1 /* vi: set sw=4 ts=4: */
5 * Copyright (C) 2010 Denys Vlasenko
7 * Licensed under GPLv2 or later, see file LICENSE in this source tree.
12 /* gcc 4.2.1 optimizes rotr64 better with inline than with macro
13 * (for rotX32, there is no difference). Why? My guess is that
14 * macro requires clever common subexpression elimination heuristics
15 * in gcc, while inline basically forces it to happen.
17 //#define rotl32(x,n) (((x) << (n)) | ((x) >> (32 - (n))))
18 static ALWAYS_INLINE uint32_t rotl32(uint32_t x, unsigned n)
20 return (x << n) | (x >> (32 - n));
22 //#define rotr32(x,n) (((x) >> (n)) | ((x) << (32 - (n))))
23 static ALWAYS_INLINE uint32_t rotr32(uint32_t x, unsigned n)
25 return (x >> n) | (x << (32 - n));
27 /* rotr64 in needed for sha512 only: */
28 //#define rotr64(x,n) (((x) >> (n)) | ((x) << (64 - (n))))
29 static ALWAYS_INLINE uint64_t rotr64(uint64_t x, unsigned n)
31 return (x >> n) | (x << (64 - n));
34 /* rotl64 only used for sha3 currently */
35 static ALWAYS_INLINE uint64_t rotl64(uint64_t x, unsigned n)
37 return (x << n) | (x >> (64 - n));
40 /* Feed data through a temporary buffer.
41 * The internal buffer remembers previous data until it has 64
42 * bytes worth to pass on.
44 static void FAST_FUNC common64_hash(md5_ctx_t *ctx, const void *buffer, size_t len)
46 unsigned bufpos = ctx->total64 & 63;
51 unsigned remaining = 64 - bufpos;
54 /* Copy data into aligned buffer */
55 memcpy(ctx->wbuffer + bufpos, buffer, remaining);
57 buffer = (const char *)buffer + remaining;
59 /* Clever way to do "if (bufpos != N) break; ... ; bufpos = 0;" */
63 /* Buffer is filled up, process it */
64 ctx->process_block(ctx);
65 /*bufpos = 0; - already is */
69 /* Process the remaining bytes in the buffer */
70 static void FAST_FUNC common64_end(md5_ctx_t *ctx, int swap_needed)
72 unsigned bufpos = ctx->total64 & 63;
73 /* Pad the buffer to the next 64-byte boundary with 0x80,0,0,0... */
74 ctx->wbuffer[bufpos++] = 0x80;
76 /* This loop iterates either once or twice, no more, no less */
78 unsigned remaining = 64 - bufpos;
79 memset(ctx->wbuffer + bufpos, 0, remaining);
80 /* Do we have enough space for the length count? */
82 /* Store the 64-bit counter of bits in the buffer */
83 uint64_t t = ctx->total64 << 3;
86 /* wbuffer is suitably aligned for this */
87 *(bb__aliased_uint64_t *) (&ctx->wbuffer[64 - 8]) = t;
89 ctx->process_block(ctx);
98 * Compute MD5 checksum of strings according to the
99 * definition of MD5 in RFC 1321 from April 1992.
101 * Written by Ulrich Drepper <drepper@gnu.ai.mit.edu>, 1995.
103 * Copyright (C) 1995-1999 Free Software Foundation, Inc.
104 * Copyright (C) 2001 Manuel Novoa III
105 * Copyright (C) 2003 Glenn L. McGrath
106 * Copyright (C) 2003 Erik Andersen
108 * Licensed under GPLv2 or later, see file LICENSE in this source tree.
111 /* 0: fastest, 3: smallest */
112 #if CONFIG_MD5_SMALL < 0
114 #elif CONFIG_MD5_SMALL > 3
117 # define MD5_SMALL CONFIG_MD5_SMALL
120 /* These are the four functions used in the four steps of the MD5 algorithm
121 * and defined in the RFC 1321. The first function is a little bit optimized
122 * (as found in Colin Plumbs public domain implementation).
123 * #define FF(b, c, d) ((b & c) | (~b & d))
129 #define FF(b, c, d) (d ^ (b & (c ^ d)))
130 #define FG(b, c, d) FF(d, b, c)
131 #define FH(b, c, d) (b ^ c ^ d)
132 #define FI(b, c, d) (c ^ (b | ~d))
134 /* Hash a single block, 64 bytes long and 4-byte aligned */
135 static void FAST_FUNC md5_process_block64(md5_ctx_t *ctx)
138 /* Before we start, one word to the strange constants.
