From mboxrd@z Thu Jan 1 00:00:00 1970 From: jamal Subject: Re: [PATCH 3/3] ifb: move tq from ifb_private Date: Sat, 04 Dec 2010 09:42:36 -0500 Message-ID: <1291473756.2159.31.camel@mojatatu> References: <1291442121-3302-1-git-send-email-xiaosuo@gmail.com> <1291442121-3302-3-git-send-email-xiaosuo@gmail.com> <1291472282.2159.8.camel@mojatatu> Reply-To: hadi@cyberus.ca Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org To: Changli Gao Return-path: Received: from mail-gx0-f180.google.com ([209.85.161.180]:34406 "EHLO mail-gx0-f180.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755114Ab0LDOmm (ORCPT ); Sat, 4 Dec 2010 09:42:42 -0500 Received: by gxk19 with SMTP id 19so5661780gxk.11 for ; Sat, 04 Dec 2010 06:42:41 -0800 (PST) In-Reply-To: <1291472282.2159.8.camel@mojatatu> Sender: netdev-owner@vger.kernel.org List-ID: On Sat, 2010-12-04 at 09:18 -0500, jamal wrote: > I like the splice idea but this patch makes me twitch > a little. What test setup did you use to check it? Ok, here's one thing you changed which is important. We do: -->XXX-->rq-->tq-->XXX--> rq is controlled by queue limit. We only load rq to tq if all of tq is empty. If it is not we dont move things over. Essentially this is a flow control scheme. We dont want many sources to be overwhelming us with packets and every time we grab a txqlen number of packets. For this reason: I would be comfortable if all you did was to add the splice after you skb_peek() - i think that would be a good improvement which is not bound to break anything else. cheers, jamal