Home | History | Annotate | Download | only in sha
      1 /* Copyright (C) 1995-1998 Eric Young (eay (at) cryptsoft.com)
      2  * All rights reserved.
      3  *
      4  * This package is an SSL implementation written
      5  * by Eric Young (eay (at) 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 (at) 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 (at) 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 (at) 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 #include "internal.h"
     64 #include "../../internal.h"
     65 
     66 
     67 // The 32-bit hash algorithms share a common byte-order neutral collector and
     68 // padding function implementations that operate on unaligned data,
     69 // ../digest/md32_common.h. SHA-512 is the only 64-bit hash algorithm, as of
     70 // this writing, so there is no need for a common collector/padding
     71 // implementation yet.
     72 
     73 int SHA384_Init(SHA512_CTX *sha) {
     74   sha->h[0] = UINT64_C(0xcbbb9d5dc1059ed8);
     75   sha->h[1] = UINT64_C(0x629a292a367cd507);
     76   sha->h[2] = UINT64_C(0x9159015a3070dd17);
     77   sha->h[3] = UINT64_C(0x152fecd8f70e5939);
     78   sha->h[4] = UINT64_C(0x67332667ffc00b31);
     79   sha->h[5] = UINT64_C(0x8eb44a8768581511);
     80   sha->h[6] = UINT64_C(0xdb0c2e0d64f98fa7);
     81   sha->h[7] = UINT64_C(0x47b5481dbefa4fa4);
     82 
     83   sha->Nl = 0;
     84   sha->Nh = 0;
     85   sha->num = 0;
     86   sha->md_len = SHA384_DIGEST_LENGTH;
     87   return 1;
     88 }
     89 
     90 
     91 int SHA512_Init(SHA512_CTX *sha) {
     92   sha->h[0] = UINT64_C(0x6a09e667f3bcc908);
     93   sha->h[1] = UINT64_C(0xbb67ae8584caa73b);
     94   sha->h[2] = UINT64_C(0x3c6ef372fe94f82b);
     95   sha->h[3] = UINT64_C(0xa54ff53a5f1d36f1);
     96   sha->h[4] = UINT64_C(0x510e527fade682d1);
     97   sha->h[5] = UINT64_C(0x9b05688c2b3e6c1f);
     98   sha->h[6] = UINT64_C(0x1f83d9abfb41bd6b);
     99   sha->h[7] = UINT64_C(0x5be0cd19137e2179);
    100 
    101   sha->Nl = 0;
    102   sha->Nh = 0;
    103   sha->num = 0;
    104   sha->md_len = SHA512_DIGEST_LENGTH;
    105   return 1;
    106 }
    107 
    108 uint8_t *SHA384(const uint8_t *data, size_t len, uint8_t *out) {
    109   SHA512_CTX ctx;
    110   SHA384_Init(&ctx);
    111   SHA384_Update(&ctx, data, len);
    112   SHA384_Final(out, &ctx);
    113   OPENSSL_cleanse(&ctx, sizeof(ctx));
    114   return out;
    115 }
    116 
    117 uint8_t *SHA512(const uint8_t *data, size_t len, uint8_t *out) {
    118   SHA512_CTX ctx;
    119   SHA512_Init(&ctx);
    120   SHA512_Update(&ctx, data, len);
    121   SHA512_Final(out, &ctx);
    122   OPENSSL_cleanse(&ctx, sizeof(ctx));
    123   return out;
    124 }
    125 
    126 #if !defined(SHA512_ASM)
    127 static void sha512_block_data_order(uint64_t *state, const uint8_t *in,
    128                                     size_t num_blocks);
    129 #endif
    130 
    131 
    132 int SHA384_Final(uint8_t *md, SHA512_CTX *sha) {
    133   return SHA512_Final(md, sha);
    134 }
    135 
    136 int SHA384_Update(SHA512_CTX *sha, const void *data, size_t len) {
    137   return SHA512_Update(sha, data, len);
    138 }
    139 
