#include "aes.h"
#include "block/qcow2.h"
#include "qemu-error.h"
+#include "qerror.h"
/*
Differences with QCOW:
uint32_t magic;
uint32_t len;
} QCowExtension;
-#define QCOW_EXT_MAGIC_END 0
-#define QCOW_EXT_MAGIC_BACKING_FORMAT 0xE2792ACA
+#define QCOW2_EXT_MAGIC_END 0
+#define QCOW2_EXT_MAGIC_BACKING_FORMAT 0xE2792ACA
-static int qcow_probe(const uint8_t *buf, int buf_size, const char *filename)
+static int qcow2_probe(const uint8_t *buf, int buf_size, const char *filename)
{
const QCowHeader *cow_header = (const void *)buf;
if (buf_size >= sizeof(QCowHeader) &&
be32_to_cpu(cow_header->magic) == QCOW_MAGIC &&
- be32_to_cpu(cow_header->version) == QCOW_VERSION)
+ be32_to_cpu(cow_header->version) >= QCOW_VERSION)
return 100;
else
return 0;
* unknown magic is skipped (future extension this version knows nothing about)
* return 0 upon success, non-0 otherwise
*/
-static int qcow_read_extensions(BlockDriverState *bs, uint64_t start_offset,
- uint64_t end_offset)
+static int qcow2_read_extensions(BlockDriverState *bs, uint64_t start_offset,
+ uint64_t end_offset)
{
QCowExtension ext;
uint64_t offset;
#ifdef DEBUG_EXT
- printf("qcow_read_extensions: start=%ld end=%ld\n", start_offset, end_offset);
+ printf("qcow2_read_extensions: start=%ld end=%ld\n", start_offset, end_offset);
#endif
offset = start_offset;
while (offset < end_offset) {
#ifdef DEBUG_EXT
+ BDRVQcowState *s = bs->opaque;
/* Sanity check */
if (offset > s->cluster_size)
- printf("qcow_handle_extension: suspicious offset %lu\n", offset);
+ printf("qcow2_read_extension: suspicious offset %lu\n", offset);
printf("attemting to read extended header in offset %lu\n", offset);
#endif
if (bdrv_pread(bs->file, offset, &ext, sizeof(ext)) != sizeof(ext)) {
- fprintf(stderr, "qcow_handle_extension: ERROR: "
+ fprintf(stderr, "qcow2_read_extension: ERROR: "
"pread fail from offset %" PRIu64 "\n",
offset);
return 1;
printf("ext.magic = 0x%x\n", ext.magic);
#endif
switch (ext.magic) {
- case QCOW_EXT_MAGIC_END:
+ case QCOW2_EXT_MAGIC_END:
return 0;
- case QCOW_EXT_MAGIC_BACKING_FORMAT:
+ case QCOW2_EXT_MAGIC_BACKING_FORMAT:
if (ext.len >= sizeof(bs->backing_format)) {
fprintf(stderr, "ERROR: ext_backing_format: len=%u too large"
" (>=%zu)\n",
}
-static int qcow_open(BlockDriverState *bs, int flags)
+static int qcow2_open(BlockDriverState *bs, int flags)
{
BDRVQcowState *s = bs->opaque;
- int len, i;
+ int len, i, ret = 0;
QCowHeader header;
uint64_t ext_end;
+ bool writethrough;
- if (bdrv_pread(bs->file, 0, &header, sizeof(header)) != sizeof(header))
+ ret = bdrv_pread(bs->file, 0, &header, sizeof(header));
+ if (ret < 0) {
goto fail;
+ }
be32_to_cpus(&header.magic);
be32_to_cpus(&header.version);
be64_to_cpus(&header.backing_file_offset);
be64_to_cpus(&header.snapshots_offset);
be32_to_cpus(&header.nb_snapshots);
- if (header.magic != QCOW_MAGIC || header.version != QCOW_VERSION)
+ if (header.magic != QCOW_MAGIC) {
+ ret = -EINVAL;
goto fail;
+ }
+ if (header.version != QCOW_VERSION) {
+ char version[64];
+ snprintf(version, sizeof(version), "QCOW version %d", header.version);
+ qerror_report(QERR_UNKNOWN_BLOCK_FORMAT_FEATURE,
+ bs->device_name, "qcow2", version);
+ ret = -ENOTSUP;
+ goto fail;
+ }
if (header.cluster_bits < MIN_CLUSTER_BITS ||
- header.cluster_bits > MAX_CLUSTER_BITS)
+ header.cluster_bits > MAX_CLUSTER_BITS) {
+ ret = -EINVAL;
goto fail;
- if (header.crypt_method > QCOW_CRYPT_AES)
+ }
+ if (header.crypt_method > QCOW_CRYPT_AES) {
+ ret = -EINVAL;
goto fail;
+ }
s->crypt_method_header = header.crypt_method;
- if (s->crypt_method_header)
+ if (s->crypt_method_header) {
bs->encrypted = 1;
+ }
s->cluster_bits = header.cluster_bits;
s->cluster_size = 1 << s->cluster_bits;
s->cluster_sectors = 1 << (s->cluster_bits - 9);
s->l1_vm_state_index = size_to_l1(s, header.size);
/* the L1 table must contain at least enough entries to put
header.size bytes */
- if (s->l1_size < s->l1_vm_state_index)
+ if (s->l1_size < s->l1_vm_state_index) {
+ ret = -EINVAL;
goto fail;
+ }
s->l1_table_offset = header.l1_table_offset;
if (s->l1_size > 0) {
- s->l1_table = qemu_mallocz(
+ s->l1_table = g_malloc0(
align_offset(s->l1_size * sizeof(uint64_t), 512));
- if (bdrv_pread(bs->file, s->l1_table_offset, s->l1_table, s->l1_size * sizeof(uint64_t)) !=
- s->l1_size * sizeof(uint64_t))
+ ret = bdrv_pread(bs->file, s->l1_table_offset, s->l1_table,
+ s->l1_size * sizeof(uint64_t));
