crypt/md5.c

   1 /* Functions to compute MD5 message digest of files or memory blocks.
   2    according to the definition of MD5 in RFC 1321 from April 1992.
   3    Copyright (C) 1995-2012 Free Software Foundation, Inc.
   4    This file is part of the GNU C Library.
   5
   6    The GNU C Library is free software; you can redistribute it and/or
   7    modify it under the terms of the GNU Lesser General Public
   8    License as published by the Free Software Foundation; either
   9    version 2.1 of the License, or (at your option) any later version.
  10
  11    The GNU C Library is distributed in the hope that it will be useful,
  12    but WITHOUT ANY WARRANTY; without even the implied warranty of
  13    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14    Lesser General Public License for more details.
  15
  16    You should have received a copy of the GNU Lesser General Public
  17    License along with the GNU C Library; if not, see
  18    <http://www.gnu.org/licenses/>.  */
  19
  20 /* Written by Ulrich Drepper <drepper@gnu.ai.mit.edu>, 1995.  */
  21
  22 #ifdef HAVE_CONFIG_H
  23 # include <config.h>
  24 #endif
  25
  26 #include <sys/types.h>
  27
  28 #if STDC_HEADERS || defined _LIBC
  29 # include <stdlib.h>
  30 # include <string.h>
  31 #else
  32 # ifndef HAVE_MEMCPY
  33 #  define memcpy(d, s, n) (bcopy ((s), (d), (n)), (d))
  34 # endif
  35 #endif
  36
  37 #include "md5.h"
  38
  39 #ifdef _LIBC
  40 # include <endian.h>
  41 # if __BYTE_ORDER == __BIG_ENDIAN
  42 #  define WORDS_BIGENDIAN 1
  43 # endif
  44 /* We need to keep the namespace clean so define the MD5 function
  45    protected using leading __ .  */
  46 # define md5_init_ctx __md5_init_ctx
  47 # define md5_process_block __md5_process_block
  48 # define md5_process_bytes __md5_process_bytes
  49 # define md5_finish_ctx __md5_finish_ctx
  50 # define md5_read_ctx __md5_read_ctx
  51 # define md5_stream __md5_stream
  52 # define md5_buffer __md5_buffer
  53 #endif
  54
  55 #ifdef WORDS_BIGENDIAN
  56 # define SWAP(n)                                                        \
  57     (((n) << 24) | (((n) & 0xff00) << 8) | (((n) >> 8) & 0xff00) | ((n) >> 24))
  58 #else
  59 # define SWAP(n) (n)
  60 #endif
  61
  62
  63 /* This array contains the bytes used to pad the buffer to the next
  64    64-byte boundary.  (RFC 1321, 3.1: Step 1)  */
  65 static const unsigned char fillbuf[64] = { 0x80, 0 /* , 0, 0, ...  */ };
  66
  67
  68 /* Initialize structure containing state of computation.
  69    (RFC 1321, 3.3: Step 3)  */
  70 void
  71 md5_init_ctx (ctx)
  72      struct md5_ctx *ctx;
  73 {
  74   ctx->A = 0x67452301;
  75   ctx->B = 0xefcdab89;
  76   ctx->C = 0x98badcfe;
  77   ctx->D = 0x10325476;
  78
  79   ctx->total[0] = ctx->total[1] = 0;
  80   ctx->buflen = 0;
  81 }
  82
  83 /* Put result from CTX in first 16 bytes following RESBUF.  The result
  84    must be in little endian byte order.
  85
  86    IMPORTANT: On some systems it is required that RESBUF is correctly
  87    aligned for a 32 bits value.  */
  88 void *
  89 md5_read_ctx (ctx, resbuf)
  90      const struct md5_ctx *ctx;
  91      void *resbuf;
  92 {
  93   ((md5_uint32 *) resbuf)[0] = SWAP (ctx->A);
  94   ((md5_uint32 *) resbuf)[1] = SWAP (ctx->B);
  95   ((md5_uint32 *) resbuf)[2] = SWAP (ctx->C);
  96   ((md5_uint32 *) resbuf)[3] = SWAP (ctx->D);
  97
  98   return resbuf;
  99 }
 100
 101 /* Process the remaining bytes in the internal buffer and the usual
 102    prolog according to the standard and write the result to RESBUF.
