From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Subject: Re: [PATCH] KVM:PPC Issue in exit timing clearance Date: Wed, 23 Mar 2011 11:20:12 +0200 Message-ID: <4D89BB4C.1060806@redhat.com> References: <1300293459-32767-1-git-send-email-Bharat.Bhushan@freescale.com> <4D88998B.6050208@redhat.com> <4D88B1BB.8070306@suse.de> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: Bharat Bhushan , kvm@vger.kernel.org, bharatb.yadav@gmail.com, Bharat Bhushan To: Alexander Graf Return-path: Received: from mx1.redhat.com ([209.132.183.28]:1622 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755943Ab1CWJUW (ORCPT ); Wed, 23 Mar 2011 05:20:22 -0400 In-Reply-To: <4D88B1BB.8070306@suse.de> Sender: kvm-owner@vger.kernel.org List-ID: On 03/22/2011 04:27 PM, Alexander Graf wrote: > >> >> What about using vcpu->requests to have the statistics cleared in >> vcpu context? >> >> What about dropping the whole thing and replacing it with tracepoints? > > That should work. The question is if it's worth the effort. The > current code is there after all :). > Tracepoints are wonderful for debugging problems in the field, from my experience. Removing the special timing code (or rather, moving it to userspace) is just a bonus. -- error compiling committee.c: too many arguments to function