bfq/mq-deadline: remove redundant check for passthrough request
authorLin Feng <linf@wangsu.com>
Thu, 15 Apr 2021 03:43:26 +0000 (11:43 +0800)
committerJens Axboe <axboe@kernel.dk>
Fri, 16 Apr 2021 12:08:52 +0000 (06:08 -0600)
commit7687b38ae470f01749e420079c36cccb24b8619a
treec650eec206afe0912baeaf2b2bdbf51533eace6c
parent8d663f34f8afcf5fc6a84c3cc4fa28cc84d58e39
bfq/mq-deadline: remove redundant check for passthrough request

Since commit 01e99aeca39796003 'blk-mq: insert passthrough request into
hctx->dispatch directly', passthrough request should not appear in
IO-scheduler any more, so blk_rq_is_passthrough checking in addon IO
schedulers is redundant.

(Notes: this patch passes generic IO load test with hdds under SAS
controller and hdds under AHCI controller but obviously not covers all.
Not sure if passthrough request can still escape into IO scheduler from
blk_mq_sched_insert_requests, which is used by blk_mq_flush_plug_list and
has lots of indirect callers.)

Signed-off-by: Lin Feng <linf@wangsu.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
block/bfq-iosched.c
block/mq-deadline.c