1 files changed, 47 insertions, 139 deletions

diff --git a/kernel/sched.c b/kernel/sched.c
index 3ee2ae45125..4d46e90f59c 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -178,13 +178,6 @@ static unsigned int task_timeslice(task_t *p)
 #define task_hot(p, now, sd) ((long long) ((now) - (p)->last_ran)	\
 				< (long long) (sd)->cache_hot_time)
 
-void __put_task_struct_cb(struct rcu_head *rhp)
-{
-	__put_task_struct(container_of(rhp, struct task_struct, rcu));
-}
-
-EXPORT_SYMBOL_GPL(__put_task_struct_cb);
-
 /*
  * These are the runqueue data structures:
  */
@@ -215,7 +208,6 @@ struct runqueue {
 	 */
 	unsigned long nr_running;
 #ifdef CONFIG_SMP
-	unsigned long prio_bias;
 	unsigned long cpu_load[3];
 #endif
 	unsigned long long nr_switches;
@@ -669,68 +661,13 @@ static int effective_prio(task_t *p)
 	return prio;
 }
 
-#ifdef CONFIG_SMP
-static inline void inc_prio_bias(runqueue_t *rq, int prio)
-{
-	rq->prio_bias += MAX_PRIO - prio;
-}
-
-static inline void dec_prio_bias(runqueue_t *rq, int prio)
-{
-	rq->prio_bias -= MAX_PRIO - prio;
-}
-
-static inline void inc_nr_running(task_t *p, runqueue_t *rq)
-{
-	rq->nr_running++;
-	if (rt_task(p)) {
-		if (p != rq->migration_thread)
-			/*
-			 * The migration thread does the actual balancing. Do
-			 * not bias by its priority as the ultra high priority
-			 * will skew balancing adversely.
-			 */
-			inc_prio_bias(rq, p->prio);
-	} else
-		inc_prio_bias(rq, p->static_prio);
-}
-
-static inline void dec_nr_running(task_t *p, runqueue_t *rq)
-{
-	rq->nr_running--;
-	if (rt_task(p)) {
-		if (p != rq->migration_thread)
-			dec_prio_bias(rq, p->prio);
-	} else
-		dec_prio_bias(rq, p->static_prio);
-}
-#else
-static inline void inc_prio_bias(runqueue_t *rq, int prio)
-{
-}
-
-static inline void dec_prio_bias(runqueue_t *rq, int prio)
-{
-}
-
-static inline void inc_nr_running(task_t *p, runqueue_t *rq)
-{
-	rq->nr_running++;
-}
-
-static inline void dec_nr_running(task_t *p, runqueue_t *rq)
-{
-	rq->nr_running--;
-}
-#endif
-
 /*
  * __activate_task - move a task to the runqueue.
  */
 static inline void __activate_task(task_t *p, runqueue_t *rq)
 {
 	enqueue_task(p, rq->active);
-	inc_nr_running(p, rq);
+	rq->nr_running++;
 }
 
 /*
@@ -739,7 +676,7 @@ static inline void __activate_task(task_t *p, runqueue_t *rq)
 static inline void __activate_idle_task(task_t *p, runqueue_t *rq)
 {
 	enqueue_task_head(p, rq->active);
-	inc_nr_running(p, rq);
+	rq->nr_running++;
 }
 
 static int recalc_task_prio(task_t *p, unsigned long long now)
@@ -863,7 +800,7 @@ static void activate_task(task_t *p, runqueue_t *rq, int local)
  */
 static void deactivate_task(struct task_struct *p, runqueue_t *rq)
 {
-	dec_nr_running(p, rq);
+	rq->nr_running--;
 	dequeue_task(p, p->array);
 	p->array = NULL;
 }
@@ -1007,61 +944,27 @@ void kick_process(task_t *p)
  * We want to under-estimate the load of migration sources, to
  * balance conservatively.
  */
-static unsigned long __source_load(int cpu, int type, enum idle_type idle)
+static inline unsigned long source_load(int cpu, int type)
 {
 	runqueue_t *rq = cpu_rq(cpu);
-	unsigned long running = rq->nr_running;
-	unsigned long source_load, cpu_load = rq->cpu_load[type-1],
-		load_now = running * SCHED_LOAD_SCALE;
-
+	unsigned long load_now = rq->nr_running * SCHED_LOAD_SCALE;
 	if (type == 0)
-		source_load = load_now;
-	else
-		source_load = min(cpu_load, load_now);
-
-	if (running > 1 || (idle == NOT_IDLE && running))
-		/*
-		 * If we are busy rebalancing the load is biased by
-		 * priority to create 'nice' support across cpus. When
-		 * idle rebalancing we should only bias the source_load if
-		 * there is more than one task running on that queue to
-		 * prevent idle rebalance from trying to pull tasks from a
-		 * queue with only one running task.
-		 */
-		source_load = source_load * rq->prio_bias / running;
-
-	return source_load;
-}
+		return load_now;
 
