[PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Srivatsa S. Bhat]

Cpufreq governors like the ondemand governor calculate the load on the CPU
periodically by employing deferrable timers. A deferrable timer won't fire
if the CPU is completely idle (and there are no other timers to be run), in
order to avoid unnecessary wakeups and thus save CPU power.

However, the load calculation logic is agnostic to all this, and this can
lead to the problem described below.


Time (ms)               CPU 1

100                Task-A running

110                Governor's timer fires, finds load as 100% in the last
                   10ms interval and increases the CPU frequency.

110.5              Task-A running

120		   Governor's timer fires, finds load as 100% in the last
		   10ms interval and increases the CPU frequency.

125		   Task-A went to sleep. With nothing else to do, CPU 1
		   went completely idle.

200		   Task-A woke up and started running again.

200.5		   Governor's deferred timer (which was originally programmed
		   to fire at time 130) fires now. It calculates load for the
		   time period 120 to 200.5, and finds the load is almost zero.
		   Hence it decreases the CPU frequency to the minimum.

210		   Governor's timer fires, finds load as 100% in the last
		   10ms interval and increases the CPU frequency.


So, after the workload woke up and started running, the frequency was suddenly
dropped to absolute minimum, and after that, there was an unnecessary delay of
10ms (sampling period) to increase the CPU frequency back to a reasonable value.
And this pattern repeats for every wake-up-from-cpu-idle for that workload.
This can be quite undesirable for latency- or response-time sensitive bursty
workloads. So we need to fix the governor's logic to detect such wake-up-from-
cpu-idle scenarios and start the workload at a reasonably high CPU frequency.

One extreme solution would be to fake a load of 100% in such scenarios. But
that might lead to undesirable side-effects such as frequency spikes (which
might also need voltage changes) especially if the previous frequency happened
to be very low.

We just want to avoid the stupidity of dropping down the frequency to a minimum
and then enduring a needless (and long) delay before ramping it up back again.
So, let us simply carry forward the previous load - that is, let us just pretend
that the 'load' for the current time-window is the same as the load for the
previous window. That way, the frequency and voltage will continue to be set
to whatever values they were set at previously. This means that bursty workloads
will get a chance to influence the CPU frequency at which they wake up from
cpu-idle, based on their past execution history. Thus, they might be able to
avoid suffering from slow wakeups and long response-times.

[ The right way to solve this problem is to teach the CPU frequency governors
to track load on a per-task basis, not a per-CPU basis, and set the appropriate
frequency on whichever CPU the task executes. But that involves redesigning
the cpufreq subsystem, so this patch should make the situation bearable until
then. ]

Experimental results:
====================

I ran a modified version of ebizzy (called 'sleeping-ebizzy') that sleeps in
between its execution such that its total utilization can be a user-defined
value, say 10% or 20% (higher the utilization specified, lesser the amount of
sleeps injected). This ebizzy was run with a single-thread, tied to CPU 8.

Behavior observed with tracing (sample taken from 40% utilization runs):
------------------------------------------------------------------------

Without patch:
~~~~~~~~~~~~~~
kworker/8:2-12137  416.335742: cpu_frequency: state=2061000 cpu_id=8
kworker/8:2-12137  416.335744: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
      <...>-40753  416.345741: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
kworker/8:2-12137  416.345744: cpu_frequency: state=4123000 cpu_id=8
kworker/8:2-12137  416.345746: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
      <...>-40753  416.355738: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
<snip>  ---------------------------------------------------------------------  <snip>
      <...>-40753  416.402202: sched_switch: prev_comm=ebizzy ==> next_comm=swapper/8
     <idle>-0      416.502130: sched_switch: prev_comm=swapper/8 ==> next_comm=ebizzy
      <...>-40753  416.505738: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
kworker/8:2-12137  416.505739: cpu_frequency: state=2061000 cpu_id=8
kworker/8:2-12137  416.505741: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
      <...>-40753  416.515739: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
kworker/8:2-12137  416.515742: cpu_frequency: state=4123000 cpu_id=8
kworker/8:2-12137  416.515744: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy

Observation: Ebizzy went idle at 416.402202, and started running again at
416.502130. But cpufreq noticed the long idle period, and dropped the frequency
at 416.505739, only to increase it back again at 416.515742, realizing that the
workload is in-fact CPU bound. Thus ebizzy needlessly ran at the lowest frequency
for almost 13 milliseconds (almost 1 full sample period), and this pattern
repeats on every sleep-wakeup. This could hurt latency-sensitive workloads quite
a lot.

With patch:
~~~~~~~~~~~

kworker/8:2-29802  464.832535: cpu_frequency: state=2061000 cpu_id=8
<snip>  ---------------------------------------------------------------------  <snip>
kworker/8:2-29802  464.962538: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
      <...>-40738  464.972533: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
kworker/8:2-29802  464.972536: cpu_frequency: state=4123000 cpu_id=8
kworker/8:2-29802  464.972538: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
      <...>-40738  464.982531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
<snip>  ---------------------------------------------------------------------  <snip>
kworker/8:2-29802  465.022533: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
      <...>-40738  465.032531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
kworker/8:2-29802  465.032532: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
      <...>-40738  465.035797: sched_switch: prev_comm=ebizzy ==> next_comm=swapper/8
     <idle>-0      465.240178: sched_switch: prev_comm=swapper/8 ==> next_comm=ebizzy
      <...>-40738  465.242533: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
kworker/8:2-29802  465.242535: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
      <...>-40738  465.252531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2

Observation: Ebizzy went idle at 465.035797, and started running again at
465.240178. Since ebizzy was the only real workload running on this CPU,
cpufreq retained the frequency at 4.1Ghz throughout the run of ebizzy, no
matter how many times ebizzy slept and woke-up in-between. Thus, ebizzy
got the 10ms worth of 4.1 Ghz benefit during every sleep-wakeup (as compared
to the run without the patch) and this boost gave a modest improvement in total
throughput, as shown below.

Sleeping-ebizzy records-per-second:
-----------------------------------

Utilization  Without patch  With patch  Difference (Absolute and % values)
    10%         274767        277046        +  2279 (+0.829%)
    20%         543429        553484        + 10055 (+1.850%)
    40%        1090744       1107959        + 17215 (+1.578%)
    60%        1634908       1662018        + 27110 (+1.658%)

A rudimentary and somewhat approximately latency-sensitive workload such as
sleeping-ebizzy itself showed a consistent, noticeable performance improvement
with this patch. Hence, workloads that are truly latency-sensitive will benefit
quite a bit from this change. Moreover, this is an overall win-win since this
patch does not hurt power-savings at all (because, this patch does not reduce
the idle time or idle residency; and the high frequency of the CPU when it goes
to cpu-idle does not affect/hurt the power-savings of deep idle states).

Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
---

 drivers/cpufreq/cpufreq_conservative.c |    2 +-
 drivers/cpufreq/cpufreq_governor.c     |   32 +++++++++++++++++++++++++++++---
 drivers/cpufreq/cpufreq_governor.h     |    4 +++-
 drivers/cpufreq/cpufreq_ondemand.c     |    9 ++++++++-
 4 files changed, 41 insertions(+), 6 deletions(-)

diff --git a/drivers/cpufreq/cpufreq_conservative.c b/drivers/cpufreq/cpufreq_conservative.c
index 25a70d0..65c9905 100644
--- a/drivers/cpufreq/cpufreq_conservative.c
+++ b/drivers/cpufreq/cpufreq_conservative.c
@@ -118,7 +118,7 @@ static void cs_dbs_timer(struct work_struct *work)
 	if (!need_load_eval(&core_dbs_info->cdbs, cs_tuners->sampling_rate))
 		modify_all = false;
 	else
-		dbs_check_cpu(dbs_data, cpu);
+		dbs_check_cpu(dbs_data, cpu, cs_tuners->sampling_rate);
 
 	gov_queue_work(dbs_data, dbs_info->cdbs.cur_policy, delay, modify_all);
 	mutex_unlock(&core_dbs_info->cdbs.timer_mutex);
diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c
index e1c6433..910d472 100644
--- a/drivers/cpufreq/cpufreq_governor.c
+++ b/drivers/cpufreq/cpufreq_governor.c
@@ -30,7 +30,8 @@ static struct attribute_group *get_sysfs_attr(struct dbs_data *dbs_data)
 		return dbs_data->cdata->attr_group_gov_sys;
 }
 
-void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
+void dbs_check_cpu(struct dbs_data *dbs_data, int cpu,
+		   unsigned int sampling_rate)
 {
 	struct cpu_dbs_common_info *cdbs = dbs_data->cdata->get_cpu_cdbs(cpu);
 	struct od_dbs_tuners *od_tuners = dbs_data->tuners;
@@ -96,7 +97,28 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
 		if (unlikely(!wall_time || wall_time < idle_time))
 			continue;
 
-		load = 100 * (wall_time - idle_time) / wall_time;
+		/*
+		 * If the CPU had gone completely idle, and a task just woke up
+		 * on this CPU now, it would be unfair to calculate 'load' the
+		 * usual way for this elapsed time-window, because it will show
+		 * near-zero load, irrespective of how CPU intensive the new
+		 * task is. This is undesirable for latency-sensitive bursty
+		 * workloads.
+		 *
+		 * To avoid this, we reuse the 'load' from the previous
+		 * time-window and give this task a chance to start with a
+		 * reasonably high CPU frequency.
+		 *
+		 * Detecting this situation is easy: the governor's deferrable
+		 * timer would not have fired during CPU-idle periods. Hence
+		 * an unusually large 'wall_time' indicates this scenario.
+		 */
+		if (unlikely(wall_time > (2 * sampling_rate))) {
+			load = j_cdbs->prev_load;
+		} else {
+			load = 100 * (wall_time - idle_time) / wall_time;
+			j_cdbs->prev_load = load;
+		}
 
 		if (load > max_load)
 			max_load = load;
@@ -323,6 +345,10 @@ int cpufreq_governor_dbs(struct cpufreq_policy *policy,
 			j_cdbs->cur_policy = policy;
 			j_cdbs->prev_cpu_idle = get_cpu_idle_time(j,
 					       &j_cdbs->prev_cpu_wall, io_busy);
+			j_cdbs->prev_load = 100 * (j_cdbs->prev_cpu_wall -
+						   j_cdbs->prev_cpu_idle) /
+						   j_cdbs->prev_cpu_wall;
+
 			if (ignore_nice)
 				j_cdbs->prev_cpu_nice =
 					kcpustat_cpu(j).cpustat[CPUTIME_NICE];
@@ -378,7 +404,7 @@ int cpufreq_governor_dbs(struct cpufreq_policy *policy,
 		else if (policy->min > cpu_cdbs->cur_policy->cur)
 			__cpufreq_driver_target(cpu_cdbs->cur_policy,
 					policy->min, CPUFREQ_RELATION_L);
-		dbs_check_cpu(dbs_data, cpu);
+		dbs_check_cpu(dbs_data, cpu, sampling_rate);
 		mutex_unlock(&cpu_cdbs->timer_mutex);
 		mutex_unlock(&dbs_data->mutex);
 		break;
diff --git a/drivers/cpufreq/cpufreq_governor.h b/drivers/cpufreq/cpufreq_governor.h
index bfb9ae1..2fbf878 100644
--- a/drivers/cpufreq/cpufreq_governor.h
+++ b/drivers/cpufreq/cpufreq_governor.h
@@ -134,6 +134,7 @@ struct cpu_dbs_common_info {
 	u64 prev_cpu_idle;
 	u64 prev_cpu_wall;
 	u64 prev_cpu_nice;
+	unsigned int prev_load;
 	struct cpufreq_policy *cur_policy;
 	struct delayed_work work;
 	/*
@@ -259,7 +260,8 @@ static ssize_t show_sampling_rate_min_gov_pol				\
 
 extern struct mutex cpufreq_governor_lock;
 
