded to the nearest integer using the specified rounding control and
///    copied to the corresponding bits of the result.
/// \param M
///    An integer value that specifies the rounding operation. \n
///    Bits [7:4] are reserved. \n
///    Bit [3] is a precision exception value: \n
///      0: A normal PE exception is used \n
///      1: The PE field is not updated \n
///    Bit [2] is the rounding control source: \n
///      0: Use bits [1:0] of \a M \n
///      1: Use the current MXCSR setting \n
///    Bits [1:0] contain the rounding control definition: \n
///      00: Nearest \n
///      01: Downward (toward negative infinity) \n
///      10: Upward (toward positive infinity) \n
///      11: Truncated
/// \returns A 128-bit vector of [2 x double] containing the copied and rounded
///    values.
#define _mm_round_sd(X, Y, M) \
  (__m128d)__builtin_ia32_roundsd((__v2df)(__m128d)(X), \
                                  (__v2df)(__m128d)(Y), (M))

/* SSE4 Packed Blending Intrinsics.  */
/// Returns a 128-bit vector of [2 x double] where the values are
///    selected from either the first or second operand as specified by the
///    third operand, the control mask.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128d _mm_blend_pd(__m128d V1, __m128d V2, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VBLENDPD / BLENDPD </c> instruction.
///
/// \param V1
///    A 128-bit vector of [2 x double].
/// \param V2
///    A 128-bit vector of [2 x double].
/// \param M
///    An immediate integer operand, with mask bits [1:0] specifying how the
///    values are to be copied. The position of the mask bit corresponds to the
///    index of a copied value. When a mask bit is 0, the corresponding 64-bit
///    element in operand \a V1 is copied to the same position in the result.
///    When a mask bit is 1, the corresponding 64-bit element in operand \a V2
///    is copied to the same position in the result.
/// \returns A 128-bit vector of [2 x double] containing the copied values.
#define _mm_blend_pd(V1, V2, M) \
  (__m128d) __builtin_ia32_blendpd ((__v2df)(__m128d)(V1), \
                                    (__v2df)(__m128d)(V2), (int)(M))

/// Returns a 128-bit vector of [4 x float] where the values are selected
///    from either the first or second operand as specified by the third
///    operand, the control mask.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128 _mm_blend_ps(__m128 V1, __m128 V2, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VBLENDPS / BLENDPS </c> instruction.
///
/// \param V1
///    A 128-bit vector of [4 x float].
/// \param V2
///    A 128-bit vector of [4 x float].
/// \param M
///    An immediate integer operand, with mask bits [3:0] specifying how the
///    values are to be copied. The position of the mask bit corresponds to the
///    index of a copied value. When a mask bit is 0, the corresponding 32-bit
///    element in operand \a V1 is copied to the same position in the result.
///    When a mask bit is 1, the corresponding 32-bit element in operand \a V2
///    is copied to the same position in the result.
/// \returns A 128-bit vector of [4 x float] containing the copied values.
#define _mm_blend_ps(V1, V2, M) \
  (__m128) __builtin_ia32_blendps ((__v4sf)(__m128)(V1), \
                                   (__v4sf)(__m128)(V2), (int)(M))

/// Returns a 128-bit vector of [2 x double] where the values are
///    selected from either the first or second operand as specified by the
///    third operand, the control mask.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VBLENDVPD / BLENDVPD </c> instruction.
///
/// \param __V1
///    A 128-bit vector of [2 x double].
/// \param __V2
///    A 128-bit vector of [2 x double].
/// \param __M
///    A 128-bit vector operand, with mask bits 127 and 63 specifying how the
///    values are to be copied. The position of the mask bit corresponds to the
///    most significant bit of a copied value. When a mask bit is 0, the
///    corresponding 64-bit element in operand \a __V1 is copied to the same
///    position in the result. When a mask bit is 1, the corresponding 64-bit
///    element in operand \a __V2 is copied to the same position in the result.
/// \returns A 128-bit vector of [2 x double] containing the copied values.
static __inline__ __m128d __DEFAULT_FN_ATTRS
_mm_blendv_pd (__m128d __V1, __m128d __V2, __m128d __M)
{
  return (__m128d) __builtin_ia32_blendvpd ((__v2df)__V1, (__v2df)__V2,
                                            (__v2df)__M);
}

/// Returns a 128-bit vector of [4 x float] where the values are
///    selected from either the first or second operand as specified by the
///    third operand, the control mask.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VBLENDVPS / BLENDVPS </c> instruction.
///
/// \param __V1
///    A 128-bit vector of [4 x float].
/// \param __V2
///    A 128-bit vector of [4 x float].
/// \param __M
///    A 128-bit vector operand, with mask bits 127, 95, 63, and 31 specifying
///    how the values are to be copied. The position of the mask bit corresponds
///    to the most significant bit of a copied value. When a mask bit is 0, the
///    corresponding 32-bit element in operand \a __V1 is copied to the same
///    position in the result. When a mask bit is 1, the corresponding 32-bit
///    element in operand \a __V2 is copied to the same position in the result.
/// \returns A 128-bit vector of [4 x float] containing the copied values.
static __inline__ __m128 __DEFAULT_FN_ATTRS
_mm_blendv_ps (__m128 __V1, __m128 __V2, __m128 __M)
{
  return (__m128) __builtin_ia32_blendvps ((__v4sf)__V1, (__v4sf)__V2,
                                           (__v4sf)__M);
}

/// Returns a 128-bit vector of [16 x i8] where the values are selected
///    from either of the first or second operand as specified by the third
///    operand, the control mask.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPBLENDVB / PBLENDVB </c> instruction.
///
/// \param __V1
///    A 128-bit vector of [16 x i8].
/// \param __V2
///    A 128-bit vector of [16 x i8].
/// \param __M
///    A 128-bit vector operand, with mask bits 127, 119, 111...7 specifying
///    how the values are to be copied. The position of the mask bit corresponds
///    to the most significant bit of a copied value. When a mask bit is 0, the
///    corresponding 8-bit element in operand \a __V1 is copied to the same
///    position in the result. When a mask bit is 1, the corresponding 8-bit
///    element in operand \a __V2 is copied to the same position in the result.
/// \returns A 128-bit vector of [16 x i8] containing the copied values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_blendv_epi8 (__m128i __V1, __m128i __V2, __m128i __M)
{
  return (__m128i) __builtin_ia32_pblendvb128 ((__v16qi)__V1, (__v16qi)__V2,
                                               (__v16qi)__M);
}

/// Returns a 128-bit vector of [8 x i16] where the values are selected
///    from either of the first or second operand as specified by the third
///    operand, the control mask.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128i _mm_blend_epi16(__m128i V1, __m128i V2, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPBLENDW / PBLENDW </c> instruction.
///
/// \param V1
///    A 128-bit vector of [8 x i16].
/// \param V2
///    A 128-bit vector of [8 x i16].
/// \param M
///    An immediate integer operand, with mask bits [7:0] specifying how the
///    values are to be copied. The position of the mask bit corresponds to the
///    index of a copied value. When a mask bit is 0, the corresponding 16-bit
///    element in operand \a V1 is copied to the same position in the result.
///    When a mask bit is 1, the corresponding 16-bit element in operand \a V2
///    is copied to the same position in the result.
/// \returns A 128-bit vector of [8 x i16] containing the copied values.
#define _mm_blend_epi16(V1, V2, M) \
  (__m128i) __builtin_ia32_pblendw128 ((__v8hi)(__m128i)(V1), \
                                       (__v8hi)(__m128i)(V2), (int)(M))

/* SSE4 Dword Multiply Instructions.  */
/// Multiples corresponding elements of two 128-bit vectors of [4 x i32]
///    and returns the lower 32 bits of the each product in a 128-bit vector of
///    [4 x i32].
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMULLD / PMULLD </c> instruction.
///
/// \param __V1
///    A 128-bit integer vector.
/// \param __V2
///    A 128-bit integer vector.
/// \returns A 128-bit integer vector containing the products of both operands.
static __inline__  __m128i __DEFAULT_FN_ATTRS
_mm_mullo_epi32 (__m128i __V1, __m128i __V2)
{
  return (__m128i) ((__v4su)__V1 * (__v4su)__V2);
}

