2 * Block driver for the QCOW version 2 format
4 * Copyright (c) 2004-2006 Fabrice Bellard
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:
13 * The above copyright notice and this permission notice shall be included in
14 * all copies or substantial portions of the Software.
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
24 #include "qemu-common.h"
25 #include "block_int.h"
31 Differences with QCOW:
33 - Support for multiple incremental snapshots.
34 - Memory management by reference counts.
35 - Clusters which have a reference count of one have the bit
36 QCOW_OFLAG_COPIED to optimize write performance.
37 - Size of compressed clusters is stored in sectors to reduce bit usage
38 in the cluster offsets.
39 - Support for storing additional data (such as the VM state) in the
41 - If a backing store is used, the cluster size is not constrained
42 (could be backported to QCOW).
43 - L2 tables have always a size of one cluster.
47 //#define DEBUG_ALLOC2
49 #define QCOW_MAGIC (('Q' << 24) | ('F' << 16) | ('I' << 8) | 0xfb)
50 #define QCOW_VERSION 2
52 #define QCOW_CRYPT_NONE 0
53 #define QCOW_CRYPT_AES 1
55 /* indicate that the refcount of the referenced cluster is exactly one. */
56 #define QCOW_OFLAG_COPIED (1LL << 63)
57 /* indicate that the cluster is compressed (they never have the copied flag) */
58 #define QCOW_OFLAG_COMPRESSED (1LL << 62)
60 #define REFCOUNT_SHIFT 1 /* refcount size is 2 bytes */
63 #define offsetof(type, field) ((size_t) &((type *)0)->field)
66 typedef struct QCowHeader {
69 uint64_t backing_file_offset;
70 uint32_t backing_file_size;
71 uint32_t cluster_bits;
72 uint64_t size; /* in bytes */
73 uint32_t crypt_method;
74 uint32_t l1_size; /* XXX: save number of clusters instead ? */
75 uint64_t l1_table_offset;
76 uint64_t refcount_table_offset;
77 uint32_t refcount_table_clusters;
78 uint32_t nb_snapshots;
79 uint64_t snapshots_offset;
82 typedef struct __attribute__((packed)) QCowSnapshotHeader {
83 /* header is 8 byte aligned */
84 uint64_t l1_table_offset;
93 uint64_t vm_clock_nsec;
95 uint32_t vm_state_size;
96 uint32_t extra_data_size; /* for extension */
97 /* extra data follows */
100 } QCowSnapshotHeader;
102 #define L2_CACHE_SIZE 16
104 typedef struct QCowSnapshot {
105 uint64_t l1_table_offset;
109 uint32_t vm_state_size;
112 uint64_t vm_clock_nsec;
115 typedef struct BDRVQcowState {
116 BlockDriverState *hd;
123 int l1_vm_state_index;
126 uint64_t cluster_offset_mask;
127 uint64_t l1_table_offset;
130 uint64_t l2_cache_offsets[L2_CACHE_SIZE];
131 uint32_t l2_cache_counts[L2_CACHE_SIZE];
132 uint8_t *cluster_cache;
133 uint8_t *cluster_data;
134 uint64_t cluster_cache_offset;
136 uint64_t *refcount_table;
137 uint64_t refcount_table_offset;
138 uint32_t refcount_table_size;
139 uint64_t refcount_block_cache_offset;
140 uint16_t *refcount_block_cache;
141 int64_t free_cluster_index;
142 int64_t free_byte_offset;
144 uint32_t crypt_method; /* current crypt method, 0 if no key yet */
145 uint32_t crypt_method_header;
146 AES_KEY aes_encrypt_key;
147 AES_KEY aes_decrypt_key;
148 uint64_t snapshots_offset;
151 QCowSnapshot *snapshots;
154 static int decompress_cluster(BDRVQcowState *s, uint64_t cluster_offset);
155 static int qcow_read(BlockDriverState *bs, int64_t sector_num,
156 uint8_t *buf, int nb_sectors);
157 static int qcow_read_snapshots(BlockDriverState *bs);
158 static void qcow_free_snapshots(BlockDriverState *bs);
159 static int refcount_init(BlockDriverState *bs);
160 static void refcount_close(BlockDriverState *bs);
161 static int get_refcount(BlockDriverState *bs, int64_t cluster_index);
162 static int update_cluster_refcount(BlockDriverState *bs,
163 int64_t cluster_index,
165 static void update_refcount(BlockDriverState *bs,
166 int64_t offset, int64_t length,
168 static int64_t alloc_clusters(BlockDriverState *bs, int64_t size);
169 static int64_t alloc_bytes(BlockDriverState *bs, int size);
170 static void free_clusters(BlockDriverState *bs,
171 int64_t offset, int64_t size);
173 static void check_refcounts(BlockDriverState *bs);
176 static int qcow_probe(const uint8_t *buf, int buf_size, const char *filename)
178 const QCowHeader *cow_header = (const void *)buf;
180 if (buf_size >= sizeof(QCowHeader) &&
181 be32_to_cpu(cow_header->magic) == QCOW_MAGIC &&
182 be32_to_cpu(cow_header->version) == QCOW_VERSION)
188 static int qcow_open(BlockDriverState *bs, const char *filename, int flags)
190 BDRVQcowState *s = bs->opaque;
191 int len, i, shift, ret;
194 ret = bdrv_file_open(&s->hd, filename, flags);
197 if (bdrv_pread(s->hd, 0, &header, sizeof(header)) != sizeof(header))
199 be32_to_cpus(&header.magic);
200 be32_to_cpus(&header.version);
201 be64_to_cpus(&header.backing_file_offset);
202 be32_to_cpus(&header.backing_file_size);
203 be64_to_cpus(&header.size);
204 be32_to_cpus(&header.cluster_bits);
205 be32_to_cpus(&header.crypt_method);
206 be64_to_cpus(&header.l1_table_offset);
207 be32_to_cpus(&header.l1_size);
208 be64_to_cpus(&header.refcount_table_offset);
209 be32_to_cpus(&header.refcount_table_clusters);
210 be64_to_cpus(&header.snapshots_offset);
211 be32_to_cpus(&header.nb_snapshots);
213 if (header.magic != QCOW_MAGIC || header.version != QCOW_VERSION)
215 if (header.size <= 1 ||
216 header.cluster_bits < 9 ||
217 header.cluster_bits > 16)
219 if (header.crypt_method > QCOW_CRYPT_AES)
221 s->crypt_method_header = header.crypt_method;
222 if (s->crypt_method_header)
224 s->cluster_bits = header.cluster_bits;
225 s->cluster_size = 1 << s->cluster_bits;
226 s->cluster_sectors = 1 << (s->cluster_bits - 9);
227 s->l2_bits = s->cluster_bits - 3; /* L2 is always one cluster */
228 s->l2_size = 1 << s->l2_bits;
229 bs->total_sectors = header.size / 512;
230 s->csize_shift = (62 - (s->cluster_bits - 8));
231 s->csize_mask = (1 << (s->cluster_bits - 8)) - 1;
232 s->cluster_offset_mask = (1LL << s->csize_shift) - 1;
233 s->refcount_table_offset = header.refcount_table_offset;
234 s->refcount_table_size =
235 header.refcount_table_clusters << (s->cluster_bits - 3);
237 s->snapshots_offset = header.snapshots_offset;
238 s->nb_snapshots = header.nb_snapshots;
240 /* read the level 1 table */
241 s->l1_size = header.l1_size;
242 shift = s->cluster_bits + s->l2_bits;
243 s->l1_vm_state_index = (header.size + (1LL << shift) - 1) >> shift;
244 /* the L1 table must contain at least enough entries to put
246 if (s->l1_size < s->l1_vm_state_index)
248 s->l1_table_offset = header.l1_table_offset;
249 s->l1_table = qemu_malloc(s->l1_size * sizeof(uint64_t));
252 if (bdrv_pread(s->hd, s->l1_table_offset, s->l1_table, s->l1_size * sizeof(uint64_t)) !=
253 s->l1_size * sizeof(uint64_t))
255 for(i = 0;i < s->l1_size; i++) {
256 be64_to_cpus(&s->l1_table[i]);
259 s->l2_cache = qemu_malloc(s->l2_size * L2_CACHE_SIZE * sizeof(uint64_t));
262 s->cluster_cache = qemu_malloc(s->cluster_size);
263 if (!s->cluster_cache)
265 /* one more sector for decompressed data alignment */
266 s->cluster_data = qemu_malloc(s->cluster_size + 512);
267 if (!s->cluster_data)
269 s->cluster_cache_offset = -1;
271 if (refcount_init(bs) < 0)
274 /* read the backing file name */
275 if (header.backing_file_offset != 0) {
276 len = header.backing_file_size;
279 if (bdrv_pread(s->hd, header.backing_file_offset, bs->backing_file, len) != len)
281 bs->backing_file[len] = '\0';
283 if (qcow_read_snapshots(bs) < 0)
292 qcow_free_snapshots(bs);
294 qemu_free(s->l1_table);
295 qemu_free(s->l2_cache);
296 qemu_free(s->cluster_cache);
297 qemu_free(s->cluster_data);
302 static int qcow_set_key(BlockDriverState *bs, const char *key)
304 BDRVQcowState *s = bs->opaque;
308 memset(keybuf, 0, 16);
312 /* XXX: we could compress the chars to 7 bits to increase
314 for(i = 0;i < len;i++) {
317 s->crypt_method = s->crypt_method_header;
319 if (AES_set_encrypt_key(keybuf, 128, &s->aes_encrypt_key) != 0)
321 if (AES_set_decrypt_key(keybuf, 128, &s->aes_decrypt_key) != 0)
