#include <zlib.h>
#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)
{
BDRVQcowState *s = bs->opaque;
QCowExtension ext;
uint64_t offset;
+ int ret;
#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);
+ printf("attempting to read extended header in offset %lu\n", offset);
#endif
- if (bdrv_pread(s->hd, offset, &ext, sizeof(ext)) != sizeof(ext)) {
- fprintf(stderr, "qcow_handle_extension: ERROR: pread fail from offset %llu\n",
- (unsigned long long)offset);
+ if (bdrv_pread(bs->file, offset, &ext, sizeof(ext)) != sizeof(ext)) {
+ fprintf(stderr, "qcow2_read_extension: ERROR: "
+ "pread fail from offset %" PRIu64 "\n",
+ offset);
return 1;
}
be32_to_cpus(&ext.magic);
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",
ext.len, sizeof(bs->backing_format));
return 2;
}
- if (bdrv_pread(s->hd, offset , bs->backing_format,
+ if (bdrv_pread(bs->file, offset , bs->backing_format,
ext.len) != ext.len)
return 3;
bs->backing_format[ext.len] = '\0';
break;
default:
- /* unknown magic -- just skip it */
- offset = ((offset + ext.len + 7) & ~7);
+ /* unknown magic - save it in case we need to rewrite the header */
+ {
+ Qcow2UnknownHeaderExtension *uext;
+
+ uext = g_malloc0(sizeof(*uext) + ext.len);
+ uext->magic = ext.magic;
+ uext->len = ext.len;
+ QLIST_INSERT_HEAD(&s->unknown_header_ext, uext, next);
+
+ ret = bdrv_pread(bs->file, offset , uext->data, uext->len);
+ if (ret < 0) {
+ return ret;
+ }
+
+ offset = ((offset + ext.len + 7) & ~7);
+ }
break;
}
}
return 0;
}
+static void cleanup_unknown_header_ext(BlockDriverState *bs)
+{
+ BDRVQcowState *s = bs->opaque;
+ Qcow2UnknownHeaderExtension *uext, *next;
+
+ QLIST_FOREACH_SAFE(uext, &s->unknown_header_ext, next, next) {
+ QLIST_REMOVE(uext, next);
+ g_free(uext);
+ }
+}
-static int qcow_open(BlockDriverState *bs, const char *filename, int flags)
+static int qcow2_open(BlockDriverState *bs, int flags)
{
BDRVQcowState *s = bs->opaque;
- int len, i, shift, ret;
+ int len, i, ret = 0;
QCowHeader header;
uint64_t ext_end;
+ bool writethrough;
- ret = bdrv_file_open(&s->hd, filename, flags);
- if (ret < 0)
- return ret;
- if (bdrv_pread(s->hd, 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);
/* read the level 1 table */
s->l1_size = header.l1_size;
- shift = s->cluster_bits + s->l2_bits;
- s->l1_vm_state_index = (header.size + (1LL << shift) - 1) >> shift;
+ 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(s->hd, 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;
+ s->flags = flags;
- 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(s->hd, 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)
+
+ ret = qcow2_read_snapshots(bs);
+ if (ret < 0) {
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:
+ cleanup_unknown_header_ext(bs);
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);
- bdrv_delete(s->hd);
- 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 coroutine_fn qcow2_co_is_allocated(BlockDriverState *bs,
+ int64_t sector_num, int nb_sectors, int *pnum)
{
+ BDRVQcowState *s = bs->opaque;
uint64_t cluster_offset;
+ int ret;
*pnum = nb_sectors;
- cluster_offset = qcow2_get_cluster_offset(bs, sector_num << 9, pnum);
+ /* FIXME We can get errors here, but the bdrv_co_is_allocated interface
+ * can't pass them on today */
+ qemu_co_mutex_lock(&s->lock);
+ ret = qcow2_get_cluster_offset(bs, sector_num << 9, pnum, &cluster_offset);
+ qemu_co_mutex_unlock(&s->lock);
+ if (ret < 0) {
+ *pnum = 0;
+ }
return (cluster_offset != 0);
}
/* handle reading after the end of the backing file */
-int qcow2_backing_read1(BlockDriverState *bs,
- int64_t sector_num, uint8_t *buf, int nb_sectors)
+int qcow2_backing_read1(BlockDriverState *bs, QEMUIOVector *qiov,
+ int64_t sector_num, int nb_sectors)
{
int n1;
if ((sector_num + nb_sectors) <= bs->total_sectors)
n1 = 0;
else
n1 = bs->total_sectors - sector_num;
- memset(buf + n1 * 512, 0, 512 * (nb_sectors - n1));
- return n1;
-}
-
-typedef struct QCowAIOCB {
- BlockDriverAIOCB common;
- int64_t sector_num;
