linux-kernel - Re: [RFC PATCH v2] sched: reflect sched_entity movement into task

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Mon, 6 Jun 2016 11:52:29 +0100
From:	Dietmar Eggemann <dietmar.eggemann@....com>
To:	Vincent Guittot <vincent.guittot@...aro.org>, peterz@...radead.org,
	mingo@...nel.org, linux-kernel@...r.kernel.org, pjt@...gle.com
Cc:	yuyang.du@...el.com, bsegall@...gle.com, Morten.Rasmussen@....com
Subject: Re: [RFC PATCH v2] sched: reflect sched_entity movement into
 task_group's utilization

On 24/05/16 10:55, Vincent Guittot wrote:

[...]

> +/* Take into account the change of the utilization of a child task group */
> +static void update_tg_cfs_util(struct sched_entity *se, int blocked)
> +{
> +	int delta;
> +	struct cfs_rq *cfs_rq;
> +	long update_util_avg;
> +	long last_update_time;
> +	long old_util_avg;
> +
> +
> +	/*
> +	 * update_blocked_average will call this function for root cfs_rq
> +	 * whose se is null. In this case just return
> +	 */
> +	if (!se)
> +		return;
> +
> +	if (entity_is_task(se))
> +		return 0;

void function

> +
> +	/* Get sched_entity of cfs rq */

You mean /* Get cfs_rq owned by this task group */?

> +	cfs_rq = group_cfs_rq(se);
> +
> +	update_util_avg = cfs_rq->diff_util_avg;
> +
> +	if (!update_util_avg)
> +		return 0;

Couldn't you not just get rid of long update_util_avg and only use
cfs_rq->diff_util_avg here (clearing pending changes after you set
se->avg.util_avg)?

> +
> +	/* Clear pending changes */
> +	cfs_rq->diff_util_avg = 0;
> +
> +	/* Add changes in sched_entity utilizaton */
> +	old_util_avg = se->avg.util_avg;
> +	se->avg.util_avg = max_t(long, se->avg.util_avg + update_util_avg, 0);
> +	se->avg.util_sum = se->avg.util_avg * LOAD_AVG_MAX;
> +
> +	/* Get parent cfs_rq */
> +	cfs_rq = cfs_rq_of(se);
> +
> +	if (blocked) {
> +		/*
> +		 * blocked utilization has to be synchronized with its parent
> +		 * cfs_rq's timestamp
> +		 */

We don't have stand-alone blocked utilization any more although the
function is still called update_blocked_averages(). Do you need this for
cpu's going through idle periods?
It's also necessary because there could be other se's which could have
been [en|de]queued at/from this cfs_rq so its last_update_time value is
more recent?

[...]