sha512.c revision 4969cc9b0ab2905ec478277f50ed3849b37a6c6b
1/* Copyright (C) 1995-1998 Eric Young (eay@cryptsoft.com) 2 * All rights reserved. 3 * 4 * This package is an SSL implementation written 5 * by Eric Young (eay@cryptsoft.com). 6 * The implementation was written so as to conform with Netscapes SSL. 7 * 8 * This library is free for commercial and non-commercial use as long as 9 * the following conditions are aheared to. The following conditions 10 * apply to all code found in this distribution, be it the RC4, RSA, 11 * lhash, DES, etc., code; not just the SSL code. The SSL documentation 12 * included with this distribution is covered by the same copyright terms 13 * except that the holder is Tim Hudson (tjh@cryptsoft.com). 14 * 15 * Copyright remains Eric Young's, and as such any Copyright notices in 16 * the code are not to be removed. 17 * If this package is used in a product, Eric Young should be given attribution 18 * as the author of the parts of the library used. 19 * This can be in the form of a textual message at program startup or 20 * in documentation (online or textual) provided with the package. 21 * 22 * Redistribution and use in source and binary forms, with or without 23 * modification, are permitted provided that the following conditions 24 * are met: 25 * 1. Redistributions of source code must retain the copyright 26 * notice, this list of conditions and the following disclaimer. 27 * 2. Redistributions in binary form must reproduce the above copyright 28 * notice, this list of conditions and the following disclaimer in the 29 * documentation and/or other materials provided with the distribution. 30 * 3. All advertising materials mentioning features or use of this software 31 * must display the following acknowledgement: 32 * "This product includes cryptographic software written by 33 * Eric Young (eay@cryptsoft.com)" 34 * The word 'cryptographic' can be left out if the rouines from the library 35 * being used are not cryptographic related :-). 36 * 4. If you include any Windows specific code (or a derivative thereof) from 37 * the apps directory (application code) you must include an acknowledgement: 38 * "This product includes software written by Tim Hudson (tjh@cryptsoft.com)" 39 * 40 * THIS SOFTWARE IS PROVIDED BY ERIC YOUNG ``AS IS'' AND 41 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 42 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 43 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE 44 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 45 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 46 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 47 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 48 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 49 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 50 * SUCH DAMAGE. 