X-Git-Url: http://ftp.safe.ca/?a=blobdiff_plain;f=kernel%2Fsched_rt.c;h=d2ea2828164e8ba34ac7e0a12ac91f08a75c0ff8;hb=217bfb5190813d97e6003f7692f8014655ffc02a;hp=9bf0d2a7304569a87ba4aa0c919bbb2531c39c06;hpb=eaa958402ea40851097d051f52ba1bb7a885efe9;p=safe%2Fjmp%2Flinux-2.6

diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c
index 9bf0d2a..d2ea282 100644
--- a/kernel/sched_rt.c
+++ b/kernel/sched_rt.c
@@ -3,13 +3,18 @@
  * policies)
  */
 
+#ifdef CONFIG_RT_GROUP_SCHED
+
+#define rt_entity_is_task(rt_se) (!(rt_se)->my_q)
+
 static inline struct task_struct *rt_task_of(struct sched_rt_entity *rt_se)
 {
+#ifdef CONFIG_SCHED_DEBUG
+	WARN_ON_ONCE(!rt_entity_is_task(rt_se));
+#endif
 	return container_of(rt_se, struct task_struct, rt);
 }
 
-#ifdef CONFIG_RT_GROUP_SCHED
-
 static inline struct rq *rq_of_rt_rq(struct rt_rq *rt_rq)
 {
 	return rt_rq->rq;
@@ -22,6 +27,13 @@ static inline struct rt_rq *rt_rq_of_se(struct sched_rt_entity *rt_se)
 
 #else /* CONFIG_RT_GROUP_SCHED */
 
+#define rt_entity_is_task(rt_se) (1)
+
+static inline struct task_struct *rt_task_of(struct sched_rt_entity *rt_se)
+{
+	return container_of(rt_se, struct task_struct, rt);
+}
+
 static inline struct rq *rq_of_rt_rq(struct rt_rq *rt_rq)
 {
 	return container_of(rt_rq, struct rq, rt);
@@ -73,7 +85,7 @@ static inline void rt_clear_overload(struct rq *rq)
 
 static void update_rt_migration(struct rt_rq *rt_rq)
 {
-	if (rt_rq->rt_nr_migratory && (rt_rq->rt_nr_running > 1)) {
+	if (rt_rq->rt_nr_migratory && rt_rq->rt_nr_total > 1) {
 		if (!rt_rq->overloaded) {
 			rt_set_overload(rq_of_rt_rq(rt_rq));
 			rt_rq->overloaded = 1;
@@ -86,6 +98,12 @@ static void update_rt_migration(struct rt_rq *rt_rq)
 
 static void inc_rt_migration(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq)
 {
+	if (!rt_entity_is_task(rt_se))
+		return;
+
+	rt_rq = &rq_of_rt_rq(rt_rq)->rt;
+
+	rt_rq->rt_nr_total++;
 	if (rt_se->nr_cpus_allowed > 1)
 		rt_rq->rt_nr_migratory++;
 
@@ -94,6 +112,12 @@ static void inc_rt_migration(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq)
 
 static void dec_rt_migration(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq)
 {
+	if (!rt_entity_is_task(rt_se))
+		return;
+
+	rt_rq = &rq_of_rt_rq(rt_rq)->rt;
+
+	rt_rq->rt_nr_total--;
 	if (rt_se->nr_cpus_allowed > 1)
 		rt_rq->rt_nr_migratory--;
 
@@ -112,6 +136,11 @@ static void dequeue_pushable_task(struct rq *rq, struct task_struct *p)
 	plist_del(&p->pushable_tasks, &rq->rt.pushable_tasks);
 }
 
+static inline int has_pushable_tasks(struct rq *rq)
+{
+	return !plist_head_empty(&rq->rt.pushable_tasks);
+}
+
 #else
 
 static inline void enqueue_pushable_task(struct rq *rq, struct task_struct *p)
@@ -298,7 +327,7 @@ static int do_balance_runtime(struct rt_rq *rt_rq)
 
 	weight = cpumask_weight(rd->span);
 
-	spin_lock(&rt_b->rt_runtime_lock);
+	raw_spin_lock(&rt_b->rt_runtime_lock);
 	rt_period = ktime_to_ns(rt_b->rt_period);
 	for_each_cpu(i, rd->span) {
 		struct rt_rq *iter = sched_rt_period_rt_rq(rt_b, i);
@@ -307,7 +336,7 @@ static int do_balance_runtime(struct rt_rq *rt_rq)
 		if (iter == rt_rq)
 			continue;
 
