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 #ifdef USE_OPENSSL
  34 #include "openssl/md4.h"
  35 #include "openssl/md5.h"
  36 #endif
  37
  38 extern int am_server;
  39 extern int whole_file;
  40 extern int checksum_seed;
  41 extern int protocol_version;
  42 extern int proper_seed_order;
  43 extern const char *checksum_choice;
  44
  45 #define CSUM_NONE 0
  46 #define CSUM_MD4_ARCHAIC 1
  47 #define CSUM_MD4_BUSTED 2
  48 #define CSUM_MD4_OLD 3
  49 #define CSUM_MD4 4
  50 #define CSUM_MD5 5
  51 #define CSUM_XXH64 6
  52
  53 struct name_num_obj valid_checksums = {
  54         "checksum", NULL, NULL, 0, 0, {
  55 #ifdef SUPPORT_XXHASH
  56                 { CSUM_XXH64, "xxh64", NULL },
  57                 { CSUM_XXH64, "xxhash", NULL },
  58 #endif
  59                 { CSUM_MD5, "md5", NULL },
  60                 { CSUM_MD4, "md4", NULL },
  61                 { CSUM_NONE, "none", NULL },
  62                 { 0, NULL, NULL }
  63         }
  64 };
  65
  66 #ifndef USE_OPENSSL
  67 #define MD5_CTX md_context
  68 #define MD5_Init md5_begin
  69 #define MD5_Update md5_update
  70 #define MD5_Final(digest, cptr) md5_result(cptr, digest)
  71 #endif
  72
  73 int xfersum_type = 0; /* used for the file transfer checksums */
  74 int checksum_type = 0; /* used for the pre-transfer (--checksum) checksums */
  75
  76 static int parse_csum_name(const char *name, int len)
  77 {
  78         struct name_num_item *nni;
  79
  80         if (len < 0 && name)
  81                 len = strlen(name);
  82
  83         if (!name || (len == 4 && strncasecmp(name, "auto", 4) == 0)) {
  84                 if (protocol_version >= 30)
  85                         return CSUM_MD5;
  86                 if (protocol_version >= 27)
  87                         return CSUM_MD4_OLD;
  88                 if (protocol_version >= 21)
  89                         return CSUM_MD4_BUSTED;
  90                 return CSUM_MD4_ARCHAIC;
  91         }
  92
  93         nni = get_nni_by_name(&valid_checksums, name, len);
  94
  95         if (!nni) {
  96                 rprintf(FERROR, "unknown checksum name: %s\n", name);
  97                 exit_cleanup(RERR_UNSUPPORTED);
  98         }
  99
 100         return nni->num;
 101 }
 102
 103 static const char *checksum_name(int num)
 104 {
 105         struct name_num_item *nni = get_nni_by_num(&valid_checksums, num);
 106
 107         return nni ? nni->name : num < CSUM_MD4 ? "MD4" : "UNKNOWN";
 108 }
 109
 110 void parse_checksum_choice(int final_call)
 111 {
 112         if (valid_checksums.negotiated_name)
 113                 xfersum_type = checksum_type = valid_checksums.negotiated_num;
 114         else {
 115                 char *cp = checksum_choice ? strchr(checksum_choice, ',') : NULL;
 116                 if (cp) {
 117                         xfersum_type = parse_csum_name(checksum_choice, cp - checksum_choice);
 118                         checksum_type = parse_csum_name(cp+1, -1);
 119                 } else
 120                         xfersum_type = checksum_type = parse_csum_name(checksum_choice, -1);
 121         }
 122
 123         if (xfersum_type == CSUM_NONE)
 124                 whole_file = 1;
 125
 126         /* Snag the checksum name for both write_batch's option output & the following debug output. */
 127         if (valid_checksums.negotiated_name)
 128                 checksum_choice = valid_checksums.negotiated_name;
 129         else if (checksum_choice == NULL)
 130                 checksum_choice = checksum_name(xfersum_type);
 131
 132         if (final_call && DEBUG_GTE(NSTR, am_server ? 3 : 1)) {