331 AES_encrypt(in, tmp, &s->aes_encrypt_key);
332 AES_decrypt(tmp, out, &s->aes_decrypt_key);
333 for(i = 0; i < 16; i++)
334 printf(" %02x", tmp[i]);
336 for(i = 0; i < 16; i++)
337 printf(" %02x", out[i]);
344 /* The crypt function is compatible with the linux cryptoloop
345 algorithm for < 4 GB images. NOTE: out_buf == in_buf is
347 static void encrypt_sectors(BDRVQcowState *s, int64_t sector_num,
348 uint8_t *out_buf, const uint8_t *in_buf,
349 int nb_sectors, int enc,
358 for(i = 0; i < nb_sectors; i++) {
359 ivec.ll[0] = cpu_to_le64(sector_num);
361 AES_cbc_encrypt(in_buf, out_buf, 512, key,
369 static int copy_sectors(BlockDriverState *bs, uint64_t start_sect,
370 uint64_t cluster_offset, int n_start, int n_end)
372 BDRVQcowState *s = bs->opaque;
378 ret = qcow_read(bs, start_sect + n_start, s->cluster_data, n);
381 if (s->crypt_method) {
382 encrypt_sectors(s, start_sect + n_start,
384 s->cluster_data, n, 1,
385 &s->aes_encrypt_key);
387 ret = bdrv_write(s->hd, (cluster_offset >> 9) + n_start,
394 static void l2_cache_reset(BlockDriverState *bs)
396 BDRVQcowState *s = bs->opaque;
398 memset(s->l2_cache, 0, s->l2_size * L2_CACHE_SIZE * sizeof(uint64_t));
399 memset(s->l2_cache_offsets, 0, L2_CACHE_SIZE * sizeof(uint64_t));
400 memset(s->l2_cache_counts, 0, L2_CACHE_SIZE * sizeof(uint32_t));
403 static inline int l2_cache_new_entry(BlockDriverState *bs)
405 BDRVQcowState *s = bs->opaque;
409 /* find a new entry in the least used one */
411 min_count = 0xffffffff;
412 for(i = 0; i < L2_CACHE_SIZE; i++) {
413 if (s->l2_cache_counts[i] < min_count) {
414 min_count = s->l2_cache_counts[i];
421 static int64_t align_offset(int64_t offset, int n)
423 offset = (offset + n - 1) & ~(n - 1);
427 static int grow_l1_table(BlockDriverState *bs, int min_size)
429 BDRVQcowState *s = bs->opaque;
430 int new_l1_size, new_l1_size2, ret, i;
431 uint64_t *new_l1_table;
432 uint64_t new_l1_table_offset;
436 new_l1_size = s->l1_size;
437 if (min_size <= new_l1_size)
439 while (min_size > new_l1_size) {
440 new_l1_size = (new_l1_size * 3 + 1) / 2;
443 printf("grow l1_table from %d to %d\n", s->l1_size, new_l1_size);
446 new_l1_size2 = sizeof(uint64_t) * new_l1_size;
447 new_l1_table = qemu_mallocz(new_l1_size2);
450 memcpy(new_l1_table, s->l1_table, s->l1_size * sizeof(uint64_t));
452 /* write new table (align to cluster) */
453 new_l1_table_offset = alloc_clusters(bs, new_l1_size2);
455 for(i = 0; i < s->l1_size; i++)
456 new_l1_table[i] = cpu_to_be64(new_l1_table[i]);
457 ret = bdrv_pwrite(s->hd, new_l1_table_offset, new_l1_table, new_l1_size2);
458 if (ret != new_l1_size2)
460 for(i = 0; i < s->l1_size; i++)
461 new_l1_table[i] = be64_to_cpu(new_l1_table[i]);
464 data64 = cpu_to_be64(new_l1_table_offset);
465 if (bdrv_pwrite(s->hd, offsetof(QCowHeader, l1_table_offset),
466 &data64, sizeof(data64)) != sizeof(data64))
468 data32 = cpu_to_be32(new_l1_size);
469 if (bdrv_pwrite(s->hd, offsetof(QCowHeader, l1_size),
470 &data32, sizeof(data32)) != sizeof(data32))
472 qemu_free(s->l1_table);
473 free_clusters(bs, s->l1_table_offset, s->l1_size * sizeof(uint64_t));
474 s->l1_table_offset = new_l1_table_offset;
475 s->l1_table = new_l1_table;
476 s->l1_size = new_l1_size;
479 qemu_free(s->l1_table);
486 * seek l2_offset in the l2_cache table
487 * if not found, return NULL,
489 * increments the l2 cache hit count of the entry,
490 * if counter overflow, divide by two all counters
491 * return the pointer to the l2 cache entry
495 static uint64_t *seek_l2_table(BDRVQcowState *s, uint64_t l2_offset)
499 for(i = 0; i < L2_CACHE_SIZE; i++) {
500 if (l2_offset == s->l2_cache_offsets[i]) {
501 /* increment the hit count */
502 if (++s->l2_cache_counts[i] == 0xffffffff) {
503 for(j = 0; j < L2_CACHE_SIZE; j++) {
504 s->l2_cache_counts[j] >>= 1;
507 return s->l2_cache + (i << s->l2_bits);
516 * Loads a L2 table into memory. If the table is in the cache, the cache
517 * is used; otherwise the L2 table is loaded from the image file.
519 * Returns a pointer to the L2 table on success, or NULL if the read from
520 * the image file failed.
523 static uint64_t *l2_load(BlockDriverState *bs, uint64_t l2_offset)
525 BDRVQcowState *s = bs->opaque;
529 /* seek if the table for the given offset is in the cache */
531 l2_table = seek_l2_table(s, l2_offset);
532 if (l2_table != NULL)
535 /* not found: load a new entry in the least used one */
537 min_index = l2_cache_new_entry(bs);
538 l2_table = s->l2_cache + (min_index << s->l2_bits);
539 if (bdrv_pread(s->hd, l2_offset, l2_table, s->l2_size * sizeof(uint64_t)) !=
540 s->l2_size * sizeof(uint64_t))
542 s->l2_cache_offsets[min_index] = l2_offset;
543 s->l2_cache_counts[min_index] = 1;
551 * Allocate a new l2 entry in the file. If l1_index points to an already
552 * used entry in the L2 table (i.e. we are doing a copy on write for the L2
553 * table) copy the contents of the old L2 table into the newly allocated one.
554 * Otherwise the new table is initialized with zeros.
558 static uint64_t *l2_allocate(BlockDriverState *bs, int l1_index)
560 BDRVQcowState *s = bs->opaque;
562 uint64_t old_l2_offset, tmp;
563 uint64_t *l2_table, l2_offset;
565 old_l2_offset = s->l1_table[l1_index];
567 /* allocate a new l2 entry */
569 l2_offset = alloc_clusters(bs, s->l2_size * sizeof(uint64_t));
571 /* update the L1 entry */
573 s->l1_table[l1_index] = l2_offset | QCOW_OFLAG_COPIED;
575 tmp = cpu_to_be64(l2_offset | QCOW_OFLAG_COPIED);
576 if (bdrv_pwrite(s->hd, s->l1_table_offset + l1_index * sizeof(tmp),
577 &tmp, sizeof(tmp)) != sizeof(tmp))
580 /* allocate a new entry in the l2 cache */
582 min_index = l2_cache_new_entry(bs);
583 l2_table = s->l2_cache + (min_index << s->l2_bits);
585 if (old_l2_offset == 0) {
586 /* if there was no old l2 table, clear the new table */
587 memset(l2_table, 0, s->l2_size * sizeof(uint64_t));
589 /* if there was an old l2 table, read it from the disk */
590 if (bdrv_pread(s->hd, old_l2_offset,
591 l2_table, s->l2_size * sizeof(uint64_t)) !=
592 s->l2_size * sizeof(uint64_t))
595 /* write the l2 table to the file */
596 if (bdrv_pwrite(s->hd, l2_offset,
597 l2_table, s->l2_size * sizeof(uint64_t)) !=
598 s->l2_size * sizeof(uint64_t))
601 /* update the l2 cache entry */
603 s->l2_cache_offsets[min_index] = l2_offset;
604 s->l2_cache_counts[min_index] = 1;
612 * For a given offset of the disk image, return cluster offset in
615 * Return 1, if the offset is found
616 * Return 0, otherwise.
620 static uint64_t get_cluster_offset(BlockDriverState *bs, uint64_t offset)
622 BDRVQcowState *s = bs->opaque;
623 int l1_index, l2_index;
624 uint64_t l2_offset, *l2_table, cluster_offset;
626 /* seek the the l2 offset in the l1 table */
628 l1_index = offset >> (s->l2_bits + s->cluster_bits);
629 if (l1_index >= s->l1_size)
632 l2_offset = s->l1_table[l1_index];
634 /* seek the l2 table of the given l2 offset */
639 /* load the l2 table in memory */
641 l2_offset &= ~QCOW_OFLAG_COPIED;
642 l2_table = l2_load(bs, l2_offset);
643 if (l2_table == NULL)
646 /* find the cluster offset for the given disk offset */
648 l2_index = (offset >> s->cluster_bits) & (s->l2_size - 1);
649 cluster_offset = be64_to_cpu(l2_table[l2_index]);
651 return cluster_offset & ~QCOW_OFLAG_COPIED;
657 * free clusters according to its type: compressed or not
661 static void free_any_clusters(BlockDriverState *bs,
662 uint64_t cluster_offset)
664 BDRVQcowState *s = bs->opaque;
666 if (cluster_offset == 0)
669 /* free the cluster */
671 if (cluster_offset & QCOW_OFLAG_COMPRESSED) {
673 nb_csectors = ((cluster_offset >> s->csize_shift) &
675 free_clusters(bs, (cluster_offset & s->cluster_offset_mask) & ~511,
680 free_clusters(bs, cluster_offset, s->cluster_size);
686 * for a given disk offset, load (and allocate if needed)
689 * the l2 table offset in the qcow2 file and the cluster index
690 * in the l2 table are given to the caller.
