From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH net-next] ipv4: tcp: dont cache unconfirmed intput dst Date: Wed, 27 Jun 2012 17:01:01 -0700 (PDT) Message-ID: <20120627.170101.99084491660488389.davem@davemloft.net> References: <1340788455.26242.67.camel@edumazet-glaptop> <20120627.153454.30398632011109264.davem@davemloft.net> <20120627.164418.1928194990434756968.davem@davemloft.net> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, hans.schillstrom@ericsson.com To: eric.dumazet@gmail.com Return-path: Received: from shards.monkeyblade.net ([149.20.54.216]:48307 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754013Ab2F1ABC (ORCPT ); Wed, 27 Jun 2012 20:01:02 -0400 In-Reply-To: <20120627.164418.1928194990434756968.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: From: David Miller Date: Wed, 27 Jun 2012 16:44:18 -0700 (PDT) > What happens now is that legitimate traffic is harmed too. If we > really go to established state, we'll cache the DST_NOCACHE route > in sk->sk_rx_dst. This change also means that all routed TCP traffic will use DST_NOCACHE routes as well. It's not a requirement to turn off early demux on a router, and I very much wanted to avoid the knob altogether. So this side effect is not acceptable. There are quite a number of unwanted side effects from this change, so I think we'll have to revert unless you can fix up all of the relevant cases quickly.