Date:   Sat, 4 Dec 2021 10:53:16 +0100
From:   Peter Zijlstra <peterz@infradead.org>
To:     Vincent Donnefort <vincent.donnefort@arm.com>
Cc:     mingo@redhat.com, vincent.guittot@linaro.org,
        linux-kernel@vger.kernel.org, mgorman@techsingularity.net,
        dietmar.eggemann@arm.com, valentin.schneider@arm.com
Subject: Re: [PATCH v2] sched/fair: Fix detection of per-CPU kthreads waking
 a task
Message-ID: <20211204095316.GQ16608@worktop.programming.kicks-ass.net>
References: <20211201143450.479472-1-vincent.donnefort@arm.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20211201143450.479472-1-vincent.donnefort@arm.com>
Precedence: bulk

On Wed, Dec 01, 2021 at 02:34:50PM +0000, Vincent Donnefort wrote:
> select_idle_sibling() has a special case for tasks woken up by a per-CPU
> kthread, where the selected CPU is the previous one. However, the current
> condition for this exit path is incomplete. A task can wake up from an
> interrupt context (e.g. hrtimer), while a per-CPU kthread is running. A
> such scenario would spuriously trigger the special case described above.
> Also, a recent change made the idle task like a regular per-CPU kthread,
> hence making that situation more likely to happen
> (is_per_cpu_kthread(swapper) being true now).
> 
> Checking for task context makes sure select_idle_sibling() will not
> interpret a wake up from any other context as a wake up by a per-CPU
> kthread.
> 
> Fixes: 52262ee567ad ("sched/fair: Allow a per-CPU kthread waking a task to stack on the same CPU, to fix XFS performance regression")
> Signed-off-by: Vincent Donnefort <vincent.donnefort@arm.com>
> ---

> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index 945d987246c5..56db4ae85995 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -6399,6 +6399,7 @@ static int select_idle_sibling(struct task_struct *p, int prev, int target)
>  	 * pattern is IO completions.
>  	 */
>  	if (is_per_cpu_kthread(current) &&
> +	    in_task() &&
>  	    prev == smp_processor_id() &&
>  	    this_rq()->nr_running <= 1) {
>  		return prev;

Hurmph, so now I have two 'trivial' patches from you that touch this
same function and they's conflicting. I've fixed it up, but perhaps it
would've been nice to have them combined in a series or somesuch :-)