From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Wed, 9 May 2001 17:21:27 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Wed, 9 May 2001 17:21:18 -0400 Received: from host-64-65-206-13.choiceone.net ([64.65.206.13]:15687 "EHLO isaiah.tpr-east.com") by vger.kernel.org with ESMTP id ; Wed, 9 May 2001 17:21:02 -0400 Message-ID: <3AF9B411.2A0F3F43@tpr.com> Date: Wed, 09 May 2001 17:18:09 -0400 From: Mark Bratcher Organization: Torrey Pines Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 To: linux-kernel@vger.kernel.org CC: bratcher@tpr.com Subject: ATAPI Tape Driver Failure in Kernel 2.4.4, More In-Reply-To: <3AF9558F.9C76953C@tpr.com> Content-Type: multipart/mixed; boundary="------------C51193CFE1F8FF2CDF30FE2F" Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org This is a multi-part message in MIME format. --------------C51193CFE1F8FF2CDF30FE2F Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, I just upgraded from kernel 2.2.17 to 2.4.4. I use a Seagate ATAPI tape drive, model STT20000A. I use dump to do backups (probably not relevant). I have more data regarding tape backup failures. Here is what I get: * In a full tape dump of about 50MB or so, everything goes smoothly. No errors. * In a full tape dump of a little over 1GB (to a 10GB tape), I get the following errors immediately after the backup completes and I try to write a file mark with "mt": May 9 16:50:49 isaiah kernel: ide-tape: Reached idetape_chrdev_open May 9 16:52:05 isaiah kernel: hdd: irq timeout: status=0xd0 { Busy } May 9 16:52:05 isaiah kernel: hdd: ATAPI reset complete May 9 16:52:05 isaiah kernel: ide-tape: Couldn't write a filemark May 9 16:52:06 isaiah kernel: ide-tape: Reached idetape_chrdev_open May 9 16:54:06 isaiah kernel: ide-tape: ht0: DSC timeout May 9 16:54:06 isaiah kernel: hdd: ATAPI reset complete May 9 16:54:06 isaiah kernel: ide-tape: ht0: I/O error, pc = 10, key = 2, asc = 4, ascq = 1 May 9 16:54:06 isaiah kernel: ide-tape: ht0: I/O error, pc = 34, key = 2, asc = 4, ascq = 1 * In a full tape dump of between 6GB and 7GB, the backup completes, I successfully write a tape file mark with "mt", rewind, then attempt to compare. I get: May 9 04:21:18 isaiah kernel: ide-tape: Reached idetape_chrdev_open May 9 04:22:35 isaiah kernel: ide-tape: bh == NULL in idetape_copy_stage_to_user May 9 04:22:35 isaiah kernel: ide-tape: bh == NULL in idetape_copy_stage_to_user May 9 04:22:36 isaiah kernel: ide-tape: Reached idetape_chrdev_open The errors are repeatable and do not occur in 2.2.17. At this rate, I'll have to go back to 2.2.17 to get successful backup scripts working. Anyone know what causes this problem? Please reply directly, as I'm not a member of this list. Thanks :-) Mark Bratcher --------------C51193CFE1F8FF2CDF30FE2F Content-Type: text/x-vcard; charset=us-ascii; name="bratcher.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Mark Bratcher Content-Disposition: attachment; filename="bratcher.vcf" begin:vcard n:Bratcher;Mark tel;fax:716/288-4604 tel;work:716/288-7220 x-mozilla-html:FALSE url:www.tpr.com org:Torrey Pines Research version:2.1 email;internet:bratcher@tpr.com title:Director of Software Development adr;quoted-printable:;;565 Blossom Road=0D=0ASuite A;Rochester;New York;14610;USA x-mozilla-cpt:;19472 fn:Bratcher, Mark end:vcard --------------C51193CFE1F8FF2CDF30FE2F--