From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from rune.pobox.com (rune.pobox.com [208.210.124.79]) by ozlabs.org (Postfix) with ESMTP id 5B166679E1 for ; Tue, 4 Apr 2006 00:18:30 +1000 (EST) Date: Mon, 3 Apr 2006 09:18:05 -0500 From: Nathan Lynch To: Andi Kleen Subject: Re: Fw: 2.6.16 crashes when running numastat on p575 Message-ID: <20060403141805.GC25663@localdomain> References: <20060402213216.2e61b74e.akpm@osdl.org> <20060402221513.96f05bdc.pj@sgi.com> <200604031349.03036.ak@suse.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <200604031349.03036.ak@suse.de> Cc: akpm@osdl.org, ak@suse.com, linux-kernel@vger.kernel.org, linuxppc-dev@ozlabs.org, Paul Jackson , Christoph Lameter List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Andi Kleen wrote: > On Monday 03 April 2006 07:15, Paul Jackson wrote: > > - for (cpu = 0; cpu < NR_CPUS; cpu++) { > > + for_each_online_cpu(cpu) { > > > > Idle curiosity -- what keeps a cpu from going offline during > > this scan, and leaving us with the same crash as before? > > > CPU hotdown uses RCU like techniques to avoid this. Only potential > problem could be on a preemptive kernel, but I hope nobody tries > cpu unplug on such a beast. I always turn on preempt when testing cpu hotplug (usually on ppc64). There were several preempt vs cpu hotplug issues until around 2.6.10 or so, iirc, but I think the situation has been stable for a while now. From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751233AbWDCOSb (ORCPT ); Mon, 3 Apr 2006 10:18:31 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751277AbWDCOSb (ORCPT ); Mon, 3 Apr 2006 10:18:31 -0400 Received: from rune.pobox.com ([208.210.124.79]:50326 "EHLO rune.pobox.com") by vger.kernel.org with ESMTP id S1751233AbWDCOSa (ORCPT ); Mon, 3 Apr 2006 10:18:30 -0400 Date: Mon, 3 Apr 2006 09:18:05 -0500 From: Nathan Lynch To: Andi Kleen Cc: Paul Jackson , akpm@osdl.org, linux-kernel@vger.kernel.org, linuxppc-dev@ozlabs.org, ak@suse.com, Christoph Lameter Subject: Re: Fw: 2.6.16 crashes when running numastat on p575 Message-ID: <20060403141805.GC25663@localdomain> References: <20060402213216.2e61b74e.akpm@osdl.org> <20060402221513.96f05bdc.pj@sgi.com> <200604031349.03036.ak@suse.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <200604031349.03036.ak@suse.de> User-Agent: Mutt/1.5.9i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Andi Kleen wrote: > On Monday 03 April 2006 07:15, Paul Jackson wrote: > > - for (cpu = 0; cpu < NR_CPUS; cpu++) { > > + for_each_online_cpu(cpu) { > > > > Idle curiosity -- what keeps a cpu from going offline during > > this scan, and leaving us with the same crash as before? > > > CPU hotdown uses RCU like techniques to avoid this. Only potential > problem could be on a preemptive kernel, but I hope nobody tries > cpu unplug on such a beast. I always turn on preempt when testing cpu hotplug (usually on ppc64). There were several preempt vs cpu hotplug issues until around 2.6.10 or so, iirc, but I think the situation has been stable for a while now.