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) 1995-2023 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   md5_uint32 swap_bytes;
 107   size_t pad;
 108
 109   /* Now count remaining bytes.  */
 110   ctx->total[0] += bytes;
 111   if (ctx->total[0] < bytes)
 112     ++ctx->total[1];
 113
 114   pad = bytes >= 56 ? 64 + 56 - bytes : 56 - bytes;
 115   memcpy (&ctx->buffer[bytes], fillbuf, pad);
 116
 117   /* Put the 64-bit file length in *bits* at the end of the buffer.
 118      Use memcpy to avoid aliasing problems.  On most systems, this
 119      will be optimized away to the same code.  */
 120   swap_bytes = SWAP (ctx->total[0] << 3);
 121   memcpy (&ctx->buffer[bytes + pad], &swap_bytes, sizeof (swap_bytes));
 122   swap_bytes = SWAP ((ctx->total[1] << 3) | (ctx->total[0] >> 29));
 123   memcpy (&ctx->buffer[bytes + pad + 4], &swap_bytes, sizeof (swap_bytes));
 124
 125   /* Process last bytes.  */
 126   md5_process_block (ctx->buffer, bytes + pad + 8, ctx);
 127
 128   return md5_read_ctx (ctx, resbuf);
 129 }
 130
 131 /* Compute MD5 message digest for bytes read from STREAM.  The
 132    resulting message digest number will be written into the 16 bytes
 133    beginning at RESBLOCK.  */
 134 int
 135 md5_stream (FILE *stream, void *resblock)
 136 {
 137   /* Important: BLOCKSIZE must be a multiple of 64.  */
 138 #define BLOCKSIZE 4096
 139   struct md5_ctx ctx;
 140   char buffer[BLOCKSIZE + 72];
 141   size_t sum;
 142
 143   /* Initialize the computation context.  */
 144   md5_init_ctx (&ctx);
 145
 146   /* Iterate over full file contents.  */
 147   while (1)
 148     {
 149       /* We read the file in blocks of BLOCKSIZE bytes.  One call of the
 150          computation function processes the whole buffer so that with the
 151          next round of the loop another block can be read.  */
 152       size_t n;
 153       sum = 0;
 154
 155       /* Read block.  Take care for partial reads.  */
 156       do
 157         {
 158           n = fread (buffer + sum, 1, BLOCKSIZE - sum, stream);
 159
 160           sum += n;
 161         }
 162       while (sum < BLOCKSIZE && n != 0);
 163       if (n == 0 && ferror (stream))
 164         return 1;
 165
 166       /* If end of file is reached, end the loop.  */
 167       if (n == 0)
 168         break;
 169
 170       /* Process buffer with BLOCKSIZE bytes.  Note that
 171                         BLOCKSIZE % 64 == 0
 172        */
 173       md5_process_block (buffer, BLOCKSIZE, &ctx);
 174     }
 175
 176   /* Add the last bytes if necessary.  */
 177   if (sum > 0)
 178     md5_process_bytes (buffer, sum, &ctx);
 179
 180   /* Construct result in desired memory.  */
 181   md5_finish_ctx (&ctx, resblock);
 182   return 0;
 183 }
 184
 185 /* Compute MD5 message digest for LEN bytes beginning at BUFFER.  The
 186    result is always in little endian byte order, so that a byte-wise
 187    output yields to the wanted ASCII representation of the message
 188    digest.  */
 189 void *
 190 md5_buffer (const char *buffer, size_t len, void *resblock)
 191 {
 192   struct md5_ctx ctx;
 193
 194   /* Initialize the computation context.  */
 195   md5_init_ctx (&ctx);
 196
 197   /* Process whole buffer but last len % 64 bytes.  */
 198   md5_process_bytes (buffer, len, &ctx);
 199
 200   /* Put result in desired memory area.  */
 201   return md5_finish_ctx (&ctx, resblock);
 202 }
 203
 204
 205 void
 206 md5_process_bytes (const void *buffer, size_t len, struct md5_ctx *ctx)
 207 {
 208   /* When we already have some bits in our internal buffer concatenate
 209      both inputs first.  */
 210   if (ctx->buflen != 0)
 211     {
 212       size_t left_over = ctx->buflen;
 213       size_t add = 128 - left_over > len ? len : 128 - left_over;
 214
 215       memcpy (&ctx->buffer[left_over], buffer, add);
 216       ctx->buflen += add;
 217
 218       if (left_over + add > 64)
 219         {
 220           md5_process_block (ctx->buffer, (left_over + add) & ~63, ctx);
 221           /* The regions in the following copy operation cannot overlap.  */
 222           memcpy (ctx->buffer, &ctx->buffer[(left_over + add) & ~63],
 223                   (left_over + add) & 63);
 224           ctx->buflen = (left_over + add) & 63;
 225         }
 226
