src/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, 1996, 1997, 1999, 2000 Free Software Foundation, Inc.
   4    This file is part of the GNU Emacs.
   5
   6    The GNU C Library is free software; you can redistribute it and/or
   7    modify it under the terms of the GNU Library General Public License as
   8    published by the Free Software Foundation; either version 2 of the
   9    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    Library General Public License for more details.
  15
  16    You should have received a copy of the GNU Library General Public
  17    License along with the GNU C Library; see the file COPYING.LIB.  If not,
  18    write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  19    Boston, MA 02111-1307, USA.  */
  20
  21 /* Written by Ulrich Drepper <drepper@gnu.ai.mit.edu>, 1995.  */
  22
  23 #ifdef HAVE_CONFIG_H
  24 # include <config.h>
  25 #endif
  26
  27 #include <sys/types.h>
  28
  29 #if STDC_HEADERS || defined _LIBC
  30 # include <stdlib.h>
  31 # include <string.h>
  32 #else
  33 # ifndef HAVE_MEMCPY
  34 #  define memcpy(d, s, n) bcopy ((s), (d), (n))
  35 # endif
  36 #endif
  37
  38 #ifdef _LIBC
  39 # include <endian.h>
  40 # if __BYTE_ORDER == __BIG_ENDIAN
  41 #  define WORDS_BIG_ENDIAN 1
  42 # endif
  43 /* We need to keep the namespace clean so define the MD5 function
  44    protected using leading __ .  */
  45 # define md5_init_ctx __md5_init_ctx
  46 # define md5_process_block __md5_process_block
  47 # define md5_process_bytes __md5_process_bytes
  48 # define md5_finish_ctx __md5_finish_ctx
  49 # define md5_read_ctx __md5_read_ctx
  50 # define md5_stream __md5_stream
  51 # define md5_buffer __md5_buffer
  52 #endif
  53
  54 #include "md5.h"
  55
  56 #ifdef WORDS_BIG_ENDIAN
  57 # define SWAP(n)                                                        \
  58     (((n) << 24) | (((n) & 0xff00) << 8) | (((n) >> 8) & 0xff00) | ((n) >> 24))
  59 #else
  60 # define SWAP(n) (n)
  61 #endif
  62
  63
  64 /* This array contains the bytes used to pad the buffer to the next
  65    64-byte boundary.  (RFC 1321, 3.1: Step 1)  */
  66 static const unsigned char fillbuf[64] = { 0x80, 0 /* , 0, 0, ...  */ };
  67
  68
  69 /* Initialize structure containing state of computation.
  70    (RFC 1321, 3.3: Step 3)  */
  71 void
  72 md5_init_ctx (ctx)
  73      struct md5_ctx *ctx;
  74 {
  75   ctx->A = 0x67452301;
  76   ctx->B = 0xefcdab89;
  77   ctx->C = 0x98badcfe;
  78   ctx->D = 0x10325476;
  79
  80   ctx->total[0] = ctx->total[1] = 0;
  81   ctx->buflen = 0;
  82 }
  83
  84 /* Put result from CTX in first 16 bytes following RESBUF.  The result
  85    must be in little endian byte order.
  86
  87    IMPORTANT: On some systems it is required that RESBUF is correctly
  88    aligned for a 32 bits value.  */
  89 void *
  90 md5_read_ctx (ctx, resbuf)
  91      const struct md5_ctx *ctx;
  92      void *resbuf;
  93 {
  94   ((md5_uint32 *) resbuf)[0] = SWAP (ctx->A);
  95   ((md5_uint32 *) resbuf)[1] = SWAP (ctx->B);
  96   ((md5_uint32 *) resbuf)[2] = SWAP (ctx->C);
  97   ((md5_uint32 *) resbuf)[3] = SWAP (ctx->D);
  98
  99   return resbuf;
 100 }
 101
 102 /* Process the remaining bytes in the internal buffer and the usual
 103    prolog according to the standard and write the result to RESBUF.
