The following changes since commit 86f4c7e05b1c44dbe1b329a51f311f10aef6ff34: Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20180302' into staging (2018-03-02 14:37:10 +0000) are available in the git repository at: git://repo.or.cz/qemu/kevin.git tags/for-upstream for you to fetch changes up to 9d9b4b640f9e583ff4b24dc762630945f3ccc16d: Merge remote-tracking branch 'mreitz/tags/pull-block-2018-03-02' into queue-block (2018-03-02 18:45:03 +0100) ---------------------------------------------------------------- Block layer patches ---------------------------------------------------------------- Alberto Garcia (3): specs/qcow2: Fix documentation of the compressed cluster descriptor docs: document how to use the l2-cache-entry-size parameter qcow2: Replace align_offset() with ROUND_UP() Anton Nefedov (2): block: fix write with zero flag set and iovector provided iotest 033: add misaligned write-zeroes test via truncate Eric Blake (21): block: Add .bdrv_co_block_status() callback nvme: Drop pointless .bdrv_co_get_block_status() block: Switch passthrough drivers to .bdrv_co_block_status() file-posix: Switch to .bdrv_co_block_status() gluster: Switch to .bdrv_co_block_status() iscsi: Switch cluster_sectors to byte-based iscsi: Switch iscsi_allocmap_update() to byte-based iscsi: Switch to .bdrv_co_block_status() null: Switch to .bdrv_co_block_status() parallels: Switch to .bdrv_co_block_status() qcow: Switch to .bdrv_co_block_status() qcow2: Switch to .bdrv_co_block_status() qed: Switch to .bdrv_co_block_status() raw: Switch to .bdrv_co_block_status() sheepdog: Switch to .bdrv_co_block_status() vdi: Avoid bitrot of debugging code vdi: Switch to .bdrv_co_block_status() vmdk: Switch to .bdrv_co_block_status() vpc: Switch to .bdrv_co_block_status() vvfat: Switch to .bdrv_co_block_status() block: Drop unused .bdrv_co_get_block_status() Kevin Wolf (2): block: test blk_aio_flush() with blk->root == NULL Merge remote-tracking branch 'mreitz/tags/pull-block-2018-03-02' into queue-block Max Reitz (4): qemu-img: Make resize error message more general block/ssh: Pull ssh_grow_file() from ssh_create() block/ssh: Make ssh_grow_file() blocking block/ssh: Add basic .bdrv_truncate() Stefan Hajnoczi (6): aio: rename aio_context_in_iothread() to in_aio_context_home_thread() block: extract AIO_WAIT_WHILE() from BlockDriverState block: add BlockBackend->in_flight counter Revert "IDE: Do not flush empty CDROM drives" block: rename .bdrv_create() to .bdrv_co_create_opts() qcow2: make qcow2_co_create2() a coroutine_fn docs/interop/qcow2.txt | 16 ++++- docs/qcow2-cache.txt | 46 ++++++++++++- block/qcow2.h | 6 -- include/block/aio-wait.h | 116 ++++++++++++++++++++++++++++++++ include/block/aio.h | 7 +- include/block/block.h | 54 ++++----------- include/block/block_int.h | 61 ++++++++++------- block.c | 11 ++- block/blkdebug.c | 20 +++--- block/block-backend.c | 60 +++++++++++++++-- block/commit.c | 2 +- block/crypto.c | 8 +-- block/file-posix.c | 79 +++++++++++----------- block/file-win32.c | 5 +- block/gluster.c | 83 ++++++++++++----------- block/io.c | 98 +++++++++++---------------- block/iscsi.c | 164 ++++++++++++++++++++++++--------------------- block/mirror.c | 2 +- block/nfs.c | 5 +- block/null.c | 23 ++++--- block/nvme.c | 14 ---- block/parallels.c | 28 +++++--- block/qcow.c | 32 +++++---- block/qcow2-bitmap.c | 4 +- block/qcow2-cluster.c | 4 +- block/qcow2-refcount.c | 4 +- block/qcow2-snapshot.c | 10 +-- block/qcow2.c | 60 +++++++++-------- block/qed.c | 82 ++++++++--------------- block/raw-format.c | 21 +++--- block/rbd.c | 6 +- block/sheepdog.c | 36 +++++----- block/ssh.c | 66 +++++++++++++++--- block/throttle.c | 2 +- block/vdi.c | 50 +++++++------- block/vhdx.c | 5 +- block/vmdk.c | 43 +++++------- block/vpc.c | 50 +++++++------- block/vvfat.c | 16 ++--- hw/ide/core.c | 10 +-- qemu-img.c | 2 +- tests/test-block-backend.c | 82 +++++++++++++++++++++++ util/aio-wait.c | 40 +++++++++++ tests/Makefile.include | 2 + tests/qemu-iotests/033 | 29 ++++++++ tests/qemu-iotests/033.out | 13 ++++ util/Makefile.objs | 2 +- 47 files changed, 973 insertions(+), 606 deletions(-) create mode 100644 include/block/aio-wait.h create mode 100644 tests/test-block-backend.c create mode 100644 util/aio-wait.c From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Now that the block layer exposes byte-based allocation, it's time to tackle the drivers. Add a new callback that operates on as small as byte boundaries. Subsequent patches will then update individual drivers, then finally remove .bdrv_co_get_block_status(). The new code also passes through the 'want_zero' hint, which will allow subsequent patches to further optimize callers that only care about how much of the image is allocated (want_zero is false), rather than full details about runs of zeroes and which offsets the allocation actually maps to (want_zero is true). As part of this effort, fix another part of the documentation: the claim in commit 4c41cb4 that BDRV_BLOCK_ALLOCATED is short for 'DATA || ZERO' is a lie at the block layer (see commit e88ae2264), even though it is how the bit is computed from the driver layer. After all, there are intentionally cases where we return ZERO but not ALLOCATED at the block layer, when we know that a read sees zero because the backing file is too short. Note that the driver interface is thus slightly different than the public interface with regards to which bits will be set, and what guarantees are provided on input. We also add an assertion that any driver using the new callback will make progress (the only time pnum will be 0 is if the block layer already handled an out-of-bounds request, or if there is an error); the old driver interface did not provide this guarantee, which could lead to some inf-loops in drastic corner-case failures. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block.h | 14 +++++++------- include/block/block_int.h | 20 +++++++++++++++----- block/io.c | 28 +++++++++++++++++++--------- 3 files changed, 41 insertions(+), 21 deletions(-) diff --git a/include/block/block.h b/include/block/block.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block.h +++ b/include/block/block.h @@ -XXX,XX +XXX,XX @@ typedef struct HDGeometry { * BDRV_BLOCK_ZERO: offset reads as zero * BDRV_BLOCK_OFFSET_VALID: an associated offset exists for accessing raw data * BDRV_BLOCK_ALLOCATED: the content of the block is determined by this - * layer (short for DATA || ZERO), set by block layer - * BDRV_BLOCK_EOF: the returned pnum covers through end of file for this layer + * layer rather than any backing, set by block layer + * BDRV_BLOCK_EOF: the returned pnum covers through end of file for this + * layer, set by block layer * * Internal flag: * BDRV_BLOCK_RAW: for use by passthrough drivers, such as raw, to request * that the block layer recompute the answer from the returned * BDS; must be accompanied by just BDRV_BLOCK_OFFSET_VALID. * - * If BDRV_BLOCK_OFFSET_VALID is set, bits 9-62 (BDRV_BLOCK_OFFSET_MASK) of - * the return value (old interface) or the entire map parameter (new - * interface) represent the offset in the returned BDS that is allocated for - * the corresponding raw data. However, whether that offset actually - * contains data also depends on BDRV_BLOCK_DATA, as follows: + * If BDRV_BLOCK_OFFSET_VALID is set, the map parameter represents the + * host offset within the returned BDS that is allocated for the + * corresponding raw guest data. However, whether that offset + * actually contains data also depends on BDRV_BLOCK_DATA, as follows: * * DATA ZERO OFFSET_VALID * t t t sectors read as zero, returned file is zero at offset diff --git a/include/block/block_int.h b/include/block/block_int.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block_int.h +++ b/include/block/block_int.h @@ -XXX,XX +XXX,XX @@ struct BlockDriver { /* * Building block for bdrv_block_status[_above] and * bdrv_is_allocated[_above]. The driver should answer only - * according to the current layer, and should not set - * BDRV_BLOCK_ALLOCATED, but may set BDRV_BLOCK_RAW. See block.h - * for the meaning of _DATA, _ZERO, and _OFFSET_VALID. The block - * layer guarantees input aligned to request_alignment, as well as - * non-NULL pnum and file. + * according to the current layer, and should only need to set + * BDRV_BLOCK_DATA, BDRV_BLOCK_ZERO, BDRV_BLOCK_OFFSET_VALID, + * and/or BDRV_BLOCK_RAW; if the current layer defers to a backing + * layer, the result should be 0 (and not BDRV_BLOCK_ZERO). See + * block.h for the overall meaning of the bits. As a hint, the + * flag want_zero is true if the caller cares more about precise + * mappings (favor accurate _OFFSET_VALID/_ZERO) or false for + * overall allocation (favor larger *pnum, perhaps by reporting + * _DATA instead of _ZERO). The block layer guarantees input + * clamped to bdrv_getlength() and aligned to request_alignment, + * as well as non-NULL pnum, map, and file; in turn, the driver + * must return an error or set pnum to an aligned non-zero value. */ int64_t coroutine_fn (*bdrv_co_get_block_status)(BlockDriverState *bs, int64_t sector_num, int nb_sectors, int *pnum, BlockDriverState **file); + int coroutine_fn (*bdrv_co_block_status)(BlockDriverState *bs, + bool want_zero, int64_t offset, int64_t bytes, int64_t *pnum, + int64_t *map, BlockDriverState **file); /* * Invalidate any cached meta-data. diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ int64_t coroutine_fn bdrv_co_get_block_status_from_backing(BlockDriverState *bs, * Drivers not implementing the functionality are assumed to not support * backing files, hence all their sectors are reported as allocated. * - * If 'want_zero' is true, the caller is querying for mapping purposes, - * and the result should include BDRV_BLOCK_OFFSET_VALID and - * BDRV_BLOCK_ZERO where possible; otherwise, the result may omit those - * bits particularly if it allows for a larger value in 'pnum'. + * If 'want_zero' is true, the caller is querying for mapping + * purposes, with a focus on valid BDRV_BLOCK_OFFSET_VALID, _DATA, and + * _ZERO where possible; otherwise, the result favors larger 'pnum', + * with a focus on accurate BDRV_BLOCK_ALLOCATED. * * If 'offset' is beyond the end of the disk image the return value is * BDRV_BLOCK_EOF and 'pnum' is set to 0. @@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_block_status(BlockDriverState *bs, /* Must be non-NULL or bdrv_getlength() would have failed */ assert(bs->drv); - if (!bs->drv->bdrv_co_get_block_status) { + if (!bs->drv->bdrv_co_get_block_status && !bs->drv->bdrv_co_block_status) { *pnum = bytes; ret = BDRV_BLOCK_DATA | BDRV_BLOCK_ALLOCATED; if (offset + bytes == total_size) { @@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_block_status(BlockDriverState *bs, bdrv_inc_in_flight(bs); /* Round out to request_alignment boundaries */ - /* TODO: until we have a byte-based driver callback, we also have to - * round out to sectors, even if that is bigger than request_alignment */ - align = MAX(bs->bl.request_alignment, BDRV_SECTOR_SIZE); + align = bs->bl.request_alignment; + if (bs->drv->bdrv_co_get_block_status && align < BDRV_SECTOR_SIZE) { + align = BDRV_SECTOR_SIZE; + } aligned_offset = QEMU_ALIGN_DOWN(offset, align); aligned_bytes = ROUND_UP(offset + bytes, align) - aligned_offset; - { + if (bs->drv->bdrv_co_get_block_status) { int count; /* sectors */ int64_t longret; @@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_block_status(BlockDriverState *bs, } ret = longret & ~BDRV_BLOCK_OFFSET_MASK; *pnum = count * BDRV_SECTOR_SIZE; + } else { + ret = bs->drv->bdrv_co_block_status(bs, want_zero, aligned_offset, + aligned_bytes, pnum, &local_map, + &local_file); + if (ret < 0) { + *pnum = 0; + goto out; + } + assert(*pnum); /* The block driver must make progress */ } /* -- 2.13.6 From: Eric Blake <eblake@redhat.com> Commit bdd6a90 has a bug: drivers should never directly set BDRV_BLOCK_ALLOCATED, but only io.c should do that (as needed). Instead, drivers should report BDRV_BLOCK_DATA if it knows that data comes from this BDS. But let's look at the bigger picture: semantically, the nvme driver is similar to the nbd, null, and raw drivers (no backing file, all data comes from this BDS). But while two of those other drivers have to supply the callback (null because it can special-case BDRV_BLOCK_ZERO, raw because it can special-case a different offset), in this case the block layer defaults are good enough without the callback at all (similar to nbd). So, fix the bug by deletion ;) Signed-off-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/nvme.c | 14 -------------- 1 file changed, 14 deletions(-) diff --git a/block/nvme.c b/block/nvme.c index XXXXXXX..XXXXXXX 100644 --- a/block/nvme.c +++ b/block/nvme.c @@ -XXX,XX +XXX,XX @@ static int nvme_reopen_prepare(BDRVReopenState *reopen_state, return 0; } -static int64_t coroutine_fn nvme_co_get_block_status(BlockDriverState *bs, - int64_t sector_num, - int nb_sectors, int *pnum, - BlockDriverState **file) -{ - *pnum = nb_sectors; - *file = bs; - - return BDRV_BLOCK_ALLOCATED | BDRV_BLOCK_OFFSET_VALID | - (sector_num << BDRV_SECTOR_BITS); -} - static void nvme_refresh_filename(BlockDriverState *bs, QDict *opts) { QINCREF(opts); @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_nvme = { .bdrv_co_flush_to_disk = nvme_co_flush, .bdrv_reopen_prepare = nvme_reopen_prepare, - .bdrv_co_get_block_status = nvme_co_get_block_status, - .bdrv_refresh_filename = nvme_refresh_filename, .bdrv_refresh_limits = nvme_refresh_limits, -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the generic helpers, and all passthrough clients (blkdebug, commit, mirror, throttle) accordingly. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block_int.h | 28 ++++++++++++++++------------ block/blkdebug.c | 20 +++++++++++--------- block/commit.c | 2 +- block/io.c | 36 ++++++++++++++++++++---------------- block/mirror.c | 2 +- block/throttle.c | 2 +- 6 files changed, 50 insertions(+), 40 deletions(-) diff --git a/include/block/block_int.h b/include/block/block_int.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block_int.h +++ b/include/block/block_int.h @@ -XXX,XX +XXX,XX @@ void bdrv_format_default_perms(BlockDriverState *bs, BdrvChild *c, uint64_t *nperm, uint64_t *nshared); /* - * Default implementation for drivers to pass bdrv_co_get_block_status() to + * Default implementation for drivers to pass bdrv_co_block_status() to * their file. */ -int64_t coroutine_fn bdrv_co_get_block_status_from_file(BlockDriverState *bs, - int64_t sector_num, - int nb_sectors, - int *pnum, - BlockDriverState **file); +int coroutine_fn bdrv_co_block_status_from_file(BlockDriverState *bs, + bool want_zero, + int64_t offset, + int64_t bytes, + int64_t *pnum, + int64_t *map, + BlockDriverState **file); /* - * Default implementation for drivers to pass bdrv_co_get_block_status() to + * Default implementation for drivers to pass bdrv_co_block_status() to * their backing file. */ -int64_t coroutine_fn bdrv_co_get_block_status_from_backing(BlockDriverState *bs, - int64_t sector_num, - int nb_sectors, - int *pnum, - BlockDriverState **file); +int coroutine_fn bdrv_co_block_status_from_backing(BlockDriverState *bs, + bool want_zero, + int64_t offset, + int64_t bytes, + int64_t *pnum, + int64_t *map, + BlockDriverState **file); const char *bdrv_get_parent_name(const BlockDriverState *bs); void blk_dev_change_media_cb(BlockBackend *blk, bool load, Error **errp); bool blk_dev_has_removable_media(BlockBackend *blk); diff --git a/block/blkdebug.c b/block/blkdebug.c index XXXXXXX..XXXXXXX 100644 --- a/block/blkdebug.c +++ b/block/blkdebug.c @@ -XXX,XX +XXX,XX @@ static int coroutine_fn blkdebug_co_pdiscard(BlockDriverState *bs, return bdrv_co_pdiscard(bs->file->bs, offset, bytes); } -static int64_t coroutine_fn blkdebug_co_get_block_status( - BlockDriverState *bs, int64_t sector_num, int nb_sectors, int *pnum, - BlockDriverState **file) +static int coroutine_fn blkdebug_co_block_status(BlockDriverState *bs, + bool want_zero, + int64_t offset, + int64_t bytes, + int64_t *pnum, + int64_t *map, + BlockDriverState **file) { - assert(QEMU_IS_ALIGNED(sector_num | nb_sectors, - DIV_ROUND_UP(bs->bl.request_alignment, - BDRV_SECTOR_SIZE))); - return bdrv_co_get_block_status_from_file(bs, sector_num, nb_sectors, - pnum, file); + assert(QEMU_IS_ALIGNED(offset | bytes, bs->bl.request_alignment)); + return bdrv_co_block_status_from_file(bs, want_zero, offset, bytes, + pnum, map, file); } static void blkdebug_close(BlockDriverState *bs) @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_blkdebug = { .bdrv_co_flush_to_disk = blkdebug_co_flush, .bdrv_co_pwrite_zeroes = blkdebug_co_pwrite_zeroes, .bdrv_co_pdiscard = blkdebug_co_pdiscard, - .bdrv_co_get_block_status = blkdebug_co_get_block_status, + .bdrv_co_block_status = blkdebug_co_block_status, .bdrv_debug_event = blkdebug_debug_event, .bdrv_debug_breakpoint = blkdebug_debug_breakpoint, diff --git a/block/commit.c b/block/commit.c index XXXXXXX..XXXXXXX 100644 --- a/block/commit.c +++ b/block/commit.c @@ -XXX,XX +XXX,XX @@ static void bdrv_commit_top_child_perm(BlockDriverState *bs, BdrvChild *c, static BlockDriver bdrv_commit_top = { .format_name = "commit_top", .bdrv_co_preadv = bdrv_commit_top_preadv, - .bdrv_co_get_block_status = bdrv_co_get_block_status_from_backing, + .bdrv_co_block_status = bdrv_co_block_status_from_backing, .bdrv_refresh_filename = bdrv_commit_top_refresh_filename, .bdrv_close = bdrv_commit_top_close, .bdrv_child_perm = bdrv_commit_top_child_perm, diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ typedef struct BdrvCoBlockStatusData { bool done; } BdrvCoBlockStatusData; -int64_t coroutine_fn bdrv_co_get_block_status_from_file(BlockDriverState *bs, - int64_t sector_num, - int nb_sectors, - int *pnum, - BlockDriverState **file) +int coroutine_fn bdrv_co_block_status_from_file(BlockDriverState *bs, + bool want_zero, + int64_t offset, + int64_t bytes, + int64_t *pnum, + int64_t *map, + BlockDriverState **file) { assert(bs->file && bs->file->bs); - *pnum = nb_sectors; + *pnum = bytes; + *map = offset; *file = bs->file->bs; - return BDRV_BLOCK_RAW | BDRV_BLOCK_OFFSET_VALID | - (sector_num << BDRV_SECTOR_BITS); + return BDRV_BLOCK_RAW | BDRV_BLOCK_OFFSET_VALID; } -int64_t coroutine_fn bdrv_co_get_block_status_from_backing(BlockDriverState *bs, - int64_t sector_num, - int nb_sectors, - int *pnum, - BlockDriverState **file) +int coroutine_fn bdrv_co_block_status_from_backing(BlockDriverState *bs, + bool want_zero, + int64_t offset, + int64_t bytes, + int64_t *pnum, + int64_t *map, + BlockDriverState **file) { assert(bs->backing && bs->backing->bs); - *pnum = nb_sectors; + *pnum = bytes; + *map = offset; *file = bs->backing->bs; - return BDRV_BLOCK_RAW | BDRV_BLOCK_OFFSET_VALID | - (sector_num << BDRV_SECTOR_BITS); + return BDRV_BLOCK_RAW | BDRV_BLOCK_OFFSET_VALID; } /* diff --git a/block/mirror.c b/block/mirror.c index XXXXXXX..XXXXXXX 100644 --- a/block/mirror.c +++ b/block/mirror.c @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_mirror_top = { .bdrv_co_pwrite_zeroes = bdrv_mirror_top_pwrite_zeroes, .bdrv_co_pdiscard = bdrv_mirror_top_pdiscard, .bdrv_co_flush = bdrv_mirror_top_flush, - .bdrv_co_get_block_status = bdrv_co_get_block_status_from_backing, + .bdrv_co_block_status = bdrv_co_block_status_from_backing, .bdrv_refresh_filename = bdrv_mirror_top_refresh_filename, .bdrv_close = bdrv_mirror_top_close, .bdrv_child_perm = bdrv_mirror_top_child_perm, diff --git a/block/throttle.c b/block/throttle.c index XXXXXXX..XXXXXXX 100644 --- a/block/throttle.c +++ b/block/throttle.c @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_throttle = { .bdrv_reopen_prepare = throttle_reopen_prepare, .bdrv_reopen_commit = throttle_reopen_commit, .bdrv_reopen_abort = throttle_reopen_abort, - .bdrv_co_get_block_status = bdrv_co_get_block_status_from_file, + .bdrv_co_block_status = bdrv_co_block_status_from_file, .bdrv_co_drain_begin = throttle_co_drain_begin, .bdrv_co_drain_end = throttle_co_drain_end, -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the file protocol driver accordingly. In want_zero mode, we continue to report fine-grained hole information (the caller wants as much mapping detail as possible); but when not in that mode, the caller prefers larger *pnum and merely cares about what offsets are allocated at this layer, rather than where the holes live. Since holes still read as zeroes at this layer (rather than deferring to a backing layer), we can take the shortcut of skipping lseek(), and merely state that all bytes are allocated. We can also drop redundant bounds checks that are already guaranteed by the block layer. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/file-posix.c | 64 +++++++++++++++++++++++++----------------------------- 1 file changed, 30 insertions(+), 34 deletions(-) diff --git a/block/file-posix.c b/block/file-posix.c index XXXXXXX..XXXXXXX 100644 --- a/block/file-posix.c +++ b/block/file-posix.c @@ -XXX,XX +XXX,XX @@ static int find_allocation(BlockDriverState *bs, off_t start, } /* - * Returns the allocation status of the specified sectors. + * Returns the allocation status of the specified offset. * - * If 'sector_num' is beyond the end of the disk image the return value is 0 - * and 'pnum' is set to 0. + * The block layer guarantees 'offset' and 'bytes' are within bounds. * - * 'pnum' is set to the number of sectors (including and immediately following - * the specified sector) that are known to be in the same + * 'pnum' is set to the number of bytes (including and immediately following + * the specified offset) that are known to be in the same * allocated/unallocated state. * - * 'nb_sectors' is the max value 'pnum' should be set to. If nb_sectors goes - * beyond the end of the disk image it will be clamped. + * 'bytes' is the max value 'pnum' should be set to. */ -static int64_t coroutine_fn raw_co_get_block_status(BlockDriverState *bs, - int64_t sector_num, - int nb_sectors, int *pnum, - BlockDriverState **file) -{ - off_t start, data = 0, hole = 0; - int64_t total_size; +static int coroutine_fn raw_co_block_status(BlockDriverState *bs, + bool want_zero, + int64_t offset, + int64_t bytes, int64_t *pnum, + int64_t *map, + BlockDriverState **file) +{ + off_t data = 0, hole = 0; int ret; ret = fd_open(bs); @@ -XXX,XX +XXX,XX @@ static int64_t coroutine_fn raw_co_get_block_status(BlockDriverState *bs, return ret; } - start = sector_num * BDRV_SECTOR_SIZE; - total_size = bdrv_getlength(bs); - if (total_size < 0) { - return total_size; - } else if (start >= total_size) { - *pnum = 0; - return 0; - } else if (start + nb_sectors * BDRV_SECTOR_SIZE > total_size) { - nb_sectors = DIV_ROUND_UP(total_size - start, BDRV_SECTOR_SIZE); + if (!want_zero) { + *pnum = bytes; + *map = offset; + *file = bs; + return BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID; } - ret = find_allocation(bs, start, &data, &hole); + ret = find_allocation(bs, offset, &data, &hole); if (ret == -ENXIO) { /* Trailing hole */ - *pnum = nb_sectors; + *pnum = bytes; ret = BDRV_BLOCK_ZERO; } else if (ret < 0) { /* No info available, so pretend there are no holes */ - *pnum = nb_sectors; + *pnum = bytes; ret = BDRV_BLOCK_DATA; - } else if (data == start) { - /* On a data extent, compute sectors to the end of the extent, + } else if (data == offset) { + /* On a data extent, compute bytes to the end of the extent, * possibly including a partial sector at EOF. */ - *pnum = MIN(nb_sectors, DIV_ROUND_UP(hole - start, BDRV_SECTOR_SIZE)); + *pnum = MIN(bytes, hole - offset); ret = BDRV_BLOCK_DATA; } else { - /* On a hole, compute sectors to the beginning of the next extent. */ - assert(hole == start); - *pnum = MIN(nb_sectors, (data - start) / BDRV_SECTOR_SIZE); + /* On a hole, compute bytes to the beginning of the next extent. */ + assert(hole == offset); + *pnum = MIN(bytes, data - offset); ret = BDRV_BLOCK_ZERO; } + *map = offset; *file = bs; - return ret | BDRV_BLOCK_OFFSET_VALID | start; + return ret | BDRV_BLOCK_OFFSET_VALID; } static coroutine_fn BlockAIOCB *raw_aio_pdiscard(BlockDriverState *bs, @@ -XXX,XX +XXX,XX @@ BlockDriver bdrv_file = { .bdrv_close = raw_close, .bdrv_create = raw_create, .bdrv_has_zero_init = bdrv_has_zero_init_1, - .bdrv_co_get_block_status = raw_co_get_block_status, + .bdrv_co_block_status = raw_co_block_status, .bdrv_co_pwrite_zeroes = raw_co_pwrite_zeroes, .bdrv_co_preadv = raw_co_preadv, -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the gluster driver accordingly. In want_zero mode, we continue to report fine-grained hole information (the caller wants as much mapping detail as possible); but when not in that mode, the caller prefers larger *pnum and merely cares about what offsets are allocated at this layer, rather than where the holes live. Since holes still read as zeroes at this layer (rather than deferring to a backing layer), we can take the shortcut of skipping find_allocation(), and merely state that all bytes are allocated. We can also drop redundant bounds checks that are already guaranteed by the block layer. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/gluster.c | 70 ++++++++++++++++++++++++++++----------------------------- 1 file changed, 34 insertions(+), 36 deletions(-) diff --git a/block/gluster.c b/block/gluster.c index XXXXXXX..XXXXXXX 100644 --- a/block/gluster.c +++ b/block/gluster.c @@ -XXX,XX +XXX,XX @@ exit: } /* - * Returns the allocation status of the specified sectors. + * Returns the allocation status of the specified offset. * - * If 'sector_num' is beyond the end of the disk image the return value is 0 - * and 'pnum' is set to 0. + * The block layer guarantees 'offset' and 'bytes' are within bounds. * - * 'pnum' is set to the number of sectors (including and immediately following - * the specified sector) that are known to be in the same + * 'pnum' is set to the number of bytes (including and immediately following + * the specified offset) that are known to be in the same * allocated/unallocated state. * - * 'nb_sectors' is the max value 'pnum' should be set to. If nb_sectors goes - * beyond the end of the disk image it will be clamped. + * 'bytes' is the max value 'pnum' should be set to. * - * (Based on raw_co_get_block_status() from file-posix.c.) + * (Based on raw_co_block_status() from file-posix.c.) */ -static int64_t coroutine_fn qemu_gluster_co_get_block_status( - BlockDriverState *bs, int64_t sector_num, int nb_sectors, int *pnum, - BlockDriverState **file) +static int coroutine_fn qemu_gluster_co_block_status(BlockDriverState *bs, + bool want_zero, + int64_t offset, + int64_t bytes, + int64_t *pnum, + int64_t *map, + BlockDriverState **file) { BDRVGlusterState *s = bs->opaque; - off_t start, data = 0, hole = 0; - int64_t total_size; + off_t data = 0, hole = 0; int ret = -EINVAL; if (!s->fd) { return ret; } - start = sector_num * BDRV_SECTOR_SIZE; - total_size = bdrv_getlength(bs); - if (total_size < 0) { - return total_size; - } else if (start >= total_size) { - *pnum = 0; - return 0; - } else if (start + nb_sectors * BDRV_SECTOR_SIZE > total_size) { - nb_sectors = DIV_ROUND_UP(total_size - start, BDRV_SECTOR_SIZE); + if (!want_zero) { + *pnum = bytes; + *map = offset; + *file = bs; + return BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID; } - ret = find_allocation(bs, start, &data, &hole); + ret = find_allocation(bs, offset, &data, &hole); if (ret == -ENXIO) { /* Trailing hole */ - *pnum = nb_sectors; + *pnum = bytes; ret = BDRV_BLOCK_ZERO; } else if (ret < 0) { /* No info available, so pretend there are no holes */ - *pnum = nb_sectors; + *pnum = bytes; ret = BDRV_BLOCK_DATA; - } else if (data == start) { - /* On a data extent, compute sectors to the end of the extent, + } else if (data == offset) { + /* On a data extent, compute bytes to the end of the extent, * possibly including a partial sector at EOF. */ - *pnum = MIN(nb_sectors, DIV_ROUND_UP(hole - start, BDRV_SECTOR_SIZE)); + *pnum = MIN(bytes, hole - offset); ret = BDRV_BLOCK_DATA; } else { - /* On a hole, compute sectors to the beginning of the next extent. */ - assert(hole == start); - *pnum = MIN(nb_sectors, (data - start) / BDRV_SECTOR_SIZE); + /* On a hole, compute bytes to the beginning of the next extent. */ + assert(hole == offset); + *pnum = MIN(bytes, data - offset); ret = BDRV_BLOCK_ZERO; } + *map = offset; *file = bs; - return ret | BDRV_BLOCK_OFFSET_VALID | start; + return ret | BDRV_BLOCK_OFFSET_VALID; } @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_gluster = { #ifdef CONFIG_GLUSTERFS_ZEROFILL .bdrv_co_pwrite_zeroes = qemu_gluster_co_pwrite_zeroes, #endif - .bdrv_co_get_block_status = qemu_gluster_co_get_block_status, + .bdrv_co_block_status = qemu_gluster_co_block_status, .create_opts = &qemu_gluster_create_opts, }; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_gluster_tcp = { #ifdef CONFIG_GLUSTERFS_ZEROFILL .bdrv_co_pwrite_zeroes = qemu_gluster_co_pwrite_zeroes, #endif - .bdrv_co_get_block_status = qemu_gluster_co_get_block_status, + .bdrv_co_block_status = qemu_gluster_co_block_status, .create_opts = &qemu_gluster_create_opts, }; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_gluster_unix = { #ifdef CONFIG_GLUSTERFS_ZEROFILL .bdrv_co_pwrite_zeroes = qemu_gluster_co_pwrite_zeroes, #endif - .bdrv_co_get_block_status = qemu_gluster_co_get_block_status, + .bdrv_co_block_status = qemu_gluster_co_block_status, .create_opts = &qemu_gluster_create_opts, }; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_gluster_rdma = { #ifdef CONFIG_GLUSTERFS_ZEROFILL .bdrv_co_pwrite_zeroes = qemu_gluster_co_pwrite_zeroes, #endif - .bdrv_co_get_block_status = qemu_gluster_co_get_block_status, + .bdrv_co_block_status = qemu_gluster_co_block_status, .create_opts = &qemu_gluster_create_opts, }; -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually converting to byte-based interfaces, as they are easier to reason about than sector-based. Convert all uses of the cluster size in sectors, along with adding assertions that we are not dividing by zero. Improve some comment grammar while in the area. Signed-off-by: Eric Blake <eblake@redhat.com> Acked-by: Paolo Bonzini <pbonzini@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/iscsi.c | 56 +++++++++++++++++++++++++++++++++++--------------------- 1 file changed, 35 insertions(+), 21 deletions(-) diff --git a/block/iscsi.c b/block/iscsi.c index XXXXXXX..XXXXXXX 100644 --- a/block/iscsi.c +++ b/block/iscsi.c @@ -XXX,XX +XXX,XX @@ typedef struct IscsiLun { unsigned long *allocmap; unsigned long *allocmap_valid; long allocmap_size; - int cluster_sectors; + int cluster_size; bool use_16_for_rw; bool write_protected; bool lbpme; @@ -XXX,XX +XXX,XX @@ static int iscsi_allocmap_init(IscsiLun *iscsilun, int open_flags) { iscsi_allocmap_free(iscsilun); + assert(iscsilun->cluster_size); iscsilun->allocmap_size = - DIV_ROUND_UP(sector_lun2qemu(iscsilun->num_blocks, iscsilun), - iscsilun->cluster_sectors); + DIV_ROUND_UP(iscsilun->num_blocks * iscsilun->block_size, + iscsilun->cluster_size); iscsilun->allocmap = bitmap_try_new(iscsilun->allocmap_size); if (!iscsilun->allocmap) { @@ -XXX,XX +XXX,XX @@ static int iscsi_allocmap_init(IscsiLun *iscsilun, int open_flags) } if (open_flags & BDRV_O_NOCACHE) { - /* in case that cache.direct = on all allocmap entries are + /* when cache.direct = on all allocmap entries are * treated as invalid to force a relookup of the block * status on every read request */ return 0; @@ -XXX,XX +XXX,XX @@ iscsi_allocmap_update(IscsiLun *iscsilun, int64_t sector_num, int nb_sectors, bool allocated, bool valid) { int64_t cl_num_expanded, nb_cls_expanded, cl_num_shrunk, nb_cls_shrunk; + int cluster_sectors = iscsilun->cluster_size >> BDRV_SECTOR_BITS; if (iscsilun->allocmap == NULL) { return; } /* expand to entirely contain all affected clusters */ - cl_num_expanded = sector_num / iscsilun->cluster_sectors; + assert(cluster_sectors); + cl_num_expanded = sector_num / cluster_sectors; nb_cls_expanded = DIV_ROUND_UP(sector_num + nb_sectors, - iscsilun->cluster_sectors) - cl_num_expanded; + cluster_sectors) - cl_num_expanded; /* shrink to touch only completely contained clusters */ - cl_num_shrunk = DIV_ROUND_UP(sector_num, iscsilun->cluster_sectors); - nb_cls_shrunk = (sector_num + nb_sectors) / iscsilun->cluster_sectors + cl_num_shrunk = DIV_ROUND_UP(sector_num, cluster_sectors); + nb_cls_shrunk = (sector_num + nb_sectors) / cluster_sectors - cl_num_shrunk; if (allocated) { bitmap_set(iscsilun->allocmap, cl_num_expanded, nb_cls_expanded); @@ -XXX,XX +XXX,XX @@ iscsi_allocmap_is_allocated(IscsiLun *iscsilun, int64_t sector_num, if (iscsilun->allocmap == NULL) { return true; } - size = DIV_ROUND_UP(sector_num + nb_sectors, iscsilun->cluster_sectors); + assert(iscsilun->cluster_size); + size = DIV_ROUND_UP(sector_num + nb_sectors, + iscsilun->cluster_size >> BDRV_SECTOR_BITS); return !