 227       buffer = (const void *) ((const char *) buffer + add);
 228       len -= add;
 229     }
 230
 231   /* Process available complete blocks.  */
 232   if (len > 64)
 233     {
 234 #if !_STRING_ARCH_unaligned || defined UBSAN_BOOTSTRAP
 235 /* To check alignment gcc has an appropriate operator.  Other
 236    compilers don't.  */
 237 # if __GNUC__ >= 2
 238 #  define UNALIGNED_P(p) (((md5_uintptr) p) % __alignof__ (md5_uint32) != 0)
 239 # else
 240 #  define UNALIGNED_P(p) (((md5_uintptr) p) % sizeof (md5_uint32) != 0)
 241 # endif
 242       if (UNALIGNED_P (buffer))
 243         while (len > 64)
 244           {
 245             memcpy (ctx->buffer, buffer, 64);
 246             md5_process_block (ctx->buffer, 64, ctx);
 247             buffer = (const char *) buffer + 64;
 248             len -= 64;
 249           }
 250       else
 251 #endif
 252         {
 253           md5_process_block (buffer, len & ~63, ctx);
 254           buffer = (const void *) ((const char *) buffer + (len & ~63));
 255           len &= 63;
 256         }
 257     }
 258
 259   /* Move remaining bytes in internal buffer.  */
 260   if (len > 0)
 261     {
 262       memcpy (ctx->buffer, buffer, len);
 263       ctx->buflen = len;
 264     }
 265 }
 266
 267
 268 /* These are the four functions used in the four steps of the MD5 algorithm
 269    and defined in the RFC 1321.  The first function is a little bit optimized
 270    (as found in Colin Plumbs public domain implementation).  */
 271 /* #define FF(b, c, d) ((b & c) | (~b & d)) */
 272 #define FF(b, c, d) (d ^ (b & (c ^ d)))
 273 #define FG(b, c, d) FF (d, b, c)
 274 #define FH(b, c, d) (b ^ c ^ d)
 275 #define FI(b, c, d) (c ^ (b | ~d))
 276
 277 /* Process LEN bytes of BUFFER, accumulating context into CTX.
 278    It is assumed that LEN % 64 == 0.  */
 279
 280 void
 281 md5_process_block (const void *buffer, size_t len, struct md5_ctx *ctx)
 282 {
 283   md5_uint32 correct_words[16];
 284   const md5_uint32 *words = (const md5_uint32 *) buffer;
 285   size_t nwords = len / sizeof (md5_uint32);
 286   const md5_uint32 *endp = words + nwords;
 287   md5_uint32 A = ctx->A;
 288   md5_uint32 B = ctx->B;
 289   md5_uint32 C = ctx->C;
 290   md5_uint32 D = ctx->D;
 291
 292   /* First increment the byte count.  RFC 1321 specifies the possible
 293      length of the file up to 2^64 bits.  Here we only compute the
 294      number of bytes.  Do a double word increment.  */
 295   ctx->total[0] += len;
 296   ctx->total[1] += ((len >> 31) >> 1) + (ctx->total[0] < len);
 297
 298   /* Process all bytes in the buffer with 64 bytes in each round of
 299      the loop.  */
 300   while (words < endp)
 301     {
 302       md5_uint32 *cwp = correct_words;
 303       md5_uint32 A_save = A;
 304       md5_uint32 B_save = B;
 305       md5_uint32 C_save = C;
 306       md5_uint32 D_save = D;
 307
 308       /* First round: using the given function, the context and a constant
 309          the next context is computed.  Because the algorithms processing
 310          unit is a 32-bit word and it is determined to work on words in
 311          little endian byte order we perhaps have to change the byte order
 312          before the computation.  To reduce the work for the next steps
 313          we store the swapped words in the array CORRECT_WORDS.  */
 314
 315 #define OP(a, b, c, d, s, T)                                            \
 316       do                                                                \
 317         {                                                               \
 318           a += FF (b, c, d) + (*cwp++ = SWAP (*words)) + T;             \
 319           ++words;                                                      \
 320           CYCLIC (a, s);                                                \
 321           a += b;                                                       \
 322         }                                                               \
 323       while (0)
 324
 325       /* It is unfortunate that C does not provide an operator for
 326          cyclic rotation.  Hope the C compiler is smart enough.  */
 327 #define CYCLIC(w, s) (w = (w << s) | (w >> (32 - s)))
 328
 329       /* Before we start, one word to the strange constants.