694 static int get_cluster_table(BlockDriverState *bs, uint64_t offset,
695 uint64_t **new_l2_table,
696 uint64_t *new_l2_offset,
699 BDRVQcowState *s = bs->opaque;
700 int l1_index, l2_index, ret;
701 uint64_t l2_offset, *l2_table;
703 /* seek the the l2 offset in the l1 table */
705 l1_index = offset >> (s->l2_bits + s->cluster_bits);
706 if (l1_index >= s->l1_size) {
707 ret = grow_l1_table(bs, l1_index + 1);
711 l2_offset = s->l1_table[l1_index];
713 /* seek the l2 table of the given l2 offset */
715 if (l2_offset & QCOW_OFLAG_COPIED) {
716 /* load the l2 table in memory */
717 l2_offset &= ~QCOW_OFLAG_COPIED;
718 l2_table = l2_load(bs, l2_offset);
719 if (l2_table == NULL)
723 free_clusters(bs, l2_offset, s->l2_size * sizeof(uint64_t));
724 l2_table = l2_allocate(bs, l1_index);
725 if (l2_table == NULL)
727 l2_offset = s->l1_table[l1_index] & ~QCOW_OFLAG_COPIED;
730 /* find the cluster offset for the given disk offset */
732 l2_index = (offset >> s->cluster_bits) & (s->l2_size - 1);
734 *new_l2_table = l2_table;
735 *new_l2_offset = l2_offset;
736 *new_l2_index = l2_index;
742 * alloc_compressed_cluster_offset
744 * For a given offset of the disk image, return cluster offset in
747 * If the offset is not found, allocate a new compressed cluster.
749 * Return the cluster offset if successful,
750 * Return 0, otherwise.
754 static uint64_t alloc_compressed_cluster_offset(BlockDriverState *bs,
758 BDRVQcowState *s = bs->opaque;
760 uint64_t l2_offset, *l2_table, cluster_offset;
763 ret = get_cluster_table(bs, offset, &l2_table, &l2_offset, &l2_index);
767 cluster_offset = be64_to_cpu(l2_table[l2_index]);
768 if (cluster_offset & QCOW_OFLAG_COPIED)
769 return cluster_offset & ~QCOW_OFLAG_COPIED;
771 free_any_clusters(bs, cluster_offset);
773 cluster_offset = alloc_bytes(bs, compressed_size);
774 nb_csectors = ((cluster_offset + compressed_size - 1) >> 9) -
775 (cluster_offset >> 9);
777 cluster_offset |= QCOW_OFLAG_COMPRESSED |
778 ((uint64_t)nb_csectors << s->csize_shift);
780 /* update L2 table */
782 /* compressed clusters never have the copied flag */
784 l2_table[l2_index] = cpu_to_be64(cluster_offset);
785 if (bdrv_pwrite(s->hd,
786 l2_offset + l2_index * sizeof(uint64_t),
788 sizeof(uint64_t)) != sizeof(uint64_t))
791 return cluster_offset;
795 * alloc_cluster_offset
797 * For a given offset of the disk image, return cluster offset in
800 * If the offset is not found, allocate a new cluster.
802 * Return the cluster offset if successful,
803 * Return 0, otherwise.
807 static uint64_t alloc_cluster_offset(BlockDriverState *bs,
809 int n_start, int n_end)
811 BDRVQcowState *s = bs->opaque;
813 uint64_t l2_offset, *l2_table, cluster_offset;
815 ret = get_cluster_table(bs, offset, &l2_table, &l2_offset, &l2_index);
819 cluster_offset = be64_to_cpu(l2_table[l2_index]);
820 if (cluster_offset & QCOW_OFLAG_COPIED)
821 return cluster_offset & ~QCOW_OFLAG_COPIED;
823 free_any_clusters(bs, cluster_offset);
825 /* allocate a new cluster */
827 cluster_offset = alloc_clusters(bs, s->cluster_size);
829 /* we must initialize the cluster content which won't be
832 if ((n_end - n_start) < s->cluster_sectors) {
835 start_sect = (offset & ~(s->cluster_size - 1)) >> 9;
836 ret = copy_sectors(bs, start_sect,
837 cluster_offset, 0, n_start);
840 ret = copy_sectors(bs, start_sect,
841 cluster_offset, n_end, s->cluster_sectors);
846 /* update L2 table */
848 l2_table[l2_index] = cpu_to_be64(cluster_offset | QCOW_OFLAG_COPIED);
849 if (bdrv_pwrite(s->hd,
850 l2_offset + l2_index * sizeof(uint64_t),
852 sizeof(uint64_t)) != sizeof(uint64_t))
855 return cluster_offset;
858 static int qcow_is_allocated(BlockDriverState *bs, int64_t sector_num,
859 int nb_sectors, int *pnum)
861 BDRVQcowState *s = bs->opaque;
862 int index_in_cluster, n;
863 uint64_t cluster_offset;
865 cluster_offset = get_cluster_offset(bs, sector_num << 9);
866 index_in_cluster = sector_num & (s->cluster_sectors - 1);
867 n = s->cluster_sectors - index_in_cluster;
871 return (cluster_offset != 0);
874 static int decompress_buffer(uint8_t *out_buf, int out_buf_size,
875 const uint8_t *buf, int buf_size)
877 z_stream strm1, *strm = &strm1;
880 memset(strm, 0, sizeof(*strm));
882 strm->next_in = (uint8_t *)buf;
883 strm->avail_in = buf_size;
884 strm->next_out = out_buf;
885 strm->avail_out = out_buf_size;
887 ret = inflateInit2(strm, -12);
890 ret = inflate(strm, Z_FINISH);
891 out_len = strm->next_out - out_buf;
892 if ((ret != Z_STREAM_END && ret != Z_BUF_ERROR) ||
893 out_len != out_buf_size) {
901 static int decompress_cluster(BDRVQcowState *s, uint64_t cluster_offset)
903 int ret, csize, nb_csectors, sector_offset;
906 coffset = cluster_offset & s->cluster_offset_mask;
907 if (s->cluster_cache_offset != coffset) {
908 nb_csectors = ((cluster_offset >> s->csize_shift) & s->csize_mask) + 1;
909 sector_offset = coffset & 511;
910 csize = nb_csectors * 512 - sector_offset;
911 ret = bdrv_read(s->hd, coffset >> 9, s->cluster_data, nb_csectors);
915 if (decompress_buffer(s->cluster_cache, s->cluster_size,
916 s->cluster_data + sector_offset, csize) < 0) {
919 s->cluster_cache_offset = coffset;
924 /* handle reading after the end of the backing file */
925 static int backing_read1(BlockDriverState *bs,
926 int64_t sector_num, uint8_t *buf, int nb_sectors)
929 if ((sector_num + nb_sectors) <= bs->total_sectors)
931 if (sector_num >= bs->total_sectors)
934 n1 = bs->total_sectors - sector_num;
935 memset(buf + n1 * 512, 0, 512 * (nb_sectors - n1));
939 static int qcow_read(BlockDriverState *bs, int64_t sector_num,
940 uint8_t *buf, int nb_sectors)
942 BDRVQcowState *s = bs->opaque;
943 int ret, index_in_cluster, n, n1;
944 uint64_t cluster_offset;
946 while (nb_sectors > 0) {
947 cluster_offset = get_cluster_offset(bs, sector_num << 9);
948 index_in_cluster = sector_num & (s->cluster_sectors - 1);
949 n = s->cluster_sectors - index_in_cluster;
952 if (!cluster_offset) {
953 if (bs->backing_hd) {
954 /* read from the base image */
955 n1 = backing_read1(bs->backing_hd, sector_num, buf, n);
957 ret = bdrv_read(bs->backing_hd, sector_num, buf, n1);
962 memset(buf, 0, 512 * n);
964 } else if (cluster_offset & QCOW_OFLAG_COMPRESSED) {
965 if (decompress_cluster(s, cluster_offset) < 0)
967 memcpy(buf, s->cluster_cache + index_in_cluster * 512, 512 * n);
969 ret = bdrv_pread(s->hd, cluster_offset + index_in_cluster * 512, buf, n * 512);
972 if (s->crypt_method) {
973 encrypt_sectors(s, sector_num, buf, buf, n, 0,
974 &s->aes_decrypt_key);
984 static int qcow_write(BlockDriverState *bs, int64_t sector_num,
985 const uint8_t *buf, int nb_sectors)
987 BDRVQcowState *s = bs->opaque;
988 int ret, index_in_cluster, n;
989 uint64_t cluster_offset;
991 while (nb_sectors > 0) {
992 index_in_cluster = sector_num & (s->cluster_sectors - 1);
993 n = s->cluster_sectors - index_in_cluster;
996 cluster_offset = alloc_cluster_offset(bs, sector_num << 9,
998 index_in_cluster + n);
1001 if (s->crypt_method) {
1002 encrypt_sectors(s, sector_num, s->cluster_data, buf, n, 1,
1003 &s->aes_encrypt_key);
1004 ret = bdrv_pwrite(s->hd, cluster_offset + index_in_cluster * 512,
1005 s->cluster_data, n * 512);
1007 ret = bdrv_pwrite(s->hd, cluster_offset + index_in_cluster * 512, buf, n * 512);
