backup-top filter driver for backup
v11 cover backup-top filter driver for backup
Hi all!
These series introduce backup-top driver. It's a filter-node, which do copy-before-write operation. Mirror uses filter-node for handling guest writes, let's move to filter-node (from write-notifiers) for backup too.
v11: based on Kevin's block branch
02,03: Add Max's r-b
04: - improve comments
- rebase on dropped backup_drain
- s/job/bcs/ in trace events
05: one hunk dropped due to rebasing
on dropped backup_drain, still,
keep Max's r-b
06: rebased on 04 changes, keep Max's r-b
08,13,14: add Max's r-b
Based-on: git://repo.or.cz/qemu/kevin.git block
Vladimir Sementsov-Ogievskiy (14):
block/backup: fix backup_cow_with_offload for last cluster block/backup: split shareable copying part from backup_do_cow block/backup: improve comment about image fleecing block/backup: introduce BlockCopyState
block/backup: fix block-comment style
block: move block_copy from block/backup.c to separate file
block: teach bdrv_debug_breakpoint skip filters with backing
iotests: prepare 124 and 257 bitmap querying for backup-top filter
iotests: 257: drop unused Drive.device field
iotests: 257: drop device_add
block/io: refactor wait_serialising_requests
block: add lock/unlock range functions
block: introduce backup-top filter driver
block/backup: use backup-top instead of write notifiers
qapi/block-core.json | 8 +- block/backup-top.h | 37 ++ include/block/block-copy.h | 84 ++++ include/block/block_int.h | 5 + block.c | 34 +- block/backup-top.c | 240 ++++++++++++ block/backup.c | 440 ++++----------------- block/block-copy.c | 337 ++++++++++++++++ block/io.c | 68 +++- block/replication.c | 2 +- blockdev.c | 1 + block/Makefile.objs | 3 + block/trace-events | 14 +- tests/qemu-iotests/056 | 8 +- tests/qemu-iotests/124 | 83 ++-- tests/qemu-iotests/257 | 91 ++--- tests/qemu-iotests/257.out | 714 ++++++++++++++-------------------- tests/qemu-iotests/iotests.py | 27 ++ 18 files changed, 1278 insertions(+), 918 deletions(-) create mode 100644 block/backup-top.h create mode 100644 include/block/block-copy.h create mode 100644 block/backup-top.c create mode 100644 block/block-copy.c
Signature · 1 line
2.18.0
PATCH v11 1/14 block/backup: fix backup_cow_with_offload for last cluster
We shouldn't try to copy bytes beyond EOF. Fix it.
--- block/backup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
block/backup.c +1 -1
@@ -161,7 +161,7 @@ static int coroutine_fn backup_cow_with_offload(BackupBlockJob *job,assert (QEMU_IS_ALIGNED (job->copy_range_size, job->cluster_size));assert (QEMU_IS_ALIGNED (start, job->cluster_size)); - nbytes =MIN (job->copy_range_size, end - start); + nbytes =MIN (job->copy_range_size,MIN (end, job->len) - start); nr_clusters =DIV_ROUND_UP (nbytes, job->cluster_size);bdrv_reset_dirty_bitmap (job->copy_bitmap, start, job->cluster_size * nr_clusters);
Signature · 1 line
2.18.0
PATCH v11 2/14 block/backup: split shareable copying part from backup_do_cow
In reply to Vladimir Sementsov-OgievskiySplit copying logic which will be shared with backup-top filter.
--- block/backup.c | 47 ++++++++++++++++++++++++++++++++--------------- 1 file changed, 32 insertions(+), 15 deletions(-)
block/backup.c +32 -15
@@ -248,26 +248,18 @@ static int64_t backup_bitmap_reset_unallocated(BackupBlockJob *s,return ret; } -static int coroutine_fn backup_do_cow(BackupBlockJob *job, -int64_t offset,uint64_t bytes, -bool *error_is_read, -bool is_write_notifier) +static int coroutine_fn backup_do_copy(BackupBlockJob *job, +int64_t start,uint64_t bytes, +bool *error_is_read, +bool is_write_notifier) { -CowRequest cow_request;int ret =0 ; -int64_t start, end;/* bytes */ +int64_t end = bytes + start;/* bytes */ void *bounce_buffer =NULL ;int64_t status_bytes; -qemu_co_rwlock_rdlock (&job->flush_rwlock); - - start =QEMU_ALIGN_DOWN (offset, job->cluster_size); - end =QEMU_ALIGN_UP (bytes + offset, job->cluster_size); - -trace_backup_do_cow_enter (job, start, offset, bytes); - -wait_for_overlapping_requests (job, start, end); -cow_request_begin (&cow_request, job, start, end); +assert (QEMU_IS_ALIGNED (start, job->cluster_size)); +assert (QEMU_IS_ALIGNED (end, job->cluster_size));while (start < end) {int64_t dirty_end;
@@ -326,6 +318,31 @@ static int coroutine_fn backup_do_cow(BackupBlockJob *job,qemu_vfree (bounce_buffer); } +return ret; +} + +static int coroutine_fn backup_do_cow(BackupBlockJob *job, +int64_t offset,uint64_t bytes, +bool *error_is_read, +bool is_write_notifier) +{ +CowRequest cow_request; +int ret =0 ; +int64_t start, end;/* bytes */ + +qemu_co_rwlock_rdlock (&job->flush_rwlock); + + start =QEMU_ALIGN_DOWN (offset, job->cluster_size); + end =QEMU_ALIGN_UP (bytes + offset, job->cluster_size); + +trace_backup_do_cow_enter (job, start, offset, bytes); + +wait_for_overlapping_requests (job, start, end); +cow_request_begin (&cow_request, job, start, end); + + ret =backup_do_copy (job, start, end - start, error_is_read, + is_write_notifier); +cow_request_end (&cow_request);trace_backup_do_cow_return (job, offset, bytes, ret);
Signature · 1 line
2.18.0
PATCH v11 3/14 block/backup: improve comment about image fleecing
In reply to Vladimir Sementsov-Ogievskiy--- block/backup.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-)
block/backup.c +12 -3
@@ -747,9 +747,18 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs, job->bitmap_mode = bitmap_mode;/* -* Set write flags: -* 1. Detect image-fleecing (and similar) schemes -* 2. Handle compression +* If source is in backing chain of target assume that target is going to be +* used for "image fleecing", i.e. it should represent a kind of snapshot of +* source at backup-start point in time. And target is going to be read by +* somebody (for example, used as NBD export) during backup job. +* +* In this case, we need to add BDRV_REQ_SERIALISING write flag to avoid +* intersection of backup writes and third party reads from target, +* otherwise reading from target we may occasionally read already updated by +* guest data. +* +* For more information see commit f8d59dfb40bb and test +* tests/qemu-iotests/222 */ job->write_flags = (bdrv_chain_contains(target, bs) ? BDRV_REQ_SERIALISING : 0) |
Signature · 1 line
2.18.0
PATCH v11 4/14 block/backup: introduce BlockCopyState
In reply to Vladimir Sementsov-OgievskiySplit copying code part from backup to "block-copy", including separate state structure and function renaming. This is needed to share it with backup-top filter driver in further commits.
Notes:
1. As BlockCopyState keeps own BlockBackend objects, remaining job->common.blk users only use it to get bs by blk_bs() call, so clear job->commen.blk permissions set in block_job_create and add job->source_bs to be used instead of blk_bs(job->common.blk), to keep it more clear which bs we use when introduce backup-top filter in further commit.
2. Rename s/initializing_bitmap/skip_unallocated/ to sound a bit better as interface to BlockCopyState
3. Split is not very clean: there left some duplicated fields, backup code uses some BlockCopyState fields directly, let's postpone it for further improvements and keep this comment simpler for review.
--- block/backup.c | 356 ++++++++++++++++++++++++++++----------------- block/trace-events | 12 +- 2 files changed, 231 insertions(+), 137 deletions(-)
block/backup.c +225 -131
@@ -35,12 +35,52 @@ typedef struct CowRequest {CoQueue wait_queue;/* coroutines blocked on this request */ } CowRequest; +typedef void (*ProgressBytesCallbackFunc )(int64_t bytes,void *opaque); +typedef void (*ProgressResetCallbackFunc )(void *opaque); +typedef struct BlockCopyState { +BlockBackend *source; +BlockBackend *target; +BdrvDirtyBitmap *copy_bitmap; +int64_t cluster_size; +bool use_copy_range; +int64_t copy_range_size; +uint64_t len; + +BdrvRequestFlags write_flags; + +/* +* skip_unallocated: +* +* Used by sync=top jobs, which first scan the source node for unallocated +* areas and clear them in the copy_bitmap. During this process, the bitmap +* is thus not fully initialized: It may still have bits set for areas that +* are unallocated and should actually not be copied. +* +* This is indicated by skip_unallocated. +* +* In this case, block_copy() will query the source’s allocation status, +* skip unallocated regions, clear them in the copy_bitmap, and invoke +* block_copy_reset_unallocated() every time it does. +*/ +bool skip_unallocated; + +/* progress_bytes_callback: called when some copying progress is done. */ +ProgressBytesCallbackFunc progress_bytes_callback; + +/* +* progress_reset_callback: called when some bytes reset from copy_bitmap +* (see @skip_unallocated above). The callee is assumed to recalculate how +* many bytes remain based on the dirty bit count of copy_bitmap. +*/ +ProgressResetCallbackFunc progress_reset_callback; +void *progress_opaque; +}BlockCopyState ; + typedef struct BackupBlockJob { BlockJob common; - BlockBackend *target; + BlockDriverState *source_bs; BdrvDirtyBitmap *sync_bitmap; - BdrvDirtyBitmap *copy_bitmap; MirrorSyncMode sync_mode; BitmapSyncMode bitmap_mode;
@@ -53,11 +93,7 @@ typedef struct BackupBlockJob {NotifierWithReturn before_write; QLIST_HEAD(, CowRequest) inflight_reqs; -bool use_copy_range; -int64_t copy_range_size; - -BdrvRequestFlags write_flags; -bool initializing_bitmap; +BlockCopyState *bcs; } BackupBlockJob;static const BlockJobDriver backup_job_driver;
@@ -99,9 +135,88 @@ static void cow_request_end(CowRequest *req)qemu_co_queue_restart_all (&req->wait_queue); } +static void block_copy_state_free (BlockCopyState *s) +{ +if (!s) { +return ; + } + +bdrv_release_dirty_bitmap (blk_bs (s->source), s->copy_bitmap); +blk_unref (s->source); +blk_unref (s->target); +g_free (s); +} + +static BlockCopyState *block_copy_state_new ( +BlockDriverState *source,BlockDriverState *target, +int64_t cluster_size,BdrvRequestFlags write_flags, +ProgressBytesCallbackFunc progress_bytes_callback, +ProgressResetCallbackFunc progress_reset_callback, +void *progress_opaque,Error **errp) +{ +BlockCopyState *s; +int ret; +uint64_t no_resize =BLK_PERM_CONSISTENT_READ |BLK_PERM_WRITE | +BLK_PERM_WRITE_UNCHANGED |BLK_PERM_GRAPH_MOD ; +BdrvDirtyBitmap *copy_bitmap; + + copy_bitmap =bdrv_create_dirty_bitmap (source, cluster_size,NULL , errp); +if (!copy_bitmap) { +return NULL ; + } +bdrv_disable_dirty_bitmap (copy_bitmap); + + s =g_new (BlockCopyState,1 ); + *s = (BlockCopyState ) { + .source =blk_new (bdrv_get_aio_context (source), +BLK_PERM_CONSISTENT_READ , no_resize), + .target =blk_new (bdrv_get_aio_context (target), +BLK_PERM_WRITE , no_resize), + .copy_bitmap = copy_bitmap, + .cluster_size = cluster_size, + .len =bdrv_dirty_bitmap_size (copy_bitmap), + .write_flags = write_flags, + .use_copy_range = !(write_flags &BDRV_REQ_WRITE_COMPRESSED ), + .progress_bytes_callback = progress_bytes_callback, + .progress_reset_callback = progress_reset_callback, + .progress_opaque = progress_opaque, + }; + + s->copy_range_size =QEMU_ALIGN_UP (MIN (blk_get_max_transfer (s->source), +blk_get_max_transfer (s->target)), + s->cluster_size); + +/* +* We just allow aio context change on our block backends. block_copy() user +* (now it's only backup) is responsible for source and target being in same +* aio context. +*/ +blk_set_disable_request_queuing (s->source, true); +blk_set_allow_aio_context_change (s->source, true); +blk_set_disable_request_queuing (s->target, true); +blk_set_allow_aio_context_change (s->target, true); + + ret =blk_insert_bs (s->source, source, errp); +if (ret <0 ) { + goto fail; + } + + ret =blk_insert_bs (s->target, target, errp); +if (ret <0 ) { + goto fail; + } + +return s; + +fail: +block_copy_state_free (s); + +return NULL ; +} +/* Copy range to target with a bounce buffer and return the bytes copied. If * error occurred, return a negative error number */ -static int coroutine_fn backup_cow_with_bounce_buffer(BackupBlockJob *job, +static int coroutine_fn block_copy_with_bounce_buffer(BlockCopyState *s, int64_t start, int64_t end, bool is_write_notifier,
