]> git.proxmox.com Git - qemu.git/blame - block-qcow2.c
projects / qemu.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
Add missing #define.
[qemu.git] / block-qcow2.c
CommitLineData
585f8587
FB		 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#include "vl.h"
25#include "block_int.h"
26#include <zlib.h>
27#include "aes.h"
28#include <assert.h>
29
30/*
31 Differences with QCOW:
32
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
40 snapshots.
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.
44*/
45
46//#define DEBUG_ALLOC
47//#define DEBUG_ALLOC2
48
49#define QCOW_MAGIC (('Q' << 24) | ('F' << 16) | ('I' << 8) | 0xfb)
50#define QCOW_VERSION 2
51
52#define QCOW_CRYPT_NONE 0
53#define QCOW_CRYPT_AES 1
54
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)
59
60#define REFCOUNT_SHIFT 1 /* refcount size is 2 bytes */
61
62#ifndef offsetof
63#define offsetof(type, field) ((size_t) &((type *)0)->field)
64#endif
65
66typedef struct QCowHeader {
67 uint32_t magic;
68 uint32_t version;
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;
80} QCowHeader;
81
82typedef struct __attribute__((packed)) QCowSnapshotHeader {
83 /* header is 8 byte aligned */
84 uint64_t l1_table_offset;
85
86 uint32_t l1_size;
87 uint16_t id_str_size;
88 uint16_t name_size;
89
90 uint32_t date_sec;
91 uint32_t date_nsec;
92
93 uint64_t vm_clock_nsec;
94
95 uint32_t vm_state_size;
96 uint32_t extra_data_size; /* for extension */
97 /* extra data follows */
98 /* id_str follows */
99 /* name follows */
100} QCowSnapshotHeader;
101
102#define L2_CACHE_SIZE 16
103
104typedef struct QCowSnapshot {
105 uint64_t l1_table_offset;
106 uint32_t l1_size;
107 char *id_str;
108 char *name;
109 uint32_t vm_state_size;
110 uint32_t date_sec;
111 uint32_t date_nsec;
112 uint64_t vm_clock_nsec;
113} QCowSnapshot;
114
115typedef struct BDRVQcowState {
116 BlockDriverState *hd;
117 int cluster_bits;
118 int cluster_size;
119 int cluster_sectors;
120 int l2_bits;
121 int l2_size;
122 int l1_size;
123 int l1_vm_state_index;
124 int csize_shift;
125 int csize_mask;
126 uint64_t cluster_offset_mask;
127 uint64_t l1_table_offset;
128 uint64_t *l1_table;
129 uint64_t *l2_cache;
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;
135
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;
143
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;
149 int snapshots_size;
150 int nb_snapshots;
151 QCowSnapshot *snapshots;
152} BDRVQcowState;
153
154static int decompress_cluster(BDRVQcowState *s, uint64_t cluster_offset);
155static int qcow_read(BlockDriverState *bs, int64_t sector_num,
156 uint8_t *buf, int nb_sectors);
157static int qcow_read_snapshots(BlockDriverState *bs);
158static void qcow_free_snapshots(BlockDriverState *bs);
159static int refcount_init(BlockDriverState *bs);
160static void refcount_close(BlockDriverState *bs);
161static int get_refcount(BlockDriverState *bs, int64_t cluster_index);
162static int update_cluster_refcount(BlockDriverState *bs,
163 int64_t cluster_index,
164 int addend);
165static void update_refcount(BlockDriverState *bs,
166 int64_t offset, int64_t length,
167 int addend);
168static int64_t alloc_clusters(BlockDriverState *bs, int64_t size);
169static int64_t alloc_bytes(BlockDriverState *bs, int size);
170static void free_clusters(BlockDriverState *bs,
171 int64_t offset, int64_t size);
172#ifdef DEBUG_ALLOC
173static void check_refcounts(BlockDriverState *bs);
174#endif
175
176static int qcow_probe(const uint8_t *buf, int buf_size, const char *filename)
177{
178 const QCowHeader *cow_header = (const void *)buf;
179
180 if (buf_size >= sizeof(QCowHeader) &&
181 be32_to_cpu(cow_header->magic) == QCOW_MAGIC &&
182 be32_to_cpu(cow_header->version) == QCOW_VERSION)
183 return 100;
184 else
185 return 0;
186}
187
188static int qcow_open(BlockDriverState *bs, const char *filename, int flags)
189{
190 BDRVQcowState *s = bs->opaque;
191 int len, i, shift, ret;
192 QCowHeader header;
193
194 ret = bdrv_file_open(&s->hd, filename, flags);
195 if (ret < 0)
196 return ret;
197 if (bdrv_pread(s->hd, 0, &header, sizeof(header)) != sizeof(header))
198 goto fail;
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);
212
213 if (header.magic != QCOW_MAGIC || header.version != QCOW_VERSION)
214 goto fail;
215 if (header.size <= 1 ||
216 header.cluster_bits < 9 ||
217 header.cluster_bits > 16)
218 goto fail;
219 if (header.crypt_method > QCOW_CRYPT_AES)
220 goto fail;
221 s->crypt_method_header = header.crypt_method;
222 if (s->crypt_method_header)
223 bs->encrypted = 1;
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);
236
237 s->snapshots_offset = header.snapshots_offset;
238 s->nb_snapshots = header.nb_snapshots;
239
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
245 header.size bytes */
246 if (s->l1_size < s->l1_vm_state_index)
247 goto fail;
248 s->l1_table_offset = header.l1_table_offset;
249 s->l1_table = qemu_malloc(s->l1_size * sizeof(uint64_t));
250 if (!s->l1_table)
251 goto fail;
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))
254 goto fail;
255 for(i = 0;i < s->l1_size; i++) {
256 be64_to_cpus(&s->l1_table[i]);
257 }
258 /* alloc L2 cache */
259 s->l2_cache = qemu_malloc(s->l2_size * L2_CACHE_SIZE * sizeof(uint64_t));
260 if (!s->l2_cache)
261 goto fail;
262 s->cluster_cache = qemu_malloc(s->cluster_size);
263 if (!s->cluster_cache)
264 goto fail;
265 /* one more sector for decompressed data alignment */
266 s->cluster_data = qemu_malloc(s->cluster_size + 512);
267 if (!s->cluster_data)
268 goto fail;
269 s->cluster_cache_offset = -1;
270
271 if (refcount_init(bs) < 0)
272 goto fail;
273
274 /* read the backing file name */
275 if (header.backing_file_offset != 0) {
276 len = header.backing_file_size;
277 if (len > 1023)
278 len = 1023;
279 if (bdrv_pread(s->hd, header.backing_file_offset, bs->backing_file, len) != len)
280 goto fail;
281 bs->backing_file[len] = '\0';
282 }
283 if (qcow_read_snapshots(bs) < 0)
284 goto fail;
285
286#ifdef DEBUG_ALLOC
287 check_refcounts(bs);
288#endif
289 return 0;
290
291 fail:
292 qcow_free_snapshots(bs);
293 refcount_close(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);
298 bdrv_delete(s->hd);
299 return -1;
300}
301
302static int qcow_set_key(BlockDriverState *bs, const char *key)
303{
304 BDRVQcowState *s = bs->opaque;
305 uint8_t keybuf[16];
306 int len, i;
307
308 memset(keybuf, 0, 16);
309 len = strlen(key);
310 if (len > 16)
311 len = 16;
312 /* XXX: we could compress the chars to 7 bits to increase
313 entropy */
314 for(i = 0;i < len;i++) {
315 keybuf[i] = key[i];
316 }
317 s->crypt_method = s->crypt_method_header;
318
319 if (AES_set_encrypt_key(keybuf, 128, &s->aes_encrypt_key) != 0)
320 return -1;
321 if (AES_set_decrypt_key(keybuf, 128, &s->aes_decrypt_key) != 0)
322 return -1;
323#if 0
324 /* test */
325 {
326 uint8_t in[16];
327 uint8_t out[16];
328 uint8_t tmp[16];
329 for(i=0;i<16;i++)
330 in[i] = i;
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]);
335 printf("\n");
336 for(i = 0; i < 16; i++)
337 printf(" %02x", out[i]);
338 printf("\n");
339 }
340#endif
341 return 0;
342}
343
344/* The crypt function is compatible with the linux cryptoloop
345 algorithm for < 4 GB images. NOTE: out_buf == in_buf is
346 supported */
347static 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,
350 const AES_KEY *key)
351{
352 union {
353 uint64_t ll[2];
354 uint8_t b[16];
355 } ivec;
356 int i;
357
358 for(i = 0; i < nb_sectors; i++) {
359 ivec.ll[0] = cpu_to_le64(sector_num);
360 ivec.ll[1] = 0;
361 AES_cbc_encrypt(in_buf, out_buf, 512, key,
362 ivec.b, enc);
363 sector_num++;
364 in_buf += 512;
365 out_buf += 512;
366 }
367}
368
369static int copy_sectors(BlockDriverState *bs, uint64_t start_sect,
370 uint64_t cluster_offset, int n_start, int n_end)
371{
372 BDRVQcowState *s = bs->opaque;
373 int n, ret;
374
375 n = n_end - n_start;
376 if (n <= 0)
377 return 0;
378 ret = qcow_read(bs, start_sect + n_start, s->cluster_data, n);
379 if (ret < 0)
380 return ret;
381 if (s->crypt_method) {
382 encrypt_sectors(s, start_sect + n_start,
383 s->cluster_data,
384 s->cluster_data, n, 1,
385 &s->aes_encrypt_key);
386 }
387 ret = bdrv_write(s->hd, (cluster_offset >> 9) + n_start,
388 s->cluster_data, n);
389 if (ret < 0)
390 return ret;
391 return 0;
392}
393
394static void l2_cache_reset(BlockDriverState *bs)
395{
396 BDRVQcowState *s = bs->opaque;
397
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));
401}
402
403static inline int l2_cache_new_entry(BlockDriverState *bs)
404{
405 BDRVQcowState *s = bs->opaque;
406 uint32_t min_count;
407 int min_index, i;
408
409 /* find a new entry in the least used one */
410 min_index = 0;
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];
415 min_index = i;
416 }
417 }
418 return min_index;
419}
420
421static int64_t align_offset(int64_t offset, int n)
422{
423 offset = (offset + n - 1) & ~(n - 1);
424 return offset;
425}
426
427static int grow_l1_table(BlockDriverState *bs, int min_size)
428{
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;
433 uint64_t data64;
434 uint32_t data32;
435
436 new_l1_size = s->l1_size;
437 if (min_size <= new_l1_size)
438 return 0;
439 while (min_size > new_l1_size) {
440 new_l1_size = (new_l1_size * 3 + 1) / 2;
441 }
442#ifdef DEBUG_ALLOC2
443 printf("grow l1_table from %d to %d\n", s->l1_size, new_l1_size);
444#endif
445
446 new_l1_size2 = sizeof(uint64_t) * new_l1_size;
447 new_l1_table = qemu_mallocz(new_l1_size2);
448 if (!new_l1_table)
449 return -ENOMEM;
450 memcpy(new_l1_table, s->l1_table, s->l1_size * sizeof(uint64_t));
451
452 /* write new table (align to cluster) */
453 new_l1_table_offset = alloc_clusters(bs, new_l1_size2);
454
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)
459 goto fail;
460 for(i = 0; i < s->l1_size; i++)
461 new_l1_table[i] = be64_to_cpu(new_l1_table[i]);
462
463 /* set new table */
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))
467 goto fail;
468 data32 = cpu_to_be32(new_l1_size);
469 if (bdrv_pwrite(s->hd, offsetof(QCowHeader, l1_size),
470 &data32, sizeof(data32)) != sizeof(data32))
471 goto fail;
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;
477 return 0;
478 fail:
479 qemu_free(s->l1_table);
480 return -EIO;
481}
482
483/* 'allocate' is:
484 *
485 * 0 not to allocate.
