On 28.10.25 19:08, Wesley Hershberger wrote:
> Thank you all for the discussion in the previous thread:
> 20251024-second-fix-3149-v1-1-d997fa3d5ce2@canonical.com
>
> I have verified that this patch resolves the issue from #3149 and passes
> the iotests for qcow2/raw/nbd that don't fail on master (for me). Thanks
> for the pointers on testing.
>
> I'm committed to finding a solution to this that is adequately
> understood and tested; I'm happy to continue to make
> adjustments/submissions as needed until we have a solution that is
> satisfying.
>
> Please let me know if any adjustments are needed; thanks for your
> patience and continued involvement.
>
> Signed-off-by: Wesley Hershberger <wesley.hershberger@canonical.com>
> ---
> Wesley Hershberger (2):
> iotests/257: Insert missing flush during blockdev-backup
> block: Drop detach_subchain for bdrv_replace_node
>
> block.c | 38 ++++----------------------------------
> tests/qemu-iotests/257 | 8 ++++++--
> tests/qemu-iotests/257.out | 14 +++++++-------
> 3 files changed, 17 insertions(+), 43 deletions(-)
> ---
> base-commit: 36076d24f04ea9dc3357c0fbe7bb14917375819c
> change-id: 20251027-third-fix-3149-582fd7382fb9
>
> Best regards,
Seems good to me:
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
(weak for test, as I don't know its logic. Still, this test-output change
doesn't look wrong in context of block permissions, so I don't care)
Also, I think the two commits should be merged into one (keep my r-b if you
do), as otherwise we break ./check.py on the first one and then fix on the
second, it may break git bisect.
--
Best regards,
Vladimir