Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752598AbaLTBGc (ORCPT ); Fri, 19 Dec 2014 20:06:32 -0500 Received: from www.linutronix.de ([62.245.132.108]:36387 "EHLO Galois.linutronix.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751003AbaLTBGb (ORCPT ); Fri, 19 Dec 2014 20:06:31 -0500 Date: Sat, 20 Dec 2014 02:06:12 +0100 (CET) From: Thomas Gleixner To: Chris Mason cc: Linus Torvalds , Dave Jones , Mike Galbraith , Ingo Molnar , Peter Zijlstra , =?ISO-8859-15?Q?D=E2niel_Fraga?= , Sasha Levin , "Paul E. McKenney" , Linux Kernel Mailing List , Suresh Siddha , Oleg Nesterov , Peter Anvin Subject: Re: frequent lockups in 3.18rc4 In-Reply-To: <1419034369.13012.8@mail.thefacebook.com> Message-ID: References: <20141218051327.GA31988@redhat.com> <1418918059.17358.6@mail.thefacebook.com> <20141218161230.GA6042@redhat.com> <20141219024549.GB1671@redhat.com> <20141219035859.GA20022@redhat.com> <20141219040308.GB20022@redhat.com> <20141219145528.GC13404@redhat.com> <20141219203135.GA1200@ret.masoncoding.com> <1419034369.13012.8@mail.thefacebook.com> User-Agent: Alpine 2.11 (DEB 23 2013-08-11) 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,URIBL_BLOCKED=0.001 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, 19 Dec 2014, Chris Mason wrote: > On Fri, Dec 19, 2014 at 6:22 PM, Thomas Gleixner wrote: > > But at the very end this would be detected by the runtime check of the > > hrtimer interrupt, which does not trigger. And it would trigger at > > some point as ALL cpus including CPU0 in that trace dump make > > progress. > > I'll admit that at some point we should be hitting one of the WARN or BUG_ON, > but it's possible to thread that needle and corrupt the timer list, without > hitting a warning (CPU 1 in my example has to enqueue last). Once the rbtree > is hosed, it can go forever. Probably not the bug we're looking for, but > still suspect in general. I surely have a close look at that, but in that case we get out of that state later on and I doubt that we have A) a corruption of the rbtree B) a self healing of the rbtree afterwards I doubt it, but who knows. Though even if A & B would happen we would still get the 'hrtimer interrupt took a gazillion of seconds' warning because CPU0 definitely leaves the timer interrupt at some point otherwise we would not see backtraces from usb, userspace and idle later on. Thanks, tglx -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/