#include "qemu-common.h"
#include "block_int.h"
#include "module.h"
+#include "migration.h"
+#include <zlib.h>
#define VMDK3_MAGIC (('C' << 24) | ('O' << 16) | ('W' << 8) | 'D')
#define VMDK4_MAGIC (('K' << 24) | ('D' << 16) | ('M' << 8) | 'V')
+#define VMDK4_COMPRESSION_DEFLATE 1
+#define VMDK4_FLAG_RGD (1 << 1)
+#define VMDK4_FLAG_COMPRESS (1 << 16)
+#define VMDK4_FLAG_MARKER (1 << 17)
typedef struct {
uint32_t version;
int64_t desc_offset;
int64_t desc_size;
int32_t num_gtes_per_gte;
- int64_t rgd_offset;
int64_t gd_offset;
+ int64_t rgd_offset;
int64_t grain_offset;
char filler[1];
char check_bytes[4];
+ uint16_t compressAlgorithm;
} QEMU_PACKED VMDK4Header;
#define L2_CACHE_SIZE 16
typedef struct VmdkExtent {
BlockDriverState *file;
bool flat;
+ bool compressed;
+ bool has_marker;
int64_t sectors;
int64_t end_sector;
int64_t flat_start_offset;
} VmdkExtent;
typedef struct BDRVVmdkState {
+ CoMutex lock;
int desc_offset;
bool cid_updated;
uint32_t parent_cid;
int num_extents;
/* Extent array with num_extents entries, ascend ordered by address */
VmdkExtent *extents;
+ Error *migration_blocker;
} BDRVVmdkState;
typedef struct VmdkMetaData {
int valid;
} VmdkMetaData;
+typedef struct VmdkGrainMarker {
+ uint64_t lba;
+ uint32_t size;
+ uint8_t data[0];
+} VmdkGrainMarker;
+
static int vmdk_probe(const uint8_t *buf, int buf_size, const char *filename)
{
uint32_t magic;
{
int i;
BDRVVmdkState *s = bs->opaque;
+ VmdkExtent *e;
for (i = 0; i < s->num_extents; i++) {
- g_free(s->extents[i].l1_table);
- g_free(s->extents[i].l2_cache);
- g_free(s->extents[i].l1_backup_table);
+ e = &s->extents[i];
+ g_free(e->l1_table);
+ g_free(e->l2_cache);
+ g_free(e->l1_backup_table);
+ if (e->file != bs->file) {
+ bdrv_delete(e->file);
+ }
}
g_free(s->extents);
}
static uint32_t vmdk_read_cid(BlockDriverState *bs, int parent)
{
char desc[DESC_SIZE];
- uint32_t cid;
+ uint32_t cid = 0xffffffff;
const char *p_name, *cid_str;
size_t cid_str_size;
BDRVVmdkState *s = bs->opaque;
+ int ret;
- if (bdrv_pread(bs->file, s->desc_offset, desc, DESC_SIZE) != DESC_SIZE) {
+ ret = bdrv_pread(bs->file, s->desc_offset, desc, DESC_SIZE);
+ if (ret < 0) {
return 0;
}
cid_str_size = sizeof("CID");
}
+ desc[DESC_SIZE - 1] = '\0';
p_name = strstr(desc, cid_str);
if (p_name != NULL) {
p_name += cid_str_size;
char desc[DESC_SIZE], tmp_desc[DESC_SIZE];
char *p_name, *tmp_str;
BDRVVmdkState *s = bs->opaque;
+ int ret;
- memset(desc, 0, sizeof(desc));
- if (bdrv_pread(bs->file, s->desc_offset, desc, DESC_SIZE) != DESC_SIZE) {
- return -EIO;
+ ret = bdrv_pread(bs->file, s->desc_offset, desc, DESC_SIZE);
+ if (ret < 0) {
+ return ret;
}
+ desc[DESC_SIZE - 1] = '\0';
tmp_str = strstr(desc, "parentCID");
+ if (tmp_str == NULL) {
+ return -EINVAL;
+ }
+
pstrcpy(tmp_desc, sizeof(tmp_desc), tmp_str);
p_name = strstr(desc, "CID");
if (p_name != NULL) {
pstrcat(desc, sizeof(desc), tmp_desc);
}
- if (bdrv_pwrite_sync(bs->file, s->desc_offset, desc, DESC_SIZE) < 0) {
- return -EIO;
+ ret = bdrv_pwrite_sync(bs->file, s->desc_offset, desc, DESC_SIZE);
+ if (ret < 0) {
+ return ret;
}
+
return 0;
}
char *p_name;
char desc[DESC_SIZE + 1];
BDRVVmdkState *s = bs->opaque;
+ int ret;
desc[DESC_SIZE] = '\0';
