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) 1995,1996,1997,1999,2000,2001,2005,2006
   4         Free Software Foundation, Inc.
   5    This file is part of the GNU C Library.
   6
   7    This program is free software; you can redistribute it and/or modify it
   8    under the terms of the GNU General Public License as published by the
   9    Free Software Foundation; either version 2, or (at your option) any
  10    later version.
  11
  12    This program is distributed in the hope that it will be useful,
  13    but WITHOUT ANY WARRANTY; without even the implied warranty of
  14    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15    GNU General Public License for more details.
  16
  17    You should have received a copy of the GNU General Public License
  18    along with this program; if not, write to the Free Software Foundation,
  19    Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.  */
  20
  21 /* Written by Ulrich Drepper <drepper@gnu.ai.mit.edu>, 1995.  */
  22
  23 #include <config.h>
  24
  25 #include "md5.h"
  26
  27 #include <stddef.h>
  28 #include <stdlib.h>
  29 #include <string.h>
  30 #include <sys/types.h>
  31
  32 #if USE_UNLOCKED_IO
  33 # include "unlocked-io.h"
  34 #endif
  35
  36 #ifdef _LIBC
  37 # include <endian.h>
  38 # if __BYTE_ORDER == __BIG_ENDIAN
  39 #  define WORDS_BIGENDIAN 1
  40 # endif
  41 /* We need to keep the namespace clean so define the MD5 function
  42    protected using leading __ .  */
  43 # define md5_init_ctx __md5_init_ctx
  44 # define md5_process_block __md5_process_block
  45 # define md5_process_bytes __md5_process_bytes
  46 # define md5_finish_ctx __md5_finish_ctx
  47 # define md5_read_ctx __md5_read_ctx
  48 # define md5_stream __md5_stream
  49 # define md5_buffer __md5_buffer
  50 #endif
  51
  52 #ifdef WORDS_BIGENDIAN
  53 # define SWAP(n)                                                        \
  54     (((n) << 24) | (((n) & 0xff00) << 8) | (((n) >> 8) & 0xff00) | ((n) >> 24))
  55 #else
  56 # define SWAP(n) (n)
  57 #endif
  58
  59 #define BLOCKSIZE 4096
  60 #if BLOCKSIZE % 64 != 0
  61 # error "invalid BLOCKSIZE"
  62 #endif
  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 (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-bit value.  */
  88 void *
  89 md5_read_ctx (const struct md5_ctx *ctx, void *resbuf)
  90 {
  91   ((uint32_t *) resbuf)[0] = SWAP (ctx->A);
  92   ((uint32_t *) resbuf)[1] = SWAP (ctx->B);
  93   ((uint32_t *) resbuf)[2] = SWAP (ctx->C);
  94   ((uint32_t *) resbuf)[3] = SWAP (ctx->D);
  95
  96   return resbuf;
  97 }
  98
  99 /* Process the remaining bytes in the internal buffer and the usual
 100    prolog according to the standard and write the result to RESBUF.
 101
 102    IMPORTANT: On some systems it is required that RESBUF is correctly
 103    aligned for a 32-bit value.  */
 104 void *
 105 md5_finish_ctx (struct md5_ctx *ctx, void *resbuf)
 106 {
 107   /* Take yet unprocessed bytes into account.  */
 108   uint32_t bytes = ctx->buflen;
 109   size_t size = (bytes < 56) ? 64 / 4 : 64 * 2 / 4;
 110
 111   /* Now count remaining bytes.  */
 112   ctx->total[0] += bytes;
 113   if (ctx->total[0] < bytes)
 114     ++ctx->total[1];
 115
 116   /* Put the 64-bit file length in *bits* at the end of the buffer.  */
 117   ctx->buffer[size - 2] = SWAP (ctx->total[0] << 3);
 118   ctx->buffer[size - 1] = SWAP ((ctx->total[1] << 3) | (ctx->total[0] >> 29));
 119
 120   memcpy (&((char *) ctx->buffer)[bytes], fillbuf, (size - 2) * 4 - bytes);
 121
 122   /* Process last bytes.  */
 123   md5_process_block (ctx->buffer, size * 4, ctx);
 124
 125   return md5_read_ctx (ctx, resbuf);
 126 }
 127
 128 /* Compute MD5 message digest for bytes read from STREAM.  The
 129    resulting message digest number will be written into the 16 bytes
 130    beginning at RESBLOCK.  */
 131 int
 132 md5_stream (FILE *stream, void *resblock)
 133 {
 134   struct md5_ctx ctx;
 135   char buffer[BLOCKSIZE + 72];
 136   size_t sum;
 137
 138   /* Initialize the computation context.  */
 139   md5_init_ctx (&ctx);
 140
 141   /* Iterate over full file contents.  */
 142   while (1)
 143     {
 144       /* We read the file in blocks of BLOCKSIZE bytes.  One call of the
 145          computation function processes the whole buffer so that with the
 146          next round of the loop another block can be read.  */
 147       size_t n;
 148       sum = 0;
 149
 150       /* Read block.  Take care for partial reads.  */
 151       while (1)
 152         {
 153           n = fread (buffer + sum, 1, BLOCKSIZE - sum, stream);
 154
 155           sum += n;
 156
 157           if (sum == BLOCKSIZE)
 158             break;
 159
 160           if (n == 0)
 161             {
 162               /* Check for the error flag IFF N == 0, so that we don't
 163                  exit the loop after a partial read due to e.g., EAGAIN
 164                  or EWOULDBLOCK.  */
 165               if (ferror (stream))
 166                 return 1;
 167               goto process_partial_block;
 168             }
 169
 170           /* We've read at least one byte, so ignore errors.  But always
 171              check for EOF, since feof may be true even though N > 0.