1015 s->cluster_cache_offset = -1; /* disable compressed cache */
1019 typedef struct QCowAIOCB {
1020 BlockDriverAIOCB common;
1025 uint64_t cluster_offset;
1026 uint8_t *cluster_data;
1027 BlockDriverAIOCB *hd_aiocb;
1030 static void qcow_aio_read_cb(void *opaque, int ret)
1032 QCowAIOCB *acb = opaque;
1033 BlockDriverState *bs = acb->common.bs;
1034 BDRVQcowState *s = bs->opaque;
1035 int index_in_cluster, n1;
1037 acb->hd_aiocb = NULL;
1040 acb->common.cb(acb->common.opaque, ret);
1041 qemu_aio_release(acb);
1046 /* post process the read buffer */
1047 if (!acb->cluster_offset) {
1049 } else if (acb->cluster_offset & QCOW_OFLAG_COMPRESSED) {
1052 if (s->crypt_method) {
1053 encrypt_sectors(s, acb->sector_num, acb->buf, acb->buf,
1055 &s->aes_decrypt_key);
1059 acb->nb_sectors -= acb->n;
1060 acb->sector_num += acb->n;
1061 acb->buf += acb->n * 512;
1063 if (acb->nb_sectors == 0) {
1064 /* request completed */
1065 acb->common.cb(acb->common.opaque, 0);
1066 qemu_aio_release(acb);
1070 /* prepare next AIO request */
1071 acb->cluster_offset = get_cluster_offset(bs, acb->sector_num << 9);
1072 index_in_cluster = acb->sector_num & (s->cluster_sectors - 1);
1073 acb->n = s->cluster_sectors - index_in_cluster;
1074 if (acb->n > acb->nb_sectors)
1075 acb->n = acb->nb_sectors;
1077 if (!acb->cluster_offset) {
1078 if (bs->backing_hd) {
1079 /* read from the base image */
1080 n1 = backing_read1(bs->backing_hd, acb->sector_num,
1083 acb->hd_aiocb = bdrv_aio_read(bs->backing_hd, acb->sector_num,
1084 acb->buf, acb->n, qcow_aio_read_cb, acb);
1085 if (acb->hd_aiocb == NULL)
1091 /* Note: in this case, no need to wait */
1092 memset(acb->buf, 0, 512 * acb->n);
1095 } else if (acb->cluster_offset & QCOW_OFLAG_COMPRESSED) {
1096 /* add AIO support for compressed blocks ? */
1097 if (decompress_cluster(s, acb->cluster_offset) < 0)
1100 s->cluster_cache + index_in_cluster * 512, 512 * acb->n);
1103 if ((acb->cluster_offset & 511) != 0) {
1107 acb->hd_aiocb = bdrv_aio_read(s->hd,
1108 (acb->cluster_offset >> 9) + index_in_cluster,
1109 acb->buf, acb->n, qcow_aio_read_cb, acb);
1110 if (acb->hd_aiocb == NULL)
1115 static QCowAIOCB *qcow_aio_setup(BlockDriverState *bs,
1116 int64_t sector_num, uint8_t *buf, int nb_sectors,
1117 BlockDriverCompletionFunc *cb, void *opaque)
1121 acb = qemu_aio_get(bs, cb, opaque);
1124 acb->hd_aiocb = NULL;
1125 acb->sector_num = sector_num;
1127 acb->nb_sectors = nb_sectors;
1129 acb->cluster_offset = 0;
1133 static BlockDriverAIOCB *qcow_aio_read(BlockDriverState *bs,
1134 int64_t sector_num, uint8_t *buf, int nb_sectors,
1135 BlockDriverCompletionFunc *cb, void *opaque)
1139 acb = qcow_aio_setup(bs, sector_num, buf, nb_sectors, cb, opaque);
1143 qcow_aio_read_cb(acb, 0);
1144 return &acb->common;
1147 static void qcow_aio_write_cb(void *opaque, int ret)
1149 QCowAIOCB *acb = opaque;
1150 BlockDriverState *bs = acb->common.bs;
1151 BDRVQcowState *s = bs->opaque;
1152 int index_in_cluster;
1153 uint64_t cluster_offset;
1154 const uint8_t *src_buf;
1156 acb->hd_aiocb = NULL;
1160 acb->common.cb(acb->common.opaque, ret);
1161 qemu_aio_release(acb);
1165 acb->nb_sectors -= acb->n;
1166 acb->sector_num += acb->n;
1167 acb->buf += acb->n * 512;
1169 if (acb->nb_sectors == 0) {
1170 /* request completed */
1171 acb->common.cb(acb->common.opaque, 0);
1172 qemu_aio_release(acb);
1176 index_in_cluster = acb->sector_num & (s->cluster_sectors - 1);
1177 acb->n = s->cluster_sectors - index_in_cluster;
1178 if (acb->n > acb->nb_sectors)
1179 acb->n = acb->nb_sectors;
1180 cluster_offset = alloc_cluster_offset(bs, acb->sector_num << 9,
1182 index_in_cluster + acb->n);
1183 if (!cluster_offset || (cluster_offset & 511) != 0) {
1187 if (s->crypt_method) {
1188 if (!acb->cluster_data) {
1189 acb->cluster_data = qemu_mallocz(s->cluster_size);
1190 if (!acb->cluster_data) {
1195 encrypt_sectors(s, acb->sector_num, acb->cluster_data, acb->buf,
1196 acb->n, 1, &s->aes_encrypt_key);
1197 src_buf = acb->cluster_data;
1201 acb->hd_aiocb = bdrv_aio_write(s->hd,
1202 (cluster_offset >> 9) + index_in_cluster,
1204 qcow_aio_write_cb, acb);
1205 if (acb->hd_aiocb == NULL)
1209 static BlockDriverAIOCB *qcow_aio_write(BlockDriverState *bs,
1210 int64_t sector_num, const uint8_t *buf, int nb_sectors,
1211 BlockDriverCompletionFunc *cb, void *opaque)
1213 BDRVQcowState *s = bs->opaque;
1216 s->cluster_cache_offset = -1; /* disable compressed cache */
1218 acb = qcow_aio_setup(bs, sector_num, (uint8_t*)buf, nb_sectors, cb, opaque);
1222 qcow_aio_write_cb(acb, 0);
1223 return &acb->common;
1226 static void qcow_aio_cancel(BlockDriverAIOCB *blockacb)
1228 QCowAIOCB *acb = (QCowAIOCB *)blockacb;
1230 bdrv_aio_cancel(acb->hd_aiocb);
1231 qemu_aio_release(acb);
1234 static void qcow_close(BlockDriverState *bs)
1236 BDRVQcowState *s = bs->opaque;
1237 qemu_free(s->l1_table);
1238 qemu_free(s->l2_cache);
1239 qemu_free(s->cluster_cache);
1240 qemu_free(s->cluster_data);
1245 /* XXX: use std qcow open function ? */
1246 typedef struct QCowCreateState {
1249 uint16_t *refcount_block;
1250 uint64_t *refcount_table;
1251 int64_t l1_table_offset;
1252 int64_t refcount_table_offset;
1253 int64_t refcount_block_offset;
1256 static void create_refcount_update(QCowCreateState *s,
1257 int64_t offset, int64_t size)
1260 int64_t start, last, cluster_offset;
1263 start = offset & ~(s->cluster_size - 1);
1264 last = (offset + size - 1) & ~(s->cluster_size - 1);
1265 for(cluster_offset = start; cluster_offset <= last;
1266 cluster_offset += s->cluster_size) {
1267 p = &s->refcount_block[cluster_offset >> s->cluster_bits];
1268 refcount = be16_to_cpu(*p);
1270 *p = cpu_to_be16(refcount);
1274 static int qcow_create(const char *filename, int64_t total_size,
1275 const char *backing_file, int flags)
1277 int fd, header_size, backing_filename_len, l1_size, i, shift, l2_bits;
1279 uint64_t tmp, offset;
1280 QCowCreateState s1, *s = &s1;
1282 memset(s, 0, sizeof(*s));
1284 fd = open(filename, O_WRONLY | O_CREAT | O_TRUNC | O_BINARY, 0644);
1287 memset(&header, 0, sizeof(header));
1288 header.magic = cpu_to_be32(QCOW_MAGIC);
1289 header.version = cpu_to_be32(QCOW_VERSION);
1290 header.size = cpu_to_be64(total_size * 512);
1291 header_size = sizeof(header);
1292 backing_filename_len = 0;
1294 header.backing_file_offset = cpu_to_be64(header_size);
1295 backing_filename_len = strlen(backing_file);
1296 header.backing_file_size = cpu_to_be32(backing_filename_len);
1297 header_size += backing_filename_len;
1299 s->cluster_bits = 12; /* 4 KB clusters */
1300 s->cluster_size = 1 << s->cluster_bits;
1301 header.cluster_bits = cpu_to_be32(s->cluster_bits);
1302 header_size = (header_size + 7) & ~7;
1303 if (flags & BLOCK_FLAG_ENCRYPT) {
1304 header.crypt_method = cpu_to_be32(QCOW_CRYPT_AES);
1306 header.crypt_method = cpu_to_be32(QCOW_CRYPT_NONE);
1308 l2_bits = s->cluster_bits - 3;
1309 shift = s->cluster_bits + l2_bits;
1310 l1_size = (((total_size * 512) + (1LL << shift) - 1) >> shift);
1311 offset = align_offset(header_size, s->cluster_size);
1312 s->l1_table_offset = offset;
1313 header.l1_table_offset = cpu_to_be64(s->l1_table_offset);
1314 header.l1_size = cpu_to_be32(l1_size);
1315 offset += align_offset(l1_size * sizeof(uint64_t), s->cluster_size);
