1 // SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0-only 2 /* 3 * xxHash - Fast Hash algorithm 4 * Copyright (c) 2012-2020, Yann Collet, Facebook, Inc. 5 * 6 * You can contact the author at : 7 * - xxHash homepage: http://www.xxhash.com 8 * - xxHash source repository : https://github.com/Cyan4973/xxHash 9 * 10 * This source code is licensed under both the BSD-style license (found in the 11 * LICENSE file in the root directory of this source tree) and the GPLv2 (found 12 * in the COPYING file in the root directory of this source tree). 13 * You may select, at your option, one of the above-listed licenses. 14 */ 15 16 17 /* ************************************* 18 * Tuning parameters 19 ***************************************/ 20 /*!XXH_FORCE_MEMORY_ACCESS : 21 * By default, access to unaligned memory is controlled by `memcpy()`, which is safe and portable. 22 * Unfortunately, on some target/compiler combinations, the generated assembly is sub-optimal. 23 * The below switch allow to select different access method for improved performance. 24 * Method 0 (default) : use `memcpy()`. Safe and portable. 25 * Method 1 : `__packed` statement. It depends on compiler extension (ie, not portable). 26 * This method is safe if your compiler supports it, and *generally* as fast or faster than `memcpy`. 27 * Method 2 : direct access. This method doesn't depend on compiler but violate C standard. 28 * It can generate buggy code on targets which do not support unaligned memory accesses. 29 * But in some circumstances, it's the only known way to get the most performance (ie GCC + ARMv6) 30 * See http://stackoverflow.com/a/32095106/646947 for details. 31 * Prefer these methods in priority order (0 > 1 > 2) 32 */ 33 #ifndef XXH_FORCE_MEMORY_ACCESS /* can be defined externally, on command line for example */ 34 # if defined(__GNUC__) && ( defined(__ARM_ARCH_6__) || defined(__ARM_ARCH_6J__) || defined(__ARM_ARCH_6K__) || defined(__ARM_ARCH_6Z__) || defined(__ARM_ARCH_6ZK__) || defined(__ARM_ARCH_6T2__) ) 35 # define XXH_FORCE_MEMORY_ACCESS 2 36 # elif (defined(__INTEL_COMPILER) && !defined(WIN32)) || \ 37 (defined(__GNUC__) && ( defined(__ARM_ARCH_7__) || defined(__ARM_ARCH_7A__) || defined(__ARM_ARCH_7R__) || defined(__ARM_ARCH_7M__) || defined(__ARM_ARCH_7S__) )) || \ 38 defined(__ICCARM__) 39 # define XXH_FORCE_MEMORY_ACCESS 1 40 # endif 41 #endif 42 43 /*!XXH_ACCEPT_NULL_INPUT_POINTER : 44 * If the input pointer is a null pointer, xxHash default behavior is to trigger a memory access error, since it is a bad pointer. 45 * When this option is enabled, xxHash output for null input pointers will be the same as a null-length input. 46 * By default, this option is disabled. To enable it, uncomment below define : 47 */ 48 /* #define XXH_ACCEPT_NULL_INPUT_POINTER 1 */ 49 50 /*!XXH_FORCE_NATIVE_FORMAT : 51 * By default, xxHash library provides endian-independent Hash values, based on little-endian convention. 52 * Results are therefore identical for little-endian and big-endian CPU. 53 * This comes at a performance cost for big-endian CPU, since some swapping is required to emulate little-endian format. 54 * Should endian-independence be of no importance for your application, you may set the #define below to 1, 55 * to improve speed for Big-endian CPU. 56 * This option has no impact on Little_Endian CPU. 57 */ 58 #ifndef XXH_FORCE_NATIVE_FORMAT /* can be defined externally */ 59 # define XXH_FORCE_NATIVE_FORMAT 0 60 #endif 61 62 /*!XXH_FORCE_ALIGN_CHECK : 63 * This is a minor performance trick, only useful with lots of very small keys. 64 * It means : check for aligned/unaligned input. 