 172              Otherwise, we could end up calling fread after EOF.  */
 173           if (feof (stream))
 174             goto process_partial_block;
 175         }
 176
 177       /* Process buffer with BLOCKSIZE bytes.  Note that
 178          BLOCKSIZE % 64 == 0
 179        */
 180       md5_process_block (buffer, BLOCKSIZE, &ctx);
 181     }
 182
 183 process_partial_block:
 184
 185   /* Process any remaining bytes.  */
 186   if (sum > 0)
 187     md5_process_bytes (buffer, sum, &ctx);
 188
 189   /* Construct result in desired memory.  */
 190   md5_finish_ctx (&ctx, resblock);
 191   return 0;
 192 }
 193
 194 /* Compute MD5 message digest for LEN bytes beginning at BUFFER.  The
 195    result is always in little endian byte order, so that a byte-wise
 196    output yields to the wanted ASCII representation of the message
 197    digest.  */
 198 void *
 199 md5_buffer (const char *buffer, size_t len, 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 (const void *buffer, size_t len, struct md5_ctx *ctx)
 216 {
 217   /* When we already have some bits in our internal buffer concatenate
 218      both inputs first.  */
 219   if (ctx->buflen != 0)
 220     {
 221       size_t left_over = ctx->buflen;
 222       size_t add = 128 - left_over > len ? len : 128 - left_over;
 223
 224       memcpy (&((char *) ctx->buffer)[left_over], buffer, add);
 225       ctx->buflen += add;
 226
 227       if (ctx->buflen > 64)
 228         {
 229           md5_process_block (ctx->buffer, ctx->buflen & ~63, ctx);
 230
 231           ctx->buflen &= 63;
 232           /* The regions in the following copy operation cannot overlap.  */
 233           memcpy (ctx->buffer,
 234                   &((char *) ctx->buffer)[(left_over + add) & ~63],
 235                   ctx->buflen);
 236         }
 237
 238       buffer = (const char *) buffer + add;
 239       len -= add;
 240     }
 241
 242   /* Process available complete blocks.  */
 243   if (len >= 64)
 244     {
 245 #if !_STRING_ARCH_unaligned
 246 # define alignof(type) offsetof (struct { char c; type x; }, x)
 247 # define UNALIGNED_P(p) (((size_t) p) % alignof (uint32_t) != 0)
 248       if (UNALIGNED_P (buffer))
 249         while (len > 64)
 250           {
 251             md5_process_block (memcpy (ctx->buffer, buffer, 64), 64, ctx);
 252             buffer = (const char *) buffer + 64;
 253             len -= 64;
 254           }
 255       else
 256 #endif
 257         {
 258           md5_process_block (buffer, len & ~63, ctx);
 259           buffer = (const char *) buffer + (len & ~63);
 260           len &= 63;
 261         }
 262     }
 263
 264   /* Move remaining bytes in internal buffer.  */
 265   if (len > 0)
 266     {
 267       size_t left_over = ctx->buflen;
 268
 269       memcpy (&((char *) ctx->buffer)[left_over], buffer, len);
 270       left_over += len;
 271       if (left_over >= 64)
 272         {
 273           md5_process_block (ctx->buffer, 64, ctx);
 274           left_over -= 64;
 275           memcpy (ctx->buffer, &ctx->buffer[16], left_over);
 276         }
 277       ctx->buflen = left_over;
 278     }
 279 }
 280
 281
 282 /* These are the four functions used in the four steps of the MD5 algorithm
 283    and defined in the RFC 1321.  The first function is a little bit optimized
 284    (as found in Colin Plumbs public domain implementation).  */
 285 /* #define FF(b, c, d) ((b & c) | (~b & d)) */
 286 #define FF(b, c, d) (d ^ (b & (c ^ d)))
 287 #define FG(b, c, d) FF (d, b, c)
 288 #define FH(b, c, d) (b ^ c ^ d)
 289 #define FI(b, c, d) (c ^ (b | ~d))
 290
 291 /* Process LEN bytes of BUFFER, accumulating context into CTX.
