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
|
/*
* xxHash - Extremely Fast Hash algorithm
* Header File
* Copyright (C) 2012-2020 Yann Collet
*
* BSD 2-Clause License (https://www.opensource.org/licenses/bsd-license.php)
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are
* met:
*
* * Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* * Redistributions in binary form must reproduce the above
* copyright notice, this list of conditions and the following disclaimer
* in the documentation and/or other materials provided with the
* distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
* A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
* OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*
* You can contact the author at:
* - xxHash homepage: https://www.xxhash.com
* - xxHash source repository: https://github.com/Cyan4973/xxHash
*/
/* TODO: update */
/* Notice extracted from xxHash homepage:
xxHash is an extremely fast hash algorithm, running at RAM speed limits.
It also successfully passes all tests from the SMHasher suite.
Comparison (single thread, Windows Seven 32 bits, using SMHasher on a Core 2 Duo @3GHz)
Name Speed Q.Score Author
xxHash 5.4 GB/s 10
CrapWow 3.2 GB/s 2 Andrew
MumurHash 3a 2.7 GB/s 10 Austin Appleby
SpookyHash 2.0 GB/s 10 Bob Jenkins
SBox 1.4 GB/s 9 Bret Mulvey
Lookup3 1.2 GB/s 9 Bob Jenkins
SuperFastHash 1.2 GB/s 1 Paul Hsieh
CityHash64 1.05 GB/s 10 Pike & Alakuijala
FNV 0.55 GB/s 5 Fowler, Noll, Vo
CRC32 0.43 GB/s 9
MD5-32 0.33 GB/s 10 Ronald L. Rivest
SHA1-32 0.28 GB/s 10
Q.Score is a measure of quality of the hash function.
It depends on successfully passing SMHasher test set.
10 is a perfect score.
Note: SMHasher's CRC32 implementation is not the fastest one.
Other speed-oriented implementations can be faster,
especially in combination with PCLMUL instruction:
https://fastcompression.blogspot.com/2019/03/presenting-xxh3.html?showComment=1552696407071#c3490092340461170735
A 64-bit version, named XXH64, is available since r35.
It offers much better speed, but for 64-bit applications only.
Name Speed on 64 bits Speed on 32 bits
XXH64 13.8 GB/s 1.9 GB/s
XXH32 6.8 GB/s 6.0 GB/s
*/
#if defined (__cplusplus)
extern "C" {
#endif
/* ****************************
* INLINE mode
******************************/
/*!
* XXH_INLINE_ALL (and XXH_PRIVATE_API)
* Use these build macros to inline xxhash into the target unit.
* Inlining improves performance on small inputs, especially when the length is
* expressed as a compile-time constant:
*
* https://fastcompression.blogspot.com/2018/03/xxhash-for-small-keys-impressive-power.html
*
* It also keeps xxHash symbols private to the unit, so they are not exported.
*
* Usage:
* #define XXH_INLINE_ALL
* #include "xxhash.h"
*
* Do not compile and link xxhash.o as a separate object, as it is not useful.
*/
#if (defined(XXH_INLINE_ALL) || defined(XXH_PRIVATE_API)) \
&& !defined(XXH_INLINE_ALL_31684351384)
/* this section should be traversed only once */
# define XXH_INLINE_ALL_31684351384
/* give access to the advanced API, required to compile implementations */
# undef XXH_STATIC_LINKING_ONLY /* avoid macro redef */
# define XXH_STATIC_LINKING_ONLY
/* make all functions private */
# undef XXH_PUBLIC_API
# if defined(__GNUC__)
# define XXH_PUBLIC_API static __inline __attribute__((unused))
# elif defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */)
# define XXH_PUBLIC_API static inline
# elif defined(_MSC_VER)
# define XXH_PUBLIC_API static __inline
# else
/* note: this version may generate warnings for unused static functions */
# define XXH_PUBLIC_API static
# endif
/*
* This part deals with the special case where a unit wants to inline xxHash,
* but "xxhash.h" has previously been included without XXH_INLINE_ALL, such
* as part of some previously included *.h header file.
* Without further action, the new include would just be ignored,
* and functions would effectively _not_ be inlined (silent failure).
* The following macros solve this situation by prefixing all inlined names,
* avoiding naming collision with previous inclusions.
*/
# ifdef XXH_NAMESPACE
# error "XXH_INLINE_ALL with XXH_NAMESPACE is not supported"
/*
* Note: Alternative: #undef all symbols (it's a pretty large list).
* Without #error: it compiles, but functions are actually not inlined.
*/
# endif
# define XXH_NAMESPACE XXH_INLINE_
/*
* Some identifiers (enums, type names) are not symbols, but they must
* still be renamed to avoid redeclaration.
* Alternative solution: do not redeclare them.
* However, this requires some #ifdefs, and is a more dispersed action.
* Meanwhile, renaming can be achieved in a single block
*/
# define XXH_IPREF(Id) XXH_INLINE_ ## Id
# define XXH_OK XXH_IPREF(XXH_OK)
# define XXH_ERROR XXH_IPREF(XXH_ERROR)
# define XXH_errorcode XXH_IPREF(XXH_errorcode)
# define XXH32_canonical_t XXH_IPREF(XXH32_canonical_t)
# define XXH64_canonical_t XXH_IPREF(XXH64_canonical_t)
# define XXH128_canonical_t XXH_IPREF(XXH128_canonical_t)
# define XXH32_state_s XXH_IPREF(XXH32_state_s)
# define XXH32_state_t XXH_IPREF(XXH32_state_t)
# define XXH64_state_s XXH_IPREF(XXH64_state_s)
# define XXH64_state_t XXH_IPREF(XXH64_state_t)
# define XXH3_state_s XXH_IPREF(XXH3_state_s)
# define XXH3_state_t XXH_IPREF(XXH3_state_t)
# define XXH128_hash_t XXH_IPREF(XXH128_hash_t)
/* Ensure the header is parsed again, even if it was previously included */
# undef XXHASH_H_5627135585666179
# undef XXHASH_H_STATIC_13879238742
#endif /* XXH_INLINE_ALL || XXH_PRIVATE_API */
/* ****************************************************************
* Stable API
*****************************************************************/
#ifndef XXHASH_H_5627135585666179
#define XXHASH_H_5627135585666179 1
/* specific declaration modes for Windows */
#if !defined(XXH_INLINE_ALL) && !defined(XXH_PRIVATE_API)
# if defined(WIN32) && defined(_MSC_VER) && (defined(XXH_IMPORT) || defined(XXH_EXPORT))
# ifdef XXH_EXPORT
# define XXH_PUBLIC_API __declspec(dllexport)
# elif XXH_IMPORT
# define XXH_PUBLIC_API __declspec(dllimport)
# endif
# else
# define XXH_PUBLIC_API /* do nothing */
# endif
#endif
/*!
* XXH_NAMESPACE, aka Namespace Emulation:
*
* If you want to include _and expose_ xxHash functions from within your own
* library, but also want to avoid symbol collisions with other libraries which
* may also include xxHash, you can use XXH_NAMESPACE to automatically prefix
* any public symbol from xxhash library with the value of XXH_NAMESPACE
* (therefore, avoid empty or numeric values).
*
* Note that no change is required within the calling program as long as it
* includes `xxhash.h`: Regular symbol names will be automatically translated
* by this header.
*/
#ifdef XXH_NAMESPACE
# define XXH_CAT(A,B) A##B
# define XXH_NAME2(A,B) XXH_CAT(A,B)
# define XXH_versionNumber XXH_NAME2(XXH_NAMESPACE, XXH_versionNumber)
# define XXH32 XXH_NAME2(XXH_NAMESPACE, XXH32)
# define XXH32_createState XXH_NAME2(XXH_NAMESPACE, XXH32_createState)
# define XXH32_freeState XXH_NAME2(XXH_NAMESPACE, XXH32_freeState)
# define XXH32_reset XXH_NAME2(XXH_NAMESPACE, XXH32_reset)
# define XXH32_update XXH_NAME2(XXH_NAMESPACE, XXH32_update)
# define XXH32_digest XXH_NAME2(XXH_NAMESPACE, XXH32_digest)
# define XXH32_copyState XXH_NAME2(XXH_NAMESPACE, XXH32_copyState)
# define XXH32_canonicalFromHash XXH_NAME2(XXH_NAMESPACE, XXH32_canonicalFromHash)
# define XXH32_hashFromCanonical XXH_NAME2(XXH_NAMESPACE, XXH32_hashFromCanonical)
# define XXH64 XXH_NAME2(XXH_NAMESPACE, XXH64)
# define XXH64_createState XXH_NAME2(XXH_NAMESPACE, XXH64_createState)
# define XXH64_freeState XXH_NAME2(XXH_NAMESPACE, XXH64_freeState)
# define XXH64_reset XXH_NAME2(XXH_NAMESPACE, XXH64_reset)
# define XXH64_update XXH_NAME2(XXH_NAMESPACE, XXH64_update)
# define XXH64_digest XXH_NAME2(XXH_NAMESPACE, XXH64_digest)
# define XXH64_copyState XXH_NAME2(XXH_NAMESPACE, XXH64_copyState)
# define XXH64_canonicalFromHash XXH_NAME2(XXH_NAMESPACE, XXH64_canonicalFromHash)
# define XXH64_hashFromCanonical XXH_NAME2(XXH_NAMESPACE, XXH64_hashFromCanonical)
#endif
/* *************************************
* Version
***************************************/
#define XXH_VERSION_MAJOR 0
#define XXH_VERSION_MINOR 7
#define XXH_VERSION_RELEASE 4
#define XXH_VERSION_NUMBER (XXH_VERSION_MAJOR *100*100 + XXH_VERSION_MINOR *100 + XXH_VERSION_RELEASE)
XXH_PUBLIC_API unsigned XXH_versionNumber (void);
/* ****************************
* Definitions
******************************/
#include <stddef.h> /* size_t */
typedef enum { XXH_OK=0, XXH_ERROR } XXH_errorcode;
/*-**********************************************************************
* 32-bit hash
************************************************************************/
#if !defined (__VMS) \
&& (defined (__cplusplus) \
|| (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) )
# include <stdint.h>
typedef uint32_t XXH32_hash_t;
#else
# include <limits.h>
# if UINT_MAX == 0xFFFFFFFFUL
typedef unsigned int XXH32_hash_t;
# else
# if ULONG_MAX == 0xFFFFFFFFUL
typedef unsigned long XXH32_hash_t;
# else
# error "unsupported platform: need a 32-bit type"
# endif
# endif
#endif
/*!
