From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [fixed] [patch] Re: [bug] stuck localhost TCP connections, v2.6.26-rc3+ Date: Tue, 10 Jun 2008 15:49:29 -0700 (PDT) Message-ID: <20080610.154929.168124808.davem@davemloft.net> References: <20080606183926.GB12651@elte.hu> <1212782937.23706.46.camel@tng> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: mingo@elte.hu, ilpo.jarvinen@helsinki.fi, peterz@infradead.org, linux-kernel@vger.kernel.org, netdev@vger.kernel.org, rjw@sisk.pl, akpm@linux-foundation.org, johnpol@2ka.mipt.ru To: mcmanus@ducksong.com Return-path: Received: from 74-93-104-97-Washington.hfc.comcastbusiness.net ([74.93.104.97]:54002 "EHLO sunset.davemloft.net" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1753657AbYFJWtd (ORCPT ); Tue, 10 Jun 2008 18:49:33 -0400 In-Reply-To: <1212782937.23706.46.camel@tng> Sender: netdev-owner@vger.kernel.org List-ID: From: Patrick McManus Date: Fri, 06 Jun 2008 16:08:57 -0400 > If we're sure of that conclusion we should just take Ilpo's DA patch as > that will narrow the field for finding Hakon's issue. While the conclusion might be accurate, Ilpo's patch adds a lock ordering problem that might introduce potential deadlocks (worst case) or at least some lockdep warnings. Both of which are unacceptable. I personally don't see any clear solution, and because this new DA code moves an operation into a code block where the locking order pretty much has to be inverted, I don't see any real short term solution other than a revert of this particular DA commit.