/// Multiplies corresponding even-indexed elements of two 128-bit
///    vectors of [4 x i32] and returns a 128-bit vector of [2 x i64]
///    containing the products.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMULDQ / PMULDQ </c> instruction.
///
/// \param __V1
///    A 128-bit vector of [4 x i32].
/// \param __V2
///    A 128-bit vector of [4 x i32].
/// \returns A 128-bit vector of [2 x i64] containing the products of both
///    operands.
static __inline__  __m128i __DEFAULT_FN_ATTRS
_mm_mul_epi32 (__m128i __V1, __m128i __V2)
{
  return (__m128i) __builtin_ia32_pmuldq128 ((__v4si)__V1, (__v4si)__V2);
}

/* SSE4 Floating Point Dot Product Instructions.  */
/// Computes the dot product of the two 128-bit vectors of [4 x float]
///    and returns it in the elements of the 128-bit result vector of
///    [4 x float].
///
///    The immediate integer operand controls which input elements
///    will contribute to the dot product, and where the final results are
///    returned.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128 _mm_dp_ps(__m128 X, __m128 Y, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VDPPS / DPPS </c> instruction.
///
/// \param X
///    A 128-bit vector of [4 x float].
/// \param Y
///    A 128-bit vector of [4 x float].
/// \param M
///    An immediate integer operand. Mask bits [7:4] determine which elements
///    of the input vectors are used, with bit [4] corresponding to the lowest
///    element and bit [7] corresponding to the highest element of each [4 x
///    float] vector. If a bit is set, the corresponding elements from the two
///    input vectors are used as an input for dot product; otherwise that input
///    is treated as zero. Bits [3:0] determine which elements of the result
///    will receive a copy of the final dot product, with bit [0] corresponding
///    to the lowest element and bit [3] corresponding to the highest element of
///    each [4 x float] subvector. If a bit is set, the dot product is returned
///    in the corresponding element; otherwise that element is set to zero.
/// \returns A 128-bit vector of [4 x float] containing the dot product.
#define _mm_dp_ps(X, Y, M) \
  (__m128) __builtin_ia32_dpps((__v4sf)(__m128)(X), \
                               (__v4sf)(__m128)(Y), (M))

/// Computes the dot product of the two 128-bit vectors of [2 x double]
///    and returns it in the elements of the 128-bit result vector of
///    [2 x double].
///
///    The immediate integer operand controls which input
///    elements will contribute to the dot product, and where the final results
///    are returned.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128d _mm_dp_pd(__m128d X, __m128d Y, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VDPPD / DPPD </c> instruction.
///
/// \param X
///    A 128-bit vector of [2 x double].
/// \param Y
///    A 128-bit vector of [2 x double].
/// \param M
///    An immediate integer operand. Mask bits [5:4] determine which elements
///    of the input vectors are used, with bit [4] corresponding to the lowest
///    element and bit [5] corresponding to the highest element of each of [2 x
///    double] vector. If a bit is set, the corresponding elements from the two
///    input vectors are used as an input for dot product; otherwise that input
///    is treated as zero. Bits [1:0] determine which elements of the result
///    will receive a copy of the final dot product, with bit [0] corresponding
///    to the lowest element and bit [1] corresponding to the highest element of
///    each [2 x double] vector. If a bit is set, the dot product is returned in
///    the corresponding element; otherwise that element is set to zero.
#define _mm_dp_pd(X, Y, M) \
  (__m128d) __builtin_ia32_dppd((__v2df)(__m128d)(X), \
                                (__v2df)(__m128d)(Y), (M))

/* SSE4 Streaming Load Hint Instruction.  */
/// Loads integer values from a 128-bit aligned memory location to a
///    128-bit integer vector.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VMOVNTDQA / MOVNTDQA </c> instruction.
///
/// \param __V
///    A pointer to a 128-bit aligned memory location that contains the integer
///    values.
/// \returns A 128-bit integer vector containing the data stored at the
///    specified memory location.
static __inline__  __m128i __DEFAULT_FN_ATTRS
_mm_stream_load_si128 (__m128i const *__V)
{
  return (__m128i) __builtin_nontemporal_load ((const __v2di *) __V);
}

/* SSE4 Packed Integer Min/Max Instructions.  */
/// Compares the corresponding elements of two 128-bit vectors of
///    [16 x i8] and returns a 128-bit vector of [16 x i8] containing the lesser
///    of the two values.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMINSB / PMINSB </c> instruction.
///
/// \param __V1
///    A 128-bit vector of [16 x i8].
/// \param __V2
///    A 128-bit vector of [16 x i8]
/// \returns A 128-bit vector of [16 x i8] containing the lesser values.
static __inline__  __m128i __DEFAULT_FN_ATTRS
_mm_min_epi8 (__m128i __V1, __m128i __V2)
{
  return (__m128i) __builtin_ia32_pminsb128 ((__v16qi) __V1, (__v16qi) __V2);
}

/// Compares the corresponding elements of two 128-bit vectors of
///    [16 x i8] and returns a 128-bit vector of [16 x i8] containing the
///    greater value of the two.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMAXSB / PMAXSB </c> instruction.
///
/// \param __V1
///    A 128-bit vector of [16 x i8].
/// \param __V2
///    A 128-bit vector of [16 x i8].
/// \returns A 128-bit vector of [16 x i8] containing the greater values.
static __inline__  __m128i __DEFAULT_FN_ATTRS
_mm_max_epi8 (__m128i __V1, __m128i __V2)
{
  return (__m128i) __builtin_ia32_pmaxsb128 ((__v16qi) __V1, (__v16qi) __V2);
}

/// Compares the corresponding elements of two 128-bit vectors of
///    [8 x u16] and returns a 128-bit vector of [8 x u16] containing the lesser
///    value of the two.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMINUW / PMINUW </c> instruction.
///
/// \param __V1
///    A 128-bit vector of [8 x u16].
/// \param __V2
///    A 128-bit vector of [8 x u16].
/// \returns A 128-bit vector of [8 x u16] containing the lesser values.
static __inline__  __m128i __DEFAULT_FN_ATTRS
_mm_min_epu16 (__m128i __V1, __m128i __V2)
{
  return (__m128i) __builtin_ia32_pminuw128 ((__v8hi) __V1, (__v8hi) __V2);
}

/// Compares the corresponding elements of two 128-bit vectors of
///    [8 x u16] and returns a 128-bit vector of [8 x u16] containing the
///    greater value of the two.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMAXUW / PMAXUW </c> instruction.
///
/// \param __V1
///    A 128-bit vector of [8 x u16].
/// \param __V2
///    A 128-bit vector of [8 x u16].
/// \returns A 128-bit vector of [8 x u16] containing the greater values.
static __inline__  __m128i __DEFAULT_FN_ATTRS
_mm_max_epu16 (__m128i __V1, __m128i __V2)
{
  return (__m128i) __builtin_ia32_pmaxuw128 ((__v8hi) __V1, (__v8hi) __V2);
}

/// Compares the corresponding elements of two 128-bit vectors of
///    [4 x i32] and returns a 128-bit vector of [4 x i32] containing the lesser
///    value of the two.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMINSD / PMINSD </c> instruction.
///
/// \param __V1
///    A 128-bit vector of [4 x i32].
/// \param __V2
///    A 128-bit vector of [4 x i32].
/// \returns A 128-bit vector of [4 x i32] containing the lesser values.
static __inline__  __m128i __DEFAULT_FN_ATTRS
_mm_min_epi32 (__m128i __V1, __m128i __V2)
{
  return (__m128i) __builtin_ia32_pminsd128 ((__v4si) __V1, (__v4si) __V2);
}

/// Compares the corresponding elements of two 128-bit vectors of
///    [4 x i32] and returns a 128-bit vector of [4 x i32] containing the
///    greater value of the two.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMAXSD / PMAXSD </c> instruction.
///
/// \param __V1
///    A 128-bit vector of [4 x i32].
/// \param __V2
///    A 128-bit vector of [4 x i32].
/// \returns A 128-bit vector of [4 x i32] containing the greater values.
static __inline__  __m128i __DEFAULT_FN_ATTRS
_mm_max_epi32 (__m128i __V1, __m128i __V2)
{
  return (__m128i) __builtin_ia32_pmaxsd128 ((__v4si) __V1, (__v4si) __V2);
}

