From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261631AbULFUBy (ORCPT ); Mon, 6 Dec 2004 15:01:54 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261621AbULFUBy (ORCPT ); Mon, 6 Dec 2004 15:01:54 -0500 Received: from holomorphy.com ([207.189.100.168]:58328 "EHLO holomorphy.com") by vger.kernel.org with ESMTP id S261631AbULFUBo (ORCPT ); Mon, 6 Dec 2004 15:01:44 -0500 Date: Mon, 6 Dec 2004 12:01:19 -0800 From: William Lee Irwin III To: Jan Panteltje Cc: linux-kernel@vger.kernel.org, davej@redhat.com Subject: Re: 2.6.9 slows down everything Message-ID: <20041206200119.GR2714@holomorphy.com> References: <41B463BA.5080008@yahoo.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <41B463BA.5080008@yahoo.com> Organization: The Domain of Holomorphy User-Agent: Mutt/1.5.6+20040722i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Dec 06, 2004 at 02:50:50PM +0100, Jan Panteltje wrote: > Slow slow slow, compile takes looooong time..../ > even with all modules removed. > Even the mouse becomes sluggy > top shops no excessive CPU usage. > I am back to 2.4.25, I've heard of several MTRR issues with 2.6.x, of which this is likely an instance. The pattern seemed to involve attempts to cover the final portion of RAM with numerous fragments decreasing in size with ascending physical address, and failure to cover the final 4-8MB RAM. Odder still, the affected systems seemed to only have 1GB RAM, and for some reason 2.4.x was not affected. I've walked about 3 or 4 users through manually correcting the MTRR settings over the past 2 weeks. -- wli