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 #include "../../internal.h"
64
65
66 // IMPLEMENTATION NOTES.
67 //
68 // The 32-bit hash algorithms share a common byte-order neutral collector and
69 // padding function implementations that operate on unaligned data,
70 // ../md32_common.h. This SHA-512 implementation does not. Reasons
71 // [in reverse order] are:
72 //
73 // - It's the only 64-bit hash algorithm for the moment of this writing,
74 // there is no need for common collector/padding implementation [yet];
75 // - By supporting only a transform function that operates on *aligned* data
76 // the collector/padding function is simpler and easier to optimize.
77
78 #if !defined(OPENSSL_NO_ASM) && \
79 (defined(OPENSSL_X86) || defined(OPENSSL_X86_64) || \
80 defined(OPENSSL_ARM) || defined(OPENSSL_AARCH64))
81 #define SHA512_ASM
82 #endif
83
84 #if defined(OPENSSL_X86) || defined(OPENSSL_X86_64) || \
85 defined(__ARM_FEATURE_UNALIGNED)
86 #define SHA512_BLOCK_CAN_MANAGE_UNALIGNED_DATA
87 #endif
88
SHA384_Init(SHA512_CTX * sha)89 int SHA384_Init(SHA512_CTX *sha) {
90 sha->h[0] = UINT64_C(0xcbbb9d5dc1059ed8);
91 sha->h[1] = UINT64_C(0x629a292a367cd507);
92 sha->h[2] = UINT64_C(0x9159015a3070dd17);
93 sha->h[3] = UINT64_C(0x152fecd8f70e5939);
94 sha->h[4] = UINT64_C(0x67332667ffc00b31);
95 sha->h[5] = UINT64_C(0x8eb44a8768581511);
96 sha->h[6] = UINT64_C(0xdb0c2e0d64f98fa7);
97 sha->h[7] = UINT64_C(0x47b5481dbefa4fa4);
98
99 sha->Nl = 0;
100 sha->Nh = 0;
101 sha->num = 0;
102 sha->md_len = SHA384_DIGEST_LENGTH;
103 return 1;
104 }
105
106
SHA512_Init(SHA512_CTX * sha)107 int SHA512_Init(SHA512_CTX *sha) {
108 sha->h[0] = UINT64_C(0x6a09e667f3bcc908);
109 sha->h[1] = UINT64_C(0xbb67ae8584caa73b);
110 sha->h[2] = UINT64_C(0x3c6ef372fe94f82b);
111 sha->h[3] = UINT64_C(0xa54ff53a5f1d36f1);
112 sha->h[4] = UINT64_C(0x510e527fade682d1);
113 sha->h[5] = UINT64_C(0x9b05688c2b3e6c1f);
114 sha->h[6] = UINT64_C(0x1f83d9abfb41bd6b);
115 sha->h[7] = UINT64_C(0x5be0cd19137e2179);
116
117 sha->Nl = 0;
118 sha->Nh = 0;
119 sha->num = 0;
120 sha->md_len = SHA512_DIGEST_LENGTH;
121 return 1;
122 }
123
SHA384(const uint8_t * data,size_t len,uint8_t * out)124 uint8_t *SHA384(const uint8_t *data, size_t len, uint8_t *out) {
125 SHA512_CTX ctx;
126 SHA384_Init(&ctx);
127 SHA384_Update(&ctx, data, len);
128 SHA384_Final(out, &ctx);
129 OPENSSL_cleanse(&ctx, sizeof(ctx));
130 return out;
131 }
132
SHA512(const uint8_t * data,size_t len,uint8_t * out)133 uint8_t *SHA512(const uint8_t *data, size_t len, uint8_t *out) {
