2009-05-28 16:07:05 +02:00
|
|
|
/*
|
|
|
|
* Block driver for the QCOW version 2 format
|
|
|
|
*
|
|
|
|
* Copyright (c) 2004-2006 Fabrice Bellard
|
|
|
|
*
|
|
|
|
* Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
|
|
* of this software and associated documentation files (the "Software"), to deal
|
|
|
|
* in the Software without restriction, including without limitation the rights
|
|
|
|
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|
|
|
* copies of the Software, and to permit persons to whom the Software is
|
|
|
|
* furnished to do so, subject to the following conditions:
|
|
|
|
*
|
|
|
|
* The above copyright notice and this permission notice shall be included in
|
|
|
|
* all copies or substantial portions of the Software.
|
|
|
|
*
|
|
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
|
|
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
|
|
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
|
|
|
|
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
|
|
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
|
|
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
|
|
|
|
* THE SOFTWARE.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <zlib.h>
|
|
|
|
|
|
|
|
#include "qemu-common.h"
|
|
|
|
#include "block_int.h"
|
|
|
|
#include "block/qcow2.h"
|
|
|
|
|
2010-10-18 17:53:53 +02:00
|
|
|
int qcow2_grow_l1_table(BlockDriverState *bs, int min_size, bool exact_size)
|
2009-05-28 16:07:05 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
int new_l1_size, new_l1_size2, ret, i;
|
|
|
|
uint64_t *new_l1_table;
|
2010-01-20 15:04:01 +01:00
|
|
|
int64_t new_l1_table_offset;
|
2009-05-28 16:07:05 +02:00
|
|
|
uint8_t data[12];
|
|
|
|
|
2010-10-18 17:53:53 +02:00
|
|
|
if (min_size <= s->l1_size)
|
2009-05-28 16:07:05 +02:00
|
|
|
return 0;
|
2010-10-18 17:53:53 +02:00
|
|
|
|
|
|
|
if (exact_size) {
|
|
|
|
new_l1_size = min_size;
|
|
|
|
} else {
|
|
|
|
/* Bump size up to reduce the number of times we have to grow */
|
|
|
|
new_l1_size = s->l1_size;
|
|
|
|
if (new_l1_size == 0) {
|
|
|
|
new_l1_size = 1;
|
|
|
|
}
|
|
|
|
while (min_size > new_l1_size) {
|
|
|
|
new_l1_size = (new_l1_size * 3 + 1) / 2;
|
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
}
|
2010-10-18 17:53:53 +02:00
|
|
|
|
2009-05-28 16:07:05 +02:00
|
|
|
#ifdef DEBUG_ALLOC2
|
2011-08-25 09:23:55 +02:00
|
|
|
fprintf(stderr, "grow l1_table from %d to %d\n", s->l1_size, new_l1_size);
|
2009-05-28 16:07:05 +02:00
|
|
|
#endif
|
|
|
|
|
|
|
|
new_l1_size2 = sizeof(uint64_t) * new_l1_size;
|
2011-08-21 05:09:37 +02:00
|
|
|
new_l1_table = g_malloc0(align_offset(new_l1_size2, 512));
|
2009-05-28 16:07:05 +02:00
|
|
|
memcpy(new_l1_table, s->l1_table, s->l1_size * sizeof(uint64_t));
|
|
|
|
|
|
|
|
/* write new table (align to cluster) */
|
2010-04-14 14:17:38 +02:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L1_GROW_ALLOC_TABLE);
|
2009-05-28 16:07:07 +02:00
|
|
|
new_l1_table_offset = qcow2_alloc_clusters(bs, new_l1_size2);
|
2010-01-20 15:04:01 +01:00
|
|
|
if (new_l1_table_offset < 0) {
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(new_l1_table);
|
2010-01-20 15:04:01 +01:00
|
|
|
return new_l1_table_offset;
|
|
|
|
}
|
2011-01-10 17:17:28 +01:00
|
|
|
|
|
|
|
ret = qcow2_cache_flush(bs, s->refcount_block_cache);
|
|
|
|
if (ret < 0) {
|
2011-06-01 10:50:00 +02:00
|
|
|
goto fail;
|
2011-01-10 17:17:28 +01:00
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
|
2010-04-14 14:17:38 +02:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L1_GROW_WRITE_TABLE);
|
2009-05-28 16:07:05 +02:00
|
|
|
for(i = 0; i < s->l1_size; i++)
|
|
|
|
new_l1_table[i] = cpu_to_be64(new_l1_table[i]);
|
2010-06-16 17:44:35 +02:00
|
|
|
ret = bdrv_pwrite_sync(bs->file, new_l1_table_offset, new_l1_table, new_l1_size2);
|
|
|
|
if (ret < 0)
|
2009-05-28 16:07:05 +02:00
|
|
|
goto fail;
|
|
|
|
for(i = 0; i < s->l1_size; i++)
|
|
|
|
new_l1_table[i] = be64_to_cpu(new_l1_table[i]);
|
|
|
|
|
|
|
|
/* set new table */
|
2010-04-14 14:17:38 +02:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L1_GROW_ACTIVATE_TABLE);
|
2009-05-28 16:07:05 +02:00
|
|
|
cpu_to_be32w((uint32_t*)data, new_l1_size);
|
2011-01-01 21:50:34 +01:00
|
|
|
cpu_to_be64wu((uint64_t*)(data + 4), new_l1_table_offset);
|
2010-06-16 17:44:35 +02:00
|
|
|
ret = bdrv_pwrite_sync(bs->file, offsetof(QCowHeader, l1_size), data,sizeof(data));
|
|
|
|
if (ret < 0) {
|
2009-05-28 16:07:05 +02:00
|
|
|
goto fail;
|
2010-01-20 15:02:58 +01:00
|
|
|
}
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(s->l1_table);
|
2009-05-28 16:07:07 +02:00
|
|
|
qcow2_free_clusters(bs, s->l1_table_offset, s->l1_size * sizeof(uint64_t));
|
2009-05-28 16:07:05 +02:00
|
|
|
s->l1_table_offset = new_l1_table_offset;
|
|
|
|
s->l1_table = new_l1_table;
|
|
|
|
s->l1_size = new_l1_size;
|
|
|
|
return 0;
|
|
|
|
fail:
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(new_l1_table);
|
2010-01-20 15:02:58 +01:00
|
|
|
qcow2_free_clusters(bs, new_l1_table_offset, new_l1_size2);
|
2010-06-16 17:44:35 +02:00
|
|
|
return ret;
|
2009-05-28 16:07:05 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* l2_load
|
|
|
|
*
|
|
|
|
* Loads a L2 table into memory. If the table is in the cache, the cache
|
|
|
|
* is used; otherwise the L2 table is loaded from the image file.
