1 files changed, 229 insertions, 502 deletions

diff --git a/drivers/cpufreq/cpufreq_ondemand.c b/drivers/cpufreq/cpufreq_ondemand.c
index 396322f2a83f..7731f7c7e79a 100644
--- a/drivers/cpufreq/cpufreq_ondemand.c
+++ b/drivers/cpufreq/cpufreq_ondemand.c
@@ -10,24 +10,23 @@
  * published by the Free Software Foundation.
  */
 
-#include <linux/kernel.h>
-#include <linux/module.h>
-#include <linux/init.h>
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
 #include <linux/cpufreq.h>
-#include <linux/cpu.h>
-#include <linux/jiffies.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
 #include <linux/kernel_stat.h>
+#include <linux/kobject.h>
+#include <linux/module.h>
 #include <linux/mutex.h>
-#include <linux/hrtimer.h>
+#include <linux/percpu-defs.h>
+#include <linux/sysfs.h>
 #include <linux/tick.h>
-#include <linux/ktime.h>
-#include <linux/sched.h>
+#include <linux/types.h>
 
-/*
- * dbs is used in this file as a shortform for demandbased switching
- * It helps to keep variable names smaller, simpler
- */
+#include "cpufreq_governor.h"
 
+/* On-demand governor macors */
 #define DEF_FREQUENCY_DOWN_DIFFERENTIAL         (10)
 #define DEF_FREQUENCY_UP_THRESHOLD              (80)
 #define DEF_SAMPLING_DOWN_FACTOR                (1)
@@ -38,80 +37,14 @@
 #define MIN_FREQUENCY_UP_THRESHOLD              (11)
 #define MAX_FREQUENCY_UP_THRESHOLD              (100)
 
-/*
- * The polling frequency of this governor depends on the capability of
- * the processor. Default polling frequency is 1000 times the transition
- * latency of the processor. The governor will work on any processor with
- * transition latency <= 10mS, using appropriate sampling
- * rate.
- * For CPUs with transition latency > 10mS (mostly drivers with CPUFREQ_ETERNAL)
- * this governor will not work.
- * All times here are in uS.
- */
-#define MIN_SAMPLING_RATE_RATIO                 (2)
-
-static unsigned int min_sampling_rate;
-
-#define LATENCY_MULTIPLIER                      (1000)
-#define MIN_LATENCY_MULTIPLIER                  (100)
-#define TRANSITION_LATENCY_LIMIT                (10 * 1000 * 1000)
-
-static void do_dbs_timer(struct work_struct *work);
-static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
-                                unsigned int event);
+static struct dbs_data od_dbs_data;
+static DEFINE_PER_CPU(struct od_cpu_dbs_info_s, od_cpu_dbs_info);
 
 #ifndef CONFIG_CPU_FREQ_DEFAULT_GOV_ONDEMAND
-static
+static struct cpufreq_governor cpufreq_gov_ondemand;
 #endif
-struct cpufreq_governor cpufreq_gov_ondemand = {
-       .name                   = "ondemand",
-       .governor               = cpufreq_governor_dbs,
-       .max_transition_latency = TRANSITION_LATENCY_LIMIT,
-       .owner                  = THIS_MODULE,
-};
 
-/* Sampling types */
-enum {DBS_NORMAL_SAMPLE, DBS_SUB_SAMPLE};
-
-struct cpu_dbs_info_s {
-        cputime64_t prev_cpu_idle;
-        cputime64_t prev_cpu_iowait;
-        cputime64_t prev_cpu_wall;
-        cputime64_t prev_cpu_nice;
-        struct cpufreq_policy *cur_policy;
-        struct delayed_work work;
-        struct cpufreq_frequency_table *freq_table;
-        unsigned int freq_lo;
-        unsigned int freq_lo_jiffies;
-        unsigned int freq_hi_jiffies;
-        unsigned int rate_mult;
-        int cpu;
-        unsigned int sample_type:1;
-        /*
-         * percpu mutex that serializes governor limit change with
-         * do_dbs_timer invocation. We do not want do_dbs_timer to run
-         * when user is changing the governor or limits.
-         */
-        struct mutex timer_mutex;
-};
-static DEFINE_PER_CPU(struct cpu_dbs_info_s, od_cpu_dbs_info);
-
-static unsigned int dbs_enable; /* number of CPUs using this policy */
-
-/*
- * dbs_mutex protects dbs_enable in governor start/stop.
- */
-static DEFINE_MUTEX(dbs_mutex);
-
-static struct dbs_tuners {
-        unsigned int sampling_rate;
-        unsigned int up_threshold;
-        unsigned int down_differential;
-        unsigned int ignore_nice;
-        unsigned int sampling_down_factor;
-        unsigned int powersave_bias;
-        unsigned int io_is_busy;
-} dbs_tuners_ins = {
+static struct od_dbs_tuners od_tuners = {
         .up_threshold = DEF_FREQUENCY_UP_THRESHOLD,
         .sampling_down_factor = DEF_SAMPLING_DOWN_FACTOR,
         .down_differential = DEF_FREQUENCY_DOWN_DIFFERENTIAL,
@@ -119,48 +52,35 @@ static struct dbs_tuners {
         .powersave_bias = 0,
 };
 