134 SHA512_CTX ctx;
135 SHA512_Init(&ctx);
136 SHA512_Update(&ctx, data, len);
137 SHA512_Final(out, &ctx);
138 OPENSSL_cleanse(&ctx, sizeof(ctx));
139 return out;
140 }
141
142 #if !defined(SHA512_ASM)
143 static
144 #endif
145 void sha512_block_data_order(uint64_t *state, const uint64_t *W, size_t num);
146
147
SHA384_Final(uint8_t * md,SHA512_CTX * sha)148 int SHA384_Final(uint8_t *md, SHA512_CTX *sha) {
149 return SHA512_Final(md, sha);
150 }
151
SHA384_Update(SHA512_CTX * sha,const void * data,size_t len)152 int SHA384_Update(SHA512_CTX *sha, const void *data, size_t len) {
153 return SHA512_Update(sha, data, len);
154 }
155
SHA512_Transform(SHA512_CTX * c,const uint8_t * block)156 void SHA512_Transform(SHA512_CTX *c, const uint8_t *block) {
157 #ifndef SHA512_BLOCK_CAN_MANAGE_UNALIGNED_DATA
158 if ((size_t)block % sizeof(c->u.d[0]) != 0) {
159 OPENSSL_memcpy(c->u.p, block, sizeof(c->u.p));
160 block = c->u.p;
161 }
162 #endif
163 sha512_block_data_order(c->h, (uint64_t *)block, 1);
164 }
165
SHA512_Update(SHA512_CTX * c,const void * in_data,size_t len)166 int SHA512_Update(SHA512_CTX *c, const void *in_data, size_t len) {
167 uint64_t l;
168 uint8_t *p = c->u.p;
169 const uint8_t *data = (const uint8_t *)in_data;
170
171 if (len == 0) {
172 return 1;
173 }
174
175 l = (c->Nl + (((uint64_t)len) << 3)) & UINT64_C(0xffffffffffffffff);
176 if (l < c->Nl) {
177 c->Nh++;
178 }
179 if (sizeof(len) >= 8) {
180 c->Nh += (((uint64_t)len) >> 61);
181 }
182 c->Nl = l;
183
184 if (c->num != 0) {
185 size_t n = sizeof(c->u) - c->num;
186
187 if (len < n) {
188 OPENSSL_memcpy(p + c->num, data, len);
189 c->num += (unsigned int)len;
190 return 1;
191 } else {
192 OPENSSL_memcpy(p + c->num, data, n), c->num = 0;
193 len -= n;
194 data += n;
195 sha512_block_data_order(c->h, (uint64_t *)p, 1);
196 }
197 }
198
199 if (len >= sizeof(c->u)) {
200 #ifndef SHA512_BLOCK_CAN_MANAGE_UNALIGNED_DATA
201 if ((size_t)data % sizeof(c->u.d[0]) != 0) {
202 while (len >= sizeof(c->u)) {
203 OPENSSL_memcpy(p, data, sizeof(c->u));
204 sha512_block_data_order(c->h, (uint64_t *)p, 1);
205 len -= sizeof(c->u);
206 data += sizeof(c->u);
207 }
208 } else
209 #endif
210 {
211 sha512_block_data_order(c->h, (uint64_t *)data, len / sizeof(c->u));
212 data += len;
213 len %= sizeof(c->u);
214 data -= len;
215 }
216 }
217
218 if (len != 0) {
219 OPENSSL_memcpy(p, data, len);
220 c->num = (int)len;
221 }
222
223 return 1;
224 }
225
SHA512_Final(uint8_t * md,SHA512_CTX * sha)226 int SHA512_Final(uint8_t *md, SHA512_CTX *sha) {