* XXH32():
* Calculate the 32-bit hash of sequence "length" bytes stored at memory address "input".
* The memory between input & input+length must be valid (allocated and read-accessible).
* "seed" can be used to alter the result predictably.
* Speed on Core 2 Duo @ 3 GHz (single thread, SMHasher benchmark): 5.4 GB/s
*
* Note: XXH3 provides competitive speed for both 32-bit and 64-bit systems,
* and offers true 64/128 bit hash results. It provides a superior level of
* dispersion, and greatly reduces the risks of collisions.
*/
XXH_PUBLIC_API XXH32_hash_t XXH32 (const void* input, size_t length, XXH32_hash_t seed);
/******* Streaming *******/
/*
* Streaming functions generate the xxHash value from an incrememtal input.
* This method is slower than single-call functions, due to state management.
* For small inputs, prefer `XXH32()` and `XXH64()`, which are better optimized.
*
* An XXH state must first be allocated using `XXH*_createState()`.
*
* Start a new hash by initializing the state with a seed using `XXH*_reset()`.
*
* Then, feed the hash state by calling `XXH*_update()` as many times as necessary.
*
* The function returns an error code, with 0 meaning OK, and any other value
* meaning there is an error.
*
* Finally, a hash value can be produced anytime, by using `XXH*_digest()`.
* This function returns the nn-bits hash as an int or long long.
*
* It's still possible to continue inserting input into the hash state after a
* digest, and generate new hash values later on by invoking `XXH*_digest()`.
*
* When done, release the state using `XXH*_freeState()`.
*/
typedef struct XXH32_state_s XXH32_state_t; /* incomplete type */
XXH_PUBLIC_API XXH32_state_t* XXH32_createState(void);
XXH_PUBLIC_API XXH_errorcode XXH32_freeState(XXH32_state_t* statePtr);
XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* dst_state, const XXH32_state_t* src_state);
XXH_PUBLIC_API XXH_errorcode XXH32_reset (XXH32_state_t* statePtr, XXH32_hash_t seed);
XXH_PUBLIC_API XXH_errorcode XXH32_update (XXH32_state_t* statePtr, const void* input, size_t length);
XXH_PUBLIC_API XXH32_hash_t XXH32_digest (const XXH32_state_t* statePtr);
/******* Canonical representation *******/
/*
* The default return values from XXH functions are unsigned 32 and 64 bit
* integers.
* This the simplest and fastest format for further post-processing.
*
* However, this leaves open the question of what is the order on the byte level,
* since little and big endian conventions will store the same number differently.
*
* The canonical representation settles this issue by mandating big-endian
* convention, the same convention as human-readable numbers (large digits first).
*
* When writing hash values to storage, sending them over a network, or printing
* them, it's highly recommended to use the canonical representation to ensure
* portability across a wider range of systems, present and future.
*
* The following functions allow transformation of hash values to and from
* canonical format.
*/
typedef struct { unsigned char digest[4]; } XXH32_canonical_t;
XXH_PUBLIC_API void XXH32_canonicalFromHash(XXH32_canonical_t* dst, XXH32_hash_t hash);
XXH_PUBLIC_API XXH32_hash_t XXH32_hashFromCanonical(const XXH32_canonical_t* src);
#ifndef XXH_NO_LONG_LONG
/*-**********************************************************************
* 64-bit hash
************************************************************************/
#if !defined (__VMS) \
&& (defined (__cplusplus) \
|| (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) )
# include <stdint.h>
typedef uint64_t XXH64_hash_t;
#else
/* the following type must have a width of 64-bit */
typedef unsigned long long XXH64_hash_t;
#endif
/*!
* XXH64():
* Returns the 64-bit hash of sequence of length @length stored at memory
* address @input.
* @seed can be used to alter the result predictably.
*
* This function usually runs faster on 64-bit systems, but slower on 32-bit
* systems (see benchmark).
*
* Note: XXH3 provides competitive speed for both 32-bit and 64-bit systems,
* and offers true 64/128 bit hash results. It provides a superior level of
* dispersion, and greatly reduces the risks of collisions.
*/
XXH_PUBLIC_API XXH64_hash_t XXH64 (const void* input, size_t length, XXH64_hash_t seed);
/******* Streaming *******/
typedef struct XXH64_state_s XXH64_state_t; /* incomplete type */
XXH_PUBLIC_API XXH64_state_t* XXH64_createState(void);
XXH_PUBLIC_API XXH_errorcode XXH64_freeState(XXH64_state_t* statePtr);
XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* dst_state, const XXH64_state_t* src_state);
XXH_PUBLIC_API XXH_errorcode XXH64_reset (XXH64_state_t* statePtr, XXH64_hash_t seed);
XXH_PUBLIC_API XXH_errorcode XXH64_update (XXH64_state_t* statePtr, const void* input, size_t length);
XXH_PUBLIC_API XXH64_hash_t XXH64_digest (const XXH64_state_t* statePtr);
/******* Canonical representation *******/
typedef struct { unsigned char digest[sizeof(XXH64_hash_t)]; } XXH64_canonical_t;
XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t hash);
XXH_PUBLIC_API XXH64_hash_t XXH64_hashFromCanonical(const XXH64_canonical_t* src);
#endif /* XXH_NO_LONG_LONG */
#endif /* XXHASH_H_5627135585666179 */
#if defined(XXH_STATIC_LINKING_ONLY) && !defined(XXHASH_H_STATIC_13879238742)
#define XXHASH_H_STATIC_13879238742
/* ****************************************************************************
* This section contains declarations which are not guaranteed to remain stable.
* They may change in future versions, becoming incompatible with a different
* version of the library.
* These declarations should only be used with static linking.
* Never use them in association with dynamic linking!
***************************************************************************** */
/*
* These definitions are only present to allow static allocation of an XXH
* state, for example, on the stack or in a struct.
* Never **ever** access members directly.
*/
struct XXH32_state_s {
XXH32_hash_t total_len_32;
XXH32_hash_t large_len;
XXH32_hash_t v1;
XXH32_hash_t v2;
XXH32_hash_t v3;
XXH32_hash_t v4;
XXH32_hash_t mem32[4];
XXH32_hash_t memsize;
XXH32_hash_t reserved; /* never read nor write, might be removed in a future version */
}; /* typedef'd to XXH32_state_t */
#ifndef XXH_NO_LONG_LONG /* defined when there is no 64-bit support */
struct XXH64_state_s {
XXH64_hash_t total_len;
XXH64_hash_t v1;
XXH64_hash_t v2;
XXH64_hash_t v3;
XXH64_hash_t v4;
XXH64_hash_t mem64[4];
XXH32_hash_t memsize;
XXH32_hash_t reserved32; /* required for padding anyway */
XXH64_hash_t reserved64; /* never read nor write, might be removed in a future version */
}; /* typedef'd to XXH64_state_t */
/*-**********************************************************************
* XXH3
* New experimental hash
************************************************************************/
/* ************************************************************************
* XXH3 is a new hash algorithm featuring:
* - Improved speed for both small and large inputs
* - True 64-bit and 128-bit outputs
* - SIMD acceleration
* - Improved 32-bit viability
*
* Speed analysis methodology is explained here:
*
* https://fastcompression.blogspot.com/2019/03/presenting-xxh3.html
*
* In general, expect XXH3 to run about ~2x faster on large inputs and >3x
* faster on small ones compared to XXH64, though exact differences depend on
* the platform.
*
* The algorithm is portable: Like XXH32 and XXH64, it generates the same hash
* on all platforms.
*
* It benefits greatly from SIMD and 64-bit arithmetic, but does not require it.
*
* Almost all 32-bit and 64-bit targets that can run XXH32 smoothly can run
* XXH3 at competitive speeds, even if XXH64 runs slowly. Further details are
* explained in the implementation.
*
* Optimized implementations are provided for AVX512, AVX2, SSE2, NEON, POWER8,
* ZVector and scalar targets. This can be controlled with the XXH_VECTOR macro.
*
* XXH3 offers 2 variants, _64bits and _128bits.
* When only 64 bits are needed, prefer calling the _64bits variant, as it
* reduces the amount of mixing, resulting in faster speed on small inputs.
*
* It's also generally simpler to manipulate a scalar return type than a struct.
*
* The 128-bit version adds additional strength, but it is slightly slower.
*
* The XXH3 algorithm is still in development.
* The results it produces may still change in future versions.
*
* Results produced by v0.7.x are not comparable with results from v0.7.y.
* However, the API is completely stable, and it can safely be used for
* ephemeral data (local sessions).
*
* Avoid storing values in long-term storage until the algorithm is finalized.
* XXH3's return values will be officially finalized upon reaching v0.8.0.
*
* After which, return values of XXH3 and XXH128 will no longer change in
* future versions.
*
* The API supports one-shot hashing, streaming mode, and custom secrets.
*/
#ifdef XXH_NAMESPACE
# define XXH3_64bits XXH_NAME2(XXH_NAMESPACE, XXH3_64bits)
# define XXH3_64bits_withSecret XXH_NAME2(XXH_NAMESPACE, XXH3_64bits_withSecret)
# define XXH3_64bits_withSeed XXH_NAME2(XXH_NAMESPACE, XXH3_64bits_withSeed)
# define XXH3_createState XXH_NAME2(XXH_NAMESPACE, XXH3_createState)
# define XXH3_freeState XXH_NAME2(XXH_NAMESPACE, XXH3_freeState)
# define XXH3_copyState XXH_NAME2(XXH_NAMESPACE, XXH3_copyState)
# define XXH3_64bits_reset XXH_NAME2(XXH_NAMESPACE, XXH3_64bits_reset)
# define XXH3_64bits_reset_withSeed XXH_NAME2(XXH_NAMESPACE, XXH3_64bits_reset_withSeed)
# define XXH3_64bits_reset_withSecret XXH_NAME2(XXH_NAMESPACE, XXH3_64bits_reset_withSecret)
# define XXH3_64bits_update XXH_NAME2(XXH_NAMESPACE, XXH3_64bits_update)
# define XXH3_64bits_digest XXH_NAME2(XXH_NAMESPACE, XXH3_64bits_digest)
# define XXH3_generateSecret XXH_NAME2(XXH_NAMESPACE, XXH3_generateSecret)
#endif
/* XXH3_64bits():
* default 64-bit variant, using default secret and default seed of 0.