-static inline unsigned long source_load(int cpu, int type)
-{
-	return __source_load(cpu, type, NOT_IDLE);
+	return min(rq->cpu_load[type-1], load_now);
 }
 
 /*
  * Return a high guess at the load of a migration-target cpu
  */
-static inline unsigned long __target_load(int cpu, int type, enum idle_type idle)
+static inline unsigned long target_load(int cpu, int type)
 {
 	runqueue_t *rq = cpu_rq(cpu);
-	unsigned long running = rq->nr_running;
-	unsigned long target_load, cpu_load = rq->cpu_load[type-1],
-		load_now = running * SCHED_LOAD_SCALE;
-
+	unsigned long load_now = rq->nr_running * SCHED_LOAD_SCALE;
 	if (type == 0)
-		target_load = load_now;
-	else
-		target_load = max(cpu_load, load_now);
-
-	if (running > 1 || (idle == NOT_IDLE && running))
-		target_load = target_load * rq->prio_bias / running;
+		return load_now;
 
-	return target_load;
-}
-
-static inline unsigned long target_load(int cpu, int type)
-{
-	return __target_load(cpu, type, NOT_IDLE);
+	return max(rq->cpu_load[type-1], load_now);
 }
 
 /*
@@ -1294,9 +1197,6 @@ static int try_to_wake_up(task_t *p, unsigned int state, int sync)
 		}
 	}
 
-	if (p->last_waker_cpu != this_cpu)
-		goto out_set_cpu;
-
 	if (unlikely(!cpu_isset(this_cpu, p->cpus_allowed)))
 		goto out_set_cpu;
 
@@ -1367,8 +1267,6 @@ out_set_cpu:
 		cpu = task_cpu(p);
 	}
 
-	p->last_waker_cpu = this_cpu;
-
 out_activate:
 #endif /* CONFIG_SMP */
 	if (old_state == TASK_UNINTERRUPTIBLE) {
@@ -1450,12 +1348,9 @@ void fastcall sched_fork(task_t *p, int clone_flags)
 #ifdef CONFIG_SCHEDSTATS
 	memset(&p->sched_info, 0, sizeof(p->sched_info));
 #endif
-#if defined(CONFIG_SMP)
-	p->last_waker_cpu = cpu;
-#if defined(__ARCH_WANT_UNLOCKED_CTXSW)
+#if defined(CONFIG_SMP) && defined(__ARCH_WANT_UNLOCKED_CTXSW)
 	p->oncpu = 0;
 #endif
-#endif
 #ifdef CONFIG_PREEMPT
 	/* Want to start with kernel preemption disabled. */
 	task_thread_info(p)->preempt_count = 1;
@@ -1530,7 +1425,7 @@ void fastcall wake_up_new_task(task_t *p, unsigned long clone_flags)
 				list_add_tail(&p->run_list, &current->run_list);
 				p->array = current->array;
 				p->array->nr_active++;
-				inc_nr_running(p, rq);
+				rq->nr_running++;
 			}
 			set_need_resched();
 		} else
@@ -1875,9 +1770,9 @@ void pull_task(runqueue_t *src_rq, prio_array_t *src_array, task_t *p,
 	       runqueue_t *this_rq, prio_array_t *this_array, int this_cpu)
 {
 	dequeue_task(p, src_array);
-	dec_nr_running(p, src_rq);
+	src_rq->nr_running--;
 	set_task_cpu(p, this_cpu);
-	inc_nr_running(p, this_rq);
+	this_rq->nr_running++;
 	enqueue_task(p, this_array);
 	p->timestamp = (p->timestamp - src_rq->timestamp_last_tick)
 				+ this_rq->timestamp_last_tick;
@@ -2056,9 +1951,9 @@ find_busiest_group(struct sched_domain *sd, int this_cpu,
 