227 uint8_t *p = (uint8_t *)sha->u.p;
228 size_t n = sha->num;
229
230 p[n] = 0x80; // There always is a room for one
231 n++;
232 if (n > (sizeof(sha->u) - 16)) {
233 OPENSSL_memset(p + n, 0, sizeof(sha->u) - n);
234 n = 0;
235 sha512_block_data_order(sha->h, (uint64_t *)p, 1);
236 }
237
238 OPENSSL_memset(p + n, 0, sizeof(sha->u) - 16 - n);
239 p[sizeof(sha->u) - 1] = (uint8_t)(sha->Nl);
240 p[sizeof(sha->u) - 2] = (uint8_t)(sha->Nl >> 8);
241 p[sizeof(sha->u) - 3] = (uint8_t)(sha->Nl >> 16);
242 p[sizeof(sha->u) - 4] = (uint8_t)(sha->Nl >> 24);
243 p[sizeof(sha->u) - 5] = (uint8_t)(sha->Nl >> 32);
244 p[sizeof(sha->u) - 6] = (uint8_t)(sha->Nl >> 40);
245 p[sizeof(sha->u) - 7] = (uint8_t)(sha->Nl >> 48);
246 p[sizeof(sha->u) - 8] = (uint8_t)(sha->Nl >> 56);
247 p[sizeof(sha->u) - 9] = (uint8_t)(sha->Nh);
248 p[sizeof(sha->u) - 10] = (uint8_t)(sha->Nh >> 8);
249 p[sizeof(sha->u) - 11] = (uint8_t)(sha->Nh >> 16);
250 p[sizeof(sha->u) - 12] = (uint8_t)(sha->Nh >> 24);
251 p[sizeof(sha->u) - 13] = (uint8_t)(sha->Nh >> 32);
252 p[sizeof(sha->u) - 14] = (uint8_t)(sha->Nh >> 40);
253 p[sizeof(sha->u) - 15] = (uint8_t)(sha->Nh >> 48);
254 p[sizeof(sha->u) - 16] = (uint8_t)(sha->Nh >> 56);
255
256 sha512_block_data_order(sha->h, (uint64_t *)p, 1);
257
258 if (md == NULL) {
259 // TODO(davidben): This NULL check is absent in other low-level hash 'final'
260 // functions and is one of the few places one can fail.
261 return 0;
262 }
263
264 switch (sha->md_len) {
265 // Let compiler decide if it's appropriate to unroll...
266 case SHA384_DIGEST_LENGTH:
267 for (n = 0; n < SHA384_DIGEST_LENGTH / 8; n++) {
268 uint64_t t = sha->h[n];
269
270 *(md++) = (uint8_t)(t >> 56);
271 *(md++) = (uint8_t)(t >> 48);
272 *(md++) = (uint8_t)(t >> 40);
273 *(md++) = (uint8_t)(t >> 32);
274 *(md++) = (uint8_t)(t >> 24);
275 *(md++) = (uint8_t)(t >> 16);
276 *(md++) = (uint8_t)(t >> 8);
277 *(md++) = (uint8_t)(t);
278 }
279 break;
280 case SHA512_DIGEST_LENGTH:
281 for (n = 0; n < SHA512_DIGEST_LENGTH / 8; n++) {
282 uint64_t t = sha->h[n];
283
284 *(md++) = (uint8_t)(t >> 56);
285 *(md++) = (uint8_t)(t >> 48);
286 *(md++) = (uint8_t)(t >> 40);
287 *(md++) = (uint8_t)(t >> 32);
288 *(md++) = (uint8_t)(t >> 24);
289 *(md++) = (uint8_t)(t >> 16);
290 *(md++) = (uint8_t)(t >> 8);
291 *(md++) = (uint8_t)(t);
292 }
293 break;
294 // ... as well as make sure md_len is not abused.
295 default:
296 // TODO(davidben): This bad |md_len| case is one of the few places a
297 // low-level hash 'final' function can fail. This should never happen.
