From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752582Ab3CAVxy (ORCPT ); Fri, 1 Mar 2013 16:53:54 -0500 Received: from e39.co.us.ibm.com ([32.97.110.160]:43317 "EHLO e39.co.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752264Ab3CAVxx (ORCPT ); Fri, 1 Mar 2013 16:53:53 -0500 Date: Fri, 1 Mar 2013 13:53:46 -0800 From: "Paul E. McKenney" To: Andrew Morton Cc: Jan Kara , LKML Subject: Re: [PATCH v3] printk: Avoid softlockups in console_unlock() Message-ID: <20130301215346.GF3285@linux.vnet.ibm.com> Reply-To: paulmck@linux.vnet.ibm.com References: <1361927819-1570-1-git-send-email-jack@suse.cz> <20130227175945.GC3336@linux.vnet.ibm.com> <20130301103027.GC26128@quack.suse.cz> <20130301134321.85e18987.akpm@linux-foundation.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20130301134321.85e18987.akpm@linux-foundation.org> User-Agent: Mutt/1.5.21 (2010-09-15) X-Content-Scanned: Fidelis XPS MAILER x-cbid: 13030121-3620-0000-0000-0000016EFAD3 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Mar 01, 2013 at 01:43:21PM -0800, Andrew Morton wrote: > On Fri, 1 Mar 2013 11:30:27 +0100 > Jan Kara wrote: > > > > If a CPU remains in your loop where you invoke rcu_cpu_stall_reset(), > > > there is no indication of the stall. Now, I do understand why having > > > RCU add a CPU stall warning into the mix would be very bad, but it > > > would be good to have some sort of indication that there is stalling. > > > Perhaps a counter visible in sysfs? Whatever it is, at least some way > > > of diagnosing the stall condition would be very good. > > I see. If others (Andrew, ping?) are OK with the rest of the patch, I can > > extend it and add a counter of printk() breaks we had to make. That looks > > like a good idea. > > hmpf, spose so. It sound lke the world's least interesting counter but > perhaps it will look more interesting when we see its sparkly > documentation. Boredom is a good thing here, given that the goal is to reduce excitement when bad things happen. Or at least to increase awareness of root causes. Thanx, Paul