Merge schedutil governor updates for v4.10.

4 files changed, 134 insertions, 46 deletions

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 94732d1ab00a..154fd689fe02 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -5192,21 +5192,14 @@ void sched_show_task(struct task_struct *p)
 	int ppid;
 	unsigned long state = p->state;
 
+	if (!try_get_task_stack(p))
+		return;
 	if (state)
 		state = __ffs(state) + 1;
 	printk(KERN_INFO "%-15.15s %c", p->comm,
 		state < sizeof(stat_nam) - 1 ? stat_nam[state] : '?');
-#if BITS_PER_LONG == 32
-	if (state == TASK_RUNNING)
-		printk(KERN_CONT " running  ");
-	else
-		printk(KERN_CONT " %08lx ", thread_saved_pc(p));
-#else
 	if (state == TASK_RUNNING)
 		printk(KERN_CONT "  running task    ");
-	else
-		printk(KERN_CONT " %016lx ", thread_saved_pc(p));
-#endif
 #ifdef CONFIG_DEBUG_STACK_USAGE
 	free = stack_not_used(p);
 #endif
@@ -5221,6 +5214,7 @@ void sched_show_task(struct task_struct *p)
 
 	print_worker_info(KERN_INFO, p);
 	show_stack(p, NULL);
+	put_task_stack(p);
 }
 
 void show_state_filter(unsigned long state_filter)
@@ -7515,11 +7509,27 @@ static struct kmem_cache *task_group_cache __read_mostly;
 DECLARE_PER_CPU(cpumask_var_t, load_balance_mask);
 DECLARE_PER_CPU(cpumask_var_t, select_idle_mask);
 
+#define WAIT_TABLE_BITS 8
+#define WAIT_TABLE_SIZE (1 << WAIT_TABLE_BITS)
+static wait_queue_head_t bit_wait_table[WAIT_TABLE_SIZE] __cacheline_aligned;
+
+wait_queue_head_t *bit_waitqueue(void *word, int bit)
+{
+	const int shift = BITS_PER_LONG == 32 ? 5 : 6;
+	unsigned long val = (unsigned long)word << shift | bit;
+
+	return bit_wait_table + hash_long(val, WAIT_TABLE_BITS);
+}
+EXPORT_SYMBOL(bit_waitqueue);
+
 void __init sched_init(void)
 {
 	int i, j;
 	unsigned long alloc_size = 0, ptr;
 
+	for (i = 0; i < WAIT_TABLE_SIZE; i++)
+		init_waitqueue_head(bit_wait_table + i);
+
 #ifdef CONFIG_FAIR_GROUP_SCHED
 	alloc_size += 2 * nr_cpu_ids * sizeof(void **);
 #endif
diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c
index 69e06898997d..fd4659313640 100644
--- a/kernel/sched/cpufreq_schedutil.c
+++ b/kernel/sched/cpufreq_schedutil.c
@@ -12,11 +12,14 @@
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
 #include <linux/cpufreq.h>
+#include <linux/kthread.h>
 #include <linux/slab.h>
 #include <trace/events/power.h>
 
 #include "sched.h"
 
+#define SUGOV_KTHREAD_PRIORITY	50
+
 struct sugov_tunables {
 	struct gov_attr_set attr_set;
 	unsigned int rate_limit_us;
@@ -35,8 +38,10 @@ struct sugov_policy {
 
 	/* The next fields are only needed if fast switch cannot be used. */
 	struct irq_work irq_work;
-	struct work_struct work;
+	struct kthread_work work;
 	struct mutex work_lock;
+	struct kthread_worker worker;
+	struct task_struct *thread;
 	bool work_in_progress;
 
 	bool need_freq_update;
@@ -291,7 +296,7 @@ static void sugov_update_shared(struct update_util_data *hook, u64 time,
 	raw_spin_unlock(&sg_policy->update_lock);
 }
 
-static void sugov_work(struct work_struct *work)
+static void sugov_work(struct kthread_work *work)
 {
 	struct sugov_policy *sg_policy = container_of(work, struct sugov_policy, work);
 
@@ -308,7 +313,21 @@ static void sugov_irq_work(struct irq_work *irq_work)
 	struct sugov_policy *sg_policy;
 
 	sg_policy = container_of(irq_work, struct sugov_policy, irq_work);
-	schedule_work_on(smp_processor_id(), &sg_policy->work);
+
+	/*
+	 * For RT and deadline tasks, the schedutil governor shoots the
+	 * frequency to maximum. Special care must be taken to ensure that this
+	 * kthread doesn't result in the same behavior.
+	 *
+	 * This is (mostly) guaranteed by the work_in_progress flag. The flag is
+	 * updated only at the end of the sugov_work() function and before that
+	 * the schedutil governor rejects all other frequency scaling requests.
+	 *
+	 * There is a very rare case though, where the RT thread yields right
+	 * after the work_in_progress flag is cleared. The effects of that are
+	 * neglected for now.
+	 */
+	kthread_queue_work(&sg_policy->worker, &sg_policy->work);
 }
 