 			/* Bias balancing toward cpus of our domain */
 			if (local_group)
-				load = __target_load(i, load_idx, idle);
+				load = target_load(i, load_idx);
 			else
-				load = __source_load(i, load_idx, idle);
+				load = source_load(i, load_idx);
 
 			avg_load += load;
 		}
@@ -2171,7 +2066,7 @@ static runqueue_t *find_busiest_queue(struct sched_group *group,
 	int i;
 
 	for_each_cpu_mask(i, group->cpumask) {
-		load = __source_load(i, 0, idle);
+		load = source_load(i, 0);
 
 		if (load > max_load) {
 			max_load = load;
@@ -3571,10 +3466,8 @@ void set_user_nice(task_t *p, long nice)
 		goto out_unlock;
 	}
 	array = p->array;
-	if (array) {
+	if (array)
 		dequeue_task(p, array);
-		dec_prio_bias(rq, p->static_prio);
-	}
 
 	old_prio = p->prio;
 	new_prio = NICE_TO_PRIO(nice);
@@ -3584,7 +3477,6 @@ void set_user_nice(task_t *p, long nice)
 
 	if (array) {
 		enqueue_task(p, array);
-		inc_prio_bias(rq, p->static_prio);
 		/*
 		 * If the task increased its priority or is running and
 		 * lowered its priority, then reschedule its CPU:
@@ -4031,7 +3923,7 @@ long sched_getaffinity(pid_t pid, cpumask_t *mask)
 		goto out_unlock;
 
 	retval = 0;
-	cpus_and(*mask, p->cpus_allowed, cpu_possible_map);
+	cpus_and(*mask, p->cpus_allowed, cpu_online_map);
 
 out_unlock:
 	read_unlock(&tasklist_lock);
@@ -4129,6 +4021,8 @@ static inline void __cond_resched(void)
 	 */
 	if (unlikely(preempt_count()))
 		return;
+	if (unlikely(system_state != SYSTEM_RUNNING))
+		return;
 	do {
 		add_preempt_count(PREEMPT_ACTIVE);
 		schedule();
@@ -4434,6 +4328,7 @@ void __devinit init_idle(task_t *idle, int cpu)
 	runqueue_t *rq = cpu_rq(cpu);
 	unsigned long flags;
 
+	idle->timestamp = sched_clock();
 	idle->sleep_avg = 0;
 	idle->array = NULL;
 	idle->prio = MAX_PRIO;
@@ -5141,7 +5036,7 @@ static void init_sched_build_groups(struct sched_group groups[], cpumask_t span,
 #define SEARCH_SCOPE		2
 #define MIN_CACHE_SIZE		(64*1024U)
 #define DEFAULT_CACHE_SIZE	(5*1024*1024U)
-#define ITERATIONS		2
+#define ITERATIONS		1
 #define SIZE_THRESH		130
 #define COST_THRESH		130
 
@@ -5159,7 +5054,18 @@ static void init_sched_build_groups(struct sched_group groups[], cpumask_t span,
 #define MAX_DOMAIN_DISTANCE 32
 
 static unsigned long long migration_cost[MAX_DOMAIN_DISTANCE] =
-		{ [ 0 ... MAX_DOMAIN_DISTANCE-1 ] = -1LL };
+		{ [ 0 ... MAX_DOMAIN_DISTANCE-1 ] =
+/*
+ * Architectures may override the migration cost and thus avoid
+ * boot-time calibration. Unit is nanoseconds. Mostly useful for
+ * virtualized hardware:
+ */
+#ifdef CONFIG_DEFAULT_MIGRATION_COST
+			CONFIG_DEFAULT_MIGRATION_COST
+#else
+			-1LL
+#endif
+};
 
 /*
  * Allow override of migration cost - in units of microseconds.
@@ -5480,9 +5386,9 @@ static unsigned long long measure_migration_cost(int cpu1, int cpu2)
 				break;
 			}
 		/*
-		 * Increase the cachesize in 5% steps:
+		 * Increase the cachesize in 10% steps:
 		 */
-		size = size * 20 / 19;
+		size = size * 10 / 9;
 	}
 
 	if (migration_debug)
@@ -5551,13 +5457,15 @@ static void calibrate_migration_costs(const cpumask_t *cpu_map)
 			-1
 #endif
 		);
-	printk("migration_cost=");
-	for (distance = 0; distance <= max_distance; distance++) {
-		if (distance)
-			printk(",");
-		printk("%ld", (long)migration_cost[distance] / 1000);
+	if (system_state == SYSTEM_BOOTING) {
+		printk("migration_cost=");
+		for (distance = 0; distance <= max_distance; distance++) {
+			if (distance)
+				printk(",");
+			printk("%ld", (long)migration_cost[distance] / 1000);
+		}
+		printk("\n");
 	}
-	printk("\n");
 	j1 = jiffies;
 	if (migration_debug)
 		printk("migration: %ld seconds\n", (j1-j0)/HZ);
@@ -6109,7 +6017,7 @@ void __init sched_init(void)
 	runqueue_t *rq;
 	int i, j, k;
 
-	for (i = 0; i < NR_CPUS; i++) {
+	for_each_cpu(i) {
 		prio_array_t *array;
 
 		rq = cpu_rq(i);