+ if (ret < 0) {
goto fail;
+ }
for(i = 0;i < s->l1_size; i++) {
be64_to_cpus(&s->l1_table[i]);
}
}
- /* alloc L2 cache */
- s->l2_cache = qemu_malloc(s->l2_size * L2_CACHE_SIZE * sizeof(uint64_t));
- s->cluster_cache = qemu_malloc(s->cluster_size);
+
+ /* alloc L2 table/refcount block cache */
+ writethrough = ((flags & BDRV_O_CACHE_WB) == 0);
+ s->l2_table_cache = qcow2_cache_create(bs, L2_CACHE_SIZE, writethrough);
+ s->refcount_block_cache = qcow2_cache_create(bs, REFCOUNT_CACHE_SIZE,
+ writethrough);
+
+ s->cluster_cache = g_malloc(s->cluster_size);
/* one more sector for decompressed data alignment */
- s->cluster_data = qemu_malloc(QCOW_MAX_CRYPT_CLUSTERS * s->cluster_size
+ s->cluster_data = qemu_blockalign(bs, QCOW_MAX_CRYPT_CLUSTERS * s->cluster_size
+ 512);
s->cluster_cache_offset = -1;
- if (qcow2_refcount_init(bs) < 0)
+ ret = qcow2_refcount_init(bs);
+ if (ret != 0) {
goto fail;
+ }
QLIST_INIT(&s->cluster_allocs);
/* read qcow2 extensions */
- if (header.backing_file_offset)
+ if (header.backing_file_offset) {
ext_end = header.backing_file_offset;
- else
+ } else {
ext_end = s->cluster_size;
- if (qcow_read_extensions(bs, sizeof(header), ext_end))
+ }
+ if (qcow2_read_extensions(bs, sizeof(header), ext_end)) {
+ ret = -EINVAL;
goto fail;
+ }
/* read the backing file name */
if (header.backing_file_offset != 0) {
len = header.backing_file_size;
- if (len > 1023)
+ if (len > 1023) {
len = 1023;
- if (bdrv_pread(bs->file, header.backing_file_offset, bs->backing_file, len) != len)
+ }
+ ret = bdrv_pread(bs->file, header.backing_file_offset,
+ bs->backing_file, len);
+ if (ret < 0) {
goto fail;
+ }
bs->backing_file[len] = '\0';
}
- if (qcow2_read_snapshots(bs) < 0)
+ if (qcow2_read_snapshots(bs) < 0) {
+ ret = -EINVAL;
goto fail;
+ }
+
+ /* Initialise locks */
+ qemu_co_mutex_init(&s->lock);
#ifdef DEBUG_ALLOC
- qcow2_check_refcounts(bs);
+ {
+ BdrvCheckResult result = {0};
+ qcow2_check_refcounts(bs, &result);
+ }
#endif
- return 0;
+ return ret;
fail:
qcow2_free_snapshots(bs);
qcow2_refcount_close(bs);
- qemu_free(s->l1_table);
- qemu_free(s->l2_cache);
- qemu_free(s->cluster_cache);
- qemu_free(s->cluster_data);
- return -1;
+ g_free(s->l1_table);
+ if (s->l2_table_cache) {
+ qcow2_cache_destroy(bs, s->l2_table_cache);
+ }
+ g_free(s->cluster_cache);
+ qemu_vfree(s->cluster_data);
+ return ret;
}
-static int qcow_set_key(BlockDriverState *bs, const char *key)
+static int qcow2_set_key(BlockDriverState *bs, const char *key)
{
BDRVQcowState *s = bs->opaque;
uint8_t keybuf[16];
return 0;
}
-static int qcow_is_allocated(BlockDriverState *bs, int64_t sector_num,
- int nb_sectors, int *pnum)
+static int qcow2_is_allocated(BlockDriverState *bs, int64_t sector_num,
+ int nb_sectors, int *pnum)
{
uint64_t cluster_offset;
int ret;
else
n1 = bs->total_sectors - sector_num;
- qemu_iovec_memset(qiov, 0, 512 * (nb_sectors - n1));
+ qemu_iovec_memset_skip(qiov, 0, 512 * (nb_sectors - n1), 512 * n1);
return n1;
}
-typedef struct QCowAIOCB {
- BlockDriverAIOCB common;
- int64_t sector_num;
- QEMUIOVector *qiov;
- int remaining_sectors;
- int cur_nr_sectors; /* number of sectors in current iteration */
- uint64_t bytes_done;
- uint64_t cluster_offset;
- uint8_t *cluster_data;
- BlockDriverAIOCB *hd_aiocb;
- QEMUIOVector hd_qiov;
- QEMUBH *bh;
- QCowL2Meta l2meta;
- QLIST_ENTRY(QCowAIOCB) next_depend;
-} QCowAIOCB;
-
-static void qcow_aio_cancel(BlockDriverAIOCB *blockacb)
-{
- QCowAIOCB *acb = container_of(blockacb, QCowAIOCB, common);
- if (acb->hd_aiocb)
- bdrv_aio_cancel(acb->hd_aiocb);
- qemu_aio_release(acb);
-}
-
-static AIOPool qcow_aio_pool = {
- .aiocb_size = sizeof(QCowAIOCB),
- .cancel = qcow_aio_cancel,
-};
-
-static void qcow_aio_read_cb(void *opaque, int ret);
-static void qcow_aio_read_bh(void *opaque)
-{
- QCowAIOCB *acb = opaque;
- qemu_bh_delete(acb->bh);
- acb->bh = NULL;
- qcow_aio_read_cb(opaque, 0);
-}
-
-static int qcow_schedule_bh(QEMUBHFunc *cb, QCowAIOCB *acb)
+static int qcow2_co_readv(BlockDriverState *bs, int64_t sector_num,
+ int remaining_sectors, QEMUIOVector *qiov)
{
- if (acb->bh)
- return -EIO;
-
- acb->bh = qemu_bh_new(cb, acb);
- if (!acb->bh)
- return -EIO;
-
- qemu_bh_schedule(acb->bh);
-
- return 0;
-}
-
-static void qcow_aio_read_cb(void *opaque, int ret)
-{
- QCowAIOCB *acb = opaque;
- BlockDriverState *bs = acb->common.bs;
BDRVQcowState *s = bs->opaque;