1317 s->refcount_table = qemu_mallocz(s->cluster_size);
1318 if (!s->refcount_table)
1320 s->refcount_block = qemu_mallocz(s->cluster_size);
1321 if (!s->refcount_block)
1324 s->refcount_table_offset = offset;
1325 header.refcount_table_offset = cpu_to_be64(offset);
1326 header.refcount_table_clusters = cpu_to_be32(1);
1327 offset += s->cluster_size;
1329 s->refcount_table[0] = cpu_to_be64(offset);
1330 s->refcount_block_offset = offset;
1331 offset += s->cluster_size;
1333 /* update refcounts */
1334 create_refcount_update(s, 0, header_size);
1335 create_refcount_update(s, s->l1_table_offset, l1_size * sizeof(uint64_t));
1336 create_refcount_update(s, s->refcount_table_offset, s->cluster_size);
1337 create_refcount_update(s, s->refcount_block_offset, s->cluster_size);
1339 /* write all the data */
1340 write(fd, &header, sizeof(header));
1342 write(fd, backing_file, backing_filename_len);
1344 lseek(fd, s->l1_table_offset, SEEK_SET);
1346 for(i = 0;i < l1_size; i++) {
1347 write(fd, &tmp, sizeof(tmp));
1349 lseek(fd, s->refcount_table_offset, SEEK_SET);
1350 write(fd, s->refcount_table, s->cluster_size);
1352 lseek(fd, s->refcount_block_offset, SEEK_SET);
1353 write(fd, s->refcount_block, s->cluster_size);
1355 qemu_free(s->refcount_table);
1356 qemu_free(s->refcount_block);
1360 qemu_free(s->refcount_table);
1361 qemu_free(s->refcount_block);
1366 static int qcow_make_empty(BlockDriverState *bs)
1369 /* XXX: not correct */
1370 BDRVQcowState *s = bs->opaque;
1371 uint32_t l1_length = s->l1_size * sizeof(uint64_t);
1374 memset(s->l1_table, 0, l1_length);
1375 if (bdrv_pwrite(s->hd, s->l1_table_offset, s->l1_table, l1_length) < 0)
1377 ret = bdrv_truncate(s->hd, s->l1_table_offset + l1_length);
1386 /* XXX: put compressed sectors first, then all the cluster aligned
1387 tables to avoid losing bytes in alignment */
1388 static int qcow_write_compressed(BlockDriverState *bs, int64_t sector_num,
1389 const uint8_t *buf, int nb_sectors)
1391 BDRVQcowState *s = bs->opaque;
1395 uint64_t cluster_offset;
1397 if (nb_sectors == 0) {
1398 /* align end of file to a sector boundary to ease reading with
1399 sector based I/Os */
1400 cluster_offset = bdrv_getlength(s->hd);
1401 cluster_offset = (cluster_offset + 511) & ~511;
1402 bdrv_truncate(s->hd, cluster_offset);
1406 if (nb_sectors != s->cluster_sectors)
1409 out_buf = qemu_malloc(s->cluster_size + (s->cluster_size / 1000) + 128);
1413 /* best compression, small window, no zlib header */
1414 memset(&strm, 0, sizeof(strm));
1415 ret = deflateInit2(&strm, Z_DEFAULT_COMPRESSION,
1417 9, Z_DEFAULT_STRATEGY);
1423 strm.avail_in = s->cluster_size;
1424 strm.next_in = (uint8_t *)buf;
1425 strm.avail_out = s->cluster_size;
1426 strm.next_out = out_buf;
1428 ret = deflate(&strm, Z_FINISH);
1429 if (ret != Z_STREAM_END && ret != Z_OK) {
1434 out_len = strm.next_out - out_buf;
1438 if (ret != Z_STREAM_END || out_len >= s->cluster_size) {
1439 /* could not compress: write normal cluster */
1440 qcow_write(bs, sector_num, buf, s->cluster_sectors);
1442 cluster_offset = alloc_compressed_cluster_offset(bs, sector_num << 9,
1444 if (!cluster_offset)
1446 cluster_offset &= s->cluster_offset_mask;
1447 if (bdrv_pwrite(s->hd, cluster_offset, out_buf, out_len) != out_len) {
1457 static void qcow_flush(BlockDriverState *bs)
1459 BDRVQcowState *s = bs->opaque;
1463 static int qcow_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
1465 BDRVQcowState *s = bs->opaque;
1466 bdi->cluster_size = s->cluster_size;
1467 bdi->vm_state_offset = (int64_t)s->l1_vm_state_index <<
1468 (s->cluster_bits + s->l2_bits);
1472 /*********************************************************/
1473 /* snapshot support */
1475 /* update the refcounts of snapshots and the copied flag */
1476 static int update_snapshot_refcount(BlockDriverState *bs,
1477 int64_t l1_table_offset,
1481 BDRVQcowState *s = bs->opaque;
1482 uint64_t *l1_table, *l2_table, l2_offset, offset, l1_size2, l1_allocated;
1483 int64_t old_offset, old_l2_offset;
1484 int l2_size, i, j, l1_modified, l2_modified, nb_csectors, refcount;
1490 l1_size2 = l1_size * sizeof(uint64_t);
1492 if (l1_table_offset != s->l1_table_offset) {
1493 l1_table = qemu_malloc(l1_size2);
1497 if (bdrv_pread(s->hd, l1_table_offset,
1498 l1_table, l1_size2) != l1_size2)
1500 for(i = 0;i < l1_size; i++)
1501 be64_to_cpus(&l1_table[i]);
1503 assert(l1_size == s->l1_size);
1504 l1_table = s->l1_table;
1508 l2_size = s->l2_size * sizeof(uint64_t);
1509 l2_table = qemu_malloc(l2_size);
1513 for(i = 0; i < l1_size; i++) {
1514 l2_offset = l1_table[i];
1516 old_l2_offset = l2_offset;
1517 l2_offset &= ~QCOW_OFLAG_COPIED;
1519 if (bdrv_pread(s->hd, l2_offset, l2_table, l2_size) != l2_size)
1521 for(j = 0; j < s->l2_size; j++) {
1522 offset = be64_to_cpu(l2_table[j]);
1524 old_offset = offset;
1525 offset &= ~QCOW_OFLAG_COPIED;
1526 if (offset & QCOW_OFLAG_COMPRESSED) {
1527 nb_csectors = ((offset >> s->csize_shift) &
1530 update_refcount(bs, (offset & s->cluster_offset_mask) & ~511,
1531 nb_csectors * 512, addend);
1532 /* compressed clusters are never modified */
1536 refcount = update_cluster_refcount(bs, offset >> s->cluster_bits, addend);
1538 refcount = get_refcount(bs, offset >> s->cluster_bits);
1542 if (refcount == 1) {
1543 offset |= QCOW_OFLAG_COPIED;
1545 if (offset != old_offset) {
1546 l2_table[j] = cpu_to_be64(offset);
1552 if (bdrv_pwrite(s->hd,
1553 l2_offset, l2_table, l2_size) != l2_size)
1558 refcount = update_cluster_refcount(bs, l2_offset >> s->cluster_bits, addend);
1560 refcount = get_refcount(bs, l2_offset >> s->cluster_bits);
1562 if (refcount == 1) {
1563 l2_offset |= QCOW_OFLAG_COPIED;
1565 if (l2_offset != old_l2_offset) {
1566 l1_table[i] = l2_offset;
1572 for(i = 0; i < l1_size; i++)
1573 cpu_to_be64s(&l1_table[i]);
1574 if (bdrv_pwrite(s->hd, l1_table_offset, l1_table,
1575 l1_size2) != l1_size2)
1577 for(i = 0; i < l1_size; i++)
1578 be64_to_cpus(&l1_table[i]);
1581 qemu_free(l1_table);
1582 qemu_free(l2_table);
1586 qemu_free(l1_table);
1587 qemu_free(l2_table);
1591 static void qcow_free_snapshots(BlockDriverState *bs)
1593 BDRVQcowState *s = bs->opaque;
1596 for(i = 0; i < s->nb_snapshots; i++) {
1597 qemu_free(s->snapshots[i].name);
1598 qemu_free(s->snapshots[i].id_str);
1600 qemu_free(s->snapshots);
1601 s->snapshots = NULL;
1602 s->nb_snapshots = 0;
1605 static int qcow_read_snapshots(BlockDriverState *bs)
1607 BDRVQcowState *s = bs->opaque;
1608 QCowSnapshotHeader h;
1610 int i, id_str_size, name_size;
1612 uint32_t extra_data_size;
1614 offset = s->snapshots_offset;
1615 s->snapshots = qemu_mallocz(s->nb_snapshots * sizeof(QCowSnapshot));
1618 for(i = 0; i < s->nb_snapshots; i++) {
1619 offset = align_offset(offset, 8);
1620 if (bdrv_pread(s->hd, offset, &h, sizeof(h)) != sizeof(h))
1622 offset += sizeof(h);
1623 sn = s->snapshots + i;
1624 sn->l1_table_offset = be64_to_cpu(h.l1_table_offset);
1625 sn->l1_size = be32_to_cpu(h.l1_size);
1626 sn->vm_state_size = be32_to_cpu(h.vm_state_size);
1627 sn->date_sec = be32_to_cpu(h.date_sec);
1628 sn->date_nsec = be32_to_cpu(h.date_nsec);
1629 sn->vm_clock_nsec = be64_to_cpu(h.vm_clock_nsec);
1630 extra_data_size = be32_to_cpu(h.extra_data_size);
1632 id_str_size = be16_to_cpu(h.id_str_size);
1633 name_size = be16_to_cpu(h.name_size);
