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) 1995-2011 Red Hat, Inc.
4 This file is part of elfutils.
5 Written by Ulrich Drepper <drepper@redhat.com>, 1995.
6
7 This file is free software; you can redistribute it and/or modify
8 it under the terms of either
9
10 * the GNU Lesser General Public License as published by the Free
11 Software Foundation; either version 3 of the License, or (at
12 your option) any later version
13
14 or
15
16 * the GNU General Public License as published by the Free
17 Software Foundation; either version 2 of the License, or (at
18 your option) any later version
19
20 or both in parallel, as here.
21
22 elfutils is distributed in the hope that it will be useful, but
23 WITHOUT ANY WARRANTY; without even the implied warranty of
24 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
25 General Public License for more details.
26
27 You should have received copies of the GNU General Public License and
28 the GNU Lesser General Public License along with this program. If
29 not, see <http://www.gnu.org/licenses/>. */
30
31 #ifdef HAVE_CONFIG_H
32 # include <config.h>
33 #endif
34
35 #include <stdlib.h>
36 #include <string.h>
37 #include <sys/types.h>
38
39 #include "md5.h"
40 #include "system.h"
41
42 #define SWAP(n) LE32 (n)
43
44 /* This array contains the bytes used to pad the buffer to the next
45 64-byte boundary. (RFC 1321, 3.1: Step 1) */
46 static const unsigned char fillbuf[64] = { 0x80, 0 /* , 0, 0, ... */ };
47
48
49 /* Initialize structure containing state of computation.
50 (RFC 1321, 3.3: Step 3) */
51 void
md5_init_ctx(ctx)52 md5_init_ctx (ctx)
53 struct md5_ctx *ctx;
54 {
55 ctx->A = 0x67452301;
56 ctx->B = 0xefcdab89;
57 ctx->C = 0x98badcfe;
58 ctx->D = 0x10325476;
59
60 ctx->total[0] = ctx->total[1] = 0;
61 ctx->buflen = 0;
62 }
63
64 /* Put result from CTX in first 16 bytes following RESBUF. The result
65 must be in little endian byte order.
66
67 IMPORTANT: On some systems it is required that RESBUF is correctly
68 aligned for a 32 bits value. */
69 void *
md5_read_ctx(ctx,resbuf)70 md5_read_ctx (ctx, resbuf)
71 const struct md5_ctx *ctx;
72 void *resbuf;
73 {
74 ((md5_uint32 *) resbuf)[0] = SWAP (ctx->A);
75 ((md5_uint32 *) resbuf)[1] = SWAP (ctx->B);
76 ((md5_uint32 *) resbuf)[2] = SWAP (ctx->C);
77 ((md5_uint32 *) resbuf)[3] = SWAP (ctx->D);
78
79 return resbuf;
80 }
81
82 static void
le64_copy(char * dest,uint64_t x)83 le64_copy (char *dest, uint64_t x)
84 {
85 for (size_t i = 0; i < 8; ++i)
86 {
87 dest[i] = (uint8_t) x;
88 x >>= 8;
89 }
90 }
91
92 /* Process the remaining bytes in the internal buffer and the usual
93 prolog according to the standard and write the result to RESBUF.
