From mboxrd@z Thu Jan 1 00:00:00 1970 From: Paolo Bonzini Subject: Re: [PATCH RFC 1/2] KVM: don't check for PF_VCPU when yielding Date: Wed, 03 Dec 2014 13:53:11 +0100 Message-ID: <547F07B7.6020902@redhat.com> References: <1416931449-24585-1-git-send-email-dahi@linux.vnet.ibm.com> <1416931449-24585-2-git-send-email-dahi@linux.vnet.ibm.com> <54758692.5000906@de.ibm.com> <20141126102319.2f92c64b@thinkpad-w530> <54759E07.5050204@de.ibm.com> <54785539.9010005@de.ibm.com> <54785F29.1060506@linux.vnet.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Cc: David Hildenbrand , KVM , Gleb Natapov , jfrei@linux.vnet.ibm.com, Cornelia Huck , Michael Mueller , raghavendra.kt.linux@gmail.com To: Raghavendra K T , Christian Borntraeger Return-path: Received: from mx1.redhat.com ([209.132.183.28]:39103 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751443AbaLCMx1 (ORCPT ); Wed, 3 Dec 2014 07:53:27 -0500 In-Reply-To: <54785F29.1060506@linux.vnet.ibm.com> Sender: kvm-owner@vger.kernel.org List-ID: On 28/11/2014 12:40, Raghavendra K T wrote: > I am seeing very small improvement in <= 1x commit cases > and for >1x overcommit, a very slight regression. But considering the > test environment noises, I do not see much effect from the > patch. I think these results are the only one that could be statisically significant: base %stdev patched %stdev %improvement kernbench 1x 53.1421 2.3086 54.6671 2.9673 -2.86966 dbench 1x 6386.4737 1.0487 6703.9113 1.2298 4.97047 and, of course :) one of them says things get worse and the other says things get better. Paolo > But I admit, I have not explored deeply about, > 1. assumption of preempted approximately equals PF_VCPU case logic, > 2. whether it helps for any future usages of yield_to against current > sole usage of virtualization. > > >