 292    It is assumed that LEN % 64 == 0.  */
 293
 294 void
 295 md5_process_block (const void *buffer, size_t len, struct md5_ctx *ctx)
 296 {
 297   uint32_t correct_words[16];
 298   const uint32_t *words = buffer;
 299   size_t nwords = len / sizeof (uint32_t);
 300   const uint32_t *endp = words + nwords;
 301   uint32_t A = ctx->A;
 302   uint32_t B = ctx->B;
 303   uint32_t C = ctx->C;
 304   uint32_t D = ctx->D;
 305
 306   /* First increment the byte count.  RFC 1321 specifies the possible
 307      length of the file up to 2^64 bits.  Here we only compute the
 308      number of bytes.  Do a double word increment.  */
 309   ctx->total[0] += len;
 310   if (ctx->total[0] < len)
 311     ++ctx->total[1];
 312
 313   /* Process all bytes in the buffer with 64 bytes in each round of
 314      the loop.  */
 315   while (words < endp)
 316     {
 317       uint32_t *cwp = correct_words;
 318       uint32_t A_save = A;
 319       uint32_t B_save = B;
 320       uint32_t C_save = C;
 321       uint32_t D_save = D;
 322
 323       /* First round: using the given function, the context and a constant
 324          the next context is computed.  Because the algorithms processing
 325          unit is a 32-bit word and it is determined to work on words in
 326          little endian byte order we perhaps have to change the byte order
 327          before the computation.  To reduce the work for the next steps
 328          we store the swapped words in the array CORRECT_WORDS.  */
 329
 330 #define OP(a, b, c, d, s, T)                                            \
 331       do                                                                \
 332         {                                                               \
 333           a += FF (b, c, d) + (*cwp++ = SWAP (*words)) + T;             \
 334           ++words;                                                      \
 335           CYCLIC (a, s);                                                \
 336           a += b;                                                       \
 337         }                                                               \
 338       while (0)
 339
 340       /* It is unfortunate that C does not provide an operator for
 341          cyclic rotation.  Hope the C compiler is smart enough.  */
 342 #define CYCLIC(w, s) (w = (w << s) | (w >> (32 - s)))
 343
 344       /* Before we start, one word to the strange constants.
 345          They are defined in RFC 1321 as
 346
 347          T[i] = (int) (4294967296.0 * fabs (sin (i))), i=1..64
 348
 349          Here is an equivalent invocation using Perl:
 350
 351          perl -e 'foreach(1..64){printf "0x%08x\n", int (4294967296 * abs (sin $_))}'
 352        */
 353
 354       /* Round 1.  */
 355       OP (A, B, C, D, 7, 0xd76aa478);
 356       OP (D, A, B, C, 12, 0xe8c7b756);
 357       OP (C, D, A, B, 17, 0x242070db);
 358       OP (B, C, D, A, 22, 0xc1bdceee);
 359       OP (A, B, C, D, 7, 0xf57c0faf);
 360       OP (D, A, B, C, 12, 0x4787c62a);
 361       OP (C, D, A, B, 17, 0xa8304613);
 362       OP (B, C, D, A, 22, 0xfd469501);
 363       OP (A, B, C, D, 7, 0x698098d8);
 364       OP (D, A, B, C, 12, 0x8b44f7af);
 365       OP (C, D, A, B, 17, 0xffff5bb1);
 366       OP (B, C, D, A, 22, 0x895cd7be);
 367       OP (A, B, C, D, 7, 0x6b901122);
 368       OP (D, A, B, C, 12, 0xfd987193);
 369       OP (C, D, A, B, 17, 0xa679438e);
 370       OP (B, C, D, A, 22, 0x49b40821);
 371
 372       /* For the second to fourth round we have the possibly swapped words
