Merge branch 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull scheduler updates from Ingo Molnar:
 "The main changes in this cycle were:

   - Optimized support for Intel "Cluster-on-Die" (CoD) topologies (Dave
     Hansen)

   - Various sched/idle refinements for better idle handling (Nicolas
     Pitre, Daniel Lezcano, Chuansheng Liu, Vincent Guittot)

   - sched/numa updates and optimizations (Rik van Riel)

   - sysbench speedup (Vincent Guittot)

   - capacity calculation cleanups/refactoring (Vincent Guittot)

   - Various cleanups to thread group iteration (Oleg Nesterov)

   - Double-rq-lock removal optimization and various refactorings
     (Kirill Tkhai)

   - various sched/deadline fixes

  ... and lots of other changes"

* 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (72 commits)
  sched/dl: Use dl_bw_of() under rcu_read_lock_sched()
  sched/fair: Delete resched_cpu() from idle_balance()
  sched, time: Fix build error with 64 bit cputime_t on 32 bit systems
  sched: Improve sysbench performance by fixing spurious active migration
  sched/x86: Fix up typo in topology detection
  x86, sched: Add new topology for multi-NUMA-node CPUs
  sched/rt: Use resched_curr() in task_tick_rt()
  sched: Use rq->rd in sched_setaffinity() under RCU read lock
  sched: cleanup: Rename 'out_unlock' to 'out_free_new_mask'
  sched: Use dl_bw_of() under RCU read lock
  sched/fair: Remove duplicate code from can_migrate_task()
  sched, mips, ia64: Remove __ARCH_WANT_UNLOCKED_CTXSW
  sched: print_rq(): Don't use tasklist_lock
  sched: normalize_rt_tasks(): Don't use _irqsave for tasklist_lock, use task_rq_lock()
  sched: Fix the task-group check in tg_has_rt_tasks()
  sched/fair: Leverage the idle state info when choosing the "idlest" cpu
  sched: Let the scheduler see CPU idle states
  sched/deadline: Fix inter- exclusive cpusets migrations
  sched/deadline: Clear dl_entity params when setscheduling to different class
  sched/numa: Kill the wrong/dead TASK_DEAD check in task_numa_fault()
  ...

1 files changed, 291 insertions, 188 deletions

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 82088b29704e..b78280c59b46 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -23,6 +23,7 @@
 #include <linux/latencytop.h>
 #include <linux/sched.h>
 #include <linux/cpumask.h>
+#include <linux/cpuidle.h>
 #include <linux/slab.h>
 #include <linux/profile.h>
 #include <linux/interrupt.h>
@@ -665,6 +666,7 @@ static u64 sched_vslice(struct cfs_rq *cfs_rq, struct sched_entity *se)
 }
 
 #ifdef CONFIG_SMP
+static int select_idle_sibling(struct task_struct *p, int cpu);
 static unsigned long task_h_load(struct task_struct *p);
 
 static inline void __update_task_entity_contrib(struct sched_entity *se);
@@ -1038,7 +1040,8 @@ struct numa_stats {
  */
 static void update_numa_stats(struct numa_stats *ns, int nid)
 {
-	int cpu, cpus = 0;
+	int smt, cpu, cpus = 0;
+	unsigned long capacity;
 
 	memset(ns, 0, sizeof(*ns));
 	for_each_cpu(cpu, cpumask_of_node(nid)) {
@@ -1062,8 +1065,12 @@ static void update_numa_stats(struct numa_stats *ns, int nid)
 	if (!cpus)
 		return;
 
-	ns->task_capacity =
-		DIV_ROUND_CLOSEST(ns->compute_capacity, SCHED_CAPACITY_SCALE);
+	/* smt := ceil(cpus / capacity), assumes: 1 < smt_power < 2 */
+	smt = DIV_ROUND_UP(SCHED_CAPACITY_SCALE * cpus, ns->compute_capacity);
+	capacity = cpus / smt; /* cores */
+
+	ns->task_capacity = min_t(unsigned, capacity,
+		DIV_ROUND_CLOSEST(ns->compute_capacity, SCHED_CAPACITY_SCALE));
 	ns->has_free_capacity = (ns->nr_running < ns->task_capacity);
 }
 
@@ -1206,7 +1213,7 @@ static void task_numa_compare(struct task_numa_env *env,
 
 	if (!cur) {
 		/* Is there capacity at our destination? */
-		if (env->src_stats.has_free_capacity &&
+		if (env->src_stats.nr_running <= env->src_stats.task_capacity &&
 		    !env->dst_stats.has_free_capacity)
 			goto unlock;
 
@@ -1252,6 +1259,13 @@ balance:
 	if (load_too_imbalanced(src_load, dst_load, env))
 		goto unlock;
 