94
95 IMPORTANT: On some systems it is required that RESBUF is correctly
96 aligned for a 32 bits value. */
97 void *
md5_finish_ctx(ctx,resbuf)98 md5_finish_ctx (ctx, resbuf)
99 struct md5_ctx *ctx;
100 void *resbuf;
101 {
102 /* Take yet unprocessed bytes into account. */
103 md5_uint32 bytes = ctx->buflen;
104 size_t pad;
105
106 /* Now count remaining bytes. */
107 ctx->total[0] += bytes;
108 if (ctx->total[0] < bytes)
109 ++ctx->total[1];
110
111 pad = bytes >= 56 ? 64 + 56 - bytes : 56 - bytes;
112 memcpy (&ctx->buffer[bytes], fillbuf, pad);
113
114 /* Put the 64-bit file length in *bits* at the end of the buffer. */
115 const uint64_t bit_length = ((ctx->total[0] << 3)
116 + ((uint64_t) ((ctx->total[1] << 3) |
117 (ctx->total[0] >> 29)) << 32));
118 le64_copy (&ctx->buffer[bytes + pad], bit_length);
119
120 /* Process last bytes. */
121 md5_process_block (ctx->buffer, bytes + pad + 8, ctx);
122
123 return md5_read_ctx (ctx, resbuf);
124 }
125
126
127 #ifdef NEED_MD5_STREAM
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
md5_stream(stream,resblock)132 md5_stream (stream, resblock)
133 FILE *stream;
134 void *resblock;
135 {
136 /* Important: BLOCKSIZE must be a multiple of 64. */
137 #define BLOCKSIZE 4096
138 struct md5_ctx ctx;
139 char buffer[BLOCKSIZE + 72];
140 size_t sum;
141
142 /* Initialize the computation context. */
143 md5_init_ctx (&ctx);
144
145 /* Iterate over full file contents. */
146 while (1)
147 {
148 /* We read the file in blocks of BLOCKSIZE bytes. One call of the
149 computation function processes the whole buffer so that with the
150 next round of the loop another block can be read. */
151 size_t n;
152 sum = 0;
153
154 /* Read block. Take care for partial reads. */
155 do
156 {
157 n = fread (buffer + sum, 1, BLOCKSIZE - sum, stream);
158
159 sum += n;
160 }
161 while (sum < BLOCKSIZE && n != 0);
162 if (n == 0 && ferror (stream))
163 return 1;
164
165 /* If end of file is reached, end the loop. */
166 if (n == 0)
167 break;
168
169 /* Process buffer with BLOCKSIZE bytes. Note that
170 BLOCKSIZE % 64 == 0
171 */
172 md5_process_block (buffer, BLOCKSIZE, &ctx);
173 }
174
175 /* Add the last bytes if necessary. */
176 if (sum > 0)
177 md5_process_bytes (buffer, sum, &ctx);
178
179 /* Construct result in desired memory. */
180 md5_finish_ctx (&ctx, resblock);
181 return 0;
182 }
183 #endif
184
185
186 #ifdef NEED_MD5_BUFFER
187 /* Compute MD5 message digest for LEN bytes beginning at BUFFER. The
188 result is always in little endian byte order, so that a byte-wise
189 output yields to the wanted ASCII representation of the message
190 digest. */
191 void *
md5_buffer(buffer,len,resblock)192 md5_buffer (buffer, len, resblock)
193 const char *buffer;
194 size_t len;
195 void *resblock;
196 {
197 struct md5_ctx ctx;
198
199 /* Initialize the computation context. */
200 md5_init_ctx (&ctx);
201
202 /* Process whole buffer but last len % 64 bytes. */
203 md5_process_bytes (buffer, len, &ctx);
204
205 /* Put result in desired memory area. */
206 return md5_finish_ctx (&ctx, resblock);
207 }
208 #endif
209
210
211 void
md5_process_bytes(buffer,len,ctx)212 md5_process_bytes (buffer, len, ctx)
213 const void *buffer;
214 size_t len;
215 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 (&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, &ctx->buffer[(left_over + add) & ~63],
234 ctx->buflen);
235 }
236
237 buffer = (const char *) buffer + add;
238 len -= add;
239 }
240
241 /* Process available complete blocks. */
242 if (len >= 64)
243 {
244 #if !_STRING_ARCH_unaligned
245 /* To check alignment gcc has an appropriate operator. Other
246 compilers don't. */
247 # if __GNUC__ >= 2
248 # define UNALIGNED_P(p) (((md5_uintptr) p) % __alignof__ (md5_uint32) != 0)
249 # else
250 # define UNALIGNED_P(p) (((md5_uintptr) p) % sizeof (md5_uint32) != 0)
251 # endif
252 if (UNALIGNED_P (buffer))
253 while (len > 64)
254 {
255 md5_process_block (memcpy (ctx->buffer, buffer, 64), 64, ctx);
256 buffer = (const char *) buffer + 64;
257 len -= 64;
258 }
259 else
260 #endif
261 {
262 md5_process_block (buffer, len & ~63, ctx);
263 buffer = (const char *) buffer + (len & ~63);
264 len &= 63;
265 }
266 }
267
268 /* Move remaining bytes in internal buffer. */
269 if (len > 0)
270 {
271 size_t left_over = ctx->buflen;
272
273 memcpy (&ctx->buffer[left_over], buffer, len);
274 left_over += len;
275 if (left_over >= 64)
276 {
277 md5_process_block (ctx->buffer, 64, ctx);
278 left_over -= 64;
279 memcpy (ctx->buffer, &ctx->buffer[64], left_over);
280 }
281 ctx->buflen = left_over;
282 }
283 }
284
285
286 /* These are the four functions used in the four steps of the MD5 algorithm
287 and defined in the RFC 1321. The first function is a little bit optimized
288 (as found in Colin Plumbs public domain implementation). */
289 /* #define FF(b, c, d) ((b & c) | (~b & d)) */
290 #define FF(b, c, d) (d ^ (b & (c ^ d)))
291 #define FG(b, c, d) FF (d, b, c)
292 #define FH(b, c, d) (b ^ c ^ d)
293 #define FI(b, c, d) (c ^ (b | ~d))
294
295 /* Process LEN bytes of BUFFER, accumulating context into CTX.