* It's the fastest variant. */
XXH_PUBLIC_API XXH64_hash_t XXH3_64bits(const void* data, size_t len);
/*
* XXH3_64bits_withSeed():
* This variant generates a custom secret on the fly based on the default
* secret, altered using the `seed` value.
* While this operation is decently fast, note that it's not completely free.
* Note: seed==0 produces the same results as XXH3_64bits().
*/
XXH_PUBLIC_API XXH64_hash_t XXH3_64bits_withSeed(const void* data, size_t len, XXH64_hash_t seed);
/*
* XXH3_64bits_withSecret():
* It's possible to provide any blob of bytes as a "secret" to generate the hash.
* This makes it more difficult for an external actor to prepare an intentional collision.
* The main condition is that secretSize *must* be large enough (>= XXH3_SECRET_SIZE_MIN).
* However, the quality of the hash highly depends on the secret's entropy.
* Technically, the secret must look like a bunch of random bytes.
* Avoid "trivial" or structured data such as repeated sequences or a text document.
* Whenever unsure about the "randonmess" of the blob of bytes,
* consider relabelling it as a "custom seed" instead,
* and employ "XXH3_generateSecret()" (see below)
* to generate a high quality secret derived from this custom seed.
*/
#define XXH3_SECRET_SIZE_MIN 136
XXH_PUBLIC_API XXH64_hash_t XXH3_64bits_withSecret(const void* data, size_t len, const void* secret, size_t secretSize);
/* streaming 64-bit */
#if defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 201112L) /* C11+ */
# include <stdalign.h>
# define XXH_ALIGN(n) alignas(n)
#elif defined(__GNUC__)
# define XXH_ALIGN(n) __attribute__ ((aligned(n)))
#elif defined(_MSC_VER)
# define XXH_ALIGN(n) __declspec(align(n))
#else
# define XXH_ALIGN(n) /* disabled */
#endif
/* Old GCC versions only accept the attribute after the type in structures. */
#if !(defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 201112L)) /* C11+ */ \
&& defined(__GNUC__)
# define XXH_ALIGN_MEMBER(align, type) type XXH_ALIGN(align)
#else
# define XXH_ALIGN_MEMBER(align, type) XXH_ALIGN(align) type
#endif
typedef struct XXH3_state_s XXH3_state_t;
#define XXH3_INTERNALBUFFER_SIZE 256
#define XXH3_SECRET_DEFAULT_SIZE 192
struct XXH3_state_s {
XXH_ALIGN_MEMBER(64, XXH64_hash_t acc[8]);
/* used to store a custom secret generated from a seed */
XXH_ALIGN_MEMBER(64, unsigned char customSecret[XXH3_SECRET_DEFAULT_SIZE]);
XXH_ALIGN_MEMBER(64, unsigned char buffer[XXH3_INTERNALBUFFER_SIZE]);
XXH32_hash_t bufferedSize;
XXH32_hash_t reserved32;
size_t nbStripesPerBlock;
size_t nbStripesSoFar;
size_t secretLimit;
XXH64_hash_t totalLen;
XXH64_hash_t seed;
XXH64_hash_t reserved64;
const unsigned char* extSecret; /* reference to external secret;
* if == NULL, use .customSecret instead */
/* note: there may be some padding at the end due to alignment on 64 bytes */
}; /* typedef'd to XXH3_state_t */
#undef XXH_ALIGN_MEMBER
/*
* Streaming requires state maintenance.
* This operation costs memory and CPU.
* As a consequence, streaming is slower than one-shot hashing.
* For better performance, prefer one-shot functions whenever possible.
*/
XXH_PUBLIC_API XXH3_state_t* XXH3_createState(void);
XXH_PUBLIC_API XXH_errorcode XXH3_freeState(XXH3_state_t* statePtr);
XXH_PUBLIC_API void XXH3_copyState(XXH3_state_t* dst_state, const XXH3_state_t* src_state);
/*
* XXH3_64bits_reset():
* Initialize with the default parameters.
* The result will be equivalent to `XXH3_64bits()`.
*/
XXH_PUBLIC_API XXH_errorcode XXH3_64bits_reset(XXH3_state_t* statePtr);
/*
* XXH3_64bits_reset_withSeed():
* Generate a custom secret from `seed`, and store it into `statePtr`.
* digest will be equivalent to `XXH3_64bits_withSeed()`.
*/
XXH_PUBLIC_API XXH_errorcode XXH3_64bits_reset_withSeed(XXH3_state_t* statePtr, XXH64_hash_t seed);
/*
* XXH3_64bits_reset_withSecret():
* `secret` is referenced, it _must outlive_ the hash streaming session.
* Similar to one-shot API, `secretSize` must be >= `XXH3_SECRET_SIZE_MIN`,
* and the quality of the hash depends on secret's entropy,
* meaning that the secret should look like a bunch of random bytes.
* When in doubt about the randomness of a candidate `secret`,
* consider employing `XXH3_generateSecret()` instead (see below).
*/
XXH_PUBLIC_API XXH_errorcode XXH3_64bits_reset_withSecret(XXH3_state_t* statePtr, const void* secret, size_t secretSize);
XXH_PUBLIC_API XXH_errorcode XXH3_64bits_update (XXH3_state_t* statePtr, const void* input, size_t length);
XXH_PUBLIC_API XXH64_hash_t XXH3_64bits_digest (const XXH3_state_t* statePtr);
/* 128-bit */
#ifdef XXH_NAMESPACE
# define XXH128 XXH_NAME2(XXH_NAMESPACE, XXH128)
# define XXH3_128bits XXH_NAME2(XXH_NAMESPACE, XXH3_128bits)
# define XXH3_128bits_withSeed XXH_NAME2(XXH_NAMESPACE, XXH3_128bits_withSeed)
# define XXH3_128bits_withSecret XXH_NAME2(XXH_NAMESPACE, XXH3_128bits_withSecret)
# define XXH3_128bits_reset XXH_NAME2(XXH_NAMESPACE, XXH3_128bits_reset)
# define XXH3_128bits_reset_withSeed XXH_NAME2(XXH_NAMESPACE, XXH3_128bits_reset_withSeed)
# define XXH3_128bits_reset_withSecret XXH_NAME2(XXH_NAMESPACE, XXH3_128bits_reset_withSecret)
# define XXH3_128bits_update XXH_NAME2(XXH_NAMESPACE, XXH3_128bits_update)
# define XXH3_128bits_digest XXH_NAME2(XXH_NAMESPACE, XXH3_128bits_digest)
# define XXH128_isEqual XXH_NAME2(XXH_NAMESPACE, XXH128_isEqual)
# define XXH128_cmp XXH_NAME2(XXH_NAMESPACE, XXH128_cmp)
# define XXH128_canonicalFromHash XXH_NAME2(XXH_NAMESPACE, XXH128_canonicalFromHash)
# define XXH128_hashFromCanonical XXH_NAME2(XXH_NAMESPACE, XXH128_hashFromCanonical)
#endif
typedef struct {
XXH64_hash_t low64;
XXH64_hash_t high64;
} XXH128_hash_t;
XXH_PUBLIC_API XXH128_hash_t XXH128(const void* data, size_t len, XXH64_hash_t seed);
XXH_PUBLIC_API XXH128_hash_t XXH3_128bits(const void* data, size_t len);
XXH_PUBLIC_API XXH128_hash_t XXH3_128bits_withSeed(const void* data, size_t len, XXH64_hash_t seed); /* == XXH128() */
XXH_PUBLIC_API XXH128_hash_t XXH3_128bits_withSecret(const void* data, size_t len, const void* secret, size_t secretSize);
XXH_PUBLIC_API XXH_errorcode XXH3_128bits_reset(XXH3_state_t* statePtr);
XXH_PUBLIC_API XXH_errorcode XXH3_128bits_reset_withSeed(XXH3_state_t* statePtr, XXH64_hash_t seed);
XXH_PUBLIC_API XXH_errorcode XXH3_128bits_reset_withSecret(XXH3_state_t* statePtr, const void* secret, size_t secretSize);
XXH_PUBLIC_API XXH_errorcode XXH3_128bits_update (XXH3_state_t* statePtr, const void* input, size_t length);
XXH_PUBLIC_API XXH128_hash_t XXH3_128bits_digest (const XXH3_state_t* statePtr);
/* Note: For better performance, these functions can be inlined using XXH_INLINE_ALL */
/*!
* XXH128_isEqual():
* Return: 1 if `h1` and `h2` are equal, 0 if they are not.
*/
XXH_PUBLIC_API int XXH128_isEqual(XXH128_hash_t h1, XXH128_hash_t h2);
/*!
* XXH128_cmp():
*
* This comparator is compatible with stdlib's `qsort()`/`bsearch()`.
*
* return: >0 if *h128_1 > *h128_2
* =0 if *h128_1 == *h128_2
* <0 if *h128_1 < *h128_2
*/
XXH_PUBLIC_API int XXH128_cmp(const void* h128_1, const void* h128_2);
/******* Canonical representation *******/
typedef struct { unsigned char digest[sizeof(XXH128_hash_t)]; } XXH128_canonical_t;
XXH_PUBLIC_API void XXH128_canonicalFromHash(XXH128_canonical_t* dst, XXH128_hash_t hash);
XXH_PUBLIC_API XXH128_hash_t XXH128_hashFromCanonical(const XXH128_canonical_t* src);
/* === Experimental API === */
/* Symbols defined below must be considered tied to a specific library version. */
/*
* XXH3_generateSecret():
*
* Derive a high-entropy secret from any user-defined content, named customSeed.