(find_next_bit(iscsilun->allocmap, size, - sector_num / iscsilun->cluster_sectors) == size); + sector_num * BDRV_SECTOR_SIZE / + iscsilun->cluster_size) == size); } static inline bool iscsi_allocmap_is_valid(IscsiLun *iscsilun, @@ -XXX,XX +XXX,XX @@ static inline bool iscsi_allocmap_is_valid(IscsiLun *iscsilun, if (iscsilun->allocmap_valid == NULL) { return false; } - size = DIV_ROUND_UP(sector_num + nb_sectors, iscsilun->cluster_sectors); + assert(iscsilun->cluster_size); + size = DIV_ROUND_UP(sector_num + nb_sectors, + iscsilun->cluster_size >> BDRV_SECTOR_BITS); return (find_next_zero_bit(iscsilun->allocmap_valid, size, - sector_num / iscsilun->cluster_sectors) == size); + sector_num * BDRV_SECTOR_SIZE / + iscsilun->cluster_size) == size); } static int coroutine_fn @@ -XXX,XX +XXX,XX @@ static int coroutine_fn iscsi_co_readv(BlockDriverState *bs, BlockDriverState *file; /* check the block status from the beginning of the cluster * containing the start sector */ - int64_t ret = iscsi_co_get_block_status(bs, - sector_num - sector_num % iscsilun->cluster_sectors, - BDRV_REQUEST_MAX_SECTORS, &pnum, &file); + int cluster_sectors = iscsilun->cluster_size >> BDRV_SECTOR_BITS; + int head; + int64_t ret; + + assert(cluster_sectors); + head = sector_num % cluster_sectors; + ret = iscsi_co_get_block_status(bs, sector_num - head, + BDRV_REQUEST_MAX_SECTORS, &pnum, + &file); if (ret < 0) { return ret; } /* if the whole request falls into an unallocated area we can avoid - * to read and directly return zeroes instead */ - if (ret & BDRV_BLOCK_ZERO && - pnum >= nb_sectors + sector_num % iscsilun->cluster_sectors) { + * reading and directly return zeroes instead */ + if (ret & BDRV_BLOCK_ZERO && pnum >= nb_sectors + head) { qemu_iovec_memset(iov, 0, 0x00, iov->size); return 0; } @@ -XXX,XX +XXX,XX @@ static int iscsi_open(BlockDriverState *bs, QDict *options, int flags, * reasonable size */ if (iscsilun->bl.opt_unmap_gran * iscsilun->block_size >= 4 * 1024 && iscsilun->bl.opt_unmap_gran * iscsilun->block_size <= 16 * 1024 * 1024) { - iscsilun->cluster_sectors = (iscsilun->bl.opt_unmap_gran * - iscsilun->block_size) >> BDRV_SECTOR_BITS; + iscsilun->cluster_size = iscsilun->bl.opt_unmap_gran * + iscsilun->block_size; if (iscsilun->lbprz) { ret = iscsi_allocmap_init(iscsilun, bs->open_flags); } @@ -XXX,XX +XXX,XX @@ static int iscsi_get_info(BlockDriverState *bs, BlockDriverInfo *bdi) { IscsiLun *iscsilun = bs->opaque; bdi->unallocated_blocks_are_zero = iscsilun->lbprz; - bdi->cluster_size = iscsilun->cluster_sectors * BDRV_SECTOR_SIZE; + bdi->cluster_size = iscsilun->cluster_size; return 0; } -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually converting to byte-based interfaces, as they are easier to reason about than sector-based. Convert all uses of the allocmap (no semantic change). Callers that already had bytes available are simpler, and callers that now scale to bytes will be easier to switch to byte-based in the future. Signed-off-by: Eric Blake <eblake@redhat.com> Acked-by: Paolo Bonzini <pbonzini@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/iscsi.c | 90 +++++++++++++++++++++++++++++------------------------------ 1 file changed, 44 insertions(+), 46 deletions(-) diff --git a/block/iscsi.c b/block/iscsi.c index XXXXXXX..XXXXXXX 100644 --- a/block/iscsi.c +++ b/block/iscsi.c @@ -XXX,XX +XXX,XX @@ static int iscsi_allocmap_init(IscsiLun *iscsilun, int open_flags) } static void -iscsi_allocmap_update(IscsiLun *iscsilun, int64_t sector_num, - int nb_sectors, bool allocated, bool valid) +iscsi_allocmap_update(IscsiLun *iscsilun, int64_t offset, + int64_t bytes, bool allocated, bool valid) { int64_t cl_num_expanded, nb_cls_expanded, cl_num_shrunk, nb_cls_shrunk; - int cluster_sectors = iscsilun->cluster_size >> BDRV_SECTOR_BITS; if (iscsilun->allocmap == NULL) { return; } /* expand to entirely contain all affected clusters */ - assert(cluster_sectors); - cl_num_expanded = sector_num / cluster_sectors; - nb_cls_expanded = DIV_ROUND_UP(sector_num + nb_sectors, - cluster_sectors) - cl_num_expanded; + assert(iscsilun->cluster_size); + cl_num_expanded = offset / iscsilun->cluster_size; + nb_cls_expanded = DIV_ROUND_UP(offset + bytes, + iscsilun->cluster_size) - cl_num_expanded; /* shrink to touch only completely contained clusters */ - cl_num_shrunk = DIV_ROUND_UP(sector_num, cluster_sectors); - nb_cls_shrunk = (sector_num + nb_sectors) / cluster_sectors - - cl_num_shrunk; + cl_num_shrunk = DIV_ROUND_UP(offset, iscsilun->cluster_size); + nb_cls_shrunk = (offset + bytes) / iscsilun->cluster_size - cl_num_shrunk; if (allocated) { bitmap_set(iscsilun->allocmap, cl_num_expanded, nb_cls_expanded); } else { @@ -XXX,XX +XXX,XX @@ iscsi_allocmap_update(IscsiLun *iscsilun, int64_t sector_num, } static void -iscsi_allocmap_set_allocated(IscsiLun *iscsilun, int64_t sector_num, - int nb_sectors) +iscsi_allocmap_set_allocated(IscsiLun *iscsilun, int64_t offset, + int64_t bytes) { - iscsi_allocmap_update(iscsilun, sector_num, nb_sectors, true, true); + iscsi_allocmap_update(iscsilun, offset, bytes, true, true); } static void -iscsi_allocmap_set_unallocated(IscsiLun *iscsilun, int64_t sector_num, - int nb_sectors) +iscsi_allocmap_set_unallocated(IscsiLun *iscsilun, int64_t offset, + int64_t bytes) { /* Note: if cache.direct=on the fifth argument to iscsi_allocmap_update * is ignored, so this will in effect be an iscsi_allocmap_set_invalid. */ - iscsi_allocmap_update(iscsilun, sector_num, nb_sectors, false, true); + iscsi_allocmap_update(iscsilun, offset, bytes, false, true); } -static void iscsi_allocmap_set_invalid(IscsiLun *iscsilun, int64_t sector_num, - int nb_sectors) +static void iscsi_allocmap_set_invalid(IscsiLun *iscsilun, int64_t offset, + int64_t bytes) { - iscsi_allocmap_update(iscsilun, sector_num, nb_sectors, false, false); + iscsi_allocmap_update(iscsilun, offset, bytes, false, false); } static void iscsi_allocmap_invalidate(IscsiLun *iscsilun) @@ -XXX,XX +XXX,XX @@ static void iscsi_allocmap_invalidate(IscsiLun *iscsilun) } static inline bool -iscsi_allocmap_is_allocated(IscsiLun *iscsilun, int64_t sector_num, - int nb_sectors) +iscsi_allocmap_is_allocated(IscsiLun *iscsilun, int64_t offset, + int64_t bytes) { unsigned long size; if (iscsilun->allocmap == NULL) { return true; } assert(iscsilun->cluster_size); - size = DIV_ROUND_UP(sector_num + nb_sectors, - iscsilun->cluster_size >> BDRV_SECTOR_BITS); + size = DIV_ROUND_UP(offset + bytes, iscsilun->cluster_size); return !(find_next_bit(iscsilun->allocmap, size, - sector_num * BDRV_SECTOR_SIZE / - iscsilun->cluster_size) == size); + offset / iscsilun->cluster_size) == size); } static inline bool iscsi_allocmap_is_valid(IscsiLun *iscsilun, - int64_t sector_num, int nb_sectors) + int64_t offset, int64_t bytes) { unsigned long size; if (iscsilun->allocmap_valid == NULL) { return false; } assert(iscsilun->cluster_size); - size = DIV_ROUND_UP(sector_num + nb_sectors, - iscsilun->cluster_size >> BDRV_SECTOR_BITS); + size = DIV_ROUND_UP(offset + bytes, iscsilun->cluster_size); return (find_next_zero_bit(iscsilun->allocmap_valid, size, - sector_num * BDRV_SECTOR_SIZE / - iscsilun->cluster_size) == size); + offset / iscsilun->cluster_size) == size); } static int coroutine_fn @@ -XXX,XX +XXX,XX @@ retry: } if (iTask.status != SCSI_STATUS_GOOD) { - iscsi_allocmap_set_invalid(iscsilun, sector_num, nb_sectors); + iscsi_allocmap_set_invalid(iscsilun, sector_num * BDRV_SECTOR_SIZE, + nb_sectors * BDRV_SECTOR_SIZE); error_report("iSCSI WRITE10/16 failed at lba %" PRIu64 ": %s", lba, iTask.err_str); r = iTask.err_code; goto out_unlock; } - iscsi_allocmap_set_allocated(iscsilun, sector_num, nb_sectors); + iscsi_allocmap_set_allocated(iscsilun, sector_num * BDRV_SECTOR_SIZE, + nb_sectors * BDRV_SECTOR_SIZE); out_unlock: qemu_mutex_unlock(&iscsilun->mutex); @@ -XXX,XX +XXX,XX @@ retry: } if (ret & BDRV_BLOCK_ZERO) { - iscsi_allocmap_set_unallocated(iscsilun, sector_num, *pnum); + iscsi_allocmap_set_unallocated(iscsilun, sector_num * BDRV_SECTOR_SIZE, + *pnum * BDRV_SECTOR_SIZE); } else { - iscsi_allocmap_set_allocated(iscsilun, sector_num, *pnum); + iscsi_allocmap_set_allocated(iscsilun, sector_num * BDRV_SECTOR_SIZE, + *pnum * BDRV_SECTOR_SIZE); } if (*pnum > nb_sectors) { @@ -XXX,XX +XXX,XX @@ static int coroutine_fn iscsi_co_readv(BlockDriverState *bs, /* if cache.direct is off and we have a valid entry in our allocation map * we can skip checking the block status and directly return zeroes if * the request falls within an unallocated area */ - if (iscsi_allocmap_is_valid(iscsilun, sector_num, nb_sectors) && - !iscsi_allocmap_is_allocated(iscsilun, sector_num, nb_sectors)) { + if (iscsi_allocmap_is_valid(iscsilun, sector_num * BDRV_SECTOR_SIZE, + nb_sectors * BDRV_SECTOR_SIZE) && + !iscsi_allocmap_is_allocated(iscsilun, sector_num * BDRV_SECTOR_SIZE, + nb_sectors * BDRV_SECTOR_SIZE)) { qemu_iovec_memset(iov, 0, 0x00, iov->size); return 0; } if (nb_sectors >= ISCSI_CHECKALLOC_THRES && - !iscsi_allocmap_is_valid(iscsilun, sector_num, nb_sectors) && - !iscsi_allocmap_is_allocated(iscsilun, sector_num, nb_sectors)) { + !iscsi_allocmap_is_valid(iscsilun, sector_num * BDRV_SECTOR_SIZE, + nb_sectors * BDRV_SECTOR_SIZE) && + !iscsi_allocmap_is_allocated(iscsilun, sector_num * BDRV_SECTOR_SIZE, + nb_sectors * BDRV_SECTOR_SIZE)) { int pnum; BlockDriverState *file; /* check the block status from the beginning of the cluster @@ -XXX,XX +XXX,XX @@ retry: goto retry; } - iscsi_allocmap_set_invalid(iscsilun, offset >> BDRV_SECTOR_BITS, - bytes >> BDRV_SECTOR_BITS); + iscsi_allocmap_set_invalid(iscsilun, offset, bytes); if (iTask.status == SCSI_STATUS_CHECK_CONDITION) { /* the target might fail with a check condition if it @@ -XXX,XX +XXX,XX @@ retry: } if (iTask.status != SCSI_STATUS_GOOD) { - iscsi_allocmap_set_invalid(iscsilun, offset >> BDRV_SECTOR_BITS, - bytes >> BDRV_SECTOR_BITS); + iscsi_allocmap_set_invalid(iscsilun, offset, bytes); error_report("iSCSI WRITESAME10/16 failed at lba %" PRIu64 ": %s", lba, iTask.err_str); r = iTask.err_code; @@ -XXX,XX +XXX,XX @@ retry: } if (flags & BDRV_REQ_MAY_UNMAP) { - iscsi_allocmap_set_invalid(iscsilun, offset >> BDRV_SECTOR_BITS, - bytes >> BDRV_SECTOR_BITS); + iscsi_allocmap_set_invalid(iscsilun, offset, bytes); } else { - iscsi_allocmap_set_allocated(iscsilun, offset >> BDRV_SECTOR_BITS, - bytes >> BDRV_SECTOR_BITS); + iscsi_allocmap_set_allocated(iscsilun, offset, bytes); } out_unlock: -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the iscsi driver accordingly. In this case, it is handy to teach iscsi_co_block_status() to handle a NULL map and file parameter, even though the block layer passes non-NULL values, because we also call the function directly. For now, there are no optimizations done based on the want_zero flag. We can also make the simplification of asserting that the block layer passed in aligned values. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/iscsi.c | 69 ++++++++++++++++++++++++++++------------------------------- 1 file changed, 33 insertions(+), 36 deletions(-) diff --git a/block/iscsi.c b/block/iscsi.c index XXXXXXX..XXXXXXX 100644 --- a/block/iscsi.c +++ b/block/iscsi.c @@ -XXX,XX +XXX,XX @@ out_unlock: -static int64_t coroutine_fn iscsi_co_get_block_status(BlockDriverState *bs, - int64_t sector_num, - int nb_sectors, int *pnum, - BlockDriverState **file) +static int coroutine_fn iscsi_co_block_status(BlockDriverState *bs, + bool want_zero, int64_t offset, + int64_t bytes, int64_t *pnum, + int64_t *map, + BlockDriverState **file) { IscsiLun *iscsilun = bs->opaque; struct scsi_get_lba_status *lbas = NULL; struct scsi_lba_status_descriptor *lbasd = NULL; struct IscsiTask iTask; uint64_t lba; - int64_t ret; + int ret; iscsi_co_init_iscsitask(iscsilun, &iTask); - if (!is_sector_request_lun_aligned(sector_num, nb_sectors, iscsilun)) { - ret = -EINVAL; - goto out; - } + assert(QEMU_IS_ALIGNED(offset | bytes, iscsilun->block_size)); /* default to all sectors allocated */ - ret = BDRV_BLOCK_DATA; - ret |= (sector_num << BDRV_SECTOR_BITS) | BDRV_BLOCK_OFFSET_VALID; - *pnum = nb_sectors; + ret = BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID; + if (map) { + *map = offset; + } + *pnum = bytes; /* LUN does not support logical block provisioning */ if (!iscsilun->lbpme) { goto out; } - lba = sector_qemu2lun(sector_num, iscsilun); + lba = offset / iscsilun->block_size; qemu_mutex_lock(&iscsilun->mutex); retry: @@ -XXX,XX +XXX,XX @@ retry: lbasd = &lbas->descriptors[0]; - if (sector_qemu2lun(sector_num, iscsilun) != lbasd->lba) { + if (lba != lbasd->lba) { ret = -EIO; goto out_unlock; } - *pnum = sector_lun2qemu(lbasd->num_blocks, iscsilun); + *pnum = lbasd->num_blocks * iscsilun->block_size; if (lbasd->provisioning == SCSI_PROVISIONING_TYPE_DEALLOCATED || lbasd->provisioning == SCSI_PROVISIONING_TYPE_ANCHORED) { @@ -XXX,XX +XXX,XX @@ retry: } if (ret & BDRV_BLOCK_ZERO) { - iscsi_allocmap_set_unallocated(iscsilun, sector_num * BDRV_SECTOR_SIZE, - *pnum * BDRV_SECTOR_SIZE); + iscsi_allocmap_set_unallocated(iscsilun, offset, *pnum); } else { - iscsi_allocmap_set_allocated(iscsilun, sector_num * BDRV_SECTOR_SIZE, - *pnum * BDRV_SECTOR_SIZE); + iscsi_allocmap_set_allocated(iscsilun, offset, *pnum); } - if (*pnum > nb_sectors) { - *pnum = nb_sectors; + if (*pnum > bytes) { + *pnum = bytes; } out_unlock: qemu_mutex_unlock(&iscsilun->mutex); @@ -XXX,XX +XXX,XX @@ out: if (iTask.task != NULL) { scsi_free_scsi_task(iTask.task); } - if (ret > 0 && ret & BDRV_BLOCK_OFFSET_VALID) { + if (ret > 0 && ret & BDRV_BLOCK_OFFSET_VALID && file) { *file = bs; } return ret; @@ -XXX,XX +XXX,XX @@ static int coroutine_fn iscsi_co_readv(BlockDriverState *bs, nb_sectors * BDRV_SECTOR_SIZE) && !iscsi_allocmap_is_allocated(iscsilun, sector_num * BDRV_SECTOR_SIZE, nb_sectors * BDRV_SECTOR_SIZE)) { - int pnum; - BlockDriverState *file; + int64_t pnum; /* check the block status from the beginning of the cluster * containing the start sector */ - int cluster_sectors = iscsilun->cluster_size >> BDRV_SECTOR_BITS; - int head; - int64_t ret; - - assert(cluster_sectors); - head = sector_num % cluster_sectors; - ret = iscsi_co_get_block_status(bs, sector_num - head, - BDRV_REQUEST_MAX_SECTORS, &pnum, - &file); + int64_t head; + int ret; + + assert(iscsilun->cluster_size); + head = (sector_num * BDRV_SECTOR_SIZE) % iscsilun->cluster_size; + ret = iscsi_co_block_status(bs, true, + sector_num * BDRV_SECTOR_SIZE - head, + BDRV_REQUEST_MAX_BYTES, &pnum, NULL, NULL); if (ret < 0) { return ret; } /* if the whole request falls into an unallocated area we can avoid * reading and directly return zeroes instead */ - if (ret & BDRV_BLOCK_ZERO && pnum >= nb_sectors + head) { + if (ret & BDRV_BLOCK_ZERO && + pnum >= nb_sectors * BDRV_SECTOR_SIZE + head) { qemu_iovec_memset(iov, 0, 0x00, iov->size); return 0; } @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_iscsi = { .bdrv_truncate = iscsi_truncate, .bdrv_refresh_limits = iscsi_refresh_limits, - .bdrv_co_get_block_status = iscsi_co_get_block_status, + .bdrv_co_block_status = iscsi_co_block_status, .bdrv_co_pdiscard = iscsi_co_pdiscard, .bdrv_co_pwrite_zeroes = iscsi_co_pwrite_zeroes, .bdrv_co_readv = iscsi_co_readv, @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_iser = { .bdrv_truncate = iscsi_truncate, .bdrv_refresh_limits = iscsi_refresh_limits, - .bdrv_co_get_block_status = iscsi_co_get_block_status, + .bdrv_co_block_status = iscsi_co_block_status, .bdrv_co_pdiscard = iscsi_co_pdiscard, .bdrv_co_pwrite_zeroes = iscsi_co_pwrite_zeroes, .bdrv_co_readv = iscsi_co_readv, -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the null driver accordingly. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/null.c | 23 ++++++++++++----------- 1 file changed, 12 insertions(+), 11 deletions(-) diff --git a/block/null.c b/block/null.c index XXXXXXX..XXXXXXX 100644 --- a/block/null.c +++ b/block/null.c @@ -XXX,XX +XXX,XX @@ static int null_reopen_prepare(BDRVReopenState *reopen_state, return 0; } -static int64_t coroutine_fn null_co_get_block_status(BlockDriverState *bs, - int64_t sector_num, - int nb_sectors, int *pnum, - BlockDriverState **file) +static int coroutine_fn null_co_block_status(BlockDriverState *bs, + bool want_zero, int64_t offset, + int64_t bytes, int64_t *pnum, + int64_t *map, + BlockDriverState **file) { BDRVNullState *s = bs->opaque; - off_t start = sector_num * BDRV_SECTOR_SIZE; + int ret = BDRV_BLOCK_OFFSET_VALID; - *pnum = nb_sectors; + *pnum = bytes; + *map = offset; *file = bs; if (s->read_zeroes) { - return BDRV_BLOCK_OFFSET_VALID | start | BDRV_BLOCK_ZERO; - } else { - return BDRV_BLOCK_OFFSET_VALID | start; + ret |= BDRV_BLOCK_ZERO; } + return ret; } static void null_refresh_filename(BlockDriverState *bs, QDict *opts) @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_null_co = { .bdrv_co_flush_to_disk = null_co_flush, .bdrv_reopen_prepare = null_reopen_prepare, - .bdrv_co_get_block_status = null_co_get_block_status, + .bdrv_co_block_status = null_co_block_status, .bdrv_refresh_filename = null_refresh_filename, }; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_null_aio = { .bdrv_aio_flush = null_aio_flush, .bdrv_reopen_prepare = null_reopen_prepare, - .bdrv_co_get_block_status = null_co_get_block_status, + .bdrv_co_block_status = null_co_block_status, .bdrv_refresh_filename = null_refresh_filename, }; -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the parallels driver accordingly. Note that the internal function block_status() is still sector-based, because it is still in use by other sector-based functions; but that's okay because request_alignment is 512 as a result of those functions. For now, no optimizations are added based on the mapping hint. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/parallels.c | 22 +++++++++++++++------- 1 file changed, 15 insertions(+), 7 deletions(-) diff --git a/block/parallels.c b/block/parallels.c index XXXXXXX..XXXXXXX 100644 --- a/block/parallels.c +++ b/block/parallels.c @@ -XXX,XX +XXX,XX @@ static coroutine_fn int parallels_co_flush_to_os(BlockDriverState *bs) } -static int64_t coroutine_fn parallels_co_get_block_status(BlockDriverState *bs, - int64_t sector_num, int nb_sectors, int *pnum, BlockDriverState **file) +static int coroutine_fn parallels_co_block_status(BlockDriverState *bs, + bool want_zero, + int64_t offset, + int64_t bytes, + int64_t *pnum, + int64_t *map, + BlockDriverState **file) { BDRVParallelsState *s = bs->opaque; - int64_t offset; + int count; + assert(QEMU_IS_ALIGNED(offset | bytes, BDRV_SECTOR_SIZE)); qemu_co_mutex_lock(&s->lock); - offset = block_status(s, sector_num, nb_sectors, pnum); + offset = block_status(s, offset >> BDRV_SECTOR_BITS, + bytes >> BDRV_SECTOR_BITS, &count); qemu_co_mutex_unlock(&s->lock); + *pnum = count * BDRV_SECTOR_SIZE; if (offset < 0) { return 0; } + *map = offset * BDRV_SECTOR_SIZE; *file = bs->file->bs; - return (offset << BDRV_SECTOR_BITS) | - BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID; + return BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID; } static coroutine_fn int parallels_co_writev(BlockDriverState *bs, @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_parallels = { .bdrv_open = parallels_open, .bdrv_close = parallels_close, .bdrv_child_perm = bdrv_format_default_perms, - .bdrv_co_get_block_status = parallels_co_get_block_status, + .bdrv_co_block_status = parallels_co_block_status, .bdrv_has_zero_init = bdrv_has_zero_init_1, .bdrv_co_flush_to_os = parallels_co_flush_to_os, .bdrv_co_readv = parallels_co_readv, -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the qcow driver accordingly. There is no intent to optimize based on the want_zero flag for this format. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/qcow.c | 27 ++++++++++++++++----------- 1 file changed, 16 insertions(+), 11 deletions(-) diff --git a/block/qcow.c b/block/qcow.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow.c +++ b/block/qcow.c @@ -XXX,XX +XXX,XX @@ static int get_cluster_offset(BlockDriverState *bs, return 1; } -static int64_t coroutine_fn qcow_co_get_block_status(BlockDriverState *bs, - int64_t sector_num, int nb_sectors, int *pnum, BlockDriverState **file) +static int coroutine_fn qcow_co_block_status(BlockDriverState *bs, + bool want_zero, + int64_t offset, int64_t bytes, + int64_t *pnum, int64_t *map, + BlockDriverState **file) { BDRVQcowState *s = bs->opaque; - int index_in_cluster, n, ret; + int index_in_cluster, ret; + int64_t n; uint64_t cluster_offset; qemu_co_mutex_lock(&s->lock); - ret = get_cluster_offset(bs, sector_num << 9, 0, 0, 0, 0, &cluster_offset); + ret = get_cluster_offset(bs, offset, 0, 0, 0, 0, &cluster_offset); qemu_co_mutex_unlock(&s->lock); if (ret < 0) { return ret; } - index_in_cluster = sector_num & (s->cluster_sectors - 1); - n = s->cluster_sectors - index_in_cluster; - if (n > nb_sectors) - n = nb_sectors; + index_in_cluster = offset & (s->cluster_size - 1); + n = s->cluster_size - index_in_cluster; + if (n > bytes) { + n = bytes; + } *pnum = n; if (!cluster_offset) { return 0; @@ -XXX,XX +XXX,XX @@ static int64_t coroutine_fn qcow_co_get_block_status(BlockDriverState *bs, if ((cluster_offset & QCOW_OFLAG_COMPRESSED) || s->crypto) { return BDRV_BLOCK_DATA; } - cluster_offset |= (index_in_cluster << BDRV_SECTOR_BITS); + *map = cluster_offset | index_in_cluster; *file = bs->file->bs; - return BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID | cluster_offset; + return BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID; } static int decompress_buffer(uint8_t *out_buf, int out_buf_size, @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_qcow = { .bdrv_co_readv = qcow_co_readv, .bdrv_co_writev = qcow_co_writev, - .bdrv_co_get_block_status = qcow_co_get_block_status, + .bdrv_co_block_status = qcow_co_block_status, .bdrv_make_empty = qcow_make_empty, .bdrv_co_pwritev_compressed = qcow_co_pwritev_compressed, -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the qcow2 driver accordingly. For now, we are ignoring the 'want_zero' hint. However, it should be relatively straightforward to honor the hint as a way to return larger *pnum values when we have consecutive clusters with the same data/zero status but which differ only in having non-consecutive mappings. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/qcow2.c | 24 +++++++++++++----------- 1 file changed, 13 insertions(+), 11 deletions(-) diff --git a/block/qcow2.c b/block/qcow2.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -XXX,XX +XXX,XX @@ static void qcow2_join_options(QDict *options, QDict *old_options) } } -static int64_t coroutine_fn qcow2_co_get_block_status(BlockDriverState *bs, - int64_t sector_num, int nb_sectors, int *pnum, BlockDriverState **file) +static int coroutine_fn qcow2_co_block_status(BlockDriverState *bs, + bool want_zero, + int64_t offset, int64_t count, + int64_t *pnum, int64_t *map, + BlockDriverState **file) { BDRVQcow2State *s = bs->opaque; uint64_t cluster_offset; int index_in_cluster, ret; unsigned int bytes; - int64_t status = 0; + int status = 0; - bytes = MIN(INT_MAX, nb_sectors * BDRV_SECTOR_SIZE); + bytes = MIN(INT_MAX, count); qemu_co_mutex_lock(&s->lock); - ret = qcow2_get_cluster_offset(bs, sector_num << BDRV_SECTOR_BITS, &bytes, - &cluster_offset); + ret = qcow2_get_cluster_offset(bs, offset, &bytes, &cluster_offset); qemu_co_mutex_unlock(&s->lock); if (ret < 0) { return ret; } - *pnum = bytes >> BDRV_SECTOR_BITS; + *pnum = bytes; if (cluster_offset != 0 && ret != QCOW2_CLUSTER_COMPRESSED && !s->crypto) { - index_in_cluster = sector_num & (s->cluster_sectors - 1); - cluster_offset |= (index_in_cluster << BDRV_SECTOR_BITS); + index_in_cluster = offset & (s->cluster_size - 1); + *map = cluster_offset | index_in_cluster; *file = bs->file->bs; - status |= BDRV_BLOCK_OFFSET_VALID | cluster_offset; + status |= BDRV_BLOCK_OFFSET_VALID; } if (ret == QCOW2_CLUSTER_ZERO_PLAIN || ret == QCOW2_CLUSTER_ZERO_ALLOC) { status |= BDRV_BLOCK_ZERO; @@ -XXX,XX +XXX,XX @@ BlockDriver bdrv_qcow2 = { .bdrv_child_perm = bdrv_format_default_perms, .bdrv_create = qcow2_create, .bdrv_has_zero_init = bdrv_has_zero_init_1, - .bdrv_co_get_block_status = qcow2_co_get_block_status, + .bdrv_co_block_status = qcow2_co_block_status, .bdrv_co_preadv = qcow2_co_preadv, .bdrv_co_pwritev = qcow2_co_pwritev, -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the qed driver accordingly, taking the opportunity to inline qed_is_allocated_cb() into its lone caller (the callback used to be important, until we switched qed to coroutines). There is no intent to optimize based on the want_zero flag for this format. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/qed.c | 76 +++++++++++++++++++------------------------------------------ 1 file changed, 24 insertions(+), 52 deletions(-) diff --git a/block/qed.c b/block/qed.c index XXXXXXX..XXXXXXX 100644 --- a/block/qed.c +++ b/block/qed.c @@ -XXX,XX +XXX,XX @@ finish: return ret; } -typedef struct { - BlockDriverState *bs; - Coroutine *co; - uint64_t pos; - int64_t status; - int *pnum; - BlockDriverState **file; -} QEDIsAllocatedCB; - -/* Called with table_lock held. */ -static void qed_is_allocated_cb(void *opaque, int ret, uint64_t offset, size_t len) +static int coroutine_fn bdrv_qed_co_block_status(BlockDriverState *bs, + bool want_zero, + int64_t pos, int64_t bytes, + int64_t *pnum, int64_t *map, + BlockDriverState **file) { - QEDIsAllocatedCB *cb = opaque; - BDRVQEDState *s = cb->bs->opaque; - *cb->pnum = len / BDRV_SECTOR_SIZE; + BDRVQEDState *s = bs->opaque; + size_t len = MIN(bytes, SIZE_MAX); + int status; + QEDRequest request = { .l2_table = NULL }; + uint64_t offset; + int ret; + + qemu_co_mutex_lock(&s->table_lock); + ret = qed_find_cluster(s, &request, pos, &len, &offset); + + *pnum = len; switch (ret) { case QED_CLUSTER_FOUND: - offset |= qed_offset_into_cluster(s, cb->pos); - cb->status = BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID | offset; - *cb->file = cb->bs->file->bs; + *map = offset | qed_offset_into_cluster(s, pos); + status = BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID; + *file = bs->file->bs; break; case QED_CLUSTER_ZERO: - cb->status = BDRV_BLOCK_ZERO; + status = BDRV_BLOCK_ZERO; break; case QED_CLUSTER_L2: case QED_CLUSTER_L1: - cb->status = 0; + status = 0; break; default: assert(ret < 0); - cb->status = ret; + status = ret; break; } - if (cb->co) { - aio_co_wake(cb->co); - } -} - -static int64_t coroutine_fn bdrv_qed_co_get_block_status(BlockDriverState *bs, - int64_t sector_num, - int nb_sectors, int *pnum, - BlockDriverState **file) -{ - BDRVQEDState *s = bs->opaque; - size_t len = (size_t)nb_sectors * BDRV_SECTOR_SIZE; - QEDIsAllocatedCB cb = { - .bs = bs, - .pos = (uint64_t)sector_num * BDRV_SECTOR_SIZE, - .status = BDRV_BLOCK_OFFSET_MASK, - .pnum = pnum, - .file = file, - }; - QEDRequest request = { .l2_table = NULL }; - uint64_t offset; - int ret; - - qemu_co_mutex_lock(&s->table_lock); - ret = qed_find_cluster(s, &request, cb.pos, &len, &offset); - qed_is_allocated_cb(&cb, ret, offset, len); - - /* The callback was invoked immediately */ - assert(cb.status != BDRV_BLOCK_OFFSET_MASK); - qed_unref_l2_cache_entry(request.l2_table); qemu_co_mutex_unlock(&s->table_lock); - return cb.status; + return status; } static BDRVQEDState *acb_to_s(QEDAIOCB *acb) @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_qed = { .bdrv_child_perm = bdrv_format_default_perms, .bdrv_create = bdrv_qed_create, .bdrv_has_zero_init = bdrv_has_zero_init_1, - .bdrv_co_get_block_status = bdrv_qed_co_get_block_status, + .bdrv_co_block_status = bdrv_qed_co_block_status, .bdrv_co_readv = bdrv_qed_co_readv, .bdrv_co_writev = bdrv_qed_co_writev, .bdrv_co_pwrite_zeroes = bdrv_qed_co_pwrite_zeroes, -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the raw driver accordingly. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/raw-format.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/block/raw-format.c b/block/raw-format.c index XXXXXXX..XXXXXXX 100644 --- a/block/raw-format.c +++ b/block/raw-format.c @@ -XXX,XX +XXX,XX @@ fail: return ret; } -static int64_t coroutine_fn raw_co_get_block_status(BlockDriverState *bs, - int64_t sector_num, - int nb_sectors, int *pnum, +static int coroutine_fn raw_co_block_status(BlockDriverState *bs, + bool want_zero, int64_t offset, + int64_t bytes, int64_t *pnum, + int64_t *map, BlockDriverState **file) { BDRVRawState *s = bs->opaque; - *pnum = nb_sectors; + *pnum = bytes; *file = bs->file->bs; - sector_num += s->offset / BDRV_SECTOR_SIZE; - return BDRV_BLOCK_RAW | BDRV_BLOCK_OFFSET_VALID | - (sector_num << BDRV_SECTOR_BITS); + *map = offset + s->offset; + return BDRV_BLOCK_RAW | BDRV_BLOCK_OFFSET_VALID; } static int coroutine_fn raw_co_pwrite_zeroes(BlockDriverState *bs, @@ -XXX,XX +XXX,XX @@ BlockDriver bdrv_raw = { .bdrv_co_pwritev = &raw_co_pwritev, .bdrv_co_pwrite_zeroes = &raw_co_pwrite_zeroes, .bdrv_co_pdiscard = &raw_co_pdiscard, - .bdrv_co_get_block_status = &raw_co_get_block_status, + .bdrv_co_block_status = &raw_co_block_status, .bdrv_truncate = &raw_truncate, .bdrv_getlength = &raw_getlength, .has_variable_length = true, -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the sheepdog driver accordingly. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Reviewed-by: Jeff Cody <jcody@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/sheepdog.c | 26 +++++++++++++------------- 1 file changed, 13 insertions(+), 13 deletions(-) diff --git a/block/sheepdog.c b/block/sheepdog.c index XXXXXXX..XXXXXXX 100644 --- a/block/sheepdog.c +++ b/block/sheepdog.c @@ -XXX,XX +XXX,XX @@ static coroutine_fn int sd_co_pdiscard(BlockDriverState *bs, int64_t offset, return acb.ret; } -static coroutine_fn int64_t -sd_co_get_block_status(BlockDriverState *bs, int64_t sector_num, int nb_sectors, - int *pnum, BlockDriverState **file) +static coroutine_fn int +sd_co_block_status(BlockDriverState *bs, bool want_zero, int64_t offset, + int64_t bytes, int64_t *pnum, int64_t *map, + BlockDriverState **file) { BDRVSheepdogState *s = bs->opaque; SheepdogInode *inode = &s->inode; uint32_t object_size = (UINT32_C(1) << inode->block_size_shift); - uint64_t offset = sector_num * BDRV_SECTOR_SIZE; unsigned long start = offset / object_size, - end = DIV_ROUND_UP((sector_num + nb_sectors) * - BDRV_SECTOR_SIZE, object_size); + end = DIV_ROUND_UP(offset + bytes, object_size); unsigned long idx; - int64_t ret = BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID | offset; + *map = offset; + int ret = BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID; for (idx = start; idx < end; idx++) { if (inode->data_vdi_id[idx] == 0) { @@ -XXX,XX +XXX,XX @@ sd_co_get_block_status(BlockDriverState *bs, int64_t sector_num, int nb_sectors, } } - *pnum = (idx - start) * object_size / BDRV_SECTOR_SIZE; - if (*pnum > nb_sectors) { - *pnum = nb_sectors; + *pnum = (idx - start) * object_size; + if (*pnum > bytes) { + *pnum = bytes; } if (ret > 0 && ret & BDRV_BLOCK_OFFSET_VALID) { *file = bs; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_sheepdog = { .bdrv_co_writev = sd_co_writev, .bdrv_co_flush_to_disk = sd_co_flush_to_disk, .bdrv_co_pdiscard = sd_co_pdiscard, - .bdrv_co_get_block_status = sd_co_get_block_status, + .bdrv_co_block_status = sd_co_block_status, .bdrv_snapshot_create = sd_snapshot_create, .bdrv_snapshot_goto = sd_snapshot_goto, @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_sheepdog_tcp = { .bdrv_co_writev = sd_co_writev, .bdrv_co_flush_to_disk = sd_co_flush_to_disk, .bdrv_co_pdiscard = sd_co_pdiscard, - .bdrv_co_get_block_status = sd_co_get_block_status, + .bdrv_co_block_status = sd_co_block_status, .bdrv_snapshot_create = sd_snapshot_create, .bdrv_snapshot_goto = sd_snapshot_goto, @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_sheepdog_unix = { .bdrv_co_writev = sd_co_writev, .bdrv_co_flush_to_disk = sd_co_flush_to_disk, .bdrv_co_pdiscard = sd_co_pdiscard, - .bdrv_co_get_block_status = sd_co_get_block_status, + .bdrv_co_block_status = sd_co_block_status, .bdrv_snapshot_create = sd_snapshot_create, .bdrv_snapshot_goto = sd_snapshot_goto, -- 2.13.6 From: Eric Blake <eblake@redhat.com> Rework the debug define so that we always get -Wformat checking, even when debugging is disabled. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Stefan Weil <sw@weilnetz.de> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/vdi.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/block/vdi.c b/block/vdi.c index XXXXXXX..XXXXXXX 100644 --- a/block/vdi.c +++ b/block/vdi.c @@ -XXX,XX +XXX,XX @@ #define DEFAULT_CLUSTER_SIZE (1 * MiB) #if defined(CONFIG_VDI_DEBUG) -#define logout(fmt, ...) \ - fprintf(stderr, "vdi\t%-24s" fmt, __func__, ##__VA_ARGS__) +#define VDI_DEBUG 1 #else -#define logout(fmt, ...) ((void)0) +#define VDI_DEBUG 0 #endif +#define logout(fmt, ...) \ + do { \ + if (VDI_DEBUG) { \ + fprintf(stderr, "vdi\t%-24s" fmt, __func__, ##__VA_ARGS__); \ + } \ + } while (0) + /* Image signature. */ #define VDI_SIGNATURE 0xbeda107f -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the vdi driver accordingly. Note that the TODO is already covered (the block layer guarantees bounds of its requests), and that we can remove the now-unused s->block_sectors. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/vdi.c | 33 +++++++++++++-------------------- 1 file changed, 13 insertions(+), 20 deletions(-) diff --git a/block/vdi.c b/block/vdi.c index XXXXXXX..XXXXXXX 100644 --- a/block/vdi.c +++ b/block/vdi.c @@ -XXX,XX +XXX,XX @@ typedef struct { uint32_t *bmap; /* Size of block (bytes). */ uint32_t block_size; - /* Size of block (sectors). */ - uint32_t block_sectors; /* First sector of block map. */ uint32_t bmap_sector; /* VDI header (converted to host endianness). */ @@ -XXX,XX +XXX,XX @@ static int vdi_open(BlockDriverState *bs, QDict *options, int flags, bs->total_sectors = header.disk_size / SECTOR_SIZE; s->block_size = header.block_size; - s->block_sectors = header.block_size / SECTOR_SIZE; s->bmap_sector = header.offset_bmap / SECTOR_SIZE; s->header = header; @@ -XXX,XX +XXX,XX @@ static int vdi_reopen_prepare(BDRVReopenState *state, return 0; } -static int64_t coroutine_fn vdi_co_get_block_status(BlockDriverState *bs, - int64_t sector_num, int nb_sectors, int *pnum, BlockDriverState **file) +static int coroutine_fn vdi_co_block_status(BlockDriverState *bs, + bool want_zero, + int64_t offset, int64_t bytes, + int64_t *pnum, int64_t *map, + BlockDriverState **file) { - /* TODO: Check for too large sector_num (in bdrv_is_allocated or here). */ BDRVVdiState *s = (BDRVVdiState *)bs->opaque; - size_t bmap_index = sector_num / s->block_sectors; - size_t sector_in_block = sector_num % s->block_sectors; - int n_sectors = s->block_sectors - sector_in_block; + size_t bmap_index = offset / s->block_size; + size_t index_in_block = offset % s->block_size; uint32_t bmap_entry = le32_to_cpu(s->bmap[bmap_index]); - uint64_t offset; int result; - logout("%p, %" PRId64 ", %d, %p\n", bs, sector_num, nb_sectors, pnum); - if (n_sectors > nb_sectors) { - n_sectors = nb_sectors; - } - *pnum = n_sectors; + logout("%p, %" PRId64 ", %" PRId64 ", %p\n", bs, offset, bytes, pnum); + *pnum = MIN(s->block_size - index_in_block, bytes); result = VDI_IS_ALLOCATED(bmap_entry); if (!result) { return 0; } - offset = s->header.offset_data + - (uint64_t)bmap_entry * s->block_size + - sector_in_block * SECTOR_SIZE; + *map = s->header.offset_data + (uint64_t)bmap_entry * s->block_size + + index_in_block; *file = bs->file->bs; - return BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID | offset; + return BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID; } static int coroutine_fn @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_vdi = { .bdrv_child_perm = bdrv_format_default_perms, .bdrv_create = vdi_create, .bdrv_has_zero_init = bdrv_has_zero_init_1, - .bdrv_co_get_block_status = vdi_co_get_block_status, + .bdrv_co_block_status = vdi_co_block_status, .bdrv_make_empty = vdi_make_empty, .bdrv_co_preadv = vdi_co_preadv, -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the vmdk driver accordingly. Drop the now-unused vmdk_find_index_in_cluster(). Also, fix a pre-existing bug: if find_extent() fails (unlikely, since the block layer did a bounds check), then we must return a failure, rather than 0. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/vmdk.c | 38 ++++++++++++++------------------------ 1 file changed, 14 insertions(+), 24 deletions(-) diff --git a/block/vmdk.c b/block/vmdk.c index XXXXXXX..XXXXXXX 100644 --- a/block/vmdk.c +++ b/block/vmdk.c @@ -XXX,XX +XXX,XX @@ static inline uint64_t vmdk_find_offset_in_cluster(VmdkExtent *extent, return extent_relative_offset % cluster_size; } -static inline uint64_t vmdk_find_index_in_cluster(VmdkExtent *extent, - int64_t sector_num) -{ - uint64_t offset; - offset = vmdk_find_offset_in_cluster(extent, sector_num * BDRV_SECTOR_SIZE); - return offset / BDRV_SECTOR_SIZE; -} - -static int64_t coroutine_fn vmdk_co_get_block_status(BlockDriverState *bs, - int64_t sector_num, int nb_sectors, int *pnum, BlockDriverState **file) +static int coroutine_fn vmdk_co_block_status(BlockDriverState *bs, + bool want_zero, + int64_t offset, int64_t bytes, + int64_t *pnum, int64_t *map, + BlockDriverState **file) { BDRVVmdkState *s = bs->opaque; int64_t index_in_cluster, n, ret; - uint64_t offset; + uint64_t cluster_offset; VmdkExtent *extent; - extent = find_extent(s, sector_num, NULL); + extent = find_extent(s, offset >> BDRV_SECTOR_BITS, NULL); if (!extent) { - return 0; + return -EIO; } qemu_co_mutex_lock(&s->lock); - ret = get_cluster_offset(bs, extent, NULL, - sector_num * 512, false, &offset, + ret = get_cluster_offset(bs, extent, NULL, offset, false, &cluster_offset, 0, 0); qemu_co_mutex_unlock(&s->lock); - index_in_cluster = vmdk_find_index_in_cluster(extent, sector_num); + index_in_cluster = vmdk_find_offset_in_cluster(extent, offset); switch (ret) { case VMDK_ERROR: ret = -EIO; @@ -XXX,XX +XXX,XX @@ static int64_t coroutine_fn vmdk_co_get_block_status(BlockDriverState *bs, ret = BDRV_BLOCK_DATA; if (!extent->compressed) { ret |= BDRV_BLOCK_OFFSET_VALID; - ret |= (offset + (index_in_cluster << BDRV_SECTOR_BITS)) - & BDRV_BLOCK_OFFSET_MASK; + *map = cluster_offset + index_in_cluster; } *file = extent->file->bs; break; } - n = extent->cluster_sectors - index_in_cluster; - if (n > nb_sectors) { - n = nb_sectors; - } - *pnum = n; + n = extent->cluster_sectors * BDRV_SECTOR_SIZE - index_in_cluster; + *pnum = MIN(n, bytes); return ret; } @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_vmdk = { .bdrv_close = vmdk_close, .bdrv_create = vmdk_create, .bdrv_co_flush_to_disk = vmdk_co_flush, - .bdrv_co_get_block_status = vmdk_co_get_block_status, + .bdrv_co_block_status = vmdk_co_block_status, .bdrv_get_allocated_file_size = vmdk_get_allocated_file_size, .bdrv_has_zero_init = vmdk_has_zero_init, .bdrv_get_specific_info = vmdk_get_specific_info, -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the vpc driver accordingly. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/vpc.c | 45 +++++++++++++++++++++++---------------------- 1 file changed, 23 insertions(+), 22 deletions(-) diff --git a/block/vpc.c b/block/vpc.c index XXXXXXX..XXXXXXX 100644 --- a/block/vpc.c +++ b/block/vpc.c @@ -XXX,XX +XXX,XX @@ fail: return ret; } -static int64_t coroutine_fn vpc_co_get_block_status(BlockDriverState *bs, - int64_t sector_num, int nb_sectors, int *pnum, BlockDriverState **file) +static int coroutine_fn vpc_co_block_status(BlockDriverState *bs, + bool want_zero, + int64_t offset, int64_t bytes, + int64_t *pnum, int64_t *map, + BlockDriverState **file) { BDRVVPCState *s = bs->opaque; VHDFooter *footer = (VHDFooter*) s->footer_buf; - int64_t start, offset; + int64_t image_offset; bool allocated; - int64_t ret; - int n; + int ret; + int64_t n; if (be32_to_cpu(footer->type) == VHD_FIXED) { - *pnum = nb_sectors; + *pnum = bytes; + *map = offset; *file = bs->file->bs; - return BDRV_BLOCK_RAW | BDRV_BLOCK_OFFSET_VALID | - (sector_num << BDRV_SECTOR_BITS); + return BDRV_BLOCK_RAW | BDRV_BLOCK_OFFSET_VALID; } qemu_co_mutex_lock(&s->lock); - offset = get_image_offset(bs, sector_num << BDRV_SECTOR_BITS, false, NULL); - start = offset; - allocated = (offset != -1); + image_offset = get_image_offset(bs, offset, false, NULL); + allocated = (image_offset != -1); *pnum = 0; ret = 0; do { /* All sectors in a block are contiguous (without using the bitmap) */ - n = ROUND_UP(sector_num + 1, s->block_size / BDRV_SECTOR_SIZE) - - sector_num; - n = MIN(n, nb_sectors); + n = ROUND_UP(offset + 1, s->block_size) - offset; + n = MIN(n, bytes); *pnum += n; - sector_num += n; - nb_sectors -= n; + offset += n; + bytes -= n; /* *pnum can't be greater than one block for allocated * sectors since there is always a bitmap in between. */ if (allocated) { *file = bs->file->bs; - ret = BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID | start; + *map = image_offset; + ret = BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID; break; } - if (nb_sectors == 0) { + if (bytes == 0) { break; } - offset = get_image_offset(bs, sector_num << BDRV_SECTOR_BITS, false, - NULL); - } while (offset == -1); + image_offset = get_image_offset(bs, offset, false, NULL); + } while (image_offset == -1); qemu_co_mutex_unlock(&s->lock); return ret; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_vpc = { .bdrv_co_preadv = vpc_co_preadv, .bdrv_co_pwritev = vpc_co_pwritev, - .bdrv_co_get_block_status = vpc_co_get_block_status, + .bdrv_co_block_status = vpc_co_block_status, .bdrv_get_info = vpc_get_info, -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Update the vvfat driver accordingly. Note that we can rely on the block driver having already clamped limits to our block size, and simplify accordingly. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/vvfat.c | 16 +++++++--------- 1 file changed, 7 insertions(+), 9 deletions(-) diff --git a/block/vvfat.c b/block/vvfat.c index XXXXXXX..XXXXXXX 100644 --- a/block/vvfat.c +++ b/block/vvfat.c @@ -XXX,XX +XXX,XX @@ vvfat_co_pwritev(BlockDriverState *bs, uint64_t offset, uint64_t bytes, return ret; } -static int64_t coroutine_fn vvfat_co_get_block_status(BlockDriverState *bs, - int64_t sector_num, int nb_sectors, int *n, BlockDriverState **file) +static int coroutine_fn vvfat_co_block_status(BlockDriverState *bs, + bool want_zero, int64_t offset, + int64_t bytes, int64_t *n, + int64_t *map, + BlockDriverState **file) { - *n = bs->total_sectors - sector_num; - if (*n > nb_sectors) { - *n = nb_sectors; - } else if (*n < 0) { - return 0; - } + *n = bytes; return BDRV_BLOCK_DATA; } @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_vvfat = { .bdrv_co_preadv = vvfat_co_preadv, .bdrv_co_pwritev = vvfat_co_pwritev, - .bdrv_co_get_block_status = vvfat_co_get_block_status, + .bdrv_co_block_status = vvfat_co_block_status, }; static void bdrv_vvfat_init(void) -- 2.13.6 From: Eric Blake <eblake@redhat.com> We are gradually moving away from sector-based interfaces, towards byte-based. Now that all drivers have been updated to provide the byte-based .bdrv_co_block_status(), we can delete the sector-based interface. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block_int.h | 3 --- block/io.c | 50 ++++++++++------------------------------------- 2 files changed, 10 insertions(+), 43 deletions(-) diff --git a/include/block/block_int.h b/include/block/block_int.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block_int.h +++ b/include/block/block_int.h @@ -XXX,XX +XXX,XX @@ struct BlockDriver { * as well as non-NULL pnum, map, and file; in turn, the driver * must return an error or set pnum to an aligned non-zero value. */ - int64_t coroutine_fn (*bdrv_co_get_block_status)(BlockDriverState *bs, - int64_t sector_num, int nb_sectors, int *pnum, - BlockDriverState **file); int coroutine_fn (*bdrv_co_block_status)(BlockDriverState *bs, bool want_zero, int64_t offset, int64_t bytes, int64_t *pnum, int64_t *map, BlockDriverState **file); diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_block_status(BlockDriverState *bs, /* Must be non-NULL or bdrv_getlength() would have failed */ assert(bs->drv); - if (!bs->drv->bdrv_co_get_block_status && !bs->drv->bdrv_co_block_status) { + if (!bs->drv->bdrv_co_block_status) { *pnum = bytes; ret = BDRV_BLOCK_DATA | BDRV_BLOCK_ALLOCATED; if (offset + bytes == total_size) { @@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_block_status(BlockDriverState *bs, /* Round out to request_alignment boundaries */ align = bs->bl.request_alignment; - if (bs->drv->bdrv_co_get_block_status && align < BDRV_SECTOR_SIZE) { - align = BDRV_SECTOR_SIZE; - } aligned_offset = QEMU_ALIGN_DOWN(offset, align); aligned_bytes = ROUND_UP(offset + bytes, align) - aligned_offset; - if (bs->drv->bdrv_co_get_block_status) { - int count; /* sectors */ - int64_t longret; - - assert(QEMU_IS_ALIGNED(aligned_offset | aligned_bytes, - BDRV_SECTOR_SIZE)); - /* - * The contract allows us to return pnum smaller than bytes, even - * if the next query would see the same status; we truncate the - * request to avoid overflowing the driver's 32-bit interface. - */ - longret = bs->drv->bdrv_co_get_block_status( - bs, aligned_offset >> BDRV_SECTOR_BITS, - MIN(INT_MAX, aligned_bytes) >> BDRV_SECTOR_BITS, &count, - &local_file); - if (longret < 0) { - assert(INT_MIN <= longret); - ret = longret; - goto out; - } - if (longret & BDRV_BLOCK_OFFSET_VALID) { - local_map = longret & BDRV_BLOCK_OFFSET_MASK; - } - ret = longret & ~BDRV_BLOCK_OFFSET_MASK; - *pnum = count * BDRV_SECTOR_SIZE; - } else { - ret = bs->drv->bdrv_co_block_status(bs, want_zero, aligned_offset, - aligned_bytes, pnum, &local_map, - &local_file); - if (ret < 0) { - *pnum = 0; - goto out; - } - assert(*pnum); /* The block driver must make progress */ + ret = bs->drv->bdrv_co_block_status(bs, want_zero, aligned_offset, + aligned_bytes, pnum, &local_map, + &local_file); + if (ret < 0) { + *pnum = 0; + goto out; } /* - * The driver's result must be a multiple of request_alignment. + * The driver's result must be a non-zero multiple of request_alignment. * Clamp pnum and adjust map to original request. */ - assert(QEMU_IS_ALIGNED(*pnum, align) && align > offset - aligned_offset); + assert(*pnum && QEMU_IS_ALIGNED(*pnum, align) && + align > offset - aligned_offset); *pnum -= offset - aligned_offset; if (*pnum > bytes) { *pnum = bytes; -- 2.13.6 From: Anton Nefedov <anton.nefedov@virtuozzo.com> The normal bdrv_co_pwritev() use is either - BDRV_REQ_ZERO_WRITE clear and iovector provided - BDRV_REQ_ZERO_WRITE set and iovector == NULL while - the flag clear and iovector == NULL is an assertion failure in bdrv_co_do_zero_pwritev() - the flag set and iovector provided is in fact allowed (the flag prevails and zeroes are written) However the alignment logic does not support the latter case so the padding areas get overwritten with zeroes. Currently, general functions like bdrv_rw_co() do provide iovector regardless of flags. So, keep it supported and use bdrv_co_do_zero_pwritev() alignment for it which also makes the code a bit more obvious anyway. Signed-off-by: Anton Nefedov <anton.nefedov@virtuozzo.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Alberto Garcia <berto@igalia.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/io.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child, */ tracked_request_begin(&req, bs, offset, bytes, BDRV_TRACKED_WRITE); - if (!qiov) { + if (flags & BDRV_REQ_ZERO_WRITE) { ret = bdrv_co_do_zero_pwritev(child, offset, bytes, flags, &req); goto out; } -- 2.13.6 From: Anton Nefedov <anton.nefedov@virtuozzo.com> This new test case only makes sense for qcow2 while iotest 033 is generic; however it matches the test purpose perfectly and also 033 contains those do_test() tricks to pass the alignment, which won't look nice being duplicated in other tests or moved to the common code. Signed-off-by: Anton Nefedov <anton.nefedov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/qemu-iotests/033 | 29 +++++++++++++++++++++++++++++ tests/qemu-iotests/033.out | 13 +++++++++++++ 2 files changed, 42 insertions(+) diff --git a/tests/qemu-iotests/033 b/tests/qemu-iotests/033 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/033 +++ b/tests/qemu-iotests/033 @@ -XXX,XX +XXX,XX @@ do_test() } | $QEMU_IO $IO_EXTRA_ARGS } +echo +echo "=== Test aligned and misaligned write zeroes operations ===" + for write_zero_cmd in "write -z" "aio_write -z"; do for align in 512 4k; do echo @@ -XXX,XX +XXX,XX @@ for align in 512 4k; do done done + +# Trigger truncate that would shrink qcow2 L1 table, which is done by +# clearing one entry (8 bytes) with bdrv_co_pwrite_zeroes() + +echo +echo "=== Test misaligned write zeroes via truncate ===" +echo + +# any size will do, but the smaller the size the smaller the required image +CLUSTER_SIZE=$((4 * 1024)) +L2_COVERAGE=$(($CLUSTER_SIZE * $CLUSTER_SIZE / 8)) +_make_test_img $(($L2_COVERAGE * 2)) + +do_test 512 "write -P 1 0 0x200" "$TEST_IMG" | _filter_qemu_io +# next L2 table +do_test 512 "write -P 1 $L2_COVERAGE 0x200" "$TEST_IMG" | _filter_qemu_io + +# only interested in qcow2 here; also other formats might respond with +# "not supported" error message +if [ $IMGFMT = "qcow2" ]; then + do_test 512 "truncate $L2_COVERAGE" "$TEST_IMG" | _filter_qemu_io +fi + +do_test 512 "read -P 1 0 0x200" "$TEST_IMG" | _filter_qemu_io + # success, all done +echo echo "*** done" rm -f $seq.full status=0 diff --git a/tests/qemu-iotests/033.out b/tests/qemu-iotests/033.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/033.out +++ b/tests/qemu-iotests/033.out @@ -XXX,XX +XXX,XX @@ QA output created by 033 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 +=== Test aligned and misaligned write zeroes operations === + == preparing image == wrote 1024/1024 bytes at offset 512 1 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) @@ -XXX,XX +XXX,XX @@ read 512/512 bytes at offset 512 read 3072/3072 bytes at offset 1024 3 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) + +=== Test misaligned write zeroes via truncate === + +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=4194304 +wrote 512/512 bytes at offset 0 +512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +wrote 512/512 bytes at offset 2097152 +512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 512/512 bytes at offset 0 +512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) + *** done -- 2.13.6 From: Alberto Garcia <berto@igalia.com> This patch fixes several mistakes in the documentation of the compressed cluster descriptor: 1) the documentation claims that the cluster descriptor contains the number of sectors used to store the compressed data, but what it actually contains is the number of sectors *minus one* or, in other words, the number of additional sectors after the first one. 2) the width of the fields is incorrectly specified. The number of bits used by each field is x = 62 - (cluster_bits - 8) for the offset field y = (cluster_bits - 8) for the size field So the offset field's location is [0, x-1], not [0, x] as stated. 3) the size field does not contain the size of the compressed data, but rather the number of sectors where that data is stored. The compressed data starts at the exact point specified in the offset field and ends when there's enough data to produce a cluster of decompressed data. Both points can be in the middle of a sector, allowing several compressed clusters to be stored next to one another, sharing sectors if necessary. Cc: qemu-stable@nongnu.org Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- docs/interop/qcow2.txt | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-) diff --git a/docs/interop/qcow2.txt b/docs/interop/qcow2.txt index XXXXXXX..XXXXXXX 100644 --- a/docs/interop/qcow2.txt +++ b/docs/interop/qcow2.txt @@ -XXX,XX +XXX,XX @@ Standard Cluster Descriptor: Compressed Clusters Descriptor (x = 62 - (cluster_bits - 8)): - Bit 0 - x: Host cluster offset. This is usually _not_ aligned to a - cluster boundary! + Bit 0 - x-1: Host cluster offset. This is usually _not_ aligned to a + cluster or sector boundary! - x+1 - 61: Compressed size of the images in sectors of 512 bytes + x - 61: Number of additional 512-byte sectors used for the + compressed data, beyond the sector containing the offset + in the previous field. Some of these sectors may reside + in the next contiguous host cluster. + + Note that the compressed data does not necessarily occupy + all of the bytes in the final sector; rather, decompression + stops when it has produced a cluster of data. + + Another compressed cluster may map to the tail of the final + sector used by this compressed cluster. If a cluster is unallocated, read requests shall read the data from the backing file (except if bit 0 in the Standard Cluster Descriptor is set). If there is -- 2.13.6 From: Alberto Garcia <berto@igalia.com> This patch updates docs/qcow2-cache.txt explaining how to use the new l2-cache-entry-size parameter. Here's a more detailed technical description of this feature: https://lists.gnu.org/archive/html/qemu-block/2017-09/msg00635.html And here are some performance numbers: https://lists.gnu.org/archive/html/qemu-block/2017-12/msg00507.html Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- docs/qcow2-cache.txt | 46 +++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 43 insertions(+), 3 deletions(-) diff --git a/docs/qcow2-cache.txt b/docs/qcow2-cache.txt index XXXXXXX..XXXXXXX 100644 --- a/docs/qcow2-cache.txt +++ b/docs/qcow2-cache.txt @@ -XXX,XX +XXX,XX @@ qcow2 L2/refcount cache configuration ===================================== -Copyright (C) 2015 Igalia, S.L. +Copyright (C) 2015, 2018 Igalia, S.L. Author: Alberto Garcia <berto@igalia.com> This work is licensed under the terms of the GNU GPL, version 2 or @@ -XXX,XX +XXX,XX @@ There are three options available, and all of them take bytes: There are two things that need to be taken into account: - - Both caches must have a size that is a multiple of the cluster - size. + - Both caches must have a size that is a multiple of the cluster size + (or the cache entry size: see "Using smaller cache sizes" below). - If you only set one of the options above, QEMU will automatically adjust the others so that the L2 cache is 4 times bigger than the @@ -XXX,XX +XXX,XX @@ much less often than the L2 cache, so it's perfectly reasonable to keep it small. +Using smaller cache entries +--------------------------- +The qcow2 L2 cache stores complete tables by default. This means that +if QEMU needs an entry from an L2 table then the whole table is read +from disk and is kept in the cache. If the cache is full then a +complete table needs to be evicted first. + +This can be inefficient with large cluster sizes since it results in +more disk I/O and wastes more cache memory. + +Since QEMU 2.12 you can change the size of the L2 cache entry and make +it smaller than the cluster size. This can be configured using the +"l2-cache-entry-size" parameter: + + -drive file=hd.qcow2,l2-cache-size=2097152,l2-cache-entry-size=4096 + +Some things to take into account: + + - The L2 cache entry size has the same restrictions as the cluster + size (power of two, at least 512 bytes). + + - Smaller entry sizes generally improve the cache efficiency and make + disk I/O faster. This is particularly true with solid state drives + so it's a good idea to reduce the entry size in those cases. With + rotating hard drives the situation is a bit more complicated so you + should test it first and stay with the default size if unsure. + + - Try different entry sizes to see which one gives faster performance + in your case. The block size of the host filesystem is generally a + good default (usually 4096 bytes in the case of ext4). + + - Only the L2 cache can be configured this way. The refcount cache + always uses the cluster size as the entry size. + + - If the L2 cache is big enough to hold all of the image's L2 tables + (as explained in the "Choosing the right cache sizes" section + earlier in this document) then none of this is necessary and you + can omit the "l2-cache-entry-size" parameter altogether. + + Reducing the memory usage ------------------------- It is possible to clean unused cache entries in order to reduce the -- 2.13.6 From: Stefan Hajnoczi <stefanha@redhat.com> The name aio_context_in_iothread() is misleading because it also returns true when called on the main AioContext from the main loop thread, which is not an IOThread. This patch renames it to in_aio_context_home_thread() and expands the doc comment to make the semantics clearer. Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/aio.h | 7 +++++-- include/block/block.h | 2 +- 2 files changed, 6 insertions(+), 3 deletions(-) diff --git a/include/block/aio.h b/include/block/aio.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/aio.h +++ b/include/block/aio.h @@ -XXX,XX +XXX,XX @@ void aio_co_enter(AioContext *ctx, struct Coroutine *co); AioContext *qemu_get_current_aio_context(void); /** + * in_aio_context_home_thread: * @ctx: the aio context * - * Return whether we are running in the I/O thread that manages @ctx. + * Return whether we are running in the thread that normally runs @ctx. Note + * that acquiring/releasing ctx does not affect the outcome, each AioContext + * still only has one home thread that is responsible for running it. */ -static inline bool aio_context_in_iothread(AioContext *ctx) +static inline bool in_aio_context_home_thread(AioContext *ctx) { return ctx == qemu_get_current_aio_context(); } diff --git a/include/block/block.h b/include/block/block.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block.h +++ b/include/block/block.h @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all(void); bool busy_ = true; \ BlockDriverState *bs_ = (bs); \ AioContext *ctx_ = bdrv_get_aio_context(bs_); \ - if (aio_context_in_iothread(ctx_)) { \ + if (in_aio_context_home_thread(ctx_)) { \ while ((cond) || busy_) { \ busy_ = aio_poll(ctx_, (cond)); \ waited_ |= !!(cond) | busy_; \ -- 2.13.6 From: Stefan Hajnoczi <stefanha@redhat.com> BlockDriverState has the BDRV_POLL_WHILE() macro to wait on event loop activity while a condition evaluates to true. This is used to implement synchronous operations where it acts as a condvar between the IOThread running the operation and the main loop waiting for the operation. It can also be called from the thread that owns the AioContext and in that case it's just a nested event loop. BlockBackend needs this behavior but doesn't always have a BlockDriverState it can use. This patch extracts BDRV_POLL_WHILE() into the AioWait abstraction, which can be used with AioContext and isn't tied to BlockDriverState anymore. This feature could be built directly into AioContext but then all users would kick the event loop even if they signal different conditions. Imagine an AioContext with many BlockDriverStates, each time a request completes any waiter would wake up and re-check their condition. It's nicer to keep a separate AioWait object for each condition instead. Please see "block/aio-wait.h" for details on the API. The name AIO_WAIT_WHILE() avoids the confusion between AIO_POLL_WHILE() and AioContext polling. Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/aio-wait.h | 116 ++++++++++++++++++++++++++++++++++++++++++++++ include/block/block.h | 40 +++------------- include/block/block_int.h | 7 ++- block.c | 5 ++ block/io.c | 10 +--- util/aio-wait.c | 40 ++++++++++++++++ util/Makefile.objs | 2 +- 7 files changed, 174 insertions(+), 46 deletions(-) create mode 100644 include/block/aio-wait.h create mode 100644 util/aio-wait.c diff --git a/include/block/aio-wait.h b/include/block/aio-wait.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/block/aio-wait.h @@ -XXX,XX +XXX,XX @@ +/* + * AioContext wait support + * + * Copyright (C) 2018 Red Hat, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#ifndef QEMU_AIO_WAIT_H +#define QEMU_AIO_WAIT_H + +#include "block/aio.h" + +/** + * AioWait: + * + * An object that facilitates synchronous waiting on a condition. The main + * loop can wait on an operation running in an IOThread as follows: + * + * AioWait *wait = ...; + * AioContext *ctx = ...; + * MyWork work = { .done = false }; + * schedule_my_work_in_iothread(ctx, &work); + * AIO_WAIT_WHILE(wait, ctx, !work.done); + * + * The IOThread must call aio_wait_kick() to notify the main loop when + * work.done changes: + * + * static void do_work(...) + * { + * ... + * work.done = true; + * aio_wait_kick(wait); + * } + */ +typedef struct { + /* Is the main loop waiting for a kick? Accessed with atomic ops. */ + bool need_kick; +} AioWait; + +/** + * AIO_WAIT_WHILE: + * @wait: the aio wait object + * @ctx: the aio context + * @cond: wait while this conditional expression is true + * + * Wait while a condition is true. Use this to implement synchronous + * operations that require event loop activity. + * + * The caller must be sure that something calls aio_wait_kick() when the value + * of @cond might have changed. + * + * The caller's thread must be the IOThread that owns @ctx or the main loop + * thread (with @ctx acquired exactly once). This function cannot be used to + * wait on conditions between two IOThreads since that could lead to deadlock, + * go via the main loop instead. + */ +#define AIO_WAIT_WHILE(wait, ctx, cond) ({ \ + bool waited_ = false; \ + bool busy_ = true; \ + AioWait *wait_ = (wait); \ + AioContext *ctx_ = (ctx); \ + if (in_aio_context_home_thread(ctx_)) { \ + while ((cond) || busy_) { \ + busy_ = aio_poll(ctx_, (cond)); \ + waited_ |= !!(cond) | busy_; \ + } \ + } else { \ + assert(qemu_get_current_aio_context() == \ + qemu_get_aio_context()); \ + assert(!wait_->need_kick); \ + /* Set wait_->need_kick before evaluating cond. */ \ + atomic_mb_set(&wait_->need_kick, true); \ + while (busy_) { \ + if ((cond)) { \ + waited_ = busy_ = true; \ + aio_context_release(ctx_); \ + aio_poll(qemu_get_aio_context(), true); \ + aio_context_acquire(ctx_); \ + } else { \ + busy_ = aio_poll(ctx_, false); \ + waited_ |= busy_; \ + } \ + } \ + atomic_set(&wait_->need_kick, false); \ + } \ + waited_; }) + +/** + * aio_wait_kick: + * @wait: the aio wait object that should re-evaluate its condition + * + * Wake up the main thread if it is waiting on AIO_WAIT_WHILE(). During + * synchronous operations performed in an IOThread, the main thread lets the + * IOThread's event loop run, waiting for the operation to complete. A + * aio_wait_kick() call will wake up the main thread. + */ +void aio_wait_kick(AioWait *wait); + +#endif /* QEMU_AIO_WAIT */ diff --git a/include/block/block.h b/include/block/block.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block.h +++ b/include/block/block.h @@ -XXX,XX +XXX,XX @@ #define BLOCK_H #include "block/aio.h" +#include "block/aio-wait.h" #include "qapi-types.h" #include "qemu/iov.h" #include "qemu/coroutine.h" @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void); void bdrv_drain_all_end(void); void bdrv_drain_all(void); +/* Returns NULL when bs == NULL */ +AioWait *bdrv_get_aio_wait(BlockDriverState *bs); + #define BDRV_POLL_WHILE(bs, cond) ({ \ - bool waited_ = false; \ - bool busy_ = true; \ BlockDriverState *bs_ = (bs); \ - AioContext *ctx_ = bdrv_get_aio_context(bs_); \ - if (in_aio_context_home_thread(ctx_)) { \ - while ((cond) || busy_) { \ - busy_ = aio_poll(ctx_, (cond)); \ - waited_ |= !!(cond) | busy_; \ - } \ - } else { \ - assert(qemu_get_current_aio_context() == \ - qemu_get_aio_context()); \ - /* Ask bdrv_dec_in_flight to wake up the main \ - * QEMU AioContext. Extra I/O threads never take \ - * other I/O threads' AioContexts (see for example \ - * block_job_defer_to_main_loop for how to do it). \ - */ \ - assert(!bs_->wakeup); \ - /* Set bs->wakeup before evaluating cond. */ \ - atomic_mb_set(&bs_->wakeup, true); \ - while (busy_) { \ - if ((cond)) { \ - waited_ = busy_ = true; \ - aio_context_release(ctx_); \ - aio_poll(qemu_get_aio_context(), true); \ - aio_context_acquire(ctx_); \ - } else { \ - busy_ = aio_poll(ctx_, false); \ - waited_ |= busy_; \ - } \ - } \ - atomic_set(&bs_->wakeup, false); \ - } \ - waited_; }) + AIO_WAIT_WHILE(bdrv_get_aio_wait(bs_), \ + bdrv_get_aio_context(bs_), \ + cond); }) int bdrv_pdiscard(BlockDriverState *bs, int64_t offset, int bytes); int bdrv_co_pdiscard(BlockDriverState *bs, int64_t offset, int bytes); diff --git a/include/block/block_int.h b/include/block/block_int.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block_int.h +++ b/include/block/block_int.h @@ -XXX,XX +XXX,XX @@ #include "block/accounting.h" #include "block/block.h" +#include "block/aio-wait.h" #include "qemu/queue.h" #include "qemu/coroutine.h" #include "qemu/stats64.h" @@ -XXX,XX +XXX,XX @@ struct BlockDriverState { unsigned int in_flight; unsigned int serialising_in_flight; - /* Internal to BDRV_POLL_WHILE and bdrv_wakeup. Accessed with atomic - * ops. - */ - bool wakeup; + /* Kicked to signal main loop when a request completes. */ + AioWait wait; /* counter for nested bdrv_io_plug. * Accessed with atomic ops. diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ AioContext *bdrv_get_aio_context(BlockDriverState *bs) return bs->aio_context; } +AioWait *bdrv_get_aio_wait(BlockDriverState *bs) +{ + return bs ? &bs->wait : NULL; +} + void bdrv_coroutine_enter(BlockDriverState *bs, Coroutine *co) { aio_co_enter(bdrv_get_aio_context(bs), co); diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "trace.h" #include "sysemu/block-backend.h" +#include "block/aio-wait.h" #include "block/blockjob.h" #include "block/blockjob_int.h" #include "block/block_int.h" @@ -XXX,XX +XXX,XX @@ void bdrv_inc_in_flight(BlockDriverState *bs) atomic_inc(&bs->in_flight); } -static void dummy_bh_cb(void *opaque) -{ -} - void bdrv_wakeup(BlockDriverState *bs) { - /* The barrier (or an atomic op) is in the caller. */ - if (atomic_read(&bs->wakeup)) { - aio_bh_schedule_oneshot(qemu_get_aio_context(), dummy_bh_cb, NULL); - } + aio_wait_kick(bdrv_get_aio_wait(bs)); } void bdrv_dec_in_flight(BlockDriverState *bs) diff --git a/util/aio-wait.c b/util/aio-wait.