1635 offset += extra_data_size;
1637 sn->id_str = qemu_malloc(id_str_size + 1);
1640 if (bdrv_pread(s->hd, offset, sn->id_str, id_str_size) != id_str_size)
1642 offset += id_str_size;
1643 sn->id_str[id_str_size] = '\0';
1645 sn->name = qemu_malloc(name_size + 1);
1648 if (bdrv_pread(s->hd, offset, sn->name, name_size) != name_size)
1650 offset += name_size;
1651 sn->name[name_size] = '\0';
1653 s->snapshots_size = offset - s->snapshots_offset;
1656 qcow_free_snapshots(bs);
1660 /* add at the end of the file a new list of snapshots */
1661 static int qcow_write_snapshots(BlockDriverState *bs)
1663 BDRVQcowState *s = bs->opaque;
1665 QCowSnapshotHeader h;
1666 int i, name_size, id_str_size, snapshots_size;
1669 int64_t offset, snapshots_offset;
1671 /* compute the size of the snapshots */
1673 for(i = 0; i < s->nb_snapshots; i++) {
1674 sn = s->snapshots + i;
1675 offset = align_offset(offset, 8);
1676 offset += sizeof(h);
1677 offset += strlen(sn->id_str);
1678 offset += strlen(sn->name);
1680 snapshots_size = offset;
1682 snapshots_offset = alloc_clusters(bs, snapshots_size);
1683 offset = snapshots_offset;
1685 for(i = 0; i < s->nb_snapshots; i++) {
1686 sn = s->snapshots + i;
1687 memset(&h, 0, sizeof(h));
1688 h.l1_table_offset = cpu_to_be64(sn->l1_table_offset);
1689 h.l1_size = cpu_to_be32(sn->l1_size);
1690 h.vm_state_size = cpu_to_be32(sn->vm_state_size);
1691 h.date_sec = cpu_to_be32(sn->date_sec);
1692 h.date_nsec = cpu_to_be32(sn->date_nsec);
1693 h.vm_clock_nsec = cpu_to_be64(sn->vm_clock_nsec);
1695 id_str_size = strlen(sn->id_str);
1696 name_size = strlen(sn->name);
1697 h.id_str_size = cpu_to_be16(id_str_size);
1698 h.name_size = cpu_to_be16(name_size);
1699 offset = align_offset(offset, 8);
1700 if (bdrv_pwrite(s->hd, offset, &h, sizeof(h)) != sizeof(h))
1702 offset += sizeof(h);
1703 if (bdrv_pwrite(s->hd, offset, sn->id_str, id_str_size) != id_str_size)
1705 offset += id_str_size;
1706 if (bdrv_pwrite(s->hd, offset, sn->name, name_size) != name_size)
1708 offset += name_size;
1711 /* update the various header fields */
1712 data64 = cpu_to_be64(snapshots_offset);
1713 if (bdrv_pwrite(s->hd, offsetof(QCowHeader, snapshots_offset),
1714 &data64, sizeof(data64)) != sizeof(data64))
1716 data32 = cpu_to_be32(s->nb_snapshots);
1717 if (bdrv_pwrite(s->hd, offsetof(QCowHeader, nb_snapshots),
1718 &data32, sizeof(data32)) != sizeof(data32))
1721 /* free the old snapshot table */
1722 free_clusters(bs, s->snapshots_offset, s->snapshots_size);
1723 s->snapshots_offset = snapshots_offset;
1724 s->snapshots_size = snapshots_size;
1730 static void find_new_snapshot_id(BlockDriverState *bs,
1731 char *id_str, int id_str_size)
1733 BDRVQcowState *s = bs->opaque;
1735 int i, id, id_max = 0;
1737 for(i = 0; i < s->nb_snapshots; i++) {
1738 sn = s->snapshots + i;
1739 id = strtoul(sn->id_str, NULL, 10);
1743 snprintf(id_str, id_str_size, "%d", id_max + 1);
1746 static int find_snapshot_by_id(BlockDriverState *bs, const char *id_str)
1748 BDRVQcowState *s = bs->opaque;
1751 for(i = 0; i < s->nb_snapshots; i++) {
1752 if (!strcmp(s->snapshots[i].id_str, id_str))
1758 static int find_snapshot_by_id_or_name(BlockDriverState *bs, const char *name)
1760 BDRVQcowState *s = bs->opaque;
1763 ret = find_snapshot_by_id(bs, name);
1766 for(i = 0; i < s->nb_snapshots; i++) {
1767 if (!strcmp(s->snapshots[i].name, name))
1773 /* if no id is provided, a new one is constructed */
1774 static int qcow_snapshot_create(BlockDriverState *bs,
1775 QEMUSnapshotInfo *sn_info)
1777 BDRVQcowState *s = bs->opaque;
1778 QCowSnapshot *snapshots1, sn1, *sn = &sn1;
1780 uint64_t *l1_table = NULL;
1782 memset(sn, 0, sizeof(*sn));
1784 if (sn_info->id_str[0] == '\0') {
1785 /* compute a new id */
1786 find_new_snapshot_id(bs, sn_info->id_str, sizeof(sn_info->id_str));
1789 /* check that the ID is unique */
1790 if (find_snapshot_by_id(bs, sn_info->id_str) >= 0)
1793 sn->id_str = qemu_strdup(sn_info->id_str);
1796 sn->name = qemu_strdup(sn_info->name);
1799 sn->vm_state_size = sn_info->vm_state_size;
1800 sn->date_sec = sn_info->date_sec;
1801 sn->date_nsec = sn_info->date_nsec;
1802 sn->vm_clock_nsec = sn_info->vm_clock_nsec;
1804 ret = update_snapshot_refcount(bs, s->l1_table_offset, s->l1_size, 1);
1808 /* create the L1 table of the snapshot */
1809 sn->l1_table_offset = alloc_clusters(bs, s->l1_size * sizeof(uint64_t));
1810 sn->l1_size = s->l1_size;
1812 l1_table = qemu_malloc(s->l1_size * sizeof(uint64_t));
1815 for(i = 0; i < s->l1_size; i++) {
1816 l1_table[i] = cpu_to_be64(s->l1_table[i]);
1818 if (bdrv_pwrite(s->hd, sn->l1_table_offset,
1819 l1_table, s->l1_size * sizeof(uint64_t)) !=
1820 (s->l1_size * sizeof(uint64_t)))
1822 qemu_free(l1_table);
1825 snapshots1 = qemu_malloc((s->nb_snapshots + 1) * sizeof(QCowSnapshot));
1828 memcpy(snapshots1, s->snapshots, s->nb_snapshots * sizeof(QCowSnapshot));
1829 s->snapshots = snapshots1;
1830 s->snapshots[s->nb_snapshots++] = *sn;
1832 if (qcow_write_snapshots(bs) < 0)
1835 check_refcounts(bs);
1839 qemu_free(sn->name);
1840 qemu_free(l1_table);
1844 /* copy the snapshot 'snapshot_name' into the current disk image */
1845 static int qcow_snapshot_goto(BlockDriverState *bs,
1846 const char *snapshot_id)
1848 BDRVQcowState *s = bs->opaque;
1850 int i, snapshot_index, l1_size2;
1852 snapshot_index = find_snapshot_by_id_or_name(bs, snapshot_id);
1853 if (snapshot_index < 0)
1855 sn = &s->snapshots[snapshot_index];
1857 if (update_snapshot_refcount(bs, s->l1_table_offset, s->l1_size, -1) < 0)
1860 if (grow_l1_table(bs, sn->l1_size) < 0)
1863 s->l1_size = sn->l1_size;
1864 l1_size2 = s->l1_size * sizeof(uint64_t);
1865 /* copy the snapshot l1 table to the current l1 table */
1866 if (bdrv_pread(s->hd, sn->l1_table_offset,
1867 s->l1_table, l1_size2) != l1_size2)
1869 if (bdrv_pwrite(s->hd, s->l1_table_offset,
1870 s->l1_table, l1_size2) != l1_size2)
1872 for(i = 0;i < s->l1_size; i++) {
1873 be64_to_cpus(&s->l1_table[i]);
1876 if (update_snapshot_refcount(bs, s->l1_table_offset, s->l1_size, 1) < 0)
1880 check_refcounts(bs);
1887 static int qcow_snapshot_delete(BlockDriverState *bs, const char *snapshot_id)
1889 BDRVQcowState *s = bs->opaque;
1891 int snapshot_index, ret;
1893 snapshot_index = find_snapshot_by_id_or_name(bs, snapshot_id);
1894 if (snapshot_index < 0)
1896 sn = &s->snapshots[snapshot_index];
1898 ret = update_snapshot_refcount(bs, sn->l1_table_offset, sn->l1_size, -1);
1901 /* must update the copied flag on the current cluster offsets */
1902 ret = update_snapshot_refcount(bs, s->l1_table_offset, s->l1_size, 0);
1905 free_clusters(bs, sn->l1_table_offset, sn->l1_size * sizeof(uint64_t));
1907 qemu_free(sn->id_str);
1908 qemu_free(sn->name);
1909 memmove(sn, sn + 1, (s->nb_snapshots - snapshot_index - 1) * sizeof(*sn));
1911 ret = qcow_write_snapshots(bs);
1913 /* XXX: restore snapshot if error ? */
1917 check_refcounts(bs);
1922 static int qcow_snapshot_list(BlockDriverState *bs,
1923 QEMUSnapshotInfo **psn_tab)
1925 BDRVQcowState *s = bs->opaque;
1926 QEMUSnapshotInfo *sn_tab, *sn_info;
1930 sn_tab = qemu_mallocz(s->nb_snapshots * sizeof(QEMUSnapshotInfo));
1933 for(i = 0; i < s->nb_snapshots; i++) {
1934 sn_info = sn_tab + i;
1935 sn = s->snapshots + i;
