From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1768620Ab2KOR75 (ORCPT ); Thu, 15 Nov 2012 12:59:57 -0500 Received: from www.linutronix.de ([62.245.132.108]:36505 "EHLO Galois.linutronix.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1768575Ab2KOR74 (ORCPT ); Thu, 15 Nov 2012 12:59:56 -0500 Date: Thu, 15 Nov 2012 18:59:54 +0100 (CET) From: Thomas Gleixner To: Grant Likely cc: LKML , linux-rt-users Subject: Re: [ANNOUNCE] 3.6.5-rt15 In-Reply-To: Message-ID: References: User-Agent: Alpine 2.02 (LFD 1266 2009-07-14) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Linutronix-Spam-Score: -1.0 X-Linutronix-Spam-Level: - X-Linutronix-Spam-Status: No , -1.0 points, 5.0 required, ALL_TRUSTED=-1,SHORTCIRCUIT=-0.0001 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 14 Nov 2012, Grant Likely wrote: > Hey Thomas, > > I get a kernel bug when going from rt14 to rt15. It appears that > preempt_lazy on my ARM platform causes (or exposes) the following > backtrace on boot. I'm basing this on the only relevant change between > rt14 and rt15 is the typo fix that enables PREEMPT_LAZY. I also see > the problem on -rt16 and -rt17. > > (full boot log also attached below) > > The actual backtrace changes every time, but it is always related to > an rt_spin_lock() call. > > This is on a TI AM3894 platform which is equivalent to a TI8160 > without the DSP. The board support isn't mainlined, but there is > nothing particularly exotic about the changes. It is primarily device > drivers and device driver data. Cortex-A8 platform. > > Any thoughts on how to debug this? Probably by staring at the patch to the ASM code in arch/arm. I bet I messed that up somehow :) Thanks, tglx