139 They are defined in RFC 1321 as
140 T[i] = (int)(4294967296.0 * fabs(sin(i))), i=1..64
142 static const uint32_t C_array[] = {
144 0xd76aa478, 0xe8c7b756, 0x242070db, 0xc1bdceee,
145 0xf57c0faf, 0x4787c62a, 0xa8304613, 0xfd469501,
146 0x698098d8, 0x8b44f7af, 0xffff5bb1, 0x895cd7be,
147 0x6b901122, 0xfd987193, 0xa679438e, 0x49b40821,
149 0xf61e2562, 0xc040b340, 0x265e5a51, 0xe9b6c7aa,
150 0xd62f105d, 0x02441453, 0xd8a1e681, 0xe7d3fbc8,
151 0x21e1cde6, 0xc33707d6, 0xf4d50d87, 0x455a14ed,
152 0xa9e3e905, 0xfcefa3f8, 0x676f02d9, 0x8d2a4c8a,
154 0xfffa3942, 0x8771f681, 0x6d9d6122, 0xfde5380c,
155 0xa4beea44, 0x4bdecfa9, 0xf6bb4b60, 0xbebfbc70,
156 0x289b7ec6, 0xeaa127fa, 0xd4ef3085, 0x4881d05,
157 0xd9d4d039, 0xe6db99e5, 0x1fa27cf8, 0xc4ac5665,
159 0xf4292244, 0x432aff97, 0xab9423a7, 0xfc93a039,
160 0x655b59c3, 0x8f0ccc92, 0xffeff47d, 0x85845dd1,
161 0x6fa87e4f, 0xfe2ce6e0, 0xa3014314, 0x4e0811a1,
162 0xf7537e82, 0xbd3af235, 0x2ad7d2bb, 0xeb86d391
164 static const char P_array[] ALIGN1 = {
166 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, /* 1 */
168 1, 6, 11, 0, 5, 10, 15, 4, 9, 14, 3, 8, 13, 2, 7, 12, /* 2 */
169 5, 8, 11, 14, 1, 4, 7, 10, 13, 0, 3, 6, 9, 12, 15, 2, /* 3 */
170 0, 7, 14, 5, 12, 3, 10, 1, 8, 15, 6, 13, 4, 11, 2, 9 /* 4 */
173 uint32_t *words = (void*) ctx->wbuffer;
174 uint32_t A = ctx->hash[0];
175 uint32_t B = ctx->hash[1];
176 uint32_t C = ctx->hash[2];
177 uint32_t D = ctx->hash[3];
179 #if MD5_SMALL >= 2 /* 2 or 3 */
181 static const char S_array[] ALIGN1 = {
194 for (i = 0; i < 16; i++)
195 words[i] = SWAP_LE32(words[i]);
202 for (i = 0; i < 64; i++) {
219 temp += words[(int) (*pp++)] + *pc++;
220 temp = rotl32(temp, ps[i & 3]);
227 # else /* MD5_SMALL == 2 */
232 for (i = 0; i < 16; i++) {
233 temp = A + FF(B, C, D) + words[(int) (*pp++)] + *pc++;
234 temp = rotl32(temp, ps[i & 3]);
242 for (i = 0; i < 16; i++) {
243 temp = A + FG(B, C, D) + words[(int) (*pp++)] + *pc++;
244 temp = rotl32(temp, ps[i & 3]);
252 for (i = 0; i < 16; i++) {
253 temp = A + FH(B, C, D) + words[(int) (*pp++)] + *pc++;
254 temp = rotl32(temp, ps[i & 3]);
262 for (i = 0; i < 16; i++) {
263 temp = A + FI(B, C, D) + words[(int) (*pp++)] + *pc++;
264 temp = rotl32(temp, ps[i & 3]);
272 /* Add checksum to the starting values */
278 #else /* MD5_SMALL == 0 or 1 */
290 /* First round: using the given function, the context and a constant
291 the next context is computed. Because the algorithm's processing
292 unit is a 32-bit word and it is determined to work on words in
293 little endian byte order we perhaps have to change the byte order
294 before the computation. To reduce the work for the next steps
295 we save swapped words in WORDS array. */
297 # define OP(a, b, c, d, s, T) \
299 a += FF(b, c, d) + (*words IF_BIG_ENDIAN(= SWAP_LE32(*words))) + T; \
308 for (i = 0; i < 4; i++) {
309 OP(A, B, C, D, 7, *pc++);
310 OP(D, A, B, C, 12, *pc++);
311 OP(C, D, A, B, 17, *pc++);
312 OP(B, C, D, A, 22, *pc++);
315 OP(A, B, C, D, 7, 0xd76aa478);
316 OP(D, A, B, C, 12, 0xe8c7b756);
317 OP(C, D, A, B, 17, 0x242070db);
318 OP(B, C, D, A, 22, 0xc1bdceee);
319 OP(A, B, C, D, 7, 0xf57c0faf);
320 OP(D, A, B, C, 12, 0x4787c62a);
321 OP(C, D, A, B, 17, 0xa8304613);
322 OP(B, C, D, A, 22, 0xfd469501);
323 OP(A, B, C, D, 7, 0x698098d8);
324 OP(D, A, B, C, 12, 0x8b44f7af);
325 OP(C, D, A, B, 17, 0xffff5bb1);
326 OP(B, C, D, A, 22, 0x895cd7be);
327 OP(A, B, C, D, 7, 0x6b901122);
328 OP(D, A, B, C, 12, 0xfd987193);
329 OP(C, D, A, B, 17, 0xa679438e);
330 OP(B, C, D, A, 22, 0x49b40821);
334 /* For the second to fourth round we have the possibly swapped words
335 in WORDS. Redefine the macro to take an additional first
336 argument specifying the function to use. */
338 # define OP(f, a, b, c, d, k, s, T) \
340 a += f(b, c, d) + words[k] + T; \
348 for (i = 0; i < 4; i++) {
349 OP(FG, A, B, C, D, (int) (*pp++), 5, *pc++);
350 OP(FG, D, A, B, C, (int) (*pp++), 9, *pc++);
351 OP(FG, C, D, A, B, (int) (*pp++), 14, *pc++);
352 OP(FG, B, C, D, A, (int) (*pp++), 20, *pc++);
355 OP(FG, A, B, C, D, 1, 5, 0xf61e2562);
356 OP(FG, D, A, B, C, 6, 9, 0xc040b340);
357 OP(FG, C, D, A, B, 11, 14, 0x265e5a51);
358 OP(FG, B, C, D, A, 0, 20, 0xe9b6c7aa);
359 OP(FG, A, B, C, D, 5, 5, 0xd62f105d);
360 OP(FG, D, A, B, C, 10, 9, 0x02441453);
361 OP(FG, C, D, A, B, 15, 14, 0xd8a1e681);
362 OP(FG, B, C, D, A, 4, 20, 0xe7d3fbc8);