+	/*
+	 * One idle CPU per node is evaluated for a task numa move.
+	 * Call select_idle_sibling to maybe find a better one.
+	 */
+	if (!cur)
+		env->dst_cpu = select_idle_sibling(env->p, env->dst_cpu);
+
 assign:
 	task_numa_assign(env, cur, imp);
 unlock:
@@ -1775,7 +1789,7 @@ void task_numa_free(struct task_struct *p)
 		list_del(&p->numa_entry);
 		grp->nr_tasks--;
 		spin_unlock_irqrestore(&grp->lock, flags);
-		rcu_assign_pointer(p->numa_group, NULL);
+		RCU_INIT_POINTER(p->numa_group, NULL);
 		put_numa_group(grp);
 	}
 
@@ -1804,10 +1818,6 @@ void task_numa_fault(int last_cpupid, int mem_node, int pages, int flags)
 	if (!p->mm)
 		return;
 
-	/* Do not worry about placement if exiting */
-	if (p->state == TASK_DEAD)
-		return;
-
 	/* Allocate buffer to track faults on a per-node basis */
 	if (unlikely(!p->numa_faults_memory)) {
 		int size = sizeof(*p->numa_faults_memory) *
@@ -2211,8 +2221,8 @@ static __always_inline u64 decay_load(u64 val, u64 n)
 
 	/*
 	 * As y^PERIOD = 1/2, we can combine
-	 *    y^n = 1/2^(n/PERIOD) * k^(n%PERIOD)
-	 * With a look-up table which covers k^n (n<PERIOD)
+	 *    y^n = 1/2^(n/PERIOD) * y^(n%PERIOD)
+	 * With a look-up table which covers y^n (n<PERIOD)
 	 *
 	 * To achieve constant time decay_load.
 	 */
@@ -2377,6 +2387,9 @@ static inline void __update_cfs_rq_tg_load_contrib(struct cfs_rq *cfs_rq,
 	tg_contrib = cfs_rq->runnable_load_avg + cfs_rq->blocked_load_avg;
 	tg_contrib -= cfs_rq->tg_load_contrib;
 
+	if (!tg_contrib)
+		return;
+
 	if (force_update || abs(tg_contrib) > cfs_rq->tg_load_contrib / 8) {
 		atomic_long_add(tg_contrib, &tg->load_avg);
 		cfs_rq->tg_load_contrib += tg_contrib;
@@ -3892,14 +3905,6 @@ static void hrtick_start_fair(struct rq *rq, struct task_struct *p)
 				resched_curr(rq);
 			return;
 		}
-
-		/*
-		 * Don't schedule slices shorter than 10000ns, that just
-		 * doesn't make sense. Rely on vruntime for fairness.
-		 */
-		if (rq->curr != p)
-			delta = max_t(s64, 10000LL, delta);
-
 		hrtick_start(rq, delta);
 	}
 }
@@ -4087,7 +4092,7 @@ static unsigned long capacity_of(int cpu)
 static unsigned long cpu_avg_load_per_task(int cpu)
 {
 	struct rq *rq = cpu_rq(cpu);
-	unsigned long nr_running = ACCESS_ONCE(rq->nr_running);
+	unsigned long nr_running = ACCESS_ONCE(rq->cfs.h_nr_running);
 	unsigned long load_avg = rq->cfs.runnable_load_avg;
 
 	if (nr_running)
@@ -4276,8 +4281,8 @@ static int wake_wide(struct task_struct *p)
 static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync)
 {
 	s64 this_load, load;
+	s64 this_eff_load, prev_eff_load;
 	int idx, this_cpu, prev_cpu;
-	unsigned long tl_per_task;
 	struct task_group *tg;
 	unsigned long weight;
 	int balanced;
@@ -4320,47 +4325,30 @@ static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync)
 	 * Otherwise check if either cpus are near enough in load to allow this
 	 * task to be woken on this_cpu.
 	 */
-	if (this_load > 0) {
-		s64 this_eff_load, prev_eff_load;
+	this_eff_load = 100;
+	this_eff_load *= capacity_of(prev_cpu);
 
-		this_eff_load = 100;
-		this_eff_load *= capacity_of(prev_cpu);
+	prev_eff_load = 100 + (sd->imbalance_pct - 100) / 2;
+	prev_eff_load *= capacity_of(this_cpu);
+
+	if (this_load > 0) {
 		this_eff_load *= this_load +
 			effective_load(tg, this_cpu, weight, weight);
 
-		prev_eff_load = 100 + (sd->imbalance_pct - 100) / 2;
-		prev_eff_load *= capacity_of(this_cpu);
 		prev_eff_load *= load + effective_load(tg, prev_cpu, 0, weight);
+	}
 
-		balanced = this_eff_load <= prev_eff_load;
-	} else
-		balanced = true;
-
-	/*
-	 * If the currently running task will sleep within
-	 * a reasonable amount of time then attract this newly
-	 * woken task:
-	 */
-	if (sync && balanced)
-		return 1;
+	balanced = this_eff_load <= prev_eff_load;
 
 	schedstat_inc(p, se.statistics.nr_wakeups_affine_attempts);
-	tl_per_task = cpu_avg_load_per_task(this_cpu);
 