51 * 52 * The licence and distribution terms for any publically available version or 53 * derivative of this code cannot be changed. i.e. this code cannot simply be 54 * copied and put under another distribution licence 55 * [including the GNU Public Licence.] */ 56 57#include <openssl/sha.h> 58 59#include <string.h> 60 61#include <openssl/mem.h> 62 63 64/* IMPLEMENTATION NOTES. 65 * 66 * The 32-bit hash algorithms share a common byte-order neutral collector and 67 * padding function implementations that operate on unaligned data, 68 * ../md32_common.h. This SHA-512 implementation does not. Reasons 69 * [in reverse order] are: 70 * 71 * - It's the only 64-bit hash algorithm for the moment of this writing, 72 * there is no need for common collector/padding implementation [yet]; 73 * - By supporting only a transform function that operates on *aligned* data 74 * the collector/padding function is simpler and easier to optimize. */ 75 76#if !defined(OPENSSL_NO_ASM) && \ 77 (defined(OPENSSL_X86) || defined(OPENSSL_X86_64) || \ 78 defined(OPENSSL_ARM) || defined(OPENSSL_AARCH64)) 79#define SHA512_ASM 80#endif 81 82#if defined(OPENSSL_X86) || defined(OPENSSL_X86_64) || \ 83 defined(__ARM_FEATURE_UNALIGNED) 84#define SHA512_BLOCK_CAN_MANAGE_UNALIGNED_DATA 85#endif 86 87int SHA384_Init(SHA512_CTX *sha) { 88 sha->h[0] = UINT64_C(0xcbbb9d5dc1059ed8); 89 sha->h[1] = UINT64_C(0x629a292a367cd507); 90 sha->h[2] = UINT64_C(0x9159015a3070dd17); 91 sha->h[3] = UINT64_C(0x152fecd8f70e5939); 92 sha->h[4] = UINT64_C(0x67332667ffc00b31); 93 sha->h[5] = UINT64_C(0x8eb44a8768581511); 94 sha->h[6] = UINT64_C(0xdb0c2e0d64f98fa7); 95 sha->h[7] = UINT64_C(0x47b5481dbefa4fa4); 96 97 sha->Nl = 0; 98 sha->Nh = 0; 99 sha->num = 0; 100 sha->md_len = SHA384_DIGEST_LENGTH; 101 return 1; 102} 103 104 105int SHA512_Init(SHA512_CTX *sha) { 106 sha->h[0] = UINT64_C(0x6a09e667f3bcc908); 107 sha->h[1] = UINT64_C(0xbb67ae8584caa73b); 108 sha->h[2] = UINT64_C(0x3c6ef372fe94f82b); 109 sha->h[3] = UINT64_C(0xa54ff53a5f1d36f1); 110 sha->h[4] = UINT64_C(0x510e527fade682d1); 111 sha->h[5] = UINT64_C(0x9b05688c2b3e6c1f); 112 sha->h[6] = UINT64_C(0x1f83d9abfb41bd6b); 113 sha->h[7] = UINT64_C(0x5be0cd19137e2179); 114 115 sha->Nl = 0; 116 sha->Nh = 0; 117 sha->num = 0; 118 sha->md_len = SHA512_DIGEST_LENGTH; 119 return 1; 120} 121 122uint8_t *SHA384(const uint8_t *data, size_t len, uint8_t *out) { 123 SHA512_CTX ctx; 124 static uint8_t buf[SHA384_DIGEST_LENGTH]; 125 126 /* TODO(fork): remove this static buffer. */ 127 if (out == NULL) { 128 out = buf; 129 } 130 131 SHA384_Init(&ctx); 132 SHA384_Update(&ctx, data, len); 133 SHA384_Final(out, &ctx); 134 OPENSSL_cleanse(&ctx, sizeof(ctx)); 135 return out; 136} 137 138uint8_t *SHA512(const uint8_t *data, size_t len, uint8_t *out) { 139 SHA512_CTX ctx; 140 static uint8_t buf[SHA512_DIGEST_LENGTH]; 141 142 /* TODO(fork): remove this static buffer. */ 143 if (out == NULL) { 144 out = buf; 145 } 146 