From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757231AbaEIR3G (ORCPT ); Fri, 9 May 2014 13:29:06 -0400 Received: from g4t3426.houston.hp.com ([15.201.208.54]:45973 "EHLO g4t3426.houston.hp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756553AbaEIR3E (ORCPT ); Fri, 9 May 2014 13:29:04 -0400 Message-ID: <1399656540.2169.20.camel@j-VirtualBox> Subject: Re: [PATCH 2/2] sched: Fix next_balance logic in rebalance_domains() and idle_balance() From: Jason Low To: Peter Zijlstra Cc: Ingo Molnar , linux-kernel@vger.kernel.org, daniel.lezcano@linaro.org, alex.shi@linaro.org, preeti@linux.vnet.ibm.com, efault@gmx.de, vincent.guittot@linaro.org, morten.rasmussen@arm.com, aswin@hp.com, jason.low2@hp.com Date: Fri, 09 May 2014 10:29:00 -0700 In-Reply-To: <20140509090805.GK30445@twins.programming.kicks-ass.net> References: <1398725155-7591-1-git-send-email-jason.low2@hp.com> <1398725155-7591-3-git-send-email-jason.low2@hp.com> <1399568364.2030.9.camel@j-VirtualBox> <20140508173835.GB9838@gmail.com> <1399587244.2030.59.camel@j-VirtualBox> <1399596562.2200.7.camel@j-VirtualBox> <20140509090805.GK30445@twins.programming.kicks-ass.net> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.2.3-0ubuntu6 Content-Transfer-Encoding: 7bit Mime-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, 2014-05-09 at 11:08 +0200, Peter Zijlstra wrote: > On Thu, May 08, 2014 at 05:49:22PM -0700, Jason Low wrote: > > On Thu, 2014-05-08 at 15:14 -0700, Jason Low wrote: > > > Hi Ingo, Peter, > > > > > > Were there NULL domains on the test system? If so, I think we can > > > address the problem by doing update_next_balance() only if the below > > > rcu_dereference_check_sched_domain() returns a non-null domain. > > > > Also, below is a patch which has the change: > > > > Thanks, and yes, NULL domains are a perfectly normal thing to have > (though not common on normal boot paths). Okay. Likewise, on my test machine, I offline all but 1 CPU in order to try reproducing the issue. Thanks, Jason