|
|
|
|
*
|
|
|
|
* Returns a pointer to the L2 table on success, or NULL if the read from
|
|
|
|
* the image file failed.
|
|
|
|
*/
|
|
|
|
|
2010-05-21 18:25:20 +02:00
|
|
|
static int l2_load(BlockDriverState *bs, uint64_t l2_offset,
|
|
|
|
uint64_t **l2_table)
|
2009-05-28 16:07:05 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
2010-05-21 18:25:20 +02:00
|
|
|
int ret;
|
2009-05-28 16:07:05 +02:00
|
|
|
|
2011-01-10 17:17:28 +01:00
|
|
|
ret = qcow2_cache_get(bs, s->l2_table_cache, l2_offset, (void**) l2_table);
|
2009-05-28 16:07:05 +02:00
|
|
|
|
2011-01-10 17:17:28 +01:00
|
|
|
return ret;
|
2009-05-28 16:07:05 +02:00
|
|
|
}
|
|
|
|
|
2009-06-16 11:31:28 +02:00
|
|
|
/*
|
|
|
|
* Writes one sector of the L1 table to the disk (can't update single entries
|
|
|
|
* and we really don't want bdrv_pread to perform a read-modify-write)
|
|
|
|
*/
|
|
|
|
#define L1_ENTRIES_PER_SECTOR (512 / 8)
|
2010-04-14 14:17:38 +02:00
|
|
|
static int write_l1_entry(BlockDriverState *bs, int l1_index)
|
2009-06-16 11:31:28 +02:00
|
|
|
{
|
2010-04-14 14:17:38 +02:00
|
|
|
BDRVQcowState *s = bs->opaque;
|
2009-06-16 11:31:28 +02:00
|
|
|
uint64_t buf[L1_ENTRIES_PER_SECTOR];
|
|
|
|
int l1_start_index;
|
2010-03-23 17:28:22 +01:00
|
|
|
int i, ret;
|
2009-06-16 11:31:28 +02:00
|
|
|
|
|
|
|
l1_start_index = l1_index & ~(L1_ENTRIES_PER_SECTOR - 1);
|
|
|
|
for (i = 0; i < L1_ENTRIES_PER_SECTOR; i++) {
|
|
|
|
buf[i] = cpu_to_be64(s->l1_table[l1_start_index + i]);
|
|
|
|
}
|
|
|
|
|
2010-04-14 14:17:38 +02:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L1_UPDATE);
|
2010-06-16 17:44:35 +02:00
|
|
|
ret = bdrv_pwrite_sync(bs->file, s->l1_table_offset + 8 * l1_start_index,
|
2010-03-23 17:28:22 +01:00
|
|
|
buf, sizeof(buf));
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
2009-06-16 11:31:28 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2009-05-28 16:07:05 +02:00
|
|
|
/*
|
|
|
|
* l2_allocate
|
|
|
|
*
|
|
|
|
* Allocate a new l2 entry in the file. If l1_index points to an already
|
|
|
|
* used entry in the L2 table (i.e. we are doing a copy on write for the L2
|
|
|
|
* table) copy the contents of the old L2 table into the newly allocated one.
|
|
|
|
* Otherwise the new table is initialized with zeros.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
2010-03-23 17:41:24 +01:00
|
|
|
static int l2_allocate(BlockDriverState *bs, int l1_index, uint64_t **table)
|
2009-05-28 16:07:05 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
2009-06-16 11:31:28 +02:00
|
|
|
uint64_t old_l2_offset;
|
2010-02-02 15:20:57 +01:00
|
|
|
uint64_t *l2_table;
|
|
|
|
int64_t l2_offset;
|
2010-03-23 17:41:24 +01:00
|
|
|
int ret;
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
old_l2_offset = s->l1_table[l1_index];
|
|
|
|
|
|
|
|
/* allocate a new l2 entry */
|
|
|
|
|
2009-05-28 16:07:07 +02:00
|
|
|
l2_offset = qcow2_alloc_clusters(bs, s->l2_size * sizeof(uint64_t));
|
2010-01-20 15:04:01 +01:00
|
|
|
if (l2_offset < 0) {
|
2010-03-23 17:41:24 +01:00
|
|
|
return l2_offset;
|
2010-01-20 15:04:01 +01:00
|
|
|
}
|
2011-01-10 17:17:28 +01:00
|
|
|
|
|
|
|
ret = qcow2_cache_flush(bs, s->refcount_block_cache);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
/* allocate a new entry in the l2 cache */
|
|
|
|
|
2011-01-10 17:17:28 +01:00
|
|
|
ret = qcow2_cache_get_empty(bs, s->l2_table_cache, l2_offset, (void**) table);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
l2_table = *table;
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
if (old_l2_offset == 0) {
|
|
|
|
/* if there was no old l2 table, clear the new table */
|
|
|
|
memset(l2_table, 0, s->l2_size * sizeof(uint64_t));
|
|
|
|
} else {
|
2011-01-10 17:17:28 +01:00
|
|
|
uint64_t* old_table;
|
|
|
|
|
2009-05-28 16:07:05 +02:00
|
|
|
/* if there was an old l2 table, read it from the disk */
|
2010-04-14 14:17:38 +02:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L2_ALLOC_COW_READ);
|
2011-01-10 17:17:28 +01:00
|
|
|
ret = qcow2_cache_get(bs, s->l2_table_cache, old_l2_offset,
|
|
|
|
(void**) &old_table);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
memcpy(l2_table, old_table, s->cluster_size);
|
|
|
|
|
|
|
|
ret = qcow2_cache_put(bs, s->l2_table_cache, (void**) &old_table);
|
2010-03-23 17:41:24 +01:00
|
|
|
if (ret < 0) {
|
2010-05-12 16:23:26 +02:00
|
|
|
goto fail;
|
2010-03-23 17:41:24 +01:00
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
}
|
2011-01-10 17:17:28 +01:00
|
|
|
|
2009-05-28 16:07:05 +02:00
|
|
|
/* write the l2 table to the file */
|
2010-04-14 14:17:38 +02:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L2_ALLOC_WRITE);
|
2011-01-10 17:17:28 +01:00
|
|
|
|
|
|
|
qcow2_cache_entry_mark_dirty(s->l2_table_cache, l2_table);
|
|
|
|
ret = qcow2_cache_flush(bs, s->l2_table_cache);
|
2010-03-23 17:41:24 +01:00
|
|
|
if (ret < 0) {
|
2010-05-12 16:23:26 +02:00
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* update the L1 entry */
|
|
|
|
s->l1_table[l1_index] = l2_offset | QCOW_OFLAG_COPIED;
|
|
|
|
ret = write_l1_entry(bs, l1_index);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
2010-03-23 17:41:24 +01:00
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
|
2010-03-23 17:41:24 +01:00
|
|
|
*table = l2_table;
|
|
|
|
return 0;
|
2010-05-12 16:23:26 +02:00
|
|
|
|
|
|
|
fail:
|
2011-01-10 17:17:28 +01:00
|
|
|
qcow2_cache_put(bs, s->l2_table_cache, (void**) table);
|
2010-06-07 16:43:22 +02:00
|