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/util/aio-wait.c @@ -XXX,XX +XXX,XX @@ +/* + * AioContext wait support + * + * Copyright (C) 2018 Red Hat, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "qemu/main-loop.h" +#include "block/aio-wait.h" + +static void dummy_bh_cb(void *opaque) +{ + /* The point is to make AIO_WAIT_WHILE()'s aio_poll() return */ +} + +void aio_wait_kick(AioWait *wait) +{ + /* The barrier (or an atomic op) is in the caller. */ + if (atomic_read(&wait->need_kick)) { + aio_bh_schedule_oneshot(qemu_get_aio_context(), dummy_bh_cb, NULL); + } +} diff --git a/util/Makefile.objs b/util/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/util/Makefile.objs +++ b/util/Makefile.objs @@ -XXX,XX +XXX,XX @@ util-obj-y = osdep.o cutils.o unicode.o qemu-timer-common.o util-obj-y += bufferiszero.o util-obj-y += lockcnt.o -util-obj-y += aiocb.o async.o thread-pool.o qemu-timer.o +util-obj-y += aiocb.o async.o aio-wait.o thread-pool.o qemu-timer.o util-obj-y += main-loop.o iohandler.o util-obj-$(CONFIG_POSIX) += aio-posix.o util-obj-$(CONFIG_POSIX) += compatfd.o -- 2.13.6 From: Stefan Hajnoczi <stefanha@redhat.com> BlockBackend currently relies on BlockDriverState->in_flight to track requests for blk_drain(). There is a corner case where BlockDriverState->in_flight cannot be used though: blk->root can be NULL when there is no medium. This results in a segfault when the NULL pointer is dereferenced. Introduce a BlockBackend->in_flight counter for aio requests so it works even when blk->root == NULL. Based on a patch by Kevin Wolf <kwolf@redhat.com>. Signed-off-by: Kevin Wolf <kwolf@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 2 +- block/block-backend.c | 60 +++++++++++++++++++++++++++++++++++++++++++++------ 2 files changed, 54 insertions(+), 8 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ out: AioContext *bdrv_get_aio_context(BlockDriverState *bs) { - return bs->aio_context; + return bs ? bs->aio_context : qemu_get_aio_context(); } AioWait *bdrv_get_aio_wait(BlockDriverState *bs) diff --git a/block/block-backend.c b/block/block-backend.c index XXXXXXX..XXXXXXX 100644 --- a/block/block-backend.c +++ b/block/block-backend.c @@ -XXX,XX +XXX,XX @@ struct BlockBackend { int quiesce_counter; VMChangeStateEntry *vmsh; bool force_allow_inactivate; + + /* Number of in-flight aio requests. BlockDriverState also counts + * in-flight requests but aio requests can exist even when blk->root is + * NULL, so we cannot rely on its counter for that case. + * Accessed with atomic ops. + */ + unsigned int in_flight; + AioWait wait; }; typedef struct BlockBackendAIOCB { @@ -XXX,XX +XXX,XX @@ int blk_make_zero(BlockBackend *blk, BdrvRequestFlags flags) return bdrv_make_zero(blk->root, flags); } +static void blk_inc_in_flight(BlockBackend *blk) +{ + atomic_inc(&blk->in_flight); +} + +static void blk_dec_in_flight(BlockBackend *blk) +{ + atomic_dec(&blk->in_flight); + aio_wait_kick(&blk->wait); +} + static void error_callback_bh(void *opaque) { struct BlockBackendAIOCB *acb = opaque; - bdrv_dec_in_flight(acb->common.bs); + blk_dec_in_flight(acb->blk); acb->common.cb(acb->common.opaque, acb->ret); qemu_aio_unref(acb); } @@ -XXX,XX +XXX,XX @@ BlockAIOCB *blk_abort_aio_request(BlockBackend *blk, { struct BlockBackendAIOCB *acb; - bdrv_inc_in_flight(blk_bs(blk)); + blk_inc_in_flight(blk); acb = blk_aio_get(&block_backend_aiocb_info, blk, cb, opaque); acb->blk = blk; acb->ret = ret; @@ -XXX,XX +XXX,XX @@ static const AIOCBInfo blk_aio_em_aiocb_info = { static void blk_aio_complete(BlkAioEmAIOCB *acb) { if (acb->has_returned) { - bdrv_dec_in_flight(acb->common.bs); + blk_dec_in_flight(acb->rwco.blk); acb->common.cb(acb->common.opaque, acb->rwco.ret); qemu_aio_unref(acb); } @@ -XXX,XX +XXX,XX @@ static BlockAIOCB *blk_aio_prwv(BlockBackend *blk, int64_t offset, int bytes, BlkAioEmAIOCB *acb; Coroutine *co; - bdrv_inc_in_flight(blk_bs(blk)); + blk_inc_in_flight(blk); acb = blk_aio_get(&blk_aio_em_aiocb_info, blk, cb, opaque); acb->rwco = (BlkRwCo) { .blk = blk, @@ -XXX,XX +XXX,XX @@ int blk_flush(BlockBackend *blk) void blk_drain(BlockBackend *blk) { - if (blk_bs(blk)) { - bdrv_drain(blk_bs(blk)); + BlockDriverState *bs = blk_bs(blk); + + if (bs) { + bdrv_drained_begin(bs); + } + + /* We may have -ENOMEDIUM completions in flight */ + AIO_WAIT_WHILE(&blk->wait, + blk_get_aio_context(blk), + atomic_mb_read(&blk->in_flight) > 0); + + if (bs) { + bdrv_drained_end(bs); } } void blk_drain_all(void) { - bdrv_drain_all(); + BlockBackend *blk = NULL; + + bdrv_drain_all_begin(); + + while ((blk = blk_all_next(blk)) != NULL) { + AioContext *ctx = blk_get_aio_context(blk); + + aio_context_acquire(ctx); + + /* We may have -ENOMEDIUM completions in flight */ + AIO_WAIT_WHILE(&blk->wait, ctx, + atomic_mb_read(&blk->in_flight) > 0); + + aio_context_release(ctx); + } + + bdrv_drain_all_end(); } void blk_set_on_error(BlockBackend *blk, BlockdevOnError on_read_error, -- 2.13.6 This patch adds test cases for the scenario where blk_aio_flush() is called on a BlockBackend with no root. Calling drain afterwards should complete the requests with -ENOMEDIUM. Signed-off-by: Kevin Wolf <kwolf@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/test-block-backend.c | 82 ++++++++++++++++++++++++++++++++++++++++++++++ tests/Makefile.include | 2 ++ 2 files changed, 84 insertions(+) create mode 100644 tests/test-block-backend.c diff --git a/tests/test-block-backend.c b/tests/test-block-backend.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/test-block-backend.c @@ -XXX,XX +XXX,XX @@ +/* + * BlockBackend tests + * + * Copyright (c) 2017 Kevin Wolf <kwolf@redhat.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "block/block.h" +#include "sysemu/block-backend.h" +#include "qapi/error.h" + +static void test_drain_aio_error_flush_cb(void *opaque, int ret) +{ + bool *completed = opaque; + + g_assert(ret == -ENOMEDIUM); + *completed = true; +} + +static void test_drain_aio_error(void) +{ + BlockBackend *blk = blk_new(BLK_PERM_ALL, BLK_PERM_ALL); + BlockAIOCB *acb; + bool completed = false; + + acb = blk_aio_flush(blk, test_drain_aio_error_flush_cb, &completed); + g_assert(acb != NULL); + g_assert(completed == false); + + blk_drain(blk); + g_assert(completed == true); + + blk_unref(blk); +} + +static void test_drain_all_aio_error(void) +{ + BlockBackend *blk = blk_new(BLK_PERM_ALL, BLK_PERM_ALL); + BlockAIOCB *acb; + bool completed = false; + + acb = blk_aio_flush(blk, test_drain_aio_error_flush_cb, &completed); + g_assert(acb != NULL); + g_assert(completed == false); + + blk_drain_all(); + g_assert(completed == true); + + blk_unref(blk); +} + +int main(int argc, char **argv) +{ + bdrv_init(); + qemu_init_main_loop(&error_abort); + + g_test_init(&argc, &argv, NULL); + + g_test_add_func("/block-backend/drain_aio_error", test_drain_aio_error); + g_test_add_func("/block-backend/drain_all_aio_error", + test_drain_all_aio_error); + + return g_test_run(); +} diff --git a/tests/Makefile.include b/tests/Makefile.include index XXXXXXX..XXXXXXX 100644 --- a/tests/Makefile.include +++ b/tests/Makefile.include @@ -XXX,XX +XXX,XX @@ gcov-files-test-hbitmap-y = blockjob.c check-unit-y += tests/test-bdrv-drain$(EXESUF) check-unit-y += tests/test-blockjob$(EXESUF) check-unit-y += tests/test-blockjob-txn$(EXESUF) +check-unit-y += tests/test-block-backend$(EXESUF) check-unit-y += tests/test-x86-cpuid$(EXESUF) # all code tested by test-x86-cpuid is inside topology.h gcov-files-test-x86-cpuid-y = @@ -XXX,XX +XXX,XX @@ tests/test-throttle$(EXESUF): tests/test-throttle.o $(test-block-obj-y) tests/test-bdrv-drain$(EXESUF): tests/test-bdrv-drain.o $(test-block-obj-y) $(test-util-obj-y) tests/test-blockjob$(EXESUF): tests/test-blockjob.o $(test-block-obj-y) $(test-util-obj-y) tests/test-blockjob-txn$(EXESUF): tests/test-blockjob-txn.o $(test-block-obj-y) $(test-util-obj-y) +tests/test-block-backend$(EXESUF): tests/test-block-backend.o $(test-block-obj-y) $(test-util-obj-y) tests/test-thread-pool$(EXESUF): tests/test-thread-pool.o $(test-block-obj-y) tests/test-iov$(EXESUF): tests/test-iov.o $(test-util-obj-y) tests/test-hbitmap$(EXESUF): tests/test-hbitmap.o $(test-util-obj-y) $(test-crypto-obj-y) -- 2.13.6 From: Stefan Hajnoczi <stefanha@redhat.com> This reverts commit 4da97120d51a4383aa96d741a2b837f8c4bbcd0b. blk_aio_flush() now handles the blk->root == NULL case, so we no longer need this workaround. Cc: John Snow <jsnow@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- hw/ide/core.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) diff --git a/hw/ide/core.c b/hw/ide/core.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ide/core.c +++ b/hw/ide/core.c @@ -XXX,XX +XXX,XX @@ static void ide_flush_cache(IDEState *s) s->status |= BUSY_STAT; ide_set_retry(s); block_acct_start(blk_get_stats(s->blk), &s->acct, 0, BLOCK_ACCT_FLUSH); - - if (blk_bs(s->blk)) { - s->pio_aiocb = blk_aio_flush(s->blk, ide_flush_cb, s); - } else { - /* XXX blk_aio_flush() crashes when blk_bs(blk) is NULL, remove this - * temporary workaround when blk_aio_*() functions handle NULL blk_bs. - */ - ide_flush_cb(s, 0); - } + s->pio_aiocb = blk_aio_flush(s->blk, ide_flush_cb, s); } static void ide_cfata_metadata_inquiry(IDEState *s) -- 2.13.6 From: Stefan Hajnoczi <stefanha@redhat.com> BlockDriver->bdrv_create() has been called from coroutine context since commit 5b7e1542cfa41a281af9629d31cef03704d976e6 ("block: make bdrv_create adopt coroutine"). Make this explicit by renaming to .bdrv_co_create_opts() and add the coroutine_fn annotation. This makes it obvious to block driver authors that they may yield, use CoMutex, or other coroutine_fn APIs. bdrv_co_create is reserved for the QAPI-based version that Kevin is working on. Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Message-Id: <20170705102231.20711-2-stefanha@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block_int.h | 3 ++- block.c | 4 ++-- block/crypto.c | 8 ++++---- block/file-posix.c | 15 ++++++++------- block/file-win32.c | 5 +++-- block/gluster.c | 13 +++++++------ block/iscsi.c | 7 ++++--- block/nfs.c | 5 +++-- block/parallels.c | 6 ++++-- block/qcow.c | 5 +++-- block/qcow2.c | 5 +++-- block/qed.c | 6 ++++-- block/raw-format.c | 5 +++-- block/rbd.c | 6 ++++-- block/sheepdog.c | 10 +++++----- block/ssh.c | 5 +++-- block/vdi.c | 5 +++-- block/vhdx.c | 5 +++-- block/vmdk.c | 5 +++-- block/vpc.c | 5 +++-- 20 files changed, 74 insertions(+), 54 deletions(-) diff --git a/include/block/block_int.h b/include/block/block_int.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block_int.h +++ b/include/block/block_int.h @@ -XXX,XX +XXX,XX @@ struct BlockDriver { int (*bdrv_file_open)(BlockDriverState *bs, QDict *options, int flags, Error **errp); void (*bdrv_close)(BlockDriverState *bs); - int (*bdrv_create)(const char *filename, QemuOpts *opts, Error **errp); + int coroutine_fn (*bdrv_co_create_opts)(const char *filename, QemuOpts *opts, + Error **errp); int (*bdrv_make_empty)(BlockDriverState *bs); void (*bdrv_refresh_filename)(BlockDriverState *bs, QDict *options); diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_create_co_entry(void *opaque) CreateCo *cco = opaque; assert(cco->drv); - ret = cco->drv->bdrv_create(cco->filename, cco->opts, &local_err); + ret = cco->drv->bdrv_co_create_opts(cco->filename, cco->opts, &local_err); error_propagate(&cco->err, local_err); cco->ret = ret; } @@ -XXX,XX +XXX,XX @@ int bdrv_create(BlockDriver *drv, const char* filename, .err = NULL, }; - if (!drv->bdrv_create) { + if (!drv->bdrv_co_create_opts) { error_setg(errp, "Driver '%s' does not support image creation", drv->format_name); ret = -ENOTSUP; goto out; diff --git a/block/crypto.c b/block/crypto.c index XXXXXXX..XXXXXXX 100644 --- a/block/crypto.c +++ b/block/crypto.c @@ -XXX,XX +XXX,XX @@ static int block_crypto_open_luks(BlockDriverState *bs, bs, options, flags, errp); } -static int block_crypto_create_luks(const char *filename, - QemuOpts *opts, - Error **errp) +static int coroutine_fn block_crypto_co_create_opts_luks(const char *filename, + QemuOpts *opts, + Error **errp) { return block_crypto_create_generic(Q_CRYPTO_BLOCK_FORMAT_LUKS, filename, opts, errp); @@ -XXX,XX +XXX,XX @@ BlockDriver bdrv_crypto_luks = { .bdrv_open = block_crypto_open_luks, .bdrv_close = block_crypto_close, .bdrv_child_perm = bdrv_format_default_perms, - .bdrv_create = block_crypto_create_luks, + .bdrv_co_create_opts = block_crypto_co_create_opts_luks, .bdrv_truncate = block_crypto_truncate, .create_opts = &block_crypto_create_opts_luks, diff --git a/block/file-posix.c b/block/file-posix.c index XXXXXXX..XXXXXXX 100644 --- a/block/file-posix.c +++ b/block/file-posix.c @@ -XXX,XX +XXX,XX @@ static int64_t raw_get_allocated_file_size(BlockDriverState *bs) return (int64_t)st.st_blocks * 512; } -static int raw_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn raw_co_create_opts(const char *filename, QemuOpts *opts, + Error **errp) { int fd; int result = 0; @@ -XXX,XX +XXX,XX @@ BlockDriver bdrv_file = { .bdrv_reopen_commit = raw_reopen_commit, .bdrv_reopen_abort = raw_reopen_abort, .bdrv_close = raw_close, - .bdrv_create = raw_create, + .bdrv_co_create_opts = raw_co_create_opts, .bdrv_has_zero_init = bdrv_has_zero_init_1, .bdrv_co_block_status = raw_co_block_status, .bdrv_co_pwrite_zeroes = raw_co_pwrite_zeroes, @@ -XXX,XX +XXX,XX @@ static coroutine_fn int hdev_co_pwrite_zeroes(BlockDriverState *bs, return -ENOTSUP; } -static int hdev_create(const char *filename, QemuOpts *opts, - Error **errp) +static int coroutine_fn hdev_co_create_opts(const char *filename, QemuOpts *opts, + Error **errp) { int fd; int ret = 0; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_host_device = { .bdrv_reopen_prepare = raw_reopen_prepare, .bdrv_reopen_commit = raw_reopen_commit, .bdrv_reopen_abort = raw_reopen_abort, - .bdrv_create = hdev_create, + .bdrv_co_create_opts = hdev_co_create_opts, .create_opts = &raw_create_opts, .bdrv_co_pwrite_zeroes = hdev_co_pwrite_zeroes, @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_host_cdrom = { .bdrv_reopen_prepare = raw_reopen_prepare, .bdrv_reopen_commit = raw_reopen_commit, .bdrv_reopen_abort = raw_reopen_abort, - .bdrv_create = hdev_create, + .bdrv_co_create_opts = hdev_co_create_opts, .create_opts = &raw_create_opts, @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_host_cdrom = { .bdrv_reopen_prepare = raw_reopen_prepare, .bdrv_reopen_commit = raw_reopen_commit, .bdrv_reopen_abort = raw_reopen_abort, - .bdrv_create = hdev_create, + .bdrv_co_create_opts = hdev_co_create_opts, .create_opts = &raw_create_opts, .bdrv_co_preadv = raw_co_preadv, diff --git a/block/file-win32.c b/block/file-win32.c index XXXXXXX..XXXXXXX 100644 --- a/block/file-win32.c +++ b/block/file-win32.c @@ -XXX,XX +XXX,XX @@ static int64_t raw_get_allocated_file_size(BlockDriverState *bs) return st.st_size; } -static int raw_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn raw_co_create_opts(const char *filename, QemuOpts *opts, + Error **errp) { int fd; int64_t total_size = 0; @@ -XXX,XX +XXX,XX @@ BlockDriver bdrv_file = { .bdrv_file_open = raw_open, .bdrv_refresh_limits = raw_probe_alignment, .bdrv_close = raw_close, - .bdrv_create = raw_create, + .bdrv_co_create_opts = raw_co_create_opts, .bdrv_has_zero_init = bdrv_has_zero_init_1, .bdrv_aio_readv = raw_aio_readv, diff --git a/block/gluster.c b/block/gluster.c index XXXXXXX..XXXXXXX 100644 --- a/block/gluster.c +++ b/block/gluster.c @@ -XXX,XX +XXX,XX @@ static int qemu_gluster_do_truncate(struct glfs_fd *fd, int64_t offset, return 0; } -static int qemu_gluster_create(const char *filename, - QemuOpts *opts, Error **errp) +static int coroutine_fn qemu_gluster_co_create_opts(const char *filename, + QemuOpts *opts, + Error **errp) { BlockdevOptionsGluster *gconf; struct glfs *glfs; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_gluster = { .bdrv_reopen_commit = qemu_gluster_reopen_commit, .bdrv_reopen_abort = qemu_gluster_reopen_abort, .bdrv_close = qemu_gluster_close, - .bdrv_create = qemu_gluster_create, + .bdrv_co_create_opts = qemu_gluster_co_create_opts, .bdrv_getlength = qemu_gluster_getlength, .bdrv_get_allocated_file_size = qemu_gluster_allocated_file_size, .bdrv_truncate = qemu_gluster_truncate, @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_gluster_tcp = { .bdrv_reopen_commit = qemu_gluster_reopen_commit, .bdrv_reopen_abort = qemu_gluster_reopen_abort, .bdrv_close = qemu_gluster_close, - .bdrv_create = qemu_gluster_create, + .bdrv_co_create_opts = qemu_gluster_co_create_opts, .bdrv_getlength = qemu_gluster_getlength, .bdrv_get_allocated_file_size = qemu_gluster_allocated_file_size, .bdrv_truncate = qemu_gluster_truncate, @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_gluster_unix = { .bdrv_reopen_commit = qemu_gluster_reopen_commit, .bdrv_reopen_abort = qemu_gluster_reopen_abort, .bdrv_close = qemu_gluster_close, - .bdrv_create = qemu_gluster_create, + .bdrv_co_create_opts = qemu_gluster_co_create_opts, .bdrv_getlength = qemu_gluster_getlength, .bdrv_get_allocated_file_size = qemu_gluster_allocated_file_size, .bdrv_truncate = qemu_gluster_truncate, @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_gluster_rdma = { .bdrv_reopen_commit = qemu_gluster_reopen_commit, .bdrv_reopen_abort = qemu_gluster_reopen_abort, .bdrv_close = qemu_gluster_close, - .bdrv_create = qemu_gluster_create, + .bdrv_co_create_opts = qemu_gluster_co_create_opts, .bdrv_getlength = qemu_gluster_getlength, .bdrv_get_allocated_file_size = qemu_gluster_allocated_file_size, .bdrv_truncate = qemu_gluster_truncate, diff --git a/block/iscsi.c b/block/iscsi.c index XXXXXXX..XXXXXXX 100644 --- a/block/iscsi.c +++ b/block/iscsi.c @@ -XXX,XX +XXX,XX @@ static int iscsi_truncate(BlockDriverState *bs, int64_t offset, return 0; } -static int iscsi_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn iscsi_co_create_opts(const char *filename, QemuOpts *opts, + Error **errp) { int ret = 0; int64_t total_size = 0; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_iscsi = { .bdrv_parse_filename = iscsi_parse_filename, .bdrv_file_open = iscsi_open, .bdrv_close = iscsi_close, - .bdrv_create = iscsi_create, + .bdrv_co_create_opts = iscsi_co_create_opts, .create_opts = &iscsi_create_opts, .bdrv_reopen_prepare = iscsi_reopen_prepare, .bdrv_reopen_commit = iscsi_reopen_commit, @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_iser = { .bdrv_parse_filename = iscsi_parse_filename, .bdrv_file_open = iscsi_open, .bdrv_close = iscsi_close, - .bdrv_create = iscsi_create, + .bdrv_co_create_opts = iscsi_co_create_opts, .create_opts = &iscsi_create_opts, .bdrv_reopen_prepare = iscsi_reopen_prepare, .bdrv_reopen_commit = iscsi_reopen_commit, diff --git a/block/nfs.c b/block/nfs.c index XXXXXXX..XXXXXXX 100644 --- a/block/nfs.c +++ b/block/nfs.c @@ -XXX,XX +XXX,XX @@ static QemuOptsList nfs_create_opts = { } }; -static int nfs_file_create(const char *url, QemuOpts *opts, Error **errp) +static int coroutine_fn nfs_file_co_create_opts(const char *url, QemuOpts *opts, + Error **errp) { int64_t ret, total_size; NFSClient *client = g_new0(NFSClient, 1); @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_nfs = { .bdrv_file_open = nfs_file_open, .bdrv_close = nfs_file_close, - .bdrv_create = nfs_file_create, + .bdrv_co_create_opts = nfs_file_co_create_opts, .bdrv_reopen_prepare = nfs_reopen_prepare, .bdrv_co_preadv = nfs_co_preadv, diff --git a/block/parallels.c b/block/parallels.c index XXXXXXX..XXXXXXX 100644 --- a/block/parallels.c +++ b/block/parallels.c @@ -XXX,XX +XXX,XX @@ static int parallels_check(BlockDriverState *bs, BdrvCheckResult *res, } -static int parallels_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn parallels_co_create_opts(const char *filename, + QemuOpts *opts, + Error **errp) { int64_t total_size, cl_size; uint8_t tmp[BDRV_SECTOR_SIZE]; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_parallels = { .bdrv_co_readv = parallels_co_readv, .bdrv_co_writev = parallels_co_writev, .supports_backing = true, - .bdrv_create = parallels_create, + .bdrv_co_create_opts = parallels_co_create_opts, .bdrv_check = parallels_check, .create_opts = &parallels_create_opts, }; diff --git a/block/qcow.c b/block/qcow.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow.c +++ b/block/qcow.c @@ -XXX,XX +XXX,XX @@ static void qcow_close(BlockDriverState *bs) error_free(s->migration_blocker); } -static int qcow_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn qcow_co_create_opts(const char *filename, QemuOpts *opts, + Error **errp) { int header_size, backing_filename_len, l1_size, shift, i; QCowHeader header; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_qcow = { .bdrv_close = qcow_close, .bdrv_child_perm = bdrv_format_default_perms, .bdrv_reopen_prepare = qcow_reopen_prepare, - .bdrv_create = qcow_create, + .bdrv_co_create_opts = qcow_co_create_opts, .bdrv_has_zero_init = bdrv_has_zero_init_1, .supports_backing = true, diff --git a/block/qcow2.c b/block/qcow2.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -XXX,XX +XXX,XX @@ out: return ret; } -static int qcow2_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn qcow2_co_create_opts(const char *filename, QemuOpts *opts, + Error **errp) { char *backing_file = NULL; char *backing_fmt = NULL; @@ -XXX,XX +XXX,XX @@ BlockDriver bdrv_qcow2 = { .bdrv_reopen_abort = qcow2_reopen_abort, .bdrv_join_options = qcow2_join_options, .bdrv_child_perm = bdrv_format_default_perms, - .bdrv_create = qcow2_create, + .bdrv_co_create_opts = qcow2_co_create_opts, .bdrv_has_zero_init = bdrv_has_zero_init_1, .bdrv_co_block_status = qcow2_co_block_status, diff --git a/block/qed.c b/block/qed.c index XXXXXXX..XXXXXXX 100644 --- a/block/qed.c +++ b/block/qed.c @@ -XXX,XX +XXX,XX @@ out: return ret; } -static int bdrv_qed_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn bdrv_qed_co_create_opts(const char *filename, + QemuOpts *opts, + Error **errp) { uint64_t image_size = 0; uint32_t cluster_size = QED_DEFAULT_CLUSTER_SIZE; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_qed = { .bdrv_close = bdrv_qed_close, .bdrv_reopen_prepare = bdrv_qed_reopen_prepare, .bdrv_child_perm = bdrv_format_default_perms, - .bdrv_create = bdrv_qed_create, + .bdrv_co_create_opts = bdrv_qed_co_create_opts, .bdrv_has_zero_init = bdrv_has_zero_init_1, .bdrv_co_block_status = bdrv_qed_co_block_status, .bdrv_co_readv = bdrv_qed_co_readv, diff --git a/block/raw-format.c b/block/raw-format.c index XXXXXXX..XXXXXXX 100644 --- a/block/raw-format.c +++ b/block/raw-format.c @@ -XXX,XX +XXX,XX @@ static int raw_has_zero_init(BlockDriverState *bs) return bdrv_has_zero_init(bs->file->bs); } -static int raw_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn raw_co_create_opts(const char *filename, QemuOpts *opts, + Error **errp) { return bdrv_create_file(filename, opts, errp); } @@ -XXX,XX +XXX,XX @@ BlockDriver bdrv_raw = { .bdrv_open = &raw_open, .bdrv_close = &raw_close, .bdrv_child_perm = bdrv_filter_default_perms, - .bdrv_create = &raw_create, + .bdrv_co_create_opts = &raw_co_create_opts, .bdrv_co_preadv = &raw_co_preadv, .bdrv_co_pwritev = &raw_co_pwritev, .bdrv_co_pwrite_zeroes = &raw_co_pwrite_zeroes, diff --git a/block/rbd.c b/block/rbd.c index XXXXXXX..XXXXXXX 100644 --- a/block/rbd.c +++ b/block/rbd.c @@ -XXX,XX +XXX,XX @@ static QemuOptsList runtime_opts = { }, }; -static int qemu_rbd_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn qemu_rbd_co_create_opts(const char *filename, + QemuOpts *opts, + Error **errp) { Error *local_err = NULL; int64_t bytes = 0; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_rbd = { .bdrv_file_open = qemu_rbd_open, .bdrv_close = qemu_rbd_close, .bdrv_reopen_prepare = qemu_rbd_reopen_prepare, - .bdrv_create = qemu_rbd_create, + .bdrv_co_create_opts = qemu_rbd_co_create_opts, .bdrv_has_zero_init = bdrv_has_zero_init_1, .bdrv_get_info = qemu_rbd_getinfo, .create_opts = &qemu_rbd_create_opts, diff --git a/block/sheepdog.c b/block/sheepdog.c index XXXXXXX..XXXXXXX 100644 --- a/block/sheepdog.c +++ b/block/sheepdog.c @@ -XXX,XX +XXX,XX @@ static int parse_block_size_shift(BDRVSheepdogState *s, QemuOpts *opt) return 0; } -static int sd_create(const char *filename, QemuOpts *opts, - Error **errp) +static int coroutine_fn sd_co_create_opts(const char *filename, QemuOpts *opts, + Error **errp) { Error *err = NULL; int ret = 0; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_sheepdog = { .bdrv_reopen_commit = sd_reopen_commit, .bdrv_reopen_abort = sd_reopen_abort, .bdrv_close = sd_close, - .bdrv_create = sd_create, + .bdrv_co_create_opts = sd_co_create_opts, .bdrv_has_zero_init = bdrv_has_zero_init_1, .bdrv_getlength = sd_getlength, .bdrv_get_allocated_file_size = sd_get_allocated_file_size, @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_sheepdog_tcp = { .bdrv_reopen_commit = sd_reopen_commit, .bdrv_reopen_abort = sd_reopen_abort, .bdrv_close = sd_close, - .bdrv_create = sd_create, + .bdrv_co_create_opts = sd_co_create_opts, .bdrv_has_zero_init = bdrv_has_zero_init_1, .bdrv_getlength = sd_getlength, .bdrv_get_allocated_file_size = sd_get_allocated_file_size, @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_sheepdog_unix = { .bdrv_reopen_commit = sd_reopen_commit, .bdrv_reopen_abort = sd_reopen_abort, .bdrv_close = sd_close, - .bdrv_create = sd_create, + .bdrv_co_create_opts = sd_co_create_opts, .bdrv_has_zero_init = bdrv_has_zero_init_1, .bdrv_getlength = sd_getlength, .bdrv_get_allocated_file_size = sd_get_allocated_file_size, diff --git a/block/ssh.c b/block/ssh.c index XXXXXXX..XXXXXXX 100644 --- a/block/ssh.c +++ b/block/ssh.c @@ -XXX,XX +XXX,XX @@ static QemuOptsList ssh_create_opts = { } }; -static int ssh_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn ssh_co_create_opts(const char *filename, QemuOpts *opts, + Error **errp) { int r, ret; int64_t total_size = 0; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_ssh = { .instance_size = sizeof(BDRVSSHState), .bdrv_parse_filename = ssh_parse_filename, .bdrv_file_open = ssh_file_open, - .bdrv_create = ssh_create, + .bdrv_co_create_opts = ssh_co_create_opts, .bdrv_close = ssh_close, .bdrv_has_zero_init = ssh_has_zero_init, .bdrv_co_readv = ssh_co_readv, diff --git a/block/vdi.c b/block/vdi.c index XXXXXXX..XXXXXXX 100644 --- a/block/vdi.c +++ b/block/vdi.c @@ -XXX,XX +XXX,XX @@ nonallocating_write: return ret; } -static int vdi_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn vdi_co_create_opts(const char *filename, QemuOpts *opts, + Error **errp) { int ret = 0; uint64_t bytes = 0; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_vdi = { .bdrv_close = vdi_close, .bdrv_reopen_prepare = vdi_reopen_prepare, .bdrv_child_perm = bdrv_format_default_perms, - .bdrv_create = vdi_create, + .bdrv_co_create_opts = vdi_co_create_opts, .bdrv_has_zero_init = bdrv_has_zero_init_1, .bdrv_co_block_status = vdi_co_block_status, .bdrv_make_empty = vdi_make_empty, diff --git a/block/vhdx.c b/block/vhdx.c index XXXXXXX..XXXXXXX 100644 --- a/block/vhdx.c +++ b/block/vhdx.c @@ -XXX,XX +XXX,XX @@ exit: * .---- ~ ----------- ~ ------------ ~ ---------------- ~ -----------. * 1MB */ -static int vhdx_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn vhdx_co_create_opts(const char *filename, QemuOpts *opts, + Error **errp) { int ret = 0; uint64_t image_size = (uint64_t) 2 * GiB; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_vhdx = { .bdrv_child_perm = bdrv_format_default_perms, .bdrv_co_readv = vhdx_co_readv, .bdrv_co_writev = vhdx_co_writev, - .bdrv_create = vhdx_create, + .bdrv_co_create_opts = vhdx_co_create_opts, .bdrv_get_info = vhdx_get_info, .bdrv_check = vhdx_check, .bdrv_has_zero_init = bdrv_has_zero_init_1, diff --git a/block/vmdk.c b/block/vmdk.c index XXXXXXX..XXXXXXX 100644 --- a/block/vmdk.c +++ b/block/vmdk.c @@ -XXX,XX +XXX,XX @@ static int filename_decompose(const char *filename, char *path, char *prefix, return VMDK_OK; } -static int vmdk_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn vmdk_co_create_opts(const char *filename, QemuOpts *opts, + Error **errp) { int idx = 0; BlockBackend *new_blk = NULL; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_vmdk = { .bdrv_co_pwritev_compressed = vmdk_co_pwritev_compressed, .bdrv_co_pwrite_zeroes = vmdk_co_pwrite_zeroes, .bdrv_close = vmdk_close, - .bdrv_create = vmdk_create, + .bdrv_co_create_opts = vmdk_co_create_opts, .bdrv_co_flush_to_disk = vmdk_co_flush, .bdrv_co_block_status = vmdk_co_block_status, .bdrv_get_allocated_file_size = vmdk_get_allocated_file_size, diff --git a/block/vpc.c b/block/vpc.c index XXXXXXX..XXXXXXX 100644 --- a/block/vpc.c +++ b/block/vpc.c @@ -XXX,XX +XXX,XX @@ static int create_fixed_disk(BlockBackend *blk, uint8_t *buf, return ret; } -static int vpc_create(const char *filename, QemuOpts *opts, Error **errp) +static int coroutine_fn vpc_co_create_opts(const char *filename, QemuOpts *opts, + Error **errp) { uint8_t buf[1024]; VHDFooter *footer = (VHDFooter *) buf; @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_vpc = { .bdrv_close = vpc_close, .bdrv_reopen_prepare = vpc_reopen_prepare, .bdrv_child_perm = bdrv_format_default_perms, - .bdrv_create = vpc_create, + .bdrv_co_create_opts = vpc_co_create_opts, .bdrv_co_preadv = vpc_co_preadv, .bdrv_co_pwritev = vpc_co_pwritev, -- 2.13.6 From: Stefan Hajnoczi <stefanha@redhat.com> qcow2_create2() calls qemu_co_mutex_lock(). Only a coroutine_fn may call another coroutine_fn. In fact, qcow2_create2 is always called from coroutine context. Rename the function to add the "co" moniker and add coroutine_fn. Reported-by: Marc-André Lureau <marcandre.lureau@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Message-Id: <20170705102231.20711-3-stefanha@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/qcow2.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) diff --git a/block/qcow2.c b/block/qcow2.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -XXX,XX +XXX,XX @@ static uint64_t qcow2_opt_get_refcount_bits_del(QemuOpts *opts, int version, return refcount_bits; } -static int qcow2_create2(const char *filename, int64_t total_size, - const char *backing_file, const char *backing_format, - int flags, size_t cluster_size, PreallocMode prealloc, - QemuOpts *opts, int version, int refcount_order, - const char *encryptfmt, Error **errp) +static int coroutine_fn +qcow2_co_create2(const char *filename, int64_t total_size, + const char *backing_file, const char *backing_format, + int flags, size_t cluster_size, PreallocMode prealloc, + QemuOpts *opts, int version, int refcount_order, + const char *encryptfmt, Error **errp) { QDict *options; @@ -XXX,XX +XXX,XX @@ static int coroutine_fn qcow2_co_create_opts(const char *filename, QemuOpts *opt refcount_order = ctz32(refcount_bits); - ret = qcow2_create2(filename, size, backing_file, backing_fmt, flags, - cluster_size, prealloc, opts, version, refcount_order, - encryptfmt, &local_err); + ret = qcow2_co_create2(filename, size, backing_file, backing_fmt, flags, + cluster_size, prealloc, opts, version, refcount_order, + encryptfmt, &local_err); error_propagate(errp, local_err); finish: -- 2.13.6 From: Max Reitz <mreitz@redhat.com> The issue: $ qemu-img resize -f qcow2 foo.qcow2 qemu-img: Expecting one image file name Try 'qemu-img --help' for more information So we gave an image file name, but we omitted the length. qemu-img thinks the last argument is always the size and removes it immediately from argv (by decrementing argc), and tries to verify that it is a valid size only at a later point. So we do not actually know whether that last argument we called "size" is indeed a size or whether the user instead forgot to specify that size but did give a file name. Therefore, the error message should be more general. Bug: https://bugzilla.redhat.com/show_bug.cgi?id=1523458 Signed-off-by: Max Reitz <mreitz@redhat.com> Message-id: 20180205162745.23650-1-mreitz@redhat.com Reviewed-by: John Snow <jsnow@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- qemu-img.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/qemu-img.c b/qemu-img.c index XXXXXXX..XXXXXXX 100644 --- a/qemu-img.c +++ b/qemu-img.c @@ -XXX,XX +XXX,XX @@ static int img_resize(int argc, char **argv) } } if (optind != argc - 1) { - error_exit("Expecting one image file name"); + error_exit("Expecting image file name and size"); } filename = argv[optind++]; -- 2.13.6 From: Max Reitz <mreitz@redhat.com> If we ever want to offer even rudimentary truncation functionality for ssh, we should put the respective code into a reusable function. Signed-off-by: Max Reitz <mreitz@redhat.com> Message-id: 20180214204915.7980-2-mreitz@redhat.com Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Richard W.M. Jones <rjones@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- block/ssh.c | 30 ++++++++++++++++++++++-------- 1 file changed, 22 insertions(+), 8 deletions(-) diff --git a/block/ssh.c b/block/ssh.c index XXXXXXX..XXXXXXX 100644 --- a/block/ssh.c +++ b/block/ssh.c @@ -XXX,XX +XXX,XX @@ static int ssh_file_open(BlockDriverState *bs, QDict *options, int bdrv_flags, return ret; } +static int ssh_grow_file(BDRVSSHState *s, int64_t offset, Error **errp) +{ + ssize_t ret; + char c[1] = { '\0' }; + + /* offset must be strictly greater than the current size so we do + * not overwrite anything */ + assert(offset > 0 && offset > s->attrs.filesize); + + libssh2_sftp_seek64(s->sftp_handle, offset - 1); + ret = libssh2_sftp_write(s->sftp_handle, c, 1); + if (ret < 0) { + sftp_error_setg(errp, s, "Failed to grow file"); + return -EIO; + } + + s->attrs.filesize = offset; + return 0; +} + static QemuOptsList ssh_create_opts = { .name = "ssh-create-opts", .head = QTAILQ_HEAD_INITIALIZER(ssh_create_opts.head), @@ -XXX,XX +XXX,XX @@ static int coroutine_fn ssh_co_create_opts(const char *filename, QemuOpts *opts, int64_t total_size = 0; QDict *uri_options = NULL; BDRVSSHState s; - ssize_t r2; - char c[1] = { '\0' }; ssh_state_init(&s); @@ -XXX,XX +XXX,XX @@ static int coroutine_fn ssh_co_create_opts(const char *filename, QemuOpts *opts, } if (total_size > 0) { - libssh2_sftp_seek64(s.sftp_handle, total_size-1); - r2 = libssh2_sftp_write(s.sftp_handle, c, 1); - if (r2 < 0) { - sftp_error_setg(errp, &s, "truncate failed"); - ret = -EINVAL; + ret = ssh_grow_file(&s, total_size, errp); + if (ret < 0) { goto out; } - s.attrs.filesize = total_size; } ret = 0; -- 2.13.6 From: Max Reitz <mreitz@redhat.com> At runtime (that is, during a future ssh_truncate()), the SSH session is non-blocking. However, ssh_truncate() (or rather, bdrv_truncate() in general) is not a coroutine, so this resize operation needs to block. For ssh_create(), that is fine, too; the session is never set to non-blocking anyway. Signed-off-by: Max Reitz <mreitz@redhat.com> Message-id: 20180214204915.7980-3-mreitz@redhat.com Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Richard W.M. Jones <rjones@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- block/ssh.