From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: small RPS cache for fragments? Date: Wed, 18 May 2011 02:37:59 -0400 (EDT) Message-ID: <20110518.023759.624266793051870646.davem@davemloft.net> References: <20110517.164929.1737248436066795381.davem@davemloft.net> Mime-Version: 1.0 Content-Type: Text/Plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: therbert@google.com, netdev@vger.kernel.org To: xiaosuo@gmail.com Return-path: Received: from shards.monkeyblade.net ([198.137.202.13]:45836 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753982Ab1ERGiE convert rfc822-to-8bit (ORCPT ); Wed, 18 May 2011 02:38:04 -0400 In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: =46rom: Changli Gao Date: Wed, 18 May 2011 07:59:05 +0800 > On Wed, May 18, 2011 at 4:49 AM, David Miller w= rote: >> >> Actually, I think it won't work. =A0Even Linux emits fragments last = to >> first, so we won't see the UDP header until the last packet where it= 's >> no longer useful. >> >=20 > No. Linux emits fragments first to last now. You should check the > current code. :) I forgot that we rearranged this, thanks :-) So maybe the original idea can indeed work.