Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Benjamin Segall <bsegall@google.com>
To:     Peng Wang <rocking@linux.alibaba.com>
Cc:     mingo@redhat.com, peterz@infradead.org, juri.lelli@redhat.com,
        vincent.guittot@linaro.org, dietmar.eggemann@arm.com,
        rostedt@goodmis.org, bristot@redhat.com,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH] sched/fair: Reorder throttle_cfs_rq() path
References: <f11dd2e3ab35cc538e2eb57bf0c99b6eaffce127.1604973978.git.rocking@linux.alibaba.com>
Date:   Wed, 11 Nov 2020 12:27:07 -0800
In-Reply-To: <f11dd2e3ab35cc538e2eb57bf0c99b6eaffce127.1604973978.git.rocking@linux.alibaba.com>
        (Peng Wang's message of "Tue, 10 Nov 2020 10:11:59 +0800")
Message-ID: <xm26imab3ahw.fsf@google.com>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/27.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
Precedence: bulk

Peng Wang <rocking@linux.alibaba.com> writes:

> As commit 39f23ce07b93 ("sched/fair: Fix unthrottle_cfs_rq() for
> leaf_cfs_rq list") does in unthrottle_cfs_rq(), throttle_cfs_rq()
> can also use the same pattern as dequeue_task_fair().
>
> There is no functional changes.

It's generally a bit more hassle and less clear, but the parallel to
dequeue_task_fair probably makes up for it.

Reviewed-by: Ben Segall <bsegall@google.com>

>
> Signed-off-by: Peng Wang <rocking@linux.alibaba.com>
> ---
>  kernel/sched/fair.c | 34 +++++++++++++++++++++++-----------
>  1 file changed, 23 insertions(+), 11 deletions(-)
>
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index 290f9e3..27a69af 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -4779,25 +4779,37 @@ static bool throttle_cfs_rq(struct cfs_rq *cfs_rq)
>  		struct cfs_rq *qcfs_rq = cfs_rq_of(se);
>  		/* throttled entity or throttle-on-deactivate */
>  		if (!se->on_rq)
> -			break;
> +			goto done;
>  
> -		if (dequeue) {
> -			dequeue_entity(qcfs_rq, se, DEQUEUE_SLEEP);
> -		} else {
> -			update_load_avg(qcfs_rq, se, 0);
> -			se_update_runnable(se);
> -		}
> +		dequeue_entity(qcfs_rq, se, DEQUEUE_SLEEP);
>  
>  		qcfs_rq->h_nr_running -= task_delta;
>  		qcfs_rq->idle_h_nr_running -= idle_task_delta;
>  
> -		if (qcfs_rq->load.weight)
> -			dequeue = 0;
> +		if (qcfs_rq->load.weight) {
> +			/* Avoid re-evaluating load for this entity: */
> +			se = parent_entity(se);
> +			break;
> +		}
>  	}
>  
> -	if (!se)
> -		sub_nr_running(rq, task_delta);
> +	for_each_sched_entity(se) {
> +		struct cfs_rq *qcfs_rq = cfs_rq_of(se);
> +		/* throttled entity or throttle-on-deactivate */
> +		if (!se->on_rq)
> +			goto done;
> +
> +		update_load_avg(qcfs_rq, se, 0);
> +		se_update_runnable(se);
>  
> +		qcfs_rq->h_nr_running -= task_delta;
> +		qcfs_rq->idle_h_nr_running -= idle_task_delta;
> +	}
> +
> +	/* At this point se is NULL and we are at root level*/
> +	sub_nr_running(rq, task_delta);
> +
> +done:
>  	/*
>  	 * Note: distribution will already see us throttled via the
>  	 * throttled-list.  rq->lock protects completion.