[PATCH] cpufreq: Record stats when fast switching is enabled

[PATCH] cpufreq: Record stats when fast switching is enabled

[Matthias Kaehlcke]

When fast switching is enabled currently no cpufreq stats are
recorded and the corresponding sysfs attributes appear empty (see
also commit 1aefc75b2449 ("cpufreq: stats: Make the stats code
non-modular")).

Record the stats after a successful fast switch and re-enable access
through sysfs when fast switching is enabled. Since
cpufreq_stats_update() can now be called in interrupt context (during
a fast switch) disable local IRQs while holding the stats spinlock.

Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
---
The change is so simple that I wonder if I'm missing some important
reason why the stats can't/shouldn't be updated during/after a fast
switch ...

I would expect that holding the stats spinlock briefly in
cpufreq_stats_update() shouldn't be a problem. In theory it would
also be an option to have a per stats lock, though it seems overkill
from my (possibly ignorant) point of view.
---
 drivers/cpufreq/cpufreq.c       |  8 +++++++-
 drivers/cpufreq/cpufreq_stats.c | 11 +++--------
 2 files changed, 10 insertions(+), 9 deletions(-)

diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
index e35a886e00bcf..63aadb0bbddfe 100644
--- a/drivers/cpufreq/cpufreq.c
+++ b/drivers/cpufreq/cpufreq.c
@@ -1857,9 +1857,15 @@ EXPORT_SYMBOL(cpufreq_unregister_notifier);
 unsigned int cpufreq_driver_fast_switch(struct cpufreq_policy *policy,
 					unsigned int target_freq)
 {
+	unsigned int freq;
+
 	target_freq = clamp_val(target_freq, policy->min, policy->max);
 
-	return cpufreq_driver->fast_switch(policy, target_freq);
+	freq = cpufreq_driver->fast_switch(policy, target_freq);
+	if (freq)
+		cpufreq_stats_record_transition(policy, freq);
+
+	return freq;
 }
 EXPORT_SYMBOL_GPL(cpufreq_driver_fast_switch);
 
diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c
index 1572129844a5b..21b919bfaeccf 100644
--- a/drivers/cpufreq/cpufreq_stats.c
+++ b/drivers/cpufreq/cpufreq_stats.c
@@ -30,11 +30,12 @@ struct cpufreq_stats {
 static void cpufreq_stats_update(struct cpufreq_stats *stats)
 {
 	unsigned long long cur_time = get_jiffies_64();
+	unsigned long flags;
 
-	spin_lock(&cpufreq_stats_lock);
+	spin_lock_irqsave(&cpufreq_stats_lock, flags);
 	stats->time_in_state[stats->last_index] += cur_time - stats->last_time;
 	stats->last_time = cur_time;
-	spin_unlock(&cpufreq_stats_lock);
+	spin_unlock_irqrestore(&cpufreq_stats_lock, flags);
 }
 
 static void cpufreq_stats_clear_table(struct cpufreq_stats *stats)
@@ -58,9 +59,6 @@ static ssize_t show_time_in_state(struct cpufreq_policy *policy, char *buf)
 	ssize_t len = 0;
 	int i;
 
-	if (policy->fast_switch_enabled)
-		return 0;
-
 	cpufreq_stats_update(stats);
 	for (i = 0; i < stats->state_num; i++) {
 		len += sprintf(buf + len, "%u %llu\n", stats->freq_table[i],
@@ -84,9 +82,6 @@ static ssize_t show_trans_table(struct cpufreq_policy *policy, char *buf)
 	ssize_t len = 0;
 	int i, j;
 
-	if (policy->fast_switch_enabled)
-		return 0;
-
 	len += snprintf(buf + len, PAGE_SIZE - len, "   From  :    To\n");
 	len += snprintf(buf + len, PAGE_SIZE - len, "         : ");
 	for (i = 0; i < stats->state_num; i++) {
-- 
2.20.1.495.gaa96b0ce6b-goog

Re: [PATCH] cpufreq: Record stats when fast switching is enabled

[Viresh Kumar]

