Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful works for CONFIG_PREEMPT_NONE

From	Cheng Chao <cs.os.kernel@gmail.com>
Newsgroups	linux.kernel
Subject	Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful works for CONFIG_PREEMPT_NONE
Date	2016-09-13 04:50 +0200
Message-ID	<sgLXj-3VL-1@gated-at.bofh.it> (permalink)
References	<sfpct-qw-1@gated-at.bofh.it> <sfMiJ-6sU-9@gated-at.bofh.it> <sgxhD-2sv-11@gated-at.bofh.it>
Organization	linux.* mail to news gateway

Show all headers | View raw

on 09/12/2016 07:03 PM, Oleg Nesterov wrote:
> On 09/10, Cheng Chao wrote:
>>
>> @@ -126,6 +126,17 @@ int stop_one_cpu(unsigned int cpu, cpu_stop_fn_t fn, void *arg)
>>  	cpu_stop_init_done(&done, 1);
>>  	if (!cpu_stop_queue_work(cpu, &work))
>>  		return -ENOENT;
>> +
>> +#if defined(CONFIG_PREEMPT_NONE)
>> +	/*
>> +	 * Makes the stopper thread run as soon as possible.
>> +	 * And if the caller is TASK_RUNNING, keeps the caller TASK_RUNNING.
>> +	 * It's special useful for some callers which are expected to be
>> +	 * TASK_ON_RQ_QUEUED.
>> +	 * sched_exec does benefit from this improvement.
>> +	 */
>> +	schedule();
>> +#endif
>>  	wait_for_completion(&done.completion);
>>  	return done.ret;
>>  }
> 
> Cheng, I already tried twice to suggest to conditionalize this schedule,
> because it can only help if cpu == smp_processor_id, and you didn't reply.
> I still think _cond_resched() makes more sense.
> 
> I won't really argue if you prefer it this way. But did you see my emails?
>

I read them, thanks. because Peter didn't receive my mails before, it took me much time
to fix my mailbox, so I didn't reply on time.

Ok, even if cpu != smp_processor_id(), to call schedule() instead _cond_resched() can
give the caller a chance not to sleep. when the caller runs on the cpu again, it may 
likely find the completion is already done. 
then the stopper thread cpu_stop_signal_done() and the caller wait_for_completion() will
actually run very soon.

I think it is trivial improvement. using cond_resched()/_cond_resched() is better for 
readability, I choose the cond_resched().

thanks again.

> Oleg.
>

Thread

[PATCH v3] stop_machine: Make migration_cpu_stop() does useful works for CONFIG_PREEMPT_NONE Cheng Chao <cs.os.kernel@gmail.com> - 2016-09-10 11:00 +0200
  Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful  works for CONFIG_PREEMPT_NONE Cheng Chao <cs.os.kernel@gmail.com> - 2016-09-10 12:00 +0200
    Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful  works for CONFIG_PREEMPT_NONE Peter Zijlstra <peterz@infradead.org> - 2016-09-10 18:40 +0200
  Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful         works for CONFIG_PREEMPT_NONE Oleg Nesterov <oleg@redhat.com> - 2016-09-12 13:10 +0200
    Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful  works for CONFIG_PREEMPT_NONE Cheng Chao <cs.os.kernel@gmail.com> - 2016-09-13 04:50 +0200
  Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful  works for CONFIG_PREEMPT_NONE Peter Zijlstra <peterz@infradead.org> - 2016-09-12 13:40 +0200
    Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful  works for CONFIG_PREEMPT_NONE Peter Zijlstra <peterz@infradead.org> - 2016-09-12 13:50 +0200
      Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful         works for CONFIG_PREEMPT_NONE Oleg Nesterov <oleg@redhat.com> - 2016-09-12 15:10 +0200
        Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful  works for CONFIG_PREEMPT_NONE Peter Zijlstra <peterz@infradead.org> - 2016-09-12 17:10 +0200
          Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful         works for CONFIG_PREEMPT_NONE Oleg Nesterov <oleg@redhat.com> - 2016-09-13 18:20 +0200
            Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful  works for CONFIG_PREEMPT_NONE Peter Zijlstra <peterz@infradead.org> - 2016-09-13 18:40 +0200
              Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful  works for CONFIG_PREEMPT_NONE Cheng Chao <cs.os.kernel@gmail.com> - 2016-09-14 04:10 +0200
                Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful  works for CONFIG_PREEMPT_NONE Peter Zijlstra <peterz@infradead.org> - 2016-09-14 10:00 +0200
              Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful         works for CONFIG_PREEMPT_NONE Oleg Nesterov <oleg@redhat.com> - 2016-09-14 17:50 +0200
              [tip:sched/core] sched/core: Optimize __schedule() tip-bot for Peter Zijlstra <tipbot@zytor.com> - 2016-09-22 16:10 +0200
      Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful  works for CONFIG_PREEMPT_NONE Cheng Chao <cs.os.kernel@gmail.com> - 2016-09-13 06:10 +0200
        Re: [PATCH v3] stop_machine: Make migration_cpu_stop() does useful  works for CONFIG_PREEMPT_NONE Peter Zijlstra <peterz@infradead.org> - 2016-09-13 10:50 +0200
  [PATCH v4] stop_machine: Avoid a sleep and wakeup in the stop_one_cpu() Cheng Chao <cs.os.kernel@gmail.com> - 2016-09-14 04:10 +0200
    Re: [PATCH v4] stop_machine: Avoid a sleep and wakeup in the         stop_one_cpu() Oleg Nesterov <oleg@redhat.com> - 2016-09-14 18:00 +0200
    Re: [PATCH v4] stop_machine: Avoid a sleep and wakeup in the  stop_one_cpu() Cheng Chao <cs.os.kernel@gmail.com> - 2016-09-18 04:10 +0200
    [tip:sched/core] stop_machine: Avoid a sleep and wakeup in  stop_one_cpu() tip-bot for Cheng Chao <tipbot@zytor.com> - 2016-09-22 16:10 +0200

csiph-web