- QEMUIOVector *qiov;
- uint8_t *buf;
- void *orig_buf;
- int nb_sectors;
- int n;
- uint64_t cluster_offset;
- uint8_t *cluster_data;
- BlockDriverAIOCB *hd_aiocb;
- struct iovec hd_iov;
- QEMUIOVector hd_qiov;
- QEMUBH *bh;
- QCowL2Meta l2meta;
- QLIST_ENTRY(QCowAIOCB) next_depend;
-} QCowAIOCB;
-
-static void qcow_aio_cancel(BlockDriverAIOCB *blockacb)
-{
- QCowAIOCB *acb = (QCowAIOCB *)blockacb;
- 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)
-{
- if (acb->bh)
- return -EIO;
-
- acb->bh = qemu_bh_new(cb, acb);
- if (!acb->bh)
- return -EIO;
- qemu_bh_schedule(acb->bh);
+ qemu_iovec_memset_skip(qiov, 0, 512 * (nb_sectors - n1), 512 * n1);
- return 0;
+ return n1;
}
-static void qcow_aio_read_cb(void *opaque, int ret)
+static coroutine_fn int qcow2_co_readv(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, 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;
+ qemu_iovec_init(&hd_qiov, qiov->niov);
- /* post process the read buffer */
- if (!acb->cluster_offset) {
- /* nothing to do */
- } else if (acb->cluster_offset & QCOW_OFLAG_COMPRESSED) {
- /* nothing to do */
- } else {
+ qemu_co_mutex_lock(&s->lock);
+
+ while (remaining_sectors != 0) {
+
+ /* prepare next request */
+ cur_nr_sectors = remaining_sectors;
if (s->crypt_method) {
- qcow2_encrypt_sectors(s, acb->sector_num, acb->buf, acb->buf,
- acb->n, 0,
- &s->aes_decrypt_key);
+ cur_nr_sectors = MIN(cur_nr_sectors,
+ QCOW_MAX_CRYPT_CLUSTERS * s->cluster_sectors);
}
- }
- acb->nb_sectors -= acb->n;
- acb->sector_num += acb->n;
- acb->buf += acb->n * 512;
-
- if (acb->nb_sectors == 0) {
- /* request completed */
- ret = 0;
- goto done;
- }
+ ret = qcow2_get_cluster_offset(bs, sector_num << 9,
+ &cur_nr_sectors, &cluster_offset);
+ if (ret < 0) {
+ goto fail;
+ }
- /* prepare next AIO request */
- acb->n = acb->nb_sectors;
- acb->cluster_offset =
- qcow2_get_cluster_offset(bs, acb->sector_num << 9, &acb->n);
- index_in_cluster = acb->sector_num & (s->cluster_sectors - 1);
-
- if (!acb->cluster_offset) {
- if (bs->backing_hd) {
- /* read from the base image */
- n1 = qcow2_backing_read1(bs->backing_hd, acb->sector_num,
- acb->buf, acb->n);
- if (n1 > 0) {
- acb->hd_iov.iov_base = (void *)acb->buf;
- acb->hd_iov.iov_len = acb->n * 512;
- qemu_iovec_init_external(&acb->hd_qiov, &acb->hd_iov, 1);
- acb->hd_aiocb = bdrv_aio_readv(bs->backing_hd, acb->sector_num,
- &acb->hd_qiov, acb->n,
- 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 */
- memset(acb->buf, 0, 512 * acb->n);
- 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(s, acb->cluster_offset) < 0)
- goto done;
- memcpy(acb->buf,
- s->cluster_cache + index_in_cluster * 512, 512 * acb->n);
- 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;
+ }
- acb->hd_iov.iov_base = (void *)acb->buf;
- acb->hd_iov.iov_len = acb->n * 512;
- qemu_iovec_init_external(&acb->hd_qiov, &acb->hd_iov, 1);
- acb->hd_aiocb = bdrv_aio_readv(s->hd,
- (acb->cluster_offset >> 9) + index_in_cluster,
- &acb->hd_qiov, acb->n, qcow_aio_read_cb, acb);
- if (acb->hd_aiocb == NULL)
- goto done;
- }
+ 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);
+ }
- return;
-done:
- if (acb->qiov->niov > 1) {
- qemu_iovec_from_buffer(acb->qiov, acb->orig_buf, acb->qiov->size);
- qemu_vfree(acb->orig_buf);
- }
- acb->common.cb(acb->common.opaque, ret);
- qemu_aio_release(acb);
-}
+ 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);
+ }
+ }
-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;
- if (qiov->niov > 1) {
- acb->buf = acb->orig_buf = qemu_blockalign(bs, qiov->size);
- if (is_write)
- qemu_iovec_to_buffer(qiov, acb->buf);
- } else {
- acb->buf = (uint8_t *)qiov->iov->iov_base;
+ remaining_sectors -= cur_nr_sectors;
+ sector_num += cur_nr_sectors;
+ bytes_done += cur_nr_sectors * 512;
}
- acb->nb_sectors = nb_sectors;
- acb->n = 0;
- acb->cluster_offset = 0;
- acb->l2meta.nb_clusters = 0;
- QLIST_INIT(&acb->l2meta.dependent_requests);
- return acb;
-}
+ ret = 0;
-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.
