1 /* 2 * Block driver for the QCOW version 2 format 3 * 4 * Copyright (c) 2004-2006 Fabrice Bellard 5 * 6 * Permission is hereby granted, free of charge, to any person obtaining a copy 7 * of this software and associated documentation files (the "Software"), to deal 8 * in the Software without restriction, including without limitation the rights 9 * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell 10 * copies of the Software, and to permit persons to whom the Software is 11 * furnished to do so, subject to the following conditions: 12 * 13 * The above copyright notice and this permission notice shall be included in 14 * all copies or substantial portions of the Software. 15 * 16 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 17 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 18 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL 19 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 20 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, 21 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN 22 * THE SOFTWARE. 23 */ 24 25 #include "qemu/osdep.h" 26 #include "qapi/error.h" 27 #include "qemu-common.h" 28 #include "block/block_int.h" 29 #include "block/qcow2.h" 30 #include "qemu/range.h" 31 #include "qemu/bswap.h" 32 33 static int64_t alloc_clusters_noref(BlockDriverState *bs, uint64_t size); 34 static int QEMU_WARN_UNUSED_RESULT update_refcount(BlockDriverState *bs, 35 int64_t offset, int64_t length, uint64_t addend, 36 bool decrease, enum qcow2_discard_type type); 37 38 static uint64_t get_refcount_ro0(const void *refcount_array, uint64_t index); 39 static uint64_t get_refcount_ro1(const void *refcount_array, uint64_t index); 40 static uint64_t get_refcount_ro2(const void *refcount_array, uint64_t index); 41 static uint64_t get_refcount_ro3(const void *refcount_array, uint64_t index); 42 static uint64_t get_refcount_ro4(const void *refcount_array, uint64_t index); 43 static uint64_t get_refcount_ro5(const void *refcount_array, uint64_t index); 44 static uint64_t get_refcount_ro6(const void *refcount_array, uint64_t index); 45 46 static void set_refcount_ro0(void *refcount_array, uint64_t index, 47 uint64_t value); 48 static void set_refcount_ro1(void *refcount_array, uint64_t index, 49 uint64_t value); 50 static void set_refcount_ro2(void *refcount_array, uint64_t index, 51 uint64_t value); 52 static void set_refcount_ro3(void *refcount_array, uint64_t index, 53 uint64_t value); 54 static void set_refcount_ro4(void *refcount_array, uint64_t index, 55 uint64_t value); 56 static void set_refcount_ro5(void *refcount_array, uint64_t index, 57 uint64_t value); 58 static void set_refcount_ro6(void *refcount_array, uint64_t index, 59 uint64_t value); 60 61 62 static Qcow2GetRefcountFunc *const get_refcount_funcs[] = { 63 &get_refcount_ro0, 64 &get_refcount_ro1, 65 &get_refcount_ro2, 66 &get_refcount_ro3, 67 &get_refcount_ro4, 68 &get_refcount_ro5, 69 &get_refcount_ro6 70 }; 71 72 static Qcow2SetRefcountFunc *const set_refcount_funcs[] = { 73 &set_refcount_ro0, 74 &set_refcount_ro1, 75 &set_refcount_ro2, 76 &set_refcount_ro3, 77 &set_refcount_ro4, 78 &set_refcount_ro5, 79 &set_refcount_ro6 80 }; 81 82 83 /*********************************************************/ 84 /* refcount handling */ 85 86 int qcow2_refcount_init(BlockDriverState *bs) 87 { 88 BDRVQcow2State *s = bs->opaque; 89 unsigned int refcount_table_size2, i; 90 int ret; 91 92 assert(s->refcount_order >= 0 && s->refcount_order <= 6); 93 94 s->get_refcount = get_refcount_funcs[s->refcount_order]; 95 s->set_refcount = set_refcount_funcs[s->refcount_order]; 96 97 assert(s->refcount_table_size <= INT_MAX / sizeof(uint64_t)); 98 refcount_table_size2 = s->refcount_table_size * sizeof(uint64_t); 99 s->refcount_table = g_try_malloc(refcount_table_size2); 100 101 if (s->refcount_table_size > 0) { 102 if (s->refcount_table == NULL) { 103 ret = -ENOMEM; 104 goto fail; 105 } 106 BLKDBG_EVENT(bs->file, BLKDBG_REFTABLE_LOAD); 107 ret = bdrv_pread(bs->file, s->refcount_table_offset, 108 s->refcount_table, refcount_table_size2); 109 if (ret < 0) { 110 goto fail; 111 } 112 for(i = 0; i < s->refcount_table_size; i++) 113 be64_to_cpus(&s->refcount_table[i]); 114 } 115 return 0; 116 fail: 117 return ret; 118 } 119 120 void qcow2_refcount_close(BlockDriverState *bs) 121 { 122 BDRVQcow2State *s = bs->opaque; 123 g_free(s->refcount_table); 124 } 125 126 127 static uint64_t get_refcount_ro0(const void *refcount_array, uint64_t index) 128 { 129 return (((const uint8_t *)refcount_array)[index / 8] >> (index % 8)) & 0x1; 130 } 131 132 static void set_refcount_ro0(void *refcount_array, uint64_t index, 133 uint64_t value) 134 { 135 assert(!(value >> 1)); 136 ((uint8_t *)refcount_array)[index / 8] &= ~(0x1 << (index % 8)); 137 ((uint8_t *)refcount_array)[index / 8] |= value << (index % 8); 138 } 139 140 static uint64_t get_refcount_ro1(const void *refcount_array, uint64_t index) 141 { 142 return (((const uint8_t *)refcount_array)[index / 4] >> (2 * (index % 4))) 143 & 0x3; 144 } 145 146 static void set_refcount_ro1(void *refcount_array, uint64_t index, 147 uint64_t value) 148 { 149 assert(!(value >> 2)); 150 ((uint8_t *)refcount_array)[index / 4] &= ~(0x3 << (2 * (index % 4))); 151 ((uint8_t *)refcount_array)[index / 4] |= value << (2 * (index % 4)); 152 } 153 154 static uint64_t get_refcount_ro2(const void *refcount_array, uint64_t index) 155 { 156 return (((const uint8_t *)refcount_array)[index / 2] >> (4 * (index % 2))) 157 & 0xf; 158 } 159 160 static void set_refcount_ro2(void *refcount_array, uint64_t index, 161 uint64_t value) 162 { 163 assert(!(value >> 4)); 164 ((uint8_t *)refcount_array)[index / 2] &= ~(0xf << (4 * (index % 2))); 165 ((uint8_t *)refcount_array)[index / 2] |= value << (4 * (index % 2)); 166 } 167 168 static uint64_t get_refcount_ro3(const void *refcount_array, uint64_t index) 169 { 170 return ((const uint8_t *)refcount_array)[index]; 171 } 172 173 static void set_refcount_ro3(void *refcount_array, uint64_t index, 174 uint64_t value) 175 { 176 assert(!(value >> 8)); 177 ((uint8_t *)refcount_array)[index] = value; 178 } 179 180 static uint64_t get_refcount_ro4(const void *refcount_array, uint64_t index) 181 { 182 return be16_to_cpu(((const uint16_t *)refcount_array)[index]); 183 } 184 185 static void set_refcount_ro4(void *refcount_array, uint64_t index, 186 uint64_t value) 187 { 188 assert(!(value >> 16)); 189 ((uint16_t *)refcount_array)[index] = cpu_to_be16(value); 190 } 191 192 static uint64_t get_refcount_ro5(const void *refcount_array, uint64_t index) 193 { 194 return be32_to_cpu(((const uint32_t *)refcount_array)[index]); 195 } 196 197 static void set_refcount_ro5(void *refcount_array, uint64_t index, 198 uint64_t value) 199 { 200 assert(!(value >> 32)); 201 ((uint32_t *)refcount_array)[index] = cpu_to_be32(value); 202 } 203 204 static uint64_t get_refcount_ro6(const void *refcount_array, uint64_t index) 205 { 206 return be64_to_cpu(((const uint64_t *)refcount_array)[index]); 207 } 208 209 static void set_refcount_ro6(void *refcount_array, uint64_t index, 210 uint64_t value) 211 { 212 ((uint64_t *)refcount_array)[index] = cpu_to_be64(value); 213 } 214 215 216 static int load_refcount_block(BlockDriverState *bs, 217 int64_t refcount_block_offset, 218 void **refcount_block) 219 { 220 BDRVQcow2State *s = bs->opaque; 221 222 BLKDBG_EVENT(bs->file, BLKDBG_REFBLOCK_LOAD); 223 return qcow2_cache_get(bs, s->refcount_block_cache, refcount_block_offset, 224 refcount_block); 225 } 226 227 /* 228 * Retrieves the refcount of the cluster given by its index and stores it in 229 * *refcount. Returns 0 on success and -errno on failure. 230 */ 231 int qcow2_get_refcount(BlockDriverState *bs, int64_t cluster_index, 232 uint64_t *refcount) 233 { 234 BDRVQcow2State *s = bs->opaque; 235 uint64_t refcount_table_index, block_index; 236 int64_t refcount_block_offset; 237 int ret; 238 void *refcount_block; 239 240 refcount_table_index = cluster_index >> s->refcount_block_bits; 241 if (refcount_table_index >= s->refcount_table_size) { 242 *refcount = 0; 243 return 0; 244 } 245 refcount_block_offset = 246 s->refcount_table[refcount_table_index] & REFT_OFFSET_MASK; 247 if (!refcount_block_offset) { 248 *refcount = 0; 249 return 0; 250 } 251 252 if (offset_into_cluster(s, refcount_block_offset)) { 253 qcow2_signal_corruption(bs, true, -1, -1, "Refblock offset %#" PRIx64 254 " unaligned (reftable index: %#" PRIx64 ")", 255 refcount_block_offset, refcount_table_index); 256 return -EIO; 257 } 258 259 ret = qcow2_cache_get(bs, s->refcount_block_cache, refcount_block_offset, 260 &refcount_block); 261 if (ret < 0) { 262 return ret; 263 } 264 265 block_index = cluster_index & (s->refcount_block_size - 1); 266 *refcount = s->get_refcount(refcount_block, block_index); 267 268 qcow2_cache_put(bs, s->refcount_block_cache, &refcount_block); 269 270 return 0; 271 } 272 273 /* 274 * Rounds the refcount table size up to avoid growing the table for each single 275 * refcount block that is allocated. 276 */ 277 static unsigned int next_refcount_table_size(BDRVQcow2State *s, 278 unsigned int min_size) 279 { 280 unsigned int min_clusters = (min_size >> (s->cluster_bits - 3)) + 1; 281 unsigned int refcount_table_clusters = 282 MAX(1, s->refcount_table_size >> (s->cluster_bits - 3)); 283 284 while (min_clusters > refcount_table_clusters) { 285 refcount_table_clusters = (refcount_table_clusters * 3 + 1) / 2; 286 } 287 288 return refcount_table_clusters << (s->cluster_bits - 3); 289 } 290 291 292 /* Checks if two offsets are described by the same refcount block */ 293 static int in_same_refcount_block(BDRVQcow2State *s, uint64_t offset_a, 294 uint64_t offset_b) 295 { 296 uint64_t block_a = offset_a >> (s->cluster_bits + s->refcount_block_bits); 297 uint64_t block_b = offset_b >> (s->cluster_bits + s->refcount_block_bits); 298 299 return (block_a == block_b); 300 } 301 302 /* 303 * Loads a refcount block. If it doesn't exist yet, it is allocated first 304 * (including growing the refcount table if needed). 305 * 306 * Returns 0 on success or -errno in error case 307 */ 308 static int alloc_refcount_block(BlockDriverState *bs, 309 int64_t cluster_index, void **refcount_block) 310 { 311 BDRVQcow2State *s = bs->opaque; 312 unsigned int refcount_table_index; 313 int ret; 314 315 BLKDBG_EVENT(bs->file, BLKDBG_REFBLOCK_ALLOC); 316 317 /* Find the refcount block for the given cluster */ 318 refcount_table_index = cluster_index >> s->refcount_block_bits; 319 320 if (refcount_table_index < s->refcount_table_size) { 321 322 uint64_t refcount_block_offset = 323 s->refcount_table[refcount_table_index] & REFT_OFFSET_MASK; 324 325 /* If it's already there, we're done */ 326 if (refcount_block_offset) { 327 if (offset_into_cluster(s, refcount_block_offset)) { 328 qcow2_signal_corruption(bs, true, -1, -1, "Refblock offset %#" 329 PRIx64 " unaligned (reftable index: " 330 "%#x)", refcount_block_offset, 331 refcount_table_index); 332 return -EIO; 333 } 334 335 return load_refcount_block(bs, refcount_block_offset, 336 refcount_block); 337 } 338 } 339 340 /* 341 * If we came here, we need to allocate something. Something is at least 342 * a cluster for the new refcount block. It may also include a new refcount 343 * table if the old refcount table is too small. 