298 return 0;
299 }
300
301 return 1;
302 }
303
304 #ifndef SHA512_ASM
305 static const uint64_t K512[80] = {
306 UINT64_C(0x428a2f98d728ae22), UINT64_C(0x7137449123ef65cd),
307 UINT64_C(0xb5c0fbcfec4d3b2f), UINT64_C(0xe9b5dba58189dbbc),
308 UINT64_C(0x3956c25bf348b538), UINT64_C(0x59f111f1b605d019),
309 UINT64_C(0x923f82a4af194f9b), UINT64_C(0xab1c5ed5da6d8118),
310 UINT64_C(0xd807aa98a3030242), UINT64_C(0x12835b0145706fbe),
311 UINT64_C(0x243185be4ee4b28c), UINT64_C(0x550c7dc3d5ffb4e2),
312 UINT64_C(0x72be5d74f27b896f), UINT64_C(0x80deb1fe3b1696b1),
313 UINT64_C(0x9bdc06a725c71235), UINT64_C(0xc19bf174cf692694),
314 UINT64_C(0xe49b69c19ef14ad2), UINT64_C(0xefbe4786384f25e3),
315 UINT64_C(0x0fc19dc68b8cd5b5), UINT64_C(0x240ca1cc77ac9c65),
316 UINT64_C(0x2de92c6f592b0275), UINT64_C(0x4a7484aa6ea6e483),
317 UINT64_C(0x5cb0a9dcbd41fbd4), UINT64_C(0x76f988da831153b5),
318 UINT64_C(0x983e5152ee66dfab), UINT64_C(0xa831c66d2db43210),
319 UINT64_C(0xb00327c898fb213f), UINT64_C(0xbf597fc7beef0ee4),
320 UINT64_C(0xc6e00bf33da88fc2), UINT64_C(0xd5a79147930aa725),
321 UINT64_C(0x06ca6351e003826f), UINT64_C(0x142929670a0e6e70),
322 UINT64_C(0x27b70a8546d22ffc), UINT64_C(0x2e1b21385c26c926),
323 UINT64_C(0x4d2c6dfc5ac42aed), UINT64_C(0x53380d139d95b3df),
324 UINT64_C(0x650a73548baf63de), UINT64_C(0x766a0abb3c77b2a8),
325 UINT64_C(0x81c2c92e47edaee6), UINT64_C(0x92722c851482353b),
326 UINT64_C(0xa2bfe8a14cf10364), UINT64_C(0xa81a664bbc423001),
327 UINT64_C(0xc24b8b70d0f89791), UINT64_C(0xc76c51a30654be30),
328 UINT64_C(0xd192e819d6ef5218), UINT64_C(0xd69906245565a910),
329 UINT64_C(0xf40e35855771202a), UINT64_C(0x106aa07032bbd1b8),
330 UINT64_C(0x19a4c116b8d2d0c8), UINT64_C(0x1e376c085141ab53),
331 UINT64_C(0x2748774cdf8eeb99), UINT64_C(0x34b0bcb5e19b48a8),
332 UINT64_C(0x391c0cb3c5c95a63), UINT64_C(0x4ed8aa4ae3418acb),
333 UINT64_C(0x5b9cca4f7763e373), UINT64_C(0x682e6ff3d6b2b8a3),
334 UINT64_C(0x748f82ee5defb2fc), UINT64_C(0x78a5636f43172f60),
335 UINT64_C(0x84c87814a1f0ab72), UINT64_C(0x8cc702081a6439ec),
336 UINT64_C(0x90befffa23631e28), UINT64_C(0xa4506cebde82bde9),
337 UINT64_C(0xbef9a3f7b2c67915), UINT64_C(0xc67178f2e372532b),
338 UINT64_C(0xca273eceea26619c), UINT64_C(0xd186b8c721c0c207),
339 UINT64_C(0xeada7dd6cde0eb1e), UINT64_C(0xf57d4f7fee6ed178),
340 UINT64_C(0x06f067aa72176fba), UINT64_C(0x0a637dc5a2c898a6),
341 UINT64_C(0x113f9804bef90dae), UINT64_C(0x1b710b35131c471b),
342 UINT64_C(0x28db77f523047d84), UINT64_C(0x32caab7b40c72493),
343 UINT64_C(0x3c9ebe0a15c9bebc), UINT64_C(0x431d67c49c100d4c),
344 UINT64_C(0x4cc5d4becb3e42b6), UINT64_C(0x597f299cfc657e2a),
345 UINT64_C(0x5fcb6fab3ad6faec), UINT64_C(0x6c44198c4a475817),
346 };
347
348 #if defined(__GNUC__) && __GNUC__ >= 2 && !defined(OPENSSL_NO_ASM)