486 *
487 * 1 to allocate a normal cluster (for sector indexes 'n_start' to
488 * 'n_end')
489 *
490 * 2 to allocate a compressed cluster of size
491 * 'compressed_size'. 'compressed_size' must be > 0 and <
492 * cluster_size
493 *
494 * return 0 if not allocated.
495 */
496static uint64_t get_cluster_offset(BlockDriverState *bs,
497 uint64_t offset, int allocate,
498 int compressed_size,
499 int n_start, int n_end)
500{
501 BDRVQcowState *s = bs->opaque;
502 int min_index, i, j, l1_index, l2_index, ret;
503 uint64_t l2_offset, *l2_table, cluster_offset, tmp, old_l2_offset;
504
505 l1_index = offset >> (s->l2_bits + s->cluster_bits);
506 if (l1_index >= s->l1_size) {
507 /* outside l1 table is allowed: we grow the table if needed */
508 if (!allocate)
509 return 0;
510 if (grow_l1_table(bs, l1_index + 1) < 0)
511 return 0;
512 }
513 l2_offset = s->l1_table[l1_index];
514 if (!l2_offset) {
515 if (!allocate)
516 return 0;
517 l2_allocate:
518 old_l2_offset = l2_offset;
519 /* allocate a new l2 entry */
520 l2_offset = alloc_clusters(bs, s->l2_size * sizeof(uint64_t));
521 /* update the L1 entry */
522 s->l1_table[l1_index] = l2_offset | QCOW_OFLAG_COPIED;
523 tmp = cpu_to_be64(l2_offset | QCOW_OFLAG_COPIED);
524 if (bdrv_pwrite(s->hd, s->l1_table_offset + l1_index * sizeof(tmp),
525 &tmp, sizeof(tmp)) != sizeof(tmp))
526 return 0;
527 min_index = l2_cache_new_entry(bs);
528 l2_table = s->l2_cache + (min_index << s->l2_bits);
529
530 if (old_l2_offset == 0) {
531 memset(l2_table, 0, s->l2_size * sizeof(uint64_t));
532 } else {
533 if (bdrv_pread(s->hd, old_l2_offset,
534 l2_table, s->l2_size * sizeof(uint64_t)) !=
535 s->l2_size * sizeof(uint64_t))
536 return 0;
537 }
538 if (bdrv_pwrite(s->hd, l2_offset,
539 l2_table, s->l2_size * sizeof(uint64_t)) !=
540 s->l2_size * sizeof(uint64_t))
541 return 0;
542 } else {
543 if (!(l2_offset & QCOW_OFLAG_COPIED)) {
544 if (allocate) {
545 free_clusters(bs, l2_offset, s->l2_size * sizeof(uint64_t));
546 goto l2_allocate;
547 }
548 } else {
549 l2_offset &= ~QCOW_OFLAG_COPIED;
550 }
551 for(i = 0; i < L2_CACHE_SIZE; i++) {
552 if (l2_offset == s->l2_cache_offsets[i]) {
553 /* increment the hit count */
554 if (++s->l2_cache_counts[i] == 0xffffffff) {
555 for(j = 0; j < L2_CACHE_SIZE; j++) {
556 s->l2_cache_counts[j] >>= 1;
557 }
558 }
559 l2_table = s->l2_cache + (i << s->l2_bits);
560 goto found;
561 }
562 }
563 /* not found: load a new entry in the least used one */
564 min_index = l2_cache_new_entry(bs);
565 l2_table = s->l2_cache + (min_index << s->l2_bits);
566 if (bdrv_pread(s->hd, l2_offset, l2_table, s->l2_size * sizeof(uint64_t)) !=
567 s->l2_size * sizeof(uint64_t))
568 return 0;
569 }
570 s->l2_cache_offsets[min_index] = l2_offset;
571 s->l2_cache_counts[min_index] = 1;
572 found:
573 l2_index = (offset >> s->cluster_bits) & (s->l2_size - 1);
574 cluster_offset = be64_to_cpu(l2_table[l2_index]);
575 if (!cluster_offset) {
576 if (!allocate)
577 return cluster_offset;
578 } else if (!(cluster_offset & QCOW_OFLAG_COPIED)) {
579 if (!allocate)
580 return cluster_offset;
581 /* free the cluster */
582 if (cluster_offset & QCOW_OFLAG_COMPRESSED) {
583 int nb_csectors;
584 nb_csectors = ((cluster_offset >> s->csize_shift) &
585 s->csize_mask) + 1;
586 free_clusters(bs, (cluster_offset & s->cluster_offset_mask) & ~511,
587 nb_csectors * 512);
588 } else {
589 free_clusters(bs, cluster_offset, s->cluster_size);
590 }
591 } else {
592 cluster_offset &= ~QCOW_OFLAG_COPIED;
593 return cluster_offset;
594 }
595 if (allocate == 1) {
596 /* allocate a new cluster */
597 cluster_offset = alloc_clusters(bs, s->cluster_size);
598
599 /* we must initialize the cluster content which won't be
600 written */
601 if ((n_end - n_start) < s->cluster_sectors) {
602 uint64_t start_sect;
603
604 start_sect = (offset & ~(s->cluster_size - 1)) >> 9;
605 ret = copy_sectors(bs, start_sect,
606 cluster_offset, 0, n_start);
607 if (ret < 0)
608 return 0;
609 ret = copy_sectors(bs, start_sect,
610 cluster_offset, n_end, s->cluster_sectors);
611 if (ret < 0)
612 return 0;
613 }
614 tmp = cpu_to_be64(cluster_offset | QCOW_OFLAG_COPIED);
615 } else {
616 int nb_csectors;
617 cluster_offset = alloc_bytes(bs, compressed_size);
618 nb_csectors = ((cluster_offset + compressed_size - 1) >> 9) -
619 (cluster_offset >> 9);
620 cluster_offset |= QCOW_OFLAG_COMPRESSED |
621 ((uint64_t)nb_csectors << s->csize_shift);
622 /* compressed clusters never have the copied flag */
623 tmp = cpu_to_be64(cluster_offset);
624 }
625 /* update L2 table */
626 l2_table[l2_index] = tmp;
627 if (bdrv_pwrite(s->hd,
628 l2_offset + l2_index * sizeof(tmp), &tmp, sizeof(tmp)) != sizeof(tmp))
629 return 0;
630 return cluster_offset;
631}
632
633static int qcow_is_allocated(BlockDriverState *bs, int64_t sector_num,
634 int nb_sectors, int *pnum)
635{
636 BDRVQcowState *s = bs->opaque;
637 int index_in_cluster, n;
638 uint64_t cluster_offset;
639
640 cluster_offset = get_cluster_offset(bs, sector_num << 9, 0, 0, 0, 0);
641 index_in_cluster = sector_num & (s->cluster_sectors - 1);
642 n = s->cluster_sectors - index_in_cluster;
643 if (n > nb_sectors)
644 n = nb_sectors;
645 *pnum = n;
646 return (cluster_offset != 0);
647}
648
649static int decompress_buffer(uint8_t *out_buf, int out_buf_size,
650 const uint8_t *buf, int buf_size)
651{
652 z_stream strm1, *strm = &strm1;
653 int ret, out_len;
654
655 memset(strm, 0, sizeof(*strm));
656
657 strm->next_in = (uint8_t *)buf;
658 strm->avail_in = buf_size;
659 strm->next_out = out_buf;
660 strm->avail_out = out_buf_size;
661
662 ret = inflateInit2(strm, -12);
663 if (ret != Z_OK)
664 return -1;
665 ret = inflate(strm, Z_FINISH);
666 out_len = strm->next_out - out_buf;
667 if ((ret != Z_STREAM_END && ret != Z_BUF_ERROR) ||
668 out_len != out_buf_size) {
669 inflateEnd(strm);
670 return -1;
671 }
672 inflateEnd(strm);
673 return 0;
674}
675
676static int decompress_cluster(BDRVQcowState *s, uint64_t cluster_offset)
677{
678 int ret, csize, nb_csectors, sector_offset;
679 uint64_t coffset;
680
681 coffset = cluster_offset & s->cluster_offset_mask;
682 if (s->cluster_cache_offset != coffset) {
683 nb_csectors = ((cluster_offset >> s->csize_shift) & s->csize_mask) + 1;
684 sector_offset = coffset & 511;
685 csize = nb_csectors * 512 - sector_offset;
686 ret = bdrv_read(s->hd, coffset >> 9, s->cluster_data, nb_csectors);
687 if (ret < 0) {
688 return -1;
689 }
690 if (decompress_buffer(s->cluster_cache, s->cluster_size,
691 s->cluster_data + sector_offset, csize) < 0) {
692 return -1;
693 }
694 s->cluster_cache_offset = coffset;
695 }
696 return 0;
697}
698
699static int qcow_read(BlockDriverState *bs, int64_t sector_num,
700 uint8_t *buf, int nb_sectors)
701{
702 BDRVQcowState *s = bs->opaque;
703 int ret, index_in_cluster, n;
704 uint64_t cluster_offset;
705
706 while (nb_sectors > 0) {
707 cluster_offset = get_cluster_offset(bs, sector_num << 9, 0, 0, 0, 0);
708 index_in_cluster = sector_num & (s->cluster_sectors - 1);
709 n = s->cluster_sectors - index_in_cluster;
710 if (n > nb_sectors)
711 n = nb_sectors;
712 if (!cluster_offset) {
713 if (bs->backing_hd) {
714 /* read from the base image */
715 ret = bdrv_read(bs->backing_hd, sector_num, buf, n);
716 if (ret < 0)
717 return -1;
718 } else {
719 memset(buf, 0, 512 * n);
720 }
721 } else if (cluster_offset & QCOW_OFLAG_COMPRESSED) {
722 if (decompress_cluster(s, cluster_offset) < 0)
723 return -1;
724 memcpy(buf, s->cluster_cache + index_in_cluster * 512, 512 * n);
725 } else {
726 ret = bdrv_pread(s->hd, cluster_offset + index_in_cluster * 512, buf, n * 512);
