From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753108AbXDLTmZ (ORCPT ); Thu, 12 Apr 2007 15:42:25 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753122AbXDLTmZ (ORCPT ); Thu, 12 Apr 2007 15:42:25 -0400 Received: from gw.goop.org ([64.81.55.164]:39233 "EHLO mail.goop.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753108AbXDLTmY (ORCPT ); Thu, 12 Apr 2007 15:42:24 -0400 Message-ID: <461E8B7B.1000006@goop.org> Date: Thu, 12 Apr 2007 12:41:47 -0700 From: Jeremy Fitzhardinge User-Agent: Thunderbird 1.5.0.10 (X11/20070302) MIME-Version: 1.0 To: Andi Kleen CC: Andrew Morton , Daniel Walker , linux-kernel@vger.kernel.org, johnstul@us.ibm.com, tglx@linutronix.de Subject: Re: [PATCH] i386 tsc: remove xtime_lock'ing around cpufreq notifier References: <20070411162904.232696302@mvista.com> <461E6FA7.5080605@goop.org> <20070412105559.385789dd.akpm@linux-foundation.org> <200704122027.14926.ak@novell.com> In-Reply-To: <200704122027.14926.ak@novell.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Andi Kleen wrote: > They should always just store the cpu too and educate the users that > only (cpu, timestamp) pairs make sense to compare. > > That said at least my new sched_clock should not normally show > large non differences between CPUs, so it can be usually ignored, but they can > happen. I believe some of the already existing sched_clocks() (like the > one used on Altix) have the same property. > > But on VMI/Xen as currently implemented the differences will be large. Yes, it's pretty much unavoidable. It seems to me that these non-scheduler uses of sched_clock should use something else. Or we could add a new clock function for the scheduler to use, but "sched_clock" seems like the best name for it. J