1 files changed, 115 insertions, 201 deletions

diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c
index 3c96c80e0992..d8c43d73e078 100644
--- a/kernel/sched/rt.c
+++ b/kernel/sched/rt.c
@@ -74,10 +74,6 @@ static void start_rt_bandwidth(struct rt_bandwidth *rt_b)
 	raw_spin_unlock(&rt_b->rt_runtime_lock);
 }
 
-#if defined(CONFIG_SMP) && defined(HAVE_RT_PUSH_IPI)
-static void push_irq_work_func(struct irq_work *work);
-#endif
-
 void init_rt_rq(struct rt_rq *rt_rq)
 {
 	struct rt_prio_array *array;
@@ -97,13 +93,6 @@ void init_rt_rq(struct rt_rq *rt_rq)
 	rt_rq->rt_nr_migratory = 0;
 	rt_rq->overloaded = 0;
 	plist_head_init(&rt_rq->pushable_tasks);
-
-#ifdef HAVE_RT_PUSH_IPI
-	rt_rq->push_flags = 0;
-	rt_rq->push_cpu = nr_cpu_ids;
-	raw_spin_lock_init(&rt_rq->push_lock);
-	init_irq_work(&rt_rq->push_work, push_irq_work_func);
-#endif
 #endif /* CONFIG_SMP */
 	/* We start is dequeued state, because no RT tasks are queued */
 	rt_rq->rt_queued = 0;
@@ -1876,241 +1865,166 @@ static void push_rt_tasks(struct rq *rq)
 }
 
 #ifdef HAVE_RT_PUSH_IPI
+
 /*
- * The search for the next cpu always starts at rq->cpu and ends
- * when we reach rq->cpu again. It will never return rq->cpu.
- * This returns the next cpu to check, or nr_cpu_ids if the loop
- * is complete.
+ * When a high priority task schedules out from a CPU and a lower priority
+ * task is scheduled in, a check is made to see if there's any RT tasks
+ * on other CPUs that are waiting to run because a higher priority RT task
+ * is currently running on its CPU. In this case, the CPU with multiple RT
+ * tasks queued on it (overloaded) needs to be notified that a CPU has opened
+ * up that may be able to run one of its non-running queued RT tasks.
+ *
+ * All CPUs with overloaded RT tasks need to be notified as there is currently
+ * no way to know which of these CPUs have the highest priority task waiting
+ * to run. Instead of trying to take a spinlock on each of these CPUs,
+ * which has shown to cause large latency when done on machines with many
+ * CPUs, sending an IPI to the CPUs to have them push off the overloaded
+ * RT tasks waiting to run.
+ *
+ * Just sending an IPI to each of the CPUs is also an issue, as on large
+ * count CPU machines, this can cause an IPI storm on a CPU, especially
+ * if its the only CPU with multiple RT tasks queued, and a large number
+ * of CPUs scheduling a lower priority task at the same time.
+ *
+ * Each root domain has its own irq work function that can iterate over
+ * all CPUs with RT overloaded tasks. Since all CPUs with overloaded RT
+ * tassk must be checked if there's one or many CPUs that are lowering
+ * their priority, there's a single irq work iterator that will try to
+ * push off RT tasks that are waiting to run.
+ *
+ * When a CPU schedules a lower priority task, it will kick off the
+ * irq work iterator that will jump to each CPU with overloaded RT tasks.
+ * As it only takes the first CPU that schedules a lower priority task
+ * to start the process, the rto_start variable is incremented and if
+ * the atomic result is one, then that CPU will try to take the rto_lock.
+ * This prevents high contention on the lock as the process handles all
+ * CPUs scheduling lower priority tasks.
+ *
+ * All CPUs that are scheduling a lower priority task will increment the
+ * rt_loop_next variable. This will make sure that the irq work iterator
+ * checks all RT overloaded CPUs whenever a CPU schedules a new lower
+ * priority task, even if the iterator is in the middle of a scan. Incrementing
+ * the rt_loop_next will cause the iterator to perform another scan.
  *
- * rq->rt.push_cpu holds the last cpu returned by this function,
- * or if this is the first instance, it must hold rq->cpu.
  */
 static int rto_next_cpu(struct rq *rq)
 {
-	int prev_cpu = rq->rt.push_cpu;
+	struct root_domain *rd = rq->rd;
+	int next;
 	int cpu;
 