727 if (ret != n * 512)
728 return -1;
729 if (s->crypt_method) {
730 encrypt_sectors(s, sector_num, buf, buf, n, 0,
731 &s->aes_decrypt_key);
732 }
733 }
734 nb_sectors -= n;
735 sector_num += n;
736 buf += n * 512;
737 }
738 return 0;
739}
740
741static int qcow_write(BlockDriverState *bs, int64_t sector_num,
742 const uint8_t *buf, int nb_sectors)
743{
744 BDRVQcowState *s = bs->opaque;
745 int ret, index_in_cluster, n;
746 uint64_t cluster_offset;
747
748 while (nb_sectors > 0) {
749 index_in_cluster = sector_num & (s->cluster_sectors - 1);
750 n = s->cluster_sectors - index_in_cluster;
751 if (n > nb_sectors)
752 n = nb_sectors;
753 cluster_offset = get_cluster_offset(bs, sector_num << 9, 1, 0,
754 index_in_cluster,
755 index_in_cluster + n);
756 if (!cluster_offset)
757 return -1;
758 if (s->crypt_method) {
759 encrypt_sectors(s, sector_num, s->cluster_data, buf, n, 1,
760 &s->aes_encrypt_key);
761 ret = bdrv_pwrite(s->hd, cluster_offset + index_in_cluster * 512,
762 s->cluster_data, n * 512);
763 } else {
764 ret = bdrv_pwrite(s->hd, cluster_offset + index_in_cluster * 512, buf, n * 512);
765 }
766 if (ret != n * 512)
767 return -1;
768 nb_sectors -= n;
769 sector_num += n;
770 buf += n * 512;
771 }
772 s->cluster_cache_offset = -1; /* disable compressed cache */
773 return 0;
774}
775
776typedef struct {
777 int64_t sector_num;
778 uint8_t *buf;
779 int nb_sectors;
780 int n;
781 uint64_t cluster_offset;
782 uint8_t *cluster_data;
783 BlockDriverAIOCB *hd_aiocb;
784 BlockDriverAIOCB *backing_hd_aiocb;
785} QCowAIOCB;
786
787static void qcow_aio_delete(BlockDriverAIOCB *acb);
788
789static int qcow_aio_new(BlockDriverAIOCB *acb)
790{
791 BlockDriverState *bs = acb->bs;
792 BDRVQcowState *s = bs->opaque;
793 QCowAIOCB *acb1;
794 acb1 = qemu_mallocz(sizeof(QCowAIOCB));
795 if (!acb1)
796 return -1;
797 acb->opaque = acb1;
798 acb1->hd_aiocb = bdrv_aio_new(s->hd);
799 if (!acb1->hd_aiocb)
800 goto fail;
801 if (bs->backing_hd) {
802 acb1->backing_hd_aiocb = bdrv_aio_new(bs->backing_hd);
803 if (!acb1->backing_hd_aiocb)
804 goto fail;
805 }
806 return 0;
807 fail:
808 qcow_aio_delete(acb);
809 return -1;
810}
811
812static void qcow_aio_read_cb(void *opaque, int ret)
813{
814 BlockDriverAIOCB *acb = opaque;
815 BlockDriverState *bs = acb->bs;
816 BDRVQcowState *s = bs->opaque;
817 QCowAIOCB *acb1 = acb->opaque;
818 int index_in_cluster;
819
820 if (ret < 0) {
821 fail:
822 acb->cb(acb->cb_opaque, ret);
823 return;
824 }
825
826 redo:
827 /* post process the read buffer */
828 if (!acb1->cluster_offset) {
829 /* nothing to do */
830 } else if (acb1->cluster_offset & QCOW_OFLAG_COMPRESSED) {
831 /* nothing to do */
832 } else {
833 if (s->crypt_method) {
834 encrypt_sectors(s, acb1->sector_num, acb1->buf, acb1->buf,
835 acb1->n, 0,
836 &s->aes_decrypt_key);
837 }
838 }
839
840 acb1->nb_sectors -= acb1->n;
841 acb1->sector_num += acb1->n;
842 acb1->buf += acb1->n * 512;
843
844 if (acb1->nb_sectors == 0) {
845 /* request completed */
846 acb->cb(acb->cb_opaque, 0);
847 return;
848 }
849
850 /* prepare next AIO request */
851 acb1->cluster_offset = get_cluster_offset(bs,
852 acb1->sector_num << 9,
853 0, 0, 0, 0);
854 index_in_cluster = acb1->sector_num & (s->cluster_sectors - 1);
855 acb1->n = s->cluster_sectors - index_in_cluster;
856 if (acb1->n > acb1->nb_sectors)
857 acb1->n = acb1->nb_sectors;
858
859 if (!acb1->cluster_offset) {
860 if (bs->backing_hd) {
861 /* read from the base image */
862 ret = bdrv_aio_read(acb1->backing_hd_aiocb, acb1->sector_num,
863 acb1->buf, acb1->n, qcow_aio_read_cb, acb);
864 if (ret < 0)
865 goto fail;
866 } else {
867 /* Note: in this case, no need to wait */
868 memset(acb1->buf, 0, 512 * acb1->n);
869 goto redo;
870 }
871 } else if (acb1->cluster_offset & QCOW_OFLAG_COMPRESSED) {
872 /* add AIO support for compressed blocks ? */
873 if (decompress_cluster(s, acb1->cluster_offset) < 0)
874 goto fail;
875 memcpy(acb1->buf,
876 s->cluster_cache + index_in_cluster * 512, 512 * acb1->n);
877 goto redo;
878 } else {
879 if ((acb1->cluster_offset & 511) != 0) {
880 ret = -EIO;
881 goto fail;
882 }
883 ret = bdrv_aio_read(acb1->hd_aiocb,
884 (acb1->cluster_offset >> 9) + index_in_cluster,
885 acb1->buf, acb1->n, qcow_aio_read_cb, acb);
886 if (ret < 0)
887 goto fail;
888 }
889}
890
891static int qcow_aio_read(BlockDriverAIOCB *acb, int64_t sector_num,
892 uint8_t *buf, int nb_sectors)
893{
894 QCowAIOCB *acb1 = acb->opaque;
895
896 acb1->sector_num = sector_num;
897 acb1->buf = buf;
898 acb1->nb_sectors = nb_sectors;
899 acb1->n = 0;
900 acb1->cluster_offset = 0;
901
902 qcow_aio_read_cb(acb, 0);
903 return 0;
904}
905
906static void qcow_aio_write_cb(void *opaque, int ret)
907{
908 BlockDriverAIOCB *acb = opaque;
909 BlockDriverState *bs = acb->bs;
910 BDRVQcowState *s = bs->opaque;
911 QCowAIOCB *acb1 = acb->opaque;
912 int index_in_cluster;
913 uint64_t cluster_offset;
914 const uint8_t *src_buf;
915
916 if (ret < 0) {
917 fail:
918 acb->cb(acb->cb_opaque, ret);
919 return;
920 }
921
922 acb1->nb_sectors -= acb1->n;
923 acb1->sector_num += acb1->n;
924 acb1->buf += acb1->n * 512;
925
926 if (acb1->nb_sectors == 0) {
927 /* request completed */
928 acb->cb(acb->cb_opaque, 0);
929 return;
930 }
931
932 index_in_cluster = acb1->sector_num & (s->cluster_sectors - 1);
933 acb1->n = s->cluster_sectors - index_in_cluster;
934 if (acb1->n > acb1->nb_sectors)
935 acb1->n = acb1->nb_sectors;
936 cluster_offset = get_cluster_offset(bs, acb1->sector_num << 9, 1, 0,
937 index_in_cluster,
938 index_in_cluster + acb1->n);
939 if (!cluster_offset || (cluster_offset & 511) != 0) {
940 ret = -EIO;
941 goto fail;
942 }
943 if (s->crypt_method) {
944 if (!acb1->cluster_data) {
945 acb1->cluster_data = qemu_mallocz(s->cluster_size);
946 if (!acb1->cluster_data) {
947 ret = -ENOMEM;
948 goto fail;
949 }
950 }
951 encrypt_sectors(s, acb1->sector_num, acb1->cluster_data, acb1->buf,
952 acb1->n, 1, &s->aes_encrypt_key);
953 src_buf = acb1->cluster_data;
954 } else {
955 src_buf = acb1->buf;
956 }
957 ret = bdrv_aio_write(acb1->hd_aiocb,
958 (cluster_offset >> 9) + index_in_cluster,
959 src_buf, acb1->n,
960 qcow_aio_write_cb, acb);
961 if (ret < 0)
962 goto fail;
963}
964
965static int qcow_aio_write(BlockDriverAIOCB *acb, int64_t sector_num,
966 const uint8_t *buf, int nb_sectors)
967{
968 QCowAIOCB *acb1 = acb->opaque;
969 BlockDriverState *bs = acb->bs;
970 BDRVQcowState *s = bs->opaque;
971
972 s->cluster_cache_offset = -1; /* disable compressed cache */
973
974 acb1->sector_num = sector_num;
975 acb1->buf = (uint8_t *)buf;
976 acb1->nb_sectors = nb_sectors;
977 acb1->n = 0;
978
979 qcow_aio_write_cb(acb, 0);
980 return 0;
981}
982
983static void qcow_aio_cancel(BlockDriverAIOCB *acb)
984{
985 QCowAIOCB *acb1 = acb->opaque;
986 if (acb1->hd_aiocb)
987 bdrv_aio_cancel(acb1->hd_aiocb);
988 if (acb1->backing_hd_aiocb)
989 bdrv_aio_cancel(acb1->backing_hd_aiocb);
990}
991
992static void qcow_aio_delete(BlockDriverAIOCB *acb)
993{
994 QCowAIOCB *acb1 = acb->opaque;
995 if (acb1->hd_aiocb)
996 bdrv_aio_delete(acb1->hd_aiocb);
997 if (acb1->backing_hd_aiocb)
998 bdrv_aio_delete(acb1->backing_hd_aiocb);
999 qemu_free(acb1->cluster_data);
1000 qemu_free(acb1);
1001}
1002
1003static void qcow_close(BlockDriverState *bs)
1004{
1005 BDRVQcowState *s = bs->opaque;
1006 qemu_free(s->l1_table);
1007 qemu_free(s->l2_cache);
1008 qemu_free(s->cluster_cache);
1009 qemu_free(s->cluster_data);
1010 refcount_close(bs);
1011 bdrv_delete(s->hd);
1012}
1013
1014/* XXX: use std qcow open function ? */
1015typedef struct QCowCreateState {
1016 int cluster_size;