* The generated secret can be used in combination with `*_withSecret()` functions.
* The `_withSecret()` variants are useful to provide a higher level of protection than 64-bit seed,
* as it becomes much more difficult for an external actor to guess how to impact the calculation logic.
*
* The function accepts as input a custom seed of any length and any content,
* and derives from it a high-entropy secret of length XXH3_SECRET_DEFAULT_SIZE
* into an already allocated buffer secretBuffer.
* The generated secret is _always_ XXH_SECRET_DEFAULT_SIZE bytes long.
*
* The generated secret can then be used with any `*_withSecret()` variant.
* Functions `XXH3_128bits_withSecret()`, `XXH3_64bits_withSecret()`,
* `XXH3_128bits_reset_withSecret()` and `XXH3_64bits_reset_withSecret()`
* are part of this list. They all accept a `secret` parameter
* which must be very long for implementation reasons (>= XXH3_SECRET_SIZE_MIN)
* _and_ feature very high entropy (consist of random-looking bytes).
* These conditions can be a high bar to meet, so
* this function can be used to generate a secret of proper quality.
*
* customSeed can be anything. It can have any size, even small ones,
* and its content can be anything, even stupidly "low entropy" source such as a bunch of zeroes.
* The resulting `secret` will nonetheless provide all expected qualities.
*
* Supplying NULL as the customSeed copies the default secret into `secretBuffer`.
* When customSeedSize > 0, supplying NULL as customSeed is undefined behavior.
*/
XXH_PUBLIC_API void XXH3_generateSecret(void* secretBuffer, const void* customSeed, size_t customSeedSize);
#endif /* XXH_NO_LONG_LONG */
#if defined(XXH_INLINE_ALL) || defined(XXH_PRIVATE_API)
# define XXH_IMPLEMENTATION
#endif
#endif /* defined(XXH_STATIC_LINKING_ONLY) && !defined(XXHASH_H_STATIC_13879238742) */
/* ======================================================================== */
/* ======================================================================== */
/* ======================================================================== */
/*-**********************************************************************
* xxHash implementation
*-**********************************************************************
* xxHash's implementation used to be found in xxhash.c.
*
* However, code inlining requires the implementation to be visible to the
* compiler, usually within the header.
*
* As a workaround, xxhash.c used to be included within xxhash.h. This caused
* some issues with some build systems, especially ones which treat .c files
* as source files.
*
* Therefore, the implementation is now directly integrated within xxhash.h.
* Another small advantage is that xxhash.c is no longer needed in /include.
************************************************************************/
#if ( defined(XXH_INLINE_ALL) || defined(XXH_PRIVATE_API) \
|| defined(XXH_IMPLEMENTATION) ) && !defined(XXH_IMPLEM_13a8737387)
# define XXH_IMPLEM_13a8737387
/* *************************************
* Tuning parameters
***************************************/
/*!
* XXH_FORCE_MEMORY_ACCESS:
* By default, access to unaligned memory is controlled by `memcpy()`, which is
* safe and portable.
*
* Unfortunately, on some target/compiler combinations, the generated assembly
* is sub-optimal.
*
* The below switch allow to select a different access method for improved
* performance.
* Method 0 (default):
* Use `memcpy()`. Safe and portable.
* Method 1:
* `__attribute__((packed))` statement. It depends on compiler extensions
* and is therefore not portable.
* This method is safe if your compiler supports it, and *generally* as
* fast or faster than `memcpy`.
* Method 2:
* Direct access via cast. This method doesn't depend on the compiler but
* violates the C standard.
* It can generate buggy code on targets which do not support unaligned
* memory accesses.
* But in some circumstances, it's the only known way to get the most
* performance (ie GCC + ARMv6)
* Method 3:
* Byteshift. This can generate the best code on old compilers which don't
* inline small `memcpy()` calls, and it might also be faster on big-endian
* systems which lack a native byteswap instruction.
* See https://stackoverflow.com/a/32095106/646947 for details.
* Prefer these methods in priority order (0 > 1 > 2 > 3)
*/
#ifndef XXH_FORCE_MEMORY_ACCESS /* can be defined externally, on command line for example */
# if !defined(__clang__) && defined(__GNUC__) && defined(__ARM_FEATURE_UNALIGNED) && defined(__ARM_ARCH) && (__ARM_ARCH == 6)
# define XXH_FORCE_MEMORY_ACCESS 2
# elif !defined(__clang__) && ((defined(__INTEL_COMPILER) && !defined(_WIN32)) || \
(defined(__GNUC__) && (defined(__ARM_ARCH) && __ARM_ARCH >= 7)))
# define XXH_FORCE_MEMORY_ACCESS 1
# endif
#endif
/*!
* XXH_ACCEPT_NULL_INPUT_POINTER:
* If the input pointer is NULL, xxHash's default behavior is to dereference it,
* triggering a segfault.
* When this macro is enabled, xxHash actively checks the input for a null pointer.
* If it is, the result for null input pointers is the same as a zero-length input.
*/
#ifndef XXH_ACCEPT_NULL_INPUT_POINTER /* can be defined externally */
# define XXH_ACCEPT_NULL_INPUT_POINTER 0
#endif
/*!
* XXH_FORCE_ALIGN_CHECK:
* This is an important performance trick
* for architectures without decent unaligned memory access performance.
* It checks for input alignment, and when conditions are met,
* uses a "fast path" employing direct 32-bit/64-bit read,
* resulting in _dramatically faster_ read speed.
*
* The check costs one initial branch per hash, which is generally negligible, but not zero.
* Moreover, it's not useful to generate binary for an additional code path
* if memory access uses same instruction for both aligned and unaligned adresses.
*
* In these cases, the alignment check can be removed by setting this macro to 0.
* Then the code will always use unaligned memory access.
* Align check is automatically disabled on x86, x64 & arm64,
* which are platforms known to offer good unaligned memory accesses performance.
*
* This option does not affect XXH3 (only XXH32 and XXH64).
*/
#ifndef XXH_FORCE_ALIGN_CHECK /* can be defined externally */
# if defined(__i386) || defined(__x86_64__) || defined(__aarch64__) \
|| defined(_M_IX86) || defined(_M_X64) || defined(_M_ARM64) /* visual */
# define XXH_FORCE_ALIGN_CHECK 0
# else
# define XXH_FORCE_ALIGN_CHECK 1
# endif
#endif
/*!
* XXH_NO_INLINE_HINTS:
*
* By default, xxHash tries to force the compiler to inline almost all internal
* functions.
*
* This can usually improve performance due to reduced jumping and improved
* constant folding, but significantly increases the size of the binary which
* might not be favorable.
*
* Additionally, sometimes the forced inlining can be detrimental to performance,
* depending on the architecture.
*
* XXH_NO_INLINE_HINTS marks all internal functions as static, giving the
* compiler full control on whether to inline or not.
*
* When not optimizing (-O0), optimizing for size (-Os, -Oz), or using
* -fno-inline with GCC or Clang, this will automatically be defined.
*/
#ifndef XXH_NO_INLINE_HINTS
# if defined(__OPTIMIZE_SIZE__) /* -Os, -Oz */ \
|| defined(__NO_INLINE__) /* -O0, -fno-inline */
# define XXH_NO_INLINE_HINTS 1
# else
# define XXH_NO_INLINE_HINTS 0
# endif
#endif
/*!
* XXH_REROLL:
* Whether to reroll XXH32_finalize, and XXH64_finalize,
* instead of using an unrolled jump table/if statement loop.
*
* This is automatically defined on -Os/-Oz on GCC and Clang.
*/
#ifndef XXH_REROLL
# if defined(__OPTIMIZE_SIZE__)
# define XXH_REROLL 1
# else
# define XXH_REROLL 0
# endif
#endif
/* *************************************
* Includes & Memory related functions
***************************************/
/*!
* Modify the local functions below should you wish to use some other memory
* routines for malloc() and free()
*/
#include <stdlib.h>
static void* XXH_malloc(size_t s) { return malloc(s); }
static void XXH_free(void* p) { free(p); }
/*! and for memcpy() */
#include <string.h>
static void* XXH_memcpy(void* dest, const void* src, size_t size)
{
return memcpy(dest,src,size);
}
#include <limits.h> /* ULLONG_MAX */
/* *************************************
* Compiler Specific Options
***************************************/
#ifdef _MSC_VER /* Visual Studio warning fix */
# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */
#endif
#if XXH_NO_INLINE_HINTS /* disable inlining hints */
# if defined(__GNUC__)
# define XXH_FORCE_INLINE static __attribute__((unused))
# else
# define XXH_FORCE_INLINE static
# endif
# define XXH_NO_INLINE static
/* enable inlining hints */
#elif defined(_MSC_VER) /* Visual Studio */
# define XXH_FORCE_INLINE static __forceinline
# define XXH_NO_INLINE static __declspec(noinline)
#elif defined(__GNUC__)
# define XXH_FORCE_INLINE static __inline__ __attribute__((always_inline, unused))
# define XXH_NO_INLINE static __attribute__((noinline))
#elif defined (__cplusplus) \
|| (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L)) /* C99 */
# define XXH_FORCE_INLINE static inline
# define XXH_NO_INLINE static
#else
# define XXH_FORCE_INLINE static
# define XXH_NO_INLINE static
#endif
/* *************************************
* Debug
***************************************/
/*
* XXH_DEBUGLEVEL is expected to be defined externally, typically via the
* compiler's command line options. The value must be a number.
