From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265438AbTLHQCN (ORCPT ); Mon, 8 Dec 2003 11:02:13 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265482AbTLHQCF (ORCPT ); Mon, 8 Dec 2003 11:02:05 -0500 Received: from marc2.theaimsgroup.com ([63.238.77.172]:11713 "EHLO mailer.progressive-comp.com") by vger.kernel.org with ESMTP id S265438AbTLHQAa (ORCPT ); Mon, 8 Dec 2003 11:00:30 -0500 Date: Mon, 8 Dec 2003 10:58:26 -0500 Message-Id: <200312081558.hB8FwQ7n027927@marc2.theaimsgroup.com> From: Hank Leininger Reply-To: Hank Leininger To: linux-kernel@vger.kernel.org Subject: Re: 2.4: mylex and > 2GB RAM X-Shameless-Plug: Check out http://marc.theaimsgroup.com/ X-Warning: This mail posted via a web gateway at marc.theaimsgroup.com X-Warning: Report any violation of list policy to abuse@progressive-comp.com X-Posted-By: Hank Leininger Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On 2003-12-08, Per Buer wrote: > I have an Supermicro Superserver (wow!) 8040 or 8060 with a two Intel > Xeon (p3-based with 1MB cache) and a Mylex AcceleRAID 352. We recently > upgraded from 2 to 4GB of memory. > There seems to a problem with IO and high memory. Suddenly IO > performance will degrade dramatically (throughput of about 50KB/s). > Booting the machine with "mem=2048" remedies this. > We have tried replacing the memory with another make - no luck. You don't mention the kernel version--I'm guessing a 2.4.x? This doesn't necessarily help you, but: I haven't seen that on a dual P3 with 3GB of RAM (configured with CONFIG_HIGHMEM4G=y, CONFIG_HIGHMEM=y and CONFIG_HIGHIO=y) and an AcceleRAID 352, in 2.4.22. (I did get oopses with this card and 3GB in 2.2 kernel days, using the old bigmem patches for 2.2. I didn't stick around to find out whose fault it was though, just dropped back to 2GB) So, try verifying you've got CONFIG_HIGHIO=y, and out of curiosity you could try mem=3072, but if that doesn't help I'd suspect the motherboard. This box is... um... an Intel serverworks board whose exact model number escapes me at the moment. -- Hank Leininger