From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Sat, 29 Sep 2001 13:04:28 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Sat, 29 Sep 2001 13:04:18 -0400 Received: from mailb.telia.com ([194.22.194.6]:26129 "EHLO mailb.telia.com") by vger.kernel.org with ESMTP id ; Sat, 29 Sep 2001 13:04:05 -0400 Message-ID: <3BB5FF17.B316A969@canit.se> Date: Sat, 29 Sep 2001 19:04:23 +0200 From: Kenneth Johansson X-Mailer: Mozilla 4.77 [en] (X11; U; Linux 2.4.10 i686) X-Accept-Language: en MIME-Version: 1.0 To: linux-kernel@vger.kernel.org Subject: Problem with IDE DMA Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org ide_dmaproc: chipset supported ide_dma_timeout func only: 14 I get alot of this both with 2.4.10 and 2.4.9-ac17. This results in dma getting turned off and a slow disk. The way to reproduce it is running "find . -type f |xargs wc >/dev/null" and wait a while. I have so far tested a few different configured kernels and never once completed the above command on a 30GB filesystem without getting the hd to drop out of dma. I have two disk one on the motherboard IDE and one on a promise ultra66. Both have the same problem. -- Intel Corporation 440FX - 82441FX PMC [Natoma] (rev 2) Promise Technology, Inc. 20262 (rev 1) processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 1 model name : Pentium Pro stepping : 6