344 * 345 * Note that allocating clusters here needs some special care: 346 * 347 * - We can't use the normal qcow2_alloc_clusters(), it would try to 348 * increase the refcount and very likely we would end up with an endless 349 * recursion. Instead we must place the refcount blocks in a way that 350 * they can describe them themselves. 351 * 352 * - We need to consider that at this point we are inside update_refcounts 353 * and potentially doing an initial refcount increase. This means that 354 * some clusters have already been allocated by the caller, but their 355 * refcount isn't accurate yet. If we allocate clusters for metadata, we 356 * need to return -EAGAIN to signal the caller that it needs to restart 357 * the search for free clusters. 358 * 359 * - alloc_clusters_noref and qcow2_free_clusters may load a different 360 * refcount block into the cache 361 */ 362 363 *refcount_block = NULL; 364 365 /* We write to the refcount table, so we might depend on L2 tables */ 366 ret = qcow2_cache_flush(bs, s->l2_table_cache); 367 if (ret < 0) { 368 return ret; 369 } 370 371 /* Allocate the refcount block itself and mark it as used */ 372 int64_t new_block = alloc_clusters_noref(bs, s->cluster_size); 373 if (new_block < 0) { 374 return new_block; 375 } 376 377 #ifdef DEBUG_ALLOC2 378 fprintf(stderr, "qcow2: Allocate refcount block %d for %" PRIx64 379 " at %" PRIx64 "\n", 380 refcount_table_index, cluster_index << s->cluster_bits, new_block); 381 #endif 382 383 if (in_same_refcount_block(s, new_block, cluster_index << s->cluster_bits)) { 384 /* Zero the new refcount block before updating it */ 385 ret = qcow2_cache_get_empty(bs, s->refcount_block_cache, new_block, 386 refcount_block); 387 if (ret < 0) { 388 goto fail_block; 389 } 390 391 memset(*refcount_block, 0, s->cluster_size); 392 393 /* The block describes itself, need to update the cache */ 394 int block_index = (new_block >> s->cluster_bits) & 395 (s->refcount_block_size - 1); 396 s->set_refcount(*refcount_block, block_index, 1); 397 } else { 398 /* Described somewhere else. This can recurse at most twice before we 399 * arrive at a block that describes itself. */ 400 ret = update_refcount(bs, new_block, s->cluster_size, 1, false, 401 QCOW2_DISCARD_NEVER); 402 if (ret < 0) { 403 goto fail_block; 404 } 405 406 ret = qcow2_cache_flush(bs, s->refcount_block_cache); 407 if (ret < 0) { 408 goto fail_block; 409 } 410 411 /* Initialize the new refcount block only after updating its refcount, 412 * update_refcount uses the refcount cache itself */ 413 ret = qcow2_cache_get_empty(bs, s->refcount_block_cache, new_block, 414 refcount_block); 415 if (ret < 0) { 416 goto fail_block; 417 } 418 419 memset(*refcount_block, 0, s->cluster_size); 420 } 421 422 /* Now the new refcount block needs to be written to disk */ 423 BLKDBG_EVENT(bs->file, BLKDBG_REFBLOCK_ALLOC_WRITE); 424 qcow2_cache_entry_mark_dirty(bs, s->refcount_block_cache, *refcount_block); 425 ret = qcow2_cache_flush(bs, s->refcount_block_cache); 426 if (ret < 0) { 427 goto fail_block; 428 } 429 430 /* If the refcount table is big enough, just hook the block up there */ 431 if (refcount_table_index < s->refcount_table_size) { 432 uint64_t data64 = cpu_to_be64(new_block); 433 BLKDBG_EVENT(bs->file, BLKDBG_REFBLOCK_ALLOC_HOOKUP); 434 ret = bdrv_pwrite_sync(bs->file, 435 s->refcount_table_offset + refcount_table_index * sizeof(uint64_t), 436 &data64, sizeof(data64)); 437 if (ret < 0) { 438 goto fail_block; 439 } 440 441 s->refcount_table[refcount_table_index] = new_block; 442 443 /* The new refcount block may be where the caller intended to put its 444 * data, so let it restart the search. */ 445 return -EAGAIN; 446 } 447 448 qcow2_cache_put(bs, s->refcount_block_cache, refcount_block); 449 450 /* 451 * If we come here, we need to grow the refcount table. Again, a new 452 * refcount table needs some space and we can't simply allocate to avoid 453 * endless recursion. 454 * 455 * Therefore let's grab new refcount blocks at the end of the image, which 456 * will describe themselves and the new refcount table. This way we can 457 * reference them only in the new table and do the switch to the new 458 * refcount table at once without producing an inconsistent state in 459 * between. 460 */ 461 BLKDBG_EVENT(bs->file, BLKDBG_REFTABLE_GROW); 462 463 /* Calculate the number of refcount blocks needed so far; this will be the 464 * basis for calculating the index of the first cluster used for the 465 * self-describing refcount structures which we are about to create. 466 * 467 * Because we reached this point, there cannot be any refcount entries for 468 * cluster_index or higher indices yet. However, because new_block has been 469 * allocated to describe that cluster (and it will assume this role later 470 * on), we cannot use that index; also, new_block may actually have a higher 471 * cluster index than cluster_index, so it needs to be taken into account 472 * here (and 1 needs to be added to its value because that cluster is used). 473 */ 474 uint64_t blocks_used = DIV_ROUND_UP(MAX(cluster_index + 1, 475 (new_block >> s->cluster_bits) + 1), 476 s->refcount_block_size); 477 478 if (blocks_used > QCOW_MAX_REFTABLE_SIZE / sizeof(uint64_t)) { 479 return -EFBIG; 480 } 481 482 /* And now we need at least one block more for the new metadata */ 483 uint64_t table_size = next_refcount_table_size(s, blocks_used + 1); 484 uint64_t last_table_size; 485 uint64_t blocks_clusters; 486 do { 487 uint64_t table_clusters = 488 size_to_clusters(s, table_size * sizeof(uint64_t)); 489 blocks_clusters = 1 + 490 DIV_ROUND_UP(table_clusters, s->refcount_block_size); 491 uint64_t meta_clusters = table_clusters + blocks_clusters; 492 493 last_table_size = table_size; 494 table_size = next_refcount_table_size(s, blocks_used + 495 DIV_ROUND_UP(meta_clusters, s->refcount_block_size)); 496 497 } while (last_table_size != table_size); 498 499 #ifdef DEBUG_ALLOC2 500 fprintf(stderr, "qcow2: Grow refcount table %" PRId32 " => %" PRId64 "\n", 501 s->refcount_table_size, table_size); 502 #endif 503 504 /* Create the new refcount table and blocks */ 505 uint64_t meta_offset = (blocks_used * s->refcount_block_size) * 506 s->cluster_size; 507 uint64_t table_offset = meta_offset + blocks_clusters * s->cluster_size; 508 uint64_t *new_table = g_try_new0(uint64_t, table_size); 509 void *new_blocks = g_try_malloc0(blocks_clusters * s->cluster_size); 510 511 assert(table_size > 0 && blocks_clusters > 0); 512 if (new_table == NULL || new_blocks == NULL) { 513 ret = -ENOMEM; 514 goto fail_table; 515 } 516 517 /* Fill the new refcount table */ 518 memcpy(new_table, s->refcount_table, 519 s->refcount_table_size * sizeof(uint64_t)); 520 new_table[refcount_table_index] = new_block; 521 522 int i; 523 for (i = 0; i < blocks_clusters; i++) { 524 new_table[blocks_used + i] = meta_offset + (i * s->cluster_size); 525 } 526 527 /* Fill the refcount blocks */ 528 uint64_t table_clusters = size_to_clusters(s, table_size * sizeof(uint64_t)); 529 int block = 0; 530 for (i = 0; i < table_clusters + blocks_clusters; i++) { 531 s->set_refcount(new_blocks, block++, 1); 532 } 533 534 /* Write refcount blocks to disk */ 535 BLKDBG_EVENT(bs->file, BLKDBG_REFBLOCK_ALLOC_WRITE_BLOCKS); 536 ret = bdrv_pwrite_sync(bs->file, meta_offset, new_blocks, 537 blocks_clusters * s->cluster_size); 538 g_free(new_blocks); 539 new_blocks = NULL; 540 if (ret < 0) { 541 goto fail_table; 542 } 543 544 /* Write refcount table to disk */ 545 for(i = 0; i < table_size; i++) { 546 cpu_to_be64s(&new_table[i]); 547 } 548 549 BLKDBG_EVENT(bs->file, BLKDBG_REFBLOCK_ALLOC_WRITE_TABLE); 550 ret = bdrv_pwrite_sync(bs->file, table_offset, new_table, 551 table_size * sizeof(uint64_t)); 552 if (ret < 0) { 553 goto fail_table; 554 } 555 556 for(i = 0; i < table_size; i++) { 557 be64_to_cpus(&new_table[i]); 558 } 559 560 /* Hook up the new refcount table in the qcow2 header */ 561 struct QEMU_PACKED { 562 uint64_t d64; 563 uint32_t d32; 564 } data; 565 data.d64 = cpu_to_be64(table_offset); 566 data.d32 = cpu_to_be32(table_clusters); 567 BLKDBG_EVENT(bs->file, BLKDBG_REFBLOCK_ALLOC_SWITCH_TABLE); 568 ret = bdrv_pwrite_sync(bs->file, 569 offsetof(QCowHeader, refcount_table_offset), 570 &data, sizeof(data)); 571 if (ret < 0) { 572 goto fail_table; 573 } 574 575 /* And switch it in memory */ 576 uint64_t old_table_offset = s->refcount_table_offset; 577 uint64_t old_table_size = s->refcount_table_size; 578 579 g_free(s->refcount_table); 580 s->refcount_table = new_table; 581 s->refcount_table_size = table_size; 582 s->refcount_table_offset = table_offset; 583 584 /* Free old table. */ 585 qcow2_free_clusters(bs, old_table_offset, old_table_size * sizeof(uint64_t), 586 QCOW2_DISCARD_OTHER); 587 588 ret = load_refcount_block(bs, new_block, refcount_block); 589 if (ret < 0) { 590 return ret; 591 } 592 593 /* If we were trying to do the initial refcount update for some cluster 594 * allocation, we might have used the same clusters to store newly 595 * allocated metadata. Make the caller search some new space. */ 596 return -EAGAIN; 597 598 fail_table: 599 g_free(new_blocks); 600 g_free(new_table); 601 fail_block: 602 if (*refcount_block != NULL) { 603 qcow2_cache_put(bs, s->refcount_block_cache, refcount_block); 604 } 605 return ret; 606 } 607 608 void qcow2_process_discards(BlockDriverState *bs, int ret) 609 { 610 BDRVQcow2State *s = bs->opaque; 611 Qcow2DiscardRegion *d, *next; 612 613 QTAILQ_FOREACH_SAFE(d, &s->discards, next, next) { 614 QTAILQ_REMOVE(&s->discards, d, next); 615 616 /* Discard is optional, ignore the return value */ 617 if (ret >= 0) { 618 bdrv_pdiscard(bs->file->bs, d->offset, d->bytes); 619 } 620 621 g_free(d); 622 } 623 } 624 625 static void update_refcount_discard(BlockDriverState *bs, 626 uint64_t offset, uint64_t length) 627 { 628 BDRVQcow2State *s = bs->opaque; 629 Qcow2DiscardRegion *d, *p, *next; 630 631 QTAILQ_FOREACH(d, &s->discards, next) { 632 uint64_t new_start = MIN(offset, d->offset); 633 uint64_t new_end = MAX(offset + length, d->offset + d->bytes); 634 635 if (new_end - new_start <= length + d->bytes) { 636 /* There can't be any overlap, areas ending up here have no 637 * references any more and therefore shouldn't get freed another 638 * time. */ 639 assert(d->bytes + length == new_end - new_start); 640 d->offset = new_start; 641 d->bytes = new_end - new_start; 642 goto found; 643 } 644 } 645 646 d = g_malloc(sizeof(*d)); 647 *d = (Qcow2DiscardRegion) { 648 .bs = bs, 649 .offset = offset, 650 .bytes = length, 651 }; 652 QTAILQ_INSERT_TAIL(&s->discards, d, next); 653 654 found: 655 /* Merge discard requests if they are adjacent now */ 656 QTAILQ_FOREACH_SAFE(p, &s->discards, next, next) { 657 if (p == d 658 || p->offset > d->offset + d->bytes 659 || d->offset > p->offset + p->bytes) 660 { 661 continue; 662 } 663 664 /* Still no overlap possible */ 665 assert(p->offset == d->offset + d->bytes 666 || d->offset == p->offset + p->bytes); 667 668 QTAILQ_REMOVE(&s->discards, p, next); 669 d->offset = MIN(d->offset, p->offset); 670 d->bytes += p->bytes; 671 g_free(p); 672 } 673 } 674 675 /* XXX: cache several refcount block clusters ? */ 676 /* @addend is the absolute value of the addend; if @decrease is set, @addend 677 * will