From mboxrd@z Thu Jan 1 00:00:00 1970 From: ebiederm@xmission.com (Eric W. Biederman) Subject: Re: [PATCH 1/1] net: race condition when removing virtual net_device Date: Fri, 13 Sep 2013 18:32:52 -0700 Message-ID: <87fvt8tcfv.fsf@xmission.com> References: <1379008796-2121-1-git-send-email-fruggeri@aristanetworks.com> <87txhp249u.fsf@xmission.com> <20130913.201626.1381750756413200302.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain Cc: fruggeri@aristanetworks.com, edumazet@google.com, jiri@resnulli.us, alexander.h.duyck@intel.com, amwang@redhat.com, netdev@vger.kernel.org To: David Miller Return-path: Received: from out03.mta.xmission.com ([166.70.13.233]:48016 "EHLO out03.mta.xmission.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755913Ab3INBdA (ORCPT ); Fri, 13 Sep 2013 21:33:00 -0400 In-Reply-To: <20130913.201626.1381750756413200302.davem@davemloft.net> (David Miller's message of "Fri, 13 Sep 2013 20:16:26 -0400 (EDT)") Sender: netdev-owner@vger.kernel.org List-ID: David Miller writes: > From: ebiederm@xmission.com (Eric W. Biederman) > Date: Thu, 12 Sep 2013 13:06:53 -0700 > >> David, Eric, do any of you know why we have this netdevice notfier >> rebroadcast? > > It's been there as long as I can remember, but I don't remember > exactly why we do that. > > If I had to guess, it's to handle something that has to retry later > due to a "spin_trylock()" or similar type of situation. > > Probably best to not remove it. :) Since it doesn't completely solve the problem, it isn't worth it to solve this problem. Sigh I have added this bit of code and the moving of references to the loopback device in dst_ifdown to my hitlist of things to sort out some day when I have plenty time. Eric