|
|
s->l1_table[l1_index] = old_l2_offset;
|
2010-05-12 16:23:26 +02:00
|
|
|
return ret;
|
2009-05-28 16:07:05 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
static int count_contiguous_clusters(uint64_t nb_clusters, int cluster_size,
|
|
|
|
uint64_t *l2_table, uint64_t start, uint64_t mask)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
uint64_t offset = be64_to_cpu(l2_table[0]) & ~mask;
|
|
|
|
|
|
|
|
if (!offset)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
for (i = start; i < start + nb_clusters; i++)
|
2009-09-15 12:30:43 +02:00
|
|
|
if (offset + (uint64_t) i * cluster_size != (be64_to_cpu(l2_table[i]) & ~mask))
|
2009-05-28 16:07:05 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
return (i - start);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int count_contiguous_free_clusters(uint64_t nb_clusters, uint64_t *l2_table)
|
|
|
|
{
|
|
|
|
int i = 0;
|
|
|
|
|
|
|
|
while(nb_clusters-- && l2_table[i] == 0)
|
|
|
|
i++;
|
|
|
|
|
|
|
|
return i;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* The crypt function is compatible with the linux cryptoloop
|
|
|
|
algorithm for < 4 GB images. NOTE: out_buf == in_buf is
|
|
|
|
supported */
|
2009-05-28 16:07:07 +02:00
|
|
|
void qcow2_encrypt_sectors(BDRVQcowState *s, int64_t sector_num,
|
|
|
|
uint8_t *out_buf, const uint8_t *in_buf,
|
|
|
|
int nb_sectors, int enc,
|
|
|
|
const AES_KEY *key)
|
2009-05-28 16:07:05 +02:00
|
|
|
{
|
|
|
|
union {
|
|
|
|
uint64_t ll[2];
|
|
|
|
uint8_t b[16];
|
|
|
|
} ivec;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for(i = 0; i < nb_sectors; i++) {
|
|
|
|
ivec.ll[0] = cpu_to_le64(sector_num);
|
|
|
|
ivec.ll[1] = 0;
|
|
|
|
AES_cbc_encrypt(in_buf, out_buf, 512, key,
|
|
|
|
ivec.b, enc);
|
|
|
|
sector_num++;
|
|
|
|
in_buf += 512;
|
|
|
|
out_buf += 512;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2010-12-17 16:02:39 +01:00
|
|
|
static int qcow2_read(BlockDriverState *bs, int64_t sector_num,
|
|
|
|
uint8_t *buf, int nb_sectors)
|
2009-05-28 16:07:05 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
int ret, index_in_cluster, n, n1;
|
|
|
|
uint64_t cluster_offset;
|
2010-09-13 18:08:52 +02:00
|
|
|
struct iovec iov;
|
|
|
|
QEMUIOVector qiov;
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
while (nb_sectors > 0) {
|
|
|
|
n = nb_sectors;
|
2010-05-21 17:59:36 +02:00
|
|
|
|
|
|
|
ret = qcow2_get_cluster_offset(bs, sector_num << 9, &n,
|
|
|
|
&cluster_offset);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2009-05-28 16:07:05 +02:00
|
|
|
index_in_cluster = sector_num & (s->cluster_sectors - 1);
|
|
|
|
if (!cluster_offset) {
|
|
|
|
if (bs->backing_hd) {
|
|
|
|
/* read from the base image */
|
2010-09-13 18:08:52 +02:00
|
|
|
iov.iov_base = buf;
|
|
|
|
iov.iov_len = n * 512;
|
|
|
|
qemu_iovec_init_external(&qiov, &iov, 1);
|
|
|
|
|
|
|
|
n1 = qcow2_backing_read1(bs->backing_hd, &qiov, sector_num, n);
|
2009-05-28 16:07:05 +02:00
|
|
|
if (n1 > 0) {
|
2010-04-14 14:17:38 +02:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_READ_BACKING);
|
2009-05-28 16:07:05 +02:00
|
|
|
ret = bdrv_read(bs->backing_hd, sector_num, buf, n1);
|
|
|
|
if (ret < 0)
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
memset(buf, 0, 512 * n);
|
|
|
|
}
|
|
|
|
} else if (cluster_offset & QCOW_OFLAG_COMPRESSED) {
|
2010-04-14 14:17:38 +02:00
|
|
|
if (qcow2_decompress_cluster(bs, cluster_offset) < 0)
|
2009-05-28 16:07:05 +02:00
|
|
|
return -1;
|
|
|
|
memcpy(buf, s->cluster_cache + index_in_cluster * 512, 512 * n);
|
|
|
|
} else {
|
2010-04-14 14:17:38 +02:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_READ);
|
|
|
|
ret = bdrv_pread(bs->file, cluster_offset + index_in_cluster * 512, buf, n * 512);
|
2009-05-28 16:07:05 +02:00
|
|
|
if (ret != n * 512)
|
|
|
|
return -1;
|
|
|
|
if (s->crypt_method) {
|
2009-05-28 16:07:07 +02:00
|
|
|
qcow2_encrypt_sectors(s, sector_num, buf, buf, n, 0,
|
2009-05-28 16:07:05 +02:00
|
|
|
&s->aes_decrypt_key);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
nb_sectors -= n;
|
|
|
|
sector_num += n;
|
|
|
|
buf += n * 512;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int copy_sectors(BlockDriverState *bs, uint64_t start_sect,
|
|
|
|
uint64_t cluster_offset, int n_start, int n_end)
|
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
int n, ret;
|
|
|
|
|
|
|
|
n = n_end - n_start;
|
|
|
|
if (n <= 0)
|
|
|
|
return 0;
|
2010-04-14 14:17:38 +02:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_COW_READ);
|
2010-12-17 16:02:39 +01:00
|
|
|
ret = qcow2_read(bs, start_sect + n_start, s->cluster_data, n);
|
2009-05-28 16:07:05 +02:00
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
|
|
|
if (s->crypt_method) {
|
2009-05-28 16:07:07 +02:00
|
|
|
qcow2_encrypt_sectors(s, start_sect + n_start,
|
2009-05-28 16:07:05 +02:00
|
|
|
s->cluster_data,
|
|
|
|
s->cluster_data, n, 1,
|
|
|
|
&s->aes_encrypt_key);
|
|
|
|
}
|
2010-04-14 14:17:38 +02:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_COW_WRITE);
|
2010-09-17 17:02:09 +02:00
|
|
|
ret = bdrv_write(bs->file, (cluster_offset >> 9) + n_start,
|
2010-06-16 17:44:35 +02:00
|
|
|
s->cluster_data, n);
|
2009-05-28 16:07:05 +02:00
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* get_cluster_offset
|
|
|
|
*
|
2010-05-21 17:59:36 +02:00
|
|
|
* For a given offset of the disk image, find the cluster offset in
|
|
|
|
* qcow2 file. The offset is stored in *cluster_offset.