On 30-01-19, 17:51, Matthias Kaehlcke wrote:
> When fast switching is enabled currently no cpufreq stats are
> recorded and the corresponding sysfs attributes appear empty (see
> also commit 1aefc75b2449 ("cpufreq: stats: Make the stats code
> non-modular")).
> 
> Record the stats after a successful fast switch and re-enable access
> through sysfs when fast switching is enabled. Since
> cpufreq_stats_update() can now be called in interrupt context (during
> a fast switch) disable local IRQs while holding the stats spinlock.
> 
> Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
> ---
> The change is so simple that I wonder if I'm missing some important
> reason why the stats can't/shouldn't be updated during/after a fast
> switch ...
> 
> I would expect that holding the stats spinlock briefly in
> cpufreq_stats_update() shouldn't be a problem. In theory it would
> also be an option to have a per stats lock, though it seems overkill
> from my (possibly ignorant) point of view.
> ---
>  drivers/cpufreq/cpufreq.c       |  8 +++++++-
>  drivers/cpufreq/cpufreq_stats.c | 11 +++--------
>  2 files changed, 10 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> index e35a886e00bcf..63aadb0bbddfe 100644
> --- a/drivers/cpufreq/cpufreq.c
> +++ b/drivers/cpufreq/cpufreq.c
> @@ -1857,9 +1857,15 @@ EXPORT_SYMBOL(cpufreq_unregister_notifier);
>  unsigned int cpufreq_driver_fast_switch(struct cpufreq_policy *policy,
>  					unsigned int target_freq)
>  {
> +	unsigned int freq;
> +
>  	target_freq = clamp_val(target_freq, policy->min, policy->max);
>  
> -	return cpufreq_driver->fast_switch(policy, target_freq);
> +	freq = cpufreq_driver->fast_switch(policy, target_freq);
> +	if (freq)
> +		cpufreq_stats_record_transition(policy, freq);
> +
> +	return freq;
>  }
>  EXPORT_SYMBOL_GPL(cpufreq_driver_fast_switch);
>  
> diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c
> index 1572129844a5b..21b919bfaeccf 100644
> --- a/drivers/cpufreq/cpufreq_stats.c
> +++ b/drivers/cpufreq/cpufreq_stats.c
> @@ -30,11 +30,12 @@ struct cpufreq_stats {
>  static void cpufreq_stats_update(struct cpufreq_stats *stats)
>  {
>  	unsigned long long cur_time = get_jiffies_64();
> +	unsigned long flags;
>  
> -	spin_lock(&cpufreq_stats_lock);
> +	spin_lock_irqsave(&cpufreq_stats_lock, flags);
>  	stats->time_in_state[stats->last_index] += cur_time - stats->last_time;
>  	stats->last_time = cur_time;
> -	spin_unlock(&cpufreq_stats_lock);
> +	spin_unlock_irqrestore(&cpufreq_stats_lock, flags);
>  }

The only problem that I can think of (or recall) is that this routine
also gets called when time_in_state sysfs file is read and that can
end up taking lock which the scheduler's hotpath will wait for.

-- 
viresh

Re: [PATCH] cpufreq: Record stats when fast switching is enabled

[Rafael J. Wysocki]

On Thu, Jan 31, 2019 at 9:30 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
>
> On 30-01-19, 17:51, Matthias Kaehlcke wrote:
> > When fast switching is enabled currently no cpufreq stats are
> > recorded and the corresponding sysfs attributes appear empty (see
> > also commit 1aefc75b2449 ("cpufreq: stats: Make the stats code
> > non-modular")).
> >
> > Record the stats after a successful fast switch and re-enable access
> > through sysfs when fast switching is enabled. Since
> > cpufreq_stats_update() can now be called in interrupt context (during
> > a fast switch) disable local IRQs while holding the stats spinlock.
> >
> > Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
> > ---
> > The change is so simple that I wonder if I'm missing some important
> > reason why the stats can't/shouldn't be updated during/after a fast
> > switch ...
> >
> > I would expect that holding the stats spinlock briefly in
> > cpufreq_stats_update() shouldn't be a problem. In theory it would
> > also be an option to have a per stats lock, though it seems overkill
> > from my (possibly ignorant) point of view.
> > ---
> >  drivers/cpufreq/cpufreq.c       |  8 +++++++-
> >  drivers/cpufreq/cpufreq_stats.c | 11 +++--------
> >  2 files changed, 10 insertions(+), 9 deletions(-)
> >
> > diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> > index e35a886e00bcf..63aadb0bbddfe 100644
> > --- a/drivers/cpufreq/cpufreq.c
> > +++ b/drivers/cpufreq/cpufreq.c
> > @@ -1857,9 +1857,15 @@ EXPORT_SYMBOL(cpufreq_unregister_notifier);
> >  unsigned int cpufreq_driver_fast_switch(struct cpufreq_policy *policy,
> >                                       unsigned int target_freq)
> >  {
> > +     unsigned int freq;
> > +
> >       target_freq = clamp_val(target_freq, policy->min, policy->max);
> >
> > -     return cpufreq_driver->fast_switch(policy, target_freq);
> > +     freq = cpufreq_driver->fast_switch(policy, target_freq);
> > +     if (freq)
> > +             cpufreq_stats_record_transition(policy, freq);
> > +
> > +     return freq;
> >  }
> >  EXPORT_SYMBOL_GPL(cpufreq_driver_fast_switch);
> >
> > diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c
> > index 1572129844a5b..21b919bfaeccf 100644
> > --- a/drivers/cpufreq/cpufreq_stats.c
> > +++ b/drivers/cpufreq/cpufreq_stats.c
> > @@ -30,11 +30,12 @@ struct cpufreq_stats {
> >  static void cpufreq_stats_update(struct cpufreq_stats *stats)
> >  {
> >       unsigned long long cur_time = get_jiffies_64();
> > +     unsigned long flags;
> >
> > -     spin_lock(&cpufreq_stats_lock);
> > +     spin_lock_irqsave(&cpufreq_stats_lock, flags);
> >       stats->time_in_state[stats->last_index] += cur_time - stats->last_time;
> >       stats->last_time = cur_time;
> > -     spin_unlock(&cpufreq_stats_lock);
> > +     spin_unlock_irqrestore(&cpufreq_stats_lock, flags);
> >  }
>
> The only problem that I can think of (or recall) is that this routine
> also gets called when time_in_state sysfs file is read and that can
> end up taking lock which the scheduler's hotpath will wait for.