be subtracted from the current refcount, otherwise it will be added */ 678 static int QEMU_WARN_UNUSED_RESULT update_refcount(BlockDriverState *bs, 679 int64_t offset, 680 int64_t length, 681 uint64_t addend, 682 bool decrease, 683 enum qcow2_discard_type type) 684 { 685 BDRVQcow2State *s = bs->opaque; 686 int64_t start, last, cluster_offset; 687 void *refcount_block = NULL; 688 int64_t old_table_index = -1; 689 int ret; 690 691 #ifdef DEBUG_ALLOC2 692 fprintf(stderr, "update_refcount: offset=%" PRId64 " size=%" PRId64 693 " addend=%s%" PRIu64 "\n", offset, length, decrease ? "-" : "", 694 addend); 695 #endif 696 if (length < 0) { 697 return -EINVAL; 698 } else if (length == 0) { 699 return 0; 700 } 701 702 if (decrease) { 703 qcow2_cache_set_dependency(bs, s->refcount_block_cache, 704 s->l2_table_cache); 705 } 706 707 start = start_of_cluster(s, offset); 708 last = start_of_cluster(s, offset + length - 1); 709 for(cluster_offset = start; cluster_offset <= last; 710 cluster_offset += s->cluster_size) 711 { 712 int block_index; 713 uint64_t refcount; 714 int64_t cluster_index = cluster_offset >> s->cluster_bits; 715 int64_t table_index = cluster_index >> s->refcount_block_bits; 716 717 /* Load the refcount block and allocate it if needed */ 718 if (table_index != old_table_index) { 719 if (refcount_block) { 720 qcow2_cache_put(bs, s->refcount_block_cache, &refcount_block); 721 } 722 ret = alloc_refcount_block(bs, cluster_index, &refcount_block); 723 if (ret < 0) { 724 goto fail; 725 } 726 } 727 old_table_index = table_index; 728 729 qcow2_cache_entry_mark_dirty(bs, s->refcount_block_cache, 730 refcount_block); 731 732 /* we can update the count and save it */ 733 block_index = cluster_index & (s->refcount_block_size - 1); 734 735 refcount = s->get_refcount(refcount_block, block_index); 736 if (decrease ? (refcount - addend > refcount) 737 : (refcount + addend < refcount || 738 refcount + addend > s->refcount_max)) 739 { 740 ret = -EINVAL; 741 goto fail; 742 } 743 if (decrease) { 744 refcount -= addend; 745 } else { 746 refcount += addend; 747 } 748 if (refcount == 0 && cluster_index < s->free_cluster_index) { 749 s->free_cluster_index = cluster_index; 750 } 751 s->set_refcount(refcount_block, block_index, refcount); 752 753 if (refcount == 0 && s->discard_passthrough[type]) { 754 update_refcount_discard(bs, cluster_offset, s->cluster_size); 755 } 756 } 757 758 ret = 0; 759 fail: 760 if (!s->cache_discards) { 761 qcow2_process_discards(bs, ret); 762 } 763 764 /* Write last changed block to disk */ 765 if (refcount_block) { 766 qcow2_cache_put(bs, s->refcount_block_cache, &refcount_block); 767 } 768 769 /* 770 * Try do undo any updates if an error is returned (This may succeed in 771 * some cases like ENOSPC for allocating a new refcount block) 772 */ 773 if (ret < 0) { 774 int dummy; 775 dummy = update_refcount(bs, offset, cluster_offset - offset, addend, 776 !decrease, QCOW2_DISCARD_NEVER); 777 (void)dummy; 778 } 779 780 return ret; 781 } 782 783 /* 784 * Increases or decreases the refcount of a given cluster. 785 * 786 * @addend is the absolute value of the addend; if @decrease is set, @addend 787 * will be subtracted from the current refcount, otherwise it will be added. 788 * 789 * On success 0 is returned; on failure -errno is returned. 790 */ 791 int qcow2_update_cluster_refcount(BlockDriverState *bs, 792 int64_t cluster_index, 793 uint64_t addend, bool decrease, 794 enum qcow2_discard_type type) 795 { 796 BDRVQcow2State *s = bs->opaque; 797 int ret; 798 799 ret = update_refcount(bs, cluster_index << s->cluster_bits, 1, addend, 800 decrease, type); 801 if (ret < 0) { 802 return ret; 803 } 804 805 return 0; 806 } 807 808 809 810 /*********************************************************/ 811 /* cluster allocation functions */ 812 813 814 815 /* return < 0 if error */ 816 static int64_t alloc_clusters_noref(BlockDriverState *bs, uint64_t size) 817 { 818 BDRVQcow2State *s = bs->opaque; 819 uint64_t i, nb_clusters, refcount; 820 int ret; 821 822 /* We can't allocate clusters if they may still be queued for discard. */ 823 if (s->cache_discards) { 824 qcow2_process_discards(bs, 0); 825 } 826 827 nb_clusters = size_to_clusters(s, size); 828 retry: 829 for(i = 0; i < nb_clusters; i++) { 830 uint64_t next_cluster_index = s->free_cluster_index++; 831 ret = qcow2_get_refcount(bs, next_cluster_index, &refcount); 832 833 if (ret < 0) { 834 return ret; 835 } else if (refcount != 0) { 836 goto retry; 837 } 838 } 839 840 /* Make sure that all offsets in the "allocated" range are representable 841 * in an int64_t */ 842 if (s->free_cluster_index > 0 && 843 s->free_cluster_index - 1 > (INT64_MAX >> s->cluster_bits)) 844 { 845 return -EFBIG; 846 } 847 848 #ifdef DEBUG_ALLOC2 849 fprintf(stderr, "alloc_clusters: size=%" PRId64 " -> %" PRId64 "\n", 850 size, 851 (s->free_cluster_index - nb_clusters) << s->cluster_bits); 852 #endif 853 return (s->free_cluster_index - nb_clusters) << s->cluster_bits; 854 } 855 856 int64_t qcow2_alloc_clusters(BlockDriverState *bs, uint64_t size) 857 { 858 int64_t offset; 859 int ret; 860 861 BLKDBG_EVENT(bs->file, BLKDBG_CLUSTER_ALLOC); 862 do { 863 offset = alloc_clusters_noref(bs, size); 864 if (offset < 0) { 865 return offset; 866 } 867 868 ret = update_refcount(bs, offset, size, 1, false, QCOW2_DISCARD_NEVER); 869 } while (ret == -EAGAIN); 870 871 if (ret < 0) { 872 return ret; 873 } 874 875 return offset; 876 } 877 878 int64_t qcow2_alloc_clusters_at(BlockDriverState *bs, uint64_t offset, 879 int64_t nb_clusters) 880 { 881 BDRVQcow2State *s = bs->opaque; 882 uint64_t cluster_index, refcount; 883 uint64_t i; 884 int ret; 885 886 assert(nb_clusters >= 0); 887 if (nb_clusters == 0) { 888 return 0; 889 } 890 891 do { 892 /* Check how many clusters there are free */ 893 cluster_index = offset >> s->cluster_bits; 894 for(i = 0; i < nb_clusters; i++) { 895 ret = qcow2_get_refcount(bs, cluster_index++, &refcount); 896 if (ret < 0) { 897 return ret; 898 } else if (refcount != 0) { 899 break; 900 } 901 } 902 903 /* And then allocate them */ 904 ret = update_refcount(bs, offset, i << s->cluster_bits, 1, false, 905 QCOW2_DISCARD_NEVER); 906 } while (ret == -EAGAIN); 907 908 if (ret < 0) { 909 return ret; 910 } 911 912 return i; 913 } 914 915 /* only used to allocate compressed sectors. We try to allocate 916 contiguous sectors. size must be <= cluster_size */ 917 int64_t qcow2_alloc_bytes(BlockDriverState *bs, int size) 918 { 919 BDRVQcow2State *s = bs->opaque; 920 int64_t offset; 921 size_t free_in_cluster; 922 int ret; 923 924 BLKDBG_EVENT(bs->file, BLKDBG_CLUSTER_ALLOC_BYTES); 925 assert(size > 0 && size <= s->cluster_size); 926 assert(!s->free_byte_offset || offset_into_cluster(s, s->free_byte_offset)); 927 928 offset = s->free_byte_offset; 929 930 if (offset) { 931 uint64_t refcount; 932 ret = qcow2_get_refcount(bs, offset >> s->cluster_bits, &refcount); 933 if (ret < 0) { 934 return ret; 935 } 936 937 if (refcount == s->refcount_max) { 938 offset = 0; 939 } 940 } 941 942 free_in_cluster = s->cluster_size - offset_into_cluster(s, offset); 943 do { 944 if (!offset || free_in_cluster < size) { 945 int64_t new_cluster = alloc_clusters_noref(bs, s->cluster_size); 946 if (new_cluster < 0) { 947 return new_cluster; 948 } 949 950 if (!offset || ROUND_UP(offset, s->cluster_size) != new_cluster) { 951 offset = new_cluster; 952 free_in_cluster = s->cluster_size; 953 } else { 954 free_in_cluster += s->cluster_size; 955 } 956 } 957 958 assert(offset); 959 ret = update_refcount(bs, offset, size, 1, false, QCOW2_DISCARD_NEVER); 960 if (ret < 0) { 961 offset = 0; 962 } 963 } while (ret == -EAGAIN); 964 if (ret < 0) { 965 return ret; 966 } 967 968 /* The cluster refcount was incremented; refcount blocks must be flushed 969 * before the caller's L2 table updates. */ 970 qcow2_cache_set_dependency(bs, s->l2_table_cache, s->refcount_block_cache); 971 972 s->free_byte_offset = offset + size; 973 if (!offset_into_cluster(s, s->free_byte_offset)) { 974 s->free_byte_offset = 0; 975 } 976 977 return offset; 978 } 979 980 void qcow2_free_clusters(BlockDriverState *bs, 981 int64_t offset, int64_t size, 982 enum qcow2_discard_type type) 983 { 984 int ret; 985 986 BLKDBG_EVENT(bs->file, BLKDBG_CLUSTER_FREE); 987 ret = update_refcount(bs, offset, size, 1, true, type); 988 if (ret < 0) { 989 fprintf(stderr, "qcow2_free_clusters failed: %s\n", strerror(-ret)); 990 /* TODO Remember the clusters to free them later and avoid leaking */ 991 } 992 } 993 994 /* 995 * Free a cluster using its L2 entry (handles clusters of all types, e.g. 996 * normal cluster, compressed cluster, etc.) 997 */ 998 void qcow2_free_any_clusters(BlockDriverState *bs, uint64_t l2_entry, 999 int nb_clusters, enum qcow2_discard_type type) 1000 { 1001 BDRVQcow2State *s = bs->opaque; 1002 1003 switch (qcow2_get_cluster_type(l2_entry)) { 1004 case QCOW2_CLUSTER_COMPRESSED: 1005 { 1006 int nb_csectors; 1007 nb_csectors = ((l2_entry >> s->csize_shift) & 1008 s->csize_mask) + 1; 1009 qcow2_free_clusters(bs, 1010 (l2_entry & s->cluster_offset_mask) & ~511, 1011 nb_csectors * 512, type); 1012 } 1013 break; 1014 case QCOW2_CLUSTER_NORMAL: 1015 case QCOW2_CLUSTER_ZERO: 1016 if (l2_entry & L2E_OFFSET_MASK) { 1017 if (offset_into_cluster(s, l2_entry & L2E_OFFSET_MASK)) { 1018 qcow2_signal_corruption(bs, false, -1, -1, 1019 "Cannot free unaligned cluster %#llx", 1020 l2_entry & L2E_OFFSET_MASK); 1021 } else { 1022 qcow2_free_clusters(bs, l2_entry & L2E_OFFSET_MASK, 1023 nb_clusters << s->cluster_bits, type); 1024 } 1025 } 1026 break; 1027 case QCOW2_CLUSTER_UNALLOCATED: 1028 break; 1029 default: 1030 abort(); 1031 } 1032 } 1033 1034 1035 1036 /*********************************************************/ 1037 /* snapshots and image creation */ 1038 1039 1040 1041 /* update the refcounts of snapshots and the copied flag */ 1042 int qcow2_update_snapshot_refcount(BlockDriverState *bs, 1043 int64_t l1_table_offset, int l1_size, int addend) 1044 { 1045 BDRVQcow2State *s = bs->opaque; 1046 uint64_t *l1_table, *l2_table, l2_offset, offset, l1_size2, refcount; 1047 bool l1_allocated = false; 1048 int64_t old_offset, old_l2_offset; 1049 int i, j, l1_modified = 0, nb_csectors; 1050 int ret; 1051 1052 assert(addend >= -1 && addend <= 1); 1053 1054 l2_table = NULL; 1055 l1_table = NULL; 1056 l1_size2 = l1_size * sizeof(uint64_t); 1057 1058 s->cache_discards = true; 1059 1060 /* WARNING: qcow2_snapshot_goto relies on this function not using the 1061 * l1_table_offset when it is the current s->l1_table_offset! Be careful 1062 * when changing this! */ 1063 if (l1_table_offset != s->l1_table_offset) { 1064 l1_table = g_try_malloc0(align_offset(l1_size2, 512)); 1065 if (l1_size2 && l1_table == NULL) { 1066 ret = -ENOMEM; 1067 goto fail; 1068 } 1069 l1_allocated = true; 1070 1071 ret = bdrv_pread(bs->file, l1_table_offset, l1_table, l1_size2); 1072 if (ret < 0) { 1073 goto fail; 1074 } 1075 1076 for(i = 0;i < l1_size; i++) 1077 be64_to_cpus(&l1_table[i]); 1078 } else { 1079 assert(l1_size == s->l1_size); 1080 l1_table = s->l1_table; 1081 l1_allocated = false; 1082 } 1083 1084 for(i = 0; i < l1_size; i++) { 1085 l2_offset = l1_table[i]; 1086 if (l2_offset) { 1087 old_l2_offset = l2_offset; 1088 l2_offset &= L1E_OFFSET_MASK; 1089 1090 if (offset_into_cluster(s, l2_offset)) { 1091 qcow2_signal_corruption(bs, true, -1, -1, "L2 table offset %#" 1092 PRIx64 " unaligned (L1 index: %#x)", 1093 l2_offset, i); 1094 ret = -EIO; 1095 goto fail; 1096 } 1097 1098 ret = qcow2_cache_get(bs, s->l2_table_cache, l2_offset, 1099 (void**) &l2_table); 1100 if (ret < 0) { 1101 goto fail; 1102 } 1103 1104 for(j = 0; j < s->l2_size; j++) { 1105 uint64_t cluster_index; 1106 1107 offset = be64_to_cpu(l2_table[j]); 1108 old_offset = offset; 1109 offset &= ~QCOW_OFLAG_COPIED; 1110 1111 switch (qcow2_get_cluster_type(offset)) { 1112 case QCOW2_CLUSTER_COMPRESSED: 1113 nb_csectors = ((offset >> s->csize_shift) & 1114 s->csize_mask) + 1; 1115 if (addend != 0) { 1116 ret = update_refcount(bs, 1117 (offset & s->cluster_offset_mask) & ~511, 1118 nb_csectors * 512, abs(addend), addend < 0, 1119 QCOW2_DISCARD_SNAPSHOT); 1120 if (ret < 0) { 1121 goto fail; 1122 } 1123 } 1124 /* compressed clusters are never modified */ 1125 refcount = 2; 1126 break; 1127 1128 case QCOW2_CLUSTER_NORMAL: 1129 case QCOW2_CLUSTER_ZERO: 1130 if (offset_into_cluster(s, offset & L2E_OFFSET_MASK)) { 1131 qcow2_signal_corruption(bs, true, -1, -1, "Data " 1132 "cluster offset %#llx " 1133 "unaligned (L2 offset: %#" 1134 PRIx64 ", L2 index: %#x)", 1135 offset & L2E_OFFSET_MASK, 1136 l2_offset, j); 1137 ret = -EIO; 1138 goto fail; 1139 } 1140 1141 cluster_index = (offset & L2E_OFFSET_MASK) >> s->cluster_bits; 1142 if (!cluster_index) { 1143 /* unallocated */ 1144 refcount = 0; 1145 break; 1146 } 1147 if (addend != 0) { 1148 ret = qcow2_update_cluster_refcount(bs, 1149 cluster_index, abs(addend), addend < 0, 1150 QCOW2_DISCARD_SNAPSHOT); 1151 if (ret < 0) { 1152 goto fail; 1153 } 1154 } 1155 1156 ret = qcow2_get_refcount(bs, cluster_index, &refcount); 1157 if (ret < 0) { 1158 goto fail; 1159 } 1160 break; 1161 1162 case QCOW2_CLUSTER_UNALLOCATED: 1163 refcount = 0; 1164 break; 1165 1166 default: 1167 abort(); 1168 } 1169 1170 if (refcount == 1) { 1171 offset |= QCOW_OFLAG_COPIED; 1172 } 1173 if (offset != old_offset) { 1174 if (addend > 0) { 1175 qcow2_cache_set_dependency(bs, s->l2_table_cache, 1176 s->refcount_block_cache); 1177 } 1178 l2_table[j] = cpu_to_be64(offset); 1179 qcow2_cache_entry_mark_dirty(bs, s->l2_table_cache, 1180 l2_table); 1181 } 1182 } 1183 1184 qcow2_cache_put(bs, s->l2_table_cache, (void **) &l2_table); 1185 1186 if (addend != 0) { 1187 ret = qcow2_update_cluster_refcount(bs, l2_offset >> 1188 s->cluster_bits, 1189 abs(addend), addend < 0, 1190 QCOW2_DISCARD_SNAPSHOT); 1191 if (ret < 0) { 1192 goto fail; 1193 } 1194 } 1195 ret = qcow2_get_refcount(bs, l2_offset >> s->cluster_bits, 1196 &refcount); 1197 if (ret < 0) { 1198 goto fail; 1199 } else if (refcount == 1) { 1200 l2_offset |= QCOW_OFLAG_COPIED; 1201 } 1202 if (l2_offset != old_l2_offset) { 1203 l1_table[i] = l2_offset; 1204 l1_modified = 1; 1205 } 1206 } 1207 } 1208 1209 ret = bdrv_flush(bs); 1210 fail: 1211 if (l2_table) { 1212 qcow2_cache_put(bs, s->l2_table_cache, (void**) &l2_table); 1213 } 1214 1215 s->cache_discards = false; 1216 qcow2_process_discards(bs, ret); 1217 1218 /* Update L1 only if it isn't deleted anyway (addend = -1) */ 1219 if (ret == 0 && addend >= 0 && l1_modified) { 1220 for (i = 0; i < l1_size; i++) { 1221 cpu_to_be64s(&l1_table[i]); 1222 } 1223 1224 ret = bdrv_pwrite_sync(bs->file, l1_table_offset, 1225 l1_table, l1_size2); 1226 1227 for (i = 0; i < l1_size; i++) { 1228 be64_to_cpus(&l1_table[i]); 1229 } 1230 } 1231 if (l1_allocated) 1232 g_free(l1_table); 1233 return ret; 1234 } 1235 1236 1237 1238 1239 /*********************************************************/ 1240 /* refcount checking functions */ 1241 1242 1243 static uint64_t refcount_array_byte_size(BDRVQcow2State *s, uint64_t entries) 1244 { 1245 /* This assertion holds because there is no way we can address more than 1246 * 2^(64 - 9) clusters at once (with cluster size 512 = 2^9, and because 1247 * offsets have to be representable in bytes); due to every cluster 1248 * corresponding to one refcount entry, we are well below that limit */ 1249 assert(entries < (UINT64_C(1) << (64 - 9))); 1250 1251 /* Thanks to the assertion this will not overflow, because 1252 * s->refcount_order < 7. 1253 * (note: x << s->refcount_order == x * s->refcount_bits) */ 1254 return DIV_ROUND_UP(entries << s->refcount_order, 8); 1255 } 1256 1257 /** 1258 * Reallocates *array so that it can hold new_size entries. *size must contain 1259 * the current number of entries in *array. If the reallocation fails, *array 1260 * and *size will not be modified and -errno will be returned. If the 1261 * reallocation is successful, *array will be set to the new buffer, *size 1262 * will be set to new_size and 0 will be returned. The size of the reallocated 1263 * refcount array buffer will be aligned to a cluster boundary, and the newly 1264 * allocated area will be zeroed. 1265 */ 1266 static int realloc_refcount_array(BDRVQcow2State *s, void **array, 1267 int64_t *size, int64_t new_size) 1268 { 1269 int64_t old_byte_size, new_byte_size; 1270 void *new_ptr; 1271 1272 /* Round to clusters so the array can be directly written to disk */ 1273 old_byte_size = size_to_clusters(s, refcount_array_byte_size(s, *size)) 1274 * s->cluster_size; 1275 new_byte_size = size_to_clusters(s, refcount_array_byte_size(s, new_size)) 1276 * s->cluster_size; 1277 1278 if (new_byte_size == old_byte_size) { 1279 *size = new_size; 1280 return 0; 1281 } 1282 1283 assert(new_byte_size > 0); 1284 1285 if (new_byte_size > SIZE_MAX) { 1286 return -ENOMEM; 1287 } 1288 1289 new_ptr = g_try_realloc(*array, new_byte_size); 1290 if (!new_ptr) { 1291 return -ENOMEM; 1292 } 1293 1294 if (new_byte_size > old_byte_size) { 1295 memset((char *)new_ptr + old_byte_size, 0, 1296 new_byte_size - old_byte_size); 1297 } 1298 1299 *array = new_ptr; 1300 *size = new_size; 1301 1302 return 0; 1303 } 1304 1305 /* 1306 * Increases the refcount for a range of clusters in a given refcount table. 1307 * This is used to construct a temporary refcount table out of L1 and L2 tables 1308 * which can be compared to the refcount table saved in the image. 1309 * 1310 * Modifies the number of errors in res. 1311 */ 1312 static int inc_refcounts(BlockDriverState *bs, 1313 BdrvCheckResult *res, 1314 void **refcount_table, 1315 int64_t *refcount_table_size, 1316 int64_t offset, int64_t size) 1317 { 1318 BDRVQcow2State *s = bs->opaque; 1319 uint64_t start, last, cluster_offset, k, refcount; 1320 int ret; 1321 1322 if (size <= 0) { 1323 return 0; 1324 } 1325 1326 start = start_of_cluster(s, offset); 1327 last = start_of_cluster(s, offset + size - 1); 1328 for(cluster_offset = start; cluster_offset <= last; 1329 cluster_offset += s->cluster_size) { 1330 k = cluster_offset >> s->cluster_bits; 1331 if (k >= *refcount_table_size) { 1332 ret = realloc_refcount_array(s, refcount_table, 1333 refcount_table_size, k + 1); 1334 if (ret < 0) { 1335 res->check_errors++; 1336 return ret; 1337 } 1338 } 1339 1340 refcount = s->get_refcount(*refcount_table, k); 1341 if (refcount == s->refcount_max) { 1342 fprintf(stderr, "ERROR: overflow cluster offset=0x%" PRIx64 1343 "\n", cluster_offset); 1344 fprintf(stderr, "Use qemu-img amend to increase the refcount entry " 1345 "width or qemu-img convert to create a clean copy if the " 1346 "image cannot be opened for writing\n"); 1347 res->corruptions++; 1348 continue; 1349 } 1350 s->set_refcount(*refcount_table, k, refcount + 1); 1351 } 1352 1353 return 0; 1354 } 1355 1356 /* Flags for check_refcounts_l1() and check_refcounts_l2() */ 1357 enum { 1358 CHECK_FRAG_INFO = 0x2, /* update BlockFragInfo counters */ 1359 }; 1360 1361 /* 1362 * Increases the refcount in the given refcount table for the all clusters 1363 * referenced in the L2 table. While doing so, performs some checks on L2 1364 * entries. 1365 * 1366 * Returns the number of errors found by the checks or -errno if an internal 1367 * error occurred. 1368 */ 1369 static int check_refcounts_l2(BlockDriverState *bs, BdrvCheckResult *res, 1370 void **refcount_table, 1371 int64_t *refcount_table_size, int64_t l2_offset, 1372 int flags) 1373 { 1374 BDRVQcow2State *s = bs->opaque; 1375 uint64_t *l2_table, l2_entry; 1376 uint64_t next_contiguous_offset = 0; 1377 int i, l2_size, nb_csectors, ret; 1378 1379 /* Read L2 table from disk */ 1380 l2_size = s->l2_size * sizeof(uint64_t); 1381 l2_table = g_malloc(l2_size); 1382 1383 ret = bdrv_pread(bs->file, l2_offset, l2_table, l2_size); 1384 if (ret < 0) { 1385 fprintf(stderr, "ERROR: I/O error in check_refcounts_l2\n"); 1386 res->check_errors++; 1387 goto fail; 1388 } 1389 1390 /* Do the actual checks */ 1391 for(i = 0; i < s->l2_size; i++) { 1392 l2_entry = be64_to_cpu(l2_table[i]); 1393 1394 switch (qcow2_get_cluster_type(l2_entry)) { 1395 case QCOW2_CLUSTER_COMPRESSED: 1396 /* Compressed clusters don't have QCOW_OFLAG_COPIED */ 1397 if (l2_entry & QCOW_OFLAG_COPIED) { 1398 fprintf(stderr, "ERROR: cluster %" PRId64 ": " 1399 "copied flag must never be set for compressed " 1400 "clusters\n", l2_entry >> s->cluster_bits); 1401 l2_entry &= ~QCOW_OFLAG_COPIED; 1402 res->corruptions++; 1403 } 1404 1405 /* Mark cluster as used */ 1406 nb_csectors = ((l2_entry >> s->csize_shift) & 1407 s->csize_mask) + 1; 1408 l2_entry &= s->cluster_offset_mask; 1409 ret = inc_refcounts(bs, res, refcount_table, refcount_table_size, 1410 l2_entry & ~511, nb_csectors * 512); 1411 if (ret < 0) { 1412 goto fail; 1413 } 1414 1415 if (flags & CHECK_FRAG_INFO) { 1416 res->bfi.allocated_clusters++; 1417 res->bfi.compressed_clusters++; 1418 1419 /* Compressed clusters are fragmented by nature. Since they 1420 * take up sub-sector space but we only have sector granularity 1421 * I/O we need to re-read the same sectors even for adjacent 1422 * compressed clusters. 1423 */ 1424 res->bfi.fragmented_clusters++; 1425 } 1426 break; 1427 1428 case QCOW2_CLUSTER_ZERO: 1429 if ((l2_entry & L2E_OFFSET_MASK) == 0) { 1430 break; 1431 } 1432 /* fall through */ 1433 1434 case QCOW2_CLUSTER_NORMAL: 1435 { 1436 uint64_t offset = l2_entry & L2E_OFFSET_MASK; 1437 1438 if (flags & CHECK_FRAG_INFO) { 1439 res->bfi.allocated_clusters++; 1440 if (next_contiguous_offset && 1441 offset != next_contiguous_offset) { 1442 res->bfi.fragmented_clusters++; 1443 } 1444 next_contiguous_offset = offset + s->cluster_size; 1445 } 1446 1447 /* Mark cluster as used */ 1448 ret = inc_refcounts(bs, res, refcount_table, refcount_table_size, 1449 offset, s->cluster_size); 1450 if (ret < 0) { 1451 goto fail; 1452 } 1453 1454 /* Correct offsets are cluster aligned */ 1455 if (offset_into_cluster(s, offset)) { 1456 fprintf(stderr, "ERROR offset=%" PRIx64 ": Cluster is not " 1457 "properly aligned; L2 entry corrupted.\n", offset); 1458 res->corruptions++; 1459 } 1460 break; 1461 } 1462 1463 case QCOW2_CLUSTER_UNALLOCATED: 1464 break; 1465 1466 default: 1467 abort(); 1468 } 1469 } 1470 1471 g_free(l2_table); 1472 return 0; 1473 1474 fail: 1475 g_free(l2_table); 1476 return ret; 1477 } 1478 1479 /* 1480 * Increases the refcount for the L1 table, its L2 tables and all referenced 1481 * clusters in the given refcount table. While doing so, performs some checks 1482 * on L1 and L2 entries. 1483 * 1484 * Returns the number of errors found by the checks or -errno if an internal 1485 * error occurred. 