1 /* vi: set sw=4 ts=4: */
3 * Based on shasum from http://www.netsw.org/crypto/hash/
4 * Majorly hacked up to use Dr Brian Gladman's sha1 code
6 * Copyright (C) 2002 Dr Brian Gladman <brg@gladman.me.uk>, Worcester, UK.
7 * Copyright (C) 2003 Glenn L. McGrath
8 * Copyright (C) 2003 Erik Andersen
10 * Licensed under GPLv2 or later, see file LICENSE in this tarball for details.
12 * ---------------------------------------------------------------------------
13 * Issue Date: 10/11/2002
15 * This is a byte oriented version of SHA1 that operates on arrays of bytes
16 * stored in memory. It runs at 22 cycles per byte on a Pentium P4 processor
18 * ---------------------------------------------------------------------------
20 * SHA256 and SHA512 parts are:
21 * Released into the Public Domain by Ulrich Drepper <drepper@redhat.com>.
22 * Shrank by Denys Vlasenko.
24 * ---------------------------------------------------------------------------
26 * The best way to test random blocksizes is to go to coreutils/md5_sha1_sum.c
27 * and replace "4096" with something like "2000 + time(NULL) % 2097",
28 * then rebuild and compare "shaNNNsum bigfile" results.
33 #define rotl32(x,n) (((x) << (n)) | ((x) >> (32 - (n))))
34 #define rotr32(x,n) (((x) >> (n)) | ((x) << (32 - (n))))
36 #define rotr64(x,n) (((x) >> (n)) | ((x) << (64 - (n))))
38 static inline uint64_t hton64(uint64_t v)
40 return (((uint64_t)htonl(v)) << 32) | htonl(v >> 32);
45 #define ntoh64(v) hton64(v)
47 /* To check alignment gcc has an appropriate operator. Other
49 #if defined(__GNUC__) && __GNUC__ >= 2
50 # define UNALIGNED_P(p,type) (((uintptr_t) p) % __alignof__(type) != 0)
52 # define UNALIGNED_P(p,type) (((uintptr_t) p) % sizeof(type) != 0)
56 static void FAST_FUNC sha1_process_block64(sha1_ctx_t *ctx)
59 uint32_t W[80], a, b, c, d, e;
60 const uint32_t *words = (uint32_t*) ctx->wbuffer;
62 for (t = 0; t < 16; ++t) {
67 for (/*t = 16*/; t < 80; ++t) {
68 uint32_t T = W[t - 3] ^ W[t - 8] ^ W[t - 14] ^ W[t - 16];
78 /* Reverse byte order in 32-bit words */
79 #define ch(x,y,z) ((z) ^ ((x) & ((y) ^ (z))))
80 #define parity(x,y,z) ((x) ^ (y) ^ (z))
81 #define maj(x,y,z) (((x) & (y)) | ((z) & ((x) | (y))))
82 /* A normal version as set out in the FIPS. This version uses */
83 /* partial loop unrolling and is optimised for the Pentium 4 */
87 a = rotl32(a, 5) + f(b, c, d) + e + k + W[t]; \
94 for (t = 0; t < 20; ++t)
97 for (/*t = 20*/; t < 40; ++t)
98 rnd(parity, 0x6ed9eba1);
100 for (/*t = 40*/; t < 60; ++t)
101 rnd(maj, 0x8f1bbcdc);
103 for (/*t = 60*/; t < 80; ++t)
104 rnd(parity, 0xca62c1d6);
117 /* Constants for SHA512 from FIPS 180-2:4.2.3.