*/
#ifndef XXH_DEBUGLEVEL
# ifdef DEBUGLEVEL /* backwards compat */
# define XXH_DEBUGLEVEL DEBUGLEVEL
# else
# define XXH_DEBUGLEVEL 0
# endif
#endif
#if (XXH_DEBUGLEVEL>=1)
# include <assert.h> /* note: can still be disabled with NDEBUG */
# define XXH_ASSERT(c) assert(c)
#else
# define XXH_ASSERT(c) ((void)0)
#endif
/* note: use after variable declarations */
#define XXH_STATIC_ASSERT(c) do { enum { XXH_sa = 1/(int)(!!(c)) }; } while (0)
/* *************************************
* Basic Types
***************************************/
#if !defined (__VMS) \
&& (defined (__cplusplus) \
|| (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) )
# include <stdint.h>
typedef uint8_t xxh_u8;
#else
typedef unsigned char xxh_u8;
#endif
typedef XXH32_hash_t xxh_u32;
#ifdef XXH_OLD_NAMES
# define BYTE xxh_u8
# define U8 xxh_u8
# define U32 xxh_u32
#endif
/* *** Memory access *** */
#if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==3))
/*
* Manual byteshift. Best for old compilers which don't inline memcpy.
* We actually directly use XXH_readLE32 and XXH_readBE32.
*/
#elif (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==2))
/*
* Force direct memory access. Only works on CPU which support unaligned memory
* access in hardware.
*/
static xxh_u32 XXH_read32(const void* memPtr) { return *(const xxh_u32*) memPtr; }
#elif (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==1))
/*
* __pack instructions are safer but compiler specific, hence potentially
* problematic for some compilers.
*
* Currently only defined for GCC and ICC.
*/
#ifdef XXH_OLD_NAMES
typedef union { xxh_u32 u32; } __attribute__((packed)) unalign;
#endif
static xxh_u32 XXH_read32(const void* ptr)
{
typedef union { xxh_u32 u32; } __attribute__((packed)) xxh_unalign;
return ((const xxh_unalign*)ptr)->u32;
}
#else
/*
* Portable and safe solution. Generally efficient.
* see: https://stackoverflow.com/a/32095106/646947
*/
static xxh_u32 XXH_read32(const void* memPtr)
{
xxh_u32 val;
memcpy(&val, memPtr, sizeof(val));
return val;
}
#endif /* XXH_FORCE_DIRECT_MEMORY_ACCESS */
/* *** Endianess *** */
typedef enum { XXH_bigEndian=0, XXH_littleEndian=1 } XXH_endianess;
/*!
* XXH_CPU_LITTLE_ENDIAN:
* Defined to 1 if the target is little endian, or 0 if it is big endian.
* It can be defined externally, for example on the compiler command line.
*
* If it is not defined, a runtime check (which is usually constant folded)
* is used instead.
*/
#ifndef XXH_CPU_LITTLE_ENDIAN
/*
* Try to detect endianness automatically, to avoid the nonstandard behavior
* in `XXH_isLittleEndian()`
*/
# if defined(_WIN32) /* Windows is always little endian */ \
|| defined(__LITTLE_ENDIAN__) \
|| (defined(__BYTE_ORDER__) && __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__)
# define XXH_CPU_LITTLE_ENDIAN 1
# elif defined(__BIG_ENDIAN__) \
|| (defined(__BYTE_ORDER__) && __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__)
# define XXH_CPU_LITTLE_ENDIAN 0
# else
/*
* runtime test, presumed to simplify to a constant by compiler
*/
static int XXH_isLittleEndian(void)
{
/*
* Portable and well-defined behavior.
* Don't use static: it is detrimental to performance.
*/
const union { xxh_u32 u; xxh_u8 c[4]; } one = { 1 };
return one.c[0];
}
# define XXH_CPU_LITTLE_ENDIAN XXH_isLittleEndian()
# endif
#endif
/* ****************************************
* Compiler-specific Functions and Macros
******************************************/
#define XXH_GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__)
#ifdef __has_builtin
# define XXH_HAS_BUILTIN(x) __has_builtin(x)
#else
# define XXH_HAS_BUILTIN(x) 0
#endif
#if !defined(NO_CLANG_BUILTIN) && XXH_HAS_BUILTIN(__builtin_rotateleft32) \
&& XXH_HAS_BUILTIN(__builtin_rotateleft64)
# define XXH_rotl32 __builtin_rotateleft32
# define XXH_rotl64 __builtin_rotateleft64
/* Note: although _rotl exists for minGW (GCC under windows), performance seems poor */
#elif defined(_MSC_VER)
# define XXH_rotl32(x,r) _rotl(x,r)
# define XXH_rotl64(x,r) _rotl64(x,r)
#else
# define XXH_rotl32(x,r) (((x) << (r)) | ((x) >> (32 - (r))))
# define XXH_rotl64(x,r) (((x) << (r)) | ((x) >> (64 - (r))))
#endif
#if defined(_MSC_VER) /* Visual Studio */
# define XXH_swap32 _byteswap_ulong
#elif XXH_GCC_VERSION >= 403
# define XXH_swap32 __builtin_bswap32
#else
static xxh_u32 XXH_swap32 (xxh_u32 x)
{
return ((x << 24) & 0xff000000 ) |
((x << 8) & 0x00ff0000 ) |
((x >> 8) & 0x0000ff00 ) |
((x >> 24) & 0x000000ff );
}
#endif
/* ***************************
* Memory reads
*****************************/
typedef enum { XXH_aligned, XXH_unaligned } XXH_alignment;
/*
* XXH_FORCE_MEMORY_ACCESS==3 is an endian-independent byteshift load.
*
* This is ideal for older compilers which don't inline memcpy.
*/
#if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==3))
XXH_FORCE_INLINE xxh_u32 XXH_readLE32(const void* memPtr)
{
const xxh_u8* bytePtr = (const xxh_u8 *)memPtr;
return bytePtr[0]
| ((xxh_u32)bytePtr[1] << 8)
| ((xxh_u32)bytePtr[2] << 16)
| ((xxh_u32)bytePtr[3] << 24);
}
XXH_FORCE_INLINE xxh_u32 XXH_readBE32(const void* memPtr)
{
const xxh_u8* bytePtr = (const xxh_u8 *)memPtr;
return bytePtr[3]
| ((xxh_u32)bytePtr[2] << 8)
| ((xxh_u32)bytePtr[1] << 16)
| ((xxh_u32)bytePtr[0] << 24);
}
#else
XXH_FORCE_INLINE xxh_u32 XXH_readLE32(const void* ptr)
{
return XXH_CPU_LITTLE_ENDIAN ? XXH_read32(ptr) : XXH_swap32(XXH_read32(ptr));
}
static xxh_u32 XXH_readBE32(const void* ptr)
{
return XXH_CPU_LITTLE_ENDIAN ? XXH_swap32(XXH_read32(ptr)) : XXH_read32(ptr);
}
#endif
XXH_FORCE_INLINE xxh_u32
XXH_readLE32_align(const void* ptr, XXH_alignment align)
{
if (align==XXH_unaligned) {
return XXH_readLE32(ptr);
} else {
return XXH_CPU_LITTLE_ENDIAN ? *(const xxh_u32*)ptr : XXH_swap32(*(const xxh_u32*)ptr);
}
}
/* *************************************
* Misc
***************************************/
XXH_PUBLIC_API unsigned XXH_versionNumber (void) { return XXH_VERSION_NUMBER; }
/* *******************************************************************
* 32-bit hash functions
*********************************************************************/
static const xxh_u32 XXH_PRIME32_1 = 0x9E3779B1U; /* 0b10011110001101110111100110110001 */
static const xxh_u32 XXH_PRIME32_2 = 0x85EBCA77U; /* 0b10000101111010111100101001110111 */
static const xxh_u32 XXH_PRIME32_3 = 0xC2B2AE3DU; /* 0b11000010101100101010111000111101 */
static const xxh_u32 XXH_PRIME32_4 = 0x27D4EB2FU; /* 0b00100111110101001110101100101111 */
static const xxh_u32 XXH_PRIME32_5 = 0x165667B1U; /* 0b00010110010101100110011110110001 */
#ifdef XXH_OLD_NAMES
# define PRIME32_1 XXH_PRIME32_1
# define PRIME32_2 XXH_PRIME32_2
# define PRIME32_3 XXH_PRIME32_3
# define PRIME32_4 XXH_PRIME32_4
# define PRIME32_5 XXH_PRIME32_5
#endif
static xxh_u32 XXH32_round(xxh_u32 acc, xxh_u32 input)
{
acc += input * XXH_PRIME32_2;
acc = XXH_rotl32(acc, 13);
acc *= XXH_PRIME32_1;
#if defined(__GNUC__) && defined(__SSE4_1__) && !defined(XXH_ENABLE_AUTOVECTORIZE)
/*
* UGLY HACK:
* This inline assembly hack forces acc into a normal register. This is the
* only thing that prevents GCC and Clang from autovectorizing the XXH32
* loop (pragmas and attributes don't work for some resason) without globally
* disabling SSE4.1.
*
* The reason we want to avoid vectorization is because despite working on
* 4 integers at a time, there are multiple factors slowing XXH32 down on
* SSE4:
* - There's a ridiculous amount of lag from pmulld (10 cycles of latency on
* newer chips!) making it slightly slower to multiply four integers at
* once compared to four integers independently. Even when pmulld was
* fastest, Sandy/Ivy Bridge, it is still not worth it to go into SSE
* just to multiply unless doing a long operation.
*
* - Four instructions are required to rotate,
* movqda tmp, v // not required with VEX encoding
* pslld tmp, 13 // tmp <<= 13
* psrld v, 19 // x >>= 19
* por v, tmp // x |= tmp
* compared to one for scalar:
* roll v, 13 // reliably fast across the board
* shldl v, v, 13 // Sandy Bridge and later prefer this for some reason
*
* - Instruction level parallelism is actually more beneficial here because
* the SIMD actually serializes this operation: While v1 is rotating, v2
* can load data, while v3 can multiply. SSE forces them to operate
* together.
*
* How this hack works:
* __asm__("" // Declare an assembly block but don't declare any instructions
* : // However, as an Input/Output Operand,
* "+r" // constrain a read/write operand (+) as a general purpose register (r).
* (acc) // and set acc as the operand
* );
*
* Because of the 'r', the compiler has promised that seed will be in a
* general purpose register and the '+' says that it will be 'read/write',
* so it has to assume it has changed. It is like volatile without all the
* loads and stores.
*
* Since the argument has to be in a normal register (not an SSE register),
* each time XXH32_round is called, it is impossible to vectorize.
