summaryrefslogtreecommitdiff
path: root/block
diff options
context:
space:
mode:
authorPaolo Valente <paolo.valente@linaro.org>2021-06-19 16:09:42 +0200
committerJens Axboe <axboe@kernel.dk>2021-06-21 15:03:41 -0600
commit511a2699237611b062df7798476bf3a1392910b9 (patch)
tree52387142d1f1dee29676e7233aafc6c531fff768 /block
parent76a8040817b4b9c69b53f9b326987fa891b4082a (diff)
block, bfq: let also stably merged queues enjoy weight raising
Merged bfq_queues are kept out of weight-raising (low-latency) mechanisms. The reason is that these queues are usually created for non-interactive and non-soft-real-time tasks. Yet this is not the case for stably-merged queues. These queues are merged just because they are created shortly after each other. So they may easily serve the I/O of an interactive or soft-real time application, if the application happens to spawn multiple processes. To address this issue, this commits lets also stably-merged queued enjoy weight raising. Signed-off-by: Paolo Valente <paolo.valente@linaro.org> Link: https://lore.kernel.org/r/20210619140948.98712-2-paolo.valente@linaro.org Signed-off-by: Jens Axboe <axboe@kernel.dk>
Diffstat (limited to 'block')
-rw-r--r--block/bfq-iosched.c15
1 files changed, 14 insertions, 1 deletions
diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c
index acd1f881273e..da2363f12e53 100644
--- a/block/bfq-iosched.c
+++ b/block/bfq-iosched.c
@@ -1729,10 +1729,23 @@ static void bfq_bfqq_handle_idle_busy_switch(struct bfq_data *bfqd,
bfqq->entity.new_weight == 40;
*interactive = !in_burst && idle_for_long_time &&
bfqq->entity.new_weight == 40;
+ /*
+ * Merged bfq_queues are kept out of weight-raising
+ * (low-latency) mechanisms. The reason is that these queues
+ * are usually created for non-interactive and
+ * non-soft-real-time tasks. Yet this is not the case for
+ * stably-merged queues. These queues are merged just because
+ * they are created shortly after each other. So they may
+ * easily serve the I/O of an interactive or soft-real time
+ * application, if the application happens to spawn multiple
+ * processes. So let also stably-merged queued enjoy weight
+ * raising.
+ */
wr_or_deserves_wr = bfqd->low_latency &&
(bfqq->wr_coeff > 1 ||
(bfq_bfqq_sync(bfqq) &&
- bfqq->bic && (*interactive || soft_rt)));
+ (bfqq->bic || RQ_BIC(rq)->stably_merged) &&
+ (*interactive || soft_rt)));
/*
* Using the last flag, update budget and check whether bfqq