From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756078Ab2DQL5y (ORCPT ); Tue, 17 Apr 2012 07:57:54 -0400 Received: from mx1.redhat.com ([209.132.183.28]:1983 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756033Ab2DQL5x (ORCPT ); Tue, 17 Apr 2012 07:57:53 -0400 Message-ID: <4F8D5ABC.4030705@redhat.com> Date: Tue, 17 Apr 2012 14:57:48 +0300 From: Avi Kivity User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:11.0) Gecko/20120329 Thunderbird/11.0.1 MIME-Version: 1.0 To: Jan Kiszka CC: Marcelo Tosatti , kvm , Linux Kernel Mailing List Subject: Re: [PATCH] KVM: x86: Run PIT work in own kthread References: <4F8C645C.7010302@siemens.com> <4F8D4598.5080902@redhat.com> <4F8D5186.7020801@siemens.com> In-Reply-To: <4F8D5186.7020801@siemens.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 04/17/2012 02:18 PM, Jan Kiszka wrote: > On 2012-04-17 12:27, Avi Kivity wrote: > > On 04/16/2012 09:26 PM, Jan Kiszka wrote: > >> We can't run PIT IRQ injection work in the interrupt context of the host > >> timer. This would allow the user to influence the handler complexity by > >> asking for a broadcast to a large number of VCPUs. Therefore, this work > >> was pushed into workqueue context in 9d244caf2e. However, this prevents > >> prioritizing the PIT injection over other task as workqueues share > >> kernel threads. > >> > >> This replaces the workqueue with a kthread worker and gives that thread > >> a name in the format "kvm-pit/". That allows to > >> identify and adjust the kthread priority according to the VM process > >> parameters. > > > > Is this a new ABI? > > Yep. Scripts will use it, maybe even QEMU. Do you want this to appear in > api.txt? You already know the answer. Please make it explicit that the thread is optional. This way, if we gain real-time workqueue support, or eliminate it through some other trick, userspace won't break. -- error compiling committee.c: too many arguments to function