From mboxrd@z Thu Jan 1 00:00:00 1970 From: Paul Gortmaker Subject: Re: Crap, ksoftirqd/0 looping forever in softirq Date: Thu, 11 Oct 2012 10:21:31 -0400 Message-ID: <20121011142130.GA22583@windriver.com> References: <20121011001637.GA18628@windriver.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Cc: linux-rt-users To: Thomas Gleixner Return-path: Received: from mail.windriver.com ([147.11.1.11]:48169 "EHLO mail.windriver.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932177Ab2JKOVm (ORCPT ); Thu, 11 Oct 2012 10:21:42 -0400 Content-Disposition: inline In-Reply-To: <20121011001637.GA18628@windriver.com> Sender: linux-rt-users-owner@vger.kernel.org List-ID: [Crap, ksoftirqd/0 looping forever in softirq] On 10/10/2012 (Wed 20:16) Paul Gortmaker wrote: > Seeing the above message, I added a trivial change to print out the > Mask and Pending, and I consistently saw it was (both p/m) in every > case `was __IRQF_TIMER (0x200). Of cource it helps if I look up the value in the right list. I guess the ffs had me thinking "bits", instead of looking up the value in the softirq enum list. When I add a change to record the last vecs used and dump them out when the Crap message is triggered (using softirq_to_name instead of looking things up in random tables!) I get this: [ 20.887137] Crap, ksoftirqd/0 looping forever in softirq [ 20.887150] Last: RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU [ 20.920168] Crap, ksoftirqd/0 looping forever in softirq [ 20.920179] Last: RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU [ 20.954200] Crap, ksoftirqd/0 looping forever in softirq [ 20.954213] Last: RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU RCU Which makes my previous guess complete garbage... Paul.