349 #if defined(__x86_64) || defined(__x86_64__)
350 #define ROTR(a, n) \
351 ({ \
352 uint64_t ret; \
353 __asm__("rorq %1, %0" : "=r"(ret) : "J"(n), "0"(a) : "cc"); \
354 ret; \
355 })
356 #define PULL64(x) \
357 ({ \
358 uint64_t ret = *((const uint64_t *)(&(x))); \
359 __asm__("bswapq %0" : "=r"(ret) : "0"(ret)); \
360 ret; \
361 })
362 #elif(defined(__i386) || defined(__i386__))
363 #define PULL64(x) \
364 ({ \
365 const unsigned int *p = (const unsigned int *)(&(x)); \
366 unsigned int hi = p[0], lo = p[1]; \
367 __asm__("bswapl %0; bswapl %1;" : "=r"(lo), "=r"(hi) : "0"(lo), "1"(hi)); \
368 ((uint64_t)hi) << 32 | lo; \
369 })
370 #elif(defined(_ARCH_PPC) && defined(__64BIT__)) || defined(_ARCH_PPC64)
371 #define ROTR(a, n) \
372 ({ \
373 uint64_t ret; \
374 __asm__("rotrdi %0, %1, %2" : "=r"(ret) : "r"(a), "K"(n)); \
375 ret; \
376 })
377 #elif defined(__aarch64__)
378 #define ROTR(a, n) \
379 ({ \
380 uint64_t ret; \
381 __asm__("ror %0, %1, %2" : "=r"(ret) : "r"(a), "I"(n)); \
382 ret; \
383 })
384 #if defined(__BYTE_ORDER__) && defined(__ORDER_LITTLE_ENDIAN__) && \
385 __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
386 #define PULL64(x) \
387 ({ \
388 uint64_t ret; \
389 __asm__("rev %0, %1" : "=r"(ret) : "r"(*((const uint64_t *)(&(x))))); \
390 ret; \
391 })
392 #endif
393 #endif
394 #elif defined(_MSC_VER)
395 #if defined(_WIN64) // applies to both IA-64 and AMD64
396 #pragma intrinsic(_rotr64)
397 #define ROTR(a, n) _rotr64((a), n)
398 #endif
399 #if defined(_M_IX86) && !defined(OPENSSL_NO_ASM)
__pull64be(const void * x)400 static uint64_t __fastcall __pull64be(const void *x) {
401 _asm mov edx, [ecx + 0]
402 _asm mov eax, [ecx + 4]
403 _asm bswap edx
404 _asm bswap eax
405 }
406 #define PULL64(x) __pull64be(&(x))
407 #if _MSC_VER <= 1200
408 #pragma inline_depth(0)
409 #endif
410 #endif
411 #endif
412
413 #ifndef PULL64
414 #define B(x, j) \
415 (((uint64_t)(*(((const uint8_t *)(&x)) + j))) << ((7 - j) * 8))
416 #define PULL64(x) \
417 (B(x, 0) | B(x, 1) | B(x, 2) | B(x, 3) | B(x, 4) | B(x, 5) | B(x, 6) | \
418 B(x, 7))
419 #endif
420
421 #ifndef ROTR
422 #define ROTR(x, s) (((x) >> s) | (x) << (64 - s))
423 #endif
424
425 #define Sigma0(x) (ROTR((x), 28) ^ ROTR((x), 34) ^ ROTR((x), 39))
426 #define Sigma1(x) (ROTR((x), 14) ^ ROTR((x), 18) ^ ROTR((x), 41))
427 #define sigma0(x) (ROTR((x), 1) ^ ROTR((x), 8) ^ ((x) >> 7))
428 #define sigma1(x) (ROTR((x), 19) ^ ROTR((x), 61) ^ ((x) >> 6))
429
430 #define Ch(x, y, z) (((x) & (y)) ^ ((~(x)) & (z)))
431 #define Maj(x, y, z) (((x) & (y)) ^ ((x) & (z)) ^ ((y) & (z)))
432
433
434 #if defined(__i386) || defined(__i386__) || defined(_M_IX86)
435 // This code should give better results on 32-bit CPU with less than
436 // ~24 registers, both size and performance wise...