-void dbs_check_cpu(struct dbs_data *dbs_data, int cpu);
+void dbs_check_cpu(struct dbs_data *dbs_data, int cpu,
+		   unsigned int sampling_rate);
 bool need_load_eval(struct cpu_dbs_common_info *cdbs,
 		unsigned int sampling_rate);
 int cpufreq_governor_dbs(struct cpufreq_policy *policy,
diff --git a/drivers/cpufreq/cpufreq_ondemand.c b/drivers/cpufreq/cpufreq_ondemand.c
index 18d4091..b1f054a 100644
--- a/drivers/cpufreq/cpufreq_ondemand.c
+++ b/drivers/cpufreq/cpufreq_ondemand.c
@@ -213,7 +213,14 @@ static void od_dbs_timer(struct work_struct *work)
 		__cpufreq_driver_target(core_dbs_info->cdbs.cur_policy,
 				core_dbs_info->freq_lo, CPUFREQ_RELATION_H);
 	} else {
-		dbs_check_cpu(dbs_data, cpu);
+		/*
+		 * Provide maximum delay as the sampling_rate hint to
+		 * dbs_check_cpu, to keep its wake-up-from-cpu-idle detection
+		 * logic a bit conservative.
+		 */
+		dbs_check_cpu(dbs_data, cpu,
+			od_tuners->sampling_rate * core_dbs_info->rate_mult);
+
 		if (core_dbs_info->freq_lo) {
 			/* Setup timer for SUB_SAMPLE */
 			core_dbs_info->sample_type = OD_SUB_SAMPLE;

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Rafael J. Wysocki]

Hi Srivatsa,

On Tuesday, May 27, 2014 02:23:38 AM Srivatsa S. Bhat wrote:
> Cpufreq governors like the ondemand governor calculate the load on the CPU
> periodically by employing deferrable timers. A deferrable timer won't fire
> if the CPU is completely idle (and there are no other timers to be run), in
> order to avoid unnecessary wakeups and thus save CPU power.
> 
> However, the load calculation logic is agnostic to all this, and this can
> lead to the problem described below.

This is subtle enough that I need some more time to chew on it, but since the
merge window is coming, I'm not sure when that's going to happen honestly.

Rafael

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Srivatsa S. Bhat]

On 05/27/2014 04:57 AM, Rafael J. Wysocki wrote:
> Hi Srivatsa,
> 
> On Tuesday, May 27, 2014 02:23:38 AM Srivatsa S. Bhat wrote:
>> Cpufreq governors like the ondemand governor calculate the load on the CPU
>> periodically by employing deferrable timers. A deferrable timer won't fire
>> if the CPU is completely idle (and there are no other timers to be run), in
>> order to avoid unnecessary wakeups and thus save CPU power.
>>
>> However, the load calculation logic is agnostic to all this, and this can
>> lead to the problem described below.
> 
> This is subtle enough that I need some more time to chew on it, but since the
> merge window is coming, I'm not sure when that's going to happen honestly.
>

Sure, I completely understand. Please take your time, no hurry!

Thank you very much!

Regards,
Srivatsa S. Bhat

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Gautham R Shenoy]

Hi,

On Tue, May 27, 2014 at 02:23:38AM +0530, Srivatsa S. Bhat wrote:

[..snip..]
> 
> Experimental results:
> ====================
> 
> I ran a modified version of ebizzy (called 'sleeping-ebizzy') that sleeps in
> between its execution such that its total utilization can be a user-defined
> value, say 10% or 20% (higher the utilization specified, lesser the amount of
> sleeps injected). This ebizzy was run with a single-thread, tied to CPU 8.
> 
> Behavior observed with tracing (sample taken from 40% utilization runs):
> ------------------------------------------------------------------------
> 
> Without patch:
> ~~~~~~~~~~~~~~
> kworker/8:2-12137  416.335742: cpu_frequency: state=2061000 cpu_id=8
> kworker/8:2-12137  416.335744: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>       <...>-40753  416.345741: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> kworker/8:2-12137  416.345744: cpu_frequency: state=4123000 cpu_id=8
> kworker/8:2-12137  416.345746: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>       <...>-40753  416.355738: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> <snip>  ---------------------------------------------------------------------  <snip>
>       <...>-40753  416.402202: sched_switch: prev_comm=ebizzy ==> next_comm=swapper/8
>      <idle>-0      416.502130: sched_switch: prev_comm=swapper/8 ==> next_comm=ebizzy
>       <...>-40753  416.505738: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> kworker/8:2-12137  416.505739: cpu_frequency: state=2061000 cpu_id=8
> kworker/8:2-12137  416.505741: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>       <...>-40753  416.515739: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> kworker/8:2-12137  416.515742: cpu_frequency: state=4123000 cpu_id=8
> kworker/8:2-12137  416.515744: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
> 
> Observation: Ebizzy went idle at 416.402202, and started running again at
> 416.502130. But cpufreq noticed the long idle period, and dropped the frequency
> at 416.505739, only to increase it back again at 416.515742, realizing that the
> workload is in-fact CPU bound. Thus ebizzy needlessly ran at the lowest frequency
> for almost 13 milliseconds (almost 1 full sample period), and this pattern
> repeats on every sleep-wakeup. This could hurt latency-sensitive workloads quite
> a lot.
> 
> With patch:
> ~~~~~~~~~~~
> 
> kworker/8:2-29802  464.832535: cpu_frequency: state=2061000 cpu_id=8
> <snip>  ---------------------------------------------------------------------  <snip>
> kworker/8:2-29802  464.962538: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>       <...>-40738  464.972533: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> kworker/8:2-29802  464.972536: cpu_frequency: state=4123000 cpu_id=8
> kworker/8:2-29802  464.972538: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>       <...>-40738  464.982531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> <snip>  ---------------------------------------------------------------------  <snip>
> kworker/8:2-29802  465.022533: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>       <...>-40738  465.032531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> kworker/8:2-29802  465.032532: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>       <...>-40738  465.035797: sched_switch: prev_comm=ebizzy ==> next_comm=swapper/8
>      <idle>-0      465.240178: sched_switch: prev_comm=swapper/8 ==> next_comm=ebizzy
>       <...>-40738  465.242533: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> kworker/8:2-29802  465.242535: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>       <...>-40738  465.252531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> 

Have the log entries emmitted by kworker/8 to report about the
cpu_frequency states been snipped out in the entries post the
"465.032531" mark ?


> Observation: Ebizzy went idle at 465.035797, and started running again at
> 465.240178. Since ebizzy was the only real workload running on this CPU,
> cpufreq retained the frequency at 4.1Ghz throughout the run of ebizzy, no
> matter how many times ebizzy slept and woke-up in-between. Thus, ebizzy
> got the 10ms worth of 4.1 Ghz benefit during every sleep-wakeup (as compared
> to the run without the patch) and this boost gave a modest improvement in total
> throughput, as shown below.
> 
> Sleeping-ebizzy records-per-second:
> -----------------------------------
> 
> Utilization  Without patch  With patch  Difference (Absolute and % values)
>     10%         274767        277046        +  2279 (+0.829%)
>     20%         543429        553484        + 10055 (+1.850%)
>     40%        1090744       1107959        + 17215 (+1.578%)
>     60%        1634908       1662018        + 27110 (+1.658%)
> 
> A rudimentary and somewhat approximately latency-sensitive workload such as
> sleeping-ebizzy itself showed a consistent, noticeable performance improvement
> with this patch. Hence, workloads that are truly latency-sensitive will benefit
> quite a bit from this change. Moreover, this is an overall win-win since this
> patch does not hurt power-savings at all (because, this patch does not reduce
> the idle time or idle residency; and the high frequency of the CPU when it goes
> to cpu-idle does not affect/hurt the power-savings of deep idle states).
> 
> Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
> ---
> 
>  drivers/cpufreq/cpufreq_conservative.c |    2 +-
>  drivers/cpufreq/cpufreq_governor.c     |   32 +++++++++++++++++++++++++++++---
>  drivers/cpufreq/cpufreq_governor.h     |    4 +++-
>  drivers/cpufreq/cpufreq_ondemand.c     |    9 ++++++++-
>  4 files changed, 41 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/cpufreq/cpufreq_conservative.c b/drivers/cpufreq/cpufreq_conservative.c
> index 25a70d0..65c9905 100644
> --- a/drivers/cpufreq/cpufreq_conservative.c
> +++ b/drivers/cpufreq/cpufreq_conservative.c
> @@ -118,7 +118,7 @@ static void cs_dbs_timer(struct work_struct *work)
>  	if (!need_load_eval(&core_dbs_info->cdbs, cs_tuners->sampling_rate))
>  		modify_all = false;
>  	else
> -		dbs_check_cpu(dbs_data, cpu);
> +		dbs_check_cpu(dbs_data, cpu, cs_tuners->sampling_rate);
> 
>  	gov_queue_work(dbs_data, dbs_info->cdbs.cur_policy, delay, modify_all);
>  	mutex_unlock(&core_dbs_info->cdbs.timer_mutex);
> diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c
> index e1c6433..910d472 100644
> --- a/drivers/cpufreq/cpufreq_governor.c
> +++ b/drivers/cpufreq/cpufreq_governor.c
> @@ -30,7 +30,8 @@ static struct attribute_group *get_sysfs_attr(struct dbs_data *dbs_data)
>  		return dbs_data->cdata->attr_group_gov_sys;
>  }
> 
> -void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
> +void dbs_check_cpu(struct dbs_data *dbs_data, int cpu,
> +		   unsigned int sampling_rate)
>  {
>  	struct cpu_dbs_common_info *cdbs = dbs_data->cdata->get_cpu_cdbs(cpu);
>  	struct od_dbs_tuners *od_tuners = dbs_data->tuners;
> @@ -96,7 +97,28 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
>  		if (unlikely(!wall_time || wall_time < idle_time))
>  			continue;
> 
> -		load = 100 * (wall_time - idle_time) / wall_time;
> +		/*
> +		 * If the CPU had gone completely idle, and a task just woke up
> +		 * on this CPU now, it would be unfair to calculate 'load' the
> +		 * usual way for this elapsed time-window, because it will show
> +		 * near-zero load, irrespective of how CPU intensive the new
> +		 * task is. This is undesirable for latency-sensitive bursty
> +		 * workloads.
> +		 *
> +		 * To avoid this, we reuse the 'load' from the previous
> +		 * time-window and give this task a chance to start with a
> +		 * reasonably high CPU frequency.
> +		 *
> +		 * Detecting this situation is easy: the governor's deferrable
> +		 * timer would not have fired during CPU-idle periods. Hence
> +		 * an unusually large 'wall_time' indicates this scenario.
> +		 */
> +		if (unlikely(wall_time > (2 * sampling_rate))) {
					     ^^^^^^^^^^^^^^^^^^