-static inline u64 get_cpu_idle_time_jiffy(unsigned int cpu, u64 *wall)
-{
-        u64 idle_time;
-        u64 cur_wall_time;
-        u64 busy_time;
-
-        cur_wall_time = jiffies64_to_cputime64(get_jiffies_64());
-
-        busy_time  = kcpustat_cpu(cpu).cpustat[CPUTIME_USER];
-        busy_time += kcpustat_cpu(cpu).cpustat[CPUTIME_SYSTEM];
-        busy_time += kcpustat_cpu(cpu).cpustat[CPUTIME_IRQ];
-        busy_time += kcpustat_cpu(cpu).cpustat[CPUTIME_SOFTIRQ];
-        busy_time += kcpustat_cpu(cpu).cpustat[CPUTIME_STEAL];
-        busy_time += kcpustat_cpu(cpu).cpustat[CPUTIME_NICE];
-
-        idle_time = cur_wall_time - busy_time;
-        if (wall)
-                *wall = jiffies_to_usecs(cur_wall_time);
-
-        return jiffies_to_usecs(idle_time);
-}
-
-static inline cputime64_t get_cpu_idle_time(unsigned int cpu, cputime64_t *wall)
+static void ondemand_powersave_bias_init_cpu(int cpu)
 {
-        u64 idle_time = get_cpu_idle_time_us(cpu, NULL);
-
-        if (idle_time == -1ULL)
-                return get_cpu_idle_time_jiffy(cpu, wall);
-        else
-                idle_time += get_cpu_iowait_time_us(cpu, wall);
+        struct od_cpu_dbs_info_s *dbs_info = &per_cpu(od_cpu_dbs_info, cpu);
 
-        return idle_time;
+        dbs_info->freq_table = cpufreq_frequency_get_table(cpu);
+        dbs_info->freq_lo = 0;
 }
 
-static inline cputime64_t get_cpu_iowait_time(unsigned int cpu, cputime64_t *wall)
+/*
+ * Not all CPUs want IO time to be accounted as busy; this depends on how
+ * efficient idling at a higher frequency/voltage is.
+ * Pavel Machek says this is not so for various generations of AMD and old
+ * Intel systems.
+ * Mike Chan (androidlcom) calis this is also not true for ARM.
+ * Because of this, whitelist specific known (series) of CPUs by default, and
+ * leave all others up to the user.
+ */
+static int should_io_be_busy(void)
 {
-        u64 iowait_time = get_cpu_iowait_time_us(cpu, wall);
-
-        if (iowait_time == -1ULL)
-                return 0;
-
-        return iowait_time;
+#if defined(CONFIG_X86)
+        /*
+         * For Intel, Core 2 (model 15) andl later have an efficient idle.
+         */
+        if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL &&
+                        boot_cpu_data.x86 == 6 &&
+                        boot_cpu_data.x86_model >= 15)
+                return 1;
+#endif
+        return 0;
 }
 
 /*
@@ -169,14 +89,13 @@ static inline cputime64_t get_cpu_iowait_time(unsigned int cpu, cputime64_t *wal
  * freq_lo, and freq_lo_jiffies in percpu area for averaging freqs.
  */
 static unsigned int powersave_bias_target(struct cpufreq_policy *policy,
-                                          unsigned int freq_next,
-                                          unsigned int relation)
+                unsigned int freq_next, unsigned int relation)
 {
         unsigned int freq_req, freq_reduc, freq_avg;
         unsigned int freq_hi, freq_lo;
         unsigned int index = 0;
         unsigned int jiffies_total, jiffies_hi, jiffies_lo;
-        struct cpu_dbs_info_s *dbs_info = &per_cpu(od_cpu_dbs_info,
+        struct od_cpu_dbs_info_s *dbs_info = &per_cpu(od_cpu_dbs_info,
                                                    policy->cpu);
 
         if (!dbs_info->freq_table) {
@@ -188,7 +107,7 @@ static unsigned int powersave_bias_target(struct cpufreq_policy *policy,
         cpufreq_frequency_table_target(policy, dbs_info->freq_table, freq_next,
                         relation, &index);
         freq_req = dbs_info->freq_table[index].frequency;
-        freq_reduc = freq_req * dbs_tuners_ins.powersave_bias / 1000;
+        freq_reduc = freq_req * od_tuners.powersave_bias / 1000;
         freq_avg = freq_req - freq_reduc;
 