-		spin_lock(&iter->rt_runtime_lock);
+		raw_spin_lock(&iter->rt_runtime_lock);
 		/*
 		 * Either all rqs have inf runtime and there's nothing to steal
 		 * or __disable_runtime() below sets a specific rq to inf to
@@ -329,14 +358,14 @@ static int do_balance_runtime(struct rt_rq *rt_rq)
 			rt_rq->rt_runtime += diff;
 			more = 1;
 			if (rt_rq->rt_runtime == rt_period) {
-				spin_unlock(&iter->rt_runtime_lock);
+				raw_spin_unlock(&iter->rt_runtime_lock);
 				break;
 			}
 		}
 next:
-		spin_unlock(&iter->rt_runtime_lock);
+		raw_spin_unlock(&iter->rt_runtime_lock);
 	}
-	spin_unlock(&rt_b->rt_runtime_lock);
+	raw_spin_unlock(&rt_b->rt_runtime_lock);
 
 	return more;
 }
@@ -357,8 +386,8 @@ static void __disable_runtime(struct rq *rq)
 		s64 want;
 		int i;
 
-		spin_lock(&rt_b->rt_runtime_lock);
-		spin_lock(&rt_rq->rt_runtime_lock);
+		raw_spin_lock(&rt_b->rt_runtime_lock);
+		raw_spin_lock(&rt_rq->rt_runtime_lock);
 		/*
 		 * Either we're all inf and nobody needs to borrow, or we're
 		 * already disabled and thus have nothing to do, or we have
@@ -367,7 +396,7 @@ static void __disable_runtime(struct rq *rq)
 		if (rt_rq->rt_runtime == RUNTIME_INF ||
 				rt_rq->rt_runtime == rt_b->rt_runtime)
 			goto balanced;
-		spin_unlock(&rt_rq->rt_runtime_lock);
+		raw_spin_unlock(&rt_rq->rt_runtime_lock);
 
 		/*
 		 * Calculate the difference between what we started out with
@@ -389,7 +418,7 @@ static void __disable_runtime(struct rq *rq)
 			if (iter == rt_rq || iter->rt_runtime == RUNTIME_INF)
 				continue;
 
-			spin_lock(&iter->rt_runtime_lock);
+			raw_spin_lock(&iter->rt_runtime_lock);
 			if (want > 0) {
 				diff = min_t(s64, iter->rt_runtime, want);
 				iter->rt_runtime -= diff;
@@ -398,13 +427,13 @@ static void __disable_runtime(struct rq *rq)
 				iter->rt_runtime -= want;
 				want -= want;
 			}
-			spin_unlock(&iter->rt_runtime_lock);
+			raw_spin_unlock(&iter->rt_runtime_lock);
 
 			if (!want)
 				break;
 		}
 
-		spin_lock(&rt_rq->rt_runtime_lock);
+		raw_spin_lock(&rt_rq->rt_runtime_lock);
 		/*
 		 * We cannot be left wanting - that would mean some runtime
 		 * leaked out of the system.
@@ -416,8 +445,8 @@ balanced:
 		 * runtime - in which case borrowing doesn't make sense.
 		 */
 		rt_rq->rt_runtime = RUNTIME_INF;
-		spin_unlock(&rt_rq->rt_runtime_lock);
-		spin_unlock(&rt_b->rt_runtime_lock);
+		raw_spin_unlock(&rt_rq->rt_runtime_lock);
+		raw_spin_unlock(&rt_b->rt_runtime_lock);
 	}
 }
 
@@ -425,9 +454,9 @@ static void disable_runtime(struct rq *rq)
 {
 	unsigned long flags;
 
-	spin_lock_irqsave(&rq->lock, flags);
+	raw_spin_lock_irqsave(&rq->lock, flags);
 	__disable_runtime(rq);
-	spin_unlock_irqrestore(&rq->lock, flags);
+	raw_spin_unlock_irqrestore(&rq->lock, flags);
 }
 
 static void __enable_runtime(struct rq *rq)
@@ -443,13 +472,13 @@ static void __enable_runtime(struct rq *rq)
 	for_each_leaf_rt_rq(rt_rq, rq) {
 		struct rt_bandwidth *rt_b = sched_rt_bandwidth(rt_rq);
 