-	if (balanced ||
-	    (this_load <= load &&
-	     this_load + target_load(prev_cpu, idx) <= tl_per_task)) {
-		/*
-		 * This domain has SD_WAKE_AFFINE and
-		 * p is cache cold in this domain, and
-		 * there is no bad imbalance.
-		 */
-		schedstat_inc(sd, ttwu_move_affine);
-		schedstat_inc(p, se.statistics.nr_wakeups_affine);
+	if (!balanced)
+		return 0;
 
-		return 1;
-	}
-	return 0;
+	schedstat_inc(sd, ttwu_move_affine);
+	schedstat_inc(p, se.statistics.nr_wakeups_affine);
+
+	return 1;
 }
 
 /*
@@ -4428,20 +4416,46 @@ static int
 find_idlest_cpu(struct sched_group *group, struct task_struct *p, int this_cpu)
 {
 	unsigned long load, min_load = ULONG_MAX;
-	int idlest = -1;
+	unsigned int min_exit_latency = UINT_MAX;
+	u64 latest_idle_timestamp = 0;
+	int least_loaded_cpu = this_cpu;
+	int shallowest_idle_cpu = -1;
 	int i;
 
 	/* Traverse only the allowed CPUs */
 	for_each_cpu_and(i, sched_group_cpus(group), tsk_cpus_allowed(p)) {
-		load = weighted_cpuload(i);
-
-		if (load < min_load || (load == min_load && i == this_cpu)) {
-			min_load = load;
-			idlest = i;
+		if (idle_cpu(i)) {
+			struct rq *rq = cpu_rq(i);
+			struct cpuidle_state *idle = idle_get_state(rq);
+			if (idle && idle->exit_latency < min_exit_latency) {
+				/*
+				 * We give priority to a CPU whose idle state
+				 * has the smallest exit latency irrespective
+				 * of any idle timestamp.
+				 */
+				min_exit_latency = idle->exit_latency;
+				latest_idle_timestamp = rq->idle_stamp;
+				shallowest_idle_cpu = i;
+			} else if ((!idle || idle->exit_latency == min_exit_latency) &&
+				   rq->idle_stamp > latest_idle_timestamp) {
+				/*
+				 * If equal or no active idle state, then
+				 * the most recently idled CPU might have
+				 * a warmer cache.
+				 */
+				latest_idle_timestamp = rq->idle_stamp;
+				shallowest_idle_cpu = i;
+			}
+		} else {
+			load = weighted_cpuload(i);
+			if (load < min_load || (load == min_load && i == this_cpu)) {
+				min_load = load;
+				least_loaded_cpu = i;
+			}
 		}
 	}
 
-	return idlest;
+	return shallowest_idle_cpu != -1 ? shallowest_idle_cpu : least_loaded_cpu;
 }
 
 /*
@@ -4513,11 +4527,8 @@ select_task_rq_fair(struct task_struct *p, int prev_cpu, int sd_flag, int wake_f
 	if (p->nr_cpus_allowed == 1)
 		return prev_cpu;
 
-	if (sd_flag & SD_BALANCE_WAKE) {
-		if (cpumask_test_cpu(cpu, tsk_cpus_allowed(p)))
-			want_affine = 1;
-		new_cpu = prev_cpu;
-	}
+	if (sd_flag & SD_BALANCE_WAKE)
+		want_affine = cpumask_test_cpu(cpu, tsk_cpus_allowed(p));
 
 	rcu_read_lock();
 	for_each_domain(cpu, tmp) {
@@ -4704,7 +4715,7 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_
 		return;
 
 	/*
-	 * This is possible from callers such as move_task(), in which we
+	 * This is possible from callers such as attach_tasks(), in which we
 	 * unconditionally check_prempt_curr() after an enqueue (which may have
 	 * lead to a throttle).  This both saves work and prevents false
 	 * next-buddy nomination below.
@@ -5112,27 +5123,18 @@ struct lb_env {
 	unsigned int		loop_max;
 
 	enum fbq_type		fbq_type;
+	struct list_head	tasks;
 };
 
 /*
- * move_task - move a task from one runqueue to another runqueue.
- * Both runqueues must be locked.
- */
-static void move_task(struct task_struct *p, struct lb_env *env)
-{
-	deactivate_task(env->src_rq, p, 0);
-	set_task_cpu(p, env->dst_cpu);
-	activate_task(env->dst_rq, p, 0);
-	check_preempt_curr(env->dst_rq, p, 0);
-}
-
-/*
  * Is this task likely cache-hot:
  */
 static int task_hot(struct task_struct *p, struct lb_env *env)
 {
 	s64 delta;
 
+	lockdep_assert_held(&env->src_rq->lock);
+
 	if (p->sched_class != &fair_sched_class)
 		return 0;
 
@@ -5252,6 +5254,9 @@ static
 int can_migrate_task(struct task_struct *p, struct lb_env *env)
 {
 	int tsk_cache_hot = 0;
+
+	lockdep_assert_held(&env->src_rq->lock);
+
 	/*
 	 * We do not migrate tasks that are:
 	 * 1) throttled_lb_pair, or
@@ -5310,24 +5315,12 @@ int can_migrate_task(struct task_struct *p, struct lb_env *env)
 	if (!tsk_cache_hot)
 		tsk_cache_hot = migrate_degrades_locality(p, env);
 
