Subject: Problems with newfs and second drive
To: None <netbsd-help@NetBSD.ORG>
From: Chad Mynhier <mynhier@cs.utk.edu>
List: netbsd-help
Date: 03/29/1996 22:57:18
I'm running NetBSD-1.1 on a 486. I have a 124M IDE
drive serving as my system drive, and I'm trying to use a second
drive for /usr/local and /home. I have two problems: (1) I can't
specify anything from the second drive in /etc/fstab, or my machine
will panic at boot during the automatic fsck, and (2) I can't dump
those filesystems, because dump reports a bad sblock magic number.
I think these two are symptoms of the same problem.
The second drive was previously mounted under NetBSD-1.0.
I did whatever the install document recommended to convert them.
I could use the second drive, but I had to mount it by hand. I
thought that redoing the drive under 1.1 might take care of it, so
I tried that, and it had no effect. I was curious what the sblock
magic number was, so I modified dump to print it out. Dump told me
that fs_magic was 0. I was a bit more curious, so I had dump spit
out some more of the superblock data, like fs_ntrak. These things
were all 0, too.
Perhaps it's that my /etc/disktab entry is wrong. Maybe it's
possible that I got the size or offset wrong and that one partition
is overwriting the superblock of the next. Here's the entry:
st3290|Seagate 249MB IDE:\
:dt=ST506:ty=winchester:se#512:nt#15:ns#34:nc#1001: \
:pc#510510:oc#0: \
:pd#510510:od#0: \
:pg#102000:og#0:tg=4.2BSD:bg#4096:fg#1024: \
:ph#408510:oh#102000:th=4.2BSD:bh#4096:fh#1024:
When redoing the drive under 1.1 didn't work, I decided to
do a low-level format of the drive. This didn't fix the problem.
Based on the assumption that I was overwriting the superblock,
I made the drive one partition as:
:ph#510510:oh#0:th=4.2BSD:bh#4096:fh#1024:
I still saw the errant behavior. I'm guessing that the
two problems I listed are symptoms of the same behavior because the
errors I got during booting were divide-by-zero errors, and it appeared
that most of the data in the superblock was zero. My questions are:
1) Has anyone else noticed this behavior?
2) Is there anything else I should check?
3) Is this worth opening a PR for? I checked the bugs database
and didn't find anything that seemed to be this problem.
Chad Mynhier <mynhier@cs.utk.edu>
Sorcerer's Apprentice, CS Department
University of Tennessee, Knoxville
WWW: http://www.cs.utk.edu/~mynhier/