From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263372AbVGOSOG (ORCPT ); Fri, 15 Jul 2005 14:14:06 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263353AbVGOSLd (ORCPT ); Fri, 15 Jul 2005 14:11:33 -0400 Received: from fmr22.intel.com ([143.183.121.14]:10202 "EHLO scsfmr002.sc.intel.com") by vger.kernel.org with ESMTP id S261997AbVGOSKM (ORCPT ); Fri, 15 Jul 2005 14:10:12 -0400 Date: Fri, 15 Jul 2005 11:09:21 -0700 From: Venkatesh Pallipadi To: Andi Kleen Cc: Venkatesh Pallipadi , "Brown, Len" , akpm@osdl.org, linux-kernel@vger.kernel.org, torvalds@osdl.org, vojtech@suse.cz, christoph@lameter.com Subject: Re: [PATCH] i386: Selectable Frequency of the Timer Interrupt Message-ID: <20050715110920.B16008@unix-os.sc.intel.com> References: <20050715102349.A15791@unix-os.sc.intel.com> <20050715175700.GE15783@wotan.suse.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5.1i In-Reply-To: <20050715175700.GE15783@wotan.suse.de>; from ak@suse.de on Fri, Jul 15, 2005 at 07:57:01PM +0200 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Jul 15, 2005 at 07:57:01PM +0200, Andi Kleen wrote: > > I wouldn't say it is totally impossible. There are ways in which Linux can work > > without a reliable Local APIC timer. One option being - make one CPU that gets > > the external timer interrupt multicast an IPI to all the other CPUs that > > wants to get periodic timer interrupt. > > That doesn't mix very well with variable ticks. And I believe > we really need them. It should work with variable ticks as we can easily add/remove CPUs from this multicast destination. > For no tick in idle you need a timer for each CPU that > can be programmed to a reasonably long interval to wake you > up after longer idleness. And all that > should work without bouncing cache lines around all the time > because that doesn't work on larger systems. And each CPU timer itself does very little writes in terms of caches. The two things that happen here are scheduler_tick and kstat_accounting. Or am I missing something? Did you have anything specific in mind wrt bouncing cachelines? Thanks, Venki