-		spin_lock(&rt_b->rt_runtime_lock);
-		spin_lock(&rt_rq->rt_runtime_lock);
+		raw_spin_lock(&rt_b->rt_runtime_lock);
+		raw_spin_lock(&rt_rq->rt_runtime_lock);
 		rt_rq->rt_runtime = rt_b->rt_runtime;
 		rt_rq->rt_time = 0;
 		rt_rq->rt_throttled = 0;
-		spin_unlock(&rt_rq->rt_runtime_lock);
-		spin_unlock(&rt_b->rt_runtime_lock);
+		raw_spin_unlock(&rt_rq->rt_runtime_lock);
+		raw_spin_unlock(&rt_b->rt_runtime_lock);
 	}
 }
 
@@ -457,9 +486,9 @@ static void enable_runtime(struct rq *rq)
 {
 	unsigned long flags;
 
-	spin_lock_irqsave(&rq->lock, flags);
+	raw_spin_lock_irqsave(&rq->lock, flags);
 	__enable_runtime(rq);
-	spin_unlock_irqrestore(&rq->lock, flags);
+	raw_spin_unlock_irqrestore(&rq->lock, flags);
 }
 
 static int balance_runtime(struct rt_rq *rt_rq)
@@ -467,9 +496,9 @@ static int balance_runtime(struct rt_rq *rt_rq)
 	int more = 0;
 
 	if (rt_rq->rt_time > rt_rq->rt_runtime) {
-		spin_unlock(&rt_rq->rt_runtime_lock);
+		raw_spin_unlock(&rt_rq->rt_runtime_lock);
 		more = do_balance_runtime(rt_rq);
-		spin_lock(&rt_rq->rt_runtime_lock);
+		raw_spin_lock(&rt_rq->rt_runtime_lock);
 	}
 
 	return more;
@@ -495,11 +524,11 @@ static int do_sched_rt_period_timer(struct rt_bandwidth *rt_b, int overrun)
 		struct rt_rq *rt_rq = sched_rt_period_rt_rq(rt_b, i);
 		struct rq *rq = rq_of_rt_rq(rt_rq);
 
-		spin_lock(&rq->lock);
+		raw_spin_lock(&rq->lock);
 		if (rt_rq->rt_time) {
 			u64 runtime;
 
-			spin_lock(&rt_rq->rt_runtime_lock);
+			raw_spin_lock(&rt_rq->rt_runtime_lock);
 			if (rt_rq->rt_throttled)
 				balance_runtime(rt_rq);
 			runtime = rt_rq->rt_runtime;
@@ -510,13 +539,13 @@ static int do_sched_rt_period_timer(struct rt_bandwidth *rt_b, int overrun)
 			}
 			if (rt_rq->rt_time || rt_rq->rt_nr_running)
 				idle = 0;
-			spin_unlock(&rt_rq->rt_runtime_lock);
+			raw_spin_unlock(&rt_rq->rt_runtime_lock);
 		} else if (rt_rq->rt_nr_running)
 			idle = 0;
 
 		if (enqueue)
 			sched_rt_rq_enqueue(rt_rq);
-		spin_unlock(&rq->lock);
+		raw_spin_unlock(&rq->lock);
 	}
 
 	return idle;
@@ -586,6 +615,8 @@ static void update_curr_rt(struct rq *rq)
 	curr->se.exec_start = rq->clock;
 	cpuacct_charge(curr, delta_exec);
 
+	sched_rt_avg_update(rq, delta_exec);
+
 	if (!rt_bandwidth_enabled())
 		return;
 
@@ -593,11 +624,11 @@ static void update_curr_rt(struct rq *rq)
 		rt_rq = rt_rq_of_se(rt_se);
 
 		if (sched_rt_runtime(rt_rq) != RUNTIME_INF) {
-			spin_lock(&rt_rq->rt_runtime_lock);
+			raw_spin_lock(&rt_rq->rt_runtime_lock);
 			rt_rq->rt_time += delta_exec;
 			if (sched_rt_runtime_exceeded(rt_rq))
 				resched_task(curr);
-			spin_unlock(&rt_rq->rt_runtime_lock);
+			raw_spin_unlock(&rt_rq->rt_runtime_lock);
 		}
 	}
 }
@@ -858,8 +889,6 @@ static void enqueue_task_rt(struct rq *rq, struct task_struct *p, int wakeup)
 
 	if (!task_current(rq, p) && p->rt.nr_cpus_allowed > 1)
 		enqueue_pushable_task(rq, p);
-
-	inc_cpu_load(rq, p->se.load.weight);
 }
 