/// Compares the corresponding elements of two 128-bit vectors of
///    [4 x u32] and returns a 128-bit vector of [4 x u32] containing the lesser
///    value of the two.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMINUD / PMINUD </c>  instruction.
///
/// \param __V1
///    A 128-bit vector of [4 x u32].
/// \param __V2
///    A 128-bit vector of [4 x u32].
/// \returns A 128-bit vector of [4 x u32] containing the lesser values.
static __inline__  __m128i __DEFAULT_FN_ATTRS
_mm_min_epu32 (__m128i __V1, __m128i __V2)
{
  return (__m128i) __builtin_ia32_pminud128((__v4si) __V1, (__v4si) __V2);
}

/// Compares the corresponding elements of two 128-bit vectors of
///    [4 x u32] and returns a 128-bit vector of [4 x u32] containing the
///    greater value of the two.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMAXUD / PMAXUD </c> instruction.
///
/// \param __V1
///    A 128-bit vector of [4 x u32].
/// \param __V2
///    A 128-bit vector of [4 x u32].
/// \returns A 128-bit vector of [4 x u32] containing the greater values.
static __inline__  __m128i __DEFAULT_FN_ATTRS
_mm_max_epu32 (__m128i __V1, __m128i __V2)
{
  return (__m128i) __builtin_ia32_pmaxud128((__v4si) __V1, (__v4si) __V2);
}

/* SSE4 Insertion and Extraction from XMM Register Instructions.  */
/// Takes the first argument \a X and inserts an element from the second
///    argument \a Y as selected by the third argument \a N. That result then
///    has elements zeroed out also as selected by the third argument \a N. The
///    resulting 128-bit vector of [4 x float] is then returned.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128 _mm_insert_ps(__m128 X, __m128 Y, const int N);
/// \endcode
///
/// This intrinsic corresponds to the <c> VINSERTPS </c> instruction.
///
/// \param X
///    A 128-bit vector source operand of [4 x float]. With the exception of
///    those bits in the result copied from parameter \a Y and zeroed by bits
///    [3:0] of \a N, all bits from this parameter are copied to the result.
/// \param Y
///    A 128-bit vector source operand of [4 x float]. One single-precision
///    floating-point element from this source, as determined by the immediate
///    parameter, is copied to the result.
/// \param N
///    Specifies which bits from operand \a Y will be copied, which bits in the
///    result they will be be copied to, and which bits in the result will be
///    cleared. The following assignments are made: \n
///    Bits [7:6] specify the bits to copy from operand \a Y: \n
///      00: Selects bits [31:0] from operand \a Y. \n
///      01: Selects bits [63:32] from operand \a Y. \n
///      10: Selects bits [95:64] from operand \a Y. \n
///      11: Selects bits [127:96] from operand \a Y. \n
///    Bits [5:4] specify the bits in the result to which the selected bits
///    from operand \a Y are copied: \n
///      00: Copies the selected bits from \a Y to result bits [31:0]. \n
///      01: Copies the selected bits from \a Y to result bits [63:32]. \n
///      10: Copies the selected bits from \a Y to result bits [95:64]. \n
///      11: Copies the selected bits from \a Y to result bits [127:96]. \n
///    Bits[3:0]: If any of these bits are set, the corresponding result
///    element is cleared.
/// \returns A 128-bit vector of [4 x float] containing the copied
///    single-precision floating point elements from the operands.
#define _mm_insert_ps(X, Y, N) __builtin_ia32_insertps128((X), (Y), (N))

/// Extracts a 32-bit integer from a 128-bit vector of [4 x float] and
///    returns it, using the immediate value parameter \a N as a selector.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_extract_ps(__m128 X, const int N);
/// \endcode
///
/// This intrinsic corresponds to the <c> VEXTRACTPS / EXTRACTPS </c>
/// instruction.
///
/// \param X
///    A 128-bit vector of [4 x float].
/// \param N
///    An immediate value. Bits [1:0] determines which bits from the argument
///    \a X are extracted and returned: \n
///    00: Bits [31:0] of parameter \a X are returned. \n
///    01: Bits [63:32] of parameter \a X are returned. \n
///    10: Bits [95:64] of parameter \a X are returned. \n
///    11: Bits [127:96] of parameter \a X are returned.
/// \returns A 32-bit integer containing the extracted 32 bits of float data.
#define _mm_extract_ps(X, N) (__extension__                      \
  ({ union { int __i; float __f; } __t;  \
     __t.__f = __builtin_ia32_vec_ext_v4sf((__v4sf)(__m128)(X), (int)(N)); \
     __t.__i;}))

/* Miscellaneous insert and extract macros.  */
/* Extract a single-precision float from X at index N into D.  */
#define _MM_EXTRACT_FLOAT(D, X, N) \
  { (D) = __builtin_ia32_vec_ext_v4sf((__v4sf)(__m128)(X), (int)(N)); }

/* Or together 2 sets of indexes (X and Y) with the zeroing bits (Z) to create
   an index suitable for _mm_insert_ps.  */
#define _MM_MK_INSERTPS_NDX(X, Y, Z) (((X) << 6) | ((Y) << 4) | (Z))

/* Extract a float from X at index N into the first index of the return.  */
#define _MM_PICK_OUT_PS(X, N) _mm_insert_ps (_mm_setzero_ps(), (X),   \
                                             _MM_MK_INSERTPS_NDX((N), 0, 0x0e))

/* Insert int into packed integer array at index.  */
/// Constructs a 128-bit vector of [16 x i8] by first making a copy of
///    the 128-bit integer vector parameter, and then inserting the lower 8 bits
///    of an integer parameter \a I into an offset specified by the immediate
///    value parameter \a N.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128i _mm_insert_epi8(__m128i X, int I, const int N);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPINSRB / PINSRB </c> instruction.
///
/// \param X
///    A 128-bit integer vector of [16 x i8]. This vector is copied to the
///    result and then one of the sixteen elements in the result vector is
///    replaced by the lower 8 bits of \a I.
/// \param I
///    An integer. The lower 8 bits of this operand are written to the result
///    beginning at the offset specified by \a N.
/// \param N
///    An immediate value. Bits [3:0] specify the bit offset in the result at
///    which the lower 8 bits of \a I are written. \n
///    0000: Bits [7:0] of the result are used for insertion. \n
///    0001: Bits [15:8] of the result are used for insertion. \n
///    0010: Bits [23:16] of the result are used for insertion. \n
///    0011: Bits [31:24] of the result are used for insertion. \n
///    0100: Bits [39:32] of the result are used for insertion. \n
///    0101: Bits [47:40] of the result are used for insertion. \n
///    0110: Bits [55:48] of the result are used for insertion. \n
///    0111: Bits [63:56] of the result are used for insertion. \n
///    1000: Bits [71:64] of the result are used for insertion. \n
///    1001: Bits [79:72] of the result are used for insertion. \n
///    1010: Bits [87:80] of the result are used for insertion. \n
///    1011: Bits [95:88] of the result are used for insertion. \n
///    1100: Bits [103:96] of the result are used for insertion. \n
///    1101: Bits [111:104] of the result are used for insertion. \n
///    1110: Bits [119:112] of the result are used for insertion. \n
///    1111: Bits [127:120] of the result are used for insertion.
/// \returns A 128-bit integer vector containing the constructed values.
#define _mm_insert_epi8(X, I, N) \
  (__m128i)__builtin_ia32_vec_set_v16qi((__v16qi)(__m128i)(X), \
                                        (int)(I), (int)(N))

