libiberty/md5.c

   1 /* md5.c - 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, 2011 Free Software Foundation, Inc.
   4
   5    NOTE: This source is derived from an old version taken from the GNU C
   6    Library (glibc).
   7
   8    This program is free software; you can redistribute it and/or modify it
   9    under the terms of the GNU General Public License as published by the
  10    Free Software Foundation; either version 2, or (at your option) any
  11    later version.
  12
  13    This program is distributed in the hope that it will be useful,
  14    but WITHOUT ANY WARRANTY; without even the implied warranty of
  15    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  16    GNU General Public License for more details.
  17
  18    You should have received a copy of the GNU General Public License
  19    along with this program; if not, write to the Free Software Foundation,
  20    Inc., 51 Franklin Street - Fifth Floor, Boston, MA 02110-1301, USA.  */
  21
  22 /* Written by Ulrich Drepper <drepper@gnu.ai.mit.edu>, 1995.  */
  23
  24 #ifdef HAVE_CONFIG_H
  25 # include <config.h>
  26 #endif
  27
  28 #include <sys/types.h>
  29
  30 #if STDC_HEADERS || defined _LIBC
  31 # include <stdlib.h>
  32 # include <string.h>
  33 #else
  34 # ifndef HAVE_MEMCPY
  35 #  define memcpy(d, s, n) bcopy ((s), (d), (n))
  36 # endif
  37 #endif
  38
  39 #include "ansidecl.h"
  40 #include "md5.h"
  41
  42 #ifdef _LIBC
  43 # include <endian.h>
  44 # if __BYTE_ORDER == __BIG_ENDIAN
  45 #  define WORDS_BIGENDIAN 1
  46 # endif
  47 #endif
  48
  49 #ifdef WORDS_BIGENDIAN
  50 # define SWAP(n)                                                        \
  51     (((n) << 24) | (((n) & 0xff00) << 8) | (((n) >> 8) & 0xff00) | ((n) >> 24))
  52 #else
  53 # define SWAP(n) (n)
  54 #endif
  55
  56
  57 /* This array contains the bytes used to pad the buffer to the next
  58    64-byte boundary.  (RFC 1321, 3.1: Step 1)  */
  59 static const unsigned char fillbuf[64] = { 0x80, 0 /* , 0, 0, ...  */ };
  60
  61
  62 /* Initialize structure containing state of computation.
  63    (RFC 1321, 3.3: Step 3)  */
  64 void
  65 md5_init_ctx (struct md5_ctx *ctx)
  66 {
  67   ctx->A = (md5_uint32) 0x67452301;
  68   ctx->B = (md5_uint32) 0xefcdab89;
  69   ctx->C = (md5_uint32) 0x98badcfe;
  70   ctx->D = (md5_uint32) 0x10325476;
  71
  72   ctx->total[0] = ctx->total[1] = 0;
  73   ctx->buflen = 0;
  74 }
  75
  76 /* Put result from CTX in first 16 bytes following RESBUF.  The result
  77    must be in little endian byte order.
  78
  79    IMPORTANT: RESBUF may not be aligned as strongly as MD5_UNIT32 so we
  80    put things in a local (aligned) buffer first, then memcpy into RESBUF.  */
  81 void *
  82 md5_read_ctx (const struct md5_ctx *ctx, void *resbuf)
  83 {
  84   md5_uint32 buffer[4];
  85
  86   buffer[0] = SWAP (ctx->A);
  87   buffer[1] = SWAP (ctx->B);
  88   buffer[2] = SWAP (ctx->C);
  89   buffer[3] = SWAP (ctx->D);
  90
  91   memcpy (resbuf, buffer, 16);
  92
  93   return resbuf;
  94 }
  95
  96 /* Process the remaining bytes in the internal buffer and the usual
  97    prolog according to the standard and write the result to RESBUF.
