blk-mq: plug request for shared sbitmap
authorMing Lei <ming.lei@redhat.com>
Fri, 14 May 2021 02:20:52 +0000 (10:20 +0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 19 May 2021 08:13:14 +0000 (10:13 +0200)
commitc9c1ed08c174c2fa88fe1badbb876a7317a8224f
tree721cefd3a0702d92f8ed794a32ce6894073e88ba
parentcde4b55cfb24522dcbba80bbdb0c082303e76c43
blk-mq: plug request for shared sbitmap

[ Upstream commit 03f26d8f11403295de445b6e4e0e57ac57755791 ]

In case of shared sbitmap, request won't be held in plug list any more
sine commit 32bc15afed04 ("blk-mq: Facilitate a shared sbitmap per
tagset"), this way makes request merge from flush plug list & batching
submission not possible, so cause performance regression.

Yanhui reports performance regression when running sequential IO
test(libaio, 16 jobs, 8 depth for each job) in VM, and the VM disk
is emulated with image stored on xfs/megaraid_sas.

Fix the issue by recovering original behavior to allow to hold request
in plug list.

Cc: Yanhui Ma <yama@redhat.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: kashyap.desai@broadcom.com
Fixes: 32bc15afed04 ("blk-mq: Facilitate a shared sbitmap per tagset")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20210514022052.1047665-1-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
block/blk-mq.c