1936 pstrcpy(sn_info->id_str, sizeof(sn_info->id_str),
1938 pstrcpy(sn_info->name, sizeof(sn_info->name),
1940 sn_info->vm_state_size = sn->vm_state_size;
1941 sn_info->date_sec = sn->date_sec;
1942 sn_info->date_nsec = sn->date_nsec;
1943 sn_info->vm_clock_nsec = sn->vm_clock_nsec;
1946 return s->nb_snapshots;
1953 /*********************************************************/
1954 /* refcount handling */
1956 static int refcount_init(BlockDriverState *bs)
1958 BDRVQcowState *s = bs->opaque;
1959 int ret, refcount_table_size2, i;
1961 s->refcount_block_cache = qemu_malloc(s->cluster_size);
1962 if (!s->refcount_block_cache)
1964 refcount_table_size2 = s->refcount_table_size * sizeof(uint64_t);
1965 s->refcount_table = qemu_malloc(refcount_table_size2);
1966 if (!s->refcount_table)
1968 if (s->refcount_table_size > 0) {
1969 ret = bdrv_pread(s->hd, s->refcount_table_offset,
1970 s->refcount_table, refcount_table_size2);
1971 if (ret != refcount_table_size2)
1973 for(i = 0; i < s->refcount_table_size; i++)
1974 be64_to_cpus(&s->refcount_table[i]);
1981 static void refcount_close(BlockDriverState *bs)
1983 BDRVQcowState *s = bs->opaque;
1984 qemu_free(s->refcount_block_cache);
1985 qemu_free(s->refcount_table);
1989 static int load_refcount_block(BlockDriverState *bs,
1990 int64_t refcount_block_offset)
1992 BDRVQcowState *s = bs->opaque;
1994 ret = bdrv_pread(s->hd, refcount_block_offset, s->refcount_block_cache,
1996 if (ret != s->cluster_size)
1998 s->refcount_block_cache_offset = refcount_block_offset;
2002 static int get_refcount(BlockDriverState *bs, int64_t cluster_index)
2004 BDRVQcowState *s = bs->opaque;
2005 int refcount_table_index, block_index;
2006 int64_t refcount_block_offset;
2008 refcount_table_index = cluster_index >> (s->cluster_bits - REFCOUNT_SHIFT);
2009 if (refcount_table_index >= s->refcount_table_size)
2011 refcount_block_offset = s->refcount_table[refcount_table_index];
2012 if (!refcount_block_offset)
2014 if (refcount_block_offset != s->refcount_block_cache_offset) {
2015 /* better than nothing: return allocated if read error */
2016 if (load_refcount_block(bs, refcount_block_offset) < 0)
2019 block_index = cluster_index &
2020 ((1 << (s->cluster_bits - REFCOUNT_SHIFT)) - 1);
2021 return be16_to_cpu(s->refcount_block_cache[block_index]);
2024 /* return < 0 if error */
2025 static int64_t alloc_clusters_noref(BlockDriverState *bs, int64_t size)
2027 BDRVQcowState *s = bs->opaque;
2030 nb_clusters = (size + s->cluster_size - 1) >> s->cluster_bits;
2032 if (get_refcount(bs, s->free_cluster_index) == 0) {
2033 s->free_cluster_index++;
2034 for(i = 1; i < nb_clusters; i++) {
2035 if (get_refcount(bs, s->free_cluster_index) != 0)
2037 s->free_cluster_index++;
2040 printf("alloc_clusters: size=%lld -> %lld\n",
2042 (s->free_cluster_index - nb_clusters) << s->cluster_bits);
2044 return (s->free_cluster_index - nb_clusters) << s->cluster_bits;
2047 s->free_cluster_index++;
2052 static int64_t alloc_clusters(BlockDriverState *bs, int64_t size)
2056 offset = alloc_clusters_noref(bs, size);
2057 update_refcount(bs, offset, size, 1);
2061 /* only used to allocate compressed sectors. We try to allocate
2062 contiguous sectors. size must be <= cluster_size */
2063 static int64_t alloc_bytes(BlockDriverState *bs, int size)
2065 BDRVQcowState *s = bs->opaque;
2066 int64_t offset, cluster_offset;
2067 int free_in_cluster;
2069 assert(size > 0 && size <= s->cluster_size);
2070 if (s->free_byte_offset == 0) {
2071 s->free_byte_offset = alloc_clusters(bs, s->cluster_size);
2074 free_in_cluster = s->cluster_size -
2075 (s->free_byte_offset & (s->cluster_size - 1));
2076 if (size <= free_in_cluster) {
2077 /* enough space in current cluster */
2078 offset = s->free_byte_offset;
2079 s->free_byte_offset += size;
2080 free_in_cluster -= size;
2081 if (free_in_cluster == 0)
2082 s->free_byte_offset = 0;
2083 if ((offset & (s->cluster_size - 1)) != 0)
2084 update_cluster_refcount(bs, offset >> s->cluster_bits, 1);
2086 offset = alloc_clusters(bs, s->cluster_size);
2087 cluster_offset = s->free_byte_offset & ~(s->cluster_size - 1);
2088 if ((cluster_offset + s->cluster_size) == offset) {
2089 /* we are lucky: contiguous data */
2090 offset = s->free_byte_offset;
2091 update_cluster_refcount(bs, offset >> s->cluster_bits, 1);
2092 s->free_byte_offset += size;
2094 s->free_byte_offset = offset;
2101 static void free_clusters(BlockDriverState *bs,
2102 int64_t offset, int64_t size)
2104 update_refcount(bs, offset, size, -1);
2107 static int grow_refcount_table(BlockDriverState *bs, int min_size)
2109 BDRVQcowState *s = bs->opaque;
2110 int new_table_size, new_table_size2, refcount_table_clusters, i, ret;
2111 uint64_t *new_table;
2112 int64_t table_offset;
2116 int64_t old_table_offset;
2118 if (min_size <= s->refcount_table_size)
2120 /* compute new table size */
2121 refcount_table_clusters = s->refcount_table_size >> (s->cluster_bits - 3);
2123 if (refcount_table_clusters == 0) {
2124 refcount_table_clusters = 1;
2126 refcount_table_clusters = (refcount_table_clusters * 3 + 1) / 2;
2128 new_table_size = refcount_table_clusters << (s->cluster_bits - 3);
2129 if (min_size <= new_table_size)
2133 printf("grow_refcount_table from %d to %d\n",
2134 s->refcount_table_size,
2137 new_table_size2 = new_table_size * sizeof(uint64_t);
2138 new_table = qemu_mallocz(new_table_size2);
2141 memcpy(new_table, s->refcount_table,
2142 s->refcount_table_size * sizeof(uint64_t));
2143 for(i = 0; i < s->refcount_table_size; i++)
2144 cpu_to_be64s(&new_table[i]);
2145 /* Note: we cannot update the refcount now to avoid recursion */
2146 table_offset = alloc_clusters_noref(bs, new_table_size2);
2147 ret = bdrv_pwrite(s->hd, table_offset, new_table, new_table_size2);
2148 if (ret != new_table_size2)
2150 for(i = 0; i < s->refcount_table_size; i++)
2151 be64_to_cpus(&new_table[i]);
2153 data64 = cpu_to_be64(table_offset);
2154 if (bdrv_pwrite(s->hd, offsetof(QCowHeader, refcount_table_offset),
2155 &data64, sizeof(data64)) != sizeof(data64))
2157 data32 = cpu_to_be32(refcount_table_clusters);
2158 if (bdrv_pwrite(s->hd, offsetof(QCowHeader, refcount_table_clusters),
2159 &data32, sizeof(data32)) != sizeof(data32))
2161 qemu_free(s->refcount_table);
2162 old_table_offset = s->refcount_table_offset;
2163 old_table_size = s->refcount_table_size;
2164 s->refcount_table = new_table;
2165 s->refcount_table_size = new_table_size;
2166 s->refcount_table_offset = table_offset;
2168 update_refcount(bs, table_offset, new_table_size2, 1);
2169 free_clusters(bs, old_table_offset, old_table_size * sizeof(uint64_t));
2172 free_clusters(bs, table_offset, new_table_size2);
2173 qemu_free(new_table);
2177 /* addend must be 1 or -1 */
2178 /* XXX: cache several refcount block clusters ? */
2179 static int update_cluster_refcount(BlockDriverState *bs,
2180 int64_t cluster_index,
2183 BDRVQcowState *s = bs->opaque;
2184 int64_t offset, refcount_block_offset;
2185 int ret, refcount_table_index, block_index, refcount;
2188 refcount_table_index = cluster_index >> (s->cluster_bits - REFCOUNT_SHIFT);
2189 if (refcount_table_index >= s->refcount_table_size) {
2192 ret = grow_refcount_table(bs, refcount_table_index + 1);
2196 refcount_block_offset = s->refcount_table[refcount_table_index];
