NetBSD Problem Report #39607

From www@NetBSD.org  Mon Sep 22 03:38:43 2008
Return-Path: <www@NetBSD.org>
Received: from mail.netbsd.org (mail.netbsd.org [204.152.190.11])
	by narn.NetBSD.org (Postfix) with ESMTP id 74CB763BC83
	for <gnats-bugs@gnats.netbsd.org>; Mon, 22 Sep 2008 03:38:43 +0000 (UTC)
Message-Id: <20080922033843.155DA63BA98@narn.NetBSD.org>
Date: Mon, 22 Sep 2008 03:38:43 +0000 (UTC)
From: Ephaeton@gmx.net
Reply-To: Ephaeton@gmx.net
To: gnats-bugs@NetBSD.org
Subject: newfs(_msdos) falsly? reports about "read-only filesystem" w/ VND
X-Send-Pr-Version: www-1.0

>Number:         39607
>Category:       bin
>Synopsis:       newfs(_msdos) falsly? reports about "read-only filesystem" w/ VND
>Confidential:   no
>Severity:       non-critical
>Priority:       low
>Responsible:    bin-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Mon Sep 22 03:40:00 +0000 2008
>Last-Modified:  Mon Sep 22 04:25:01 +0000 2008
>Originator:     Martin S. Weber
>Release:        4.99.72 / i386
>Organization:
me again inc.
>Environment:
NetBSD agamemnon.entropie.local 4.99.72 NetBSD 4.99.72 (AGAMEMNON) #4: Sat Sep 20 20:45:58 EDT 2008  root@agamemnon.entropie.local:/home/netbsd/obj/sys/arch/i386/compile/AGAMEMNON i386
>Description:
Maybe this is too obvious for you people... newfs_msdos complains on a vnconfig'd file that it's a read-only filesystem. It would then fail on the block device but work on the character device (you maybe expect that). It then in both cases complains about the r/o fs (huh?) and exits with status 1, although it worked in the latter case (as you are then able to mount). [[yes yes I'll use newfs -ZFs from now on]] See also below.
>How-To-Repeat:
cd /tmp
mkdir -p mnt
dd if=/dev/zero of=msdos.bin bs=1m count=10
vnconfig vnd0 /tmp/msdos.bin
newfs_msdos /dev/vnd0a
#newfs_msdos: warning: /dev/vnd0a is not a character device
# /dev/vnd0a: 20424 sectors in 2553 FAT12 clusters (4096 bytes/cluster)
# MBR type: 1
# bps=512 spc=8 res=1 nft=2 rde=512 sec=20480 mid=0xf0 spf=8 spt=32 hds=64 hid=0
# newfs_msdos: /dev/vnd0a: Read-only file system
echo $?
# 1
mount -t msdos /dev/vnd0a mnt
# mount_msdos: /dev/vnd0a on /tmp/mnt: Invalid argument
newfs_msdos /dev/rvnd0a
# /dev/rvnd0a: 20424 sectors in 2553 FAT12 clusters (4096 bytes/cluster)
# MBR type: 1
# bps=512 spc=8 res=1 nft=2 rde=512 sec=20480 mid=0xf0 spf=8 spt=32 hds=64 hid=0
# newfs_msdos: /dev/rvnd0a: Read-only file system
echo $?
# 1
mount -t msdos /dev/vnd0a mnt
# works.
>Fix:
dunno.

>Audit-Trail:
From: Quentin Garnier <cube@cubidou.net>
To: gnats-bugs@NetBSD.org
Cc: Ephaeton@gmx.net
Subject: Re: bin/39607: newfs(_msdos) falsly? reports about "read-only
	filesystem" w/ VND
Date: Mon, 22 Sep 2008 06:07:34 +0200

 --QHhm1I6mwQR20oIa
 Content-Type: text/plain; charset=us-ascii
 Content-Disposition: inline
 Content-Transfer-Encoding: quoted-printable