The sampling rate that you've passed is already multiplied by
core_dbs_info->rate_mult. Wouldn't that be sufficient ?

The reason why I am mentioning this is that we could have performed
all the scaling-up at one place.


Other than this, the patch looks good.

Reviewed-by: Gautham R. Shenoy <ego@linux.vnet.ibm.com>

> +			load = j_cdbs->prev_load;
> +		} else {
> +			load = 100 * (wall_time - idle_time) / wall_time;
> +			j_cdbs->prev_load = load;
> +		}
> 
>  		if (load > max_load)
>  			max_load = load;
> @@ -323,6 +345,10 @@ int cpufreq_governor_dbs(struct cpufreq_policy *policy,
>  			j_cdbs->cur_policy = policy;
>  			j_cdbs->prev_cpu_idle = get_cpu_idle_time(j,
>  					       &j_cdbs->prev_cpu_wall, io_busy);
> +			j_cdbs->prev_load = 100 * (j_cdbs->prev_cpu_wall -
> +						   j_cdbs->prev_cpu_idle) /
> +						   j_cdbs->prev_cpu_wall;
> +
>  			if (ignore_nice)
>  				j_cdbs->prev_cpu_nice =
>  					kcpustat_cpu(j).cpustat[CPUTIME_NICE];
> @@ -378,7 +404,7 @@ int cpufreq_governor_dbs(struct cpufreq_policy *policy,
>  		else if (policy->min > cpu_cdbs->cur_policy->cur)
>  			__cpufreq_driver_target(cpu_cdbs->cur_policy,
>  					policy->min, CPUFREQ_RELATION_L);
> -		dbs_check_cpu(dbs_data, cpu);
> +		dbs_check_cpu(dbs_data, cpu, sampling_rate);
>  		mutex_unlock(&cpu_cdbs->timer_mutex);
>  		mutex_unlock(&dbs_data->mutex);
>  		break;
> diff --git a/drivers/cpufreq/cpufreq_governor.h b/drivers/cpufreq/cpufreq_governor.h
> index bfb9ae1..2fbf878 100644
> --- a/drivers/cpufreq/cpufreq_governor.h
> +++ b/drivers/cpufreq/cpufreq_governor.h
> @@ -134,6 +134,7 @@ struct cpu_dbs_common_info {
>  	u64 prev_cpu_idle;
>  	u64 prev_cpu_wall;
>  	u64 prev_cpu_nice;
> +	unsigned int prev_load;
>  	struct cpufreq_policy *cur_policy;
>  	struct delayed_work work;
>  	/*
> @@ -259,7 +260,8 @@ static ssize_t show_sampling_rate_min_gov_pol				\
> 
>  extern struct mutex cpufreq_governor_lock;
> 
> -void dbs_check_cpu(struct dbs_data *dbs_data, int cpu);
> +void dbs_check_cpu(struct dbs_data *dbs_data, int cpu,
> +		   unsigned int sampling_rate);
>  bool need_load_eval(struct cpu_dbs_common_info *cdbs,
>  		unsigned int sampling_rate);
>  int cpufreq_governor_dbs(struct cpufreq_policy *policy,
> diff --git a/drivers/cpufreq/cpufreq_ondemand.c b/drivers/cpufreq/cpufreq_ondemand.c
> index 18d4091..b1f054a 100644
> --- a/drivers/cpufreq/cpufreq_ondemand.c
> +++ b/drivers/cpufreq/cpufreq_ondemand.c
> @@ -213,7 +213,14 @@ static void od_dbs_timer(struct work_struct *work)
>  		__cpufreq_driver_target(core_dbs_info->cdbs.cur_policy,
>  				core_dbs_info->freq_lo, CPUFREQ_RELATION_H);
>  	} else {
> -		dbs_check_cpu(dbs_data, cpu);
> +		/*
> +		 * Provide maximum delay as the sampling_rate hint to
> +		 * dbs_check_cpu, to keep its wake-up-from-cpu-idle detection
> +		 * logic a bit conservative.
> +		 */
> +		dbs_check_cpu(dbs_data, cpu,
> +			od_tuners->sampling_rate * core_dbs_info->rate_mult);
> +
>  		if (core_dbs_info->freq_lo) {
>  			/* Setup timer for SUB_SAMPLE */
>  			core_dbs_info->sample_type = OD_SUB_SAMPLE;
> 

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Srivatsa S. Bhat]

On 06/02/2014 01:03 PM, Gautham R Shenoy wrote:
> Hi,
> 
> On Tue, May 27, 2014 at 02:23:38AM +0530, Srivatsa S. Bhat wrote:
> 
> [..snip..]
>>
>> Experimental results:
>> ====================
>>
>> I ran a modified version of ebizzy (called 'sleeping-ebizzy') that sleeps in
>> between its execution such that its total utilization can be a user-defined
>> value, say 10% or 20% (higher the utilization specified, lesser the amount of
>> sleeps injected). This ebizzy was run with a single-thread, tied to CPU 8.
>>
>> Behavior observed with tracing (sample taken from 40% utilization runs):
>> ------------------------------------------------------------------------
>>
>> Without patch:
>> ~~~~~~~~~~~~~~
>> kworker/8:2-12137  416.335742: cpu_frequency: state=2061000 cpu_id=8
>> kworker/8:2-12137  416.335744: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>       <...>-40753  416.345741: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>> kworker/8:2-12137  416.345744: cpu_frequency: state=4123000 cpu_id=8
>> kworker/8:2-12137  416.345746: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>       <...>-40753  416.355738: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>> <snip>  ---------------------------------------------------------------------  <snip>
>>       <...>-40753  416.402202: sched_switch: prev_comm=ebizzy ==> next_comm=swapper/8
>>      <idle>-0      416.502130: sched_switch: prev_comm=swapper/8 ==> next_comm=ebizzy
>>       <...>-40753  416.505738: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>> kworker/8:2-12137  416.505739: cpu_frequency: state=2061000 cpu_id=8
>> kworker/8:2-12137  416.505741: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>       <...>-40753  416.515739: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>> kworker/8:2-12137  416.515742: cpu_frequency: state=4123000 cpu_id=8
>> kworker/8:2-12137  416.515744: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>
>> Observation: Ebizzy went idle at 416.402202, and started running again at
>> 416.502130. But cpufreq noticed the long idle period, and dropped the frequency
>> at 416.505739, only to increase it back again at 416.515742, realizing that the
>> workload is in-fact CPU bound. Thus ebizzy needlessly ran at the lowest frequency
>> for almost 13 milliseconds (almost 1 full sample period), and this pattern
>> repeats on every sleep-wakeup. This could hurt latency-sensitive workloads quite
>> a lot.
>>
>> With patch:
>> ~~~~~~~~~~~
>>
>> kworker/8:2-29802  464.832535: cpu_frequency: state=2061000 cpu_id=8
>> <snip>  ---------------------------------------------------------------------  <snip>
>> kworker/8:2-29802  464.962538: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>       <...>-40738  464.972533: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>> kworker/8:2-29802  464.972536: cpu_frequency: state=4123000 cpu_id=8
>> kworker/8:2-29802  464.972538: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>       <...>-40738  464.982531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>> <snip>  ---------------------------------------------------------------------  <snip>
>> kworker/8:2-29802  465.022533: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>       <...>-40738  465.032531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>> kworker/8:2-29802  465.032532: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>       <...>-40738  465.035797: sched_switch: prev_comm=ebizzy ==> next_comm=swapper/8
>>      <idle>-0      465.240178: sched_switch: prev_comm=swapper/8 ==> next_comm=ebizzy
>>       <...>-40738  465.242533: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>> kworker/8:2-29802  465.242535: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>       <...>-40738  465.252531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>>
> 
> Have the log entries emmitted by kworker/8 to report about the
> cpu_frequency states been snipped out in the entries post the
> "465.032531" mark ?
> 

No, why? Anything looks odd at that point?

Note that the CPU went idle from 465.035797 to 465.240178, and hence cpufreq's
deferrable timer didn't fire (and hence kworker didn't run). But once the CPU
became busy again at 465.240178, the kworker got scheduled on the CPU within
2 ms (at 465.242533).

