Subject: Re: Help! Kernel Panic on Boot - Please!
To: David Brownlee <abs@netbsd.org>
From: Brian Heath <brians_lists@bkheath.com>
List: port-mac68k
Date: 01/30/2003 11:30:59
David,
Thanks alot for the suggestion. I'm looking around on the NetBSD website,
and I don't see where I can get the sysinst RAMDisk install from. I think
this was the exact type of thing that I was looking for. Are there
instructions someplace that you can point me to?
Thanks...
Brian Heath
> From: David Brownlee <abs@netbsd.org>
> Date: Thu, 30 Jan 2003 10:48:59 +0000 (GMT)
> To: Brian Heath <brians-lists@bkheath.com>
> Cc: MacBSD <port-mac68k@netbsd.org>
> Subject: Re: Help! Kernel Panic on Boot - Please!
>
> On Thu, 30 Jan 2003, Brian Heath wrote:
> > 36 Hours ago I received a kernel panic in the middle of the night. This
> > occurred on a Quadra 700 running NetBSD 1.5.2 that has been running
> > relatively reliably for the last year and a half. Now I can no longer
boot
> > the machine. I am trying to figure out what my options are at this
point.
> > The message I get when it drops into the debugger is:
> >
> > root file system type: ffs
> > panic: ffs_read: type 0
> > stopped in sh at _cpu_Debugger+0x6: unlk a6
> > db>
> >
> > When I do a trace, I get messages involving ffs_read, vn_read,
dofileread,
> > sys_read, syscall, and trap0. I don't have anyway currently to copy and
past
> > the trace messages which is why I'm not reporting all the trace info. If
> > there's more there that will help decipher what is going on, let me know
and
> > I'll gladly type it all into a message and send it out.
> It sounds like your root partition has become corrupted.
> You could try booting a 1.6 install kernel with sysinst
> ramdisk, exiting to the shell and running fsck.
> >
> > The machine boots into the MacOS just fine. I then run the booter which
> > seems to go along fine until it hits the above section. I can use mini
shell
> > in the installer to look at the drive and everything there seems fine. I
> > even tried reloading the kernel.tgz and base.tgz to see if that would
> > replace some corrupt files. But that didn't make any difference.
> That is genrally a very risky move if the filesystem is corrupt -
> it can end up significantly increasing the corruption.
> --
> David/absolute -- www.netbsd.org: No hype required --
>
>