 103
 104    IMPORTANT: On some systems it is required that RESBUF is correctly
 105    aligned for a 32 bits value.  */
 106 void *
 107 md5_finish_ctx (ctx, resbuf)
 108      struct md5_ctx *ctx;
 109      void *resbuf;
 110 {
 111   /* Take yet unprocessed bytes into account.  */
 112   md5_uint32 bytes = ctx->buflen;
 113   size_t pad;
 114
 115   /* Now count remaining bytes.  */
 116   ctx->total[0] += bytes;
 117   if (ctx->total[0] < bytes)
 118     ++ctx->total[1];
 119
 120   pad = bytes >= 56 ? 64 + 56 - bytes : 56 - bytes;
 121   memcpy (&ctx->buffer[bytes], fillbuf, pad);
 122
 123   /* Put the 64-bit file length in *bits* at the end of the buffer.  */
 124   ctx->buffer32[(bytes + pad) / 4] = SWAP (ctx->total[0] << 3);
 125   ctx->buffer32[(bytes + pad + 4) / 4] = SWAP ((ctx->total[1] << 3) |
 126                                                (ctx->total[0] >> 29));
 127
 128   /* Process last bytes.  */
 129   md5_process_block (ctx->buffer, bytes + pad + 8, ctx);
 130
 131   return md5_read_ctx (ctx, resbuf);
 132 }
 133
 134 /* Compute MD5 message digest for bytes read from STREAM.  The
 135    resulting message digest number will be written into the 16 bytes
 136    beginning at RESBLOCK.  */
 137 int
 138 md5_stream (stream, resblock)
 139      FILE *stream;
 140      void *resblock;
 141 {
 142   /* Important: BLOCKSIZE must be a multiple of 64.  */
 143 #define BLOCKSIZE 4096
 144   struct md5_ctx ctx;
 145   char buffer[BLOCKSIZE + 72];
 146   size_t sum;
 147
 148   /* Initialize the computation context.  */
 149   md5_init_ctx (&ctx);
 150
 151   /* Iterate over full file contents.  */
 152   while (1)
 153     {
 154       /* We read the file in blocks of BLOCKSIZE bytes.  One call of the
 155          computation function processes the whole buffer so that with the
 156          next round of the loop another block can be read.  */
 157       size_t n;
 158       sum = 0;
 159
 160       /* Read block.  Take care for partial reads.  */
 161       do
 162         {
 163           n = fread (buffer + sum, 1, BLOCKSIZE - sum, stream);
 164
 165           sum += n;
 166         }
 167       while (sum < BLOCKSIZE && n != 0);
 168       if (n == 0 && ferror (stream))
 169         return 1;
 170
 171       /* If end of file is reached, end the loop.  */
 172       if (n == 0)
 173         break;
 174
 175       /* Process buffer with BLOCKSIZE bytes.  Note that
 176                         BLOCKSIZE % 64 == 0
 177        */
 178       md5_process_block (buffer, BLOCKSIZE, &ctx);
 179     }
 180
 181   /* Add the last bytes if necessary.  */
 182   if (sum > 0)
 183     md5_process_bytes (buffer, sum, &ctx);
 184
 185   /* Construct result in desired memory.  */
 186   md5_finish_ctx (&ctx, resblock);
 187   return 0;
 188 }
 189
 190 /* Compute MD5 message digest for LEN bytes beginning at BUFFER.  The
 191    result is always in little endian byte order, so that a byte-wise
 192    output yields to the wanted ASCII representation of the message
 193    digest.  */
 194 void *
 195 md5_buffer (buffer, len, resblock)
 196      const char *buffer;
 197      size_t len;
 198      void *resblock;
 199 {
 200   struct md5_ctx ctx;