SHA512_Init(&ctx); 147 SHA512_Update(&ctx, data, len); 148 SHA512_Final(out, &ctx); 149 OPENSSL_cleanse(&ctx, sizeof(ctx)); 150 return out; 151} 152 153#if !defined(SHA512_ASM) 154static 155#endif 156void sha512_block_data_order(uint64_t *state, const uint64_t *W, size_t num); 157 158 159int SHA384_Final(uint8_t *md, SHA512_CTX *sha) { 160 return SHA512_Final(md, sha); 161} 162 163int SHA384_Update(SHA512_CTX *sha, const void *data, size_t len) { 164 return SHA512_Update(sha, data, len); 165} 166 167void SHA512_Transform(SHA512_CTX *c, const uint8_t *data) { 168#ifndef SHA512_BLOCK_CAN_MANAGE_UNALIGNED_DATA 169 if ((size_t)data % sizeof(c->u.d[0]) != 0) { 170 memcpy(c->u.p, data, sizeof(c->u.p)); 171 data = c->u.p; 172 } 173#endif 174 sha512_block_data_order(c->h, (uint64_t *)data, 1); 175} 176 177int SHA512_Update(SHA512_CTX *c, const void *in_data, size_t len) { 178 uint64_t l; 179 uint8_t *p = c->u.p; 180 const uint8_t *data = (const uint8_t *)in_data; 181 182 if (len == 0) { 183 return 1; 184 } 185 186 l = (c->Nl + (((uint64_t)len) << 3)) & UINT64_C(0xffffffffffffffff); 187 if (l < c->Nl) { 188 c->Nh++; 189 } 190 if (sizeof(len) >= 8) { 191 c->Nh += (((uint64_t)len) >> 61); 192 } 193 c->Nl = l; 194 195 if (c->num != 0) { 196 size_t n = sizeof(c->u) - c->num; 197 198 if (len < n) { 199 memcpy(p + c->num, data, len); 200 c->num += (unsigned int)len; 201 return 1; 202 } else { 203 memcpy(p + c->num, data, n), c->num = 0; 204 len -= n; 205 data += n; 206 sha512_block_data_order(c->h, (uint64_t *)p, 1); 207 } 208 } 209 210 if (len >= sizeof(c->u)) { 211#ifndef SHA512_BLOCK_CAN_MANAGE_UNALIGNED_DATA 212 if ((size_t)data % sizeof(c->u.d[0]) != 0) { 213 while (len >= sizeof(c->u)) { 214 memcpy(p, data, sizeof(c->u)); 215 sha512_block_data_order(c->h, (uint64_t *)p, 1); 216 len -= sizeof(c->u); 217 data += sizeof(c->u); 218 } 219 } else 220#endif 221 { 222 sha512_block_data_order(c->h, (uint64_t *)data, len / sizeof(c->u)); 223 data += len; 224 len %= sizeof(c->u); 225 data -= len; 226 } 227 } 228 229 if (len != 0) { 230 memcpy(p, data, len); 231 c->num = (int)len; 232 } 233 234 return 1; 235} 236 237int SHA512_Final(uint8_t *md, SHA512_CTX *sha) { 238 uint8_t *p = (uint8_t *)sha->u.p; 239 size_t n = sha->num; 240 241 p[n] = 0x80; /* There always is a room for one */ 242 n++; 243 if (n > (sizeof(sha->u) - 16)) { 244 memset(p + n, 0, sizeof(sha->u) - n); 245 n = 0; 246 sha512_block_data_order(sha->h, (uint64_t *)p, 1); 247 } 248 249 memset(p + n, 0, sizeof(sha->u) - 16 - n); 250 p[sizeof(sha->u) - 1] = (uint8_t)(sha->Nl); 251 p[sizeof(sha->u) - 2] = (uint8_t)(sha->Nl >> 8); 252 p[sizeof(sha->u) - 3] = (uint8_t)(sha->Nl >> 16); 253 p[sizeof(sha->u) - 4] = (uint8_t)(sha->Nl >> 24); 254 p[sizeof(sha->u) - 5] = (uint8_t)(sha->Nl >> 32); 255 p[sizeof(sha->u) - 6] = (uint8_t)(sha->Nl >> 40); 256 p[sizeof(sha->u) - 7] = (uint8_t)(sha->Nl >> 48); 257 