1017 int cluster_bits;
1018 uint16_t *refcount_block;
1019 uint64_t *refcount_table;
1020 int64_t l1_table_offset;
1021 int64_t refcount_table_offset;
1022 int64_t refcount_block_offset;
1023} QCowCreateState;
1024
1025static void create_refcount_update(QCowCreateState *s,
1026 int64_t offset, int64_t size)
1027{
1028 int refcount;
1029 int64_t start, last, cluster_offset;
1030 uint16_t *p;
1031
1032 start = offset & ~(s->cluster_size - 1);
1033 last = (offset + size - 1) & ~(s->cluster_size - 1);
1034 for(cluster_offset = start; cluster_offset <= last;
1035 cluster_offset += s->cluster_size) {
1036 p = &s->refcount_block[cluster_offset >> s->cluster_bits];
1037 refcount = be16_to_cpu(*p);
1038 refcount++;
1039 *p = cpu_to_be16(refcount);
1040 }
1041}
1042
1043static int qcow_create(const char *filename, int64_t total_size,
1044 const char *backing_file, int flags)
1045{
1046 int fd, header_size, backing_filename_len, l1_size, i, shift, l2_bits;
1047 QCowHeader header;
1048 uint64_t tmp, offset;
1049 QCowCreateState s1, *s = &s1;
1050
1051 memset(s, 0, sizeof(*s));
1052
1053 fd = open(filename, O_WRONLY | O_CREAT | O_TRUNC | O_BINARY, 0644);
1054 if (fd < 0)
1055 return -1;
1056 memset(&header, 0, sizeof(header));
1057 header.magic = cpu_to_be32(QCOW_MAGIC);
1058 header.version = cpu_to_be32(QCOW_VERSION);
1059 header.size = cpu_to_be64(total_size * 512);
1060 header_size = sizeof(header);
1061 backing_filename_len = 0;
1062 if (backing_file) {
1063 header.backing_file_offset = cpu_to_be64(header_size);
1064 backing_filename_len = strlen(backing_file);
1065 header.backing_file_size = cpu_to_be32(backing_filename_len);
1066 header_size += backing_filename_len;
1067 }
1068 s->cluster_bits = 12; /* 4 KB clusters */
1069 s->cluster_size = 1 << s->cluster_bits;
1070 header.cluster_bits = cpu_to_be32(s->cluster_bits);
1071 header_size = (header_size + 7) & ~7;
1072 if (flags) {
1073 header.crypt_method = cpu_to_be32(QCOW_CRYPT_AES);
1074 } else {
1075 header.crypt_method = cpu_to_be32(QCOW_CRYPT_NONE);
1076 }
1077 l2_bits = s->cluster_bits - 3;
1078 shift = s->cluster_bits + l2_bits;
1079 l1_size = (((total_size * 512) + (1LL << shift) - 1) >> shift);
1080 offset = align_offset(header_size, s->cluster_size);
1081 s->l1_table_offset = offset;
1082 header.l1_table_offset = cpu_to_be64(s->l1_table_offset);
1083 header.l1_size = cpu_to_be32(l1_size);
15e6690a 1084 offset += align_offset(l1_size * sizeof(uint64_t), s->cluster_size);
585f8587
FB		 1085
1086 s->refcount_table = qemu_mallocz(s->cluster_size);
1087 if (!s->refcount_table)
1088 goto fail;
1089 s->refcount_block = qemu_mallocz(s->cluster_size);
1090 if (!s->refcount_block)
1091 goto fail;
15e6690a 1092
585f8587
FB		 1093 s->refcount_table_offset = offset;
1094 header.refcount_table_offset = cpu_to_be64(offset);
1095 header.refcount_table_clusters = cpu_to_be32(1);
1096 offset += s->cluster_size;
1097
1098 s->refcount_table[0] = cpu_to_be64(offset);
1099 s->refcount_block_offset = offset;
1100 offset += s->cluster_size;
1101
1102 /* update refcounts */
1103 create_refcount_update(s, 0, header_size);
15e6690a 1104 create_refcount_update(s, s->l1_table_offset, l1_size * sizeof(uint64_t));
585f8587
FB		 1105 create_refcount_update(s, s->refcount_table_offset, s->cluster_size);
1106 create_refcount_update(s, s->refcount_block_offset, s->cluster_size);
1107
1108 /* write all the data */
1109 write(fd, &header, sizeof(header));
1110 if (backing_file) {
1111 write(fd, backing_file, backing_filename_len);
1112 }
1113 lseek(fd, s->l1_table_offset, SEEK_SET);
1114 tmp = 0;
1115 for(i = 0;i < l1_size; i++) {
1116 write(fd, &tmp, sizeof(tmp));
1117 }
1118 lseek(fd, s->refcount_table_offset, SEEK_SET);
1119 write(fd, s->refcount_table, s->cluster_size);
1120
1121 lseek(fd, s->refcount_block_offset, SEEK_SET);
1122 write(fd, s->refcount_block, s->cluster_size);
1123
1124 qemu_free(s->refcount_table);
1125 qemu_free(s->refcount_block);
1126 close(fd);
1127 return 0;
1128 fail:
1129 qemu_free(s->refcount_table);
1130 qemu_free(s->refcount_block);
1131 close(fd);
1132 return -ENOMEM;
1133}
1134
1135static int qcow_make_empty(BlockDriverState *bs)
1136{
1137#if 0
1138 /* XXX: not correct */
1139 BDRVQcowState *s = bs->opaque;
1140 uint32_t l1_length = s->l1_size * sizeof(uint64_t);
1141 int ret;
1142
1143 memset(s->l1_table, 0, l1_length);
1144 if (bdrv_pwrite(s->hd, s->l1_table_offset, s->l1_table, l1_length) < 0)
1145 return -1;
1146 ret = bdrv_truncate(s->hd, s->l1_table_offset + l1_length);
1147 if (ret < 0)
1148 return ret;
1149
1150 l2_cache_reset(bs);
1151#endif
1152 return 0;
1153}
1154
1155/* XXX: put compressed sectors first, then all the cluster aligned
1156 tables to avoid losing bytes in alignment */
1157static int qcow_write_compressed(BlockDriverState *bs, int64_t sector_num,
1158 const uint8_t *buf, int nb_sectors)
1159{
1160 BDRVQcowState *s = bs->opaque;
1161 z_stream strm;
1162 int ret, out_len;
1163 uint8_t *out_buf;
1164 uint64_t cluster_offset;
1165
1166 if (nb_sectors == 0) {
1167 /* align end of file to a sector boundary to ease reading with
1168 sector based I/Os */
1169 cluster_offset = bdrv_getlength(s->hd);
1170 cluster_offset = (cluster_offset + 511) & ~511;
1171 bdrv_truncate(s->hd, cluster_offset);
1172 return 0;
1173 }
1174
1175 if (nb_sectors != s->cluster_sectors)
1176 return -EINVAL;
1177
1178 out_buf = qemu_malloc(s->cluster_size + (s->cluster_size / 1000) + 128);
1179 if (!out_buf)
1180 return -ENOMEM;
1181
1182 /* best compression, small window, no zlib header */
1183 memset(&strm, 0, sizeof(strm));
1184 ret = deflateInit2(&strm, Z_DEFAULT_COMPRESSION,
1185 Z_DEFLATED, -12,
1186 9, Z_DEFAULT_STRATEGY);
1187 if (ret != 0) {
1188 qemu_free(out_buf);
1189 return -1;
1190 }
1191
1192 strm.avail_in = s->cluster_size;
1193 strm.next_in = (uint8_t *)buf;
1194 strm.avail_out = s->cluster_size;
1195 strm.next_out = out_buf;
1196
1197 ret = deflate(&strm, Z_FINISH);
1198 if (ret != Z_STREAM_END && ret != Z_OK) {
1199 qemu_free(out_buf);
1200 deflateEnd(&strm);
1201 return -1;
1202 }
1203 out_len = strm.next_out - out_buf;
1204
1205 deflateEnd(&strm);
1206
1207 if (ret != Z_STREAM_END || out_len >= s->cluster_size) {
1208 /* could not compress: write normal cluster */
1209 qcow_write(bs, sector_num, buf, s->cluster_sectors);
1210 } else {
1211 cluster_offset = get_cluster_offset(bs, sector_num << 9, 2,
1212 out_len, 0, 0);
1213 cluster_offset &= s->cluster_offset_mask;
1214 if (bdrv_pwrite(s->hd, cluster_offset, out_buf, out_len) != out_len) {
1215 qemu_free(out_buf);
1216 return -1;
1217 }
1218 }
1219
1220 qemu_free(out_buf);
1221 return 0;
1222}
1223
1224static void qcow_flush(BlockDriverState *bs)
1225{
1226 BDRVQcowState *s = bs->opaque;
1227 bdrv_flush(s->hd);
1228}
1229
1230static int qcow_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
1231{
1232 BDRVQcowState *s = bs->opaque;
1233 bdi->cluster_size = s->cluster_size;
1234 bdi->vm_state_offset = (int64_t)s->l1_vm_state_index <<
1235 (s->cluster_bits + s->l2_bits);
1236 return 0;
1237}
1238
1239/*********************************************************/
1240/* snapshot support */
1241
1242/* update the refcounts of snapshots and the copied flag */
1243static int update_snapshot_refcount(BlockDriverState *bs,
1244 int64_t l1_table_offset,
1245 int l1_size,
1246 int addend)
1247{
1248 BDRVQcowState *s = bs->opaque;
1249 uint64_t *l1_table, *l2_table, l2_offset, offset, l1_size2, l1_allocated;
1250 int64_t old_offset, old_l2_offset;
1251 int l2_size, i, j, l1_modified, l2_modified, nb_csectors, refcount;
1252
1253 l2_cache_reset(bs);
