NetBSD Problem Report #296

From gnats  Mon Jun 20 19:14:02 1994
Received: from bloom-beacon.mit.edu (BLOOM-BEACON.MIT.EDU [18.181.0.26]) by sun-lamp.cs.berkeley.edu (8.6.9/8.6.9) with SMTP id TAA00553 for <gnats-bugs@sun-lamp.cs.berkeley.edu>; Mon, 20 Jun 1994 19:14:00 -0700
Message-Id: <199406201618.MAA00535@orchard.medford.ma.us>
Date: Mon, 20 Jun 1994 12:18:35 -0400
From: sommerfeld@orchard.medford.ma.us
Reply-To: sommerfeld@orchard.medford.ma.us
To: gnats-bugs@sun-lamp.cs.berkeley.edu
Subject: problems converting filesystems with fsck -c 2
X-Send-Pr-Version: 3.2

>Number:         296
>Category:       port-i386
>Synopsis:       problems converting filesystems with fsck -c 2
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    gnats-admin (GNATS administrator)
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Mon Jun 20 19:20:03 +0000 1994
>Closed-Date:    Thu Jun 30 06:42:25 +0000 1994
>Last-Modified:  
>Originator:     Bill Sommerfeld
>Release:        0.9C
>Organization:
	none
>Environment:

System: NetBSD orchard.medford.ma.us 0.9C NetBSD 0.9C (ORCHARD) #1: Fri Jun 17 10:35:34 EDT 1994 root@orchard.medford.ma.us:/u1/nsys/sys/arch/i386/compile/ORCHARD i386


>Description:
	I was able to successfully convert a small filesystem (~16MB)
with fsck -c 2; when converting a second, larger file system (~370MB),
the system rebooted suddenly after a hard reset.  After this failure,
FSCK reports a superblock mis-match error which it cannot fix.

Details:

Bug I (hard reset on fsck -c 2 on raw device)

	File system successfully converted was "/var" on /dev/rwd0e;
file system which failed was "/usr" on /dev/rwd0f.  This disk is
shared between DOS and NetBSD; /dev/rwd0g is the DOS partition.

The exact command used was:

fsck -c 2 /dev/rwd0f

The hard reset came *immediately* (less than one second) after
answering the "CONVERT TO NEW INODE FORMAT" question.  This hard reset is reproduceable (I've done it twice)..

Bug II (fsck doesn't fix the mess it made..)

After the reboot, fsck reports:

VALUES IN SUPER BLOCK DISAGREE WITH THOSE IN FIRST ALTERNATE

If fsck is being run manually, it falls back to using alternate
superblock 32.  At the end, it asks

UPDATE STANDARD SUPERBLOCK 

and I answer yes.  If I repeat the fsck, it continues to report
"VALUES IN SUPER BLOCK DISAGREE".. etc. as if the superblock were not
correctly updated.

I have not yet had a change to debug this part of the problem.

The disk label on the disk in question:

# /dev/rwd0e:
type: ESDI
disk: ESDI/IDE
label: 
flags:
bytes/sector: 512
sectors/track: 63
tracks/cylinder: 16
sectors/cylinder: 1008
cylinders: 1048
rpm: 3600
interleave: 1
trackskew: 0
cylinderskew: 0
headswitch: 0		# milliseconds
track-to-track seek: 0	# milliseconds
drivedata: 0 

7 partitions:
#        size   offset    fstype   [fsize bsize   cpg]
  a:    32256   161280    4.2BSD     1024  8192    16 	# (Cyl.  160 - 191)
  b:    64512   193536    unused        0     0       	# (Cyl.  192 - 255)
  c:   870912   161280    unused        0     0       	# (Cyl.  160 - 1023)
  d:  1056384        0    unused        0     0       	# (Cyl.    0 - 1047)
  e:    32256   258048    4.2BSD     1024  8192    16 	# (Cyl.  256 - 287)
  f:   766080   290304    4.2BSD     1024  8192    16 	# (Cyl.  288 - 1047)
  g:   161217       63    unused        0     0       	# (Cyl.    0*- 159*)

If you need more information, just let me know.

>How-To-Repeat:
	???
>Fix:
	???
>Release-Note:
>Audit-Trail:
State-Changed-From-To: open->closed 
State-Changed-By: cgd 
State-Changed-When: Wed Jun 29 23:42:25 PDT 1994 
State-Changed-Why:  
fixed, thanks to WS.  verified that it solves the problem, on boat-anchor. 
>Unformatted:

Home	PR Database Search