- * Can't use QLIST_FOREACH here - the next link might not be the same
- * any more after the callback (request could depend on a different
- * request now)
- */
- for (req = m->dependent_requests.lh_first; req != NULL; req = next) {
- next = req->next_depend.le_next;
- qcow_aio_write_cb(req, 0);
+ /* Restart all dependent requests */
+ if (!qemu_co_queue_empty(&m->dependent_requests)) {
+ qemu_co_mutex_unlock(&s->lock);
+ qemu_co_queue_restart_all(&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 coroutine_fn 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;
- const uint8_t *src_buf;
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->cluster_offset, &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->nb_sectors -= acb->n;
- acb->sector_num += acb->n;
- acb->buf += acb->n * 512;
+ while (remaining_sectors != 0) {
- if (acb->nb_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->nb_sectors;
- if (s->crypt_method &&
- n_end > QCOW_MAX_CRYPT_CLUSTERS * s->cluster_sectors)
- n_end = QCOW_MAX_CRYPT_CLUSTERS * s->cluster_sectors;
-
- acb->cluster_offset = qcow2_alloc_cluster_offset(bs, acb->sector_num << 9,
- index_in_cluster,
- n_end, &acb->n, &acb->l2meta);
-
- /* Need to wait for another request? If so, we are done for now. */
- if (!acb->cluster_offset && acb->l2meta.depends_on != NULL) {
- QLIST_INSERT_HEAD(&acb->l2meta.depends_on->dependent_requests,
- acb, next_depend);
- return;
- }
+ ret = qcow2_alloc_cluster_offset(bs, sector_num << 9,
+ index_in_cluster, n_end, &cur_nr_sectors, &l2meta);
+ if (ret < 0) {
+ goto fail;
+ }
- if (!acb->cluster_offset || (acb->cluster_offset & 511) != 0) {
- ret = -EIO;
- goto done;
- }
- if (s->crypt_method) {
- if (!acb->cluster_data) {
- acb->cluster_data = qemu_mallocz(QCOW_MAX_CRYPT_CLUSTERS *
- s->cluster_size);
+ cluster_offset = l2meta.cluster_offset;
+ assert((cluster_offset & 511) == 0);
+
+ qemu_iovec_reset(&hd_qiov);
+ qemu_iovec_copy(&hd_qiov, qiov, bytes_done,
+ cur_nr_sectors * 512);
+
+ if (s->crypt_method) {
+ if (!cluster_data) {
+ cluster_data = qemu_blockalign(bs, QCOW_MAX_CRYPT_CLUSTERS *
+ s->cluster_size);
+ }
+
+ assert(hd_qiov.size <=
+ QCOW_MAX_CRYPT_CLUSTERS * s->cluster_size);
+ qemu_iovec_to_buffer(&hd_qiov, cluster_data);
+
+ qcow2_encrypt_sectors(s, sector_num, cluster_data,
+ cluster_data, cur_nr_sectors, 1, &s->aes_encrypt_key);
+
+ qemu_iovec_reset(&hd_qiov);
+ qemu_iovec_add(&hd_qiov, cluster_data,
+ cur_nr_sectors * 512);
}
- qcow2_encrypt_sectors(s, acb->sector_num, acb->cluster_data, acb->buf,
- acb->n, 1, &s->aes_encrypt_key);
- src_buf = acb->cluster_data;
- } else {
- src_buf = acb->buf;
+
+ 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;
+ }
+
+ ret = qcow2_alloc_cluster_link_l2(bs, &l2meta);
+ if (ret < 0) {
+ goto fail;
+ }
+
+ run_dependent_requests(s, &l2meta);
+
+ remaining_sectors -= cur_nr_sectors;
+ sector_num += cur_nr_sectors;
+ bytes_done += cur_nr_sectors * 512;
}
- acb->hd_iov.iov_base = (void *)src_buf;
- acb->hd_iov.iov_len = acb->n * 512;
- qemu_iovec_init_external(&acb->hd_qiov, &acb->hd_iov, 1);
- acb->hd_aiocb = bdrv_aio_writev(s->hd,
- (acb->cluster_offset >> 9) + index_in_cluster,
- &acb->hd_qiov, acb->n,
- qcow_aio_write_cb, acb);
- if (acb->hd_aiocb == NULL)
- goto done;
-
- return;
-
-done:
- if (acb->qiov->niov > 1)
- qemu_vfree(acb->orig_buf);
- acb->common.cb(acb->common.opaque, ret);
- qemu_aio_release(acb);
+ ret = 0;
+
+fail:
+ run_dependent_requests(s, &l2meta);
+
+ qemu_co_mutex_unlock(&s->lock);
+
+ qemu_iovec_destroy(&hd_qiov);
+ qemu_vfree(cluster_data);
+
+ return ret;
}
-static BlockDriverAIOCB *qcow_aio_writev(BlockDriverState *bs,
- int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
- BlockDriverCompletionFunc *cb, void *opaque)
+static void qcow2_close(BlockDriverState *bs)
{
BDRVQcowState *s = bs->opaque;
- QCowAIOCB *acb;
+ g_free(s->l1_table);
- s->cluster_cache_offset = -1; /* disable compressed cache */
+ qcow2_cache_flush(bs, s->l2_table_cache);
+ qcow2_cache_flush(bs, s->refcount_block_cache);
- acb = qcow_aio_setup(bs, sector_num, qiov, nb_sectors, cb, opaque, 1);
- if (!acb)
- return NULL;
+ qcow2_cache_destroy(bs, s->l2_table_cache);
+ qcow2_cache_destroy(bs, s->refcount_block_cache);
- qcow_aio_write_cb(acb, 0);
- return &acb->common;
+ cleanup_unknown_header_ext(bs);
+ g_free(s->cluster_cache);
+ qemu_vfree(s->cluster_data);
+ qcow2_refcount_close(bs);
+ qcow2_free_snapshots(bs);
}
-static void qcow_close(BlockDriverState *bs)
+static void qcow2_invalidate_cache(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);
- qcow2_refcount_close(bs);
- bdrv_delete(s->hd);
+ int flags = s->flags;
+ AES_KEY aes_encrypt_key;
+ AES_KEY aes_decrypt_key;
+ uint32_t crypt_method = 0;
+
+ /*
+ * Backing files are read-only which makes all of their metadata immutable,
+ * that means we don't have to worry about reopening them here.