1254
1255 l2_table = NULL;
1256 l1_table = NULL;
1257 l1_size2 = l1_size * sizeof(uint64_t);
1258 l1_allocated = 0;
1259 if (l1_table_offset != s->l1_table_offset) {
1260 l1_table = qemu_malloc(l1_size2);
1261 if (!l1_table)
1262 goto fail;
1263 l1_allocated = 1;
1264 if (bdrv_pread(s->hd, l1_table_offset,
1265 l1_table, l1_size2) != l1_size2)
1266 goto fail;
1267 for(i = 0;i < l1_size; i++)
1268 be64_to_cpus(&l1_table[i]);
1269 } else {
1270 assert(l1_size == s->l1_size);
1271 l1_table = s->l1_table;
1272 l1_allocated = 0;
1273 }
1274
1275 l2_size = s->l2_size * sizeof(uint64_t);
1276 l2_table = qemu_malloc(l2_size);
1277 if (!l2_table)
1278 goto fail;
1279 l1_modified = 0;
1280 for(i = 0; i < l1_size; i++) {
1281 l2_offset = l1_table[i];
1282 if (l2_offset) {
1283 old_l2_offset = l2_offset;
1284 l2_offset &= ~QCOW_OFLAG_COPIED;
1285 l2_modified = 0;
1286 if (bdrv_pread(s->hd, l2_offset, l2_table, l2_size) != l2_size)
1287 goto fail;
1288 for(j = 0; j < s->l2_size; j++) {
1289 offset = be64_to_cpu(l2_table[j]);
1290 if (offset != 0) {
1291 old_offset = offset;
1292 offset &= ~QCOW_OFLAG_COPIED;
1293 if (offset & QCOW_OFLAG_COMPRESSED) {
1294 nb_csectors = ((offset >> s->csize_shift) &
1295 s->csize_mask) + 1;
1296 if (addend != 0)
1297 update_refcount(bs, (offset & s->cluster_offset_mask) & ~511,
1298 nb_csectors * 512, addend);
1299 /* compressed clusters are never modified */
1300 refcount = 2;
1301 } else {
1302 if (addend != 0) {
1303 refcount = update_cluster_refcount(bs, offset >> s->cluster_bits, addend);
1304 } else {
1305 refcount = get_refcount(bs, offset >> s->cluster_bits);
1306 }
1307 }
1308
1309 if (refcount == 1) {
1310 offset |= QCOW_OFLAG_COPIED;
1311 }
1312 if (offset != old_offset) {
1313 l2_table[j] = cpu_to_be64(offset);
1314 l2_modified = 1;
1315 }
1316 }
1317 }
1318 if (l2_modified) {
1319 if (bdrv_pwrite(s->hd,
1320 l2_offset, l2_table, l2_size) != l2_size)
1321 goto fail;
1322 }
1323
1324 if (addend != 0) {
1325 refcount = update_cluster_refcount(bs, l2_offset >> s->cluster_bits, addend);
1326 } else {
1327 refcount = get_refcount(bs, l2_offset >> s->cluster_bits);
1328 }
1329 if (refcount == 1) {
1330 l2_offset |= QCOW_OFLAG_COPIED;
1331 }
1332 if (l2_offset != old_l2_offset) {
1333 l1_table[i] = l2_offset;
1334 l1_modified = 1;
1335 }
1336 }
1337 }
1338 if (l1_modified) {
1339 for(i = 0; i < l1_size; i++)
1340 cpu_to_be64s(&l1_table[i]);
1341 if (bdrv_pwrite(s->hd, l1_table_offset, l1_table,
1342 l1_size2) != l1_size2)
1343 goto fail;
1344 for(i = 0; i < l1_size; i++)
1345 be64_to_cpus(&l1_table[i]);
1346 }
1347 if (l1_allocated)
1348 qemu_free(l1_table);
1349 qemu_free(l2_table);
1350 return 0;
1351 fail:
1352 if (l1_allocated)
1353 qemu_free(l1_table);
1354 qemu_free(l2_table);
1355 return -EIO;
1356}
1357
1358static void qcow_free_snapshots(BlockDriverState *bs)
1359{
1360 BDRVQcowState *s = bs->opaque;
1361 int i;
1362
1363 for(i = 0; i < s->nb_snapshots; i++) {
1364 qemu_free(s->snapshots[i].name);
1365 qemu_free(s->snapshots[i].id_str);
1366 }
1367 qemu_free(s->snapshots);
1368 s->snapshots = NULL;
1369 s->nb_snapshots = 0;
1370}
1371
1372static int qcow_read_snapshots(BlockDriverState *bs)
1373{
1374 BDRVQcowState *s = bs->opaque;
1375 QCowSnapshotHeader h;
1376 QCowSnapshot *sn;
1377 int i, id_str_size, name_size;
1378 int64_t offset;
1379 uint32_t extra_data_size;
1380
1381 offset = s->snapshots_offset;
1382 s->snapshots = qemu_mallocz(s->nb_snapshots * sizeof(QCowSnapshot));
1383 if (!s->snapshots)
1384 goto fail;
1385 for(i = 0; i < s->nb_snapshots; i++) {
1386 offset = align_offset(offset, 8);
1387 if (bdrv_pread(s->hd, offset, &h, sizeof(h)) != sizeof(h))
1388 goto fail;
1389 offset += sizeof(h);
1390 sn = s->snapshots + i;
1391 sn->l1_table_offset = be64_to_cpu(h.l1_table_offset);
1392 sn->l1_size = be32_to_cpu(h.l1_size);
1393 sn->vm_state_size = be32_to_cpu(h.vm_state_size);
1394 sn->date_sec = be32_to_cpu(h.date_sec);
1395 sn->date_nsec = be32_to_cpu(h.date_nsec);
1396 sn->vm_clock_nsec = be64_to_cpu(h.vm_clock_nsec);
1397 extra_data_size = be32_to_cpu(h.extra_data_size);
1398
1399 id_str_size = be16_to_cpu(h.id_str_size);
1400 name_size = be16_to_cpu(h.name_size);
1401
1402 offset += extra_data_size;
1403
1404 sn->id_str = qemu_malloc(id_str_size + 1);
1405 if (!sn->id_str)
1406 goto fail;
1407 if (bdrv_pread(s->hd, offset, sn->id_str, id_str_size) != id_str_size)
1408 goto fail;
1409 offset += id_str_size;
1410 sn->id_str[id_str_size] = '\0';
1411
1412 sn->name = qemu_malloc(name_size + 1);
1413 if (!sn->name)
1414 goto fail;
1415 if (bdrv_pread(s->hd, offset, sn->name, name_size) != name_size)
1416 goto fail;
1417 offset += name_size;
1418 sn->name[name_size] = '\0';
1419 }
1420 s->snapshots_size = offset - s->snapshots_offset;
1421 return 0;
1422 fail:
1423 qcow_free_snapshots(bs);
1424 return -1;
1425}
1426
1427/* add at the end of the file a new list of snapshots */
1428static int qcow_write_snapshots(BlockDriverState *bs)
1429{
1430 BDRVQcowState *s = bs->opaque;
1431 QCowSnapshot *sn;
1432 QCowSnapshotHeader h;
1433 int i, name_size, id_str_size, snapshots_size;
1434 uint64_t data64;
1435 uint32_t data32;
1436 int64_t offset, snapshots_offset;
1437
1438 /* compute the size of the snapshots */
1439 offset = 0;
1440 for(i = 0; i < s->nb_snapshots; i++) {
1441 sn = s->snapshots + i;
1442 offset = align_offset(offset, 8);
1443 offset += sizeof(h);
1444 offset += strlen(sn->id_str);
1445 offset += strlen(sn->name);
1446 }
1447 snapshots_size = offset;
1448
1449 snapshots_offset = alloc_clusters(bs, snapshots_size);
1450 offset = snapshots_offset;
1451
1452 for(i = 0; i < s->nb_snapshots; i++) {
1453 sn = s->snapshots + i;
1454 memset(&h, 0, sizeof(h));
1455 h.l1_table_offset = cpu_to_be64(sn->l1_table_offset);
1456 h.l1_size = cpu_to_be32(sn->l1_size);
1457 h.vm_state_size = cpu_to_be32(sn->vm_state_size);
1458 h.date_sec = cpu_to_be32(sn->date_sec);
1459 h.date_nsec = cpu_to_be32(sn->date_nsec);
1460 h.vm_clock_nsec = cpu_to_be64(sn->vm_clock_nsec);
1461
1462 id_str_size = strlen(sn->id_str);
1463 name_size = strlen(sn->name);
1464 h.id_str_size = cpu_to_be16(id_str_size);
1465 h.name_size = cpu_to_be16(name_size);
1466 offset = align_offset(offset, 8);
1467 if (bdrv_pwrite(s->hd, offset, &h, sizeof(h)) != sizeof(h))
1468 goto fail;
1469 offset += sizeof(h);
1470 if (bdrv_pwrite(s->hd, offset, sn->id_str, id_str_size) != id_str_size)
1471 goto fail;
1472 offset += id_str_size;
1473 if (bdrv_pwrite(s->hd, offset, sn->name, name_size) != name_size)
1474 goto fail;
1475 offset += name_size;
1476 }
1477
1478 /* update the various header fields */
1479 data64 = cpu_to_be64(snapshots_offset);
1480 if (bdrv_pwrite(s->hd, offsetof(QCowHeader, snapshots_offset),
1481 &data64, sizeof(data64)) != sizeof(data64))
1482 goto fail;
1483 data32 = cpu_to_be32(s->nb_snapshots);
1484 if (bdrv_pwrite(s->hd, offsetof(QCowHeader, nb_snapshots),
1485 &data32, sizeof(data32)) != sizeof(data32))
1486 goto fail;
1487
1488 /* free the old snapshot table */
1489 free_clusters(bs, s->snapshots_offset, s->snapshots_size);
1490 s->snapshots_offset = snapshots_offset;
1491 s->snapshots_size = snapshots_size;
1492 return 0;
1493 fail:
1494 return -1;
1495}
1496
1497static void find_new_snapshot_id(BlockDriverState *bs,
1498 char *id_str, int id_str_size)
1499{
1500 BDRVQcowState *s = bs->opaque;
1501 QCowSnapshot *sn;
1502 int i, id, id_max = 0;
1503
1504 for(i = 0; i < s->nb_snapshots; i++) {
1505 sn = s->snapshots + i;
1506 id = strtoul(sn->id_str, NULL, 10);
1507 if (id > id_max)
1508 id_max = id;
1509 }