int index_in_cluster, n1;
+ int ret;
+ int cur_nr_sectors; /* number of sectors in current iteration */
+ uint64_t cluster_offset = 0;
+ uint64_t bytes_done = 0;
+ QEMUIOVector hd_qiov;
+ uint8_t *cluster_data = NULL;
- acb->hd_aiocb = NULL;
- if (ret < 0)
- goto done;
-
- /* post process the read buffer */
- if (!acb->cluster_offset) {
- /* nothing to do */
- } else if (acb->cluster_offset & QCOW_OFLAG_COMPRESSED) {
- /* nothing to do */
- } else {
- if (s->crypt_method) {
- qcow2_encrypt_sectors(s, acb->sector_num, acb->cluster_data,
- acb->cluster_data, acb->cur_nr_sectors, 0, &s->aes_decrypt_key);
- qemu_iovec_reset(&acb->hd_qiov);
- qemu_iovec_copy(&acb->hd_qiov, acb->qiov, acb->bytes_done,
- acb->cur_nr_sectors * 512);
- qemu_iovec_from_buffer(&acb->hd_qiov, acb->cluster_data,
- 512 * acb->cur_nr_sectors);
- }
- }
+ qemu_iovec_init(&hd_qiov, qiov->niov);
- acb->remaining_sectors -= acb->cur_nr_sectors;
- acb->sector_num += acb->cur_nr_sectors;
- acb->bytes_done += acb->cur_nr_sectors * 512;
+ qemu_co_mutex_lock(&s->lock);
- if (acb->remaining_sectors == 0) {
- /* request completed */
- ret = 0;
- goto done;
- }
+ while (remaining_sectors != 0) {
- /* prepare next AIO request */
- acb->cur_nr_sectors = acb->remaining_sectors;
- if (s->crypt_method) {
- acb->cur_nr_sectors = MIN(acb->cur_nr_sectors,
- QCOW_MAX_CRYPT_CLUSTERS * s->cluster_sectors);
- }
+ /* prepare next request */
+ cur_nr_sectors = remaining_sectors;
+ if (s->crypt_method) {
+ cur_nr_sectors = MIN(cur_nr_sectors,
+ QCOW_MAX_CRYPT_CLUSTERS * s->cluster_sectors);
+ }
- ret = qcow2_get_cluster_offset(bs, acb->sector_num << 9,
- &acb->cur_nr_sectors, &acb->cluster_offset);
- if (ret < 0) {
- goto done;
- }
+ ret = qcow2_get_cluster_offset(bs, sector_num << 9,
+ &cur_nr_sectors, &cluster_offset);
+ if (ret < 0) {
+ goto fail;
+ }
- index_in_cluster = acb->sector_num & (s->cluster_sectors - 1);
-
- qemu_iovec_reset(&acb->hd_qiov);
- qemu_iovec_copy(&acb->hd_qiov, acb->qiov, acb->bytes_done,
- acb->cur_nr_sectors * 512);
-
- if (!acb->cluster_offset) {
-
- if (bs->backing_hd) {
- /* read from the base image */
- n1 = qcow2_backing_read1(bs->backing_hd, &acb->hd_qiov,
- acb->sector_num, acb->cur_nr_sectors);
- if (n1 > 0) {
- BLKDBG_EVENT(bs->file, BLKDBG_READ_BACKING_AIO);
- acb->hd_aiocb = bdrv_aio_readv(bs->backing_hd, acb->sector_num,
- &acb->hd_qiov, acb->cur_nr_sectors,
- qcow_aio_read_cb, acb);
- if (acb->hd_aiocb == NULL)
- goto done;
+ index_in_cluster = sector_num & (s->cluster_sectors - 1);
+
+ qemu_iovec_reset(&hd_qiov);
+ qemu_iovec_copy(&hd_qiov, qiov, bytes_done,
+ cur_nr_sectors * 512);
+
+ if (!cluster_offset) {
+
+ if (bs->backing_hd) {
+ /* read from the base image */
+ n1 = qcow2_backing_read1(bs->backing_hd, &hd_qiov,
+ sector_num, cur_nr_sectors);
+ if (n1 > 0) {
+ BLKDBG_EVENT(bs->file, BLKDBG_READ_BACKING_AIO);
+ qemu_co_mutex_unlock(&s->lock);
+ ret = bdrv_co_readv(bs->backing_hd, sector_num,
+ n1, &hd_qiov);
+ qemu_co_mutex_lock(&s->lock);
+ if (ret < 0) {
+ goto fail;
+ }
+ }
} else {
- ret = qcow_schedule_bh(qcow_aio_read_bh, acb);
- if (ret < 0)
- goto done;
+ /* Note: in this case, no need to wait */
+ qemu_iovec_memset(&hd_qiov, 0, 512 * cur_nr_sectors);
}
+ } else if (cluster_offset & QCOW_OFLAG_COMPRESSED) {
+ /* add AIO support for compressed blocks ? */
+ ret = qcow2_decompress_cluster(bs, cluster_offset);
+ if (ret < 0) {
+ goto fail;
+ }
+
+ qemu_iovec_from_buffer(&hd_qiov,
+ s->cluster_cache + index_in_cluster * 512,
+ 512 * cur_nr_sectors);
} else {
- /* Note: in this case, no need to wait */
- qemu_iovec_memset(&acb->hd_qiov, 0, 512 * acb->cur_nr_sectors);
- ret = qcow_schedule_bh(qcow_aio_read_bh, acb);
- if (ret < 0)
- goto done;
- }
- } else if (acb->cluster_offset & QCOW_OFLAG_COMPRESSED) {
- /* add AIO support for compressed blocks ? */
- if (qcow2_decompress_cluster(bs, acb->cluster_offset) < 0)
- goto done;
-
- qemu_iovec_from_buffer(&acb->hd_qiov,
- s->cluster_cache + index_in_cluster * 512,
- 512 * acb->cur_nr_sectors);
-
- ret = qcow_schedule_bh(qcow_aio_read_bh, acb);
- if (ret < 0)
- goto done;
- } else {
- if ((acb->cluster_offset & 511) != 0) {
- ret = -EIO;
- goto done;
- }
+ if ((cluster_offset & 511) != 0) {
+ ret = -EIO;
+ goto fail;
+ }
- if (s->crypt_method) {
- /*
- * For encrypted images, read everything into a temporary
- * contiguous buffer on which the AES functions can work.