@@ -109,30 +224,29 @@ static int coroutine_fn backup_cow_with_bounce_buffer(BackupBlockJob *job, void **bounce_buffer) {int ret; -BlockBackend *blk = job->common.blk;int nbytes;int read_flags = is_write_notifier ?BDRV_REQ_NO_SERIALISING :0 ; -assert (QEMU_IS_ALIGNED (start, job->cluster_size)); -bdrv_reset_dirty_bitmap (job->copy_bitmap, start, job->cluster_size); - nbytes =MIN (job->cluster_size, job->len - start); +assert (QEMU_IS_ALIGNED (start, s->cluster_size)); +bdrv_reset_dirty_bitmap (s->copy_bitmap, start, s->cluster_size); + nbytes =MIN (s->cluster_size, s->len - start);if (!*bounce_buffer) { - *bounce_buffer =blk_blockalign (blk, job->cluster_size); + *bounce_buffer =blk_blockalign (s->source, s->cluster_size); } - ret =blk_co_pread (blk, start, nbytes, *bounce_buffer, read_flags); + ret =blk_co_pread (s->source, start, nbytes, *bounce_buffer, read_flags);if (ret <0 ) { -trace_backup_do_cow_read_fail (job, start, ret); +trace_block_copy_with_bounce_buffer_read_fail (s, start, ret);if (error_is_read) { *error_is_read = true; } goto fail; } - ret =blk_co_pwrite (job->target, start, nbytes, *bounce_buffer, - job->write_flags); + ret =blk_co_pwrite (s->target, start, nbytes, *bounce_buffer, + s->write_flags);if (ret <0 ) { -trace_backup_do_cow_write_fail (job, start, ret); +trace_block_copy_with_bounce_buffer_write_fail (s, start, ret);if (error_is_read) { *error_is_read = false; }
@@ -141,36 +255,35 @@ static int coroutine_fn backup_cow_with_bounce_buffer(BackupBlockJob *job,return nbytes; fail: -bdrv_set_dirty_bitmap (job->copy_bitmap, start, job->cluster_size); +bdrv_set_dirty_bitmap (s->copy_bitmap, start, s->cluster_size);return ret; }/* Copy range to target and return the bytes copied. If error occurred, return a * negative error number. */ -static int coroutine_fn backup_cow_with_offload(BackupBlockJob *job, +static int coroutine_fn block_copy_with_offload(BlockCopyState *s,int64_t start,int64_t end,bool is_write_notifier) {int ret;int nr_clusters; -BlockBackend *blk = job->common.blk;int nbytes;int read_flags = is_write_notifier ?BDRV_REQ_NO_SERIALISING :0 ; -assert (QEMU_IS_ALIGNED (job->copy_range_size, job->cluster_size)); -assert (QEMU_IS_ALIGNED (start, job->cluster_size)); - nbytes =MIN (job->copy_range_size,MIN (end, job->len) - start); - nr_clusters =DIV_ROUND_UP (nbytes, job->cluster_size); -bdrv_reset_dirty_bitmap (job->copy_bitmap, start, - job->cluster_size * nr_clusters); - ret =blk_co_copy_range (blk, start, job->target, start, nbytes, - read_flags, job->write_flags); +assert (QEMU_IS_ALIGNED (s->copy_range_size, s->cluster_size)); +assert (QEMU_IS_ALIGNED (start, s->cluster_size)); + nbytes =MIN (s->copy_range_size,MIN (end, s->len) - start); + nr_clusters =DIV_ROUND_UP (nbytes, s->cluster_size); +bdrv_reset_dirty_bitmap (s->copy_bitmap, start, + s->cluster_size * nr_clusters); + ret =blk_co_copy_range (s->source, start, s->target, start, nbytes, + read_flags, s->write_flags);if (ret <0 ) { -trace_backup_do_cow_copy_range_fail (job, start, ret); -bdrv_set_dirty_bitmap (job->copy_bitmap, start, - job->cluster_size * nr_clusters); +trace_block_copy_with_offload_fail (s, start, ret); +bdrv_set_dirty_bitmap (s->copy_bitmap, start, + s->cluster_size * nr_clusters);return ret; }
@@ -181,10 +294,10 @@ static int coroutine_fn backup_cow_with_offload(BackupBlockJob *job,* Check if the cluster starting at offset is allocated or not. * return via pnum the number of contiguous clusters sharing this allocation. */ -static int backup_is_cluster_allocated (BackupBlockJob *s,int64_t offset, -int64_t *pnum) +static int block_copy_is_cluster_allocated (BlockCopyState *s,int64_t offset, +int64_t *pnum) { -BlockDriverState *bs =blk_bs (s->common.blk); +BlockDriverState *bs =blk_bs (s->source);int64_t count, total_count =0 ;int64_t bytes = s->len - offset;int ret;
@@ -225,13 +338,13 @@ static int backup_is_cluster_allocated(BackupBlockJob *s, int64_t offset,* @return 0 when the cluster at @offset was unallocated, * 1 otherwise, and -ret on error. */ -static int64_t backup_bitmap_reset_unallocated (BackupBlockJob *s, -int64_t offset,int64_t *count) +static int64_t block_copy_reset_unallocated (BlockCopyState *s, +int64_t offset,int64_t *count) {int ret; -int64_t clusters, bytes, estimate; +int64_t clusters, bytes; - ret =backup_is_cluster_allocated (s, offset, &clusters); + ret =block_copy_is_cluster_allocated (s, offset, &clusters);if (ret <0 ) {return ret; }
@@ -240,46 +353,51 @@ static int64_t backup_bitmap_reset_unallocated(BackupBlockJob *s,if (!ret) {bdrv_reset_dirty_bitmap (s->copy_bitmap, offset, bytes); - estimate =bdrv_get_dirty_count (s->copy_bitmap); -job_progress_set_remaining (&s->common.job, estimate); + s->progress_reset_callback (s->progress_opaque); } *count = bytes;return ret; } -static int coroutine_fn backup_do_copy(BackupBlockJob *job, -int64_t start,uint64_t bytes, -bool *error_is_read, -bool is_write_notifier) +static int coroutine_fn block_copy(BlockCopyState *s, +int64_t start,uint64_t bytes, +bool *error_is_read, +bool is_write_notifier) {int ret =0 ;int64_t end = bytes + start;/* bytes */ void *bounce_buffer =NULL ;int64_t status_bytes; -assert (QEMU_IS_ALIGNED (start, job->cluster_size)); -assert (QEMU_IS_ALIGNED (end, job->cluster_size)); +/* +* block_copy() user is responsible for keeping source and target in same +* aio context +*/ +assert (blk_get_aio_context (s->source) ==blk_get_aio_context (s->target)); + +assert (QEMU_IS_ALIGNED (start, s->cluster_size)); +assert (QEMU_IS_ALIGNED (end, s->cluster_size));while (start < end) {int64_t dirty_end; -if (!bdrv_dirty_bitmap_get (job->copy_bitmap, start)) { -trace_backup_do_cow_skip (job, start); - start += job->cluster_size; +if (!bdrv_dirty_bitmap_get (s->copy_bitmap, start)) { +trace_block_copy_skip (s, start); + start += s->cluster_size;continue ;/* already copied */ } - dirty_end =bdrv_dirty_bitmap_next_zero (job->copy_bitmap, start, + dirty_end =bdrv_dirty_bitmap_next_zero (s->copy_bitmap, start, (end - start));if (dirty_end <0 ) { dirty_end = end; } -if (job->initializing_bitmap) { - ret =backup_bitmap_reset_unallocated (job, start, &status_bytes); +if (s->skip_unallocated) { + ret =block_copy_reset_unallocated (s, start, &status_bytes);if (ret ==0 ) { -trace_backup_do_cow_skip_range (job, start, status_bytes); +trace_block_copy_skip_range (s, start, status_bytes); start += status_bytes;continue ; }
@@ -287,17 +405,17 @@ static int coroutine_fn backup_do_copy(BackupBlockJob *job, dirty_end =MIN (dirty_end, start + status_bytes); } -trace_backup_do_cow_process (job, start); +trace_block_copy_process (s, start); -if (job->use_copy_range) { - ret =backup_cow_with_offload (job, start, dirty_end, +if (s->use_copy_range) { + ret =block_copy_with_offload (s, start, dirty_end, is_write_notifier);if (ret <0 ) { - job->use_copy_range = false; + s->use_copy_range = false; } } -if (!job->use_copy_range) { - ret =backup_cow_with_bounce_buffer (job, start, dirty_end, +if (!s->use_copy_range) { + ret =block_copy_with_bounce_buffer (s, start, dirty_end, is_write_notifier, error_is_read, &bounce_buffer); }
@@ -305,12 +423,8 @@ static int coroutine_fn backup_do_copy(BackupBlockJob *job,break ; } -/* Publish progress, guest I/O counts as progress too. Note that the -* offset field is an opaque progress value, it is not a disk offset. -*/ start += ret; - job->bytes_read += ret; -job_progress_update (&job->common.job, ret); + s->progress_bytes_callback (ret, s->progress_opaque); ret =0 ; }
@@ -321,6 +435,22 @@ static int coroutine_fn backup_do_copy(BackupBlockJob *job,return ret; } +static void backup_progress_bytes_callback (int64_t bytes,void *opaque) +{ +BackupBlockJob *s = opaque; + + s->bytes_read += bytes; +job_progress_update (&s->common.job, bytes); +} + +static void backup_progress_reset_callback (void *opaque) +{ +BackupBlockJob *s = opaque; +uint64_t estimate =bdrv_get_dirty_count (s->bcs->copy_bitmap); + +job_progress_set_remaining (&s->common.job, estimate); +} + static int coroutine_fn backup_do_cow(BackupBlockJob *job, int64_t offset, uint64_t bytes, bool *error_is_read,
@@ -340,8 +470,8 @@ static int coroutine_fn backup_do_cow(BackupBlockJob *job,wait_for_overlapping_requests (job, start, end);cow_request_begin (&cow_request, job, start, end); - ret =backup_do_copy (job, start, end - start, error_is_read, - is_write_notifier); + ret =block_copy (job->bcs, start, end - start, error_is_read, + is_write_notifier);cow_request_end (&cow_request);
@@ -359,7 +489,7 @@ static int coroutine_fn backup_before_write_notify(BackupBlockJob *job =container_of (notifier, BackupBlockJob, before_write);BdrvTrackedRequest *req = opaque; -assert (req->bs ==blk_bs (job->common.blk)); +assert (req->bs == job->source_bs);assert (QEMU_IS_ALIGNED (req->offset,BDRV_SECTOR_SIZE ));assert (QEMU_IS_ALIGNED (req->bytes,BDRV_SECTOR_SIZE ));
@@ -369,7 +499,6 @@ static int coroutine_fn backup_before_write_notify(static void backup_cleanup_sync_bitmap (BackupBlockJob *job,int ret) {BdrvDirtyBitmap *bm; -BlockDriverState *bs =blk_bs (job->common.blk);bool sync = (((ret ==0 ) || (job->bitmap_mode ==BITMAP_SYNC_MODE_ALWAYS )) \ && (job->bitmap_mode !=BITMAP_SYNC_MODE_NEVER ));
@@ -378,20 +507,20 @@ static void backup_cleanup_sync_bitmap(BackupBlockJob *job, int ret)* We succeeded, or we always intended to sync the bitmap. * Delete this bitmap and install the child. */ - bm =bdrv_dirty_bitmap_abdicate (bs, job->sync_bitmap,NULL ); + bm =bdrv_dirty_bitmap_abdicate (job->source_bs, job->sync_bitmap,NULL ); }else {/* * We failed, or we never intended to sync the bitmap anyway. * Merge the successor back into the parent, keeping all data. */ - bm =bdrv_reclaim_dirty_bitmap (bs, job->sync_bitmap,NULL ); + bm =bdrv_reclaim_dirty_bitmap (job->source_bs, job->sync_bitmap,NULL ); }assert (bm);if (ret <0 && job->bitmap_mode ==BITMAP_SYNC_MODE_ALWAYS ) {/* If we failed and synced, merge in the bits we didn't copy: */ -bdrv_dirty_bitmap_merge_internal (bm, job->copy_bitmap, +bdrv_dirty_bitmap_merge_internal (bm, job->bcs->copy_bitmap,NULL , true); } }
