Received-SPF: pass (google.com: domain of linux-kernel+bounces-14233-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.48.161 as permitted sender) client-ip=147.75.48.161;
From: Valentin Schneider <vschneid@redhat.com>
To: alexs@kernel.org, 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>, linux-kernel@vger.kernel.org
Cc: curuwang@tencent.com, Alex Shi <alexs@kernel.org>
Subject: Re: [PATCH] sched/tracing: correct the task blocking state
In-Reply-To: <20240102073351.1527503-1-alexs@kernel.org>
References: <20240102073351.1527503-1-alexs@kernel.org>
Date: Tue, 02 Jan 2024 11:18:52 +0100
Message-ID: <xhsmhplykjawj.mognet@vschneid-thinkpadt14sgen2i.remote.csb>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain

On 02/01/24 15:33, alexs@kernel.org wrote:
> From: Alex Shi <alexs@kernel.org>
>
> commit 80ed87c8a9ca ("sched/wait: Introduce TASK_NOLOAD and TASK_IDLE")
> stopped the idle kthreads contribution to loadavg. Also task idle should
> separated from blocked state too, otherwise we will get incorrect task
> blocking state from event tracing sched:sched_stat_blocked.
>

Why is that incorrect? AFAICT we have mapped the (schedstat) 'blocked'
meaning to TASK_UNINTERRUPTIBLE. TASK_IDLE tasks don't contribute to
loadavg yes, but they are still in an UNINTERRUPTIBLE wait.

> diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c
> index b28114478b82..b6afa596f071 100644
> --- a/kernel/sched/deadline.c
> +++ b/kernel/sched/deadline.c
> @@ -1570,7 +1570,7 @@ update_stats_dequeue_dl(struct dl_rq *dl_rq, struct sched_dl_entity *dl_se,
>                       __schedstat_set(p->stats.sleep_start,
>                                       rq_clock(rq_of_dl_rq(dl_rq)));
>
> -		if (state & TASK_UNINTERRUPTIBLE)
> +		if (is_blocked_task_state(state))
>                       __schedstat_set(p->stats.block_start,
>                                       rq_clock(rq_of_dl_rq(dl_rq)));

This change makes it so tasks waiting in TASK_IDLE have their waiting
ignored by schedstat (they are seen as neither INTERRUPTIBLE nor UNINTERRUPTIBLE).