65 * The check costs one initial branch per hash; set to 0 when the input data 66 * is guaranteed to be aligned. 67 */ 68 #ifndef XXH_FORCE_ALIGN_CHECK /* can be defined externally */ 69 # if defined(__i386) || defined(_M_IX86) || defined(__x86_64__) || defined(_M_X64) 70 # define XXH_FORCE_ALIGN_CHECK 0 71 # else 72 # define XXH_FORCE_ALIGN_CHECK 1 73 # endif 74 #endif 75 76 77 /* ************************************* 78 * Includes & Memory related functions 79 ***************************************/ 80 /* Modify the local functions below should you wish to use some other memory routines */ 81 /* for malloc(), free() */ 82 #include <stdlib.h> 83 #include <stddef.h> /* size_t */ 84 static void* XXH_malloc(size_t s) { return malloc(s); } 85 static void XXH_free (void* p) { free(p); } 86 /* for memcpy() */ 87 #include <string.h> 88 static void* XXH_memcpy(void* dest, const void* src, size_t size) { return memcpy(dest,src,size); } 89 90 #ifndef XXH_STATIC_LINKING_ONLY 91 # define XXH_STATIC_LINKING_ONLY 92 #endif 93 #include "xxhash.h" 94 95 96 /* ************************************* 97 * Compiler Specific Options 98 ***************************************/ 99 #if (defined(__GNUC__) && !defined(__STRICT_ANSI__)) || defined(__cplusplus) || defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ 100 # define INLINE_KEYWORD inline 101 #else 102 # define INLINE_KEYWORD 103 #endif 104 105 #if defined(__GNUC__) || defined(__ICCARM__) 106 # define FORCE_INLINE_ATTR __attribute__((always_inline)) 107 #elif defined(_MSC_VER) 108 # define FORCE_INLINE_ATTR __forceinline 109 #else 110 # define FORCE_INLINE_ATTR 111 #endif 112 113 #define FORCE_INLINE_TEMPLATE static INLINE_KEYWORD FORCE_INLINE_ATTR 114 115 116 #ifdef _MSC_VER 117 # pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ 118 #endif 119 120 121 /* ************************************* 122 * Basic Types 123 ***************************************/ 124 #ifndef MEM_MODULE 125 # define MEM_MODULE 126 # if !defined (__VMS) && (defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) ) 127 # include <stdint.h> 128 typedef uint8_t BYTE; 129 typedef uint16_t U16; 130 typedef uint32_t U32; 131 typedef int32_t S32; 132 typedef uint64_t U64; 133 # else 134 typedef unsigned char BYTE; 135 typedef unsigned short U16; 136 typedef unsigned int U32; 137 typedef signed int S32; 138 typedef unsigned long long U64; /* if your compiler doesn't support unsigned long long, replace by another 64-bit type here. Note that xxhash.h will also need to be updated. */ 139 # endif 140 #endif 141 142 143 #if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==2)) 144 145 /* Force direct memory access. Only works on CPU which support unaligned memory access in hardware */ 146 static U32 XXH_read32(const void* memPtr) { return *(const U32*) memPtr; } 147 static U64 XXH_read64(const void* memPtr) { return *(const U64*) memPtr; } 148 149 #elif (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==1)) 150 151 /* __pack instructions are safer, but compiler specific, hence potentially problematic for some compilers */ 152 /* currently only defined for gcc and icc */ 153 typedef union { U32 u32; U64 u64; } __attribute__((packed)) unalign; 154 155 static U32 XXH_read32(const void* ptr) { return ((const unalign*)ptr)->u32; } 156 static U64 XXH_read64(const void* ptr) { return ((const unalign*)ptr)->u64; } 157 158 #else 159 160 /* portable and safe solution. Generally efficient. 