363 OP(FG, A, B, C, D, 9, 5, 0x21e1cde6);
364 OP(FG, D, A, B, C, 14, 9, 0xc33707d6);
365 OP(FG, C, D, A, B, 3, 14, 0xf4d50d87);
366 OP(FG, B, C, D, A, 8, 20, 0x455a14ed);
367 OP(FG, A, B, C, D, 13, 5, 0xa9e3e905);
368 OP(FG, D, A, B, C, 2, 9, 0xfcefa3f8);
369 OP(FG, C, D, A, B, 7, 14, 0x676f02d9);
370 OP(FG, B, C, D, A, 12, 20, 0x8d2a4c8a);
375 for (i = 0; i < 4; i++) {
376 OP(FH, A, B, C, D, (int) (*pp++), 4, *pc++);
377 OP(FH, D, A, B, C, (int) (*pp++), 11, *pc++);
378 OP(FH, C, D, A, B, (int) (*pp++), 16, *pc++);
379 OP(FH, B, C, D, A, (int) (*pp++), 23, *pc++);
382 OP(FH, A, B, C, D, 5, 4, 0xfffa3942);
383 OP(FH, D, A, B, C, 8, 11, 0x8771f681);
384 OP(FH, C, D, A, B, 11, 16, 0x6d9d6122);
385 OP(FH, B, C, D, A, 14, 23, 0xfde5380c);
386 OP(FH, A, B, C, D, 1, 4, 0xa4beea44);
387 OP(FH, D, A, B, C, 4, 11, 0x4bdecfa9);
388 OP(FH, C, D, A, B, 7, 16, 0xf6bb4b60);
389 OP(FH, B, C, D, A, 10, 23, 0xbebfbc70);
390 OP(FH, A, B, C, D, 13, 4, 0x289b7ec6);
391 OP(FH, D, A, B, C, 0, 11, 0xeaa127fa);
392 OP(FH, C, D, A, B, 3, 16, 0xd4ef3085);
393 OP(FH, B, C, D, A, 6, 23, 0x04881d05);
394 OP(FH, A, B, C, D, 9, 4, 0xd9d4d039);
395 OP(FH, D, A, B, C, 12, 11, 0xe6db99e5);
396 OP(FH, C, D, A, B, 15, 16, 0x1fa27cf8);
397 OP(FH, B, C, D, A, 2, 23, 0xc4ac5665);
402 for (i = 0; i < 4; i++) {
403 OP(FI, A, B, C, D, (int) (*pp++), 6, *pc++);
404 OP(FI, D, A, B, C, (int) (*pp++), 10, *pc++);
405 OP(FI, C, D, A, B, (int) (*pp++), 15, *pc++);
406 OP(FI, B, C, D, A, (int) (*pp++), 21, *pc++);
409 OP(FI, A, B, C, D, 0, 6, 0xf4292244);
410 OP(FI, D, A, B, C, 7, 10, 0x432aff97);
411 OP(FI, C, D, A, B, 14, 15, 0xab9423a7);
412 OP(FI, B, C, D, A, 5, 21, 0xfc93a039);
413 OP(FI, A, B, C, D, 12, 6, 0x655b59c3);
414 OP(FI, D, A, B, C, 3, 10, 0x8f0ccc92);
415 OP(FI, C, D, A, B, 10, 15, 0xffeff47d);
416 OP(FI, B, C, D, A, 1, 21, 0x85845dd1);
417 OP(FI, A, B, C, D, 8, 6, 0x6fa87e4f);
418 OP(FI, D, A, B, C, 15, 10, 0xfe2ce6e0);
419 OP(FI, C, D, A, B, 6, 15, 0xa3014314);
420 OP(FI, B, C, D, A, 13, 21, 0x4e0811a1);
421 OP(FI, A, B, C, D, 4, 6, 0xf7537e82);
422 OP(FI, D, A, B, C, 11, 10, 0xbd3af235);
423 OP(FI, C, D, A, B, 2, 15, 0x2ad7d2bb);
424 OP(FI, B, C, D, A, 9, 21, 0xeb86d391);
427 /* Add checksum to the starting values */
428 ctx->hash[0] = A_save + A;
429 ctx->hash[1] = B_save + B;
430 ctx->hash[2] = C_save + C;
431 ctx->hash[3] = D_save + D;
439 /* Initialize structure containing state of computation.
440 * (RFC 1321, 3.3: Step 3)
442 void FAST_FUNC md5_begin(md5_ctx_t *ctx)
444 ctx->hash[0] = 0x67452301;
445 ctx->hash[1] = 0xefcdab89;
446 ctx->hash[2] = 0x98badcfe;
447 ctx->hash[3] = 0x10325476;
449 ctx->process_block = md5_process_block64;
452 /* Used also for sha1 and sha256 */
453 void FAST_FUNC md5_hash(md5_ctx_t *ctx, const void *buffer, size_t len)
455 common64_hash(ctx, buffer, len);
458 /* Process the remaining bytes in the buffer and put result from CTX
459 * in first 16 bytes following RESBUF. The result is always in little
460 * endian byte order, so that a byte-wise output yields to the wanted
461 * ASCII representation of the message digest.
463 void FAST_FUNC md5_end(md5_ctx_t *ctx, void *resbuf)
465 /* MD5 stores total in LE, need to swap on BE arches: */
466 common64_end(ctx, /*swap_needed:*/ BB_BIG_ENDIAN);
468 /* The MD5 result is in little endian byte order */
470 ctx->hash[0] = SWAP_LE32(ctx->hash[0]);
471 ctx->hash[1] = SWAP_LE32(ctx->hash[1]);
472 ctx->hash[2] = SWAP_LE32(ctx->hash[2]);
473 ctx->hash[3] = SWAP_LE32(ctx->hash[3]);
476 memcpy(resbuf, ctx->hash, sizeof(ctx->hash[0]) * 4);
482 * Copyright 2007 Rob Landley <rob@landley.net>
484 * Based on the public domain SHA-1 in C by Steve Reid <steve@edmweb.com>
485 * from http://www.mirrors.wiretapped.net/security/cryptography/hashes/sha1/
487 * Licensed under GPLv2, see file LICENSE in this source tree.
489 * ---------------------------------------------------------------------------
491 * SHA256 and SHA512 parts are:
492 * Released into the Public Domain by Ulrich Drepper <drepper@redhat.com>.
493 * Shrank by Denys Vlasenko.
495 * ---------------------------------------------------------------------------
497 * The best way to test random blocksizes is to go to coreutils/md5_sha1_sum.c
498 * and replace "4096" with something like "2000 + time(NULL) % 2097",
499 * then rebuild and compare "shaNNNsum bigfile" results.