What about the extra locking overhead in the scheduler context?

Re: [PATCH] cpufreq: Record stats when fast switching is enabled

[Viresh Kumar]

On 31-01-19, 11:03, Rafael J. Wysocki wrote:
> On Thu, Jan 31, 2019 at 9:30 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> >
> > On 30-01-19, 17:51, Matthias Kaehlcke wrote:
> > > When fast switching is enabled currently no cpufreq stats are
> > > recorded and the corresponding sysfs attributes appear empty (see
> > > also commit 1aefc75b2449 ("cpufreq: stats: Make the stats code
> > > non-modular")).
> > >
> > > Record the stats after a successful fast switch and re-enable access
> > > through sysfs when fast switching is enabled. Since
> > > cpufreq_stats_update() can now be called in interrupt context (during
> > > a fast switch) disable local IRQs while holding the stats spinlock.
> > >
> > > Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
> > > ---
> > > The change is so simple that I wonder if I'm missing some important
> > > reason why the stats can't/shouldn't be updated during/after a fast
> > > switch ...
> > >
> > > I would expect that holding the stats spinlock briefly in
> > > cpufreq_stats_update() shouldn't be a problem. In theory it would
> > > also be an option to have a per stats lock, though it seems overkill
> > > from my (possibly ignorant) point of view.
> > > ---
> > >  drivers/cpufreq/cpufreq.c       |  8 +++++++-
> > >  drivers/cpufreq/cpufreq_stats.c | 11 +++--------
> > >  2 files changed, 10 insertions(+), 9 deletions(-)
> > >
> > > diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> > > index e35a886e00bcf..63aadb0bbddfe 100644
> > > --- a/drivers/cpufreq/cpufreq.c
> > > +++ b/drivers/cpufreq/cpufreq.c
> > > @@ -1857,9 +1857,15 @@ EXPORT_SYMBOL(cpufreq_unregister_notifier);
> > >  unsigned int cpufreq_driver_fast_switch(struct cpufreq_policy *policy,
> > >                                       unsigned int target_freq)
> > >  {
> > > +     unsigned int freq;
> > > +
> > >       target_freq = clamp_val(target_freq, policy->min, policy->max);
> > >
> > > -     return cpufreq_driver->fast_switch(policy, target_freq);
> > > +     freq = cpufreq_driver->fast_switch(policy, target_freq);
> > > +     if (freq)
> > > +             cpufreq_stats_record_transition(policy, freq);
> > > +
> > > +     return freq;
> > >  }
> > >  EXPORT_SYMBOL_GPL(cpufreq_driver_fast_switch);
> > >
> > > diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c
> > > index 1572129844a5b..21b919bfaeccf 100644
> > > --- a/drivers/cpufreq/cpufreq_stats.c
> > > +++ b/drivers/cpufreq/cpufreq_stats.c
> > > @@ -30,11 +30,12 @@ struct cpufreq_stats {
> > >  static void cpufreq_stats_update(struct cpufreq_stats *stats)
> > >  {
> > >       unsigned long long cur_time = get_jiffies_64();
> > > +     unsigned long flags;
> > >
> > > -     spin_lock(&cpufreq_stats_lock);
> > > +     spin_lock_irqsave(&cpufreq_stats_lock, flags);
> > >       stats->time_in_state[stats->last_index] += cur_time - stats->last_time;
> > >       stats->last_time = cur_time;
> > > -     spin_unlock(&cpufreq_stats_lock);
> > > +     spin_unlock_irqrestore(&cpufreq_stats_lock, flags);
> > >  }
> >
> > The only problem that I can think of (or recall) is that this routine
> > also gets called when time_in_state sysfs file is read and that can
> > end up taking lock which the scheduler's hotpath will wait for.
> 
> What about the extra locking overhead in the scheduler context?

What about using READ_ONCE/WRITE_ONCE here ? Not sure if we really
need locking in this particular case.