161 * see : http://stackoverflow.com/a/32095106/646947 162 */ 163 164 static U32 XXH_read32(const void* memPtr) 165 { 166 U32 val; 167 memcpy(&val, memPtr, sizeof(val)); 168 return val; 169 } 170 171 static U64 XXH_read64(const void* memPtr) 172 { 173 U64 val; 174 memcpy(&val, memPtr, sizeof(val)); 175 return val; 176 } 177 178 #endif /* XXH_FORCE_DIRECT_MEMORY_ACCESS */ 179 180 181 /* **************************************** 182 * Compiler-specific Functions and Macros 183 ******************************************/ 184 #define GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__) 185 186 /* Note : although _rotl exists for minGW (GCC under windows), performance seems poor */ 187 #if defined(_MSC_VER) 188 # define XXH_rotl32(x,r) _rotl(x,r) 189 # define XXH_rotl64(x,r) _rotl64(x,r) 190 #else 191 #if defined(__ICCARM__) 192 # include <intrinsics.h> 193 # define XXH_rotl32(x,r) __ROR(x,(32 - r)) 194 #else 195 # define XXH_rotl32(x,r) ((x << r) | (x >> (32 - r))) 196 #endif 197 # define XXH_rotl64(x,r) ((x << r) | (x >> (64 - r))) 198 #endif 199 200 #if defined(_MSC_VER) /* Visual Studio */ 201 # define XXH_swap32 _byteswap_ulong 202 # define XXH_swap64 _byteswap_uint64 203 #elif GCC_VERSION >= 403 204 # define XXH_swap32 __builtin_bswap32 205 # define XXH_swap64 __builtin_bswap64 206 #else 207 static U32 XXH_swap32 (U32 x) 208 { 209 return ((x << 24) & 0xff000000 ) | 210 ((x << 8) & 0x00ff0000 ) | 211 ((x >> 8) & 0x0000ff00 ) | 212 ((x >> 24) & 0x000000ff ); 213 } 214 static U64 XXH_swap64 (U64 x) 215 { 216 return ((x << 56) & 0xff00000000000000ULL) | 217 ((x << 40) & 0x00ff000000000000ULL) | 218 ((x << 24) & 0x0000ff0000000000ULL) | 219 ((x << 8) & 0x000000ff00000000ULL) | 220 ((x >> 8) & 0x00000000ff000000ULL) | 221 ((x >> 24) & 0x0000000000ff0000ULL) | 222 ((x >> 40) & 0x000000000000ff00ULL) | 223 ((x >> 56) & 0x00000000000000ffULL); 224 } 225 #endif 226 227 228 /* ************************************* 229 * Architecture Macros 230 ***************************************/ 231 typedef enum { XXH_bigEndian=0, XXH_littleEndian=1 } XXH_endianess; 232 233 /* XXH_CPU_LITTLE_ENDIAN can be defined externally, for example on the compiler command line */ 234 #ifndef XXH_CPU_LITTLE_ENDIAN 235 static const int g_one = 1; 236 # define XXH_CPU_LITTLE_ENDIAN (*(const char*)(&g_one)) 237 #endif 238 239 240 /* *************************** 241 * Memory reads 242 *****************************/ 243 typedef enum { XXH_aligned, XXH_unaligned } XXH_alignment; 244 245 FORCE_INLINE_TEMPLATE U32 XXH_readLE32_align(const void* ptr, XXH_endianess endian, XXH_alignment align) 246 { 247 if (align==XXH_unaligned) 248 return endian==XXH_littleEndian ? XXH_read32(ptr) : XXH_swap32(XXH_read32(ptr)); 249 else 250 return endian==XXH_littleEndian ? *(const U32*)ptr : XXH_swap32(*(const U32*)ptr); 251 } 252 253 FORCE_INLINE_TEMPLATE U32 XXH_readLE32(const void* ptr, XXH_endianess endian) 254 { 255 return XXH_readLE32_align(ptr, endian, XXH_unaligned); 256 } 257 258 static U32 XXH_readBE32(const void* ptr) 259 { 260 return XXH_CPU_LITTLE_ENDIAN ? XXH_swap32(XXH_read32(ptr)) : XXH_read32(ptr); 261 } 262 263 FORCE_INLINE_TEMPLATE U64 XXH_readLE64_align(const void* ptr, XXH_endianess endian, XXH_alignment align) 264 { 265 if (align==XXH_unaligned) 266 return endian==XXH_littleEndian ? XXH_read64(ptr) : XXH_swap64(XXH_read64(ptr)); 267 else 268 return endian==XXH_littleEndian ? *(const U64*)ptr : XXH_swap64(*(const U64*)ptr); 269 } 270 271 FORCE_INLINE_TEMPLATE U64 XXH_readLE64(const void* ptr, XXH_endianess endian) 272 { 273 return XXH_readLE64_align(ptr, endian, XXH_unaligned); 274 } 275 276 static U64 XXH_readBE64(const void* ptr) 277 { 278 return XXH_CPU_LITTLE_ENDIAN ? XXH_swap64(XXH_read64(ptr)) : XXH_read64(ptr); 279 } 280 281 282 /* ************************************* 283 * Macros 284 ***************************************/ 285 #define XXH_STATIC_ASSERT(c) { enum { XXH_static_assert = 1/(int)(!!