From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sheng Yang Subject: Re: KVM Test report, kernel e6a9246... qemu 94f964d... Date: Tue, 7 Sep 2010 16:54:26 +0800 Message-ID: <201009071654.26156.sheng@linux.intel.com> References: <4C85F489.306@redhat.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Cc: kvm@vger.kernel.org, "Hao, Xudong" To: Avi Kivity Return-path: Received: from mga09.intel.com ([134.134.136.24]:17100 "EHLO mga09.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752441Ab0IGIyR (ORCPT ); Tue, 7 Sep 2010 04:54:17 -0400 In-Reply-To: <4C85F489.306@redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: On Tuesday 07 September 2010 16:15:05 Avi Kivity wrote: > On 09/07/2010 10:22 AM, Hao, Xudong wrote: > >> I never assigned a pci device in my life... > > > > Avi, I assumed your platform has no EPT hardware feature, right? > > Correct. > > > I tried sevral platforms, if I used no EPT suported machine, everything > > is OK as yours; if I used EPT supported machine, and then disable EPT > > with "modprobe kvm_intel ept=0", then issue happen. > > That's what I thought - probably vpid+shadow interaction. I'll test it > out. > > I usually test on an old machine since shadow is much more stressful > than ept. But shadow+vpid is even more stressful, I'll move my testing > there. BTW: the bad commit is: c94c3ba901f87115cf574044e7faabfb1d0251fb "KVM: Don't save/restore MSR_IA32_PERF_STATUS", but I haven't figured out the reason yet. -- regards Yang, Sheng