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;
Date:   Fri, 11 Dec 2020 09:45:40 +0000
From:   Mel Gorman <mgorman@techsingularity.net>
To:     Hillf Danton <hdanton@sina.com>
Cc:     Peter Ziljstra <peterz@infradead.org>,
        Ingo Molnar <mingo@redhat.com>,
        LKML <linux-kernel@vger.kernel.org>,
        Aubrey Li <aubrey.li@linux.intel.com>,
        Barry Song <song.bao.hua@hisilicon.com>,
        Juri Lelli <juri.lelli@redhat.com>,
        Vincent Guittot <vincent.guittot@linaro.org>,
        Valentin Schneider <valentin.schneider@arm.com>,
        Linux-ARM <linux-arm-kernel@lists.infradead.org>
Subject: Re: [PATCH 3/4] sched/fair: Do not replace recent_used_cpu with the
 new target
Message-ID: <20201211094540.GV3371@techsingularity.net>
References: <20201208153501.1467-1-mgorman@techsingularity.net>
 <20201211062542.3082-1-hdanton@sina.com>
 <20201211093443.3259-1-hdanton@sina.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-15
Content-Disposition: inline
In-Reply-To: <20201211093443.3259-1-hdanton@sina.com>
User-Agent: Mutt/1.10.1 (2018-07-13)
Precedence: bulk

On Fri, Dec 11, 2020 at 05:34:43PM +0800, Hillf Danton wrote:
> On Fri, 11 Dec 2020 09:02:28 +0000 Mel Gorman wrote:
> >On Fri, Dec 11, 2020 at 02:25:42PM +0800, Hillf Danton wrote:
> >> On Tue,  8 Dec 2020 15:35:00 +0000 Mel Gorman wrote:
> >> > @@ -6277,17 +6277,13 @@ static int select_idle_sibling(struct task_struct *p, int prev, int target)
> >> >  
> >> >  	/* Check a recently used CPU as a potential idle candidate: */
> >> >  	recent_used_cpu = p->recent_used_cpu;
> >> > +	p->recent_used_cpu = prev;
> >> >  	if (recent_used_cpu != prev &&
> >> >  	    recent_used_cpu != target &&
> >> >  	    cpus_share_cache(recent_used_cpu, target) &&
> >> >  	    (available_idle_cpu(recent_used_cpu) || sched_idle_cpu(recent_used_cpu)) &&
> >> >  	    cpumask_test_cpu(p->recent_used_cpu, p->cpus_ptr) &&
> >> 
> >> Typo? Fix it in spin if so.
> >> 
> >
> >What typo?
> 
> After your change it is prev that we check against p->cpus_ptr instead of
> the recent CPU. Wonder the point to do such a check for returning the
> recent one.

Ah... yes, this is indeed wrong. It wouldn't affect Vincent's case
that showed a problem with a hackbench configuration (which I'm still
disappointed about as it's a trade-off depending on machine and workload)
but it allows a task to run on the wrong cpu if sched_setscheduler()
was called between wakeup events.

-- 
Mel Gorman
SUSE Labs