On 03/07/2024 14:19, Christoph Hellwig wrote:
On Wed, Jul 03, 2024 at 01:20:26PM +0100, John Garry wrote:
So if we stop validating the limits in a., there is a user-visible change in
behaviour (as we stop rejecting invalid limits from the NBD_SET_BLKSIZE
ioctl).
We could add a "dryrun" option to queue_limits_commit_update() (and call
that instead of blk_validate_block_size(), which is effectively the same as
calling blk_validate_block_size()). Or we can keep
nbd as the only blk_validate_limits() user (outside the block layer).
I'd just keep the extra external blk_validate_block_size call in nbd.c.
Maybe add a comment to the blk_validate_block_size declaration that
drivers should not bother with it as it's already done by
blk_validate_limits.
ok, fine. It's a bit unfortunate that blk_validate_block_size() won't be
internal to the block layer.