296 It is assumed that LEN % 64 == 0. */
297
298 void
md5_process_block(buffer,len,ctx)299 md5_process_block (buffer, len, ctx)
300 const void *buffer;
301 size_t len;
302 struct md5_ctx *ctx;
303 {
304 md5_uint32 correct_words[16];
305 const md5_uint32 *words = buffer;
306 size_t nwords = len / sizeof (md5_uint32);
307 const md5_uint32 *endp = words + nwords;
308 md5_uint32 A = ctx->A;
309 md5_uint32 B = ctx->B;
310 md5_uint32 C = ctx->C;
311 md5_uint32 D = ctx->D;
312
313 /* First increment the byte count. RFC 1321 specifies the possible
314 length of the file up to 2^64 bits. Here we only compute the
315 number of bytes. Do a double word increment. */
316 ctx->total[0] += len;
317 if (ctx->total[0] < len)
318 ++ctx->total[1];
319
320 /* Process all bytes in the buffer with 64 bytes in each round of
321 the loop. */
322 while (words < endp)
323 {
324 md5_uint32 *cwp = correct_words;
325 md5_uint32 A_save = A;
326 md5_uint32 B_save = B;
327 md5_uint32 C_save = C;
328 md5_uint32 D_save = D;
329
330 /* First round: using the given function, the context and a constant
331 the next context is computed. Because the algorithms processing
332 unit is a 32-bit word and it is determined to work on words in
333 little endian byte order we perhaps have to change the byte order
334 before the computation. To reduce the work for the next steps
335 we store the swapped words in the array CORRECT_WORDS. */
336
337 #define OP(a, b, c, d, s, T) \
338 do \
339 { \
340 a += FF (b, c, d) + (*cwp++ = SWAP (*words)) + T; \
341 ++words; \
342 CYCLIC (a, s); \
343 a += b; \
344 } \
345 while (0)
346
347 /* It is unfortunate that C does not provide an operator for
348 cyclic rotation. Hope the C compiler is smart enough. */
349 #define CYCLIC(w, s) (w = (w << s) | (w >> (32 - s)))
350
351 /* Before we start, one word to the strange constants.
352 They are defined in RFC 1321 as
353
354 T[i] = (int) (4294967296.0 * fabs (sin (i))), i=1..64
355 */
356
357 /* Round 1. */
358 OP (A, B, C, D, 7, 0xd76aa478);
359 OP (D, A, B, C, 12, 0xe8c7b756);
360 OP (C, D, A, B, 17, 0x242070db);
361 OP (B, C, D, A, 22, 0xc1bdceee);
362 OP (A, B, C, D, 7, 0xf57c0faf);
363 OP (D, A, B, C, 12, 0x4787c62a);
364 OP (C, D, A, B, 17, 0xa8304613);
365 OP (B, C, D, A, 22, 0xfd469501);
366 OP (A, B, C, D, 7, 0x698098d8);
367 OP (D, A, B, C, 12, 0x8b44f7af);
368 OP (C, D, A, B, 17, 0xffff5bb1);
369 OP (B, C, D, A, 22, 0x895cd7be);
370 OP (A, B, C, D, 7, 0x6b901122);
371 OP (D, A, B, C, 12, 0xfd987193);
372 OP (C, D, A, B, 17, 0xa679438e);
373 OP (B, C, D, A, 22, 0x49b40821);
374
375 /* For the second to fourth round we have the possibly swapped words
376 in CORRECT_WORDS. Redefine the macro to take an additional first
