Skip to contentSkip to conversation

sched: Move blk_schedule_flush_plug() out of __schedule()

tip:sched/urgent tip-bot for Thomas Gleixner 1 participant 1 message Aug 29, 2011

lore
Share & tools
lore
#0 lore raw
Commit-ID:  9c40cef2b799f9b5e7fa5de4d2ad3a0168ba118c
Gitweb:     http://git.kernel.org/tip/9c40cef2b799f9b5e7fa5de4d2ad3a0168ba118c
Author:     Thomas Gleixner <[email protected]>
AuthorDate: Wed, 22 Jun 2011 19:47:01 +0200
Committer:  Ingo Molnar <[email protected]>
CommitDate: Mon, 29 Aug 2011 12:26:59 +0200

sched: Move blk_schedule_flush_plug() out of __schedule()

There is no real reason to run blk_schedule_flush_plug() with interrupts and preemption disabled.

Move it into schedule() and call it when the task is going voluntarily to sleep. There might be false positives when the task is woken between that call and actually scheduling, but that's not really different from being woken immediately after switching away.

This fixes a deadlock in the scheduler where the blk_schedule_flush_plug() callchain enables interrupts and thereby allows a wakeup to happen of the task that's going to sleep.

---
 kernel/sched.c |   25 +++++++++++++++----------
 1 files changed, 15 insertions(+), 10 deletions(-)
kernel/sched.c +15 -10
index ec15e81..511732c 100644
@@ -4322,16 +4322,6 @@ need_resched:
 				if (to_wakeup)
 					try_to_wake_up_local(to_wakeup);
 			}
-
-			/*
-			 * If we are going to sleep and we have plugged IO
-			 * queued, make sure to submit it to avoid deadlocks.
-			 */
-			if (blk_needs_flush_plug(prev)) {
-				raw_spin_unlock(&rq->lock);
-				blk_schedule_flush_plug(prev);
-				raw_spin_lock(&rq->lock);
-			}
 		}
 		switch_count = &prev->nvcsw;
 	}
@@ -4370,8 +4360,23 @@ need_resched:
 		goto need_resched;
 }
 
+static inline void sched_submit_work(struct task_struct *tsk)
+{
+	if (!tsk->state)
+		return;
+	/*
+	 * If we are going to sleep and we have plugged IO queued,
+	 * make sure to submit it to avoid deadlocks.
+	 */
+	if (blk_needs_flush_plug(tsk))
+		blk_schedule_flush_plug(tsk);
+}
+
 asmlinkage void schedule(void)
 {
+	struct task_struct *tsk = current;
+
+	sched_submit_work(tsk);
 	__schedule();
 }
 EXPORT_SYMBOL(schedule);