+ */
+
+ if (s->crypt_method) {
+ crypt_method = s->crypt_method;
+ memcpy(&aes_encrypt_key, &s->aes_encrypt_key, sizeof(aes_encrypt_key));
+ memcpy(&aes_decrypt_key, &s->aes_decrypt_key, sizeof(aes_decrypt_key));
+ }
+
+ qcow2_close(bs);
+
+ memset(s, 0, sizeof(BDRVQcowState));
+ qcow2_open(bs, flags);
+
+ if (crypt_method) {
+ s->crypt_method = crypt_method;
+ memcpy(&s->aes_encrypt_key, &aes_encrypt_key, sizeof(aes_encrypt_key));
+ memcpy(&s->aes_decrypt_key, &aes_decrypt_key, sizeof(aes_decrypt_key));
+ }
+}
+
+static size_t header_ext_add(char *buf, uint32_t magic, const void *s,
+ size_t len, size_t buflen)
+{
+ QCowExtension *ext_backing_fmt = (QCowExtension*) buf;
+ size_t ext_len = sizeof(QCowExtension) + ((len + 7) & ~7);
+
+ if (buflen < ext_len) {
+ return -ENOSPC;
+ }
+
+ *ext_backing_fmt = (QCowExtension) {
+ .magic = cpu_to_be32(magic),
+ .len = cpu_to_be32(len),
+ };
+ memcpy(buf + sizeof(QCowExtension), s, len);
+
+ return ext_len;
}
/*
- * Updates the variable length parts of the qcow2 header, i.e. the backing file
- * name and all extensions. qcow2 was not designed to allow such changes, so if
- * we run out of space (we can only use the first cluster) this function may
- * fail.
+ * Updates the qcow2 header, including the variable length parts of it, i.e.
+ * the backing file name and all extensions. qcow2 was not designed to allow
+ * such changes, so if we run out of space (we can only use the first cluster)
+ * this function may fail.
*
* Returns 0 on success, -errno in error cases.
*/
-static int qcow2_update_ext_header(BlockDriverState *bs,
- const char *backing_file, const char *backing_fmt)
+int qcow2_update_header(BlockDriverState *bs)
{
- size_t backing_file_len = 0;
- size_t backing_fmt_len = 0;
BDRVQcowState *s = bs->opaque;
- QCowExtension ext_backing_fmt = {0, 0};
+ QCowHeader *header;
+ char *buf;
+ size_t buflen = s->cluster_size;
int ret;
+ uint64_t total_size;
+ uint32_t refcount_table_clusters;
+ Qcow2UnknownHeaderExtension *uext;
- /* Backing file format doesn't make sense without a backing file */
- if (backing_fmt && !backing_file) {
- return -EINVAL;
- }
+ buf = qemu_blockalign(bs, buflen);
+ memset(buf, 0, s->cluster_size);
- /* 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);
- backing_fmt_len = ((sizeof(ext_backing_fmt)
- + strlen(backing_fmt) + 7) & ~7);
- }
+ /* Header structure */
+ header = (QCowHeader*) buf;
- /* Check if we can fit the new header into the first cluster */
- if (backing_file) {
- backing_file_len = strlen(backing_file);
+ if (buflen < sizeof(*header)) {
+ ret = -ENOSPC;
+ goto fail;
}
- size_t header_size = sizeof(QCowHeader) + backing_file_len
- + backing_fmt_len;
+ total_size = bs->total_sectors * BDRV_SECTOR_SIZE;
+ refcount_table_clusters = s->refcount_table_size >> (s->cluster_bits - 3);
+
+ *header = (QCowHeader) {
+ .magic = cpu_to_be32(QCOW_MAGIC),
+ .version = cpu_to_be32(QCOW_VERSION),
+ .backing_file_offset = 0,
+ .backing_file_size = 0,
+ .cluster_bits = cpu_to_be32(s->cluster_bits),
+ .size = cpu_to_be64(total_size),
+ .crypt_method = cpu_to_be32(s->crypt_method_header),
+ .l1_size = cpu_to_be32(s->l1_size),
+ .l1_table_offset = cpu_to_be64(s->l1_table_offset),
+ .refcount_table_offset = cpu_to_be64(s->refcount_table_offset),
+ .refcount_table_clusters = cpu_to_be32(refcount_table_clusters),
+ .nb_snapshots = cpu_to_be32(s->nb_snapshots),
+ .snapshots_offset = cpu_to_be64(s->snapshots_offset),
+ };
+
+ buf += sizeof(*header);
+ buflen -= sizeof(*header);
+
+ /* Backing file format header extension */
+ if (*bs->backing_format) {
+ ret = header_ext_add(buf, QCOW2_EXT_MAGIC_BACKING_FORMAT,
+ bs->backing_format, strlen(bs->backing_format),
+ buflen);
+ if (ret < 0) {
+ goto fail;
+ }
- if (header_size > s->cluster_size) {
- return -ENOSPC;
+ buf += ret;