  98
  99    IMPORTANT: On some systems it is required that RESBUF is correctly
 100    aligned for a 32 bits value.  */
 101 void *
 102 md5_finish_ctx (struct md5_ctx *ctx, void *resbuf)
 103 {
 104   /* Take yet unprocessed bytes into account.  */
 105   md5_uint32 bytes = ctx->buflen;
 106   size_t pad;
 107
 108   /* Now count remaining bytes.  */
 109   ctx->total[0] += bytes;
 110   if (ctx->total[0] < bytes)
 111     ++ctx->total[1];
 112
 113   pad = bytes >= 56 ? 64 + 56 - bytes : 56 - bytes;
 114   memcpy (&ctx->buffer[bytes], fillbuf, pad);
 115
 116   /* Put the 64-bit file length in *bits* at the end of the buffer.  */
 117   *(md5_uint32 *) &ctx->buffer[bytes + pad] = SWAP (ctx->total[0] << 3);
 118   *(md5_uint32 *) &ctx->buffer[bytes + pad + 4] = SWAP ((ctx->total[1] << 3) |
 119                                                         (ctx->total[0] >> 29));
 120
 121   /* Process last bytes.  */
 122   md5_process_block (ctx->buffer, bytes + pad + 8, ctx);
 123
 124   return md5_read_ctx (ctx, resbuf);
 125 }
 126
 127 /* Compute MD5 message digest for bytes read from STREAM.  The
 128    resulting message digest number will be written into the 16 bytes
 129    beginning at RESBLOCK.  */
 130 int
 131 md5_stream (FILE *stream, void *resblock)
 132 {
 133   /* Important: BLOCKSIZE must be a multiple of 64.  */
 134 #define BLOCKSIZE 4096
 135   struct md5_ctx ctx;
 136   char buffer[BLOCKSIZE + 72];
 137   size_t sum;
 138
 139   /* Initialize the computation context.  */
 140   md5_init_ctx (&ctx);
 141
 142   /* Iterate over full file contents.  */
 143   while (1)
 144     {
 145       /* We read the file in blocks of BLOCKSIZE bytes.  One call of the
 146          computation function processes the whole buffer so that with the
 147          next round of the loop another block can be read.  */
 148       size_t n;
 149       sum = 0;
 150
 151       /* Read block.  Take care for partial reads.  */
 152       do
 153         {
 154           n = fread (buffer + sum, 1, BLOCKSIZE - sum, stream);
 155
 156           sum += n;
 157         }
 158       while (sum < BLOCKSIZE && n != 0);
 159       if (n == 0 && ferror (stream))
 160         return 1;
 161
 162       /* If end of file is reached, end the loop.  */
 163       if (n == 0)
 164         break;
 165
 166       /* Process buffer with BLOCKSIZE bytes.  Note that
 167                         BLOCKSIZE % 64 == 0
 168        */
 169       md5_process_block (buffer, BLOCKSIZE, &ctx);
 170     }
 171
 172   /* Add the last bytes if necessary.  */
 173   if (sum > 0)
 174     md5_process_bytes (buffer, sum, &ctx);
 175
 176   /* Construct result in desired memory.  */
 177   md5_finish_ctx (&ctx, resblock);
 178   return 0;
 179 }
 180
 181 /* Compute MD5 message digest for LEN bytes beginning at BUFFER.  The
 182    result is always in little endian byte order, so that a byte-wise
 183    output yields to the wanted ASCII representation of the message
 184    digest.  */
 185 void *
 186 md5_buffer (const char *buffer, size_t len, void *resblock)
 187 {
 188   struct md5_ctx ctx;
 189
 190   /* Initialize the computation context.  */
 191   md5_init_ctx (&ctx);
 192
 193   /* Process whole buffer but last len % 64 bytes.  */
 194   md5_process_bytes (buffer, len, &ctx);
 195
 196   /* Put result in desired memory area.  */
 197   return md5_finish_ctx (&ctx, resblock);
 198 }
 199
 200
 201 void
 202 md5_process_bytes (const void *buffer, size_t len, struct md5_ctx *ctx)
 203 {
 204   /* When we already have some bits in our internal buffer concatenate
 205      both inputs first.  */
 206   if (ctx->buflen != 0)
 207     {
 208       size_t left_over = ctx->buflen;
 209       size_t add = 128 - left_over > len ? len : 128 - left_over;
 210
 211       memcpy (&ctx->buffer[left_over], buffer, add);
 212       ctx->buflen += add;
 213
 214       if (left_over + add > 64)
 215         {
 216           md5_process_block (ctx->buffer, (left_over + add) & ~63, ctx);
 217           /* The regions in the following copy operation cannot overlap.  */
 218           memcpy (ctx->buffer, &ctx->buffer[(left_over + add) & ~63],
 219                   (left_over + add) & 63);
 220           ctx->buflen = (left_over + add) & 63;