- */
- if (!acb->cluster_data) {
- acb->cluster_data =
- qemu_mallocz(QCOW_MAX_CRYPT_CLUSTERS * s->cluster_size);
+ if (s->crypt_method) {
+ /*
+ * For encrypted images, read everything into a temporary
+ * contiguous buffer on which the AES functions can work.
+ */
+ if (!cluster_data) {
+ cluster_data =
+ qemu_blockalign(bs, QCOW_MAX_CRYPT_CLUSTERS * s->cluster_size);
+ }
+
+ assert(cur_nr_sectors <=
+ QCOW_MAX_CRYPT_CLUSTERS * s->cluster_sectors);
+ qemu_iovec_reset(&hd_qiov);
+ qemu_iovec_add(&hd_qiov, cluster_data,
+ 512 * cur_nr_sectors);
}
- assert(acb->cur_nr_sectors <=
- QCOW_MAX_CRYPT_CLUSTERS * s->cluster_sectors);
- qemu_iovec_reset(&acb->hd_qiov);
- qemu_iovec_add(&acb->hd_qiov, acb->cluster_data,
- 512 * acb->cur_nr_sectors);
+ BLKDBG_EVENT(bs->file, BLKDBG_READ_AIO);
+ qemu_co_mutex_unlock(&s->lock);
+ ret = bdrv_co_readv(bs->file,
+ (cluster_offset >> 9) + index_in_cluster,
+ cur_nr_sectors, &hd_qiov);
+ qemu_co_mutex_lock(&s->lock);
+ if (ret < 0) {
+ goto fail;
+ }
+ if (s->crypt_method) {
+ qcow2_encrypt_sectors(s, sector_num, cluster_data,
+ cluster_data, cur_nr_sectors, 0, &s->aes_decrypt_key);
+ qemu_iovec_reset(&hd_qiov);
+ qemu_iovec_copy(&hd_qiov, qiov, bytes_done,
+ cur_nr_sectors * 512);
+ qemu_iovec_from_buffer(&hd_qiov, cluster_data,
+ 512 * cur_nr_sectors);
+ }
}
- BLKDBG_EVENT(bs->file, BLKDBG_READ_AIO);
- acb->hd_aiocb = bdrv_aio_readv(bs->file,
- (acb->cluster_offset >> 9) + index_in_cluster,
- &acb->hd_qiov, acb->cur_nr_sectors,
- qcow_aio_read_cb, acb);
- if (acb->hd_aiocb == NULL) {
- ret = -EIO;
- goto done;
- }
+ remaining_sectors -= cur_nr_sectors;
+ sector_num += cur_nr_sectors;
+ bytes_done += cur_nr_sectors * 512;
}
+ ret = 0;
- return;
-done:
- acb->common.cb(acb->common.opaque, ret);
- qemu_iovec_destroy(&acb->hd_qiov);
- qemu_aio_release(acb);
-}
-
-static QCowAIOCB *qcow_aio_setup(BlockDriverState *bs,
- int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
- BlockDriverCompletionFunc *cb, void *opaque, int is_write)
-{
- QCowAIOCB *acb;
-
- acb = qemu_aio_get(&qcow_aio_pool, bs, cb, opaque);
- if (!acb)
- return NULL;
- acb->hd_aiocb = NULL;
- acb->sector_num = sector_num;
- acb->qiov = qiov;
-
- qemu_iovec_init(&acb->hd_qiov, qiov->niov);
-
- acb->bytes_done = 0;
- acb->remaining_sectors = nb_sectors;
- acb->cur_nr_sectors = 0;
- acb->cluster_offset = 0;
- acb->l2meta.nb_clusters = 0;
- QLIST_INIT(&acb->l2meta.dependent_requests);
- return acb;
-}
-
-static BlockDriverAIOCB *qcow_aio_readv(BlockDriverState *bs,
- int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
- BlockDriverCompletionFunc *cb, void *opaque)
-{
- QCowAIOCB *acb;
+fail:
+ qemu_co_mutex_unlock(&s->lock);
- acb = qcow_aio_setup(bs, sector_num, qiov, nb_sectors, cb, opaque, 0);
- if (!acb)
- return NULL;
+ qemu_iovec_destroy(&hd_qiov);
+ qemu_vfree(cluster_data);
- qcow_aio_read_cb(acb, 0);
- return &acb->common;
+ return ret;
}
-static void qcow_aio_write_cb(void *opaque, int ret);
-
-static void run_dependent_requests(QCowL2Meta *m)
+static void run_dependent_requests(BDRVQcowState *s, QCowL2Meta *m)
{
- QCowAIOCB *req;
- QCowAIOCB *next;
-
/* Take the request off the list of running requests */
if (m->nb_clusters != 0) {
QLIST_REMOVE(m, next_in_flight);
}
/* Restart all dependent requests */
- QLIST_FOREACH_SAFE(req, &m->dependent_requests, next_depend, next) {
- qcow_aio_write_cb(req, 0);
+ if (!qemu_co_queue_empty(&m->dependent_requests)) {
+ qemu_co_mutex_unlock(&s->lock);
+ while(qemu_co_queue_next(&m->dependent_requests));
+ qemu_co_mutex_lock(&s->lock);
}
-
- /* Empty the list for the next part of the request */
- QLIST_INIT(&m->dependent_requests);
}
-static void qcow_aio_write_cb(void *opaque, int ret)
+static int qcow2_co_writev(BlockDriverState *bs,
+ int64_t sector_num,
+ int remaining_sectors,