    140 void SHA512_Transform(SHA512_CTX *c, const uint8_t *block) {
    141   sha512_block_data_order(c->h, block, 1);
    142 }
    143 
    144 int SHA512_Update(SHA512_CTX *c, const void *in_data, size_t len) {
    145   uint64_t l;
    146   uint8_t *p = c->p;
    147   const uint8_t *data = in_data;
    148 
    149   if (len == 0) {
    150     return 1;
    151   }
    152 
    153   l = (c->Nl + (((uint64_t)len) << 3)) & UINT64_C(0xffffffffffffffff);
    154   if (l < c->Nl) {
    155     c->Nh++;
    156   }
    157   if (sizeof(len) >= 8) {
    158     c->Nh += (((uint64_t)len) >> 61);
    159   }
    160   c->Nl = l;
    161 
    162   if (c->num != 0) {
    163     size_t n = sizeof(c->p) - c->num;
    164 
    165     if (len < n) {
    166       OPENSSL_memcpy(p + c->num, data, len);
    167       c->num += (unsigned int)len;
    168       return 1;
    169     } else {
    170       OPENSSL_memcpy(p + c->num, data, n), c->num = 0;
    171       len -= n;
    172       data += n;
    173       sha512_block_data_order(c->h, p, 1);
    174     }
    175   }
    176 
    177   if (len >= sizeof(c->p)) {
    178     sha512_block_data_order(c->h, data, len / sizeof(c->p));
    179     data += len;
    180     len %= sizeof(c->p);
    181     data -= len;
    182   }
    183 
    184   if (len != 0) {
    185     OPENSSL_memcpy(p, data, len);
    186     c->num = (int)len;
    187   }
    188 
    189   return 1;
    190 }
    191 
    192 int SHA512_Final(uint8_t *md, SHA512_CTX *sha) {
    193   uint8_t *p = sha->p;
    194   size_t n = sha->num;
    195 
    196   p[n] = 0x80;  // There always is a room for one
    197   n++;
    198   if (n > (sizeof(sha->p) - 16)) {
    199     OPENSSL_memset(p + n, 0, sizeof(sha->p) - n);
    200     n = 0;
    201     sha512_block_data_order(sha->h, p, 1);
    202   }
    203 
    204   OPENSSL_memset(p + n, 0, sizeof(sha->p) - 16 - n);
    205   p[sizeof(sha->p) - 1] = (uint8_t)(sha->Nl);
    206   p[sizeof(sha->p) - 2] = (uint8_t)(sha->Nl >> 8);
    207   p[sizeof(sha->p) - 3] = (uint8_t)(sha->Nl >> 16);
    208   p[sizeof(sha->p) - 4] = (uint8_t)(sha->Nl >> 24);
    209   p[sizeof(sha->p) - 5] = (uint8_t)(sha->Nl >> 32);
    210   p[sizeof(sha->p) - 6] = (uint8_t)(sha->Nl >> 40);
    211   p[sizeof(sha->p) - 7] = (uint8_t)(sha->Nl >> 48);
    212   p[sizeof(sha->p) - 8] = (uint8_t)(sha->Nl >> 56);
    213   p[sizeof(sha->p) - 9] = (uint8_t)(sha->Nh);
    214   p[sizeof(sha->p) - 10] = (uint8_t)(sha->Nh >> 8);
    215   p[sizeof(sha->p) - 11] = (uint8_t)(sha->Nh >> 16);
    216   p[sizeof(sha->p) - 12] = (uint8_t)(sha->Nh >> 24);
    217   p[sizeof(sha->p) - 13] = (uint8_t)(sha->Nh >> 32);
    218   p[sizeof(sha->p) - 14] = (uint8_t)(sha->Nh >> 40);
    219   p[sizeof(sha->p) - 15] = (uint8_t)(sha->Nh >> 48);
    220   p[sizeof(sha->p) - 16] = (uint8_t)(sha->Nh >> 56);
    221 
    222   sha512_block_data_order(sha->h, p, 1);
    223 
    224   if (md == NULL) {
    225     // TODO(davidben): This NULL check is absent in other low-level hash 'final'
    226     // functions and is one of the few places one can fail.
    227     return 0;
    228   }
    229 
    230   switch (sha->md_len) {
    231     // Let compiler decide if it's appropriate to unroll...