 104
 105    IMPORTANT: On some systems it is required that RESBUF is correctly
 106    aligned for a 32 bits value.  */
 107 void *
 108 md5_finish_ctx (ctx, resbuf)
 109      struct md5_ctx *ctx;
 110      void *resbuf;
 111 {
 112   /* Take yet unprocessed bytes into account.  */
 113   md5_uint32 bytes = ctx->buflen;
 114   size_t pad;
 115
 116   /* Now count remaining bytes.  */
 117   ctx->total[0] += bytes;
 118   if (ctx->total[0] < bytes)
 119     ++ctx->total[1];
 120
 121   pad = bytes >= 56 ? 64 + 56 - bytes : 56 - bytes;
 122   memcpy (&ctx->buffer[bytes], fillbuf, pad);
 123
 124   /* Put the 64-bit file length in *bits* at the end of the buffer.  */
 125   *(md5_uint32 *) &ctx->buffer[bytes + pad] = SWAP (ctx->total[0] << 3);
 126   *(md5_uint32 *) &ctx->buffer[bytes + pad + 4] = SWAP ((ctx->total[1] << 3) |
 127                                                         (ctx->total[0] >> 29));
 128
 129   /* Process last bytes.  */
 130   md5_process_block (ctx->buffer, bytes + pad + 8, ctx);
 131
 132   return md5_read_ctx (ctx, resbuf);
 133 }
 134
 135 /* Compute MD5 message digest for bytes read from STREAM.  The
 136    resulting message digest number will be written into the 16 bytes
 137    beginning at RESBLOCK.  */
 138 int
 139 md5_stream (stream, resblock)
 140      FILE *stream;
 141      void *resblock;
 142 {
 143   /* Important: BLOCKSIZE must be a multiple of 64.  */
 144 #define BLOCKSIZE 4096
 145   struct md5_ctx ctx;
 146   char buffer[BLOCKSIZE + 72];
 147   size_t sum;
 148
 149   /* Initialize the computation context.  */
 150   md5_init_ctx (&ctx);
 151
 152   /* Iterate over full file contents.  */
 153   while (1)
 154     {
 155       /* We read the file in blocks of BLOCKSIZE bytes.  One call of the
 156          computation function processes the whole buffer so that with the
 157          next round of the loop another block can be read.  */
 158       size_t n;
 159       sum = 0;
 160
 161       /* Read block.  Take care for partial reads.  */
 162       do
 163         {
 164           n = fread (buffer + sum, 1, BLOCKSIZE - sum, stream);
 165
 166           sum += n;
 167         }
 168       while (sum < BLOCKSIZE && n != 0);
 169       if (n == 0 && ferror (stream))
 170         return 1;
 171
 172       /* If end of file is reached, end the loop.  */
 173       if (n == 0)
 174         break;
 175
 176       /* Process buffer with BLOCKSIZE bytes.  Note that
 177                         BLOCKSIZE % 64 == 0
 178        */
 179       md5_process_block (buffer, BLOCKSIZE, &ctx);
 180     }
 181
 182   /* Add the last bytes if necessary.  */
 183   if (sum > 0)
 184     md5_process_bytes (buffer, sum, &ctx);
 185
 186   /* Construct result in desired memory.  */
 187   md5_finish_ctx (&ctx, resblock);
 188   return 0;
 189 }
 190
 191 /* Compute MD5 message digest for LEN bytes beginning at BUFFER.  The
 192    result is always in little endian byte order, so that a byte-wise
