This is a note to let you know that I've just added the patch titled
intel_pstate: add sample time scaling
to the 3.14-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary
The filename of the patch is:
intel_pstate-add-sample-time-scaling.patch
and it can be found in the queue-3.14 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <[email protected]> know about it.
>From c4ee841f602e5eef8eab673295c49c5b49d7732b Mon Sep 17 00:00:00 2001
From: Dirk Brandewie <[email protected]>
Date: Thu, 29 May 2014 09:32:24 -0700
Subject: intel_pstate: add sample time scaling
From: Dirk Brandewie <[email protected]>
commit c4ee841f602e5eef8eab673295c49c5b49d7732b upstream.
The PID assumes that samples are of equal time, which for a deferable
timers this is not true when the system goes idle. This causes the
PID to take a long time to converge to the min P state and depending
on the pattern of the idle load can make the P state appear stuck.
The hold-off value of three sample times before using the scaling is
to give a grace period for applications that have high performance
requirements and spend a lot of time idle, The poster child for this
behavior is the ffmpeg benchmark in the Phoronix test suite.
Signed-off-by: Dirk Brandewie <[email protected]>
Signed-off-by: Rafael J. Wysocki <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
---
drivers/cpufreq/intel_pstate.c | 18 +++++++++++++++++-
1 file changed, 17 insertions(+), 1 deletion(-)
--- a/drivers/cpufreq/intel_pstate.c
+++ b/drivers/cpufreq/intel_pstate.c
@@ -60,6 +60,7 @@ struct sample {
u64 aperf;
u64 mperf;
int freq;
+ ktime_t time;
};
struct pstate_data {
@@ -97,6 +98,7 @@ struct cpudata {
struct vid_data vid;
struct _pid pid;
+ ktime_t last_sample_time;
u64 prev_aperf;
u64 prev_mperf;
struct sample sample;
@@ -583,6 +585,8 @@ static inline void intel_pstate_sample(s
aperf = aperf >> FRAC_BITS;
mperf = mperf >> FRAC_BITS;
+ cpu->last_sample_time = cpu->sample.time;
+ cpu->sample.time = ktime_get();
cpu->sample.aperf = aperf;
cpu->sample.mperf = mperf;
cpu->sample.aperf -= cpu->prev_aperf;
@@ -605,12 +609,24 @@ static inline void intel_pstate_set_samp
static inline int32_t intel_pstate_get_scaled_busy(struct cpudata *cpu)
{
- int32_t core_busy, max_pstate, current_pstate;
+ int32_t core_busy, max_pstate, current_pstate, sample_ratio;
+ u32 duration_us;
+ u32 sample_time;
core_busy = cpu->sample.core_pct_busy;
max_pstate = int_tofp(cpu->pstate.max_pstate);
current_pstate = int_tofp(cpu->pstate.current_pstate);
core_busy = mul_fp(core_busy, div_fp(max_pstate, current_pstate));
+
+ sample_time = (pid_params.sample_rate_ms * USEC_PER_MSEC);
+ duration_us = (u32) ktime_us_delta(cpu->sample.time,
+ cpu->last_sample_time);
+ if (duration_us > sample_time * 3) {
+ sample_ratio = div_fp(int_tofp(sample_time),
+ int_tofp(duration_us));
+ core_busy = mul_fp(core_busy, sample_ratio);
+ }
+
return core_busy;
}
Patches currently in stable-queue which might be from
[email protected] are
queue-3.14/intel_pstate-remove-unneeded-sample-buffers.patch
queue-3.14/intel_pstate-improve-initial-busy-calculation.patch
queue-3.14/intel_pstate-remove-c0-tracking.patch
queue-3.14/intel_pstate-correct-rounding-in-busy-calculation.patch
queue-3.14/intel_pstate-add-sample-time-scaling.patch
--
To unsubscribe from this list: send the line "unsubscribe stable" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