 373          in CORRECT_WORDS.  Redefine the macro to take an additional first
 374          argument specifying the function to use.  */
 375 #undef OP
 376 #define OP(f, a, b, c, d, k, s, T)                                      \
 377       do                                                                \
 378         {                                                               \
 379           a += f (b, c, d) + correct_words[k] + T;                      \
 380           CYCLIC (a, s);                                                \
 381           a += b;                                                       \
 382         }                                                               \
 383       while (0)
 384
 385       /* Round 2.  */
 386       OP (FG, A, B, C, D, 1, 5, 0xf61e2562);
 387       OP (FG, D, A, B, C, 6, 9, 0xc040b340);
 388       OP (FG, C, D, A, B, 11, 14, 0x265e5a51);
 389       OP (FG, B, C, D, A, 0, 20, 0xe9b6c7aa);
 390       OP (FG, A, B, C, D, 5, 5, 0xd62f105d);
 391       OP (FG, D, A, B, C, 10, 9, 0x02441453);
 392       OP (FG, C, D, A, B, 15, 14, 0xd8a1e681);
 393       OP (FG, B, C, D, A, 4, 20, 0xe7d3fbc8);
 394       OP (FG, A, B, C, D, 9, 5, 0x21e1cde6);
 395       OP (FG, D, A, B, C, 14, 9, 0xc33707d6);
 396       OP (FG, C, D, A, B, 3, 14, 0xf4d50d87);
 397       OP (FG, B, C, D, A, 8, 20, 0x455a14ed);
 398       OP (FG, A, B, C, D, 13, 5, 0xa9e3e905);
 399       OP (FG, D, A, B, C, 2, 9, 0xfcefa3f8);
 400       OP (FG, C, D, A, B, 7, 14, 0x676f02d9);
 401       OP (FG, B, C, D, A, 12, 20, 0x8d2a4c8a);
 402
 403       /* Round 3.  */
 404       OP (FH, A, B, C, D, 5, 4, 0xfffa3942);
 405       OP (FH, D, A, B, C, 8, 11, 0x8771f681);
 406       OP (FH, C, D, A, B, 11, 16, 0x6d9d6122);
 407       OP (FH, B, C, D, A, 14, 23, 0xfde5380c);
 408       OP (FH, A, B, C, D, 1, 4, 0xa4beea44);
 409       OP (FH, D, A, B, C, 4, 11, 0x4bdecfa9);
 410       OP (FH, C, D, A, B, 7, 16, 0xf6bb4b60);
 411       OP (FH, B, C, D, A, 10, 23, 0xbebfbc70);
 412       OP (FH, A, B, C, D, 13, 4, 0x289b7ec6);
 413       OP (FH, D, A, B, C, 0, 11, 0xeaa127fa);
 414       OP (FH, C, D, A, B, 3, 16, 0xd4ef3085);
 415       OP (FH, B, C, D, A, 6, 23, 0x04881d05);
 416       OP (FH, A, B, C, D, 9, 4, 0xd9d4d039);
 417       OP (FH, D, A, B, C, 12, 11, 0xe6db99e5);
 418       OP (FH, C, D, A, B, 15, 16, 0x1fa27cf8);
 419       OP (FH, B, C, D, A, 2, 23, 0xc4ac5665);
 420
 421       /* Round 4.  */
 422       OP (FI, A, B, C, D, 0, 6, 0xf4292244);
 423       OP (FI, D, A, B, C, 7, 10, 0x432aff97);
 424       OP (FI, C, D, A, B, 14, 15, 0xab9423a7);
 425       OP (FI, B, C, D, A, 5, 21, 0xfc93a039);
 426       OP (FI, A, B, C, D, 12, 6, 0x655b59c3);
 427       OP (FI, D, A, B, C, 3, 10, 0x8f0ccc92);
 428       OP (FI, C, D, A, B, 10, 15, 0xffeff47d);
 429       OP (FI, B, C, D, A, 1, 21, 0x85845dd1);
 430       OP (FI, A, B, C, D, 8, 6, 0x6fa87e4f);
 431       OP (FI, D, A, B, C, 15, 10, 0xfe2ce6e0);
 432       OP (FI, C, D, A, B, 6, 15, 0xa3014314);
 433       OP (FI, B, C, D, A, 13, 21, 0x4e0811a1);
 434       OP (FI, A, B, C, D, 4, 6, 0xf7537e82);
 435       OP (FI, D, A, B, C, 11, 10, 0xbd3af235);
 436       OP (FI, C, D, A, B, 2, 15, 0x2ad7d2bb);
 437       OP (FI, B, C, D, A, 9, 21, 0xeb86d391);
 438
 439       /* Add the starting values of the context.  */
 440       A += A_save;
 441       B += B_save;
 442       C += C_save;
 443       D += D_save;
 444     }
 445
 446   /* Put checksum in context given as argument.  */
 447   ctx->A = A;
 448   ctx->B = B;
 449   ctx->C = C;
 450   ctx->D = D;
 451 }