p[sizeof(sha->u) - 8] = (uint8_t)(sha->Nl >> 56); 258 p[sizeof(sha->u) - 9] = (uint8_t)(sha->Nh); 259 p[sizeof(sha->u) - 10] = (uint8_t)(sha->Nh >> 8); 260 p[sizeof(sha->u) - 11] = (uint8_t)(sha->Nh >> 16); 261 p[sizeof(sha->u) - 12] = (uint8_t)(sha->Nh >> 24); 262 p[sizeof(sha->u) - 13] = (uint8_t)(sha->Nh >> 32); 263 p[sizeof(sha->u) - 14] = (uint8_t)(sha->Nh >> 40); 264 p[sizeof(sha->u) - 15] = (uint8_t)(sha->Nh >> 48); 265 p[sizeof(sha->u) - 16] = (uint8_t)(sha->Nh >> 56); 266 267 sha512_block_data_order(sha->h, (uint64_t *)p, 1); 268 269 if (md == NULL) { 270 /* TODO(davidben): This NULL check is absent in other low-level hash 'final' 271 * functions and is one of the few places one can fail. */ 272 return 0; 273 } 274 275 switch (sha->md_len) { 276 /* Let compiler decide if it's appropriate to unroll... */ 277 case SHA384_DIGEST_LENGTH: 278 for (n = 0; n < SHA384_DIGEST_LENGTH / 8; n++) { 279 uint64_t t = sha->h[n]; 280 281 *(md++) = (uint8_t)(t >> 56); 282 *(md++) = (uint8_t)(t >> 48); 283 *(md++) = (uint8_t)(t >> 40); 284 *(md++) = (uint8_t)(t >> 32); 285 *(md++) = (uint8_t)(t >> 24); 286 *(md++) = (uint8_t)(t >> 16); 287 *(md++) = (uint8_t)(t >> 8); 288 *(md++) = (uint8_t)(t); 289 } 290 break; 291 case SHA512_DIGEST_LENGTH: 292 for (n = 0; n < SHA512_DIGEST_LENGTH / 8; n++) { 293 uint64_t t = sha->h[n]; 294 295 *(md++) = (uint8_t)(t >> 56); 296 *(md++) = (uint8_t)(t >> 48); 297 *(md++) = (uint8_t)(t >> 40); 298 *(md++) = (uint8_t)(t >> 32); 299 *(md++) = (uint8_t)(t >> 24); 300 *(md++) = (uint8_t)(t >> 16); 301 *(md++) = (uint8_t)(t >> 8); 302 *(md++) = (uint8_t)(t); 303 } 304 break; 305 /* ... as well as make sure md_len is not abused. */ 306 default: 307 /* TODO(davidben): This bad |md_len| case is one of the few places a 308 * low-level hash 'final' function can fail. This should never happen. */ 309 return 0; 310 } 311 312 return 1; 313} 314 315#ifndef SHA512_ASM 316static const uint64_t K512[80] = { 317 UINT64_C(0x428a2f98d728ae22), UINT64_C(0x7137449123ef65cd), 318 UINT64_C(0xb5c0fbcfec4d3b2f), UINT64_C(0xe9b5dba58189dbbc), 319 UINT64_C(0x3956c25bf348b538), UINT64_C(0x59f111f1b605d019), 320 UINT64_C(0x923f82a4af194f9b), UINT64_C(0xab1c5ed5da6d8118), 321 UINT64_C(0xd807aa98a3030242), UINT64_C(0x12835b0145706fbe), 322 UINT64_C(0x243185be4ee4b28c), UINT64_C(0x550c7dc3d5ffb4e2), 323 UINT64_C(0x72be5d74f27b896f), UINT64_C(0x80deb1fe3b1696b1), 324 UINT64_C(0x9bdc06a725c71235), UINT64_C(0xc19bf174cf692694), 325 UINT64_C(0xe49b69c19ef14ad2), UINT64_C(0xefbe4786384f25e3), 326 UINT64_C(0x0fc19dc68b8cd5b5), UINT64_C(0x240ca1cc77ac9c65), 327 UINT64_C(0x2de92c6f592b0275), UINT64_C(0x4a7484aa6ea6e483), 328 UINT64_C(0x5cb0a9dcbd41fbd4), UINT64_C(0x76f988da831153b5), 329 UINT64_C(0x983e5152ee66dfab), UINT64_C(0xa831c66d2db43210), 330 UINT64_C(0xb00327c898fb213f), UINT64_C(0xbf597fc7beef0ee4), 331 UINT64_C(0xc6e00bf33da88fc2), UINT64_C(0xd5a79147930aa725), 