From mboxrd@z Thu Jan 1 00:00:00 1970 From: Patrick McHardy Subject: Re: [PATCH net-next-2.6] macvlan: add multiqueue capability Date: Thu, 03 Sep 2009 20:27:45 +0200 Message-ID: <4AA00AA1.6040504@trash.net> References: <4A9F9661.7020301@gmail.com> <4AA002BC.3050507@trash.net> <4AA008C9.4000805@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: "David S. Miller" , Linux Netdev List To: Eric Dumazet Return-path: Received: from stinky.trash.net ([213.144.137.162]:45120 "EHLO stinky.trash.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754006AbZICS1r (ORCPT ); Thu, 3 Sep 2009 14:27:47 -0400 In-Reply-To: <4AA008C9.4000805@gmail.com> Sender: netdev-owner@vger.kernel.org List-ID: Eric Dumazet wrote: > Patrick McHardy a =E9crit : >> The patch looks fine, but it just occured to me that this won't >> have any effect since both VLAN and macvlan use a tx_queue_len of 0, >> so they will by default have queueing disabled. In fact this >> will increase costs for the default case since we're now hashing >> every packet. >=20 > Just read again dev_queue_xmit(), in case we have no queueing > on macvlan/vlan >=20 > Having mutiple txq should help multi flow / multi cpus setups, > since hashing will provide more chances to hit different txq/locks, > and let several cpus run concurrently, each one on a different queue. You're right, I missed that we're also perfoming locking in the noqueue case. Sorry :)