118 * SHA256 constants from FIPS 180-2:4.2.2
119 * are the most significant half of first 64 elements
122 static const uint64_t sha_K[80] = {
123 0x428a2f98d728ae22ULL, 0x7137449123ef65cdULL,
124 0xb5c0fbcfec4d3b2fULL, 0xe9b5dba58189dbbcULL,
125 0x3956c25bf348b538ULL, 0x59f111f1b605d019ULL,
126 0x923f82a4af194f9bULL, 0xab1c5ed5da6d8118ULL,
127 0xd807aa98a3030242ULL, 0x12835b0145706fbeULL,
128 0x243185be4ee4b28cULL, 0x550c7dc3d5ffb4e2ULL,
129 0x72be5d74f27b896fULL, 0x80deb1fe3b1696b1ULL,
130 0x9bdc06a725c71235ULL, 0xc19bf174cf692694ULL,
131 0xe49b69c19ef14ad2ULL, 0xefbe4786384f25e3ULL,
132 0x0fc19dc68b8cd5b5ULL, 0x240ca1cc77ac9c65ULL,
133 0x2de92c6f592b0275ULL, 0x4a7484aa6ea6e483ULL,
134 0x5cb0a9dcbd41fbd4ULL, 0x76f988da831153b5ULL,
135 0x983e5152ee66dfabULL, 0xa831c66d2db43210ULL,
136 0xb00327c898fb213fULL, 0xbf597fc7beef0ee4ULL,
137 0xc6e00bf33da88fc2ULL, 0xd5a79147930aa725ULL,
138 0x06ca6351e003826fULL, 0x142929670a0e6e70ULL,
139 0x27b70a8546d22ffcULL, 0x2e1b21385c26c926ULL,
140 0x4d2c6dfc5ac42aedULL, 0x53380d139d95b3dfULL,
141 0x650a73548baf63deULL, 0x766a0abb3c77b2a8ULL,
142 0x81c2c92e47edaee6ULL, 0x92722c851482353bULL,
143 0xa2bfe8a14cf10364ULL, 0xa81a664bbc423001ULL,
144 0xc24b8b70d0f89791ULL, 0xc76c51a30654be30ULL,
145 0xd192e819d6ef5218ULL, 0xd69906245565a910ULL,
146 0xf40e35855771202aULL, 0x106aa07032bbd1b8ULL,
147 0x19a4c116b8d2d0c8ULL, 0x1e376c085141ab53ULL,
148 0x2748774cdf8eeb99ULL, 0x34b0bcb5e19b48a8ULL,
149 0x391c0cb3c5c95a63ULL, 0x4ed8aa4ae3418acbULL,
150 0x5b9cca4f7763e373ULL, 0x682e6ff3d6b2b8a3ULL,
151 0x748f82ee5defb2fcULL, 0x78a5636f43172f60ULL,
152 0x84c87814a1f0ab72ULL, 0x8cc702081a6439ecULL,
153 0x90befffa23631e28ULL, 0xa4506cebde82bde9ULL,
154 0xbef9a3f7b2c67915ULL, 0xc67178f2e372532bULL,
155 0xca273eceea26619cULL, 0xd186b8c721c0c207ULL, /* [64]+ are used for sha512 only */
156 0xeada7dd6cde0eb1eULL, 0xf57d4f7fee6ed178ULL,
157 0x06f067aa72176fbaULL, 0x0a637dc5a2c898a6ULL,
158 0x113f9804bef90daeULL, 0x1b710b35131c471bULL,
159 0x28db77f523047d84ULL, 0x32caab7b40c72493ULL,
160 0x3c9ebe0a15c9bebcULL, 0x431d67c49c100d4cULL,
161 0x4cc5d4becb3e42b6ULL, 0x597f299cfc657e2aULL,
162 0x5fcb6fab3ad6faecULL, 0x6c44198c4a475817ULL
165 static void FAST_FUNC sha256_process_block64(sha256_ctx_t *ctx)
168 uint32_t W[64], a, b, c, d, e, f, g, h;
169 const uint32_t *words = (uint32_t*) ctx->wbuffer;
171 /* Operators defined in FIPS 180-2:4.1.2. */
172 #define Ch(x, y, z) ((x & y) ^ (~x & z))
173 #define Maj(x, y, z) ((x & y) ^ (x & z) ^ (y & z))
174 #define S0(x) (rotr32(x, 2) ^ rotr32(x, 13) ^ rotr32(x, 22))
175 #define S1(x) (rotr32(x, 6) ^ rotr32(x, 11) ^ rotr32(x, 25))
176 #define R0(x) (rotr32(x, 7) ^ rotr32(x, 18) ^ (x >> 3))
