From owner-freebsd-geom@FreeBSD.ORG Tue Jan 18 02:33:43 2005 Return-Path: Delivered-To: freebsd-geom@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id AD2DD16A4CE for ; Tue, 18 Jan 2005 02:33:43 +0000 (GMT) Received: from hollin.esperi.net (hollin.esperi.net [194.153.168.44]) by mx1.FreeBSD.org (Postfix) with ESMTP id 2D64F43D2F for ; Tue, 18 Jan 2005 02:33:43 +0000 (GMT) (envelope-from nicolai@hollin.esperi.net) Received: from localhost ([127.0.0.1] helo=hollin.esperi.net) by hollin.esperi.net with esmtp (Exim 4.10) id 1CqjBd-0003IG-00 for freebsd-geom@freebsd.org; Tue, 18 Jan 2005 02:33:41 +0000 Received: (from nicolai@localhost) by hollin.esperi.net (8.12.3/8.12.3/Submit) id j0I2XfmU012661; Tue, 18 Jan 2005 02:33:41 GMT MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <16876.30085.273846.833564@hollin.esperi.net> Date: Tue, 18 Jan 2005 02:33:41 +0000 From: Nicolai E M Plum To: freebsd-geom@freebsd.org X-Mailer: VM 7.07 under Emacs 21.3.1 Subject: geom_mirror panic with pair of firewire disc X-BeenThere: freebsd-geom@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: GEOM-specific discussions and implementations List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 18 Jan 2005 02:33:43 -0000 Hi Using FreeBSD 5-STABLE as of about 2 days ago and these discs: # dmesg | grep da[56] da5 at sbp1 bus 0 target 1 lun 0 da5: Fixed Simplified Direct Access SCSI-4 device da5: 50.000MB/s transfers da5: 57259MB (117266688 512 byte sectors: 255H 63S/T 7299C) da6 at sbp1 bus 0 target 1 lun 1 da6: Fixed Simplified Direct Access SCSI-4 device da6: 50.000MB/s transfers da6: 57259MB (117266688 512 byte sectors: 255H 63S/T 7299C) with this partitioning: # fdisk /dev/da5 ******* Working on device /dev/da5 ******* parameters extracted from in-core disklabel are: cylinders=7299 heads=255 sectors/track=63 (16065 blks/cyl) Figures below won't work with BIOS for partitions not in cyl 1 parameters to be used for BIOS calculations are: cylinders=7299 heads=255 sectors/track=63 (16065 blks/cyl) Media sector size is 512 Warning: BIOS sector numbering starts with sector 1 Information from DOS bootblock is: The data for partition 1 is: sysid 165 (0xa5),(FreeBSD/NetBSD/386BSD) start 63, size 116728227 (56996 Meg), flag 80 (active) beg: cyl 0/ head 1/ sector 1; end: cyl 1023/ head 254/ sector 63 The data for partition 2 is: The data for partition 3 is: The data for partition 4 is: # fdisk /dev/da6 ******* Working on device /dev/da6 ******* parameters extracted from in-core disklabel are: cylinders=7299 heads=255 sectors/track=63 (16065 blks/cyl) Figures below won't work with BIOS for partitions not in cyl 1 parameters to be used for BIOS calculations are: cylinders=7299 heads=255 sectors/track=63 (16065 blks/cyl) Media sector size is 512 Warning: BIOS sector numbering starts with sector 1 Information from DOS bootblock is: The data for partition 1 is: sysid 165 (0xa5),(FreeBSD/NetBSD/386BSD) start 63, size 116728227 (56996 Meg), flag 80 (active) beg: cyl 0/ head 1/ sector 1; end: cyl 1023/ head 254/ sector 63 The data for partition 2 is: The data for partition 3 is: The data for partition 4 is: and this command to create a mirrored pair: # gmirror label m1 /dev/da5s1 /dev/da6s1 I immediately get this result: (syslog) ---- Jan 16 23:19:13 nurn kernel: GEOM_MIRROR: Device m1 created (id=2113314100). Jan 16 23:19:13 nurn kernel: GEOM_MIRROR: Device m1: provider da5s1 detected. Jan 16 23:19:13 nurn kernel: GEOM_MIRROR: Cannot add disk da5s1c to m1 (error=17). Jan 16 23:19:14 nurn kernel: GEOM_MIRROR: Device m1: provider da6s1 detected. Jan 16 23:19:14 nurn kernel: GEOM_MIRROR: Device m1: provider da6s1 activated. Jan 16 23:19:14 nurn kernel: GEOM_MIRROR: Device m1: provider da5s1 activated. Jan 16 23:19:14 nurn kernel: GEOM_MIRROR: Device m1: provider mirror/m1 launched. Jan 16 23:19:14 nurn kernel: WARNING: Expected rawoffset 0, found 63 ---- and on the console: Fatal double fault: eip = 0xc07a4acd esp = 0xd4226000 ebp = 0xd4226014 panic: double fault Uptime: 2h45m29s In fact, this happens now if I run "gmirror load", I get the same immediate panic. I don't think is supposed to happen. Has anyone seen it before? My intent is to mirror these two slices which I have created to be slightly smaller than the entire disc. This is so that if I have to replace a disc with another type that is slightly smaller I don't have a problem. (I know that I can just get a bigger disc than 60GB, but this is a test setup; in production I want to use the largest discs that are economically reasonable so I cannot rely on always replacing a failed disc with a larger one). Nicolai