[CPUFREQ] make ondemand governor aware of CPU domains
The following patch makes ondemand governor aware of policy->cpus.
policy->cpus mask lets multiple cpu use the same policy (useful where cpus
share the frequency state), a recent change in cpufreq core.
Now ondemand governor looks at all cpus in policy->cpus and takes its
frequency increase/decrease decisions based on most lightly loaded cpu in
the group. This patch will not affect systems where policy->cpus contain
only one cpu.
Signed-off-by:: "Venkatesh Pallipadi" <venkatesh.pallipadi@intel.com>
Signed-off-by: Dominik Brodowski <linux@brodo.de>
diff -ruN linux-original/drivers/cpufreq/cpufreq_ondemand.c linux/drivers/cpufreq/cpufreq_ondemand.c
--- linux-original/drivers/cpufreq/cpufreq_ondemand.c 2004-11-27 12:06:53.600614160 +0100
+++ linux/drivers/cpufreq/cpufreq_ondemand.c 2004-11-27 12:30:34.153657272 +0100
@@ -229,10 +229,14 @@
static int down_skip[NR_CPUS];
struct cpu_dbs_info_s *this_dbs_info;
+ 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;
/*
* The default safe range is 20% to 80%
* Every sampling_rate, we check
@@ -246,12 +250,33 @@
* Frequency reduction happens at minimum steps of
* 5% of max_frequency
*/
+
/* Check for frequency increase */
total_idle_ticks = kstat_cpu(cpu).cpustat.idle +
kstat_cpu(cpu).cpustat.iowait;
idle_ticks = total_idle_ticks -
this_dbs_info->prev_cpu_idle_up;
this_dbs_info->prev_cpu_idle_up = total_idle_ticks;
+
+
+ for_each_cpu_mask(j, policy->cpus) {
+ unsigned int tmp_idle_ticks;
+ struct cpu_dbs_info_s *j_dbs_info;
+
+ if (j == cpu)
+ continue;
+
+ j_dbs_info = &per_cpu(cpu_dbs_info, j);
+ /* Check for frequency increase */
+ total_idle_ticks = kstat_cpu(j).cpustat.idle +
+ kstat_cpu(j).cpustat.iowait;
+ tmp_idle_ticks = total_idle_ticks -
+ j_dbs_info->prev_cpu_idle_up;
+ j_dbs_info->prev_cpu_idle_up = total_idle_ticks;
+
+ if (tmp_idle_ticks < idle_ticks)
+ idle_ticks = tmp_idle_ticks;
+ }
/* Scale idle ticks by 100 and compare with up and down ticks */
idle_ticks *= 100;
@@ -259,8 +284,7 @@
sampling_rate_in_HZ(dbs_tuners_ins.sampling_rate);
if (idle_ticks < up_idle_ticks) {
- __cpufreq_driver_target(this_dbs_info->cur_policy,
- this_dbs_info->cur_policy->max,
+ __cpufreq_driver_target(policy, policy->max,
CPUFREQ_RELATION_H);
down_skip[cpu] = 0;
this_dbs_info->prev_cpu_idle_down = total_idle_ticks;
@@ -272,12 +296,34 @@
if (down_skip[cpu] < dbs_tuners_ins.sampling_down_factor)
return;
+ total_idle_ticks = kstat_cpu(cpu).cpustat.idle +
+ kstat_cpu(cpu).cpustat.iowait;
idle_ticks = total_idle_ticks -
this_dbs_info->prev_cpu_idle_down;
+ this_dbs_info->prev_cpu_idle_down = total_idle_ticks;
+
+ for_each_cpu_mask(j, policy->cpus) {
+ unsigned int tmp_idle_ticks;
+ struct cpu_dbs_info_s *j_dbs_info;
+
+ if (j == cpu)
+ continue;
+
+ j_dbs_info = &per_cpu(cpu_dbs_info, j);
+ /* Check for frequency increase */
+ total_idle_ticks = kstat_cpu(j).cpustat.idle +
+ kstat_cpu(j).cpustat.iowait;
+ 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;
+ }
+
/* Scale idle ticks by 100 and compare with up and down ticks */
idle_ticks *= 100;
down_skip[cpu] = 0;
- this_dbs_info->prev_cpu_idle_down = total_idle_ticks;
freq_down_sampling_rate = dbs_tuners_ins.sampling_rate *
dbs_tuners_ins.sampling_down_factor;
@@ -285,14 +331,14 @@
sampling_rate_in_HZ(freq_down_sampling_rate);
if (idle_ticks > down_idle_ticks ) {
- freq_down_step = (5 * this_dbs_info->cur_policy->max) / 100;
+ freq_down_step = (5 * policy->max) / 100;
/* max freq cannot be less than 100. But who knows.... */
if (unlikely(freq_down_step == 0))
freq_down_step = 5;
- __cpufreq_driver_target(this_dbs_info->cur_policy,
- this_dbs_info->cur_policy->cur - freq_down_step,
+ __cpufreq_driver_target(policy,
+ policy->cur - freq_down_step,
CPUFREQ_RELATION_H);
return;
}
@@ -328,6 +374,7 @@
{
unsigned int cpu = policy->cpu;
struct cpu_dbs_info_s *this_dbs_info;
+ unsigned int j;
this_dbs_info = &per_cpu(cpu_dbs_info, cpu);
@@ -344,14 +391,18 @@
break;
down(&dbs_sem);
- this_dbs_info->cur_policy = policy;
+ 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;
- this_dbs_info->prev_cpu_idle_up =
- kstat_cpu(cpu).cpustat.idle +
- kstat_cpu(cpu).cpustat.iowait;
- this_dbs_info->prev_cpu_idle_down =
- kstat_cpu(cpu).cpustat.idle +
- kstat_cpu(cpu).cpustat.iowait;
+ j_dbs_info->prev_cpu_idle_up =
+ kstat_cpu(j).cpustat.idle +
+ kstat_cpu(j).cpustat.iowait;
+ j_dbs_info->prev_cpu_idle_down =
+ kstat_cpu(j).cpustat.idle +
+ kstat_cpu(j).cpustat.iowait;
+ }
this_dbs_info->enable = 1;
sysfs_create_group(&policy->kobj, &dbs_attr_group);
dbs_enable++;