From mboxrd@z Thu Jan  1 00:00:00 1970
From: Marcelo Tosatti <mtosatti@redhat.com>
Subject: Re: [patch 2/2] KVM: close timer injection race window in
	__vcpu_run
Date: Sun, 8 Jun 2008 12:08:34 -0300
Message-ID: <20080608150834.GB1408@dmt.cnet>
References: <20080606193734.495417169@localhost.localdomain> <20080606194010.961258871@localhost.localdomain> <484B877B.1080108@qumranet.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: Chris Wright <chrisw@redhat.com>, kvm@vger.kernel.org
To: Avi Kivity <avi@qumranet.com>
Return-path: <kvm-owner@vger.kernel.org>
Received: from mx1.redhat.com ([66.187.233.31]:43287 "EHLO mx1.redhat.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1754078AbYFHPJX (ORCPT <rfc822;kvm@vger.kernel.org>);
	Sun, 8 Jun 2008 11:09:23 -0400
Content-Disposition: inline
In-Reply-To: <484B877B.1080108@qumranet.com>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

On Sun, Jun 08, 2008 at 10:17:15AM +0300, Avi Kivity wrote:
>> Index: kvm/arch/x86/kvm/i8254.c
>> ===================================================================
>> --- kvm.orig/arch/x86/kvm/i8254.c
>> +++ kvm/arch/x86/kvm/i8254.c
>> @@ -200,9 +200,12 @@ static int __pit_timer_fn(struct kvm_kpi
>>   	atomic_inc(&pt->pending);
>>  	smp_mb__after_atomic_inc();
>> -	if (vcpu0 && waitqueue_active(&vcpu0->wq)) {
>> -		vcpu0->arch.mp_state = KVM_MP_STATE_RUNNABLE;
>> -		wake_up_interruptible(&vcpu0->wq);
>> +	if (vcpu0) {
>> +		set_bit(KVM_REQ_PENDING_TIMER, &vcpu0->requests);
>> +		if (waitqueue_active(&vcpu0->wq)) {
>> +			vcpu0->arch.mp_state = KVM_MP_STATE_RUNNABLE;
>> +			wake_up_interruptible(&vcpu0->wq);
>> +		}
>>  	}
>>    
>
> We probably ought to wakeup only if pt->pending was zero, no?

Yep, same for LAPIC.