+ QEMUIOVector *qiov)
{
- QCowAIOCB *acb = opaque;
- BlockDriverState *bs = acb->common.bs;
BDRVQcowState *s = bs->opaque;
int index_in_cluster;
int n_end;
+ int ret;
+ int cur_nr_sectors; /* number of sectors in current iteration */
+ uint64_t cluster_offset;
+ QEMUIOVector hd_qiov;
+ uint64_t bytes_done = 0;
+ uint8_t *cluster_data = NULL;
+ QCowL2Meta l2meta = {
+ .nb_clusters = 0,
+ };
- acb->hd_aiocb = NULL;
+ qemu_co_queue_init(&l2meta.dependent_requests);
- if (ret >= 0) {
- ret = qcow2_alloc_cluster_link_l2(bs, &acb->l2meta);
- }
+ qemu_iovec_init(&hd_qiov, qiov->niov);
- run_dependent_requests(&acb->l2meta);
+ s->cluster_cache_offset = -1; /* disable compressed cache */
- if (ret < 0)
- goto done;
+ qemu_co_mutex_lock(&s->lock);
- acb->remaining_sectors -= acb->cur_nr_sectors;
- acb->sector_num += acb->cur_nr_sectors;
- acb->bytes_done += acb->cur_nr_sectors * 512;
+ while (remaining_sectors != 0) {
- if (acb->remaining_sectors == 0) {
- /* request completed */
- ret = 0;
- goto done;
- }
+ index_in_cluster = sector_num & (s->cluster_sectors - 1);
+ n_end = index_in_cluster + remaining_sectors;
+ if (s->crypt_method &&
+ n_end > QCOW_MAX_CRYPT_CLUSTERS * s->cluster_sectors) {
+ n_end = QCOW_MAX_CRYPT_CLUSTERS * s->cluster_sectors;
+ }
- index_in_cluster = acb->sector_num & (s->cluster_sectors - 1);
- n_end = index_in_cluster + acb->remaining_sectors;
- if (s->crypt_method &&
- n_end > QCOW_MAX_CRYPT_CLUSTERS * s->cluster_sectors)
- n_end = QCOW_MAX_CRYPT_CLUSTERS * s->cluster_sectors;
+ ret = qcow2_alloc_cluster_offset(bs, sector_num << 9,
+ index_in_cluster, n_end, &cur_nr_sectors, &l2meta);
+ if (ret < 0) {
+ goto fail;
+ }
- ret = qcow2_alloc_cluster_offset(bs, acb->sector_num << 9,
- index_in_cluster, n_end, &acb->cur_nr_sectors, &acb->l2meta);
- if (ret < 0) {
- goto done;
- }
+ cluster_offset = l2meta.cluster_offset;
+ assert((cluster_offset & 511) == 0);
- acb->cluster_offset = acb->l2meta.cluster_offset;
+ qemu_iovec_reset(&hd_qiov);
+ qemu_iovec_copy(&hd_qiov, qiov, bytes_done,
+ cur_nr_sectors * 512);
- /* Need to wait for another request? If so, we are done for now. */
- if (acb->l2meta.nb_clusters == 0 && acb->l2meta.depends_on != NULL) {
- QLIST_INSERT_HEAD(&acb->l2meta.depends_on->dependent_requests,
- acb, next_depend);
- return;
- }
+ if (s->crypt_method) {
+ if (!cluster_data) {
+ cluster_data = qemu_blockalign(bs, QCOW_MAX_CRYPT_CLUSTERS *
+ s->cluster_size);
+ }
- assert((acb->cluster_offset & 511) == 0);
+ assert(hd_qiov.size <=
+ QCOW_MAX_CRYPT_CLUSTERS * s->cluster_size);
+ qemu_iovec_to_buffer(&hd_qiov, cluster_data);
- qemu_iovec_reset(&acb->hd_qiov);
- qemu_iovec_copy(&acb->hd_qiov, acb->qiov, acb->bytes_done,
- acb->cur_nr_sectors * 512);
+ qcow2_encrypt_sectors(s, sector_num, cluster_data,
+ cluster_data, cur_nr_sectors, 1, &s->aes_encrypt_key);
- if (s->crypt_method) {
- if (!acb->cluster_data) {
- acb->cluster_data = qemu_mallocz(QCOW_MAX_CRYPT_CLUSTERS *
- s->cluster_size);
+ qemu_iovec_reset(&hd_qiov);
+ qemu_iovec_add(&hd_qiov, cluster_data,
+ cur_nr_sectors * 512);
}
- assert(acb->hd_qiov.size <= QCOW_MAX_CRYPT_CLUSTERS * s->cluster_size);
- qemu_iovec_to_buffer(&acb->hd_qiov, acb->cluster_data);
+ BLKDBG_EVENT(bs->file, BLKDBG_WRITE_AIO);
+ qemu_co_mutex_unlock(&s->lock);
+ ret = bdrv_co_writev(bs->file,
+ (cluster_offset >> 9) + index_in_cluster,
+ cur_nr_sectors, &hd_qiov);
+ qemu_co_mutex_lock(&s->lock);
+ if (ret < 0) {
+ goto fail;
+ }
- qcow2_encrypt_sectors(s, acb->sector_num, acb->cluster_data,
- acb->cluster_data, acb->cur_nr_sectors, 1, &s->aes_encrypt_key);
+ ret = qcow2_alloc_cluster_link_l2(bs, &l2meta);
+ if (ret < 0) {
+ goto fail;
+ }
- qemu_iovec_reset(&acb->hd_qiov);
- qemu_iovec_add(&acb->hd_qiov, acb->cluster_data,
- acb->cur_nr_sectors * 512);
- }
+ run_dependent_requests(s, &l2meta);
