The qcow2 backend caches L2 table and refcount updates in memory and
only writes them back on a guest-initiated flush, clean shutdown or
drop. A paused VM therefore leaves the on-disk image without the
mappings for any cluster allocated since the last guest flush: the
data clusters are present in the file, but nothing references them.
Anything that reads the image while the VM is paused sees those
writes as missing. Copying the disk alongside vm.snapshot (the
documented snapshot workflow) captures a stale image, and live
migration releases the disk locks after pausing so the destination
reopens the file with the same stale metadata. In both cases writes
the guest has completed, and may later read back, silently disappear.
Add a MetadataSync capability trait with a no-op default, fold it
into FullDiskFile, implement it for the qcow2 backend as a metadata
cache flush, and call it from the virtio-block pause path after
in-flight requests have drained. Pause is the quiesce point both
flows rely on, and it is a cold path, so the extra flush does not
affect runtime I/O.
Reproduced by writing to a qcow2 disk from the guest with O_DIRECT
and no explicit flush, pausing the VM and copying the image: qemu-img
map on the copy shows no mapped clusters and reads return zeros. With
this change the copy contains every completed write. A unit test
covers the same sequence at the format level: a completed write is
invisible to a fresh reader until sync_metadata, and visible after.
Signed-off-by: CMGS <ilskdw@gmail.com>
All disk format backends now implement AsyncFullDiskFile directly.
The DiskBackend enum that dispatched between Legacy and Next arms
is no longer needed since the factory returns trait objects and vmm
no longer constructs format types manually.
Replace DiskBackend with Box<dyn AsyncFullDiskFile> in the Block
struct and its constructor. Remove the DiskBackend::Next wrapping
in device_manager and the fuzz target.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
The new_ prefix in Rust conventionally denotes constructors that return
Self (e.g. Vec::new(), File::new()). AsyncDiskFile::new_async_io does
not return Self. It is a factory method that constructs and returns a
Box<dyn AsyncIo> worker bound to the disk file descriptor and
metadata. The create_ prefix communicates this: the caller receives
a freshly constructed object of a different type.
This rename touches every format backend in block plus two external
callers in virtio-devices and performance-metrics. Every change is a
mechanical s/new_async_io/create_async_io/ substitution. No functional
change.
Ref: #7877 (task 3.2.8)
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Introduce DiskBackend with two variants:
- Legacy: wraps Box<dyn async_io::DiskFile> for existing formats
- Next: wraps Box<dyn disk_file::AsyncFullDiskFile>
Methods return BlockResult, with DiskFileError converted up to
BlockError on the Legacy path. The Next path passes through
directly with zero conversion overhead.
This is a transitional type. Once all formats implement
AsyncFullDiskFile, DiskBackend and Legacy are removed and
callers hold Box<dyn AsyncFullDiskFile> directly.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Marker trait combining FullDiskFile and AsyncDiskFile. Blanket impl
covers any type implementing both supertraits.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Marker trait bundling all optional capabilities (PhysicalSize, DiskFd,
SparseCapable, Resizable) on top of DiskFile. Blanket impl covers any
type implementing all constituent traits.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Extend DiskFile with async I/O construction for virtio queue workers.
AsyncDiskFile adds try_clone() for creating independent handles to
the same backing storage, and new_async_io() for constructing an
async I/O engine at the given ring depth.
Bounds: DiskFile + Unpin. Unpin ensures trait objects can be moved
freely (all concrete disk file types are naturally Unpin since they
hold no self referential state).
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Bundles DiskSize and Geometry as the universal disk
capabilities every format must implement. Adds Sync so
that Arc<dyn DiskFile> can be shared across threads for
concurrent readonly access.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Live disk resize support. Single method resize() taking
&mut self and the new size in bytes. Implementations may
return an error if the backend does not support resizing.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Sparse and zero flag support for thin provisioned disk
images. Two methods with false defaults: sparse operations
(punch hole, write zeroes, discard) and zero flag
optimization in WRITE_ZEROES.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Sector and cluster geometry of a disk image. Returns
DiskTopology with a default implementation providing
512B logical and physical block sizes. Formats that
probe the underlying device override this.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Backing file descriptor access for disk images backed by
a file. Returns a BorrowedDiskFd that wraps the raw fd
with lifetime tracking. Not available for network or
memory backed disk formats.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Host allocation size for file-backed disk images. Reports
actual bytes occupied on the host filesystem. Not every
format supports this, e.g. network or memory backed disks.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Reported capacity of a disk image. Every format, be it
file backed, network, memory, exposes a logical size.
Single method logical_size() returning the virtual size
in bytes.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Composable disk capability traits with DiskFile as a supertrait
bundling DiskSize and Geometry. Optional capabilities are
separate traits: PhysicalSize, DiskFd, SparseCapable, Resizable.
AsyncDiskFile extends DiskFile with async I/O construction.
Empty module with doc comment, trait definitions follow.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
