[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20190822152037.15413-5-paolo.valente@linaro.org>
Date: Thu, 22 Aug 2019 17:20:37 +0200
From: Paolo Valente <paolo.valente@...aro.org>
To: Jens Axboe <axboe@...nel.dk>
Cc: linux-block@...r.kernel.org, linux-kernel@...r.kernel.org,
ulf.hansson@...aro.org, linus.walleij@...aro.org,
bfq-iosched@...glegroups.com, oleksandr@...alenko.name,
Paolo Valente <paolo.valente@...aro.org>
Subject: [PATCH 4/4] block, bfq: push up injection only after setting service time
If equal to 0, the injection limit for a bfq_queue is pushed to 1
after a first sample of the total service time of the I/O requests of
the queue is computed (to allow injection to start). Yet, because of a
mistake in the branch that performs this action, the push may happen
also in some other case. This commit fixes this issue.
Signed-off-by: Paolo Valente <paolo.valente@...aro.org>
---
block/bfq-iosched.c | 12 +++++++-----
1 file changed, 7 insertions(+), 5 deletions(-)
diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c
index ddac93e910fa..0319d6339822 100644
--- a/block/bfq-iosched.c
+++ b/block/bfq-iosched.c
@@ -5823,12 +5823,14 @@ static void bfq_update_inject_limit(struct bfq_data *bfqd,
*/
if ((bfqq->last_serv_time_ns == 0 && bfqd->rq_in_driver == 1) ||
tot_time_ns < bfqq->last_serv_time_ns) {
+ if (bfqq->last_serv_time_ns == 0) {
+ /*
+ * Now we certainly have a base value: make sure we
+ * start trying injection.
+ */
+ bfqq->inject_limit = max_t(unsigned int, 1, old_limit);
+ }
bfqq->last_serv_time_ns = tot_time_ns;
- /*
- * Now we certainly have a base value: make sure we
- * start trying injection.
- */
- bfqq->inject_limit = max_t(unsigned int, 1, old_limit);
} else if (!bfqd->rqs_injected && bfqd->rq_in_driver == 1)
/*
* No I/O injected and no request still in service in
--
2.20.1
Powered by blists - more mailing lists