[<prev] [next>] [day] [month] [year] [list]
Message-Id: <1539632566-4440-1-git-send-email-peng.hao2@zte.com.cn>
Date: Tue, 16 Oct 2018 03:42:46 +0800
From: Peng Hao <peng.hao2@....com.cn>
To: mingo@...hat.com, peterz@...radead.org
Cc: rostedt@...dmis.org, andrea.parri@...rulasolutions.com,
linux-kernel@...r.kernel.org, juri.lelli@...il.com,
Peng Hao <peng.hao2@....com.cn>
Subject: [PATCH v2] sched/dl: Reduce push_dl_task() retries
Improve push_dl_task() by propagating the double_lock_balance()
usage from find_lock_later_rq(), thereby reducing the number
of cases where we have to assume rq->lock was dropped.
Signed-off-by: Peng Hao <peng.hao2@....com.cn>
---
kernel/sched/deadline.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c
index 997ea7b..8e875ec 100644
--- a/kernel/sched/deadline.c
+++ b/kernel/sched/deadline.c
@@ -1979,7 +1979,7 @@ static struct rq *find_lock_later_rq(struct task_struct *task, struct rq *rq)
!dl_task(task) ||
!task_on_rq_queued(task))) {
double_unlock_balance(rq, later_rq);
- later_rq = NULL;
+ later_rq = RETRY_TASK;
break;
}
}
@@ -2063,7 +2063,9 @@ static int push_dl_task(struct rq *rq)
/* Will lock the rq it'll find */
later_rq = find_lock_later_rq(next_task, rq);
- if (!later_rq) {
+ if (!later_rq)
+ goto out;
+ if (later_rq == RETRY_TASK) {
struct task_struct *task;
/*
--
1.8.3.1
Powered by blists - more mailing lists