Re: [Patch v5 3/6] sched/fair: Enable periodic update of average thermal pressure

[Thara Gopinath <thara.gopinath@linaro.org>]

On 11/06/2019 03:32 AM, Vincent Guittot wrote:
> On Tue, 5 Nov 2019 at 19:49, Thara Gopinath <thara.gopinath@linaro.org> wrote:
>>
>> Introduce support in CFS periodic tick and other bookkeeping apis
>> to trigger the process of computing average thermal pressure for a
>> cpu. Also consider avg_thermal.load_avg in others_have_blocked
>> which allows for decay of pelt signals.
>>
>> Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
>> ---
>>
>> v4->v5:
>>         - Updated both versions of update_blocked_averages to trigger the
>>           process of computing average thermal pressure.
>>         - Updated others_have_blocked to considerd avg_thermal.load_avg.
>>
>>  kernel/sched/fair.c | 10 ++++++++++
>>  1 file changed, 10 insertions(+)
>>
>> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
>> index 2e907cc..9fb0494 100644
>> --- a/kernel/sched/fair.c
>> +++ b/kernel/sched/fair.c
>> @@ -92,6 +92,8 @@ const_debug unsigned int sysctl_sched_migration_cost  = 500000UL;
>>   */
>>  static DEFINE_PER_CPU(unsigned long, thermal_pressure);
>>
>> +static void trigger_thermal_pressure_average(struct rq *rq);
>> +
>>  #ifdef CONFIG_SMP
>>  /*
>>   * For asym packing, by default the lower numbered CPU has higher priority.
>> @@ -7493,6 +7495,9 @@ static inline bool others_have_blocked(struct rq *rq)
>>         if (READ_ONCE(rq->avg_dl.util_avg))
>>                 return true;
>>
>> +       if (READ_ONCE(rq->avg_thermal.load_avg))
>> +               return true;
>> +
>>  #ifdef CONFIG_HAVE_SCHED_AVG_IRQ
>>         if (READ_ONCE(rq->avg_irq.util_avg))
>>                 return true;
>> @@ -7580,6 +7585,8 @@ static void update_blocked_averages(int cpu)
>>                 done = false;
>>
>>         update_blocked_load_status(rq, !done);
>> +
>> +       trigger_thermal_pressure_average(rq);
> 
> This must be called before others_have_blocked() to take into account
> the latest update

It is a bug. I agree. Will fix it.
> 
>>         rq_unlock_irqrestore(rq, &rf);
>>  }
>>
>> @@ -7646,6 +7653,7 @@ static inline void update_blocked_averages(int cpu)
>>         update_dl_rq_load_avg(rq_clock_pelt(rq), rq, curr_class == &dl_sched_class);
>>         update_irq_load_avg(rq, 0);
>>         update_blocked_load_status(rq, cfs_rq_has_blocked(cfs_rq) || others_have_blocked(rq));
>> +       trigger_thermal_pressure_average(rq);
> 
> idem
> 
>>         rq_unlock_irqrestore(rq, &rf);
>>  }
>>
>> @@ -9939,6 +9947,8 @@ static void task_tick_fair(struct rq *rq, struct task_struct *curr, int queued)
>>
>>         update_misfit_status(curr, rq);
>>         update_overutilized_status(task_rq(curr));
>> +
> 
> remove blank line
> 
>> +       trigger_thermal_pressure_average(rq);
>>  }
>>
>>  /*
>> --
>> 2.1.4
>>


-- 
Warm Regards
Thara