502 static void FAST_FUNC sha1_process_block64(sha1_ctx_t *ctx)
504 static const uint32_t rconsts[] = {
505 0x5A827999, 0x6ED9EBA1, 0x8F1BBCDC, 0xCA62C1D6
510 uint32_t a, b, c, d, e;
512 /* On-stack work buffer frees up one register in the main loop
513 * which otherwise will be needed to hold ctx pointer */
514 for (i = 0; i < 16; i++)
515 W[i] = W[i+16] = SWAP_BE32(((uint32_t*)ctx->wbuffer)[i]);
523 /* 4 rounds of 20 operations each */
525 for (i = 0; i < 4; i++) {
532 work = (work & b) ^ d;
535 /* Used to do SWAP_BE32 here, but this
536 * requires ctx (see comment above) */
540 work = ((b | c) & d) | (b & c);
541 else /* i = 1 or 3 */
544 W[cnt] = W[cnt+16] = rotl32(W[cnt+13] ^ W[cnt+8] ^ W[cnt+2] ^ W[cnt], 1);
547 work += e + rotl32(a, 5) + rconsts[i];
549 /* Rotate by one for next time */
552 c = /* b = */ rotl32(b, 30);
555 cnt = (cnt + 1) & 15;
566 /* Constants for SHA512 from FIPS 180-2:4.2.3.
567 * SHA256 constants from FIPS 180-2:4.2.2
568 * are the most significant half of first 64 elements
571 static const uint64_t sha_K[80] = {
572 0x428a2f98d728ae22ULL, 0x7137449123ef65cdULL,
573 0xb5c0fbcfec4d3b2fULL, 0xe9b5dba58189dbbcULL,
574 0x3956c25bf348b538ULL, 0x59f111f1b605d019ULL,
575 0x923f82a4af194f9bULL, 0xab1c5ed5da6d8118ULL,
576 0xd807aa98a3030242ULL, 0x12835b0145706fbeULL,
577 0x243185be4ee4b28cULL, 0x550c7dc3d5ffb4e2ULL,
578 0x72be5d74f27b896fULL, 0x80deb1fe3b1696b1ULL,
579 0x9bdc06a725c71235ULL, 0xc19bf174cf692694ULL,
580 0xe49b69c19ef14ad2ULL, 0xefbe4786384f25e3ULL,
581 0x0fc19dc68b8cd5b5ULL, 0x240ca1cc77ac9c65ULL,
582 0x2de92c6f592b0275ULL, 0x4a7484aa6ea6e483ULL,
583 0x5cb0a9dcbd41fbd4ULL, 0x76f988da831153b5ULL,
584 0x983e5152ee66dfabULL, 0xa831c66d2db43210ULL,
585 0xb00327c898fb213fULL, 0xbf597fc7beef0ee4ULL,
586 0xc6e00bf33da88fc2ULL, 0xd5a79147930aa725ULL,
587 0x06ca6351e003826fULL, 0x142929670a0e6e70ULL,
588 0x27b70a8546d22ffcULL, 0x2e1b21385c26c926ULL,
589 0x4d2c6dfc5ac42aedULL, 0x53380d139d95b3dfULL,
590 0x650a73548baf63deULL, 0x766a0abb3c77b2a8ULL,
591 0x81c2c92e47edaee6ULL, 0x92722c851482353bULL,
592 0xa2bfe8a14cf10364ULL, 0xa81a664bbc423001ULL,
593 0xc24b8b70d0f89791ULL, 0xc76c51a30654be30ULL,
594 0xd192e819d6ef5218ULL, 0xd69906245565a910ULL,
595 0xf40e35855771202aULL, 0x106aa07032bbd1b8ULL,
596 0x19a4c116b8d2d0c8ULL, 0x1e376c085141ab53ULL,
597 0x2748774cdf8eeb99ULL, 0x34b0bcb5e19b48a8ULL,
598 0x391c0cb3c5c95a63ULL, 0x4ed8aa4ae3418acbULL,
599 0x5b9cca4f7763e373ULL, 0x682e6ff3d6b2b8a3ULL,
600 0x748f82ee5defb2fcULL, 0x78a5636f43172f60ULL,
601 0x84c87814a1f0ab72ULL, 0x8cc702081a6439ecULL,
602 0x90befffa23631e28ULL, 0xa4506cebde82bde9ULL,
603 0xbef9a3f7b2c67915ULL, 0xc67178f2e372532bULL,
604 0xca273eceea26619cULL, 0xd186b8c721c0c207ULL, /* [64]+ are used for sha512 only */
605 0xeada7dd6cde0eb1eULL, 0xf57d4f7fee6ed178ULL,
606 0x06f067aa72176fbaULL, 0x0a637dc5a2c898a6ULL,
607 0x113f9804bef90daeULL, 0x1b710b35131c471bULL,
608 0x28db77f523047d84ULL, 0x32caab7b40c72493ULL,
609 0x3c9ebe0a15c9bebcULL, 0x431d67c49c100d4cULL,
610 0x4cc5d4becb3e42b6ULL, 0x597f299cfc657e2aULL,
611 0x5fcb6fab3ad6faecULL, 0x6c44198c4a475817ULL
621 static void FAST_FUNC sha256_process_block64(sha256_ctx_t *ctx)
624 uint32_t W[64], a, b, c, d, e, f, g, h;
625 const uint32_t *words = (uint32_t*) ctx->wbuffer;
627 /* Operators defined in FIPS 180-2:4.1.2. */
628 #define Ch(x, y, z) ((x & y) ^ (~x & z))
629 #define Maj(x, y, z) ((x & y) ^ (x & z) ^ (y & z))
630 #define S0(x) (rotr32(x, 2) ^ rotr32(x, 13) ^ rotr32(x, 22))