 193    output yields to the wanted ASCII representation of the message
 194    digest.  */
 195 void *
 196 md5_buffer (buffer, len, resblock)
 197      const char *buffer;
 198      size_t len;
 199      void *resblock;
 200 {
 201   struct md5_ctx ctx;
 202
 203   /* Initialize the computation context.  */
 204   md5_init_ctx (&ctx);
 205
 206   /* Process whole buffer but last len % 64 bytes.  */
 207   md5_process_bytes (buffer, len, &ctx);
 208
 209   /* Put result in desired memory area.  */
 210   return md5_finish_ctx (&ctx, resblock);
 211 }
 212
 213
 214 void
 215 md5_process_bytes (buffer, len, ctx)
 216      const void *buffer;
 217      size_t len;
 218      struct md5_ctx *ctx;
 219 {
 220   /* const void aligned_buffer = buffer; */
 221
 222   /* When we already have some bits in our internal buffer concatenate
 223      both inputs first.  */
 224   if (ctx->buflen != 0)
 225     {
 226       size_t left_over = ctx->buflen;
 227       size_t add = 128 - left_over > len ? len : 128 - left_over;
 228
 229       /* Only put full words in the buffer.  */
 230       add -= add % __alignof__ (md5_uint32);
 231
 232       memcpy (&ctx->buffer[left_over], buffer, add);
 233       ctx->buflen += add;
 234
 235       if (ctx->buflen > 64)
 236         {
 237           md5_process_block (ctx->buffer, ctx->buflen & ~63, ctx);
 238
 239           ctx->buflen &= 63;
 240           /* The regions in the following copy operation cannot overlap.  */
 241           memcpy (ctx->buffer, &ctx->buffer[(left_over + add) & ~63],
 242                   ctx->buflen);
 243         }
 244
 245       buffer = (const char *) buffer + add;
 246       len -= add;
 247     }
 248
 249   /* Process available complete blocks.  */
 250   if (len > 64)
 251     {
 252       md5_process_block (buffer, len & ~63, ctx);
 253       buffer = (const char *) buffer + (len & ~63);
 254       len &= 63;
 255     }
 256
 257   /* Move remaining bytes in internal buffer.  */
 258   if (len > 0)
 259     {
 260       size_t left_over = ctx->buflen;
 261
 262       memcpy (&ctx->buffer[left_over], buffer, len);
 263       left_over += len;
 264       if (left_over >= 64)
 265         {
 266           md5_process_block (ctx->buffer, 64, ctx);
 267           left_over -= 64;
 268           memcpy (ctx->buffer, &ctx->buffer[64], left_over);
 269         }
 270       ctx->buflen = left_over;
 271     }
 272 }
 273
 274
 275 /* These are the four functions used in the four steps of the MD5 algorithm
 276    and defined in the RFC 1321.  The first function is a little bit optimized
 277    (as found in Colin Plumbs public domain implementation).  */
 278 /* #define FF(b, c, d) ((b & c) | (~b & d)) */
 279 #define FF(b, c, d) (d ^ (b & (c ^ d)))
 280 #define FG(b, c, d) FF (d, b, c)
 281 #define FH(b, c, d) (b ^ c ^ d)
 282 #define FI(b, c, d) (c ^ (b | ~d))
 283
 284 /* Process LEN bytes of BUFFER, accumulating context into CTX.