 /************************** sysfs interface ************************/
@@ -371,19 +390,64 @@ static struct sugov_policy *sugov_policy_alloc(struct cpufreq_policy *policy)
 		return NULL;
 
 	sg_policy->policy = policy;
-	init_irq_work(&sg_policy->irq_work, sugov_irq_work);
-	INIT_WORK(&sg_policy->work, sugov_work);
-	mutex_init(&sg_policy->work_lock);
 	raw_spin_lock_init(&sg_policy->update_lock);
 	return sg_policy;
 }
 
 static void sugov_policy_free(struct sugov_policy *sg_policy)
 {
-	mutex_destroy(&sg_policy->work_lock);
 	kfree(sg_policy);
 }
 
+static int sugov_kthread_create(struct sugov_policy *sg_policy)
+{
+	struct task_struct *thread;
+	struct sched_param param = { .sched_priority = MAX_USER_RT_PRIO / 2 };
+	struct cpufreq_policy *policy = sg_policy->policy;
+	int ret;
+
+	/* kthread only required for slow path */
+	if (policy->fast_switch_enabled)
+		return 0;
+
+	kthread_init_work(&sg_policy->work, sugov_work);
+	kthread_init_worker(&sg_policy->worker);
+	thread = kthread_create(kthread_worker_fn, &sg_policy->worker,
+				"sugov:%d",
+				cpumask_first(policy->related_cpus));
+	if (IS_ERR(thread)) {
+		pr_err("failed to create sugov thread: %ld\n", PTR_ERR(thread));
+		return PTR_ERR(thread);
+	}
+
+	ret = sched_setscheduler_nocheck(thread, SCHED_FIFO, &param);
+	if (ret) {
+		kthread_stop(thread);
+		pr_warn("%s: failed to set SCHED_FIFO\n", __func__);
+		return ret;
+	}
+
+	sg_policy->thread = thread;
+	kthread_bind_mask(thread, policy->related_cpus);
+	init_irq_work(&sg_policy->irq_work, sugov_irq_work);
+	mutex_init(&sg_policy->work_lock);
+
+	wake_up_process(thread);
+
+	return 0;
+}
+
+static void sugov_kthread_stop(struct sugov_policy *sg_policy)
+{
+	/* kthread only required for slow path */
+	if (sg_policy->policy->fast_switch_enabled)
+		return;
+
+	kthread_flush_worker(&sg_policy->worker);
+	kthread_stop(sg_policy->thread);
+	mutex_destroy(&sg_policy->work_lock);
+}
+
 static struct sugov_tunables *sugov_tunables_alloc(struct sugov_policy *sg_policy)
 {
 	struct sugov_tunables *tunables;
@@ -416,16 +480,24 @@ static int sugov_init(struct cpufreq_policy *policy)
 	if (policy->governor_data)
 		return -EBUSY;
 
+	cpufreq_enable_fast_switch(policy);
+
 	sg_policy = sugov_policy_alloc(policy);
-	if (!sg_policy)
-		return -ENOMEM;
+	if (!sg_policy) {
+		ret = -ENOMEM;
+		goto disable_fast_switch;
+	}
+
+	ret = sugov_kthread_create(sg_policy);
+	if (ret)
+		goto free_sg_policy;
 
 	mutex_lock(&global_tunables_lock);
 
 	if (global_tunables) {
 		if (WARN_ON(have_governor_per_policy())) {
 			ret = -EINVAL;
-			goto free_sg_policy;
+			goto stop_kthread;
 		}
 		policy->governor_data = sg_policy;
 		sg_policy->tunables = global_tunables;
@@ -437,7 +509,7 @@ static int sugov_init(struct cpufreq_policy *policy)
 	tunables = sugov_tunables_alloc(sg_policy);
 	if (!tunables) {
 		ret = -ENOMEM;
-		goto free_sg_policy;
+		goto stop_kthread;
 	}
 
 	tunables->rate_limit_us = LATENCY_MULTIPLIER;
@@ -454,20 +526,25 @@ static int sugov_init(struct cpufreq_policy *policy)
 	if (ret)
 		goto fail;
 
- out:
+out:
 	mutex_unlock(&global_tunables_lock);
-
-	cpufreq_enable_fast_switch(policy);
 	return 0;
 
- fail:
+fail:
 	policy->governor_data = NULL;
 	sugov_tunables_free(tunables);
 
- free_sg_policy:
+stop_kthread:
+	sugov_kthread_stop(sg_policy);
+
+free_sg_policy:
 	mutex_unlock(&global_tunables_lock);
 
 	sugov_policy_free(sg_policy);
+
+disable_fast_switch:
+	cpufreq_disable_fast_switch(policy);
+
 	pr_err("initialization failed (error %d)\n", ret);
 	return ret;
 }
@@ -478,8 +555,6 @@ static void sugov_exit(struct cpufreq_policy *policy)
 	struct sugov_tunables *tunables = sg_policy->tunables;
 	unsigned int count;
 