631 #define S1(x) (rotr32(x, 6) ^ rotr32(x, 11) ^ rotr32(x, 25))
632 #define R0(x) (rotr32(x, 7) ^ rotr32(x, 18) ^ (x >> 3))
633 #define R1(x) (rotr32(x, 17) ^ rotr32(x, 19) ^ (x >> 10))
635 /* Compute the message schedule according to FIPS 180-2:6.2.2 step 2. */
636 for (t = 0; t < 16; ++t)
637 W[t] = SWAP_BE32(words[t]);
638 for (/*t = 16*/; t < 64; ++t)
639 W[t] = R1(W[t - 2]) + W[t - 7] + R0(W[t - 15]) + W[t - 16];
650 /* The actual computation according to FIPS 180-2:6.2.2 step 3. */
651 for (t = 0; t < 64; ++t) {
652 /* Need to fetch upper half of sha_K[t]
653 * (I hope compiler is clever enough to just fetch
656 uint32_t K_t = sha_K[t] >> 32;
657 uint32_t T1 = h + S1(e) + Ch(e, f, g) + K_t + W[t];
658 uint32_t T2 = S0(a) + Maj(a, b, c);
674 /* Add the starting values of the context according to FIPS 180-2:6.2.2
686 static void FAST_FUNC sha512_process_block128(sha512_ctx_t *ctx)
690 /* On i386, having assignments here (not later as sha256 does)
691 * produces 99 bytes smaller code with gcc 4.3.1
693 uint64_t a = ctx->hash[0];
694 uint64_t b = ctx->hash[1];
695 uint64_t c = ctx->hash[2];
696 uint64_t d = ctx->hash[3];
697 uint64_t e = ctx->hash[4];
698 uint64_t f = ctx->hash[5];
699 uint64_t g = ctx->hash[6];
700 uint64_t h = ctx->hash[7];
701 const uint64_t *words = (uint64_t*) ctx->wbuffer;
703 /* Operators defined in FIPS 180-2:4.1.2. */
704 #define Ch(x, y, z) ((x & y) ^ (~x & z))
705 #define Maj(x, y, z) ((x & y) ^ (x & z) ^ (y & z))
706 #define S0(x) (rotr64(x, 28) ^ rotr64(x, 34) ^ rotr64(x, 39))
707 #define S1(x) (rotr64(x, 14) ^ rotr64(x, 18) ^ rotr64(x, 41))
708 #define R0(x) (rotr64(x, 1) ^ rotr64(x, 8) ^ (x >> 7))
709 #define R1(x) (rotr64(x, 19) ^ rotr64(x, 61) ^ (x >> 6))
711 /* Compute the message schedule according to FIPS 180-2:6.3.2 step 2. */
712 for (t = 0; t < 16; ++t)
713 W[t] = SWAP_BE64(words[t]);
714 for (/*t = 16*/; t < 80; ++t)
715 W[t] = R1(W[t - 2]) + W[t - 7] + R0(W[t - 15]) + W[t - 16];
717 /* The actual computation according to FIPS 180-2:6.3.2 step 3. */
718 for (t = 0; t < 80; ++t) {
719 uint64_t T1 = h + S1(e) + Ch(e, f, g) + sha_K[t] + W[t];
720 uint64_t T2 = S0(a) + Maj(a, b, c);
736 /* Add the starting values of the context according to FIPS 180-2:6.3.2
749 void FAST_FUNC sha1_begin(sha1_ctx_t *ctx)
751 ctx->hash[0] = 0x67452301;
752 ctx->hash[1] = 0xefcdab89;
753 ctx->hash[2] = 0x98badcfe;
754 ctx->hash[3] = 0x10325476;
755 ctx->hash[4] = 0xc3d2e1f0;
757 ctx->process_block = sha1_process_block64;
760 static const uint32_t init256[] = {
772 static const uint32_t init512_lo[] = {
785 /* Initialize structure containing state of computation.
786 (FIPS 180-2:5.3.2) */
787 void FAST_FUNC sha256_begin(sha256_ctx_t *ctx)
789 memcpy(&ctx->total64, init256, sizeof(init256));
790 /*ctx->total64 = 0; - done by prepending two 32-bit zeros to init256 */
791 ctx->process_block = sha256_process_block64;
794 /* Initialize structure containing state of computation.
795 (FIPS 180-2:5.3.3) */
796 void FAST_FUNC sha512_begin(sha512_ctx_t *ctx)
799 /* Two extra iterations zero out ctx->total64[2] */
800 uint64_t *tp = ctx->total64;
801 for (i = 0; i < 2+8; i++)
802 tp[i] = ((uint64_t)(init256[i]) << 32) + init512_lo[i];
803 /*ctx->total64[0] = ctx->total64[1] = 0; - already done */
806 void FAST_FUNC sha512_hash(sha512_ctx_t *ctx, const void *buffer, size_t len)
808 unsigned bufpos = ctx->total64[0] & 127;
811 /* First increment the byte count. FIPS 180-2 specifies the possible
812 length of the file up to 2^128 _bits_.