 201
 202   /* Initialize the computation context.  */
 203   md5_init_ctx (&ctx);
 204
 205   /* Process whole buffer but last len % 64 bytes.  */
 206   md5_process_bytes (buffer, len, &ctx);
 207
 208   /* Put result in desired memory area.  */
 209   return md5_finish_ctx (&ctx, resblock);
 210 }
 211
 212
 213 void
 214 md5_process_bytes (buffer, len, ctx)
 215      const void *buffer;
 216      size_t len;
 217      struct md5_ctx *ctx;
 218 {
 219   /* When we already have some bits in our internal buffer concatenate
 220      both inputs first.  */
 221   if (ctx->buflen != 0)
 222     {
 223       size_t left_over = ctx->buflen;
 224       size_t add = 128 - left_over > len ? len : 128 - left_over;
 225
 226       memcpy (&ctx->buffer[left_over], buffer, add);
 227       ctx->buflen += add;
 228
 229       if (ctx->buflen > 64)
 230         {
 231           md5_process_block (ctx->buffer, ctx->buflen & ~63, ctx);
 232
 233           ctx->buflen &= 63;
 234           /* The regions in the following copy operation cannot overlap.  */
 235           memcpy (ctx->buffer, &ctx->buffer[(left_over + add) & ~63],
 236                   ctx->buflen);
 237         }
 238
 239       buffer = (const char *) buffer + add;
 240       len -= add;
 241     }
 242
 243   /* Process available complete blocks.  */
 244   if (len >= 64)
 245     {
 246 #if !_STRING_ARCH_unaligned
 247 /* To check alignment gcc has an appropriate operator.  Other
 248    compilers don't.  */
 249 # if __GNUC__ >= 2
 250 #  define UNALIGNED_P(p) (((md5_uintptr) p) % __alignof__ (md5_uint32) != 0)
 251 # else
 252 #  define UNALIGNED_P(p) (((md5_uintptr) p) % sizeof (md5_uint32) != 0)
 253 # endif
 254       if (UNALIGNED_P (buffer))
 255         while (len > 64)
 256           {
 257             md5_process_block (memcpy (ctx->buffer, buffer, 64), 64, ctx);
 258             buffer = (const char *) buffer + 64;
 259             len -= 64;
 260           }
 261       else
 262 #endif
 263         {
 264           md5_process_block (buffer, len & ~63, ctx);
 265           buffer = (const char *) buffer + (len & ~63);
 266           len &= 63;
 267         }
 268     }
 269
 270   /* Move remaining bytes in internal buffer.  */
 271   if (len > 0)
 272     {
 273       size_t left_over = ctx->buflen;
 274
 275       memcpy (&ctx->buffer[left_over], buffer, len);
 276       left_over += len;
 277       if (left_over >= 64)
 278         {
 279           md5_process_block (ctx->buffer, 64, ctx);
 280           left_over -= 64;
 281           memcpy (ctx->buffer, &ctx->buffer[64], left_over);
 282         }
 283       ctx->buflen = left_over;
 284     }
 285 }
 286
 287
 288 /* These are the four functions used in the four steps of the MD5 algorithm
 289    and defined in the RFC 1321.  The first function is a little bit optimized
 290    (as found in Colin Plumbs public domain implementation).  */
 291 /* #define FF(b, c, d) ((b & c) | (~b & d)) */
 292 #define FF(b, c, d) (d ^ (b & (c ^ d)))
 293 #define FG(b, c, d) FF (d, b, c)
 294 #define FH(b, c, d) (b ^ c ^ d)
 295 #define FI(b, c, d) (c ^ (b | ~d))
 296
 297 /* Process LEN bytes of BUFFER, accumulating context into CTX.