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/block/ssh.c b/block/ssh.c index XXXXXXX..XXXXXXX 100644 --- a/block/ssh.c +++ b/block/ssh.c @@ -XXX,XX +XXX,XX @@ static int ssh_file_open(BlockDriverState *bs, QDict *options, int bdrv_flags, return ret; } +/* Note: This is a blocking operation */ static int ssh_grow_file(BDRVSSHState *s, int64_t offset, Error **errp) { ssize_t ret; char c[1] = { '\0' }; + int was_blocking = libssh2_session_get_blocking(s->session); /* offset must be strictly greater than the current size so we do * not overwrite anything */ assert(offset > 0 && offset > s->attrs.filesize); + libssh2_session_set_blocking(s->session, 1); + libssh2_sftp_seek64(s->sftp_handle, offset - 1); ret = libssh2_sftp_write(s->sftp_handle, c, 1); + + libssh2_session_set_blocking(s->session, was_blocking); + if (ret < 0) { sftp_error_setg(errp, s, "Failed to grow file"); return -EIO; -- 2.13.6 From: Max Reitz <mreitz@redhat.com> libssh2 does not seem to offer real truncation support, so we can only grow files -- but that is better than nothing. Signed-off-by: Max Reitz <mreitz@redhat.com> Message-id: 20180214204915.7980-4-mreitz@redhat.com Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Richard W.M. Jones <rjones@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- block/ssh.c | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) diff --git a/block/ssh.c b/block/ssh.c index XXXXXXX..XXXXXXX 100644 --- a/block/ssh.c +++ b/block/ssh.c @@ -XXX,XX +XXX,XX @@ static int64_t ssh_getlength(BlockDriverState *bs) return length; } +static int ssh_truncate(BlockDriverState *bs, int64_t offset, + PreallocMode prealloc, Error **errp) +{ + BDRVSSHState *s = bs->opaque; + + if (prealloc != PREALLOC_MODE_OFF) { + error_setg(errp, "Unsupported preallocation mode '%s'", + PreallocMode_str(prealloc)); + return -ENOTSUP; + } + + if (offset < s->attrs.filesize) { + error_setg(errp, "ssh driver does not support shrinking files"); + return -ENOTSUP; + } + + if (offset == s->attrs.filesize) { + return 0; + } + + return ssh_grow_file(s, offset, errp); +} + static BlockDriver bdrv_ssh = { .format_name = "ssh", .protocol_name = "ssh", @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_ssh = { .bdrv_co_readv = ssh_co_readv, .bdrv_co_writev = ssh_co_writev, .bdrv_getlength = ssh_getlength, + .bdrv_truncate = ssh_truncate, .bdrv_co_flush_to_disk = ssh_co_flush, .create_opts = &ssh_create_opts, }; -- 2.13.6 From: Alberto Garcia <berto@igalia.com> The align_offset() function is equivalent to the ROUND_UP() macro so there's no need to use the former. The ROUND_UP() name is also a bit more explicit. This patch uses ROUND_UP() instead of the slower QEMU_ALIGN_UP() because align_offset() already requires that the second parameter is a power of two. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20180215131008.5153-1-berto@igalia.com Signed-off-by: Max Reitz <mreitz@redhat.com> --- block/qcow2.h | 6 ------ block/qcow2-bitmap.c | 4 ++-- block/qcow2-cluster.c | 4 ++-- block/qcow2-refcount.c | 4 ++-- block/qcow2-snapshot.c | 10 +++++----- block/qcow2.c | 14 +++++++------- 6 files changed, 18 insertions(+), 24 deletions(-) diff --git a/block/qcow2.h b/block/qcow2.h index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2.h +++ b/block/qcow2.h @@ -XXX,XX +XXX,XX @@ static inline int offset_to_l2_slice_index(BDRVQcow2State *s, int64_t offset) return (offset >> s->cluster_bits) & (s->l2_slice_size - 1); } -static inline int64_t align_offset(int64_t offset, int n) -{ - offset = (offset + n - 1) & ~(n - 1); - return offset; -} - static inline int64_t qcow2_vm_state_offset(BDRVQcow2State *s) { return (int64_t)s->l1_vm_state_index << (s->cluster_bits + s->l2_bits); diff --git a/block/qcow2-bitmap.c b/block/qcow2-bitmap.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2-bitmap.c +++ b/block/qcow2-bitmap.c @@ -XXX,XX +XXX,XX @@ static inline void bitmap_dir_entry_to_be(Qcow2BitmapDirEntry *entry) static inline int calc_dir_entry_size(size_t name_size, size_t extra_data_size) { - return align_offset(sizeof(Qcow2BitmapDirEntry) + - name_size + extra_data_size, 8); + int size = sizeof(Qcow2BitmapDirEntry) + name_size + extra_data_size; + return ROUND_UP(size, 8); } static inline int dir_entry_size(Qcow2BitmapDirEntry *entry) diff --git a/block/qcow2-cluster.c b/block/qcow2-cluster.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2-cluster.c +++ b/block/qcow2-cluster.c @@ -XXX,XX +XXX,XX @@ int qcow2_grow_l1_table(BlockDriverState *bs, uint64_t min_size, new_l1_size2 = sizeof(uint64_t) * new_l1_size; new_l1_table = qemu_try_blockalign(bs->file->bs, - align_offset(new_l1_size2, 512)); + ROUND_UP(new_l1_size2, 512)); if (new_l1_table == NULL) { return -ENOMEM; } - memset(new_l1_table, 0, align_offset(new_l1_size2, 512)); + memset(new_l1_table, 0, ROUND_UP(new_l1_size2, 512)); if (s->l1_size) { memcpy(new_l1_table, s->l1_table, s->l1_size * sizeof(uint64_t)); diff --git a/block/qcow2-refcount.c b/block/qcow2-refcount.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2-refcount.c +++ b/block/qcow2-refcount.c @@ -XXX,XX +XXX,XX @@ int qcow2_update_snapshot_refcount(BlockDriverState *bs, * l1_table_offset when it is the current s->l1_table_offset! Be careful * when changing this! */ if (l1_table_offset != s->l1_table_offset) { - l1_table = g_try_malloc0(align_offset(l1_size2, 512)); + l1_table = g_try_malloc0(ROUND_UP(l1_size2, 512)); if (l1_size2 && l1_table == NULL) { ret = -ENOMEM; goto fail; @@ -XXX,XX +XXX,XX @@ int qcow2_check_metadata_overlap(BlockDriverState *bs, int ign, int64_t offset, } /* align range to test to cluster boundaries */ - size = align_offset(offset_into_cluster(s, offset) + size, s->cluster_size); + size = ROUND_UP(offset_into_cluster(s, offset) + size, s->cluster_size); offset = start_of_cluster(s, offset); if ((chk & QCOW2_OL_ACTIVE_L1) && s->l1_size) { diff --git a/block/qcow2-snapshot.c b/block/qcow2-snapshot.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2-snapshot.c +++ b/block/qcow2-snapshot.c @@ -XXX,XX +XXX,XX @@ int qcow2_read_snapshots(BlockDriverState *bs) for(i = 0; i < s->nb_snapshots; i++) { /* Read statically sized part of the snapshot header */ - offset = align_offset(offset, 8); + offset = ROUND_UP(offset, 8); ret = bdrv_pread(bs->file, offset, &h, sizeof(h)); if (ret < 0) { goto fail; @@ -XXX,XX +XXX,XX @@ static int qcow2_write_snapshots(BlockDriverState *bs) offset = 0; for(i = 0; i < s->nb_snapshots; i++) { sn = s->snapshots + i; - offset = align_offset(offset, 8); + offset = ROUND_UP(offset, 8); offset += sizeof(h); offset += sizeof(extra); offset += strlen(sn->id_str); @@ -XXX,XX +XXX,XX @@ static int qcow2_write_snapshots(BlockDriverState *bs) assert(id_str_size <= UINT16_MAX && name_size <= UINT16_MAX); h.id_str_size = cpu_to_be16(id_str_size); h.name_size = cpu_to_be16(name_size); - offset = align_offset(offset, 8); + offset = ROUND_UP(offset, 8); ret = bdrv_pwrite(bs->file, offset, &h, sizeof(h)); if (ret < 0) { @@ -XXX,XX +XXX,XX @@ int qcow2_snapshot_create(BlockDriverState *bs, QEMUSnapshotInfo *sn_info) /* The VM state isn't needed any more in the active L1 table; in fact, it * hurts by causing expensive COW for the next snapshot. */ qcow2_cluster_discard(bs, qcow2_vm_state_offset(s), - align_offset(sn->vm_state_size, s->cluster_size), + ROUND_UP(sn->vm_state_size, s->cluster_size), QCOW2_DISCARD_NEVER, false); #ifdef DEBUG_ALLOC @@ -XXX,XX +XXX,XX @@ int qcow2_snapshot_load_tmp(BlockDriverState *bs, } new_l1_bytes = sn->l1_size * sizeof(uint64_t); new_l1_table = qemu_try_blockalign(bs->file->bs, - align_offset(new_l1_bytes, 512)); + ROUND_UP(new_l1_bytes, 512)); if (new_l1_table == NULL) { return -ENOMEM; } diff --git a/block/qcow2.c b/block/qcow2.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -XXX,XX +XXX,XX @@ static int qcow2_do_open(BlockDriverState *bs, QDict *options, int flags, if (s->l1_size > 0) { s->l1_table = qemu_try_blockalign(bs->file->bs, - align_offset(s->l1_size * sizeof(uint64_t), 512)); + ROUND_UP(s->l1_size * sizeof(uint64_t), 512)); if (s->l1_table == NULL) { error_setg(errp, "Could not allocate L1 table"); ret = -ENOMEM; @@ -XXX,XX +XXX,XX @@ static int64_t qcow2_calc_prealloc_size(int64_t total_size, { int64_t meta_size = 0; uint64_t nl1e, nl2e; - int64_t aligned_total_size = align_offset(total_size, cluster_size); + int64_t aligned_total_size = ROUND_UP(total_size, cluster_size); /* header: 1 cluster */ meta_size += cluster_size; /* total size of L2 tables */ nl2e = aligned_total_size / cluster_size; - nl2e = align_offset(nl2e, cluster_size / sizeof(uint64_t)); + nl2e = ROUND_UP(nl2e, cluster_size / sizeof(uint64_t)); meta_size += nl2e * sizeof(uint64_t); /* total size of L1 tables */ nl1e = nl2e * sizeof(uint64_t) / cluster_size; - nl1e = align_offset(nl1e, cluster_size / sizeof(uint64_t)); + nl1e = ROUND_UP(nl1e, cluster_size / sizeof(uint64_t)); meta_size += nl1e * sizeof(uint64_t); /* total size of refcount table and blocks */ @@ -XXX,XX +XXX,XX @@ static BlockMeasureInfo *qcow2_measure(QemuOpts *opts, BlockDriverState *in_bs, has_backing_file = !!optstr; g_free(optstr); - virtual_size = align_offset(qemu_opt_get_size_del(opts, BLOCK_OPT_SIZE, 0), - cluster_size); + virtual_size = qemu_opt_get_size_del(opts, BLOCK_OPT_SIZE, 0); + virtual_size = ROUND_UP(virtual_size, cluster_size); /* Check that virtual disk size is valid */ l2_tables = DIV_ROUND_UP(virtual_size / cluster_size, @@ -XXX,XX +XXX,XX @@ static BlockMeasureInfo *qcow2_measure(QemuOpts *opts, BlockDriverState *in_bs, goto err; } - virtual_size = align_offset(ssize, cluster_size); + virtual_size = ROUND_UP(ssize, cluster_size); if (has_backing_file) { /* We don't how much of the backing chain is shared by the input -- 2.13.6 The following changes since commit 281f327487c9c9b1599f93c589a408bbf4a651b8: Merge remote-tracking branch 'remotes/vivier/tags/m68k-for-2.12-pull-request' into staging (2017-12-22 00:11:36 +0000) are available in the git repository at: git://repo.or.cz/qemu/kevin.git tags/for-upstream for you to fetch changes up to 1a63a907507fbbcfaee3f622907ec244b7eabda8: block: Keep nodes drained between reopen_queue/multiple (2017-12-22 15:05:32 +0100) ---------------------------------------------------------------- Block layer patches ---------------------------------------------------------------- Doug Gale (1): nvme: Add tracing Edgar Kaziakhmedov (1): qcow2: get rid of qcow2_backing_read1 routine Fam Zheng (2): block: Open backing image in force share mode for size probe block: Remove unused bdrv_requests_pending John Snow (1): iotests: fix 197 for vpc Kevin Wolf (27): block: Formats don't need CONSISTENT_READ with NO_IO block: Make bdrv_drain_invoke() recursive block: Call .drain_begin only once in bdrv_drain_all_begin() test-bdrv-drain: Test BlockDriver callbacks for drain block: bdrv_drain_recurse(): Remove unused begin parameter block: Don't wait for requests in bdrv_drain*_end() block: Unify order in drain functions block: Don't acquire AioContext in hmp_qemu_io() block: Document that x-blockdev-change breaks quorum children list block: Assert drain_all is only called from main AioContext block: Make bdrv_drain() driver callbacks non-recursive test-bdrv-drain: Test callback for bdrv_drain test-bdrv-drain: Test bs->quiesce_counter blockjob: Pause job on draining any job BDS test-bdrv-drain: Test drain vs. block jobs block: Don't block_job_pause_all() in bdrv_drain_all() block: Nested drain_end must still call callbacks test-bdrv-drain: Test nested drain sections block: Don't notify parents in drain call chain block: Add bdrv_subtree_drained_begin/end() test-bdrv-drain: Tests for bdrv_subtree_drain test-bdrv-drain: Test behaviour in coroutine context test-bdrv-drain: Recursive draining with multiple parents block: Allow graph changes in subtree drained section test-bdrv-drain: Test graph changes in drained section commit: Simplify reopen of base block: Keep nodes drained between reopen_queue/multiple Thomas Huth (3): block: Remove the obsolete -drive boot=on|off parameter block: Remove the deprecated -hdachs option block: Mention -drive cyls/heads/secs/trans/serial/addr in deprecation chapter qapi/block-core.json | 4 + block/qcow2.h | 3 - include/block/block.h | 15 +- include/block/block_int.h | 6 +- block.c | 75 ++++- block/commit.c | 8 +- block/io.c | 164 +++++++--- block/qcow2.c | 51 +-- block/replication.c | 6 + blockdev.c | 11 - blockjob.c | 22 +- hmp.c | 6 - hw/block/nvme.c | 349 +++++++++++++++++---- qemu-io-cmds.c | 3 + tests/test-bdrv-drain.c | 651 +++++++++++++++++++++++++++++++++++++++ vl.c | 86 +----- hw/block/trace-events | 93 ++++++ qemu-doc.texi | 29 +- qemu-options.hx | 19 +- tests/Makefile.include | 2 + tests/qemu-iotests/197 | 4 + tests/qemu-iotests/common.filter | 3 +- 22 files changed, 1294 insertions(+), 316 deletions(-) create mode 100644 tests/test-bdrv-drain.c Commit 1f4ad7d fixed 'qemu-img info' for raw images that are currently in use as a mirror target. It is not enough for image formats, though, as these still unconditionally request BLK_PERM_CONSISTENT_READ. As this permission is geared towards whether the guest-visible data is consistent, and has no impact on whether the metadata is sane, and 'qemu-img info' does not read guest-visible data (except for the raw format), it makes sense to not require BLK_PERM_CONSISTENT_READ if there is not going to be any guest I/O performed, regardless of image format. Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ void bdrv_format_default_perms(BlockDriverState *bs, BdrvChild *c, assert(role == &child_backing || role == &child_file); if (!backing) { + int flags = bdrv_reopen_get_flags(reopen_queue, bs); + /* Apart from the modifications below, the same permissions are * forwarded and left alone as for filters */ bdrv_filter_default_perms(bs, c, role, reopen_queue, perm, shared, @@ -XXX,XX +XXX,XX @@ void bdrv_format_default_perms(BlockDriverState *bs, BdrvChild *c, /* bs->file always needs to be consistent because of the metadata. We * can never allow other users to resize or write to it. */ - perm |= BLK_PERM_CONSISTENT_READ; + if (!(flags & BDRV_O_NO_IO)) { + perm |= BLK_PERM_CONSISTENT_READ; + } shared &= ~(BLK_PERM_WRITE | BLK_PERM_RESIZE); } else { /* We want consistent read from backing files if the parent needs it. -- 2.13.6 From: John Snow <jsnow@redhat.com> VPC has some difficulty creating geometries of particular size. However, we can indeed force it to use a literal one, so let's do that for the sake of test 197, which is testing some specific offsets. Signed-off-by: John Snow <jsnow@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Lukáš Doktor <ldoktor@redhat.com> --- tests/qemu-iotests/197 | 4 ++++ tests/qemu-iotests/common.filter | 3 ++- 2 files changed, 6 insertions(+), 1 deletion(-) diff --git a/tests/qemu-iotests/197 b/tests/qemu-iotests/197 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/197 +++ b/tests/qemu-iotests/197 @@ -XXX,XX +XXX,XX @@ echo '=== Copy-on-read ===' echo # Prep the images +# VPC rounds image sizes to a specific geometry, force a specific size. +if [ "$IMGFMT" = "vpc" ]; then + IMGOPTS=$(_optstr_add "$IMGOPTS" "force_size") +fi _make_test_img 4G $QEMU_IO -c "write -P 55 3G 1k" "$TEST_IMG" | _filter_qemu_io IMGPROTO=file IMGFMT=qcow2 IMGOPTS= TEST_IMG_FILE="$TEST_WRAP" \ diff --git a/tests/qemu-iotests/common.filter b/tests/qemu-iotests/common.filter index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/common.filter +++ b/tests/qemu-iotests/common.filter @@ -XXX,XX +XXX,XX @@ _filter_img_create() -e "s# log_size=[0-9]\\+##g" \ -e "s# refcount_bits=[0-9]\\+##g" \ -e "s# key-secret=[a-zA-Z0-9]\\+##g" \ - -e "s# iter-time=[0-9]\\+##g" + -e "s# iter-time=[0-9]\\+##g" \ + -e "s# force_size=\\(on\\|off\\)##g" } _filter_img_info() -- 2.13.6 This change separates bdrv_drain_invoke(), which calls the BlockDriver drain callbacks, from bdrv_drain_recurse(). Instead, the function performs its own recursion now. One reason for this is that bdrv_drain_recurse() can be called multiple times by bdrv_drain_all_begin(), but the callbacks may only be called once. The separation is necessary to fix this bug. The other reason is that we intend to go to a model where we call all driver callbacks first, and only then start polling. This is not fully achieved yet with this patch, as bdrv_drain_invoke() contains a BDRV_POLL_WHILE() loop for the block driver callbacks, which can still call callbacks for any unrelated event. It's a step in this direction anyway. Cc: qemu-stable@nongnu.org Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> --- block/io.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_drain_invoke_entry(void *opaque) bdrv_wakeup(bs); } +/* Recursively call BlockDriver.bdrv_co_drain_begin/end callbacks */ static void bdrv_drain_invoke(BlockDriverState *bs, bool begin) { + BdrvChild *child, *tmp; BdrvCoDrainData data = { .bs = bs, .done = false, .begin = begin}; if (!bs->drv || (begin && !bs->drv->bdrv_co_drain_begin) || @@ -XXX,XX +XXX,XX @@ static void bdrv_drain_invoke(BlockDriverState *bs, bool begin) data.co = qemu_coroutine_create(bdrv_drain_invoke_entry, &data); bdrv_coroutine_enter(bs, data.co); BDRV_POLL_WHILE(bs, !data.done); + + QLIST_FOREACH_SAFE(child, &bs->children, next, tmp) { + bdrv_drain_invoke(child->bs, begin); + } } static bool bdrv_drain_recurse(BlockDriverState *bs, bool begin) @@ -XXX,XX +XXX,XX @@ static bool bdrv_drain_recurse(BlockDriverState *bs, bool begin) BdrvChild *child, *tmp; bool waited; - /* Ensure any pending metadata writes are submitted to bs->file. */ - bdrv_drain_invoke(bs, begin); - /* Wait for drained requests to finish */ waited = BDRV_POLL_WHILE(bs, atomic_read(&bs->in_flight) > 0); @@ -XXX,XX +XXX,XX @@ void bdrv_drained_begin(BlockDriverState *bs) bdrv_parent_drained_begin(bs); } + bdrv_drain_invoke(bs, true); bdrv_drain_recurse(bs, true); } @@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs) } bdrv_parent_drained_end(bs); + bdrv_drain_invoke(bs, false); bdrv_drain_recurse(bs, false); aio_enable_external(bdrv_get_aio_context(bs)); } @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void) aio_context_acquire(aio_context); for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) { if (aio_context == bdrv_get_aio_context(bs)) { + /* FIXME Calling this multiple times is wrong */ + bdrv_drain_invoke(bs, true); waited |= bdrv_drain_recurse(bs, true); } } @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void) aio_context_acquire(aio_context); aio_enable_external(aio_context); bdrv_parent_drained_end(bs); + bdrv_drain_invoke(bs, false); bdrv_drain_recurse(bs, false); aio_context_release(aio_context); } -- 2.13.6 bdrv_drain_all_begin() used to call the .bdrv_co_drain_begin() driver callback inside its polling loop. This means that how many times it got called for each node depended on long it had to poll the event loop. This is obviously not right and results in nodes that stay drained even after bdrv_drain_all_end(), which calls .bdrv_co_drain_begin() once per node. Fix bdrv_drain_all_begin() to call the callback only once, too. Cc: qemu-stable@nongnu.org Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> --- block/io.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void) aio_context_acquire(aio_context); bdrv_parent_drained_begin(bs); aio_disable_external(aio_context); + bdrv_drain_invoke(bs, true); aio_context_release(aio_context); if (!g_slist_find(aio_ctxs, aio_context)) { @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void) aio_context_acquire(aio_context); for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) { if (aio_context == bdrv_get_aio_context(bs)) { - /* FIXME Calling this multiple times is wrong */ - bdrv_drain_invoke(bs, true); waited |= bdrv_drain_recurse(bs, true); } } -- 2.13.6 This adds a test case that the BlockDriver callbacks for drain are called in bdrv_drained_all_begin/end(), and that both of them are called exactly once. Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> --- tests/test-bdrv-drain.c | 137 ++++++++++++++++++++++++++++++++++++++++++++++++ tests/Makefile.include | 2 + 2 files changed, 139 insertions(+) create mode 100644 tests/test-bdrv-drain.c diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/test-bdrv-drain.c @@ -XXX,XX +XXX,XX @@ +/* + * Block node draining tests + * + * Copyright (c) 2017 Kevin Wolf <kwolf@redhat.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "block/block.h" +#include "sysemu/block-backend.h" +#include "qapi/error.h" + +typedef struct BDRVTestState { + int drain_count; +} BDRVTestState; + +static void coroutine_fn bdrv_test_co_drain_begin(BlockDriverState *bs) +{ + BDRVTestState *s = bs->opaque; + s->drain_count++; +} + +static void coroutine_fn bdrv_test_co_drain_end(BlockDriverState *bs) +{ + BDRVTestState *s = bs->opaque; + s->drain_count--; +} + +static void bdrv_test_close(BlockDriverState *bs) +{ + BDRVTestState *s = bs->opaque; + g_assert_cmpint(s->drain_count, >, 0); +} + +static int coroutine_fn bdrv_test_co_preadv(BlockDriverState *bs, + uint64_t offset, uint64_t bytes, + QEMUIOVector *qiov, int flags) +{ + /* We want this request to stay until the polling loop in drain waits for + * it to complete. We need to sleep a while as bdrv_drain_invoke() comes + * first and polls its result, too, but it shouldn't accidentally complete + * this request yet. */ + qemu_co_sleep_ns(QEMU_CLOCK_REALTIME, 100000); + + return 0; +} + +static BlockDriver bdrv_test = { + .format_name = "test", + .instance_size = sizeof(BDRVTestState), + + .bdrv_close = bdrv_test_close, + .bdrv_co_preadv = bdrv_test_co_preadv, + + .bdrv_co_drain_begin = bdrv_test_co_drain_begin, + .bdrv_co_drain_end = bdrv_test_co_drain_end, +}; + +static void aio_ret_cb(void *opaque, int ret) +{ + int *aio_ret = opaque; + *aio_ret = ret; +} + +static void test_drv_cb_drain_all(void) +{ + BlockBackend *blk; + BlockDriverState *bs; + BDRVTestState *s; + BlockAIOCB *acb; + int aio_ret; + + QEMUIOVector qiov; + struct iovec iov = { + .iov_base = NULL, + .iov_len = 0, + }; + qemu_iovec_init_external(&qiov, &iov, 1); + + blk = blk_new(BLK_PERM_ALL, BLK_PERM_ALL); + bs = bdrv_new_open_driver(&bdrv_test, "test-node", BDRV_O_RDWR, + &error_abort); + s = bs->opaque; + blk_insert_bs(blk, bs, &error_abort); + + /* Simple bdrv_drain_all_begin/end pair, check that CBs are called */ + g_assert_cmpint(s->drain_count, ==, 0); + bdrv_drain_all_begin(); + g_assert_cmpint(s->drain_count, ==, 1); + bdrv_drain_all_end(); + g_assert_cmpint(s->drain_count, ==, 0); + + /* Now do the same while a request is pending */ + aio_ret = -EINPROGRESS; + acb = blk_aio_preadv(blk, 0, &qiov, 0, aio_ret_cb, &aio_ret); + g_assert(acb != NULL); + g_assert_cmpint(aio_ret, ==, -EINPROGRESS); + + g_assert_cmpint(s->drain_count, ==, 0); + bdrv_drain_all_begin(); + g_assert_cmpint(aio_ret, ==, 0); + g_assert_cmpint(s->drain_count, ==, 1); + bdrv_drain_all_end(); + g_assert_cmpint(s->drain_count, ==, 0); + + bdrv_unref(bs); + blk_unref(blk); +} + +int main(int argc, char **argv) +{ + bdrv_init(); + qemu_init_main_loop(&error_abort); + + g_test_init(&argc, &argv, NULL); + + g_test_add_func("/bdrv-drain/driver-cb/drain_all", test_drv_cb_drain_all); + + return g_test_run(); +} diff --git a/tests/Makefile.include b/tests/Makefile.include index XXXXXXX..XXXXXXX 100644 --- a/tests/Makefile.include +++ b/tests/Makefile.include @@ -XXX,XX +XXX,XX @@ gcov-files-test-thread-pool-y = thread-pool.c gcov-files-test-hbitmap-y = util/hbitmap.c check-unit-y += tests/test-hbitmap$(EXESUF) gcov-files-test-hbitmap-y = blockjob.c +check-unit-y += tests/test-bdrv-drain$(EXESUF) check-unit-y += tests/test-blockjob$(EXESUF) check-unit-y += tests/test-blockjob-txn$(EXESUF) check-unit-y += tests/test-x86-cpuid$(EXESUF) @@ -XXX,XX +XXX,XX @@ tests/test-coroutine$(EXESUF): tests/test-coroutine.o $(test-block-obj-y) tests/test-aio$(EXESUF): tests/test-aio.o $(test-block-obj-y) tests/test-aio-multithread$(EXESUF): tests/test-aio-multithread.o $(test-block-obj-y) tests/test-throttle$(EXESUF): tests/test-throttle.o $(test-block-obj-y) +tests/test-bdrv-drain$(EXESUF): tests/test-bdrv-drain.o $(test-block-obj-y) $(test-util-obj-y) tests/test-blockjob$(EXESUF): tests/test-blockjob.o $(test-block-obj-y) $(test-util-obj-y) tests/test-blockjob-txn$(EXESUF): tests/test-blockjob-txn.o $(test-block-obj-y) $(test-util-obj-y) tests/test-thread-pool$(EXESUF): tests/test-thread-pool.o $(test-block-obj-y) -- 2.13.6 Now that the bdrv_drain_invoke() calls are pulled up to the callers of bdrv_drain_recurse(), the 'begin' parameter isn't needed any more. Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> --- block/io.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ static void bdrv_drain_invoke(BlockDriverState *bs, bool begin) } } -static bool bdrv_drain_recurse(BlockDriverState *bs, bool begin) +static bool bdrv_drain_recurse(BlockDriverState *bs) { BdrvChild *child, *tmp; bool waited; @@ -XXX,XX +XXX,XX @@ static bool bdrv_drain_recurse(BlockDriverState *bs, bool begin) */ bdrv_ref(bs); } - waited |= bdrv_drain_recurse(bs, begin); + waited |= bdrv_drain_recurse(bs); if (in_main_loop) { bdrv_unref(bs); } @@ -XXX,XX +XXX,XX @@ void bdrv_drained_begin(BlockDriverState *bs) } bdrv_drain_invoke(bs, true); - bdrv_drain_recurse(bs, true); + bdrv_drain_recurse(bs); } void bdrv_drained_end(BlockDriverState *bs) @@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs) bdrv_parent_drained_end(bs); bdrv_drain_invoke(bs, false); - bdrv_drain_recurse(bs, false); + bdrv_drain_recurse(bs); aio_enable_external(bdrv_get_aio_context(bs)); } @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void) aio_context_acquire(aio_context); for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) { if (aio_context == bdrv_get_aio_context(bs)) { - waited |= bdrv_drain_recurse(bs, true); + waited |= bdrv_drain_recurse(bs); } } aio_context_release(aio_context); @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void) aio_enable_external(aio_context); bdrv_parent_drained_end(bs); bdrv_drain_invoke(bs, false); - bdrv_drain_recurse(bs, false); + bdrv_drain_recurse(bs); aio_context_release(aio_context); } -- 2.13.6 The device is drained, so there is no point in waiting for requests at the end of the drained section. Remove the bdrv_drain_recurse() calls there. The bdrv_drain_recurse() calls were introduced in commit 481cad48e5e in order to call the .bdrv_co_drain_end() driver callback. This is now done by a separate bdrv_drain_invoke() call. Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> --- block/io.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs) bdrv_parent_drained_end(bs); bdrv_drain_invoke(bs, false); - bdrv_drain_recurse(bs); aio_enable_external(bdrv_get_aio_context(bs)); } @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void) aio_enable_external(aio_context); bdrv_parent_drained_end(bs); bdrv_drain_invoke(bs, false); - bdrv_drain_recurse(bs); aio_context_release(aio_context); } -- 2.13.6 Drain requests are propagated to child nodes, parent nodes and directly to the AioContext. The order in which this happened was different between all combinations of drain/drain_all and begin/end. The correct order is to keep children only drained when their parents are also drained. This means that at the start of a drained section, the AioContext needs to be drained first, the parents second and only then the children. The correct order for the end of a drained section is the opposite. This patch changes the three other functions to follow the example of bdrv_drained_begin(), which is the only one that got it right. Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> --- block/io.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ void bdrv_drained_begin(BlockDriverState *bs) return; } + /* Stop things in parent-to-child order */ if (atomic_fetch_inc(&bs->quiesce_counter) == 0) { aio_disable_external(bdrv_get_aio_context(bs)); bdrv_parent_drained_begin(bs); @@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs) return; } - bdrv_parent_drained_end(bs); + /* Re-enable things in child-to-parent order */ bdrv_drain_invoke(bs, false); + bdrv_parent_drained_end(bs); aio_enable_external(bdrv_get_aio_context(bs)); } @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void) for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) { AioContext *aio_context = bdrv_get_aio_context(bs); + /* Stop things in parent-to-child order */ aio_context_acquire(aio_context); - bdrv_parent_drained_begin(bs); aio_disable_external(aio_context); + bdrv_parent_drained_begin(bs); bdrv_drain_invoke(bs, true); aio_context_release(aio_context); @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void) for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) { AioContext *aio_context = bdrv_get_aio_context(bs); + /* Re-enable things in child-to-parent order */ aio_context_acquire(aio_context); - aio_enable_external(aio_context); - bdrv_parent_drained_end(bs); bdrv_drain_invoke(bs, false); + bdrv_parent_drained_end(bs); + aio_enable_external(aio_context); aio_context_release(aio_context); } -- 2.13.6 Commit 15afd94a047 added code to acquire and release the AioContext in qemuio_command(). This means that the lock is taken twice now in the call path from hmp_qemu_io(). This causes BDRV_POLL_WHILE() to hang for any requests issued to nodes in a non-mainloop AioContext. Dropping the first locking from hmp_qemu_io() fixes the problem. Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> --- hmp.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/hmp.c b/hmp.c index XXXXXXX..XXXXXXX 100644 --- a/hmp.c +++ b/hmp.c @@ -XXX,XX +XXX,XX @@ void hmp_qemu_io(Monitor *mon, const QDict *qdict) { BlockBackend *blk; BlockBackend *local_blk = NULL; - AioContext *aio_context; const char* device = qdict_get_str(qdict, "device"); const char* command = qdict_get_str(qdict, "command"); Error *err = NULL; @@ -XXX,XX +XXX,XX @@ void hmp_qemu_io(Monitor *mon, const QDict *qdict) } } - aio_context = blk_get_aio_context(blk); - aio_context_acquire(aio_context); - /* * Notably absent: Proper permission management. This is sad, but it seems * almost impossible to achieve without changing the semantics and thereby @@ -XXX,XX +XXX,XX @@ void hmp_qemu_io(Monitor *mon, const QDict *qdict) */ qemuio_command(blk, command); - aio_context_release(aio_context); - fail: blk_unref(local_blk); hmp_handle_error(mon, &err); -- 2.13.6 From: Edgar Kaziakhmedov <edgar.kaziakhmedov@virtuozzo.com> Since bdrv_co_preadv does all neccessary checks including reading after the end of the backing file, avoid duplication of verification before bdrv_co_preadv call. Signed-off-by: Edgar Kaziakhmedov <edgar.kaziakhmedov@virtuozzo.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/qcow2.h | 3 --- block/qcow2.c | 51 ++++++++------------------------------------------- 2 files changed, 8 insertions(+), 46 deletions(-) diff --git a/block/qcow2.h b/block/qcow2.h index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2.h +++ b/block/qcow2.h @@ -XXX,XX +XXX,XX @@ uint32_t offset_to_reftable_index(BDRVQcow2State *s, uint64_t offset) } /* qcow2.c functions */ -int qcow2_backing_read1(BlockDriverState *bs, QEMUIOVector *qiov, - int64_t sector_num, int nb_sectors); - int64_t qcow2_refcount_metadata_size(int64_t clusters, size_t cluster_size, int refcount_order, bool generous_increase, uint64_t *refblock_count); diff --git a/block/qcow2.c b/block/qcow2.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -XXX,XX +XXX,XX @@ static int64_t coroutine_fn qcow2_co_get_block_status(BlockDriverState *bs, return status; } -/* handle reading after the end of the backing file */ -int qcow2_backing_read1(BlockDriverState *bs, QEMUIOVector *qiov, - int64_t offset, int bytes) -{ - uint64_t bs_size = bs->total_sectors * BDRV_SECTOR_SIZE; - int n1; - - if ((offset + bytes) <= bs_size) { - return bytes; - } - - if (offset >= bs_size) { - n1 = 0; - } else { - n1 = bs_size - offset; - } - - qemu_iovec_memset(qiov, n1, 0, bytes - n1); - - return n1; -} - static coroutine_fn int qcow2_co_preadv(BlockDriverState *bs, uint64_t offset, uint64_t bytes, QEMUIOVector *qiov, int flags) { BDRVQcow2State *s = bs->opaque; - int offset_in_cluster, n1; + int offset_in_cluster; int ret; unsigned int cur_bytes; /* number of bytes in current iteration */ uint64_t cluster_offset = 0; @@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow2_co_preadv(BlockDriverState *bs, uint64_t offset, case QCOW2_CLUSTER_UNALLOCATED: if (bs->backing) { - /* read from the base image */ - n1 = qcow2_backing_read1(bs->backing->bs, &hd_qiov, - offset, cur_bytes); - if (n1 > 0) { - QEMUIOVector local_qiov; - - qemu_iovec_init(&local_qiov, hd_qiov.niov); - qemu_iovec_concat(&local_qiov, &hd_qiov, 0, n1); - - BLKDBG_EVENT(bs->file, BLKDBG_READ_BACKING_AIO); - qemu_co_mutex_unlock(&s->lock); - ret = bdrv_co_preadv(bs->backing, offset, n1, - &local_qiov, 0); - qemu_co_mutex_lock(&s->lock); - - qemu_iovec_destroy(&local_qiov); - - if (ret < 0) { - goto fail; - } + BLKDBG_EVENT(bs->file, BLKDBG_READ_BACKING_AIO); + qemu_co_mutex_unlock(&s->lock); + ret = bdrv_co_preadv(bs->backing, offset, cur_bytes, + &hd_qiov, 0); + qemu_co_mutex_lock(&s->lock); + if (ret < 0) { + goto fail; } } else { /* Note: in this case, no need to wait */ -- 2.13.6 Removing a quorum child node with x-blockdev-change results in a quorum driver state that cannot be recreated with create options because it would require a list with gaps. This causes trouble in at least .bdrv_refresh_filename(). Document this problem so that we won't accidentally mark the command stable without having addressed it. Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Alberto Garcia <berto@igalia.com> --- qapi/block-core.json | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/qapi/block-core.json b/qapi/block-core.json index XXXXXXX..XXXXXXX 100644 --- a/qapi/block-core.json +++ b/qapi/block-core.json @@ -XXX,XX +XXX,XX @@ # does not support all kinds of operations, all kinds of children, nor # all block drivers. # +# FIXME Removing children from a quorum node means introducing gaps in the +# child indices. This cannot be represented in the 'children' list of +# BlockdevOptionsQuorum, as returned by .bdrv_refresh_filename(). +# # Warning: The data in a new quorum child MUST be consistent with that of # the rest of the array. # -- 2.13.6 From: Doug Gale <doug16k@gmail.com> Add trace output for commands, errors, and undefined behavior. Add guest error log output for undefined behavior. Report invalid undefined accesses to MMIO. Annotate unlikely error checks with unlikely. Signed-off-by: Doug Gale <doug16k@gmail.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- hw/block/nvme.c | 349 ++++++++++++++++++++++++++++++++++++++++++-------- hw/block/trace-events | 93 ++++++++++++++ 2 files changed, 390 insertions(+), 52 deletions(-) diff --git a/hw/block/nvme.c b/hw/block/nvme.c index XXXXXXX..XXXXXXX 100644 --- a/hw/block/nvme.c +++ b/hw/block/nvme.c @@ -XXX,XX +XXX,XX @@ #include "qapi/visitor.h" #include "sysemu/block-backend.h" +#include "qemu/log.h" +#include "trace.h" #include "nvme.h" +#define NVME_GUEST_ERR(trace, fmt, ...) \ + do { \ + (trace_##trace)(__VA_ARGS__); \ + qemu_log_mask(LOG_GUEST_ERROR, #trace \ + " in %s: " fmt "\n", __func__, ## __VA_ARGS__); \ + } while (0) + static void nvme_process_sq(void *opaque); static void nvme_addr_read(NvmeCtrl *n, hwaddr addr, void *buf, int size) @@ -XXX,XX +XXX,XX @@ static void nvme_isr_notify(NvmeCtrl *n, NvmeCQueue *cq) { if (cq->irq_enabled) { if (msix_enabled(&(n->parent_obj))) { + trace_nvme_irq_msix(cq->vector); msix_notify(&(n->parent_obj), cq->vector); } else { + trace_nvme_irq_pin(); pci_irq_pulse(&n->parent_obj); } + } else { + trace_nvme_irq_masked(); } } @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1, trans_len = MIN(len, trans_len); int num_prps = (len >> n->page_bits) + 1; - if (!prp1) { + if (unlikely(!prp1)) { + trace_nvme_err_invalid_prp(); return NVME_INVALID_FIELD | NVME_DNR; } else if (n->cmbsz && prp1 >= n->ctrl_mem.addr && prp1 < n->ctrl_mem.addr + int128_get64(n->ctrl_mem.size)) { @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1, } len -= trans_len; if (len) { - if (!prp2) { + if (unlikely(!prp2)) { + trace_nvme_err_invalid_prp2_missing(); goto unmap; } if (len > n->page_size) { @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1, uint64_t prp_ent = le64_to_cpu(prp_list[i]); if (i == n->max_prp_ents - 1 && len > n->page_size) { - if (!prp_ent || prp_ent & (n->page_size - 1)) { + if (unlikely(!prp_ent || prp_ent & (n->page_size - 1))) { + trace_nvme_err_invalid_prplist_ent(prp_ent); goto unmap; } @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1, prp_ent = le64_to_cpu(prp_list[i]); } - if (!prp_ent || prp_ent & (n->page_size - 1)) { + if (unlikely(!prp_ent || prp_ent & (n->page_size - 1))) { + trace_nvme_err_invalid_prplist_ent(prp_ent); goto unmap; } @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1, i++; } } else { - if (prp2 & (n->page_size - 1)) { + if (unlikely(prp2 & (n->page_size - 1))) { + trace_nvme_err_invalid_prp2_align(prp2); goto unmap; } if (qsg->nsg) { @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_dma_read_prp(NvmeCtrl *n, uint8_t *ptr, uint32_t len, QEMUIOVector iov; uint16_t status = NVME_SUCCESS; + trace_nvme_dma_read(prp1, prp2); + if (nvme_map_prp(&qsg, &iov, prp1, prp2, len, n)) { return NVME_INVALID_FIELD | NVME_DNR; } if (qsg.nsg > 0) { - if (dma_buf_read(ptr, len, &qsg)) { + if (unlikely(dma_buf_read(ptr, len, &qsg))) { + trace_nvme_err_invalid_dma(); status = NVME_INVALID_FIELD | NVME_DNR; } qemu_sglist_destroy(&qsg); } else { - if (qemu_iovec_to_buf(&iov, 0, ptr, len) != len) { + if (unlikely(qemu_iovec_to_buf(&iov, 0, ptr, len) != len)) { + trace_nvme_err_invalid_dma(); status = NVME_INVALID_FIELD | NVME_DNR; } qemu_iovec_destroy(&iov); @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_write_zeros(NvmeCtrl *n, NvmeNamespace *ns, NvmeCmd *cmd, uint64_t aio_slba = slba << (data_shift - BDRV_SECTOR_BITS); uint32_t aio_nlb = nlb << (data_shift - BDRV_SECTOR_BITS); - if (slba + nlb > ns->id_ns.nsze) { + if (unlikely(slba + nlb > ns->id_ns.nsze)) { + trace_nvme_err_invalid_lba_range(slba, nlb, ns->id_ns.nsze); return NVME_LBA_RANGE | NVME_DNR; } @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_rw(NvmeCtrl *n, NvmeNamespace *ns, NvmeCmd *cmd, int is_write = rw->opcode == NVME_CMD_WRITE ? 1 : 0; enum BlockAcctType acct = is_write ? BLOCK_ACCT_WRITE : BLOCK_ACCT_READ; - if ((slba + nlb) > ns->id_ns.nsze) { + trace_nvme_rw(is_write ? "write" : "read", nlb, data_size, slba); + + if (unlikely((slba + nlb) > ns->id_ns.nsze)) { block_acct_invalid(blk_get_stats(n->conf.blk), acct); + trace_nvme_err_invalid_lba_range(slba, nlb, ns->id_ns.nsze); return NVME_LBA_RANGE | NVME_DNR; } @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_io_cmd(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req) NvmeNamespace *ns; uint32_t nsid = le32_to_cpu(cmd->nsid); - if (nsid == 0 || nsid > n->num_namespaces) { + if (unlikely(nsid == 0 || nsid > n->num_namespaces)) { + trace_nvme_err_invalid_ns(nsid, n->num_namespaces); return NVME_INVALID_NSID | NVME_DNR; } @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_io_cmd(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req) case NVME_CMD_READ: return nvme_rw(n, ns, cmd, req); default: + trace_nvme_err_invalid_opc(cmd->opcode); return NVME_INVALID_OPCODE | NVME_DNR; } } @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_del_sq(NvmeCtrl *n, NvmeCmd *cmd) NvmeCQueue *cq; uint16_t qid = le16_to_cpu(c->qid); - if (!qid || nvme_check_sqid(n, qid)) { + if (unlikely(!qid || nvme_check_sqid(n, qid))) { + trace_nvme_err_invalid_del_sq(qid); return NVME_INVALID_QID | NVME_DNR; } + trace_nvme_del_sq(qid); + sq = n->sq[qid]; while (!QTAILQ_EMPTY(&sq->out_req_list)) { req = QTAILQ_FIRST(&sq->out_req_list); @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_create_sq(NvmeCtrl *n, NvmeCmd *cmd) uint16_t qflags = le16_to_cpu(c->sq_flags); uint64_t prp1 = le64_to_cpu(c->prp1); - if (!cqid || nvme_check_cqid(n, cqid)) { + trace_nvme_create_sq(prp1, sqid, cqid, qsize, qflags); + + if (unlikely(!cqid || nvme_check_cqid(n, cqid))) { + trace_nvme_err_invalid_create_sq_cqid(cqid); return NVME_INVALID_CQID | NVME_DNR; } - if (!sqid || !nvme_check_sqid(n, sqid)) { + if (unlikely(!sqid || !nvme_check_sqid(n, sqid))) { + trace_nvme_err_invalid_create_sq_sqid(sqid); return NVME_INVALID_QID | NVME_DNR; } - if (!qsize || qsize > NVME_CAP_MQES(n->bar.cap)) { + if (unlikely(!qsize || qsize > NVME_CAP_MQES(n->bar.cap))) { + trace_nvme_err_invalid_create_sq_size(qsize); return NVME_MAX_QSIZE_EXCEEDED | NVME_DNR; } - if (!prp1 || prp1 & (n->page_size - 1)) { + if (unlikely(!prp1 || prp1 & (n->page_size - 1))) { + trace_nvme_err_invalid_create_sq_addr(prp1); return NVME_INVALID_FIELD | NVME_DNR; } - if (!(NVME_SQ_FLAGS_PC(qflags))) { + if (unlikely(!(NVME_SQ_FLAGS_PC(qflags)))) { + trace_nvme_err_invalid_create_sq_qflags(NVME_SQ_FLAGS_PC(qflags)); return NVME_INVALID_FIELD | NVME_DNR; } sq = g_malloc0(sizeof(*sq)); @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_del_cq(NvmeCtrl *n, NvmeCmd *cmd) NvmeCQueue *cq; uint16_t qid = le16_to_cpu(c->qid); - if (!qid || nvme_check_cqid(n, qid)) { + if (unlikely(!qid || nvme_check_cqid(n, qid))) { + trace_nvme_err_invalid_del_cq_cqid(qid); return NVME_INVALID_CQID | NVME_DNR; } cq = n->cq[qid]; - if (!QTAILQ_EMPTY(&cq->sq_list)) { + if (unlikely(!QTAILQ_EMPTY(&cq->sq_list))) { + trace_nvme_err_invalid_del_cq_notempty(qid); return NVME_INVALID_QUEUE_DEL; } + trace_nvme_del_cq(qid); nvme_free_cq(cq, n); return NVME_SUCCESS; } @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_create_cq(NvmeCtrl *n, NvmeCmd *cmd) uint16_t qflags = le16_to_cpu(c->cq_flags); uint64_t prp1 = le64_to_cpu(c->prp1); - if (!cqid || !nvme_check_cqid(n, cqid)) { + trace_nvme_create_cq(prp1, cqid, vector, qsize, qflags, + NVME_CQ_FLAGS_IEN(qflags) != 0); + + if (unlikely(!cqid || !nvme_check_cqid(n, cqid))) { + trace_nvme_err_invalid_create_cq_cqid(cqid); return NVME_INVALID_CQID | NVME_DNR; } - if (!qsize || qsize > NVME_CAP_MQES(n->bar.cap)) { + if (unlikely(!qsize || qsize > NVME_CAP_MQES(n->bar.cap))) { + trace_nvme_err_invalid_create_cq_size(qsize); return NVME_MAX_QSIZE_EXCEEDED | NVME_DNR; } - if (!prp1) { + if (unlikely(!prp1)) { + trace_nvme_err_invalid_create_cq_addr(prp1); return NVME_INVALID_FIELD | NVME_DNR; } - if (vector > n->num_queues) { + if (unlikely(vector > n->num_queues)) { + trace_nvme_err_invalid_create_cq_vector(vector); return NVME_INVALID_IRQ_VECTOR | NVME_DNR; } - if (!(NVME_CQ_FLAGS_PC(qflags))) { + if (unlikely(!(NVME_CQ_FLAGS_PC(qflags)))) { + trace_nvme_err_invalid_create_cq_qflags(NVME_CQ_FLAGS_PC(qflags)); return NVME_INVALID_FIELD | NVME_DNR; } @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_identify_ctrl(NvmeCtrl *n, NvmeIdentify *c) uint64_t prp1 = le64_to_cpu(c->prp1); uint64_t prp2 = le64_to_cpu(c->prp2); + trace_nvme_identify_ctrl(); + return nvme_dma_read_prp(n, (uint8_t *)&n->id_ctrl, sizeof(n->id_ctrl), prp1, prp2); } @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_identify_ns(NvmeCtrl *n, NvmeIdentify *c) uint64_t prp1 = le64_to_cpu(c->prp1); uint64_t prp2 = le64_to_cpu(c->prp2); - if (nsid == 0 || nsid > n->num_namespaces) { + trace_nvme_identify_ns(nsid); + + if (unlikely(nsid == 0 || nsid > n->num_namespaces)) { + trace_nvme_err_invalid_ns(nsid, n->num_namespaces); return NVME_INVALID_NSID | NVME_DNR; } ns = &n->namespaces[nsid - 1]; + return nvme_dma_read_prp(n, (uint8_t *)&ns->id_ns, sizeof(ns->id_ns), prp1, prp2); } @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_identify_nslist(NvmeCtrl *n, NvmeIdentify *c) uint16_t ret; int i, j = 0; + trace_nvme_identify_nslist(min_nsid); + list = g_malloc0(data_len); for (i = 0; i < n->num_namespaces; i++) { if (i < min_nsid) { @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_identify(NvmeCtrl *n, NvmeCmd *cmd) case 0x02: return nvme_identify_nslist(n, c); default: + trace_nvme_err_invalid_identify_cns(le32_to_cpu(c->cns)); return NVME_INVALID_FIELD | NVME_DNR; } } @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_get_feature(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req) switch (dw10) { case NVME_VOLATILE_WRITE_CACHE: result = blk_enable_write_cache(n->conf.blk); + trace_nvme_getfeat_vwcache(result ? "enabled" : "disabled"); break; case NVME_NUMBER_OF_QUEUES: result = cpu_to_le32((n->num_queues - 2) | ((n->num_queues - 2) << 16)); + trace_nvme_getfeat_numq(result); break; default: + trace_nvme_err_invalid_getfeat(dw10); return NVME_INVALID_FIELD | NVME_DNR; } @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_set_feature(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req) blk_set_enable_write_cache(n->conf.blk, dw11 & 1); break; case NVME_NUMBER_OF_QUEUES: + trace_nvme_setfeat_numq((dw11 & 0xFFFF) + 1, + ((dw11 >> 16) & 0xFFFF) + 1, + n->num_queues - 1, n->num_queues - 1); req->cqe.result = cpu_to_le32((n->num_queues - 2) | ((n->num_queues - 2) << 16)); break; default: + trace_nvme_err_invalid_setfeat(dw10); return NVME_INVALID_FIELD | NVME_DNR; } return NVME_SUCCESS; @@ -XXX,XX +XXX,XX @@ static uint16_t nvme_admin_cmd(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req) case NVME_ADM_CMD_GET_FEATURES: return nvme_get_feature(n, cmd, req); default: + trace_nvme_err_invalid_admin_opc(cmd->opcode); return NVME_INVALID_OPCODE | NVME_DNR; } } @@ -XXX,XX +XXX,XX @@ static int nvme_start_ctrl(NvmeCtrl *n) uint32_t page_bits = NVME_CC_MPS(n->bar.cc) + 12; uint32_t page_size = 1 << page_bits; - if (n->cq[0] || n->sq[0] || !n->bar.asq || !n->bar.acq || - n->bar.asq & (page_size - 1) || n->bar.acq & (page_size - 1) || - NVME_CC_MPS(n->bar.cc) < NVME_CAP_MPSMIN(n->bar.cap) || - NVME_CC_MPS(n->bar.cc) > NVME_CAP_MPSMAX(n->bar.cap) || - NVME_CC_IOCQES(n->bar.cc) < NVME_CTRL_CQES_MIN(n->id_ctrl.cqes) || - NVME_CC_IOCQES(n->bar.cc) > NVME_CTRL_CQES_MAX(n->id_ctrl.cqes) || - NVME_CC_IOSQES(n->bar.cc) < NVME_CTRL_SQES_MIN(n->id_ctrl.sqes) || - NVME_CC_IOSQES(n->bar.cc) > NVME_CTRL_SQES_MAX(n->id_ctrl.sqes) || - !NVME_AQA_ASQS(n->bar.aqa) || !NVME_AQA_ACQS(n->bar.aqa)) { + if (unlikely(n->cq[0])) { + trace_nvme_err_startfail_cq(); + return -1; + } + if (unlikely(n->sq[0])) { + trace_nvme_err_startfail_sq(); + return -1; + } + if (unlikely(!n->bar.asq)) { + trace_nvme_err_startfail_nbarasq(); + return -1; + } + if (unlikely(!n->bar.acq)) { + trace_nvme_err_startfail_nbaracq(); + return -1; + } + if (unlikely(n->bar.asq & (page_size - 1))) { + trace_nvme_err_startfail_asq_misaligned(n->bar.asq); + return -1; + } + if (unlikely(n->bar.acq & (page_size - 1))) { + trace_nvme_err_startfail_acq_misaligned(n->bar.acq); + return -1; + } + if (unlikely(NVME_CC_MPS(n->bar.cc) < + NVME_CAP_MPSMIN(n->bar.cap))) { + trace_nvme_err_startfail_page_too_small( + NVME_CC_MPS(n->bar.cc), + NVME_CAP_MPSMIN(n->bar.cap)); + return -1; + } + if (unlikely(NVME_CC_MPS(n->bar.cc) > + NVME_CAP_MPSMAX(n->bar.cap))) { + trace_nvme_err_startfail_page_too_large( + NVME_CC_MPS(n->bar.cc), + NVME_CAP_MPSMAX(n->bar.cap)); + return -1; + } + if (unlikely(NVME_CC_IOCQES(n->bar.cc) < + NVME_CTRL_CQES_MIN(n->id_ctrl.cqes))) { + trace_nvme_err_startfail_cqent_too_small( + NVME_CC_IOCQES(n->bar.cc), + NVME_CTRL_CQES_MIN(n->bar.cap)); + return -1; + } + if (unlikely(NVME_CC_IOCQES(n->bar.cc) > + NVME_CTRL_CQES_MAX(n->id_ctrl.cqes))) { + trace_nvme_err_startfail_cqent_too_large( + NVME_CC_IOCQES(n->bar.cc), + NVME_CTRL_CQES_MAX(n->bar.cap)); + return -1; + } + if (unlikely(NVME_CC_IOSQES(n->bar.cc) < + NVME_CTRL_SQES_MIN(n->id_ctrl.sqes))) { + trace_nvme_err_startfail_sqent_too_small( + NVME_CC_IOSQES(n->bar.cc), + NVME_CTRL_SQES_MIN(n->bar.cap)); + return -1; + } + if (unlikely(NVME_CC_IOSQES(n->bar.cc) > + NVME_CTRL_SQES_MAX(n->id_ctrl.sqes))) { + trace_nvme_err_startfail_sqent_too_large( + NVME_CC_IOSQES(n->bar.cc), + NVME_CTRL_SQES_MAX(n->bar.cap)); + return -1; + } + if (unlikely(!NVME_AQA_ASQS(n->bar.aqa))) { + trace_nvme_err_startfail_asqent_sz_zero(); + return -1; + } + if (unlikely(!NVME_AQA_ACQS(n->bar.aqa))) { + trace_nvme_err_startfail_acqent_sz_zero(); return -1; } @@ -XXX,XX +XXX,XX @@ static int nvme_start_ctrl(NvmeCtrl *n) static void nvme_write_bar(NvmeCtrl *n, hwaddr offset, uint64_t data, unsigned size) { + if (unlikely(offset & (sizeof(uint32_t) - 1))) { + NVME_GUEST_ERR(nvme_ub_mmiowr_misaligned32, + "MMIO write not 32-bit aligned," + " offset=0x%"PRIx64"", offset); + /* should be ignored, fall through for now */ + } + + if (unlikely(size < sizeof(uint32_t))) { + NVME_GUEST_ERR(nvme_ub_mmiowr_toosmall, + "MMIO write smaller than 32-bits," + " offset=0x%"PRIx64", size=%u", + offset, size); + /* should be ignored, fall through for now */ + } + switch (offset) { - case 0xc: + case 0xc: /* INTMS */ + if (unlikely(msix_enabled(&(n->parent_obj)))) { + NVME_GUEST_ERR(nvme_ub_mmiowr_intmask_with_msix, + "undefined access to interrupt mask set" + " when MSI-X is enabled"); + /* should be ignored, fall through for now */ + } n->bar.intms |= data & 0xffffffff; n->bar.intmc = n->bar.intms; + trace_nvme_mmio_intm_set(data & 0xffffffff, + n->bar.intmc); break; - case 0x10: + case 0x10: /* INTMC */ + if (unlikely(msix_enabled(&(n->parent_obj)))) { + NVME_GUEST_ERR(nvme_ub_mmiowr_intmask_with_msix, + "undefined access to interrupt mask clr" + " when MSI-X is enabled"); + /* should be ignored, fall through for now */ + } n->bar.intms &= ~(data & 0xffffffff); n->bar.intmc = n->bar.intms; + trace_nvme_mmio_intm_clr(data & 0xffffffff, + n->bar.intmc); break; - case 0x14: + case 0x14: /* CC */ + trace_nvme_mmio_cfg(data & 0xffffffff); /* Windows first sends data, then sends enable bit */ if (!NVME_CC_EN(data) && !NVME_CC_EN(n->bar.cc) && !NVME_CC_SHN(data) && !NVME_CC_SHN(n->bar.cc)) @@ -XXX,XX +XXX,XX @@ static void nvme_write_bar(NvmeCtrl *n, hwaddr offset, uint64_t data, if (NVME_CC_EN(data) && !NVME_CC_EN(n->bar.cc)) { n->bar.cc = data; - if (nvme_start_ctrl(n)) { + if (unlikely(nvme_start_ctrl(n))) { + trace_nvme_err_startfail(); n->bar.csts = NVME_CSTS_FAILED; } else { + trace_nvme_mmio_start_success(); n->bar.csts = NVME_CSTS_READY; } } else if (!NVME_CC_EN(data) && NVME_CC_EN(n->bar.cc)) { + trace_nvme_mmio_stopped(); nvme_clear_ctrl(n); n->bar.csts &= ~NVME_CSTS_READY; } if (NVME_CC_SHN(data) && !(NVME_CC_SHN(n->bar.cc))) { - nvme_clear_ctrl(n); - n->bar.cc = data; - n->bar.csts |= NVME_CSTS_SHST_COMPLETE; + trace_nvme_mmio_shutdown_set(); + nvme_clear_ctrl(n); + n->bar.cc = data; + n->bar.csts |= NVME_CSTS_SHST_COMPLETE; } else if (!NVME_CC_SHN(data) && NVME_CC_SHN(n->bar.cc)) { - n->bar.csts &= ~NVME_CSTS_SHST_COMPLETE; - n->bar.cc = data; + trace_nvme_mmio_shutdown_cleared(); + n->bar.csts &= ~NVME_CSTS_SHST_COMPLETE; + n->bar.cc = data; + } + break; + case 0x1C: /* CSTS */ + if (data & (1 << 4)) { + NVME_GUEST_ERR(nvme_ub_mmiowr_ssreset_w1c_unsupported, + "attempted to W1C CSTS.NSSRO" + " but CAP.NSSRS is zero (not supported)"); + } else if (data != 0) { + NVME_GUEST_ERR(nvme_ub_mmiowr_ro_csts, + "attempted to set a read only bit" + " of controller status"); + } + break; + case 0x20: /* NSSR */ + if (data == 0x4E564D65) { + trace_nvme_ub_mmiowr_ssreset_unsupported(); + } else { + /* The spec says that writes of other values have no effect */ + return; } break; - case 0x24: + case 0x24: /* AQA */ n->bar.aqa = data & 0xffffffff; + trace_nvme_mmio_aqattr(data & 0xffffffff); break; - case 0x28: + case 0x28: /* ASQ */ n->bar.asq = data; + trace_nvme_mmio_asqaddr(data); break; - case 0x2c: + case 0x2c: /* ASQ hi */ n->bar.asq |= data << 32; + trace_nvme_mmio_asqaddr_hi(data, n->bar.asq); break; - case 0x30: + case 0x30: /* ACQ */ + trace_nvme_mmio_acqaddr(data); n->bar.acq = data; break; - case 0x34: + case 0x34: /* ACQ hi */ n->bar.acq |= data << 32; + trace_nvme_mmio_acqaddr_hi(data, n->bar.acq); break; + case 0x38: /* CMBLOC */ + NVME_GUEST_ERR(nvme_ub_mmiowr_cmbloc_reserved, + "invalid write to reserved CMBLOC" + " when CMBSZ is zero, ignored"); + return; + case 0x3C: /* CMBSZ */ + NVME_GUEST_ERR(nvme_ub_mmiowr_cmbsz_readonly, + "invalid write to read only CMBSZ, ignored"); + return; default: + NVME_GUEST_ERR(nvme_ub_mmiowr_invalid, + "invalid MMIO write," + " offset=0x%"PRIx64", data=%"PRIx64"", + offset, data); break; } } @@ -XXX,XX +XXX,XX @@ static uint64_t nvme_mmio_read(void *opaque, hwaddr addr, unsigned size) uint8_t *ptr = (uint8_t *)&n->bar; uint64_t val = 0; + if (unlikely(addr & (sizeof(uint32_t) - 1))) { + NVME_GUEST_ERR(nvme_ub_mmiord_misaligned32, + "MMIO read not 32-bit aligned," + " offset=0x%"PRIx64"", addr); + /* should RAZ, fall through for now */ + } else if (unlikely(size < sizeof(uint32_t))) { + NVME_GUEST_ERR(nvme_ub_mmiord_toosmall, + "MMIO read smaller than 32-bits," + " offset=0x%"PRIx64"", addr); + /* should RAZ, fall through for now */ + } + if (addr < sizeof(n->bar)) { memcpy(&val, ptr + addr, size); + } else { + NVME_GUEST_ERR(nvme_ub_mmiord_invalid_ofs, + "MMIO read beyond last register," + " offset=0x%"PRIx64", returning 0", addr); } + return val; } @@ -XXX,XX +XXX,XX @@ static void nvme_process_db(NvmeCtrl *n, hwaddr addr, int val) { uint32_t qid; - if (addr & ((1 << 2) - 1)) { + if (unlikely(addr & ((1 << 2) - 1))) { + NVME_GUEST_ERR(nvme_ub_db_wr_misaligned, + "doorbell write not 32-bit aligned," + " offset=0x%"PRIx64", ignoring", addr); return; } if (((addr - 0x1000) >> 2) & 1) { + /* Completion queue doorbell write */ + uint16_t new_head = val & 0xffff; int start_sqs; NvmeCQueue *cq; qid = (addr - (0x1000 + (1 << 2))) >> 3; - if (nvme_check_cqid(n, qid)) { + if (unlikely(nvme_check_cqid(n, qid))) { + NVME_GUEST_ERR(nvme_ub_db_wr_invalid_cq, + "completion queue doorbell write" + " for nonexistent queue," + " sqid=%"PRIu32", ignoring", qid); return; } cq = n->cq[qid]; - if (new_head >= cq->size) { + if (unlikely(new_head >= cq->size)) { + NVME_GUEST_ERR(nvme_ub_db_wr_invalid_cqhead, + "completion queue doorbell write value" + " beyond queue size, sqid=%"PRIu32"," + " new_head=%"PRIu16", ignoring", + qid, new_head); return; } @@ -XXX,XX +XXX,XX @@ static void nvme_process_db(NvmeCtrl *n, hwaddr addr, int val) nvme_isr_notify(n, cq); } } else { + /* Submission queue doorbell write */ + uint16_t new_tail = val & 0xffff; NvmeSQueue *sq; qid = (addr - 0x1000) >> 3; - if (nvme_check_sqid(n, qid)) { + if (unlikely(nvme_check_sqid(n, qid))) { + NVME_GUEST_ERR(nvme_ub_db_wr_invalid_sq, + "submission queue doorbell write" + " for nonexistent queue," + " sqid=%"PRIu32", ignoring", qid); return; } sq = n->sq[qid]; - if (new_tail >= sq->size) { + if (unlikely(new_tail >= sq->size)) { + NVME_GUEST_ERR(nvme_ub_db_wr_invalid_sqtail, + "submission queue doorbell write value" + " beyond queue size, sqid=%"PRIu32"," + " new_tail=%"PRIu16", ignoring", + qid, new_tail); return; } diff --git a/hw/block/trace-events b/hw/block/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/block/trace-events +++ b/hw/block/trace-events @@ -XXX,XX +XXX,XX @@ virtio_blk_submit_multireq(void *vdev, void *mrb, int start, int num_reqs, uint6 hd_geometry_lchs_guess(void *blk, int cyls, int heads, int secs) "blk %p LCHS %d %d %d" hd_geometry_guess(void *blk, uint32_t cyls, uint32_t heads, uint32_t secs, int trans) "blk %p CHS %u %u %u trans %d" +# hw/block/nvme.c +# nvme traces for successful events +nvme_irq_msix(uint32_t vector) "raising MSI-X IRQ vector %u" +nvme_irq_pin(void) "pulsing IRQ pin" +nvme_irq_masked(void) "IRQ is masked" +nvme_dma_read(uint64_t prp1, uint64_t prp2) "DMA read, prp1=0x%"PRIx64" prp2=0x%"PRIx64"" +nvme_rw(char const *verb, uint32_t blk_count, uint64_t byte_count, uint64_t lba) "%s %"PRIu32" blocks (%"PRIu64" bytes) from LBA %"PRIu64"" +nvme_create_sq(uint64_t addr, uint16_t sqid, uint16_t cqid, uint16_t qsize, uint16_t qflags) "create submission queue, addr=0x%"PRIx64", sqid=%"PRIu16", cqid=%"PRIu16", qsize=%"PRIu16", qflags=%"PRIu16"" +nvme_create_cq(uint64_t addr, uint16_t cqid, uint16_t vector, uint16_t size, uint16_t qflags, int ien) "create completion queue, addr=0x%"PRIx64", cqid=%"PRIu16", vector=%"PRIu16", qsize=%"PRIu16", qflags=%"PRIu16", ien=%d" +nvme_del_sq(uint16_t qid) "deleting submission queue sqid=%"PRIu16"" +nvme_del_cq(uint16_t cqid) "deleted completion queue, sqid=%"PRIu16"" +nvme_identify_ctrl(void) "identify controller" +nvme_identify_ns(uint16_t ns) "identify namespace, nsid=%"PRIu16"" +nvme_identify_nslist(uint16_t ns) "identify namespace list, nsid=%"PRIu16"" +nvme_getfeat_vwcache(char const* result) "get feature volatile write cache, result=%s" +nvme_getfeat_numq(int result) "get feature number of queues, result=%d" +nvme_setfeat_numq(int reqcq, int reqsq, int gotcq, int gotsq) "requested cq_count=%d sq_count=%d, responding with cq_count=%d sq_count=%d" +nvme_mmio_intm_set(uint64_t data, uint64_t new_mask) "wrote MMIO, interrupt mask set, data=0x%"PRIx64", new_mask=0x%"PRIx64"" +nvme_mmio_intm_clr(uint64_t data, uint64_t new_mask) "wrote MMIO, interrupt mask clr, data=0x%"PRIx64", new_mask=0x%"PRIx64"" +nvme_mmio_cfg(uint64_t data) "wrote MMIO, config controller config=0x%"PRIx64"" +nvme_mmio_aqattr(uint64_t data) "wrote MMIO, admin queue attributes=0x%"PRIx64"" +nvme_mmio_asqaddr(uint64_t data) "wrote MMIO, admin submission queue address=0x%"PRIx64"" +nvme_mmio_acqaddr(uint64_t data) "wrote MMIO, admin completion queue address=0x%"PRIx64"" +nvme_mmio_asqaddr_hi(uint64_t data, uint64_t new_addr) "wrote MMIO, admin submission queue high half=0x%"PRIx64", new_address=0x%"PRIx64"" +nvme_mmio_acqaddr_hi(uint64_t data, uint64_t new_addr) "wrote MMIO, admin completion queue high half=0x%"PRIx64", new_address=0x%"PRIx64"" +nvme_mmio_start_success(void) "setting controller enable bit succeeded" +nvme_mmio_stopped(void) "cleared controller enable bit" +nvme_mmio_shutdown_set(void) "shutdown bit set" +nvme_mmio_shutdown_cleared(void) "shutdown bit cleared" + +# nvme traces for error conditions +nvme_err_invalid_dma(void) "PRP/SGL is too small for transfer size" +nvme_err_invalid_prplist_ent(uint64_t prplist) "PRP list entry is null or not page aligned: 0x%"PRIx64"" +nvme_err_invalid_prp2_align(uint64_t prp2) "PRP2 is not page aligned: 0x%"PRIx64"" +nvme_err_invalid_prp2_missing(void) "PRP2 is null and more data to be transferred" +nvme_err_invalid_field(void) "invalid field" +nvme_err_invalid_prp(void) "invalid PRP" +nvme_err_invalid_sgl(void) "invalid SGL" +nvme_err_invalid_ns(uint32_t ns, uint32_t limit) "invalid namespace %u not within 1-%u" +nvme_err_invalid_opc(uint8_t opc) "invalid opcode 0x%"PRIx8"" +nvme_err_invalid_admin_opc(uint8_t opc) "invalid admin opcode 0x%"PRIx8"" +nvme_err_invalid_lba_range(uint64_t start, uint64_t len, uint64_t limit) "Invalid LBA start=%"PRIu64" len=%"PRIu64" limit=%"PRIu64"" +nvme_err_invalid_del_sq(uint16_t qid) "invalid submission queue deletion, sid=%"PRIu16"" +nvme_err_invalid_create_sq_cqid(uint16_t cqid) "failed creating submission queue, invalid cqid=%"PRIu16"" +nvme_err_invalid_create_sq_sqid(uint16_t sqid) "failed creating submission queue, invalid sqid=%"PRIu16"" +nvme_err_invalid_create_sq_size(uint16_t qsize) "failed creating submission queue, invalid qsize=%"PRIu16"" +nvme_err_invalid_create_sq_addr(uint64_t addr) "failed creating submission queue, addr=0x%"PRIx64"" +nvme_err_invalid_create_sq_qflags(uint16_t qflags) "failed creating submission queue, qflags=%"PRIu16"" +nvme_err_invalid_del_cq_cqid(uint16_t cqid) "failed deleting completion queue, cqid=%"PRIu16"" +nvme_err_invalid_del_cq_notempty(uint16_t cqid) "failed deleting completion queue, it is not empty, cqid=%"PRIu16"" +nvme_err_invalid_create_cq_cqid(uint16_t cqid) "failed creating completion queue, cqid=%"PRIu16"" +nvme_err_invalid_create_cq_size(uint16_t size) "failed creating completion queue, size=%"PRIu16"" +nvme_err_invalid_create_cq_addr(uint64_t addr) "failed creating completion queue, addr=0x%"PRIx64"" +nvme_err_invalid_create_cq_vector(uint16_t vector) "failed creating completion queue, vector=%"PRIu16"" +nvme_err_invalid_create_cq_qflags(uint16_t qflags) "failed creating completion queue, qflags=%"PRIu16"" +nvme_err_invalid_identify_cns(uint16_t cns) "identify, invalid cns=0x%"PRIx16"" +nvme_err_invalid_getfeat(int dw10) "invalid get features, dw10=0x%"PRIx32"" +nvme_err_invalid_setfeat(uint32_t dw10) "invalid set features, dw10=0x%"PRIx32"" +nvme_err_startfail_cq(void) "nvme_start_ctrl failed because there are non-admin completion queues" +nvme_err_startfail_sq(void) "nvme_start_ctrl failed because there are non-admin submission queues" +nvme_err_startfail_nbarasq(void) "nvme_start_ctrl failed because the admin submission queue address is null" +nvme_err_startfail_nbaracq(void) "nvme_start_ctrl failed because the admin completion queue address is null" +nvme_err_startfail_asq_misaligned(uint64_t addr) "nvme_start_ctrl failed because the admin submission queue address is misaligned: 0x%"PRIx64"" +nvme_err_startfail_acq_misaligned(uint64_t addr) "nvme_start_ctrl failed because the admin completion queue address is misaligned: 0x%"PRIx64"" +nvme_err_startfail_page_too_small(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the page size is too small: log2size=%u, min=%u" +nvme_err_startfail_page_too_large(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the page size is too large: log2size=%u, max=%u" +nvme_err_startfail_cqent_too_small(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the completion queue entry size is too small: log2size=%u, min=%u" +nvme_err_startfail_cqent_too_large(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the completion queue entry size is too large: log2size=%u, max=%u" +nvme_err_startfail_sqent_too_small(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the submission queue entry size is too small: log2size=%u, min=%u" +nvme_err_startfail_sqent_too_large(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the submission queue entry size is too large: log2size=%u, max=%u" +nvme_err_startfail_asqent_sz_zero(void) "nvme_start_ctrl failed because the admin submission queue size is zero" +nvme_err_startfail_acqent_sz_zero(void) "nvme_start_ctrl failed because the admin completion queue size is zero" +nvme_err_startfail(void) "setting controller enable bit failed" + +# Traces for undefined behavior +nvme_ub_mmiowr_misaligned32(uint64_t offset) "MMIO write not 32-bit aligned, offset=0x%"PRIx64"" +nvme_ub_mmiowr_toosmall(uint64_t offset, unsigned size) "MMIO write smaller than 32 bits, offset=0x%"PRIx64", size=%u" +nvme_ub_mmiowr_intmask_with_msix(void) "undefined access to interrupt mask set when MSI-X is enabled" +nvme_ub_mmiowr_ro_csts(void) "attempted to set a read only bit of controller status" +nvme_ub_mmiowr_ssreset_w1c_unsupported(void) "attempted to W1C CSTS.NSSRO but CAP.NSSRS is zero (not supported)" +nvme_ub_mmiowr_ssreset_unsupported(void) "attempted NVM subsystem reset but CAP.NSSRS is zero (not supported)" +nvme_ub_mmiowr_cmbloc_reserved(void) "invalid write to reserved CMBLOC when CMBSZ is zero, ignored" +nvme_ub_mmiowr_cmbsz_readonly(void) "invalid write to read only CMBSZ, ignored" +nvme_ub_mmiowr_invalid(uint64_t offset, uint64_t data) "invalid MMIO write, offset=0x%"PRIx64", data=0x%"PRIx64"" +nvme_ub_mmiord_misaligned32(uint64_t offset) "MMIO read not 32-bit aligned, offset=0x%"PRIx64"" +nvme_ub_mmiord_toosmall(uint64_t offset) "MMIO read smaller than 32-bits, offset=0x%"PRIx64"" +nvme_ub_mmiord_invalid_ofs(uint64_t offset) "MMIO read beyond last register, offset=0x%"PRIx64", returning 0" +nvme_ub_db_wr_misaligned(uint64_t offset) "doorbell write not 32-bit aligned, offset=0x%"PRIx64", ignoring" +nvme_ub_db_wr_invalid_cq(uint32_t qid) "completion queue doorbell write for nonexistent queue, cqid=%"PRIu32", ignoring" +nvme_ub_db_wr_invalid_cqhead(uint32_t qid, uint16_t new_head) "completion queue doorbell write value beyond queue size, cqid=%"PRIu32", new_head=%"PRIu16", ignoring" +nvme_ub_db_wr_invalid_sq(uint32_t qid) "submission queue doorbell write for nonexistent queue, sqid=%"PRIu32", ignoring" +nvme_ub_db_wr_invalid_sqtail(uint32_t qid, uint16_t new_tail) "submission queue doorbell write value beyond queue size, sqid=%"PRIu32", new_head=%"PRIu16", ignoring" + # hw/block/xen_disk.c xen_disk_alloc(char *name) "%s" xen_disk_init(char *name) "%s" -- 2.13.6 From: Fam Zheng <famz@redhat.com> Management tools create overlays of running guests with qemu-img: $ qemu-img create -b /image/in/use.qcow2 -f qcow2 /overlay/image.qcow2 but this doesn't work anymore due to image locking: qemu-img: /overlay/image.qcow2: Failed to get shared "write" lock Is another process using the image? Could not open backing image to determine size. Use the force share option to allow this use case again. Cc: qemu-stable@nongnu.org Signed-off-by: Fam Zheng <famz@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ void bdrv_img_create(const char *filename, const char *fmt, back_flags = flags; back_flags &= ~(BDRV_O_RDWR | BDRV_O_SNAPSHOT | BDRV_O_NO_BACKING); + backing_options = qdict_new(); if (backing_fmt) { - backing_options = qdict_new(); qdict_put_str(backing_options, "driver", backing_fmt); } + qdict_put_bool(backing_options, BDRV_OPT_FORCE_SHARE, true); bs = bdrv_open(full_backing, NULL, backing_options, back_flags, &local_err); -- 2.13.6 From: Thomas Huth <thuth@redhat.com> It's not working anymore since QEMU v1.3.0 - time to remove it now. Signed-off-by: Thomas Huth <thuth@redhat.com> Reviewed-by: John Snow <jsnow@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- blockdev.c | 11 ----------- qemu-doc.texi | 6 ------ 2 files changed, 17 deletions(-) diff --git a/blockdev.c b/blockdev.c index XXXXXXX..XXXXXXX 100644 --- a/blockdev.c +++ b/blockdev.c @@ -XXX,XX +XXX,XX @@ QemuOptsList qemu_legacy_drive_opts = { .type = QEMU_OPT_STRING, .help = "chs translation (auto, lba, none)", },{ - .name = "boot", - .type = QEMU_OPT_BOOL, - .help = "(deprecated, ignored)", - },{ .name = "addr", .type = QEMU_OPT_STRING, .help = "pci address (virtio only)", @@ -XXX,XX +XXX,XX @@ DriveInfo *drive_new(QemuOpts *all_opts, BlockInterfaceType block_default_type) goto fail; } - /* Deprecated option boot=[on|off] */ - if (qemu_opt_get(legacy_opts, "boot") != NULL) { - fprintf(stderr, "qemu-kvm: boot=on|off is deprecated and will be " - "ignored. Future versions will reject this parameter. Please " - "update your scripts.