Merge master.kernel.org:/pub/scm/linux/kernel/git/davej/cpufreq

* master.kernel.org:/pub/scm/linux/kernel/git/davej/cpufreq:
  Move workqueue exports to where the functions are defined.
  [CPUFREQ] Misc cleanups in ondemand.
  [CPUFREQ] Make ondemand sampling per CPU and remove the mutex usage in sampling path.
  [CPUFREQ] Add queue_delayed_work_on() interface for workqueues.
  [CPUFREQ] Remove slowdown from ondemand sampling path.

1 files changed, 91 insertions, 169 deletions

diff --git a/drivers/cpufreq/cpufreq_ondemand.c b/drivers/cpufreq/cpufreq_ondemand.c
index 693e540481b4..87299924e735 100644
--- a/drivers/cpufreq/cpufreq_ondemand.c
+++ b/drivers/cpufreq/cpufreq_ondemand.c
@@ -12,22 +12,11 @@
 
 #include <linux/kernel.h>
 #include <linux/module.h>
-#include <linux/smp.h>
 #include <linux/init.h>
-#include <linux/interrupt.h>
-#include <linux/ctype.h>
 #include <linux/cpufreq.h>
-#include <linux/sysctl.h>
-#include <linux/types.h>
-#include <linux/fs.h>
-#include <linux/sysfs.h>
 #include <linux/cpu.h>
-#include <linux/sched.h>
-#include <linux/kmod.h>
-#include <linux/workqueue.h>
 #include <linux/jiffies.h>
 #include <linux/kernel_stat.h>
-#include <linux/percpu.h>
 #include <linux/mutex.h>
 
 /*
@@ -56,16 +45,15 @@ static unsigned int def_sampling_rate;
 #define MIN_SAMPLING_RATE			(def_sampling_rate / MIN_SAMPLING_RATE_RATIO)
 #define MAX_SAMPLING_RATE			(500 * def_sampling_rate)
 #define DEF_SAMPLING_RATE_LATENCY_MULTIPLIER	(1000)
-#define DEF_SAMPLING_DOWN_FACTOR		(1)
-#define MAX_SAMPLING_DOWN_FACTOR		(10)
 #define TRANSITION_LATENCY_LIMIT		(10 * 1000)
 
 static void do_dbs_timer(void *data);
 
 struct cpu_dbs_info_s {
+	cputime64_t prev_cpu_idle;
+	cputime64_t prev_cpu_wall;
 	struct cpufreq_policy *cur_policy;
-	unsigned int prev_cpu_idle_up;
-	unsigned int prev_cpu_idle_down;
+ 	struct work_struct work;
 	unsigned int enable;
 };
 static DEFINE_PER_CPU(struct cpu_dbs_info_s, cpu_dbs_info);
@@ -80,31 +68,32 @@ static unsigned int dbs_enable;	/* number of CPUs using this policy */
  * cpu_hotplug lock should be taken before that. Note that cpu_hotplug lock
  * is recursive for the same process. -Venki
  */
-static DEFINE_MUTEX (dbs_mutex);
-static DECLARE_WORK	(dbs_work, do_dbs_timer, NULL);
+static DEFINE_MUTEX(dbs_mutex);
 
-static struct workqueue_struct *dbs_workq;
+static struct workqueue_struct	*kondemand_wq;
 
 struct dbs_tuners {
 	unsigned int sampling_rate;
-	unsigned int sampling_down_factor;
 	unsigned int up_threshold;
 	unsigned int ignore_nice;
 };
 
 static struct dbs_tuners dbs_tuners_ins = {
 	.up_threshold = DEF_FREQUENCY_UP_THRESHOLD,
-	.sampling_down_factor = DEF_SAMPLING_DOWN_FACTOR,
 	.ignore_nice = 0,
 };
 
-static inline unsigned int get_cpu_idle_time(unsigned int cpu)
+static inline cputime64_t get_cpu_idle_time(unsigned int cpu)
 {
-	return	kstat_cpu(cpu).cpustat.idle +
-		kstat_cpu(cpu).cpustat.iowait +
-		( dbs_tuners_ins.ignore_nice ?
-		  kstat_cpu(cpu).cpustat.nice :
-		  0);
+	cputime64_t retval;
+
+	retval = cputime64_add(kstat_cpu(cpu).cpustat.idle,
+			kstat_cpu(cpu).cpustat.iowait);
+
+	if (dbs_tuners_ins.ignore_nice)
+		retval = cputime64_add(retval, kstat_cpu(cpu).cpustat.nice);
+
+	return retval;
 }
 