/// Constructs a 128-bit vector of [4 x i32] by first making a copy of
///    the 128-bit integer vector parameter, and then inserting the 32-bit
///    integer parameter \a I at the offset specified by the immediate value
///    parameter \a N.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128i _mm_insert_epi32(__m128i X, int I, const int N);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPINSRD / PINSRD </c> instruction.
///
/// \param X
///    A 128-bit integer vector of [4 x i32]. This vector is copied to the
///    result and then one of the four elements in the result vector is
///    replaced by \a I.
/// \param I
///    A 32-bit integer that is written to the result beginning at the offset
///    specified by \a N.
/// \param N
///    An immediate value. Bits [1:0] specify the bit offset in the result at
///    which the integer \a I is written. \n
///    00: Bits [31:0] of the result are used for insertion. \n
///    01: Bits [63:32] of the result are used for insertion. \n
///    10: Bits [95:64] of the result are used for insertion. \n
///    11: Bits [127:96] of the result are used for insertion.
/// \returns A 128-bit integer vector containing the constructed values.
#define _mm_insert_epi32(X, I, N) \
  (__m128i)__builtin_ia32_vec_set_v4si((__v4si)(__m128i)(X), \
                                       (int)(I), (int)(N))

#ifdef __x86_64__
/// Constructs a 128-bit vector of [2 x i64] by first making a copy of
///    the 128-bit integer vector parameter, and then inserting the 64-bit
///    integer parameter \a I, using the immediate value parameter \a N as an
///    insertion location selector.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128i _mm_insert_epi64(__m128i X, long long I, const int N);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPINSRQ / PINSRQ </c> instruction.
///
/// \param X
///    A 128-bit integer vector of [2 x i64]. This vector is copied to the
///    result and then one of the two elements in the result vector is replaced
///    by \a I.
/// \param I
///    A 64-bit integer that is written to the result beginning at the offset
///    specified by \a N.
/// \param N
///    An immediate value. Bit [0] specifies the bit offset in the result at
///    which the integer \a I is written. \n
///    0: Bits [63:0] of the result are used for insertion. \n
///    1: Bits [127:64] of the result are used for insertion. \n
/// \returns A 128-bit integer vector containing the constructed values.
#define _mm_insert_epi64(X, I, N) \
  (__m128i)__builtin_ia32_vec_set_v2di((__v2di)(__m128i)(X), \
                                       (long long)(I), (int)(N))
#endif /* __x86_64__ */

/* Extract int from packed integer array at index.  This returns the element
 * as a zero extended value, so it is unsigned.
 */
/// Extracts an 8-bit element from the 128-bit integer vector of
///    [16 x i8], using the immediate value parameter \a N as a selector.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_extract_epi8(__m128i X, const int N);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPEXTRB / PEXTRB </c> instruction.
///
/// \param X
///    A 128-bit integer vector.
/// \param N
///    An immediate value. Bits [3:0] specify which 8-bit vector element from
///    the argument \a X to extract and copy to the result. \n
///    0000: Bits [7:0] of parameter \a X are extracted. \n
///    0001: Bits [15:8] of the parameter \a X are extracted. \n
///    0010: Bits [23:16] of the parameter \a X are extracted. \n
///    0011: Bits [31:24] of the parameter \a X are extracted. \n
///    0100: Bits [39:32] of the parameter \a X are extracted. \n
///    0101: Bits [47:40] of the parameter \a X are extracted. \n
///    0110: Bits [55:48] of the parameter \a X are extracted. \n
///    0111: Bits [63:56] of the parameter \a X are extracted. \n
///    1000: Bits [71:64] of the parameter \a X are extracted. \n
///    1001: Bits [79:72] of the parameter \a X are extracted. \n
///    1010: Bits [87:80] of the parameter \a X are extracted. \n
///    1011: Bits [95:88] of the parameter \a X are extracted. \n
///    1100: Bits [103:96] of the parameter \a X are extracted. \n
///    1101: Bits [111:104] of the parameter \a X are extracted. \n
///    1110: Bits [119:112] of the parameter \a X are extracted. \n
///    1111: Bits [127:120] of the parameter \a X are extracted.
/// \returns  An unsigned integer, whose lower 8 bits are selected from the
///    128-bit integer vector parameter and the remaining bits are assigned
///    zeros.
#define _mm_extract_epi8(X, N) \
  (int)(unsigned char)__builtin_ia32_vec_ext_v16qi((__v16qi)(__m128i)(X), \
                                                   (int)(N))

/// Extracts a 32-bit element from the 128-bit integer vector of
///    [4 x i32], using the immediate value parameter \a N as a selector.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_extract_epi32(__m128i X, const int N);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPEXTRD / PEXTRD </c> instruction.
///
/// \param X
///    A 128-bit integer vector.
/// \param N
///    An immediate value. Bits [1:0] specify which 32-bit vector element from
///    the argument \a X to extract and copy to the result. \n
///    00: Bits [31:0] of the parameter \a X are extracted. \n
///    01: Bits [63:32] of the parameter \a X are extracted. \n
///    10: Bits [95:64] of the parameter \a X are extracted. \n
///    11: Bits [127:96] of the parameter \a X are exracted.
/// \returns  An integer, whose lower 32 bits are selected from the 128-bit
///    integer vector parameter and the remaining bits are assigned zeros.
#define _mm_extract_epi32(X, N) \
  (int)__builtin_ia32_vec_ext_v4si((__v4si)(__m128i)(X), (int)(N))

#ifdef __x86_64__
/// Extracts a 64-bit element from the 128-bit integer vector of
///    [2 x i64], using the immediate value parameter \a N as a selector.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// long long _mm_extract_epi64(__m128i X, const int N);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPEXTRQ / PEXTRQ </c> instruction.
///
/// \param X
///    A 128-bit integer vector.
/// \param N
///    An immediate value. Bit [0] specifies which 64-bit vector element from
///    the argument \a X to return. \n
///    0: Bits [63:0] are returned. \n
///    1: Bits [127:64] are returned. \n
/// \returns  A 64-bit integer.
#define _mm_extract_epi64(X, N) \
  (long long)__builtin_ia32_vec_ext_v2di((__v2di)(__m128i)(X), (int)(N))
#endif /* __x86_64 */

/* SSE4 128-bit Packed Integer Comparisons.  */
/// Tests whether the specified bits in a 128-bit integer vector are all
///    zeros.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPTEST / PTEST </c> instruction.
///
/// \param __M
///    A 128-bit integer vector containing the bits to be tested.
/// \param __V
///    A 128-bit integer vector selecting which bits to test in operand \a __M.
/// \returns TRUE if the specified bits are all zeros; FALSE otherwise.
static __inline__ int __DEFAULT_FN_ATTRS
_mm_testz_si128(__m128i __M, __m128i __V)
{
  return __builtin_ia32_ptestz128((__v2di)__M, (__v2di)__V);
}

/// Tests whether the specified bits in a 128-bit integer vector are all
///    ones.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPTEST / PTEST </c> instruction.
///
/// \param __M
///    A 128-bit integer vector containing the bits to be tested.
/// \param __V
///    A 128-bit integer vector selecting which bits to test in operand \a __M.
/// \returns TRUE if the specified bits are all ones; FALSE otherwise.
static __inline__ int __DEFAULT_FN_ATTRS
_mm_testc_si128(__m128i __M, __m128i __V)
{
  return __builtin_ia32_ptestc128((__v2di)__M, (__v2di)__V);
}

/// Tests whether the specified bits in a 128-bit integer vector are
///    neither all zeros nor all ones.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPTEST / PTEST </c> instruction.
///
/// \param __M
///    A 128-bit integer vector containing the bits to be tested.
/// \param __V
///    A 128-bit integer vector selecting which bits to test in operand \a __M.
/// \returns TRUE if the specified bits are neither all zeros nor all ones;
///    FALSE otherwise.
static __inline__ int __DEFAULT_FN_ATTRS
_mm_testnzc_si128(__m128i __M, __m128i __V)
{
  return __builtin_ia32_ptestnzc128((__v2di)__M, (__v2di)__V);
}

/// Tests whether the specified bits in a 128-bit integer vector are all
///    ones.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_test_all_ones(__m128i V);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPTEST / PTEST </c> instruction.
///
/// \param V
///    A 128-bit integer vector containing the bits to be tested.
/// \returns TRUE if the bits specified in the operand are all set to 1; FALSE
///    otherwise.
#define _mm_test_all_ones(V) _mm_testc_si128((V), _mm_cmpeq_epi32((V), (V)))