> 
>> Observation: Ebizzy went idle at 465.035797, and started running again at
>> 465.240178. Since ebizzy was the only real workload running on this CPU,
>> cpufreq retained the frequency at 4.1Ghz throughout the run of ebizzy, no
>> matter how many times ebizzy slept and woke-up in-between. Thus, ebizzy
>> got the 10ms worth of 4.1 Ghz benefit during every sleep-wakeup (as compared
>> to the run without the patch) and this boost gave a modest improvement in total
>> throughput, as shown below.
>>
>> Sleeping-ebizzy records-per-second:
>> -----------------------------------
>>
>> Utilization  Without patch  With patch  Difference (Absolute and % values)
>>     10%         274767        277046        +  2279 (+0.829%)
>>     20%         543429        553484        + 10055 (+1.850%)
>>     40%        1090744       1107959        + 17215 (+1.578%)
>>     60%        1634908       1662018        + 27110 (+1.658%)
>>
>> A rudimentary and somewhat approximately latency-sensitive workload such as
>> sleeping-ebizzy itself showed a consistent, noticeable performance improvement
>> with this patch. Hence, workloads that are truly latency-sensitive will benefit
>> quite a bit from this change. Moreover, this is an overall win-win since this
>> patch does not hurt power-savings at all (because, this patch does not reduce
>> the idle time or idle residency; and the high frequency of the CPU when it goes
>> to cpu-idle does not affect/hurt the power-savings of deep idle states).
>>
>> Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
>> ---
>>
>>  drivers/cpufreq/cpufreq_conservative.c |    2 +-
>>  drivers/cpufreq/cpufreq_governor.c     |   32 +++++++++++++++++++++++++++++---
>>  drivers/cpufreq/cpufreq_governor.h     |    4 +++-
>>  drivers/cpufreq/cpufreq_ondemand.c     |    9 ++++++++-
>>  4 files changed, 41 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/cpufreq/cpufreq_conservative.c b/drivers/cpufreq/cpufreq_conservative.c
>> index 25a70d0..65c9905 100644
>> --- a/drivers/cpufreq/cpufreq_conservative.c
>> +++ b/drivers/cpufreq/cpufreq_conservative.c
>> @@ -118,7 +118,7 @@ static void cs_dbs_timer(struct work_struct *work)
>>  	if (!need_load_eval(&core_dbs_info->cdbs, cs_tuners->sampling_rate))
>>  		modify_all = false;
>>  	else
>> -		dbs_check_cpu(dbs_data, cpu);
>> +		dbs_check_cpu(dbs_data, cpu, cs_tuners->sampling_rate);
>>
>>  	gov_queue_work(dbs_data, dbs_info->cdbs.cur_policy, delay, modify_all);
>>  	mutex_unlock(&core_dbs_info->cdbs.timer_mutex);
>> diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c
>> index e1c6433..910d472 100644
>> --- a/drivers/cpufreq/cpufreq_governor.c
>> +++ b/drivers/cpufreq/cpufreq_governor.c
>> @@ -30,7 +30,8 @@ static struct attribute_group *get_sysfs_attr(struct dbs_data *dbs_data)
>>  		return dbs_data->cdata->attr_group_gov_sys;
>>  }
>>
>> -void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
>> +void dbs_check_cpu(struct dbs_data *dbs_data, int cpu,
>> +		   unsigned int sampling_rate)
>>  {
>>  	struct cpu_dbs_common_info *cdbs = dbs_data->cdata->get_cpu_cdbs(cpu);
>>  	struct od_dbs_tuners *od_tuners = dbs_data->tuners;
>> @@ -96,7 +97,28 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
>>  		if (unlikely(!wall_time || wall_time < idle_time))
>>  			continue;
>>
>> -		load = 100 * (wall_time - idle_time) / wall_time;
>> +		/*
>> +		 * If the CPU had gone completely idle, and a task just woke up
>> +		 * on this CPU now, it would be unfair to calculate 'load' the
>> +		 * usual way for this elapsed time-window, because it will show
>> +		 * near-zero load, irrespective of how CPU intensive the new
>> +		 * task is. This is undesirable for latency-sensitive bursty
>> +		 * workloads.
>> +		 *
>> +		 * To avoid this, we reuse the 'load' from the previous
>> +		 * time-window and give this task a chance to start with a
>> +		 * reasonably high CPU frequency.
>> +		 *
>> +		 * Detecting this situation is easy: the governor's deferrable
>> +		 * timer would not have fired during CPU-idle periods. Hence
>> +		 * an unusually large 'wall_time' indicates this scenario.
>> +		 */
>> +		if (unlikely(wall_time > (2 * sampling_rate))) {
> 					     ^^^^^^^^^^^^^^^^^^
> 
> The sampling rate that you've passed is already multiplied by
> core_dbs_info->rate_mult. Wouldn't that be sufficient ?
>

Hmm, no. Strictly speaking, the sampling rate is not a constant in the ondemand
governor. It gets multiplied with ->rate_mult to dynamically change the sampling
rate (whenever that is desired). So in the core cpufreq governor code, we should
use this value (i.e., whatever the ondemand governor passes to dbs_check_cpu()
as the sampling_rate, _after_ multiplying with ->rate_mult), since that represents
the maximum sampling rate of the ondemand cpufreq governor.

However, the additional multiplication by 2 in this code is to ensure that we
give sufficient buffer or gap, before we start the new "were we idle for very
long?" logic that this patch introduces.

I didn't want to check for 'wall_time > sampling_rate' because the timer can fire
slightly after the expiration of the sampling period, during regular operation
as well. So we don't want to treat those instances as "very long idle periods",
because that would be incorrect. Hence I decided to give an additional buffer
of a full sampling-period.
 
> The reason why I am mentioning this is that we could have performed
> all the scaling-up at one place.
> 

Indeed, the scaling-up is performed at only this one place: dbs_check_cpu().
This is not "scaling-up" per-se, but rather a means to ensure that we don't
kick off the "copy-previous-load" logic if the CPU wasn't actually idle for long.

And the multiplication with ->rate_mult in the ondemand governor is to avoid the
complication arising from the dynamic behaviour of the sampling-period, by simply
assuming the worst-case (largest) sampling period, all the time. This helps us
accurately recognize the true idle periods in dbs_check_cpu().

> 
> Other than this, the patch looks good.
> 
> Reviewed-by: Gautham R. Shenoy <ego@linux.vnet.ibm.com>

Thank you!

Regards,
Srivatsa S. Bhat

> 
>> +			load = j_cdbs->prev_load;
>> +		} else {
>> +			load = 100 * (wall_time - idle_time) / wall_time;
>> +			j_cdbs->prev_load = load;
>> +		}
>>
>>  		if (load > max_load)
>>  			max_load = load;
>> @@ -323,6 +345,10 @@ int cpufreq_governor_dbs(struct cpufreq_policy *policy,
>>  			j_cdbs->cur_policy = policy;
>>  			j_cdbs->prev_cpu_idle = get_cpu_idle_time(j,
>>  					       &j_cdbs->prev_cpu_wall, io_busy);
>> +			j_cdbs->prev_load = 100 * (j_cdbs->prev_cpu_wall -
>> +						   j_cdbs->prev_cpu_idle) /
>> +						   j_cdbs->prev_cpu_wall;
>> +
>>  			if (ignore_nice)
>>  				j_cdbs->prev_cpu_nice =
>>  					kcpustat_cpu(j).cpustat[CPUTIME_NICE];
>> @@ -378,7 +404,7 @@ int cpufreq_governor_dbs(struct cpufreq_policy *policy,
>>  		else if (policy->min > cpu_cdbs->cur_policy->cur)
>>  			__cpufreq_driver_target(cpu_cdbs->cur_policy,
>>  					policy->min, CPUFREQ_RELATION_L);
>> -		dbs_check_cpu(dbs_data, cpu);
>> +		dbs_check_cpu(dbs_data, cpu, sampling_rate);
>>  		mutex_unlock(&cpu_cdbs->timer_mutex);
>>  		mutex_unlock(&dbs_data->mutex);
>>  		break;
>> diff --git a/drivers/cpufreq/cpufreq_governor.h b/drivers/cpufreq/cpufreq_governor.h
>> index bfb9ae1..2fbf878 100644
>> --- a/drivers/cpufreq/cpufreq_governor.h
>> +++ b/drivers/cpufreq/cpufreq_governor.h
>> @@ -134,6 +134,7 @@ struct cpu_dbs_common_info {
>>  	u64 prev_cpu_idle;
>>  	u64 prev_cpu_wall;
>>  	u64 prev_cpu_nice;
>> +	unsigned int prev_load;
>>  	struct cpufreq_policy *cur_policy;
>>  	struct delayed_work work;
>>  	/*
>> @@ -259,7 +260,8 @@ static ssize_t show_sampling_rate_min_gov_pol				\
>>
>>  extern struct mutex cpufreq_governor_lock;
>>
>> -void dbs_check_cpu(struct dbs_data *dbs_data, int cpu);
>> +void dbs_check_cpu(struct dbs_data *dbs_data, int cpu,
>> +		   unsigned int sampling_rate);
>>  bool need_load_eval(struct cpu_dbs_common_info *cdbs,
>>  		unsigned int sampling_rate);
>>  int cpufreq_governor_dbs(struct cpufreq_policy *policy,
>> diff --git a/drivers/cpufreq/cpufreq_ondemand.c b/drivers/cpufreq/cpufreq_ondemand.c
>> index 18d4091..b1f054a 100644
>> --- a/drivers/cpufreq/cpufreq_ondemand.c
>> +++ b/drivers/cpufreq/cpufreq_ondemand.c
>> @@ -213,7 +213,14 @@ static void od_dbs_timer(struct work_struct *work)
>>  		__cpufreq_driver_target(core_dbs_info->cdbs.cur_policy,
>>  				core_dbs_info->freq_lo, CPUFREQ_RELATION_H);
>>  	} else {
>> -		dbs_check_cpu(dbs_data, cpu);
>> +		/*
>> +		 * Provide maximum delay as the sampling_rate hint to
>> +		 * dbs_check_cpu, to keep its wake-up-from-cpu-idle detection
>> +		 * logic a bit conservative.
>> +		 */
>> +		dbs_check_cpu(dbs_data, cpu,
>> +			od_tuners->sampling_rate * core_dbs_info->rate_mult);
>> +
>>  		if (core_dbs_info->freq_lo) {
>>  			/* Setup timer for SUB_SAMPLE */
>>  			core_dbs_info->sample_type = OD_SUB_SAMPLE;
>>


-- 
Regards,
Srivatsa S. Bhat

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Gautham R Shenoy]