\n"); - } - /* Other deprecated options */ if (!qtest_enabled()) { for (i = 0; i < ARRAY_SIZE(deprecated); i++) { diff --git a/qemu-doc.texi b/qemu-doc.texi index XXXXXXX..XXXXXXX 100644 --- a/qemu-doc.texi +++ b/qemu-doc.texi @@ -XXX,XX +XXX,XX @@ deprecated. @section System emulator command line arguments -@subsection -drive boot=on|off (since 1.3.0) - -The ``boot=on|off'' option to the ``-drive'' argument is -ignored. Applications should use the ``bootindex=N'' parameter -to set an absolute ordering between devices instead. - @subsection -tdf (since 1.3.0) The ``-tdf'' argument is ignored. The behaviour implemented -- 2.13.6 From: Thomas Huth <thuth@redhat.com> It's been marked as deprecated since QEMU v2.10.0, and so far nobody complained that we should keep it, so let's remove this legacy option now to simplify the code quite a bit. Signed-off-by: Thomas Huth <thuth@redhat.com> Reviewed-by: John Snow <jsnow@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- vl.c | 86 ++------------------------------------------------------- qemu-doc.texi | 8 ------ qemu-options.hx | 19 ++----------- 3 files changed, 4 insertions(+), 109 deletions(-) diff --git a/vl.c b/vl.c index XXXXXXX..XXXXXXX 100644 --- a/vl.c +++ b/vl.c @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp) const char *boot_order = NULL; const char *boot_once = NULL; DisplayState *ds; - int cyls, heads, secs, translation; QemuOpts *opts, *machine_opts; - QemuOpts *hda_opts = NULL, *icount_opts = NULL, *accel_opts = NULL; + QemuOpts *icount_opts = NULL, *accel_opts = NULL; QemuOptsList *olist; int optind; const char *optarg; @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp) cpu_model = NULL; snapshot = 0; - cyls = heads = secs = 0; - translation = BIOS_ATA_TRANSLATION_AUTO; nb_nics = 0; @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp) if (optind >= argc) break; if (argv[optind][0] != '-') { - hda_opts = drive_add(IF_DEFAULT, 0, argv[optind++], HD_OPTS); + drive_add(IF_DEFAULT, 0, argv[optind++], HD_OPTS); } else { const QEMUOption *popt; @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp) cpu_model = optarg; break; case QEMU_OPTION_hda: - { - char buf[256]; - if (cyls == 0) - snprintf(buf, sizeof(buf), "%s", HD_OPTS); - else - snprintf(buf, sizeof(buf), - "%s,cyls=%d,heads=%d,secs=%d%s", - HD_OPTS , cyls, heads, secs, - translation == BIOS_ATA_TRANSLATION_LBA ? - ",trans=lba" : - translation == BIOS_ATA_TRANSLATION_NONE ? - ",trans=none" : ""); - drive_add(IF_DEFAULT, 0, optarg, buf); - break; - } case QEMU_OPTION_hdb: case QEMU_OPTION_hdc: case QEMU_OPTION_hdd: @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp) case QEMU_OPTION_snapshot: snapshot = 1; break; - case QEMU_OPTION_hdachs: - { - const char *p; - p = optarg; - cyls = strtol(p, (char **)&p, 0); - if (cyls < 1 || cyls > 16383) - goto chs_fail; - if (*p != ',') - goto chs_fail; - p++; - heads = strtol(p, (char **)&p, 0); - if (heads < 1 || heads > 16) - goto chs_fail; - if (*p != ',') - goto chs_fail; - p++; - secs = strtol(p, (char **)&p, 0); - if (secs < 1 || secs > 63) - goto chs_fail; - if (*p == ',') { - p++; - if (!strcmp(p, "large")) { - translation = BIOS_ATA_TRANSLATION_LARGE; - } else if (!strcmp(p, "rechs")) { - translation = BIOS_ATA_TRANSLATION_RECHS; - } else if (!strcmp(p, "none")) { - translation = BIOS_ATA_TRANSLATION_NONE; - } else if (!strcmp(p, "lba")) { - translation = BIOS_ATA_TRANSLATION_LBA; - } else if (!strcmp(p, "auto")) { - translation = BIOS_ATA_TRANSLATION_AUTO; - } else { - goto chs_fail; - } - } else if (*p != '\0') { - chs_fail: - error_report("invalid physical CHS format"); - exit(1); - } - if (hda_opts != NULL) { - qemu_opt_set_number(hda_opts, "cyls", cyls, - &error_abort); - qemu_opt_set_number(hda_opts, "heads", heads, - &error_abort); - qemu_opt_set_number(hda_opts, "secs", secs, - &error_abort); - if (translation == BIOS_ATA_TRANSLATION_LARGE) { - qemu_opt_set(hda_opts, "trans", "large", - &error_abort); - } else if (translation == BIOS_ATA_TRANSLATION_RECHS) { - qemu_opt_set(hda_opts, "trans", "rechs", - &error_abort); - } else if (translation == BIOS_ATA_TRANSLATION_LBA) { - qemu_opt_set(hda_opts, "trans", "lba", - &error_abort); - } else if (translation == BIOS_ATA_TRANSLATION_NONE) { - qemu_opt_set(hda_opts, "trans", "none", - &error_abort); - } - } - } - error_report("'-hdachs' is deprecated, please use '-device" - " ide-hd,cyls=c,heads=h,secs=s,...' instead"); - break; case QEMU_OPTION_numa: opts = qemu_opts_parse_noisily(qemu_find_opts("numa"), optarg, true); diff --git a/qemu-doc.texi b/qemu-doc.texi index XXXXXXX..XXXXXXX 100644 --- a/qemu-doc.texi +++ b/qemu-doc.texi @@ -XXX,XX +XXX,XX @@ The ``--net dump'' argument is now replaced with the ``-object filter-dump'' argument which works in combination with the modern ``-netdev`` backends instead. -@subsection -hdachs (since 2.10.0) - -The ``-hdachs'' argument is now a synonym for setting -the ``cyls'', ``heads'', ``secs'', and ``trans'' properties -on the ``ide-hd'' device using the ``-device'' argument. -The new syntax allows different settings to be provided -per disk. - @subsection -usbdevice (since 2.10.0) The ``-usbdevice DEV'' argument is now a synonym for setting diff --git a/qemu-options.hx b/qemu-options.hx index XXXXXXX..XXXXXXX 100644 --- a/qemu-options.hx +++ b/qemu-options.hx @@ -XXX,XX +XXX,XX @@ of available connectors of a given interface type. @item media=@var{media} This option defines the type of the media: disk or cdrom. @item cyls=@var{c},heads=@var{h},secs=@var{s}[,trans=@var{t}] -These options have the same definition as they have in @option{-hdachs}. -These parameters are deprecated, use the corresponding parameters +Force disk physical geometry and the optional BIOS translation (trans=none or +lba). These parameters are deprecated, use the corresponding parameters of @code{-device} instead. @item snapshot=@var{snapshot} @var{snapshot} is "on" or "off" and controls snapshot mode for the given drive @@ -XXX,XX +XXX,XX @@ the raw disk image you use is not written back. You can however force the write back by pressing @key{C-a s} (@pxref{disk_images}). ETEXI -DEF("hdachs", HAS_ARG, QEMU_OPTION_hdachs, \ - "-hdachs c,h,s[,t]\n" \ - " force hard disk 0 physical geometry and the optional BIOS\n" \ - " translation (t=none or lba) (usually QEMU can guess them)\n", - QEMU_ARCH_ALL) -STEXI -@item -hdachs @var{c},@var{h},@var{s},[,@var{t}] -@findex -hdachs -Force hard disk 0 physical geometry (1 <= @var{c} <= 16383, 1 <= -@var{h} <= 16, 1 <= @var{s} <= 63) and optionally force the BIOS -translation mode (@var{t}=none, lba or auto). Usually QEMU can guess -all those parameters. This option is deprecated, please use -@code{-device ide-hd,cyls=c,heads=h,secs=s,...} instead. -ETEXI - DEF("fsdev", HAS_ARG, QEMU_OPTION_fsdev, "-fsdev fsdriver,id=id[,path=path,][security_model={mapped-xattr|mapped-file|passthrough|none}]\n" " [,writeout=immediate][,readonly][,socket=socket|sock_fd=sock_fd][,fmode=fmode][,dmode=dmode]\n" -- 2.13.6 From: Thomas Huth <thuth@redhat.com> Looks like we forgot to announce the deprecation of these options in the corresponding chapter of the qemu-doc text, so let's do that now. Signed-off-by: Thomas Huth <thuth@redhat.com> Reviewed-by: John Snow <jsnow@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- qemu-doc.texi | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/qemu-doc.texi b/qemu-doc.texi index XXXXXXX..XXXXXXX 100644 --- a/qemu-doc.texi +++ b/qemu-doc.texi @@ -XXX,XX +XXX,XX @@ longer be directly supported in QEMU. The ``-drive if=scsi'' argument is replaced by the the ``-device BUS-TYPE'' argument combined with ``-drive if=none''. +@subsection -drive cyls=...,heads=...,secs=...,trans=... (since 2.10.0) + +The drive geometry arguments are replaced by the the geometry arguments +that can be specified with the ``-device'' parameter. + +@subsection -drive serial=... (since 2.10.0) + +The drive serial argument is replaced by the the serial argument +that can be specified with the ``-device'' parameter. + +@subsection -drive addr=... (since 2.10.0) + +The drive addr argument is replaced by the the addr argument +that can be specified with the ``-device'' parameter. + @subsection -net dump (since 2.10.0) The ``--net dump'' argument is now replaced with the -- 2.13.6 From: Fam Zheng <famz@redhat.com> Signed-off-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block_int.h | 1 - block/io.c | 18 ------------------ 2 files changed, 19 deletions(-) diff --git a/include/block/block_int.h b/include/block/block_int.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block_int.h +++ b/include/block/block_int.h @@ -XXX,XX +XXX,XX @@ bool blk_dev_is_tray_open(BlockBackend *blk); bool blk_dev_is_medium_locked(BlockBackend *blk); void bdrv_set_dirty(BlockDriverState *bs, int64_t offset, int64_t bytes); -bool bdrv_requests_pending(BlockDriverState *bs); void bdrv_clear_dirty_bitmap(BdrvDirtyBitmap *bitmap, HBitmap **out); void bdrv_undo_clear_dirty_bitmap(BdrvDirtyBitmap *bitmap, HBitmap *in); diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ void bdrv_disable_copy_on_read(BlockDriverState *bs) assert(old >= 1); } -/* Check if any requests are in-flight (including throttled requests) */ -bool bdrv_requests_pending(BlockDriverState *bs) -{ - BdrvChild *child; - - if (atomic_read(&bs->in_flight)) { - return true; - } - - QLIST_FOREACH(child, &bs->children, next) { - if (bdrv_requests_pending(child->bs)) { - return true; - } - } - - return false; -} - typedef struct { Coroutine *co; BlockDriverState *bs; -- 2.13.6 Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> --- block/io.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void) BdrvNextIterator it; GSList *aio_ctxs = NULL, *ctx; + /* BDRV_POLL_WHILE() for a node can only be called from its own I/O thread + * or the main loop AioContext. We potentially use BDRV_POLL_WHILE() on + * nodes in several different AioContexts, so make sure we're in the main + * context. */ + assert(qemu_get_current_aio_context() == qemu_get_aio_context()); + block_job_pause_all(); for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) { -- 2.13.6 bdrv_drained_begin() doesn't increase bs->quiesce_counter recursively and also doesn't notify other parent nodes of children, which both means that the child nodes are not actually drained, and bdrv_drained_begin() is providing useful functionality only on a single node. To keep things consistent, we also shouldn't call the block driver callbacks recursively. A proper recursive drain version that provides an actually working drained section for child nodes will be introduced later. Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> --- block/io.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_drain_invoke_entry(void *opaque) } /* Recursively call BlockDriver.bdrv_co_drain_begin/end callbacks */ -static void bdrv_drain_invoke(BlockDriverState *bs, bool begin) +static void bdrv_drain_invoke(BlockDriverState *bs, bool begin, bool recursive) { BdrvChild *child, *tmp; BdrvCoDrainData data = { .bs = bs, .done = false, .begin = begin}; @@ -XXX,XX +XXX,XX @@ static void bdrv_drain_invoke(BlockDriverState *bs, bool begin) bdrv_coroutine_enter(bs, data.co); BDRV_POLL_WHILE(bs, !data.done); - QLIST_FOREACH_SAFE(child, &bs->children, next, tmp) { - bdrv_drain_invoke(child->bs, begin); + if (recursive) { + QLIST_FOREACH_SAFE(child, &bs->children, next, tmp) { + bdrv_drain_invoke(child->bs, begin, true); + } } } @@ -XXX,XX +XXX,XX @@ void bdrv_drained_begin(BlockDriverState *bs) bdrv_parent_drained_begin(bs); } - bdrv_drain_invoke(bs, true); + bdrv_drain_invoke(bs, true, false); bdrv_drain_recurse(bs); } @@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs) } /* Re-enable things in child-to-parent order */ - bdrv_drain_invoke(bs, false); + bdrv_drain_invoke(bs, false, false); bdrv_parent_drained_end(bs); aio_enable_external(bdrv_get_aio_context(bs)); } @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void) aio_context_acquire(aio_context); aio_disable_external(aio_context); bdrv_parent_drained_begin(bs); - bdrv_drain_invoke(bs, true); + bdrv_drain_invoke(bs, true, true); aio_context_release(aio_context); if (!g_slist_find(aio_ctxs, aio_context)) { @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void) /* Re-enable things in child-to-parent order */ aio_context_acquire(aio_context); - bdrv_drain_invoke(bs, false); + bdrv_drain_invoke(bs, false, true); bdrv_parent_drained_end(bs); aio_enable_external(aio_context); aio_context_release(aio_context); -- 2.13.6 The existing test is for bdrv_drain_all_begin/end() only. Generalise the test case so that it can be run for the other variants as well. At the moment this is only bdrv_drain_begin/end(), but in a while, we'll add another one. Also, add a backing file to the test node to test whether the operations work recursively. Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/test-bdrv-drain.c | 69 ++++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 62 insertions(+), 7 deletions(-) diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c index XXXXXXX..XXXXXXX 100644 --- a/tests/test-bdrv-drain.c +++ b/tests/test-bdrv-drain.c @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_test = { .bdrv_co_drain_begin = bdrv_test_co_drain_begin, .bdrv_co_drain_end = bdrv_test_co_drain_end, + + .bdrv_child_perm = bdrv_format_default_perms, }; static void aio_ret_cb(void *opaque, int ret) @@ -XXX,XX +XXX,XX @@ static void aio_ret_cb(void *opaque, int ret) *aio_ret = ret; } -static void test_drv_cb_drain_all(void) +enum drain_type { + BDRV_DRAIN_ALL, + BDRV_DRAIN, +}; + +static void do_drain_begin(enum drain_type drain_type, BlockDriverState *bs) +{ + switch (drain_type) { + case BDRV_DRAIN_ALL: bdrv_drain_all_begin(); break; + case BDRV_DRAIN: bdrv_drained_begin(bs); break; + default: g_assert_not_reached(); + } +} + +static void do_drain_end(enum drain_type drain_type, BlockDriverState *bs) +{ + switch (drain_type) { + case BDRV_DRAIN_ALL: bdrv_drain_all_end(); break; + case BDRV_DRAIN: bdrv_drained_end(bs); break; + default: g_assert_not_reached(); + } +} + +static void test_drv_cb_common(enum drain_type drain_type, bool recursive) { BlockBackend *blk; - BlockDriverState *bs; - BDRVTestState *s; + BlockDriverState *bs, *backing; + BDRVTestState *s, *backing_s; BlockAIOCB *acb; int aio_ret; @@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain_all(void) s = bs->opaque; blk_insert_bs(blk, bs, &error_abort); + backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort); + backing_s = backing->opaque; + bdrv_set_backing_hd(bs, backing, &error_abort); + /* Simple bdrv_drain_all_begin/end pair, check that CBs are called */ g_assert_cmpint(s->drain_count, ==, 0); - bdrv_drain_all_begin(); + g_assert_cmpint(backing_s->drain_count, ==, 0); + + do_drain_begin(drain_type, bs); + g_assert_cmpint(s->drain_count, ==, 1); - bdrv_drain_all_end(); + g_assert_cmpint(backing_s->drain_count, ==, !!recursive); + + do_drain_end(drain_type, bs); + g_assert_cmpint(s->drain_count, ==, 0); + g_assert_cmpint(backing_s->drain_count, ==, 0); /* Now do the same while a request is pending */ aio_ret = -EINPROGRESS; @@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain_all(void) g_assert_cmpint(aio_ret, ==, -EINPROGRESS); g_assert_cmpint(s->drain_count, ==, 0); - bdrv_drain_all_begin(); + g_assert_cmpint(backing_s->drain_count, ==, 0); + + do_drain_begin(drain_type, bs); + g_assert_cmpint(aio_ret, ==, 0); g_assert_cmpint(s->drain_count, ==, 1); - bdrv_drain_all_end(); + g_assert_cmpint(backing_s->drain_count, ==, !!recursive); + + do_drain_end(drain_type, bs); + g_assert_cmpint(s->drain_count, ==, 0); + g_assert_cmpint(backing_s->drain_count, ==, 0); + bdrv_unref(backing); bdrv_unref(bs); blk_unref(blk); } +static void test_drv_cb_drain_all(void) +{ + test_drv_cb_common(BDRV_DRAIN_ALL, true); +} + +static void test_drv_cb_drain(void) +{ + test_drv_cb_common(BDRV_DRAIN, false); +} + int main(int argc, char **argv) { bdrv_init(); @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv) g_test_init(&argc, &argv, NULL); g_test_add_func("/bdrv-drain/driver-cb/drain_all", test_drv_cb_drain_all); + g_test_add_func("/bdrv-drain/driver-cb/drain", test_drv_cb_drain); return g_test_run(); } -- 2.13.6 This is currently only working correctly for bdrv_drain(), not for bdrv_drain_all(). Leave a comment for the drain_all case, we'll address it later. Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/test-bdrv-drain.c | 45 +++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 45 insertions(+) diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c index XXXXXXX..XXXXXXX 100644 --- a/tests/test-bdrv-drain.c +++ b/tests/test-bdrv-drain.c @@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain(void) test_drv_cb_common(BDRV_DRAIN, false); } +static void test_quiesce_common(enum drain_type drain_type, bool recursive) +{ + BlockBackend *blk; + BlockDriverState *bs, *backing; + + blk = blk_new(BLK_PERM_ALL, BLK_PERM_ALL); + bs = bdrv_new_open_driver(&bdrv_test, "test-node", BDRV_O_RDWR, + &error_abort); + blk_insert_bs(blk, bs, &error_abort); + + backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort); + bdrv_set_backing_hd(bs, backing, &error_abort); + + g_assert_cmpint(bs->quiesce_counter, ==, 0); + g_assert_cmpint(backing->quiesce_counter, ==, 0); + + do_drain_begin(drain_type, bs); + + g_assert_cmpint(bs->quiesce_counter, ==, 1); + g_assert_cmpint(backing->quiesce_counter, ==, !!recursive); + + do_drain_end(drain_type, bs); + + g_assert_cmpint(bs->quiesce_counter, ==, 0); + g_assert_cmpint(backing->quiesce_counter, ==, 0); + + bdrv_unref(backing); + bdrv_unref(bs); + blk_unref(blk); +} + +static void test_quiesce_drain_all(void) +{ + // XXX drain_all doesn't quiesce + //test_quiesce_common(BDRV_DRAIN_ALL, true); +} + +static void test_quiesce_drain(void) +{ + test_quiesce_common(BDRV_DRAIN, false); +} + int main(int argc, char **argv) { bdrv_init(); @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv) g_test_add_func("/bdrv-drain/driver-cb/drain_all", test_drv_cb_drain_all); g_test_add_func("/bdrv-drain/driver-cb/drain", test_drv_cb_drain); + g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all); + g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain); + return g_test_run(); } -- 2.13.6 Block jobs already paused themselves when their main BlockBackend entered a drained section. This is not good enough: We also want to pause a block job and may not submit new requests if, for example, the mirror target node should be drained. This implements .drained_begin/end callbacks in child_job in order to consider all block nodes related to the job, and removes the BlockBackend callbacks which are unnecessary now because the root of the job main BlockBackend is always referenced with a child_job, too. Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- blockjob.c | 22 +++++++++------------- 1 file changed, 9 insertions(+), 13 deletions(-) diff --git a/blockjob.c b/blockjob.c index XXXXXXX..XXXXXXX 100644 --- a/blockjob.c +++ b/blockjob.c @@ -XXX,XX +XXX,XX @@ static char *child_job_get_parent_desc(BdrvChild *c) job->id); } -static const BdrvChildRole child_job = { - .get_parent_desc = child_job_get_parent_desc, - .stay_at_node = true, -}; - -static void block_job_drained_begin(void *opaque) +static void child_job_drained_begin(BdrvChild *c) { - BlockJob *job = opaque; + BlockJob *job = c->opaque; block_job_pause(job); } -static void block_job_drained_end(void *opaque) +static void child_job_drained_end(BdrvChild *c) { - BlockJob *job = opaque; + BlockJob *job = c->opaque; block_job_resume(job); } -static const BlockDevOps block_job_dev_ops = { - .drained_begin = block_job_drained_begin, - .drained_end = block_job_drained_end, +static const BdrvChildRole child_job = { + .get_parent_desc = child_job_get_parent_desc, + .drained_begin = child_job_drained_begin, + .drained_end = child_job_drained_end, + .stay_at_node = true, }; void block_job_remove_all_bdrv(BlockJob *job) @@ -XXX,XX +XXX,XX @@ void *block_job_create(const char *job_id, const BlockJobDriver *driver, block_job_add_bdrv(job, "main node", bs, 0, BLK_PERM_ALL, &error_abort); bs->job = job; - blk_set_dev_ops(blk, &block_job_dev_ops, job); bdrv_op_unblock(bs, BLOCK_OP_TYPE_DATAPLANE, job->blocker); QLIST_INSERT_HEAD(&block_jobs, job, job_list); -- 2.13.6 Block jobs must be paused if any of the involved nodes are drained. Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/test-bdrv-drain.c | 121 ++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 121 insertions(+) diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c index XXXXXXX..XXXXXXX 100644 --- a/tests/test-bdrv-drain.c +++ b/tests/test-bdrv-drain.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "block/block.h" +#include "block/blockjob_int.h" #include "sysemu/block-backend.h" #include "qapi/error.h" @@ -XXX,XX +XXX,XX @@ static void test_quiesce_drain(void) test_quiesce_common(BDRV_DRAIN, false); } + +typedef struct TestBlockJob { + BlockJob common; + bool should_complete; +} TestBlockJob; + +static void test_job_completed(BlockJob *job, void *opaque) +{ + block_job_completed(job, 0); +} + +static void coroutine_fn test_job_start(void *opaque) +{ + TestBlockJob *s = opaque; + + while (!s->should_complete) { + block_job_sleep_ns(&s->common, 100000); + } + + block_job_defer_to_main_loop(&s->common, test_job_completed, NULL); +} + +static void test_job_complete(BlockJob *job, Error **errp) +{ + TestBlockJob *s = container_of(job, TestBlockJob, common); + s->should_complete = true; +} + +BlockJobDriver test_job_driver = { + .instance_size = sizeof(TestBlockJob), + .start = test_job_start, + .complete = test_job_complete, +}; + +static void test_blockjob_common(enum drain_type drain_type) +{ + BlockBackend *blk_src, *blk_target; + BlockDriverState *src, *target; + BlockJob *job; + int ret; + + src = bdrv_new_open_driver(&bdrv_test, "source", BDRV_O_RDWR, + &error_abort); + blk_src = blk_new(BLK_PERM_ALL, BLK_PERM_ALL); + blk_insert_bs(blk_src, src, &error_abort); + + target = bdrv_new_open_driver(&bdrv_test, "target", BDRV_O_RDWR, + &error_abort); + blk_target = blk_new(BLK_PERM_ALL, BLK_PERM_ALL); + blk_insert_bs(blk_target, target, &error_abort); + + job = block_job_create("job0", &test_job_driver, src, 0, BLK_PERM_ALL, 0, + 0, NULL, NULL, &error_abort); + block_job_add_bdrv(job, "target", target, 0, BLK_PERM_ALL, &error_abort); + block_job_start(job); + + g_assert_cmpint(job->pause_count, ==, 0); + g_assert_false(job->paused); + g_assert_false(job->busy); /* We're in block_job_sleep_ns() */ + + do_drain_begin(drain_type, src); + + if (drain_type == BDRV_DRAIN_ALL) { + /* bdrv_drain_all() drains both src and target, and involves an + * additional block_job_pause_all() */ + g_assert_cmpint(job->pause_count, ==, 3); + } else { + g_assert_cmpint(job->pause_count, ==, 1); + } + /* XXX We don't wait until the job is actually paused. Is this okay? */ + /* g_assert_true(job->paused); */ + g_assert_false(job->busy); /* The job is paused */ + + do_drain_end(drain_type, src); + + g_assert_cmpint(job->pause_count, ==, 0); + g_assert_false(job->paused); + g_assert_false(job->busy); /* We're in block_job_sleep_ns() */ + + do_drain_begin(drain_type, target); + + if (drain_type == BDRV_DRAIN_ALL) { + /* bdrv_drain_all() drains both src and target, and involves an + * additional block_job_pause_all() */ + g_assert_cmpint(job->pause_count, ==, 3); + } else { + g_assert_cmpint(job->pause_count, ==, 1); + } + /* XXX We don't wait until the job is actually paused. Is this okay? */ + /* g_assert_true(job->paused); */ + g_assert_false(job->busy); /* The job is paused */ + + do_drain_end(drain_type, target); + + g_assert_cmpint(job->pause_count, ==, 0); + g_assert_false(job->paused); + g_assert_false(job->busy); /* We're in block_job_sleep_ns() */ + + ret = block_job_complete_sync(job, &error_abort); + g_assert_cmpint(ret, ==, 0); + + blk_unref(blk_src); + blk_unref(blk_target); + bdrv_unref(src); + bdrv_unref(target); +} + +static void test_blockjob_drain_all(void) +{ + test_blockjob_common(BDRV_DRAIN_ALL); +} + +static void test_blockjob_drain(void) +{ + test_blockjob_common(BDRV_DRAIN); +} + int main(int argc, char **argv) { bdrv_init(); @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv) g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all); g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain); + g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all); + g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain); + return g_test_run(); } -- 2.13.6 Block jobs are already paused using the BdrvChildRole drain callbacks, so we don't need an additional block_job_pause_all() call. Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/io.c | 4 ---- tests/test-bdrv-drain.c | 10 ++++------ 2 files changed, 4 insertions(+), 10 deletions(-) diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void) * context. */ assert(qemu_get_current_aio_context() == qemu_get_aio_context()); - block_job_pause_all(); - for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) { AioContext *aio_context = bdrv_get_aio_context(bs); @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void) aio_enable_external(aio_context); aio_context_release(aio_context); } - - block_job_resume_all(); } void bdrv_drain_all(void) diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c index XXXXXXX..XXXXXXX 100644 --- a/tests/test-bdrv-drain.c +++ b/tests/test-bdrv-drain.c @@ -XXX,XX +XXX,XX @@ static void test_blockjob_common(enum drain_type drain_type) do_drain_begin(drain_type, src); if (drain_type == BDRV_DRAIN_ALL) { - /* bdrv_drain_all() drains both src and target, and involves an - * additional block_job_pause_all() */ - g_assert_cmpint(job->pause_count, ==, 3); + /* bdrv_drain_all() drains both src and target */ + g_assert_cmpint(job->pause_count, ==, 2); } else { g_assert_cmpint(job->pause_count, ==, 1); } @@ -XXX,XX +XXX,XX @@ static void test_blockjob_common(enum drain_type drain_type) do_drain_begin(drain_type, target); if (drain_type == BDRV_DRAIN_ALL) { - /* bdrv_drain_all() drains both src and target, and involves an - * additional block_job_pause_all() */ - g_assert_cmpint(job->pause_count, ==, 3); + /* bdrv_drain_all() drains both src and target */ + g_assert_cmpint(job->pause_count, ==, 2); } else { g_assert_cmpint(job->pause_count, ==, 1); } -- 2.13.6 bdrv_do_drained_begin() restricts the call of parent callbacks and aio_disable_external() to the outermost drain section, but the block driver callbacks are always called. bdrv_do_drained_end() must match this behaviour, otherwise nodes stay drained even if begin/end calls were balanced. Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/io.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ void bdrv_drained_begin(BlockDriverState *bs) void bdrv_drained_end(BlockDriverState *bs) { + int old_quiesce_counter; + if (qemu_in_coroutine()) { bdrv_co_yield_to_drain(bs, false); return; } assert(bs->quiesce_counter > 0); - if (atomic_fetch_dec(&bs->quiesce_counter) > 1) { - return; - } + old_quiesce_counter = atomic_fetch_dec(&bs->quiesce_counter); /* Re-enable things in child-to-parent order */ bdrv_drain_invoke(bs, false, false); - bdrv_parent_drained_end(bs); - aio_enable_external(bdrv_get_aio_context(bs)); + if (old_quiesce_counter == 1) { + bdrv_parent_drained_end(bs); + aio_enable_external(bdrv_get_aio_context(bs)); + } } /* -- 2.13.6 Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/test-bdrv-drain.c | 57 +++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 57 insertions(+) diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c index XXXXXXX..XXXXXXX 100644 --- a/tests/test-bdrv-drain.c +++ b/tests/test-bdrv-drain.c @@ -XXX,XX +XXX,XX @@ static void aio_ret_cb(void *opaque, int ret) enum drain_type { BDRV_DRAIN_ALL, BDRV_DRAIN, + DRAIN_TYPE_MAX, }; static void do_drain_begin(enum drain_type drain_type, BlockDriverState *bs) @@ -XXX,XX +XXX,XX @@ static void test_quiesce_drain(void) test_quiesce_common(BDRV_DRAIN, false); } +static void test_nested(void) +{ + BlockBackend *blk; + BlockDriverState *bs, *backing; + BDRVTestState *s, *backing_s; + enum drain_type outer, inner; + + blk = blk_new(BLK_PERM_ALL, BLK_PERM_ALL); + bs = bdrv_new_open_driver(&bdrv_test, "test-node", BDRV_O_RDWR, + &error_abort); + s = bs->opaque; + blk_insert_bs(blk, bs, &error_abort); + + backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort); + backing_s = backing->opaque; + bdrv_set_backing_hd(bs, backing, &error_abort); + + for (outer = 0; outer < DRAIN_TYPE_MAX; outer++) { + for (inner = 0; inner < DRAIN_TYPE_MAX; inner++) { + /* XXX bdrv_drain_all() doesn't increase the quiesce_counter */ + int bs_quiesce = (outer != BDRV_DRAIN_ALL) + + (inner != BDRV_DRAIN_ALL); + int backing_quiesce = 0; + int backing_cb_cnt = (outer != BDRV_DRAIN) + + (inner != BDRV_DRAIN); + + g_assert_cmpint(bs->quiesce_counter, ==, 0); + g_assert_cmpint(backing->quiesce_counter, ==, 0); + g_assert_cmpint(s->drain_count, ==, 0); + g_assert_cmpint(backing_s->drain_count, ==, 0); + + do_drain_begin(outer, bs); + do_drain_begin(inner, bs); + + g_assert_cmpint(bs->quiesce_counter, ==, bs_quiesce); + g_assert_cmpint(backing->quiesce_counter, ==, backing_quiesce); + g_assert_cmpint(s->drain_count, ==, 2); + g_assert_cmpint(backing_s->drain_count, ==, backing_cb_cnt); + + do_drain_end(inner, bs); + do_drain_end(outer, bs); + + g_assert_cmpint(bs->quiesce_counter, ==, 0); + g_assert_cmpint(backing->quiesce_counter, ==, 0); + g_assert_cmpint(s->drain_count, ==, 0); + g_assert_cmpint(backing_s->drain_count, ==, 0); + } + } + + bdrv_unref(backing); + bdrv_unref(bs); + blk_unref(blk); +} + typedef struct TestBlockJob { BlockJob common; @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv) g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all); g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain); + g_test_add_func("/bdrv-drain/nested", test_nested); + g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all); g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain); -- 2.13.6 This is in preparation for subtree drains, i.e. drained sections that affect not only a single node, but recursively all child nodes, too. Calling the parent callbacks for drain is pointless when we just came from that parent node recursively and leads to multiple increases of bs->quiesce_counter in a single drain call. Don't do it. In order for this to work correctly, the parent callback must be called for every bdrv_drain_begin/end() call, not only for the outermost one: If we have a node N with two parents A and B, recursive draining of A should cause the quiesce_counter of B to increase because its child N is drained independently of B. If now B is recursively drained, too, A must increase its quiesce_counter because N is drained independently of A only now, even if N is going from quiesce_counter 1 to 2. Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block.h | 4 ++-- block.c | 13 +++++++++---- block/io.c | 47 ++++++++++++++++++++++++++++++++++------------- 3 files changed, 45 insertions(+), 19 deletions(-) diff --git a/include/block/block.h b/include/block/block.