177 #define R1(x) (rotr32(x, 17) ^ rotr32(x, 19) ^ (x >> 10))
179 /* Compute the message schedule according to FIPS 180-2:6.2.2 step 2. */
180 for (t = 0; t < 16; ++t) {
181 W[t] = ntohl(*words);
185 for (/*t = 16*/; t < 64; ++t)
186 W[t] = R1(W[t - 2]) + W[t - 7] + R0(W[t - 15]) + W[t - 16];
197 /* The actual computation according to FIPS 180-2:6.2.2 step 3. */
198 for (t = 0; t < 64; ++t) {
199 /* Need to fetch upper half of sha_K[t]
200 * (I hope compiler is clever enough to just fetch
203 uint32_t K_t = sha_K[t] >> 32;
204 uint32_t T1 = h + S1(e) + Ch(e, f, g) + K_t + W[t];
205 uint32_t T2 = S0(a) + Maj(a, b, c);
221 /* Add the starting values of the context according to FIPS 180-2:6.2.2
233 static void FAST_FUNC sha512_process_block128(sha512_ctx_t *ctx)
237 /* On i386, having assignments here (not later as sha256 does)
238 * produces 99 bytes smaller code with gcc 4.3.1
240 uint64_t a = ctx->hash[0];
241 uint64_t b = ctx->hash[1];
242 uint64_t c = ctx->hash[2];
243 uint64_t d = ctx->hash[3];
244 uint64_t e = ctx->hash[4];
245 uint64_t f = ctx->hash[5];
246 uint64_t g = ctx->hash[6];
247 uint64_t h = ctx->hash[7];
248 const uint64_t *words = (uint64_t*) ctx->wbuffer;
250 /* Operators defined in FIPS 180-2:4.1.2. */
251 #define Ch(x, y, z) ((x & y) ^ (~x & z))
252 #define Maj(x, y, z) ((x & y) ^ (x & z) ^ (y & z))
253 #define S0(x) (rotr64(x, 28) ^ rotr64(x, 34) ^ rotr64(x, 39))
254 #define S1(x) (rotr64(x, 14) ^ rotr64(x, 18) ^ rotr64(x, 41))
255 #define R0(x) (rotr64(x, 1) ^ rotr64(x, 8) ^ (x >> 7))
256 #define R1(x) (rotr64(x, 19) ^ rotr64(x, 61) ^ (x >> 6))
258 /* Compute the message schedule according to FIPS 180-2:6.3.2 step 2. */
259 for (t = 0; t < 16; ++t) {
260 W[t] = ntoh64(*words);
263 for (/*t = 16*/; t < 80; ++t)
264 W[t] = R1(W[t - 2]) + W[t - 7] + R0(W[t - 15]) + W[t - 16];
266 /* The actual computation according to FIPS 180-2:6.3.2 step 3. */
267 for (t = 0; t < 80; ++t) {
268 uint64_t T1 = h + S1(e) + Ch(e, f, g) + sha_K[t] + W[t];
269 uint64_t T2 = S0(a) + Maj(a, b, c);
285 /* Add the starting values of the context according to FIPS 180-2:6.3.2
298 void FAST_FUNC sha1_begin(sha1_ctx_t *ctx)
300 ctx->hash[0] = 0x67452301;
301 ctx->hash[1] = 0xefcdab89;
302 ctx->hash[2] = 0x98badcfe;
303 ctx->hash[3] = 0x10325476;
304 ctx->hash[4] = 0xc3d2e1f0;
306 ctx->process_block = sha1_process_block64;
309 static const uint32_t init256[] = {
319 static const uint32_t init512_lo[] = {
330 /* Initialize structure containing state of computation.
331 (FIPS 180-2:5.3.2) */
332 void FAST_FUNC sha256_begin(sha256_ctx_t *ctx)
334 memcpy(ctx->hash, init256, sizeof(init256));
336 ctx->process_block = sha256_process_block64;
339 /* Initialize structure containing state of computation.