On Mon, Jun 02, 2014 at 01:45:38PM +0530, Srivatsa S. Bhat wrote:
> On 06/02/2014 01:03 PM, Gautham R Shenoy wrote:
> > Hi,
> > 
> > On Tue, May 27, 2014 at 02:23:38AM +0530, Srivatsa S. Bhat wrote:
> > 
> > [..snip..]
> >>
> >> Experimental results:
> >> ====================
> >>
> >> I ran a modified version of ebizzy (called 'sleeping-ebizzy') that sleeps in
> >> between its execution such that its total utilization can be a user-defined
> >> value, say 10% or 20% (higher the utilization specified, lesser the amount of
> >> sleeps injected). This ebizzy was run with a single-thread, tied to CPU 8.
> >>
> >> Behavior observed with tracing (sample taken from 40% utilization runs):
> >> ------------------------------------------------------------------------
> >>
> >> Without patch:
> >> ~~~~~~~~~~~~~~
> >> kworker/8:2-12137  416.335742: cpu_frequency: state=2061000 cpu_id=8
> >> kworker/8:2-12137  416.335744: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
> >>       <...>-40753  416.345741: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> >> kworker/8:2-12137  416.345744: cpu_frequency: state=4123000 cpu_id=8
> >> kworker/8:2-12137  416.345746: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
> >>       <...>-40753  416.355738: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> >> <snip>  ---------------------------------------------------------------------  <snip>
> >>       <...>-40753  416.402202: sched_switch: prev_comm=ebizzy ==> next_comm=swapper/8
> >>      <idle>-0      416.502130: sched_switch: prev_comm=swapper/8 ==> next_comm=ebizzy
> >>       <...>-40753  416.505738: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> >> kworker/8:2-12137  416.505739: cpu_frequency: state=2061000 cpu_id=8
> >> kworker/8:2-12137  416.505741: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
> >>       <...>-40753  416.515739: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> >> kworker/8:2-12137  416.515742: cpu_frequency: state=4123000 cpu_id=8
> >> kworker/8:2-12137  416.515744: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
> >>
> >> Observation: Ebizzy went idle at 416.402202, and started running again at
> >> 416.502130. But cpufreq noticed the long idle period, and dropped the frequency
> >> at 416.505739, only to increase it back again at 416.515742, realizing that the
> >> workload is in-fact CPU bound. Thus ebizzy needlessly ran at the lowest frequency
> >> for almost 13 milliseconds (almost 1 full sample period), and this pattern
> >> repeats on every sleep-wakeup. This could hurt latency-sensitive workloads quite
> >> a lot.
> >>
> >> With patch:
> >> ~~~~~~~~~~~
> >>
> >> kworker/8:2-29802  464.832535: cpu_frequency: state=2061000 cpu_id=8
> >> <snip>  ---------------------------------------------------------------------  <snip>
> >> kworker/8:2-29802  464.962538: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
> >>       <...>-40738  464.972533: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> >> kworker/8:2-29802  464.972536: cpu_frequency: state=4123000 cpu_id=8
> >> kworker/8:2-29802  464.972538: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
> >>       <...>-40738  464.982531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> >> <snip>  ---------------------------------------------------------------------  <snip>
> >> kworker/8:2-29802  465.022533: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
> >>       <...>-40738  465.032531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> >> kworker/8:2-29802  465.032532: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
> >>       <...>-40738  465.035797: sched_switch: prev_comm=ebizzy ==> next_comm=swapper/8
> >>      <idle>-0      465.240178: sched_switch: prev_comm=swapper/8 ==> next_comm=ebizzy
> >>       <...>-40738  465.242533: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> >> kworker/8:2-29802  465.242535: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
> >>       <...>-40738  465.252531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
> >>
> > 
> > Have the log entries emmitted by kworker/8 to report about the
> > cpu_frequency states been snipped out in the entries post the
> > "465.032531" mark ?
> > 
> 
> No, why? Anything looks odd at that point?

I was expecting to see log messages of the following kind after a
kworker thread is scheduled in.

  "kworker/8:2-12137  416.505739: cpu_frequency: state=2061000 cpu_id=8"

> 
> Note that the CPU went idle from 465.035797 to 465.240178, and hence cpufreq's
> deferrable timer didn't fire (and hence kworker didn't run). But once the CPU
> became busy again at 465.240178, the kworker got scheduled on the CPU within
> 2 ms (at 465.242533).

Yes, but the logs don't show the frequency that the kworker thread
would have set on that cpu.

> [..snip...]
> >>
> >> -		load = 100 * (wall_time - idle_time) / wall_time;
> >> +		/*
> >> +		 * If the CPU had gone completely idle, and a task just woke up
> >> +		 * on this CPU now, it would be unfair to calculate 'load' the
> >> +		 * usual way for this elapsed time-window, because it will show
> >> +		 * near-zero load, irrespective of how CPU intensive the new
> >> +		 * task is. This is undesirable for latency-sensitive bursty
> >> +		 * workloads.
> >> +		 *
> >> +		 * To avoid this, we reuse the 'load' from the previous
> >> +		 * time-window and give this task a chance to start with a
> >> +		 * reasonably high CPU frequency.
> >> +		 *
> >> +		 * Detecting this situation is easy: the governor's deferrable
> >> +		 * timer would not have fired during CPU-idle periods. Hence
> >> +		 * an unusually large 'wall_time' indicates this scenario.
> >> +		 */
> >> +		if (unlikely(wall_time > (2 * sampling_rate))) {
> > 					     ^^^^^^^^^^^^^^^^^^
> > 
> > The sampling rate that you've passed is already multiplied by
> > core_dbs_info->rate_mult. Wouldn't that be sufficient ?
> >
> 
> Hmm, no. Strictly speaking, the sampling rate is not a constant in the ondemand
> governor. It gets multiplied with ->rate_mult to dynamically change the sampling
> rate (whenever that is desired). So in the core cpufreq governor code, we should
> use this value (i.e., whatever the ondemand governor passes to dbs_check_cpu()
> as the sampling_rate, _after_ multiplying with ->rate_mult), since that represents
> the maximum sampling rate of the ondemand cpufreq governor.
> 
> However, the additional multiplication by 2 in this code is to ensure that we
> give sufficient buffer or gap, before we start the new "were we idle for very
> long?" logic that this patch introduces.
> 
> I didn't want to check for 'wall_time > sampling_rate' because the timer can fire
> slightly after the expiration of the sampling period, during regular operation
> as well. So we don't want to treat those instances as "very long idle periods",
> because that would be incorrect. Hence I decided to give an additional buffer
> of a full sampling-period.

Ok this makes sense. Thanks for the explanation.


> 
> > The reason why I am mentioning this is that we could have performed
> > all the scaling-up at one place.
> > 
> 
> Indeed, the scaling-up is performed at only this one place: dbs_check_cpu().
> This is not "scaling-up" per-se, but rather a means to ensure that we don't
> kick off the "copy-previous-load" logic if the CPU wasn't actually
> idle for long.

Ok. Understood. 

> 
> And the multiplication with ->rate_mult in the ondemand governor is to avoid the
> complication arising from the dynamic behaviour of the sampling-period, by simply
> assuming the worst-case (largest) sampling period, all the time. This helps us
> accurately recognize the true idle periods in dbs_check_cpu().
> 
> > 
> > Other than this, the patch looks good.
> > 
> > Reviewed-by: Gautham R. Shenoy <ego@linux.vnet.ibm.com>
> 
> Thank you!
> 
> Regards,
> Srivatsa S. Bhat
> 
--
Thanks and Regards
gautham.

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Srivatsa S. Bhat]

On 06/03/2014 10:46 AM, Gautham R Shenoy wrote:
> On Mon, Jun 02, 2014 at 01:45:38PM +0530, Srivatsa S. Bhat wrote:
>> On 06/02/2014 01:03 PM, Gautham R Shenoy wrote:
>>> Hi,
>>>
>>> On Tue, May 27, 2014 at 02:23:38AM +0530, Srivatsa S. Bhat wrote:
>>>
>>> [..snip..]
>>>>
>>>> Experimental results:
>>>> ====================
>>>>
>>>> I ran a modified version of ebizzy (called 'sleeping-ebizzy') that sleeps in
>>>> between its execution such that its total utilization can be a user-defined
>>>> value, say 10% or 20% (higher the utilization specified, lesser the amount of
>>>> sleeps injected). This ebizzy was run with a single-thread, tied to CPU 8.
>>>>
>>>> Behavior observed with tracing (sample taken from 40% utilization runs):
>>>> ------------------------------------------------------------------------
>>>>
>>>> Without patch:
>>>> ~~~~~~~~~~~~~~
>>>> kworker/8:2-12137  416.335742: cpu_frequency: state=2061000 cpu_id=8
>>>> kworker/8:2-12137  416.335744: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>>>       <...>-40753  416.345741: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>>>> kworker/8:2-12137  416.345744: cpu_frequency: state=4123000 cpu_id=8
>>>> kworker/8:2-12137  416.345746: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>>>       <...>-40753  416.355738: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>>>> <snip>  ---------------------------------------------------------------------  <snip>
>>>>       <...>-40753  416.402202: sched_switch: prev_comm=ebizzy ==> next_comm=swapper/8
>>>>      <idle>-0      416.502130: sched_switch: prev_comm=swapper/8 ==> next_comm=ebizzy
>>>>       <...>-40753  416.505738: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>>>> kworker/8:2-12137  416.505739: cpu_frequency: state=2061000 cpu_id=8
>>>> kworker/8:2-12137  416.505741: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>>>       <...>-40753  416.515739: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>>>> kworker/8:2-12137  416.515742: cpu_frequency: state=4123000 cpu_id=8
>>>> kworker/8:2-12137  416.515744: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>>>
>>>> Observation: Ebizzy went idle at 416.402202, and started running again at
>>>> 416.502130. But cpufreq noticed the long idle period, and dropped the frequency
>>>> at 416.505739, only to increase it back again at 416.515742, realizing that the
>>>> workload is in-fact CPU bound. Thus ebizzy needlessly ran at the lowest frequency
>>>> for almost 13 milliseconds (almost 1 full sample period), and this pattern
>>>> repeats on every sleep-wakeup. This could hurt latency-sensitive workloads quite
>>>> a lot.
>>>>
>>>> With patch:
>>>> ~~~~~~~~~~~
>>>>
>>>> kworker/8:2-29802  464.832535: cpu_frequency: state=2061000 cpu_id=8
>>>> <snip>  ---------------------------------------------------------------------  <snip>
>>>> kworker/8:2-29802  464.962538: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>>>       <...>-40738  464.972533: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>>>> kworker/8:2-29802  464.972536: cpu_frequency: state=4123000 cpu_id=8
>>>> kworker/8:2-29802  464.972538: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>>>       <...>-40738  464.982531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>>>> <snip>  ---------------------------------------------------------------------  <snip>
>>>> kworker/8:2-29802  465.022533: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>>>       <...>-40738  465.032531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>>>> kworker/8:2-29802  465.032532: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>>>       <...>-40738  465.035797: sched_switch: prev_comm=ebizzy ==> next_comm=swapper/8
>>>>      <idle>-0      465.240178: sched_switch: prev_comm=swapper/8 ==> next_comm=ebizzy
>>>>       <...>-40738  465.242533: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>>>> kworker/8:2-29802  465.242535: sched_switch: prev_comm=kworker/8:2 ==> next_comm=ebizzy
>>>>       <...>-40738  465.252531: sched_switch: prev_comm=ebizzy ==> next_comm=kworker/8:2
>>>>
>>>
>>> Have the log entries emmitted by kworker/8 to report about the
>>> cpu_frequency states been snipped out in the entries post the
>>> "465.032531" mark ?
>>>
>>
>> No, why? Anything looks odd at that point?
> 
> I was expecting to see log messages of the following kind after a
> kworker thread is scheduled in.
> 
>   "kworker/8:2-12137  416.505739: cpu_frequency: state=2061000 cpu_id=8"
>

But this gets printed only if the frequency is changed. If the frequency is left at the
same value as it was previously set at (that's the point of this patch), then we won't
get this print. [Note that these logs are with the patch applied.]
 
