From mboxrd@z Thu Jan  1 00:00:00 1970
From: Arjan Koers <0h61vkll2ly8@xutrox.com>
Subject: Re: 2.6.35-rc1 regression with pvclock and smp guests
Date: Fri, 08 Oct 2010 02:12:51 +0200
Message-ID: <4CAE6203.6040902@xutrox.com>
References: <4CA2F8A3.80400@redhat.com> <4CA30424.9030007@msgid.tls.msk.ru> <4CA30493.6090503@msgid.tls.msk.ru> <4CA392FE.5090009@xutrox.com> <4CA4427C.9090304@msgid.tls.msk.ru> <4CA45F7B.8050806@msgid.tls.msk.ru> <4CA4968F.9050402@redhat.com> <4CA4A8C5.3030407@msgid.tls.msk.ru> <4CA4AD87.8060502@redhat.com> <4CA4DBC8.6070606@xutrox.com> <20100930190507.GA1111@amt.cnet> <4CA51715.1070507@msgid.tls.msk.ru> <4CA51847.5060208@msgid.tls.msk.ru> <4CA6C4BB.5020004@redhat.com> <4CA6E0BF.90605@msgid.tls.msk.ru> <4CA75969.1080405@xutrox.com> <4CA7C34C.4040000@redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Cc: Zachary Amsden <zamsden@redhat.com>,
	Marcelo Tosatti <mtosatti@redhat.com>,
	Michael Tokarev <mjt@tls.msk.ru>, Avi Kivity <avi@redhat.com>,
	Glauber Costa <glommer@redhat.com>,
	Andre Przywara <andre.przywara@amd.com>
To: kvm@vger.kernel.org
Return-path: <kvm-owner@vger.kernel.org>
Received: from smtp-out2.tiscali.nl ([195.241.79.177]:60686 "EHLO
	smtp-out2.tiscali.nl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1752126Ab0JHAM7 (ORCPT <rfc822;kvm@vger.kernel.org>);
	Thu, 7 Oct 2010 20:12:59 -0400
In-Reply-To: <4CA7C34C.4040000@redhat.com>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

On 2010-10-03 01:42, Zachary Amsden wrote:
...
> 
> Umm...  do you guys have this commit?  This is supposed to address the
> issue where the guest keeps resetting the TSC.  A guest which does that
> will break kvmclock.  It only happens on SMP, and it's much worse on AMD
> CPUs...
> 
> sound like your scenario.
> 
> commit bd59fc8ff95126f27b7a0df1b6cc602aa428812d
> Author: Zachary Amsden <zamsden@redhat.com>
> Date:   Thu Aug 19 22:07:26 2010 -1000


This commit fixes the problem:

commit aad07c4f92bae2edaa42bcef84c2afdd0d082458
Author: Zachary Amsden <zamsden@redhat.com>
Date:   Thu Aug 19 22:07:19 2010 -1000

    KVM: x86: Move TSC reset out of vmcb_init

    The VMCB is reset whenever we receive a startup IPI, so Linux is setting
    TSC back to zero happens very late in the boot process and destabilizing
    the TSC.  Instead, just set TSC to zero once at VCPU creation time.

    Why the separate patch?  So git-bisect is your friend.