|
2009-05-28 16:07:05 +02:00
|
|
|
*
|
2011-08-08 01:47:36 +02:00
|
|
|
* on entry, *num is the number of contiguous sectors we'd like to
|
2009-05-28 16:07:05 +02:00
|
|
|
* access following offset.
|
|
|
|
*
|
2011-08-08 01:47:36 +02:00
|
|
|
* on exit, *num is the number of contiguous sectors we can read.
|
2009-05-28 16:07:05 +02:00
|
|
|
*
|
2010-05-21 17:59:36 +02:00
|
|
|
* Return 0, if the offset is found
|
|
|
|
* Return -errno, otherwise.
|
2009-05-28 16:07:05 +02:00
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
2010-05-21 17:59:36 +02:00
|
|
|
int qcow2_get_cluster_offset(BlockDriverState *bs, uint64_t offset,
|
|
|
|
int *num, uint64_t *cluster_offset)
|
2009-05-28 16:07:05 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
2009-09-15 12:30:43 +02:00
|
|
|
unsigned int l1_index, l2_index;
|
2010-05-21 17:59:36 +02:00
|
|
|
uint64_t l2_offset, *l2_table;
|
2009-05-28 16:07:05 +02:00
|
|
|
int l1_bits, c;
|
2009-09-15 12:30:43 +02:00
|
|
|
unsigned int index_in_cluster, nb_clusters;
|
|
|
|
uint64_t nb_available, nb_needed;
|
2010-05-21 18:25:20 +02:00
|
|
|
int ret;
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
index_in_cluster = (offset >> 9) & (s->cluster_sectors - 1);
|
|
|
|
nb_needed = *num + index_in_cluster;
|
|
|
|
|
|
|
|
l1_bits = s->l2_bits + s->cluster_bits;
|
|
|
|
|
|
|
|
/* compute how many bytes there are between the offset and
|
|
|
|
* the end of the l1 entry
|
|
|
|
*/
|
|
|
|
|
2009-09-15 12:30:43 +02:00
|
|
|
nb_available = (1ULL << l1_bits) - (offset & ((1ULL << l1_bits) - 1));
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
/* compute the number of available sectors */
|
|
|
|
|
|
|
|
nb_available = (nb_available >> 9) + index_in_cluster;
|
|
|
|
|
|
|
|
if (nb_needed > nb_available) {
|
|
|
|
nb_needed = nb_available;
|
|
|
|
}
|
|
|
|
|
2010-05-21 17:59:36 +02:00
|
|
|
*cluster_offset = 0;
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
/* seek the the l2 offset in the l1 table */
|
|
|
|
|
|
|
|
l1_index = offset >> l1_bits;
|
|
|
|
if (l1_index >= s->l1_size)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
l2_offset = s->l1_table[l1_index];
|
|
|
|
|
|
|
|
/* seek the l2 table of the given l2 offset */
|
|
|
|
|
|
|
|
if (!l2_offset)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
/* load the l2 table in memory */
|
|
|
|
|
|
|
|
l2_offset &= ~QCOW_OFLAG_COPIED;
|
2010-05-21 18:25:20 +02:00
|
|
|
ret = l2_load(bs, l2_offset, &l2_table);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
2010-05-21 17:59:36 +02:00
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
/* find the cluster offset for the given disk offset */
|
|
|
|
|
|
|
|
l2_index = (offset >> s->cluster_bits) & (s->l2_size - 1);
|
2010-05-21 17:59:36 +02:00
|
|
|
*cluster_offset = be64_to_cpu(l2_table[l2_index]);
|
2009-05-28 16:07:05 +02:00
|
|
|
nb_clusters = size_to_clusters(s, nb_needed << 9);
|
|
|
|
|
2010-05-21 17:59:36 +02:00
|
|
|
if (!*cluster_offset) {
|
2009-05-28 16:07:05 +02:00
|
|
|
/* how many empty clusters ? */
|
|
|
|
c = count_contiguous_free_clusters(nb_clusters, &l2_table[l2_index]);
|
|
|
|
} else {
|
|
|
|
/* how many allocated clusters ? */
|
|
|
|
c = count_contiguous_clusters(nb_clusters, s->cluster_size,
|
|
|
|
&l2_table[l2_index], 0, QCOW_OFLAG_COPIED);
|
|
|
|
}
|
|
|
|
|
2011-01-10 17:17:28 +01:00
|
|
|
qcow2_cache_put(bs, s->l2_table_cache, (void**) &l2_table);
|
|
|
|
|
2009-05-28 16:07:05 +02:00
|
|
|
nb_available = (c * s->cluster_sectors);
|
|
|
|
out:
|
|
|
|
if (nb_available > nb_needed)
|
|
|
|
nb_available = nb_needed;
|
|
|
|
|
|
|
|
*num = nb_available - index_in_cluster;
|
|
|
|
|
2010-05-21 17:59:36 +02:00
|
|
|
*cluster_offset &=~QCOW_OFLAG_COPIED;
|
|
|
|
return 0;
|
2009-05-28 16:07:05 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* get_cluster_table
|
|
|
|
*
|
|
|
|
* for a given disk offset, load (and allocate if needed)
|
|
|
|
* the l2 table.
|
|
|
|
*
|
|
|
|
* the l2 table offset in the qcow2 file and the cluster index
|
|
|
|
* in the l2 table are given to the caller.