 133                 rprintf(FINFO, "%s%s checksum: %s\n",
 134                         am_server ? "Server" : "Client",
 135                         valid_checksums.negotiated_name ? " negotiated" : "",
 136                         checksum_choice);
 137         }
 138 }
 139
 140 int csum_len_for_type(int cst, BOOL flist_csum)
 141 {
 142         switch (cst) {
 143           case CSUM_NONE:
 144                 return 1;
 145           case CSUM_MD4_ARCHAIC:
 146                 /* The oldest checksum code is rather weird: the file-list code only sent
 147                  * 2-byte checksums, but all other checksums were full MD4 length. */
 148                 return flist_csum ? 2 : MD4_DIGEST_LEN;
 149           case CSUM_MD4:
 150           case CSUM_MD4_OLD:
 151           case CSUM_MD4_BUSTED:
 152                 return MD4_DIGEST_LEN;
 153           case CSUM_MD5:
 154                 return MD5_DIGEST_LEN;
 155 #ifdef SUPPORT_XXHASH
 156           case CSUM_XXH64:
 157                 return 64/8;
 158 #endif
 159           default: /* paranoia to prevent missing case values */
 160                 exit_cleanup(RERR_UNSUPPORTED);
 161         }
 162         return 0;
 163 }
 164
 165 /* Returns 0 if the checksum is not canonical (i.e. it includes a seed value).
 166  * Returns 1 if the public sum order matches our internal sum order.
 167  * Returns -1 if the public sum order is the reverse of our internal sum order.
 168  */
 169 int canonical_checksum(int csum_type)
 170 {
 171         switch (csum_type) {
 172           case CSUM_NONE:
 173           case CSUM_MD4_ARCHAIC:
 174           case CSUM_MD4_OLD:
 175           case CSUM_MD4_BUSTED:
 176                 break;
 177           case CSUM_MD4:
 178           case CSUM_MD5:
 179                 return -1;
 180 #ifdef SUPPORT_XXHASH
 181           case CSUM_XXH64:
 182                 return 1;
 183 #endif
 184           default: /* paranoia to prevent missing case values */
 185                 exit_cleanup(RERR_UNSUPPORTED);
 186         }
 187         return 0;
 188 }
 189
 190 #ifndef HAVE_SIMD /* See simd-checksum-*.cpp. */
 191 /*
 192   a simple 32 bit checksum that can be updated from either end
 193   (inspired by Mark Adler's Adler-32 checksum)
 194   */
 195 uint32 get_checksum1(char *buf1, int32 len)
 196 {
 197         int32 i;
 198         uint32 s1, s2;
 199         schar *buf = (schar *)buf1;
 200
 201         s1 = s2 = 0;
 202         for (i = 0; i < (len-4); i+=4) {
 203                 s2 += 4*(s1 + buf[i]) + 3*buf[i+1] + 2*buf[i+2] + buf[i+3] + 10*CHAR_OFFSET;
 204                 s1 += (buf[i+0] + buf[i+1] + buf[i+2] + buf[i+3] + 4*CHAR_OFFSET);
 205         }
 206         for (; i < len; i++) {
 207                 s1 += (buf[i]+CHAR_OFFSET); s2 += s1;
 208         }
 209         return (s1 & 0xffff) + (s2 << 16);
 210 }
 211 #endif
 212
 213 void get_checksum2(char *buf, int32 len, char *sum)
 214 {
 215         switch (xfersum_type) {
 216           case CSUM_MD5: {
 217                 MD5_CTX m5;
 218                 uchar seedbuf[4];
 219                 MD5_Init(&m5);
 220                 if (proper_seed_order) {
 221                         if (checksum_seed) {
 222                                 SIVALu(seedbuf, 0, checksum_seed);
 223                                 MD5_Update(&m5, seedbuf, 4);
 224                         }
 225                         MD5_Update(&m5, (uchar *)buf, len);
 226                 } else {
 227                         MD5_Update(&m5, (uchar *)buf, len);
 228                         if (checksum_seed) {