@@ -415,16 +544,8 @@ static void backup_abort(Job *job)static void backup_clean (Job *job) {BackupBlockJob *s =container_of (job, BackupBlockJob, common.job); -BlockDriverState *bs =blk_bs (s->common.blk); -if (s->copy_bitmap) { -bdrv_release_dirty_bitmap (bs, s->copy_bitmap); - s->copy_bitmap =NULL ; - } - -assert (s->target); -blk_unref (s->target); - s->target =NULL ; +block_copy_state_free (s->bcs); }void backup_do_checkpoint (BlockJob *job,Error **errp)
@@ -439,7 +560,7 @@ void backup_do_checkpoint(BlockJob *job, Error **errp)return ; } -bdrv_set_dirty_bitmap (backup_job->copy_bitmap,0 , backup_job->len); +bdrv_set_dirty_bitmap (backup_job->bcs->copy_bitmap,0 , backup_job->len); } static BlockErrorAction backup_error_action(BackupBlockJob *job,
@@ -482,7 +603,7 @@ static int coroutine_fn backup_loop(BackupBlockJob *job)BdrvDirtyBitmapIter *bdbi;int ret =0 ; - bdbi =bdrv_dirty_iter_new (job->copy_bitmap); + bdbi =bdrv_dirty_iter_new (job->bcs->copy_bitmap);while ((offset =bdrv_dirty_iter_next (bdbi)) !=-1 ) { do { if (yield_and_check(job)) {
@@ -509,7 +630,7 @@ static void backup_init_copy_bitmap(BackupBlockJob *job)uint64_t estimate;if (job->sync_mode ==MIRROR_SYNC_MODE_BITMAP ) { - ret =bdrv_dirty_bitmap_merge_internal (job->copy_bitmap, + ret =bdrv_dirty_bitmap_merge_internal (job->bcs->copy_bitmap, job->sync_bitmap,NULL , true);assert (ret);
@@ -519,19 +640,18 @@ static void backup_init_copy_bitmap(BackupBlockJob *job)* We can't hog the coroutine to initialize this thoroughly. * Set a flag and resume work when we are able to yield safely. */ - job->initializing_bitmap = true; + job->bcs->skip_unallocated = true; } -bdrv_set_dirty_bitmap (job->copy_bitmap,0 , job->len); +bdrv_set_dirty_bitmap (job->bcs->copy_bitmap,0 , job->len); } - estimate =bdrv_get_dirty_count (job->copy_bitmap); + estimate =bdrv_get_dirty_count (job->bcs->copy_bitmap);job_progress_set_remaining (&job->common.job, estimate); } static int coroutine_fn backup_run(Job *job, Error **errp) {BackupBlockJob *s =container_of (job, BackupBlockJob, common.job); -BlockDriverState *bs =blk_bs (s->common.blk);int ret =0 ;QLIST_INIT (&s->inflight_reqs);
@@ -540,7 +660,7 @@ static int coroutine_fn backup_run(Job *job, Error **errp)backup_init_copy_bitmap (s); s->before_write.notify = backup_before_write_notify; -bdrv_add_before_write_notifier (bs, &s->before_write); +bdrv_add_before_write_notifier (s->source_bs, &s->before_write);if (s->sync_mode ==MIRROR_SYNC_MODE_TOP ) {int64_t offset =0 ;
@@ -552,14 +672,14 @@ static int coroutine_fn backup_run(Job *job, Error **errp) goto out; } - ret =backup_bitmap_reset_unallocated (s, offset, &count); + ret =block_copy_reset_unallocated (s->bcs, offset, &count);if (ret <0 ) { goto out; } offset += count; } - s->initializing_bitmap = false; + s->bcs->skip_unallocated = false; }if (s->sync_mode ==MIRROR_SYNC_MODE_NONE ) {
@@ -646,9 +766,8 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs, {int64_t len;BackupBlockJob *job =NULL ; -int ret;int64_t cluster_size; -BdrvDirtyBitmap *copy_bitmap =NULL ; +BdrvRequestFlags write_flags;assert (bs);assert (target);
@@ -713,33 +832,14 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs, goto error; } - copy_bitmap =bdrv_create_dirty_bitmap (bs, cluster_size,NULL , errp); -if (!copy_bitmap) { - goto error; - } -bdrv_disable_dirty_bitmap (copy_bitmap); -/* job->len is fixed, so we can't allow resize */ - job =block_job_create (job_id, &backup_job_driver, txn, bs, -BLK_PERM_CONSISTENT_READ , -BLK_PERM_CONSISTENT_READ |BLK_PERM_WRITE | -BLK_PERM_WRITE_UNCHANGED |BLK_PERM_GRAPH_MOD , + job =block_job_create (job_id, &backup_job_driver, txn, bs,0 ,BLK_PERM_ALL , speed, creation_flags, cb, opaque, errp);if (!job) { goto error; } -/* The target must match the source in size, so no resize here either */ - job->target =blk_new (job->common.job.aio_context, -BLK_PERM_WRITE , -BLK_PERM_CONSISTENT_READ |BLK_PERM_WRITE | -BLK_PERM_WRITE_UNCHANGED |BLK_PERM_GRAPH_MOD ); - ret =blk_insert_bs (job->target, target, errp); -if (ret <0 ) { - goto error; - } -blk_set_disable_request_queuing (job->target, true); - + job->source_bs = bs; job->on_source_error = on_source_error; job->on_target_error = on_target_error; job->sync_mode = sync_mode;
@@ -760,21 +860,19 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,* For more information see commit f8d59dfb40bb and test * tests/qemu-iotests/222 */ - job->write_flags = - (bdrv_chain_contains (target, bs) ?BDRV_REQ_SERIALISING :0 ) | - (compress ?BDRV_REQ_WRITE_COMPRESSED :0 ); + write_flags = (bdrv_chain_contains (target, bs) ?BDRV_REQ_SERIALISING :0 ) | + (compress ?BDRV_REQ_WRITE_COMPRESSED :0 ), + + job->bcs =block_copy_state_new (bs, target, cluster_size, write_flags, + backup_progress_bytes_callback, + backup_progress_reset_callback, job, errp); +if (!job->bcs) { + goto error; + } job->cluster_size = cluster_size; - job->copy_bitmap = copy_bitmap; - copy_bitmap =NULL ; - job->use_copy_range = !compress;/* compression isn't supported for it */ - job->copy_range_size =MIN_NON_ZERO (blk_get_max_transfer (job->common.blk), -blk_get_max_transfer (job->target)); - job->copy_range_size =MAX (job->cluster_size, -QEMU_ALIGN_UP (job->copy_range_size, - job->cluster_size)); - -/* Required permissions are already taken with target's blk_new() */ + +/* Required permissions are already taken by block-copy-state target */ block_job_add_bdrv (&job->common,"target" , target,0 ,BLK_PERM_ALL , &error_abort); job->len = len;
@@ -782,10 +880,6 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,return &job->common; error: -if (copy_bitmap) { -assert (!job || !job->copy_bitmap); -bdrv_release_dirty_bitmap (bs, copy_bitmap); - }if (sync_bitmap) {bdrv_reclaim_dirty_bitmap (bs, sync_bitmap,NULL ); }
block/trace-events +6 -6
@@ -40,12 +40,12 @@ mirror_yield_in_flight(void *s, int64_t offset, int in_flight) "s %p offset %" P # backup.c backup_do_cow_enter(void *job, int64_t start, int64_t offset, uint64_t bytes) "job %p start %" PRId64 " offset %" PRId64 " bytes %" PRIu64 backup_do_cow_return(void *job, int64_t offset, uint64_t bytes, int ret) "job %p offset %" PRId64 " bytes %" PRIu64 " ret %d" -backup_do_cow_skip(void *job, int64_t start) "job %p start %"PRId64 -backup_do_cow_skip_range(void *job, int64_t start, uint64_t bytes) "job %p start %"PRId64" bytes %"PRId64 -backup_do_cow_process(void *job, int64_t start) "job %p start %"PRId64 -backup_do_cow_read_fail(void *job, int64_t start, int ret) "job %p start %"PRId64" ret %d" -backup_do_cow_write_fail(void *job, int64_t start, int ret) "job %p start %"PRId64" ret %d" -backup_do_cow_copy_range_fail(void *job, int64_t start, int ret) "job %p start %"PRId64" ret %d" +block_copy_skip(void *bcs, int64_t start) "bcs %p start %"PRId64 +block_copy_skip_range(void *bcs, int64_t start, uint64_t bytes) "bcs %p start %"PRId64" bytes %"PRId64 +block_copy_process(void *bcs, int64_t start) "bcs %p start %"PRId64 +block_copy_with_bounce_buffer_read_fail(void *bcs, int64_t start, int ret) "bcs %p start %"PRId64" ret %d" +block_copy_with_bounce_buffer_write_fail(void *bcs, int64_t start, int ret) "bcs %p start %"PRId64" ret %d" +block_copy_with_offload_fail(void *bcs, int64_t start, int ret) "bcs %p start %"PRId64" ret %d" # ../blockdev.c qmp_block_job_cancel(void *job) "job %p"
Signature · 1 line
2.18.0
Re: [Qemu-devel] [PATCH v11 04/14] block/backup: introduce BlockCopyState
Quoted 22 lines · Vladimir Sementsov-Ogievskiy
Split copying code part from backup to "block-copy", including separate state structure and function renaming. This is needed to share it with backup-top filter driver in further commits.
Notes:
1. As BlockCopyState keeps own BlockBackend objects, remaining
job->common.blkusers only use it to get bs byblk_bs()call, so clearjob->commen.blkpermissions set inblock_job_createand addjob->source_bsto be used instead ofblk_bs(job->common.blk), to keep it more clear which bs we use when introduce backup-top filter in further commit.2. Rename s/initializing_bitmap/skip_unallocated/ to sound a bit better as interface to BlockCopyState
3. Split is not very clean: there left some duplicated fields, backup code uses some BlockCopyState fields directly, let's postpone it for further improvements and keep this comment simpler for review.
Signed-off-by Vladimir Sementsov-Ogievskiy <[email protected]>---
[..]
Quoted 29 lines
block/backup.c +29
-0+ +static BlockCopyState *block_copy_state_new ( +BlockDriverState *source,BlockDriverState *target, +int64_t cluster_size,BdrvRequestFlags write_flags, +ProgressBytesCallbackFunc progress_bytes_callback, +ProgressResetCallbackFunc progress_reset_callback, +void *progress_opaque,Error **errp) +{ +BlockCopyState *s; +int ret; +uint64_t no_resize =BLK_PERM_CONSISTENT_READ |BLK_PERM_WRITE | +BLK_PERM_WRITE_UNCHANGED |BLK_PERM_GRAPH_MOD ; +BdrvDirtyBitmap *copy_bitmap; + + copy_bitmap =bdrv_create_dirty_bitmap (source, cluster_size,NULL , errp); +if (!copy_bitmap) { +return NULL ; + } +bdrv_disable_dirty_bitmap (copy_bitmap); + + s =g_new (BlockCopyState,1 ); + *s = (BlockCopyState ) { + .source =blk_new (bdrv_get_aio_context (source), +BLK_PERM_CONSISTENT_READ , no_resize), + .target =blk_new (bdrv_get_aio_context (target), +BLK_PERM_WRITE , no_resize), + .copy_bitmap = copy_bitmap, + .cluster_size = cluster_size, + .len =bdrv_dirty_bitmap_size (copy_bitmap),
block/backup.c static void cow_request_end(CowRequest *req) · line 187
Quoted 10 lines
block/backup.c +10
-0+ .write_flags = write_flags, + .use_copy_range = !(write_flags & BDRV_REQ_WRITE_COMPRESSED), + .progress_bytes_callback = progress_bytes_callback, + .progress_reset_callback = progress_reset_callback, + .progress_opaque = progress_opaque, + }; + + s->copy_range_size =QEMU_ALIGN_UP (MIN (blk_get_max_transfer (s->source), +blk_get_max_transfer (s->target)), + s->cluster_size);
preexistent, but it obviously should be QEMU_ALIGN_DOWN. I can resend with a separate fix, it may be fixed while queuing (if resend is not needed for other reasons) or I'll send a follow-up fix later, whichever you prefer.
Quoted 11 lines
block/backup.c +11
-0+ +/* +* We just allow aio context change on our block backends. block_copy() user +* (now it's only backup) is responsible for source and target being in same +* aio context. +*/ +blk_set_disable_request_queuing (s->source, true); +blk_set_allow_aio_context_change (s->source, true); +blk_set_disable_request_queuing (s->target, true); +blk_set_allow_aio_context_change (s->target, true); +
[..]