sha512_block_data_order(uint64_t * state,const uint64_t * W,size_t num)437 static void sha512_block_data_order(uint64_t *state, const uint64_t *W,
438 size_t num) {
439 uint64_t A, E, T;
440 uint64_t X[9 + 80], *F;
441 int i;
442
443 while (num--) {
444 F = X + 80;
445 A = state[0];
446 F[1] = state[1];
447 F[2] = state[2];
448 F[3] = state[3];
449 E = state[4];
450 F[5] = state[5];
451 F[6] = state[6];
452 F[7] = state[7];
453
454 for (i = 0; i < 16; i++, F--) {
455 T = PULL64(W[i]);
456 F[0] = A;
457 F[4] = E;
458 F[8] = T;
459 T += F[7] + Sigma1(E) + Ch(E, F[5], F[6]) + K512[i];
460 E = F[3] + T;
461 A = T + Sigma0(A) + Maj(A, F[1], F[2]);
462 }
463
464 for (; i < 80; i++, F--) {
465 T = sigma0(F[8 + 16 - 1]);
466 T += sigma1(F[8 + 16 - 14]);
467 T += F[8 + 16] + F[8 + 16 - 9];
468
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 state[0] += A;
478 state[1] += F[1];
479 state[2] += F[2];
480 state[3] += F[3];
481 state[4] += E;
482 state[5] += F[5];
483 state[6] += F[6];
484 state[7] += F[7];
485
486 W += 16;
487 }
488 }
489
490 #else
491
492 #define ROUND_00_15(i, a, b, c, d, e, f, g, h) \
493 do { \
494 T1 += h + Sigma1(e) + Ch(e, f, g) + K512[i]; \
495 h = Sigma0(a) + Maj(a, b, c); \
496 d += T1; \
497 h += T1; \
498 } while (0)
499
500 #define ROUND_16_80(i, j, a, b, c, d, e, f, g, h, X) \
501 do { \
502 s0 = X[(j + 1) & 0x0f]; \
503 s0 = sigma0(s0); \
504 s1 = X[(j + 14) & 0x0f]; \
505 s1 = sigma1(s1); \
506 T1 = X[(j) & 0x0f] += s0 + s1 + X[(j + 9) & 0x0f]; \
507 ROUND_00_15(i + j, a, b, c, d, e, f, g, h); \
508 } while (0)
509
sha512_block_data_order(uint64_t * state,const uint64_t * W,size_t num)510 static void sha512_block_data_order(uint64_t *state, const uint64_t *W,
511 size_t num) {
512 uint64_t a, b, c, d, e, f, g, h, s0, s1, T1;
513 uint64_t X[16];
514 int i;
515
516 while (num--) {
517
518 a = state[0];
519 b = state[1];
520 c = state[2];
521 d = state[3];
522 e = state[4];
523 f = state[5];
524 g = state[6];
525 h = state[7];
526
527 T1 = X[0] = PULL64(W[0]);
528 ROUND_00_15(0, a, b, c, d, e, f, g, h);
529 T1 = X[1] = PULL64(W[1]);
530 ROUND_00_15(1, h, a, b, c, d, e, f, g);
531 T1 = X[2] = PULL64(W[2]);
532 ROUND_00_15(2, g, h, a, b, c, d, e, f);
533 T1 = X[3] = PULL64(W[3]);