 229                                 SIVALu(seedbuf, 0, checksum_seed);
 230                                 MD5_Update(&m5, seedbuf, 4);
 231                         }
 232                 }
 233                 MD5_Final((uchar *)sum, &m5);
 234                 break;
 235           }
 236           case CSUM_MD4:
 237 #ifdef USE_OPENSSL
 238           {
 239                 MD4_CTX m4;
 240                 MD4_Init(&m4);
 241                 MD4_Update(&m4, (uchar *)buf, len);
 242                 if (checksum_seed) {
 243                         uchar seedbuf[4];
 244                         SIVALu(seedbuf, 0, checksum_seed);
 245                         MD4_Update(&m4, seedbuf, 4);
 246                 }
 247                 MD4_Final((uchar *)sum, &m4);
 248                 break;
 249           }
 250 #endif
 251           case CSUM_MD4_OLD:
 252           case CSUM_MD4_BUSTED:
 253           case CSUM_MD4_ARCHAIC: {
 254                 md_context m;
 255                 int32 i;
 256                 static char *buf1;
 257                 static int32 len1;
 258
 259                 mdfour_begin(&m);
 260
 261                 if (len > len1) {
 262                         if (buf1)
 263                                 free(buf1);
 264                         buf1 = new_array(char, len+4);
 265                         len1 = len;
 266                         if (!buf1)
 267                                 out_of_memory("get_checksum2");
 268                 }
 269
 270                 memcpy(buf1, buf, len);
 271                 if (checksum_seed) {
 272                         SIVAL(buf1,len,checksum_seed);
 273                         len += 4;
 274                 }
 275
 276                 for (i = 0; i + CSUM_CHUNK <= len; i += CSUM_CHUNK)
 277                         mdfour_update(&m, (uchar *)(buf1+i), CSUM_CHUNK);
 278
 279                 /*
 280                  * Prior to version 27 an incorrect MD4 checksum was computed
 281                  * by failing to call mdfour_tail() for block sizes that
 282                  * are multiples of 64.  This is fixed by calling mdfour_update()
 283                  * even when there are no more bytes.
 284                  */
 285                 if (len - i > 0 || xfersum_type > CSUM_MD4_BUSTED)
 286                         mdfour_update(&m, (uchar *)(buf1+i), len-i);
 287
 288                 mdfour_result(&m, (uchar *)sum);
 289                 break;
 290           }
 291 #ifdef SUPPORT_XXHASH
 292           case CSUM_XXH64:
 293                 SIVAL64(sum, 0, XXH64(buf, len, checksum_seed));
 294                 break;
 295 #endif
 296           default: /* paranoia to prevent missing case values */
 297                 exit_cleanup(RERR_UNSUPPORTED);
 298         }
 299 }
 300
 301 void file_checksum(const char *fname, const STRUCT_STAT *st_p, char *sum)
 302 {
 303         struct map_struct *buf;
 304         OFF_T i, len = st_p->st_size;
 305         int32 remainder;
 306         int fd;
 307
 308         memset(sum, 0, MAX_DIGEST_LEN);
 309
 310         fd = do_open(fname, O_RDONLY, 0);
 311         if (fd == -1)
 312                 return;
 313
 314         buf = map_file(fd, len, MAX_MAP_SIZE, CHUNK_SIZE);
 315
 316         switch (checksum_type) {
 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 #ifdef SUPPORT_XXHASH
 372           case CSUM_XXH64: {
 373                 XXH64_state_t* state = XXH64_createState();
 374                 if (state == NULL)
 375                         out_of_memory("file_checksum XXH64");
 376
 377                 if (XXH64_reset(state, 0) == XXH_ERROR) {