(c)) }; } /* use only *after* variable declarations */ 286 287 288 /* ************************************* 289 * Constants 290 ***************************************/ 291 static const U32 PRIME32_1 = 2654435761U; 292 static const U32 PRIME32_2 = 2246822519U; 293 static const U32 PRIME32_3 = 3266489917U; 294 static const U32 PRIME32_4 = 668265263U; 295 static const U32 PRIME32_5 = 374761393U; 296 297 static const U64 PRIME64_1 = 11400714785074694791ULL; 298 static const U64 PRIME64_2 = 14029467366897019727ULL; 299 static const U64 PRIME64_3 = 1609587929392839161ULL; 300 static const U64 PRIME64_4 = 9650029242287828579ULL; 301 static const U64 PRIME64_5 = 2870177450012600261ULL; 302 303 XXH_PUBLIC_API unsigned XXH_versionNumber (void) { return XXH_VERSION_NUMBER; } 304 305 306 /* ************************** 307 * Utils 308 ****************************/ 309 XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* restrict dstState, const XXH32_state_t* restrict srcState) 310 { 311 memcpy(dstState, srcState, sizeof(*dstState)); 312 } 313 314 XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* restrict dstState, const XXH64_state_t* restrict srcState) 315 { 316 memcpy(dstState, srcState, sizeof(*dstState)); 317 } 318 319 320 /* *************************** 321 * Simple Hash Functions 322 *****************************/ 323 324 static U32 XXH32_round(U32 seed, U32 input) 325 { 326 seed += input * PRIME32_2; 327 seed = XXH_rotl32(seed, 13); 328 seed *= PRIME32_1; 329 return seed; 330 } 331 332 FORCE_INLINE_TEMPLATE U32 XXH32_endian_align(const void* input, size_t len, U32 seed, XXH_endianess endian, XXH_alignment align) 333 { 334 const BYTE* p = (const BYTE*)input; 335 const BYTE* bEnd = p + len; 336 U32 h32; 337 #define XXH_get32bits(p) XXH_readLE32_align(p, endian, align) 338 339 #ifdef XXH_ACCEPT_NULL_INPUT_POINTER 340 if (p==NULL) { 341 len=0; 342 bEnd=p=(const BYTE*)(size_t)16; 343 } 344 #endif 345 346 if (len>=16) { 347 const BYTE* const limit = bEnd - 16; 348 U32 v1 = seed + PRIME32_1 + PRIME32_2; 349 U32 v2 = seed + PRIME32_2; 350 U32 v3 = seed + 0; 351 U32 v4 = seed - PRIME32_1; 352 353 do { 354 v1 = XXH32_round(v1, XXH_get32bits(p)); p+=4; 355 v2 = XXH32_round(v2, XXH_get32bits(p)); p+=4; 356 v3 = XXH32_round(v3, XXH_get32bits(p)); p+=4; 357 v4 = XXH32_round(v4, XXH_get32bits(p)); p+=4; 358 } while (p<=limit); 359 360 h32 = XXH_rotl32(v1, 1) + XXH_rotl32(v2, 7) + XXH_rotl32(v3, 12) + XXH_rotl32(v4, 18); 361 } else { 362 h32 = seed + PRIME32_5; 363 } 364 365 h32 += (U32) len; 366 367 while (p+4<=bEnd) { 368 h32 += XXH_get32bits(p) * PRIME32_3; 369 h32 = XXH_rotl32(h32, 17) * PRIME32_4 ; 370 p+=4; 371 } 372 373 while (p<bEnd) { 374 h32 += (*p) * PRIME32_5; 375 h32 = XXH_rotl32(h32, 11) * PRIME32_1 ; 376 p++; 377 } 378 379 h32 ^= h32 >> 15; 380 h32 *= PRIME32_2; 381 h32 ^= h32 >> 13; 382 h32 *= PRIME32_3; 383 h32 ^= h32 >> 16; 384 385 return h32; 386 } 387 388 389 XXH_PUBLIC_API unsigned int XXH32 (const void* input, size_t len, unsigned int seed) 390 { 391 #if 0 392 /* Simple version, good for code maintenance, but unfortunately slow for small inputs */ 393 XXH32_CREATESTATE_STATIC(state); 394 XXH32_reset(state, seed); 395 XXH32_update(state, input, len); 