 221         }
 222
 223       buffer = (const void *) ((const char *) buffer + add);
 224       len -= add;
 225     }
 226
 227   /* Process available complete blocks.  */
 228   if (len > 64)
 229     {
 230 #if !_STRING_ARCH_unaligned
 231 /* To check alignment gcc has an appropriate operator.  Other
 232    compilers don't.  */
 233 # if __GNUC__ >= 2
 234 #  define UNALIGNED_P(p) (((md5_uintptr) p) % __alignof__ (md5_uint32) != 0)
 235 # else
 236 #  define UNALIGNED_P(p) (((md5_uintptr) p) % sizeof (md5_uint32) != 0)
 237 # endif
 238       if (UNALIGNED_P (buffer))
 239         while (len > 64)
 240           {
 241             memcpy (ctx->buffer, buffer, 64);
 242             md5_process_block (ctx->buffer, 64, ctx);
 243             buffer = (const char *) buffer + 64;
 244             len -= 64;
 245           }
 246       else
 247 #endif
 248         {
 249           md5_process_block (buffer, len & ~63, ctx);
 250           buffer = (const void *) ((const char *) buffer + (len & ~63));
 251           len &= 63;
 252         }
 253     }
 254
 255   /* Move remaining bytes in internal buffer.  */
 256   if (len > 0)
 257     {
 258       memcpy (ctx->buffer, buffer, len);
 259       ctx->buflen = len;
 260     }
 261 }
 262
 263
 264 /* These are the four functions used in the four steps of the MD5 algorithm
 265    and defined in the RFC 1321.  The first function is a little bit optimized
 266    (as found in Colin Plumbs public domain implementation).  */
 267 /* #define FF(b, c, d) ((b & c) | (~b & d)) */
 268 #define FF(b, c, d) (d ^ (b & (c ^ d)))
 269 #define FG(b, c, d) FF (d, b, c)
 270 #define FH(b, c, d) (b ^ c ^ d)
 271 #define FI(b, c, d) (c ^ (b | ~d))
 272
 273 /* Process LEN bytes of BUFFER, accumulating context into CTX.
 274    It is assumed that LEN % 64 == 0.  */
 275
 276 void
 277 md5_process_block (const void *buffer, size_t len, struct md5_ctx *ctx)
 278 {
 279   md5_uint32 correct_words[16];
 280   const md5_uint32 *words = (const md5_uint32 *) buffer;
 281   size_t nwords = len / sizeof (md5_uint32);
 282   const md5_uint32 *endp = words + nwords;
 283   md5_uint32 A = ctx->A;
 284   md5_uint32 B = ctx->B;
 285   md5_uint32 C = ctx->C;
 286   md5_uint32 D = ctx->D;
 287
 288   /* First increment the byte count.  RFC 1321 specifies the possible
 289      length of the file up to 2^64 bits.  Here we only compute the
 290      number of bytes.  Do a double word increment.  */
 291   ctx->total[0] += len;
 292   if (ctx->total[0] < len)
 293     ++ctx->total[1];
 294
 295   /* Process all bytes in the buffer with 64 bytes in each round of
 296      the loop.  */
 297   while (words < endp)
 298     {
 299       md5_uint32 *cwp = correct_words;
 300       md5_uint32 A_save = A;
 301       md5_uint32 B_save = B;
 302       md5_uint32 C_save = C;
 303       md5_uint32 D_save = D;
 304
 305       /* First round: using the given function, the context and a constant
 306          the next context is computed.  Because the algorithms processing
 307          unit is a 32-bit word and it is determined to work on words in
 308          little endian byte order we perhaps have to change the byte order
 309          before the computation.  To reduce the work for the next steps
 310          we store the swapped words in the array CORRECT_WORDS.  */
 311
 312 #define OP(a, b, c, d, s, T)                                            \
 313       do                                                                \
 314         {                                                               \
 315           a += FF (b, c, d) + (*cwp++ = SWAP (*words)) + T;             \
 316           ++words;                                                      \
 317           CYCLIC (a, s);                                                \
 318           a += b;                                                       \
 319         }                                                               \
 320       while (0)
 321
 322       /* It is unfortunate that C does not provide an operator for
 323          cyclic rotation.  Hope the C compiler is smart enough.  */
 324 #define CYCLIC(w, s) (w = (w << s) | (w >> (32 - s)))
 325
 326       /* Before we start, one word to the strange constants.
