Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Tue, 1 May 2018 13:31:32 +0200
From:   Peter Zijlstra <peterz@infradead.org>
To:     "Kohli, Gaurav" <gkohli@codeaurora.org>
Cc:     tglx@linutronix.de, mpe@ellerman.id.au, mingo@kernel.org,
        bigeasy@linutronix.de, linux-kernel@vger.kernel.org,
        linux-arm-msm@vger.kernel.org,
        Neeraj Upadhyay <neeraju@codeaurora.org>,
        Will Deacon <will.deacon@arm.com>,
        Oleg Nesterov <oleg@redhat.com>
Subject: Re: [PATCH v1] kthread/smpboot: Serialize kthread parking against
 wakeup
Message-ID: <20180501113132.GF12217@hirez.programming.kicks-ass.net>
References: <1524645199-5596-1-git-send-email-gkohli@codeaurora.org>
 <20180425200917.GZ4082@hirez.programming.kicks-ass.net>
 <20180426084131.GV4129@hirez.programming.kicks-ass.net>
 <20180426085719.GW4129@hirez.programming.kicks-ass.net>
 <4d3f68f8-e599-6b27-a2e8-9e96b401d57a@codeaurora.org>
 <20180430111744.GE4082@hirez.programming.kicks-ass.net>
 <3af3365b-4e3f-e388-8e90-45a3bd4120fd@codeaurora.org>
 <20180501101845.GE12217@hirez.programming.kicks-ass.net>
 <f134c8d8-e56f-bf29-0177-e6702ae79f2b@codeaurora.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <f134c8d8-e56f-bf29-0177-e6702ae79f2b@codeaurora.org>
User-Agent: Mutt/1.9.5 (2018-04-13)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, May 01, 2018 at 04:10:53PM +0530, Kohli, Gaurav wrote:
> Yes with loop, it will reset TASK_PARKED but that is not happening in the
> dumps we have seen.

But was that with or without the fixed wait-loop? I don't care about
stuff you might have seen with the current code, that is clearly broken.

> > takedown_cpu() can proceed beyond smpboot_park_threads() and kill the
> > CPU before any of the threads are parked -- per having the complete()
> > before hitting schedule().
> > 
> > And, afaict, that is harmless. When we go offline, sched_cpu_dying() ->
> > migrate_tasks() will migrate any still runnable threads off the cpu.
> > But because at this point the thread must be in the PARKED wait-loop, it
> > will hit schedule() and go to sleep eventually.
> > 
> > Also note that kthread_unpark() does __kthread_bind() to rebind the
> > threads.
> > 
> > Aaaah... I think I've spotted a problem there. We clear SHOULD_PARK
> > before we rebind, so if the thread lost the first PARKED store,
> > does the completion, gets migrated, cycles through the loop and now
> > observes !SHOULD_PARK and bails the wait-loop, then __kthread_bind()
> > will forever wait.
> > 
> 
> So during next unpark
> __kthread_unpark -> __kthread_bind -> wait_task_inactive (this got failed,
> as current state is running so failed on below call:

Aah, yes, I seem to have mis-remembered how wait_task_inactive() works.
And it is indeed still a problem..

Let me ponder what the best solution is, it's a bit of a mess.