 378                         rprintf(FERROR, "error resetting XXH64 seed");
 379                         exit_cleanup(RERR_STREAMIO);
 380                 }
 381
 382                 for (i = 0; i + CHUNK_SIZE <= len; i += CHUNK_SIZE) {
 383                         XXH_errorcode const updateResult =
 384                             XXH64_update(state, (uchar *)map_ptr(buf, i, CHUNK_SIZE), CHUNK_SIZE);
 385                         if (updateResult == XXH_ERROR) {
 386                                 rprintf(FERROR, "error computing XXH64 hash");
 387                                 exit_cleanup(RERR_STREAMIO);
 388                         }
 389                 }
 390
 391                 remainder = (int32)(len - i);
 392                 if (remainder > 0)
 393                         XXH64_update(state, (uchar *)map_ptr(buf, i, remainder), remainder);
 394                 SIVAL64(sum, 0, XXH64_digest(state));
 395
 396                 XXH64_freeState(state);
 397                 break;
 398           }
 399 #endif
 400           default:
 401                 rprintf(FERROR, "Invalid checksum-choice for --checksum: %s (%d)\n",
 402                         checksum_name(checksum_type), checksum_type);
 403                 exit_cleanup(RERR_UNSUPPORTED);
 404         }
 405
 406         close(fd);
 407         unmap_file(buf);
 408 }
 409
 410 static int32 sumresidue;
 411 static union {
 412         md_context md;
 413 #ifdef USE_OPENSSL
 414         MD4_CTX m4;
 415 #endif
 416         MD5_CTX m5;
 417 } ctx;
 418 #ifdef SUPPORT_XXHASH
 419 static XXH64_state_t* xxh64_state;
 420 #endif
 421 static int cursum_type;
 422
 423 void sum_init(int csum_type, int seed)
 424 {
 425         char s[4];
 426
 427         if (csum_type < 0)
 428                 csum_type = parse_csum_name(NULL, 0);
 429         cursum_type = csum_type;
 430
 431         switch (csum_type) {
 432           case CSUM_MD5:
 433                 MD5_Init(&ctx.m5);
 434                 break;
 435           case CSUM_MD4:
 436 #ifdef USE_OPENSSL
 437                 MD4_Init(&ctx.m4);
 438 #else
 439                 mdfour_begin(&ctx.md);
 440                 sumresidue = 0;
 441 #endif
 442                 break;
 443           case CSUM_MD4_OLD:
 444           case CSUM_MD4_BUSTED:
 445           case CSUM_MD4_ARCHAIC:
 446                 mdfour_begin(&ctx.md);
 447                 sumresidue = 0;
 448                 SIVAL(s, 0, seed);
 449                 sum_update(s, 4);
 450                 break;
 451 #ifdef SUPPORT_XXHASH
 452           case CSUM_XXH64:
 453                 if (xxh64_state == NULL) {
 454                         xxh64_state = XXH64_createState();
 455                         if (xxh64_state == NULL)
 456                                 out_of_memory("sum_init xxh64");
 457                 }
 458                 if (XXH64_reset(xxh64_state, 0) == XXH_ERROR) {
 459                         rprintf(FERROR, "error resetting XXH64 state");
 460                         exit_cleanup(RERR_STREAMIO);
 461                 }
 462                 break;
 463 #endif
 464           case CSUM_NONE:
 465                 break;
 466           default: /* paranoia to prevent missing case values */
 467                 exit_cleanup(RERR_UNSUPPORTED);
 468         }
 469 }
 470
 471 /**
 472  * Feed data into an MD4 accumulator, md.  The results may be
 473  * retrieved using sum_end().  md is used for different purposes at
 474  * different points during execution.
 475  *
 476  * @todo Perhaps get rid of md and just pass in the address each time.
 477  * Very slightly clearer and slower.
