checksum.c

   1 /*
   2  * Routines to support checksumming of bytes.
   3  *
   4  * Copyright (C) 1996 Andrew Tridgell
   5  * Copyright (C) 1996 Paul Mackerras
   6  * Copyright (C) 2004-2020 Wayne Davison
   7  *
   8  * This program is free software; you can redistribute it and/or modify
   9  * it under the terms of the GNU General Public License as published by
  10  * the Free Software Foundation; either version 3 of the License, or
  11  * (at your option) any later version.
  12  *
  13  * In addition, as a special exception, the copyright holders give
  14  * permission to dynamically link rsync with the OpenSSL and xxhash
  15  * libraries when those libraries are being distributed in compliance
  16  * with their license terms, and to distribute a dynamically linked
  17  * combination of rsync and these libraries.  This is also considered
  18  * to be covered under the GPL's System Libraries exception.
  19  *
  20  * This program is distributed in the hope that it will be useful,
  21  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  22  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  23  * GNU General Public License for more details.
  24  *
  25  * You should have received a copy of the GNU General Public License along
  26  * with this program; if not, visit the http://fsf.org website.
  27  */
  28
  29 #include "rsync.h"
  30 #ifdef SUPPORT_XXHASH
  31 #include "xxhash.h"
  32 #endif
  33
  34 extern int am_server;
  35 extern int whole_file;
  36 extern int checksum_seed;
  37 extern int protocol_version;
  38 extern int proper_seed_order;
  39 extern const char *checksum_choice;
  40
  41 struct name_num_obj valid_checksums = {
  42         "checksum", NULL, NULL, 0, 0, {
  43 #ifdef SUPPORT_XXHASH
  44                 { CSUM_XXH64, "xxh64", NULL },
  45                 { CSUM_XXH64, "xxhash", NULL },
  46 #endif
  47                 { CSUM_MD5, "md5", NULL },
  48                 { CSUM_MD4, "md4", NULL },
  49                 { CSUM_NONE, "none", NULL },
  50                 { 0, NULL, NULL }
  51         }
  52 };
  53
  54 int xfersum_type = 0; /* used for the file transfer checksums */
  55 int checksum_type = 0; /* used for the pre-transfer (--checksum) checksums */
  56
  57 int parse_csum_name(const char *name, int len)
  58 {
  59         struct name_num_item *nni;
  60
  61         if (len < 0 && name)
  62                 len = strlen(name);
  63
  64         if (!name || (len == 4 && strncasecmp(name, "auto", 4) == 0)) {
  65                 if (protocol_version >= 30)
  66                         return CSUM_MD5;
  67                 if (protocol_version >= 27)
  68                         return CSUM_MD4_OLD;
  69                 if (protocol_version >= 21)
  70                         return CSUM_MD4_BUSTED;
  71                 return CSUM_MD4_ARCHAIC;
  72         }
  73
  74         nni = get_nni_by_name(&valid_checksums, name, len);
  75
  76         if (!nni) {
  77                 rprintf(FERROR, "unknown checksum name: %s\n", name);
  78                 exit_cleanup(RERR_UNSUPPORTED);
  79         }
  80
  81         return nni->num;
  82 }
  83
  84 static const char *checksum_name(int num)
  85 {
  86         struct name_num_item *nni = get_nni_by_num(&valid_checksums, num);
  87
  88         return nni ? nni->name : num < CSUM_MD4 ? "MD4" : "UNKNOWN";
  89 }
  90
  91 void parse_checksum_choice(int final_call)
  92 {
  93         if (valid_checksums.negotiated_name)
  94                 xfersum_type = checksum_type = valid_checksums.negotiated_num;
  95         else {
  96                 char *cp = checksum_choice ? strchr(checksum_choice, ',') : NULL;
  97                 if (cp) {
  98                         xfersum_type = parse_csum_name(checksum_choice, cp - checksum_choice);
  99                         checksum_type = parse_csum_name(cp+1, -1);
 100                 } else
 101                         xfersum_type = checksum_type = parse_csum_name(checksum_choice, -1);
 102         }
 103
 104         if (xfersum_type == CSUM_NONE)
 105                 whole_file = 1;
 106