Quoted 3 lines
block/backup.c +0
-0@@ -760,21 +860,19 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,* For more information see commit f8d59dfb40bb and test * tests/qemu-iotests/222
[..]
Quoted 16 lines
job->cluster_size = cluster_size;block/backup.c +2
-10- job->copy_bitmap = copy_bitmap; - copy_bitmap =NULL ; - job->use_copy_range = !compress;/* compression isn't supported for it */ - job->copy_range_size =MIN_NON_ZERO (blk_get_max_transfer (job->common.blk), -blk_get_max_transfer (job->target)); - job->copy_range_size =MAX (job->cluster_size, -QEMU_ALIGN_UP (job->copy_range_size, - job->cluster_size)); - -/* Required permissions are already taken with target's blk_new() */ + +/* Required permissions are already taken by block-copy-state target */ block_job_add_bdrv (&job->common,"target" , target,0 ,BLK_PERM_ALL , &error_abort); job->len = len;
Signature · 2 lines
Best regards, Vladimir
Re: [PATCH v11 04/14] block/backup: introduce BlockCopyState
Quoted 62 lines · Vladimir Sementsov-Ogievskiy
Quoted 22 lines · Vladimir Sementsov-Ogievskiy
Split copying code part from backup to "block-copy", including separate state structure and function renaming. This is needed to share it with backup-top filter driver in further commits.
Notes:
1. As BlockCopyState keeps own BlockBackend objects, remaining
job->common.blkusers only use it to get bs byblk_bs()call, so clearjob->commen.blkpermissions set inblock_job_createand addjob->source_bsto be used instead ofblk_bs(job->common.blk), to keep it more clear which bs we use when introduce backup-top filter in further commit.2. Rename s/initializing_bitmap/skip_unallocated/ to sound a bit better as interface to BlockCopyState
3. Split is not very clean: there left some duplicated fields, backup code uses some BlockCopyState fields directly, let's postpone it for further improvements and keep this comment simpler for review.
Signed-off-by Vladimir Sementsov-Ogievskiy <[email protected]>---
[..]
Quoted 35 lines
block/backup.c +35
-0+ +static BlockCopyState *block_copy_state_new ( +BlockDriverState *source,BlockDriverState *target, +int64_t cluster_size,BdrvRequestFlags write_flags, +ProgressBytesCallbackFunc progress_bytes_callback, +ProgressResetCallbackFunc progress_reset_callback, +void *progress_opaque,Error **errp) +{ +BlockCopyState *s; +int ret; +uint64_t no_resize =BLK_PERM_CONSISTENT_READ |BLK_PERM_WRITE | +BLK_PERM_WRITE_UNCHANGED |BLK_PERM_GRAPH_MOD ; +BdrvDirtyBitmap *copy_bitmap; + + copy_bitmap =bdrv_create_dirty_bitmap (source, cluster_size,NULL , errp); +if (!copy_bitmap) { +return NULL ; + } +bdrv_disable_dirty_bitmap (copy_bitmap); + + s =g_new (BlockCopyState,1 ); + *s = (BlockCopyState ) { + .source =blk_new (bdrv_get_aio_context (source), +BLK_PERM_CONSISTENT_READ , no_resize), + .target =blk_new (bdrv_get_aio_context (target), +BLK_PERM_WRITE , no_resize), + .copy_bitmap = copy_bitmap, + .cluster_size = cluster_size, + .len =bdrv_dirty_bitmap_size (copy_bitmap), + .write_flags = write_flags, + .use_copy_range = !(write_flags &BDRV_REQ_WRITE_COMPRESSED ), + .progress_bytes_callback = progress_bytes_callback, + .progress_reset_callback = progress_reset_callback, + .progress_opaque = progress_opaque, + };
Quoted 8 lines
block/backup.c static void cow_request_end(CowRequest *req) · line 187
Quoted 4 lines
block/backup.c +4
-0+ + s->copy_range_size =QEMU_ALIGN_UP (MIN (blk_get_max_transfer (s->source), +blk_get_max_transfer (s->target)), + s->cluster_size);preexistent, but it obviously should be
QEMU_ALIGN_DOWN. I can resend with a separate fix, it may be fixed while queuing (if resend is not needed for other reasons) or I'll send a follow-up fix later, whichever you prefer.
Hm, true. But then we’ll also need to handle the (unlikely, admittedly) case where max_transfer < cluster_size so this would then return 0 (by setting use_copy_range = false). So how about this:
Quoted 11 lines
block/backup.c +0
-3@@ -182,9 +182,13 @@ static BlockCopyState *block_copy_state_new( .progress_opaque = progress_opaque, }; - s->copy_range_size =QEMU_ALIGN_UP (MIN (blk_get_max_transfer (s->source), -blk_get_max_transfer (s->target)), - s->cluster_size);
block/backup.c static BlockCopyState *block_copy_state_new( · line 194
Quoted 10 lines
block/backup.c +7
-0+ s->copy_range_size =QEMU_ALIGN_DOWN (MIN (blk_get_max_transfer (s->source), +blk_get_max_transfer (s->target)), + s->cluster_size); +if (s->copy_range_size ==0 ) { +/* max_transfer < cluster_size */ + s->use_copy_range = false; + }/* * We just allow aio context change on our block backends. block_copy() user
Max
Re: [PATCH v11 04/14] block/backup: introduce BlockCopyState
Quoted 72 lines · Max Reitz
Quoted 62 lines · Vladimir Sementsov-Ogievskiy
Quoted 22 lines · Vladimir Sementsov-Ogievskiy
Split copying code part from backup to "block-copy", including separate state structure and function renaming. This is needed to share it with backup-top filter driver in further commits.
Notes:
1. As BlockCopyState keeps own BlockBackend objects, remaining
job->common.blkusers only use it to get bs byblk_bs()call, so clearjob->commen.blkpermissions set inblock_job_createand addjob->source_bsto be used instead ofblk_bs(job->common.blk), to keep it more clear which bs we use when introduce backup-top filter in further commit.2. Rename s/initializing_bitmap/skip_unallocated/ to sound a bit better as interface to BlockCopyState
3. Split is not very clean: there left some duplicated fields, backup code uses some BlockCopyState fields directly, let's postpone it for further improvements and keep this comment simpler for review.
Signed-off-by Vladimir Sementsov-Ogievskiy <[email protected]>---
[..]
Quoted 35 lines
block/backup.c +35
-0+ +static BlockCopyState *block_copy_state_new ( +BlockDriverState *source,BlockDriverState *target, +int64_t cluster_size,BdrvRequestFlags write_flags, +ProgressBytesCallbackFunc progress_bytes_callback, +ProgressResetCallbackFunc progress_reset_callback, +void *progress_opaque,Error **errp) +{ +BlockCopyState *s; +int ret; +uint64_t no_resize =BLK_PERM_CONSISTENT_READ |BLK_PERM_WRITE | +BLK_PERM_WRITE_UNCHANGED |BLK_PERM_GRAPH_MOD ; +BdrvDirtyBitmap *copy_bitmap; + + copy_bitmap =bdrv_create_dirty_bitmap (source, cluster_size,NULL , errp); +if (!copy_bitmap) { +return NULL ; + } +bdrv_disable_dirty_bitmap (copy_bitmap); + + s =g_new (BlockCopyState,1 ); + *s = (BlockCopyState ) { + .source =blk_new (bdrv_get_aio_context (source), +BLK_PERM_CONSISTENT_READ , no_resize), + .target =blk_new (bdrv_get_aio_context (target), +BLK_PERM_WRITE , no_resize), + .copy_bitmap = copy_bitmap, + .cluster_size = cluster_size, + .len =bdrv_dirty_bitmap_size (copy_bitmap), + .write_flags = write_flags, + .use_copy_range = !(write_flags &BDRV_REQ_WRITE_COMPRESSED ), + .progress_bytes_callback = progress_bytes_callback, + .progress_reset_callback = progress_reset_callback, + .progress_opaque = progress_opaque, + };Quoted 8 lines
block/backup.c static void cow_request_end(CowRequest *req) · line 187
Quoted 4 lines
block/backup.c +4
-0+ + s->copy_range_size =QEMU_ALIGN_UP (MIN (blk_get_max_transfer (s->source), +blk_get_max_transfer (s->target)), + s->cluster_size);preexistent, but it obviously should be
QEMU_ALIGN_DOWN. I can resend with a separate fix, it may be fixed while queuing (if resend is not needed for other reasons) or I'll send a follow-up fix later, whichever you prefer.
Quoted 3 lines
Hm, true. But then we’ll also need to handle the (unlikely, admittedly) case where
max_transfer<cluster_sizeso this would then return 0 (by settinguse_copy_range= false). So how about this:
Done in [PATCH v12 0/2] backup: copy_range fixes.
If it is convenient I'll rebase these series on "[PATCH v12 0/2] backup: copy_range fixes"
Quoted 25 lines
Quoted 12 lines
block/backup.c +0
-3@@ -182,9 +182,13 @@ static BlockCopyState *block_copy_state_new( .progress_opaque = progress_opaque, }; - s->copy_range_size =QEMU_ALIGN_UP (MIN (blk_get_max_transfer (s->source), -blk_get_max_transfer (s->target)), - s->cluster_size);block/backup.c static BlockCopyState *block_copy_state_new( · line 194
Quoted 10 lines
block/backup.c +7
-0+ s->copy_range_size =QEMU_ALIGN_DOWN (MIN (blk_get_max_transfer (s->source), +blk_get_max_transfer (s->target)), + s->cluster_size); +if (s->copy_range_size ==0 ) { +/* max_transfer < cluster_size */ + s->use_copy_range = false; + }/* * We just allow aio context change on our block backends. block_copy() user Max
Signature · 2 lines
Best regards, Vladimir
Re: [PATCH v11 04/14] block/backup: introduce BlockCopyState
Quoted 73 lines · Vladimir Sementsov-Ogievskiy
Quoted 72 lines · Max Reitz
Quoted 70 lines · Vladimir Sementsov-Ogievskiy
Quoted 22 lines · Vladimir Sementsov-Ogievskiy
Split copying code part from backup to "block-copy", including separate state structure and function renaming. This is needed to share it with backup-top filter driver in further commits.
Notes:
1. As BlockCopyState keeps own BlockBackend objects, remaining
job->common.blkusers only use it to get bs byblk_bs()call, so clearjob->commen.blkpermissions set inblock_job_createand addjob->source_bsto be used instead ofblk_bs(job->common.blk), to keep it more clear which bs we use when introduce backup-top filter in further commit.2. Rename s/initializing_bitmap/skip_unallocated/ to sound a bit better as interface to BlockCopyState
3. Split is not very clean: there left some duplicated fields, backup code uses some BlockCopyState fields directly, let's postpone it for further improvements and keep this comment simpler for review.
Signed-off-by Vladimir Sementsov-Ogievskiy <[email protected]>---
[..]
Quoted 29 lines
block/backup.c +29
-0+ +static BlockCopyState *block_copy_state_new ( + BlockDriverState *source, BlockDriverState *target, + int64_t cluster_size, BdrvRequestFlags write_flags, + ProgressBytesCallbackFunc progress_bytes_callback, + ProgressResetCallbackFunc progress_reset_callback, + void *progress_opaque, Error **errp) +{ + BlockCopyState *s; + int ret; + uint64_t no_resize = BLK_PERM_CONSISTENT_READ | BLK_PERM_WRITE | + BLK_PERM_WRITE_UNCHANGED | BLK_PERM_GRAPH_MOD; + BdrvDirtyBitmap *copy_bitmap; + + copy_bitmap = bdrv_create_dirty_bitmap(source, cluster_size, NULL, errp); + if (!copy_bitmap) { + return NULL; + } + bdrv_disable_dirty_bitmap(copy_bitmap); + + s = g_new(BlockCopyState, 1); + *s = (BlockCopyState) { + .source = blk_new(bdrv_get_aio_context(source), + BLK_PERM_CONSISTENT_READ, no_resize), + .target = blk_new(bdrv_get_aio_context(target), + BLK_PERM_WRITE, no_resize), + .copy_bitmap = copy_bitmap, + .cluster_size = cluster_size, + .len = bdrv_dirty_bitmap_size(copy_bitmap),block/backup.c static void cow_request_end(CowRequest *req) · line 187
Quoted 10 lines
block/backup.c +10
-0+ .write_flags = write_flags, + .use_copy_range = !(write_flags & BDRV_REQ_WRITE_COMPRESSED), + .progress_bytes_callback = progress_bytes_callback, + .progress_reset_callback = progress_reset_callback, + .progress_opaque = progress_opaque, + }; + + s->copy_range_size = QEMU_ALIGN_UP(MIN(blk_get_max_transfer(s->source), + blk_get_max_transfer(s->target)), + s->cluster_size);preexistent, but it obviously should be
QEMU_ALIGN_DOWN. I can resend with a separate fix, it may be fixed while queuing (if resend is not needed for other reasons) or I'll send a follow-up fix later, whichever you prefer.