         /* Find freq bounds for freq_avg in freq_table */
@@ -207,7 +126,7 @@ static unsigned int powersave_bias_target(struct cpufreq_policy *policy,
                 dbs_info->freq_lo_jiffies = 0;
                 return freq_lo;
         }
-        jiffies_total = usecs_to_jiffies(dbs_tuners_ins.sampling_rate);
+        jiffies_total = usecs_to_jiffies(od_tuners.sampling_rate);
         jiffies_hi = (freq_avg - freq_lo) * jiffies_total;
         jiffies_hi += ((freq_hi - freq_lo) / 2);
         jiffies_hi /= (freq_hi - freq_lo);
@@ -218,13 +137,6 @@ static unsigned int powersave_bias_target(struct cpufreq_policy *policy,
         return freq_hi;
 }
 
-static void ondemand_powersave_bias_init_cpu(int cpu)
-{
-        struct cpu_dbs_info_s *dbs_info = &per_cpu(od_cpu_dbs_info, cpu);
-        dbs_info->freq_table = cpufreq_frequency_get_table(cpu);
-        dbs_info->freq_lo = 0;
-}
-
 static void ondemand_powersave_bias_init(void)
 {
         int i;
@@ -233,83 +145,173 @@ static void ondemand_powersave_bias_init(void)
         }
 }
 
-/************************** sysfs interface ************************/
+static void dbs_freq_increase(struct cpufreq_policy *p, unsigned int freq)
+{
+        if (od_tuners.powersave_bias)
+                freq = powersave_bias_target(p, freq, CPUFREQ_RELATION_H);
+        else if (p->cur == p->max)
+                return;
 
-static ssize_t show_sampling_rate_min(struct kobject *kobj,
-                                      struct attribute *attr, char *buf)
+        __cpufreq_driver_target(p, freq, od_tuners.powersave_bias ?
+                        CPUFREQ_RELATION_L : CPUFREQ_RELATION_H);
+}
+
+/*
+ * Every sampling_rate, we check, if current idle time is less than 20%
+ * (default), then we try to increase frequency 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.
+ *
+ * Any frequency increase takes it to the maximum frequency. Frequency reduction
+ * happens at minimum steps of 5% (default) of current frequency
+ */
+static void od_check_cpu(int cpu, unsigned int load_freq)
 {
-        return sprintf(buf, "%u\n", min_sampling_rate);
+        struct od_cpu_dbs_info_s *dbs_info = &per_cpu(od_cpu_dbs_info, cpu);
+        struct cpufreq_policy *policy = dbs_info->cdbs.cur_policy;
+
+        dbs_info->freq_lo = 0;
+
+        /* Check for frequency increase */
+        if (load_freq > od_tuners.up_threshold * policy->cur) {
+                /* If switching to max speed, apply sampling_down_factor */
+                if (policy->cur < policy->max)
+                        dbs_info->rate_mult =
+                                od_tuners.sampling_down_factor;
+                dbs_freq_increase(policy, policy->max);
+                return;
+        }
+
+        /* Check for frequency decrease */
+        /* if we cannot reduce the frequency anymore, break out early */
+        if (policy->cur == policy->min)
+                return;
+
+        /*
+         * 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.
+         */
+        if (load_freq < (od_tuners.up_threshold - od_tuners.down_differential) *
+                        policy->cur) {
+                unsigned int freq_next;
+                freq_next = load_freq / (od_tuners.up_threshold -
+                                od_tuners.down_differential);
+
+                /* No longer fully busy, reset rate_mult */
+                dbs_info->rate_mult = 1;
+
+                if (freq_next < policy->min)
+                        freq_next = policy->min;
+
+                if (!od_tuners.powersave_bias) {
+                        __cpufreq_driver_target(policy, freq_next,
+                                        CPUFREQ_RELATION_L);
+                } else {
+                        int freq = powersave_bias_target(policy, freq_next,
+                                        CPUFREQ_RELATION_L);
+                        __cpufreq_driver_target(policy, freq,
+                                        CPUFREQ_RELATION_L);
+                }
+        }
 }
 
-define_one_global_ro(sampling_rate_min);
+static void od_dbs_timer(struct work_struct *work)
+{
+        struct od_cpu_dbs_info_s *dbs_info =
+                container_of(work, struct od_cpu_dbs_info_s, cdbs.work.work);
+        unsigned int cpu = dbs_info->cdbs.cpu;
+        int delay, sample_type = dbs_info->sample_type;
+
+        mutex_lock(&dbs_info->cdbs.timer_mutex);
+
+        /* Common NORMAL_SAMPLE setup */
+        dbs_info->sample_type = OD_NORMAL_SAMPLE;
+        if (sample_type == OD_SUB_SAMPLE) {
+                delay = dbs_info->freq_lo_jiffies;
+                __cpufreq_driver_target(dbs_info->cdbs.cur_policy,
+                        dbs_info->freq_lo, CPUFREQ_RELATION_H);
+        } else {
+                dbs_check_cpu(&od_dbs_data, cpu);
+                if (dbs_info->freq_lo) {
+                        /* Setup timer for SUB_SAMPLE */
+                        dbs_info->sample_type = OD_SUB_SAMPLE;
+                        delay = dbs_info->freq_hi_jiffies;
+                } else {
+                        delay = delay_for_sampling_rate(od_tuners.sampling_rate
+                                                * dbs_info->rate_mult);
+                }
+        }
+
+        schedule_delayed_work_on(cpu, &dbs_info->cdbs.work, delay);
+        mutex_unlock(&dbs_info->cdbs.timer_mutex);
+}
+
+/************************** sysfs interface ************************/
 