+ buflen -= ret;
}
- /* Rewrite backing file name and qcow2 extensions */
- size_t ext_size = header_size - sizeof(QCowHeader);
- uint8_t buf[ext_size];
- size_t offset = 0;
- size_t backing_file_offset = 0;
-
- if (backing_file) {
- if (backing_fmt) {
- int padding = backing_fmt_len -
- (sizeof(ext_backing_fmt) + strlen(backing_fmt));
-
- memcpy(buf + offset, &ext_backing_fmt, sizeof(ext_backing_fmt));
- offset += sizeof(ext_backing_fmt);
-
- memcpy(buf + offset, backing_fmt, strlen(backing_fmt));
- offset += strlen(backing_fmt);
-
- memset(buf + offset, 0, padding);
- offset += padding;
+ /* Keep unknown header extensions */
+ QLIST_FOREACH(uext, &s->unknown_header_ext, next) {
+ ret = header_ext_add(buf, uext->magic, uext->data, uext->len, buflen);
+ if (ret < 0) {
+ goto fail;
}
- memcpy(buf + offset, backing_file, backing_file_len);
- backing_file_offset = sizeof(QCowHeader) + offset;
+ buf += ret;
+ buflen -= ret;
}
- ret = bdrv_pwrite(s->hd, sizeof(QCowHeader), buf, ext_size);
+ /* End of header extensions */
+ ret = header_ext_add(buf, QCOW2_EXT_MAGIC_END, NULL, 0, buflen);
if (ret < 0) {
goto fail;
}
- /* Update header fields */
- uint64_t be_backing_file_offset = cpu_to_be64(backing_file_offset);
- uint32_t be_backing_file_size = cpu_to_be32(backing_file_len);
+ buf += ret;
+ buflen -= ret;
- ret = bdrv_pwrite(s->hd, offsetof(QCowHeader, backing_file_offset),
- &be_backing_file_offset, sizeof(uint64_t));
- if (ret < 0) {
- goto fail;
+ /* Backing file name */
+ if (*bs->backing_file) {
+ size_t backing_file_len = strlen(bs->backing_file);
+
+ if (buflen < backing_file_len) {
+ ret = -ENOSPC;
+ goto fail;
+ }
+
+ strncpy(buf, bs->backing_file, buflen);
+
+ header->backing_file_offset = cpu_to_be64(buf - ((char*) header));
+ header->backing_file_size = cpu_to_be32(backing_file_len);
}
- ret = bdrv_pwrite(s->hd, offsetof(QCowHeader, backing_file_size),
- &be_backing_file_size, sizeof(uint32_t));
+ /* Write the new header */
+ ret = bdrv_pwrite(bs->file, 0, header, s->cluster_size);
if (ret < 0) {
goto fail;
}
ret = 0;
fail:
+ qemu_vfree(header);
return ret;
}
static int qcow2_change_backing_file(BlockDriverState *bs,
const char *backing_file, const char *backing_fmt)
{
- return qcow2_update_ext_header(bs, backing_file, backing_fmt);
-}
-
-static int get_bits_from_size(size_t size)
-{
- int res = 0;
-
- if (size == 0) {
- return -1;
+ /* Backing file format doesn't make sense without a backing file */
+ if (backing_fmt && !backing_file) {
+ return -EINVAL;
}
- while (size != 1) {
- /* Not a power of two */
- if (size & 1) {
- return -1;
- }
-
- size >>= 1;
- res++;
- }
+ pstrcpy(bs->backing_file, sizeof(bs->backing_file), backing_file ?: "");
+ pstrcpy(bs->backing_format, sizeof(bs->backing_format), backing_fmt ?: "");
- return res;
+ return qcow2_update_header(bs);
}
-
static int preallocate(BlockDriverState *bs)
{
- BDRVQcowState *s = bs->opaque;
- uint64_t cluster_offset = 0;
uint64_t nb_sectors;
uint64_t offset;
int num;
+ int ret;
QCowL2Meta meta;
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) {
num = MIN(nb_sectors, INT_MAX >> 9);
- cluster_offset = qcow2_alloc_cluster_offset(bs, offset, 0, num, &num,
- &meta);
-
- if (cluster_offset == 0) {
- return -1;
+ ret = qcow2_alloc_cluster_offset(bs, offset, 0, num, &num, &meta);
+ if (ret < 0) {
+ return ret;
}
- if (qcow2_alloc_cluster_link_l2(bs, cluster_offset, &meta) < 0) {
- qcow2_free_any_clusters(bs, cluster_offset, meta.nb_clusters);
- return -1;
+ ret = qcow2_alloc_cluster_link_l2(bs, &meta);
+ if (ret < 0) {
+ qcow2_free_any_clusters(bs, meta.cluster_offset, meta.nb_clusters);
+ return ret;
}
/* 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 */
* all of the allocated clusters (otherwise we get failing reads after
* EOF). Extend the image to the last allocated sector.