 /************************** sysfs interface ************************/
@@ -133,35 +122,15 @@ static ssize_t show_##file_name						\
 	return sprintf(buf, "%u\n", dbs_tuners_ins.object);		\
 }
 show_one(sampling_rate, sampling_rate);
-show_one(sampling_down_factor, sampling_down_factor);
 show_one(up_threshold, up_threshold);
 show_one(ignore_nice_load, ignore_nice);
 
-static ssize_t store_sampling_down_factor(struct cpufreq_policy *unused,
-		const char *buf, size_t count)
-{
-	unsigned int input;
-	int ret;
-	ret = sscanf (buf, "%u", &input);
-	if (ret != 1 )
-		return -EINVAL;
-
-	if (input > MAX_SAMPLING_DOWN_FACTOR || input < 1)
-		return -EINVAL;
-
-	mutex_lock(&dbs_mutex);
-	dbs_tuners_ins.sampling_down_factor = input;
-	mutex_unlock(&dbs_mutex);
-
-	return count;
-}
-
 static ssize_t store_sampling_rate(struct cpufreq_policy *unused,
 		const char *buf, size_t count)
 {
 	unsigned int input;
 	int ret;
-	ret = sscanf (buf, "%u", &input);
+	ret = sscanf(buf, "%u", &input);
 
 	mutex_lock(&dbs_mutex);
 	if (ret != 1 || input > MAX_SAMPLING_RATE || input < MIN_SAMPLING_RATE) {
@@ -180,7 +149,7 @@ static ssize_t store_up_threshold(struct cpufreq_policy *unused,
 {
 	unsigned int input;
 	int ret;
-	ret = sscanf (buf, "%u", &input);
+	ret = sscanf(buf, "%u", &input);
 
 	mutex_lock(&dbs_mutex);
 	if (ret != 1 || input > MAX_FREQUENCY_UP_THRESHOLD ||
@@ -203,7 +172,7 @@ static ssize_t store_ignore_nice_load(struct cpufreq_policy *policy,
 
 	unsigned int j;
 
-	ret = sscanf (buf, "%u", &input);
+	ret = sscanf(buf, "%u", &input);
 	if ( ret != 1 )
 		return -EINVAL;
 
@@ -217,12 +186,12 @@ static ssize_t store_ignore_nice_load(struct cpufreq_policy *policy,
 	}
 	dbs_tuners_ins.ignore_nice = input;
 
-	/* we need to re-evaluate prev_cpu_idle_up and prev_cpu_idle_down */
+	/* we need to re-evaluate prev_cpu_idle */
 	for_each_online_cpu(j) {
-		struct cpu_dbs_info_s *j_dbs_info;
-		j_dbs_info = &per_cpu(cpu_dbs_info, j);
-		j_dbs_info->prev_cpu_idle_up = get_cpu_idle_time(j);
-		j_dbs_info->prev_cpu_idle_down = j_dbs_info->prev_cpu_idle_up;
+		struct cpu_dbs_info_s *dbs_info;
+		dbs_info = &per_cpu(cpu_dbs_info, j);
+		dbs_info->prev_cpu_idle = get_cpu_idle_time(j);
+		dbs_info->prev_cpu_wall = get_jiffies_64();
 	}
 	mutex_unlock(&dbs_mutex);
 
@@ -234,7 +203,6 @@ static struct freq_attr _name = \
 __ATTR(_name, 0644, show_##_name, store_##_name)
 
 define_one_rw(sampling_rate);
-define_one_rw(sampling_down_factor);
 define_one_rw(up_threshold);
 define_one_rw(ignore_nice_load);
 
@@ -242,7 +210,6 @@ static struct attribute * dbs_attributes[] = {
 	&sampling_rate_max.attr,
 	&sampling_rate_min.attr,
 	&sampling_rate.attr,
-	&sampling_down_factor.attr,
 	&up_threshold.attr,
 	&ignore_nice_load.attr,
 	NULL
@@ -255,26 +222,27 @@ static struct attribute_group dbs_attr_group = {
 
