From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Subject: Re: [PATCH 3/3] KVM: x86: Optimize NMI watchdog delivery Date: Sun, 19 Oct 2008 13:15:20 +0200 Message-ID: <48FB16C8.2010503@redhat.com> References: <20081015142748.385784583@mchn012c.ww002.siemens.net> <20081015142748.931910158@mchn012c.ww002.siemens.net> <20081017170654.GA22408@yukikaze> <48F8C9F5.5070500@web.de> <20081017173414.GD22408@yukikaze> <48F8CDF0.4070902@web.de> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: sheng@linux.intel.com, Jan Kiszka , kvm@vger.kernel.org, jiajun.xu@intel.com To: Jan Kiszka Return-path: Received: from mx2.redhat.com ([66.187.237.31]:52127 "EHLO mx2.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750886AbYJSLP0 (ORCPT ); Sun, 19 Oct 2008 07:15:26 -0400 In-Reply-To: <48F8CDF0.4070902@web.de> Sender: kvm-owner@vger.kernel.org List-ID: Jan Kiszka wrote: > If you have dozens of lapics, you don't want to check them all if they > are ALL switched of anyway. That information is better encoded in a > single, (virtual) system-wide bool. That's the most common case we want > to speed up. And it is the core of the optimization Avi suggested > (unless I totally misunderstood him). > > Yes, that's what I meant. Touching vcpu data is expensive since it's likely to be in MESI (or whatever it's called now) exclusive mode on some other cpu. -- error compiling committee.c: too many arguments to function