/// Tests whether the specified bits in a 128-bit integer vector are
///    neither all zeros nor all ones.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_test_mix_ones_zeros(__m128i M, __m128i V);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPTEST / PTEST </c> instruction.
///
/// \param M
///    A 128-bit integer vector containing the bits to be tested.
/// \param V
///    A 128-bit integer vector selecting which bits to test in operand \a M.
/// \returns TRUE if the specified bits are neither all zeros nor all ones;
///    FALSE otherwise.
#define _mm_test_mix_ones_zeros(M, V) _mm_testnzc_si128((M), (V))

/// Tests whether the specified bits in a 128-bit integer vector are all
///    zeros.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_test_all_zeros(__m128i M, __m128i V);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPTEST / PTEST </c> instruction.
///
/// \param M
///    A 128-bit integer vector containing the bits to be tested.
/// \param V
///    A 128-bit integer vector selecting which bits to test in operand \a M.
/// \returns TRUE if the specified bits are all zeros; FALSE otherwise.
#define _mm_test_all_zeros(M, V) _mm_testz_si128 ((M), (V))

/* SSE4 64-bit Packed Integer Comparisons.  */
/// Compares each of the corresponding 64-bit values of the 128-bit
///    integer vectors for equality.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPCMPEQQ / PCMPEQQ </c> instruction.
///
/// \param __V1
///    A 128-bit integer vector.
/// \param __V2
///    A 128-bit integer vector.
/// \returns A 128-bit integer vector containing the comparison results.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cmpeq_epi64(__m128i __V1, __m128i __V2)
{
  return (__m128i)((__v2di)__V1 == (__v2di)__V2);
}

/* SSE4 Packed Integer Sign-Extension.  */
/// Sign-extends each of the lower eight 8-bit integer elements of a
///    128-bit vector of [16 x i8] to 16-bit values and returns them in a
///    128-bit vector of [8 x i16]. The upper eight elements of the input vector
///    are unused.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMOVSXBW / PMOVSXBW </c> instruction.
///
/// \param __V
///    A 128-bit vector of [16 x i8]. The lower eight 8-bit elements are sign-
///    extended to 16-bit values.
/// \returns A 128-bit vector of [8 x i16] containing the sign-extended values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cvtepi8_epi16(__m128i __V)
{
  /* This function always performs a signed extension, but __v16qi is a char
     which may be signed or unsigned, so use __v16qs. */
  return (__m128i)__builtin_convertvector(__builtin_shufflevector((__v16qs)__V, (__v16qs)__V, 0, 1, 2, 3, 4, 5, 6, 7), __v8hi);
}

/// Sign-extends each of the lower four 8-bit integer elements of a
///    128-bit vector of [16 x i8] to 32-bit values and returns them in a
///    128-bit vector of [4 x i32]. The upper twelve elements of the input
///    vector are unused.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMOVSXBD / PMOVSXBD </c> instruction.
///
/// \param __V
///    A 128-bit vector of [16 x i8]. The lower four 8-bit elements are
///    sign-extended to 32-bit values.
/// \returns A 128-bit vector of [4 x i32] containing the sign-extended values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cvtepi8_epi32(__m128i __V)
{
  /* This function always performs a signed extension, but __v16qi is a char
     which may be signed or unsigned, so use __v16qs. */
  return (__m128i)__builtin_convertvector(__builtin_shufflevector((__v16qs)__V, (__v16qs)__V, 0, 1, 2, 3), __v4si);
}

/// Sign-extends each of the lower two 8-bit integer elements of a
///    128-bit integer vector of [16 x i8] to 64-bit values and returns them in
///    a 128-bit vector of [2 x i64]. The upper fourteen elements of the input
///    vector are unused.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMOVSXBQ / PMOVSXBQ </c> instruction.
///
/// \param __V
///    A 128-bit vector of [16 x i8]. The lower two 8-bit elements are
///    sign-extended to 64-bit values.
/// \returns A 128-bit vector of [2 x i64] containing the sign-extended values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cvtepi8_epi64(__m128i __V)
{
  /* This function always performs a signed extension, but __v16qi is a char
     which may be signed or unsigned, so use __v16qs. */
  return (__m128i)__builtin_convertvector(__builtin_shufflevector((__v16qs)__V, (__v16qs)__V, 0, 1), __v2di);
}

/// Sign-extends each of the lower four 16-bit integer elements of a
///    128-bit integer vector of [8 x i16] to 32-bit values and returns them in
///    a 128-bit vector of [4 x i32]. The upper four elements of the input
///    vector are unused.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMOVSXWD / PMOVSXWD </c> instruction.
///
/// \param __V
///    A 128-bit vector of [8 x i16]. The lower four 16-bit elements are
///    sign-extended to 32-bit values.
/// \returns A 128-bit vector of [4 x i32] containing the sign-extended values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cvtepi16_epi32(__m128i __V)
{
  return (__m128i)__builtin_convertvector(__builtin_shufflevector((__v8hi)__V, (__v8hi)__V, 0, 1, 2, 3), __v4si);
}

/// Sign-extends each of the lower two 16-bit integer elements of a
///    128-bit integer vector of [8 x i16] to 64-bit values and returns them in
///    a 128-bit vector of [2 x i64]. The upper six elements of the input
///    vector are unused.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMOVSXWQ / PMOVSXWQ </c> instruction.
///
/// \param __V
///    A 128-bit vector of [8 x i16]. The lower two 16-bit elements are
///     sign-extended to 64-bit values.
/// \returns A 128-bit vector of [2 x i64] containing the sign-extended values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cvtepi16_epi64(__m128i __V)
{
  return (__m128i)__builtin_convertvector(__builtin_shufflevector((__v8hi)__V, (__v8hi)__V, 0, 1), __v2di);
}

/// Sign-extends each of the lower two 32-bit integer elements of a
///    128-bit integer vector of [4 x i32] to 64-bit values and returns them in
///    a 128-bit vector of [2 x i64]. The upper two elements of the input vector
///    are unused.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMOVSXDQ / PMOVSXDQ </c> instruction.
///
/// \param __V
///    A 128-bit vector of [4 x i32]. The lower two 32-bit elements are
///    sign-extended to 64-bit values.
/// \returns A 128-bit vector of [2 x i64] containing the sign-extended values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cvtepi32_epi64(__m128i __V)
{
  return (__m128i)__builtin_convertvector(__builtin_shufflevector((__v4si)__V, (__v4si)__V, 0, 1), __v2di);
}

/* SSE4 Packed Integer Zero-Extension.  */
/// Zero-extends each of the lower eight 8-bit integer elements of a
///    128-bit vector of [16 x i8] to 16-bit values and returns them in a
///    128-bit vector of [8 x i16]. The upper eight elements of the input vector
///    are unused.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMOVZXBW / PMOVZXBW </c> instruction.
///
/// \param __V
///    A 128-bit vector of [16 x i8]. The lower eight 8-bit elements are
///    zero-extended to 16-bit values.
/// \returns A 128-bit vector of [8 x i16] containing the zero-extended values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cvtepu8_epi16(__m128i __V)
{
  return (__m128i)__builtin_convertvector(__builtin_shufflevector((__v16qu)__V, (__v16qu)__V, 0, 1, 2, 3, 4, 5, 6, 7), __v8hi);
}

/// Zero-extends each of the lower four 8-bit integer elements of a
///    128-bit vector of [16 x i8] to 32-bit values and returns them in a
///    128-bit vector of [4 x i32]. The upper twelve elements of the input
///    vector are unused.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMOVZXBD / PMOVZXBD </c> instruction.
///
/// \param __V
///    A 128-bit vector of [16 x i8]. The lower four 8-bit elements are
///    zero-extended to 32-bit values.
/// \returns A 128-bit vector of [4 x i32] containing the zero-extended values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cvtepu8_epi32(__m128i __V)
{
  return (__m128i)__builtin_convertvector(__builtin_shufflevector((__v16qu)__V, (__v16qu)__V, 0, 1, 2, 3), __v4si);
}