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block.h +++ b/include/block/block.h @@ -XXX,XX +XXX,XX @@ void bdrv_io_unplug(BlockDriverState *bs); * Begin a quiesced section of all users of @bs. This is part of * bdrv_drained_begin. */ -void bdrv_parent_drained_begin(BlockDriverState *bs); +void bdrv_parent_drained_begin(BlockDriverState *bs, BdrvChild *ignore); /** * bdrv_parent_drained_end: @@ -XXX,XX +XXX,XX @@ void bdrv_parent_drained_begin(BlockDriverState *bs); * End a quiesced section of all users of @bs. This is part of * bdrv_drained_end. */ -void bdrv_parent_drained_end(BlockDriverState *bs); +void bdrv_parent_drained_end(BlockDriverState *bs, BdrvChild *ignore); /** * bdrv_drained_begin: diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_noperm(BdrvChild *child, BlockDriverState *new_bs) { BlockDriverState *old_bs = child->bs; + int i; if (old_bs && new_bs) { assert(bdrv_get_aio_context(old_bs) == bdrv_get_aio_context(new_bs)); } if (old_bs) { if (old_bs->quiesce_counter && child->role->drained_end) { - child->role->drained_end(child); + for (i = 0; i < old_bs->quiesce_counter; i++) { + child->role->drained_end(child); + } } if (child->role->detach) { child->role->detach(child); @@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_noperm(BdrvChild *child, if (new_bs) { QLIST_INSERT_HEAD(&new_bs->parents, child, next_parent); if (new_bs->quiesce_counter && child->role->drained_begin) { - child->role->drained_begin(child); + for (i = 0; i < new_bs->quiesce_counter; i++) { + child->role->drained_begin(child); + } } if (child->role->attach) { @@ -XXX,XX +XXX,XX @@ void bdrv_set_aio_context(BlockDriverState *bs, AioContext *new_context) AioContext *ctx = bdrv_get_aio_context(bs); aio_disable_external(ctx); - bdrv_parent_drained_begin(bs); + bdrv_parent_drained_begin(bs, NULL); bdrv_drain(bs); /* ensure there are no in-flight requests */ while (aio_poll(ctx, false)) { @@ -XXX,XX +XXX,XX @@ void bdrv_set_aio_context(BlockDriverState *bs, AioContext *new_context) */ aio_context_acquire(new_context); bdrv_attach_aio_context(bs, new_context); - bdrv_parent_drained_end(bs); + bdrv_parent_drained_end(bs, NULL); aio_enable_external(ctx); aio_context_release(new_context); } diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_pwrite_zeroes(BlockDriverState *bs, int64_t offset, int bytes, BdrvRequestFlags flags); -void bdrv_parent_drained_begin(BlockDriverState *bs) +void bdrv_parent_drained_begin(BlockDriverState *bs, BdrvChild *ignore) { BdrvChild *c, *next; QLIST_FOREACH_SAFE(c, &bs->parents, next_parent, next) { + if (c == ignore) { + continue; + } if (c->role->drained_begin) { c->role->drained_begin(c); } } } -void bdrv_parent_drained_end(BlockDriverState *bs) +void bdrv_parent_drained_end(BlockDriverState *bs, BdrvChild *ignore) { BdrvChild *c, *next; QLIST_FOREACH_SAFE(c, &bs->parents, next_parent, next) { + if (c == ignore) { + continue; + } if (c->role->drained_end) { c->role->drained_end(c); } @@ -XXX,XX +XXX,XX @@ typedef struct { BlockDriverState *bs; bool done; bool begin; + BdrvChild *parent; } BdrvCoDrainData; static void coroutine_fn bdrv_drain_invoke_entry(void *opaque) @@ -XXX,XX +XXX,XX @@ static bool bdrv_drain_recurse(BlockDriverState *bs) return waited; } +static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent); +static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent); + static void bdrv_co_drain_bh_cb(void *opaque) { BdrvCoDrainData *data = opaque; @@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque) bdrv_dec_in_flight(bs); if (data->begin) { - bdrv_drained_begin(bs); + bdrv_do_drained_begin(bs, data->parent); } else { - bdrv_drained_end(bs); + bdrv_do_drained_end(bs, data->parent); } data->done = true; @@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque) } static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs, - bool begin) + bool begin, BdrvChild *parent) { BdrvCoDrainData data; @@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs, .bs = bs, .done = false, .begin = begin, + .parent = parent, }; bdrv_inc_in_flight(bs); aio_bh_schedule_oneshot(bdrv_get_aio_context(bs), @@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs, assert(data.done); } -void bdrv_drained_begin(BlockDriverState *bs) +static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent) { if (qemu_in_coroutine()) { - bdrv_co_yield_to_drain(bs, true); + bdrv_co_yield_to_drain(bs, true, parent); return; } /* Stop things in parent-to-child order */ if (atomic_fetch_inc(&bs->quiesce_counter) == 0) { aio_disable_external(bdrv_get_aio_context(bs)); - bdrv_parent_drained_begin(bs); } + bdrv_parent_drained_begin(bs, parent); bdrv_drain_invoke(bs, true, false); bdrv_drain_recurse(bs); } -void bdrv_drained_end(BlockDriverState *bs) +void bdrv_drained_begin(BlockDriverState *bs) +{ + bdrv_do_drained_begin(bs, NULL); +} + +static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent) { int old_quiesce_counter; if (qemu_in_coroutine()) { - bdrv_co_yield_to_drain(bs, false); + bdrv_co_yield_to_drain(bs, false, parent); return; } assert(bs->quiesce_counter > 0); @@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs) /* Re-enable things in child-to-parent order */ bdrv_drain_invoke(bs, false, false); + bdrv_parent_drained_end(bs, parent); if (old_quiesce_counter == 1) { - bdrv_parent_drained_end(bs); aio_enable_external(bdrv_get_aio_context(bs)); } } +void bdrv_drained_end(BlockDriverState *bs) +{ + bdrv_do_drained_end(bs, NULL); +} + /* * Wait for pending requests to complete on a single BlockDriverState subtree, * and suspend block driver's internal I/O until next request arrives. @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void) /* Stop things in parent-to-child order */ aio_context_acquire(aio_context); aio_disable_external(aio_context); - bdrv_parent_drained_begin(bs); + bdrv_parent_drained_begin(bs, NULL); bdrv_drain_invoke(bs, true, true); aio_context_release(aio_context); @@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void) /* Re-enable things in child-to-parent order */ aio_context_acquire(aio_context); bdrv_drain_invoke(bs, false, true); - bdrv_parent_drained_end(bs); + bdrv_parent_drained_end(bs, NULL); aio_enable_external(aio_context); aio_context_release(aio_context); } -- 2.13.6 bdrv_drained_begin() waits for the completion of requests in the whole subtree, but it only actually keeps its immediate bs parameter quiesced until bdrv_drained_end(). Add a version that keeps the whole subtree drained. As of this commit, graph changes cannot be allowed during a subtree drained section, but this will be fixed soon. Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block.h | 13 +++++++++++++ block/io.c | 54 ++++++++++++++++++++++++++++++++++++++++----------- 2 files changed, 56 insertions(+), 11 deletions(-) diff --git a/include/block/block.h b/include/block/block.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block.h +++ b/include/block/block.h @@ -XXX,XX +XXX,XX @@ void bdrv_parent_drained_end(BlockDriverState *bs, BdrvChild *ignore); void bdrv_drained_begin(BlockDriverState *bs); /** + * Like bdrv_drained_begin, but recursively begins a quiesced section for + * exclusive access to all child nodes as well. + * + * Graph changes are not allowed during a subtree drain section. + */ +void bdrv_subtree_drained_begin(BlockDriverState *bs); + +/** * bdrv_drained_end: * * End a quiescent section started by bdrv_drained_begin(). */ void bdrv_drained_end(BlockDriverState *bs); +/** + * End a quiescent section started by bdrv_subtree_drained_begin(). + */ +void bdrv_subtree_drained_end(BlockDriverState *bs); + void bdrv_add_child(BlockDriverState *parent, BlockDriverState *child, Error **errp); void bdrv_del_child(BlockDriverState *parent, BdrvChild *child, Error **errp); diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ typedef struct { BlockDriverState *bs; bool done; bool begin; + bool recursive; BdrvChild *parent; } BdrvCoDrainData; @@ -XXX,XX +XXX,XX @@ static bool bdrv_drain_recurse(BlockDriverState *bs) return waited; } -static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent); -static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent); +static void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive, + BdrvChild *parent); +static void bdrv_do_drained_end(BlockDriverState *bs, bool recursive, + BdrvChild *parent); static void bdrv_co_drain_bh_cb(void *opaque) { @@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque) bdrv_dec_in_flight(bs); if (data->begin) { - bdrv_do_drained_begin(bs, data->parent); + bdrv_do_drained_begin(bs, data->recursive, data->parent); } else { - bdrv_do_drained_end(bs, data->parent); + bdrv_do_drained_end(bs, data->recursive, data->parent); } data->done = true; @@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque) } static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs, - bool begin, BdrvChild *parent) + bool begin, bool recursive, + BdrvChild *parent) { BdrvCoDrainData data; @@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs, .bs = bs, .done = false, .begin = begin, + .recursive = recursive, .parent = parent, }; bdrv_inc_in_flight(bs); @@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs, assert(data.done); } -static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent) +static void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive, + BdrvChild *parent) { + BdrvChild *child, *next; + if (qemu_in_coroutine()) { - bdrv_co_yield_to_drain(bs, true, parent); + bdrv_co_yield_to_drain(bs, true, recursive, parent); return; } @@ -XXX,XX +XXX,XX @@ static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent) bdrv_parent_drained_begin(bs, parent); bdrv_drain_invoke(bs, true, false); bdrv_drain_recurse(bs); + + if (recursive) { + QLIST_FOREACH_SAFE(child, &bs->children, next, next) { + bdrv_do_drained_begin(child->bs, true, child); + } + } } void bdrv_drained_begin(BlockDriverState *bs) { - bdrv_do_drained_begin(bs, NULL); + bdrv_do_drained_begin(bs, false, NULL); +} + +void bdrv_subtree_drained_begin(BlockDriverState *bs) +{ + bdrv_do_drained_begin(bs, true, NULL); } -static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent) +static void bdrv_do_drained_end(BlockDriverState *bs, bool recursive, + BdrvChild *parent) { + BdrvChild *child, *next; int old_quiesce_counter; if (qemu_in_coroutine()) { - bdrv_co_yield_to_drain(bs, false, parent); + bdrv_co_yield_to_drain(bs, false, recursive, parent); return; } assert(bs->quiesce_counter > 0); @@ -XXX,XX +XXX,XX @@ static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent) if (old_quiesce_counter == 1) { aio_enable_external(bdrv_get_aio_context(bs)); } + + if (recursive) { + QLIST_FOREACH_SAFE(child, &bs->children, next, next) { + bdrv_do_drained_end(child->bs, true, child); + } + } } void bdrv_drained_end(BlockDriverState *bs) { - bdrv_do_drained_end(bs, NULL); + bdrv_do_drained_end(bs, false, NULL); +} + +void bdrv_subtree_drained_end(BlockDriverState *bs) +{ + bdrv_do_drained_end(bs, true, NULL); } /* -- 2.13.6 Add a subtree drain version to the existing test cases. Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/test-bdrv-drain.c | 27 ++++++++++++++++++++++++++- 1 file changed, 26 insertions(+), 1 deletion(-) diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c index XXXXXXX..XXXXXXX 100644 --- a/tests/test-bdrv-drain.c +++ b/tests/test-bdrv-drain.c @@ -XXX,XX +XXX,XX @@ static void aio_ret_cb(void *opaque, int ret) enum drain_type { BDRV_DRAIN_ALL, BDRV_DRAIN, + BDRV_SUBTREE_DRAIN, DRAIN_TYPE_MAX, }; @@ -XXX,XX +XXX,XX @@ static void do_drain_begin(enum drain_type drain_type, BlockDriverState *bs) switch (drain_type) { case BDRV_DRAIN_ALL: bdrv_drain_all_begin(); break; case BDRV_DRAIN: bdrv_drained_begin(bs); break; + case BDRV_SUBTREE_DRAIN: bdrv_subtree_drained_begin(bs); break; default: g_assert_not_reached(); } } @@ -XXX,XX +XXX,XX @@ static void do_drain_end(enum drain_type drain_type, BlockDriverState *bs) switch (drain_type) { case BDRV_DRAIN_ALL: bdrv_drain_all_end(); break; case BDRV_DRAIN: bdrv_drained_end(bs); break; + case BDRV_SUBTREE_DRAIN: bdrv_subtree_drained_end(bs); break; default: g_assert_not_reached(); } } @@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain(void) test_drv_cb_common(BDRV_DRAIN, false); } +static void test_drv_cb_drain_subtree(void) +{ + test_drv_cb_common(BDRV_SUBTREE_DRAIN, true); +} + static void test_quiesce_common(enum drain_type drain_type, bool recursive) { BlockBackend *blk; @@ -XXX,XX +XXX,XX @@ static void test_quiesce_drain(void) test_quiesce_common(BDRV_DRAIN, false); } +static void test_quiesce_drain_subtree(void) +{ + test_quiesce_common(BDRV_SUBTREE_DRAIN, true); +} + static void test_nested(void) { BlockBackend *blk; @@ -XXX,XX +XXX,XX @@ static void test_nested(void) /* XXX bdrv_drain_all() doesn't increase the quiesce_counter */ int bs_quiesce = (outer != BDRV_DRAIN_ALL) + (inner != BDRV_DRAIN_ALL); - int backing_quiesce = 0; + int backing_quiesce = (outer == BDRV_SUBTREE_DRAIN) + + (inner == BDRV_SUBTREE_DRAIN); int backing_cb_cnt = (outer != BDRV_DRAIN) + (inner != BDRV_DRAIN); @@ -XXX,XX +XXX,XX @@ static void test_blockjob_drain(void) test_blockjob_common(BDRV_DRAIN); } +static void test_blockjob_drain_subtree(void) +{ + test_blockjob_common(BDRV_SUBTREE_DRAIN); +} + int main(int argc, char **argv) { bdrv_init(); @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv) g_test_add_func("/bdrv-drain/driver-cb/drain_all", test_drv_cb_drain_all); g_test_add_func("/bdrv-drain/driver-cb/drain", test_drv_cb_drain); + g_test_add_func("/bdrv-drain/driver-cb/drain_subtree", + test_drv_cb_drain_subtree); g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all); g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain); + g_test_add_func("/bdrv-drain/quiesce/drain_subtree", + test_quiesce_drain_subtree); g_test_add_func("/bdrv-drain/nested", test_nested); g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all); g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain); + g_test_add_func("/bdrv-drain/blockjob/drain_subtree", + test_blockjob_drain_subtree); return g_test_run(); } -- 2.13.6 If bdrv_do_drained_begin/end() are called in coroutine context, they first use a BH to get out of the coroutine context. Call some existing tests again from a coroutine to cover this code path. Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/test-bdrv-drain.c | 59 +++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 59 insertions(+) diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c index XXXXXXX..XXXXXXX 100644 --- a/tests/test-bdrv-drain.c +++ b/tests/test-bdrv-drain.c @@ -XXX,XX +XXX,XX @@ static void aio_ret_cb(void *opaque, int ret) *aio_ret = ret; } +typedef struct CallInCoroutineData { + void (*entry)(void); + bool done; +} CallInCoroutineData; + +static coroutine_fn void call_in_coroutine_entry(void *opaque) +{ + CallInCoroutineData *data = opaque; + + data->entry(); + data->done = true; +} + +static void call_in_coroutine(void (*entry)(void)) +{ + Coroutine *co; + CallInCoroutineData data = { + .entry = entry, + .done = false, + }; + + co = qemu_coroutine_create(call_in_coroutine_entry, &data); + qemu_coroutine_enter(co); + while (!data.done) { + aio_poll(qemu_get_aio_context(), true); + } +} + enum drain_type { BDRV_DRAIN_ALL, BDRV_DRAIN, @@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain_subtree(void) test_drv_cb_common(BDRV_SUBTREE_DRAIN, true); } +static void test_drv_cb_co_drain(void) +{ + call_in_coroutine(test_drv_cb_drain); +} + +static void test_drv_cb_co_drain_subtree(void) +{ + call_in_coroutine(test_drv_cb_drain_subtree); +} + static void test_quiesce_common(enum drain_type drain_type, bool recursive) { BlockBackend *blk; @@ -XXX,XX +XXX,XX @@ static void test_quiesce_drain_subtree(void) test_quiesce_common(BDRV_SUBTREE_DRAIN, true); } +static void test_quiesce_co_drain(void) +{ + call_in_coroutine(test_quiesce_drain); +} + +static void test_quiesce_co_drain_subtree(void) +{ + call_in_coroutine(test_quiesce_drain_subtree); +} + static void test_nested(void) { BlockBackend *blk; @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv) g_test_add_func("/bdrv-drain/driver-cb/drain_subtree", test_drv_cb_drain_subtree); + // XXX bdrv_drain_all() doesn't work in coroutine context + g_test_add_func("/bdrv-drain/driver-cb/co/drain", test_drv_cb_co_drain); + g_test_add_func("/bdrv-drain/driver-cb/co/drain_subtree", + test_drv_cb_co_drain_subtree); + + g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all); g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain); g_test_add_func("/bdrv-drain/quiesce/drain_subtree", test_quiesce_drain_subtree); + // XXX bdrv_drain_all() doesn't work in coroutine context + g_test_add_func("/bdrv-drain/quiesce/co/drain", test_quiesce_co_drain); + g_test_add_func("/bdrv-drain/quiesce/co/drain_subtree", + test_quiesce_co_drain_subtree); + g_test_add_func("/bdrv-drain/nested", test_nested); g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all); -- 2.13.6 Test that drain sections are correctly propagated through the graph. Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/test-bdrv-drain.c | 74 +++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 74 insertions(+) diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c index XXXXXXX..XXXXXXX 100644 --- a/tests/test-bdrv-drain.c +++ b/tests/test-bdrv-drain.c @@ -XXX,XX +XXX,XX @@ static void test_nested(void) blk_unref(blk); } +static void test_multiparent(void) +{ + BlockBackend *blk_a, *blk_b; + BlockDriverState *bs_a, *bs_b, *backing; + BDRVTestState *a_s, *b_s, *backing_s; + + blk_a = blk_new(BLK_PERM_ALL, BLK_PERM_ALL); + bs_a = bdrv_new_open_driver(&bdrv_test, "test-node-a", BDRV_O_RDWR, + &error_abort); + a_s = bs_a->opaque; + blk_insert_bs(blk_a, bs_a, &error_abort); + + blk_b = blk_new(BLK_PERM_ALL, BLK_PERM_ALL); + bs_b = bdrv_new_open_driver(&bdrv_test, "test-node-b", BDRV_O_RDWR, + &error_abort); + b_s = bs_b->opaque; + blk_insert_bs(blk_b, bs_b, &error_abort); + + backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort); + backing_s = backing->opaque; + bdrv_set_backing_hd(bs_a, backing, &error_abort); + bdrv_set_backing_hd(bs_b, backing, &error_abort); + + g_assert_cmpint(bs_a->quiesce_counter, ==, 0); + g_assert_cmpint(bs_b->quiesce_counter, ==, 0); + g_assert_cmpint(backing->quiesce_counter, ==, 0); + g_assert_cmpint(a_s->drain_count, ==, 0); + g_assert_cmpint(b_s->drain_count, ==, 0); + g_assert_cmpint(backing_s->drain_count, ==, 0); + + do_drain_begin(BDRV_SUBTREE_DRAIN, bs_a); + + g_assert_cmpint(bs_a->quiesce_counter, ==, 1); + g_assert_cmpint(bs_b->quiesce_counter, ==, 1); + g_assert_cmpint(backing->quiesce_counter, ==, 1); + g_assert_cmpint(a_s->drain_count, ==, 1); + g_assert_cmpint(b_s->drain_count, ==, 1); + g_assert_cmpint(backing_s->drain_count, ==, 1); + + do_drain_begin(BDRV_SUBTREE_DRAIN, bs_b); + + g_assert_cmpint(bs_a->quiesce_counter, ==, 2); + g_assert_cmpint(bs_b->quiesce_counter, ==, 2); + g_assert_cmpint(backing->quiesce_counter, ==, 2); + g_assert_cmpint(a_s->drain_count, ==, 2); + g_assert_cmpint(b_s->drain_count, ==, 2); + g_assert_cmpint(backing_s->drain_count, ==, 2); + + do_drain_end(BDRV_SUBTREE_DRAIN, bs_b); + + g_assert_cmpint(bs_a->quiesce_counter, ==, 1); + g_assert_cmpint(bs_b->quiesce_counter, ==, 1); + g_assert_cmpint(backing->quiesce_counter, ==, 1); + g_assert_cmpint(a_s->drain_count, ==, 1); + g_assert_cmpint(b_s->drain_count, ==, 1); + g_assert_cmpint(backing_s->drain_count, ==, 1); + + do_drain_end(BDRV_SUBTREE_DRAIN, bs_a); + + g_assert_cmpint(bs_a->quiesce_counter, ==, 0); + g_assert_cmpint(bs_b->quiesce_counter, ==, 0); + g_assert_cmpint(backing->quiesce_counter, ==, 0); + g_assert_cmpint(a_s->drain_count, ==, 0); + g_assert_cmpint(b_s->drain_count, ==, 0); + g_assert_cmpint(backing_s->drain_count, ==, 0); + + bdrv_unref(backing); + bdrv_unref(bs_a); + bdrv_unref(bs_b); + blk_unref(blk_a); + blk_unref(blk_b); +} + typedef struct TestBlockJob { BlockJob common; @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv) test_quiesce_co_drain_subtree); g_test_add_func("/bdrv-drain/nested", test_nested); + g_test_add_func("/bdrv-drain/multiparent", test_multiparent); g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all); g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain); -- 2.13.6 We need to remember how many of the drain sections in which a node is were recursive (i.e. subtree drain rather than node drain), so that they can be correctly applied when children are added or removed during the drained section. With this change, it is safe to modify the graph even inside a bdrv_subtree_drained_begin/end() section. Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block.h | 2 -- include/block/block_int.h | 5 +++++ block.c | 32 +++++++++++++++++++++++++++++--- block/io.c | 28 ++++++++++++++++++++++++---- 4 files changed, 58 insertions(+), 9 deletions(-) diff --git a/include/block/block.h b/include/block/block.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block.h +++ b/include/block/block.h @@ -XXX,XX +XXX,XX @@ void bdrv_drained_begin(BlockDriverState *bs); /** * Like bdrv_drained_begin, but recursively begins a quiesced section for * exclusive access to all child nodes as well. - * - * Graph changes are not allowed during a subtree drain section. */ void bdrv_subtree_drained_begin(BlockDriverState *bs); diff --git a/include/block/block_int.h b/include/block/block_int.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block_int.h +++ b/include/block/block_int.h @@ -XXX,XX +XXX,XX @@ struct BlockDriverState { /* Accessed with atomic ops. */ int quiesce_counter; + int recursive_quiesce_counter; + unsigned int write_gen; /* Current data generation */ /* Protected by reqs_lock. */ @@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child, int64_t offset, unsigned int bytes, QEMUIOVector *qiov, BdrvRequestFlags flags); +void bdrv_apply_subtree_drain(BdrvChild *child, BlockDriverState *new_parent); +void bdrv_unapply_subtree_drain(BdrvChild *child, BlockDriverState *old_parent); + int get_tmp_filename(char *filename, int size); BlockDriver *bdrv_probe_all(const uint8_t *buf, int buf_size, const char *filename); diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static void bdrv_child_cb_drained_end(BdrvChild *child) bdrv_drained_end(bs); } +static void bdrv_child_cb_attach(BdrvChild *child) +{ + BlockDriverState *bs = child->opaque; + bdrv_apply_subtree_drain(child, bs); +} + +static void bdrv_child_cb_detach(BdrvChild *child) +{ + BlockDriverState *bs = child->opaque; + bdrv_unapply_subtree_drain(child, bs); +} + static int bdrv_child_cb_inactivate(BdrvChild *child) { BlockDriverState *bs = child->opaque; @@ -XXX,XX +XXX,XX @@ const BdrvChildRole child_file = { .inherit_options = bdrv_inherited_options, .drained_begin = bdrv_child_cb_drained_begin, .drained_end = bdrv_child_cb_drained_end, + .attach = bdrv_child_cb_attach, + .detach = bdrv_child_cb_detach, .inactivate = bdrv_child_cb_inactivate, }; @@ -XXX,XX +XXX,XX @@ const BdrvChildRole child_format = { .inherit_options = bdrv_inherited_fmt_options, .drained_begin = bdrv_child_cb_drained_begin, .drained_end = bdrv_child_cb_drained_end, + .attach = bdrv_child_cb_attach, + .detach = bdrv_child_cb_detach, .inactivate = bdrv_child_cb_inactivate, }; @@ -XXX,XX +XXX,XX @@ static void bdrv_backing_attach(BdrvChild *c) parent->backing_blocker); bdrv_op_unblock(backing_hd, BLOCK_OP_TYPE_BACKUP_TARGET, parent->backing_blocker); + + bdrv_child_cb_attach(c); } static void bdrv_backing_detach(BdrvChild *c) @@ -XXX,XX +XXX,XX @@ static void bdrv_backing_detach(BdrvChild *c) bdrv_op_unblock_all(c->bs, parent->backing_blocker); error_free(parent->backing_blocker); parent->backing_blocker = NULL; + + bdrv_child_cb_detach(c); } /* @@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_noperm(BdrvChild *child, assert(bdrv_get_aio_context(old_bs) == bdrv_get_aio_context(new_bs)); } if (old_bs) { + /* Detach first so that the recursive drain sections coming from @child + * are already gone and we only end the drain sections that came from + * elsewhere. */ + if (child->role->detach) { + child->role->detach(child); + } if (old_bs->quiesce_counter && child->role->drained_end) { for (i = 0; i < old_bs->quiesce_counter; i++) { child->role->drained_end(child); } } - if (child->role->detach) { - child->role->detach(child); - } QLIST_REMOVE(child, next_parent); } @@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_noperm(BdrvChild *child, } } + /* Attach only after starting new drained sections, so that recursive + * drain sections coming from @child don't get an extra .drained_begin + * callback. */ if (child->role->attach) { child->role->attach(child); } diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs, assert(data.done); } -static void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive, - BdrvChild *parent) +void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive, + BdrvChild *parent) { BdrvChild *child, *next; @@ -XXX,XX +XXX,XX @@ static void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive, bdrv_drain_recurse(bs); if (recursive) { + bs->recursive_quiesce_counter++; QLIST_FOREACH_SAFE(child, &bs->children, next, next) { bdrv_do_drained_begin(child->bs, true, child); } @@ -XXX,XX +XXX,XX @@ void bdrv_subtree_drained_begin(BlockDriverState *bs) bdrv_do_drained_begin(bs, true, NULL); } -static void bdrv_do_drained_end(BlockDriverState *bs, bool recursive, - BdrvChild *parent) +void bdrv_do_drained_end(BlockDriverState *bs, bool recursive, + BdrvChild *parent) { BdrvChild *child, *next; int old_quiesce_counter; @@ -XXX,XX +XXX,XX @@ static void bdrv_do_drained_end(BlockDriverState *bs, bool recursive, } if (recursive) { + bs->recursive_quiesce_counter--; QLIST_FOREACH_SAFE(child, &bs->children, next, next) { bdrv_do_drained_end(child->bs, true, child); } @@ -XXX,XX +XXX,XX @@ void bdrv_subtree_drained_end(BlockDriverState *bs) bdrv_do_drained_end(bs, true, NULL); } +void bdrv_apply_subtree_drain(BdrvChild *child, BlockDriverState *new_parent) +{ + int i; + + for (i = 0; i < new_parent->recursive_quiesce_counter; i++) { + bdrv_do_drained_begin(child->bs, true, child); + } +} + +void bdrv_unapply_subtree_drain(BdrvChild *child, BlockDriverState *old_parent) +{ + int i; + + for (i = 0; i < old_parent->recursive_quiesce_counter; i++) { + bdrv_do_drained_end(child->bs, true, child); + } +} + /* * Wait for pending requests to complete on a single BlockDriverState subtree, * and suspend block driver's internal I/O until next request arrives. -- 2.13.6 Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/test-bdrv-drain.c | 80 +++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 80 insertions(+) diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c index XXXXXXX..XXXXXXX 100644 --- a/tests/test-bdrv-drain.c +++ b/tests/test-bdrv-drain.c @@ -XXX,XX +XXX,XX @@ static void test_multiparent(void) blk_unref(blk_b); } +static void test_graph_change(void) +{ + BlockBackend *blk_a, *blk_b; + BlockDriverState *bs_a, *bs_b, *backing; + BDRVTestState *a_s, *b_s, *backing_s; + + blk_a = blk_new(BLK_PERM_ALL, BLK_PERM_ALL); + bs_a = bdrv_new_open_driver(&bdrv_test, "test-node-a", BDRV_O_RDWR, + &error_abort); + a_s = bs_a->opaque; + blk_insert_bs(blk_a, bs_a, &error_abort); + + blk_b = blk_new(BLK_PERM_ALL, BLK_PERM_ALL); + bs_b = bdrv_new_open_driver(&bdrv_test, "test-node-b", BDRV_O_RDWR, + &error_abort); + b_s = bs_b->opaque; + blk_insert_bs(blk_b, bs_b, &error_abort); + + backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort); + backing_s = backing->opaque; + bdrv_set_backing_hd(bs_a, backing, &error_abort); + + g_assert_cmpint(bs_a->quiesce_counter, ==, 0); + g_assert_cmpint(bs_b->quiesce_counter, ==, 0); + g_assert_cmpint(backing->quiesce_counter, ==, 0); + g_assert_cmpint(a_s->drain_count, ==, 0); + g_assert_cmpint(b_s->drain_count, ==, 0); + g_assert_cmpint(backing_s->drain_count, ==, 0); + + do_drain_begin(BDRV_SUBTREE_DRAIN, bs_a); + do_drain_begin(BDRV_SUBTREE_DRAIN, bs_a); + do_drain_begin(BDRV_SUBTREE_DRAIN, bs_a); + do_drain_begin(BDRV_SUBTREE_DRAIN, bs_b); + do_drain_begin(BDRV_SUBTREE_DRAIN, bs_b); + + bdrv_set_backing_hd(bs_b, backing, &error_abort); + g_assert_cmpint(bs_a->quiesce_counter, ==, 5); + g_assert_cmpint(bs_b->quiesce_counter, ==, 5); + g_assert_cmpint(backing->quiesce_counter, ==, 5); + g_assert_cmpint(a_s->drain_count, ==, 5); + g_assert_cmpint(b_s->drain_count, ==, 5); + g_assert_cmpint(backing_s->drain_count, ==, 5); + + bdrv_set_backing_hd(bs_b, NULL, &error_abort); + g_assert_cmpint(bs_a->quiesce_counter, ==, 3); + g_assert_cmpint(bs_b->quiesce_counter, ==, 2); + g_assert_cmpint(backing->quiesce_counter, ==, 3); + g_assert_cmpint(a_s->drain_count, ==, 3); + g_assert_cmpint(b_s->drain_count, ==, 2); + g_assert_cmpint(backing_s->drain_count, ==, 3); + + bdrv_set_backing_hd(bs_b, backing, &error_abort); + g_assert_cmpint(bs_a->quiesce_counter, ==, 5); + g_assert_cmpint(bs_b->quiesce_counter, ==, 5); + g_assert_cmpint(backing->quiesce_counter, ==, 5); + g_assert_cmpint(a_s->drain_count, ==, 5); + g_assert_cmpint(b_s->drain_count, ==, 5); + g_assert_cmpint(backing_s->drain_count, ==, 5); + + do_drain_end(BDRV_SUBTREE_DRAIN, bs_b); + do_drain_end(BDRV_SUBTREE_DRAIN, bs_b); + do_drain_end(BDRV_SUBTREE_DRAIN, bs_a); + do_drain_end(BDRV_SUBTREE_DRAIN, bs_a); + do_drain_end(BDRV_SUBTREE_DRAIN, bs_a); + + g_assert_cmpint(bs_a->quiesce_counter, ==, 0); + g_assert_cmpint(bs_b->quiesce_counter, ==, 0); + g_assert_cmpint(backing->quiesce_counter, ==, 0); + g_assert_cmpint(a_s->drain_count, ==, 0); + g_assert_cmpint(b_s->drain_count, ==, 0); + g_assert_cmpint(backing_s->drain_count, ==, 0); + + bdrv_unref(backing); + bdrv_unref(bs_a); + bdrv_unref(bs_b); + blk_unref(blk_a); + blk_unref(blk_b); +} + typedef struct TestBlockJob { BlockJob common; @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv) g_test_add_func("/bdrv-drain/nested", test_nested); g_test_add_func("/bdrv-drain/multiparent", test_multiparent); + g_test_add_func("/bdrv-drain/graph-change", test_graph_change); g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all); g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain); -- 2.13.6 Since commit bde70715, base is the only node that is reopened in commit_start(). This means that the code, which still involves an explicit BlockReopenQueue, can now be simplified by using bdrv_reopen(). Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> --- block/commit.c | 8 +------- 1 file changed, 1 insertion(+), 7 deletions(-) diff --git a/block/commit.c b/block/commit.c index XXXXXXX..XXXXXXX 100644 --- a/block/commit.c +++ b/block/commit.c @@ -XXX,XX +XXX,XX @@ void commit_start(const char *job_id, BlockDriverState *bs, const char *filter_node_name, Error **errp) { CommitBlockJob *s; - BlockReopenQueue *reopen_queue = NULL; int orig_base_flags; BlockDriverState *iter; BlockDriverState *commit_top_bs = NULL; @@ -XXX,XX +XXX,XX @@ void commit_start(const char *job_id, BlockDriverState *bs, /* convert base to r/w, if necessary */ orig_base_flags = bdrv_get_flags(base); if (!(orig_base_flags & BDRV_O_RDWR)) { - reopen_queue = bdrv_reopen_queue(reopen_queue, base, NULL, - orig_base_flags | BDRV_O_RDWR); - } - - if (reopen_queue) { - bdrv_reopen_multiple(bdrv_get_aio_context(bs), reopen_queue, &local_err); + bdrv_reopen(base, orig_base_flags | BDRV_O_RDWR, &local_err); if (local_err != NULL) { error_propagate(errp, local_err); goto fail; -- 2.13.6 The bdrv_reopen*() implementation doesn't like it if the graph is changed between queuing nodes for reopen and actually reopening them (one of the reasons is that queuing can be recursive). So instead of draining the device only in bdrv_reopen_multiple(), require that callers already drained all affected nodes, and assert this in bdrv_reopen_queue(). Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> --- block.c | 23 ++++++++++++++++------- block/replication.c | 6 ++++++ qemu-io-cmds.c | 3 +++ 3 files changed, 25 insertions(+), 7 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ BlockDriverState *bdrv_open(const char *filename, const char *reference, * returns a pointer to bs_queue, which is either the newly allocated * bs_queue, or the existing bs_queue being used. * + * bs must be drained between bdrv_reopen_queue() and bdrv_reopen_multiple(). */ static BlockReopenQueue *bdrv_reopen_queue_child(BlockReopenQueue *bs_queue, BlockDriverState *bs, @@ -XXX,XX +XXX,XX @@ static BlockReopenQueue *bdrv_reopen_queue_child(BlockReopenQueue *bs_queue, BdrvChild *child; QDict *old_options, *explicit_options; + /* Make sure that the caller remembered to use a drained section. This is + * important to avoid graph changes between the recursive queuing here and + * bdrv_reopen_multiple(). */ + assert(bs->quiesce_counter > 0); + if (bs_queue == NULL) { bs_queue = g_new0(BlockReopenQueue, 1); QSIMPLEQ_INIT(bs_queue); @@ -XXX,XX +XXX,XX @@ BlockReopenQueue *bdrv_reopen_queue(BlockReopenQueue *bs_queue, * If all devices prepare successfully, then the changes are committed * to all devices. * + * All affected nodes must be drained between bdrv_reopen_queue() and + * bdrv_reopen_multiple(). */ int bdrv_reopen_multiple(AioContext *ctx, BlockReopenQueue *bs_queue, Error **errp) { @@ -XXX,XX +XXX,XX @@ int bdrv_reopen_multiple(AioContext *ctx, BlockReopenQueue *bs_queue, Error **er assert(bs_queue != NULL); - aio_context_release(ctx); - bdrv_drain_all_begin(); - aio_context_acquire(ctx); - QSIMPLEQ_FOREACH(bs_entry, bs_queue, entry) { + assert(bs_entry->state.bs->quiesce_counter > 0); if (bdrv_reopen_prepare(&bs_entry->state, bs_queue, &local_err)) { error_propagate(errp, local_err); goto cleanup; @@ -XXX,XX +XXX,XX @@ cleanup: } g_free(bs_queue); - bdrv_drain_all_end(); - return ret; } @@ -XXX,XX +XXX,XX @@ int bdrv_reopen(BlockDriverState *bs, int bdrv_flags, Error **errp) { int ret = -1; Error *local_err = NULL; - BlockReopenQueue *queue = bdrv_reopen_queue(NULL, bs, NULL, bdrv_flags); + BlockReopenQueue *queue; + bdrv_subtree_drained_begin(bs); + + queue = bdrv_reopen_queue(NULL, bs, NULL, bdrv_flags); ret = bdrv_reopen_multiple(bdrv_get_aio_context(bs), queue, &local_err); if (local_err != NULL) { error_propagate(errp, local_err); } + + bdrv_subtree_drained_end(bs); + return ret; } diff --git a/block/replication.c b/block/replication.c index XXXXXXX..XXXXXXX 100644 --- a/block/replication.c +++ b/block/replication.c @@ -XXX,XX +XXX,XX @@ static void reopen_backing_file(BlockDriverState *bs, bool writable, new_secondary_flags = s->orig_secondary_flags; } + bdrv_subtree_drained_begin(s->hidden_disk->bs); + bdrv_subtree_drained_begin(s->secondary_disk->bs); + if (orig_hidden_flags != new_hidden_flags) { reopen_queue = bdrv_reopen_queue(reopen_queue, s->hidden_disk->bs, NULL, new_hidden_flags); @@ -XXX,XX +XXX,XX @@ static void reopen_backing_file(BlockDriverState *bs, bool writable, reopen_queue, &local_err); error_propagate(errp, local_err); } + + bdrv_subtree_drained_end(s->hidden_disk->bs); + bdrv_subtree_drained_end(s->secondary_disk->bs); } static void backup_job_cleanup(BlockDriverState *bs) diff --git a/qemu-io-cmds.c b/qemu-io-cmds.c index XXXXXXX..XXXXXXX 100644 --- a/qemu-io-cmds.c +++ b/qemu-io-cmds.c @@ -XXX,XX +XXX,XX @@ static int reopen_f(BlockBackend *blk, int argc, char **argv) opts = qopts ? qemu_opts_to_qdict(qopts, NULL) : NULL; qemu_opts_reset(&reopen_opts); + bdrv_subtree_drained_begin(bs); brq = bdrv_reopen_queue(NULL, bs, opts, flags); bdrv_reopen_multiple(bdrv_get_aio_context(bs), brq, &local_err); + bdrv_subtree_drained_end(bs); + if (local_err) { error_report_err(local_err); } else { -- 2.13.6