-	if (migrate_improves_locality(p, env)) {
-#ifdef CONFIG_SCHEDSTATS
+	if (migrate_improves_locality(p, env) || !tsk_cache_hot ||
+	    env->sd->nr_balance_failed > env->sd->cache_nice_tries) {
 		if (tsk_cache_hot) {
 			schedstat_inc(env->sd, lb_hot_gained[env->idle]);
 			schedstat_inc(p, se.statistics.nr_forced_migrations);
 		}
-#endif
-		return 1;
-	}
-
-	if (!tsk_cache_hot ||
-		env->sd->nr_balance_failed > env->sd->cache_nice_tries) {
-
-		if (tsk_cache_hot) {
-			schedstat_inc(env->sd, lb_hot_gained[env->idle]);
-			schedstat_inc(p, se.statistics.nr_forced_migrations);
-		}
-
 		return 1;
 	}
 
@@ -5336,47 +5329,63 @@ int can_migrate_task(struct task_struct *p, struct lb_env *env)
 }
 
 /*
- * move_one_task tries to move exactly one task from busiest to this_rq, as
+ * detach_task() -- detach the task for the migration specified in env
+ */
+static void detach_task(struct task_struct *p, struct lb_env *env)
+{
+	lockdep_assert_held(&env->src_rq->lock);
+
+	deactivate_task(env->src_rq, p, 0);
+	p->on_rq = TASK_ON_RQ_MIGRATING;
+	set_task_cpu(p, env->dst_cpu);
+}
+
+/*
+ * detach_one_task() -- tries to dequeue exactly one task from env->src_rq, as
  * part of active balancing operations within "domain".
- * Returns 1 if successful and 0 otherwise.
  *
- * Called with both runqueues locked.
+ * Returns a task if successful and NULL otherwise.
  */
-static int move_one_task(struct lb_env *env)
+static struct task_struct *detach_one_task(struct lb_env *env)
 {
 	struct task_struct *p, *n;
 
+	lockdep_assert_held(&env->src_rq->lock);
+
 	list_for_each_entry_safe(p, n, &env->src_rq->cfs_tasks, se.group_node) {
 		if (!can_migrate_task(p, env))
 			continue;
 
-		move_task(p, env);
+		detach_task(p, env);
+
 		/*
-		 * Right now, this is only the second place move_task()
-		 * is called, so we can safely collect move_task()
-		 * stats here rather than inside move_task().
+		 * Right now, this is only the second place where
+		 * lb_gained[env->idle] is updated (other is detach_tasks)
+		 * so we can safely collect stats here rather than
+		 * inside detach_tasks().
 		 */
 		schedstat_inc(env->sd, lb_gained[env->idle]);
-		return 1;
+		return p;
 	}
-	return 0;
+	return NULL;
 }
 
 static const unsigned int sched_nr_migrate_break = 32;
 
 /*
- * move_tasks tries to move up to imbalance weighted load from busiest to
- * this_rq, as part of a balancing operation within domain "sd".
- * Returns 1 if successful and 0 otherwise.
+ * detach_tasks() -- tries to detach up to imbalance weighted load from
+ * busiest_rq, as part of a balancing operation within domain "sd".
  *
- * Called with both runqueues locked.
+ * Returns number of detached tasks if successful and 0 otherwise.
  */
-static int move_tasks(struct lb_env *env)
+static int detach_tasks(struct lb_env *env)
 {
 	struct list_head *tasks = &env->src_rq->cfs_tasks;
 	struct task_struct *p;
 	unsigned long load;
-	int pulled = 0;
+	int detached = 0;
+
+	lockdep_assert_held(&env->src_rq->lock);
 
 	if (env->imbalance <= 0)
 		return 0;
@@ -5407,14 +5416,16 @@ static int move_tasks(struct lb_env *env)
 		if ((load / 2) > env->imbalance)
 			goto next;
 
-		move_task(p, env);
-		pulled++;
+		detach_task(p, env);
+		list_add(&p->se.group_node, &env->tasks);
+
+		detached++;
 		env->imbalance -= load;
 
 #ifdef CONFIG_PREEMPT
 		/*
 		 * NEWIDLE balancing is a source of latency, so preemptible
-		 * kernels will stop after the first task is pulled to minimize
+		 * kernels will stop after the first task is detached to minimize
 		 * the critical section.
 		 */
 		if (env->idle == CPU_NEWLY_IDLE)
@@ -5434,13 +5445,58 @@ next:
 	}
 