1486 */ 1487 static int check_refcounts_l1(BlockDriverState *bs, 1488 BdrvCheckResult *res, 1489 void **refcount_table, 1490 int64_t *refcount_table_size, 1491 int64_t l1_table_offset, int l1_size, 1492 int flags) 1493 { 1494 BDRVQcow2State *s = bs->opaque; 1495 uint64_t *l1_table = NULL, l2_offset, l1_size2; 1496 int i, ret; 1497 1498 l1_size2 = l1_size * sizeof(uint64_t); 1499 1500 /* Mark L1 table as used */ 1501 ret = inc_refcounts(bs, res, refcount_table, refcount_table_size, 1502 l1_table_offset, l1_size2); 1503 if (ret < 0) { 1504 goto fail; 1505 } 1506 1507 /* Read L1 table entries from disk */ 1508 if (l1_size2 > 0) { 1509 l1_table = g_try_malloc(l1_size2); 1510 if (l1_table == NULL) { 1511 ret = -ENOMEM; 1512 res->check_errors++; 1513 goto fail; 1514 } 1515 ret = bdrv_pread(bs->file, l1_table_offset, l1_table, l1_size2); 1516 if (ret < 0) { 1517 fprintf(stderr, "ERROR: I/O error in check_refcounts_l1\n"); 1518 res->check_errors++; 1519 goto fail; 1520 } 1521 for(i = 0;i < l1_size; i++) 1522 be64_to_cpus(&l1_table[i]); 1523 } 1524 1525 /* Do the actual checks */ 1526 for(i = 0; i < l1_size; i++) { 1527 l2_offset = l1_table[i]; 1528 if (l2_offset) { 1529 /* Mark L2 table as used */ 1530 l2_offset &= L1E_OFFSET_MASK; 1531 ret = inc_refcounts(bs, res, refcount_table, refcount_table_size, 1532 l2_offset, s->cluster_size); 1533 if (ret < 0) { 1534 goto fail; 1535 } 1536 1537 /* L2 tables are cluster aligned */ 1538 if (offset_into_cluster(s, l2_offset)) { 1539 fprintf(stderr, "ERROR l2_offset=%" PRIx64 ": Table is not " 1540 "cluster aligned; L1 entry corrupted\n", l2_offset); 1541 res->corruptions++; 1542 } 1543 1544 /* Process and check L2 entries */ 1545 ret = check_refcounts_l2(bs, res, refcount_table, 1546 refcount_table_size, l2_offset, flags); 1547 if (ret < 0) { 1548 goto fail; 1549 } 1550 } 1551 } 1552 g_free(l1_table); 1553 return 0; 1554 1555 fail: 1556 g_free(l1_table); 1557 return ret; 1558 } 1559 1560 /* 1561 * Checks the OFLAG_COPIED flag for all L1 and L2 entries. 1562 * 1563 * This function does not print an error message nor does it increment 1564 * check_errors if qcow2_get_refcount fails (this is because such an error will 1565 * have been already detected and sufficiently signaled by the calling function 1566 * (qcow2_check_refcounts) by the time this function is called). 1567 */ 1568 static int check_oflag_copied(BlockDriverState *bs, BdrvCheckResult *res, 1569 BdrvCheckMode fix) 1570 { 1571 BDRVQcow2State *s = bs->opaque; 1572 uint64_t *l2_table = qemu_blockalign(bs, s->cluster_size); 1573 int ret; 1574 uint64_t refcount; 1575 int i, j; 1576 1577 for (i = 0; i < s->l1_size; i++) { 1578 uint64_t l1_entry = s->l1_table[i]; 1579 uint64_t l2_offset = l1_entry & L1E_OFFSET_MASK; 1580 bool l2_dirty = false; 1581 1582 if (!l2_offset) { 1583 continue; 1584 } 1585 1586 ret = qcow2_get_refcount(bs, l2_offset >> s->cluster_bits, 1587 &refcount); 1588 if (ret < 0) { 1589 /* don't print message nor increment check_errors */ 1590 continue; 1591 } 1592 if ((refcount == 1) != ((l1_entry & QCOW_OFLAG_COPIED) != 0)) { 1593 fprintf(stderr, "%s OFLAG_COPIED L2 cluster: l1_index=%d " 1594 "l1_entry=%" PRIx64 " refcount=%" PRIu64 "\n", 1595 fix & BDRV_FIX_ERRORS ? "Repairing" : 1596 "ERROR", 1597 i, l1_entry, refcount); 1598 if (fix & BDRV_FIX_ERRORS) { 1599 s->l1_table[i] = refcount == 1 1600 ? l1_entry | QCOW_OFLAG_COPIED 1601 : l1_entry & ~QCOW_OFLAG_COPIED; 1602 ret = qcow2_write_l1_entry(bs, i); 1603 if (ret < 0) { 1604 res->check_errors++; 1605 goto fail; 1606 } 1607 res->corruptions_fixed++; 1608 } else { 1609 res->corruptions++; 1610 } 1611 } 1612 1613 ret = bdrv_pread(bs->file, l2_offset, l2_table, 1614 s->l2_size * sizeof(uint64_t)); 1615 if (ret < 0) { 1616 fprintf(stderr, "ERROR: Could not read L2 table: %s\n", 1617 strerror(-ret)); 1618 res->check_errors++; 1619 goto fail; 1620 } 1621 1622 for (j = 0; j < s->l2_size; j++) { 1623 uint64_t l2_entry = be64_to_cpu(l2_table[j]); 1624 uint64_t data_offset = l2_entry & L2E_OFFSET_MASK; 1625 int cluster_type = qcow2_get_cluster_type(l2_entry); 1626 1627 if ((cluster_type == QCOW2_CLUSTER_NORMAL) || 1628 ((cluster_type == QCOW2_CLUSTER_ZERO) && (data_offset != 0))) { 1629 ret = qcow2_get_refcount(bs, 1630 data_offset >> s->cluster_bits, 1631 &refcount); 1632 if (ret < 0) { 1633 /* don't print message nor increment check_errors */ 1634 continue; 1635 } 1636 if ((refcount == 1) != ((l2_entry & QCOW_OFLAG_COPIED) != 0)) { 1637 fprintf(stderr, "%s OFLAG_COPIED data cluster: " 1638 "l2_entry=%" PRIx64 " refcount=%" PRIu64 "\n", 1639 fix & BDRV_FIX_ERRORS ? "Repairing" : 1640 "ERROR", 1641 l2_entry, refcount); 1642 if (fix & BDRV_FIX_ERRORS) { 1643 l2_table[j] = cpu_to_be64(refcount == 1 1644 ? l2_entry | QCOW_OFLAG_COPIED 1645 : l2_entry & ~QCOW_OFLAG_COPIED); 1646 l2_dirty = true; 1647 res->corruptions_fixed++; 1648 } else { 1649 res->corruptions++; 1650 } 1651 } 1652 } 1653 } 1654 1655 if (l2_dirty) { 1656 ret = qcow2_pre_write_overlap_check(bs, QCOW2_OL_ACTIVE_L2, 1657 l2_offset, s->cluster_size); 1658 if (ret < 0) { 1659 fprintf(stderr, "ERROR: Could not write L2 table; metadata " 1660 "overlap check failed: %s\n", strerror(-ret)); 1661 res->check_errors++; 1662 goto fail; 1663 } 1664 1665 ret = bdrv_pwrite(bs->file, l2_offset, l2_table, 1666 s->cluster_size); 1667 if (ret < 0) { 1668 fprintf(stderr, "ERROR: Could not write L2 table: %s\n", 1669 strerror(-ret)); 1670 res->check_errors++; 1671 goto fail; 1672 } 1673 } 1674 } 1675 1676 ret = 0; 1677 1678 fail: 1679 qemu_vfree(l2_table); 1680 return ret; 1681 } 1682 1683 /* 1684 * Checks consistency of refblocks and accounts for each refblock in 1685 * *refcount_table. 1686 */ 1687 static int check_refblocks(BlockDriverState *bs, BdrvCheckResult *res, 1688 BdrvCheckMode fix, bool *rebuild, 1689 void **refcount_table, int64_t *nb_clusters) 1690 { 1691 BDRVQcow2State *s = bs->opaque; 1692 int64_t i, size; 1693 int ret; 1694 1695 for(i = 0; i < s->refcount_table_size; i++) { 1696 uint64_t offset, cluster; 1697 offset = s->refcount_table[i]; 1698 cluster = offset >> s->cluster_bits; 1699 1700 /* Refcount blocks are cluster aligned */ 1701 if (offset_into_cluster(s, offset)) { 1702 fprintf(stderr, "ERROR refcount block %" PRId64 " is not " 1703 "cluster aligned; refcount table entry corrupted\n", i); 1704 res->corruptions++; 1705 *rebuild = true; 1706 continue; 1707 } 1708 1709 if (cluster >= *nb_clusters) { 1710 fprintf(stderr, "%s refcount block %" PRId64 " is outside image\n", 1711 fix & BDRV_FIX_ERRORS ? "Repairing" : "ERROR", i); 1712 1713 if (fix & BDRV_FIX_ERRORS) { 1714 int64_t new_nb_clusters; 1715 1716 if (offset > INT64_MAX - s->cluster_size) { 1717 ret = -EINVAL; 1718 goto resize_fail; 1719 } 1720 1721 ret = bdrv_truncate(bs->file->bs, offset + s->cluster_size); 1722 if (ret < 0) { 1723 goto resize_fail; 1724 } 1725 size = bdrv_getlength(bs->file->bs); 1726 if (size < 0) { 1727 ret = size; 1728 goto resize_fail; 1729 } 1730 1731 new_nb_clusters = size_to_clusters(s, size); 1732 assert(new_nb_clusters >= *nb_clusters); 1733 1734 ret = realloc_refcount_array(s, refcount_table, 1735 nb_clusters, new_nb_clusters); 1736 if (ret < 0) { 1737 res->check_errors++; 1738 return ret; 1739 } 1740 1741 if (cluster >= *nb_clusters) { 1742 ret = -EINVAL; 1743 goto resize_fail; 1744 } 1745 1746 res->corruptions_fixed++; 1747 ret = inc_refcounts(bs, res, refcount_table, nb_clusters, 1748 offset, s->cluster_size); 1749 if (ret < 0) { 1750 return ret; 1751 } 1752 /* No need to check whether the refcount is now greater than 1: 1753 * This area was just allocated and zeroed, so it can only be 1754 * exactly 1 after inc_refcounts() */ 1755 continue; 1756 1757 resize_fail: 1758 res->corruptions++; 1759 *rebuild = true; 1760 fprintf(stderr, "ERROR could not resize image: %s\n", 1761 strerror(-ret)); 1762 } else { 1763 res->corruptions++; 1764 } 1765 continue; 1766 } 1767 1768 if (offset != 0) { 1769 ret = inc_refcounts(bs, res, refcount_table, nb_clusters, 1770 offset, s->cluster_size); 1771 if (ret < 0) { 1772 return ret; 1773 } 1774 if (s->get_refcount(*refcount_table, cluster) != 1) { 1775 fprintf(stderr, "ERROR refcount block %" PRId64 1776 " refcount=%" PRIu64 "\n", i, 1777 s->get_refcount(*refcount_table, cluster)); 1778 res->corruptions++; 1779 *rebuild = true; 1780 } 1781 } 1782 } 1783 1784 return 0; 1785 } 1786 1787 /* 1788 * Calculates an in-memory refcount table. 1789 */ 1790 static int calculate_refcounts(BlockDriverState *bs, BdrvCheckResult *res, 1791 BdrvCheckMode fix, bool *rebuild, 1792 void **refcount_table, int64_t *nb_clusters) 1793 { 1794 BDRVQcow2State *s = bs->opaque; 1795 int64_t i; 1796 QCowSnapshot *sn; 1797 int ret; 1798 1799 if (!*refcount_table) { 1800 int64_t old_size = 0; 1801 ret = realloc_refcount_array(s, refcount_table, 1802 &old_size, *nb_clusters); 1803 if (ret < 0) { 1804 res->check_errors++; 1805 return ret; 1806 } 1807 } 1808 1809 /* header */ 1810 ret = inc_refcounts(bs, res, refcount_table, nb_clusters, 1811 0, s->cluster_size); 1812 if (ret < 0) { 1813 return ret; 1814 } 1815 1816 /* current L1 table */ 1817 ret = check_refcounts_l1(bs, res, refcount_table, nb_clusters, 1818 s->l1_table_offset, s->l1_size, CHECK_FRAG_INFO); 1819 if (ret < 0) { 1820 return ret; 1821 } 1822 1823 /* snapshots */ 1824 for (i = 0; i < s->nb_snapshots; i++) { 1825 sn = s->snapshots + i; 1826 ret = check_refcounts_l1(bs, res, refcount_table, nb_clusters, 1827 sn->l1_table_offset, sn->l1_size, 0); 1828 if (ret < 0) { 1829 return ret; 1830 } 1831 } 1832 ret = inc_refcounts(bs, res, refcount_table, nb_clusters, 1833 s->snapshots_offset, s->snapshots_size); 1834 if (ret < 0) { 1835 return ret; 1836 } 1837 1838 /* refcount data */ 1839 ret = inc_refcounts(bs, res, refcount_table, nb_clusters, 1840 s->refcount_table_offset, 1841 s->refcount_table_size * sizeof(uint64_t)); 1842 if (ret < 0) { 1843 return ret; 1844 } 1845 1846 return check_refblocks(bs, res, fix, rebuild, refcount_table, nb_clusters); 1847 } 1848 1849 /* 1850 * Compares the actual reference count for each cluster in the image against the 1851 * refcount as reported by the refcount structures on-disk. 1852 */ 1853 static void compare_refcounts(BlockDriverState *bs, BdrvCheckResult *res, 1854 BdrvCheckMode fix, bool *rebuild, 1855 int64_t *highest_cluster, 1856 void *refcount_table, int64_t nb_clusters) 1857 { 1858 BDRVQcow2State *s = bs->opaque; 1859 int64_t i; 1860 uint64_t refcount1, refcount2; 1861 int ret; 1862 1863 for (i = 0, *highest_cluster = 0; i < nb_clusters; i++) { 1864 ret = qcow2_get_refcount(bs, i, &refcount1); 1865 if (ret < 0) { 1866 fprintf(stderr, "Can't get refcount for cluster %" PRId64 ": %s\n", 1867 i, strerror(-ret)); 1868 res->check_errors++; 1869 continue; 1870 } 1871 1872 refcount2 = s->get_refcount(refcount_table, i); 1873 1874 if (refcount1 > 0 || refcount2 > 0) { 1875 *highest_cluster = i; 1876 } 1877 1878 if (refcount1 != refcount2) { 1879 /* Check if we're allowed to fix the mismatch */ 1880 int *num_fixed = NULL; 1881 if (refcount1 == 0) { 1882 *rebuild = true; 1883 } else if (refcount1 > refcount2 && (fix & BDRV_FIX_LEAKS)) { 1884 num_fixed = &res->leaks_fixed; 1885 } else if (refcount1 < refcount2 && (fix & BDRV_FIX_ERRORS)) { 1886 num_fixed = &res->corruptions_fixed; 1887 } 1888 1889 fprintf(stderr, "%s cluster %" PRId64 " refcount=%" PRIu64 1890 " reference=%" PRIu64 "\n", 1891 num_fixed != NULL ? "Repairing" : 1892 refcount1 < refcount2 ? "ERROR" : 1893 "Leaked", 1894 i, refcount1, refcount2); 1895 1896 if (num_fixed) { 1897 ret = update_refcount(bs, i << s->cluster_bits, 1, 1898 refcount_diff(refcount1, refcount2), 1899 refcount1 > refcount2, 1900 QCOW2_DISCARD_ALWAYS); 1901 if (ret >= 0) { 1902 (*num_fixed)++; 1903 continue; 1904 } 1905 } 1906 1907 /* And if we couldn't, print an error */ 1908 if (refcount1 < refcount2) { 1909 res->corruptions++; 1910 } else { 1911 res->leaks++; 1912 } 1913 } 1914 } 1915 } 1916 1917 /* 1918 * Allocates clusters using an in-memory refcount table (IMRT) in contrast to 1919 * the on-disk refcount structures. 