    232     case SHA384_DIGEST_LENGTH:
    233       for (n = 0; n < SHA384_DIGEST_LENGTH / 8; n++) {
    234         uint64_t t = sha->h[n];
    235 
    236         *(md++) = (uint8_t)(t >> 56);
    237         *(md++) = (uint8_t)(t >> 48);
    238         *(md++) = (uint8_t)(t >> 40);
    239         *(md++) = (uint8_t)(t >> 32);
    240         *(md++) = (uint8_t)(t >> 24);
    241         *(md++) = (uint8_t)(t >> 16);
    242         *(md++) = (uint8_t)(t >> 8);
    243         *(md++) = (uint8_t)(t);
    244       }
    245       break;
    246     case SHA512_DIGEST_LENGTH:
    247       for (n = 0; n < SHA512_DIGEST_LENGTH / 8; n++) {
    248         uint64_t t = sha->h[n];
    249 
    250         *(md++) = (uint8_t)(t >> 56);
    251         *(md++) = (uint8_t)(t >> 48);
    252         *(md++) = (uint8_t)(t >> 40);
    253         *(md++) = (uint8_t)(t >> 32);
    254         *(md++) = (uint8_t)(t >> 24);
    255         *(md++) = (uint8_t)(t >> 16);
    256         *(md++) = (uint8_t)(t >> 8);
    257         *(md++) = (uint8_t)(t);
    258       }
    259       break;
    260     // ... as well as make sure md_len is not abused.
    261     default:
    262       // TODO(davidben): This bad |md_len| case is one of the few places a
    263       // low-level hash 'final' function can fail. This should never happen.
    264       return 0;
    265   }
    266 
    267   return 1;
    268 }
    269 
    270 #ifndef SHA512_ASM
    271 static const uint64_t K512[80] = {
    272     UINT64_C(0x428a2f98d728ae22), UINT64_C(0x7137449123ef65cd),
    273     UINT64_C(0xb5c0fbcfec4d3b2f), UINT64_C(0xe9b5dba58189dbbc),
    274     UINT64_C(0x3956c25bf348b538), UINT64_C(0x59f111f1b605d019),
    275     UINT64_C(0x923f82a4af194f9b), UINT64_C(0xab1c5ed5da6d8118),
    276     UINT64_C(0xd807aa98a3030242), UINT64_C(0x12835b0145706fbe),
    277     UINT64_C(0x243185be4ee4b28c), UINT64_C(0x550c7dc3d5ffb4e2),
    278     UINT64_C(0x72be5d74f27b896f), UINT64_C(0x80deb1fe3b1696b1),
    279     UINT64_C(0x9bdc06a725c71235), UINT64_C(0xc19bf174cf692694),
    280     UINT64_C(0xe49b69c19ef14ad2), UINT64_C(0xefbe4786384f25e3),
    281     UINT64_C(0x0fc19dc68b8cd5b5), UINT64_C(0x240ca1cc77ac9c65),
    282     UINT64_C(0x2de92c6f592b0275), UINT64_C(0x4a7484aa6ea6e483),
    283     UINT64_C(0x5cb0a9dcbd41fbd4), UINT64_C(0x76f988da831153b5),
    284     UINT64_C(0x983e5152ee66dfab), UINT64_C(0xa831c66d2db43210),
    285     UINT64_C(0xb00327c898fb213f), UINT64_C(0xbf597fc7beef0ee4),
    286     UINT64_C(0xc6e00bf33da88fc2), UINT64_C(0xd5a79147930aa725),
    287     UINT64_C(0x06ca6351e003826f), UINT64_C(0x142929670a0e6e70),
    288     UINT64_C(0x27b70a8546d22ffc), UINT64_C(0x2e1b21385c26c926),
    289     UINT64_C(0x4d2c6dfc5ac42aed), UINT64_C(0x53380d139d95b3df),
    290     UINT64_C(0x650a73548baf63de), UINT64_C(0x766a0abb3c77b2a8),
    291     UINT64_C(0x81c2c92e47edaee6), UINT64_C(0x92722c851482353b),
    292     UINT64_C(0xa2bfe8a14cf10364), UINT64_C(0xa81a664bbc423001),
    293     UINT64_C(0xc24b8b70d0f89791), UINT64_C(0xc76c51a30654be30),
    294     UINT64_C(0xd192e819d6ef5218), UINT64_C(0xd69906245565a910),
    295     UINT64_C(0xf40e35855771202a), UINT64_C(0x106aa07032bbd1b8),
    296     UINT64_C(0x19a4c116b8d2d0c8), UINT64_C(0x1e376c085141ab53),
    297     UINT64_C(0x2748774cdf8eeb99), UINT64_C(0x34b0bcb5e19b48a8),
    298     UINT64_C(0x391c0cb3c5c95a63), UINT64_C(0x4ed8aa4ae3418acb),
