Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit 057f3fad authored by Peter Zijlstra's avatar Peter Zijlstra Committed by Ingo Molnar
Browse files

sched: Fix sched_domain iterations vs. RCU 



Vladis Kletnieks reported a new RCU debug warning in the scheduler.

Since commit dce840a0 ("sched: Dynamically allocate sched_domain/
sched_group data-structures") the sched_domain trees are protected by
RCU instead of RCU-sched.

This means that we need to include rcu_read_lock() protection when we
iterate them since disabling preemption doesn't suffice anymore.

Reported-by: default avatar <Valdis.Kletnieks@vt.edu>
Signed-off-by: default avatarPeter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1302882741.2388.241.camel@twins


Signed-off-by: default avatarIngo Molnar <mingo@elte.hu>
parent 2f36825b

kernel/sched.c

+11 −3
Original line number Diff line number Diff line
@@ -1208,11 +1208,17 @@ int get_nohz_timer_target(void) 
	int i;

	struct sched_domain *sd;



	rcu_read_lock();

	for_each_domain(cpu, sd) {

		for_each_cpu(i, sched_domain_span(sd))

			if (!idle_cpu(i))

				return i;

		for_each_cpu(i, sched_domain_span(sd)) {

			if (!idle_cpu(i)) {

				cpu = i;

				goto unlock;

			}

		}

	}

unlock:

	rcu_read_unlock();

	return cpu;

}

/*
@@ -2415,12 +2421,14 @@ ttwu_stat(struct task_struct *p, int cpu, int wake_flags) 
		struct sched_domain *sd;



		schedstat_inc(p, se.statistics.nr_wakeups_remote);

		rcu_read_lock();

		for_each_domain(this_cpu, sd) {

			if (cpumask_test_cpu(cpu, sched_domain_span(sd))) {

				schedstat_inc(sd, ttwu_wake_remote);

				break;

			}

		}

		rcu_read_unlock();

	}

#endif /* CONFIG_SMP */