1920 * 1921 * On input, *first_free_cluster tells where to start looking, and need not 1922 * actually be a free cluster; the returned offset will not be before that 1923 * cluster. On output, *first_free_cluster points to the first gap found, even 1924 * if that gap was too small to be used as the returned offset. 1925 * 1926 * Note that *first_free_cluster is a cluster index whereas the return value is 1927 * an offset. 1928 */ 1929 static int64_t alloc_clusters_imrt(BlockDriverState *bs, 1930 int cluster_count, 1931 void **refcount_table, 1932 int64_t *imrt_nb_clusters, 1933 int64_t *first_free_cluster) 1934 { 1935 BDRVQcow2State *s = bs->opaque; 1936 int64_t cluster = *first_free_cluster, i; 1937 bool first_gap = true; 1938 int contiguous_free_clusters; 1939 int ret; 1940 1941 /* Starting at *first_free_cluster, find a range of at least cluster_count 1942 * continuously free clusters */ 1943 for (contiguous_free_clusters = 0; 1944 cluster < *imrt_nb_clusters && 1945 contiguous_free_clusters < cluster_count; 1946 cluster++) 1947 { 1948 if (!s->get_refcount(*refcount_table, cluster)) { 1949 contiguous_free_clusters++; 1950 if (first_gap) { 1951 /* If this is the first free cluster found, update 1952 * *first_free_cluster accordingly */ 1953 *first_free_cluster = cluster; 1954 first_gap = false; 1955 } 1956 } else if (contiguous_free_clusters) { 1957 contiguous_free_clusters = 0; 1958 } 1959 } 1960 1961 /* If contiguous_free_clusters is greater than zero, it contains the number 1962 * of continuously free clusters until the current cluster; the first free 1963 * cluster in the current "gap" is therefore 1964 * cluster - contiguous_free_clusters */ 1965 1966 /* If no such range could be found, grow the in-memory refcount table 1967 * accordingly to append free clusters at the end of the image */ 1968 if (contiguous_free_clusters < cluster_count) { 1969 /* contiguous_free_clusters clusters are already empty at the image end; 1970 * we need cluster_count clusters; therefore, we have to allocate 1971 * cluster_count - contiguous_free_clusters new clusters at the end of 1972 * the image (which is the current value of cluster; note that cluster 1973 * may exceed old_imrt_nb_clusters if *first_free_cluster pointed beyond 1974 * the image end) */ 1975 ret = realloc_refcount_array(s, refcount_table, imrt_nb_clusters, 1976 cluster + cluster_count 1977 - contiguous_free_clusters); 1978 if (ret < 0) { 1979 return ret; 1980 } 1981 } 1982 1983 /* Go back to the first free cluster */ 1984 cluster -= contiguous_free_clusters; 1985 for (i = 0; i < cluster_count; i++) { 1986 s->set_refcount(*refcount_table, cluster + i, 1); 1987 } 1988 1989 return cluster << s->cluster_bits; 1990 } 1991 1992 /* 1993 * Creates a new refcount structure based solely on the in-memory information 1994 * given through *refcount_table. All necessary allocations will be reflected 1995 * in that array. 1996 * 1997 * On success, the old refcount structure is leaked (it will be covered by the 1998 * new refcount structure). 1999 */ 2000 static int rebuild_refcount_structure(BlockDriverState *bs, 2001 BdrvCheckResult *res, 2002 void **refcount_table, 2003 int64_t *nb_clusters) 2004 { 2005 BDRVQcow2State *s = bs->opaque; 2006 int64_t first_free_cluster = 0, reftable_offset = -1, cluster = 0; 2007 int64_t refblock_offset, refblock_start, refblock_index; 2008 uint32_t reftable_size = 0; 2009 uint64_t *on_disk_reftable = NULL; 2010 void *on_disk_refblock; 2011 int ret = 0; 2012 struct { 2013 uint64_t reftable_offset; 2014 uint32_t reftable_clusters; 2015 } QEMU_PACKED reftable_offset_and_clusters; 2016 2017 qcow2_cache_empty(bs, s->refcount_block_cache); 2018 2019 write_refblocks: 2020 for (; cluster < *nb_clusters; cluster++) { 2021 if (!s->get_refcount(*refcount_table, cluster)) { 2022 continue; 2023 } 2024 2025 refblock_index = cluster >> s->refcount_block_bits; 2026 refblock_start = refblock_index << s->refcount_block_bits; 2027 2028 /* Don't allocate a cluster in a refblock already written to disk */ 2029 if (first_free_cluster < refblock_start) { 2030 first_free_cluster = refblock_start; 2031 } 2032 refblock_offset = alloc_clusters_imrt(bs, 1, refcount_table, 2033 nb_clusters, &first_free_cluster); 2034 if (refblock_offset < 0) { 2035 fprintf(stderr, "ERROR allocating refblock: %s\n", 2036 strerror(-refblock_offset)); 2037 res->check_errors++; 2038 ret = refblock_offset; 2039 goto fail; 2040 } 2041 2042 if (reftable_size <= refblock_index) { 2043 uint32_t old_reftable_size = reftable_size; 2044 uint64_t *new_on_disk_reftable; 2045 2046 reftable_size = ROUND_UP((refblock_index + 1) * sizeof(uint64_t), 2047 s->cluster_size) / sizeof(uint64_t); 2048 new_on_disk_reftable = g_try_realloc(on_disk_reftable, 2049 reftable_size * 2050 sizeof(uint64_t)); 2051 if (!new_on_disk_reftable) { 2052 res->check_errors++; 2053 ret = -ENOMEM; 2054 goto fail; 2055 } 2056 on_disk_reftable = new_on_disk_reftable; 2057 2058 memset(on_disk_reftable + old_reftable_size, 0, 2059 (reftable_size - old_reftable_size) * sizeof(uint64_t)); 2060 2061 /* The offset we have for the reftable is now no longer valid; 2062 * this will leak that range, but we can easily fix that by running 2063 * a leak-fixing check after this rebuild operation */ 2064 reftable_offset = -1; 2065 } 2066 on_disk_reftable[refblock_index] = refblock_offset; 2067 2068 /* If this is apparently the last refblock (for now), try to squeeze the 2069 * reftable in */ 2070 if (refblock_index == (*nb_clusters - 1) >> s->refcount_block_bits && 2071 reftable_offset < 0) 2072 { 2073 uint64_t reftable_clusters = size_to_clusters(s, reftable_size * 2074 sizeof(uint64_t)); 2075 reftable_offset = alloc_clusters_imrt(bs, reftable_clusters, 2076 refcount_table, nb_clusters, 2077 &first_free_cluster); 2078 if (reftable_offset < 0) { 2079 fprintf(stderr, "ERROR allocating reftable: %s\n", 2080 strerror(-reftable_offset)); 2081 res->check_errors++; 2082 ret = reftable_offset; 2083 goto fail; 2084 } 2085 } 2086 2087 ret = qcow2_pre_write_overlap_check(bs, 0, refblock_offset, 2088 s->cluster_size); 2089 if (ret < 0) { 2090 fprintf(stderr, "ERROR writing refblock: %s\n", strerror(-ret)); 2091 goto fail; 2092 } 2093 2094 /* The size of *refcount_table is always cluster-aligned, therefore the 2095 * write operation will not overflow */ 2096 on_disk_refblock = (void *)((char *) *refcount_table + 2097 refblock_index * s->cluster_size); 2098 2099 ret = bdrv_write(bs->file, refblock_offset / BDRV_SECTOR_SIZE, 2100 on_disk_refblock, s->cluster_sectors); 2101 if (ret < 0) { 2102 fprintf(stderr, "ERROR writing refblock: %s\n", strerror(-ret)); 2103 goto fail; 2104 } 2105 2106 /* Go to the end of this refblock */ 2107 cluster = refblock_start + s->refcount_block_size - 1; 2108 } 2109 2110 if (reftable_offset < 0) { 2111 uint64_t post_refblock_start, reftable_clusters; 2112 2113 post_refblock_start = ROUND_UP(*nb_clusters, s->refcount_block_size); 2114 reftable_clusters = size_to_clusters(s, 2115 reftable_size * sizeof(uint64_t)); 2116 /* Not pretty but simple */ 2117 if (first_free_cluster < post_refblock_start) { 2118 first_free_cluster = post_refblock_start; 2119 } 2120 reftable_offset = alloc_clusters_imrt(bs, reftable_clusters, 2121 refcount_table, nb_clusters, 2122 &first_free_cluster); 2123 if (reftable_offset < 0) { 2124 fprintf(stderr, "ERROR allocating reftable: %s\n", 2125 strerror(-reftable_offset)); 2126 res->check_errors++; 2127 ret = reftable_offset; 2128 goto fail; 2129 } 2130 2131 goto write_refblocks; 2132 } 2133 2134 assert(on_disk_reftable); 2135 2136 for (refblock_index = 0; refblock_index < reftable_size; refblock_index++) { 2137 cpu_to_be64s(&on_disk_reftable[refblock_index]); 2138 } 2139 2140 ret = qcow2_pre_write_overlap_check(bs, 0, reftable_offset, 2141 reftable_size * sizeof(uint64_t)); 2142 if (ret < 0) { 2143 fprintf(stderr, "ERROR writing reftable: %s\n", strerror(-ret)); 2144 goto fail; 2145 } 2146 2147 assert(reftable_size < INT_MAX / sizeof(uint64_t)); 2148 ret = bdrv_pwrite(bs->file, reftable_offset, on_disk_reftable, 2149 reftable_size * sizeof(uint64_t)); 2150 if (ret < 0) { 2151 fprintf(stderr, "ERROR writing reftable: %s\n", strerror(-ret)); 2152 goto fail; 2153 } 2154 2155 /* Enter new reftable into the image header */ 2156 reftable_offset_and_clusters.reftable_offset = cpu_to_be64(reftable_offset); 2157 reftable_offset_and_clusters.reftable_clusters = 2158 cpu_to_be32(size_to_clusters(s, reftable_size * sizeof(uint64_t))); 2159 ret = bdrv_pwrite_sync(bs->file, 2160 offsetof(QCowHeader, refcount_table_offset), 2161 &reftable_offset_and_clusters, 2162 sizeof(reftable_offset_and_clusters)); 2163 if (ret < 0) { 2164 fprintf(stderr, "ERROR setting reftable: %s\n", strerror(-ret)); 2165 goto fail; 2166 } 2167 2168 for (refblock_index = 0; refblock_index < reftable_size; refblock_index++) { 2169 be64_to_cpus(&on_disk_reftable[refblock_index]); 2170 } 2171 s->refcount_table = on_disk_reftable; 2172 s->refcount_table_offset = reftable_offset; 2173 s->refcount_table_size = reftable_size; 2174 2175 return 0; 2176 2177 fail: 2178 g_free(on_disk_reftable); 2179 return ret; 2180 } 2181 2182 /* 2183 * Checks an image for refcount consistency. 2184 * 2185 * Returns 0 if no errors are found, the number of errors in case the image is 2186 * detected as corrupted, and -errno when an internal error occurred. 