[Devel] [PATCH hci-8.1 v11 00/10] dm-qcow2: truncate unused tail of image on suspend

[Konstantin Khorenko]

I won't apply this version, please send v12 with the changes below.

On 11/9/26 15:01, Alexander Ivanov wrote:
> Alexander Ivanov (10):
>   dm-qcow2: add qcow2_md_page_read_sync() helper
>   dm-qcow2: drop duplicate prefix from QC_ERR/QC_INFO messages
>   dm-qcow2: extend the file for a new refblock in allocate_cluster()
>   dm-qcow2: complete wbd before ending the data qio
>   dm-qcow2: flush lower md before clearing dirty bit on merge failure
>   dm-qcow2: keep the dirty bit of an image that needs check
>   dm-qcow2: flag the image for check if reftable relocation fails late
>   dm-qcow2: leak clusters allocated for a failed L1/L2 writeback
>   dm-qcow2: retry R1/R2 writeback that failed with ENOSPC
>   dm-qcow2: truncate unused tail of image on suspend

What to change for v12:

1. Patch 8, code: in prepare_l1l2_allocation(), queue the qio that
   allocates an L2-table cluster for an L1 update on
   l1.md->wbd->dependent_list instead of l1.md->wait_list, so a failed
   L1 writeback ends it with the error instead of sending it back to
   allocate another cluster. The new loop in complete_wbd() then has to
   skip a qio without ext.
2. Patch 8, commit message: say what is leaked: one cluster per failed
   L1/L2 write of a request.
3. Patch 9, comment and commit message: a page that failed before the
   flush gets up to two attempts during it, not "once" or "one more
   try". MD_DIRTY is dropped when the write is submitted, not by the
   failure.
4. Patch 9, commit message: say that with patch 8 a request whose L2
   write keeps failing with ENOSPC leaks one data cluster on every
   resubmission, until the space is back.
5. Patch 10, commit message: the scan is skipped on md_writeback_error,
   which a late reftable relocation failure (patch 7) sets too, not only
   a failed R1/R2 writeback.
6. Cover letter: the patch 9 line of the changelog says "once" as well,
   fix it as in item 3.

Separate patches, not needed for v12:

1. A REQ_OP_FLUSH whose fsync fails with ENOSPC ends with
   BLK_STS_NOSPC, and qrq_endio() hits WARN_ON_ONCE(!op_is_write())
   for it. This is older than the series.

-- 
Konstantin Khorenko,
Virtuozzo Linux Kernel Team