 478  **/
 479 void sum_update(const char *p, int32 len)
 480 {
 481         switch (cursum_type) {
 482           case CSUM_MD5:
 483                 MD5_Update(&ctx.m5, (uchar *)p, len);
 484                 break;
 485           case CSUM_MD4:
 486 #ifdef USE_OPENSSL
 487                 MD4_Update(&ctx.m4, (uchar *)p, len);
 488                 break;
 489 #endif
 490           case CSUM_MD4_OLD:
 491           case CSUM_MD4_BUSTED:
 492           case CSUM_MD4_ARCHAIC:
 493                 if (len + sumresidue < CSUM_CHUNK) {
 494                         memcpy(ctx.md.buffer + sumresidue, p, len);
 495                         sumresidue += len;
 496                         break;
 497                 }
 498
 499                 if (sumresidue) {
 500                         int32 i = CSUM_CHUNK - sumresidue;
 501                         memcpy(ctx.md.buffer + sumresidue, p, i);
 502                         mdfour_update(&ctx.md, (uchar *)ctx.md.buffer, CSUM_CHUNK);
 503                         len -= i;
 504                         p += i;
 505                 }
 506
 507                 while (len >= CSUM_CHUNK) {
 508                         mdfour_update(&ctx.md, (uchar *)p, CSUM_CHUNK);
 509                         len -= CSUM_CHUNK;
 510                         p += CSUM_CHUNK;
 511                 }
 512
 513                 sumresidue = len;
 514                 if (sumresidue)
 515                         memcpy(ctx.md.buffer, p, sumresidue);
 516                 break;
 517 #ifdef SUPPORT_XXHASH
 518           case CSUM_XXH64:
 519                 if (XXH64_update(xxh64_state, p, len) == XXH_ERROR) {
 520                         rprintf(FERROR, "error computing XXH64 hash");
 521                         exit_cleanup(RERR_STREAMIO);
 522                 }
 523                 break;
 524 #endif
 525           case CSUM_NONE:
 526                 break;
 527           default: /* paranoia to prevent missing case values */
 528                 exit_cleanup(RERR_UNSUPPORTED);
 529         }
 530 }
 531
 532 /* NOTE: all the callers of sum_end() pass in a pointer to a buffer that is
 533  * MAX_DIGEST_LEN in size, so even if the csum-len is shorter that that (i.e.
 534  * CSUM_MD4_ARCHAIC), we don't have to worry about limiting the data we write
 535  * into the "sum" buffer. */
 536 int sum_end(char *sum)
 537 {
 538         switch (cursum_type) {
 539           case CSUM_MD5:
 540                 MD5_Final((uchar *)sum, &ctx.m5);
 541                 break;
 542           case CSUM_MD4:
 543 #ifdef USE_OPENSSL
 544                 MD4_Final((uchar *)sum, &ctx.m4);
 545                 break;
 546 #endif
 547           case CSUM_MD4_OLD:
 548                 mdfour_update(&ctx.md, (uchar *)ctx.md.buffer, sumresidue);
 549                 mdfour_result(&ctx.md, (uchar *)sum);
 550                 break;
 551           case CSUM_MD4_BUSTED:
 552           case CSUM_MD4_ARCHAIC:
 553                 if (sumresidue)
 554                         mdfour_update(&ctx.md, (uchar *)ctx.md.buffer, sumresidue);
 555                 mdfour_result(&ctx.md, (uchar *)sum);
 556                 break;
 557 #ifdef SUPPORT_XXHASH
 558           case CSUM_XXH64:
 559                 SIVAL64(sum, 0, XXH64_digest(xxh64_state));
 560                 break;
 561 #endif
 562           case CSUM_NONE:
 563                 *sum = '\0';
 564                 break;
 565           default: /* paranoia to prevent missing case values */
 566                 exit_cleanup(RERR_UNSUPPORTED);
 567         }
 568
 569         return csum_len_for_type(cursum_type, 0);
 570 }