-- 
viresh

Re: [PATCH] cpufreq: Record stats when fast switching is enabled

[Rafael J. Wysocki]

On Thu, Jan 31, 2019 at 11:07 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
>
> On 31-01-19, 11:03, Rafael J. Wysocki wrote:
> > On Thu, Jan 31, 2019 at 9:30 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> > >
> > > On 30-01-19, 17:51, Matthias Kaehlcke wrote:
> > > > When fast switching is enabled currently no cpufreq stats are
> > > > recorded and the corresponding sysfs attributes appear empty (see
> > > > also commit 1aefc75b2449 ("cpufreq: stats: Make the stats code
> > > > non-modular")).
> > > >
> > > > Record the stats after a successful fast switch and re-enable access
> > > > through sysfs when fast switching is enabled. Since
> > > > cpufreq_stats_update() can now be called in interrupt context (during
> > > > a fast switch) disable local IRQs while holding the stats spinlock.
> > > >
> > > > Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
> > > > ---
> > > > The change is so simple that I wonder if I'm missing some important
> > > > reason why the stats can't/shouldn't be updated during/after a fast
> > > > switch ...
> > > >
> > > > I would expect that holding the stats spinlock briefly in
> > > > cpufreq_stats_update() shouldn't be a problem. In theory it would
> > > > also be an option to have a per stats lock, though it seems overkill
> > > > from my (possibly ignorant) point of view.
> > > > ---
> > > >  drivers/cpufreq/cpufreq.c       |  8 +++++++-
> > > >  drivers/cpufreq/cpufreq_stats.c | 11 +++--------
> > > >  2 files changed, 10 insertions(+), 9 deletions(-)
> > > >
> > > > diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> > > > index e35a886e00bcf..63aadb0bbddfe 100644
> > > > --- a/drivers/cpufreq/cpufreq.c
> > > > +++ b/drivers/cpufreq/cpufreq.c
> > > > @@ -1857,9 +1857,15 @@ EXPORT_SYMBOL(cpufreq_unregister_notifier);
> > > >  unsigned int cpufreq_driver_fast_switch(struct cpufreq_policy *policy,
> > > >                                       unsigned int target_freq)
> > > >  {
> > > > +     unsigned int freq;
> > > > +
> > > >       target_freq = clamp_val(target_freq, policy->min, policy->max);
> > > >
> > > > -     return cpufreq_driver->fast_switch(policy, target_freq);
> > > > +     freq = cpufreq_driver->fast_switch(policy, target_freq);
> > > > +     if (freq)
> > > > +             cpufreq_stats_record_transition(policy, freq);
> > > > +
> > > > +     return freq;
> > > >  }
> > > >  EXPORT_SYMBOL_GPL(cpufreq_driver_fast_switch);
> > > >
> > > > diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c
> > > > index 1572129844a5b..21b919bfaeccf 100644
> > > > --- a/drivers/cpufreq/cpufreq_stats.c
> > > > +++ b/drivers/cpufreq/cpufreq_stats.c
> > > > @@ -30,11 +30,12 @@ struct cpufreq_stats {
> > > >  static void cpufreq_stats_update(struct cpufreq_stats *stats)
> > > >  {
> > > >       unsigned long long cur_time = get_jiffies_64();
> > > > +     unsigned long flags;
> > > >
> > > > -     spin_lock(&cpufreq_stats_lock);
> > > > +     spin_lock_irqsave(&cpufreq_stats_lock, flags);
> > > >       stats->time_in_state[stats->last_index] += cur_time - stats->last_time;
> > > >       stats->last_time = cur_time;
> > > > -     spin_unlock(&cpufreq_stats_lock);
> > > > +     spin_unlock_irqrestore(&cpufreq_stats_lock, flags);
> > > >  }
> > >
> > > The only problem that I can think of (or recall) is that this routine
> > > also gets called when time_in_state sysfs file is read and that can
> > > end up taking lock which the scheduler's hotpath will wait for.
> >
> > What about the extra locking overhead in the scheduler context?
>
> What about using READ_ONCE/WRITE_ONCE here ? Not sure if we really
> need locking in this particular case.

If that works, then fine, but ISTR some synchronization issues related to that.

Re: [PATCH] cpufreq: Record stats when fast switching is enabled

[Matthias Kaehlcke]

