From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Tue, 29 May 2001 08:59:11 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Tue, 29 May 2001 08:59:02 -0400 Received: from zeus.kernel.org ([209.10.41.242]:23221 "EHLO zeus.kernel.org") by vger.kernel.org with ESMTP id ; Tue, 29 May 2001 08:58:46 -0400 Subject: Re: Status of ALi MAGiK 1 support in 2.4.? To: bdbryant@mail.utexas.edu (Bobby D. Bryant) Date: Tue, 29 May 2001 13:35:08 +0100 (BST) Cc: linux-kernel@vger.kernel.org (linux-kernel@vger.kernel.org), mfrisch@saturn.tlug.org, Axel.Thimm@physik.fu-berlin.de In-Reply-To: <3B12DCCF.6524A99@mail.utexas.edu> from "Bobby D. Bryant" at May 29, 2001 05:18:39 AM X-Mailer: ELM [version 2.5 PL3] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-Id: From: Alan Cox Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org > I actually saw a consistent 2.7% speedup on a FP-intensive benchmark > that my research group uses, comparing the A7A266 to the 8KTA3 after the > switch, using the same hardware for everything else and with the BIOS > settings matched as closely as possible. One thing you also want to do is to go through your benchmark adding prefetch instructions about 320 bytes ahead of current data. Even with 133MHz SDRAM on an Athlon that can add 40% to your throughput. Prefetch data, prefetchw data you will be writing too. On DDR it should be an even bigger win. > May 22 21:45:07 pollux kernel: PCI: No IRQ known for interrupt pin A of > device 00:04.0. Please try using pci=biosirq. Yes that is a bit odd. > May 22 21:45:07 pollux kernel: ALI15X3: simplex device: DMA disabled > May 22 21:45:07 pollux kernel: ide0: ALI15X3 Bus-Master DMA disabled > (BIOS) > May 22 21:45:07 pollux kernel: ALI15X3: simplex device: DMA disabled > May 22 21:45:07 pollux kernel: ide1: ALI15X3 Bus-Master DMA disabled The DMA was off because the BIOS left it off.