    299     UINT64_C(0x5b9cca4f7763e373), UINT64_C(0x682e6ff3d6b2b8a3),
    300     UINT64_C(0x748f82ee5defb2fc), UINT64_C(0x78a5636f43172f60),
    301     UINT64_C(0x84c87814a1f0ab72), UINT64_C(0x8cc702081a6439ec),
    302     UINT64_C(0x90befffa23631e28), UINT64_C(0xa4506cebde82bde9),
    303     UINT64_C(0xbef9a3f7b2c67915), UINT64_C(0xc67178f2e372532b),
    304     UINT64_C(0xca273eceea26619c), UINT64_C(0xd186b8c721c0c207),
    305     UINT64_C(0xeada7dd6cde0eb1e), UINT64_C(0xf57d4f7fee6ed178),
    306     UINT64_C(0x06f067aa72176fba), UINT64_C(0x0a637dc5a2c898a6),
    307     UINT64_C(0x113f9804bef90dae), UINT64_C(0x1b710b35131c471b),
    308     UINT64_C(0x28db77f523047d84), UINT64_C(0x32caab7b40c72493),
    309     UINT64_C(0x3c9ebe0a15c9bebc), UINT64_C(0x431d67c49c100d4c),
    310     UINT64_C(0x4cc5d4becb3e42b6), UINT64_C(0x597f299cfc657e2a),
    311     UINT64_C(0x5fcb6fab3ad6faec), UINT64_C(0x6c44198c4a475817),
    312 };
    313 
    314 #if defined(__GNUC__) && __GNUC__ >= 2 && !defined(OPENSSL_NO_ASM)
    315 #if defined(__x86_64) || defined(__x86_64__)
    316 #define ROTR(a, n)                                              \
    317   ({                                                            \
    318     uint64_t ret;                                               \
    319     __asm__("rorq %1, %0" : "=r"(ret) : "J"(n), "0"(a) : "cc"); \
    320     ret;                                                        \
    321   })
    322 #elif(defined(_ARCH_PPC) && defined(__64BIT__)) || defined(_ARCH_PPC64)
    323 #define ROTR(a, n)                                             \
    324   ({                                                           \
    325     uint64_t ret;                                              \
    326     __asm__("rotrdi %0, %1, %2" : "=r"(ret) : "r"(a), "K"(n)); \
    327     ret;                                                       \
    328   })
    329 #elif defined(__aarch64__)
    330 #define ROTR(a, n)                                          \
    331   ({                                                        \
    332     uint64_t ret;                                           \
    333     __asm__("ror %0, %1, %2" : "=r"(ret) : "r"(a), "I"(n)); \
    334     ret;                                                    \
    335   })
    336 #endif
    337 #elif defined(_MSC_VER) && defined(_WIN64)
    338 #pragma intrinsic(_rotr64)
    339 #define ROTR(a, n) _rotr64((a), n)
    340 #endif
    341 
    342 #ifndef ROTR
    343 #define ROTR(x, s) (((x) >> s) | (x) << (64 - s))
    344 #endif
    345 
    346 static inline uint64_t load_u64_be(const void *ptr) {
    347   uint64_t ret;
    348   OPENSSL_memcpy(&ret, ptr, sizeof(ret));
    349   return CRYPTO_bswap8(ret);
    350 }
    351 
    352 #define Sigma0(x) (ROTR((x), 28) ^ ROTR((x), 34) ^ ROTR((x), 39))
    353 #define Sigma1(x) (ROTR((x), 14) ^ ROTR((x), 18) ^ ROTR((x), 41))
    354 #define sigma0(x) (ROTR((x), 1) ^ ROTR((x), 8) ^ ((x) >> 7))
    355 #define sigma1(x) (ROTR((x), 19) ^ ROTR((x), 61) ^ ((x) >> 6))
    356 
    357 #define Ch(x, y, z) (((x) & (y)) ^ ((~(x)) & (z)))
    358 #define Maj(x, y, z) (((x) & (y)) ^ ((x) & (z)) ^ ((y) & (z)))
    359 
    360 
    361 #if defined(__i386) || defined(__i386__) || defined(_M_IX86)
    362 // This code should give better results on 32-bit CPU with less than
    363 // ~24 registers, both size and performance wise...