 On Mon, Sep 22, 2008 at 03:40:01AM +0000, Ephaeton@gmx.net wrote:
 > >Number:         39607
 > >Category:       bin
 > >Synopsis:       newfs(_msdos) falsly? reports about "read-only filesyste=
 m" w/ VND
 > >Confidential:   no
 > >Severity:       non-critical
 > >Priority:       low
 > >Responsible:    bin-bug-people
 > >State:          open
 > >Class:          sw-bug
 > >Submitter-Id:   net
 > >Arrival-Date:   Mon Sep 22 03:40:00 +0000 2008
 > >Originator:     Martin S. Weber
 > >Release:        4.99.72 / i386
 > >Organization:
 > me again inc.
 > >Environment:
 > NetBSD agamemnon.entropie.local 4.99.72 NetBSD 4.99.72 (AGAMEMNON) #4: Sa=
 t Sep 20 20:45:58 EDT 2008  root@agamemnon.entropie.local:/home/netbsd/obj/=
 sys/arch/i386/compile/AGAMEMNON i386
 > >Description:
 > Maybe this is too obvious for you people... newfs_msdos complains on a vn=
 config'd file that it's a read-only filesystem. It would then fail on the b=
 lock device but work on the character device (you maybe expect that). It th=
 en in both cases complains about the r/o fs (huh?) and exits with status 1,=
  although it worked in the latter case (as you are then able to mount). [[y=
 es yes I'll use newfs -ZFs from now on]] See also below.

 1.  Yes you should use the character device.
 2.  The EROFS is not a false error.  If you look at the disklabel of
     vnd0 after you configure it, you'll see that having no MBR, it
     defaults with a 4.2BSD partition in it.  The EROFS comes from
     newfs_msdos trying to write over the position where the disklabel
     is normally stored.

 --=20
 Quentin Garnier - cube@cubidou.net - cube@NetBSD.org
 "See the look on my face from staying too long in one place
 [...] every time the morning breaks I know I'm closer to falling"
 KT Tunstall, Saving My Face, Drastic Fantastic, 2007.

 --QHhm1I6mwQR20oIa
 Content-Type: application/pgp-signature
 Content-Disposition: inline

 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1.4.9 (NetBSD)

 iQEcBAEBAgAGBQJI1xoGAAoJENgoQloHrPnodOUIALPDM42xieWd/AXKbeEuPk7a
 1LLMxlUJ3ifNWahSndh88v+Bm5/3g3I9dIDCk3/rUb1pRjwMiCxKm0Q6o0J33FHZ
 GYzRUqG12pCYz2cLrKrSf2+EUdY0Q4ryZwYiB1188HsZ4E3L2kenvaa14nxqSjxS
 HxXBHAgHZXbOZGIChgJffaOJGu/HOjZyiiERBZoqB+JN3ZXy5MEs7O/tW722iOym
 uvqRLBZkqWjK3ufwHKuNPF2hK5U0qaUo8Xeg1qfyjPYDdcdBC5Jp1qZBRyebeXGg
 hzfbZ2N3FpqW8wxVCmK0201K3sXDVV3V+i42nyoEeXB8aQOARajiiC0YzoLw19E=
 =ihdW
 -----END PGP SIGNATURE-----

 --QHhm1I6mwQR20oIa--

From: "Martin S. Weber" <Ephaeton@gmx.net>
To: Quentin Garnier <cube@cubidou.net>
Cc: gnats-bugs@NetBSD.org
Subject: Re: bin/39607: newfs(_msdos) falsly? reports about "read-only filesystem" w/ VND
Date: Mon, 22 Sep 2008 00:20:31 -0400

 On Mon, Sep 22, 2008 at 06:07:34AM +0200, Quentin Garnier wrote:
 > (.. myself cut ..)
 > 1.  Yes you should use the character device.
 > 2.  The EROFS is not a false error. (...)

 Ok, so "wrong alarm" so to speak. To make something useful out of this PR,
 may I suggest that it gets documented to use the character device? I see
 newfs_msods' manpage's example is using the character devices but I don't 
 see it in NOTES of newfs (where of course yet another way is noted how I
 should build my filesystem images) or newfs_msdos.

 Sorry for the noise.

 And while you maybe work on the newfs manpage, you could also fix this
 parse error for me (in NOTES section):

      The block size and fragment size will be written back to the
      disklabel only if the last character of special references the same par-
      tition as the minor device number.  that provide disk like block and
      character devices.

 Aside of that, feel free to close this.

 Thanks tho,

 -Martin

Home	PR Database Search