On Thu, Jan 31, 2019 at 11:14:03AM +0100, Rafael J. Wysocki wrote:
> On Thu, Jan 31, 2019 at 11:07 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> >
> > On 31-01-19, 11:03, Rafael J. Wysocki wrote:
> > > On Thu, Jan 31, 2019 at 9:30 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> > > >
> > > > On 30-01-19, 17:51, Matthias Kaehlcke wrote:
> > > > > When fast switching is enabled currently no cpufreq stats are
> > > > > recorded and the corresponding sysfs attributes appear empty (see
> > > > > also commit 1aefc75b2449 ("cpufreq: stats: Make the stats code
> > > > > non-modular")).
> > > > >
> > > > > Record the stats after a successful fast switch and re-enable access
> > > > > through sysfs when fast switching is enabled. Since
> > > > > cpufreq_stats_update() can now be called in interrupt context (during
> > > > > a fast switch) disable local IRQs while holding the stats spinlock.
> > > > >
> > > > > Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
> > > > > ---
> > > > > The change is so simple that I wonder if I'm missing some important
> > > > > reason why the stats can't/shouldn't be updated during/after a fast
> > > > > switch ...
> > > > >
> > > > > I would expect that holding the stats spinlock briefly in
> > > > > cpufreq_stats_update() shouldn't be a problem. In theory it would
> > > > > also be an option to have a per stats lock, though it seems overkill
> > > > > from my (possibly ignorant) point of view.
> > > > > ---
> > > > >  drivers/cpufreq/cpufreq.c       |  8 +++++++-
> > > > >  drivers/cpufreq/cpufreq_stats.c | 11 +++--------
> > > > >  2 files changed, 10 insertions(+), 9 deletions(-)
> > > > >
> > > > > diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> > > > > index e35a886e00bcf..63aadb0bbddfe 100644
> > > > > --- a/drivers/cpufreq/cpufreq.c
> > > > > +++ b/drivers/cpufreq/cpufreq.c
> > > > > @@ -1857,9 +1857,15 @@ EXPORT_SYMBOL(cpufreq_unregister_notifier);
> > > > >  unsigned int cpufreq_driver_fast_switch(struct cpufreq_policy *policy,
> > > > >                                       unsigned int target_freq)
> > > > >  {
> > > > > +     unsigned int freq;
> > > > > +
> > > > >       target_freq = clamp_val(target_freq, policy->min, policy->max);
> > > > >
> > > > > -     return cpufreq_driver->fast_switch(policy, target_freq);
> > > > > +     freq = cpufreq_driver->fast_switch(policy, target_freq);
> > > > > +     if (freq)
> > > > > +             cpufreq_stats_record_transition(policy, freq);
> > > > > +
> > > > > +     return freq;
> > > > >  }
> > > > >  EXPORT_SYMBOL_GPL(cpufreq_driver_fast_switch);
> > > > >
> > > > > diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c
> > > > > index 1572129844a5b..21b919bfaeccf 100644
> > > > > --- a/drivers/cpufreq/cpufreq_stats.c
> > > > > +++ b/drivers/cpufreq/cpufreq_stats.c
> > > > > @@ -30,11 +30,12 @@ struct cpufreq_stats {
> > > > >  static void cpufreq_stats_update(struct cpufreq_stats *stats)
> > > > >  {
> > > > >       unsigned long long cur_time = get_jiffies_64();
> > > > > +     unsigned long flags;
> > > > >
> > > > > -     spin_lock(&cpufreq_stats_lock);
> > > > > +     spin_lock_irqsave(&cpufreq_stats_lock, flags);
> > > > >       stats->time_in_state[stats->last_index] += cur_time - stats->last_time;
> > > > >       stats->last_time = cur_time;
> > > > > -     spin_unlock(&cpufreq_stats_lock);
> > > > > +     spin_unlock_irqrestore(&cpufreq_stats_lock, flags);
> > > > >  }
> > > >
> > > > The only problem that I can think of (or recall) is that this routine
> > > > also gets called when time_in_state sysfs file is read and that can
> > > > end up taking lock which the scheduler's hotpath will wait for.
> > >
> > > What about the extra locking overhead in the scheduler context?
> >
> > What about using READ_ONCE/WRITE_ONCE here ? Not sure if we really
> > need locking in this particular case.
> 
> If that works, then fine, but ISTR some synchronization issues related to that.

I also think there would be synchronization issues :(

Is your main concern with the spin lock the contention case or the
general overhead of locking?

It would be really nice to have cpufreq stats with schedutil. We
initially considered a sysfs attribute to allow to temporarily disable
fast switching, but at closer sight this seems messy (would require
quite some rework in cpufreq_schedutil.c), besides not recording the
actual behavior.

If another (rarely and only shortly held) lock in scheduler context is
a no-go deferred recording could be an option, if that can be
implemented without locks in scheduler context.

Re: [PATCH] cpufreq: Record stats when fast switching is enabled

[Rafael J. Wysocki]

