Date: Wed, 14 Jan 2009 12:30:06 +0100
From: Ingo Molnar <mingo@elte.hu>
To: Rusty Russell <rusty@rustcorp.com.au>
Cc: Maciej Rutecki <maciej.rutecki@gmail.com>, Dieter Ries <clip2@gmx.de>,
       travis@sgi.com, linux-kernel@vger.kernel.org
Subject: Re: 2.6.29-rc1 does not boot
Message-ID: <20090114113006.GE8625@elte.hu>
References: <496A085E.8020604@gmx.de> <8db1092f0901120322x5e453fd0x61a78cc1a55982aa@mail.gmail.com> <20090112112608.GB19388@elte.hu> <200901141146.04563.rusty@rustcorp.com.au>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <200901141146.04563.rusty@rustcorp.com.au>
User-Agent: Mutt/1.5.18 (2008-05-17)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1021
Lines: 25


* Rusty Russell <rusty@rustcorp.com.au> wrote:

> On Monday 12 January 2009 21:56:08 Ingo Molnar wrote:
> > work_on_cpu() needs to be reworked to be more generally usable.
> 
> Yes, but how?  If we don't take the hotplug lock, we are relying on the 
> callers to "prove" the cpu can't go away.  But we've already shown that 
> we can't find this by inspection :(

it appears the problem wasnt even the hotplug lock, but the fact that 
work_on_cpu() uses the generic schedule_work() method, which goes into a 
queue with all other keventsd worklets.

and keventd is used by other aspects of cpufreq as well -> creating 
workqueue-driven lock inversion scenarios - see Dieter's lockdep dump 
earlier in this thread. ( I'm amazed lockdep sees through that dependency
:-)

	Ingo
--
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/