*/
- if (cluster_offset != 0) {
+ if (meta.cluster_offset != 0) {
uint8_t buf[512];
memset(buf, 0, 512);
- bdrv_write(s->hd, (cluster_offset >> 9) + num - 1, buf, 1);
+ ret = bdrv_write(bs->file, (meta.cluster_offset >> 9) + num - 1, buf, 1);
+ if (ret < 0) {
+ return ret;
+ }
}
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)
+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)
{
+ /* Calculate cluster_bits */
+ int cluster_bits;
+ cluster_bits = ffs(cluster_size) - 1;
+ if (cluster_bits < MIN_CLUSTER_BITS || cluster_bits > MAX_CLUSTER_BITS ||
+ (1 << cluster_bits) != cluster_size)
+ {
+ error_report(
+ "Cluster size must be a power of two between %d and %dk",
+ 1 << MIN_CLUSTER_BITS, 1 << (MAX_CLUSTER_BITS - 10));
+ return -EINVAL;
+ }
- int fd, header_size, backing_filename_len, l1_size, i, shift, l2_bits;
- int ref_clusters, backing_format_len = 0;
- int rounded_ext_bf_len = 0;
+ /*
+ * Open the image file and write a minimal qcow2 header.
+ *
+ * We keep things simple and start with a zero-sized image. We also
+ * do without refcount blocks or a L1 table for now. We'll fix the
+ * inconsistency later.
+ *
+ * We do need a refcount table because growing the refcount table means
+ * allocating two new refcount blocks - the seconds of which would be at
+ * 2 GB for 64k clusters, and we don't want to have a 2 GB initial file
+ * size for any qcow2 image.
+ */
+ BlockDriverState* bs;
QCowHeader header;
- uint64_t tmp, offset;
- QCowCreateState s1, *s = &s1;
- QCowExtension ext_bf = {0, 0};
+ uint8_t* refcount_table;
+ int ret;
+ ret = bdrv_create_file(filename, options);
+ if (ret < 0) {
+ return ret;
+ }
- memset(s, 0, sizeof(*s));
+ ret = bdrv_file_open(&bs, filename, BDRV_O_RDWR);
+ if (ret < 0) {
+ return ret;
+ }
- fd = open(filename, O_WRONLY | O_CREAT | O_TRUNC | O_BINARY, 0644);
- if (fd < 0)
- return -1;
+ /* Write the header */
memset(&header, 0, sizeof(header));
header.magic = cpu_to_be32(QCOW_MAGIC);
header.version = cpu_to_be32(QCOW_VERSION);
- header.size = cpu_to_be64(total_size * 512);
- header_size = sizeof(header);
- backing_filename_len = 0;
- if (backing_file) {
- if (backing_format) {
- ext_bf.magic = QCOW_EXT_MAGIC_BACKING_FORMAT;
- backing_format_len = strlen(backing_format);
- ext_bf.len = backing_format_len;
- rounded_ext_bf_len = (sizeof(ext_bf) + ext_bf.len + 7) & ~7;
- header_size += rounded_ext_bf_len;
- }
- header.backing_file_offset = cpu_to_be64(header_size);
- backing_filename_len = strlen(backing_file);
- header.backing_file_size = cpu_to_be32(backing_filename_len);
- header_size += backing_filename_len;
- }
-
- /* Cluster size */
- s->cluster_bits = get_bits_from_size(cluster_size);
- if (s->cluster_bits < MIN_CLUSTER_BITS ||
- s->cluster_bits > MAX_CLUSTER_BITS)
- {
- fprintf(stderr, "Cluster size must be a power of two between "
- "%d and %dk\n",
- 1 << MIN_CLUSTER_BITS,
- 1 << (MAX_CLUSTER_BITS - 10));
- return -EINVAL;
- }
- s->cluster_size = 1 << s->cluster_bits;
+ header.cluster_bits = cpu_to_be32(cluster_bits);
+ header.size = cpu_to_be64(0);
+ header.l1_table_offset = cpu_to_be64(0);
+ header.l1_size = cpu_to_be32(0);
+ header.refcount_table_offset = cpu_to_be64(cluster_size);
+ header.refcount_table_clusters = cpu_to_be32(1);
- header.cluster_bits = cpu_to_be32(s->cluster_bits);
- header_size = (header_size + 7) & ~7;
if (flags & BLOCK_FLAG_ENCRYPT) {
header.crypt_method = cpu_to_be32(QCOW_CRYPT_AES);
} else {
header.crypt_method = cpu_to_be32(QCOW_CRYPT_NONE);
}
- l2_bits = s->cluster_bits - 3;
- shift = s->cluster_bits + l2_bits;
- l1_size = (((total_size * 512) + (1LL << shift) - 1) >> shift);
- offset = align_offset(header_size, s->cluster_size);