 	/*
-	 * Right now, this is one of only two places move_task() is called,
-	 * so we can safely collect move_task() stats here rather than
-	 * inside move_task().
+	 * Right now, this is one of only two places we collect this stat
+	 * so we can safely collect detach_one_task() stats here rather
+	 * than inside detach_one_task().
 	 */
-	schedstat_add(env->sd, lb_gained[env->idle], pulled);
+	schedstat_add(env->sd, lb_gained[env->idle], detached);
+
+	return detached;
+}
+
+/*
+ * attach_task() -- attach the task detached by detach_task() to its new rq.
+ */
+static void attach_task(struct rq *rq, struct task_struct *p)
+{
+	lockdep_assert_held(&rq->lock);
+
+	BUG_ON(task_rq(p) != rq);
+	p->on_rq = TASK_ON_RQ_QUEUED;
+	activate_task(rq, p, 0);
+	check_preempt_curr(rq, p, 0);
+}
+
+/*
+ * attach_one_task() -- attaches the task returned from detach_one_task() to
+ * its new rq.
+ */
+static void attach_one_task(struct rq *rq, struct task_struct *p)
+{
+	raw_spin_lock(&rq->lock);
+	attach_task(rq, p);
+	raw_spin_unlock(&rq->lock);
+}
+
+/*
+ * attach_tasks() -- attaches all tasks detached by detach_tasks() to their
+ * new rq.
+ */
+static void attach_tasks(struct lb_env *env)
+{
+	struct list_head *tasks = &env->tasks;
+	struct task_struct *p;
+
+	raw_spin_lock(&env->dst_rq->lock);
+
+	while (!list_empty(tasks)) {
+		p = list_first_entry(tasks, struct task_struct, se.group_node);
+		list_del_init(&p->se.group_node);
 
-	return pulled;
+		attach_task(env->dst_rq, p);
+	}
+
+	raw_spin_unlock(&env->dst_rq->lock);
 }
 
 #ifdef CONFIG_FAIR_GROUP_SCHED
@@ -5559,6 +5615,13 @@ static unsigned long task_h_load(struct task_struct *p)
 #endif
 
 /********** Helpers for find_busiest_group ************************/
+
+enum group_type {
+	group_other = 0,
+	group_imbalanced,
+	group_overloaded,
+};
+
 /*
  * sg_lb_stats - stats of a sched_group required for load_balancing
  */
@@ -5572,7 +5635,7 @@ struct sg_lb_stats {
 	unsigned int group_capacity_factor;
 	unsigned int idle_cpus;
 	unsigned int group_weight;
-	int group_imb; /* Is there an imbalance in the group ? */
+	enum group_type group_type;
 	int group_has_free_capacity;
 #ifdef CONFIG_NUMA_BALANCING
 	unsigned int nr_numa_running;
@@ -5610,6 +5673,8 @@ static inline void init_sd_lb_stats(struct sd_lb_stats *sds)
 		.total_capacity = 0UL,
 		.busiest_stat = {
 			.avg_load = 0UL,
+			.sum_nr_running = 0,
+			.group_type = group_other,
 		},
 	};
 }
@@ -5652,19 +5717,17 @@ unsigned long __weak arch_scale_freq_capacity(struct sched_domain *sd, int cpu)
 	return default_scale_capacity(sd, cpu);
 }
 
-static unsigned long default_scale_smt_capacity(struct sched_domain *sd, int cpu)
+static unsigned long default_scale_cpu_capacity(struct sched_domain *sd, int cpu)
 {
-	unsigned long weight = sd->span_weight;
-	unsigned long smt_gain = sd->smt_gain;
+	if ((sd->flags & SD_SHARE_CPUCAPACITY) && (sd->span_weight > 1))
+		return sd->smt_gain / sd->span_weight;
 
-	smt_gain /= weight;
-
-	return smt_gain;
+	return SCHED_CAPACITY_SCALE;
 }
 
-unsigned long __weak arch_scale_smt_capacity(struct sched_domain *sd, int cpu)
+unsigned long __weak arch_scale_cpu_capacity(struct sched_domain *sd, int cpu)
 {
-	return default_scale_smt_capacity(sd, cpu);
+	return default_scale_cpu_capacity(sd, cpu);
 }
 
 static unsigned long scale_rt_capacity(int cpu)
@@ -5703,18 +5766,15 @@ static unsigned long scale_rt_capacity(int cpu)
 
 static void update_cpu_capacity(struct sched_domain *sd, int cpu)
 {
-	unsigned long weight = sd->span_weight;
 	unsigned long capacity = SCHED_CAPACITY_SCALE;
 	struct sched_group *sdg = sd->groups;
 
-	if ((sd->flags & SD_SHARE_CPUCAPACITY) && weight > 1) {
-		if (sched_feat(ARCH_CAPACITY))
-			capacity *= arch_scale_smt_capacity(sd, cpu);
-		else
-			capacity *= default_scale_smt_capacity(sd, cpu);
+	if (sched_feat(ARCH_CAPACITY))
+		capacity *= arch_scale_cpu_capacity(sd, cpu);
+	else
+		capacity *= default_scale_cpu_capacity(sd, cpu);
 
-		capacity >>= SCHED_CAPACITY_SHIFT;
-	}
+	capacity >>= SCHED_CAPACITY_SHIFT;
 
 	sdg->sgc->capacity_orig = capacity;
 