2187 */ 2188 int qcow2_check_refcounts(BlockDriverState *bs, BdrvCheckResult *res, 2189 BdrvCheckMode fix) 2190 { 2191 BDRVQcow2State *s = bs->opaque; 2192 BdrvCheckResult pre_compare_res; 2193 int64_t size, highest_cluster, nb_clusters; 2194 void *refcount_table = NULL; 2195 bool rebuild = false; 2196 int ret; 2197 2198 size = bdrv_getlength(bs->file->bs); 2199 if (size < 0) { 2200 res->check_errors++; 2201 return size; 2202 } 2203 2204 nb_clusters = size_to_clusters(s, size); 2205 if (nb_clusters > INT_MAX) { 2206 res->check_errors++; 2207 return -EFBIG; 2208 } 2209 2210 res->bfi.total_clusters = 2211 size_to_clusters(s, bs->total_sectors * BDRV_SECTOR_SIZE); 2212 2213 ret = calculate_refcounts(bs, res, fix, &rebuild, &refcount_table, 2214 &nb_clusters); 2215 if (ret < 0) { 2216 goto fail; 2217 } 2218 2219 /* In case we don't need to rebuild the refcount structure (but want to fix 2220 * something), this function is immediately called again, in which case the 2221 * result should be ignored */ 2222 pre_compare_res = *res; 2223 compare_refcounts(bs, res, 0, &rebuild, &highest_cluster, refcount_table, 2224 nb_clusters); 2225 2226 if (rebuild && (fix & BDRV_FIX_ERRORS)) { 2227 BdrvCheckResult old_res = *res; 2228 int fresh_leaks = 0; 2229 2230 fprintf(stderr, "Rebuilding refcount structure\n"); 2231 ret = rebuild_refcount_structure(bs, res, &refcount_table, 2232 &nb_clusters); 2233 if (ret < 0) { 2234 goto fail; 2235 } 2236 2237 res->corruptions = 0; 2238 res->leaks = 0; 2239 2240 /* Because the old reftable has been exchanged for a new one the 2241 * references have to be recalculated */ 2242 rebuild = false; 2243 memset(refcount_table, 0, refcount_array_byte_size(s, nb_clusters)); 2244 ret = calculate_refcounts(bs, res, 0, &rebuild, &refcount_table, 2245 &nb_clusters); 2246 if (ret < 0) { 2247 goto fail; 2248 } 2249 2250 if (fix & BDRV_FIX_LEAKS) { 2251 /* The old refcount structures are now leaked, fix it; the result 2252 * can be ignored, aside from leaks which were introduced by 2253 * rebuild_refcount_structure() that could not be fixed */ 2254 BdrvCheckResult saved_res = *res; 2255 *res = (BdrvCheckResult){ 0 }; 2256 2257 compare_refcounts(bs, res, BDRV_FIX_LEAKS, &rebuild, 2258 &highest_cluster, refcount_table, nb_clusters); 2259 if (rebuild) { 2260 fprintf(stderr, "ERROR rebuilt refcount structure is still " 2261 "broken\n"); 2262 } 2263 2264 /* Any leaks accounted for here were introduced by 2265 * rebuild_refcount_structure() because that function has created a 2266 * new refcount structure from scratch */ 2267 fresh_leaks = res->leaks; 2268 *res = saved_res; 2269 } 2270 2271 if (res->corruptions < old_res.corruptions) { 2272 res->corruptions_fixed += old_res.corruptions - res->corruptions; 2273 } 2274 if (res->leaks < old_res.leaks) { 2275 res->leaks_fixed += old_res.leaks - res->leaks; 2276 } 2277 res->leaks += fresh_leaks; 2278 } else if (fix) { 2279 if (rebuild) { 2280 fprintf(stderr, "ERROR need to rebuild refcount structures\n"); 2281 res->check_errors++; 2282 ret = -EIO; 2283 goto fail; 2284 } 2285 2286 if (res->leaks || res->corruptions) { 2287 *res = pre_compare_res; 2288 compare_refcounts(bs, res, fix, &rebuild, &highest_cluster, 2289 refcount_table, nb_clusters); 2290 } 2291 } 2292 2293 /* check OFLAG_COPIED */ 2294 ret = check_oflag_copied(bs, res, fix); 2295 if (ret < 0) { 2296 goto fail; 2297 } 2298 2299 res->image_end_offset = (highest_cluster + 1) * s->cluster_size; 2300 ret = 0; 2301 2302 fail: 2303 g_free(refcount_table); 2304 2305 return ret; 2306 } 2307 2308 #define overlaps_with(ofs, sz) \ 2309 ranges_overlap(offset, size, ofs, sz) 2310 2311 /* 2312 * Checks if the given offset into the image file is actually free to use by 2313 * looking for overlaps with important metadata sections (L1/L2 tables etc.), 2314 * i.e. a sanity check without relying on the refcount tables. 2315 * 2316 * The ign parameter specifies what checks not to perform (being a bitmask of 2317 * QCow2MetadataOverlap values), i.e., what sections to ignore. 2318 * 2319 * Returns: 2320 * - 0 if writing to this offset will not affect the mentioned metadata 2321 * - a positive QCow2MetadataOverlap value indicating one overlapping section 2322 * - a negative value (-errno) indicating an error while performing a check, 2323 * e.g. when bdrv_read failed on QCOW2_OL_INACTIVE_L2 2324 */ 2325 int qcow2_check_metadata_overlap(BlockDriverState *bs, int ign, int64_t offset, 2326 int64_t size) 2327 { 2328 BDRVQcow2State *s = bs->opaque; 2329 int chk = s->overlap_check & ~ign; 2330 int i, j; 2331 2332 if (!size) { 2333 return 0; 2334 } 2335 2336 if (chk & QCOW2_OL_MAIN_HEADER) { 2337 if (offset < s->cluster_size) { 2338 return QCOW2_OL_MAIN_HEADER; 2339 } 2340 } 2341 2342 /* align range to test to cluster boundaries */ 2343 size = align_offset(offset_into_cluster(s, offset) + size, s->cluster_size); 2344 offset = start_of_cluster(s, offset); 2345 2346 if ((chk & QCOW2_OL_ACTIVE_L1) && s->l1_size) { 2347 if (overlaps_with(s->l1_table_offset, s->l1_size * sizeof(uint64_t))) { 2348 return QCOW2_OL_ACTIVE_L1; 2349 } 2350 } 2351 2352 if ((chk & QCOW2_OL_REFCOUNT_TABLE) && s->refcount_table_size) { 2353 if (overlaps_with(s->refcount_table_offset, 2354 s->refcount_table_size * sizeof(uint64_t))) { 2355 return QCOW2_OL_REFCOUNT_TABLE; 2356 } 2357 } 2358 2359 if ((chk & QCOW2_OL_SNAPSHOT_TABLE) && s->snapshots_size) { 2360 if (overlaps_with(s->snapshots_offset, s->snapshots_size)) { 2361 return QCOW2_OL_SNAPSHOT_TABLE; 2362 } 2363 } 2364 2365 if ((chk & QCOW2_OL_INACTIVE_L1) && s->snapshots) { 2366 for (i = 0; i < s->nb_snapshots; i++) { 2367 if (s->snapshots[i].l1_size && 2368 overlaps_with(s->snapshots[i].l1_table_offset, 2369 s->snapshots[i].l1_size * sizeof(uint64_t))) { 2370 return QCOW2_OL_INACTIVE_L1; 2371 } 2372 } 2373 } 2374 2375 if ((chk & QCOW2_OL_ACTIVE_L2) && s->l1_table) { 2376 for (i = 0; i < s->l1_size; i++) { 2377 if ((s->l1_table[i] & L1E_OFFSET_MASK) && 2378 overlaps_with(s->l1_table[i] & L1E_OFFSET_MASK, 2379 s->cluster_size)) { 2380 return QCOW2_OL_ACTIVE_L2; 2381 } 2382 } 2383 } 2384 2385 if ((chk & QCOW2_OL_REFCOUNT_BLOCK) && s->refcount_table) { 2386 for (i = 0; i < s->refcount_table_size; i++) { 2387 if ((s->refcount_table[i] & REFT_OFFSET_MASK) && 2388 overlaps_with(s->refcount_table[i] & REFT_OFFSET_MASK, 2389 s->cluster_size)) { 2390 return QCOW2_OL_REFCOUNT_BLOCK; 2391 } 2392 } 2393 } 2394 2395 if ((chk & QCOW2_OL_INACTIVE_L2) && s->snapshots) { 2396 for (i = 0; i < s->nb_snapshots; i++) { 2397 uint64_t l1_ofs = s->snapshots[i].l1_table_offset; 2398 uint32_t l1_sz = s->snapshots[i].l1_size; 2399 uint64_t l1_sz2 = l1_sz * sizeof(uint64_t); 2400 uint64_t *l1 = g_try_malloc(l1_sz2); 2401 int ret; 2402 2403 if (l1_sz2 && l1 == NULL) { 2404 return -ENOMEM; 2405 } 2406 2407 ret = bdrv_pread(bs->file, l1_ofs, l1, l1_sz2); 2408 if (ret < 0) { 2409 g_free(l1); 2410 return ret; 2411 } 2412 2413 for (j = 0; j < l1_sz; j++) { 2414 uint64_t l2_ofs = be64_to_cpu(l1[j]) & L1E_OFFSET_MASK; 2415 if (l2_ofs && overlaps_with(l2_ofs, s->cluster_size)) { 2416 g_free(l1); 2417 return QCOW2_OL_INACTIVE_L2; 2418 } 2419 } 2420 2421 g_free(l1); 2422 } 2423 } 2424 2425 return 0; 2426 } 2427 2428 static const char *metadata_ol_names[] = { 2429 [QCOW2_OL_MAIN_HEADER_BITNR] = "qcow2_header", 2430 [QCOW2_OL_ACTIVE_L1_BITNR] = "active L1 table", 2431 [QCOW2_OL_ACTIVE_L2_BITNR] = "active L2 table", 2432 [QCOW2_OL_REFCOUNT_TABLE_BITNR] = "refcount table", 2433 [QCOW2_OL_REFCOUNT_BLOCK_BITNR] = "refcount block", 2434 [QCOW2_OL_SNAPSHOT_TABLE_BITNR] = "snapshot table", 2435 [QCOW2_OL_INACTIVE_L1_BITNR] = "inactive L1 table", 2436 [QCOW2_OL_INACTIVE_L2_BITNR] = "inactive L2 table", 2437 }; 2438 2439 /* 2440 * First performs a check for metadata overlaps (through 2441 * qcow2_check_metadata_overlap); if that fails with a negative value (error 2442 * while performing a check), that value is returned. If an impending overlap 2443 * is detected, the BDS will be made unusable, the qcow2 file marked corrupt 2444 * and -EIO returned. 2445 * 2446 * Returns 0 if there were neither overlaps nor errors while checking for 2447 * overlaps; or a negative value (-errno) on error. 2448 */ 2449 int qcow2_pre_write_overlap_check(BlockDriverState *bs, int ign, int64_t offset, 2450 int64_t size) 2451 { 2452 int ret = qcow2_check_metadata_overlap(bs, ign, offset, size); 2453 2454 if (ret < 0) { 2455 return ret; 2456 } else if (ret > 0) { 2457 int metadata_ol_bitnr = ctz32(ret); 2458 assert(metadata_ol_bitnr < QCOW2_OL_MAX_BITNR); 2459 2460 qcow2_signal_corruption(bs, true, offset, size, "Preventing invalid " 2461 "write on metadata (overlaps with %s)", 2462 metadata_ol_names[metadata_ol_bitnr]); 2463 return -EIO; 2464 } 2465 2466 return 0; 2467 } 2468 2469 /* A pointer to a function of this type is given to walk_over_reftable(). That 2470 * function will create refblocks and pass them to a RefblockFinishOp once they 2471 * are completed (@refblock). @refblock_empty is set if the refblock is 2472 * completely empty. 2473 * 2474 * Along with the refblock, a corresponding reftable entry is passed, in the 2475 * reftable @reftable (which may be reallocated) at @reftable_index. 2476 * 2477 * @allocated should be set to true if a new cluster has been allocated. 2478 */ 2479 typedef int (RefblockFinishOp)(BlockDriverState *bs, uint64_t **reftable, 2480 uint64_t reftable_index, uint64_t *reftable_size, 2481 void *refblock, bool refblock_empty, 2482 bool *allocated, Error **errp); 2483 2484 /** 2485 * This "operation" for walk_over_reftable() allocates the refblock on disk (if 2486 * it is not empty) and inserts its offset into the new reftable. The size of 2487 * this new reftable is increased as required. 2488 */ 2489 static int alloc_refblock(BlockDriverState *bs, uint64_t **reftable, 2490 uint64_t reftable_index, uint64_t *reftable_size, 2491 void *refblock, bool refblock_empty, bool *allocated, 2492 Error **errp) 2493 { 2494 BDRVQcow2State *s = bs->opaque; 2495 int64_t offset; 2496 2497 if (!refblock_empty && reftable_index >= *reftable_size) { 2498 uint64_t *new_reftable; 2499 uint64_t new_reftable_size; 2500 2501 new_reftable_size = ROUND_UP(reftable_index + 1, 2502 s->cluster_size / sizeof(uint64_t)); 2503 if (new_reftable_size > QCOW_MAX_REFTABLE_SIZE / sizeof(uint64_t)) { 2504 error_setg(errp, 2505 "This operation would make the refcount table grow " 2506 "beyond the maximum size supported by QEMU, aborting"); 2507 return -ENOTSUP; 2508 } 2509 2510 new_reftable = g_try_realloc(*reftable, new_reftable_size * 2511 sizeof(uint64_t)); 2512 if (!new_reftable) { 2513 error_setg(errp, "Failed to increase reftable buffer size"); 2514 return -ENOMEM; 2515 } 2516 2517 memset(new_reftable + *reftable_size, 0, 2518 (new_reftable_size - *reftable_size) * sizeof(uint64_t)); 2519 2520 *reftable = new_reftable; 2521 *reftable_size = new_reftable_size; 2522 } 2523 2524 if (!refblock_empty && !(*reftable)[reftable_index]) { 2525 offset = qcow2_alloc_clusters(bs, s->cluster_size); 2526 if (offset < 0) { 2527 error_setg_errno(errp, -offset, "Failed to allocate refblock"); 2528 return offset; 2529 } 2530 (*reftable)[reftable_index] = offset; 2531 *allocated = true; 2532 } 2533 2534 return 0; 2535 } 2536 2537 /** 2538 * This "operation" for walk_over_reftable() writes the refblock to disk at the 2539 * offset specified by the new reftable's entry. It does not modify the new 2540 * reftable or change any refcounts. 2541 */ 2542 static int flush_refblock(BlockDriverState *bs, uint64_t **reftable, 2543 uint64_t reftable_index, uint64_t *reftable_size, 2544 void *refblock, bool refblock_empty, bool *allocated, 2545 Error **errp) 2546 { 2547 BDRVQcow2State *s = bs->opaque; 2548 int64_t offset; 2549 int ret; 2550 2551 if (reftable_index < *reftable_size && (*reftable)[reftable_index]) { 2552 offset = (*reftable)[reftable_index]; 2553 2554 ret = qcow2_pre_write_overlap_check(bs, 0, offset, s->cluster_size); 2555 if (ret < 0) { 2556 error_setg_errno(errp, -ret, "Overlap check failed"); 2557 return ret; 2558 } 2559 2560 ret = bdrv_pwrite(bs->file, offset, refblock, s->cluster_size); 2561 if (ret < 0) { 2562 error_setg_errno(errp, -ret, "Failed to write refblock"); 2563 return ret; 2564 } 2565 } else { 2566 assert(refblock_empty); 2567 } 2568 2569 return 0; 2570 } 2571 2572 /** 2573 * This function walks over the existing reftable and every referenced refblock; 2574 * if @new_set_refcount is non-NULL, it is called for every refcount entry to 2575 * create an equal new entry in the passed @new_refblock. Once that 2576 * @new_refblock is completely filled, @operation will be called. 