- s->l1_table_offset = offset;
- header.l1_table_offset = cpu_to_be64(s->l1_table_offset);
- header.l1_size = cpu_to_be32(l1_size);
- offset += align_offset(l1_size * sizeof(uint64_t), s->cluster_size);
-
- s->refcount_table = qemu_mallocz(s->cluster_size);
-
- s->refcount_table_offset = offset;
- header.refcount_table_offset = cpu_to_be64(offset);
- header.refcount_table_clusters = cpu_to_be32(1);
- offset += s->cluster_size;
- s->refcount_block_offset = offset;
-
- /* count how many refcount blocks needed */
- tmp = offset >> s->cluster_bits;
- ref_clusters = (tmp >> (s->cluster_bits - REFCOUNT_SHIFT)) + 1;
- for (i=0; i < ref_clusters; i++) {
- s->refcount_table[i] = cpu_to_be64(offset);
- offset += s->cluster_size;
+
+ ret = bdrv_pwrite(bs, 0, &header, sizeof(header));
+ if (ret < 0) {
+ goto out;
}
- s->refcount_block = qemu_mallocz(ref_clusters * s->cluster_size);
+ /* Write an empty refcount table */
+ refcount_table = g_malloc0(cluster_size);
+ ret = bdrv_pwrite(bs, cluster_size, refcount_table, cluster_size);
+ g_free(refcount_table);
- /* update refcounts */
- qcow2_create_refcount_update(s, 0, header_size);
- qcow2_create_refcount_update(s, s->l1_table_offset,
- l1_size * sizeof(uint64_t));
- qcow2_create_refcount_update(s, s->refcount_table_offset, s->cluster_size);
- qcow2_create_refcount_update(s, s->refcount_block_offset,
- ref_clusters * s->cluster_size);
+ if (ret < 0) {
+ goto out;
+ }
- /* write all the data */
- write(fd, &header, sizeof(header));
- if (backing_file) {
- if (backing_format_len) {
- char zero[16];
- int padding = rounded_ext_bf_len - (ext_bf.len + sizeof(ext_bf));
-
- memset(zero, 0, sizeof(zero));
- cpu_to_be32s(&ext_bf.magic);
- cpu_to_be32s(&ext_bf.len);
- write(fd, &ext_bf, sizeof(ext_bf));
- write(fd, backing_format, backing_format_len);
- if (padding > 0) {
- write(fd, zero, padding);
- }
- }
- write(fd, backing_file, backing_filename_len);
+ bdrv_close(bs);
+
+ /*
+ * And now open the image and make it consistent first (i.e. increase the
+ * refcount of the cluster that is occupied by the header and the refcount
+ * table)
+ */
+ BlockDriver* drv = bdrv_find_format("qcow2");
+ assert(drv != NULL);
+ ret = bdrv_open(bs, filename,
+ BDRV_O_RDWR | BDRV_O_CACHE_WB | BDRV_O_NO_FLUSH, drv);
+ if (ret < 0) {
+ goto out;
}
- lseek(fd, s->l1_table_offset, SEEK_SET);
- tmp = 0;
- for(i = 0;i < l1_size; i++) {
- write(fd, &tmp, sizeof(tmp));
+
+ ret = qcow2_alloc_clusters(bs, 2 * cluster_size);
+ if (ret < 0) {
+ goto out;
+
+ } else if (ret != 0) {
+ error_report("Huh, first cluster in empty image is already in use?");
+ abort();
}
- lseek(fd, s->refcount_table_offset, SEEK_SET);
- write(fd, s->refcount_table, s->cluster_size);
- lseek(fd, s->refcount_block_offset, SEEK_SET);
- write(fd, s->refcount_block, ref_clusters * s->cluster_size);
+ /* Okay, now that we have a valid image, let's give it the right size */
+ ret = bdrv_truncate(bs, total_size * BDRV_SECTOR_SIZE);
+ if (ret < 0) {
+ goto out;
+ }
- qemu_free(s->refcount_table);
- qemu_free(s->refcount_block);
- close(fd);
+ /* Want a backing file? There you go.*/
+ if (backing_file) {
+ ret = bdrv_change_backing_file(bs, backing_file, backing_format);
+ if (ret < 0) {
+ goto out;
+ }
+ }
- /* Preallocate metadata */
+ /* And if we're supposed to preallocate metadata, do that now */
if (prealloc) {
- BlockDriverState *bs;
- bs = bdrv_new("");
- bdrv_open(bs, filename, BDRV_O_CACHE_WB);
- preallocate(bs);
- bdrv_close(bs);
+ ret = preallocate(bs);
+ if (ret < 0) {
+ goto out;
+ }
}
- return 0;
+ ret = 0;
+out:
+ bdrv_delete(bs);
+ 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);
+ 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 */