-/* cpufreq_ondemand Governor Tunables */
-#define show_one(file_name, object)                                     \
-static ssize_t show_##file_name                                         \
-(struct kobject *kobj, struct attribute *attr, char *buf)              \
-{                                                                       \
-        return sprintf(buf, "%u\n", dbs_tuners_ins.object);             \
+static ssize_t show_sampling_rate_min(struct kobject *kobj,
+                                      struct attribute *attr, char *buf)
+{
+        return sprintf(buf, "%u\n", od_dbs_data.min_sampling_rate);
 }
-show_one(sampling_rate, sampling_rate);
-show_one(io_is_busy, io_is_busy);
-show_one(up_threshold, up_threshold);
-show_one(sampling_down_factor, sampling_down_factor);
-show_one(ignore_nice_load, ignore_nice);
-show_one(powersave_bias, powersave_bias);
 
 /**
  * update_sampling_rate - update sampling rate effective immediately if needed.
  * @new_rate: new sampling rate
  *
  * If new rate is smaller than the old, simply updaing
- * dbs_tuners_int.sampling_rate might not be appropriate. For example,
- * if the original sampling_rate was 1 second and the requested new sampling
- * rate is 10 ms because the user needs immediate reaction from ondemand
- * governor, but not sure if higher frequency will be required or not,
- * then, the governor may change the sampling rate too late; up to 1 second
- * later. Thus, if we are reducing the sampling rate, we need to make the
- * new value effective immediately.
+ * dbs_tuners_int.sampling_rate might not be appropriate. For example, if the
+ * original sampling_rate was 1 second and the requested new sampling rate is 10
+ * ms because the user needs immediate reaction from ondemand governor, but not
+ * sure if higher frequency will be required or not, then, the governor may
+ * change the sampling rate too late; up to 1 second later. Thus, if we are
+ * reducing the sampling rate, we need to make the new value effective
+ * immediately.
  */
 static void update_sampling_rate(unsigned int new_rate)
 {
         int cpu;
 
-        dbs_tuners_ins.sampling_rate = new_rate
-                                     = max(new_rate, min_sampling_rate);
+        od_tuners.sampling_rate = new_rate = max(new_rate,
+                        od_dbs_data.min_sampling_rate);
 
         for_each_online_cpu(cpu) {
                 struct cpufreq_policy *policy;
-                struct cpu_dbs_info_s *dbs_info;
+                struct od_cpu_dbs_info_s *dbs_info;
                 unsigned long next_sampling, appointed_at;
 
                 policy = cpufreq_cpu_get(cpu);
                 if (!policy)
                         continue;
+                if (policy->governor != &cpufreq_gov_ondemand) {
+                        cpufreq_cpu_put(policy);
+                        continue;
+                }
                 dbs_info = &per_cpu(od_cpu_dbs_info, policy->cpu);
                 cpufreq_cpu_put(policy);
 
-                mutex_lock(&dbs_info->timer_mutex);
+                mutex_lock(&dbs_info->cdbs.timer_mutex);
 
-                if (!delayed_work_pending(&dbs_info->work)) {
-                        mutex_unlock(&dbs_info->timer_mutex);
+                if (!delayed_work_pending(&dbs_info->cdbs.work)) {
+                        mutex_unlock(&dbs_info->cdbs.timer_mutex);
                         continue;
                 }
 
-                next_sampling  = jiffies + usecs_to_jiffies(new_rate);
-                appointed_at = dbs_info->work.timer.expires;
-
+                next_sampling = jiffies + usecs_to_jiffies(new_rate);
+                appointed_at = dbs_info->cdbs.work.timer.expires;
 
                 if (time_before(next_sampling, appointed_at)) {
 
-                        mutex_unlock(&dbs_info->timer_mutex);
-                        cancel_delayed_work_sync(&dbs_info->work);
-                        mutex_lock(&dbs_info->timer_mutex);
+                        mutex_unlock(&dbs_info->cdbs.timer_mutex);
+                        cancel_delayed_work_sync(&dbs_info->cdbs.work);
+                        mutex_lock(&dbs_info->cdbs.timer_mutex);
 
-                        schedule_delayed_work_on(dbs_info->cpu, &dbs_info->work,
-                                                 usecs_to_jiffies(new_rate));
+                        schedule_delayed_work_on(dbs_info->cdbs.cpu,
+                                        &dbs_info->cdbs.work,
+                                        usecs_to_jiffies(new_rate));
 
                 }
-                mutex_unlock(&dbs_info->timer_mutex);
+                mutex_unlock(&dbs_info->cdbs.timer_mutex);
         }
 }
 