396 return XXH32_digest(state); 397 #else 398 XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; 399 400 if (XXH_FORCE_ALIGN_CHECK) { 401 if ((((size_t)input) & 3) == 0) { /* Input is 4-bytes aligned, leverage the speed benefit */ 402 if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) 403 return XXH32_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned); 404 else 405 return XXH32_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned); 406 } } 407 408 if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) 409 return XXH32_endian_align(input, len, seed, XXH_littleEndian, XXH_unaligned); 410 else 411 return XXH32_endian_align(input, len, seed, XXH_bigEndian, XXH_unaligned); 412 #endif 413 } 414 415 416 static U64 XXH64_round(U64 acc, U64 input) 417 { 418 acc += input * PRIME64_2; 419 acc = XXH_rotl64(acc, 31); 420 acc *= PRIME64_1; 421 return acc; 422 } 423 424 static U64 XXH64_mergeRound(U64 acc, U64 val) 425 { 426 val = XXH64_round(0, val); 427 acc ^= val; 428 acc = acc * PRIME64_1 + PRIME64_4; 429 return acc; 430 } 431 432 FORCE_INLINE_TEMPLATE U64 XXH64_endian_align(const void* input, size_t len, U64 seed, XXH_endianess endian, XXH_alignment align) 433 { 434 const BYTE* p = (const BYTE*)input; 435 const BYTE* const bEnd = p + len; 436 U64 h64; 437 #define XXH_get64bits(p) XXH_readLE64_align(p, endian, align) 438 439 #ifdef XXH_ACCEPT_NULL_INPUT_POINTER 440 if (p==NULL) { 441 len=0; 442 bEnd=p=(const BYTE*)(size_t)32; 443 } 444 #endif 445 446 if (len>=32) { 447 const BYTE* const limit = bEnd - 32; 448 U64 v1 = seed + PRIME64_1 + PRIME64_2; 449 U64 v2 = seed + PRIME64_2; 450 U64 v3 = seed + 0; 451 U64 v4 = seed - PRIME64_1; 452 453 do { 454 v1 = XXH64_round(v1, XXH_get64bits(p)); p+=8; 455 v2 = XXH64_round(v2, XXH_get64bits(p)); p+=8; 456 v3 = XXH64_round(v3, XXH_get64bits(p)); p+=8; 457 v4 = XXH64_round(v4, XXH_get64bits(p)); p+=8; 458 } while (p<=limit); 459 460 h64 = XXH_rotl64(v1, 1) + XXH_rotl64(v2, 7) + XXH_rotl64(v3, 12) + XXH_rotl64(v4, 18); 461 h64 = XXH64_mergeRound(h64, v1); 462 h64 = XXH64_mergeRound(h64, v2); 463 h64 = XXH64_mergeRound(h64, v3); 464 h64 = XXH64_mergeRound(h64, v4); 465 466 } else { 467 h64 = seed + PRIME64_5; 468 } 469 470 h64 += (U64) len; 471 472 while (p+8<=bEnd) { 473 U64 const k1 = XXH64_round(0, XXH_get64bits(p)); 474 h64 ^= k1; 475 h64 = XXH_rotl64(h64,27) * PRIME64_1 + PRIME64_4; 476 p+=8; 477 } 478 479 if (p+4<=bEnd) { 480 h64 ^= (U64)(XXH_get32bits(p)) * PRIME64_1; 481 h64 = XXH_rotl64(h64, 23) * PRIME64_2 + PRIME64_3; 482 p+=4; 483 } 484 485 while (p<bEnd) { 486 h64 ^= (*p) * PRIME64_5; 487 h64 = XXH_rotl64(h64, 11) * PRIME64_1; 488 p++; 489 } 490 491 h64 ^= h64 >> 33; 492 h64 *= PRIME64_2; 493 h64 ^= h64 >> 29; 494 h64 *= PRIME64_3; 495 h64 ^= h64 >> 32; 496 497 return h64; 498 } 499 500 501 XXH_PUBLIC_API unsigned long long XXH64 (const void* input, size_t len, unsigned long long seed) 502 { 503 #if 0 504 /* Simple version, good for code maintenance, but unfortunately slow for small inputs */ 505 XXH64_CREATESTATE_STATIC(state); 506 XXH64_reset(state, seed); 507 XXH64_update(state, input, len); 508 return XXH64_digest(state); 509 #else 510 XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; 511 512 if (XXH_FORCE_ALIGN_CHECK) { 513 if ((((size_t)input) & 7)==0) { /* Input is aligned, let's leverage the speed advantage */ 514 if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) 515 return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned); 516 else 517 return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned); 