On Thursday, January 31, 2019 7:37:30 PM CET Matthias Kaehlcke wrote:
> On Thu, Jan 31, 2019 at 11:14:03AM +0100, Rafael J. Wysocki wrote:
> > On Thu, Jan 31, 2019 at 11:07 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> > >
> > > On 31-01-19, 11:03, Rafael J. Wysocki wrote:
> > > > On Thu, Jan 31, 2019 at 9:30 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> > > > >
> > > > > On 30-01-19, 17:51, Matthias Kaehlcke wrote:
> > > > > > When fast switching is enabled currently no cpufreq stats are
> > > > > > recorded and the corresponding sysfs attributes appear empty (see
> > > > > > also commit 1aefc75b2449 ("cpufreq: stats: Make the stats code
> > > > > > non-modular")).
> > > > > >
> > > > > > Record the stats after a successful fast switch and re-enable access
> > > > > > through sysfs when fast switching is enabled. Since
> > > > > > cpufreq_stats_update() can now be called in interrupt context (during
> > > > > > a fast switch) disable local IRQs while holding the stats spinlock.
> > > > > >
> > > > > > Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
> > > > > > ---
> > > > > > The change is so simple that I wonder if I'm missing some important
> > > > > > reason why the stats can't/shouldn't be updated during/after a fast
> > > > > > switch ...
> > > > > >
> > > > > > I would expect that holding the stats spinlock briefly in
> > > > > > cpufreq_stats_update() shouldn't be a problem. In theory it would
> > > > > > also be an option to have a per stats lock, though it seems overkill
> > > > > > from my (possibly ignorant) point of view.
> > > > > > ---
> > > > > >  drivers/cpufreq/cpufreq.c       |  8 +++++++-
> > > > > >  drivers/cpufreq/cpufreq_stats.c | 11 +++--------
> > > > > >  2 files changed, 10 insertions(+), 9 deletions(-)
> > > > > >
> > > > > > diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> > > > > > index e35a886e00bcf..63aadb0bbddfe 100644
> > > > > > --- a/drivers/cpufreq/cpufreq.c
> > > > > > +++ b/drivers/cpufreq/cpufreq.c
> > > > > > @@ -1857,9 +1857,15 @@ EXPORT_SYMBOL(cpufreq_unregister_notifier);
> > > > > >  unsigned int cpufreq_driver_fast_switch(struct cpufreq_policy *policy,
> > > > > >                                       unsigned int target_freq)
> > > > > >  {
> > > > > > +     unsigned int freq;
> > > > > > +
> > > > > >       target_freq = clamp_val(target_freq, policy->min, policy->max);
> > > > > >
> > > > > > -     return cpufreq_driver->fast_switch(policy, target_freq);
> > > > > > +     freq = cpufreq_driver->fast_switch(policy, target_freq);
> > > > > > +     if (freq)
> > > > > > +             cpufreq_stats_record_transition(policy, freq);
> > > > > > +
> > > > > > +     return freq;
> > > > > >  }
> > > > > >  EXPORT_SYMBOL_GPL(cpufreq_driver_fast_switch);
> > > > > >
> > > > > > diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c
> > > > > > index 1572129844a5b..21b919bfaeccf 100644
> > > > > > --- a/drivers/cpufreq/cpufreq_stats.c
> > > > > > +++ b/drivers/cpufreq/cpufreq_stats.c
> > > > > > @@ -30,11 +30,12 @@ struct cpufreq_stats {
> > > > > >  static void cpufreq_stats_update(struct cpufreq_stats *stats)
> > > > > >  {
> > > > > >       unsigned long long cur_time = get_jiffies_64();
> > > > > > +     unsigned long flags;
> > > > > >
> > > > > > -     spin_lock(&cpufreq_stats_lock);
> > > > > > +     spin_lock_irqsave(&cpufreq_stats_lock, flags);
> > > > > >       stats->time_in_state[stats->last_index] += cur_time - stats->last_time;
> > > > > >       stats->last_time = cur_time;
> > > > > > -     spin_unlock(&cpufreq_stats_lock);
> > > > > > +     spin_unlock_irqrestore(&cpufreq_stats_lock, flags);
> > > > > >  }
> > > > >
> > > > > The only problem that I can think of (or recall) is that this routine
> > > > > also gets called when time_in_state sysfs file is read and that can
> > > > > end up taking lock which the scheduler's hotpath will wait for.
> > > >
> > > > What about the extra locking overhead in the scheduler context?
> > >
> > > What about using READ_ONCE/WRITE_ONCE here ? Not sure if we really
> > > need locking in this particular case.
> > 
> > If that works, then fine, but ISTR some synchronization issues related to that.
> 
> I also think there would be synchronization issues :(
> 
> Is your main concern with the spin lock the contention case or the
> general overhead of locking?

The general overhead is bad enough.  The contention case would be a
disaster.

> It would be really nice to have cpufreq stats with schedutil. We
> initially considered a sysfs attribute to allow to temporarily disable
> fast switching, but at closer sight this seems messy (would require
> quite some rework in cpufreq_schedutil.c), besides not recording the
> actual behavior.
> 
> If another (rarely and only shortly held) lock in scheduler context

This is a global spinlock and you'd like to take it on every frequency
change for each policy.  On x86, as a rule, there is a policy per logical
CPU and systems with hundreds of these are not uncommon.  Come on.

> is a no-go deferred recording could be an option, if that can be
> implemented without locks in scheduler context.

Why do you need the stats at all in the fast switch case?

There is the cpu_frequency tracepoint that can be used to callect
all data that you need.  Why can't that be used?