 285    It is assumed that LEN % 64 == 0.  */
 286
 287 void
 288 md5_process_block (buffer, len, ctx)
 289      const void *buffer;
 290      size_t len;
 291      struct md5_ctx *ctx;
 292 {
 293   md5_uint32 correct_words[16];
 294   const md5_uint32 *words = buffer;
 295   size_t nwords = len / sizeof (md5_uint32);
 296   const md5_uint32 *endp = words + nwords;
 297   md5_uint32 A = ctx->A;
 298   md5_uint32 B = ctx->B;
 299   md5_uint32 C = ctx->C;
 300   md5_uint32 D = ctx->D;
 301
 302   /* First increment the byte count.  RFC 1321 specifies the possible
 303      length of the file up to 2^64 bits.  Here we only compute the
 304      number of bytes.  Do a double word increment.  */
 305   ctx->total[0] += len;
 306   if (ctx->total[0] < len)
 307     ++ctx->total[1];
 308
 309   /* Process all bytes in the buffer with 64 bytes in each round of
 310      the loop.  */
 311   while (words < endp)
 312     {
 313       md5_uint32 *cwp = correct_words;
 314       md5_uint32 A_save = A;
 315       md5_uint32 B_save = B;
 316       md5_uint32 C_save = C;
 317       md5_uint32 D_save = D;
 318
 319       /* First round: using the given function, the context and a constant
 320          the next context is computed.  Because the algorithms processing
 321          unit is a 32-bit word and it is determined to work on words in
 322          little endian byte order we perhaps have to change the byte order
 323          before the computation.  To reduce the work for the next steps
 324          we store the swapped words in the array CORRECT_WORDS.  */
 325
 326 #define OP(a, b, c, d, s, T)                                            \
 327       do                                                                \
 328         {                                                               \
 329           a += FF (b, c, d) + (*cwp++ = SWAP (*words)) + T;             \
 330           ++words;                                                      \
 331           CYCLIC (a, s);                                                \
 332           a += b;                                                       \
 333         }                                                               \
 334       while (0)
 335
 336       /* It is unfortunate that C does not provide an operator for
 337          cyclic rotation.  Hope the C compiler is smart enough.  */
 338 #define CYCLIC(w, s) (w = (w << s) | (w >> (32 - s)))
 339
 340       /* Before we start, one word to the strange constants.
 341          They are defined in RFC 1321 as
 342
 343          T[i] = (int) (4294967296.0 * fabs (sin (i))), i=1..64
 344        */
 345
 346       /* Round 1.  */
 347       OP (A, B, C, D,  7, 0xd76aa478);
 348       OP (D, A, B, C, 12, 0xe8c7b756);
 349       OP (C, D, A, B, 17, 0x242070db);
 350       OP (B, C, D, A, 22, 0xc1bdceee);
 351       OP (A, B, C, D,  7, 0xf57c0faf);
 352       OP (D, A, B, C, 12, 0x4787c62a);
 353       OP (C, D, A, B, 17, 0xa8304613);
 354       OP (B, C, D, A, 22, 0xfd469501);
 355       OP (A, B, C, D,  7, 0x698098d8);
 356       OP (D, A, B, C, 12, 0x8b44f7af);
 357       OP (C, D, A, B, 17, 0xffff5bb1);
 358       OP (B, C, D, A, 22, 0x895cd7be);
 359       OP (A, B, C, D,  7, 0x6b901122);
 360       OP (D, A, B, C, 12, 0xfd987193);
 361       OP (C, D, A, B, 17, 0xa679438e);
 362       OP (B, C, D, A, 22, 0x49b40821);
 363
 364       /* For the second to fourth round we have the possibly swapped words
 365          in CORRECT_WORDS.  Redefine the macro to take an additional first