 327          They are defined in RFC 1321 as
 328
 329          T[i] = (int) (4294967296.0 * fabs (sin (i))), i=1..64
 330        */
 331
 332       /* Round 1.  */
 333       OP (A, B, C, D,  7, (md5_uint32) 0xd76aa478);
 334       OP (D, A, B, C, 12, (md5_uint32) 0xe8c7b756);
 335       OP (C, D, A, B, 17, (md5_uint32) 0x242070db);
 336       OP (B, C, D, A, 22, (md5_uint32) 0xc1bdceee);
 337       OP (A, B, C, D,  7, (md5_uint32) 0xf57c0faf);
 338       OP (D, A, B, C, 12, (md5_uint32) 0x4787c62a);
 339       OP (C, D, A, B, 17, (md5_uint32) 0xa8304613);
 340       OP (B, C, D, A, 22, (md5_uint32) 0xfd469501);
 341       OP (A, B, C, D,  7, (md5_uint32) 0x698098d8);
 342       OP (D, A, B, C, 12, (md5_uint32) 0x8b44f7af);
 343       OP (C, D, A, B, 17, (md5_uint32) 0xffff5bb1);
 344       OP (B, C, D, A, 22, (md5_uint32) 0x895cd7be);
 345       OP (A, B, C, D,  7, (md5_uint32) 0x6b901122);
 346       OP (D, A, B, C, 12, (md5_uint32) 0xfd987193);
 347       OP (C, D, A, B, 17, (md5_uint32) 0xa679438e);
 348       OP (B, C, D, A, 22, (md5_uint32) 0x49b40821);
 349
 350       /* For the second to fourth round we have the possibly swapped words
 351          in CORRECT_WORDS.  Redefine the macro to take an additional first
 352          argument specifying the function to use.  */
 353 #undef OP
 354 #define OP(a, b, c, d, k, s, T)                                         \
 355       do                                                                \
 356         {                                                               \
 357           a += FX (b, c, d) + correct_words[k] + T;                     \
 358           CYCLIC (a, s);                                                \
 359           a += b;                                                       \
 360         }                                                               \
 361       while (0)
 362
 363 #define FX(b, c, d) FG (b, c, d)
 364
 365       /* Round 2.  */
 366       OP (A, B, C, D,  1,  5, (md5_uint32) 0xf61e2562);
 367       OP (D, A, B, C,  6,  9, (md5_uint32) 0xc040b340);
 368       OP (C, D, A, B, 11, 14, (md5_uint32) 0x265e5a51);
 369       OP (B, C, D, A,  0, 20, (md5_uint32) 0xe9b6c7aa);
 370       OP (A, B, C, D,  5,  5, (md5_uint32) 0xd62f105d);
 371       OP (D, A, B, C, 10,  9, (md5_uint32) 0x02441453);
 372       OP (C, D, A, B, 15, 14, (md5_uint32) 0xd8a1e681);
 373       OP (B, C, D, A,  4, 20, (md5_uint32) 0xe7d3fbc8);
 374       OP (A, B, C, D,  9,  5, (md5_uint32) 0x21e1cde6);
 375       OP (D, A, B, C, 14,  9, (md5_uint32) 0xc33707d6);
 376       OP (C, D, A, B,  3, 14, (md5_uint32) 0xf4d50d87);
 377       OP (B, C, D, A,  8, 20, (md5_uint32) 0x455a14ed);