|
|
|
|
*
|
2010-01-20 15:03:00 +01:00
|
|
|
* Returns 0 on success, -errno in failure case
|
2009-05-28 16:07:05 +02:00
|
|
|
*/
|
|
|
|
static int get_cluster_table(BlockDriverState *bs, uint64_t offset,
|
|
|
|
uint64_t **new_l2_table,
|
|
|
|
uint64_t *new_l2_offset,
|
|
|
|
int *new_l2_index)
|
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
2009-09-15 12:30:43 +02:00
|
|
|
unsigned int l1_index, l2_index;
|
2010-03-23 17:41:24 +01:00
|
|
|
uint64_t l2_offset;
|
|
|
|
uint64_t *l2_table = NULL;
|
2009-09-15 12:30:43 +02:00
|
|
|
int ret;
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
/* seek the the l2 offset in the l1 table */
|
|
|
|
|
|
|
|
l1_index = offset >> (s->l2_bits + s->cluster_bits);
|
|
|
|
if (l1_index >= s->l1_size) {
|
2010-10-18 17:53:53 +02:00
|
|
|
ret = qcow2_grow_l1_table(bs, l1_index + 1, false);
|
2010-01-20 15:03:00 +01:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
}
|
|
|
|
l2_offset = s->l1_table[l1_index];
|
|
|
|
|
|
|
|
/* seek the l2 table of the given l2 offset */
|
|
|
|
|
|
|
|
if (l2_offset & QCOW_OFLAG_COPIED) {
|
|
|
|
/* load the l2 table in memory */
|
|
|
|
l2_offset &= ~QCOW_OFLAG_COPIED;
|
2010-05-21 18:25:20 +02:00
|
|
|
ret = l2_load(bs, l2_offset, &l2_table);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
2010-01-20 15:03:00 +01:00
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
} else {
|
2011-02-09 17:36:19 +01:00
|
|
|
/* First allocate a new L2 table (and do COW if needed) */
|
2010-03-23 17:41:24 +01:00
|
|
|
ret = l2_allocate(bs, l1_index, &l2_table);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
2010-01-20 15:03:00 +01:00
|
|
|
}
|
2011-02-09 17:36:19 +01:00
|
|
|
|
|
|
|
/* Then decrease the refcount of the old table */
|
|
|
|
if (l2_offset) {
|
|
|
|
qcow2_free_clusters(bs, l2_offset, s->l2_size * sizeof(uint64_t));
|
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
l2_offset = s->l1_table[l1_index] & ~QCOW_OFLAG_COPIED;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* find the cluster offset for the given disk offset */
|
|
|
|
|
|
|
|
l2_index = (offset >> s->cluster_bits) & (s->l2_size - 1);
|
|
|
|
|
|
|
|
*new_l2_table = l2_table;
|
|
|
|
*new_l2_offset = l2_offset;
|
|
|
|
*new_l2_index = l2_index;
|
|
|
|
|
2010-01-20 15:03:00 +01:00
|
|
|
return 0;
|
2009-05-28 16:07:05 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* alloc_compressed_cluster_offset
|
|
|
|
*
|
|
|
|
* For a given offset of the disk image, return cluster offset in
|
|
|
|
* qcow2 file.
|
|
|
|
*
|
|
|
|
* If the offset is not found, allocate a new compressed cluster.
|
|
|
|
*
|
|
|
|
* Return the cluster offset if successful,
|
|
|
|
* Return 0, otherwise.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
2009-05-28 16:07:07 +02:00
|
|
|
uint64_t qcow2_alloc_compressed_cluster_offset(BlockDriverState *bs,
|
|
|
|
uint64_t offset,
|
|
|
|
int compressed_size)
|
2009-05-28 16:07:05 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
int l2_index, ret;
|
2010-02-02 15:20:57 +01:00
|
|
|
uint64_t l2_offset, *l2_table;
|
|
|
|
int64_t cluster_offset;
|
2009-05-28 16:07:05 +02:00
|
|
|
int nb_csectors;
|
|
|
|
|
|
|
|
ret = get_cluster_table(bs, offset, &l2_table, &l2_offset, &l2_index);
|
2010-01-20 15:03:00 +01:00
|
|
|
if (ret < 0) {
|
2009-05-28 16:07:05 +02:00
|
|
|
return 0;
|
2010-01-20 15:03:00 +01:00
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
cluster_offset = be64_to_cpu(l2_table[l2_index]);
|
|
|
|
if (cluster_offset & QCOW_OFLAG_COPIED)
|
|
|
|
return cluster_offset & ~QCOW_OFLAG_COPIED;
|
|
|
|
|
|
|
|
if (cluster_offset)
|
2009-05-28 16:07:07 +02:00
|
|
|
qcow2_free_any_clusters(bs, cluster_offset, 1);
|
2009-05-28 16:07:05 +02:00
|
|
|
|
2009-05-28 16:07:07 +02:00
|
|
|
cluster_offset = qcow2_alloc_bytes(bs, compressed_size);
|
2010-01-20 15:04:01 +01:00
|
|
|
if (cluster_offset < 0) {
|
2011-01-10 17:17:28 +01:00
|
|
|
qcow2_cache_put(bs, s->l2_table_cache, (void**) &l2_table);
|
2010-01-20 15:04:01 +01:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2009-05-28 16:07:05 +02:00
|
|
|
nb_csectors = ((cluster_offset + compressed_size - 1) >> 9) -
|
|
|
|
(cluster_offset >> 9);
|
|
|
|
|
|
|
|
cluster_offset |= QCOW_OFLAG_COMPRESSED |
|
|
|
|
((uint64_t)nb_csectors << s->csize_shift);
|
|
|
|
|
|
|
|
/* update L2 table */
|
|
|
|
|
|
|
|
/* compressed clusters never have the copied flag */
|
|
|
|
|
2010-04-14 14:17:38 +02:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L2_UPDATE_COMPRESSED);
|
2011-01-10 17:17:28 +01:00
|
|
|
qcow2_cache_entry_mark_dirty(s->l2_table_cache, l2_table);
|
2009-05-28 16:07:05 +02:00
|
|
|
l2_table[l2_index] = cpu_to_be64(cluster_offset);
|
2011-01-10 17:17:28 +01:00
|
|
|
ret = qcow2_cache_put(bs, s->l2_table_cache, (void**) &l2_table);
|
2010-03-23 12:49:17 +01:00
|
|
|
if (ret < 0) {
|
2011-01-10 17:17:28 +01:00
|
|
|
return 0;
|
2009-06-16 11:31:29 +02:00
|
|
|
}
|
|
|
|
|
2011-01-10 17:17:28 +01:00
|
|
|
return cluster_offset;
|
2009-06-16 11:31:29 +02:00
|
|
|
}
|
|
|
|
|
2010-01-20 15:03:01 +01:00
|
|
|
int qcow2_alloc_cluster_link_l2(BlockDriverState *bs, QCowL2Meta *m)
|
2009-05-28 16:07:05 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
int i, j = 0, l2_index, ret;
|
|
|
|
uint64_t *old_cluster, start_sect, l2_offset, *l2_table;
|
2010-01-20 15:03:01 +01:00
|
|
|
uint64_t cluster_offset = m->cluster_offset;
|
2011-01-10 17:17:28 +01:00
|
|
|
bool cow = false;
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
if (m->nb_clusters == 0)
|
|
|
|
return 0;
|
|
|
|
|
2011-08-21 05:09:37 +02:00
|
|
|
old_cluster = g_malloc(m->nb_clusters * sizeof(uint64_t));
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
/* copy content of unmodified sectors */
|
|
|
|
start_sect = (m->offset & ~(s->cluster_size - 1)) >> 9;
|
|
|
|
if (m->n_start) {
|
2011-01-10 17:17:28 +01:00
|
|
|
cow = true;
|
2009-05-28 16:07:05 +02:00
|
|
|
ret = copy_sectors(bs, start_sect, cluster_offset, 0, m->n_start);
|
|
|
|
if (ret < 0)
|
|
|
|
goto err;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (m->nb_available & (s->cluster_sectors - 1)) {
|
|
|
|
uint64_t end = m->nb_available & ~(uint64_t)(s->cluster_sectors - 1);
|
2011-01-10 17:17:28 +01:00
|
|
|
cow = true;
|
2009-05-28 16:07:05 +02:00
|
|
|
ret = copy_sectors(bs, start_sect + end, cluster_offset + (end << 9),
|
|
|
|
m->nb_available - end, s->cluster_sectors);
|
|
|
|
if (ret < 0)
|
|
|
|
goto err;
|
|
|
|
}
|
|
|
|
|
2011-01-10 17:17:28 +01:00
|
|
|
/*
|
|
|
|
* Update L2 table.
