From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andi Kleen Subject: Re: [PATCH] Disable TSO for non standard qdiscs Date: Fri, 1 Feb 2008 15:34:21 +0100 Message-ID: <20080201143421.GB16630@one.firstfloor.org> References: <20080131100846.00934e25@extreme> <20080131185328.GD4671@one.firstfloor.org> <47A211A0.1040502@trash.net> <20080131190125.GE4671@one.firstfloor.org> <20080131193406.GH4671@one.firstfloor.org> <1201847738.17656.5.camel@roma.44ansell.gdt.id.au> <20080201074600.GA12644@one.firstfloor.org> <47A2C985.30409@trash.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Patrick McHardy , Andi Kleen , Glen Turner , Stephen Hemminger , netdev@vger.kernel.org To: "Waskiewicz Jr, Peter P" Return-path: Received: from one.firstfloor.org ([213.235.205.2]:42701 "EHLO one.firstfloor.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753252AbYBAN7u (ORCPT ); Fri, 1 Feb 2008 08:59:50 -0500 Content-Disposition: inline In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: > The TSO defer logic is based on your congestion window and current > window size. So the actual frame sizes hitting your NIC attached to > your DSL probably aren't anywhere near 64KB, but probably more in line > with whatever your window size is for DSL. DSL windows can be quite large because a lot of DSL lines have a quite long latency due to error correction. And with ADSL2 we have upto 16Mbit now. > I think we're having more of a disagreement of what is considered the > "normal case" user. If you are on a slow link, such as a DSL/cable > line, your TCP window/congestion window aren't going to be big enough to > generate large TSO's, so what is the issue? But disabling TSO, say on a 64k TSOs are likely even with DSL. Anyways even with smaller TSOs the change still makes sense because each increase makes packet scheduling less smooth. -Andi