From mboxrd@z Thu Jan 1 00:00:00 1970 From: jamal Subject: Re: [PATCH] NET: Multiqueue network device support. Date: Tue, 26 Jun 2007 09:27:28 -0400 Message-ID: <1182864448.5186.34.camel@localhost> References: <1182440388.5017.31.camel@localhost> <1182475614.30969.102.camel@debian.sh.intel.com> <1182790051.5184.29.camel@localhost> <20070625.134741.129320846.davem@davemloft.net> Reply-To: hadi@cyberus.ca Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Cc: yi.zhu@intel.com, kaber@trash.net, peter.p.waskiewicz.jr@intel.com, netdev@vger.kernel.org, jeff@garzik.org, auke-jan.h.kok@intel.com To: David Miller Return-path: Received: from wx-out-0506.google.com ([66.249.82.237]:7242 "EHLO wx-out-0506.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752046AbXFZN1d (ORCPT ); Tue, 26 Jun 2007 09:27:33 -0400 Received: by wx-out-0506.google.com with SMTP id t15so1768727wxc for ; Tue, 26 Jun 2007 06:27:32 -0700 (PDT) In-Reply-To: <20070625.134741.129320846.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Mon, 2007-25-06 at 13:47 -0700, David Miller wrote: > tg3 uses 1/4: > > #define TG3_TX_WAKEUP_THRESH(tp) ((tp)->tx_pending / 4) > Sorry - meant tg3 uses a much higher value (default is 128) - say relative to e1000 (default of 32). My tests with batching on e1000 indicate 128 gives the best results which is 1/2 if you look at it from tx ring size perspective. So i have hardcoded it in the git tree. > tp->tx_pending is the current configured ring size, configurable > via ethtool. Thats what e1000 needs as well - currently it hardcodes things. I think ethtool is the way to go - I will update the batch tree with that fix for e1000. Back to the question: Do you recall how this number was arrived at? 128 packets will be sent out at GiGe in about 80 microsecs, so from a feel-the-wind-direction perspective it seems reasonable. cheers, jamal