Re: very bad unmounts from 2.0.34/5?

Stephen C. Tweedie (sct@redhat.com)
Mon, 14 Sep 1998 12:35:46 +0100


Hi,

On Wed, 09 Sep 1998 17:21:20 +0200, Anders Henriksson <andhe@ida.liu.se>
said:

> Then I recompiled the 2.0.32 kernel on hdb (the second drive), installed it
> and made it work fine.

> Then I tried the 2.0.34 kernel.
> Initially it works fine. The first boot is ok and I can run X and basicly do
> anything I want.
> Then I shutdown the system with: shutdown -r now
> Shutdown looks normal to me (with 6 months linux experience)
> The reboot halts with the message: kernel panic! failed to mount root on 3.10

3.10? Why are you trying to mount the root from /dev/hda16?

The 2.0.34 kernel upgrade added code to the ide driver to correctly deal
with drive geometries on some large disks. I'm not sure why you should
be getting such bad corruption, but it is certainly conceivable that the
geometry mismatch between the two kernels is upsetting the partitioning.
When you do boot from the newer kernel, can you do an fdisk there and
verify the partition table on /dev/hdb to make sure it still looks
sensible? If not, then I think a solution is to partition the second
disk using geometry information which makes sense to the .34 kernel.
The 2.0.34 boot message will tell you what geometry it detects, and
expert mode in fdisk will let you set the CHS geometry manually while
you partition it.

--Stephen

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/faq.html