@@ -334,7 +336,7 @@ static ssize_t store_io_is_busy(struct kobject *a, struct attribute *b,
         ret = sscanf(buf, "%u", &input);
         if (ret != 1)
                 return -EINVAL;
-        dbs_tuners_ins.io_is_busy = !!input;
+        od_tuners.io_is_busy = !!input;
         return count;
 }
 
@@ -349,7 +351,7 @@ static ssize_t store_up_threshold(struct kobject *a, struct attribute *b,
                         input < MIN_FREQUENCY_UP_THRESHOLD) {
                 return -EINVAL;
         }
-        dbs_tuners_ins.up_threshold = input;
+        od_tuners.up_threshold = input;
         return count;
 }
 
@@ -362,12 +364,12 @@ static ssize_t store_sampling_down_factor(struct kobject *a,
 
         if (ret != 1 || input > MAX_SAMPLING_DOWN_FACTOR || input < 1)
                 return -EINVAL;
-        dbs_tuners_ins.sampling_down_factor = input;
+        od_tuners.sampling_down_factor = input;
 
         /* Reset down sampling multiplier in case it was active */
         for_each_online_cpu(j) {
-                struct cpu_dbs_info_s *dbs_info;
-                dbs_info = &per_cpu(od_cpu_dbs_info, j);
+                struct od_cpu_dbs_info_s *dbs_info = &per_cpu(od_cpu_dbs_info,
+                                j);
                 dbs_info->rate_mult = 1;
         }
         return count;
@@ -388,19 +390,20 @@ static ssize_t store_ignore_nice_load(struct kobject *a, struct attribute *b,
         if (input > 1)
                 input = 1;
 
-        if (input == dbs_tuners_ins.ignore_nice) { /* nothing to do */
+        if (input == od_tuners.ignore_nice) { /* nothing to do */
                 return count;
         }
-        dbs_tuners_ins.ignore_nice = input;
+        od_tuners.ignore_nice = input;
 
         /* we need to re-evaluate prev_cpu_idle */
         for_each_online_cpu(j) {
-                struct cpu_dbs_info_s *dbs_info;
+                struct od_cpu_dbs_info_s *dbs_info;
                 dbs_info = &per_cpu(od_cpu_dbs_info, j);
-                dbs_info->prev_cpu_idle = get_cpu_idle_time(j,
-                                                &dbs_info->prev_cpu_wall);
-                if (dbs_tuners_ins.ignore_nice)
-                        dbs_info->prev_cpu_nice = kcpustat_cpu(j).cpustat[CPUTIME_NICE];
+                dbs_info->cdbs.prev_cpu_idle = get_cpu_idle_time(j,
+                                                &dbs_info->cdbs.prev_cpu_wall);
+                if (od_tuners.ignore_nice)
+                        dbs_info->cdbs.prev_cpu_nice =
+                                kcpustat_cpu(j).cpustat[CPUTIME_NICE];
 
         }
         return count;
@@ -419,17 +422,25 @@ static ssize_t store_powersave_bias(struct kobject *a, struct attribute *b,
         if (input > 1000)
                 input = 1000;
 
-        dbs_tuners_ins.powersave_bias = input;
+        od_tuners.powersave_bias = input;
         ondemand_powersave_bias_init();
         return count;
 }
 
+show_one(od, sampling_rate, sampling_rate);
+show_one(od, io_is_busy, io_is_busy);
+show_one(od, up_threshold, up_threshold);
+show_one(od, sampling_down_factor, sampling_down_factor);
+show_one(od, ignore_nice_load, ignore_nice);
+show_one(od, powersave_bias, powersave_bias);
+
 define_one_global_rw(sampling_rate);
 define_one_global_rw(io_is_busy);
 define_one_global_rw(up_threshold);
 define_one_global_rw(sampling_down_factor);
 define_one_global_rw(ignore_nice_load);
 define_one_global_rw(powersave_bias);
+define_one_global_ro(sampling_rate_min);
 
 static struct attribute *dbs_attributes[] = {
         &sampling_rate_min.attr,
@@ -442,354 +453,71 @@ static struct attribute *dbs_attributes[] = {
         NULL
 };
 
-static struct attribute_group dbs_attr_group = {
+static struct attribute_group od_attr_group = {
         .attrs = dbs_attributes,
         .name = "ondemand",
 };
 