>>
>> Note that the CPU went idle from 465.035797 to 465.240178, and hence cpufreq's
>> deferrable timer didn't fire (and hence kworker didn't run). But once the CPU
>> became busy again at 465.240178, the kworker got scheduled on the CPU within
>> 2 ms (at 465.242533).
> 
> Yes, but the logs don't show the frequency that the kworker thread
> would have set on that cpu.
> 

Yes, and that's expected, because we copy the previous load to the present interval,
and hence kworker won't change the frequency (in most cases), because it finds that
the frequency is already set suitably for the perceived load in this interval.
Hence we don't see any prints in the logs indicating a change in frequency.

Regards,
Srivatsa S. Bhat

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Viresh Kumar]

On 27 May 2014 02:23, Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> wrote:

Looks fine, some nits..

> diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c
> -void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
> +void dbs_check_cpu(struct dbs_data *dbs_data, int cpu,
> +                  unsigned int sampling_rate)

We don't need to pass a new argument, we can get all the information from
dbs_data alone. Its already done for multiple routines. Let me know if you
find it difficult to figure out..

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Srivatsa S. Bhat]

On 06/03/2014 01:48 PM, Viresh Kumar wrote:
> On 27 May 2014 02:23, Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> wrote:
> 
> Looks fine, some nits..
> 
>> diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c
>> -void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
>> +void dbs_check_cpu(struct dbs_data *dbs_data, int cpu,
>> +                  unsigned int sampling_rate)
> 
> We don't need to pass a new argument, we can get all the information from
> dbs_data alone. Its already done for multiple routines. Let me know if you
> find it difficult to figure out..
> 

Sure, that would be a good improvement. Does something like the patch below
look good? I have only compile-tested it. I'll send out the patch with changelog
once I finish testing it.

Thank you!

Regards,
Srivatsa S. Bhat

----------------------------------------------------------------------------

diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c
index e1c6433..3e8588f 100644
--- a/drivers/cpufreq/cpufreq_governor.c
+++ b/drivers/cpufreq/cpufreq_governor.c
@@ -36,14 +36,29 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
 	struct od_dbs_tuners *od_tuners = dbs_data->tuners;
 	struct cs_dbs_tuners *cs_tuners = dbs_data->tuners;
 	struct cpufreq_policy *policy;
+	unsigned int sampling_rate;
 	unsigned int max_load = 0;
 	unsigned int ignore_nice;
 	unsigned int j;
 
-	if (dbs_data->cdata->governor == GOV_ONDEMAND)
+	if (dbs_data->cdata->governor == GOV_ONDEMAND) {
+		struct od_cpu_dbs_info_s *od_dbs_info;
+
+		/*
+		 * Sometimes, the ondemand governor uses an additional
+		 * multiplier to give long delays. So apply this multiplier to
+		 * the 'sampling_rate', so as to keep the wake-up-from-idle
+		 * detection logic a bit conservative.
+		 */
+		sampling_rate = od_tuners->sampling_rate;
+		od_dbs_info = dbs_data->cdata->get_cpu_dbs_info_s(cpu);
+		sampling_rate *= od_dbs_info->rate_mult;
+
 		ignore_nice = od_tuners->ignore_nice_load;
-	else
+	} else {
+		sampling_rate = cs_tuners->sampling_rate;
 		ignore_nice = cs_tuners->ignore_nice_load;
+	}
 
 	policy = cdbs->cur_policy;
 
@@ -96,7 +111,29 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
 		if (unlikely(!wall_time || wall_time < idle_time))
 			continue;
 
-		load = 100 * (wall_time - idle_time) / wall_time;
+		/*
+		 * If the CPU had gone completely idle, and a task just woke up
+		 * on this CPU now, it would be unfair to calculate 'load' the
+		 * usual way for this elapsed time-window, because it will show
+		 * near-zero load, irrespective of how CPU intensive the new
+		 * task is. This is undesirable for latency-sensitive bursty
+		 * workloads.
+		 *
+		 * To avoid this, we reuse the 'load' from the previous
+		 * time-window and give this task a chance to start with a
+		 * reasonably high CPU frequency.
+		 *
+		 * Detecting this situation is easy: the governor's deferrable
+		 * timer would not have fired during CPU-idle periods. Hence
+		 * an unusually large 'wall_time' (as compared to the sampling
+		 * rate) indicates this scenario.
+		 */
+		if (unlikely(wall_time > (2 * sampling_rate))) {
+			load = j_cdbs->prev_load;
+		} else {
+			load = 100 * (wall_time - idle_time) / wall_time;
+			j_cdbs->prev_load = load;
+		}
 
 		if (load > max_load)
 			max_load = load;
@@ -323,6 +360,10 @@ int cpufreq_governor_dbs(struct cpufreq_policy *policy,
 			j_cdbs->cur_policy = policy;
 			j_cdbs->prev_cpu_idle = get_cpu_idle_time(j,
 					       &j_cdbs->prev_cpu_wall, io_busy);
+			j_cdbs->prev_load = 100 * (j_cdbs->prev_cpu_wall -
+						   j_cdbs->prev_cpu_idle) /
+						   j_cdbs->prev_cpu_wall;
+
 			if (ignore_nice)
 				j_cdbs->prev_cpu_nice =
 					kcpustat_cpu(j).cpustat[CPUTIME_NICE];
diff --git a/drivers/cpufreq/cpufreq_governor.h b/drivers/cpufreq/cpufreq_governor.h
index bfb9ae1..b56552b 100644
--- a/drivers/cpufreq/cpufreq_governor.h
+++ b/drivers/cpufreq/cpufreq_governor.h
@@ -134,6 +134,7 @@ struct cpu_dbs_common_info {
 	u64 prev_cpu_idle;
 	u64 prev_cpu_wall;
 	u64 prev_cpu_nice;
+	unsigned int prev_load;
 	struct cpufreq_policy *cur_policy;
 	struct delayed_work work;
 	/*

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Viresh Kumar]

On 3 June 2014 15:02, Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> wrote:
> diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c
> index e1c6433..3e8588f 100644
> --- a/drivers/cpufreq/cpufreq_governor.c
> +++ b/drivers/cpufreq/cpufreq_governor.c
> @@ -36,14 +36,29 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
>         struct od_dbs_tuners *od_tuners = dbs_data->tuners;
>         struct cs_dbs_tuners *cs_tuners = dbs_data->tuners;
>         struct cpufreq_policy *policy;
> +       unsigned int sampling_rate;
>         unsigned int max_load = 0;
>         unsigned int ignore_nice;
>         unsigned int j;
>
> -       if (dbs_data->cdata->governor == GOV_ONDEMAND)
> +       if (dbs_data->cdata->governor == GOV_ONDEMAND) {
> +               struct od_cpu_dbs_info_s *od_dbs_info;
> +
> +               /*
> +                * Sometimes, the ondemand governor uses an additional
> +                * multiplier to give long delays. So apply this multiplier to
> +                * the 'sampling_rate', so as to keep the wake-up-from-idle
> +                * detection logic a bit conservative.
> +                */
> +               sampling_rate = od_tuners->sampling_rate;
> +               od_dbs_info = dbs_data->cdata->get_cpu_dbs_info_s(cpu);

Probably do both above right after definition of od_dbs_info or merge
above with it. and just keep below after the comment ?

> +               sampling_rate *= od_dbs_info->rate_mult;

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Srivatsa S. Bhat]

On 06/03/2014 03:09 PM, Viresh Kumar wrote:
> On 3 June 2014 15:02, Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> wrote:
>> diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c
>> index e1c6433..3e8588f 100644
>> --- a/drivers/cpufreq/cpufreq_governor.c
>> +++ b/drivers/cpufreq/cpufreq_governor.c
>> @@ -36,14 +36,29 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
>>         struct od_dbs_tuners *od_tuners = dbs_data->tuners;
>>         struct cs_dbs_tuners *cs_tuners = dbs_data->tuners;
>>         struct cpufreq_policy *policy;
>> +       unsigned int sampling_rate;
>>         unsigned int max_load = 0;
>>         unsigned int ignore_nice;
>>         unsigned int j;
>>
>> -       if (dbs_data->cdata->governor == GOV_ONDEMAND)
>> +       if (dbs_data->cdata->governor == GOV_ONDEMAND) {
>> +               struct od_cpu_dbs_info_s *od_dbs_info;
>> +
>> +               /*
>> +                * Sometimes, the ondemand governor uses an additional
>> +                * multiplier to give long delays. So apply this multiplier to
>> +                * the 'sampling_rate', so as to keep the wake-up-from-idle
>> +                * detection logic a bit conservative.
>> +                */
>> +               sampling_rate = od_tuners->sampling_rate;
>> +               od_dbs_info = dbs_data->cdata->get_cpu_dbs_info_s(cpu);
> 
> Probably do both above right after definition of od_dbs_info or merge
> above with it. and just keep below after the comment ?
> 
>> +               sampling_rate *= od_dbs_info->rate_mult;
> 

Well, the method I used keeps the organization such that the code following
the comment does precisely what the comment says (i.e, get the sampling_rate,
fetch the multiplier, and then multiply). So I feel it makes it easier to
understand.

