Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:5 as permitted sender) client-ip=2620:137:e000::3:5;
Message-ID: <471d24f71cc531448572e2fc5964f1231e7ef7b3.camel@linux.intel.com>
Subject: Re: [PATCH] sched/fair: merge same code in enqueue_task_fair
From:   Tim Chen <tim.c.chen@linux.intel.com>
To:     WangJinchao <wangjinchao@xfusion.com>,
        Ingo Molnar <mingo@redhat.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Juri Lelli <juri.lelli@redhat.com>,
        Vincent Guittot <vincent.guittot@linaro.org>,
        Dietmar Eggemann <dietmar.eggemann@arm.com>,
        Steven Rostedt <rostedt@goodmis.org>,
        Ben Segall <bsegall@google.com>, Mel Gorman <mgorman@suse.de>,
        Daniel Bristot de Oliveira <bristot@redhat.com>,
        Valentin Schneider <vschneid@redhat.com>,
        linux-kernel@vger.kernel.org
Cc:     stone.xulei@xfusion.com
Date:   Mon, 11 Dec 2023 14:22:07 -0800
In-Reply-To: <202312101719+0800-wangjinchao@xfusion.com>
References: <202312101719+0800-wangjinchao@xfusion.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.4 (3.44.4-2.fc36) 
MIME-Version: 1.0
Precedence: bulk

On Sun, 2023-12-10 at 17:21 +0800, WangJinchao wrote:
> 1. The code below is duplicated in two for loops and need to be
>    consolidated
> 2. Fix the bug where a se's on_rq is true but its parent is not

In the current code, If a se is already on a rq, all the parents would have=
 already been
on rq too.  I don't think there's a case where se's on_rq and parent
is not for the current code before your patch.  Otherwise the child
would never get scheduled. I don't think we have seen such bug being
reported.

>=20
> ```c
> 		cfs_rq->h_nr_running++;
> 		cfs_rq->idle_h_nr_running +=3D idle_h_nr_running;
>=20
> 		if (cfs_rq_is_idle(cfs_rq))
> 			idle_h_nr_running =3D 1;
>=20
> 		/* end evaluation on encountering a throttled cfs_rq */
> 		if (cfs_rq_throttled(cfs_rq))
> 			goto enqueue_throttle;
> ```
>=20
> Signed-off-by: WangJinchao <wangjinchao@xfusion.com>
> ---
>  kernel/sched/fair.c | 31 ++++++++-----------------------
>  1 file changed, 8 insertions(+), 23 deletions(-)
>=20
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index d7a3c63a2171..e1373bfd4f2e 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -6681,30 +6681,15 @@ enqueue_task_fair(struct rq *rq, struct task_stru=
ct *p, int flags)
>  		cpufreq_update_util(rq, SCHED_CPUFREQ_IOWAIT);
> =20
>  	for_each_sched_entity(se) {
> -		if (se->on_rq)
> -			break;
>  		cfs_rq =3D cfs_rq_of(se);
> -		enqueue_entity(cfs_rq, se, flags);
> -
> -		cfs_rq->h_nr_running++;
> -		cfs_rq->idle_h_nr_running +=3D idle_h_nr_running;
> -
> -		if (cfs_rq_is_idle(cfs_rq))
> -			idle_h_nr_running =3D 1;
> -
> -		/* end evaluation on encountering a throttled cfs_rq */
> -		if (cfs_rq_throttled(cfs_rq))
> -			goto enqueue_throttle;
> -
> -		flags =3D ENQUEUE_WAKEUP;
> -	}
> -
> -	for_each_sched_entity(se) {
> -		cfs_rq =3D cfs_rq_of(se);
> -
> -		update_load_avg(cfs_rq, se, UPDATE_TG);
> -		se_update_runnable(se);
> -		update_cfs_group(se);
> +		if (se->on_rq) {
> +			update_load_avg(cfs_rq, se, UPDATE_TG);
> +			se_update_runnable(se);
> +			update_cfs_group(se);
> +		} else {
> +			enqueue_entity(cfs_rq, se, flags);
> +			flags =3D ENQUEUE_WAKEUP;
> +		}

The code change looks like a reasonable simplification.  Logic
is the same as the old code, assuming that once a se's on_rq flag
is true, all its parents are too.

Thanks.

Tim

> =20
>  		cfs_rq->h_nr_running++;
>  		cfs_rq->idle_h_nr_running +=3D idle_h_nr_running;