 static void dequeue_task_rt(struct rq *rq, struct task_struct *p, int sleep)
@@ -870,8 +899,6 @@ static void dequeue_task_rt(struct rq *rq, struct task_struct *p, int sleep)
 	dequeue_rt_entity(rt_se);
 
 	dequeue_pushable_task(rq, p);
-
-	dec_cpu_load(rq, p->se.load.weight);
 }
 
 /*
@@ -911,10 +938,13 @@ static void yield_task_rt(struct rq *rq)
 #ifdef CONFIG_SMP
 static int find_lowest_rq(struct task_struct *task);
 
-static int select_task_rq_rt(struct task_struct *p, int sync)
+static int select_task_rq_rt(struct task_struct *p, int sd_flag, int flags)
 {
 	struct rq *rq = task_rq(p);
 
+	if (sd_flag != SD_BALANCE_WAKE)
+		return smp_processor_id();
+
 	/*
 	 * If the current task is an RT task, then
 	 * try to see if we can wake this RT task up on another
@@ -972,7 +1002,7 @@ static void check_preempt_equal_prio(struct rq *rq, struct task_struct *p)
 /*
  * Preempt the current task with a newly woken task if needed:
  */
-static void check_preempt_curr_rt(struct rq *rq, struct task_struct *p, int sync)
+static void check_preempt_curr_rt(struct rq *rq, struct task_struct *p, int flags)
 {
 	if (p->prio < rq->curr->prio) {
 		resched_task(rq->curr);
@@ -1048,6 +1078,14 @@ static struct task_struct *pick_next_task_rt(struct rq *rq)
 	if (p)
 		dequeue_pushable_task(rq, p);
 
+#ifdef CONFIG_SMP
+	/*
+	 * We detect this state here so that we can avoid taking the RQ
+	 * lock again later if there is no need to push
+	 */
+	rq->post_schedule = has_pushable_tasks(rq);
+#endif
+
 	return p;
 }
 
@@ -1115,29 +1153,12 @@ static struct task_struct *pick_next_highest_task_rt(struct rq *rq, int cpu)
 
 static DEFINE_PER_CPU(cpumask_var_t, local_cpu_mask);
 
-static inline int pick_optimal_cpu(int this_cpu,
-				   const struct cpumask *mask)
-{
-	int first;
-
-	/* "this_cpu" is cheaper to preempt than a remote processor */
-	if ((this_cpu != -1) && cpumask_test_cpu(this_cpu, mask))
-		return this_cpu;
-
-	first = cpumask_first(mask);
-	if (first < nr_cpu_ids)
-		return first;
-
-	return -1;
-}
-
 static int find_lowest_rq(struct task_struct *task)
 {
 	struct sched_domain *sd;
 	struct cpumask *lowest_mask = __get_cpu_var(local_cpu_mask);
 	int this_cpu = smp_processor_id();
 	int cpu      = task_cpu(task);
-	cpumask_var_t domain_mask;
 
 	if (task->rt.nr_cpus_allowed == 1)
 		return -1; /* No other targets possible */
@@ -1146,13 +1167,6 @@ static int find_lowest_rq(struct task_struct *task)
 		return -1; /* No targets found */
 
 	/*
-	 * Only consider CPUs that are usable for migration.
-	 * I guess we might want to change cpupri_find() to ignore those
-	 * in the first place.
-	 */
-	cpumask_and(lowest_mask, lowest_mask, cpu_active_mask);
-
-	/*
 	 * At this point we have built a mask of cpus representing the
 	 * lowest priority tasks in the system.  Now we want to elect
 	 * the best one based on our affinity and topology.
@@ -1167,28 +1181,26 @@ static int find_lowest_rq(struct task_struct *task)
 	 * Otherwise, we consult the sched_domains span maps to figure
 	 * out which cpu is logically closest to our hot cache data.
 	 */
-	if (this_cpu == cpu)
-		this_cpu = -1; /* Skip this_cpu opt if the same */
-
-	if (alloc_cpumask_var(&domain_mask, GFP_ATOMIC)) {
-		for_each_domain(cpu, sd) {
-			if (sd->flags & SD_WAKE_AFFINE) {
-				int best_cpu;
-
-				cpumask_and(domain_mask,
-					    sched_domain_span(sd),
-					    lowest_mask);
+	if (!cpumask_test_cpu(this_cpu, lowest_mask))
+		this_cpu = -1; /* Skip this_cpu opt if not among lowest */
 