Quoted 7 lines
Quoted 3 lines
Hm, true. But then we’ll also need to handle the (unlikely, admittedly) case where
max_transfer<cluster_sizeso this would then return 0 (by settinguse_copy_range= false). So how about this:Done in [PATCH v12 0/2] backup:
copy_rangefixes.
If it is convenient I'll rebase these series on "[PATCH v12 0/2] backup:copy_rangefixes"
Or vice versa
Signature · 2 lines
Best regards, Vladimir
Re: [PATCH v11 04/14] block/backup: introduce BlockCopyState
Quoted 74 lines · Vladimir Sementsov-Ogievskiy
Quoted 73 lines · Vladimir Sementsov-Ogievskiy
Quoted 72 lines · Max Reitz
Quoted 70 lines · Vladimir Sementsov-Ogievskiy
Quoted 22 lines · Vladimir Sementsov-Ogievskiy
Split copying code part from backup to "block-copy", including separate state structure and function renaming. This is needed to share it with backup-top filter driver in further commits.
Notes:
1. As BlockCopyState keeps own BlockBackend objects, remaining
job->common.blkusers only use it to get bs byblk_bs()call, so clearjob->commen.blkpermissions set inblock_job_createand addjob->source_bsto be used instead ofblk_bs(job->common.blk), to keep it more clear which bs we use when introduce backup-top filter in further commit.2. Rename s/initializing_bitmap/skip_unallocated/ to sound a bit better as interface to BlockCopyState
3. Split is not very clean: there left some duplicated fields, backup code uses some BlockCopyState fields directly, let's postpone it for further improvements and keep this comment simpler for review.
Signed-off-by Vladimir Sementsov-Ogievskiy <[email protected]>---
[..]
Quoted 29 lines
block/backup.c +29
-0+ +static BlockCopyState *block_copy_state_new ( + BlockDriverState *source, BlockDriverState *target, + int64_t cluster_size, BdrvRequestFlags write_flags, + ProgressBytesCallbackFunc progress_bytes_callback, + ProgressResetCallbackFunc progress_reset_callback, + void *progress_opaque, Error **errp) +{ + BlockCopyState *s; + int ret; + uint64_t no_resize = BLK_PERM_CONSISTENT_READ | BLK_PERM_WRITE | + BLK_PERM_WRITE_UNCHANGED | BLK_PERM_GRAPH_MOD; + BdrvDirtyBitmap *copy_bitmap; + + copy_bitmap = bdrv_create_dirty_bitmap(source, cluster_size, NULL, errp); + if (!copy_bitmap) { + return NULL; + } + bdrv_disable_dirty_bitmap(copy_bitmap); + + s = g_new(BlockCopyState, 1); + *s = (BlockCopyState) { + .source = blk_new(bdrv_get_aio_context(source), + BLK_PERM_CONSISTENT_READ, no_resize), + .target = blk_new(bdrv_get_aio_context(target), + BLK_PERM_WRITE, no_resize), + .copy_bitmap = copy_bitmap, + .cluster_size = cluster_size, + .len = bdrv_dirty_bitmap_size(copy_bitmap),block/backup.c static void cow_request_end(CowRequest *req) · line 187
Quoted 10 lines
block/backup.c +10
-0+ .write_flags = write_flags, + .use_copy_range = !(write_flags & BDRV_REQ_WRITE_COMPRESSED), + .progress_bytes_callback = progress_bytes_callback, + .progress_reset_callback = progress_reset_callback, + .progress_opaque = progress_opaque, + }; + + s->copy_range_size = QEMU_ALIGN_UP(MIN(blk_get_max_transfer(s->source), + blk_get_max_transfer(s->target)), + s->cluster_size);preexistent, but it obviously should be
QEMU_ALIGN_DOWN. I can resend with a separate fix, it may be fixed while queuing (if resend is not needed for other reasons) or I'll send a follow-up fix later, whichever you prefer.
Quoted 6 lines
Quoted 6 lines
Quoted 3 lines
Hm, true. But then we’ll also need to handle the (unlikely, admittedly) case where
max_transfer<cluster_sizeso this would then return 0 (by settinguse_copy_range= false). So how about this:Done in [PATCH v12 0/2] backup:
copy_rangefixes.
If it is convenient I'll rebase these series on "[PATCH v12 0/2] backup:copy_rangefixes"
Oh, good.
I think taking copy_range fixes first would make more sense. It seems that John still had some suggestion for it...?
Max
Re: [PATCH v11 04/14] block/backup: introduce BlockCopyState
Quoted 82 lines · Max Reitz
Quoted 74 lines · Vladimir Sementsov-Ogievskiy
Quoted 73 lines · Vladimir Sementsov-Ogievskiy
Quoted 72 lines · Max Reitz
Quoted 70 lines · Vladimir Sementsov-Ogievskiy
Quoted 22 lines · Vladimir Sementsov-Ogievskiy
Split copying code part from backup to "block-copy", including separate state structure and function renaming. This is needed to share it with backup-top filter driver in further commits.
Notes:
1. As BlockCopyState keeps own BlockBackend objects, remaining
job->common.blkusers only use it to get bs byblk_bs()call, so clearjob->commen.blkpermissions set inblock_job_createand addjob->source_bsto be used instead ofblk_bs(job->common.blk), to keep it more clear which bs we use when introduce backup-top filter in further commit.2. Rename s/initializing_bitmap/skip_unallocated/ to sound a bit better as interface to BlockCopyState
3. Split is not very clean: there left some duplicated fields, backup code uses some BlockCopyState fields directly, let's postpone it for further improvements and keep this comment simpler for review.
Signed-off-by Vladimir Sementsov-Ogievskiy <[email protected]>---
[..]
Quoted 29 lines
block/backup.c +29
-0+ +static BlockCopyState *block_copy_state_new ( + BlockDriverState *source, BlockDriverState *target, + int64_t cluster_size, BdrvRequestFlags write_flags, + ProgressBytesCallbackFunc progress_bytes_callback, + ProgressResetCallbackFunc progress_reset_callback, + void *progress_opaque, Error **errp) +{ + BlockCopyState *s; + int ret; + uint64_t no_resize = BLK_PERM_CONSISTENT_READ | BLK_PERM_WRITE | + BLK_PERM_WRITE_UNCHANGED | BLK_PERM_GRAPH_MOD; + BdrvDirtyBitmap *copy_bitmap; + + copy_bitmap = bdrv_create_dirty_bitmap(source, cluster_size, NULL, errp); + if (!copy_bitmap) { + return NULL; + } + bdrv_disable_dirty_bitmap(copy_bitmap); + + s = g_new(BlockCopyState, 1); + *s = (BlockCopyState) { + .source = blk_new(bdrv_get_aio_context(source), + BLK_PERM_CONSISTENT_READ, no_resize), + .target = blk_new(bdrv_get_aio_context(target), + BLK_PERM_WRITE, no_resize), + .copy_bitmap = copy_bitmap, + .cluster_size = cluster_size, + .len = bdrv_dirty_bitmap_size(copy_bitmap),block/backup.c static void cow_request_end(CowRequest *req) · line 187
Quoted 10 lines
block/backup.c +10
-0+ .write_flags = write_flags, + .use_copy_range = !(write_flags & BDRV_REQ_WRITE_COMPRESSED), + .progress_bytes_callback = progress_bytes_callback, + .progress_reset_callback = progress_reset_callback, + .progress_opaque = progress_opaque, + }; + + s->copy_range_size = QEMU_ALIGN_UP(MIN(blk_get_max_transfer(s->source), + blk_get_max_transfer(s->target)), + s->cluster_size);preexistent, but it obviously should be
QEMU_ALIGN_DOWN. I can resend with a separate fix, it may be fixed while queuing (if resend is not needed for other reasons) or I'll send a follow-up fix later, whichever you prefer.Quoted 6 lines
Quoted 6 lines
Quoted 3 lines
Hm, true. But then we’ll also need to handle the (unlikely, admittedly) case where
max_transfer<cluster_sizeso this would then return 0 (by settinguse_copy_range= false). So how about this:Done in [PATCH v12 0/2] backup:
copy_rangefixes.
If it is convenient I'll rebase these series on "[PATCH v12 0/2] backup:copy_rangefixes"
Quoted 7 lines
Oh, good.
I think taking
copy_rangefixes first would make more sense. It seems that John still had some suggestion for it...?Max
Finally, only wording of the comment. I think, I can resend combined series with my final suggestion of the comment (a bit tweaked John's one).
Signature · 2 lines
Best regards, Vladimir
PATCH v11 5/14 block/backup: fix block-comment style
In reply to Vladimir Sementsov-OgievskiyWe need to fix comment style around block-copy functions before further moving them to separate file to satisfy checkpatch. But do more: fix all comments style. Also, seems like doubled first asterisk is not forbidden, but drop it too for consistency.
--- block/backup.c | 32 +++++++++++++++++++++----------- 1 file changed, 21 insertions(+), 11 deletions(-)
block/backup.c +21 -11
@@ -214,8 +214,10 @@ fail:return NULL ; } -/* Copy range to target with a bounce buffer and return the bytes copied. If -* error occurred, return a negative error number */ +/* +* Copy range to target with a bounce buffer and return the bytes copied. If +* error occurred, return a negative error number +*/ static int coroutine_fn block_copy_with_bounce_buffer(BlockCopyState *s, int64_t start, int64_t end,
@@ -260,8 +262,10 @@ fail: } -/* Copy range to target and return the bytes copied. If error occurred, return a -* negative error number. */ +/* +* Copy range to target and return the bytes copied. If error occurred, return a +* negative error number. +*/ static int coroutine_fn block_copy_with_offload(BlockCopyState *s, int64_t start, int64_t end,
@@ -332,7 +336,7 @@ static int block_copy_is_cluster_allocated(BlockCopyState *s, int64_t offset, } } -/** +/* * Reset bits in copy_bitmap starting at offset if they represent unallocated * data in the image. May reset subsequent contiguous bits. * @return 0 when the cluster at @offset was unallocated,
@@ -583,8 +587,10 @@ static bool coroutine_fn yield_and_check(BackupBlockJob *job)return true; } -/* We need to yield even for delay_ns = 0 so that bdrv_drain_all() can -* return. Without a yield, the VM would not reboot. */ +/* +* We need to yield even for delay_ns = 0 so that bdrv_drain_all() can +* return. Without a yield, the VM would not reboot. +*/ delay_ns =block_job_ratelimit_get_delay (&job->common, job->bytes_read); job->bytes_read =0 ;job_sleep_ns (&job->common.job, delay_ns);
@@ -683,11 +689,15 @@ static int coroutine_fn backup_run(Job *job, Error **errp) }if (s->sync_mode ==MIRROR_SYNC_MODE_NONE ) { -/* All bits are set in copy_bitmap to allow any cluster to be copied. -* This does not actually require them to be copied. */ +/* +* All bits are set in copy_bitmap to allow any cluster to be copied. +* This does not actually require them to be copied. +*/ while (!job_is_cancelled (job)) { -/* Yield until the job is cancelled. We just let our before_write -* notify callback service CoW requests. */ +/* +* Yield until the job is cancelled. We just let our before_write +* notify callback service CoW requests. +*/ job_yield (job); } }else {
Signature · 1 line
2.18.0
PATCH v11 6/14 block: move block_copy from block/backup.c to separate file
In reply to Vladimir Sementsov-OgievskiySplit block_copy to separate file, to be cleanly shared with backup-top filter driver in further commits.
It's a clean movement, the only change is drop "static" from interface functions.