    364 static void sha512_block_data_order(uint64_t *state, const uint8_t *in,
    365                                     size_t num) {
    366   uint64_t A, E, T;
    367   uint64_t X[9 + 80], *F;
    368   int i;
    369 
    370   while (num--) {
    371     F = X + 80;
    372     A = state[0];
    373     F[1] = state[1];
    374     F[2] = state[2];
    375     F[3] = state[3];
    376     E = state[4];
    377     F[5] = state[5];
    378     F[6] = state[6];
    379     F[7] = state[7];
    380 
    381     for (i = 0; i < 16; i++, F--) {
    382       T = load_u64_be(in + i * 8);
    383       F[0] = A;
    384       F[4] = E;
    385       F[8] = T;
    386       T += F[7] + Sigma1(E) + Ch(E, F[5], F[6]) + K512[i];
    387       E = F[3] + T;
    388       A = T + Sigma0(A) + Maj(A, F[1], F[2]);
    389     }
    390 
    391     for (; i < 80; i++, F--) {
    392       T = sigma0(F[8 + 16 - 1]);
    393       T += sigma1(F[8 + 16 - 14]);
    394       T += F[8 + 16] + F[8 + 16 - 9];
    395 
    396       F[0] = A;
    397       F[4] = E;
    398       F[8] = T;
    399       T += F[7] + Sigma1(E) + Ch(E, F[5], F[6]) + K512[i];
    400       E = F[3] + T;
    401       A = T + Sigma0(A) + Maj(A, F[1], F[2]);
    402     }
    403 
    404     state[0] += A;
    405     state[1] += F[1];
    406     state[2] += F[2];
    407     state[3] += F[3];
    408     state[4] += E;
    409     state[5] += F[5];
    410     state[6] += F[6];
    411     state[7] += F[7];
    412 
    413     in += 16 * 8;
    414   }
    415 }
    416 
    417 #else
    418 
    419 #define ROUND_00_15(i, a, b, c, d, e, f, g, h)   \
    420   do {                                           \
    421     T1 += h + Sigma1(e) + Ch(e, f, g) + K512[i]; \
    422     h = Sigma0(a) + Maj(a, b, c);                \
    423     d += T1;                                     \
    424     h += T1;                                     \
    425   } while (0)
    426 
    427 #define ROUND_16_80(i, j, a, b, c, d, e, f, g, h, X)   \
    428   do {                                                 \
    429     s0 = X[(j + 1) & 0x0f];                            \
    430     s0 = sigma0(s0);                                   \
    431     s1 = X[(j + 14) & 0x0f];                           \
    432     s1 = sigma1(s1);                                   \
    433     T1 = X[(j) & 0x0f] += s0 + s1 + X[(j + 9) & 0x0f]; \
    434     ROUND_00_15(i + j, a, b, c, d, e, f, g, h);        \
    435   } while (0)
    436 
    437 static void sha512_block_data_order(uint64_t *state, const uint8_t *in,
    438                                     size_t num) {
    439   uint64_t a, b, c, d, e, f, g, h, s0, s1, T1;
    440   uint64_t X[16];
    441   int i;
    442 
    443   while (num--) {
    444 
    445     a = state[0];
    446     b = state[1];
    447     c = state[2];
    448     d = state[3];
    449     e = state[4];
    450     f = state[5];
    451     g = state[6];
    452     h = state[7];
    453 
    454     T1 = X[0] = load_u64_be(in);
    455     ROUND_00_15(0, a, b, c, d, e, f, g, h);