813 We compute the number of _bytes_ and convert to bits later. */
814 ctx->total64[0] += len;
815 if (ctx->total64[0] < len)
818 remaining = 128 - bufpos;
820 /* Hash whole blocks */
821 while (len >= remaining) {
822 memcpy(ctx->wbuffer + bufpos, buffer, remaining);
823 buffer = (const char *)buffer + remaining;
827 sha512_process_block128(ctx);
830 /* Save last, partial blosk */
831 memcpy(ctx->wbuffer + bufpos, buffer, len);
834 remaining = 128 - bufpos;
837 /* Copy data into aligned buffer */
838 memcpy(ctx->wbuffer + bufpos, buffer, remaining);
840 buffer = (const char *)buffer + remaining;
842 /* Clever way to do "if (bufpos != N) break; ... ; bufpos = 0;" */
846 /* Buffer is filled up, process it */
847 sha512_process_block128(ctx);
848 /*bufpos = 0; - already is */
853 /* Used also for sha256 */
854 void FAST_FUNC sha1_end(sha1_ctx_t *ctx, void *resbuf)
858 /* SHA stores total in BE, need to swap on LE arches: */
859 common64_end(ctx, /*swap_needed:*/ BB_LITTLE_ENDIAN);
861 hash_size = (ctx->process_block == sha1_process_block64) ? 5 : 8;
862 /* This way we do not impose alignment constraints on resbuf: */
863 if (BB_LITTLE_ENDIAN) {
865 for (i = 0; i < hash_size; ++i)
866 ctx->hash[i] = SWAP_BE32(ctx->hash[i]);
868 memcpy(resbuf, ctx->hash, sizeof(ctx->hash[0]) * hash_size);
871 void FAST_FUNC sha512_end(sha512_ctx_t *ctx, void *resbuf)
873 unsigned bufpos = ctx->total64[0] & 127;
875 /* Pad the buffer to the next 128-byte boundary with 0x80,0,0,0... */
876 ctx->wbuffer[bufpos++] = 0x80;
879 unsigned remaining = 128 - bufpos;
880 memset(ctx->wbuffer + bufpos, 0, remaining);
881 if (remaining >= 16) {
882 /* Store the 128-bit counter of bits in the buffer in BE format */
884 t = ctx->total64[0] << 3;
886 *(bb__aliased_uint64_t *) (&ctx->wbuffer[128 - 8]) = t;
887 t = (ctx->total64[1] << 3) | (ctx->total64[0] >> 61);
889 *(bb__aliased_uint64_t *) (&ctx->wbuffer[128 - 16]) = t;
891 sha512_process_block128(ctx);
897 if (BB_LITTLE_ENDIAN) {
899 for (i = 0; i < ARRAY_SIZE(ctx->hash); ++i)
900 ctx->hash[i] = SWAP_BE64(ctx->hash[i]);
902 memcpy(resbuf, ctx->hash, sizeof(ctx->hash));
907 * The Keccak sponge function, designed by Guido Bertoni, Joan Daemen,
908 * Michael Peeters and Gilles Van Assche. For more information, feedback or
909 * questions, please refer to our website: http://keccak.noekeon.org/
911 * Implementation by Ronny Van Keer,
912 * hereby denoted as "the implementer".
914 * To the extent possible under law, the implementer has waived all copyright
915 * and related or neighboring rights to the source code in this file.
916 * http://creativecommons.org/publicdomain/zero/1.0/
918 * Busybox modifications (C) Lauri Kasanen, under the GPLv2.
921 #if CONFIG_SHA3_SMALL < 0
922 # define SHA3_SMALL 0
923 #elif CONFIG_SHA3_SMALL > 1
924 # define SHA3_SMALL 1
926 # define SHA3_SMALL CONFIG_SHA3_SMALL
929 #define OPTIMIZE_SHA3_FOR_32 0
931 * SHA3 can be optimized for 32-bit CPUs with bit-slicing:
932 * every 64-bit word of state[] can be split into two 32-bit words
933 * by even/odd bits. In this form, all rotations of sha3 round
934 * are 32-bit - and there are lots of them.
935 * However, it requires either splitting/combining state words
936 * before/after sha3 round (code does this now)
937 * or shuffling bits before xor'ing them into state and in sha3_end.
938 * Without shuffling, bit-slicing results in -130 bytes of code
939 * and marginal speedup (but of course it gives wrong result).
940 * With shuffling it works, but +260 code bytes, and slower.
943 #if 0 /* LONG_MAX == 0x7fffffff */
944 # undef OPTIMIZE_SHA3_FOR_32
945 # define OPTIMIZE_SHA3_FOR_32 1
949 SHA3_IBLK_BYTES = 72, /* 576 bits / 8 */
952 #if OPTIMIZE_SHA3_FOR_32
953 /* This splits every 64-bit word into a pair of 32-bit words,
954 * even bits go into first word, odd bits go to second one.
955 * The conversion is done in-place.
957 static void split_halves(uint64_t *state)
959 /* Credit: Henry S. Warren, Hacker's Delight, Addison-Wesley, 2002 */
960 uint32_t *s32 = (uint32_t*)state;
963 for (i = 24; i >= 0; --i) {
965 t = (x0 ^ (x0 >> 1)) & 0x22222222; x0 = x0 ^ t ^ (t << 1);
966 t = (x0 ^ (x0 >> 2)) & 0x0C0C0C0C; x0 = x0 ^ t ^ (t << 2);
967 t = (x0 ^ (x0 >> 4)) & 0x00F000F0; x0 = x0 ^ t ^ (t << 4);
968 t = (x0 ^ (x0 >> 8)) & 0x0000FF00; x0 = x0 ^ t ^ (t << 8);
970 t = (x1 ^ (x1 >> 1)) & 0x22222222; x1 = x1 ^ t ^ (t << 1);
971 t = (x1 ^ (x1 >> 2)) & 0x0C0C0C0C; x1 = x1 ^ t ^ (t << 2);
972 t = (x1 ^ (x1 >> 4)) & 0x00F000F0; x1 = x1 ^ t ^ (t << 4);
973 t = (x1 ^ (x1 >> 8)) & 0x0000FF00; x1 = x1 ^ t ^ (t << 8);
974 *s32++ = (x0 & 0x0000FFFF) | (x1 << 16);
975 *s32++ = (x0 >> 16) | (x1 & 0xFFFF0000);
978 /* The reverse operation */
979 static void combine_halves(uint64_t *state)
981 uint32_t *s32 = (uint32_t*)state;
984 for (i = 24; i >= 0; --i) {
987 t = (x0 & 0x0000FFFF) | (x1 << 16);
988 x1 = (x0 >> 16) | (x1 & 0xFFFF0000);
990 t = (x0 ^ (x0 >> 8)) & 0x0000FF00; x0 = x0 ^ t ^ (t << 8);
991 t = (x0 ^ (x0 >> 4)) & 0x00F000F0; x0 = x0 ^ t ^ (t << 4);
992 t = (x0 ^ (x0 >> 2)) & 0x0C0C0C0C; x0 = x0 ^ t ^ (t << 2);
993 t = (x0 ^ (x0 >> 1)) & 0x22222222; x0 = x0 ^ t ^ (t << 1);
995 t = (x1 ^ (x1 >> 8)) & 0x0000FF00; x1 = x1 ^ t ^ (t << 8);
996 t = (x1 ^ (x1 >> 4)) & 0x00F000F0; x1 = x1 ^ t ^ (t << 4);
997 t = (x1 ^ (x1 >> 2)) & 0x0C0C0C0C; x1 = x1 ^ t ^ (t << 2);
998 t = (x1 ^ (x1 >> 1)) & 0x22222222; x1 = x1 ^ t ^ (t << 1);
1005 * In the crypto literature this function is usually called Keccak-f().