/// Zero-extends each of the lower two 8-bit integer elements of a
///    128-bit integer vector of [16 x i8] to 64-bit values and returns them in
///    a 128-bit vector of [2 x i64]. The upper fourteen elements of the input
///    vector are unused.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMOVZXBQ / PMOVZXBQ </c> instruction.
///
/// \param __V
///    A 128-bit vector of [16 x i8]. The lower two 8-bit elements are
///    zero-extended to 64-bit values.
/// \returns A 128-bit vector of [2 x i64] containing the zero-extended values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cvtepu8_epi64(__m128i __V)
{
  return (__m128i)__builtin_convertvector(__builtin_shufflevector((__v16qu)__V, (__v16qu)__V, 0, 1), __v2di);
}

/// Zero-extends each of the lower four 16-bit integer elements of a
///    128-bit integer vector of [8 x i16] to 32-bit values and returns them in
///    a 128-bit vector of [4 x i32]. The upper four elements of the input
///    vector are unused.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMOVZXWD / PMOVZXWD </c> instruction.
///
/// \param __V
///    A 128-bit vector of [8 x i16]. The lower four 16-bit elements are
///    zero-extended to 32-bit values.
/// \returns A 128-bit vector of [4 x i32] containing the zero-extended values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cvtepu16_epi32(__m128i __V)
{
  return (__m128i)__builtin_convertvector(__builtin_shufflevector((__v8hu)__V, (__v8hu)__V, 0, 1, 2, 3), __v4si);
}

/// Zero-extends each of the lower two 16-bit integer elements of a
///    128-bit integer vector of [8 x i16] to 64-bit values and returns them in
///    a 128-bit vector of [2 x i64]. The upper six elements of the input vector
///    are unused.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMOVZXWQ / PMOVZXWQ </c> instruction.
///
/// \param __V
///    A 128-bit vector of [8 x i16]. The lower two 16-bit elements are
///    zero-extended to 64-bit values.
/// \returns A 128-bit vector of [2 x i64] containing the zero-extended values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cvtepu16_epi64(__m128i __V)
{
  return (__m128i)__builtin_convertvector(__builtin_shufflevector((__v8hu)__V, (__v8hu)__V, 0, 1), __v2di);
}

/// Zero-extends each of the lower two 32-bit integer elements of a
///    128-bit integer vector of [4 x i32] to 64-bit values and returns them in
///    a 128-bit vector of [2 x i64]. The upper two elements of the input vector
///    are unused.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPMOVZXDQ / PMOVZXDQ </c> instruction.
///
/// \param __V
///    A 128-bit vector of [4 x i32]. The lower two 32-bit elements are
///    zero-extended to 64-bit values.
/// \returns A 128-bit vector of [2 x i64] containing the zero-extended values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cvtepu32_epi64(__m128i __V)
{
  return (__m128i)__builtin_convertvector(__builtin_shufflevector((__v4su)__V, (__v4su)__V, 0, 1), __v2di);
}

/* SSE4 Pack with Unsigned Saturation.  */
/// Converts 32-bit signed integers from both 128-bit integer vector
///    operands into 16-bit unsigned integers, and returns the packed result.
///    Values greater than 0xFFFF are saturated to 0xFFFF. Values less than
///    0x0000 are saturated to 0x0000.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPACKUSDW / PACKUSDW </c> instruction.
///
/// \param __V1
///    A 128-bit vector of [4 x i32]. Each 32-bit element is treated as a
///    signed integer and is converted to a 16-bit unsigned integer with
///    saturation. Values greater than 0xFFFF are saturated to 0xFFFF. Values
///    less than 0x0000 are saturated to 0x0000. The converted [4 x i16] values
///    are written to the lower 64 bits of the result.
/// \param __V2
///    A 128-bit vector of [4 x i32]. Each 32-bit element is treated as a
///    signed integer and is converted to a 16-bit unsigned integer with
///    saturation. Values greater than 0xFFFF are saturated to 0xFFFF. Values
///    less than 0x0000 are saturated to 0x0000. The converted [4 x i16] values
///    are written to the higher 64 bits of the result.
/// \returns A 128-bit vector of [8 x i16] containing the converted values.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_packus_epi32(__m128i __V1, __m128i __V2)
{
  return (__m128i) __builtin_ia32_packusdw128((__v4si)__V1, (__v4si)__V2);
}

/* SSE4 Multiple Packed Sums of Absolute Difference.  */
/// Subtracts 8-bit unsigned integer values and computes the absolute
///    values of the differences to the corresponding bits in the destination.
///    Then sums of the absolute differences are returned according to the bit
///    fields in the immediate operand.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128i _mm_mpsadbw_epu8(__m128i X, __m128i Y, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VMPSADBW / MPSADBW </c> instruction.
///
/// \param X
///    A 128-bit vector of [16 x i8].
/// \param Y
///    A 128-bit vector of [16 x i8].
/// \param M
///    An 8-bit immediate operand specifying how the absolute differences are to
///    be calculated, according to the following algorithm:
///    \code
///    // M2 represents bit 2 of the immediate operand
///    // M10 represents bits [1:0] of the immediate operand
///    i = M2 * 4;
///    j = M10 * 4;
///    for (k = 0; k < 8; k = k + 1) {
///      d0 = abs(X[i + k + 0] - Y[j + 0]);
///      d1 = abs(X[i + k + 1] - Y[j + 1]);
///      d2 = abs(X[i + k + 2] - Y[j + 2]);
///      d3 = abs(X[i + k + 3] - Y[j + 3]);
///      r[k] = d0 + d1 + d2 + d3;
///    }
///    \endcode
/// \returns A 128-bit integer vector containing the sums of the sets of
///    absolute differences between both operands.
#define _mm_mpsadbw_epu8(X, Y, M) \
  (__m128i) __builtin_ia32_mpsadbw128((__v16qi)(__m128i)(X), \
                                      (__v16qi)(__m128i)(Y), (M))

/// Finds the minimum unsigned 16-bit element in the input 128-bit
///    vector of [8 x u16] and returns it and along with its index.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPHMINPOSUW / PHMINPOSUW </c>
/// instruction.
///
/// \param __V
///    A 128-bit vector of [8 x u16].
/// \returns A 128-bit value where bits [15:0] contain the minimum value found
///    in parameter \a __V, bits [18:16] contain the index of the minimum value
///    and the remaining bits are set to 0.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_minpos_epu16(__m128i __V)
{
  return (__m128i) __builtin_ia32_phminposuw128((__v8hi)__V);
}

/* Handle the sse4.2 definitions here. */

/* These definitions are normally in nmmintrin.h, but gcc puts them in here
   so we'll do the same.  */

#undef __DEFAULT_FN_ATTRS
#define __DEFAULT_FN_ATTRS __attribute__((__always_inline__, __nodebug__, __target__("sse4.2")))

/* These specify the type of data that we're comparing.  */
#define _SIDD_UBYTE_OPS                 0x00
#define _SIDD_UWORD_OPS                 0x01
#define _SIDD_SBYTE_OPS                 0x02
#define _SIDD_SWORD_OPS                 0x03

/* These specify the type of comparison operation.  */
#define _SIDD_CMP_EQUAL_ANY             0x00
#define _SIDD_CMP_RANGES                0x04
#define _SIDD_CMP_EQUAL_EACH            0x08
#define _SIDD_CMP_EQUAL_ORDERED         0x0c

/* These macros specify the polarity of the operation.  */
#define _SIDD_POSITIVE_POLARITY         0x00
#define _SIDD_NEGATIVE_POLARITY         0x10
#define _SIDD_MASKED_POSITIVE_POLARITY  0x20
#define _SIDD_MASKED_NEGATIVE_POLARITY  0x30

/* These macros are used in _mm_cmpXstri() to specify the return.  */
#define _SIDD_LEAST_SIGNIFICANT         0x00
#define _SIDD_MOST_SIGNIFICANT          0x40

/* These macros are used in _mm_cmpXstri() to specify the return.  */
#define _SIDD_BIT_MASK                  0x00
#define _SIDD_UNIT_MASK                 0x40