Regards,
Srivatsa S. Bhat

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Viresh Kumar]

On 3 June 2014 15:34, Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> wrote:
> Well, the method I used keeps the organization such that the code following
> the comment does precisely what the comment says (i.e, get the sampling_rate,
> fetch the multiplier, and then multiply). So I feel it makes it easier to
> understand.

It looked like the comment is there only for this special statement:

>>> +               sampling_rate *= od_dbs_info->rate_mult;

And so suggested that :)

Anyway move this up as it doesn't belong to comment for sure.
>> +               od_dbs_info = dbs_data->cdata->get_cpu_dbs_info_s(cpu);

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Srivatsa S. Bhat]

On 06/03/2014 03:38 PM, Viresh Kumar wrote:
> On 3 June 2014 15:34, Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> wrote:
>> Well, the method I used keeps the organization such that the code following
>> the comment does precisely what the comment says (i.e, get the sampling_rate,
>> fetch the multiplier, and then multiply). So I feel it makes it easier to
>> understand.
> 
> It looked like the comment is there only for this special statement:
> 
>>>> +               sampling_rate *= od_dbs_info->rate_mult;
> 
> And so suggested that :)
> 
> Anyway move this up as it doesn't belong to comment for sure.
>>> +               od_dbs_info = dbs_data->cdata->get_cpu_dbs_info_s(cpu);
> 

Fair enough :) Here it is:


diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c
index e1c6433..2597bbe 100644
--- a/drivers/cpufreq/cpufreq_governor.c
+++ b/drivers/cpufreq/cpufreq_governor.c
@@ -36,14 +36,29 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
 	struct od_dbs_tuners *od_tuners = dbs_data->tuners;
 	struct cs_dbs_tuners *cs_tuners = dbs_data->tuners;
 	struct cpufreq_policy *policy;
+	unsigned int sampling_rate;
 	unsigned int max_load = 0;
 	unsigned int ignore_nice;
 	unsigned int j;
 
-	if (dbs_data->cdata->governor == GOV_ONDEMAND)
+	if (dbs_data->cdata->governor == GOV_ONDEMAND) {
+		struct od_cpu_dbs_info_s *od_dbs_info =
+				dbs_data->cdata->get_cpu_dbs_info_s(cpu);
+
+		/*
+		 * Sometimes, the ondemand governor uses an additional
+		 * multiplier to give long delays. So apply this multiplier to
+		 * the 'sampling_rate', so as to keep the wake-up-from-idle
+		 * detection logic a bit conservative.
+		 */
+		sampling_rate = od_tuners->sampling_rate;
+		sampling_rate *= od_dbs_info->rate_mult;
+
 		ignore_nice = od_tuners->ignore_nice_load;
-	else
+	} else {
+		sampling_rate = cs_tuners->sampling_rate;
 		ignore_nice = cs_tuners->ignore_nice_load;
+	}
 
 	policy = cdbs->cur_policy;
 
@@ -96,7 +111,29 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
 		if (unlikely(!wall_time || wall_time < idle_time))
 			continue;
 
-		load = 100 * (wall_time - idle_time) / wall_time;
+		/*
+		 * If the CPU had gone completely idle, and a task just woke up
+		 * on this CPU now, it would be unfair to calculate 'load' the
+		 * usual way for this elapsed time-window, because it will show
+		 * near-zero load, irrespective of how CPU intensive the new
+		 * task is. This is undesirable for latency-sensitive bursty
+		 * workloads.
+		 *
+		 * To avoid this, we reuse the 'load' from the previous
+		 * time-window and give this task a chance to start with a
+		 * reasonably high CPU frequency.
+		 *
+		 * Detecting this situation is easy: the governor's deferrable
+		 * timer would not have fired during CPU-idle periods. Hence
+		 * an unusually large 'wall_time' (as compared to the sampling
+		 * rate) indicates this scenario.
+		 */
+		if (unlikely(wall_time > (2 * sampling_rate))) {
+			load = j_cdbs->prev_load;
+		} else {
+			load = 100 * (wall_time - idle_time) / wall_time;
+			j_cdbs->prev_load = load;
+		}
 
 		if (load > max_load)
 			max_load = load;
@@ -323,6 +360,10 @@ int cpufreq_governor_dbs(struct cpufreq_policy *policy,
 			j_cdbs->cur_policy = policy;
 			j_cdbs->prev_cpu_idle = get_cpu_idle_time(j,
 					       &j_cdbs->prev_cpu_wall, io_busy);
+			j_cdbs->prev_load = 100 * (j_cdbs->prev_cpu_wall -
+						   j_cdbs->prev_cpu_idle) /
+						   j_cdbs->prev_cpu_wall;
+
 			if (ignore_nice)
 				j_cdbs->prev_cpu_nice =
 					kcpustat_cpu(j).cpustat[CPUTIME_NICE];
diff --git a/drivers/cpufreq/cpufreq_governor.h b/drivers/cpufreq/cpufreq_governor.h
index bfb9ae1..b56552b 100644
--- a/drivers/cpufreq/cpufreq_governor.h
+++ b/drivers/cpufreq/cpufreq_governor.h
@@ -134,6 +134,7 @@ struct cpu_dbs_common_info {
 	u64 prev_cpu_idle;
 	u64 prev_cpu_wall;
 	u64 prev_cpu_nice;
+	unsigned int prev_load;
 	struct cpufreq_policy *cur_policy;
 	struct delayed_work work;
 	/*

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Viresh Kumar]

On 3 June 2014 15:43, Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> wrote:
> diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c
> index e1c6433..2597bbe 100644
> --- a/drivers/cpufreq/cpufreq_governor.c
> +++ b/drivers/cpufreq/cpufreq_governor.c
> @@ -36,14 +36,29 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
>         struct od_dbs_tuners *od_tuners = dbs_data->tuners;
>         struct cs_dbs_tuners *cs_tuners = dbs_data->tuners;
>         struct cpufreq_policy *policy;
> +       unsigned int sampling_rate;
>         unsigned int max_load = 0;
>         unsigned int ignore_nice;
>         unsigned int j;
>
> -       if (dbs_data->cdata->governor == GOV_ONDEMAND)
> +       if (dbs_data->cdata->governor == GOV_ONDEMAND) {
> +               struct od_cpu_dbs_info_s *od_dbs_info =
> +                               dbs_data->cdata->get_cpu_dbs_info_s(cpu);
> +
> +               /*
> +                * Sometimes, the ondemand governor uses an additional
> +                * multiplier to give long delays. So apply this multiplier to
> +                * the 'sampling_rate', so as to keep the wake-up-from-idle
> +                * detection logic a bit conservative.
> +                */
> +               sampling_rate = od_tuners->sampling_rate;
> +               sampling_rate *= od_dbs_info->rate_mult;
> +
>                 ignore_nice = od_tuners->ignore_nice_load;
> -       else
> +       } else {
> +               sampling_rate = cs_tuners->sampling_rate;
>                 ignore_nice = cs_tuners->ignore_nice_load;
> +       }
>
>         policy = cdbs->cur_policy;
>
> @@ -96,7 +111,29 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
>                 if (unlikely(!wall_time || wall_time < idle_time))
>                         continue;
>
> -               load = 100 * (wall_time - idle_time) / wall_time;
> +               /*
> +                * If the CPU had gone completely idle, and a task just woke up
> +                * on this CPU now, it would be unfair to calculate 'load' the
> +                * usual way for this elapsed time-window, because it will show
> +                * near-zero load, irrespective of how CPU intensive the new
> +                * task is. This is undesirable for latency-sensitive bursty
> +                * workloads.
> +                *
> +                * To avoid this, we reuse the 'load' from the previous
> +                * time-window and give this task a chance to start with a
> +                * reasonably high CPU frequency.
> +                *
> +                * Detecting this situation is easy: the governor's deferrable
> +                * timer would not have fired during CPU-idle periods. Hence
> +                * an unusually large 'wall_time' (as compared to the sampling
> +                * rate) indicates this scenario.
> +                */
> +               if (unlikely(wall_time > (2 * sampling_rate))) {
> +                       load = j_cdbs->prev_load;
> +               } else {
> +                       load = 100 * (wall_time - idle_time) / wall_time;
> +                       j_cdbs->prev_load = load;
> +               }
>
>                 if (load > max_load)
>                         max_load = load;
> @@ -323,6 +360,10 @@ int cpufreq_governor_dbs(struct cpufreq_policy *policy,
>                         j_cdbs->cur_policy = policy;
>                         j_cdbs->prev_cpu_idle = get_cpu_idle_time(j,
>                                                &j_cdbs->prev_cpu_wall, io_busy);
> +                       j_cdbs->prev_load = 100 * (j_cdbs->prev_cpu_wall -
> +                                                  j_cdbs->prev_cpu_idle) /
> +                                                  j_cdbs->prev_cpu_wall;
> +
>                         if (ignore_nice)
>                                 j_cdbs->prev_cpu_nice =
>                                         kcpustat_cpu(j).cpustat[CPUTIME_NICE];
> diff --git a/drivers/cpufreq/cpufreq_governor.h b/drivers/cpufreq/cpufreq_governor.h
> index bfb9ae1..b56552b 100644
> --- a/drivers/cpufreq/cpufreq_governor.h
> +++ b/drivers/cpufreq/cpufreq_governor.h
> @@ -134,6 +134,7 @@ struct cpu_dbs_common_info {
>         u64 prev_cpu_idle;
>         u64 prev_cpu_wall;
>         u64 prev_cpu_nice;
> +       unsigned int prev_load;
>         struct cpufreq_policy *cur_policy;
>         struct delayed_work work;
>         /*

Acked-by: Viresh Kumar <viresh.kumar@linaro.org>

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Srivatsa S. Bhat]

On 06/03/2014 03:46 PM, Viresh Kumar wrote:
> On 3 June 2014 15:43, Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> wrote:
>> diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c
>> index e1c6433..2597bbe 100644
>> --- a/drivers/cpufreq/cpufreq_governor.c
>> +++ b/drivers/cpufreq/cpufreq_governor.c
>> @@ -36,14 +36,29 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
>>         struct od_dbs_tuners *od_tuners = dbs_data->tuners;
>>         struct cs_dbs_tuners *cs_tuners = dbs_data->tuners;
>>         struct cpufreq_policy *policy;
>> +       unsigned int sampling_rate;
>>         unsigned int max_load = 0;
>>         unsigned int ignore_nice;
>>         unsigned int j;
[...]
>> diff --git a/drivers/cpufreq/cpufreq_governor.h b/drivers/cpufreq/cpufreq_governor.h
>> index bfb9ae1..b56552b 100644
>> --- a/drivers/cpufreq/cpufreq_governor.h
>> +++ b/drivers/cpufreq/cpufreq_governor.h
>> @@ -134,6 +134,7 @@ struct cpu_dbs_common_info {
>>         u64 prev_cpu_idle;
>>         u64 prev_cpu_wall;
>>         u64 prev_cpu_nice;
>> +       unsigned int prev_load;
>>         struct cpufreq_policy *cur_policy;
>>         struct delayed_work work;
>>         /*
> 
> Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
> 

Thanks a lot!
 