 378       OP (A, B, C, D, 13,  5, (md5_uint32) 0xa9e3e905);
 379       OP (D, A, B, C,  2,  9, (md5_uint32) 0xfcefa3f8);
 380       OP (C, D, A, B,  7, 14, (md5_uint32) 0x676f02d9);
 381       OP (B, C, D, A, 12, 20, (md5_uint32) 0x8d2a4c8a);
 382
 383 #undef FX
 384 #define FX(b, c, d) FH (b, c, d)
 385
 386       /* Round 3.  */
 387       OP (A, B, C, D,  5,  4, (md5_uint32) 0xfffa3942);
 388       OP (D, A, B, C,  8, 11, (md5_uint32) 0x8771f681);
 389       OP (C, D, A, B, 11, 16, (md5_uint32) 0x6d9d6122);
 390       OP (B, C, D, A, 14, 23, (md5_uint32) 0xfde5380c);
 391       OP (A, B, C, D,  1,  4, (md5_uint32) 0xa4beea44);
 392       OP (D, A, B, C,  4, 11, (md5_uint32) 0x4bdecfa9);
 393       OP (C, D, A, B,  7, 16, (md5_uint32) 0xf6bb4b60);
 394       OP (B, C, D, A, 10, 23, (md5_uint32) 0xbebfbc70);
 395       OP (A, B, C, D, 13,  4, (md5_uint32) 0x289b7ec6);
 396       OP (D, A, B, C,  0, 11, (md5_uint32) 0xeaa127fa);
 397       OP (C, D, A, B,  3, 16, (md5_uint32) 0xd4ef3085);
 398       OP (B, C, D, A,  6, 23, (md5_uint32) 0x04881d05);
 399       OP (A, B, C, D,  9,  4, (md5_uint32) 0xd9d4d039);
 400       OP (D, A, B, C, 12, 11, (md5_uint32) 0xe6db99e5);
 401       OP (C, D, A, B, 15, 16, (md5_uint32) 0x1fa27cf8);
 402       OP (B, C, D, A,  2, 23, (md5_uint32) 0xc4ac5665);
 403
 404 #undef FX
 405 #define FX(b, c, d) FI (b, c, d)
 406
 407       /* Round 4.  */
 408       OP (A, B, C, D,  0,  6, (md5_uint32) 0xf4292244);
 409       OP (D, A, B, C,  7, 10, (md5_uint32) 0x432aff97);
 410       OP (C, D, A, B, 14, 15, (md5_uint32) 0xab9423a7);
 411       OP (B, C, D, A,  5, 21, (md5_uint32) 0xfc93a039);
 412       OP (A, B, C, D, 12,  6, (md5_uint32) 0x655b59c3);
 413       OP (D, A, B, C,  3, 10, (md5_uint32) 0x8f0ccc92);
 414       OP (C, D, A, B, 10, 15, (md5_uint32) 0xffeff47d);
 415       OP (B, C, D, A,  1, 21, (md5_uint32) 0x85845dd1);
 416       OP (A, B, C, D,  8,  6, (md5_uint32) 0x6fa87e4f);
 417       OP (D, A, B, C, 15, 10, (md5_uint32) 0xfe2ce6e0);
 418       OP (C, D, A, B,  6, 15, (md5_uint32) 0xa3014314);
 419       OP (B, C, D, A, 13, 21, (md5_uint32) 0x4e0811a1);
 420       OP (A, B, C, D,  4,  6, (md5_uint32) 0xf7537e82);
 421       OP (D, A, B, C, 11, 10, (md5_uint32) 0xbd3af235);
 422       OP (C, D, A, B,  2, 15, (md5_uint32) 0x2ad7d2bb);
 423       OP (B, C, D, A,  9, 21, (md5_uint32) 0xeb86d391);
 424
 425       /* Add the starting values of the context.  */
 426       A += A_save;
 427       B += B_save;
 428       C += C_save;
 429       D += D_save;
 430     }
 431
 432   /* Put checksum in context given as argument.  */
 433   ctx->A = A;
 434   ctx->B = B;
 435   ctx->C = C;
 436   ctx->D = D;
 437 }