1007 static void sha3_process_block72(uint64_t *state)
1009 enum { NROUNDS = 24 };
1011 #if OPTIMIZE_SHA3_FOR_32
1013 static const uint32_t IOTA_CONST_0[NROUNDS] = {
1039 ** bits are in lsb: 0101 0000 1111 0100 1111 0001
1041 uint32_t IOTA_CONST_0bits = (uint32_t)(0x0050f4f1);
1042 static const uint32_t IOTA_CONST_1[NROUNDS] = {
1069 uint32_t *const s32 = (uint32_t*)state;
1072 split_halves(state);
1074 for (round = 0; round < NROUNDS; round++) {
1080 for (x = 0; x < 10; ++x) {
1081 BC[x+10] = BC[x] = s32[x]^s32[x+10]^s32[x+20]^s32[x+30]^s32[x+40];
1083 for (x = 0; x < 10; x += 2) {
1085 ta = BC[x+8] ^ rotl32(BC[x+3], 1);
1086 tb = BC[x+9] ^ BC[x+2];
1101 uint32_t t0a,t0b, t1a,t1b;
1105 #define RhoPi(PI_LANE, ROT_CONST) \
1106 t0a = s32[PI_LANE*2+0];\
1107 t0b = s32[PI_LANE*2+1];\
1108 if (ROT_CONST & 1) {\
1109 s32[PI_LANE*2+0] = rotl32(t1b, ROT_CONST/2+1);\
1110 s32[PI_LANE*2+1] = ROT_CONST == 1 ? t1a : rotl32(t1a, ROT_CONST/2+0);\
1112 s32[PI_LANE*2+0] = rotl32(t1a, ROT_CONST/2);\
1113 s32[PI_LANE*2+1] = rotl32(t1b, ROT_CONST/2);\
1115 t1a = t0a; t1b = t0b;
1144 for (x = 0; x <= 40;) {
1145 uint32_t BC0, BC1, BC2, BC3, BC4;
1149 s32[x + 0*2] = BC0 ^ ((~BC1) & BC2);
1151 s32[x + 1*2] = BC1 ^ ((~BC2) & BC3);
1153 s32[x + 2*2] = BC2 ^ ((~BC3) & BC4);
1154 s32[x + 3*2] = BC3 ^ ((~BC4) & BC0);
1155 s32[x + 4*2] = BC4 ^ ((~BC0) & BC1);
1160 s32[x + 0*2] = BC0 ^ ((~BC1) & BC2);
1162 s32[x + 1*2] = BC1 ^ ((~BC2) & BC3);
1164 s32[x + 2*2] = BC2 ^ ((~BC3) & BC4);
1165 s32[x + 3*2] = BC3 ^ ((~BC4) & BC0);
1166 s32[x + 4*2] = BC4 ^ ((~BC0) & BC1);
1170 s32[0] ^= IOTA_CONST_0bits & 1;
1171 IOTA_CONST_0bits >>= 1;
1172 s32[1] ^= IOTA_CONST_1[round];
1175 combine_halves(state);
1177 /* Native 64-bit algorithm */
1178 static const uint16_t IOTA_CONST[NROUNDS] = {
1179 /* Elements should be 64-bit, but top half is always zero
1180 * or 0x80000000. We encode 63rd bits in a separate word below.
1181 * Same is true for 31th bits, which lets us use 16-bit table
1182 * instead of 64-bit. The speed penalty is lost in the noise.
1209 /* bit for CONST[0] is in msb: 0011 0011 0000 0111 1101 1101 */
1210 const uint32_t IOTA_CONST_bit63 = (uint32_t)(0x3307dd00);
1211 /* bit for CONST[0] is in msb: 0001 0110 0011 1000 0001 1011 */
1212 const uint32_t IOTA_CONST_bit31 = (uint32_t)(0x16381b00);
1214 static const uint8_t ROT_CONST[24] = {
1215 1, 3, 6, 10, 15, 21, 28, 36, 45, 55, 2, 14,
1216 27, 41, 56, 8, 25, 43, 62, 18, 39, 61, 20, 44,
1218 static const uint8_t PI_LANE[24] = {
1219 10, 7, 11, 17, 18, 3, 5, 16, 8, 21, 24, 4,
1220 15, 23, 19, 13, 12, 2, 20, 14, 22, 9, 6, 1,
1222 /*static const uint8_t MOD5[10] = { 0, 1, 2, 3, 4, 0, 1, 2, 3, 4, };*/
1227 if (BB_BIG_ENDIAN) {
1228 for (x = 0; x < 25; x++) {
1229 state[x] = SWAP_LE64(state[x]);
1233 for (round = 0; round < NROUNDS; ++round) {
1237 for (x = 0; x < 5; ++x) {
1238 BC[x + 5] = BC[x] = state[x]
1239 ^ state[x + 5] ^ state[x + 10]
1240 ^ state[x + 15] ^ state[x + 20];
1242 /* Using 2x5 vector above eliminates the need to use
1243 * BC[MOD5[x+N]] trick below to fetch BC[(x+N) % 5],
1244 * and the code is a bit _smaller_.
1246 for (x = 0; x < 5; ++x) {
1247 uint64_t temp = BC[x + 4] ^ rotl64(BC[x + 1], 1);
1249 state[x + 5] ^= temp;
1250 state[x + 10] ^= temp;
1251 state[x + 15] ^= temp;
1252 state[x + 20] ^= temp;
1258 uint64_t t1 = state[1];
1259 for (x = 0; x < 24; ++x) {
1260 uint64_t t0 = state[PI_LANE[x]];
1261 state[PI_LANE[x]] = rotl64(t1, ROT_CONST[x]);
1265 /* Especially large benefit for 32-bit arch (75% faster):
1266 * 64-bit rotations by non-constant usually are SLOW on those.