-	cpu = cpumask_next(prev_cpu, rq->rd->rto_mask);
-
 	/*
-	 * If the previous cpu is less than the rq's CPU, then it already
-	 * passed the end of the mask, and has started from the beginning.
-	 * We end if the next CPU is greater or equal to rq's CPU.
+	 * When starting the IPI RT pushing, the rto_cpu is set to -1,
+	 * rt_next_cpu() will simply return the first CPU found in
+	 * the rto_mask.
+	 *
+	 * If rto_next_cpu() is called with rto_cpu is a valid cpu, it
+	 * will return the next CPU found in the rto_mask.
+	 *
+	 * If there are no more CPUs left in the rto_mask, then a check is made
+	 * against rto_loop and rto_loop_next. rto_loop is only updated with
+	 * the rto_lock held, but any CPU may increment the rto_loop_next
+	 * without any locking.
 	 */
-	if (prev_cpu < rq->cpu) {
-		if (cpu >= rq->cpu)
-			return nr_cpu_ids;
+	for (;;) {
 
-	} else if (cpu >= nr_cpu_ids) {
-		/*
-		 * We passed the end of the mask, start at the beginning.
-		 * If the result is greater or equal to the rq's CPU, then
-		 * the loop is finished.
-		 */
-		cpu = cpumask_first(rq->rd->rto_mask);
-		if (cpu >= rq->cpu)
-			return nr_cpu_ids;
-	}
-	rq->rt.push_cpu = cpu;
+		/* When rto_cpu is -1 this acts like cpumask_first() */
+		cpu = cpumask_next(rd->rto_cpu, rd->rto_mask);
 
-	/* Return cpu to let the caller know if the loop is finished or not */
-	return cpu;
-}
+		rd->rto_cpu = cpu;
 
-static int find_next_push_cpu(struct rq *rq)
-{
-	struct rq *next_rq;
-	int cpu;
+		if (cpu < nr_cpu_ids)
+			return cpu;
 
-	while (1) {
-		cpu = rto_next_cpu(rq);
-		if (cpu >= nr_cpu_ids)
-			break;
-		next_rq = cpu_rq(cpu);
+		rd->rto_cpu = -1;
+
+		/*
+		 * ACQUIRE ensures we see the @rto_mask changes
+		 * made prior to the @next value observed.
+		 *
+		 * Matches WMB in rt_set_overload().
+		 */
+		next = atomic_read_acquire(&rd->rto_loop_next);
 
-		/* Make sure the next rq can push to this rq */
-		if (next_rq->rt.highest_prio.next < rq->rt.highest_prio.curr)
+		if (rd->rto_loop == next)
 			break;
+
+		rd->rto_loop = next;
 	}
 
-	return cpu;
+	return -1;
 }
 
-#define RT_PUSH_IPI_EXECUTING		1
-#define RT_PUSH_IPI_RESTART		2
+static inline bool rto_start_trylock(atomic_t *v)
+{
+	return !atomic_cmpxchg_acquire(v, 0, 1);
+}
 