 /************************** sysfs end ************************/
 
-static void dbs_freq_increase(struct cpufreq_policy *p, unsigned int freq)
-{
-        if (dbs_tuners_ins.powersave_bias)
-                freq = powersave_bias_target(p, freq, CPUFREQ_RELATION_H);
-        else if (p->cur == p->max)
-                return;
-
-        __cpufreq_driver_target(p, freq, dbs_tuners_ins.powersave_bias ?
-                        CPUFREQ_RELATION_L : CPUFREQ_RELATION_H);
-}
-
-static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info)
-{
-        unsigned int max_load_freq;
-
-        struct cpufreq_policy *policy;
-        unsigned int j;
-
-        this_dbs_info->freq_lo = 0;
-        policy = this_dbs_info->cur_policy;
-
-        /*
-         * Every sampling_rate, we check, if current idle time is less
-         * than 20% (default), then we try to increase frequency
-         * 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.
-         *
-         * Any frequency increase takes it to the maximum frequency.
-         * Frequency reduction happens at minimum steps of
-         * 5% (default) of current frequency
-         */
-
-        /* Get Absolute Load - in terms of freq */
-        max_load_freq = 0;
-
-        for_each_cpu(j, policy->cpus) {
-                struct cpu_dbs_info_s *j_dbs_info;
-                cputime64_t cur_wall_time, cur_idle_time, cur_iowait_time;
-                unsigned int idle_time, wall_time, iowait_time;
-                unsigned int load, load_freq;
-                int freq_avg;
-
-                j_dbs_info = &per_cpu(od_cpu_dbs_info, j);
-
-                cur_idle_time = get_cpu_idle_time(j, &cur_wall_time);
-                cur_iowait_time = get_cpu_iowait_time(j, &cur_wall_time);
-
-                wall_time = (unsigned int)
-                        (cur_wall_time - j_dbs_info->prev_cpu_wall);
-                j_dbs_info->prev_cpu_wall = cur_wall_time;
-
-                idle_time = (unsigned int)
-                        (cur_idle_time - j_dbs_info->prev_cpu_idle);
-                j_dbs_info->prev_cpu_idle = cur_idle_time;
-
-                iowait_time = (unsigned int)
-                        (cur_iowait_time - j_dbs_info->prev_cpu_iowait);
-                j_dbs_info->prev_cpu_iowait = cur_iowait_time;
-
-                if (dbs_tuners_ins.ignore_nice) {
-                        u64 cur_nice;
-                        unsigned long cur_nice_jiffies;
-
-                        cur_nice = kcpustat_cpu(j).cpustat[CPUTIME_NICE] -
-                                         j_dbs_info->prev_cpu_nice;
-                        /*
-                         * Assumption: nice time between sampling periods will
-                         * be less than 2^32 jiffies for 32 bit sys
-                         */
-                        cur_nice_jiffies = (unsigned long)
-                                        cputime64_to_jiffies64(cur_nice);
-
-                        j_dbs_info->prev_cpu_nice = kcpustat_cpu(j).cpustat[CPUTIME_NICE];
-                        idle_time += jiffies_to_usecs(cur_nice_jiffies);
-                }
-
-                /*
-                 * For the purpose of ondemand, waiting for disk IO is an
-                 * indication that you're performance critical, and not that
-                 * the system is actually idle. So subtract the iowait time
-                 * from the cpu idle time.
-                 */
-
-                if (dbs_tuners_ins.io_is_busy && idle_time >= iowait_time)
-                        idle_time -= iowait_time;
-
-                if (unlikely(!wall_time || wall_time < idle_time))
-                        continue;
-
-                load = 100 * (wall_time - idle_time) / wall_time;
-
-                freq_avg = __cpufreq_driver_getavg(policy, j);
-                if (freq_avg <= 0)
-                        freq_avg = policy->cur;
-
-                load_freq = load * freq_avg;
-                if (load_freq > max_load_freq)
-                        max_load_freq = load_freq;
-        }
+define_get_cpu_dbs_routines(od_cpu_dbs_info);
 
