[v9,0/2] Better discard support for block devices

Message

Andrzej Pietrasiewicz April 3, 2020, 2:43 p.m. UTC

The series is a v8 with small issues eliminated (see changelog below)
and more Reviewed-by tags - from Christoph.

This series addresses some errors seen when using the loop
device directly backed by a block device. The first change plumbs
out the correct error message, and the second change prevents the
error from occurring in many cases.

The errors look like this:
[   90.880875] print_req_error: I/O error, dev loop5, sector 0

The errors occur when trying to do a discard or write zeroes operation
on a loop device backed by a block device that does not support write zeroes.
Firstly, the error itself is incorrectly reported as I/O error, but is
actually EOPNOTSUPP. The first patch plumbs out EOPNOTSUPP to properly
report the error.

The second patch prevents these errors from occurring by mirroring the
zeroing capabilities of the underlying block device into the loop device.
Before this change, discard was always reported as being supported, and
the loop device simply turns around and does an fallocate operation on the
backing device. After this change, backing block devices that do support
zeroing will continue to work as before, and continue to get all the
benefits of doing that. Backing devices that do not support zeroing will
fail earlier, avoiding hitting the loop device at all and ultimately
avoiding this error in the logs.

I can also confirm that this fixes test block/003 in the blktests, when
running blktests on a loop device backed by a block device.

Changes in v9:
- omitted redundant braces in a conditional statement
- moved "backingq" variable to a more local scope because it is used
only there

Changes in v8:
- Improved comment in loop_config_discard() (Darrick/Evan)

Changes in v7:
- Use errno_to_blk_status() (Christoph)
- Rebase on top of Darrick's patch
- Tweak opening line of commit description (Darrick)

Changes in v6:
- Updated tags

Changes in v5:
- Don't mirror discard if lo_encrypt_key_size is non-zero (Gwendal)

Changes in v4:
- Mirror blkdev's write_zeroes into loopdev's discard_sectors.

Changes in v3:
- Updated tags
- Updated commit description

Changes in v2:
- Unnested error if statement (Bart)

Evan Green (2):
  loop: Report EOPNOTSUPP properly
  loop: Better discard support for block devices

 drivers/block/loop.c | 49 ++++++++++++++++++++++++++++++++------------
 1 file changed, 36 insertions(+), 13 deletions(-)

Comments

Jens Axboe April 3, 2020, 7:45 p.m. UTC | #1

On 4/3/20 8:43 AM, Andrzej Pietrasiewicz wrote:
> The series is a v8 with small issues eliminated (see changelog below)
> and more Reviewed-by tags - from Christoph.
> 
> This series addresses some errors seen when using the loop
> device directly backed by a block device. The first change plumbs
> out the correct error message, and the second change prevents the
> error from occurring in many cases.
> 
> The errors look like this:
> [   90.880875] print_req_error: I/O error, dev loop5, sector 0
> 
> The errors occur when trying to do a discard or write zeroes operation
> on a loop device backed by a block device that does not support write zeroes.
> Firstly, the error itself is incorrectly reported as I/O error, but is
> actually EOPNOTSUPP. The first patch plumbs out EOPNOTSUPP to properly
> report the error.
> 
> The second patch prevents these errors from occurring by mirroring the
> zeroing capabilities of the underlying block device into the loop device.
> Before this change, discard was always reported as being supported, and
> the loop device simply turns around and does an fallocate operation on the
> backing device. After this change, backing block devices that do support
> zeroing will continue to work as before, and continue to get all the
> benefits of doing that. Backing devices that do not support zeroing will
> fail earlier, avoiding hitting the loop device at all and ultimately
> avoiding this error in the logs.
> 
> I can also confirm that this fixes test block/003 in the blktests, when
> running blktests on a loop device backed by a block device.

Applied, thanks.