1510 snprintf(id_str, id_str_size, "%d", id_max + 1);
1511}
1512
1513static int find_snapshot_by_id(BlockDriverState *bs, const char *id_str)
1514{
1515 BDRVQcowState *s = bs->opaque;
1516 int i;
1517
1518 for(i = 0; i < s->nb_snapshots; i++) {
1519 if (!strcmp(s->snapshots[i].id_str, id_str))
1520 return i;
1521 }
1522 return -1;
1523}
1524
1525static int find_snapshot_by_id_or_name(BlockDriverState *bs, const char *name)
1526{
1527 BDRVQcowState *s = bs->opaque;
1528 int i, ret;
1529
1530 ret = find_snapshot_by_id(bs, name);
1531 if (ret >= 0)
1532 return ret;
1533 for(i = 0; i < s->nb_snapshots; i++) {
1534 if (!strcmp(s->snapshots[i].name, name))
1535 return i;
1536 }
1537 return -1;
1538}
1539
1540/* if no id is provided, a new one is constructed */
1541static int qcow_snapshot_create(BlockDriverState *bs,
1542 QEMUSnapshotInfo *sn_info)
1543{
1544 BDRVQcowState *s = bs->opaque;
1545 QCowSnapshot *snapshots1, sn1, *sn = &sn1;
1546 int i, ret;
1547 uint64_t *l1_table = NULL;
1548
1549 memset(sn, 0, sizeof(*sn));
1550
1551 if (sn_info->id_str[0] == '\0') {
1552 /* compute a new id */
1553 find_new_snapshot_id(bs, sn_info->id_str, sizeof(sn_info->id_str));
1554 }
1555
1556 /* check that the ID is unique */
1557 if (find_snapshot_by_id(bs, sn_info->id_str) >= 0)
1558 return -ENOENT;
1559
1560 sn->id_str = qemu_strdup(sn_info->id_str);
1561 if (!sn->id_str)
1562 goto fail;
1563 sn->name = qemu_strdup(sn_info->name);
1564 if (!sn->name)
1565 goto fail;
1566 sn->vm_state_size = sn_info->vm_state_size;
1567 sn->date_sec = sn_info->date_sec;
1568 sn->date_nsec = sn_info->date_nsec;
1569 sn->vm_clock_nsec = sn_info->vm_clock_nsec;
1570
1571 ret = update_snapshot_refcount(bs, s->l1_table_offset, s->l1_size, 1);
1572 if (ret < 0)
1573 goto fail;
1574
1575 /* create the L1 table of the snapshot */
1576 sn->l1_table_offset = alloc_clusters(bs, s->l1_size * sizeof(uint64_t));
1577 sn->l1_size = s->l1_size;
1578
1579 l1_table = qemu_malloc(s->l1_size * sizeof(uint64_t));
1580 if (!l1_table)
1581 goto fail;
1582 for(i = 0; i < s->l1_size; i++) {
1583 l1_table[i] = cpu_to_be64(s->l1_table[i]);
1584 }
1585 if (bdrv_pwrite(s->hd, sn->l1_table_offset,
1586 l1_table, s->l1_size * sizeof(uint64_t)) !=
1587 (s->l1_size * sizeof(uint64_t)))
1588 goto fail;
1589 qemu_free(l1_table);
1590 l1_table = NULL;
1591
1592 snapshots1 = qemu_malloc((s->nb_snapshots + 1) * sizeof(QCowSnapshot));
1593 if (!snapshots1)
1594 goto fail;
1595 memcpy(snapshots1, s->snapshots, s->nb_snapshots * sizeof(QCowSnapshot));
1596 s->snapshots = snapshots1;
1597 s->snapshots[s->nb_snapshots++] = *sn;
1598
1599 if (qcow_write_snapshots(bs) < 0)
1600 goto fail;
1601#ifdef DEBUG_ALLOC
1602 check_refcounts(bs);
1603#endif
1604 return 0;
1605 fail:
1606 qemu_free(sn->name);
1607 qemu_free(l1_table);
1608 return -1;
1609}
1610
1611/* copy the snapshot 'snapshot_name' into the current disk image */
1612static int qcow_snapshot_goto(BlockDriverState *bs,
1613 const char *snapshot_id)
1614{
1615 BDRVQcowState *s = bs->opaque;
1616 QCowSnapshot *sn;
1617 int i, snapshot_index, l1_size2;
1618
1619 snapshot_index = find_snapshot_by_id_or_name(bs, snapshot_id);
1620 if (snapshot_index < 0)
1621 return -ENOENT;
1622 sn = &s->snapshots[snapshot_index];
1623
1624 if (update_snapshot_refcount(bs, s->l1_table_offset, s->l1_size, -1) < 0)
1625 goto fail;
1626
1627 if (grow_l1_table(bs, sn->l1_size) < 0)
1628 goto fail;
1629
1630 s->l1_size = sn->l1_size;
1631 l1_size2 = s->l1_size * sizeof(uint64_t);
1632 /* copy the snapshot l1 table to the current l1 table */
1633 if (bdrv_pread(s->hd, sn->l1_table_offset,
1634 s->l1_table, l1_size2) != l1_size2)
1635 goto fail;
1636 if (bdrv_pwrite(s->hd, s->l1_table_offset,
1637 s->l1_table, l1_size2) != l1_size2)
1638 goto fail;
1639 for(i = 0;i < s->l1_size; i++) {
1640 be64_to_cpus(&s->l1_table[i]);
1641 }
1642
1643 if (update_snapshot_refcount(bs, s->l1_table_offset, s->l1_size, 1) < 0)
1644 goto fail;
1645
1646#ifdef DEBUG_ALLOC
1647 check_refcounts(bs);
1648#endif
1649 return 0;
1650 fail:
1651 return -EIO;
1652}
1653
1654static int qcow_snapshot_delete(BlockDriverState *bs, const char *snapshot_id)
1655{
1656 BDRVQcowState *s = bs->opaque;
1657 QCowSnapshot *sn;
1658 int snapshot_index, ret;
1659
1660 snapshot_index = find_snapshot_by_id_or_name(bs, snapshot_id);
1661 if (snapshot_index < 0)
1662 return -ENOENT;
1663 sn = &s->snapshots[snapshot_index];
1664
1665 ret = update_snapshot_refcount(bs, sn->l1_table_offset, sn->l1_size, -1);
1666 if (ret < 0)
1667 return ret;
1668 /* must update the copied flag on the current cluster offsets */
1669 ret = update_snapshot_refcount(bs, s->l1_table_offset, s->l1_size, 0);
1670 if (ret < 0)
1671 return ret;
1672 free_clusters(bs, sn->l1_table_offset, sn->l1_size * sizeof(uint64_t));
1673
1674 qemu_free(sn->id_str);
1675 qemu_free(sn->name);
1676 memmove(sn, sn + 1, (s->nb_snapshots - snapshot_index - 1) * sizeof(*sn));
1677 s->nb_snapshots--;
1678 ret = qcow_write_snapshots(bs);
1679 if (ret < 0) {
1680 /* XXX: restore snapshot if error ? */
1681 return ret;
1682 }
1683#ifdef DEBUG_ALLOC
1684 check_refcounts(bs);
1685#endif
1686 return 0;
1687}
1688
1689static int qcow_snapshot_list(BlockDriverState *bs,
1690 QEMUSnapshotInfo **psn_tab)
1691{
1692 BDRVQcowState *s = bs->opaque;
1693 QEMUSnapshotInfo *sn_tab, *sn_info;
1694 QCowSnapshot *sn;
1695 int i;
1696
1697 sn_tab = qemu_mallocz(s->nb_snapshots * sizeof(QEMUSnapshotInfo));
1698 if (!sn_tab)
1699 goto fail;
1700 for(i = 0; i < s->nb_snapshots; i++) {
1701 sn_info = sn_tab + i;
1702 sn = s->snapshots + i;
1703 pstrcpy(sn_info->id_str, sizeof(sn_info->id_str),
1704 sn->id_str);
1705 pstrcpy(sn_info->name, sizeof(sn_info->name),
1706 sn->name);
1707 sn_info->vm_state_size = sn->vm_state_size;
1708 sn_info->date_sec = sn->date_sec;
1709 sn_info->date_nsec = sn->date_nsec;
1710 sn_info->vm_clock_nsec = sn->vm_clock_nsec;
1711 }
1712 *psn_tab = sn_tab;
1713 return s->nb_snapshots;
1714 fail:
1715 qemu_free(sn_tab);
1716 *psn_tab = NULL;
1717 return -ENOMEM;
1718}
1719
1720/*********************************************************/
1721/* refcount handling */
1722
1723static int refcount_init(BlockDriverState *bs)
1724{
1725 BDRVQcowState *s = bs->opaque;
1726 int ret, refcount_table_size2, i;
1727
1728 s->refcount_block_cache = qemu_malloc(s->cluster_size);
1729 if (!s->refcount_block_cache)
1730 goto fail;
1731 refcount_table_size2 = s->refcount_table_size * sizeof(uint64_t);
1732 s->refcount_table = qemu_malloc(refcount_table_size2);
1733 if (!s->refcount_table)
1734 goto fail;
1735 if (s->refcount_table_size > 0) {
1736 ret = bdrv_pread(s->hd, s->refcount_table_offset,
1737 s->refcount_table, refcount_table_size2);
1738 if (ret != refcount_table_size2)
1739 goto fail;
1740 for(i = 0; i < s->refcount_table_size; i++)
1741 be64_to_cpus(&s->refcount_table[i]);
1742 }
1743 return 0;
1744 fail:
1745 return -ENOMEM;
1746}
1747
1748static void refcount_close(BlockDriverState *bs)
1749{
1750 BDRVQcowState *s = bs->opaque;
1751 qemu_free(s->refcount_block_cache);
1752 qemu_free(s->refcount_table);
1753}
1754
1755
1756static int load_refcount_block(BlockDriverState *bs,
1757 int64_t refcount_block_offset)
1758{
1759 BDRVQcowState *s = bs->opaque;
1760 int ret;
1761 ret = bdrv_pread(s->hd, refcount_block_offset, s->refcount_block_cache,
1762 s->cluster_size);
1763 if (ret != s->cluster_size)
1764 return -EIO;
1765 s->refcount_block_cache_offset = refcount_block_offset;
1766 return 0;
1767}
1768
1769static int get_refcount(BlockDriverState *bs, int64_t cluster_index)
1770{
1771 BDRVQcowState *s = bs->opaque;
