From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Thu, 27 Feb 2003 15:55:55 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Thu, 27 Feb 2003 15:55:55 -0500 Received: from web20709.mail.yahoo.com ([216.136.226.182]:25725 "HELO web20709.mail.yahoo.com") by vger.kernel.org with SMTP id ; Thu, 27 Feb 2003 15:55:54 -0500 Message-ID: <20030227210613.17243.qmail@web20709.mail.yahoo.com> Date: Thu, 27 Feb 2003 13:06:13 -0800 (PST) From: Paul Adams Subject: PCI bridge disconnects and retries To: linux-kernel MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I have a DMA Bus Master device on a PCI-X or PCI 64/66 bus under Linux/x86 (UP and SMP). The device is capable of long bursts without wait states, but the various host-PCI bridges I have tried will disconnect after very short bursts or will require many retries before starting a burst. I would be interested in any suggestions for configuring the CPU cache or the bridge or ways to tune the behavior of the device to optimize sustained throughput. I am looking at a number of standard chipsets such as Serverworks Grand Champion, Intel E7500, and AMD-760 MPX. For example, how can the prefetch behavior on each of these bridges be tuned? Are specs available for any of the bridges? Can the driver do anything to make better use of caches and prefetching? Paul Adams __________________________________________________ Do you Yahoo!? Yahoo! Tax Center - forms, calculators, tips, more http://taxes.yahoo.com/