-	cpufreq_disable_fast_switch(policy);
-
 	mutex_lock(&global_tunables_lock);
 
 	count = gov_attr_set_put(&tunables->attr_set, &sg_policy->tunables_hook);
@@ -489,7 +564,9 @@ static void sugov_exit(struct cpufreq_policy *policy)
 
 	mutex_unlock(&global_tunables_lock);
 
+	sugov_kthread_stop(sg_policy);
 	sugov_policy_free(sg_policy);
+	cpufreq_disable_fast_switch(policy);
 }
 
 static int sugov_start(struct cpufreq_policy *policy)
@@ -535,8 +612,10 @@ static void sugov_stop(struct cpufreq_policy *policy)
 
 	synchronize_sched();
 
-	irq_work_sync(&sg_policy->irq_work);
-	cancel_work_sync(&sg_policy->work);
+	if (!policy->fast_switch_enabled) {
+		irq_work_sync(&sg_policy->irq_work);
+		kthread_cancel_work_sync(&sg_policy->work);
+	}
 }
 
 static void sugov_limits(struct cpufreq_policy *policy)
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 2d4ad72f8f3c..c242944f5cbd 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -690,7 +690,14 @@ void init_entity_runnable_average(struct sched_entity *se)
 	 * will definitely be update (after enqueue).
 	 */
 	sa->period_contrib = 1023;
-	sa->load_avg = scale_load_down(se->load.weight);
+	/*
+	 * Tasks are intialized with full load to be seen as heavy tasks until
+	 * they get a chance to stabilize to their real load level.
+	 * Group entities are intialized with zero load to reflect the fact that
+	 * nothing has been attached to the task group yet.
+	 */
+	if (entity_is_task(se))
+		sa->load_avg = scale_load_down(se->load.weight);
 	sa->load_sum = sa->load_avg * LOAD_AVG_MAX;
 	/*
 	 * At this point, util_avg won't be used in select_task_rq_fair anyway
@@ -5471,13 +5478,18 @@ static inline int select_idle_smt(struct task_struct *p, struct sched_domain *sd
  */
 static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int target)
 {
-	struct sched_domain *this_sd = rcu_dereference(*this_cpu_ptr(&sd_llc));
-	u64 avg_idle = this_rq()->avg_idle;
-	u64 avg_cost = this_sd->avg_scan_cost;
+	struct sched_domain *this_sd;
+	u64 avg_cost, avg_idle = this_rq()->avg_idle;
 	u64 time, cost;
 	s64 delta;
 	int cpu, wrap;
 
+	this_sd = rcu_dereference(*this_cpu_ptr(&sd_llc));
+	if (!this_sd)
+		return -1;
+
+	avg_cost = this_sd->avg_scan_cost;
+
 	/*
 	 * Due to large variance we need a large fuzz factor; hackbench in
 	 * particularly is sensitive here.
@@ -8827,7 +8839,6 @@ int alloc_fair_sched_group(struct task_group *tg, struct task_group *parent)
 {
 	struct sched_entity *se;
 	struct cfs_rq *cfs_rq;
-	struct rq *rq;
 	int i;
 
 	tg->cfs_rq = kzalloc(sizeof(cfs_rq) * nr_cpu_ids, GFP_KERNEL);
@@ -8842,8 +8853,6 @@ int alloc_fair_sched_group(struct task_group *tg, struct task_group *parent)
 	init_cfs_bandwidth(tg_cfs_bandwidth(tg));
 
 	for_each_possible_cpu(i) {
-		rq = cpu_rq(i);
-
 		cfs_rq = kzalloc_node(sizeof(struct cfs_rq),
 				      GFP_KERNEL, cpu_to_node(i));
 		if (!cfs_rq)
diff --git a/kernel/sched/wait.c b/kernel/sched/wait.c
index 4f7053579fe3..9453efe9b25a 100644
--- a/kernel/sched/wait.c
+++ b/kernel/sched/wait.c
@@ -480,16 +480,6 @@ void wake_up_bit(void *word, int bit)
 }
 EXPORT_SYMBOL(wake_up_bit);
 
-wait_queue_head_t *bit_waitqueue(void *word, int bit)
-{
-	const int shift = BITS_PER_LONG == 32 ? 5 : 6;
-	const struct zone *zone = page_zone(virt_to_page(word));
-	unsigned long val = (unsigned long)word << shift | bit;
-
-	return &zone->wait_table[hash_long(val, zone->wait_table_bits)];
-}
-EXPORT_SYMBOL(bit_waitqueue);
-
 /*
  * Manipulate the atomic_t address to produce a better bit waitqueue table hash
  * index (we're keying off bit -1, but that would produce a horrible hash