*/
__asm__("" : "+r" (acc));
#endif
return acc;
}
/* mix all bits */
static xxh_u32 XXH32_avalanche(xxh_u32 h32)
{
h32 ^= h32 >> 15;
h32 *= XXH_PRIME32_2;
h32 ^= h32 >> 13;
h32 *= XXH_PRIME32_3;
h32 ^= h32 >> 16;
return(h32);
}
#define XXH_get32bits(p) XXH_readLE32_align(p, align)
static xxh_u32
XXH32_finalize(xxh_u32 h32, const xxh_u8* ptr, size_t len, XXH_alignment align)
{
#define XXH_PROCESS1 do { \
h32 += (*ptr++) * XXH_PRIME32_5; \
h32 = XXH_rotl32(h32, 11) * XXH_PRIME32_1; \
} while (0)
#define XXH_PROCESS4 do { \
h32 += XXH_get32bits(ptr) * XXH_PRIME32_3; \
ptr += 4; \
h32 = XXH_rotl32(h32, 17) * XXH_PRIME32_4; \
} while (0)
/* Compact rerolled version */
if (XXH_REROLL) {
len &= 15;
while (len >= 4) {
XXH_PROCESS4;
len -= 4;
}
while (len > 0) {
XXH_PROCESS1;
--len;
}
return XXH32_avalanche(h32);
} else {
switch(len&15) /* or switch(bEnd - p) */ {
case 12: XXH_PROCESS4;
/* fallthrough */
case 8: XXH_PROCESS4;
/* fallthrough */
case 4: XXH_PROCESS4;
return XXH32_avalanche(h32);
case 13: XXH_PROCESS4;
/* fallthrough */
case 9: XXH_PROCESS4;
/* fallthrough */
case 5: XXH_PROCESS4;
XXH_PROCESS1;
return XXH32_avalanche(h32);
case 14: XXH_PROCESS4;
/* fallthrough */
case 10: XXH_PROCESS4;
/* fallthrough */
case 6: XXH_PROCESS4;
XXH_PROCESS1;
XXH_PROCESS1;
return XXH32_avalanche(h32);
case 15: XXH_PROCESS4;
/* fallthrough */
case 11: XXH_PROCESS4;
/* fallthrough */
case 7: XXH_PROCESS4;
/* fallthrough */
case 3: XXH_PROCESS1;
/* fallthrough */
case 2: XXH_PROCESS1;
/* fallthrough */
case 1: XXH_PROCESS1;
/* fallthrough */
case 0: return XXH32_avalanche(h32);
}
XXH_ASSERT(0);
return h32; /* reaching this point is deemed impossible */
}
}
#ifdef XXH_OLD_NAMES
# define PROCESS1 XXH_PROCESS1
# define PROCESS4 XXH_PROCESS4
#else
# undef XXH_PROCESS1
# undef XXH_PROCESS4
#endif
XXH_FORCE_INLINE xxh_u32
XXH32_endian_align(const xxh_u8* input, size_t len, xxh_u32 seed, XXH_alignment align)
{
const xxh_u8* bEnd = input + len;
xxh_u32 h32;
#if defined(XXH_ACCEPT_NULL_INPUT_POINTER) && (XXH_ACCEPT_NULL_INPUT_POINTER>=1)
if (input==NULL) {
len=0;
bEnd=input=(const xxh_u8*)(size_t)16;
}
#endif
if (len>=16) {
const xxh_u8* const limit = bEnd - 15;
xxh_u32 v1 = seed + XXH_PRIME32_1 + XXH_PRIME32_2;
xxh_u32 v2 = seed + XXH_PRIME32_2;
xxh_u32 v3 = seed + 0;
xxh_u32 v4 = seed - XXH_PRIME32_1;
do {
v1 = XXH32_round(v1, XXH_get32bits(input)); input += 4;
v2 = XXH32_round(v2, XXH_get32bits(input)); input += 4;
v3 = XXH32_round(v3, XXH_get32bits(input)); input += 4;
v4 = XXH32_round(v4, XXH_get32bits(input)); input += 4;
} while (input < limit);
h32 = XXH_rotl32(v1, 1) + XXH_rotl32(v2, 7)
+ XXH_rotl32(v3, 12) + XXH_rotl32(v4, 18);
} else {
h32 = seed + XXH_PRIME32_5;
}
h32 += (xxh_u32)len;
return XXH32_finalize(h32, input, len&15, align);
}
XXH_PUBLIC_API XXH32_hash_t XXH32 (const void* input, size_t len, XXH32_hash_t seed)
{
#if 0
/* Simple version, good for code maintenance, but unfortunately slow for small inputs */
XXH32_state_t state;
XXH32_reset(&state, seed);
XXH32_update(&state, (const xxh_u8*)input, len);
return XXH32_digest(&state);
#else
if (XXH_FORCE_ALIGN_CHECK) {
if ((((size_t)input) & 3) == 0) { /* Input is 4-bytes aligned, leverage the speed benefit */
return XXH32_endian_align((const xxh_u8*)input, len, seed, XXH_aligned);
} }
return XXH32_endian_align((const xxh_u8*)input, len, seed, XXH_unaligned);
#endif
}
/******* Hash streaming *******/
XXH_PUBLIC_API XXH32_state_t* XXH32_createState(void)
{
return (XXH32_state_t*)XXH_malloc(sizeof(XXH32_state_t));
}
XXH_PUBLIC_API XXH_errorcode XXH32_freeState(XXH32_state_t* statePtr)
{
XXH_free(statePtr);
return XXH_OK;
}
XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* dstState, const XXH32_state_t* srcState)
{
memcpy(dstState, srcState, sizeof(*dstState));
}
XXH_PUBLIC_API XXH_errorcode XXH32_reset(XXH32_state_t* statePtr, XXH32_hash_t seed)
{
XXH32_state_t state; /* using a local state to memcpy() in order to avoid strict-aliasing warnings */
memset(&state, 0, sizeof(state));
state.v1 = seed + XXH_PRIME32_1 + XXH_PRIME32_2;
state.v2 = seed + XXH_PRIME32_2;
state.v3 = seed + 0;
state.v4 = seed - XXH_PRIME32_1;
/* do not write into reserved, planned to be removed in a future version */
memcpy(statePtr, &state, sizeof(state) - sizeof(state.reserved));
return XXH_OK;
}
XXH_PUBLIC_API XXH_errorcode
XXH32_update(XXH32_state_t* state, const void* input, size_t len)
{
if (input==NULL)
#if defined(XXH_ACCEPT_NULL_INPUT_POINTER) && (XXH_ACCEPT_NULL_INPUT_POINTER>=1)
return XXH_OK;
#else
return XXH_ERROR;
#endif
{ const xxh_u8* p = (const xxh_u8*)input;
const xxh_u8* const bEnd = p + len;
state->total_len_32 += (XXH32_hash_t)len;
state->large_len |= (XXH32_hash_t)((len>=16) | (state->total_len_32>=16));
if (state->memsize + len < 16) { /* fill in tmp buffer */
XXH_memcpy((xxh_u8*)(state->mem32) + state->memsize, input, len);
state->memsize += (XXH32_hash_t)len;
return XXH_OK;
}
if (state->memsize) { /* some data left from previous update */
XXH_memcpy((xxh_u8*)(state->mem32) + state->memsize, input, 16-state->memsize);
{ const xxh_u32* p32 = state->mem32;
state->v1 = XXH32_round(state->v1, XXH_readLE32(p32)); p32++;
state->v2 = XXH32_round(state->v2, XXH_readLE32(p32)); p32++;
state->v3 = XXH32_round(state->v3, XXH_readLE32(p32)); p32++;
state->v4 = XXH32_round(state->v4, XXH_readLE32(p32));
}
p += 16-state->memsize;
state->memsize = 0;
}
if (p <= bEnd-16) {
const xxh_u8* const limit = bEnd - 16;
xxh_u32 v1 = state->v1;
xxh_u32 v2 = state->v2;
xxh_u32 v3 = state->v3;
xxh_u32 v4 = state->v4;
do {
v1 = XXH32_round(v1, XXH_readLE32(p)); p+=4;
v2 = XXH32_round(v2, XXH_readLE32(p)); p+=4;
v3 = XXH32_round(v3, XXH_readLE32(p)); p+=4;
v4 = XXH32_round(v4, XXH_readLE32(p)); p+=4;
} while (p<=limit);
state->v1 = v1;
state->v2 = v2;
state->v3 = v3;
state->v4 = v4;
}
if (p < bEnd) {
XXH_memcpy(state->mem32, p, (size_t)(bEnd-p));
state->memsize = (unsigned)(bEnd-p);
}
}
return XXH_OK;
}
XXH_PUBLIC_API XXH32_hash_t XXH32_digest (const XXH32_state_t* state)
{
xxh_u32 h32;
if (state->large_len) {
h32 = XXH_rotl32(state->v1, 1)
+ XXH_rotl32(state->v2, 7)
+ XXH_rotl32(state->v3, 12)
+ XXH_rotl32(state->v4, 18);
} else {
h32 = state->v3 /* == seed */ + XXH_PRIME32_5;
}
h32 += state->total_len_32;
return XXH32_finalize(h32, (const xxh_u8*)state->mem32, state->memsize, XXH_aligned);
}
/******* Canonical representation *******/
/*
* The default return values from XXH functions are unsigned 32 and 64 bit
* integers.
*
* The canonical representation uses big endian convention, the same convention
* as human-readable numbers (large digits first).
*
* This way, hash values can be written into a file or buffer, remaining
* comparable across different systems.
*
* The following functions allow transformation of hash values to and from their
* canonical format.
*/
XXH_PUBLIC_API void XXH32_canonicalFromHash(XXH32_canonical_t* dst, XXH32_hash_t hash)
{
XXH_STATIC_ASSERT(sizeof(XXH32_canonical_t) == sizeof(XXH32_hash_t));
if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap32(hash);
memcpy(dst, &hash, sizeof(*dst));
}
XXH_PUBLIC_API XXH32_hash_t XXH32_hashFromCanonical(const XXH32_canonical_t* src)
{
return XXH_readBE32(src);
}
#ifndef XXH_NO_LONG_LONG
/* *******************************************************************
* 64-bit hash functions
*********************************************************************/
/******* Memory access *******/
typedef XXH64_hash_t xxh_u64;
#ifdef XXH_OLD_NAMES
# define U64 xxh_u64
#endif
/*!
