From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH net-next-2.6] rps: consistent rxhash Date: Tue, 20 Apr 2010 16:38:53 -0700 (PDT) Message-ID: <20100420.163853.258272672.davem@davemloft.net> References: <1271775421.7895.19.camel@edumazet-laptop> <20100420.144106.118596093.davem@davemloft.net> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: eric.dumazet@gmail.com, franco@lastsummer.de, therbert@google.com, netdev@vger.kernel.org To: xiaosuo@gmail.com Return-path: Received: from 74-93-104-97-Washington.hfc.comcastbusiness.net ([74.93.104.97]:55671 "EHLO sunset.davemloft.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752004Ab0DTXis (ORCPT ); Tue, 20 Apr 2010 19:38:48 -0400 In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: From: Changli Gao Date: Wed, 21 Apr 2010 07:35:48 +0800 > On Wed, Apr 21, 2010 at 5:41 AM, David Miller wrote: >> Eric, do you remember that "TCP friends" rough patch I sent you last >> year that essentailly made TCP sockets over loopback behave like >> AF_UNIX ones and just queue the SKBs directly to the destination >> socket without doing any protocol work? > > I think it will break some benchmark tools. Other systems already do this optimization, so if things break, this breakage is already pervasive. We should be able to tell people that they can use TCP solely in their applications and it will perform optimally regardless of transport. People already code their applications this way, and ignoring this issue would just makes us stupid.