From mboxrd@z Thu Jan  1 00:00:00 1970
From: Avi Kivity <avi@redhat.com>
Subject: Re: [PATCH 08/13] xen/pvticketlock: disable interrupts while blocking
Date: Wed, 14 Sep 2011 18:01:18 +0300
Message-ID: <4E70C1BE.7060209@redhat.com>
References: <20110907134411.GV5795@redhat.com> <4E678992.5050709@redhat.com> <20110907155657.GX5795@redhat.com> <4E679AF4.50209@redhat.com> <20110907165203.GQ6838@redhat.com> <4E67A551.4000502@redhat.com> <20110913184044.GN5795@redhat.com> <20110913190320.GR7761@one.firstfloor.org> <20110913192152.GO5795@redhat.com> <4E7050F7.3000208@redhat.com> <20110914144926.GU7761@one.firstfloor.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Return-path: <linux-kernel-owner@vger.kernel.org>
In-Reply-To: <20110914144926.GU7761@one.firstfloor.org>
Sender: linux-kernel-owner@vger.kernel.org
To: Andi Kleen <andi@firstfloor.org>
Cc: Don Zickus <dzickus@redhat.com>, Jeremy Fitzhardinge <jeremy@goop.org>, Peter Zijlstra <peterz@infradead.org>, "H. Peter Anvin" <hpa@zytor.com>, Linus Torvalds <torvalds@linux-foundation.org>, Ingo Molnar <mingo@elte.hu>, the arch/x86 maintainers <x86@kernel.org>, Linux Kernel Mailing List <linux-kernel@vger.kernel.org>, Nick Piggin <npiggin@kernel.dk>, Marcelo Tosatti <mtosatti@redhat.com>, KVM <kvm@vger.kernel.org>, Xen Devel <xen-devel@lists.xensource.com>, Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>, Stefano Stabellini <stefano.stabellini@eu.citrix.com>
List-Id: xen-devel@lists.xenproject.org

On 09/14/2011 05:49 PM, Andi Kleen wrote:
> On Wed, Sep 14, 2011 at 10:00:07AM +0300, Avi Kivity wrote:
> >  On 09/13/2011 10:21 PM, Don Zickus wrote:
> >  >Or are you saying an NMI in an idle system will have the same %rip thus
> >  >falsely detecting a back-to-back NMI?
> >  >
> >  >
> >
> >  That's easy to avoid - insert an instruction zeroing the last nmi_rip
> >  somewhere before or after hlt.  It's always okay to execute such an
> >  instruction (outside the nmi handler itself), since nmi_rip is meant to
> >  detect a "no instructions executed" condition.
>
> At least for classic hlt there is no simple "after hlt" because it's all
> interrupt handlers and exceptions and everything else that can interrupt
> combined.

If an NMI hits in an interrupt handler, or in the "after hlt" section 
before the write-to-last-nmi-rip, then we'll see that %rip has changed.  
If it hits after the write-to-last-nmi-rip instruction (or in the hlt 
itself), then we'll also see that %rip has changed, due to the effect of 
that instruction.

-- 
error compiling committee.c: too many arguments to function