377 argument specifying the function to use. */
378 #undef OP
379 #define OP(f, a, b, c, d, k, s, T) \
380 do \
381 { \
382 a += f (b, c, d) + correct_words[k] + T; \
383 CYCLIC (a, s); \
384 a += b; \
385 } \
386 while (0)
387
388 /* Round 2. */
389 OP (FG, A, B, C, D, 1, 5, 0xf61e2562);
390 OP (FG, D, A, B, C, 6, 9, 0xc040b340);
391 OP (FG, C, D, A, B, 11, 14, 0x265e5a51);
392 OP (FG, B, C, D, A, 0, 20, 0xe9b6c7aa);
393 OP (FG, A, B, C, D, 5, 5, 0xd62f105d);
394 OP (FG, D, A, B, C, 10, 9, 0x02441453);
395 OP (FG, C, D, A, B, 15, 14, 0xd8a1e681);
396 OP (FG, B, C, D, A, 4, 20, 0xe7d3fbc8);
397 OP (FG, A, B, C, D, 9, 5, 0x21e1cde6);
398 OP (FG, D, A, B, C, 14, 9, 0xc33707d6);
399 OP (FG, C, D, A, B, 3, 14, 0xf4d50d87);
400 OP (FG, B, C, D, A, 8, 20, 0x455a14ed);
401 OP (FG, A, B, C, D, 13, 5, 0xa9e3e905);
402 OP (FG, D, A, B, C, 2, 9, 0xfcefa3f8);
403 OP (FG, C, D, A, B, 7, 14, 0x676f02d9);
404 OP (FG, B, C, D, A, 12, 20, 0x8d2a4c8a);
405
406 /* Round 3. */
407 OP (FH, A, B, C, D, 5, 4, 0xfffa3942);
408 OP (FH, D, A, B, C, 8, 11, 0x8771f681);
409 OP (FH, C, D, A, B, 11, 16, 0x6d9d6122);
410 OP (FH, B, C, D, A, 14, 23, 0xfde5380c);
411 OP (FH, A, B, C, D, 1, 4, 0xa4beea44);
412 OP (FH, D, A, B, C, 4, 11, 0x4bdecfa9);
413 OP (FH, C, D, A, B, 7, 16, 0xf6bb4b60);
414 OP (FH, B, C, D, A, 10, 23, 0xbebfbc70);
415 OP (FH, A, B, C, D, 13, 4, 0x289b7ec6);
416 OP (FH, D, A, B, C, 0, 11, 0xeaa127fa);
417 OP (FH, C, D, A, B, 3, 16, 0xd4ef3085);
418 OP (FH, B, C, D, A, 6, 23, 0x04881d05);
419 OP (FH, A, B, C, D, 9, 4, 0xd9d4d039);
420 OP (FH, D, A, B, C, 12, 11, 0xe6db99e5);
421 OP (FH, C, D, A, B, 15, 16, 0x1fa27cf8);
422 OP (FH, B, C, D, A, 2, 23, 0xc4ac5665);
423
424 /* Round 4. */
425 OP (FI, A, B, C, D, 0, 6, 0xf4292244);
426 OP (FI, D, A, B, C, 7, 10, 0x432aff97);
427 OP (FI, C, D, A, B, 14, 15, 0xab9423a7);
428 OP (FI, B, C, D, A, 5, 21, 0xfc93a039);
429 OP (FI, A, B, C, D, 12, 6, 0x655b59c3);
430 OP (FI, D, A, B, C, 3, 10, 0x8f0ccc92);
431 OP (FI, C, D, A, B, 10, 15, 0xffeff47d);
432 OP (FI, B, C, D, A, 1, 21, 0x85845dd1);
433 OP (FI, A, B, C, D, 8, 6, 0x6fa87e4f);
434 OP (FI, D, A, B, C, 15, 10, 0xfe2ce6e0);
435 OP (FI, C, D, A, B, 6, 15, 0xa3014314);
436 OP (FI, B, C, D, A, 13, 21, 0x4e0811a1);
437 OP (FI, A, B, C, D, 4, 6, 0xf7537e82);
438 OP (FI, D, A, B, C, 11, 10, 0xbd3af235);
439 OP (FI, C, D, A, B, 2, 15, 0x2ad7d2bb);
440 OP (FI, B, C, D, A, 9, 21, 0xeb86d391);
441
442 /* Add the starting values of the context. */
443 A += A_save;
444 B += B_save;
445 C += C_save;
446 D += D_save;
447 }
448
449 /* Put checksum in context given as argument. */
450 ctx->A = A;
451 ctx->B = B;
452 ctx->C = C;
453 ctx->D = D;
454 }
455