Regards,
Srivatsa S. Bhat

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Pavel Machek]

Hi!

> We just want to avoid the stupidity of dropping down the frequency to a minimum
> and then enduring a needless (and long) delay before ramping it up back again.
> So, let us simply carry forward the previous load - that is, let us just pretend
> that the 'load' for the current time-window is the same as the load for the
> previous window. That way, the frequency and voltage will continue to be set
> to whatever values they were set at previously. This means that bursty workloads
> will get a chance to influence the CPU frequency at which they wake up from
> cpu-idle, based on their past execution history. Thus, they might be able to
> avoid suffering from slow wakeups and long response-times.
> 
> [ The right way to solve this problem is to teach the CPU frequency governors
> to track load on a per-task basis, not a per-CPU basis, and set the appropriate
> frequency on whichever CPU the task executes. But that involves redesigning
> the cpufreq subsystem, so this patch should make the situation bearable until
> then. ]

Are you sure? For example "./configure" load consists of a lot of
short-lived tasks. Per-task basis may not be option for that.

> A rudimentary and somewhat approximately latency-sensitive workload such as
> sleeping-ebizzy itself showed a consistent, noticeable performance improvement
> with this patch. Hence, workloads that are truly latency-sensitive will benefit
> quite a bit from this change. Moreover, this is an overall win-win since this
> patch does not hurt power-savings at all (because, this patch does not reduce
> the idle time or idle residency; and the high frequency of the CPU when it goes
> to cpu-idle does not affect/hurt the power-savings of deep idle
> states).

Are you sure about win-win?

AFAICT, your patch helps

##########.........#########.........###########...........##########............

case (not surprising, that's why you wrote the patch :-), but what happens in

##########.........#.................#.....................#.....................

case? (That is idle system, with some tasks taking very small ammounts
of CPU).

AFAICT, it will remember the (high) prev_load over the idle period,
and use too high frequency for mostly idle system, no?

									Pavel

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads

[Srivatsa S. Bhat]

On 06/07/2014 03:25 PM, Pavel Machek wrote:
> Hi!
> 
>> We just want to avoid the stupidity of dropping down the frequency to a minimum
>> and then enduring a needless (and long) delay before ramping it up back again.
>> So, let us simply carry forward the previous load - that is, let us just pretend
>> that the 'load' for the current time-window is the same as the load for the
>> previous window. That way, the frequency and voltage will continue to be set
>> to whatever values they were set at previously. This means that bursty workloads
>> will get a chance to influence the CPU frequency at which they wake up from
>> cpu-idle, based on their past execution history. Thus, they might be able to
>> avoid suffering from slow wakeups and long response-times.
>>
>> [ The right way to solve this problem is to teach the CPU frequency governors
>> to track load on a per-task basis, not a per-CPU basis, and set the appropriate
>> frequency on whichever CPU the task executes. But that involves redesigning
>> the cpufreq subsystem, so this patch should make the situation bearable until
>> then. ]
> 
> Are you sure? For example "./configure" load consists of a lot of
> short-lived tasks. Per-task basis may not be option for that.
> 

True, but then there is no guarantee that all the tasks spawned by ./configure
run on a single CPU. So per-CPU tracking may or may not help, depending on the
scheduling pattern.

For very-short-lived tasks, per-task tracking won't give much benefit, but it
will for medium to long-lived tasks. So I guess we might need some sort of a
combination of both methods of tracking, to handle all the scenarios well.

>> A rudimentary and somewhat approximately latency-sensitive workload such as
>> sleeping-ebizzy itself showed a consistent, noticeable performance improvement
>> with this patch. Hence, workloads that are truly latency-sensitive will benefit
>> quite a bit from this change. Moreover, this is an overall win-win since this
>> patch does not hurt power-savings at all (because, this patch does not reduce
>> the idle time or idle residency; and the high frequency of the CPU when it goes
>> to cpu-idle does not affect/hurt the power-savings of deep idle
>> states).
> 
> Are you sure about win-win?
> 
> AFAICT, your patch helps
> 
> ##########.........#########.........###########...........##########............
> 
> case (not surprising, that's why you wrote the patch :-), but what happens in
> 
> ##########.........#.................#.....................#.....................
> 
> case? (That is idle system, with some tasks taking very small ammounts
> of CPU).
> 
> AFAICT, it will remember the (high) prev_load over the idle period,
> and use too high frequency for mostly idle system, no?
> 

Wow, great catch! I think the problem is that the prev-load is getting copied
over too many times; we need to restrict it to just one-copy and leave the
next sampling interval to the cpufreq governor's regular ramp-up, even if it
was an idle interval.

Below is an untested patch that implements that logic (and it also fixes the
64 bit division problem reported by Fengguang's build robot in the v2 of the
patch).

I'll test this patch and then send out the v3 later today. Please let me
know your thoughts!

Regards,
Srivatsa S. Bhat

---------------------------------------------------------------------------

diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c
index e1c6433..3856a6b 100644
--- a/drivers/cpufreq/cpufreq_governor.c
+++ b/drivers/cpufreq/cpufreq_governor.c
@@ -36,14 +36,29 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
 	struct od_dbs_tuners *od_tuners = dbs_data->tuners;
 	struct cs_dbs_tuners *cs_tuners = dbs_data->tuners;
 	struct cpufreq_policy *policy;
+	unsigned int sampling_rate;
 	unsigned int max_load = 0;
 	unsigned int ignore_nice;
 	unsigned int j;
 
-	if (dbs_data->cdata->governor == GOV_ONDEMAND)
+	if (dbs_data->cdata->governor == GOV_ONDEMAND) {
+		struct od_cpu_dbs_info_s *od_dbs_info =
+				dbs_data->cdata->get_cpu_dbs_info_s(cpu);
+
+		/*
+		 * Sometimes, the ondemand governor uses an additional
+		 * multiplier to give long delays. So apply this multiplier to
+		 * the 'sampling_rate', so as to keep the wake-up-from-idle
+		 * detection logic a bit conservative.
+		 */
+		sampling_rate = od_tuners->sampling_rate;
+		sampling_rate *= od_dbs_info->rate_mult;
+
 		ignore_nice = od_tuners->ignore_nice_load;
-	else
+	} else {
+		sampling_rate = cs_tuners->sampling_rate;
 		ignore_nice = cs_tuners->ignore_nice_load;
+	}
 
 	policy = cdbs->cur_policy;
 
@@ -96,7 +111,36 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu)
 		if (unlikely(!wall_time || wall_time < idle_time))
 			continue;
 
-		load = 100 * (wall_time - idle_time) / wall_time;
+		/*
+		 * If the CPU had gone completely idle, and a task just woke up
+		 * on this CPU now, it would be unfair to calculate 'load' the
+		 * usual way for this elapsed time-window, because it will show
+		 * near-zero load, irrespective of how CPU intensive the new
+		 * task is. This is undesirable for latency-sensitive bursty
+		 * workloads.
+		 *
+		 * To avoid this, we reuse the 'load' from the previous
+		 * time-window and give this task a chance to start with a
+		 * reasonably high CPU frequency. (However, we shouldn't over-do
+		 * this copy, lest we get stuck at a high load (high frequency)
+		 * for too long, even when the current system load has actually
+		 * dropped down. So we perform the copy only once, upon the
+		 * first wake-up from idle.)
+		 *
+		 * Detecting this situation is easy: the governor's deferrable
+		 * timer would not have fired during CPU-idle periods. Hence
+		 * an unusually large 'wall_time' (as compared to the sampling
+		 * rate) indicates this scenario.
+		 */
+		if (unlikely(wall_time > (2 * sampling_rate)) &&
+						j_cdbs->copy_prev_load) {
+			load = j_cdbs->prev_load;
+			j_cdbs->copy_prev_load = false;
+		} else {
+			load = 100 * (wall_time - idle_time) / wall_time;
+			j_cdbs->prev_load = load;
+			j_cdbs->copy_prev_load = true;
+		}
 
 		if (load > max_load)
 			max_load = load;
@@ -318,11 +362,19 @@ int cpufreq_governor_dbs(struct cpufreq_policy *policy,
 		for_each_cpu(j, policy->cpus) {
 			struct cpu_dbs_common_info *j_cdbs =
 				dbs_data->cdata->get_cpu_cdbs(j);
+			unsigned int prev_load;
 
 			j_cdbs->cpu = j;
 			j_cdbs->cur_policy = policy;
 			j_cdbs->prev_cpu_idle = get_cpu_idle_time(j,
 					       &j_cdbs->prev_cpu_wall, io_busy);
+
+			prev_load = (unsigned int)
+				(j_cdbs->prev_cpu_wall - j_cdbs->prev_cpu_idle);
+			j_cdbs->prev_load = 100 * prev_load /
+					(unsigned int) j_cdbs->prev_cpu_wall;
+			j_cdbs->copy_prev_load = true;
+
 			if (ignore_nice)
 				j_cdbs->prev_cpu_nice =
 					kcpustat_cpu(j).cpustat[CPUTIME_NICE];
diff --git a/drivers/cpufreq/cpufreq_governor.h b/drivers/cpufreq/cpufreq_governor.h
index bfb9ae1..c2a5b7e 100644
--- a/drivers/cpufreq/cpufreq_governor.h
+++ b/drivers/cpufreq/cpufreq_governor.h
@@ -134,6 +134,12 @@ struct cpu_dbs_common_info {
 	u64 prev_cpu_idle;
 	u64 prev_cpu_wall;
 	u64 prev_cpu_nice;
+	unsigned int prev_load;
+	/*
+	 * Flag to ensure that we copy the previous load only once, upon the
+	 * first wake-up from idle.
+	 */
+	bool copy_prev_load;
 	struct cpufreq_policy *cur_policy;
 	struct delayed_work work;
 	/*