- BLKDBG_EVENT(bs->file, BLKDBG_WRITE_AIO);
- acb->hd_aiocb = bdrv_aio_writev(bs->file,
- (acb->cluster_offset >> 9) + index_in_cluster,
- &acb->hd_qiov, acb->cur_nr_sectors,
- qcow_aio_write_cb, acb);
- if (acb->hd_aiocb == NULL) {
- ret = -EIO;
- goto fail;
+ remaining_sectors -= cur_nr_sectors;
+ sector_num += cur_nr_sectors;
+ bytes_done += cur_nr_sectors * 512;
}
-
- return;
+ ret = 0;
fail:
- if (acb->l2meta.nb_clusters != 0) {
- QLIST_REMOVE(&acb->l2meta, next_in_flight);
- }
-done:
- acb->common.cb(acb->common.opaque, ret);
- qemu_iovec_destroy(&acb->hd_qiov);
- qemu_aio_release(acb);
-}
-
-static BlockDriverAIOCB *qcow_aio_writev(BlockDriverState *bs,
- int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
- BlockDriverCompletionFunc *cb, void *opaque)
-{
- BDRVQcowState *s = bs->opaque;
- QCowAIOCB *acb;
+ run_dependent_requests(s, &l2meta);
- s->cluster_cache_offset = -1; /* disable compressed cache */
+ qemu_co_mutex_unlock(&s->lock);
- acb = qcow_aio_setup(bs, sector_num, qiov, nb_sectors, cb, opaque, 1);
- if (!acb)
- return NULL;
+ qemu_iovec_destroy(&hd_qiov);
+ qemu_vfree(cluster_data);
- qcow_aio_write_cb(acb, 0);
- return &acb->common;
+ return ret;
}
-static void qcow_close(BlockDriverState *bs)
+static void qcow2_close(BlockDriverState *bs)
{
BDRVQcowState *s = bs->opaque;
- qemu_free(s->l1_table);
- qemu_free(s->l2_cache);
- qemu_free(s->cluster_cache);
- qemu_free(s->cluster_data);
+ g_free(s->l1_table);
+
+ qcow2_cache_flush(bs, s->l2_table_cache);
+ qcow2_cache_flush(bs, s->refcount_block_cache);
+
+ qcow2_cache_destroy(bs, s->l2_table_cache);
+ qcow2_cache_destroy(bs, s->refcount_block_cache);
+
+ g_free(s->cluster_cache);
+ qemu_vfree(s->cluster_data);
qcow2_refcount_close(bs);
}
/* Prepare the backing file format extension if needed */
if (backing_fmt) {
ext_backing_fmt.len = cpu_to_be32(strlen(backing_fmt));
- ext_backing_fmt.magic = cpu_to_be32(QCOW_EXT_MAGIC_BACKING_FORMAT);
+ ext_backing_fmt.magic = cpu_to_be32(QCOW2_EXT_MAGIC_BACKING_FORMAT);
backing_fmt_len = ((sizeof(ext_backing_fmt)
+ strlen(backing_fmt) + 7) & ~7);
}
nb_sectors = bdrv_getlength(bs) >> 9;
offset = 0;
- QLIST_INIT(&meta.dependent_requests);
+ qemu_co_queue_init(&meta.dependent_requests);
meta.cluster_offset = 0;
while (nb_sectors) {
/* There are no dependent requests, but we need to remove our request
* from the list of in-flight requests */
- run_dependent_requests(&meta);
+ run_dependent_requests(bs->opaque, &meta);
/* TODO Preallocate data if requested */
return 0;
}
-static int qcow_create2(const char *filename, int64_t total_size,
- const char *backing_file, const char *backing_format,
- int flags, size_t cluster_size, int prealloc,
- QEMUOptionParameter *options)
+static int qcow2_create2(const char *filename, int64_t total_size,
+ const char *backing_file, const char *backing_format,
+ int flags, size_t cluster_size, int prealloc,
+ QEMUOptionParameter *options)
{
/* Calulate cluster_bits */
int cluster_bits;
(1 << cluster_bits) != cluster_size)
{
error_report(
- "Cluster size must be a power of two between %d and %dk\n",
+ "Cluster size must be a power of two between %d and %dk",
1 << MIN_CLUSTER_BITS, 1 << (MAX_CLUSTER_BITS - 10));
return -EINVAL;
}
}
/* Write an empty refcount table */
- refcount_table = qemu_mallocz(cluster_size);
+ refcount_table = g_malloc0(cluster_size);
ret = bdrv_pwrite(bs, cluster_size, refcount_table, cluster_size);
- qemu_free(refcount_table);
+ g_free(refcount_table);
if (ret < 0) {
goto out;
*/
BlockDriver* drv = bdrv_find_format("qcow2");
assert(drv != NULL);
- ret = bdrv_open(bs, filename, BDRV_O_RDWR | BDRV_O_NO_FLUSH, drv);
+ ret = bdrv_open(bs, filename,
+ BDRV_O_RDWR | BDRV_O_CACHE_WB | BDRV_O_NO_FLUSH, drv);
if (ret < 0) {
goto out;
}
return ret;
}
-static int qcow_create(const char *filename, QEMUOptionParameter *options)