 366          argument specifying the function to use.  */
 367 #undef OP
 368 #define OP(f, a, b, c, d, k, s, T)                                      \
 369       do                                                                \
 370         {                                                               \
 371           a += f (b, c, d) + correct_words[k] + T;                      \
 372           CYCLIC (a, s);                                                \
 373           a += b;                                                       \
 374         }                                                               \
 375       while (0)
 376
 377       /* Round 2.  */
 378       OP (FG, A, B, C, D,  1,  5, 0xf61e2562);
 379       OP (FG, D, A, B, C,  6,  9, 0xc040b340);
 380       OP (FG, C, D, A, B, 11, 14, 0x265e5a51);
 381       OP (FG, B, C, D, A,  0, 20, 0xe9b6c7aa);
 382       OP (FG, A, B, C, D,  5,  5, 0xd62f105d);
 383       OP (FG, D, A, B, C, 10,  9, 0x02441453);
 384       OP (FG, C, D, A, B, 15, 14, 0xd8a1e681);
 385       OP (FG, B, C, D, A,  4, 20, 0xe7d3fbc8);
 386       OP (FG, A, B, C, D,  9,  5, 0x21e1cde6);
 387       OP (FG, D, A, B, C, 14,  9, 0xc33707d6);
 388       OP (FG, C, D, A, B,  3, 14, 0xf4d50d87);
 389       OP (FG, B, C, D, A,  8, 20, 0x455a14ed);
 390       OP (FG, A, B, C, D, 13,  5, 0xa9e3e905);
 391       OP (FG, D, A, B, C,  2,  9, 0xfcefa3f8);
 392       OP (FG, C, D, A, B,  7, 14, 0x676f02d9);
 393       OP (FG, B, C, D, A, 12, 20, 0x8d2a4c8a);
 394
 395       /* Round 3.  */
 396       OP (FH, A, B, C, D,  5,  4, 0xfffa3942);
 397       OP (FH, D, A, B, C,  8, 11, 0x8771f681);
 398       OP (FH, C, D, A, B, 11, 16, 0x6d9d6122);
 399       OP (FH, B, C, D, A, 14, 23, 0xfde5380c);
 400       OP (FH, A, B, C, D,  1,  4, 0xa4beea44);
 401       OP (FH, D, A, B, C,  4, 11, 0x4bdecfa9);
 402       OP (FH, C, D, A, B,  7, 16, 0xf6bb4b60);
 403       OP (FH, B, C, D, A, 10, 23, 0xbebfbc70);
 404       OP (FH, A, B, C, D, 13,  4, 0x289b7ec6);
 405       OP (FH, D, A, B, C,  0, 11, 0xeaa127fa);
 406       OP (FH, C, D, A, B,  3, 16, 0xd4ef3085);
 407       OP (FH, B, C, D, A,  6, 23, 0x04881d05);
 408       OP (FH, A, B, C, D,  9,  4, 0xd9d4d039);
 409       OP (FH, D, A, B, C, 12, 11, 0xe6db99e5);
 410       OP (FH, C, D, A, B, 15, 16, 0x1fa27cf8);
 411       OP (FH, B, C, D, A,  2, 23, 0xc4ac5665);
 412
 413       /* Round 4.  */
 414       OP (FI, A, B, C, D,  0,  6, 0xf4292244);
 415       OP (FI, D, A, B, C,  7, 10, 0x432aff97);
 416       OP (FI, C, D, A, B, 14, 15, 0xab9423a7);
 417       OP (FI, B, C, D, A,  5, 21, 0xfc93a039);
 418       OP (FI, A, B, C, D, 12,  6, 0x655b59c3);
 419       OP (FI, D, A, B, C,  3, 10, 0x8f0ccc92);
 420       OP (FI, C, D, A, B, 10, 15, 0xffeff47d);
 421       OP (FI, B, C, D, A,  1, 21, 0x85845dd1);
 422       OP (FI, A, B, C, D,  8,  6, 0x6fa87e4f);
 423       OP (FI, D, A, B, C, 15, 10, 0xfe2ce6e0);
 424       OP (FI, C, D, A, B,  6, 15, 0xa3014314);
 425       OP (FI, B, C, D, A, 13, 21, 0x4e0811a1);
 426       OP (FI, A, B, C, D,  4,  6, 0xf7537e82);
 427       OP (FI, D, A, B, C, 11, 10, 0xbd3af235);
 428       OP (FI, C, D, A, B,  2, 15, 0x2ad7d2bb);
 429       OP (FI, B, C, D, A,  9, 21, 0xeb86d391);
 430
 431       /* Add the starting values of the context.  */
 432       A += A_save;
 433       B += B_save;
 434       C += C_save;
 435       D += D_save;
 436     }
 437
 438   /* Put checksum in context given as argument.  */
 439   ctx->A = A;
 440   ctx->B = B;
 441   ctx->C = C;
 442   ctx->D = D;
 443 }