 107         /* Snag the checksum name for both write_batch's option output & the following debug output. */
 108         if (valid_checksums.negotiated_name)
 109                 checksum_choice = valid_checksums.negotiated_name;
 110         else if (checksum_choice == NULL)
 111                 checksum_choice = checksum_name(xfersum_type);
 112
 113         if (final_call && DEBUG_GTE(NSTR, am_server ? 3 : 1)) {
 114                 rprintf(FINFO, "%s%s checksum: %s\n",
 115                         am_server ? "Server" : "Client",
 116                         valid_checksums.negotiated_name ? " negotiated" : "",
 117                         checksum_choice);
 118         }
 119 }
 120
 121 int csum_len_for_type(int cst, BOOL flist_csum)
 122 {
 123         switch (cst) {
 124           case CSUM_NONE:
 125                 return 1;
 126           case CSUM_MD4_ARCHAIC:
 127                 /* The oldest checksum code is rather weird: the file-list code only sent
 128                  * 2-byte checksums, but all other checksums were full MD4 length. */
 129                 return flist_csum ? 2 : MD4_DIGEST_LEN;
 130           case CSUM_MD4:
 131           case CSUM_MD4_OLD:
 132           case CSUM_MD4_BUSTED:
 133                 return MD4_DIGEST_LEN;
 134           case CSUM_MD5:
 135                 return MD5_DIGEST_LEN;
 136 #ifdef SUPPORT_XXHASH
 137           case CSUM_XXH64:
 138                 return 64/8;
 139 #endif
 140           default: /* paranoia to prevent missing case values */
 141                 exit_cleanup(RERR_UNSUPPORTED);
 142         }
 143         return 0;
 144 }
 145
 146 /* Returns 0 if the checksum is not canonical (i.e. it includes a seed value).
 147  * Returns 1 if the public sum order matches our internal sum order.
 148  * Returns -1 if the public sum order is the reverse of our internal sum order.
 149  */
 150 int canonical_checksum(int csum_type)
 151 {
 152         switch (csum_type) {
 153           case CSUM_NONE:
 154           case CSUM_MD4_ARCHAIC:
 155           case CSUM_MD4_OLD:
 156           case CSUM_MD4_BUSTED:
 157                 break;
 158           case CSUM_MD4:
 159           case CSUM_MD5:
 160                 return -1;
 161 #ifdef SUPPORT_XXHASH
 162           case CSUM_XXH64:
 163                 return 1;
 164 #endif
 165           default: /* paranoia to prevent missing case values */
 166                 exit_cleanup(RERR_UNSUPPORTED);
 167         }
 168         return 0;
 169 }
 170
 171 #ifndef HAVE_SIMD /* See simd-checksum-*.cpp. */
 172 /*
 173   a simple 32 bit checksum that can be updated from either end
 174   (inspired by Mark Adler's Adler-32 checksum)
 175   */
 176 uint32 get_checksum1(char *buf1, int32 len)
 177 {
 178         int32 i;
 179         uint32 s1, s2;
 180         schar *buf = (schar *)buf1;
 181
 182         s1 = s2 = 0;
 183         for (i = 0; i < (len-4); i+=4) {
 184                 s2 += 4*(s1 + buf[i]) + 3*buf[i+1] + 2*buf[i+2] + buf[i+3] + 10*CHAR_OFFSET;
 185                 s1 += (buf[i+0] + buf[i+1] + buf[i+2] + buf[i+3] + 4*CHAR_OFFSET);
 186         }
 187         for (; i < len; i++) {
 188                 s1 += (buf[i]+CHAR_OFFSET); s2 += s1;
 189         }
 190         return (s1 & 0xffff) + (s2 << 16);
 191 }
 192 #endif
 193
 194 void get_checksum2(char *buf, int32 len, char *sum)
 195 {
 196         switch (xfersum_type) {
 197 #ifdef SUPPORT_XXHASH
 198           case CSUM_XXH64:
 199                 SIVAL64(sum, 0, XXH64(buf, len, checksum_seed));
 200                 break;
 201 #endif
 202           case CSUM_MD5: {
 203                 MD5_CTX m5;
 204                 uchar seedbuf[4];
 205                 MD5_Init(&m5);
 206                 if (proper_seed_order) {
 207                         if (checksum_seed) {
 208                                 SIVALu(seedbuf, 0, checksum_seed);
 209                                 MD5_Update(&m5, seedbuf, 4);
 210                         }
 211                         MD5_Update(&m5, (uchar *)buf, len);
 212                 } else {
 213                         MD5_Update(&m5, (uchar *)buf, len);
 214                         if (checksum_seed) {