 /************************** sysfs end ************************/
 
-static void dbs_check_cpu(int cpu)
+static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info)
 {
-	unsigned int idle_ticks, up_idle_ticks, total_ticks;
-	unsigned int freq_next;
-	unsigned int freq_down_sampling_rate;
-	static int down_skip[NR_CPUS];
-	struct cpu_dbs_info_s *this_dbs_info;
+	unsigned int idle_ticks, total_ticks;
+	unsigned int load;
+	cputime64_t cur_jiffies;
 
 	struct cpufreq_policy *policy;
 	unsigned int j;
 
-	this_dbs_info = &per_cpu(cpu_dbs_info, cpu);
 	if (!this_dbs_info->enable)
 		return;
 
 	policy = this_dbs_info->cur_policy;
+	cur_jiffies = jiffies64_to_cputime64(get_jiffies_64());
+	total_ticks = (unsigned int) cputime64_sub(cur_jiffies,
+			this_dbs_info->prev_cpu_wall);
+	this_dbs_info->prev_cpu_wall = cur_jiffies;
 	/*
 	 * Every sampling_rate, we check, if current idle time is less
 	 * than 20% (default), then we try to increase frequency
-	 * Every sampling_rate*sampling_down_factor, we look for a the lowest
+	 * Every sampling_rate, we look for a the lowest
 	 * frequency which can sustain the load while keeping idle time over
 	 * 30%. If such a frequency exist, we try to decrease to this frequency.
 	 *
@@ -283,36 +251,26 @@ static void dbs_check_cpu(int cpu)
 	 * 5% (default) of current frequency
 	 */
 
-	/* Check for frequency increase */
+	/* Get Idle Time */
 	idle_ticks = UINT_MAX;
 	for_each_cpu_mask(j, policy->cpus) {
-		unsigned int tmp_idle_ticks, total_idle_ticks;
+		cputime64_t total_idle_ticks;
+		unsigned int tmp_idle_ticks;
 		struct cpu_dbs_info_s *j_dbs_info;
 
 		j_dbs_info = &per_cpu(cpu_dbs_info, j);
 		total_idle_ticks = get_cpu_idle_time(j);
-		tmp_idle_ticks = total_idle_ticks -
-			j_dbs_info->prev_cpu_idle_up;
-		j_dbs_info->prev_cpu_idle_up = total_idle_ticks;
+		tmp_idle_ticks = (unsigned int) cputime64_sub(total_idle_ticks,
+				j_dbs_info->prev_cpu_idle);
+		j_dbs_info->prev_cpu_idle = total_idle_ticks;
 
 		if (tmp_idle_ticks < idle_ticks)
 			idle_ticks = tmp_idle_ticks;
 	}
+	load = (100 * (total_ticks - idle_ticks)) / total_ticks;
 
-	/* Scale idle ticks by 100 and compare with up and down ticks */
-	idle_ticks *= 100;
-	up_idle_ticks = (100 - dbs_tuners_ins.up_threshold) *
-			usecs_to_jiffies(dbs_tuners_ins.sampling_rate);
-
-	if (idle_ticks < up_idle_ticks) {
-		down_skip[cpu] = 0;
-		for_each_cpu_mask(j, policy->cpus) {
-			struct cpu_dbs_info_s *j_dbs_info;
-
-			j_dbs_info = &per_cpu(cpu_dbs_info, j);
-			j_dbs_info->prev_cpu_idle_down =
-					j_dbs_info->prev_cpu_idle_up;
-		}
+	/* Check for frequency increase */
+	if (load > dbs_tuners_ins.up_threshold) {
 		/* if we are already at full speed then break out early */
 		if (policy->cur == policy->max)
 			return;
@@ -323,83 +281,49 @@ static void dbs_check_cpu(int cpu)
 	}
 
 	/* Check for frequency decrease */
-	down_skip[cpu]++;
-	if (down_skip[cpu] < dbs_tuners_ins.sampling_down_factor)
-		return;
-
-	idle_ticks = UINT_MAX;
-	for_each_cpu_mask(j, policy->cpus) {
-		unsigned int tmp_idle_ticks, total_idle_ticks;
-		struct cpu_dbs_info_s *j_dbs_info;
-
-		j_dbs_info = &per_cpu(cpu_dbs_info, j);
-		/* Check for frequency decrease */
-		total_idle_ticks = j_dbs_info->prev_cpu_idle_up;
-		tmp_idle_ticks = total_idle_ticks -
-			j_dbs_info->prev_cpu_idle_down;
-		j_dbs_info->prev_cpu_idle_down = total_idle_ticks;
-
-		if (tmp_idle_ticks < idle_ticks)
-			idle_ticks = tmp_idle_ticks;
-	}
-
-	down_skip[cpu] = 0;
 	/* if we cannot reduce the frequency anymore, break out early */
 	if (policy->cur == policy->min)
 		return;
 
