On 6/26/23 19:08, Andrey Drobyshev wrote:
> This series makes IO requests performed with copy-on-read to be aligned
> to subclusters rather than clusters. It also affects mirror job requests
> alignment.
>
> The initial reason for that change is the following crash discovered:
>
> qemu-img create -f qcow2 base.qcow2 64K
> qemu-img create -f qcow2 -o extended_l2=on,backing_file=base.qcow2,backing_fmt=qcow2 img.qcow2 64K
> qemu-io -c "write -P 0xaa 0 2K" img.qcow2
> qemu-io -C -c "read -P 0x00 2K 62K" img.qcow2
>
> qemu-io: ../block/io.c:1236: bdrv_co_do_copy_on_readv: Assertion `skip_bytes < pnum' failed.
>
> This change in alignment fixes the crash and adds test case covering it.
>
> Andrey Drobyshev (3):
> block: add subcluster_size field to BlockDriverInfo
> block/io: align requests to subcluster_size
> tests/qemu-iotests/197: add testcase for CoR with subclusters
>
> block.c | 6 +++++
> block/io.c | 50 ++++++++++++++++++------------------
> block/mirror.c | 8 +++---
> block/qcow2.c | 1 +
> include/block/block-common.h | 4 +++
> include/block/block-io.h | 2 +-
> tests/qemu-iotests/197 | 29 +++++++++++++++++++++
> tests/qemu-iotests/197.out | 24 +++++++++++++++++
> 8 files changed, 94 insertions(+), 30 deletions(-)
>
Ping