From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH] e100: Fix the TX workqueue race Date: Fri, 23 Apr 2010 16:31:27 -0700 (PDT) Message-ID: <20100423.163127.62645049.davem@davemloft.net> References: <20100423143356.7092.45260.stgit@localhost.localdomain> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, e1000-devel@lists.sourceforge.net To: alan@linux.intel.com Return-path: Received: from 74-93-104-97-Washington.hfc.comcastbusiness.net ([74.93.104.97]:45973 "EHLO sunset.davemloft.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755449Ab0DWXbX (ORCPT ); Fri, 23 Apr 2010 19:31:23 -0400 In-Reply-To: <20100423143356.7092.45260.stgit@localhost.localdomain> Sender: netdev-owner@vger.kernel.org List-ID: From: Alan Cox Date: Fri, 23 Apr 2010 15:34:43 +0100 > I'd assumed someone would have picked up on this and fixed it using rtnl_lock > as was suggested but it seems to have fallen through the cracks ? > > Anyway this is I assume what was meant ? I hope this doesn't deadlock with linkwatch, as that's generally a problem we hit with trying to take RTNL from workqueues in the networking. Linkwatch takes the RTNL lock, and then can make calls into the driver in it's main work loop. But since you don't hold any driver locks here (you can't as if we did we couldn't take the RTNL lock here at all) so it should be OK. I'll apply this to net-2.6, thanks Alan.