Re: [PATCH v3 0/6] Support for Zoned UFS

[Niklas Cassel <cassel@kernel.org>]

Hello Daejun,

On Tue, Oct 20, 2026 at 09:16:47AM +0900, Daejun Park wrote:
> This series lets a ufs-lu be a host-managed zoned block device (ZBC)
> with zoned=on, so that the zoned storage stack, for example f2fs on
> host-managed zones, can be tested on UFS without zoned UFS hardware.
> Patches 1-3 fix the command path that the zoned code relies on and can
> be merged first; patch 4 documents zoned logical units, patch 5 adds
> them and patch 6 the tests.
> 
> Not modelled: persistent zone state, conventional and gap zones, and the
> zoned UFS descriptor bits.
> 
> Tested with the UFS qtests (also under ASan and UBSan) and with Linux
> 7.3-rc7 and android17-6.18 guests, legacy doorbell and MCQ: fio
> --zonemode=zbd with verification, blkzone, and f2fs on android17-6.18.

This series does not appear to be making use of the zoned block layer APIs
that currently exist in QEMU (see e.g. include/block/block-common.h and
block/io.c).

I know that NVMe currently does not use it, but I don't think we should
allow adding zoned storage support in additional storage protocols, if
they do not make use of the QEMU zoned block layer APIs.

Sam Li has a series that converts NVMe to use the QEMU zoned block layer
APIs, even though that series seems to have stalled:
https://lore.kernel.org/qemu-devel/20240122190013.41302-1-faithilikerun@gmail.com/


Currently, it is only virtio-blk which is making use of the QEMU zoned
block layer APIs, and the only supported backend is file-posix.c.

However, there is work in progress to add qcow2 backend support for
the QEMU zoned block layer APIs:
https://lore.kernel.org/qemu-devel/20261026121259.3354561-1-cassel@kernel.org/
Which requires the (still not merged...) prep series:
https://lore.kernel.org/qemu-devel/20260921142819.2916043-1-cassel@kernel.org/


If you make use of the QEMU zoned block layer APIs, you will get support
for conventional zones, and will not need to add a UFS property for
max_open_zones. Once the qcow2 series lands, you will also get persistence.


If you want inspiration, you can see how virtio_blk_zone_report_complete()
(hw/block/virtio-blk.c) converts between the QEMU zoned block layer zone
states BLK_ZS_* and the virtio-blk VIRTIO_BLK_ZS_* defined zone states.

If you want futher inspiration, you could look at my (work in progress)
branch (note: this branch is going to be rebased multiple times), which
converts between the QEMU zoned block layer zone states BLK_ZS_* and the
zone states defined by the Zone ATA Commands (ZAC) specification:
https://github.com/floatious/qemu/tree/zac/


Kind regards,
Niklas