sched: Remove rcu_read_lock() from wake_affine()
Nikunj A. Dadhania [Tue, 7 Jun 2011 10:13:22 +0000 (15:13 +0530)]
wake_affine() is only called from one path: select_task_rq_fair(),
which already has the RCU read lock held.

Signed-off-by: Nikunj A. Dadhania <nikunj@linux.vnet.ibm.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Link: http://lkml.kernel.org/r/20110607101251.777.34547.stgit@IBM-009124035060.in.ibm.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>

kernel/sched_fair.c

index 433491c..eb98f77 100644 (file)
@@ -1481,7 +1481,6 @@ static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync)
         * effect of the currently running task from the load
         * of the current CPU:
         */
-       rcu_read_lock();
        if (sync) {
                tg = task_group(current);
                weight = current->se.load.weight;
@@ -1517,7 +1516,6 @@ static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync)
                balanced = this_eff_load <= prev_eff_load;
        } else
                balanced = true;
-       rcu_read_unlock();
 
        /*
         * If the currently running task will sleep within