From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <benh@kernel.crashing.org>
Received: from gate.crashing.org (gate.crashing.org [63.228.1.57])
 (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits))
 (Client did not present a certificate)
 by ozlabs.org (Postfix) with ESMTPS id DFBF4B6FBB
 for <linuxppc-dev@lists.ozlabs.org>; Fri,  4 May 2012 10:10:26 +1000 (EST)
Message-ID: <1336090207.18712.9.camel@pasglop>
Subject: Re: [PATCH] powerpc: use local var instead of
 local_paca->irq_happened directly in __check_irq_replay
From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
To: Wang Sheng-Hui <shhuiw@gmail.com>
Date: Fri, 04 May 2012 10:10:07 +1000
In-Reply-To: <4FA31653.5090103@gmail.com>
References: <4FA1E527.1090807@gmail.com> <1336011306.2653.3.camel@pasglop>
 <1336018961.2653.11.camel@pasglop> <4FA226A8.1080903@gmail.com>
 <4FA22CC5.3010107@gmail.com> <1336032561.18712.1.camel@pasglop>
 <4FA31653.5090103@gmail.com>
Content-Type: text/plain; charset="UTF-8"
Mime-Version: 1.0
Cc: Stephen Rothwell <sfr@canb.auug.org.au>, linux-kernel@vger.kernel.org,
 Milton Miller <miltonm@bga.com>, Anton Blanchard <anton@samba.org>,
 linuxppc-dev@lists.ozlabs.org
List-Id: Linux on PowerPC Developers Mail List <linuxppc-dev.lists.ozlabs.org>
List-Unsubscribe: <https://lists.ozlabs.org/options/linuxppc-dev>,
 <mailto:linuxppc-dev-request@lists.ozlabs.org?subject=unsubscribe>
List-Archive: <http://lists.ozlabs.org/pipermail/linuxppc-dev>
List-Post: <mailto:linuxppc-dev@lists.ozlabs.org>
List-Help: <mailto:linuxppc-dev-request@lists.ozlabs.org?subject=help>
List-Subscribe: <https://lists.ozlabs.org/listinfo/linuxppc-dev>,
 <mailto:linuxppc-dev-request@lists.ozlabs.org?subject=subscribe>

On Fri, 2012-05-04 at 07:35 +0800, Wang Sheng-Hui wrote:
> > No, Only comment the test, you must absolutely leave the
> > __hard_irq_disable() call ! That's the whole point of the test, make
> > sure we unconditionally disable to see if that fixes the problem, in
> > which case that will tell us that we somewhere accidentally leave
> > irq_happened set to 0x01 while irqs are hard enabled.
> 
> It can work.
> My system has been running for about 15 hours without crash.

Ok, so now we need to understand under what circumstances we end up
in a situation where paca->irq_happened is 0x01 and IRQs are hard
enabled. I have a few ideas of things to look at but I'm also off
for the week-end.

I'll have a look next week.

Cheers,
Ben.