2197 if (!refcount_block_offset) {
2200 /* create a new refcount block */
2201 /* Note: we cannot update the refcount now to avoid recursion */
2202 offset = alloc_clusters_noref(bs, s->cluster_size);
2203 memset(s->refcount_block_cache, 0, s->cluster_size);
2204 ret = bdrv_pwrite(s->hd, offset, s->refcount_block_cache, s->cluster_size);
2205 if (ret != s->cluster_size)
2207 s->refcount_table[refcount_table_index] = offset;
2208 data64 = cpu_to_be64(offset);
2209 ret = bdrv_pwrite(s->hd, s->refcount_table_offset +
2210 refcount_table_index * sizeof(uint64_t),
2211 &data64, sizeof(data64));
2212 if (ret != sizeof(data64))
2215 refcount_block_offset = offset;
2216 s->refcount_block_cache_offset = offset;
2217 update_refcount(bs, offset, s->cluster_size, 1);
2219 if (refcount_block_offset != s->refcount_block_cache_offset) {
2220 if (load_refcount_block(bs, refcount_block_offset) < 0)
2224 /* we can update the count and save it */
2225 block_index = cluster_index &
2226 ((1 << (s->cluster_bits - REFCOUNT_SHIFT)) - 1);
2227 refcount = be16_to_cpu(s->refcount_block_cache[block_index]);
2229 if (refcount < 0 || refcount > 0xffff)
2231 if (refcount == 0 && cluster_index < s->free_cluster_index) {
2232 s->free_cluster_index = cluster_index;
2234 s->refcount_block_cache[block_index] = cpu_to_be16(refcount);
2235 if (bdrv_pwrite(s->hd,
2236 refcount_block_offset + (block_index << REFCOUNT_SHIFT),
2237 &s->refcount_block_cache[block_index], 2) != 2)
2242 static void update_refcount(BlockDriverState *bs,
2243 int64_t offset, int64_t length,
2246 BDRVQcowState *s = bs->opaque;
2247 int64_t start, last, cluster_offset;
2250 printf("update_refcount: offset=%lld size=%lld addend=%d\n",
2251 offset, length, addend);
2255 start = offset & ~(s->cluster_size - 1);
2256 last = (offset + length - 1) & ~(s->cluster_size - 1);
2257 for(cluster_offset = start; cluster_offset <= last;
2258 cluster_offset += s->cluster_size) {
2259 update_cluster_refcount(bs, cluster_offset >> s->cluster_bits, addend);
2264 static void inc_refcounts(BlockDriverState *bs,
2265 uint16_t *refcount_table,
2266 int refcount_table_size,
2267 int64_t offset, int64_t size)
2269 BDRVQcowState *s = bs->opaque;
2270 int64_t start, last, cluster_offset;
2276 start = offset & ~(s->cluster_size - 1);
2277 last = (offset + size - 1) & ~(s->cluster_size - 1);
2278 for(cluster_offset = start; cluster_offset <= last;
2279 cluster_offset += s->cluster_size) {
2280 k = cluster_offset >> s->cluster_bits;
2281 if (k < 0 || k >= refcount_table_size) {
2282 printf("ERROR: invalid cluster offset=0x%llx\n", cluster_offset);
2284 if (++refcount_table[k] == 0) {
2285 printf("ERROR: overflow cluster offset=0x%llx\n", cluster_offset);
2291 static int check_refcounts_l1(BlockDriverState *bs,
2292 uint16_t *refcount_table,
2293 int refcount_table_size,
2294 int64_t l1_table_offset, int l1_size,
2297 BDRVQcowState *s = bs->opaque;
2298 uint64_t *l1_table, *l2_table, l2_offset, offset, l1_size2;
2299 int l2_size, i, j, nb_csectors, refcount;
2302 l1_size2 = l1_size * sizeof(uint64_t);
2304 inc_refcounts(bs, refcount_table, refcount_table_size,
2305 l1_table_offset, l1_size2);
2307 l1_table = qemu_malloc(l1_size2);
2310 if (bdrv_pread(s->hd, l1_table_offset,
2311 l1_table, l1_size2) != l1_size2)
2313 for(i = 0;i < l1_size; i++)
2314 be64_to_cpus(&l1_table[i]);
2316 l2_size = s->l2_size * sizeof(uint64_t);
2317 l2_table = qemu_malloc(l2_size);
2320 for(i = 0; i < l1_size; i++) {
2321 l2_offset = l1_table[i];
2324 refcount = get_refcount(bs, (l2_offset & ~QCOW_OFLAG_COPIED) >> s->cluster_bits);
2325 if ((refcount == 1) != ((l2_offset & QCOW_OFLAG_COPIED) != 0)) {
2326 printf("ERROR OFLAG_COPIED: l2_offset=%llx refcount=%d\n",
2327 l2_offset, refcount);
2330 l2_offset &= ~QCOW_OFLAG_COPIED;
2331 if (bdrv_pread(s->hd, l2_offset, l2_table, l2_size) != l2_size)
2333 for(j = 0; j < s->l2_size; j++) {
2334 offset = be64_to_cpu(l2_table[j]);
2336 if (offset & QCOW_OFLAG_COMPRESSED) {
2337 if (offset & QCOW_OFLAG_COPIED) {
2338 printf("ERROR: cluster %lld: copied flag must never be set for compressed clusters\n",
2339 offset >> s->cluster_bits);
2340 offset &= ~QCOW_OFLAG_COPIED;
2342 nb_csectors = ((offset >> s->csize_shift) &
2344 offset &= s->cluster_offset_mask;
2345 inc_refcounts(bs, refcount_table,
2346 refcount_table_size,
2347 offset & ~511, nb_csectors * 512);
2350 refcount = get_refcount(bs, (offset & ~QCOW_OFLAG_COPIED) >> s->cluster_bits);
2351 if ((refcount == 1) != ((offset & QCOW_OFLAG_COPIED) != 0)) {
2352 printf("ERROR OFLAG_COPIED: offset=%llx refcount=%d\n",
2356 offset &= ~QCOW_OFLAG_COPIED;
2357 inc_refcounts(bs, refcount_table,
2358 refcount_table_size,
2359 offset, s->cluster_size);
2363 inc_refcounts(bs, refcount_table,
2364 refcount_table_size,
2369 qemu_free(l1_table);
2370 qemu_free(l2_table);
2373 printf("ERROR: I/O error in check_refcounts_l1\n");
2374 qemu_free(l1_table);
2375 qemu_free(l2_table);
2379 static void check_refcounts(BlockDriverState *bs)
2381 BDRVQcowState *s = bs->opaque;
2383 int nb_clusters, refcount1, refcount2, i;
2385 uint16_t *refcount_table;
2387 size = bdrv_getlength(s->hd);
2388 nb_clusters = (size + s->cluster_size - 1) >> s->cluster_bits;
2389 refcount_table = qemu_mallocz(nb_clusters * sizeof(uint16_t));
2392 inc_refcounts(bs, refcount_table, nb_clusters,
2393 0, s->cluster_size);
2395 check_refcounts_l1(bs, refcount_table, nb_clusters,
2396 s->l1_table_offset, s->l1_size, 1);
2399 for(i = 0; i < s->nb_snapshots; i++) {
2400 sn = s->snapshots + i;
2401 check_refcounts_l1(bs, refcount_table, nb_clusters,
2402 sn->l1_table_offset, sn->l1_size, 0);
2404 inc_refcounts(bs, refcount_table, nb_clusters,
2405 s->snapshots_offset, s->snapshots_size);
2408 inc_refcounts(bs, refcount_table, nb_clusters,
2409 s->refcount_table_offset,
2410 s->refcount_table_size * sizeof(uint64_t));
2411 for(i = 0; i < s->refcount_table_size; i++) {
2413 offset = s->refcount_table[i];
2415 inc_refcounts(bs, refcount_table, nb_clusters,
2416 offset, s->cluster_size);
2420 /* compare ref counts */
2421 for(i = 0; i < nb_clusters; i++) {
2422 refcount1 = get_refcount(bs, i);
2423 refcount2 = refcount_table[i];
2424 if (refcount1 != refcount2)
2425 printf("ERROR cluster %d refcount=%d reference=%d\n",
2426 i, refcount1, refcount2);
2429 qemu_free(refcount_table);
2433 static void dump_refcounts(BlockDriverState *bs)
2435 BDRVQcowState *s = bs->opaque;
2436 int64_t nb_clusters, k, k1, size;
2439 size = bdrv_getlength(s->hd);
2440 nb_clusters = (size + s->cluster_size - 1) >> s->cluster_bits;
2441 for(k = 0; k < nb_clusters;) {
2443 refcount = get_refcount(bs, k);
2445 while (k < nb_clusters && get_refcount(bs, k) == refcount)
2447 printf("%lld: refcount=%d nb=%lld\n", k, refcount, k - k1);
2453 BlockDriver bdrv_qcow2 = {
2455 sizeof(BDRVQcowState),
2467 .bdrv_aio_read = qcow_aio_read,
2468 .bdrv_aio_write = qcow_aio_write,
2469 .bdrv_aio_cancel = qcow_aio_cancel,
2470 .aiocb_size = sizeof(QCowAIOCB),
2471 .bdrv_write_compressed = qcow_write_compressed,
2473 .bdrv_snapshot_create = qcow_snapshot_create,
2474 .bdrv_snapshot_goto = qcow_snapshot_goto,
2475 .bdrv_snapshot_delete = qcow_snapshot_delete,
2476 .bdrv_snapshot_list = qcow_snapshot_list,
2477 .bdrv_get_info = qcow_get_info,