- if (bdrv_pread(bs->file, s->desc_offset, desc, DESC_SIZE) != DESC_SIZE) {
- return -1;
+ ret = bdrv_pread(bs->file, s->desc_offset, desc, DESC_SIZE);
+ if (ret < 0) {
+ return ret;
}
p_name = strstr(desc, "parentFileNameHint");
p_name += sizeof("parentFileNameHint") + 1;
end_name = strchr(p_name, '\"');
if (end_name == NULL) {
- return -1;
+ return -EINVAL;
}
if ((end_name - p_name) > sizeof(bs->backing_file) - 1) {
- return -1;
+ return -EINVAL;
}
pstrcpy(bs->backing_file, end_name - p_name + 1, p_name);
return ret;
}
+static int vmdk_open_desc_file(BlockDriverState *bs, int flags,
+ int64_t desc_offset);
+
static int vmdk_open_vmdk4(BlockDriverState *bs,
BlockDriverState *file,
int flags)
uint32_t l1_size, l1_entry_sectors;
VMDK4Header header;
VmdkExtent *extent;
+ int64_t l1_backup_offset = 0;
ret = bdrv_pread(file, sizeof(magic), &header, sizeof(header));
if (ret < 0) {
return ret;
}
+ if (header.capacity == 0 && header.desc_offset) {
+ return vmdk_open_desc_file(bs, flags, header.desc_offset << 9);
+ }
l1_entry_sectors = le32_to_cpu(header.num_gtes_per_gte)
* le64_to_cpu(header.granularity);
- if (l1_entry_sectors <= 0) {
+ if (l1_entry_sectors == 0) {
return -EINVAL;
}
l1_size = (le64_to_cpu(header.capacity) + l1_entry_sectors - 1)
/ l1_entry_sectors;
+ if (le32_to_cpu(header.flags) & VMDK4_FLAG_RGD) {
+ l1_backup_offset = le64_to_cpu(header.rgd_offset) << 9;
+ }
extent = vmdk_add_extent(bs, file, false,
le64_to_cpu(header.capacity),
le64_to_cpu(header.gd_offset) << 9,
- le64_to_cpu(header.rgd_offset) << 9,
+ l1_backup_offset,
l1_size,
le32_to_cpu(header.num_gtes_per_gte),
le64_to_cpu(header.granularity));
+ extent->compressed =
+ le16_to_cpu(header.compressAlgorithm) == VMDK4_COMPRESSION_DEFLATE;
+ extent->has_marker = le32_to_cpu(header.flags) & VMDK4_FLAG_MARKER;
ret = vmdk_init_tables(bs, extent);
if (ret) {
/* free extent allocated by vmdk_add_extent */
extent = vmdk_add_extent(bs, extent_file, true, sectors,
0, 0, 0, 0, sectors);
- extent->flat_start_offset = flat_offset;
+ extent->flat_start_offset = flat_offset << 9;
} else if (!strcmp(type, "SPARSE")) {
/* SPARSE extent */
ret = vmdk_open_sparse(bs, extent_file, bs->open_flags);
return 0;
}
-static int vmdk_open_desc_file(BlockDriverState *bs, int flags)
+static int vmdk_open_desc_file(BlockDriverState *bs, int flags,
+ int64_t desc_offset)
{
int ret;
char buf[2048];
char ct[128];
BDRVVmdkState *s = bs->opaque;
- ret = bdrv_pread(bs->file, 0, buf, sizeof(buf));
+ ret = bdrv_pread(bs->file, desc_offset, buf, sizeof(buf));
if (ret < 0) {
return ret;
}
return -ENOTSUP;
}
s->desc_offset = 0;
- ret = vmdk_parse_extents(buf, bs, bs->file->filename);
- if (ret) {
- return ret;
- }
-
- /* try to open parent images, if exist */
- if (vmdk_parent_open(bs)) {
- g_free(s->extents);
- return -EINVAL;
- }
- s->parent_cid = vmdk_read_cid(bs, 1);
- return 0;
+ return vmdk_parse_extents(buf, bs, bs->file->filename);
}
static int vmdk_open(BlockDriverState *bs, int flags)
if (vmdk_open_sparse(bs, bs->file, flags) == 0) {
s->desc_offset = 0x200;
- /* try to open parent images, if exist */
- ret = vmdk_parent_open(bs);
+ } else {
+ ret = vmdk_open_desc_file(bs, flags, 0);
if (ret) {
- vmdk_free_extents(bs);
- return ret;