 215                                 SIVALu(seedbuf, 0, checksum_seed);
 216                                 MD5_Update(&m5, seedbuf, 4);
 217                         }
 218                 }
 219                 MD5_Final((uchar *)sum, &m5);
 220                 break;
 221           }
 222           case CSUM_MD4:
 223 #ifdef USE_OPENSSL
 224           {
 225                 MD4_CTX m4;
 226                 MD4_Init(&m4);
 227                 MD4_Update(&m4, (uchar *)buf, len);
 228                 if (checksum_seed) {
 229                         uchar seedbuf[4];
 230                         SIVALu(seedbuf, 0, checksum_seed);
 231                         MD4_Update(&m4, seedbuf, 4);
 232                 }
 233                 MD4_Final((uchar *)sum, &m4);
 234                 break;
 235           }
 236 #endif
 237           case CSUM_MD4_OLD:
 238           case CSUM_MD4_BUSTED:
 239           case CSUM_MD4_ARCHAIC: {
 240                 md_context m;
 241                 int32 i;
 242                 static char *buf1;
 243                 static int32 len1;
 244
 245                 mdfour_begin(&m);
 246
 247                 if (len > len1) {
 248                         if (buf1)
 249                                 free(buf1);
 250                         buf1 = new_array(char, len+4);
 251                         len1 = len;
 252                         if (!buf1)
 253                                 out_of_memory("get_checksum2");
 254                 }
 255
 256                 memcpy(buf1, buf, len);
 257                 if (checksum_seed) {
 258                         SIVAL(buf1,len,checksum_seed);
 259                         len += 4;
 260                 }
 261
 262                 for (i = 0; i + CSUM_CHUNK <= len; i += CSUM_CHUNK)
 263                         mdfour_update(&m, (uchar *)(buf1+i), CSUM_CHUNK);
 264
 265                 /*
 266                  * Prior to version 27 an incorrect MD4 checksum was computed
 267                  * by failing to call mdfour_tail() for block sizes that
 268                  * are multiples of 64.  This is fixed by calling mdfour_update()
 269                  * even when there are no more bytes.
 270                  */
 271                 if (len - i > 0 || xfersum_type > CSUM_MD4_BUSTED)
 272                         mdfour_update(&m, (uchar *)(buf1+i), len-i);
 273
 274                 mdfour_result(&m, (uchar *)sum);
 275                 break;
 276           }
 277           default: /* paranoia to prevent missing case values */
 278                 exit_cleanup(RERR_UNSUPPORTED);
 279         }
 280 }
 281
 282 void file_checksum(const char *fname, const STRUCT_STAT *st_p, char *sum)
 283 {
 284         struct map_struct *buf;
 285         OFF_T i, len = st_p->st_size;
 286         int32 remainder;
 287         int fd;
 288
 289         memset(sum, 0, MAX_DIGEST_LEN);
 290
 291         fd = do_open(fname, O_RDONLY, 0);
 292         if (fd == -1)
 293                 return;
 294
 295         buf = map_file(fd, len, MAX_MAP_SIZE, CHUNK_SIZE);
 296
 297         switch (checksum_type) {
 298 #ifdef SUPPORT_XXHASH
 299           case CSUM_XXH64: {
 300                 static XXH64_state_t* state = NULL;
 301                 if (!state && !(state = XXH64_createState()))
 302                         out_of_memory("file_checksum");
 303
 304                 XXH64_reset(state, 0);
 305
 306                 for (i = 0; i + CHUNK_SIZE <= len; i += CHUNK_SIZE)
 307                         XXH64_update(state, (uchar *)map_ptr(buf, i, CHUNK_SIZE), CHUNK_SIZE);
 308
 309                 remainder = (int32)(len - i);
 310                 if (remainder > 0)
 311                         XXH64_update(state, (uchar *)map_ptr(buf, i, remainder), remainder);
 312
 313                 SIVAL64(sum, 0, XXH64_digest(state));
 314                 break;
 315           }
 316 #endif
 317           case CSUM_MD5: {
 318                 MD5_CTX m5;
 319
 320                 MD5_Init(&m5);
 321
 322                 for (i = 0; i + CHUNK_SIZE <= len; i += CHUNK_SIZE)
