From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stephen Warren Subject: Re: cpufreq_stats NULL deref on second system suspend Date: Tue, 10 Sep 2013 14:53:01 -0600 Message-ID: <522F86AD.6010603@wwwdotorg.org> References: <522E1FEF.6080803@wwwdotorg.org> <6639035.alELXsX8ZI@vostro.rjw.lan> <522E3DA2.9020707@wwwdotorg.org> <1775778.MeiRhuYy7o@vostro.rjw.lan> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: Received: from avon.wwwdotorg.org ([70.85.31.133]:57749 "EHLO avon.wwwdotorg.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751413Ab3IJUxG (ORCPT ); Tue, 10 Sep 2013 16:53:06 -0400 In-Reply-To: <1775778.MeiRhuYy7o@vostro.rjw.lan> Sender: linux-pm-owner@vger.kernel.org List-Id: linux-pm@vger.kernel.org To: "Rafael J. Wysocki" Cc: Viresh Kumar , "linux-pm@vger.kernel.org" , "Srivatsa S. Bhat" On 09/09/2013 05:14 PM, Rafael J. Wysocki wrote: > On Monday, September 09, 2013 03:29:06 PM Stephen Warren wrote: >> On 09/09/2013 02:24 PM, Rafael J. Wysocki wrote: >>> On Monday, September 09, 2013 02:01:32 PM Stephen Warren wrote: >>>> On 09/09/2013 02:01 PM, Rafael J. Wysocki wrote: >>>>> On Monday, September 09, 2013 01:22:23 PM Stephen Warren wrote: >>>>>> Viresh, >>>>>> >>>>>> I'm seeing the crash below when suspending my system for the second time. >>>>>> >>>>>> I can avoid this with the following patch, which adds a check which >>>>>> already exists in all-but-one other places that the same lookup is made: >>>>> >>>>> Which kernel did you test? >>>> >>>> next-20130909. >>> >>> Is it reproducible with the current mainline? >> >> This does not affect v3.11, but does affect current HEAD; 300893b "Merge >> tag 'xfs-for-linus-v3.12-rc1' of git://oss.sgi.com/xfs/xfs". > > What system does it break on? A dual-core ARM system (NVIDIA Tegra20 SoC, Harmony board). > Any chance to bisect cpufreq changes between 3.11 and the current HEAD? Sure, it's due to 5302c3f "cpufreq: Perform light-weight init/teardown during suspend/resume".