2577 * 2578 * @status_cb and @cb_opaque are used for the amend operation's status callback. 2579 * @index is the index of the walk_over_reftable() calls and @total is the total 2580 * number of walk_over_reftable() calls per amend operation. Both are used for 2581 * calculating the parameters for the status callback. 2582 * 2583 * @allocated is set to true if a new cluster has been allocated. 2584 */ 2585 static int walk_over_reftable(BlockDriverState *bs, uint64_t **new_reftable, 2586 uint64_t *new_reftable_index, 2587 uint64_t *new_reftable_size, 2588 void *new_refblock, int new_refblock_size, 2589 int new_refcount_bits, 2590 RefblockFinishOp *operation, bool *allocated, 2591 Qcow2SetRefcountFunc *new_set_refcount, 2592 BlockDriverAmendStatusCB *status_cb, 2593 void *cb_opaque, int index, int total, 2594 Error **errp) 2595 { 2596 BDRVQcow2State *s = bs->opaque; 2597 uint64_t reftable_index; 2598 bool new_refblock_empty = true; 2599 int refblock_index; 2600 int new_refblock_index = 0; 2601 int ret; 2602 2603 for (reftable_index = 0; reftable_index < s->refcount_table_size; 2604 reftable_index++) 2605 { 2606 uint64_t refblock_offset = s->refcount_table[reftable_index] 2607 & REFT_OFFSET_MASK; 2608 2609 status_cb(bs, (uint64_t)index * s->refcount_table_size + reftable_index, 2610 (uint64_t)total * s->refcount_table_size, cb_opaque); 2611 2612 if (refblock_offset) { 2613 void *refblock; 2614 2615 if (offset_into_cluster(s, refblock_offset)) { 2616 qcow2_signal_corruption(bs, true, -1, -1, "Refblock offset %#" 2617 PRIx64 " unaligned (reftable index: %#" 2618 PRIx64 ")", refblock_offset, 2619 reftable_index); 2620 error_setg(errp, 2621 "Image is corrupt (unaligned refblock offset)"); 2622 return -EIO; 2623 } 2624 2625 ret = qcow2_cache_get(bs, s->refcount_block_cache, refblock_offset, 2626 &refblock); 2627 if (ret < 0) { 2628 error_setg_errno(errp, -ret, "Failed to retrieve refblock"); 2629 return ret; 2630 } 2631 2632 for (refblock_index = 0; refblock_index < s->refcount_block_size; 2633 refblock_index++) 2634 { 2635 uint64_t refcount; 2636 2637 if (new_refblock_index >= new_refblock_size) { 2638 /* new_refblock is now complete */ 2639 ret = operation(bs, new_reftable, *new_reftable_index, 2640 new_reftable_size, new_refblock, 2641 new_refblock_empty, allocated, errp); 2642 if (ret < 0) { 2643 qcow2_cache_put(bs, s->refcount_block_cache, &refblock); 2644 return ret; 2645 } 2646 2647 (*new_reftable_index)++; 2648 new_refblock_index = 0; 2649 new_refblock_empty = true; 2650 } 2651 2652 refcount = s->get_refcount(refblock, refblock_index); 2653 if (new_refcount_bits < 64 && refcount >> new_refcount_bits) { 2654 uint64_t offset; 2655 2656 qcow2_cache_put(bs, s->refcount_block_cache, &refblock); 2657 2658 offset = ((reftable_index << s->refcount_block_bits) 2659 + refblock_index) << s->cluster_bits; 2660 2661 error_setg(errp, "Cannot decrease refcount entry width to " 2662 "%i bits: Cluster at offset %#" PRIx64 " has a " 2663 "refcount of %" PRIu64, new_refcount_bits, 2664 offset, refcount); 2665 return -EINVAL; 2666 } 2667 2668 if (new_set_refcount) { 2669 new_set_refcount(new_refblock, new_refblock_index++, 2670 refcount); 2671 } else { 2672 new_refblock_index++; 2673 } 2674 new_refblock_empty = new_refblock_empty && refcount == 0; 2675 } 2676 2677 qcow2_cache_put(bs, s->refcount_block_cache, &refblock); 2678 } else { 2679 /* No refblock means every refcount is 0 */ 2680 for (refblock_index = 0; refblock_index < s->refcount_block_size; 2681 refblock_index++) 2682 { 2683 if (new_refblock_index >= new_refblock_size) { 2684 /* new_refblock is now complete */ 2685 ret = operation(bs, new_reftable, *new_reftable_index, 2686 new_reftable_size, new_refblock, 2687 new_refblock_empty, allocated, errp); 2688 if (ret < 0) { 2689 return ret; 2690 } 2691 2692 (*new_reftable_index)++; 2693 new_refblock_index = 0; 2694 new_refblock_empty = true; 2695 } 2696 2697 if (new_set_refcount) { 2698 new_set_refcount(new_refblock, new_refblock_index++, 0); 2699 } else { 2700 new_refblock_index++; 2701 } 2702 } 2703 } 2704 } 2705 2706 if (new_refblock_index > 0) { 2707 /* Complete the potentially existing partially filled final refblock */ 2708 if (new_set_refcount) { 2709 for (; new_refblock_index < new_refblock_size; 2710 new_refblock_index++) 2711 { 2712 new_set_refcount(new_refblock, new_refblock_index, 0); 2713 } 2714 } 2715 2716 ret = operation(bs, new_reftable, *new_reftable_index, 2717 new_reftable_size, new_refblock, new_refblock_empty, 2718 allocated, errp); 2719 if (ret < 0) { 2720 return ret; 2721 } 2722 2723 (*new_reftable_index)++; 2724 } 2725 2726 status_cb(bs, (uint64_t)(index + 1) * s->refcount_table_size, 2727 (uint64_t)total * s->refcount_table_size, cb_opaque); 2728 2729 return 0; 2730 } 2731 2732 int qcow2_change_refcount_order(BlockDriverState *bs, int refcount_order, 2733 BlockDriverAmendStatusCB *status_cb, 2734 void *cb_opaque, Error **errp) 2735 { 2736 BDRVQcow2State *s = bs->opaque; 2737 Qcow2GetRefcountFunc *new_get_refcount; 2738 Qcow2SetRefcountFunc *new_set_refcount; 2739 void *new_refblock = qemu_blockalign(bs->file->bs, s->cluster_size); 2740 uint64_t *new_reftable = NULL, new_reftable_size = 0; 2741 uint64_t *old_reftable, old_reftable_size, old_reftable_offset; 2742 uint64_t new_reftable_index = 0; 2743 uint64_t i; 2744 int64_t new_reftable_offset = 0, allocated_reftable_size = 0; 2745 int new_refblock_size, new_refcount_bits = 1 << refcount_order; 2746 int old_refcount_order; 2747 int walk_index = 0; 2748 int ret; 2749 bool new_allocation; 2750 2751 assert(s->qcow_version >= 3); 2752 assert(refcount_order >= 0 && refcount_order <= 6); 2753 2754 /* see qcow2_open() */ 2755 new_refblock_size = 1 << (s->cluster_bits - (refcount_order - 3)); 2756 2757 new_get_refcount = get_refcount_funcs[refcount_order]; 2758 new_set_refcount = set_refcount_funcs[refcount_order]; 2759 2760 2761 do { 2762 int total_walks; 2763 2764 new_allocation = false; 2765 2766 /* At least we have to do this walk and the one which writes the 2767 * refblocks; also, at least we have to do this loop here at least 2768 * twice (normally), first to do the allocations, and second to 2769 * determine that everything is correctly allocated, this then makes 2770 * three walks in total */ 2771 total_walks = MAX(walk_index + 2, 3); 2772 2773 /* First, allocate the structures so they are present in the refcount 2774 * structures */ 2775 ret = walk_over_reftable(bs, &new_reftable, &new_reftable_index, 2776 &new_reftable_size, NULL, new_refblock_size, 2777 new_refcount_bits, &alloc_refblock, 2778 &new_allocation, NULL, status_cb, cb_opaque, 2779 walk_index++, total_walks, errp); 2780 if (ret < 0) { 2781 goto done; 2782 } 2783 2784 new_reftable_index = 0; 2785 2786 if (new_allocation) { 2787 if (new_reftable_offset) { 2788 qcow2_free_clusters(bs, new_reftable_offset, 2789 allocated_reftable_size * sizeof(uint64_t), 2790 QCOW2_DISCARD_NEVER); 2791 } 2792 2793 new_reftable_offset = qcow2_alloc_clusters(bs, new_reftable_size * 2794 sizeof(uint64_t)); 2795 if (new_reftable_offset < 0) { 2796 error_setg_errno(errp, -new_reftable_offset, 2797 "Failed to allocate the new reftable"); 2798 ret = new_reftable_offset; 2799 goto done; 2800 } 2801 allocated_reftable_size = new_reftable_size; 2802 } 2803 } while (new_allocation); 2804 2805 /* Second, write the new refblocks */ 2806 ret = walk_over_reftable(bs, &new_reftable, &new_reftable_index, 2807 &new_reftable_size, new_refblock, 2808 new_refblock_size, new_refcount_bits, 2809 &flush_refblock, &new_allocation, new_set_refcount, 2810 status_cb, cb_opaque, walk_index, walk_index + 1, 2811 errp); 2812 if (ret < 0) { 2813 goto done; 2814 } 2815 assert(!new_allocation); 2816 2817 2818 /* Write the new reftable */ 2819 ret = qcow2_pre_write_overlap_check(bs, 0, new_reftable_offset, 2820 new_reftable_size * sizeof(uint64_t)); 2821 if (ret < 0) { 2822 error_setg_errno(errp, -ret, "Overlap check failed"); 2823 goto done; 2824 } 2825 2826 for (i = 0; i < new_reftable_size; i++) { 2827 cpu_to_be64s(&new_reftable[i]); 2828 } 2829 2830 ret = bdrv_pwrite(bs->file, new_reftable_offset, new_reftable, 2831 new_reftable_size * sizeof(uint64_t)); 2832 2833 for (i = 0; i < new_reftable_size; i++) { 2834 be64_to_cpus(&new_reftable[i]); 2835 } 2836 2837 if (ret < 0) { 2838 error_setg_errno(errp, -ret, "Failed to write the new reftable"); 2839 goto done; 2840 } 2841 2842 2843 /* Empty the refcount cache */ 2844 ret = qcow2_cache_flush(bs, s->refcount_block_cache); 2845 if (ret < 0) { 2846 error_setg_errno(errp, -ret, "Failed to flush the refblock cache"); 2847 goto done; 2848 } 2849 2850 /* Update the image header to point to the new reftable; this only updates 2851 * the fields which are relevant to qcow2_update_header(); other fields 2852 * such as s->refcount_table or s->refcount_bits stay stale for now 2853 * (because we have to restore everything if qcow2_update_header() fails) */ 2854 old_refcount_order = s->refcount_order; 2855 old_reftable_size = s->refcount_table_size; 2856 old_reftable_offset = s->refcount_table_offset; 2857 2858 s->refcount_order = refcount_order; 2859 s->refcount_table_size = new_reftable_size; 2860 s->refcount_table_offset = new_reftable_offset; 2861 2862 ret = qcow2_update_header(bs); 2863 if (ret < 0) { 2864 s->refcount_order = old_refcount_order; 2865 s->refcount_table_size = old_reftable_size; 2866 s->refcount_table_offset = old_reftable_offset; 2867 error_setg_errno(errp, -ret, "Failed to update the qcow2 header"); 2868 goto done; 2869 } 2870 2871 /* Now update the rest of the in-memory information */ 2872 old_reftable = s->refcount_table; 2873 s->refcount_table = new_reftable; 2874 2875 s->refcount_bits = 1 << refcount_order; 2876 s->refcount_max = UINT64_C(1) << (s->refcount_bits - 1); 2877 s->refcount_max += s->refcount_max - 1; 2878 2879 s->refcount_block_bits = s->cluster_bits - (refcount_order - 3); 2880 s->refcount_block_size = 1 << s->refcount_block_bits; 2881 2882 s->get_refcount = new_get_refcount; 2883 s->set_refcount = new_set_refcount; 2884 2885 /* For cleaning up all old refblocks and the old reftable below the "done" 2886 * label */ 2887 new_reftable = old_reftable; 2888 new_reftable_size = old_reftable_size; 2889 new_reftable_offset = old_reftable_offset; 2890 2891 done: 2892 if (new_reftable) { 2893 /* On success, new_reftable actually points to the old reftable (and 2894 * new_reftable_size is the old reftable's size); but that is just 2895 * fine */ 2896 for (i = 0; i < new_reftable_size; i++) { 2897 uint64_t offset = new_reftable[i] & REFT_OFFSET_MASK; 2898 if (offset) { 2899 qcow2_free_clusters(bs, offset, s->cluster_size, 2900 QCOW2_DISCARD_OTHER); 2901 } 2902 } 2903 g_free(new_reftable); 2904 2905 if (new_reftable_offset > 0) { 2906 qcow2_free_clusters(bs, new_reftable_offset, 2907 new_reftable_size * sizeof(uint64_t), 2908 QCOW2_DISCARD_OTHER); 2909 } 2910 } 2911 2912 qemu_vfree(new_refblock); 2913 return ret; 2914 } 2915