-	/* Compute how many ticks there are between two measurements */
-	freq_down_sampling_rate = dbs_tuners_ins.sampling_rate *
-		dbs_tuners_ins.sampling_down_factor;
-	total_ticks = usecs_to_jiffies(freq_down_sampling_rate);
-
 	/*
 	 * The optimal frequency is the frequency that is the lowest that
 	 * can support the current CPU usage without triggering the up
 	 * policy. To be safe, we focus 10 points under the threshold.
 	 */
-	freq_next = ((total_ticks - idle_ticks) * 100) / total_ticks;
-	freq_next = (freq_next * policy->cur) /
+	if (load < (dbs_tuners_ins.up_threshold - 10)) {
+		unsigned int freq_next;
+		freq_next = (policy->cur * load) /
 			(dbs_tuners_ins.up_threshold - 10);
 
-	if (freq_next < policy->min)
-		freq_next = policy->min;
-
-	if (freq_next <= ((policy->cur * 95) / 100))
 		__cpufreq_driver_target(policy, freq_next, CPUFREQ_RELATION_L);
+	}
 }
 
 static void do_dbs_timer(void *data)
 {
-	int i;
-	lock_cpu_hotplug();
-	mutex_lock(&dbs_mutex);
-	for_each_online_cpu(i)
-		dbs_check_cpu(i);
-	queue_delayed_work(dbs_workq, &dbs_work,
-			   usecs_to_jiffies(dbs_tuners_ins.sampling_rate));
-	mutex_unlock(&dbs_mutex);
-	unlock_cpu_hotplug();
+	unsigned int cpu = smp_processor_id();
+	struct cpu_dbs_info_s *dbs_info = &per_cpu(cpu_dbs_info, cpu);
+
+	dbs_check_cpu(dbs_info);
+	queue_delayed_work_on(cpu, kondemand_wq, &dbs_info->work,
+			usecs_to_jiffies(dbs_tuners_ins.sampling_rate));
 }
 
-static inline void dbs_timer_init(void)
+static inline void dbs_timer_init(unsigned int cpu)
 {
-	INIT_WORK(&dbs_work, do_dbs_timer, NULL);
-	if (!dbs_workq)
-		dbs_workq = create_singlethread_workqueue("ondemand");
-	if (!dbs_workq) {
-		printk(KERN_ERR "ondemand: Cannot initialize kernel thread\n");
-		return;
-	}
-	queue_delayed_work(dbs_workq, &dbs_work,
-			   usecs_to_jiffies(dbs_tuners_ins.sampling_rate));
+	struct cpu_dbs_info_s *dbs_info = &per_cpu(cpu_dbs_info, cpu);
+
+	INIT_WORK(&dbs_info->work, do_dbs_timer, 0);
+	queue_delayed_work_on(cpu, kondemand_wq, &dbs_info->work,
+			usecs_to_jiffies(dbs_tuners_ins.sampling_rate));
 	return;
 }
 
-static inline void dbs_timer_exit(void)
+static inline void dbs_timer_exit(unsigned int cpu)
 {
-	if (dbs_workq)
-		cancel_rearming_delayed_workqueue(dbs_workq, &dbs_work);
+	struct cpu_dbs_info_s *dbs_info = &per_cpu(cpu_dbs_info, cpu);
+
+	cancel_rearming_delayed_workqueue(kondemand_wq, &dbs_info->work);
 }
 
 static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
@@ -413,8 +337,7 @@ static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
 
 	switch (event) {
 	case CPUFREQ_GOV_START:
-		if ((!cpu_online(cpu)) ||
-		    (!policy->cur))
+		if ((!cpu_online(cpu)) || (!policy->cur))
 			return -EINVAL;
 
 		if (policy->cpuinfo.transition_latency >
@@ -427,18 +350,26 @@ static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
 			break;
 
 		mutex_lock(&dbs_mutex);
+		dbs_enable++;
+		if (dbs_enable == 1) {
+			kondemand_wq = create_workqueue("kondemand");
+			if (!kondemand_wq) {
+				printk(KERN_ERR "Creation of kondemand failed\n");
+				dbs_enable--;
+				mutex_unlock(&dbs_mutex);
+				return -ENOSPC;
+			}
+		}
 		for_each_cpu_mask(j, policy->cpus) {
 			struct cpu_dbs_info_s *j_dbs_info;
 			j_dbs_info = &per_cpu(cpu_dbs_info, j);
 			j_dbs_info->cur_policy = policy;
 