1772 int refcount_table_index, block_index;
1773 int64_t refcount_block_offset;
1774
1775 refcount_table_index = cluster_index >> (s->cluster_bits - REFCOUNT_SHIFT);
1776 if (refcount_table_index >= s->refcount_table_size)
1777 return 0;
1778 refcount_block_offset = s->refcount_table[refcount_table_index];
1779 if (!refcount_block_offset)
1780 return 0;
1781 if (refcount_block_offset != s->refcount_block_cache_offset) {
1782 /* better than nothing: return allocated if read error */
1783 if (load_refcount_block(bs, refcount_block_offset) < 0)
1784 return 1;
1785 }
1786 block_index = cluster_index &
1787 ((1 << (s->cluster_bits - REFCOUNT_SHIFT)) - 1);
1788 return be16_to_cpu(s->refcount_block_cache[block_index]);
1789}
1790
1791/* return < 0 if error */
1792static int64_t alloc_clusters_noref(BlockDriverState *bs, int64_t size)
1793{
1794 BDRVQcowState *s = bs->opaque;
1795 int i, nb_clusters;
1796
1797 nb_clusters = (size + s->cluster_size - 1) >> s->cluster_bits;
1798 for(;;) {
1799 if (get_refcount(bs, s->free_cluster_index) == 0) {
1800 s->free_cluster_index++;
1801 for(i = 1; i < nb_clusters; i++) {
1802 if (get_refcount(bs, s->free_cluster_index) != 0)
1803 goto not_found;
1804 s->free_cluster_index++;
1805 }
1806#ifdef DEBUG_ALLOC2
1807 printf("alloc_clusters: size=%lld -> %lld\n",
1808 size,
1809 (s->free_cluster_index - nb_clusters) << s->cluster_bits);
1810#endif
1811 return (s->free_cluster_index - nb_clusters) << s->cluster_bits;
1812 } else {
1813 not_found:
1814 s->free_cluster_index++;
1815 }
1816 }
1817}
1818
1819static int64_t alloc_clusters(BlockDriverState *bs, int64_t size)
1820{
1821 int64_t offset;
1822
1823 offset = alloc_clusters_noref(bs, size);
1824 update_refcount(bs, offset, size, 1);
1825 return offset;
1826}
1827
1828/* only used to allocate compressed sectors. We try to allocate
1829 contiguous sectors. size must be <= cluster_size */
1830static int64_t alloc_bytes(BlockDriverState *bs, int size)
1831{
1832 BDRVQcowState *s = bs->opaque;
1833 int64_t offset, cluster_offset;
1834 int free_in_cluster;
1835
1836 assert(size > 0 && size <= s->cluster_size);
1837 if (s->free_byte_offset == 0) {
1838 s->free_byte_offset = alloc_clusters(bs, s->cluster_size);
1839 }
1840 redo:
1841 free_in_cluster = s->cluster_size -
1842 (s->free_byte_offset & (s->cluster_size - 1));
1843 if (size <= free_in_cluster) {
1844 /* enough space in current cluster */
1845 offset = s->free_byte_offset;
1846 s->free_byte_offset += size;
1847 free_in_cluster -= size;
1848 if (free_in_cluster == 0)
1849 s->free_byte_offset = 0;
1850 if ((offset & (s->cluster_size - 1)) != 0)
1851 update_cluster_refcount(bs, offset >> s->cluster_bits, 1);
1852 } else {
1853 offset = alloc_clusters(bs, s->cluster_size);
1854 cluster_offset = s->free_byte_offset & ~(s->cluster_size - 1);
1855 if ((cluster_offset + s->cluster_size) == offset) {
1856 /* we are lucky: contiguous data */
1857 offset = s->free_byte_offset;
1858 update_cluster_refcount(bs, offset >> s->cluster_bits, 1);
1859 s->free_byte_offset += size;
1860 } else {
1861 s->free_byte_offset = offset;
1862 goto redo;
1863 }
1864 }
1865 return offset;
1866}
1867
1868static void free_clusters(BlockDriverState *bs,
1869 int64_t offset, int64_t size)
1870{
1871 update_refcount(bs, offset, size, -1);
1872}
1873
1874static int grow_refcount_table(BlockDriverState *bs, int min_size)
1875{
1876 BDRVQcowState *s = bs->opaque;
1877 int new_table_size, new_table_size2, refcount_table_clusters, i, ret;
1878 uint64_t *new_table;
1879 int64_t table_offset;
1880 uint64_t data64;
1881 uint32_t data32;
1882
1883 if (min_size <= s->refcount_table_size)
1884 return 0;
1885 /* compute new table size */
1886 refcount_table_clusters = s->refcount_table_size >> (s->cluster_bits - 3);
1887 for(;;) {
1888 if (refcount_table_clusters == 0) {
1889 refcount_table_clusters = 1;
1890 } else {
1891 refcount_table_clusters = (refcount_table_clusters * 3 + 1) / 2;
1892 }
1893 new_table_size = refcount_table_clusters << (s->cluster_bits - 3);
1894 if (min_size <= new_table_size)
1895 break;
1896 }
15e6690a
FB		 1897#ifdef DEBUG_ALLOC2
1898 printf("grow_refcount_table from %d to %d\n",
1899 s->refcount_table_size,
1900 new_table_size);
1901#endif
585f8587
FB		 1902 new_table_size2 = new_table_size * sizeof(uint64_t);
1903 new_table = qemu_mallocz(new_table_size2);
1904 if (!new_table)
1905 return -ENOMEM;
1906 memcpy(new_table, s->refcount_table,
1907 s->refcount_table_size * sizeof(uint64_t));
1908 for(i = 0; i < s->refcount_table_size; i++)
1909 cpu_to_be64s(&new_table[i]);
1910 /* Note: we cannot update the refcount now to avoid recursion */
1911 table_offset = alloc_clusters_noref(bs, new_table_size2);
1912 ret = bdrv_pwrite(s->hd, table_offset, new_table, new_table_size2);
1913 if (ret != new_table_size2)
1914 goto fail;
1915 for(i = 0; i < s->refcount_table_size; i++)
1916 be64_to_cpus(&new_table[i]);
1917
1918 data64 = cpu_to_be64(table_offset);
1919 if (bdrv_pwrite(s->hd, offsetof(QCowHeader, refcount_table_offset),
1920 &data64, sizeof(data64)) != sizeof(data64))
1921 goto fail;
1922 data32 = cpu_to_be32(refcount_table_clusters);
1923 if (bdrv_pwrite(s->hd, offsetof(QCowHeader, refcount_table_clusters),
1924 &data32, sizeof(data32)) != sizeof(data32))
1925 goto fail;
1926 qemu_free(s->refcount_table);
1927 s->refcount_table = new_table;
1928 s->refcount_table_size = new_table_size;
1929
1930 update_refcount(bs, table_offset, new_table_size2, 1);
1931 return 0;
1932 fail:
1933 free_clusters(bs, table_offset, new_table_size2);
1934 qemu_free(new_table);
1935 return -EIO;
1936}
1937
1938/* addend must be 1 or -1 */
1939/* XXX: cache several refcount block clusters ? */
1940static int update_cluster_refcount(BlockDriverState *bs,
1941 int64_t cluster_index,
1942 int addend)
1943{
1944 BDRVQcowState *s = bs->opaque;
1945 int64_t offset, refcount_block_offset;
1946 int ret, refcount_table_index, block_index, refcount;
1947 uint64_t data64;
1948
1949 refcount_table_index = cluster_index >> (s->cluster_bits - REFCOUNT_SHIFT);
1950 if (refcount_table_index >= s->refcount_table_size) {
1951 if (addend < 0)
1952 return -EINVAL;
1953 ret = grow_refcount_table(bs, refcount_table_index + 1);
1954 if (ret < 0)
1955 return ret;
1956 }
1957 refcount_block_offset = s->refcount_table[refcount_table_index];
1958 if (!refcount_block_offset) {
1959 if (addend < 0)
1960 return -EINVAL;
1961 /* create a new refcount block */
1962 /* Note: we cannot update the refcount now to avoid recursion */
1963 offset = alloc_clusters_noref(bs, s->cluster_size);
1964 memset(s->refcount_block_cache, 0, s->cluster_size);
1965 ret = bdrv_pwrite(s->hd, offset, s->refcount_block_cache, s->cluster_size);
1966 if (ret != s->cluster_size)
1967 return -EINVAL;
1968 s->refcount_table[refcount_table_index] = offset;
1969 data64 = cpu_to_be64(offset);
1970 ret = bdrv_pwrite(s->hd, s->refcount_table_offset +
1971 refcount_table_index * sizeof(uint64_t),
1972 &data64, sizeof(data64));
1973 if (ret != sizeof(data64))
1974 return -EINVAL;
1975
1976 refcount_block_offset = offset;
1977 s->refcount_block_cache_offset = offset;
1978 update_refcount(bs, offset, s->cluster_size, 1);
1979 } else {
1980 if (refcount_block_offset != s->refcount_block_cache_offset) {
1981 if (load_refcount_block(bs, refcount_block_offset) < 0)
1982 return -EIO;
1983 }
1984 }
1985 /* we can update the count and save it */
1986 block_index = cluster_index &
1987 ((1 << (s->cluster_bits - REFCOUNT_SHIFT)) - 1);
1988 refcount = be16_to_cpu(s->refcount_block_cache[block_index]);
1989 refcount += addend;