-				best_cpu = pick_optimal_cpu(this_cpu,
-							    domain_mask);
+	for_each_domain(cpu, sd) {
+		if (sd->flags & SD_WAKE_AFFINE) {
+			int best_cpu;
 
-				if (best_cpu != -1) {
-					free_cpumask_var(domain_mask);
-					return best_cpu;
-				}
-			}
+			/*
+			 * "this_cpu" is cheaper to preempt than a
+			 * remote processor.
+			 */
+			if (this_cpu != -1 &&
+			    cpumask_test_cpu(this_cpu, sched_domain_span(sd)))
+				return this_cpu;
+
+			best_cpu = cpumask_first_and(lowest_mask,
+						     sched_domain_span(sd));
+			if (best_cpu < nr_cpu_ids)
+				return best_cpu;
 		}
-		free_cpumask_var(domain_mask);
 	}
 
 	/*
@@ -1196,7 +1208,13 @@ static int find_lowest_rq(struct task_struct *task)
 	 * just give the caller *something* to work with from the compatible
 	 * locations.
 	 */
-	return pick_optimal_cpu(this_cpu, lowest_mask);
+	if (this_cpu != -1)
+		return this_cpu;
+
+	cpu = cpumask_any(lowest_mask);
+	if (cpu < nr_cpu_ids)
+		return cpu;
+	return -1;
 }
 
 /* Will lock the rq it finds */
@@ -1228,7 +1246,7 @@ static struct rq *find_lock_lowest_rq(struct task_struct *task, struct rq *rq)
 				     task_running(rq, task) ||
 				     !task->se.on_rq)) {
 
-				spin_unlock(&lowest_rq->lock);
+				raw_spin_unlock(&lowest_rq->lock);
 				lowest_rq = NULL;
 				break;
 			}
@@ -1246,11 +1264,6 @@ static struct rq *find_lock_lowest_rq(struct task_struct *task, struct rq *rq)
 	return lowest_rq;
 }
 
-static inline int has_pushable_tasks(struct rq *rq)
-{
-	return !plist_head_empty(&rq->rt.pushable_tasks);
-}
-
 static struct task_struct *pick_next_pushable_task(struct rq *rq)
 {
 	struct task_struct *p;
@@ -1450,23 +1463,9 @@ static void pre_schedule_rt(struct rq *rq, struct task_struct *prev)
 		pull_rt_task(rq);
 }
 
-/*
- * assumes rq->lock is held
- */
-static int needs_post_schedule_rt(struct rq *rq)
-{
-	return has_pushable_tasks(rq);
-}
-
 static void post_schedule_rt(struct rq *rq)
 {
-	/*
-	 * This is only called if needs_post_schedule_rt() indicates that
-	 * we need to push tasks away
-	 */
-	spin_lock_irq(&rq->lock);
 	push_rt_tasks(rq);
-	spin_unlock_irq(&rq->lock);
 }
 
 /*
@@ -1722,6 +1721,17 @@ static void set_curr_task_rt(struct rq *rq)
 	dequeue_pushable_task(rq, p);
 }
 
+unsigned int get_rr_interval_rt(struct rq *rq, struct task_struct *task)
+{
+	/*
+	 * Time slice is 0 for SCHED_FIFO tasks
+	 */
+	if (task->policy == SCHED_RR)
+		return DEF_TIMESLICE;
+	else
+		return 0;
+}
+
 static const struct sched_class rt_sched_class = {
 	.next			= &fair_sched_class,
 	.enqueue_task		= enqueue_task_rt,
@@ -1742,7 +1752,6 @@ static const struct sched_class rt_sched_class = {
 	.rq_online              = rq_online_rt,
 	.rq_offline             = rq_offline_rt,
 	.pre_schedule		= pre_schedule_rt,
-	.needs_post_schedule	= needs_post_schedule_rt,
 	.post_schedule		= post_schedule_rt,
 	.task_wake_up		= task_wake_up_rt,
 	.switched_from		= switched_from_rt,
@@ -1751,6 +1760,8 @@ static const struct sched_class rt_sched_class = {
 	.set_curr_task          = set_curr_task_rt,
 	.task_tick		= task_tick_rt,
 
+	.get_rr_interval	= get_rr_interval_rt,
+
 	.prio_changed		= prio_changed_rt,
 	.switched_to		= switched_to_rt,
 };