From mboxrd@z Thu Jan  1 00:00:00 1970
From: Gleb Natapov <gleb@redhat.com>
Subject: Re: kvm: RCU warning in async pf
Date: Mon, 16 Apr 2012 17:17:35 +0300
Message-ID: <20120416141735.GA18613@redhat.com>
References: <1333414472.11455.7.camel@lappy>
 <20120403105226.GF14939@redhat.com>
 <20120404123033.GD11918@redhat.com>
 <1334396693.2528.45.camel@twins>
 <20120416102855.GP11918@redhat.com>
 <20120416125848.GT11918@redhat.com>
 <1334581548.28150.40.camel@twins>
 <20120416130806.GV11918@redhat.com>
 <1334582281.28150.41.camel@twins>
 <1334582840.28150.43.camel@twins>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: Sasha Levin <levinsasha928@gmail.com>, Avi Kivity <avi@redhat.com>,
	Marcelo Tosatti <mtosatti@redhat.com>,
	Thomas Gleixner <tglx@linutronix.de>,
	Ingo Molnar <mingo@elte.hu>,
	"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>,
	kvm <kvm@vger.kernel.org>,
	linux-kernel <linux-kernel@vger.kernel.org>,
	Dave Jones <davej@redhat.com>
To: Peter Zijlstra <peterz@infradead.org>
Return-path: <linux-kernel-owner@vger.kernel.org>
Content-Disposition: inline
In-Reply-To: <1334582840.28150.43.camel@twins>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: kvm.vger.kernel.org

On Mon, Apr 16, 2012 at 03:27:20PM +0200, Peter Zijlstra wrote:
> On Mon, 2012-04-16 at 15:18 +0200, Peter Zijlstra wrote:
> > On Mon, 2012-04-16 at 16:08 +0300, Gleb Natapov wrote:
> > > On Mon, Apr 16, 2012 at 03:05:48PM +0200, Peter Zijlstra wrote:
> > > > On Mon, 2012-04-16 at 15:58 +0300, Gleb Natapov wrote:
> > > > > -               rcu_irq_enter();
> > > > > +               irq_enter();
> > > > >                 exit_idle(); 
> > > > 
> > > > Do we really need the exit_idle()? I can't remember other interrupt
> > > > handlers doing that.
> > > They do. That's where I got the idea.
> > 
> > Some do.. some don't.. /me goes have a look what this exit_idle nonsense
> > is all about. Looks to be something broken.
> 
> Yeah, its broken and the whole implementation is crap anyway. There's
> only one user (drivers/idle/i7300_idle.c) so it likely doesn't matter
> (much) anyway.
> 
Should I resend the patch with removal of exit_idle() call or it will
be removed by a patch that removes exit_idle() completely later?

> Thomas, can we rip that stuff out? or do we have to like actually fix
> it?

--
			Gleb.