-			j_dbs_info->prev_cpu_idle_up = get_cpu_idle_time(j);
-			j_dbs_info->prev_cpu_idle_down
-				= j_dbs_info->prev_cpu_idle_up;
+			j_dbs_info->prev_cpu_idle = get_cpu_idle_time(j);
+			j_dbs_info->prev_cpu_wall = get_jiffies_64();
 		}
 		this_dbs_info->enable = 1;
 		sysfs_create_group(&policy->kobj, &dbs_attr_group);
-		dbs_enable++;
 		/*
 		 * Start the timerschedule work, when this governor
 		 * is used for first time
@@ -457,23 +388,20 @@ static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
 				def_sampling_rate = MIN_STAT_SAMPLING_RATE;
 
 			dbs_tuners_ins.sampling_rate = def_sampling_rate;
-			dbs_timer_init();
 		}
+		dbs_timer_init(policy->cpu);
 
 		mutex_unlock(&dbs_mutex);
 		break;
 
 	case CPUFREQ_GOV_STOP:
 		mutex_lock(&dbs_mutex);
+		dbs_timer_exit(policy->cpu);
 		this_dbs_info->enable = 0;
 		sysfs_remove_group(&policy->kobj, &dbs_attr_group);
 		dbs_enable--;
-		/*
-		 * Stop the timerschedule work, when this governor
-		 * is used for first time
-		 */
 		if (dbs_enable == 0)
-			dbs_timer_exit();
+			destroy_workqueue(kondemand_wq);
 
 		mutex_unlock(&dbs_mutex);
 
@@ -483,13 +411,13 @@ static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
 		lock_cpu_hotplug();
 		mutex_lock(&dbs_mutex);
 		if (policy->max < this_dbs_info->cur_policy->cur)
-			__cpufreq_driver_target(
-					this_dbs_info->cur_policy,
-					policy->max, CPUFREQ_RELATION_H);
+			__cpufreq_driver_target(this_dbs_info->cur_policy,
+			                        policy->max,
+			                        CPUFREQ_RELATION_H);
 		else if (policy->min > this_dbs_info->cur_policy->cur)
-			__cpufreq_driver_target(
-					this_dbs_info->cur_policy,
-					policy->min, CPUFREQ_RELATION_L);
+			__cpufreq_driver_target(this_dbs_info->cur_policy,
+			                        policy->min,
+			                        CPUFREQ_RELATION_L);
 		mutex_unlock(&dbs_mutex);
 		unlock_cpu_hotplug();
 		break;
@@ -498,9 +426,9 @@ static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
 }
 
 static struct cpufreq_governor cpufreq_gov_dbs = {
-	.name		= "ondemand",
-	.governor	= cpufreq_governor_dbs,
-	.owner		= THIS_MODULE,
+	.name = "ondemand",
+	.governor = cpufreq_governor_dbs,
+	.owner = THIS_MODULE,
 };
 
 static int __init cpufreq_gov_dbs_init(void)
@@ -510,21 +438,15 @@ static int __init cpufreq_gov_dbs_init(void)
 
 static void __exit cpufreq_gov_dbs_exit(void)
 {
-	/* Make sure that the scheduled work is indeed not running.
-	   Assumes the timer has been cancelled first. */
-	if (dbs_workq) {
-		flush_workqueue(dbs_workq);
-		destroy_workqueue(dbs_workq);
-	}
-
 	cpufreq_unregister_governor(&cpufreq_gov_dbs);
 }
 
 
-MODULE_AUTHOR ("Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>");
-MODULE_DESCRIPTION ("'cpufreq_ondemand' - A dynamic cpufreq governor for "
-		"Low Latency Frequency Transition capable processors");
-MODULE_LICENSE ("GPL");
+MODULE_AUTHOR("Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>");
+MODULE_AUTHOR("Alexey Starikovskiy <alexey.y.starikovskiy@intel.com>");
+MODULE_DESCRIPTION("'cpufreq_ondemand' - A dynamic cpufreq governor for "
+                   "Low Latency Frequency Transition capable processors");
+MODULE_LICENSE("GPL");
 
 module_init(cpufreq_gov_dbs_init);
 module_exit(cpufreq_gov_dbs_exit);