|
|
|
|
*
|
|
|
|
* Before we update the L2 table to actually point to the new cluster, we
|
|
|
|
* need to be sure that the refcounts have been increased and COW was
|
|
|
|
* handled.
|
|
|
|
*/
|
|
|
|
if (cow) {
|
2011-01-14 15:55:38 +01:00
|
|
|
qcow2_cache_depends_on_flush(s->l2_table_cache);
|
2011-01-10 17:17:28 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
qcow2_cache_set_dependency(bs, s->l2_table_cache, s->refcount_block_cache);
|
2010-01-20 15:03:00 +01:00
|
|
|
ret = get_cluster_table(bs, m->offset, &l2_table, &l2_offset, &l2_index);
|
|
|
|
if (ret < 0) {
|
2009-05-28 16:07:05 +02:00
|
|
|
goto err;
|
2010-01-20 15:03:00 +01:00
|
|
|
}
|
2011-01-10 17:17:28 +01:00
|
|
|
qcow2_cache_entry_mark_dirty(s->l2_table_cache, l2_table);
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
for (i = 0; i < m->nb_clusters; i++) {
|
|
|
|
/* if two concurrent writes happen to the same unallocated cluster
|
|
|
|
* each write allocates separate cluster and writes data concurrently.
|
|
|
|
* The first one to complete updates l2 table with pointer to its
|
|
|
|
* cluster the second one has to do RMW (which is done above by
|
|
|
|
* copy_sectors()), update l2 table with its cluster pointer and free
|
|
|
|
* old cluster. This is what this loop does */
|
|
|
|
if(l2_table[l2_index + i] != 0)
|
|
|
|
old_cluster[j++] = l2_table[l2_index + i];
|
|
|
|
|
|
|
|
l2_table[l2_index + i] = cpu_to_be64((cluster_offset +
|
|
|
|
(i << s->cluster_bits)) | QCOW_OFLAG_COPIED);
|
|
|
|
}
|
|
|
|
|
2010-09-17 17:02:09 +02:00
|
|
|
|
2011-01-10 17:17:28 +01:00
|
|
|
ret = qcow2_cache_put(bs, s->l2_table_cache, (void**) &l2_table);
|
2010-03-23 12:53:47 +01:00
|
|
|
if (ret < 0) {
|
2009-05-28 16:07:05 +02:00
|
|
|
goto err;
|
2009-06-16 11:31:29 +02:00
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
|
2010-09-01 12:40:52 +02:00
|
|
|
/*
|
|
|
|
* If this was a COW, we need to decrease the refcount of the old cluster.
|
|
|
|
* Also flush bs->file to get the right order for L2 and refcount update.
|
|
|
|
*/
|
|
|
|
if (j != 0) {
|
|
|
|
for (i = 0; i < j; i++) {
|
|
|
|
qcow2_free_any_clusters(bs,
|
|
|
|
be64_to_cpu(old_cluster[i]) & ~QCOW_OFLAG_COPIED, 1);
|
|
|
|
}
|
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
ret = 0;
|
|
|
|
err:
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(old_cluster);
|
2009-05-28 16:07:05 +02:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* alloc_cluster_offset
|
|
|
|
*
|
2010-01-20 15:03:01 +01:00
|
|
|
* For a given offset of the disk image, return cluster offset in qcow2 file.
|
2009-05-28 16:07:05 +02:00
|
|
|
* If the offset is not found, allocate a new cluster.
|
|
|
|
*
|
2010-01-20 15:03:01 +01:00
|
|
|
* If the cluster was already allocated, m->nb_clusters is set to 0,
|
2011-09-07 16:19:17 +02:00
|
|
|
* other fields in m are meaningless.
|
2010-01-20 15:03:01 +01:00
|
|
|
*
|
|
|
|
* If the cluster is newly allocated, m->nb_clusters is set to the number of
|
2011-06-30 17:42:09 +02:00
|
|
|
* contiguous clusters that have been allocated. In this case, the other
|
|
|
|
* fields of m are valid and contain information about the first allocated
|
|
|
|
* cluster.
|
2009-05-28 16:07:05 +02:00
|
|
|
*
|
2011-06-30 17:42:09 +02:00
|
|
|
* If the request conflicts with another write request in flight, the coroutine
|
|
|
|
* is queued and will be reentered when the dependency has completed.