518 } } 519 520 if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) 521 return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_unaligned); 522 else 523 return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_unaligned); 524 #endif 525 } 526 527 528 /* ************************************************** 529 * Advanced Hash Functions 530 ****************************************************/ 531 532 XXH_PUBLIC_API XXH32_state_t* XXH32_createState(void) 533 { 534 return (XXH32_state_t*)XXH_malloc(sizeof(XXH32_state_t)); 535 } 536 XXH_PUBLIC_API XXH_errorcode XXH32_freeState(XXH32_state_t* statePtr) 537 { 538 XXH_free(statePtr); 539 return XXH_OK; 540 } 541 542 XXH_PUBLIC_API XXH64_state_t* XXH64_createState(void) 543 { 544 return (XXH64_state_t*)XXH_malloc(sizeof(XXH64_state_t)); 545 } 546 XXH_PUBLIC_API XXH_errorcode XXH64_freeState(XXH64_state_t* statePtr) 547 { 548 XXH_free(statePtr); 549 return XXH_OK; 550 } 551 552 553 /*** Hash feed ***/ 554 555 XXH_PUBLIC_API XXH_errorcode XXH32_reset(XXH32_state_t* statePtr, unsigned int seed) 556 { 557 XXH32_state_t state; /* using a local state to memcpy() in order to avoid strict-aliasing warnings */ 558 memset(&state, 0, sizeof(state)-4); /* do not write into reserved, for future removal */ 559 state.v1 = seed + PRIME32_1 + PRIME32_2; 560 state.v2 = seed + PRIME32_2; 561 state.v3 = seed + 0; 562 state.v4 = seed - PRIME32_1; 563 memcpy(statePtr, &state, sizeof(state)); 564 return XXH_OK; 565 } 566 567 568 XXH_PUBLIC_API XXH_errorcode XXH64_reset(XXH64_state_t* statePtr, unsigned long long seed) 569 { 570 XXH64_state_t state; /* using a local state to memcpy() in order to avoid strict-aliasing warnings */ 571 memset(&state, 0, sizeof(state)-8); /* do not write into reserved, for future removal */ 572 state.v1 = seed + PRIME64_1 + PRIME64_2; 573 state.v2 = seed + PRIME64_2; 574 state.v3 = seed + 0; 575 state.v4 = seed - PRIME64_1; 576 memcpy(statePtr, &state, sizeof(state)); 577 return XXH_OK; 578 } 579 580 581 FORCE_INLINE_TEMPLATE XXH_errorcode XXH32_update_endian (XXH32_state_t* state, const void* input, size_t len, XXH_endianess endian) 582 { 583 const BYTE* p = (const BYTE*)input; 584 const BYTE* const bEnd = p + len; 585 586 #ifdef XXH_ACCEPT_NULL_INPUT_POINTER 587 if (input==NULL) return XXH_ERROR; 588 #endif 589 590 state->total_len_32 += (unsigned)len; 591 state->large_len |= (len>=16) | (state->total_len_32>=16); 592 593 if (state->memsize + len < 16) { /* fill in tmp buffer */ 594 XXH_memcpy((BYTE*)(state->mem32) + state->memsize, input, len); 595 state->memsize += (unsigned)len; 596 return XXH_OK; 597 } 598 599 if (state->memsize) { /* some data left from previous update */ 600 XXH_memcpy((BYTE*)(state->mem32) + state->memsize, input, 16-state->memsize); 601 { const U32* p32 = state->mem32; 602 state->v1 = XXH32_round(state->v1, XXH_readLE32(p32, endian)); p32++; 603 state->v2 = XXH32_round(state->v2, XXH_readLE32(p32, endian)); p32++; 604 state->v3 = XXH32_round(state->v3, XXH_readLE32(p32, endian)); p32++; 605 state->v4 = XXH32_round(state->v4, XXH_readLE32(p32, endian)); p32++; 606 } 607 p += 16-state->memsize; 608 state->memsize = 0; 609 } 610 611 if (p <= bEnd-16) { 612 const BYTE* const limit = bEnd - 16; 613 U32 v1 = state->v1; 614 U32 v2 = state->v2; 615 U32 v3 = state->v3; 616 U32 v4 = state->v4; 617 618 do { 619 v1 = XXH32_round(v1, XXH_readLE32(p, endian)); p+=4; 620 v2 = XXH32_round(v2, XXH_readLE32(p, endian)); p+=4; 621 v3 = XXH32_round(v3, XXH_readLE32(p, endian)); p+=4; 622 v4 = XXH32_round(v4, XXH_readLE32(p, endian)); p+=4; 