Date: Thu, 20 Feb 2014 16:22:39 -0500
From: Steven Rostedt <rostedt@goodmis.org>
To: Kirill Tkhai <tkhai@yandex.ru>
Cc: "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        Juri Lelli <juri.lelli@gmail.com>,
        Peter Zijlstra <peterz@infradead.org>, Ingo Molnar <mingo@redhat.com>
Subject: Re: [RFC] sched/deadline: Prevent rt_time growth to infinity
Message-ID: <20140220162239.3efe86a7@gandalf.local.home>
In-Reply-To: <230991392848160@web13m.yandex.ru>
References: <230991392848160@web13m.yandex.ru>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org

Peter, Juri,

Have you seen this?

-- Steve


On Thu, 20 Feb 2014 02:16:00 +0400
Kirill Tkhai <tkhai@yandex.ru> wrote:

> Since deadline tasks share rt bandwidth, we must care about
> bandwidth timer set. Otherwise rt_time may grow up to infinity
> in update_curr_dl(), if there are no other available RT tasks
> on top level bandwidth.
> 
> I'm going to decide the problem the way below. Almost untested
> because of I skipped almost all of recent patches which haveto be applied from lkml.
> 
> Please say, if I skipped anything in idea. Maybe better put
> start_top_rt_bandwidth() into set_curr_task_dl()?
> 
> ---
>  kernel/sched/deadline.c |  1 +
>  kernel/sched/rt.c       | 16 +++++++++++++++-
>  kernel/sched/sched.h    |  1 +
>  3 files changed, 17 insertions(+), 1 deletion(-)
> diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c
> index ed31ef6..f1d2304 100644
> --- a/kernel/sched/deadline.c
> +++ b/kernel/sched/deadline.c
> @@ -720,6 +720,7 @@ void inc_dl_tasks(struct sched_dl_entity *dl_se, struct dl_rq *dl_rq)
>  
>  	inc_dl_deadline(dl_rq, deadline);
>  	inc_dl_migration(dl_se, dl_rq);
> +	start_top_rt_bandwidth(rq_of_dl_rq(dl_rq));
>  }
>  
>  static inline
> diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c
> index 72f9ec7..a187eb8 100644
> --- a/kernel/sched/rt.c
> +++ b/kernel/sched/rt.c
> @@ -726,7 +726,7 @@ static inline int balance_runtime(struct rt_rq *rt_rq)
>  
>  static int do_sched_rt_period_timer(struct rt_bandwidth *rt_b, int overrun)
>  {
> -	int i, idle = 1, throttled = 0;
> +	int i, idle = 1, throttled = 0, has_dl_tasks = 0;
>  	const struct cpumask *span;
>  
>  	span = sched_rt_period_mask();
> @@ -781,9 +781,15 @@ static int do_sched_rt_period_timer(struct rt_bandwidth *rt_b, int overrun)
>  
>  		if (enqueue)
>  			sched_rt_rq_enqueue(rt_rq);
> +
> +		if (rt_rq == &rq->rt && rq->dl.dl_nr_running)
> +			has_dl_tasks = 1;
>  		raw_spin_unlock(&rq->lock);
>  	}
>  
> +	if (has_dl_tasks)
> +		return 0;
> +
>  	if (!throttled && (!rt_bandwidth_enabled() || rt_b->rt_runtime == RUNTIME_INF))
>  		return 1;
>  
> @@ -1005,6 +1011,10 @@ dec_rt_group(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq)
>  	WARN_ON(!rt_rq->rt_nr_running && rt_rq->rt_nr_boosted);
>  }
>  
> +void start_top_rt_bandwidth(struct rq *rq)
> +{
> +	start_rt_bandwidth(&rq->rt.tg->rt_bandwidth);
> +}
>  #else /* CONFIG_RT_GROUP_SCHED */
>  
>  static void
> @@ -1016,6 +1026,10 @@ inc_rt_group(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq)
>  static inline
>  void dec_rt_group(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq) {}
>  
> +void start_top_rt_bandwidth(struct rq *rq)
> +{
> +	start_rt_bandwidth(&def_rt_bandwidth);
> +}
>  #endif /* CONFIG_RT_GROUP_SCHED */
>  
>  static inline
> diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
> index 1bf34c2..a20f8f7 100644
> --- a/kernel/sched/sched.h
> +++ b/kernel/sched/sched.h
> @@ -1294,6 +1294,7 @@ static inline void sched_avg_update(struct rq *rq) { }
>  #endif
>  
>  extern void start_bandwidth_timer(struct hrtimer *period_timer, ktime_t period);
> +extern void start_top_rt_bandwidth(struct rq *rq);
>  
>  #ifdef CONFIG_SMP
>  #ifdef CONFIG_PREEMPT

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