From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262236AbVAJNUX (ORCPT ); Mon, 10 Jan 2005 08:20:23 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262234AbVAJNUX (ORCPT ); Mon, 10 Jan 2005 08:20:23 -0500 Received: from smtp.raid.ru ([212.33.224.148]:26378 "EHLO smtp.raid.ru") by vger.kernel.org with ESMTP id S262232AbVAJNUS (ORCPT ); Mon, 10 Jan 2005 08:20:18 -0500 Message-ID: <41E28115.4010002@permonline.ru> Date: Mon, 10 Jan 2005 18:20:21 +0500 From: "Artem S. Tashkinov" User-Agent: Mozilla Thunderbird 1.0 (Windows/20041206) X-Accept-Language: en-us, en MIME-Version: 1.0 To: B.Zolnierkiewicz@elka.pw.edu.pl Subject: 100% CPU load on nforce2/3 motherboards with kernels >= 2.6.8 Content-Type: text/plain; charset=KOI8-R; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Good day, Bartlomiej Zolnierkiewicz! I've got a severe problem with linux >= 2.6.9 kernels on my PC: any read/write HDD IO operations consume 100% of CPU. Kernel is compiled with nforce IDE support. Kernel is vanilla from kernel.org. No proprietary drivers are installed. My chipset is NForce3 250, CPU is Sempron 3100. My MB is GigaByte K8NS Pro. My friend (who has an nforce2 based motherboard Asus A7N8X-X) has the same problem with >= 2.6.9 kernels. 2.6.7 at the same time works fine for him. His and mine kernels say that the HDD is in UDMA5 mode. I tried switching off ACPI/IO APIC and even Plug'n'Play code with no success. Kernel 2.4.28 works perfectly With respect, Artem