-/*
- * When a high priority task schedules out from a CPU and a lower priority
- * task is scheduled in, a check is made to see if there's any RT tasks
- * on other CPUs that are waiting to run because a higher priority RT task
- * is currently running on its CPU. In this case, the CPU with multiple RT
- * tasks queued on it (overloaded) needs to be notified that a CPU has opened
- * up that may be able to run one of its non-running queued RT tasks.
- *
- * On large CPU boxes, there's the case that several CPUs could schedule
- * a lower priority task at the same time, in which case it will look for
- * any overloaded CPUs that it could pull a task from. To do this, the runqueue
- * lock must be taken from that overloaded CPU. Having 10s of CPUs all fighting
- * for a single overloaded CPU's runqueue lock can produce a large latency.
- * (This has actually been observed on large boxes running cyclictest).
- * Instead of taking the runqueue lock of the overloaded CPU, each of the
- * CPUs that scheduled a lower priority task simply sends an IPI to the
- * overloaded CPU. An IPI is much cheaper than taking an runqueue lock with
- * lots of contention. The overloaded CPU will look to push its non-running
- * RT task off, and if it does, it can then ignore the other IPIs coming
- * in, and just pass those IPIs off to any other overloaded CPU.
- *
- * When a CPU schedules a lower priority task, it only sends an IPI to
- * the "next" CPU that has overloaded RT tasks. This prevents IPI storms,
- * as having 10 CPUs scheduling lower priority tasks and 10 CPUs with
- * RT overloaded tasks, would cause 100 IPIs to go out at once.
- *
- * The overloaded RT CPU, when receiving an IPI, will try to push off its
- * overloaded RT tasks and then send an IPI to the next CPU that has
- * overloaded RT tasks. This stops when all CPUs with overloaded RT tasks
- * have completed. Just because a CPU may have pushed off its own overloaded
- * RT task does not mean it should stop sending the IPI around to other
- * overloaded CPUs. There may be another RT task waiting to run on one of
- * those CPUs that are of higher priority than the one that was just
- * pushed.
- *
- * An optimization that could possibly be made is to make a CPU array similar
- * to the cpupri array mask of all running RT tasks, but for the overloaded
- * case, then the IPI could be sent to only the CPU with the highest priority
- * RT task waiting, and that CPU could send off further IPIs to the CPU with
- * the next highest waiting task. Since the overloaded case is much less likely
- * to happen, the complexity of this implementation may not be worth it.
- * Instead, just send an IPI around to all overloaded CPUs.
- *
- * The rq->rt.push_flags holds the status of the IPI that is going around.
- * A run queue can only send out a single IPI at a time. The possible flags
- * for rq->rt.push_flags are:
- *
- *    (None or zero):		No IPI is going around for the current rq
- *    RT_PUSH_IPI_EXECUTING:	An IPI for the rq is being passed around
- *    RT_PUSH_IPI_RESTART:	The priority of the running task for the rq
- *				has changed, and the IPI should restart
- *				circulating the overloaded CPUs again.
- *
- * rq->rt.push_cpu contains the CPU that is being sent the IPI. It is updated
- * before sending to the next CPU.
- *
- * Instead of having all CPUs that schedule a lower priority task send
- * an IPI to the same "first" CPU in the RT overload mask, they send it
- * to the next overloaded CPU after their own CPU. This helps distribute
- * the work when there's more than one overloaded CPU and multiple CPUs
- * scheduling in lower priority tasks.
- *
- * When a rq schedules a lower priority task than what was currently
- * running, the next CPU with overloaded RT tasks is examined first.
- * That is, if CPU 1 and 5 are overloaded, and CPU 3 schedules a lower
- * priority task, it will send an IPI first to CPU 5, then CPU 5 will
- * send to CPU 1 if it is still overloaded. CPU 1 will clear the
- * rq->rt.push_flags if RT_PUSH_IPI_RESTART is not set.
- *
- * The first CPU to notice IPI_RESTART is set, will clear that flag and then
- * send an IPI to the next overloaded CPU after the rq->cpu and not the next
- * CPU after push_cpu. That is, if CPU 1, 4 and 5 are overloaded when CPU 3
- * schedules a lower priority task, and the IPI_RESTART gets set while the
- * handling is being done on CPU 5, it will clear the flag and send it back to
- * CPU 4 instead of CPU 1.
- *
- * Note, the above logic can be disabled by turning off the sched_feature
- * RT_PUSH_IPI. Then the rq lock of the overloaded CPU will simply be
- * taken by the CPU requesting a pull and the waiting RT task will be pulled
- * by that CPU. This may be fine for machines with few CPUs.
- */
-static void tell_cpu_to_push(struct rq *rq)
+static inline void rto_start_unlock(atomic_t *v)
 {
-	int cpu;
+	atomic_set_release(v, 0);
+}
 
-	if (rq->rt.push_flags & RT_PUSH_IPI_EXECUTING) {
-		raw_spin_lock(&rq->rt.push_lock);
-		/* Make sure it's still executing */
-		if (rq->rt.push_flags & RT_PUSH_IPI_EXECUTING) {
-			/*
-			 * Tell the IPI to restart the loop as things have
-			 * changed since it started.
-			 */
-			rq->rt.push_flags |= RT_PUSH_IPI_RESTART;
-			raw_spin_unlock(&rq->rt.push_lock);
-			return;
-		}
-		raw_spin_unlock(&rq->rt.push_lock);
-	}
+static void tell_cpu_to_push(struct rq *rq)
+{
+	int cpu = -1;
 