 298    It is assumed that LEN % 64 == 0.  */
 299
 300 void
 301 md5_process_block (buffer, len, ctx)
 302      const void *buffer;
 303      size_t len;
 304      struct md5_ctx *ctx;
 305 {
 306   md5_uint32 correct_words[16];
 307   const md5_uint32 *words = buffer;
 308   size_t nwords = len / sizeof (md5_uint32);
 309   const md5_uint32 *endp = words + nwords;
 310   md5_uint32 A = ctx->A;
 311   md5_uint32 B = ctx->B;
 312   md5_uint32 C = ctx->C;
 313   md5_uint32 D = ctx->D;
 314   md5_uint32 lolen = len;
 315
 316   /* First increment the byte count.  RFC 1321 specifies the possible
 317      length of the file up to 2^64 bits.  Here we only compute the
 318      number of bytes.  Do a double word increment.  */
 319   ctx->total[0] += lolen;
 320   ctx->total[1] += (len >> 31 >> 1) + (ctx->total[0] < lolen);
 321
 322   /* Process all bytes in the buffer with 64 bytes in each round of
 323      the loop.  */
 324   while (words < endp)
 325     {
 326       md5_uint32 *cwp = correct_words;
 327       md5_uint32 A_save = A;
 328       md5_uint32 B_save = B;
 329       md5_uint32 C_save = C;
 330       md5_uint32 D_save = D;
 331
 332       /* First round: using the given function, the context and a constant
 333          the next context is computed.  Because the algorithms processing
 334          unit is a 32-bit word and it is determined to work on words in
 335          little endian byte order we perhaps have to change the byte order
 336          before the computation.  To reduce the work for the next steps
 337          we store the swapped words in the array CORRECT_WORDS.  */
 338
 339 #define OP(a, b, c, d, s, T)                                            \
 340       do                                                                \
 341         {                                                               \
 342           a += FF (b, c, d) + (*cwp++ = SWAP (*words)) + T;             \
 343           ++words;                                                      \
 344           CYCLIC (a, s);                                                \
 345           a += b;                                                       \
 346         }                                                               \
 347       while (0)
 348
 349       /* It is unfortunate that C does not provide an operator for
 350          cyclic rotation.  Hope the C compiler is smart enough.  */
 351 #define CYCLIC(w, s) (w = (w << s) | (w >> (32 - s)))
 352
 353       /* Before we start, one word to the strange constants.
 354          They are defined in RFC 1321 as
 355
 356          T[i] = (int) (4294967296.0 * fabs (sin (i))), i=1..64
 357        */
 358
 359       /* Round 1.  */
 360       OP (A, B, C, D,  7, 0xd76aa478);
 361       OP (D, A, B, C, 12, 0xe8c7b756);
 362       OP (C, D, A, B, 17, 0x242070db);
 363       OP (B, C, D, A, 22, 0xc1bdceee);
 364       OP (A, B, C, D,  7, 0xf57c0faf);
 365       OP (D, A, B, C, 12, 0x4787c62a);
 366       OP (C, D, A, B, 17, 0xa8304613);
 367       OP (B, C, D, A, 22, 0xfd469501);
 368       OP (A, B, C, D,  7, 0x698098d8);
 369       OP (D, A, B, C, 12, 0x8b44f7af);
 370       OP (C, D, A, B, 17, 0xffff5bb1);
 371       OP (B, C, D, A, 22, 0x895cd7be);
 372       OP (A, B, C, D,  7, 0x6b901122);
 373       OP (D, A, B, C, 12, 0xfd987193);
 374       OP (C, D, A, B, 17, 0xa679438e);
 375       OP (B, C, D, A, 22, 0x49b40821);
 376
 377       /* For the second to fourth round we have the possibly swapped words
 378          in CORRECT_WORDS.  Redefine the macro to take an additional first
