From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stephen Hemminger Subject: Re: [Bugme-new] [Bug 32832] New: shutdown(2) does not fully shut down socket any more Date: Wed, 13 Apr 2011 11:47:09 -0700 Message-ID: <20110413114709.32f5931e@nehalam> References: <1302663327.2811.55.camel@edumazet-laptop> <1302663608.2811.62.camel@edumazet-laptop> <20110413.104317.71110417.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: dbaluta@ixiacom.com, eric.dumazet@gmail.com, akpm@linux-foundation.org, netdev@vger.kernel.org, bugzilla-daemon@bugzilla.kernel.org, bugme-daemon@bugzilla.kernel.org, kees@outflux.net To: David Miller Return-path: Received: from mail.vyatta.com ([76.74.103.46]:38709 "EHLO mail.vyatta.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756652Ab1DMSrM (ORCPT ); Wed, 13 Apr 2011 14:47:12 -0400 In-Reply-To: <20110413.104317.71110417.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: On Wed, 13 Apr 2011 10:43:17 -0700 (PDT) David Miller wrote: > From: Daniel Baluta > Date: Wed, 13 Apr 2011 14:57:18 +0300 > > > Cyril's use case looks suspect. I don't think that this is a good > > reason for reverting this commit. > > I complete disagree. > > Something that worked perfectly fine, probably for years, we broke. > > We simply cannot do that, especially since we do not have a reasonable > alternative at this time. > > Adding SO_REUSEPORT is a long range option, and not something that > will provide a fix for users right now. > > So please don't even pretend to suggest that we shouldn't fix this > with a revert unless a simple, obvious, kernel fix presents itself. Just to echo what Dave said. Even though the semantics of this is not documented in some standard, applications have been built on Linux expecting a certain behavior. If you want to change what happens in this case, you have to have a really good reason (like crash, security hole, or standards violation).