-	/* When here, there's no IPI going around */
+	/* Keep the loop going if the IPI is currently active */
+	atomic_inc(&rq->rd->rto_loop_next);
 
-	rq->rt.push_cpu = rq->cpu;
-	cpu = find_next_push_cpu(rq);
-	if (cpu >= nr_cpu_ids)
+	/* Only one CPU can initiate a loop at a time */
+	if (!rto_start_trylock(&rq->rd->rto_loop_start))
 		return;
 
-	rq->rt.push_flags = RT_PUSH_IPI_EXECUTING;
+	raw_spin_lock(&rq->rd->rto_lock);
+
+	/*
+	 * The rto_cpu is updated under the lock, if it has a valid cpu
+	 * then the IPI is still running and will continue due to the
+	 * update to loop_next, and nothing needs to be done here.
+	 * Otherwise it is finishing up and an ipi needs to be sent.
+	 */
+	if (rq->rd->rto_cpu < 0)
+		cpu = rto_next_cpu(rq);
 
-	irq_work_queue_on(&rq->rt.push_work, cpu);
+	raw_spin_unlock(&rq->rd->rto_lock);
+
+	rto_start_unlock(&rq->rd->rto_loop_start);
+
+	if (cpu >= 0)
+		irq_work_queue_on(&rq->rd->rto_push_work, cpu);
 }
 
 /* Called from hardirq context */
-static void try_to_push_tasks(void *arg)
+void rto_push_irq_work_func(struct irq_work *work)
 {
-	struct rt_rq *rt_rq = arg;
-	struct rq *rq, *src_rq;
-	int this_cpu;
+	struct rq *rq;
 	int cpu;
 
-	this_cpu = rt_rq->push_cpu;
+	rq = this_rq();
 
-	/* Paranoid check */
-	BUG_ON(this_cpu != smp_processor_id());
-
-	rq = cpu_rq(this_cpu);
-	src_rq = rq_of_rt_rq(rt_rq);
-
-again:
+	/*
+	 * We do not need to grab the lock to check for has_pushable_tasks.
+	 * When it gets updated, a check is made if a push is possible.
+	 */
 	if (has_pushable_tasks(rq)) {
 		raw_spin_lock(&rq->lock);
-		push_rt_task(rq);
+		push_rt_tasks(rq);
 		raw_spin_unlock(&rq->lock);
 	}
 
-	/* Pass the IPI to the next rt overloaded queue */
-	raw_spin_lock(&rt_rq->push_lock);
-	/*
-	 * If the source queue changed since the IPI went out,
-	 * we need to restart the search from that CPU again.
-	 */
-	if (rt_rq->push_flags & RT_PUSH_IPI_RESTART) {
-		rt_rq->push_flags &= ~RT_PUSH_IPI_RESTART;
-		rt_rq->push_cpu = src_rq->cpu;
-	}
+	raw_spin_lock(&rq->rd->rto_lock);
 
-	cpu = find_next_push_cpu(src_rq);
+	/* Pass the IPI to the next rt overloaded queue */
+	cpu = rto_next_cpu(rq);
 
-	if (cpu >= nr_cpu_ids)
-		rt_rq->push_flags &= ~RT_PUSH_IPI_EXECUTING;
-	raw_spin_unlock(&rt_rq->push_lock);
+	raw_spin_unlock(&rq->rd->rto_lock);
 
-	if (cpu >= nr_cpu_ids)
+	if (cpu < 0)
 		return;
 
-	/*
-	 * It is possible that a restart caused this CPU to be
-	 * chosen again. Don't bother with an IPI, just see if we
-	 * have more to push.
-	 */
-	if (unlikely(cpu == rq->cpu))
-		goto again;
-
 	/* Try the next RT overloaded CPU */
-	irq_work_queue_on(&rt_rq->push_work, cpu);
-}
-
-static void push_irq_work_func(struct irq_work *work)
-{
-	struct rt_rq *rt_rq = container_of(work, struct rt_rq, push_work);
-
-	try_to_push_tasks(rt_rq);
+	irq_work_queue_on(&rq->rd->rto_push_work, cpu);
 }
 #endif /* HAVE_RT_PUSH_IPI */