332 UINT64_C(0x06ca6351e003826f), UINT64_C(0x142929670a0e6e70), 333 UINT64_C(0x27b70a8546d22ffc), UINT64_C(0x2e1b21385c26c926), 334 UINT64_C(0x4d2c6dfc5ac42aed), UINT64_C(0x53380d139d95b3df), 335 UINT64_C(0x650a73548baf63de), UINT64_C(0x766a0abb3c77b2a8), 336 UINT64_C(0x81c2c92e47edaee6), UINT64_C(0x92722c851482353b), 337 UINT64_C(0xa2bfe8a14cf10364), UINT64_C(0xa81a664bbc423001), 338 UINT64_C(0xc24b8b70d0f89791), UINT64_C(0xc76c51a30654be30), 339 UINT64_C(0xd192e819d6ef5218), UINT64_C(0xd69906245565a910), 340 UINT64_C(0xf40e35855771202a), UINT64_C(0x106aa07032bbd1b8), 341 UINT64_C(0x19a4c116b8d2d0c8), UINT64_C(0x1e376c085141ab53), 342 UINT64_C(0x2748774cdf8eeb99), UINT64_C(0x34b0bcb5e19b48a8), 343 UINT64_C(0x391c0cb3c5c95a63), UINT64_C(0x4ed8aa4ae3418acb), 344 UINT64_C(0x5b9cca4f7763e373), UINT64_C(0x682e6ff3d6b2b8a3), 345 UINT64_C(0x748f82ee5defb2fc), UINT64_C(0x78a5636f43172f60), 346 UINT64_C(0x84c87814a1f0ab72), UINT64_C(0x8cc702081a6439ec), 347 UINT64_C(0x90befffa23631e28), UINT64_C(0xa4506cebde82bde9), 348 UINT64_C(0xbef9a3f7b2c67915), UINT64_C(0xc67178f2e372532b), 349 UINT64_C(0xca273eceea26619c), UINT64_C(0xd186b8c721c0c207), 350 UINT64_C(0xeada7dd6cde0eb1e), UINT64_C(0xf57d4f7fee6ed178), 351 UINT64_C(0x06f067aa72176fba), UINT64_C(0x0a637dc5a2c898a6), 352 UINT64_C(0x113f9804bef90dae), UINT64_C(0x1b710b35131c471b), 353 UINT64_C(0x28db77f523047d84), UINT64_C(0x32caab7b40c72493), 354 UINT64_C(0x3c9ebe0a15c9bebc), UINT64_C(0x431d67c49c100d4c), 355 UINT64_C(0x4cc5d4becb3e42b6), UINT64_C(0x597f299cfc657e2a), 356 UINT64_C(0x5fcb6fab3ad6faec), UINT64_C(0x6c44198c4a475817), 357}; 358 359#if defined(__GNUC__) && __GNUC__ >= 2 && !defined(OPENSSL_NO_ASM) 360#if defined(__x86_64) || defined(__x86_64__) 361#define ROTR(a, n) \ 362 ({ \ 363 uint64_t ret; \ 364 __asm__("rorq %1, %0" : "=r"(ret) : "J"(n), "0"(a) : "cc"); \ 365 ret; \ 366 }) 367#define PULL64(x) \ 368 ({ \ 369 uint64_t ret = *((const uint64_t *)(&(x))); \ 370 __asm__("bswapq %0" : "=r"(ret) : "0"(ret)); \ 371 ret; \ 372 }) 373#elif(defined(__i386) || defined(__i386__)) 374#define PULL64(x) \ 375 ({ \ 376 const unsigned int *p = (const unsigned int *)(&(x)); \ 377 unsigned int hi = p[0], lo = p[1]; \ 378 __asm__("bswapl %0; bswapl %1;" : "=r"(lo), "=r"(hi) : "0"(lo), "1"(hi)); \ 379 ((uint64_t)hi) << 32 | lo; \ 380 }) 381#elif(defined(_ARCH_PPC) && defined(__64BIT__)) || defined(_ARCH_PPC64) 382#define ROTR(a, n) \ 383 ({ \ 384 uint64_t ret; \ 385 __asm__("rotrdi %0, %1, %2" : "=r"(ret) : "r"(a), "K"(n)); \ 386 ret; \ 387 }) 388#elif defined(__aarch64__) 389#define ROTR(a, n) \ 390 ({ \ 391 uint64_t ret; \ 392 __asm__("ror %0, %1, %2" : "=r"(ret) : "r"(a), "I"(n)); \ 393 ret; \ 394 }) 395#if defined(__BYTE_ORDER__) && defined(__ORDER_LITTLE_ENDIAN__) && \ 396 __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__ 397#define PULL64(x) \ 398 ({ \ 399 uint64_t ret; \ 