sched: Allow interrupts to be enabled in idle balance

The idle_balance() code is treated special in the scheduler for no real apparent reason. There's an unlikely check for !rq->nr_running in the hot path of the scheduler, for just the idle load balancing.

As sched classes can have their own post schedule routine, and the idle task has its own scheduler class, we can move the idle_balance() into the idle task's post schedule routine, and out of the hot path of the scheduler (no sepecial check).

An added benefit for doing this is that this enables us to enable interrupts in the idle balance code. Shrinking the time interrupts are disabled while doing load balancing.

These patches have been thoroughly tested (on a 40 core box as well as a 4 core and UP config), and are prime for inclusion (for 3.9). But I'm posted them now as an RFC for comments.

Thanks!

-- Steve

The idle_balance() code is called to do task load balancing just before going to idle. This makes sense as the CPU is about to sleep anyway. But currently it's called in the middle of the scheduler and in a place that must have interrupts disabled. That means, while the load balancing is going on, if a task wakes up on this CPU, it wont get to run while the interrupts are disabled. The idle task doing the balancing will be clueless about it.

There's no real reason that the idle_balance() needs to be called in the middle of schedule anyway. The only benefit is that if a task is pulled to this CPU, it can be scheduled without the need to schedule the idle task. But load balancing and migrating the task makes a switch to idle and back negligible.

By using the post_schedule function pointer of the sched class, the unlikely branch in the hot path of the scheduler can be removed, and the idle task itself can do the load balancing.

Another advantage of this, is that by moving the idle_balance to the post_schedule routine, interrupts can now be enabled in the load balance allowing for interrupts and wakeups to still occur on that CPU while a balance is taking place. The enabling of interrupts will come as a separate patch.

Index: linux-trace.git/kernel/sched/core.c
===================================================================
@@ -2926,9 +2926,6 @@ need_resched:
 
 	pre_schedule(rq, prev);
 
-	if (unlikely(!rq->nr_running))
-		idle_balance(cpu, rq);
-
 	put_prev_task(rq, prev);
 	next = pick_next_task(rq);
 	clear_tsk_need_resched(prev);

Index: linux-trace.git/kernel/sched/idle_task.c
===================================================================
@@ -25,6 +25,7 @@ static void check_preempt_curr_idle(stru
 static struct task_struct *pick_next_task_idle(struct rq *rq)
 {
 	schedstat_inc(rq, sched_goidle);
+	rq->post_schedule = 1;
 	return rq->idle;
 }
 

@@ -69,6 +70,12 @@ static unsigned int get_rr_interval_idle
 	return 0;
 }
 
+
+static void post_schedule_idle(struct rq *rq)
+{
+	idle_balance(smp_processor_id(), rq);
+}
+
 /*
  * Simple, special scheduling class for the per-CPU idle tasks:
  */

@@ -91,6 +98,8 @@ const struct sched_class idle_sched_clas
 	.set_curr_task          = set_curr_task_idle,
 	.task_tick		= task_tick_idle,
 
+	.post_schedule		= post_schedule_idle,
+
 	.get_rr_interval	= get_rr_interval_idle,
 
 	.prio_changed		= prio_changed_idle,

---
So much for saying it was well tested. As my email box is separate from the development boxes, I didn't copy the write set of patches over. This patch was tested thoroughly on 40 core and 4 core, but not UP. The UP had some bugs that were fixed (but not sent :-p)

Here's the correct patch with the UP fixes. The other patch shouldn't be affected.
---

sched: Move idle_balance() to post_schedule

The idle_balance() code is called to do task load balancing just before going to idle. This makes sense as the CPU is about to sleep anyway. But currently it's called in the middle of the scheduler and in a place that must have interrupts disabled. That means, while the load balancing is going on, if a task wakes up on this CPU, it wont get to run while the interrupts are disabled. The idle task doing the balancing will be clueless about it.

There's no real reason that the idle_balance() needs to be called in the middle of schedule anyway. The only benefit is that if a task is pulled to this CPU, it can be scheduled without the need to schedule the idle task. But load balancing and migrating the task makes a switch to idle and back negligible.

By using the post_schedule function pointer of the sched class, the unlikely branch in the hot path of the scheduler can be removed, and the idle task itself can do the load balancing.

Another advantage of this, is that by moving the idle_balance to the post_schedule routine, interrupts can now be enabled in the load balance allowing for interrupts and wakeups to still occur on that CPU while a balance is taking place. The enabling of interrupts will come as a separate patch.

Index: linux-trace.git/kernel/sched/core.c
===================================================================
@@ -2926,9 +2926,6 @@ need_resched:
 
 	pre_schedule(rq, prev);
 
-	if (unlikely(!rq->nr_running))
-		idle_balance(cpu, rq);
-
 	put_prev_task(rq, prev);
 	next = pick_next_task(rq);
 	clear_tsk_need_resched(prev);

Index: linux-trace.git/kernel/sched/idle_task.c
===================================================================
@@ -13,6 +13,11 @@ select_task_rq_idle(struct task_struct *
 {
 	return task_cpu(p); /* IDLE tasks as never migrated */
 }
+
+static void post_schedule_idle(struct rq *rq)
+{
+	idle_balance(smp_processor_id(), rq);
+}
 #endif /* CONFIG_SMP */
 /*
  * Idle tasks are unconditionally rescheduled:

@@ -25,6 +30,10 @@ static void check_preempt_curr_idle(stru
 static struct task_struct *pick_next_task_idle(struct rq *rq)
 {
 	schedstat_inc(rq, sched_goidle);
+#ifdef CONFIG_SMP
+	/* Trigger the post schedule to do an idle_balance */
+	rq->post_schedule = 1;
+#endif
 	return rq->idle;
 }
 

@@ -86,6 +95,7 @@ const struct sched_class idle_sched_clas
 
 #ifdef CONFIG_SMP
 	.select_task_rq		= select_task_rq_idle,
+	.post_schedule		= post_schedule_idle,
 #endif
 
 	.set_curr_task          = set_curr_task_idle,

2012/12/22 Steven Rostedt <[email protected]>:

This cleanup looks nice as it does not only let us enable interrupts there but also debloats a bit the schedule() code from idle specific code. So it would be a pity if the optimization that goes away with your cleanup has any measurable impact. Is there any sensible benchmark that can be run against this patch? Something that may involve a lot of back and forth to idle with some bunch of tasks running around on other CPUs?

Now that the idle_balance is called from the post_schedule of the idle task sched class, it is safe to enable interrupts. This allows for better interaction of tasks waking up and other interrupts that are triggered while the idle balance is in process.

Preemption is still disabled, but perhaps that can change as well. That may need some more investigation.

Index: linux-trace.git/kernel/sched/fair.c
===================================================================
@@ -5231,9 +5231,10 @@ void idle_balance(int this_cpu, struct r
 	update_rq_runnable_avg(this_rq, 1);
 
 	/*
-	 * Drop the rq->lock, but keep IRQ/preempt disabled.
+	 * Drop the rq->lock, but keep preempt disabled.
 	 */
-	raw_spin_unlock(&this_rq->lock);
+	preempt_disable();
+	raw_spin_unlock_irq(&this_rq->lock);
 
 	update_blocked_averages(this_cpu);
 	rcu_read_lock();

@@ -5260,7 +5261,8 @@ void idle_balance(int this_cpu, struct r
 	}
 	rcu_read_unlock();
 
-	raw_spin_lock(&this_rq->lock);
+	raw_spin_lock_irq(&this_rq->lock);
+	preempt_enable();
 
 	if (pulled_task || time_after(jiffies, this_rq->next_balance)) {
 		/*