--- include/block/block-copy.h | 76 ++++++++ block/backup.c | 346 +------------------------------------ block/block-copy.c | 324 ++++++++++++++++++++++++++++++++++ block/Makefile.objs | 1 + block/trace-events | 2 + 5 files changed, 404 insertions(+), 345 deletions(-) create mode 100644 include/block/block-copy.h create mode 100644 block/block-copy.c
include/block/block-copy.h new file +76 -0
@@ -0,0 +1,76 @@ +/* +* block_copy API +* +* Copyright (C) 2013 Proxmox Server Solutions +* Copyright (c) 2019 Virtuozzo International GmbH. +* +* Authors: +* Dietmar Maurer ([email protected]) +* Vladimir Sementsov-Ogievskiy <[email protected]> +* +* This work is licensed under the terms of the GNU GPL, version 2 or later. +* See the COPYING file in the top-level directory. +*/ + +#ifndef BLOCK_COPY_H +#define BLOCK_COPY_H + +#include "block/block.h" + +typedef void (*ProgressBytesCallbackFunc )(int64_t bytes,void *opaque); +typedef void (*ProgressResetCallbackFunc )(void *opaque); +typedef struct BlockCopyState { +BlockBackend *source; +BlockBackend *target; +BdrvDirtyBitmap *copy_bitmap; +int64_t cluster_size; +bool use_copy_range; +int64_t copy_range_size; +uint64_t len; + +BdrvRequestFlags write_flags; + +/* +* skip_unallocated: +* +* Used by sync=top jobs, which first scan the source node for unallocated +* areas and clear them in the copy_bitmap. During this process, the bitmap +* is thus not fully initialized: It may still have bits set for areas that +* are unallocated and should actually not be copied. +* +* This is indicated by skip_unallocated. +* +* In this case, block_copy() will query the source’s allocation status, +* skip unallocated regions, clear them in the copy_bitmap, and invoke +* block_copy_reset_unallocated() every time it does. +*/ +bool skip_unallocated; + +/* progress_bytes_callback: called when some copying progress is done. */ +ProgressBytesCallbackFunc progress_bytes_callback; + +/* +* progress_reset_callback: called when some bytes reset from copy_bitmap +* (see @skip_unallocated above). The callee is assumed to recalculate how +* many bytes remain based on the dirty bit count of copy_bitmap. +*/ +ProgressResetCallbackFunc progress_reset_callback; +void *progress_opaque; +}BlockCopyState ; + +BlockCopyState *block_copy_state_new ( +BlockDriverState *source,BlockDriverState *target, +int64_t cluster_size,BdrvRequestFlags write_flags, +ProgressBytesCallbackFunc progress_bytes_callback, +ProgressResetCallbackFunc progress_reset_callback, +void *progress_opaque,Error **errp); + +void block_copy_state_free (BlockCopyState *s); + +int64_t block_copy_reset_unallocated (BlockCopyState *s, +int64_t offset,int64_t *count); + +int coroutine_fn block_copy(BlockCopyState *s, int64_t start, uint64_t bytes, +bool *error_is_read,bool is_write_notifier); + +#endif /* BLOCK_COPY_H */
block/backup.c +1 -345
364 lines not on this page · raw message
block/block-copy.c new file +324 -0
324 lines not on this page · raw message
block/Makefile.objs +1 -0
@@ -37,6 +37,7 @@ block-obj-y += write-threshold.o block-obj-y += backup.o block-obj-$(CONFIG_REPLICATION) += replication.o block-obj-y += throttle.o copy-on-read.o +block-obj-y += block-copy.o block-obj-y += crypto.o
block/trace-events +2 -0
@@ -40,6 +40,8 @@ mirror_yield_in_flight(void *s, int64_t offset, int in_flight) "s %p offset %" P # backup.c backup_do_cow_enter(void *job, int64_t start, int64_t offset, uint64_t bytes) "job %p start %" PRId64 " offset %" PRId64 " bytes %" PRIu64 backup_do_cow_return(void *job, int64_t offset, uint64_t bytes, int ret) "job %p offset %" PRId64 " bytes %" PRIu64 " ret %d" + +# block-copy.c block_copy_skip(void *bcs, int64_t start) "bcs %p start %"PRId64 block_copy_skip_range(void *bcs, int64_t start, uint64_t bytes) "bcs %p start %"PRId64" bytes %"PRId64 block_copy_process(void *bcs, int64_t start) "bcs %p start %"PRId64
Signature · 1 line
2.18.0
PATCH v11 7/14 block: teach bdrv_debug_breakpoint skip filters with backing
In reply to Vladimir Sementsov-OgievskiyTeach bdrv_debug_breakpoint and bdrv_debug_remove_breakpoint skip filters with backing. This is needed to implement and use in backup job it's own backup_top filter driver (like mirror already has one), and without this improvement, breakpoint removal will fail at least in 55 iotest.
--- block.c | 34 ++++++++++++++++++++++++++-------- 1 file changed, 26 insertions(+), 8 deletions(-)
block.c +26 -8
@@ -5165,14 +5165,35 @@ void bdrv_debug_event(BlockDriverState *bs, BlkdebugEvent event) bs->drv->bdrv_debug_event (bs, event); } -int bdrv_debug_breakpoint (BlockDriverState *bs,const char *event, -const char *tag) +static BlockDriverState *bdrv_find_debug_node (BlockDriverState *bs) {while (bs && bs->drv && !bs->drv->bdrv_debug_breakpoint) { - bs = bs->file ? bs->file->bs :NULL ; +if (bs->file) { + bs = bs->file->bs; +continue ; + } + +if (bs->drv->is_filter && bs->backing) { + bs = bs->backing->bs; +continue ; + } + +break ; }if (bs && bs->drv && bs->drv->bdrv_debug_breakpoint) { +assert (bs->drv->bdrv_debug_remove_breakpoint); +return bs; + } + +return NULL ; +} + +int bdrv_debug_breakpoint (BlockDriverState *bs,const char *event, +const char *tag) +{ + bs =bdrv_find_debug_node (bs); +if (bs) {return bs->drv->bdrv_debug_breakpoint (bs, event, tag); }
@@ -5181,11 +5202,8 @@ int bdrv_debug_breakpoint(BlockDriverState *bs, const char *event,int bdrv_debug_remove_breakpoint (BlockDriverState *bs,const char *tag) { -while (bs && bs->drv && !bs->drv->bdrv_debug_remove_breakpoint) { - bs = bs->file ? bs->file->bs :NULL ; - } - -if (bs && bs->drv && bs->drv->bdrv_debug_remove_breakpoint) { + bs =bdrv_find_debug_node (bs); +if (bs) {return bs->drv->bdrv_debug_remove_breakpoint (bs, tag); }
Signature · 1 line
2.18.0
PATCH v11 8/14 iotests: prepare 124 and 257 bitmap querying for backup-top filter
In reply to Vladimir Sementsov-OgievskiyAfter backup-top filter appearing it's not possible to see dirty bitmaps in top node, so use node-name instead.
--- tests/qemu-iotests/124 | 83 ++++---- tests/qemu-iotests/257 | 49 ++--- tests/qemu-iotests/257.out | 364 +++++++++++++--------------------- tests/qemu-iotests/iotests.py | 27 +++ 4 files changed, 219 insertions(+), 304 deletions(-)
tests/qemu-iotests/124 +36 -47
@@ -105,7 +105,7 @@ class TestIncrementalBackupBase(iotests.QMPTestCase): # Create a base image with a distinctive patterning drive0 = self.add_node('drive0') self.img_create(drive0['file'], drive0['fmt']) - self.vm.add_drive(drive0['file']) + self.vm.add_drive(drive0['file'], opts='node-name=node0') self.write_default_pattern(drive0['file']) self.vm.launch()
@@ -348,12 +348,14 @@ class TestIncrementalBackup(TestIncrementalBackupBase): ('0xfe', '16M', '256k'), ('0x64', '32736k', '64k'))) # Check the dirty bitmap stats - result = self.vm.qmp('query-block') - self.assert_qmp(result, 'return[0]/dirty-bitmaps[0]/name', 'bitmap0') - self.assert_qmp(result, 'return[0]/dirty-bitmaps[0]/count', 458752) - self.assert_qmp(result, 'return[0]/dirty-bitmaps[0]/granularity', 65536) - self.assert_qmp(result, 'return[0]/dirty-bitmaps[0]/status', 'active') - self.assert_qmp(result, 'return[0]/dirty-bitmaps[0]/persistent', False) + self.assertTrue(self.vm.check_bitmap_status( + 'node0', bitmap0.name, { + 'name': 'bitmap0', + 'count': 458752, + 'granularity': 65536, + 'status': 'active', + 'persistent': False + })) # Prepare a cluster_size=128k backup target without a backing file. (target, _) = bitmap0.new_target()
@@ -670,9 +672,8 @@ class TestIncrementalBackupBlkdebug(TestIncrementalBackupBase): """ drive0 = self.drives[0] - # NB: The blkdebug script here looks for a "flush, read, read" pattern. - # The flush occurs in hmp_io_writes, the first read in device_add, and - # the last read during the block job. + # NB: The blkdebug script here looks for a "flush, read" pattern. + # The flush occurs in hmp_io_writes, and the read during the block job. result = self.vm.qmp('blockdev-add', node_name=drive0['id'], driver=drive0['fmt'],
@@ -686,15 +687,11 @@ class TestIncrementalBackupBlkdebug(TestIncrementalBackupBase): 'event': 'flush_to_disk', 'state': 1, 'new_state': 2 - },{ - 'event': 'read_aio', - 'state': 2, - 'new_state': 3 }], 'inject-error': [{ 'event': 'read_aio', 'errno': 5, - 'state': 3, + 'state': 2, 'immediately': False, 'once': True }],
@@ -708,23 +705,15 @@ class TestIncrementalBackupBlkdebug(TestIncrementalBackupBase): ('0xfe', '16M', '256k'), ('0x64', '32736k', '64k'))) - # For the purposes of query-block visibility of bitmaps, add a drive - # frontend after we've written data; otherwise we can't use hmp-io - result = self.vm.qmp("device_add", - id="device0", - drive=drive0['id'], - driver="virtio-blk") - self.assert_qmp(result, 'return', {}) - # Bitmap Status Check - query = self.vm.qmp('query-block') - ret = [bmap for bmap in query['return'][0]['dirty-bitmaps'] - if bmap.get('name') == bitmap.name][0] - self.assert_qmp(ret, 'count', 458752) - self.assert_qmp(ret, 'granularity', 65536) - self.assert_qmp(ret, 'status', 'active') - self.assert_qmp(ret, 'busy', False) - self.assert_qmp(ret, 'recording', True) + self.assertTrue(self.vm.check_bitmap_status( + drive0['id'], bitmap.name, { + 'count': 458752, + 'granularity': 65536, + 'status': 'active', + 'busy': False, + 'recording': True + })) # Start backup parent, _ = bitmap.last_target()
@@ -748,14 +737,14 @@ class TestIncrementalBackupBlkdebug(TestIncrementalBackupBase): 'operation': 'read'}) # Bitmap Status Check - query = self.vm.qmp('query-block') - ret = [bmap for bmap in query['return'][0]['dirty-bitmaps'] - if bmap.get('name') == bitmap.name][0] - self.assert_qmp(ret, 'count', 458752) - self.assert_qmp(ret, 'granularity', 65536) - self.assert_qmp(ret, 'status', 'frozen') - self.assert_qmp(ret, 'busy', True) - self.assert_qmp(ret, 'recording', True) + self.assertTrue(self.vm.check_bitmap_status( + drive0['id'], bitmap.name, { + 'count': 458752, + 'granularity': 65536, + 'status': 'frozen', + 'busy': True, + 'recording': True + })) # Resume and check incremental backup for consistency res = self.vm.qmp('block-job-resume', device=bitmap.drive['id'])
@@ -763,14 +752,14 @@ class TestIncrementalBackupBlkdebug(TestIncrementalBackupBase): self.wait_qmp_backup(bitmap.drive['id']) # Bitmap Status Check - query = self.vm.qmp('query-block') - ret = [bmap for bmap in query['return'][0]['dirty-bitmaps'] - if bmap.get('name') == bitmap.name][0] - self.assert_qmp(ret, 'count', 0) - self.assert_qmp(ret, 'granularity', 65536) - self.assert_qmp(ret, 'status', 'active') - self.assert_qmp(ret, 'busy', False) - self.assert_qmp(ret, 'recording', True) + self.assertTrue(self.vm.check_bitmap_status( + drive0['id'], bitmap.name, { + 'count': 0, + 'granularity': 65536, + 'status': 'active', + 'busy': False, + 'recording': True + })) # Finalize / Cleanup self.make_reference_backup(bitmap)
tests/qemu-iotests/257 +16 -33
105 lines not on this page · raw message
tests/qemu-iotests/257.out +140 -224
1,204 lines not on this page · raw message
tests/qemu-iotests/iotests.py +27 -0
@@ -643,6 +643,33 @@ class VM(qtest.QEMUQtestMachine): return x return None + def query_bitmaps(self): + res = self.qmp("query-named-block-nodes") + return {device['node-name']: device['dirty-bitmaps'] + for device in res['return'] if 'dirty-bitmaps' in device} + + def get_bitmap(self, node_name, bitmap_name, recording=None, bitmaps=None): + """ + get a specific bitmap from the object returned by query_bitmaps. + :param recording: If specified, filter results by the specified value. + :param bitmaps: If specified, use it instead of call query_bitmaps() + """ + if bitmaps is None: + bitmaps = self.query_bitmaps() + + for bitmap in bitmaps[node_name]: + if bitmap.get('name', '') == bitmap_name: + if recording is None: + return bitmap + elif bitmap.get('recording') == recording: + return bitmap + return None + + def check_bitmap_status(self, node_name, bitmap_name, fields): + ret = self.get_bitmap(node_name, bitmap_name) + + return fields.items() <= ret.items() + index_re = re.compile(r'([^\[]+)\[([^\]]+)\]')
Signature · 1 line
2.18.0
PATCH v11 9/14 iotests: 257: drop unused Drive.device field
In reply to Vladimir Sementsov-OgievskiyAfter previous commit Drive.device is actually unused. Drop it together with .name property. While being here reuse .node in qmp commands instead of writing 'drive0' twice.