400 __asm__("rev %0, %1" : "=r"(ret) : "r"(*((const uint64_t *)(&(x))))); \ 401 ret; \ 402 }) 403#endif 404#endif 405#elif defined(_MSC_VER) 406#if defined(_WIN64) /* applies to both IA-64 and AMD64 */ 407#pragma intrinsic(_rotr64) 408#define ROTR(a, n) _rotr64((a), n) 409#endif 410#if defined(_M_IX86) && !defined(OPENSSL_NO_ASM) 411static uint64_t __fastcall __pull64be(const void *x) { 412 _asm mov edx, [ecx + 0] 413 _asm mov eax, [ecx + 4] 414 _asm bswap edx 415 _asm bswap eax 416} 417#define PULL64(x) __pull64be(&(x)) 418#if _MSC_VER <= 1200 419#pragma inline_depth(0) 420#endif 421#endif 422#endif 423 424#ifndef PULL64 425#define B(x, j) \ 426 (((uint64_t)(*(((const uint8_t *)(&x)) + j))) << ((7 - j) * 8)) 427#define PULL64(x) \ 428 (B(x, 0) | B(x, 1) | B(x, 2) | B(x, 3) | B(x, 4) | B(x, 5) | B(x, 6) | \ 429 B(x, 7)) 430#endif 431 432#ifndef ROTR 433#define ROTR(x, s) (((x) >> s) | (x) << (64 - s)) 434#endif 435 436#define Sigma0(x) (ROTR((x), 28) ^ ROTR((x), 34) ^ ROTR((x), 39)) 437#define Sigma1(x) (ROTR((x), 14) ^ ROTR((x), 18) ^ ROTR((x), 41)) 438#define sigma0(x) (ROTR((x), 1) ^ ROTR((x), 8) ^ ((x) >> 7)) 439#define sigma1(x) (ROTR((x), 19) ^ ROTR((x), 61) ^ ((x) >> 6)) 440 441#define Ch(x, y, z) (((x) & (y)) ^ ((~(x)) & (z))) 442#define Maj(x, y, z) (((x) & (y)) ^ ((x) & (z)) ^ ((y) & (z))) 443 444 445#if defined(__i386) || defined(__i386__) || defined(_M_IX86) 446/* 447 * This code should give better results on 32-bit CPU with less than 448 * ~24 registers, both size and performance wise... 449 */ 450static void sha512_block_data_order(uint64_t *state, const uint64_t *W, 451 size_t num) { 452 uint64_t A, E, T; 453 uint64_t X[9 + 80], *F; 454 int i; 455 456 while (num--) { 457 F = X + 80; 458 A = state[0]; 459 F[1] = state[1]; 460 F[2] = state[2]; 461 F[3] = state[3]; 462 E = state[4]; 463 F[5] = state[5]; 464 F[6] = state[6]; 465 F[7] = state[7]; 466 467 for (i = 0; i < 16; i++, F--) { 468 T = PULL64(W[i]); 469 F[0] = A; 470 F[4] = E; 471 F[8] = T; 472 T += F[7] + Sigma1(E) + Ch(E, F[5], F[6]) + K512[i]; 473 E = F[3] + T; 474 A = T + Sigma0(A) + Maj(A, F[1], F[2]); 475 } 476 477 for (; i < 80; i++, F--) { 478 T = sigma0(F[8 + 16 - 1]); 479 T += sigma1(F[8 + 16 - 14]); 480 T += F[8 + 16] + F[8 + 16 - 9]; 481 482 F[0] = A; 483 F[4] = E; 484 F[8] = T; 485 T += F[7] + Sigma1(E) + Ch(E, F[5], F[6]) + K512[i]; 486 E = F[3] + T; 487 A = T + Sigma0(A) + Maj(A, F[1], F[2]); 488 } 489 490 state[0] += A; 491 state[1] += F[1]; 492 state[2] += F[2]; 493 state[3] += F[3]; 494 state[4] += E; 495 state[5] += F[5]; 496 state[6] += F[6]; 497 state[7] += F[7]; 498 499 W += 16; 500 } 501} 502 503#else 504 505#define ROUND_00_15(i, a, b, c, d, e, f, g, h) \ 506 do { \ 507 T1 += h + Sigma1(e) + Ch(e, f, g) + K512[i]; \ 508 h = Sigma0(a) + Maj(a, b, c); \ 509 d += T1; \ 510 h += T1; \ 511 } while (0) 512 513#define ROUND_16_80(i, j, a, b, c, d, e, f, g, h, X) \ 