|
2010-01-20 15:03:01 +01:00
|
|
|
*
|
|
|
|
* Return 0 on success and -errno in error cases
|
2009-05-28 16:07:05 +02:00
|
|
|
*/
|
2010-02-02 15:20:57 +01:00
|
|
|
int qcow2_alloc_cluster_offset(BlockDriverState *bs, uint64_t offset,
|
|
|
|
int n_start, int n_end, int *num, QCowL2Meta *m)
|
2009-05-28 16:07:05 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
int l2_index, ret;
|
2010-01-20 15:04:01 +01:00
|
|
|
uint64_t l2_offset, *l2_table;
|
|
|
|
int64_t cluster_offset;
|
2009-09-15 12:30:43 +02:00
|
|
|
unsigned int nb_clusters, i = 0;
|
2009-08-31 16:48:49 +02:00
|
|
|
QCowL2Meta *old_alloc;
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
ret = get_cluster_table(bs, offset, &l2_table, &l2_offset, &l2_index);
|
2010-01-20 15:03:00 +01:00
|
|
|
if (ret < 0) {
|
2010-01-20 15:03:01 +01:00
|
|
|
return ret;
|
2010-01-20 15:03:00 +01:00
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
|
2011-06-30 17:42:09 +02:00
|
|
|
again:
|
2009-05-28 16:07:05 +02:00
|
|
|
nb_clusters = size_to_clusters(s, n_end << 9);
|
|
|
|
|
|
|
|
nb_clusters = MIN(nb_clusters, s->l2_size - l2_index);
|
|
|
|
|
|
|
|
cluster_offset = be64_to_cpu(l2_table[l2_index]);
|
|
|
|
|
|
|
|
/* We keep all QCOW_OFLAG_COPIED clusters */
|
|
|
|
|
|
|
|
if (cluster_offset & QCOW_OFLAG_COPIED) {
|
|
|
|
nb_clusters = count_contiguous_clusters(nb_clusters, s->cluster_size,
|
|
|
|
&l2_table[l2_index], 0, 0);
|
|
|
|
|
|
|
|
cluster_offset &= ~QCOW_OFLAG_COPIED;
|
|
|
|
m->nb_clusters = 0;
|
|
|
|
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* for the moment, multiple compressed clusters are not managed */
|
|
|
|
|
|
|
|
if (cluster_offset & QCOW_OFLAG_COMPRESSED)
|
|
|
|
nb_clusters = 1;
|
|
|
|
|
|
|
|
/* how many available clusters ? */
|
|
|
|
|
|
|
|
while (i < nb_clusters) {
|
|
|
|
i += count_contiguous_clusters(nb_clusters - i, s->cluster_size,
|
|
|
|
&l2_table[l2_index], i, 0);
|
2010-02-16 15:54:49 +01:00
|
|
|
if ((i >= nb_clusters) || be64_to_cpu(l2_table[l2_index + i])) {
|
2009-05-28 16:07:05 +02:00
|
|
|
break;
|
2010-02-16 15:54:49 +01:00
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
i += count_contiguous_free_clusters(nb_clusters - i,
|
|
|
|
&l2_table[l2_index + i]);
|
2010-02-16 15:54:49 +01:00
|
|
|
if (i >= nb_clusters) {
|
|
|
|
break;
|
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
cluster_offset = be64_to_cpu(l2_table[l2_index + i]);
|
|
|
|
|
|
|
|
if ((cluster_offset & QCOW_OFLAG_COPIED) ||
|
|
|
|
(cluster_offset & QCOW_OFLAG_COMPRESSED))
|
|
|
|
break;
|
|
|
|
}
|
2010-02-16 15:54:49 +01:00
|
|
|
assert(i <= nb_clusters);
|
2009-05-28 16:07:05 +02:00
|
|
|
nb_clusters = i;
|
|
|
|
|
2009-08-31 16:48:49 +02:00
|
|
|
/*
|
|
|
|
* Check if there already is an AIO write request in flight which allocates
|
|
|
|
* the same cluster. In this case we need to wait until the previous
|
|
|
|
* request has completed and updated the L2 table accordingly.
|
|
|
|
*/
|
2009-09-12 09:36:22 +02:00
|
|
|
QLIST_FOREACH(old_alloc, &s->cluster_allocs, next_in_flight) {
|
2009-08-31 16:48:49 +02:00
|
|
|
|
2011-09-10 10:23:56 +02:00
|
|
|
uint64_t start = offset >> s->cluster_bits;
|
|
|
|
uint64_t end = start + nb_clusters;
|
|
|
|
uint64_t old_start = old_alloc->offset >> s->cluster_bits;
|
|
|
|
uint64_t old_end = old_start + old_alloc->nb_clusters;
|
2009-08-31 16:48:49 +02:00
|
|
|
|
2011-09-10 10:23:56 +02:00
|
|
|
if (end < old_start || start > old_end) {
|
2009-08-31 16:48:49 +02:00
|
|
|
/* No intersection */
|
|
|
|
} else {
|
2011-09-10 10:23:56 +02:00
|
|
|
if (start < old_start) {
|
2009-08-31 16:48:49 +02:00
|
|
|
/* Stop at the start of a running allocation */
|
2011-09-10 10:23:56 +02:00
|
|
|
nb_clusters = old_start - start;
|
2009-08-31 16:48:49 +02:00
|
|
|
} else {
|
|
|
|
nb_clusters = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (nb_clusters == 0) {
|
2011-06-30 17:42:09 +02:00
|
|
|
/* Wait for the dependency to complete. We need to recheck
|
|
|
|
* the free/allocated clusters when we continue. */
|
|
|
|
qemu_co_mutex_unlock(&s->lock);
|
|
|
|
qemu_co_queue_wait(&old_alloc->dependent_requests);
|
|
|
|
qemu_co_mutex_lock(&s->lock);
|
|
|
|
goto again;
|
2009-08-31 16:48:49 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!nb_clusters) {
|
|
|
|
abort();
|
|
|
|
}
|
|
|
|
|
2011-09-08 13:38:42 +02:00
|
|
|
/* save info needed for meta data update */
|
|
|
|
m->offset = offset;
|
|
|
|
m->n_start = n_start;
|
|
|
|
m->nb_clusters = nb_clusters;
|
|
|
|
|
2009-09-12 09:36:22 +02:00
|
|
|
QLIST_INSERT_HEAD(&s->cluster_allocs, m, next_in_flight);
|
2009-08-31 16:48:49 +02:00
|
|
|
|
2009-05-28 16:07:05 +02:00
|
|
|
/* allocate a new cluster */
|
|
|
|
|
2009-05-28 16:07:07 +02:00
|
|
|
cluster_offset = qcow2_alloc_clusters(bs, nb_clusters * s->cluster_size);
|
2010-01-20 15:04:01 +01:00
|
|
|
if (cluster_offset < 0) {
|
2011-01-10 17:17:28 +01:00
|
|
|
ret = cluster_offset;
|
|
|
|
goto fail;
|
2010-01-20 15:04:01 +01:00
|
|
|
}
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
out:
|
2011-01-10 17:17:28 +01:00
|
|
|
ret = qcow2_cache_put(bs, s->l2_table_cache, (void**) &l2_table);
|
|
|
|
if (ret < 0) {
|
2011-05-24 16:40:02 +02:00