--- tests/qemu-iotests/257 | 37 +++++++++++++++---------------------- 1 file changed, 15 insertions(+), 22 deletions(-)
tests/qemu-iotests/257 +15 -22
@@ -148,11 +148,6 @@ class Drive: self.fmt = None self.size = None self.node = None - self.device = None - - @property - def name(self): - return self.node or self.device def img_create(self, fmt, size): self.fmt = fmt
@@ -201,7 +196,7 @@ def blockdev_backup(vm, device, target, sync, **kwargs): def blockdev_backup_mktarget(drive, target_id, filepath, sync, **kwargs): target_drive = Drive(filepath, vm=drive.vm) target_drive.create_target(target_id, drive.fmt, drive.size) - blockdev_backup(drive.vm, drive.name, target_id, sync, **kwargs) + blockdev_backup(drive.vm, drive.node, target_id, sync, **kwargs) def reference_backup(drive, n, filepath): log("--- Reference Backup #{:d} ---\n".format(n))
@@ -229,7 +224,7 @@ def perform_writes(drive, n): pattern.offset, pattern.size) log(cmd) - log(drive.vm.hmp_qemu_io(drive.name, cmd)) + log(drive.vm.hmp_qemu_io(drive.node, cmd)) bitmaps = drive.vm.query_bitmaps() log({'bitmaps': bitmaps}, indent=2) log('')
@@ -324,18 +319,17 @@ def test_bitmap_sync(bsync_mode, msync_mode='bitmap', failure=None): }] } + drive0.node = 'drive0' vm.qmp_log('blockdev-add', filters=[iotests.filter_qmp_testfiles], - node_name="drive0", + node_name=drive0.node, driver=drive0.fmt, file=file_config) - drive0.node = 'drive0' - drive0.device = 'device0' # Use share-rw to allow writes directly to the node; # The anonymous block-backend for this configuration prevents us # from using HMP's qemu-io commands to address the device. - vm.qmp_log("device_add", id=drive0.device, - drive=drive0.name, driver="scsi-hd", + vm.qmp_log("device_add", id='device0', + drive=drive0.node, driver="scsi-hd", share_rw=True) log('')
@@ -343,7 +337,7 @@ def test_bitmap_sync(bsync_mode, msync_mode='bitmap', failure=None): perform_writes(drive0, 0) reference_backup(drive0, 0, fbackup0) log('--- Add Bitmap ---\n') - vm.qmp_log("block-dirty-bitmap-add", node=drive0.name, + vm.qmp_log("block-dirty-bitmap-add", node=drive0.node, name="bitmap0", granularity=GRANULARITY) log('') ebitmap = EmulatedBitmap()
@@ -358,7 +352,7 @@ def test_bitmap_sync(bsync_mode, msync_mode='bitmap', failure=None): # 1 - Test Backup (w/ Optional induced failure) if failure == 'intermediate': # Activate blkdebug induced failure for second-to-next read - log(vm.hmp_qemu_io(drive0.name, 'flush')) + log(vm.hmp_qemu_io(drive0.node, 'flush')) log('') job = backup(drive0, 1, bsync1, msync_mode, bitmap="bitmap0", bitmap_mode=bsync_mode)
@@ -426,7 +420,7 @@ def test_bitmap_sync(bsync_mode, msync_mode='bitmap', failure=None): log('--- Cleanup ---\n') vm.qmp_log("block-dirty-bitmap-remove", - node=drive0.name, name="bitmap0") + node=drive0.node, name="bitmap0") bitmaps = vm.query_bitmaps() log({'bitmaps': bitmaps}, indent=2) vm.shutdown()
@@ -467,22 +461,21 @@ def test_backup_api(): 'filename': drive0.path } + drive0.node = 'drive0' vm.qmp_log('blockdev-add', filters=[iotests.filter_qmp_testfiles], - node_name="drive0", + node_name=drive0.node, driver=drive0.fmt, file=file_config) - drive0.node = 'drive0' - drive0.device = 'device0' - vm.qmp_log("device_add", id=drive0.device, - drive=drive0.name, driver="scsi-hd") + vm.qmp_log("device_add", id='device0', + drive=drive0.node, driver="scsi-hd") log('') target0 = Drive(backup_path, vm=vm) target0.create_target("backup_target", drive0.fmt, drive0.size) log('') - vm.qmp_log("block-dirty-bitmap-add", node=drive0.name, + vm.qmp_log("block-dirty-bitmap-add", node=drive0.node, name="bitmap0", granularity=GRANULARITY) log('')
@@ -521,7 +514,7 @@ def test_backup_api(): log("-- Sync mode {:s} tests --\n".format(sync_mode)) for bitmap in (None, 'bitmap404', 'bitmap0'): for policy in error_cases[sync_mode][bitmap]: - blockdev_backup(drive0.vm, drive0.name, "backup_target", + blockdev_backup(drive0.vm, drive0.node, "backup_target", sync_mode, job_id='api_job', bitmap=bitmap, bitmap_mode=policy) log('')
Signature · 1 line
2.18.0
PATCH v11 10/14 iotests: 257: drop device_add
In reply to Vladimir Sementsov-OgievskiySCSI devices are unused in test, drop them.
--- tests/qemu-iotests/257 | 8 ------- tests/qemu-iotests/257.out | 44 -------------------------------------- 2 files changed, 52 deletions(-)
tests/qemu-iotests/257 +0 -8
@@ -325,12 +325,6 @@ def test_bitmap_sync(bsync_mode, msync_mode='bitmap', failure=None): node_name=drive0.node, driver=drive0.fmt, file=file_config) - # Use share-rw to allow writes directly to the node; - # The anonymous block-backend for this configuration prevents us - # from using HMP's qemu-io commands to address the device. - vm.qmp_log("device_add", id='device0', - drive=drive0.node, driver="scsi-hd", - share_rw=True) log('') # 0 - Writes and Reference Backup
@@ -467,8 +461,6 @@ def test_backup_api(): node_name=drive0.node, driver=drive0.fmt, file=file_config) - vm.qmp_log("device_add", id='device0', - drive=drive0.node, driver="scsi-hd") log('') target0 = Drive(backup_path, vm=vm)
tests/qemu-iotests/257.out +0 -44
176 lines not on this page · raw message
Signature · 1 line
2.18.0
PATCH v11 11/14 block/io: refactor wait_serialising_requests
In reply to Vladimir Sementsov-OgievskiySplit out do_wait_serialising_requests with additional possibility to not actually wait but just check, that there is something to wait for.
--- block/io.c | 24 ++++++++++++++++-------- 1 file changed, 16 insertions(+), 8 deletions(-)
block/io.c +16 -8
@@ -786,12 +786,13 @@ void bdrv_dec_in_flight(BlockDriverState *bs)bdrv_wakeup (bs); } -static bool coroutine_fn wait_serialising_requests(BdrvTrackedRequest *self) +static bool coroutine_fn do_wait_serialising_requests(BdrvTrackedRequest *self, +bool wait) {BlockDriverState *bs = self->bs;BdrvTrackedRequest *req;bool retry; -bool waited = false; +bool found = false;if (!atomic_read (&bs->serialising_in_flight)) {return false;
@@ -817,11 +818,13 @@ static bool coroutine_fn wait_serialising_requests(BdrvTrackedRequest *self)* will wait for us as soon as it wakes up, then just go on * (instead of producing a deadlock in the former case). */ if (!req->waiting_for) { - self->waiting_for = req; -qemu_co_queue_wait (&req->wait_queue, &bs->reqs_lock); - self->waiting_for =NULL ; - retry = true; - waited = true; + found = true; +if (wait) { + self->waiting_for = req; +qemu_co_queue_wait (&req->wait_queue, &bs->reqs_lock); + self->waiting_for =NULL ; + retry = true; + }break ; } }
@@ -829,7 +832,12 @@ static bool coroutine_fn wait_serialising_requests(BdrvTrackedRequest *self)qemu_co_mutex_unlock (&bs->reqs_lock); }while (retry); -return waited; +return found; +} + +static bool coroutine_fn wait_serialising_requests(BdrvTrackedRequest *self) +{ +return do_wait_serialising_requests (self, true); } static int bdrv_check_byte_request(BlockDriverState *bs, int64_t offset,
Signature · 1 line
2.18.0
PATCH v11 12/14 block: add lock/unlock range functions
In reply to Vladimir Sementsov-OgievskiyIntroduce lock/unlock range functionality, based on serialized requests. This is needed to refactor backup, dropping local tracked-request-like synchronization.
--- include/block/block_int.h | 4 ++++ block/io.c | 44 ++++++++++++++++++++++++++++++++++++++- 2 files changed, 47 insertions(+), 1 deletion(-)
include/block/block_int.h +4 -0
@@ -65,6 +65,7 @@ enum BdrvTrackedRequestType {BDRV_TRACKED_WRITE ,BDRV_TRACKED_DISCARD ,BDRV_TRACKED_TRUNCATE , + BDRV_TRACKED_LOCK, }; typedef struct BdrvTrackedRequest {
@@ -968,6 +969,9 @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child, int coroutine_fn bdrv_co_pwritev_part(BdrvChild *child,int64_t offset,unsigned int bytes,QEMUIOVector *qiov,size_t qiov_offset,BdrvRequestFlags flags); +void *coroutine_fn bdrv_co_try_lock(BlockDriverState *bs, +int64_t offset,unsigned int bytes); +void coroutine_fn bdrv_co_unlock(void *opaque); static inline int coroutine_fn bdrv_co_pread(BdrvChild *child,int64_t offset,unsigned int bytes,void *buf,BdrvRequestFlags flags)
block/io.c +43 -1
65 lines not on this page · raw message
Signature · 1 line
2.18.0
PATCH v11 13/14 block: introduce backup-top filter driver
In reply to Vladimir Sementsov-OgievskiyBackup-top filter caches write operations and does copy-before-write operations.
The driver will be used in backup instead of write-notifiers.
--- block/backup-top.h | 37 +++++++ block/backup-top.c | 244 ++++++++++++++++++++++++++++++++++++++++++++ block/Makefile.objs | 2 + 3 files changed, 283 insertions(+) create mode 100644 block/backup-top.h create mode 100644 block/backup-top.c
block/backup-top.h new file +37 -0
@@ -0,0 +1,37 @@ +/* +* backup-top filter driver +* +* The driver performs Copy-Before-Write (CBW) operation: it is injected above +* some node, and before each write it copies _old_ data to the target node. +* +* Copyright (c) 2018-2019 Virtuozzo International GmbH. +* +* Author: +* Sementsov-Ogievskiy Vladimir <[email protected]> +* +* This program is free software; you can redistribute it and/or modify +* it under the terms of the GNU General Public License as published by +* the Free Software Foundation; either version 2 of the License, or +* (at your option) any later version. +* +* This program is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +* GNU General Public License for more details. +* +* You should have received a copy of the GNU General Public License +* along with this program. If not, see <http://www.gnu.org/licenses/>. +*/ + +#ifndef BACKUP_TOP_H +#define BACKUP_TOP_H + +#include "block/block_int.h" + +BlockDriverState *bdrv_backup_top_append (BlockDriverState *source, +const char *filter_node_name, +Error **errp); +void bdrv_backup_top_set_bcs (BlockDriverState *bs,BlockCopyState *copy_state); +void bdrv_backup_top_drop (BlockDriverState *bs); + +#endif /* BACKUP_TOP_H */
block/backup-top.c new file +244 -0
244 lines not on this page · raw message
block/Makefile.objs +2 -0
@@ -41,6 +41,8 @@ block-obj-y += block-copy.o block-obj-y += crypto.o +block-obj-y += backup-top.o + common-obj-y += stream.o nfs.o-libs := $(LIBNFS_LIBS)
Signature · 1 line
2.18.0
PATCH v11 14/14 block/backup: use backup-top instead of write notifiers
In reply to Vladimir Sementsov-OgievskiyDrop write notifiers and use filter node instead.