Re: [PATCH] cpufreq: Record stats when fast switching is enabled

[Matthias Kaehlcke]

On Fri, Feb 01, 2019 at 12:34:32AM +0100, Rafael J. Wysocki wrote:
> On Thursday, January 31, 2019 7:37:30 PM CET Matthias Kaehlcke wrote:
> > On Thu, Jan 31, 2019 at 11:14:03AM +0100, Rafael J. Wysocki wrote:
> > > On Thu, Jan 31, 2019 at 11:07 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> > > >
> > > > On 31-01-19, 11:03, Rafael J. Wysocki wrote:
> > > > > On Thu, Jan 31, 2019 at 9:30 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> > > > > >
> > > > > > On 30-01-19, 17:51, Matthias Kaehlcke wrote:
> > > > > > > When fast switching is enabled currently no cpufreq stats are
> > > > > > > recorded and the corresponding sysfs attributes appear empty (see
> > > > > > > also commit 1aefc75b2449 ("cpufreq: stats: Make the stats code
> > > > > > > non-modular")).
> > > > > > >
> > > > > > > Record the stats after a successful fast switch and re-enable access
> > > > > > > through sysfs when fast switching is enabled. Since
> > > > > > > cpufreq_stats_update() can now be called in interrupt context (during
> > > > > > > a fast switch) disable local IRQs while holding the stats spinlock.
> > > > > > >
> > > > > > > Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
> > > > > > > ---
> > > > > > > The change is so simple that I wonder if I'm missing some important
> > > > > > > reason why the stats can't/shouldn't be updated during/after a fast
> > > > > > > switch ...
> > > > > > >
> > > > > > > I would expect that holding the stats spinlock briefly in
> > > > > > > cpufreq_stats_update() shouldn't be a problem. In theory it would
> > > > > > > also be an option to have a per stats lock, though it seems overkill
> > > > > > > from my (possibly ignorant) point of view.
> > > > > > > ---
> > > > > > >  drivers/cpufreq/cpufreq.c       |  8 +++++++-
> > > > > > >  drivers/cpufreq/cpufreq_stats.c | 11 +++--------
> > > > > > >  2 files changed, 10 insertions(+), 9 deletions(-)
> > > > > > >
> > > > > > > diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> > > > > > > index e35a886e00bcf..63aadb0bbddfe 100644
> > > > > > > --- a/drivers/cpufreq/cpufreq.c
> > > > > > > +++ b/drivers/cpufreq/cpufreq.c
> > > > > > > @@ -1857,9 +1857,15 @@ EXPORT_SYMBOL(cpufreq_unregister_notifier);
> > > > > > >  unsigned int cpufreq_driver_fast_switch(struct cpufreq_policy *policy,
> > > > > > >                                       unsigned int target_freq)
> > > > > > >  {
> > > > > > > +     unsigned int freq;
> > > > > > > +
> > > > > > >       target_freq = clamp_val(target_freq, policy->min, policy->max);
> > > > > > >
> > > > > > > -     return cpufreq_driver->fast_switch(policy, target_freq);
> > > > > > > +     freq = cpufreq_driver->fast_switch(policy, target_freq);
> > > > > > > +     if (freq)
> > > > > > > +             cpufreq_stats_record_transition(policy, freq);
> > > > > > > +
> > > > > > > +     return freq;
> > > > > > >  }
> > > > > > >  EXPORT_SYMBOL_GPL(cpufreq_driver_fast_switch);
> > > > > > >
> > > > > > > diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c
> > > > > > > index 1572129844a5b..21b919bfaeccf 100644
> > > > > > > --- a/drivers/cpufreq/cpufreq_stats.c
> > > > > > > +++ b/drivers/cpufreq/cpufreq_stats.c
> > > > > > > @@ -30,11 +30,12 @@ struct cpufreq_stats {
> > > > > > >  static void cpufreq_stats_update(struct cpufreq_stats *stats)
> > > > > > >  {
> > > > > > >       unsigned long long cur_time = get_jiffies_64();
> > > > > > > +     unsigned long flags;
> > > > > > >
> > > > > > > -     spin_lock(&cpufreq_stats_lock);
> > > > > > > +     spin_lock_irqsave(&cpufreq_stats_lock, flags);
> > > > > > >       stats->time_in_state[stats->last_index] += cur_time - stats->last_time;
> > > > > > >       stats->last_time = cur_time;
> > > > > > > -     spin_unlock(&cpufreq_stats_lock);
> > > > > > > +     spin_unlock_irqrestore(&cpufreq_stats_lock, flags);
> > > > > > >  }
> > > > > >
> > > > > > The only problem that I can think of (or recall) is that this routine
> > > > > > also gets called when time_in_state sysfs file is read and that can
> > > > > > end up taking lock which the scheduler's hotpath will wait for.
> > > > >
> > > > > What about the extra locking overhead in the scheduler context?
> > > >
> > > > What about using READ_ONCE/WRITE_ONCE here ? Not sure if we really
> > > > need locking in this particular case.
> > > 
> > > If that works, then fine, but ISTR some synchronization issues related to that.
> > 
> > I also think there would be synchronization issues :(
> > 
> > Is your main concern with the spin lock the contention case or the
> > general overhead of locking?
> 
> The general overhead is bad enough.  The contention case would be a
> disaster.
> 
> > It would be really nice to have cpufreq stats with schedutil. We
> > initially considered a sysfs attribute to allow to temporarily disable
> > fast switching, but at closer sight this seems messy (would require
> > quite some rework in cpufreq_schedutil.c), besides not recording the
> > actual behavior.
> > 
> > If another (rarely and only shortly held) lock in scheduler context
> 
> This is a global spinlock and you'd like to take it on every frequency
> change for each policy.  On x86, as a rule, there is a policy per logical
> CPU and systems with hundreds of these are not uncommon.  Come on.