|
|
|
goto fail_put;
|
2011-01-10 17:17:28 +01:00
|
|
|
}
|
|
|
|
|
2009-05-28 16:07:05 +02:00
|
|
|
m->nb_available = MIN(nb_clusters << (s->cluster_bits - 9), n_end);
|
2010-01-20 15:03:01 +01:00
|
|
|
m->cluster_offset = cluster_offset;
|
2009-05-28 16:07:05 +02:00
|
|
|
|
|
|
|
*num = m->nb_available - n_start;
|
|
|
|
|
2010-01-20 15:03:01 +01:00
|
|
|
return 0;
|
2011-01-10 17:17:28 +01:00
|
|
|
|
|
|
|
fail:
|
|
|
|
qcow2_cache_put(bs, s->l2_table_cache, (void**) &l2_table);
|
2011-05-24 16:40:02 +02:00
|
|
|
fail_put:
|
|
|
|
QLIST_REMOVE(m, next_in_flight);
|
2011-01-10 17:17:28 +01:00
|
|
|
return ret;
|
2009-05-28 16:07:05 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
static int decompress_buffer(uint8_t *out_buf, int out_buf_size,
|
|
|
|
const uint8_t *buf, int buf_size)
|
|
|
|
{
|
|
|
|
z_stream strm1, *strm = &strm1;
|
|
|
|
int ret, out_len;
|
|
|
|
|
|
|
|
memset(strm, 0, sizeof(*strm));
|
|
|
|
|
|
|
|
strm->next_in = (uint8_t *)buf;
|
|
|
|
strm->avail_in = buf_size;
|
|
|
|
strm->next_out = out_buf;
|
|
|
|
strm->avail_out = out_buf_size;
|
|
|
|
|
|
|
|
ret = inflateInit2(strm, -12);
|
|
|
|
if (ret != Z_OK)
|
|
|
|
return -1;
|
|
|
|
ret = inflate(strm, Z_FINISH);
|
|
|
|
out_len = strm->next_out - out_buf;
|
|
|
|
if ((ret != Z_STREAM_END && ret != Z_BUF_ERROR) ||
|
|
|
|
out_len != out_buf_size) {
|
|
|
|
inflateEnd(strm);
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
inflateEnd(strm);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2010-04-14 14:17:38 +02:00
|
|
|
int qcow2_decompress_cluster(BlockDriverState *bs, uint64_t cluster_offset)
|
2009-05-28 16:07:05 +02:00
|
|
|
{
|
2010-04-14 14:17:38 +02:00
|
|
|
BDRVQcowState *s = bs->opaque;
|
2009-05-28 16:07:05 +02:00
|
|
|
int ret, csize, nb_csectors, sector_offset;
|
|
|
|
uint64_t coffset;
|
|
|
|
|
|
|
|
coffset = cluster_offset & s->cluster_offset_mask;
|
|
|
|
if (s->cluster_cache_offset != coffset) {
|
|
|
|
nb_csectors = ((cluster_offset >> s->csize_shift) & s->csize_mask) + 1;
|
|
|
|
sector_offset = coffset & 511;
|
|
|
|
csize = nb_csectors * 512 - sector_offset;
|
2010-04-14 14:17:38 +02:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_READ_COMPRESSED);
|
|
|
|
ret = bdrv_read(bs->file, coffset >> 9, s->cluster_data, nb_csectors);
|
2009-05-28 16:07:05 +02:00
|
|
|
if (ret < 0) {
|
2011-02-09 10:26:06 +01:00
|
|
|
return ret;
|
2009-05-28 16:07:05 +02:00
|
|
|
}
|
|
|
|
if (decompress_buffer(s->cluster_cache, s->cluster_size,
|
|
|
|
s->cluster_data + sector_offset, csize) < 0) {
|
2011-02-09 10:26:06 +01:00
|
|
|
return -EIO;
|
2009-05-28 16:07:05 +02:00
|
|
|
}
|
|
|
|
s->cluster_cache_offset = coffset;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
2011-01-26 16:56:48 +01:00
|
|
|
|
|
|
|
/*
|
|
|
|
* This discards as many clusters of nb_clusters as possible at once (i.e.
|
|
|
|
* all clusters in the same L2 table) and returns the number of discarded
|
|
|
|
* clusters.
|
|
|
|
*/
|
|
|
|
static int discard_single_l2(BlockDriverState *bs, uint64_t offset,
|
|
|
|
unsigned int nb_clusters)
|
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
uint64_t l2_offset, *l2_table;
|
|
|
|
int l2_index;
|
|
|
|
int ret;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
ret = get_cluster_table(bs, offset, &l2_table, &l2_offset, &l2_index);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Limit nb_clusters to one L2 table */
|
|
|
|
nb_clusters = MIN(nb_clusters, s->l2_size - l2_index);
|
|
|
|
|
|
|
|
for (i = 0; i < nb_clusters; i++) {
|
|
|
|
uint64_t old_offset;
|
|
|
|
|
|
|
|
old_offset = be64_to_cpu(l2_table[l2_index + i]);
|
|
|
|
old_offset &= ~QCOW_OFLAG_COPIED;
|
|
|
|
|
|
|
|
if (old_offset == 0) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* First remove L2 entries */
|
|
|
|
qcow2_cache_entry_mark_dirty(s->l2_table_cache, l2_table);
|
|
|
|
l2_table[l2_index + i] = cpu_to_be64(0);
|
|
|
|
|
|
|
|
/* Then decrease the refcount */
|
|
|
|
qcow2_free_any_clusters(bs, old_offset, 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = qcow2_cache_put(bs, s->l2_table_cache, (void**) &l2_table);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
return nb_clusters;
|
|
|
|
}
|
|
|
|
|
|
|
|
int qcow2_discard_clusters(BlockDriverState *bs, uint64_t offset,
|
|
|
|
int nb_sectors)
|
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
uint64_t end_offset;
|
|
|
|
unsigned int nb_clusters;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
end_offset = offset + (nb_sectors << BDRV_SECTOR_BITS);
|
|
|
|
|
|
|
|
/* Round start up and end down */
|
|
|
|
offset = align_offset(offset, s->cluster_size);
|
|
|
|
end_offset &= ~(s->cluster_size - 1);
|
|
|
|
|
|
|
|
if (offset > end_offset) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
nb_clusters = size_to_clusters(s, end_offset - offset);
|
|
|
|
|
|
|
|
/* Each L2 table is handled by its own loop iteration */
|
|
|
|
while (nb_clusters > 0) {
|
|
|
|
ret = discard_single_l2(bs, offset, nb_clusters);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
nb_clusters -= ret;
|
|
|
|
offset += (ret * s->cluster_size);
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|