int ret;
memset(s->l1_table, 0, l1_length);
- if (bdrv_pwrite(s->hd, s->l1_table_offset, s->l1_table, l1_length) < 0)
+ if (bdrv_pwrite(bs->file, s->l1_table_offset, s->l1_table, l1_length) < 0)
return -1;
- ret = bdrv_truncate(s->hd, s->l1_table_offset + l1_length);
+ ret = bdrv_truncate(bs->file, s->l1_table_offset + l1_length);
if (ret < 0)
return ret;
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;
+ int ret, new_l1_size;
+
+ if (offset & 511) {
+ return -EINVAL;
+ }
+
+ /* cannot proceed if image has snapshots */
+ if (s->nb_snapshots) {
+ return -ENOTSUP;
+ }
+
+ /* shrinking is currently not supported */
+ if (offset < bs->total_sectors * 512) {
+ return -ENOTSUP;
+ }
+
+ new_l1_size = size_to_l1(s, offset);
+ ret = qcow2_grow_l1_table(bs, new_l1_size, true);
+ if (ret < 0) {
+ return ret;
+ }
+
+ /* write updated header.size */
+ offset = cpu_to_be64(offset);
+ ret = bdrv_pwrite_sync(bs->file, offsetof(QCowHeader, size),
+ &offset, sizeof(uint64_t));
+ if (ret < 0) {
+ return ret;
+ }
+
+ s->l1_vm_state_index = new_l1_size;
+ return 0;
+}
+
/* 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 == 0) {
/* align end of file to a sector boundary to ease reading with
sector based I/Os */
- cluster_offset = bdrv_getlength(s->hd);
+ cluster_offset = bdrv_getlength(bs->file);
cluster_offset = (cluster_offset + 511) & ~511;
- bdrv_truncate(s->hd, cluster_offset);
+ bdrv_truncate(bs->file, cluster_offset);
return 0;
}
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;
- if (bdrv_pwrite(s->hd, cluster_offset, out_buf, out_len) != out_len) {
- qemu_free(out_buf);
- return -1;
+ BLKDBG_EVENT(bs->file, BLKDBG_WRITE_COMPRESSED);
+ 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 coroutine_fn int qcow2_co_flush_to_os(BlockDriverState *bs)
{
BDRVQcowState *s = bs->opaque;
- bdrv_flush(s->hd);
+ 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 coroutine_fn int qcow2_co_flush_to_disk(BlockDriverState *bs)
{
- BDRVQcowState *s = bs->opaque;
-
- return bdrv_aio_flush(s->hd, 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)
+static int qcow2_check(BlockDriverState *bs, BdrvCheckResult *result)
{
- return qcow2_check_refcounts(bs);
+ return qcow2_check_refcounts(bs, result);
}
#if 0
int64_t nb_clusters, k, k1, size;
int refcount;
- size = bdrv_getlength(s->hd);
+ size = bdrv_getlength(bs->file);
nb_clusters = size_to_clusters(s, size);
for(k = 0; k < nb_clusters;) {
k1 = k;
k++;
while (k < nb_clusters && get_refcount(bs, k) == refcount)
k++;
- printf("%lld: refcount=%d nb=%lld\n", k, refcount, k - k1);
+ printf("%" PRId64 ": refcount=%d nb=%" PRId64 "\n", k, refcount,
+ k - k1);
}
}
#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;
+ int ret;
+ BLKDBG_EVENT(bs->file, BLKDBG_VMSTATE_SAVE);
bs->growable = 1;
- 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 size;
+ 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;
int ret;
+ 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,
- .bdrv_write_compressed = qcow_write_compressed,
+ .format_name = "qcow2",
+ .instance_size = sizeof(BDRVQcowState),
+ .bdrv_probe = qcow2_probe,
+ .bdrv_open = qcow2_open,
+ .bdrv_close = qcow2_close,
+ .bdrv_create = qcow2_create,
+ .bdrv_co_is_allocated = qcow2_co_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 = 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_get_info = qcow_get_info,
+ .bdrv_snapshot_load_tmp = qcow2_snapshot_load_tmp,
+ .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,
+ .bdrv_invalidate_cache = qcow2_invalidate_cache,
+
+ .create_options = qcow2_create_options,
+ .bdrv_check = qcow2_check,
};
static void bdrv_qcow2_init(void)
projects / qemu.git / blobdiff