@@ -5891,6 +5951,18 @@ static inline int sg_capacity_factor(struct lb_env *env, struct sched_group *gro
 	return capacity_factor;
 }
 
+static enum group_type
+group_classify(struct sched_group *group, struct sg_lb_stats *sgs)
+{
+	if (sgs->sum_nr_running > sgs->group_capacity_factor)
+		return group_overloaded;
+
+	if (sg_imbalanced(group))
+		return group_imbalanced;
+
+	return group_other;
+}
+
 /**
  * update_sg_lb_stats - Update sched_group's statistics for load balancing.
  * @env: The load balancing environment.
@@ -5920,7 +5992,7 @@ static inline void update_sg_lb_stats(struct lb_env *env,
 			load = source_load(i, load_idx);
 
 		sgs->group_load += load;
-		sgs->sum_nr_running += rq->nr_running;
+		sgs->sum_nr_running += rq->cfs.h_nr_running;
 
 		if (rq->nr_running > 1)
 			*overload = true;
@@ -5942,9 +6014,8 @@ static inline void update_sg_lb_stats(struct lb_env *env,
 		sgs->load_per_task = sgs->sum_weighted_load / sgs->sum_nr_running;
 
 	sgs->group_weight = group->group_weight;
-
-	sgs->group_imb = sg_imbalanced(group);
 	sgs->group_capacity_factor = sg_capacity_factor(env, group);
+	sgs->group_type = group_classify(group, sgs);
 
 	if (sgs->group_capacity_factor > sgs->sum_nr_running)
 		sgs->group_has_free_capacity = 1;
@@ -5968,13 +6039,19 @@ static bool update_sd_pick_busiest(struct lb_env *env,
 				   struct sched_group *sg,
 				   struct sg_lb_stats *sgs)
 {
-	if (sgs->avg_load <= sds->busiest_stat.avg_load)
-		return false;
+	struct sg_lb_stats *busiest = &sds->busiest_stat;
 
-	if (sgs->sum_nr_running > sgs->group_capacity_factor)
+	if (sgs->group_type > busiest->group_type)
 		return true;
 
-	if (sgs->group_imb)
+	if (sgs->group_type < busiest->group_type)
+		return false;
+
+	if (sgs->avg_load <= busiest->avg_load)
+		return false;
+
+	/* This is the busiest node in its class. */
+	if (!(env->sd->flags & SD_ASYM_PACKING))
 		return true;
 
 	/*
@@ -5982,8 +6059,7 @@ static bool update_sd_pick_busiest(struct lb_env *env,
 	 * numbered CPUs in the group, therefore mark all groups
 	 * higher than ourself as busy.
 	 */
-	if ((env->sd->flags & SD_ASYM_PACKING) && sgs->sum_nr_running &&
-	    env->dst_cpu < group_first_cpu(sg)) {
+	if (sgs->sum_nr_running && env->dst_cpu < group_first_cpu(sg)) {
 		if (!sds->busiest)
 			return true;
 
@@ -6228,7 +6304,7 @@ static inline void calculate_imbalance(struct lb_env *env, struct sd_lb_stats *s
 	local = &sds->local_stat;
 	busiest = &sds->busiest_stat;
 
-	if (busiest->group_imb) {
+	if (busiest->group_type == group_imbalanced) {
 		/*
 		 * In the group_imb case we cannot rely on group-wide averages
 		 * to ensure cpu-load equilibrium, look at wider averages. XXX
@@ -6248,12 +6324,11 @@ static inline void calculate_imbalance(struct lb_env *env, struct sd_lb_stats *s
 		return fix_small_imbalance(env, sds);
 	}
 
-	if (!busiest->group_imb) {
-		/*
-		 * Don't want to pull so many tasks that a group would go idle.
-		 * Except of course for the group_imb case, since then we might
-		 * have to drop below capacity to reach cpu-load equilibrium.
-		 */
+	/*
+	 * If there aren't any idle cpus, avoid creating some.
+	 */
+	if (busiest->group_type == group_overloaded &&
+	    local->group_type   == group_overloaded) {
 		load_above_capacity =
 			(busiest->sum_nr_running - busiest->group_capacity_factor);
 
@@ -6337,7 +6412,7 @@ static struct sched_group *find_busiest_group(struct lb_env *env)
 	 * work because they assume all things are equal, which typically
 	 * isn't true due to cpus_allowed constraints and the like.
 	 */
-	if (busiest->group_imb)
+	if (busiest->group_type == group_imbalanced)
 		goto force_balance;
 
 	/* SD_BALANCE_NEWIDLE trumps SMP nice when underutilized */
@@ -6346,7 +6421,7 @@ static struct sched_group *find_busiest_group(struct lb_env *env)
 		goto force_balance;
 
