Unable to boot after upgrading from FC2 to FC3

[John Swartzentruber]

On 11/11/2004 10:20 AM John Swartzentruber wrote:

> I just ran an upgrade from FC2 to FC3. The upgrade went smoothly, 
> without any issues or problems, but when it reboots, I see this:
>
> mounting root filesystem
> mount: error 6 mounting ext3
> mount: error 2 mounting none
> Switching to new root
> switchroot: mount failed: 22
> umount /initrd/dev failed: 2
> Kernel panic - not syncing: Attempting to kill init!
>
>
>
> The only drive in this computer is SATA. Under FC2, this was 
> recognized as /dev/hda. Now, under FC3, it is /dev/sda. I edited my 
> /etc/fstab and edit my grub line to reflect this, but the problem 
> persists.
>
> I can boot from the emergency disc, and mount my drive to edit and 
> look at it. I verified that both partitions (/dev/sda1 and /dev/sda2 
> exist and are properly labeled -- /boot and / respectively).
>
> I googled and looked at many different posts and messages, none of 
> which seemed to directly apply to this situation. The best that I can 
> figure is that the support for SATA isn't loaded when the system needs 
> to boot. That doesn't quite make sense, however, because this computer 
> was working fine with FC2, and it seems that a clean upgrade would 
> make sure that this would work.
>
> Can anyone give me some help, or at least some good pointers?
>
After a day of digging and experimenting and learning, I've come up with 
some pointers for myself, which I will post here so that others having 
the same problem might get some ideas.

The main problem appears to be that support for the SATA hardware wasn't 
being loaded in the initrd. When I created a new image using mkinitrd 
with "--with=sata_via", I got past the initial kernel panic and the 
system booted.

Or mostly. Unfortunately, it eventually had lots of IRQ 10 errors and 
then died. At some point (correct or not), I copied my old 
/etc/modprobe.conf file back to the current one, which didn't seem to 
have much of anything in it. I also added "acpi=off" to my  kernel line 
in /etc/grub.conf. The latter seems to have allowed eth0 to work again. 
While running in single user mode without the network (this was actually 
before I tried acpi=off), I made a CD-RW disc with the new udev rpm and 
updated that.

It looks like the system is mostly working now. I'm running yum update, 
which is taking awhile to get all of the updates. Then I'll start 
figuring out whether everything is really working and try to see what 
the implications of "acpi=off" really are. It clearly solves a big 
problem, but does it cause other problems?