+static int qcow2_create(const char *filename, QEMUOptionParameter *options)
{
const char *backing_file = NULL;
const char *backing_fmt = NULL;
uint64_t sectors = 0;
int flags = 0;
- size_t cluster_size = 65536;
+ size_t cluster_size = DEFAULT_CLUSTER_SIZE;
int prealloc = 0;
/* Read out options */
return -EINVAL;
}
- return qcow_create2(filename, sectors, backing_file, backing_fmt, flags,
- cluster_size, prealloc, options);
+ return qcow2_create2(filename, sectors, backing_file, backing_fmt, flags,
+ cluster_size, prealloc, options);
}
-static int qcow_make_empty(BlockDriverState *bs)
+static int qcow2_make_empty(BlockDriverState *bs)
{
#if 0
/* XXX: not correct */
return 0;
}
+static coroutine_fn int qcow2_co_discard(BlockDriverState *bs,
+ int64_t sector_num, int nb_sectors)
+{
+ int ret;
+ BDRVQcowState *s = bs->opaque;
+
+ qemu_co_mutex_lock(&s->lock);
+ ret = qcow2_discard_clusters(bs, sector_num << BDRV_SECTOR_BITS,
+ nb_sectors);
+ qemu_co_mutex_unlock(&s->lock);
+ return ret;
+}
+
static int qcow2_truncate(BlockDriverState *bs, int64_t offset)
{
BDRVQcowState *s = bs->opaque;
/* XXX: put compressed sectors first, then all the cluster aligned
tables to avoid losing bytes in alignment */
-static int qcow_write_compressed(BlockDriverState *bs, int64_t sector_num,
- const uint8_t *buf, int nb_sectors)
+static int qcow2_write_compressed(BlockDriverState *bs, int64_t sector_num,
+ const uint8_t *buf, int nb_sectors)
{
BDRVQcowState *s = bs->opaque;
z_stream strm;
if (nb_sectors != s->cluster_sectors)
return -EINVAL;
- out_buf = qemu_malloc(s->cluster_size + (s->cluster_size / 1000) + 128);
+ out_buf = g_malloc(s->cluster_size + (s->cluster_size / 1000) + 128);
/* best compression, small window, no zlib header */
memset(&strm, 0, sizeof(strm));
Z_DEFLATED, -12,
9, Z_DEFAULT_STRATEGY);
if (ret != 0) {
- qemu_free(out_buf);
- return -1;
+ ret = -EINVAL;
+ goto fail;
}
strm.avail_in = s->cluster_size;
ret = deflate(&strm, Z_FINISH);
if (ret != Z_STREAM_END && ret != Z_OK) {
- qemu_free(out_buf);
deflateEnd(&strm);
- return -1;
+ ret = -EINVAL;
+ goto fail;
}
out_len = strm.next_out - out_buf;
if (ret != Z_STREAM_END || out_len >= s->cluster_size) {
/* could not compress: write normal cluster */
- bdrv_write(bs, sector_num, buf, s->cluster_sectors);
+ ret = bdrv_write(bs, sector_num, buf, s->cluster_sectors);
+ if (ret < 0) {
+ goto fail;
+ }
} else {
cluster_offset = qcow2_alloc_compressed_cluster_offset(bs,
sector_num << 9, out_len);
- if (!cluster_offset)
- return -1;
+ if (!cluster_offset) {
+ ret = -EIO;
+ goto fail;
+ }
cluster_offset &= s->cluster_offset_mask;
BLKDBG_EVENT(bs->file, BLKDBG_WRITE_COMPRESSED);
- if (bdrv_pwrite(bs->file, cluster_offset, out_buf, out_len) != out_len) {
- qemu_free(out_buf);
- return -1;
+ ret = bdrv_pwrite(bs->file, cluster_offset, out_buf, out_len);
+ if (ret < 0) {
+ goto fail;
}
}
- qemu_free(out_buf);
- return 0;
+ ret = 0;
+fail:
+ g_free(out_buf);
+ return ret;
}
-static void qcow_flush(BlockDriverState *bs)
+static int qcow2_co_flush_to_os(BlockDriverState *bs)
{
- bdrv_flush(bs->file);
+ BDRVQcowState *s = bs->opaque;
+ int ret;
+
+ qemu_co_mutex_lock(&s->lock);
+ ret = qcow2_cache_flush(bs, s->l2_table_cache);
+ if (ret < 0) {
+ qemu_co_mutex_unlock(&s->lock);
+ return ret;
+ }
+
+ ret = qcow2_cache_flush(bs, s->refcount_block_cache);
+ if (ret < 0) {
+ qemu_co_mutex_unlock(&s->lock);
+ return ret;
+ }
+ qemu_co_mutex_unlock(&s->lock);
+
+ return 0;
}
-static BlockDriverAIOCB *qcow_aio_flush(BlockDriverState *bs,
- BlockDriverCompletionFunc *cb, void *opaque)
+static int qcow2_co_flush_to_disk(BlockDriverState *bs)
{
- return bdrv_aio_flush(bs->file, cb, opaque);
+ return bdrv_co_flush(bs->file);
}
-static int64_t qcow_vm_state_offset(BDRVQcowState *s)