 323                         MD5_Update(&m5, (uchar *)map_ptr(buf, i, CHUNK_SIZE), CHUNK_SIZE);
 324
 325                 remainder = (int32)(len - i);
 326                 if (remainder > 0)
 327                         MD5_Update(&m5, (uchar *)map_ptr(buf, i, remainder), remainder);
 328
 329                 MD5_Final((uchar *)sum, &m5);
 330                 break;
 331           }
 332           case CSUM_MD4:
 333 #ifdef USE_OPENSSL
 334           {
 335                 MD4_CTX m4;
 336
 337                 MD4_Init(&m4);
 338
 339                 for (i = 0; i + CHUNK_SIZE <= len; i += CHUNK_SIZE)
 340                         MD4_Update(&m4, (uchar *)map_ptr(buf, i, CHUNK_SIZE), CHUNK_SIZE);
 341
 342                 remainder = (int32)(len - i);
 343                 if (remainder > 0)
 344                         MD4_Update(&m4, (uchar *)map_ptr(buf, i, remainder), remainder);
 345
 346                 MD4_Final((uchar *)sum, &m4);
 347                 break;
 348           }
 349 #endif
 350           case CSUM_MD4_OLD:
 351           case CSUM_MD4_BUSTED:
 352           case CSUM_MD4_ARCHAIC: {
 353                 md_context m;
 354
 355                 mdfour_begin(&m);
 356
 357                 for (i = 0; i + CHUNK_SIZE <= len; i += CHUNK_SIZE)
 358                         mdfour_update(&m, (uchar *)map_ptr(buf, i, CHUNK_SIZE), CHUNK_SIZE);
 359
 360                 /* Prior to version 27 an incorrect MD4 checksum was computed
 361                  * by failing to call mdfour_tail() for block sizes that
 362                  * are multiples of 64.  This is fixed by calling mdfour_update()
 363                  * even when there are no more bytes. */
 364                 remainder = (int32)(len - i);
 365                 if (remainder > 0 || checksum_type > CSUM_MD4_BUSTED)
 366                         mdfour_update(&m, (uchar *)map_ptr(buf, i, remainder), remainder);
 367
 368                 mdfour_result(&m, (uchar *)sum);
 369                 break;
 370           }
 371           default:
 372                 rprintf(FERROR, "Invalid checksum-choice for --checksum: %s (%d)\n",
 373                         checksum_name(checksum_type), checksum_type);
 374                 exit_cleanup(RERR_UNSUPPORTED);
 375         }
 376
 377         close(fd);
 378         unmap_file(buf);
 379 }
 380
 381 static int32 sumresidue;
 382 static union {
 383         md_context md;
 384 #ifdef USE_OPENSSL
 385         MD4_CTX m4;
 386 #endif
 387         MD5_CTX m5;
 388 } ctx;
 389 #ifdef SUPPORT_XXHASH
 390 static XXH64_state_t* xxh64_state;
 391 #endif
 392 static int cursum_type;
 393
 394 void sum_init(int csum_type, int seed)
 395 {
 396         char s[4];
 397
 398         if (csum_type < 0)
 399                 csum_type = parse_csum_name(NULL, 0);
 400         cursum_type = csum_type;
 401
 402         switch (csum_type) {
 403 #ifdef SUPPORT_XXHASH
 404           case CSUM_XXH64:
 405                 if (!xxh64_state && !(xxh64_state = XXH64_createState()))
 406                         out_of_memory("sum_init");
 407                 XXH64_reset(xxh64_state, 0);
 408                 break;
 409 #endif
 410           case CSUM_MD5:
 411                 MD5_Init(&ctx.m5);
 412                 break;
 413           case CSUM_MD4:
 414 #ifdef USE_OPENSSL
 415                 MD4_Init(&ctx.m4);
 416 #else
 417                 mdfour_begin(&ctx.md);
 418                 sumresidue = 0;
 419 #endif
 420                 break;
 421           case CSUM_MD4_OLD:
 422           case CSUM_MD4_BUSTED:
 423           case CSUM_MD4_ARCHAIC:
 424                 mdfour_begin(&ctx.md);
 425                 sumresidue = 0;
 426                 SIVAL(s, 0, seed);
 427                 sum_update(s, 4);
 428                 break;
 429           case CSUM_NONE:
 430                 break;
 431           default: /* paranoia to prevent missing case values */
 432                 exit_cleanup(RERR_UNSUPPORTED);
 433         }
 434 }
 435
 436 /**
 437  * Feed data into an MD4 accumulator, md.  The results may be
 438  * retrieved using sum_end().  md is used for different purposes at
 439  * different points during execution.
