diff options
| author | Jan Kara <jack@suse.cz> | 2020-06-05 16:16:16 +0200 |
|---|---|---|
| committer | Jens Axboe <axboe@kernel.dk> | 2021-01-27 09:15:38 -0700 |
| commit | 41e76c85660c022c6bf5713bfb6c21e64a487cec (patch) | |
| tree | 470e919856dcf53ea851f883e705983580b7b131 /block | |
| parent | 49d1822bc05e702be1665ffc2092ec5711e77491 (diff) | |
| download | linux-41e76c85660c022c6bf5713bfb6c21e64a487cec.tar.bz2 | |
bfq: Avoid false bfq queue merging
bfq_setup_cooperator() uses bfqd->in_serv_last_pos so detect whether it
makes sense to merge current bfq queue with the in-service queue.
However if the in-service queue is freshly scheduled and didn't dispatch
any requests yet, bfqd->in_serv_last_pos is stale and contains value
from the previously scheduled bfq queue which can thus result in a bogus
decision that the two queues should be merged. This bug can be observed
for example with the following fio jobfile:
[global]
direct=0
ioengine=sync
invalidate=1
size=1g
rw=read
[reader]
numjobs=4
directory=/mnt
where the 4 processes will end up in the one shared bfq queue although
they do IO to physically very distant files (for some reason I was able to
observe this only with slice_idle=1ms setting).
Fix the problem by invalidating bfqd->in_serv_last_pos when switching
in-service queue.
Fixes: 058fdecc6de7 ("block, bfq: fix in-service-queue check for queue merging")
CC: stable@vger.kernel.org
Signed-off-by: Jan Kara <jack@suse.cz>
Acked-by: Paolo Valente <paolo.valente@linaro.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Diffstat (limited to 'block')
| -rw-r--r-- | block/bfq-iosched.c | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index 23e293d2943c..4157cfe99ae2 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -3007,6 +3007,7 @@ static void __bfq_set_in_service_queue(struct bfq_data *bfqd, } bfqd->in_service_queue = bfqq; + bfqd->in_serv_last_pos = 0; } /* |