1990 if (refcount < 0 || refcount > 0xffff)
1991 return -EINVAL;
1992 if (refcount == 0 && cluster_index < s->free_cluster_index) {
1993 s->free_cluster_index = cluster_index;
1994 }
1995 s->refcount_block_cache[block_index] = cpu_to_be16(refcount);
1996 if (bdrv_pwrite(s->hd,
1997 refcount_block_offset + (block_index << REFCOUNT_SHIFT),
1998 &s->refcount_block_cache[block_index], 2) != 2)
1999 return -EIO;
2000 return refcount;
2001}
2002
2003static void update_refcount(BlockDriverState *bs,
2004 int64_t offset, int64_t length,
2005 int addend)
2006{
2007 BDRVQcowState *s = bs->opaque;
2008 int64_t start, last, cluster_offset;
2009
2010#ifdef DEBUG_ALLOC2
2011 printf("update_refcount: offset=%lld size=%lld addend=%d\n",
2012 offset, length, addend);
2013#endif
2014 if (length <= 0)
2015 return;
2016 start = offset & ~(s->cluster_size - 1);
2017 last = (offset + length - 1) & ~(s->cluster_size - 1);
2018 for(cluster_offset = start; cluster_offset <= last;
2019 cluster_offset += s->cluster_size) {
2020 update_cluster_refcount(bs, cluster_offset >> s->cluster_bits, addend);
2021 }
2022}
2023
2024#ifdef DEBUG_ALLOC
2025static void inc_refcounts(BlockDriverState *bs,
2026 uint16_t *refcount_table,
2027 int refcount_table_size,
2028 int64_t offset, int64_t size)
2029{
2030 BDRVQcowState *s = bs->opaque;
2031 int64_t start, last, cluster_offset;
2032 int k;
2033
2034 if (size <= 0)
2035 return;
2036
2037 start = offset & ~(s->cluster_size - 1);
2038 last = (offset + size - 1) & ~(s->cluster_size - 1);
2039 for(cluster_offset = start; cluster_offset <= last;
2040 cluster_offset += s->cluster_size) {
2041 k = cluster_offset >> s->cluster_bits;
2042 if (k < 0 || k >= refcount_table_size) {
2043 printf("ERROR: invalid cluster offset=0x%llx\n", cluster_offset);
2044 } else {
2045 if (++refcount_table[k] == 0) {
2046 printf("ERROR: overflow cluster offset=0x%llx\n", cluster_offset);
2047 }
2048 }
2049 }
2050}
2051
2052static int check_refcounts_l1(BlockDriverState *bs,
2053 uint16_t *refcount_table,
2054 int refcount_table_size,
2055 int64_t l1_table_offset, int l1_size,
2056 int check_copied)
2057{
2058 BDRVQcowState *s = bs->opaque;
2059 uint64_t *l1_table, *l2_table, l2_offset, offset, l1_size2;
2060 int l2_size, i, j, nb_csectors, refcount;
2061
2062 l2_table = NULL;
2063 l1_size2 = l1_size * sizeof(uint64_t);
2064
2065 inc_refcounts(bs, refcount_table, refcount_table_size,
2066 l1_table_offset, l1_size2);
2067
2068 l1_table = qemu_malloc(l1_size2);
2069 if (!l1_table)
2070 goto fail;
2071 if (bdrv_pread(s->hd, l1_table_offset,
2072 l1_table, l1_size2) != l1_size2)
2073 goto fail;
2074 for(i = 0;i < l1_size; i++)
2075 be64_to_cpus(&l1_table[i]);
2076
2077 l2_size = s->l2_size * sizeof(uint64_t);
2078 l2_table = qemu_malloc(l2_size);
2079 if (!l2_table)
2080 goto fail;
2081 for(i = 0; i < l1_size; i++) {
2082 l2_offset = l1_table[i];
2083 if (l2_offset) {
2084 if (check_copied) {
2085 refcount = get_refcount(bs, (l2_offset & ~QCOW_OFLAG_COPIED) >> s->cluster_bits);
2086 if ((refcount == 1) != ((l2_offset & QCOW_OFLAG_COPIED) != 0)) {
2087 printf("ERROR OFLAG_COPIED: l2_offset=%llx refcount=%d\n",
2088 l2_offset, refcount);
2089 }
2090 }
2091 l2_offset &= ~QCOW_OFLAG_COPIED;
2092 if (bdrv_pread(s->hd, l2_offset, l2_table, l2_size) != l2_size)
2093 goto fail;
2094 for(j = 0; j < s->l2_size; j++) {
2095 offset = be64_to_cpu(l2_table[j]);
2096 if (offset != 0) {
2097 if (offset & QCOW_OFLAG_COMPRESSED) {
2098 if (offset & QCOW_OFLAG_COPIED) {
2099 printf("ERROR: cluster %lld: copied flag must never be set for compressed clusters\n",
2100 offset >> s->cluster_bits);
2101 offset &= ~QCOW_OFLAG_COPIED;
2102 }
2103 nb_csectors = ((offset >> s->csize_shift) &
2104 s->csize_mask) + 1;
2105 offset &= s->cluster_offset_mask;
2106 inc_refcounts(bs, refcount_table,
2107 refcount_table_size,
2108 offset & ~511, nb_csectors * 512);
2109 } else {
2110 if (check_copied) {
2111 refcount = get_refcount(bs, (offset & ~QCOW_OFLAG_COPIED) >> s->cluster_bits);
2112 if ((refcount == 1) != ((offset & QCOW_OFLAG_COPIED) != 0)) {
2113 printf("ERROR OFLAG_COPIED: offset=%llx refcount=%d\n",
2114 offset, refcount);
2115 }
2116 }
2117 offset &= ~QCOW_OFLAG_COPIED;
2118 inc_refcounts(bs, refcount_table,
2119 refcount_table_size,
2120 offset, s->cluster_size);
2121 }
2122 }
2123 }
2124 inc_refcounts(bs, refcount_table,
2125 refcount_table_size,
2126 l2_offset,
2127 s->cluster_size);
2128 }
2129 }
2130 qemu_free(l1_table);
2131 qemu_free(l2_table);
2132 return 0;
2133 fail:
2134 printf("ERROR: I/O error in check_refcounts_l1\n");
2135 qemu_free(l1_table);
2136 qemu_free(l2_table);
2137 return -EIO;
2138}
2139
2140static void check_refcounts(BlockDriverState *bs)
2141{
2142 BDRVQcowState *s = bs->opaque;
2143 int64_t size;
2144 int nb_clusters, refcount1, refcount2, i;
2145 QCowSnapshot *sn;
2146 uint16_t *refcount_table;
2147
2148 size = bdrv_getlength(s->hd);
2149 nb_clusters = (size + s->cluster_size - 1) >> s->cluster_bits;
2150 refcount_table = qemu_mallocz(nb_clusters * sizeof(uint16_t));
15e6690a 2151
585f8587
FB		 2152 /* header */
2153 inc_refcounts(bs, refcount_table, nb_clusters,
2154 0, s->cluster_size);
2155
2156 check_refcounts_l1(bs, refcount_table, nb_clusters,
2157 s->l1_table_offset, s->l1_size, 1);
2158
2159 /* snapshots */
2160 for(i = 0; i < s->nb_snapshots; i++) {
2161 sn = s->snapshots + i;
2162 check_refcounts_l1(bs, refcount_table, nb_clusters,
2163 sn->l1_table_offset, sn->l1_size, 0);
2164 }
2165 inc_refcounts(bs, refcount_table, nb_clusters,
2166 s->snapshots_offset, s->snapshots_size);
2167
2168 /* refcount data */
2169 inc_refcounts(bs, refcount_table, nb_clusters,
2170 s->refcount_table_offset,
2171 s->refcount_table_size * sizeof(uint64_t));
2172 for(i = 0; i < s->refcount_table_size; i++) {
2173 int64_t offset;
2174 offset = s->refcount_table[i];
2175 if (offset != 0) {
2176 inc_refcounts(bs, refcount_table, nb_clusters,
2177 offset, s->cluster_size);
2178 }
2179 }
2180
2181 /* compare ref counts */
2182 for(i = 0; i < nb_clusters; i++) {
2183 refcount1 = get_refcount(bs, i);
2184 refcount2 = refcount_table[i];
2185 if (refcount1 != refcount2)
2186 printf("ERROR cluster %d refcount=%d reference=%d\n",
2187 i, refcount1, refcount2);
2188 }
2189
2190 qemu_free(refcount_table);
2191}
2192
2193#if 0
2194static void dump_refcounts(BlockDriverState *bs)
2195{
2196 BDRVQcowState *s = bs->opaque;
2197 int64_t nb_clusters, k, k1, size;
2198 int refcount;
2199
2200 size = bdrv_getlength(s->hd);
2201 nb_clusters = (size + s->cluster_size - 1) >> s->cluster_bits;
2202 for(k = 0; k < nb_clusters;) {
2203 k1 = k;
2204 refcount = get_refcount(bs, k);
2205 k++;
2206 while (k < nb_clusters && get_refcount(bs, k) == refcount)
2207 k++;
2208 printf("%lld: refcount=%d nb=%lld\n", k, refcount, k - k1);
2209 }
2210}
2211#endif
2212#endif
2213
2214BlockDriver bdrv_qcow2 = {
2215 "qcow2",
2216 sizeof(BDRVQcowState),
2217 qcow_probe,
2218 qcow_open,
2219 NULL,
2220 NULL,
2221 qcow_close,
2222 qcow_create,
2223 qcow_flush,
2224 qcow_is_allocated,
2225 qcow_set_key,
2226 qcow_make_empty,
2227
2228 .bdrv_aio_new = qcow_aio_new,
2229 .bdrv_aio_read = qcow_aio_read,
2230 .bdrv_aio_write = qcow_aio_write,
2231 .bdrv_aio_cancel = qcow_aio_cancel,
2232 .bdrv_aio_delete = qcow_aio_delete,
2233 .bdrv_write_compressed = qcow_write_compressed,
2234
2235 .bdrv_snapshot_create = qcow_snapshot_create,
2236 .bdrv_snapshot_goto = qcow_snapshot_goto,
2237 .bdrv_snapshot_delete = qcow_snapshot_delete,
2238 .bdrv_snapshot_list = qcow_snapshot_list,
2239 .bdrv_get_info = qcow_get_info,
2240};
Qemu copy for testing