From owner-freebsd-hackers Sun Jan 29 22:57:52 1995 Return-Path: hackers-owner Received: (from root@localhost) by freefall.cdrom.com (8.6.9/8.6.6) id WAA21209 for hackers-outgoing; Sun, 29 Jan 1995 22:57:52 -0800 Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.34]) by freefall.cdrom.com (8.6.9/8.6.6) with ESMTP id WAA21191 for ; Sun, 29 Jan 1995 22:57:33 -0800 Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.9/8.6.9) id RAA29733; Mon, 30 Jan 1995 17:53:36 +1100 Date: Mon, 30 Jan 1995 17:53:36 +1100 From: Bruce Evans Message-Id: <199501300653.RAA29733@godzilla.zeta.org.au> To: freebsd-hackers@FreeBSD.org, j@uriah.sax.de Subject: Re: SyQuest works with FreeBSD 2.0R ! Sender: hackers-owner@FreeBSD.org Precedence: bulk >As Peter Dufault wrote: >| >| I'd like a description of what happens when you unmount a file system, >| change the media, and try to mount a filesystem on a removable drive, >| including dmesg output. >Btw., there's another annoying thing besides the UA: a message >indicating ``disk does not have an MBR'' is logged, even if i'm using >/dev/rsd*d -- where it does not make any sense to insist on having an >MBR at all. This is from isa/readMBR.c. The SCSI driver is apparently doing a good job of reinitializing the label after a disk change. readMBRtolabel() is probably being too verbose about the missing MBR. Don't do any more work on readMBR.c. It will go away when slices are used. The corresponding file:routine is isa/diskslice_machdep.c: dsinit(). This is quiet about missing MBRs but has verbose debugging code about what it thinks are invalid DOSpartitions. dsinit() is called by dsopen() in kern/subr_diskslice.c. This has verbose non-debugging code about invalid and adjusted labels. The slice code has some support for removable media (dsgone()) but it hasn't been tested much. It is only implemented for virtual (vn) media and probably only works right if all minors for the device are closed and in one or two other cases (for vn you have to have a minor open to do the ioctl that removes the virtual media :-). It could be tested using floppies if the floppy driver had support for removed disks. Bruce