From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757428AbXGLFnT (ORCPT ); Thu, 12 Jul 2007 01:43:19 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751494AbXGLFnL (ORCPT ); Thu, 12 Jul 2007 01:43:11 -0400 Received: from il.qumranet.com ([82.166.9.18]:47239 "EHLO il.qumranet.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750971AbXGLFnK (ORCPT ); Thu, 12 Jul 2007 01:43:10 -0400 Message-ID: <4695BF6C.9000501@qumranet.com> Date: Thu, 12 Jul 2007 08:43:08 +0300 From: Avi Kivity User-Agent: Thunderbird 2.0.0.4 (X11/20070615) MIME-Version: 1.0 To: Luca CC: kvm-devel@lists.sourceforge.net, linux-kernel@vger.kernel.org Subject: Re: [kvm-devel] [PATCH 2/2] KVM: Use the scheduler preemption hooks to make kvm preemptible References: <11841693332609-git-send-email-avi@qumranet.com> <1184169333786-git-send-email-avi@qumranet.com> <68676e00707111310v4862974am37dc7c723486b96c@mail.gmail.com> In-Reply-To: <68676e00707111310v4862974am37dc7c723486b96c@mail.gmail.com> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0 (firebolt.argo.co.il [0.0.0.0]); Thu, 12 Jul 2007 08:43:09 +0300 (IDT) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Luca wrote: > On 7/11/07, Avi Kivity wrote: >> Current kvm disables preemption while the new virtualization >> registers are >> in use. This of course is not very good for latency sensitive >> workloads (one >> use of virtualization is to offload user interface and other latency >> insensitive stuff to a container, so that it is easier to analyze the >> remaining workload). This patch re-enables preemption for kvm; >> preemption >> is now only disabled when switching the registers in and out, and during >> the switch to guest mode and back. > [...] >> diff --git a/drivers/kvm/Kconfig b/drivers/kvm/Kconfig >> index 33fa28a..92644b5 100644 >> --- a/drivers/kvm/Kconfig >> +++ b/drivers/kvm/Kconfig >> @@ -10,7 +10,7 @@ if VIRTUALIZATION >> >> config KVM >> tristate "Kernel-based Virtual Machine (KVM) support" >> - depends on X86 && EXPERIMENTAL >> + depends on X86 && EXPERIMENTAL && PREEMPT_HOOKS >> depends on X86_CMPXCHG64 || 64BIT > > select PREEMPT_HOOKS maybe? > Yes. And default PREEMPT_HOOKS to n in the first patch. I'll make both these changes. -- Do not meddle in the internals of kernels, for they are subtle and quick to panic.