From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263705AbTKXK3C (ORCPT ); Mon, 24 Nov 2003 05:29:02 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263717AbTKXK3B (ORCPT ); Mon, 24 Nov 2003 05:29:01 -0500 Received: from gate.corvil.net ([213.94.219.177]:14094 "EHLO corvil.com") by vger.kernel.org with ESMTP id S263705AbTKXK3A (ORCPT ); Mon, 24 Nov 2003 05:29:00 -0500 Message-ID: <3FC1DD45.9070107@draigBrady.com> Date: Mon, 24 Nov 2003 10:28:21 +0000 From: P@draigBrady.com User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.5) Gecko/20031016 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Oded Comay CC: linux-kernel@vger.kernel.org Subject: Re: Is PACKET_RX_RING broken in 2.6.0-test? References: In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Oded Comay wrote: > The PACKET_RX_RING mechanism (enabled by CONFIG_PACKET_MMAP) appears to be > somewhat broken in 2.6.0-test* (2.6 for short). There are 3 issues: > > 1. Performance in 2.6 is much lower than in 2.4. As an example, using the > same hardware and under the same traffic load, the attached sample program > could process 177K packets/sec in 2.4, but only 70K packets/sec in 2.6. I'm guessing the scheduling changes are the cause? Try messing with sched_setscheduler (like: http://awgn.antifork.org/codes/brute.c) Anyone else notice that the packet buffer messes up if packets are being received while it's being created. I worked around it here by bringing the interfaces up AFTER the buffer is created. Pádraig.