From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262608AbVE1APj (ORCPT ); Fri, 27 May 2005 20:15:39 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262626AbVE1APj (ORCPT ); Fri, 27 May 2005 20:15:39 -0400 Received: from fire.osdl.org ([65.172.181.4]:434 "EHLO smtp.osdl.org") by vger.kernel.org with ESMTP id S262608AbVE1APa (ORCPT ); Fri, 27 May 2005 20:15:30 -0400 Date: Fri, 27 May 2005 17:16:13 -0700 From: Andrew Morton To: Alexander Fieroch Cc: linux-kernel@vger.kernel.org, Jens Axboe , Bartlomiej Zolnierkiewicz Subject: Re: [2.6.12rc4] PROBLEM: "drive appears confused" and "irq 18: nobody cared!" Message-Id: <20050527171613.5f949683.akpm@osdl.org> In-Reply-To: References: X-Mailer: Sylpheed version 1.0.0 (GTK+ 1.2.10; i386-vine-linux-gnu) Mime-Version: 1.0 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 Alexander Fieroch wrote: > > Problem 1: > while booting the message "hdb: cdrom_pc_intr: The drive appears > confused (ireason = 0x01)" repeats several times and the system hangs > for some seconds. After booting sometimes I get the same message and > linux hangs. > > Problem 2: > syslog message "irq 18: nobody cared!" followed by a call trace repeats > while booting and while running linux. > > On the bottom of this mail you'll find an extract of /var/log/syslog > showing this error. > > > My system is a P4 630 with em64t, ASUS P5GD2 Premium board (ICH6 > chipset) and nvidia 6600GT PCIe graphic card. > I'm running a self compiled kernel 2.6.12rc4 but I've had these bugs in > self compiled kernel 2.6.11.8 and default debian kernel > 2.6.11-9-em64t-p4-smp too. > The default debian kernel 2.6.8-11-em64t-p4-smp seems to be the only one > where problem 1 does not occur but problem 2 does. Does the thing work correctly under any versions of Linux? If so, which? It appears that you entered cdrom_pc_intr() with IDE_IREASON_REG = 0x01. * ATAPI Interrupt Reason Register. * * cod : Information transferred is command (1) or data (0) * io : The device requests us to read (1) or write (0) io = 0 cod = 1 I'm not sure what this means. It generated an interrupt, expecting the kernel to send a write command? That sounds odd. If the answer to my first queston is "no" then perhaps the hardware is busted. Try swapping out cables, check power supplies, try a different drive, etc. If none of that helps then perhaps there's something we can do in cdrom_pc_intr() to work around this? (Should cdrom_pc_intr() be using atapi_ireason_t?)