Re: [tip:timers/core] hrtimer: Allow concurrent hrtimer_start() for self restarting timers

[bsegall]

Peter Zijlstra <peterz@xxxxxxxxxxxxx> writes:

> @@ -4038,7 +4041,7 @@ void init_cfs_bandwidth(struct cfs_bandwidth *cfs_b)
>  	cfs_b->period = ns_to_ktime(default_cfs_period());
>  
>  	INIT_LIST_HEAD(&cfs_b->throttled_cfs_rq);
> -	hrtimer_init(&cfs_b->period_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
> +	hrtimer_init(&cfs_b->period_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS_PINNED);
>  	cfs_b->period_timer.function = sched_cfs_period_timer;
>  	hrtimer_init(&cfs_b->slack_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
>  	cfs_b->slack_timer.function = sched_cfs_slack_timer;
> @@ -4052,7 +4055,13 @@ static void init_cfs_rq_runtime(struct cfs_rq *cfs_rq)
>  
>  void start_cfs_bandwidth(struct cfs_bandwidth *cfs_b)
>  {
> -	start_bandwidth_timer(&cfs_b->period_timer, cfs_b->period);
> +	lockdep_assert_held(&cfs_b->lock);
> +
> +	if (!cfs_b->period_active) {
> +		cfs_b->period_active = 1;
> +		hrtimer_forward_now(&cfs_b->period_timer, cfs_b->period);
> +		hrtimer_start_expires(&cfs_b->period_timer, HRTIMER_MODE_ABS_PINNED);
> +	}
>  }
>  

I don't think the cfsb period timer actually needs to be pinned, even
though it has generally run as such due to start_bandwidth_timer.

The period_active code looks fine.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/