 379          argument specifying the function to use.  */
 380 #undef OP
 381 #define OP(f, a, b, c, d, k, s, T)                                      \
 382       do                                                                \
 383         {                                                               \
 384           a += f (b, c, d) + correct_words[k] + T;                      \
 385           CYCLIC (a, s);                                                \
 386           a += b;                                                       \
 387         }                                                               \
 388       while (0)
 389
 390       /* Round 2.  */
 391       OP (FG, A, B, C, D,  1,  5, 0xf61e2562);
 392       OP (FG, D, A, B, C,  6,  9, 0xc040b340);
 393       OP (FG, C, D, A, B, 11, 14, 0x265e5a51);
 394       OP (FG, B, C, D, A,  0, 20, 0xe9b6c7aa);
 395       OP (FG, A, B, C, D,  5,  5, 0xd62f105d);
 396       OP (FG, D, A, B, C, 10,  9, 0x02441453);
 397       OP (FG, C, D, A, B, 15, 14, 0xd8a1e681);
 398       OP (FG, B, C, D, A,  4, 20, 0xe7d3fbc8);
 399       OP (FG, A, B, C, D,  9,  5, 0x21e1cde6);
 400       OP (FG, D, A, B, C, 14,  9, 0xc33707d6);
 401       OP (FG, C, D, A, B,  3, 14, 0xf4d50d87);
 402       OP (FG, B, C, D, A,  8, 20, 0x455a14ed);
 403       OP (FG, A, B, C, D, 13,  5, 0xa9e3e905);
 404       OP (FG, D, A, B, C,  2,  9, 0xfcefa3f8);
 405       OP (FG, C, D, A, B,  7, 14, 0x676f02d9);
 406       OP (FG, B, C, D, A, 12, 20, 0x8d2a4c8a);
 407
 408       /* Round 3.  */
 409       OP (FH, A, B, C, D,  5,  4, 0xfffa3942);
 410       OP (FH, D, A, B, C,  8, 11, 0x8771f681);
 411       OP (FH, C, D, A, B, 11, 16, 0x6d9d6122);
 412       OP (FH, B, C, D, A, 14, 23, 0xfde5380c);
 413       OP (FH, A, B, C, D,  1,  4, 0xa4beea44);
 414       OP (FH, D, A, B, C,  4, 11, 0x4bdecfa9);
 415       OP (FH, C, D, A, B,  7, 16, 0xf6bb4b60);
 416       OP (FH, B, C, D, A, 10, 23, 0xbebfbc70);
 417       OP (FH, A, B, C, D, 13,  4, 0x289b7ec6);
 418       OP (FH, D, A, B, C,  0, 11, 0xeaa127fa);
 419       OP (FH, C, D, A, B,  3, 16, 0xd4ef3085);
 420       OP (FH, B, C, D, A,  6, 23, 0x04881d05);
 421       OP (FH, A, B, C, D,  9,  4, 0xd9d4d039);
 422       OP (FH, D, A, B, C, 12, 11, 0xe6db99e5);
 423       OP (FH, C, D, A, B, 15, 16, 0x1fa27cf8);
 424       OP (FH, B, C, D, A,  2, 23, 0xc4ac5665);
 425
 426       /* Round 4.  */
 427       OP (FI, A, B, C, D,  0,  6, 0xf4292244);
 428       OP (FI, D, A, B, C,  7, 10, 0x432aff97);
 429       OP (FI, C, D, A, B, 14, 15, 0xab9423a7);
 430       OP (FI, B, C, D, A,  5, 21, 0xfc93a039);
 431       OP (FI, A, B, C, D, 12,  6, 0x655b59c3);
 432       OP (FI, D, A, B, C,  3, 10, 0x8f0ccc92);
 433       OP (FI, C, D, A, B, 10, 15, 0xffeff47d);
 434       OP (FI, B, C, D, A,  1, 21, 0x85845dd1);
 435       OP (FI, A, B, C, D,  8,  6, 0x6fa87e4f);
 436       OP (FI, D, A, B, C, 15, 10, 0xfe2ce6e0);
 437       OP (FI, C, D, A, B,  6, 15, 0xa3014314);
 438       OP (FI, B, C, D, A, 13, 21, 0x4e0811a1);
 439       OP (FI, A, B, C, D,  4,  6, 0xf7537e82);
 440       OP (FI, D, A, B, C, 11, 10, 0xbd3af235);
 441       OP (FI, C, D, A, B,  2, 15, 0x2ad7d2bb);
 442       OP (FI, B, C, D, A,  9, 21, 0xeb86d391);
 443
 444       /* Add the starting values of the context.  */
 445       A += A_save;
 446       B += B_save;
 447       C += C_save;
 448       D += D_save;
 449     }
 450
 451   /* Put checksum in context given as argument.  */
 452   ctx->A = A;
 453   ctx->B = B;
 454   ctx->C = C;
 455   ctx->D = D;
 456 }