623 } while (p<=limit); 624 625 state->v1 = v1; 626 state->v2 = v2; 627 state->v3 = v3; 628 state->v4 = v4; 629 } 630 631 if (p < bEnd) { 632 XXH_memcpy(state->mem32, p, (size_t)(bEnd-p)); 633 state->memsize = (unsigned)(bEnd-p); 634 } 635 636 return XXH_OK; 637 } 638 639 XXH_PUBLIC_API XXH_errorcode XXH32_update (XXH32_state_t* state_in, const void* input, size_t len) 640 { 641 XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; 642 643 if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) 644 return XXH32_update_endian(state_in, input, len, XXH_littleEndian); 645 else 646 return XXH32_update_endian(state_in, input, len, XXH_bigEndian); 647 } 648 649 650 651 FORCE_INLINE_TEMPLATE U32 XXH32_digest_endian (const XXH32_state_t* state, XXH_endianess endian) 652 { 653 const BYTE * p = (const BYTE*)state->mem32; 654 const BYTE* const bEnd = (const BYTE*)(state->mem32) + state->memsize; 655 U32 h32; 656 657 if (state->large_len) { 658 h32 = XXH_rotl32(state->v1, 1) + XXH_rotl32(state->v2, 7) + XXH_rotl32(state->v3, 12) + XXH_rotl32(state->v4, 18); 659 } else { 660 h32 = state->v3 /* == seed */ + PRIME32_5; 661 } 662 663 h32 += state->total_len_32; 664 665 while (p+4<=bEnd) { 666 h32 += XXH_readLE32(p, endian) * PRIME32_3; 667 h32 = XXH_rotl32(h32, 17) * PRIME32_4; 668 p+=4; 669 } 670 671 while (p<bEnd) { 672 h32 += (*p) * PRIME32_5; 673 h32 = XXH_rotl32(h32, 11) * PRIME32_1; 674 p++; 675 } 676 677 h32 ^= h32 >> 15; 678 h32 *= PRIME32_2; 679 h32 ^= h32 >> 13; 680 h32 *= PRIME32_3; 681 h32 ^= h32 >> 16; 682 683 return h32; 684 } 685 686 687 XXH_PUBLIC_API unsigned int XXH32_digest (const XXH32_state_t* state_in) 688 { 689 XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; 690 691 if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) 692 return XXH32_digest_endian(state_in, XXH_littleEndian); 693 else 694 return XXH32_digest_endian(state_in, XXH_bigEndian); 695 } 696 697 698 699 /* **** XXH64 **** */ 700 701 FORCE_INLINE_TEMPLATE XXH_errorcode XXH64_update_endian (XXH64_state_t* state, const void* input, size_t len, XXH_endianess endian) 702 { 703 const BYTE* p = (const BYTE*)input; 704 const BYTE* const bEnd = p + len; 705 706 #ifdef XXH_ACCEPT_NULL_INPUT_POINTER 707 if (input==NULL) return XXH_ERROR; 708 #endif 709 710 state->total_len += len; 711 712 if (state->memsize + len < 32) { /* fill in tmp buffer */ 713 if (input != NULL) { 714 XXH_memcpy(((BYTE*)state->mem64) + state->memsize, input, len); 715 } 716 state->memsize += (U32)len; 717 return XXH_OK; 718 } 719 720 if (state->memsize) { /* tmp buffer is full */ 721 XXH_memcpy(((BYTE*)state->mem64) + state->memsize, input, 32-state->memsize); 722 state->v1 = XXH64_round(state->v1, XXH_readLE64(state->mem64+0, endian)); 723 state->v2 = XXH64_round(state->v2, XXH_readLE64(state->mem64+1, endian)); 724 state->v3 = XXH64_round(state->v3, XXH_readLE64(state->mem64+2, endian)); 725 state->v4 = XXH64_round(state->v4, XXH_readLE64(state->mem64+3, endian)); 726 p += 32-state->memsize; 727 state->memsize = 0; 728 } 729 730 if (p+32 <= bEnd) { 731 const BYTE* const limit = bEnd - 32; 732 U64 v1 = state->v1; 733 U64 v2 = state->v2; 734 U64 v3 = state->v3; 735 U64 v4 = state->v4; 736 737 do { 738 v1 = XXH64_round(v1, XXH_readLE64(p, endian)); p+=8; 739 v2 = XXH64_round(v2, XXH_readLE64(p, endian)); p+=8; 740 v3 = XXH64_round(v3, XXH_readLE64(p, endian)); p+=8; 741 v4 = XXH64_round(v4, XXH_readLE64(p, endian)); p+=8; 742 } while (p<=limit); 743 744 state->v1 = v1; 745 state->v2 = v2; 746 state->v3 = v3; 747 state->v4 = v4; 748 } 749 