+ goto fail;
}
- s->parent_cid = vmdk_read_cid(bs, 1);
- return 0;
- } else {
- return vmdk_open_desc_file(bs, flags);
}
+ /* try to open parent images, if exist */
+ ret = vmdk_parent_open(bs);
+ if (ret) {
+ goto fail;
+ }
+ s->parent_cid = vmdk_read_cid(bs, 1);
+ qemu_co_mutex_init(&s->lock);
+
+ /* Disable migration when VMDK images are used */
+ error_set(&s->migration_blocker,
+ QERR_BLOCK_FORMAT_FEATURE_NOT_SUPPORTED,
+ "vmdk", bs->device_name, "live migration");
+ migrate_add_blocker(s->migration_blocker);
+
+ return 0;
+
+fail:
+ vmdk_free_extents(bs);
+ return ret;
}
static int get_whole_cluster(BlockDriverState *bs,
/* Avoid the L2 tables update for the images that have snapshots. */
*cluster_offset = bdrv_getlength(extent->file);
- bdrv_truncate(
- extent->file,
- *cluster_offset + (extent->cluster_sectors << 9)
- );
+ if (!extent->compressed) {
+ bdrv_truncate(
+ extent->file,
+ *cluster_offset + (extent->cluster_sectors << 9)
+ );
+ }
*cluster_offset >>= 9;
tmp = cpu_to_le32(*cluster_offset);
return NULL;
}
-static int vmdk_is_allocated(BlockDriverState *bs, int64_t sector_num,
- int nb_sectors, int *pnum)
+static int coroutine_fn vmdk_co_is_allocated(BlockDriverState *bs,
+ int64_t sector_num, int nb_sectors, int *pnum)
{
BDRVVmdkState *s = bs->opaque;
int64_t index_in_cluster, n, ret;
if (!extent) {
return 0;
}
+ qemu_co_mutex_lock(&s->lock);
ret = get_cluster_offset(bs, extent, NULL,
sector_num * 512, 0, &offset);
+ qemu_co_mutex_unlock(&s->lock);
/* get_cluster_offset returning 0 means success */
ret = !ret;
return ret;
}
+static int vmdk_write_extent(VmdkExtent *extent, int64_t cluster_offset,
+ int64_t offset_in_cluster, const uint8_t *buf,
+ int nb_sectors, int64_t sector_num)
+{
+ int ret;
+ VmdkGrainMarker *data = NULL;
+ uLongf buf_len;
+ const uint8_t *write_buf = buf;
+ int write_len = nb_sectors * 512;
+
+ if (extent->compressed) {
+ if (!extent->has_marker) {
+ ret = -EINVAL;
+ goto out;
+ }
+ buf_len = (extent->cluster_sectors << 9) * 2;
+ data = g_malloc(buf_len + sizeof(VmdkGrainMarker));
+ if (compress(data->data, &buf_len, buf, nb_sectors << 9) != Z_OK ||
+ buf_len == 0) {
+ ret = -EINVAL;
+ goto out;
+ }
+ data->lba = sector_num;
+ data->size = buf_len;
+ write_buf = (uint8_t *)data;
+ write_len = buf_len + sizeof(VmdkGrainMarker);
+ }
+ ret = bdrv_pwrite(extent->file,
+ cluster_offset + offset_in_cluster,
+ write_buf,
+ write_len);
+ if (ret != write_len) {
+ ret = ret < 0 ? ret : -EIO;
+ goto out;
+ }
+ ret = 0;
+ out:
+ g_free(data);
+ return ret;
+}
+
+static int vmdk_read_extent(VmdkExtent *extent, int64_t cluster_offset,
+ int64_t offset_in_cluster, uint8_t *buf,
+ int nb_sectors)
+{
+ int ret;
+ int cluster_bytes, buf_bytes;
+ uint8_t *cluster_buf, *compressed_data;
+ uint8_t *uncomp_buf;
+ uint32_t data_len;
+ VmdkGrainMarker *marker;
+ uLongf buf_len;
+
+
+ if (!extent->compressed) {
+ ret = bdrv_pread(extent->file,
+ cluster_offset + offset_in_cluster,
+ buf, nb_sectors * 512);
+ if (ret == nb_sectors * 512) {
+ return 0;
+ } else {
+ return -EIO;
+ }
+ }
+ cluster_bytes = extent->cluster_sectors * 512;
+ /* Read two clusters in case GrainMarker + compressed data > one cluster */
+ buf_bytes = cluster_bytes * 2;