 	/*
-	 * If the local group is more busy than the selected busiest group
+	 * If the local group is busier than the selected busiest group
 	 * don't try and pull any tasks.
 	 */
 	if (local->avg_load >= busiest->avg_load)
@@ -6361,13 +6436,14 @@ static struct sched_group *find_busiest_group(struct lb_env *env)
 
 	if (env->idle == CPU_IDLE) {
 		/*
-		 * This cpu is idle. If the busiest group load doesn't
-		 * have more tasks than the number of available cpu's and
-		 * there is no imbalance between this and busiest group
-		 * wrt to idle cpu's, it is balanced.
+		 * This cpu is idle. If the busiest group is not overloaded
+		 * and there is no imbalance between this and busiest group
+		 * wrt idle cpus, it is balanced. The imbalance becomes
+		 * significant if the diff is greater than 1 otherwise we
+		 * might end up to just move the imbalance on another group
 		 */
-		if ((local->idle_cpus < busiest->idle_cpus) &&
-		    busiest->sum_nr_running <= busiest->group_weight)
+		if ((busiest->group_type != group_overloaded) &&
+				(local->idle_cpus <= (busiest->idle_cpus + 1)))
 			goto out_balanced;
 	} else {
 		/*
@@ -6550,6 +6626,7 @@ static int load_balance(int this_cpu, struct rq *this_rq,
 		.loop_break	= sched_nr_migrate_break,
 		.cpus		= cpus,
 		.fbq_type	= all,
+		.tasks		= LIST_HEAD_INIT(env.tasks),
 	};
 
 	/*
@@ -6599,23 +6676,30 @@ redo:
 		env.loop_max  = min(sysctl_sched_nr_migrate, busiest->nr_running);
 
 more_balance:
-		local_irq_save(flags);
-		double_rq_lock(env.dst_rq, busiest);
+		raw_spin_lock_irqsave(&busiest->lock, flags);
 
 		/*
 		 * cur_ld_moved - load moved in current iteration
 		 * ld_moved     - cumulative load moved across iterations
 		 */
-		cur_ld_moved = move_tasks(&env);
-		ld_moved += cur_ld_moved;
-		double_rq_unlock(env.dst_rq, busiest);
-		local_irq_restore(flags);
+		cur_ld_moved = detach_tasks(&env);
 
 		/*
-		 * some other cpu did the load balance for us.
+		 * We've detached some tasks from busiest_rq. Every
+		 * task is masked "TASK_ON_RQ_MIGRATING", so we can safely
+		 * unlock busiest->lock, and we are able to be sure
+		 * that nobody can manipulate the tasks in parallel.
+		 * See task_rq_lock() family for the details.
 		 */
-		if (cur_ld_moved && env.dst_cpu != smp_processor_id())
-			resched_cpu(env.dst_cpu);
+
+		raw_spin_unlock(&busiest->lock);
+
+		if (cur_ld_moved) {
+			attach_tasks(&env);
+			ld_moved += cur_ld_moved;
+		}
+
+		local_irq_restore(flags);
 
 		if (env.flags & LBF_NEED_BREAK) {
 			env.flags &= ~LBF_NEED_BREAK;
@@ -6665,10 +6749,8 @@ more_balance:
 		if (sd_parent) {
 			int *group_imbalance = &sd_parent->groups->sgc->imbalance;
 
-			if ((env.flags & LBF_SOME_PINNED) && env.imbalance > 0) {
+			if ((env.flags & LBF_SOME_PINNED) && env.imbalance > 0)
 				*group_imbalance = 1;
-			} else if (*group_imbalance)
-				*group_imbalance = 0;
 		}
 
 		/* All tasks on this runqueue were pinned by CPU affinity */
@@ -6679,7 +6761,7 @@ more_balance:
 				env.loop_break = sched_nr_migrate_break;
 				goto redo;
 			}
-			goto out_balanced;
+			goto out_all_pinned;
 		}
 	}
 
@@ -6744,7 +6826,7 @@ more_balance:
 		 * If we've begun active balancing, start to back off. This
 		 * case may not be covered by the all_pinned logic if there
 		 * is only 1 task on the busy runqueue (because we don't call
-		 * move_tasks).
+		 * detach_tasks).
 		 */
 		if (sd->balance_interval < sd->max_interval)
 			sd->balance_interval *= 2;
@@ -6753,6 +6835,23 @@ more_balance:
 	goto out;
 
 out_balanced:
+	/*
+	 * We reach balance although we may have faced some affinity
+	 * constraints. Clear the imbalance flag if it was set.
+	 */
+	if (sd_parent) {
+		int *group_imbalance = &sd_parent->groups->sgc->imbalance;
+
+		if (*group_imbalance)
+			*group_imbalance = 0;
+	}
+
+out_all_pinned:
+	/*
+	 * We reach balance because all tasks are pinned at this level so
+	 * we can't migrate them. Let the imbalance flag set so parent level
+	 * can try to migrate them.
+	 */
 	schedstat_inc(sd, lb_balanced[idle]);
 
 	sd->nr_balance_failed = 0;
@@ -6914,6 +7013,7 @@ static int active_load_balance_cpu_stop(void *data)
 	int target_cpu = busiest_rq->push_cpu;
 	struct rq *target_rq = cpu_rq(target_cpu);
 	struct sched_domain *sd;
+	struct task_struct *p = NULL;
 
 	raw_spin_lock_irq(&busiest_rq->lock);
 