-        /* Check for frequency increase */
-        if (max_load_freq > dbs_tuners_ins.up_threshold * policy->cur) {
-                /* If switching to max speed, apply sampling_down_factor */
-                if (policy->cur < policy->max)
-                        this_dbs_info->rate_mult =
-                                dbs_tuners_ins.sampling_down_factor;
-                dbs_freq_increase(policy, policy->max);
-                return;
-        }
-
-        /* Check for frequency decrease */
-        /* if we cannot reduce the frequency anymore, break out early */
-        if (policy->cur == policy->min)
-                return;
-
-        /*
-         * 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.
-         */
-        if (max_load_freq <
-            (dbs_tuners_ins.up_threshold - dbs_tuners_ins.down_differential) *
-             policy->cur) {
-                unsigned int freq_next;
-                freq_next = max_load_freq /
-                                (dbs_tuners_ins.up_threshold -
-                                 dbs_tuners_ins.down_differential);
-
-                /* No longer fully busy, reset rate_mult */
-                this_dbs_info->rate_mult = 1;
-
-                if (freq_next < policy->min)
-                        freq_next = policy->min;
-
-                if (!dbs_tuners_ins.powersave_bias) {
-                        __cpufreq_driver_target(policy, freq_next,
-                                        CPUFREQ_RELATION_L);
-                } else {
-                        int freq = powersave_bias_target(policy, freq_next,
-                                        CPUFREQ_RELATION_L);
-                        __cpufreq_driver_target(policy, freq,
-                                CPUFREQ_RELATION_L);
-                }
-        }
-}
-
-static void do_dbs_timer(struct work_struct *work)
-{
-        struct cpu_dbs_info_s *dbs_info =
-                container_of(work, struct cpu_dbs_info_s, work.work);
-        unsigned int cpu = dbs_info->cpu;
-        int sample_type = dbs_info->sample_type;
-
-        int delay;
-
-        mutex_lock(&dbs_info->timer_mutex);
-
-        /* Common NORMAL_SAMPLE setup */
-        dbs_info->sample_type = DBS_NORMAL_SAMPLE;
-        if (!dbs_tuners_ins.powersave_bias ||
-            sample_type == DBS_NORMAL_SAMPLE) {
-                dbs_check_cpu(dbs_info);
-                if (dbs_info->freq_lo) {
-                        /* Setup timer for SUB_SAMPLE */
-                        dbs_info->sample_type = DBS_SUB_SAMPLE;
-                        delay = dbs_info->freq_hi_jiffies;
-                } else {
-                        /* We want all CPUs to do sampling nearly on
-                         * same jiffy
-                         */
-                        delay = usecs_to_jiffies(dbs_tuners_ins.sampling_rate
-                                * dbs_info->rate_mult);
-
-                        if (num_online_cpus() > 1)
-                                delay -= jiffies % delay;
-                }
-        } else {
-                __cpufreq_driver_target(dbs_info->cur_policy,
-                        dbs_info->freq_lo, CPUFREQ_RELATION_H);
-                delay = dbs_info->freq_lo_jiffies;
-        }
-        schedule_delayed_work_on(cpu, &dbs_info->work, delay);
-        mutex_unlock(&dbs_info->timer_mutex);
-}
-
-static inline void dbs_timer_init(struct cpu_dbs_info_s *dbs_info)
-{
-        /* We want all CPUs to do sampling nearly on same jiffy */
-        int delay = usecs_to_jiffies(dbs_tuners_ins.sampling_rate);
-
-        if (num_online_cpus() > 1)
-                delay -= jiffies % delay;
+static struct od_ops od_ops = {
+        .io_busy = should_io_be_busy,
+        .powersave_bias_init_cpu = ondemand_powersave_bias_init_cpu,
+        .powersave_bias_target = powersave_bias_target,
+        .freq_increase = dbs_freq_increase,
+};
 
-        dbs_info->sample_type = DBS_NORMAL_SAMPLE;
-        INIT_DEFERRABLE_WORK(&dbs_info->work, do_dbs_timer);
-        schedule_delayed_work_on(dbs_info->cpu, &dbs_info->work, delay);
-}
+static struct dbs_data od_dbs_data = {
+        .governor = GOV_ONDEMAND,
+        .attr_group = &od_attr_group,
+        .tuners = &od_tuners,
+        .get_cpu_cdbs = get_cpu_cdbs,
+        .get_cpu_dbs_info_s = get_cpu_dbs_info_s,
+        .gov_dbs_timer = od_dbs_timer,
+        .gov_check_cpu = od_check_cpu,
+        .gov_ops = &od_ops,
+};
 
-static inline void dbs_timer_exit(struct cpu_dbs_info_s *dbs_info)
+static int od_cpufreq_governor_dbs(struct cpufreq_policy *policy,
+                unsigned int event)
 {
-        cancel_delayed_work_sync(&dbs_info->work);
+        return cpufreq_governor_dbs(&od_dbs_data, policy, event);
 }
 