+static int64_t qcow2_vm_state_offset(BDRVQcowState *s)
{
return (int64_t)s->l1_vm_state_index << (s->cluster_bits + s->l2_bits);
}
-static int qcow_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
+static int qcow2_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
{
BDRVQcowState *s = bs->opaque;
bdi->cluster_size = s->cluster_size;
- bdi->vm_state_offset = qcow_vm_state_offset(s);
+ bdi->vm_state_offset = qcow2_vm_state_offset(s);
return 0;
}
-static int qcow_check(BlockDriverState *bs, BdrvCheckResult *result)
+static int qcow2_check(BlockDriverState *bs, BdrvCheckResult *result)
{
return qcow2_check_refcounts(bs, result);
}
}
#endif
-static int qcow_save_vmstate(BlockDriverState *bs, const uint8_t *buf,
- int64_t pos, int size)
+static int qcow2_save_vmstate(BlockDriverState *bs, const uint8_t *buf,
+ int64_t pos, int size)
{
BDRVQcowState *s = bs->opaque;
int growable = bs->growable;
BLKDBG_EVENT(bs->file, BLKDBG_VMSTATE_SAVE);
bs->growable = 1;
- ret = bdrv_pwrite(bs, qcow_vm_state_offset(s) + pos, buf, size);
+ ret = bdrv_pwrite(bs, qcow2_vm_state_offset(s) + pos, buf, size);
bs->growable = growable;
return ret;
}
-static int qcow_load_vmstate(BlockDriverState *bs, uint8_t *buf,
- int64_t pos, int size)
+static int qcow2_load_vmstate(BlockDriverState *bs, uint8_t *buf,
+ int64_t pos, int size)
{
BDRVQcowState *s = bs->opaque;
int growable = bs->growable;
BLKDBG_EVENT(bs->file, BLKDBG_VMSTATE_LOAD);
bs->growable = 1;
- ret = bdrv_pread(bs, qcow_vm_state_offset(s) + pos, buf, size);
+ ret = bdrv_pread(bs, qcow2_vm_state_offset(s) + pos, buf, size);
bs->growable = growable;
return ret;
}
-static QEMUOptionParameter qcow_create_options[] = {
+static QEMUOptionParameter qcow2_create_options[] = {
{
.name = BLOCK_OPT_SIZE,
.type = OPT_SIZE,
{
.name = BLOCK_OPT_CLUSTER_SIZE,
.type = OPT_SIZE,
- .help = "qcow2 cluster size"
+ .help = "qcow2 cluster size",
+ .value = { .n = DEFAULT_CLUSTER_SIZE },
},
{
.name = BLOCK_OPT_PREALLOC,
};
static BlockDriver bdrv_qcow2 = {
- .format_name = "qcow2",
- .instance_size = sizeof(BDRVQcowState),
- .bdrv_probe = qcow_probe,
- .bdrv_open = qcow_open,
- .bdrv_close = qcow_close,
- .bdrv_create = qcow_create,
- .bdrv_flush = qcow_flush,
- .bdrv_is_allocated = qcow_is_allocated,
- .bdrv_set_key = qcow_set_key,
- .bdrv_make_empty = qcow_make_empty,
-
- .bdrv_aio_readv = qcow_aio_readv,
- .bdrv_aio_writev = qcow_aio_writev,
- .bdrv_aio_flush = qcow_aio_flush,
-
+ .format_name = "qcow2",
+ .instance_size = sizeof(BDRVQcowState),
+ .bdrv_probe = qcow2_probe,
+ .bdrv_open = qcow2_open,
+ .bdrv_close = qcow2_close,
+ .bdrv_create = qcow2_create,
+ .bdrv_is_allocated = qcow2_is_allocated,
+ .bdrv_set_key = qcow2_set_key,
+ .bdrv_make_empty = qcow2_make_empty,
+
+ .bdrv_co_readv = qcow2_co_readv,
+ .bdrv_co_writev = qcow2_co_writev,
+ .bdrv_co_flush_to_os = qcow2_co_flush_to_os,
+ .bdrv_co_flush_to_disk = qcow2_co_flush_to_disk,
+
+ .bdrv_co_discard = qcow2_co_discard,
.bdrv_truncate = qcow2_truncate,
- .bdrv_write_compressed = qcow_write_compressed,
+ .bdrv_write_compressed = qcow2_write_compressed,
.bdrv_snapshot_create = qcow2_snapshot_create,
.bdrv_snapshot_goto = qcow2_snapshot_goto,
.bdrv_snapshot_delete = qcow2_snapshot_delete,
.bdrv_snapshot_list = qcow2_snapshot_list,
.bdrv_snapshot_load_tmp = qcow2_snapshot_load_tmp,
- .bdrv_get_info = qcow_get_info,
+ .bdrv_get_info = qcow2_get_info,
- .bdrv_save_vmstate = qcow_save_vmstate,
- .bdrv_load_vmstate = qcow_load_vmstate,
+ .bdrv_save_vmstate = qcow2_save_vmstate,
+ .bdrv_load_vmstate = qcow2_load_vmstate,
.bdrv_change_backing_file = qcow2_change_backing_file,
- .create_options = qcow_create_options,
- .bdrv_check = qcow_check,
+ .create_options = qcow2_create_options,
+ .bdrv_check = qcow2_check,
};
static void bdrv_qcow2_init(void)
projects / qemu.git / blobdiff