 330          They are defined in RFC 1321 as
 331
 332          T[i] = (int) (4294967296.0 * fabs (sin (i))), i=1..64
 333        */
 334
 335       /* Round 1.  */
 336       OP (A, B, C, D,  7, (md5_uint32) 0xd76aa478);
 337       OP (D, A, B, C, 12, (md5_uint32) 0xe8c7b756);
 338       OP (C, D, A, B, 17, (md5_uint32) 0x242070db);
 339       OP (B, C, D, A, 22, (md5_uint32) 0xc1bdceee);
 340       OP (A, B, C, D,  7, (md5_uint32) 0xf57c0faf);
 341       OP (D, A, B, C, 12, (md5_uint32) 0x4787c62a);
 342       OP (C, D, A, B, 17, (md5_uint32) 0xa8304613);
 343       OP (B, C, D, A, 22, (md5_uint32) 0xfd469501);
 344       OP (A, B, C, D,  7, (md5_uint32) 0x698098d8);
 345       OP (D, A, B, C, 12, (md5_uint32) 0x8b44f7af);
 346       OP (C, D, A, B, 17, (md5_uint32) 0xffff5bb1);
 347       OP (B, C, D, A, 22, (md5_uint32) 0x895cd7be);
 348       OP (A, B, C, D,  7, (md5_uint32) 0x6b901122);
 349       OP (D, A, B, C, 12, (md5_uint32) 0xfd987193);
 350       OP (C, D, A, B, 17, (md5_uint32) 0xa679438e);
 351       OP (B, C, D, A, 22, (md5_uint32) 0x49b40821);
 352
 353       /* For the second to fourth round we have the possibly swapped words
 354          in CORRECT_WORDS.  Redefine the macro to take an additional first
 355          argument specifying the function to use.  */
 356 #undef OP
 357 #define OP(a, b, c, d, k, s, T)                                         \
 358       do                                                                \
 359         {                                                               \
 360           a += FX (b, c, d) + correct_words[k] + T;                     \
 361           CYCLIC (a, s);                                                \
 362           a += b;                                                       \
 363         }                                                               \
 364       while (0)
 365
 366 #define FX(b, c, d) FG (b, c, d)
 367
 368       /* Round 2.  */
 369       OP (A, B, C, D,  1,  5, (md5_uint32) 0xf61e2562);
 370       OP (D, A, B, C,  6,  9, (md5_uint32) 0xc040b340);
 371       OP (C, D, A, B, 11, 14, (md5_uint32) 0x265e5a51);
 372       OP (B, C, D, A,  0, 20, (md5_uint32) 0xe9b6c7aa);
 373       OP (A, B, C, D,  5,  5, (md5_uint32) 0xd62f105d);
 374       OP (D, A, B, C, 10,  9, (md5_uint32) 0x02441453);
 375       OP (C, D, A, B, 15, 14, (md5_uint32) 0xd8a1e681);
 376       OP (B, C, D, A,  4, 20, (md5_uint32) 0xe7d3fbc8);
 377       OP (A, B, C, D,  9,  5, (md5_uint32) 0x21e1cde6);
 378       OP (D, A, B, C, 14,  9, (md5_uint32) 0xc33707d6);
 379       OP (C, D, A, B,  3, 14, (md5_uint32) 0xf4d50d87);
 380       OP (B, C, D, A,  8, 20, (md5_uint32) 0x455a14ed);
