> Wouldn't it make sense to backport the changes to set the virt_boundary
> (which probably still is the SG_GAPS flag in such an old kernel)?
We can make storvsc use SG_GAPS. But the following patch is missing in 4.1 stable block layer to make this work on some I/O situations. Backporting is more difficult and affect other code.
commit 5e7c4274a70aa2d6f485996d0ca1dad52d0039ca
Author: Jens Axboe <axboe@xxxxxx>
Date: Thu Sep 3 19:28:20 2015 +0300
block: Check for gaps on front and back merges
We are checking for gaps to previous bio_vec, which can
only detect back merges gaps. Moreover, at the point where
we check for a gap, we don't know if we will attempt a back
or a front merge. Thus, check for gap to prev in a back merge
attempt and check for a gap to next in a front merge attempt.
Signed-off-by: Jens Axboe <axboe@xxxxxx>
[sagig: Minor rename change]
Signed-off-by: Sagi Grimberg <sagig@xxxxxxxxxxxx>