+ cluster_buf = g_malloc(buf_bytes);
+ uncomp_buf = g_malloc(cluster_bytes);
+ ret = bdrv_pread(extent->file,
+ cluster_offset,
+ cluster_buf, buf_bytes);
+ if (ret < 0) {
+ goto out;
+ }
+ compressed_data = cluster_buf;
+ buf_len = cluster_bytes;
+ data_len = cluster_bytes;
+ if (extent->has_marker) {
+ marker = (VmdkGrainMarker *)cluster_buf;
+ compressed_data = marker->data;
+ data_len = le32_to_cpu(marker->size);
+ }
+ if (!data_len || data_len > buf_bytes) {
+ ret = -EINVAL;
+ goto out;
+ }
+ ret = uncompress(uncomp_buf, &buf_len, compressed_data, data_len);
+ if (ret != Z_OK) {
+ ret = -EINVAL;
+ goto out;
+
+ }
+ if (offset_in_cluster < 0 ||
+ offset_in_cluster + nb_sectors * 512 > buf_len) {
+ ret = -EINVAL;
+ goto out;
+ }
+ memcpy(buf, uncomp_buf + offset_in_cluster, nb_sectors * 512);
+ ret = 0;
+
+ out:
+ g_free(uncomp_buf);
+ g_free(cluster_buf);
+ return ret;
+}
+
static int vmdk_read(BlockDriverState *bs, int64_t sector_num,
uint8_t *buf, int nb_sectors)
{
memset(buf, 0, 512 * n);
}
} else {
- ret = bdrv_pread(extent->file,
- cluster_offset + index_in_cluster * 512,
- buf, n * 512);
- if (ret < 0) {
+ ret = vmdk_read_extent(extent,
+ cluster_offset, index_in_cluster * 512,
+ buf, n);
+ if (ret) {
return ret;
}
}
return 0;
}
+static coroutine_fn int vmdk_co_read(BlockDriverState *bs, int64_t sector_num,
+ uint8_t *buf, int nb_sectors)
+{
+ int ret;
+ BDRVVmdkState *s = bs->opaque;
+ qemu_co_mutex_lock(&s->lock);
+ ret = vmdk_read(bs, sector_num, buf, nb_sectors);
+ qemu_co_mutex_unlock(&s->lock);
+ return ret;
+}
+
static int vmdk_write(BlockDriverState *bs, int64_t sector_num,
const uint8_t *buf, int nb_sectors)
{
bs,
extent,
&m_data,
- sector_num << 9, 1,
+ sector_num << 9, !extent->compressed,
&cluster_offset);
+ if (extent->compressed) {
+ if (ret == 0) {
+ /* Refuse write to allocated cluster for streamOptimized */
+ fprintf(stderr,
+ "VMDK: can't write to allocated cluster"
+ " for streamOptimized\n");
+ return -EIO;
+ } else {
+ /* allocate */
+ ret = get_cluster_offset(
+ bs,
+ extent,
+ &m_data,
+ sector_num << 9, 1,
+ &cluster_offset);
+ }
+ }
if (ret) {
return -EINVAL;
}
n = nb_sectors;
}
- ret = bdrv_pwrite(extent->file,
- cluster_offset + index_in_cluster * 512,
- buf,
- n * 512);
- if (ret < 0) {
+ ret = vmdk_write_extent(extent,
+ cluster_offset, index_in_cluster * 512,
+ buf, n, sector_num);
+ if (ret) {
return ret;
}
if (m_data.valid) {
/* update CID on the first write every time the virtual disk is
* opened */
if (!s->cid_updated) {
- vmdk_write_cid(bs, time(NULL));
+ ret = vmdk_write_cid(bs, time(NULL));
+ if (ret < 0) {
+ return ret;
+ }
s->cid_updated = true;
}
}
return 0;
}
+static coroutine_fn int vmdk_co_write(BlockDriverState *bs, int64_t sector_num,
+ const uint8_t *buf, int nb_sectors)
+{
+ int ret;
+ BDRVVmdkState *s = bs->opaque;
+ qemu_co_mutex_lock(&s->lock);
+ ret = vmdk_write(bs, sector_num, buf, nb_sectors);
+ qemu_co_mutex_unlock(&s->lock);
+ return ret;
+}
+
-static int vmdk_create_extent(const char *filename, int64_t filesize, bool flat)
+static int vmdk_create_extent(const char *filename, int64_t filesize,
+ bool flat, bool compress)
{
int ret, i;
int fd = 0;
magic = cpu_to_be32(VMDK4_MAGIC);
