Re: [PATCH v2 0/2] Fix the initialization of max_hw_wzeroes_unmap_sectors for stacking drivers

On 12/09/2025 07:16, Zhang Yi wrote:

Hi, Jens!

Can you take this patch set through the linux-block tree?

md raid maintainers,

please try to get this picked up ASAP. As things stand, all these RAID 
personalities will be broken for 6.17 on drives supporting/reporting 
write zeroes.

Thanks

Thanks,
Yi.

On 9/10/2025 7:11 PM, Zhang Yi wrote:

quoted

From: Zhang Yi <yi.zhang@huawei.com>

Changes since v1:
  - Improve commit messages in patch 1 by adding a simple reproduction
    case as Paul suggested and explaining the implementation differences
    between RAID 0 and RAID 1/10/5, no code changes.

v1: https://urldefense.com/v3/__https://lore.kernel.org/linux-block/20250825083320.797165-1-yi.zhang@huaweicloud.com/__;!!ACWV5N9M2RV99hQ!JAN4eq3ePrspeto0Hn7563lg3392Lh44jM1oTbgxbDoClxVOh5B73QhGD53f9tiLxuxfJCr51PyAP55COV2TTZAt$

This series fixes the initialization of max_hw_wzeroes_unmap_sectors in
queue_limits for all md raid and drbd drivers, preventing
blk_validate_limits() failures on underlying devices that support the
unmap write zeroes command.

Best regards,
Yi.

Zhang Yi (2):
   md: init queue_limits->max_hw_wzeroes_unmap_sectors parameter
   drbd: init queue_limits->max_hw_wzeroes_unmap_sectors parameter

  drivers/block/drbd/drbd_nl.c | 1 +
  drivers/md/md-linear.c       | 1 +
  drivers/md/raid0.c           | 1 +
  drivers/md/raid1.c           | 1 +
  drivers/md/raid10.c          | 1 +
  drivers/md/raid5.c           | 1 +
  6 files changed, 6 insertions(+)