* XXH_REROLL_XXH64:
* Whether to reroll the XXH64_finalize() loop.
*
* Just like XXH32, we can unroll the XXH64_finalize() loop. This can be a
* performance gain on 64-bit hosts, as only one jump is required.
*
* However, on 32-bit hosts, because arithmetic needs to be done with two 32-bit
* registers, and 64-bit arithmetic needs to be simulated, it isn't beneficial
* to unroll. The code becomes ridiculously large (the largest function in the
* binary on i386!), and rerolling it saves anywhere from 3kB to 20kB. It is
* also slightly faster because it fits into cache better and is more likely
* to be inlined by the compiler.
*
* If XXH_REROLL is defined, this is ignored and the loop is always rerolled.
*/
#ifndef XXH_REROLL_XXH64
# if (defined(__ILP32__) || defined(_ILP32)) /* ILP32 is often defined on 32-bit GCC family */ \
|| !(defined(__x86_64__) || defined(_M_X64) || defined(_M_AMD64) /* x86-64 */ \
|| defined(_M_ARM64) || defined(__aarch64__) || defined(__arm64__) /* aarch64 */ \
|| defined(__PPC64__) || defined(__PPC64LE__) || defined(__ppc64__) || defined(__powerpc64__) /* ppc64 */ \
|| defined(__mips64__) || defined(__mips64)) /* mips64 */ \
|| (!defined(SIZE_MAX) || SIZE_MAX < ULLONG_MAX) /* check limits */
# define XXH_REROLL_XXH64 1
# else
# define XXH_REROLL_XXH64 0
# endif
#endif /* !defined(XXH_REROLL_XXH64) */
#if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==3))
/*
* Manual byteshift. Best for old compilers which don't inline memcpy.
* We actually directly use XXH_readLE64 and XXH_readBE64.
*/
#elif (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==2))
/* Force direct memory access. Only works on CPU which support unaligned memory access in hardware */
static xxh_u64 XXH_read64(const void* memPtr) { return *(const xxh_u64*) memPtr; }
#elif (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==1))
/*
* __pack instructions are safer, but compiler specific, hence potentially
* problematic for some compilers.
*
* Currently only defined for GCC and ICC.
*/
#ifdef XXH_OLD_NAMES
typedef union { xxh_u32 u32; xxh_u64 u64; } __attribute__((packed)) unalign64;
#endif
static xxh_u64 XXH_read64(const void* ptr)
{
typedef union { xxh_u32 u32; xxh_u64 u64; } __attribute__((packed)) xxh_unalign64;
return ((const xxh_unalign64*)ptr)->u64;
}
#else
/*
* Portable and safe solution. Generally efficient.
* see: https://stackoverflow.com/a/32095106/646947
*/
static xxh_u64 XXH_read64(const void* memPtr)
{
xxh_u64 val;
memcpy(&val, memPtr, sizeof(val));
return val;
}
#endif /* XXH_FORCE_DIRECT_MEMORY_ACCESS */
#if defined(_MSC_VER) /* Visual Studio */
# define XXH_swap64 _byteswap_uint64
#elif XXH_GCC_VERSION >= 403
# define XXH_swap64 __builtin_bswap64
#else
static xxh_u64 XXH_swap64 (xxh_u64 x)
{
return ((x << 56) & 0xff00000000000000ULL) |
((x << 40) & 0x00ff000000000000ULL) |
((x << 24) & 0x0000ff0000000000ULL) |
((x << 8) & 0x000000ff00000000ULL) |
((x >> 8) & 0x00000000ff000000ULL) |
((x >> 24) & 0x0000000000ff0000ULL) |
((x >> 40) & 0x000000000000ff00ULL) |
((x >> 56) & 0x00000000000000ffULL);
}
#endif
/* XXH_FORCE_MEMORY_ACCESS==3 is an endian-independent byteshift load. */
#if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==3))
XXH_FORCE_INLINE xxh_u64 XXH_readLE64(const void* memPtr)
{
const xxh_u8* bytePtr = (const xxh_u8 *)memPtr;
return bytePtr[0]
| ((xxh_u64)bytePtr[1] << 8)
| ((xxh_u64)bytePtr[2] << 16)
| ((xxh_u64)bytePtr[3] << 24)
| ((xxh_u64)bytePtr[4] << 32)
| ((xxh_u64)bytePtr[5] << 40)
| ((xxh_u64)bytePtr[6] << 48)
| ((xxh_u64)bytePtr[7] << 56);
}
XXH_FORCE_INLINE xxh_u64 XXH_readBE64(const void* memPtr)
{
const xxh_u8* bytePtr = (const xxh_u8 *)memPtr;
return bytePtr[7]
| ((xxh_u64)bytePtr[6] << 8)
| ((xxh_u64)bytePtr[5] << 16)
| ((xxh_u64)bytePtr[4] << 24)
| ((xxh_u64)bytePtr[3] << 32)
| ((xxh_u64)bytePtr[2] << 40)
| ((xxh_u64)bytePtr[1] << 48)
| ((xxh_u64)bytePtr[0] << 56);
}
#else
XXH_FORCE_INLINE xxh_u64 XXH_readLE64(const void* ptr)
{
return XXH_CPU_LITTLE_ENDIAN ? XXH_read64(ptr) : XXH_swap64(XXH_read64(ptr));
}
static xxh_u64 XXH_readBE64(const void* ptr)
{
return XXH_CPU_LITTLE_ENDIAN ? XXH_swap64(XXH_read64(ptr)) : XXH_read64(ptr);
}
#endif
XXH_FORCE_INLINE xxh_u64
XXH_readLE64_align(const void* ptr, XXH_alignment align)
{
if (align==XXH_unaligned)
return XXH_readLE64(ptr);
else
return XXH_CPU_LITTLE_ENDIAN ? *(const xxh_u64*)ptr : XXH_swap64(*(const xxh_u64*)ptr);
}
/******* xxh64 *******/
static const xxh_u64 XXH_PRIME64_1 = 0x9E3779B185EBCA87ULL; /* 0b1001111000110111011110011011000110000101111010111100101010000111 */
static const xxh_u64 XXH_PRIME64_2 = 0xC2B2AE3D27D4EB4FULL; /* 0b1100001010110010101011100011110100100111110101001110101101001111 */
static const xxh_u64 XXH_PRIME64_3 = 0x165667B19E3779F9ULL; /* 0b0001011001010110011001111011000110011110001101110111100111111001 */
static const xxh_u64 XXH_PRIME64_4 = 0x85EBCA77C2B2AE63ULL; /* 0b1000010111101011110010100111011111000010101100101010111001100011 */
static const xxh_u64 XXH_PRIME64_5 = 0x27D4EB2F165667C5ULL; /* 0b0010011111010100111010110010111100010110010101100110011111000101 */
#ifdef XXH_OLD_NAMES
# define PRIME64_1 XXH_PRIME64_1
# define PRIME64_2 XXH_PRIME64_2
# define PRIME64_3 XXH_PRIME64_3
# define PRIME64_4 XXH_PRIME64_4
# define PRIME64_5 XXH_PRIME64_5
#endif
static xxh_u64 XXH64_round(xxh_u64 acc, xxh_u64 input)
{
acc += input * XXH_PRIME64_2;
acc = XXH_rotl64(acc, 31);
acc *= XXH_PRIME64_1;
return acc;
}
static xxh_u64 XXH64_mergeRound(xxh_u64 acc, xxh_u64 val)
{
val = XXH64_round(0, val);
acc ^= val;
acc = acc * XXH_PRIME64_1 + XXH_PRIME64_4;
return acc;
}
static xxh_u64 XXH64_avalanche(xxh_u64 h64)
{
h64 ^= h64 >> 33;
h64 *= XXH_PRIME64_2;
h64 ^= h64 >> 29;
h64 *= XXH_PRIME64_3;
h64 ^= h64 >> 32;
return h64;
}
#define XXH_get64bits(p) XXH_readLE64_align(p, align)
static xxh_u64
XXH64_finalize(xxh_u64 h64, const xxh_u8* ptr, size_t len, XXH_alignment align)
{
#define XXH_PROCESS1_64 do { \
h64 ^= (*ptr++) * XXH_PRIME64_5; \
h64 = XXH_rotl64(h64, 11) * XXH_PRIME64_1; \
} while (0)
#define XXH_PROCESS4_64 do { \
h64 ^= (xxh_u64)(XXH_get32bits(ptr)) * XXH_PRIME64_1; \
ptr += 4; \
h64 = XXH_rotl64(h64, 23) * XXH_PRIME64_2 + XXH_PRIME64_3; \
} while (0)
#define XXH_PROCESS8_64 do { \
xxh_u64 const k1 = XXH64_round(0, XXH_get64bits(ptr)); \
ptr += 8; \
h64 ^= k1; \
h64 = XXH_rotl64(h64,27) * XXH_PRIME64_1 + XXH_PRIME64_4; \
} while (0)
/* Rerolled version for 32-bit targets is faster and much smaller. */
if (XXH_REROLL || XXH_REROLL_XXH64) {
len &= 31;
while (len >= 8) {
XXH_PROCESS8_64;
len -= 8;
}
if (len >= 4) {
XXH_PROCESS4_64;
len -= 4;
}
while (len > 0) {
XXH_PROCESS1_64;
--len;
}
return XXH64_avalanche(h64);
} else {
switch(len & 31) {
case 24: XXH_PROCESS8_64;
/* fallthrough */
case 16: XXH_PROCESS8_64;
/* fallthrough */
case 8: XXH_PROCESS8_64;
return XXH64_avalanche(h64);
case 28: XXH_PROCESS8_64;
/* fallthrough */
case 20: XXH_PROCESS8_64;
/* fallthrough */
case 12: XXH_PROCESS8_64;
/* fallthrough */
case 4: XXH_PROCESS4_64;
return XXH64_avalanche(h64);
case 25: XXH_PROCESS8_64;
/* fallthrough */
