Subject: RE: PATCH: Fix race in cpu_down (hotplug cpu)
From: Nigel Cunningham <ncunningham@cyclades.com>
Reply-To: ncunningham@cyclades.com
To: Li Shaohua <shaohua.li@intel.com>
Cc: vatsa@in.ibm.com, Andrew Morton <akpm@osdl.org>,
       Linus Torvalds <torvalds@osdl.org>,
       Zwane Mwaikambo <zwane@arm.linux.org.uk>,
       Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
       Rusty Russell <rusty@rustcorp.com.au>
In-Reply-To: <59D45D057E9702469E5775CBB56411F171F7E0@pdsmsx406>
References: <59D45D057E9702469E5775CBB56411F171F7E0@pdsmsx406>
Content-Type: text/plain
Organization: Cyclades
Message-Id: <1127107381.9696.85.camel@localhost>
Mime-Version: 1.0
Date: Mon, 19 Sep 2005 15:23:01 +1000
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1521
Lines: 44

Hi.

On Mon, 2005-09-19 at 14:48, Li, Shaohua wrote:
> Hi,
> >
> >On Mon, Sep 19, 2005 at 01:28:38PM +1000, Nigel Cunningham wrote:
> >> There is a race condition in taking down a cpu
> (kernel/cpu.c::cpu_down).
> >> A cpu can already be idling when we clear its online flag, and we do
> not
> >> force the idle task to reschedule. This results in __cpu_die timing
> out.
> >
> >"when we clear its online flag" - This happens in take_cpu_down in the
> >context of stopmachine thread. take_cpu_down also ensures that idle
> >thread runs when it returns (sched_idle_next). So when idle thread
> runs,
> >it should notice that it is offline and invoke play_dead.  So I don't
> >understand why __cpu_die should time out.
> I guess Nigel's point is cpu_idle is preempted before take_cpu_down. If
> the preempt occurs after the cpu_is_offline check, when the cpu (after
> sched_idle_next) goes into idle again, nobody can wake it up. Nigel,
> isn't it?

Maybe I'm just an ignoramus, but I was thinking (without being a
scheduler expert at all) that if the idle thread was already running,
trying to set it up to run next might possibly have zero effect. I've
added a bit of debugging code to try and see in better detail what's
happening.

Regards,

Nigel

> Thanks,
> Shaohua
-- 


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/