750 if (p < bEnd) { 751 XXH_memcpy(state->mem64, p, (size_t)(bEnd-p)); 752 state->memsize = (unsigned)(bEnd-p); 753 } 754 755 return XXH_OK; 756 } 757 758 XXH_PUBLIC_API XXH_errorcode XXH64_update (XXH64_state_t* state_in, const void* input, size_t len) 759 { 760 XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; 761 762 if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) 763 return XXH64_update_endian(state_in, input, len, XXH_littleEndian); 764 else 765 return XXH64_update_endian(state_in, input, len, XXH_bigEndian); 766 } 767 768 769 770 FORCE_INLINE_TEMPLATE U64 XXH64_digest_endian (const XXH64_state_t* state, XXH_endianess endian) 771 { 772 const BYTE * p = (const BYTE*)state->mem64; 773 const BYTE* const bEnd = (const BYTE*)state->mem64 + state->memsize; 774 U64 h64; 775 776 if (state->total_len >= 32) { 777 U64 const v1 = state->v1; 778 U64 const v2 = state->v2; 779 U64 const v3 = state->v3; 780 U64 const v4 = state->v4; 781 782 h64 = XXH_rotl64(v1, 1) + XXH_rotl64(v2, 7) + XXH_rotl64(v3, 12) + XXH_rotl64(v4, 18); 783 h64 = XXH64_mergeRound(h64, v1); 784 h64 = XXH64_mergeRound(h64, v2); 785 h64 = XXH64_mergeRound(h64, v3); 786 h64 = XXH64_mergeRound(h64, v4); 787 } else { 788 h64 = state->v3 + PRIME64_5; 789 } 790 791 h64 += (U64) state->total_len; 792 793 while (p+8<=bEnd) { 794 U64 const k1 = XXH64_round(0, XXH_readLE64(p, endian)); 795 h64 ^= k1; 796 h64 = XXH_rotl64(h64,27) * PRIME64_1 + PRIME64_4; 797 p+=8; 798 } 799 800 if (p+4<=bEnd) { 801 h64 ^= (U64)(XXH_readLE32(p, endian)) * PRIME64_1; 802 h64 = XXH_rotl64(h64, 23) * PRIME64_2 + PRIME64_3; 803 p+=4; 804 } 805 806 while (p<bEnd) { 807 h64 ^= (*p) * PRIME64_5; 808 h64 = XXH_rotl64(h64, 11) * PRIME64_1; 809 p++; 810 } 811 812 h64 ^= h64 >> 33; 813 h64 *= PRIME64_2; 814 h64 ^= h64 >> 29; 815 h64 *= PRIME64_3; 816 h64 ^= h64 >> 32; 817 818 return h64; 819 } 820 821 822 XXH_PUBLIC_API unsigned long long XXH64_digest (const XXH64_state_t* state_in) 823 { 824 XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; 825 826 if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) 827 return XXH64_digest_endian(state_in, XXH_littleEndian); 828 else 829 return XXH64_digest_endian(state_in, XXH_bigEndian); 830 } 831 832 833 /* ************************** 834 * Canonical representation 835 ****************************/ 836 837 /*! Default XXH result types are basic unsigned 32 and 64 bits. 838 * The canonical representation follows human-readable write convention, aka big-endian (large digits first). 839 * These functions allow transformation of hash result into and from its canonical format. 840 * This way, hash values can be written into a file or buffer, and remain comparable across different systems and programs. 841 */ 842 843 XXH_PUBLIC_API void XXH32_canonicalFromHash(XXH32_canonical_t* dst, XXH32_hash_t hash) 844 { 845 XXH_STATIC_ASSERT(sizeof(XXH32_canonical_t) == sizeof(XXH32_hash_t)); 846 if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap32(hash); 847 memcpy(dst, &hash, sizeof(*dst)); 848 } 849 850 XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t hash) 851 { 852 XXH_STATIC_ASSERT(sizeof(XXH64_canonical_t) == sizeof(XXH64_hash_t)); 853 if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap64(hash); 854 memcpy(dst, &hash, sizeof(*dst)); 855 } 856 857 XXH_PUBLIC_API XXH32_hash_t XXH32_hashFromCanonical(const XXH32_canonical_t* src) 858 { 859 return XXH_readBE32(src); 860 } 861 862 XXH_PUBLIC_API XXH64_hash_t XXH64_hashFromCanonical(const XXH64_canonical_t* src) 863 { 864 return XXH_readBE64(src); 865 } 866