 440  *
 441  * @todo Perhaps get rid of md and just pass in the address each time.
 442  * Very slightly clearer and slower.
 443  **/
 444 void sum_update(const char *p, int32 len)
 445 {
 446         switch (cursum_type) {
 447 #ifdef SUPPORT_XXHASH
 448           case CSUM_XXH64:
 449                 XXH64_update(xxh64_state, p, len);
 450                 break;
 451 #endif
 452           case CSUM_MD5:
 453                 MD5_Update(&ctx.m5, (uchar *)p, len);
 454                 break;
 455           case CSUM_MD4:
 456 #ifdef USE_OPENSSL
 457                 MD4_Update(&ctx.m4, (uchar *)p, len);
 458                 break;
 459 #endif
 460           case CSUM_MD4_OLD:
 461           case CSUM_MD4_BUSTED:
 462           case CSUM_MD4_ARCHAIC:
 463                 if (len + sumresidue < CSUM_CHUNK) {
 464                         memcpy(ctx.md.buffer + sumresidue, p, len);
 465                         sumresidue += len;
 466                         break;
 467                 }
 468
 469                 if (sumresidue) {
 470                         int32 i = CSUM_CHUNK - sumresidue;
 471                         memcpy(ctx.md.buffer + sumresidue, p, i);
 472                         mdfour_update(&ctx.md, (uchar *)ctx.md.buffer, CSUM_CHUNK);
 473                         len -= i;
 474                         p += i;
 475                 }
 476
 477                 while (len >= CSUM_CHUNK) {
 478                         mdfour_update(&ctx.md, (uchar *)p, CSUM_CHUNK);
 479                         len -= CSUM_CHUNK;
 480                         p += CSUM_CHUNK;
 481                 }
 482
 483                 sumresidue = len;
 484                 if (sumresidue)
 485                         memcpy(ctx.md.buffer, p, sumresidue);
 486                 break;
 487           case CSUM_NONE:
 488                 break;
 489           default: /* paranoia to prevent missing case values */
 490                 exit_cleanup(RERR_UNSUPPORTED);
 491         }
 492 }
 493
 494 /* NOTE: all the callers of sum_end() pass in a pointer to a buffer that is
 495  * MAX_DIGEST_LEN in size, so even if the csum-len is shorter that that (i.e.
 496  * CSUM_MD4_ARCHAIC), we don't have to worry about limiting the data we write
 497  * into the "sum" buffer. */
 498 int sum_end(char *sum)
 499 {
 500         switch (cursum_type) {
 501 #ifdef SUPPORT_XXHASH
 502           case CSUM_XXH64:
 503                 SIVAL64(sum, 0, XXH64_digest(xxh64_state));
 504                 break;
 505 #endif
 506           case CSUM_MD5:
 507                 MD5_Final((uchar *)sum, &ctx.m5);
 508                 break;
 509           case CSUM_MD4:
 510 #ifdef USE_OPENSSL
 511                 MD4_Final((uchar *)sum, &ctx.m4);
 512                 break;
 513 #endif
 514           case CSUM_MD4_OLD:
 515                 mdfour_update(&ctx.md, (uchar *)ctx.md.buffer, sumresidue);
 516                 mdfour_result(&ctx.md, (uchar *)sum);
 517                 break;
 518           case CSUM_MD4_BUSTED:
 519           case CSUM_MD4_ARCHAIC:
 520                 if (sumresidue)
 521                         mdfour_update(&ctx.md, (uchar *)ctx.md.buffer, sumresidue);
 522                 mdfour_result(&ctx.md, (uchar *)sum);
 523                 break;
 524           case CSUM_NONE:
 525                 *sum = '\0';
 526                 break;
 527           default: /* paranoia to prevent missing case values */
 528                 exit_cleanup(RERR_UNSUPPORTED);
 529         }
 530
 531         return csum_len_for_type(cursum_type, 0);
 532 }