diff options
| author | Harshad Shirwadkar <harshadshirwadkar@gmail.com> | 2019-10-05 11:59:27 -0700 |
|---|---|---|
| committer | Jens Axboe <axboe@kernel.dk> | 2019-10-06 09:26:41 -0600 |
| commit | b84477d3ebb96294f87dc3161e53fa8fe22d9bfd (patch) | |
| tree | c69d5489e1f70ed4b22a7753ac61c26660b8454a /block/blk-wbt.c | |
| parent | 0e48f51cbbfbdb79149806de14dcb8bf0f01ca94 (diff) | |
blk-wbt: fix performance regression in wbt scale_up/scale_down
scale_up wakes up waiters after scaling up. But after scaling max, it
should not wake up more waiters as waiters will not have anything to
do. This patch fixes this by making scale_up (and also scale_down)
return when threshold is reached.
This bug causes increased fdatasync latency when fdatasync and dd
conv=sync are performed in parallel on 4.19 compared to 4.14. This
bug was introduced during refactoring of blk-wbt code.
Fixes: a79050434b45 ("blk-rq-qos: refactor out common elements of blk-wbt")
Cc: stable@vger.kernel.org
Cc: Josef Bacik <jbacik@fb.com>
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Diffstat (limited to 'block/blk-wbt.c')
| -rw-r--r-- | block/blk-wbt.c | 6 |
1 files changed, 4 insertions, 2 deletions
diff --git a/block/blk-wbt.c b/block/blk-wbt.c index 8af553a0ba00..8641ba9793c5 100644 --- a/block/blk-wbt.c +++ b/block/blk-wbt.c @@ -308,7 +308,8 @@ static void calc_wb_limits(struct rq_wb *rwb) static void scale_up(struct rq_wb *rwb) { - rq_depth_scale_up(&rwb->rq_depth); + if (!rq_depth_scale_up(&rwb->rq_depth)) + return; calc_wb_limits(rwb); rwb->unknown_cnt = 0; rwb_wake_all(rwb); @@ -317,7 +318,8 @@ static void scale_up(struct rq_wb *rwb) static void scale_down(struct rq_wb *rwb, bool hard_throttle) { - rq_depth_scale_down(&rwb->rq_depth, hard_throttle); + if (!rq_depth_scale_down(&rwb->rq_depth, hard_throttle)) + return; calc_wb_limits(rwb); rwb->unknown_cnt = 0; rwb_trace_step(rwb, "scale down"); |