 381       OP (A, B, C, D, 13,  5, (md5_uint32) 0xa9e3e905);
 382       OP (D, A, B, C,  2,  9, (md5_uint32) 0xfcefa3f8);
 383       OP (C, D, A, B,  7, 14, (md5_uint32) 0x676f02d9);
 384       OP (B, C, D, A, 12, 20, (md5_uint32) 0x8d2a4c8a);
 385
 386 #undef FX
 387 #define FX(b, c, d) FH (b, c, d)
 388
 389       /* Round 3.  */
 390       OP (A, B, C, D,  5,  4, (md5_uint32) 0xfffa3942);
 391       OP (D, A, B, C,  8, 11, (md5_uint32) 0x8771f681);
 392       OP (C, D, A, B, 11, 16, (md5_uint32) 0x6d9d6122);
 393       OP (B, C, D, A, 14, 23, (md5_uint32) 0xfde5380c);
 394       OP (A, B, C, D,  1,  4, (md5_uint32) 0xa4beea44);
 395       OP (D, A, B, C,  4, 11, (md5_uint32) 0x4bdecfa9);
 396       OP (C, D, A, B,  7, 16, (md5_uint32) 0xf6bb4b60);
 397       OP (B, C, D, A, 10, 23, (md5_uint32) 0xbebfbc70);
 398       OP (A, B, C, D, 13,  4, (md5_uint32) 0x289b7ec6);
 399       OP (D, A, B, C,  0, 11, (md5_uint32) 0xeaa127fa);
 400       OP (C, D, A, B,  3, 16, (md5_uint32) 0xd4ef3085);
 401       OP (B, C, D, A,  6, 23, (md5_uint32) 0x04881d05);
 402       OP (A, B, C, D,  9,  4, (md5_uint32) 0xd9d4d039);
 403       OP (D, A, B, C, 12, 11, (md5_uint32) 0xe6db99e5);
 404       OP (C, D, A, B, 15, 16, (md5_uint32) 0x1fa27cf8);
 405       OP (B, C, D, A,  2, 23, (md5_uint32) 0xc4ac5665);
 406
 407 #undef FX
 408 #define FX(b, c, d) FI (b, c, d)
 409
 410       /* Round 4.  */
 411       OP (A, B, C, D,  0,  6, (md5_uint32) 0xf4292244);
 412       OP (D, A, B, C,  7, 10, (md5_uint32) 0x432aff97);
 413       OP (C, D, A, B, 14, 15, (md5_uint32) 0xab9423a7);
 414       OP (B, C, D, A,  5, 21, (md5_uint32) 0xfc93a039);
 415       OP (A, B, C, D, 12,  6, (md5_uint32) 0x655b59c3);
 416       OP (D, A, B, C,  3, 10, (md5_uint32) 0x8f0ccc92);
 417       OP (C, D, A, B, 10, 15, (md5_uint32) 0xffeff47d);
 418       OP (B, C, D, A,  1, 21, (md5_uint32) 0x85845dd1);
 419       OP (A, B, C, D,  8,  6, (md5_uint32) 0x6fa87e4f);
 420       OP (D, A, B, C, 15, 10, (md5_uint32) 0xfe2ce6e0);
 421       OP (C, D, A, B,  6, 15, (md5_uint32) 0xa3014314);
 422       OP (B, C, D, A, 13, 21, (md5_uint32) 0x4e0811a1);
 423       OP (A, B, C, D,  4,  6, (md5_uint32) 0xf7537e82);
 424       OP (D, A, B, C, 11, 10, (md5_uint32) 0xbd3af235);
 425       OP (C, D, A, B,  2, 15, (md5_uint32) 0x2ad7d2bb);
 426       OP (B, C, D, A,  9, 21, (md5_uint32) 0xeb86d391);
 427
 428       /* Add the starting values of the context.  */
 429       A += A_save;
 430       B += B_save;
 431       C += C_save;
 432       D += D_save;
 433     }
 434
 435   /* Put checksum in context given as argument.  */
 436   ctx->A = A;
 437   ctx->B = B;
 438   ctx->C = C;
 439   ctx->D = D;
 440 }