Thanks for helping me to get a better understanding of the problem. If
the global spinlock was the main issue, this could be fixed by having
a per stats/policy lock, but it seems there's more than that.

> > is a no-go deferred recording could be an option, if that can be
> > implemented without locks in scheduler context.
> 
> Why do you need the stats at all in the fast switch case?

For the same reason as in the non-fast switch case, easy access to the
stats with existing tooling (or no tooling at all).

> There is the cpu_frequency tracepoint that can be used to callect
> all data that you need.  Why can't that be used?

It could be used, but requires non-standard tooling to process the
data and tracing must be enabled.

Could a CONFIG option make sense to enable it (off by default),
or is the overhead (with a per stats lock) so high that it would be
unreasonable to use it (I really don't have a good sense on this)?

Thanks

Matthias

Re: [PATCH] cpufreq: Record stats when fast switching is enabled

[Danny Lin]

On Thu, Jan 31, 2019 at 2:14 AM, Rafael J. Wysocki wrote:
> On Thu, Jan 31, 2019 at 11:07 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> >
> > On 31-01-19, 11:03, Rafael J. Wysocki wrote:
> > > On Thu, Jan 31, 2019 at 9:30 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> > > >
> > > > The only problem that I can think of (or recall) is that this routine
> > > > also gets called when time_in_state sysfs file is read and that can
> > > > end up taking lock which the scheduler's hotpath will wait for.
> > >
> > > What about the extra locking overhead in the scheduler context?
> >
> > What about using READ_ONCE/WRITE_ONCE here ? Not sure if we really
> > need locking in this particular case.
> 
> If that works, then fine, but ISTR some synchronization issues related to that.

Maybe using READ/WRITE_ONCE for time_in_state is problematic, but is
there any reason why atomics wouldn't work for this? As far as I can
tell, atomics are necessary to protect time_in_state due to its
multi-step add operation, and READ/WRITE_ONCE can be used for last_time
because all operations on it are single-op sets/gets.

I've been using the setup described above on a downstream arm64 4.14
kernel for nearly a year with no issues. I haven't noticed any
significant anomalies in the stats so far. The system in question has 8
CPUs split into 3 cpufreq policies and fast switch is used with the
schedutil governor, so it should be exercising the stats update path
enough.

Sorry for bumping an old thread.

Re: [PATCH] cpufreq: Record stats when fast switching is enabled

[Rafael J. Wysocki]

On Thu, Jan 31, 2019 at 2:51 AM Matthias Kaehlcke <mka@chromium.org> wrote:
>
> When fast switching is enabled currently no cpufreq stats are
> recorded and the corresponding sysfs attributes appear empty (see
> also commit 1aefc75b2449 ("cpufreq: stats: Make the stats code
> non-modular")).
>
> Record the stats after a successful fast switch and re-enable access
> through sysfs when fast switching is enabled. Since
> cpufreq_stats_update() can now be called in interrupt context (during
> a fast switch) disable local IRQs while holding the stats spinlock.
>
> Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
> ---
> The change is so simple that I wonder if I'm missing some important
> reason why the stats can't/shouldn't be updated during/after a fast
> switch ...
>
> I would expect that holding the stats spinlock briefly in
> cpufreq_stats_update() shouldn't be a problem. In theory it would
> also be an option to have a per stats lock, though it seems overkill
> from my (possibly ignorant) point of view.
> ---
>  drivers/cpufreq/cpufreq.c       |  8 +++++++-
>  drivers/cpufreq/cpufreq_stats.c | 11 +++--------
>  2 files changed, 10 insertions(+), 9 deletions(-)
>
> diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> index e35a886e00bcf..63aadb0bbddfe 100644
> --- a/drivers/cpufreq/cpufreq.c
> +++ b/drivers/cpufreq/cpufreq.c
> @@ -1857,9 +1857,15 @@ EXPORT_SYMBOL(cpufreq_unregister_notifier);
>  unsigned int cpufreq_driver_fast_switch(struct cpufreq_policy *policy,
>                                         unsigned int target_freq)
>  {
> +       unsigned int freq;
> +
>         target_freq = clamp_val(target_freq, policy->min, policy->max);
>
> -       return cpufreq_driver->fast_switch(policy, target_freq);
> +       freq = cpufreq_driver->fast_switch(policy, target_freq);
> +       if (freq)
> +               cpufreq_stats_record_transition(policy, freq);

No extra locking in the fast switch path, pretty please.

This runs in the scheduler context and it adds enough overhead in there already.

> +
> +       return freq;
>  }
>  EXPORT_SYMBOL_GPL(cpufreq_driver_fast_switch);