memset(&header, 0, sizeof(header));
header.version = 1;
- header.flags = 3; /* ?? */
+ header.flags =
+ 3 | (compress ? VMDK4_FLAG_COMPRESS | VMDK4_FLAG_MARKER : 0);
+ header.compressAlgorithm = compress ? VMDK4_COMPRESSION_DEFLATE : 0;
header.capacity = filesize / 512;
header.granularity = 128;
header.num_gtes_per_gte = 512;
header.rgd_offset = cpu_to_le64(header.rgd_offset);
header.gd_offset = cpu_to_le64(header.gd_offset);
header.grain_offset = cpu_to_le64(header.grain_offset);
+ header.compressAlgorithm = cpu_to_le16(header.compressAlgorithm);
header.check_bytes[0] = 0xa;
header.check_bytes[1] = 0x20;
const char *fmt = NULL;
int flags = 0;
int ret = 0;
- bool flat, split;
+ bool flat, split, compress;
char ext_desc_lines[BUF_SIZE] = "";
char path[PATH_MAX], prefix[PATH_MAX], postfix[PATH_MAX];
const int64_t split_size = 0x80000000; /* VMDK has constant split size */
} else if (strcmp(fmt, "monolithicFlat") &&
strcmp(fmt, "monolithicSparse") &&
strcmp(fmt, "twoGbMaxExtentSparse") &&
- strcmp(fmt, "twoGbMaxExtentFlat")) {
+ strcmp(fmt, "twoGbMaxExtentFlat") &&
+ strcmp(fmt, "streamOptimized")) {
fprintf(stderr, "VMDK: Unknown subformat: %s\n", fmt);
return -EINVAL;
}
strcmp(fmt, "twoGbMaxExtentSparse"));
flat = !(strcmp(fmt, "monolithicFlat") &&
strcmp(fmt, "twoGbMaxExtentFlat"));
+ compress = !strcmp(fmt, "streamOptimized");
if (flat) {
desc_extent_line = "RW %lld FLAT \"%s\" 0\n";
} else {
bdrv_delete(bs);
return -EINVAL;
}
- filesize = bdrv_getlength(bs);
parent_cid = vmdk_read_cid(bs, 0);
bdrv_delete(bs);
relative_path(parent_filename, sizeof(parent_filename),
snprintf(ext_filename, sizeof(ext_filename), "%s%s",
path, desc_filename);
- if (vmdk_create_extent(ext_filename, size, flat)) {
+ if (vmdk_create_extent(ext_filename, size, flat, compress)) {
return -EINVAL;
}
filesize -= size;
static void vmdk_close(BlockDriverState *bs)
{
+ BDRVVmdkState *s = bs->opaque;
+
vmdk_free_extents(bs);
+
+ migrate_del_blocker(s->migration_blocker);
+ error_free(s->migration_blocker);
}
-static int vmdk_flush(BlockDriverState *bs)
+static coroutine_fn int vmdk_co_flush(BlockDriverState *bs)
{
- int i, ret, err;
BDRVVmdkState *s = bs->opaque;
+ int i, err;
+ int ret = 0;
- ret = bdrv_flush(bs->file);
for (i = 0; i < s->num_extents; i++) {
- err = bdrv_flush(s->extents[i].file);
+ err = bdrv_co_flush(s->extents[i].file);
if (err < 0) {
ret = err;
}
.type = OPT_STRING,
.help =
"VMDK flat extent format, can be one of "
- "{monolithicSparse (default) | monolithicFlat | twoGbMaxExtentSparse | twoGbMaxExtentFlat} "
+ "{monolithicSparse (default) | monolithicFlat | twoGbMaxExtentSparse | twoGbMaxExtentFlat | streamOptimized} "
},
{ NULL }
};
.instance_size = sizeof(BDRVVmdkState),
.bdrv_probe = vmdk_probe,
.bdrv_open = vmdk_open,
- .bdrv_read = vmdk_read,
- .bdrv_write = vmdk_write,
+ .bdrv_read = vmdk_co_read,
+ .bdrv_write = vmdk_co_write,
.bdrv_close = vmdk_close,
.bdrv_create = vmdk_create,
- .bdrv_flush = vmdk_flush,
- .bdrv_is_allocated = vmdk_is_allocated,
+ .bdrv_co_flush_to_disk = vmdk_co_flush,
+ .bdrv_co_is_allocated = vmdk_co_is_allocated,
.bdrv_get_allocated_file_size = vmdk_get_allocated_file_size,
.create_options = vmdk_create_options,
projects / qemu.git / blobdiff