    456     T1 = X[1] = load_u64_be(in + 8);
    457     ROUND_00_15(1, h, a, b, c, d, e, f, g);
    458     T1 = X[2] = load_u64_be(in + 2 * 8);
    459     ROUND_00_15(2, g, h, a, b, c, d, e, f);
    460     T1 = X[3] = load_u64_be(in + 3 * 8);
    461     ROUND_00_15(3, f, g, h, a, b, c, d, e);
    462     T1 = X[4] = load_u64_be(in + 4 * 8);
    463     ROUND_00_15(4, e, f, g, h, a, b, c, d);
    464     T1 = X[5] = load_u64_be(in + 5 * 8);
    465     ROUND_00_15(5, d, e, f, g, h, a, b, c);
    466     T1 = X[6] = load_u64_be(in + 6 * 8);
    467     ROUND_00_15(6, c, d, e, f, g, h, a, b);
    468     T1 = X[7] = load_u64_be(in + 7 * 8);
    469     ROUND_00_15(7, b, c, d, e, f, g, h, a);
    470     T1 = X[8] = load_u64_be(in + 8 * 8);
    471     ROUND_00_15(8, a, b, c, d, e, f, g, h);
    472     T1 = X[9] = load_u64_be(in + 9 * 8);
    473     ROUND_00_15(9, h, a, b, c, d, e, f, g);
    474     T1 = X[10] = load_u64_be(in + 10 * 8);
    475     ROUND_00_15(10, g, h, a, b, c, d, e, f);
    476     T1 = X[11] = load_u64_be(in + 11 * 8);
    477     ROUND_00_15(11, f, g, h, a, b, c, d, e);
    478     T1 = X[12] = load_u64_be(in + 12 * 8);
    479     ROUND_00_15(12, e, f, g, h, a, b, c, d);
    480     T1 = X[13] = load_u64_be(in + 13 * 8);
    481     ROUND_00_15(13, d, e, f, g, h, a, b, c);
    482     T1 = X[14] = load_u64_be(in + 14 * 8);
    483     ROUND_00_15(14, c, d, e, f, g, h, a, b);
    484     T1 = X[15] = load_u64_be(in + 15 * 8);
    485     ROUND_00_15(15, b, c, d, e, f, g, h, a);
    486 
    487     for (i = 16; i < 80; i += 16) {
    488       ROUND_16_80(i, 0, a, b, c, d, e, f, g, h, X);
    489       ROUND_16_80(i, 1, h, a, b, c, d, e, f, g, X);
    490       ROUND_16_80(i, 2, g, h, a, b, c, d, e, f, X);
    491       ROUND_16_80(i, 3, f, g, h, a, b, c, d, e, X);
    492       ROUND_16_80(i, 4, e, f, g, h, a, b, c, d, X);
    493       ROUND_16_80(i, 5, d, e, f, g, h, a, b, c, X);
    494       ROUND_16_80(i, 6, c, d, e, f, g, h, a, b, X);
    495       ROUND_16_80(i, 7, b, c, d, e, f, g, h, a, X);
    496       ROUND_16_80(i, 8, a, b, c, d, e, f, g, h, X);
    497       ROUND_16_80(i, 9, h, a, b, c, d, e, f, g, X);
    498       ROUND_16_80(i, 10, g, h, a, b, c, d, e, f, X);
    499       ROUND_16_80(i, 11, f, g, h, a, b, c, d, e, X);
    500       ROUND_16_80(i, 12, e, f, g, h, a, b, c, d, X);
    501       ROUND_16_80(i, 13, d, e, f, g, h, a, b, c, X);
    502       ROUND_16_80(i, 14, c, d, e, f, g, h, a, b, X);
    503       ROUND_16_80(i, 15, b, c, d, e, f, g, h, a, X);
    504     }
    505 
    506     state[0] += a;
    507     state[1] += b;
    508     state[2] += c;
    509     state[3] += d;
    510     state[4] += e;
    511     state[5] += f;
    512     state[6] += g;
    513     state[7] += h;
    514 
    515     in += 16 * 8;
    516   }
    517 }
    518 
    519 #endif
    520 
    521 #endif  // !SHA512_ASM
    522 
    523 #undef ROTR
    524 #undef Sigma0
    525 #undef Sigma1
    526 #undef sigma0
    527 #undef sigma1
    528 #undef Ch
    529 #undef Maj
    530 #undef ROUND_00_15
    531 #undef ROUND_16_80
    532