case 17: XXH_PROCESS8_64;
/* fallthrough */
case 9: XXH_PROCESS8_64;
XXH_PROCESS1_64;
return XXH64_avalanche(h64);
case 29: XXH_PROCESS8_64;
/* fallthrough */
case 21: XXH_PROCESS8_64;
/* fallthrough */
case 13: XXH_PROCESS8_64;
/* fallthrough */
case 5: XXH_PROCESS4_64;
XXH_PROCESS1_64;
return XXH64_avalanche(h64);
case 26: XXH_PROCESS8_64;
/* fallthrough */
case 18: XXH_PROCESS8_64;
/* fallthrough */
case 10: XXH_PROCESS8_64;
XXH_PROCESS1_64;
XXH_PROCESS1_64;
return XXH64_avalanche(h64);
case 30: XXH_PROCESS8_64;
/* fallthrough */
case 22: XXH_PROCESS8_64;
/* fallthrough */
case 14: XXH_PROCESS8_64;
/* fallthrough */
case 6: XXH_PROCESS4_64;
XXH_PROCESS1_64;
XXH_PROCESS1_64;
return XXH64_avalanche(h64);
case 27: XXH_PROCESS8_64;
/* fallthrough */
case 19: XXH_PROCESS8_64;
/* fallthrough */
case 11: XXH_PROCESS8_64;
XXH_PROCESS1_64;
XXH_PROCESS1_64;
XXH_PROCESS1_64;
return XXH64_avalanche(h64);
case 31: XXH_PROCESS8_64;
/* fallthrough */
case 23: XXH_PROCESS8_64;
/* fallthrough */
case 15: XXH_PROCESS8_64;
/* fallthrough */
case 7: XXH_PROCESS4_64;
/* fallthrough */
case 3: XXH_PROCESS1_64;
/* fallthrough */
case 2: XXH_PROCESS1_64;
/* fallthrough */
case 1: XXH_PROCESS1_64;
/* fallthrough */
case 0: return XXH64_avalanche(h64);
}
}
/* impossible to reach */
XXH_ASSERT(0);
return 0; /* unreachable, but some compilers complain without it */
}
#ifdef XXH_OLD_NAMES
# define PROCESS1_64 XXH_PROCESS1_64
# define PROCESS4_64 XXH_PROCESS4_64
# define PROCESS8_64 XXH_PROCESS8_64
#else
# undef XXH_PROCESS1_64
# undef XXH_PROCESS4_64
# undef XXH_PROCESS8_64
#endif
XXH_FORCE_INLINE xxh_u64
XXH64_endian_align(const xxh_u8* input, size_t len, xxh_u64 seed, XXH_alignment align)
{
const xxh_u8* bEnd = input + len;
xxh_u64 h64;
#if defined(XXH_ACCEPT_NULL_INPUT_POINTER) && (XXH_ACCEPT_NULL_INPUT_POINTER>=1)
if (input==NULL) {
len=0;
bEnd=input=(const xxh_u8*)(size_t)32;
}
#endif
if (len>=32) {
const xxh_u8* const limit = bEnd - 32;
xxh_u64 v1 = seed + XXH_PRIME64_1 + XXH_PRIME64_2;
xxh_u64 v2 = seed + XXH_PRIME64_2;
xxh_u64 v3 = seed + 0;
xxh_u64 v4 = seed - XXH_PRIME64_1;
do {
v1 = XXH64_round(v1, XXH_get64bits(input)); input+=8;
v2 = XXH64_round(v2, XXH_get64bits(input)); input+=8;
v3 = XXH64_round(v3, XXH_get64bits(input)); input+=8;
v4 = XXH64_round(v4, XXH_get64bits(input)); input+=8;
} while (input<=limit);
h64 = XXH_rotl64(v1, 1) + XXH_rotl64(v2, 7) + XXH_rotl64(v3, 12) + XXH_rotl64(v4, 18);
h64 = XXH64_mergeRound(h64, v1);
h64 = XXH64_mergeRound(h64, v2);
h64 = XXH64_mergeRound(h64, v3);
h64 = XXH64_mergeRound(h64, v4);
} else {
h64 = seed + XXH_PRIME64_5;
}
h64 += (xxh_u64) len;
return XXH64_finalize(h64, input, len, align);
}
XXH_PUBLIC_API XXH64_hash_t XXH64 (const void* input, size_t len, XXH64_hash_t seed)
{
#if 0
/* Simple version, good for code maintenance, but unfortunately slow for small inputs */
XXH64_state_t state;
XXH64_reset(&state, seed);
XXH64_update(&state, (const xxh_u8*)input, len);
return XXH64_digest(&state);
#else
if (XXH_FORCE_ALIGN_CHECK) {
if ((((size_t)input) & 7)==0) { /* Input is aligned, let's leverage the speed advantage */
return XXH64_endian_align((const xxh_u8*)input, len, seed, XXH_aligned);
} }
return XXH64_endian_align((const xxh_u8*)input, len, seed, XXH_unaligned);
#endif
}
/******* Hash Streaming *******/
XXH_PUBLIC_API XXH64_state_t* XXH64_createState(void)
{
return (XXH64_state_t*)XXH_malloc(sizeof(XXH64_state_t));
}
XXH_PUBLIC_API XXH_errorcode XXH64_freeState(XXH64_state_t* statePtr)
{
XXH_free(statePtr);
return XXH_OK;
}
XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* dstState, const XXH64_state_t* srcState)
{
memcpy(dstState, srcState, sizeof(*dstState));
}
XXH_PUBLIC_API XXH_errorcode XXH64_reset(XXH64_state_t* statePtr, XXH64_hash_t seed)
{
XXH64_state_t state; /* use a local state to memcpy() in order to avoid strict-aliasing warnings */
memset(&state, 0, sizeof(state));
state.v1 = seed + XXH_PRIME64_1 + XXH_PRIME64_2;
state.v2 = seed + XXH_PRIME64_2;
state.v3 = seed + 0;
state.v4 = seed - XXH_PRIME64_1;
/* do not write into reserved64, might be removed in a future version */
memcpy(statePtr, &state, sizeof(state) - sizeof(state.reserved64));
return XXH_OK;
}
XXH_PUBLIC_API XXH_errorcode
XXH64_update (XXH64_state_t* state, const void* input, size_t len)
{
if (input==NULL)
#if defined(XXH_ACCEPT_NULL_INPUT_POINTER) && (XXH_ACCEPT_NULL_INPUT_POINTER>=1)
return XXH_OK;
#else
return XXH_ERROR;
#endif
{ const xxh_u8* p = (const xxh_u8*)input;
const xxh_u8* const bEnd = p + len;
state->total_len += len;
if (state->memsize + len < 32) { /* fill in tmp buffer */
XXH_memcpy(((xxh_u8*)state->mem64) + state->memsize, input, len);
state->memsize += (xxh_u32)len;
return XXH_OK;
}
if (state->memsize) { /* tmp buffer is full */
XXH_memcpy(((xxh_u8*)state->mem64) + state->memsize, input, 32-state->memsize);
state->v1 = XXH64_round(state->v1, XXH_readLE64(state->mem64+0));
state->v2 = XXH64_round(state->v2, XXH_readLE64(state->mem64+1));
state->v3 = XXH64_round(state->v3, XXH_readLE64(state->mem64+2));
state->v4 = XXH64_round(state->v4, XXH_readLE64(state->mem64+3));
p += 32-state->memsize;
state->memsize = 0;
}
if (p+32 <= bEnd) {
const xxh_u8* const limit = bEnd - 32;
xxh_u64 v1 = state->v1;
xxh_u64 v2 = state->v2;
xxh_u64 v3 = state->v3;
xxh_u64 v4 = state->v4;
do {
v1 = XXH64_round(v1, XXH_readLE64(p)); p+=8;
v2 = XXH64_round(v2, XXH_readLE64(p)); p+=8;
v3 = XXH64_round(v3, XXH_readLE64(p)); p+=8;
v4 = XXH64_round(v4, XXH_readLE64(p)); p+=8;
} while (p<=limit);
state->v1 = v1;
state->v2 = v2;
state->v3 = v3;
state->v4 = v4;
}
if (p < bEnd) {
XXH_memcpy(state->mem64, p, (size_t)(bEnd-p));
state->memsize = (unsigned)(bEnd-p);
}
}
return XXH_OK;
}
XXH_PUBLIC_API XXH64_hash_t XXH64_digest (const XXH64_state_t* state)
{
xxh_u64 h64;
if (state->total_len >= 32) {
xxh_u64 const v1 = state->v1;
xxh_u64 const v2 = state->v2;
xxh_u64 const v3 = state->v3;
xxh_u64 const v4 = state->v4;
h64 = XXH_rotl64(v1, 1) + XXH_rotl64(v2, 7) + XXH_rotl64(v3, 12) + XXH_rotl64(v4, 18);
h64 = XXH64_mergeRound(h64, v1);
h64 = XXH64_mergeRound(h64, v2);
h64 = XXH64_mergeRound(h64, v3);
h64 = XXH64_mergeRound(h64, v4);
} else {
h64 = state->v3 /*seed*/ + XXH_PRIME64_5;
}
h64 += (xxh_u64) state->total_len;
return XXH64_finalize(h64, (const xxh_u8*)state->mem64, (size_t)state->total_len, XXH_aligned);
}
/******* Canonical representation *******/
XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t hash)
{
XXH_STATIC_ASSERT(sizeof(XXH64_canonical_t) == sizeof(XXH64_hash_t));
if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap64(hash);
memcpy(dst, &hash, sizeof(*dst));
}
XXH_PUBLIC_API XXH64_hash_t XXH64_hashFromCanonical(const XXH64_canonical_t* src)
{
return XXH_readBE64(src);
}
/* *********************************************************************
* XXH3
* New generation hash designed for speed on small keys and vectorization
************************************************************************ */
#include "xxh3.h"
#endif /* XXH_NO_LONG_LONG */
#endif /* XXH_IMPLEMENTATION */
#if defined (__cplusplus)
}
#endif
|