From mboxrd@z Thu Jan 1 00:00:00 1970 From: Cong Wang Subject: Re: [PATCH] net/802/mrp: fix lockdep splat Date: Tue, 14 May 2013 09:12:20 +0800 Message-ID: <1368493940.30953.3.camel@cr0> References: <6901a7785249816fe13a98e3c0b35127@visp.net.lb> <1368447851.13473.119.camel@edumazet-glaptop> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: Denys Fedoryshchenko , David Ward , hayeswang , Francois Romieu , netdev@vger.kernel.org To: Eric Dumazet Return-path: Received: from mx1.redhat.com ([209.132.183.28]:3553 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752642Ab3ENBMg (ORCPT ); Mon, 13 May 2013 21:12:36 -0400 In-Reply-To: <1368447851.13473.119.camel@edumazet-glaptop> Sender: netdev-owner@vger.kernel.org List-ID: On Mon, 2013-05-13 at 05:24 -0700, Eric Dumazet wrote: > del_timer_sync(&app->join_timer); > > - spin_lock(&app->lock); > + spin_lock_bh(&app->lock); But we just have removed the join timer before taking this spin lock, so it is not possible to race with timer, but we still have race condition with mrp_rcv() which is called in BH context too. However, this is not what Denys reported, so I have no idea on this.