1267 * We resort to unrolling here.
1268 * This optimizes out PI_LANE[] and ROT_CONST[],
1269 * but generates 300-500 more bytes of code.
1272 uint64_t t1 = state[1];
1273 #define RhoPi_twice(x) \
1274 t0 = state[PI_LANE[x ]]; \
1275 state[PI_LANE[x ]] = rotl64(t1, ROT_CONST[x ]); \
1276 t1 = state[PI_LANE[x+1]]; \
1277 state[PI_LANE[x+1]] = rotl64(t0, ROT_CONST[x+1]);
1278 RhoPi_twice(0); RhoPi_twice(2);
1279 RhoPi_twice(4); RhoPi_twice(6);
1280 RhoPi_twice(8); RhoPi_twice(10);
1281 RhoPi_twice(12); RhoPi_twice(14);
1282 RhoPi_twice(16); RhoPi_twice(18);
1283 RhoPi_twice(20); RhoPi_twice(22);
1287 # if LONG_MAX > 0x7fffffff
1288 for (x = 0; x <= 20; x += 5) {
1289 uint64_t BC0, BC1, BC2, BC3, BC4;
1293 state[x + 0] = BC0 ^ ((~BC1) & BC2);
1295 state[x + 1] = BC1 ^ ((~BC2) & BC3);
1297 state[x + 2] = BC2 ^ ((~BC3) & BC4);
1298 state[x + 3] = BC3 ^ ((~BC4) & BC0);
1299 state[x + 4] = BC4 ^ ((~BC0) & BC1);
1302 /* Reduced register pressure version
1303 * for register-starved 32-bit arches
1304 * (i386: -95 bytes, and it is _faster_)
1306 for (x = 0; x <= 40;) {
1307 uint32_t BC0, BC1, BC2, BC3, BC4;
1308 uint32_t *const s32 = (uint32_t*)state;
1315 s32[x + 0*2] = BC0 ^ ((~BC1) & BC2);
1317 s32[x + 1*2] = BC1 ^ ((~BC2) & BC3);
1319 s32[x + 2*2] = BC2 ^ ((~BC3) & BC4);
1320 s32[x + 3*2] = BC3 ^ ((~BC4) & BC0);
1321 s32[x + 4*2] = BC4 ^ ((~BC0) & BC1);
1331 s32[x + 0*2] = BC0 ^ ((~BC1) & BC2);
1333 s32[x + 1*2] = BC1 ^ ((~BC2) & BC3);
1335 s32[x + 2*2] = BC2 ^ ((~BC3) & BC4);
1336 s32[x + 3*2] = BC3 ^ ((~BC4) & BC0);
1337 s32[x + 4*2] = BC4 ^ ((~BC0) & BC1);
1341 # endif /* long is 32-bit */
1343 state[0] ^= IOTA_CONST[round]
1344 | (uint32_t)((IOTA_CONST_bit31 << round) & 0x80000000)
1345 | (uint64_t)((IOTA_CONST_bit63 << round) & 0x80000000) << 32;
1348 if (BB_BIG_ENDIAN) {
1349 for (x = 0; x < 25; x++) {
1350 state[x] = SWAP_LE64(state[x]);
1356 void FAST_FUNC sha3_begin(sha3_ctx_t *ctx)
1358 memset(ctx, 0, sizeof(*ctx));
1361 void FAST_FUNC sha3_hash(sha3_ctx_t *ctx, const void *buffer, size_t len)
1364 const uint8_t *data = buffer;
1365 unsigned bufpos = ctx->bytes_queued;
1368 unsigned remaining = SHA3_IBLK_BYTES - bufpos;
1369 if (remaining > len)
1372 /* XOR data into buffer */
1373 while (remaining != 0) {
1374 uint8_t *buf = (uint8_t*)ctx->state;
1375 buf[bufpos] ^= *data++;
1379 /* Clever way to do "if (bufpos != N) break; ... ; bufpos = 0;" */
1380 bufpos -= SHA3_IBLK_BYTES;
1383 /* Buffer is filled up, process it */
1384 sha3_process_block72(ctx->state);
1385 /*bufpos = 0; - already is */
1387 ctx->bytes_queued = bufpos + SHA3_IBLK_BYTES;
1389 /* +50 bytes code size, but a bit faster because of long-sized XORs */
1390 const uint8_t *data = buffer;
1391 unsigned bufpos = ctx->bytes_queued;
1393 /* If already data in queue, continue queuing first */
1394 while (len != 0 && bufpos != 0) {
1395 uint8_t *buf = (uint8_t*)ctx->state;
1396 buf[bufpos] ^= *data++;
1399 if (bufpos == SHA3_IBLK_BYTES) {
1405 /* Absorb complete blocks */
1406 while (len >= SHA3_IBLK_BYTES) {
1407 /* XOR data onto beginning of state[].
1408 * We try to be efficient - operate one word at a time, not byte.
1409 * Careful wrt unaligned access: can't just use "*(long*)data"!
1411 unsigned count = SHA3_IBLK_BYTES / sizeof(long);
1412 long *buf = (long*)ctx->state;
1415 move_from_unaligned_long(v, (long*)data);
1417 data += sizeof(long);
1419 len -= SHA3_IBLK_BYTES;
1421 sha3_process_block72(ctx->state);
1424 /* Queue remaining data bytes */
1426 uint8_t *buf = (uint8_t*)ctx->state;
1427 buf[bufpos] ^= *data++;
1432 ctx->bytes_queued = bufpos;
1436 void FAST_FUNC sha3_end(sha3_ctx_t *ctx, void *resbuf)
1439 uint8_t *buf = (uint8_t*)ctx->state;
1440 buf[ctx->bytes_queued] ^= 1;
1441 buf[SHA3_IBLK_BYTES - 1] ^= 0x80;
1443 sha3_process_block72(ctx->state);
1446 memcpy(resbuf, ctx->state, 64);