534 ROUND_00_15(3, f, g, h, a, b, c, d, e);
535 T1 = X[4] = PULL64(W[4]);
536 ROUND_00_15(4, e, f, g, h, a, b, c, d);
537 T1 = X[5] = PULL64(W[5]);
538 ROUND_00_15(5, d, e, f, g, h, a, b, c);
539 T1 = X[6] = PULL64(W[6]);
540 ROUND_00_15(6, c, d, e, f, g, h, a, b);
541 T1 = X[7] = PULL64(W[7]);
542 ROUND_00_15(7, b, c, d, e, f, g, h, a);
543 T1 = X[8] = PULL64(W[8]);
544 ROUND_00_15(8, a, b, c, d, e, f, g, h);
545 T1 = X[9] = PULL64(W[9]);
546 ROUND_00_15(9, h, a, b, c, d, e, f, g);
547 T1 = X[10] = PULL64(W[10]);
548 ROUND_00_15(10, g, h, a, b, c, d, e, f);
549 T1 = X[11] = PULL64(W[11]);
550 ROUND_00_15(11, f, g, h, a, b, c, d, e);
551 T1 = X[12] = PULL64(W[12]);
552 ROUND_00_15(12, e, f, g, h, a, b, c, d);
553 T1 = X[13] = PULL64(W[13]);
554 ROUND_00_15(13, d, e, f, g, h, a, b, c);
555 T1 = X[14] = PULL64(W[14]);
556 ROUND_00_15(14, c, d, e, f, g, h, a, b);
557 T1 = X[15] = PULL64(W[15]);
558 ROUND_00_15(15, b, c, d, e, f, g, h, a);
559
560 for (i = 16; i < 80; i += 16) {
561 ROUND_16_80(i, 0, a, b, c, d, e, f, g, h, X);
562 ROUND_16_80(i, 1, h, a, b, c, d, e, f, g, X);
563 ROUND_16_80(i, 2, g, h, a, b, c, d, e, f, X);
564 ROUND_16_80(i, 3, f, g, h, a, b, c, d, e, X);
565 ROUND_16_80(i, 4, e, f, g, h, a, b, c, d, X);
566 ROUND_16_80(i, 5, d, e, f, g, h, a, b, c, X);
567 ROUND_16_80(i, 6, c, d, e, f, g, h, a, b, X);
568 ROUND_16_80(i, 7, b, c, d, e, f, g, h, a, X);
569 ROUND_16_80(i, 8, a, b, c, d, e, f, g, h, X);
570 ROUND_16_80(i, 9, h, a, b, c, d, e, f, g, X);
571 ROUND_16_80(i, 10, g, h, a, b, c, d, e, f, X);
572 ROUND_16_80(i, 11, f, g, h, a, b, c, d, e, X);
573 ROUND_16_80(i, 12, e, f, g, h, a, b, c, d, X);
574 ROUND_16_80(i, 13, d, e, f, g, h, a, b, c, X);
575 ROUND_16_80(i, 14, c, d, e, f, g, h, a, b, X);
576 ROUND_16_80(i, 15, b, c, d, e, f, g, h, a, X);
577 }
578
579 state[0] += a;
580 state[1] += b;
581 state[2] += c;
582 state[3] += d;
583 state[4] += e;
584 state[5] += f;
585 state[6] += g;
586 state[7] += h;
587
588 W += 16;
589 }
590 }
591
592 #endif
593
594 #endif // !SHA512_ASM
595
596 #undef ROTR
597 #undef PULL64
598 #undef B
599 #undef Sigma0
600 #undef Sigma1
601 #undef sigma0
602 #undef sigma1
603 #undef Ch
604 #undef Maj
605 #undef ROUND_00_15
606 #undef ROUND_16_80
607 #undef HOST_c2l
608 #undef HOST_l2c
609