514 do { \ 515 s0 = X[(j + 1) & 0x0f]; \ 516 s0 = sigma0(s0); \ 517 s1 = X[(j + 14) & 0x0f]; \ 518 s1 = sigma1(s1); \ 519 T1 = X[(j) & 0x0f] += s0 + s1 + X[(j + 9) & 0x0f]; \ 520 ROUND_00_15(i + j, a, b, c, d, e, f, g, h); \ 521 } while (0) 522 523static void sha512_block_data_order(uint64_t *state, const uint64_t *W, 524 size_t num) { 525 uint64_t a, b, c, d, e, f, g, h, s0, s1, T1; 526 uint64_t X[16]; 527 int i; 528 529 while (num--) { 530 531 a = state[0]; 532 b = state[1]; 533 c = state[2]; 534 d = state[3]; 535 e = state[4]; 536 f = state[5]; 537 g = state[6]; 538 h = state[7]; 539 540 T1 = X[0] = PULL64(W[0]); 541 ROUND_00_15(0, a, b, c, d, e, f, g, h); 542 T1 = X[1] = PULL64(W[1]); 543 ROUND_00_15(1, h, a, b, c, d, e, f, g); 544 T1 = X[2] = PULL64(W[2]); 545 ROUND_00_15(2, g, h, a, b, c, d, e, f); 546 T1 = X[3] = PULL64(W[3]); 547 ROUND_00_15(3, f, g, h, a, b, c, d, e); 548 T1 = X[4] = PULL64(W[4]); 549 ROUND_00_15(4, e, f, g, h, a, b, c, d); 550 T1 = X[5] = PULL64(W[5]); 551 ROUND_00_15(5, d, e, f, g, h, a, b, c); 552 T1 = X[6] = PULL64(W[6]); 553 ROUND_00_15(6, c, d, e, f, g, h, a, b); 554 T1 = X[7] = PULL64(W[7]); 555 ROUND_00_15(7, b, c, d, e, f, g, h, a); 556 T1 = X[8] = PULL64(W[8]); 557 ROUND_00_15(8, a, b, c, d, e, f, g, h); 558 T1 = X[9] = PULL64(W[9]); 559 ROUND_00_15(9, h, a, b, c, d, e, f, g); 560 T1 = X[10] = PULL64(W[10]); 561 ROUND_00_15(10, g, h, a, b, c, d, e, f); 562 T1 = X[11] = PULL64(W[11]); 563 ROUND_00_15(11, f, g, h, a, b, c, d, e); 564 T1 = X[12] = PULL64(W[12]); 565 ROUND_00_15(12, e, f, g, h, a, b, c, d); 566 T1 = X[13] = PULL64(W[13]); 567 ROUND_00_15(13, d, e, f, g, h, a, b, c); 568 T1 = X[14] = PULL64(W[14]); 569 ROUND_00_15(14, c, d, e, f, g, h, a, b); 570 T1 = X[15] = PULL64(W[15]); 571 ROUND_00_15(15, b, c, d, e, f, g, h, a); 572 573 for (i = 16; i < 80; i += 16) { 574 ROUND_16_80(i, 0, a, b, c, d, e, f, g, h, X); 575 ROUND_16_80(i, 1, h, a, b, c, d, e, f, g, X); 576 ROUND_16_80(i, 2, g, h, a, b, c, d, e, f, X); 577 ROUND_16_80(i, 3, f, g, h, a, b, c, d, e, X); 578 ROUND_16_80(i, 4, e, f, g, h, a, b, c, d, X); 579 ROUND_16_80(i, 5, d, e, f, g, h, a, b, c, X); 580 ROUND_16_80(i, 6, c, d, e, f, g, h, a, b, X); 581 ROUND_16_80(i, 7, b, c, d, e, f, g, h, a, X); 582 ROUND_16_80(i, 8, a, b, c, d, e, f, g, h, X); 583 ROUND_16_80(i, 9, h, a, b, c, d, e, f, g, X); 584 ROUND_16_80(i, 10, g, h, a, b, c, d, e, f, X); 585 ROUND_16_80(i, 11, f, g, h, a, b, c, d, e, X); 586 ROUND_16_80(i, 12, e, f, g, h, a, b, c, d, X); 587 ROUND_16_80(i, 13, d, e, f, g, h, a, b, c, X); 588 ROUND_16_80(i, 14, c, d, e, f, g, h, a, b, X); 589 ROUND_16_80(i, 15, b, c, d, e, f, g, h, a, X); 590 } 591 592 state[0] += a; 593 state[1] += b; 594 state[2] += c; 595 state[3] += d; 596 state[4] += e; 597 state[5] += f; 598 state[6] += g; 599 state[7] += h; 600 601 W += 16; 602 } 603} 604 605#endif 606 607#endif /* SHA512_ASM */ 608