/* SSE4.2 Packed Comparison Intrinsics.  */
/// Uses the immediate operand \a M to perform a comparison of string
///    data with implicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns a 128-bit integer vector representing the result
///    mask of the comparison.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128i _mm_cmpistrm(__m128i A, __m128i B, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPISTRM / PCMPISTRM </c>
/// instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words, the type of comparison to perform, and the format of the return
///    value. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search \a B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement on the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B. \n
///    Bit [6]: Determines whether the result is zero-extended or expanded to 16
///             bytes. \n
///      0: The result is zero-extended to 16 bytes. \n
///      1: The result is expanded to 16 bytes (this expansion is performed by
///         repeating each bit 8 or 16 times).
/// \returns Returns a 128-bit integer vector representing the result mask of
///    the comparison.
#define _mm_cmpistrm(A, B, M) \
  (__m128i)__builtin_ia32_pcmpistrm128((__v16qi)(__m128i)(A), \
                                       (__v16qi)(__m128i)(B), (int)(M))

/// Uses the immediate operand \a M to perform a comparison of string
///    data with implicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns an integer representing the result index of the
///    comparison.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_cmpistri(__m128i A, __m128i B, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPISTRI / PCMPISTRI </c>
/// instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words, the type of comparison to perform, and the format of the return
///    value. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement on the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B. \n
///    Bit [6]: Determines whether the index of the lowest set bit or the
///             highest set bit is returned. \n
///      0: The index of the least significant set bit. \n
///      1: The index of the most significant set bit. \n
/// \returns Returns an integer representing the result index of the comparison.
#define _mm_cmpistri(A, B, M) \
  (int)__builtin_ia32_pcmpistri128((__v16qi)(__m128i)(A), \
                                   (__v16qi)(__m128i)(B), (int)(M))

/// Uses the immediate operand \a M to perform a comparison of string
///    data with explicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns a 128-bit integer vector representing the result
///    mask of the comparison.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// __m128i _mm_cmpestrm(__m128i A, int LA, __m128i B, int LB, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPESTRM / PCMPESTRM </c>
/// instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LA
///    An integer that specifies the length of the string in \a A.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LB
///    An integer that specifies the length of the string in \a B.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words, the type of comparison to perform, and the format of the return
///    value. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search \a B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement on the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B. \n
///    Bit [6]: Determines whether the result is zero-extended or expanded to 16
///             bytes. \n
///      0: The result is zero-extended to 16 bytes. \n
///      1: The result is expanded to 16 bytes (this expansion is performed by
///         repeating each bit 8 or 16 times). \n
/// \returns Returns a 128-bit integer vector representing the result mask of
///    the comparison.
#define _mm_cmpestrm(A, LA, B, LB, M) \
  (__m128i)__builtin_ia32_pcmpestrm128((__v16qi)(__m128i)(A), (int)(LA), \
                                       (__v16qi)(__m128i)(B), (int)(LB), \
                                       (int)(M))

/// Uses the immediate operand \a M to perform a comparison of string
///    data with explicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns an integer representing the result index of the
///    comparison.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_cmpestri(__m128i A, int LA, __m128i B, int LB, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPESTRI / PCMPESTRI </c>
/// instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LA
///    An integer that specifies the length of the string in \a A.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LB
///    An integer that specifies the length of the string in \a B.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words, the type of comparison to perform, and the format of the return
///    value. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement on the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B. \n
///    Bit [6]: Determines whether the index of the lowest set bit or the
///             highest set bit is returned. \n
///      0: The index of the least significant set bit. \n
///      1: The index of the most significant set bit. \n
/// \returns Returns an integer representing the result index of the comparison.
#define _mm_cmpestri(A, LA, B, LB, M) \
  (int)__builtin_ia32_pcmpestri128((__v16qi)(__m128i)(A), (int)(LA), \
                                   (__v16qi)(__m128i)(B), (int)(LB), \
                                   (int)(M))

/* SSE4.2 Packed Comparison Intrinsics and EFlag Reading.  */
/// Uses the immediate operand \a M to perform a comparison of string
///    data with implicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns 1 if the bit mask is zero and the length of the
///    string in \a B is the maximum, otherwise, returns 0.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_cmpistra(__m128i A, __m128i B, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPISTRI / PCMPISTRI </c>
/// instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words and the type of comparison to perform. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search \a B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement on the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B. \n
/// \returns Returns 1 if the bit mask is zero and the length of the string in
///    \a B is the maximum; otherwise, returns 0.
#define _mm_cmpistra(A, B, M) \
  (int)__builtin_ia32_pcmpistria128((__v16qi)(__m128i)(A), \
                                    (__v16qi)(__m128i)(B), (int)(M))

/// Uses the immediate operand \a M to perform a comparison of string
///    data with implicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns 1 if the bit mask is non-zero, otherwise, returns
///    0.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_cmpistrc(__m128i A, __m128i B, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPISTRI / PCMPISTRI </c>
/// instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words and the type of comparison to perform. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement on the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B.
/// \returns Returns 1 if the bit mask is non-zero, otherwise, returns 0.
#define _mm_cmpistrc(A, B, M) \
  (int)__builtin_ia32_pcmpistric128((__v16qi)(__m128i)(A), \
                                    (__v16qi)(__m128i)(B), (int)(M))

/// Uses the immediate operand \a M to perform a comparison of string
///    data with implicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns bit 0 of the resulting bit mask.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_cmpistro(__m128i A, __m128i B, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPISTRI / PCMPISTRI </c>
/// instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words and the type of comparison to perform. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement on the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B. \n
/// \returns Returns bit 0 of the resulting bit mask.
#define _mm_cmpistro(A, B, M) \
  (int)__builtin_ia32_pcmpistrio128((__v16qi)(__m128i)(A), \
                                    (__v16qi)(__m128i)(B), (int)(M))

/// Uses the immediate operand \a M to perform a comparison of string
///    data with implicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns 1 if the length of the string in \a A is less than
///    the maximum, otherwise, returns 0.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_cmpistrs(__m128i A, __m128i B, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPISTRI / PCMPISTRI </c>
/// instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words and the type of comparison to perform. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search \a B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement on the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B. \n
/// \returns Returns 1 if the length of the string in \a A is less than the
///    maximum, otherwise, returns 0.
#define _mm_cmpistrs(A, B, M) \
  (int)__builtin_ia32_pcmpistris128((__v16qi)(__m128i)(A), \
                                    (__v16qi)(__m128i)(B), (int)(M))

/// Uses the immediate operand \a M to perform a comparison of string
///    data with implicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns 1 if the length of the string in \a B is less than
///    the maximum, otherwise, returns 0.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_cmpistrz(__m128i A, __m128i B, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPISTRI / PCMPISTRI </c>
/// instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words and the type of comparison to perform. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search \a B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement on the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B.
/// \returns Returns 1 if the length of the string in \a B is less than the
///    maximum, otherwise, returns 0.
#define _mm_cmpistrz(A, B, M) \
  (int)__builtin_ia32_pcmpistriz128((__v16qi)(__m128i)(A), \
                                    (__v16qi)(__m128i)(B), (int)(M))

/// Uses the immediate operand \a M to perform a comparison of string
///    data with explicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns 1 if the bit mask is zero and the length of the
///    string in \a B is the maximum, otherwise, returns 0.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_cmpestra(__m128i A, int LA, __m128i B, int LB, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPESTRI / PCMPESTRI </c>
/// instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LA
///    An integer that specifies the length of the string in \a A.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LB
///    An integer that specifies the length of the string in \a B.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words and the type of comparison to perform. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search \a B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement on the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B.
/// \returns Returns 1 if the bit mask is zero and the length of the string in
///    \a B is the maximum, otherwise, returns 0.
#define _mm_cmpestra(A, LA, B, LB, M) \
  (int)__builtin_ia32_pcmpestria128((__v16qi)(__m128i)(A), (int)(LA), \
                                    (__v16qi)(__m128i)(B), (int)(LB), \
                                    (int)(M))