340 (FIPS 180-2:5.3.3) */
341 void FAST_FUNC sha512_begin(sha512_ctx_t *ctx)
344 for (i = 0; i < 8; i++)
345 ctx->hash[i] = ((uint64_t)(init256[i]) << 32) + init512_lo[i];
346 ctx->total64[0] = ctx->total64[1] = 0;
350 /* Used also for sha256 */
351 void FAST_FUNC sha1_hash(const void *buffer, size_t len, sha1_ctx_t *ctx)
353 unsigned in_buf = ctx->total64 & 63;
354 unsigned add = 64 - in_buf;
358 while (len >= add) { /* transfer whole blocks while possible */
359 memcpy(ctx->wbuffer + in_buf, buffer, add);
360 buffer = (const char *)buffer + add;
364 ctx->process_block(ctx);
367 memcpy(ctx->wbuffer + in_buf, buffer, len);
370 void FAST_FUNC sha512_hash(const void *buffer, size_t len, sha512_ctx_t *ctx)
372 unsigned in_buf = ctx->total64[0] & 127;
373 unsigned add = 128 - in_buf;
375 /* First increment the byte count. FIPS 180-2 specifies the possible
376 length of the file up to 2^128 _bits_.
377 We compute the number of _bytes_ and convert to bits later. */
378 ctx->total64[0] += len;
379 if (ctx->total64[0] < len)
382 while (len >= add) { /* transfer whole blocks while possible */
383 memcpy(ctx->wbuffer + in_buf, buffer, add);
384 buffer = (const char *)buffer + add;
388 sha512_process_block128(ctx);
391 memcpy(ctx->wbuffer + in_buf, buffer, len);
395 /* Used also for sha256 */
396 void FAST_FUNC sha1_end(void *resbuf, sha1_ctx_t *ctx)
398 unsigned i, pad, in_buf;
400 in_buf = ctx->total64 & 63;
401 /* Pad the buffer to the next 64-byte boundary with 0x80,0,0,0... */
402 ctx->wbuffer[in_buf++] = 0x80;
404 /* This loop iterates either once or twice, no more, no less */
407 memset(ctx->wbuffer + in_buf, 0, pad);
409 /* Do we have enough space for the length count? */
411 /* Store the 64-bit counter of bits in the buffer in BE format */
412 uint64_t t = ctx->total64 << 3;
414 /* wbuffer is suitably aligned for this */
415 *(uint64_t *) (&ctx->wbuffer[64 - 8]) = t;
417 ctx->process_block(ctx);
422 in_buf = (ctx->process_block == sha1_process_block64) ? 5 : 8;
423 /* This way we do not impose alignment constraints on resbuf: */
425 for (i = 0; i < in_buf; ++i)
426 ctx->hash[i] = htonl(ctx->hash[i]);
428 memcpy(resbuf, ctx->hash, sizeof(ctx->hash[0]) * in_buf);
431 void FAST_FUNC sha512_end(void *resbuf, sha512_ctx_t *ctx)
433 unsigned i, pad, in_buf;
435 in_buf = ctx->total64[0] & 127;
436 /* Pad the buffer to the next 128-byte boundary with 0x80,0,0,0...
439 ctx->wbuffer[in_buf++] = 0x80;
443 memset(ctx->wbuffer + in_buf, 0, pad);
446 /* Store the 128-bit counter of bits in the buffer in BE format */
448 t = ctx->total64[0] << 3;
450 *(uint64_t *) (&ctx->wbuffer[128 - 8]) = t;
451 t = (ctx->total64[1] << 3) | (ctx->total64[0] >> 61);
453 *(uint64_t *) (&ctx->wbuffer[128 - 16]) = t;
455 sha512_process_block128(ctx);
461 for (i = 0; i < ARRAY_SIZE(ctx->hash); ++i)
462 ctx->hash[i] = hton64(ctx->hash[i]);
464 memcpy(resbuf, ctx->hash, sizeof(ctx->hash));