@@ -6933,9 +7033,6 @@ static int active_load_balance_cpu_stop(void *data)
 	 */
 	BUG_ON(busiest_rq == target_rq);
 
-	/* move a task from busiest_rq to target_rq */
-	double_lock_balance(busiest_rq, target_rq);
-
 	/* Search for an sd spanning us and the target CPU. */
 	rcu_read_lock();
 	for_each_domain(target_cpu, sd) {
@@ -6956,16 +7053,22 @@ static int active_load_balance_cpu_stop(void *data)
 
 		schedstat_inc(sd, alb_count);
 
-		if (move_one_task(&env))
+		p = detach_one_task(&env);
+		if (p)
 			schedstat_inc(sd, alb_pushed);
 		else
 			schedstat_inc(sd, alb_failed);
 	}
 	rcu_read_unlock();
-	double_unlock_balance(busiest_rq, target_rq);
 out_unlock:
 	busiest_rq->active_balance = 0;
-	raw_spin_unlock_irq(&busiest_rq->lock);
+	raw_spin_unlock(&busiest_rq->lock);
+
+	if (p)
+		attach_one_task(target_rq, p);
+
+	local_irq_enable();
+
 	return 0;
 }
 
@@ -7465,7 +7568,7 @@ static void task_fork_fair(struct task_struct *p)
 static void
 prio_changed_fair(struct rq *rq, struct task_struct *p, int oldprio)
 {
-	if (!p->se.on_rq)
+	if (!task_on_rq_queued(p))
 		return;
 
 	/*
@@ -7490,11 +7593,11 @@ static void switched_from_fair(struct rq *rq, struct task_struct *p)
 	 * switched back to the fair class the enqueue_entity(.flags=0) will
 	 * do the right thing.
 	 *
-	 * If it's on_rq, then the dequeue_entity(.flags=0) will already
-	 * have normalized the vruntime, if it's !on_rq, then only when
+	 * If it's queued, then the dequeue_entity(.flags=0) will already
+	 * have normalized the vruntime, if it's !queued, then only when
 	 * the task is sleeping will it still have non-normalized vruntime.
 	 */
-	if (!p->on_rq && p->state != TASK_RUNNING) {
+	if (!task_on_rq_queued(p) && p->state != TASK_RUNNING) {
 		/*
 		 * Fix up our vruntime so that the current sleep doesn't
 		 * cause 'unlimited' sleep bonus.
@@ -7521,15 +7624,15 @@ static void switched_from_fair(struct rq *rq, struct task_struct *p)
  */
 static void switched_to_fair(struct rq *rq, struct task_struct *p)
 {
-	struct sched_entity *se = &p->se;
 #ifdef CONFIG_FAIR_GROUP_SCHED
+	struct sched_entity *se = &p->se;
 	/*
 	 * Since the real-depth could have been changed (only FAIR
 	 * class maintain depth value), reset depth properly.
 	 */
 	se->depth = se->parent ? se->parent->depth + 1 : 0;
 #endif
-	if (!se->on_rq)
+	if (!task_on_rq_queued(p))
 		return;
 
 	/*
@@ -7575,7 +7678,7 @@ void init_cfs_rq(struct cfs_rq *cfs_rq)
 }
 
 #ifdef CONFIG_FAIR_GROUP_SCHED
-static void task_move_group_fair(struct task_struct *p, int on_rq)
+static void task_move_group_fair(struct task_struct *p, int queued)
 {
 	struct sched_entity *se = &p->se;
 	struct cfs_rq *cfs_rq;
@@ -7594,7 +7697,7 @@ static void task_move_group_fair(struct task_struct *p, int on_rq)
 	 * fair sleeper stuff for the first placement, but who cares.
 	 */
 	/*
-	 * When !on_rq, vruntime of the task has usually NOT been normalized.
+	 * When !queued, vruntime of the task has usually NOT been normalized.
 	 * But there are some cases where it has already been normalized:
 	 *
 	 * - Moving a forked child which is waiting for being woken up by
@@ -7605,14 +7708,14 @@ static void task_move_group_fair(struct task_struct *p, int on_rq)
 	 * To prevent boost or penalty in the new cfs_rq caused by delta
 	 * min_vruntime between the two cfs_rqs, we skip vruntime adjustment.
 	 */
-	if (!on_rq && (!se->sum_exec_runtime || p->state == TASK_WAKING))
-		on_rq = 1;
+	if (!queued && (!se->sum_exec_runtime || p->state == TASK_WAKING))
+		queued = 1;
 
-	if (!on_rq)
+	if (!queued)
 		se->vruntime -= cfs_rq_of(se)->min_vruntime;
 	set_task_rq(p, task_cpu(p));
 	se->depth = se->parent ? se->parent->depth + 1 : 0;
-	if (!on_rq) {
+	if (!queued) {
 		cfs_rq = cfs_rq_of(se);
 		se->vruntime += cfs_rq->min_vruntime;
 #ifdef CONFIG_SMP