= Changes =
1. Add filter-node-name argument for backup qmp api. We have to do it in this commit, as 257 needs to be fixed.
2. There are no more write notifiers here, so is_write_notifier parameter is dropped from block-copy paths.
3. Intersecting requests handling changed, now synchronization between backup-top, backup and guest writes are all done in block/block-copy.c and works as follows:
On copy operation, we work only with dirty areas. If bits are dirty it means that there are no requests intersecting with this area. We clear dirty bits and take bdrv range lock (bdrv_co_try_lock) on this area to prevent further operations from interaction with guest (only with guest, as neither backup nor backup-top will touch non-dirty area). If copy-operation failed we set dirty bits back together with releasing the lock.
The actual difference with old scheme is that on guest writes we don't lock the whole region but only dirty-parts, and to be more
precise: only dirty-part we are currently operate on. In old scheme guest write to non-dirty area (which may be safely ignored by backup) may wait for intersecting request, touching some other area which is dirty.
4. To sync with in-flight requests at job finish we now have drained removing of the filter, we don't need rw-lock.
= Notes =
Note the consequence of three objects appearing: backup-top, backup job and block-copy-state:
1. We want to insert backup-top before job creation, to behave similar with mirror and commit, where job is started upon filter.
2. We also have to create block-copy-state after filter injection, as we don't want its source child be replaced by filter. Instead we want to keep BCS.source to be real source node, as we want to use bdrv_co_try_lock in CBW operations and it can't be used on filter, as on filter we already have in-flight (write) request from upper layer.
So, we firstly create inject backup-top, then create job and BCS. BCS is the latest just to not create extra variable for it. Finally we set bcs for backup-top filter.
= Iotest changes =
56: op-blocker doesn't shoot now, as we set it on source, but then check on filter, when trying to start second backup.
To keep the test we instead can catch another collision: both jobs will get 'drive0' job-id, as job-id parameter is unspecified. To prevent interleaving with file-posix locks (as they are dependent on config) let's use another target for second backup.
Also, it's obvious now that we'd like to drop this op-blocker at all and add a test-case for two backups from one node (to different destinations) actually works. But not in these series.
257: The test wants to emulate guest write during backup. They should go to filter node, not to original source node, of course. Therefore we need to specify filter node name and use it.
--- qapi/block-core.json | 8 +- include/block/block-copy.h | 10 +- include/block/block_int.h | 1 + block/backup-top.c | 14 +- block/backup.c | 112 +++----------- block/block-copy.c | 45 ++++-- block/replication.c | 2 +- blockdev.c | 1 + tests/qemu-iotests/056 | 8 +- tests/qemu-iotests/257 | 7 +- tests/qemu-iotests/257.out | 306 ++++++++++++++++++------------------- 11 files changed, 237 insertions(+), 277 deletions(-)
qapi/block-core.json +7 -1
@@ -1391,6 +1391,11 @@ # list without user intervention. # Defaults to true. (Since 2.12) # +# @filter-node-name: the node name that should be assigned to the +# filter driver that the backup job inserts into the graph +# above node specified by @drive. If this option is not given, +# a node name is autogenerated. (Since: 4.2) +# # Note: @on-source-error and @on-target-error only affect background # I/O. If an error occurs during a guest write request, the device's # rerror/werror actions will be used.
@@ -1404,7 +1409,8 @@ '*compress': 'bool', '*on-source-error': 'BlockdevOnError', '*on-target-error': 'BlockdevOnError', - '*auto-finalize': 'bool', '*auto-dismiss': 'bool' } } + '*auto-finalize': 'bool', '*auto-dismiss': 'bool', + '*filter-node-name': 'str' } } ## # @DriveBackup:
include/block/block-copy.h +9 -1
@@ -20,6 +20,14 @@typedef void (*ProgressBytesCallbackFunc )(int64_t bytes,void *opaque);typedef void (*ProgressResetCallbackFunc )(void *opaque); typedef struct BlockCopyState { + /* + * block-copy shares WRITE permission on source, but it relies on user to + * guarantee that nobody is touching dirty areas. + * + * For backup job this is guaranteed by backup-top filter, which don't share + * WRITE permission and calls block_copy on any write, so it will handle + * dirty areas by itself. + */ BlockBackend *source; BlockBackend *target; BdrvDirtyBitmap *copy_bitmap;
@@ -71,6 +79,6 @@ int64_t block_copy_reset_unallocated(BlockCopyState *s, int64_t offset, int64_t *count); int coroutine_fn block_copy(BlockCopyState *s, int64_t start, uint64_t bytes, -bool *error_is_read,bool is_write_notifier); +bool *error_is_read);#endif /* BLOCK_COPY_H */
include/block/block_int.h +1 -0
@@ -1200,6 +1200,7 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,BdrvDirtyBitmap *sync_bitmap, BitmapSyncMode bitmap_mode, bool compress, + const char *filter_node_name, BlockdevOnError on_source_error, BlockdevOnError on_target_error, int creation_flags,
block/backup-top.c +5 -9
@@ -49,15 +49,11 @@ static coroutine_fn int backup_top_co_preadv( static coroutine_fn int backup_top_cbw(BlockDriverState *bs, uint64_t offset,uint64_t bytes) { -/* -* Here we'd like to use block_copy(), but it needs some additional -* synchronization mechanism to prevent intersecting guest writes during -* copy operation. The will appear in further commit (it should be done -* together with moving backup to using of backup-top and to the same -* synchronization mechanism), and for now it is a TODO. -*/ - -abort (); +BDRVBackupTopState *s = bs->opaque; +uint64_t end =QEMU_ALIGN_UP (offset + bytes, s->bcs->cluster_size); +uint64_t off =QEMU_ALIGN_DOWN (offset, s->bcs->cluster_size); + +return block_copy (s->bcs, off, end - off,NULL ); } static int coroutine_fn backup_top_co_pdiscard(BlockDriverState *bs,
block/backup.c +22 -90
226 lines not on this page · raw message
block/block-copy.c +29 -16
113 lines not on this page · raw message
block/replication.c +1 -1
@@ -543,7 +543,7 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode, s->backup_job =backup_job_create (NULL , s->secondary_disk->bs, s->hidden_disk->bs, -0 ,MIRROR_SYNC_MODE_NONE ,NULL ,0 , false, +0 ,MIRROR_SYNC_MODE_NONE ,NULL ,0 , false,NULL ,BLOCKDEV_ON_ERROR_REPORT ,BLOCKDEV_ON_ERROR_REPORT ,JOB_INTERNAL , backup_job_completed, bs,NULL , &local_err);
blockdev.c +1 -0
@@ -3601,6 +3601,7 @@ static BlockJob *do_backup_common(BackupCommon *backup, job =backup_job_create (backup->job_id, bs, target_bs, backup->speed, backup->sync, bmap, backup->bitmap_mode, backup->compress, + backup->filter_node_name, backup->on_source_error, backup->on_target_error, job_flags,NULL ,NULL , txn, errp);
tests/qemu-iotests/056 +5 -3
@@ -133,6 +133,7 @@ class BackupTest(iotests.QMPTestCase): self.vm = iotests.VM() self.test_img = img_create('test') self.dest_img = img_create('dest') + self.dest_img2 = img_create('dest2') self.ref_img = img_create('ref') self.vm.add_drive(self.test_img) self.vm.launch()
@@ -141,6 +142,7 @@ class BackupTest(iotests.QMPTestCase): self.vm.shutdown() try_remove(self.test_img) try_remove(self.dest_img) + try_remove(self.dest_img2) try_remove(self.ref_img) def hmp_io_writes(self, drive, patterns):
@@ -253,9 +255,9 @@ class BackupTest(iotests.QMPTestCase): res = self.vm.qmp('query-block-jobs') self.assert_qmp(res, 'return[0]/status', 'concluded') # Leave zombie job un-dismissed, observe a failure: - res = self.qmp_backup_and_wait(serror="Node 'drive0' is busy: block device is in use by block job: backup", + res = self.qmp_backup_and_wait(serror="Job ID 'drive0' already in use", device='drive0', format=iotests.imgfmt, - sync='full', target=self.dest_img, + sync='full', target=self.dest_img2, auto_dismiss=False) self.assertEqual(res, False) # OK, dismiss the zombie.
@@ -265,7 +267,7 @@ class BackupTest(iotests.QMPTestCase): self.assert_qmp(res, 'return', []) # Ensure it's really gone. self.qmp_backup_and_wait(device='drive0', format=iotests.imgfmt, - sync='full', target=self.dest_img, + sync='full', target=self.dest_img2, auto_dismiss=False) def dismissal_failure(self, dismissal_opt):
tests/qemu-iotests/257 +4 -3
@@ -190,6 +190,7 @@ def blockdev_backup(vm, device, target, sync, **kwargs): device=device, target=target, sync=sync, + filter_node_name='backup-top', **kwargs) return result
@@ -216,7 +217,7 @@ def backup(drive, n, filepath, sync, **kwargs): job_id=job_id, **kwargs) return job_id -def perform_writes(drive, n): +def perform_writes(drive, n, filter_node_name=None): log("--- Write #{:d} ---\n".format(n)) for pattern in GROUPS[n].patterns: cmd = "write -P{:s} 0x{:07x} 0x{:x}".format(
@@ -224,7 +225,7 @@ def perform_writes(drive, n): pattern.offset, pattern.size) log(cmd) - log(drive.vm.hmp_qemu_io(drive.node, cmd)) + log(drive.vm.hmp_qemu_io(filter_node_name or drive.node, cmd)) bitmaps = drive.vm.query_bitmaps() log({'bitmaps': bitmaps}, indent=2) log('')
@@ -355,7 +356,7 @@ def test_bitmap_sync(bsync_mode, msync_mode='bitmap', failure=None): """Issue writes while the job is open to test bitmap divergence.""" # Note: when `failure` is 'intermediate', this isn't called. log('') - bitmaps = perform_writes(drive0, 2) + bitmaps = perform_writes(drive0, 2, filter_node_name='backup-top') # Named bitmap (static, should be unchanged) ebitmap.compare(vm.get_bitmap(drive0.node, 'bitmap0', bitmaps=bitmaps))
tests/qemu-iotests/257.out +153 -153
1,043 lines not on this page · raw message
Signature · 1 line
2.18.0
Quoted 8 lines · Vladimir Sementsov-Ogievskiy
Hi all!
These series introduce backup-top driver. It's a filter-node, which do copy-before-write operation. Mirror uses filter-node for handling guest writes, let's move to filter-node (from write-notifiers) for backup too.
v11: based on Kevin's block branch
Now based on master
Quoted 53 lines
02,03: Add Max's r-b
04: - improve comments
- rebase on droppedbackup_drain
- s/job/bcs/ in trace events
05: one hunk dropped due to rebasing
on droppedbackup_drain, still,
keep Max's r-b
06: rebased on 04 changes, keep Max's r-b
08,13,14: add Max's r-bBased-on:
git://repo.or.cz/qemu/kevin.gitblockVladimir Sementsov-Ogievskiy (14):
block/backup: fixbackup_cow_with_offloadfor last cluster block/backup: split shareable copying part frombackup_do_cowblock/backup: improve comment about image fleecing block/backup: introduce BlockCopyState
block/backup: fix block-comment style
block: moveblock_copyfromblock/backup.cto separate file
block: teachbdrv_debug_breakpointskip filters with backing
iotests: prepare 124 and 257 bitmap querying for backup-top filter
iotests: 257: drop unused Drive.device field
iotests: 257: dropdevice_add
block/io: refactorwait_serialising_requests
block: add lock/unlock range functions
block: introduce backup-top filter driver
block/backup: use backup-top instead of write notifiersqapi/block-core.json | 8 +- block/backup-top.h | 37 ++ include/block/block-copy.h | 84 ++++ include/block/block_int.h | 5 + block.c | 34 +- block/backup-top.c | 240 ++++++++++++ block/backup.c | 440 ++++----------------- block/block-copy.c | 337 ++++++++++++++++ block/io.c | 68 +++- block/replication.c | 2 +- blockdev.c | 1 + block/Makefile.objs | 3 + block/trace-events | 14 +- tests/qemu-iotests/056 | 8 +- tests/qemu-iotests/124 | 83 ++-- tests/qemu-iotests/257 | 91 ++--- tests/qemu-iotests/257.out | 714 ++++++++++++++-------------------- tests/qemu-iotests/iotests.py | 27 ++ 18 files changed, 1278 insertions(+), 918 deletions(-)create mode 100644
block/backup-top.h
create mode 100644include/block/block-copy.h
create mode 100644block/backup-top.c
create mode 100644block/block-copy.c
Signature · 2 lines
Best regards, Vladimir