/// Uses the immediate operand \a M to perform a comparison of string
///    data with explicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns 1 if the resulting mask is non-zero, otherwise,
///    returns 0.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_cmpestrc(__m128i A, int LA, __m128i B, int LB, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPESTRI / PCMPESTRI </c>
/// instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LA
///    An integer that specifies the length of the string in \a A.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LB
///    An integer that specifies the length of the string in \a B.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words and the type of comparison to perform. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search \a B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement on the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B. \n
/// \returns Returns 1 if the resulting mask is non-zero, otherwise, returns 0.
#define _mm_cmpestrc(A, LA, B, LB, M) \
  (int)__builtin_ia32_pcmpestric128((__v16qi)(__m128i)(A), (int)(LA), \
                                    (__v16qi)(__m128i)(B), (int)(LB), \
                                    (int)(M))

/// Uses the immediate operand \a M to perform a comparison of string
///    data with explicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns bit 0 of the resulting bit mask.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_cmpestro(__m128i A, int LA, __m128i B, int LB, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPESTRI / PCMPESTRI </c>
/// instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LA
///    An integer that specifies the length of the string in \a A.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LB
///    An integer that specifies the length of the string in \a B.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words and the type of comparison to perform. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search \a B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement on the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B.
/// \returns Returns bit 0 of the resulting bit mask.
#define _mm_cmpestro(A, LA, B, LB, M) \
  (int)__builtin_ia32_pcmpestrio128((__v16qi)(__m128i)(A), (int)(LA), \
                                    (__v16qi)(__m128i)(B), (int)(LB), \
                                    (int)(M))

/// Uses the immediate operand \a M to perform a comparison of string
///    data with explicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns 1 if the length of the string in \a A is less than
///    the maximum, otherwise, returns 0.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_cmpestrs(__m128i A, int LA, __m128i B, int LB, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPESTRI / PCMPESTRI </c>
/// instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LA
///    An integer that specifies the length of the string in \a A.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LB
///    An integer that specifies the length of the string in \a B.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words and the type of comparison to perform. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search \a B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement in the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B. \n
/// \returns Returns 1 if the length of the string in \a A is less than the
///    maximum, otherwise, returns 0.
#define _mm_cmpestrs(A, LA, B, LB, M) \
  (int)__builtin_ia32_pcmpestris128((__v16qi)(__m128i)(A), (int)(LA), \
                                    (__v16qi)(__m128i)(B), (int)(LB), \
                                    (int)(M))

/// Uses the immediate operand \a M to perform a comparison of string
///    data with explicitly defined lengths that is contained in source operands
///    \a A and \a B. Returns 1 if the length of the string in \a B is less than
///    the maximum, otherwise, returns 0.
///
/// \headerfile <x86intrin.h>
///
/// \code
/// int _mm_cmpestrz(__m128i A, int LA, __m128i B, int LB, const int M);
/// \endcode
///
/// This intrinsic corresponds to the <c> VPCMPESTRI </c> instruction.
///
/// \param A
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LA
///    An integer that specifies the length of the string in \a A.
/// \param B
///    A 128-bit integer vector containing one of the source operands to be
///    compared.
/// \param LB
///    An integer that specifies the length of the string in \a B.
/// \param M
///    An 8-bit immediate operand specifying whether the characters are bytes or
///    words and the type of comparison to perform. \n
///    Bits [1:0]: Determine source data format. \n
///      00: 16 unsigned bytes  \n
///      01: 8 unsigned words \n
///      10: 16 signed bytes \n
///      11: 8 signed words \n
///    Bits [3:2]: Determine comparison type and aggregation method. \n
///      00: Subset: Each character in \a B is compared for equality with all
///          the characters in \a A. \n
///      01: Ranges: Each character in \a B is compared to \a A. The comparison
///          basis is greater than or equal for even-indexed elements in \a A,
///          and less than or equal for odd-indexed elements in \a A. \n
///      10: Match: Compare each pair of corresponding characters in \a A and
///          \a B for equality. \n
///      11: Substring: Search \a B for substring matches of \a A. \n
///    Bits [5:4]: Determine whether to perform a one's complement on the bit
///                mask of the comparison results. \n
///      00: No effect. \n
///      01: Negate the bit mask. \n
///      10: No effect. \n
///      11: Negate the bit mask only for bits with an index less than or equal
///          to the size of \a A or \a B.
/// \returns Returns 1 if the length of the string in \a B is less than the
///    maximum, otherwise, returns 0.
#define _mm_cmpestrz(A, LA, B, LB, M) \
  (int)__builtin_ia32_pcmpestriz128((__v16qi)(__m128i)(A), (int)(LA), \
                                    (__v16qi)(__m128i)(B), (int)(LB), \
                                    (int)(M))

/* SSE4.2 Compare Packed Data -- Greater Than.  */
/// Compares each of the corresponding 64-bit values of the 128-bit
///    integer vectors to determine if the values in the first operand are
///    greater than those in the second operand.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> VPCMPGTQ / PCMPGTQ </c> instruction.
///
/// \param __V1
///    A 128-bit integer vector.
/// \param __V2
///    A 128-bit integer vector.
/// \returns A 128-bit integer vector containing the comparison results.
static __inline__ __m128i __DEFAULT_FN_ATTRS
_mm_cmpgt_epi64(__m128i __V1, __m128i __V2)
{
  return (__m128i)((__v2di)__V1 > (__v2di)__V2);
}

/* SSE4.2 Accumulate CRC32.  */
/// Adds the unsigned integer operand to the CRC-32C checksum of the
///    unsigned char operand.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> CRC32B </c> instruction.
///
/// \param __C
///    An unsigned integer operand to add to the CRC-32C checksum of operand
///    \a  __D.
/// \param __D
///    An unsigned 8-bit integer operand used to compute the CRC-32C checksum.
/// \returns The result of adding operand \a __C to the CRC-32C checksum of
///    operand \a __D.
static __inline__ unsigned int __DEFAULT_FN_ATTRS
_mm_crc32_u8(unsigned int __C, unsigned char __D)
{
  return __builtin_ia32_crc32qi(__C, __D);
}

/// Adds the unsigned integer operand to the CRC-32C checksum of the
///    unsigned short operand.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> CRC32W </c> instruction.
///
/// \param __C
///    An unsigned integer operand to add to the CRC-32C checksum of operand
///    \a __D.
/// \param __D
///    An unsigned 16-bit integer operand used to compute the CRC-32C checksum.
/// \returns The result of adding operand \a __C to the CRC-32C checksum of
///    operand \a __D.
static __inline__ unsigned int __DEFAULT_FN_ATTRS
_mm_crc32_u16(unsigned int __C, unsigned short __D)
{
  return __builtin_ia32_crc32hi(__C, __D);
}

/// Adds the first unsigned integer operand to the CRC-32C checksum of
///    the second unsigned integer operand.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> CRC32L </c> instruction.
///
/// \param __C
///    An unsigned integer operand to add to the CRC-32C checksum of operand
///    \a __D.
/// \param __D
///    An unsigned 32-bit integer operand used to compute the CRC-32C checksum.
/// \returns The result of adding operand \a __C to the CRC-32C checksum of
///    operand \a __D.
static __inline__ unsigned int __DEFAULT_FN_ATTRS
_mm_crc32_u32(unsigned int __C, unsigned int __D)
{
  return __builtin_ia32_crc32si(__C, __D);
}

#ifdef __x86_64__
/// Adds the unsigned integer operand to the CRC-32C checksum of the
///    unsigned 64-bit integer operand.
///
/// \headerfile <x86intrin.h>
///
/// This intrinsic corresponds to the <c> CRC32Q </c> instruction.
///
/// \param __C
///    An unsigned integer operand to add to the CRC-32C checksum of operand
///    \a __D.
/// \param __D
///    An unsigned 64-bit integer operand used to compute the CRC-32C checksum.
/// \returns The result of adding operand \a __C to the CRC-32C checksum of
///    operand \a __D.
static __inline__ unsigned long long __DEFAULT_FN_ATTRS
_mm_crc32_u64(unsigned long long __C, unsigned long long __D)
{
  return __builtin_ia32_crc32di(__C, __D);
}
#endif /* __x86_64__ */

#undef __DEFAULT_FN_ATTRS

#include <popcntintrin.h>

#endif /* __SMMINTRIN_H */