-/*
- * Not all CPUs want IO time to be accounted as busy; this dependson how
- * efficient idling at a higher frequency/voltage is.
- * Pavel Machek says this is not so for various generations of AMD and old
- * Intel systems.
- * Mike Chan (androidlcom) calis this is also not true for ARM.
- * Because of this, whitelist specific known (series) of CPUs by default, and
- * leave all others up to the user.
- */
-static int should_io_be_busy(void)
-{
-#if defined(CONFIG_X86)
-        /*
-         * For Intel, Core 2 (model 15) andl later have an efficient idle.
-         */
-        if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL &&
-            boot_cpu_data.x86 == 6 &&
-            boot_cpu_data.x86_model >= 15)
-                return 1;
+#ifndef CONFIG_CPU_FREQ_DEFAULT_GOV_ONDEMAND
+static
 #endif
-        return 0;
-}
-
-static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
-                                   unsigned int event)
-{
-        unsigned int cpu = policy->cpu;
-        struct cpu_dbs_info_s *this_dbs_info;
-        unsigned int j;
-        int rc;
-
-        this_dbs_info = &per_cpu(od_cpu_dbs_info, cpu);
-
-        switch (event) {
-        case CPUFREQ_GOV_START:
-                if ((!cpu_online(cpu)) || (!policy->cur))
-                        return -EINVAL;
-
-                mutex_lock(&dbs_mutex);
-
-                dbs_enable++;
-                for_each_cpu(j, policy->cpus) {
-                        struct cpu_dbs_info_s *j_dbs_info;
-                        j_dbs_info = &per_cpu(od_cpu_dbs_info, j);
-                        j_dbs_info->cur_policy = policy;
-
-                        j_dbs_info->prev_cpu_idle = get_cpu_idle_time(j,
-                                                &j_dbs_info->prev_cpu_wall);
-                        if (dbs_tuners_ins.ignore_nice)
-                                j_dbs_info->prev_cpu_nice =
-                                                kcpustat_cpu(j).cpustat[CPUTIME_NICE];
-                }
-                this_dbs_info->cpu = cpu;
-                this_dbs_info->rate_mult = 1;
-                ondemand_powersave_bias_init_cpu(cpu);
-                /*
-                 * Start the timerschedule work, when this governor
-                 * is used for first time
-                 */
-                if (dbs_enable == 1) {
-                        unsigned int latency;
-
-                        rc = sysfs_create_group(cpufreq_global_kobject,
-                                                &dbs_attr_group);
-                        if (rc) {
-                                mutex_unlock(&dbs_mutex);
-                                return rc;
-                        }
-
-                        /* policy latency is in nS. Convert it to uS first */
-                        latency = policy->cpuinfo.transition_latency / 1000;
-                        if (latency == 0)
-                                latency = 1;
-                        /* Bring kernel and HW constraints together */
-                        min_sampling_rate = max(min_sampling_rate,
-                                        MIN_LATENCY_MULTIPLIER * latency);
-                        dbs_tuners_ins.sampling_rate =
-                                max(min_sampling_rate,
-                                    latency * LATENCY_MULTIPLIER);
-                        dbs_tuners_ins.io_is_busy = should_io_be_busy();
-                }
-                mutex_unlock(&dbs_mutex);
-
-                mutex_init(&this_dbs_info->timer_mutex);
-                dbs_timer_init(this_dbs_info);
-                break;
-
-        case CPUFREQ_GOV_STOP:
-                dbs_timer_exit(this_dbs_info);
-
-                mutex_lock(&dbs_mutex);
-                mutex_destroy(&this_dbs_info->timer_mutex);
-                dbs_enable--;
-                mutex_unlock(&dbs_mutex);
-                if (!dbs_enable)
-                        sysfs_remove_group(cpufreq_global_kobject,
-                                           &dbs_attr_group);
-
-                break;
-
-        case CPUFREQ_GOV_LIMITS:
-                mutex_lock(&this_dbs_info->timer_mutex);
-                if (policy->max < this_dbs_info->cur_policy->cur)
-                        __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);
-                dbs_check_cpu(this_dbs_info);
-                mutex_unlock(&this_dbs_info->timer_mutex);
-                break;
-        }
-        return 0;
-}
+struct cpufreq_governor cpufreq_gov_ondemand = {
+        .name                   = "ondemand",
+        .governor               = od_cpufreq_governor_dbs,
+        .max_transition_latency = TRANSITION_LATENCY_LIMIT,
+        .owner                  = THIS_MODULE,
+};
 
 static int __init cpufreq_gov_dbs_init(void)
 {
         u64 idle_time;
         int cpu = get_cpu();
 
+        mutex_init(&od_dbs_data.mutex);
         idle_time = get_cpu_idle_time_us(cpu, NULL);
         put_cpu();
         if (idle_time != -1ULL) {
                 /* Idle micro accounting is supported. Use finer thresholds */
-                dbs_tuners_ins.up_threshold = MICRO_FREQUENCY_UP_THRESHOLD;
-                dbs_tuners_ins.down_differential =
-                                        MICRO_FREQUENCY_DOWN_DIFFERENTIAL;
+                od_tuners.up_threshold = MICRO_FREQUENCY_UP_THRESHOLD;
+                od_tuners.down_differential = MICRO_FREQUENCY_DOWN_DIFFERENTIAL;
                 /*
                  * In nohz/micro accounting case we set the minimum frequency
                  * not depending on HZ, but fixed (very low). The deferred
                  * timer might skip some samples if idle/sleeping as needed.
                 */
-                min_sampling_rate = MICRO_FREQUENCY_MIN_SAMPLE_RATE;
+                od_dbs_data.min_sampling_rate = MICRO_FREQUENCY_MIN_SAMPLE_RATE;
         } else {
                 /* For correct statistics, we need 10 ticks for each measure */
-                min_sampling_rate =
-                        MIN_SAMPLING_RATE_RATIO * jiffies_to_usecs(10);
+                od_dbs_data.min_sampling_rate = MIN_SAMPLING_RATE_RATIO *
+                        jiffies_to_usecs(10);
         }
 
         return cpufreq_register_governor(&cpufreq_gov_ondemand);
@@ -800,7 +528,6 @@ static void __exit cpufreq_gov_dbs_exit(void)
         cpufreq_unregister_governor(&cpufreq_gov_ondemand);
 }
 
-
 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 "