Hi Everyone,

I just recently updated my version 13.37 installation using slackpkg upgrade-all.
It notified me to run lilo, and I blindly did and rebooted just for fun, but now the system no longer boots up.

My installation is entirely on LVM, including boot. I got up and running on the install disk, mounted up the partitions and tried to run lilo (after a chroot) and this is the error I'm seeing:

Linux experimental device 0x04x needs to be defined

I'm not sure what is going on here or what i need to do to fix it.
Any help would be greatly appreciated.

thanks
steven

A quick Google on the error you get gives this as 1st hit:
http://comments.gmane.org/gmane.comp...vm.devel/13832
Maybe it solves it for you as well?

Thanks for the reply. I found that page as well, but it's hard to tell if it's the same issue.
I tried adding disk=0x80 and max-partitions=15 in lilo.conf but it didn't help.

Hi.

Run the following before chroot: Make sure you rebuild your initrd next time!

or eliminate the need.

Unfortunately, that isn't working either. I rebuilt the initrd pretty much using the instruction in the README,
but I added the -L, and lilo still gives the same error. Even removing initrd entry from lilo.conf, still gives the same error.

Not sure where to go next...

Well, I got lilo to run successfully, but the system still doesn't boot.
After booting into the install disk (usb stick), I started setup and mounted the LVM partitions for /, /home and 2 others I have in /opt.
I then did but did not chroot. I ran lilo in /mnt/etc like this: and it ran fine, no error (Added Linux). Unfortunately, the systems still doesn't boot. It will startup up, I have the splash screen and the OS chooser (there's only linux), and when it launches into linux, I get the first line of dots print out and it halts. I usually would get a few lines and then things would start loading.

So, this is in a VM of some type? Is this still a 13.37 system or did you upgrade to a newer version?

I got it figured out and I'm back in business

My system has 3 disks, a 40G, 1TB and 2TB. The 40G is my system disk, the other 2 are data disks plugged on a different controller.
The 40G is plugged into sata0 on the motherboard, so thats the disk that gets booted, but it was showing up as /dev/sdc, not sda (which is what I would have expected). My lilo.conf had /dev/sda as the boot entry, so it wasn't getting written to the proper MBR !

There's two things I'm not sure of. Is it normal, we can't count on the ordering of the disk names to follow (what I think is) the physical ordering of the controllers ? And, because it's been a long time since I originally setup this system, did my lilo.conf change during the upgrade, meaning I originally had the "out of order" disk specified on the boot entry or did the ordering of the disks change in linux ?

Hope I'm explaining myself well enough, in case someone else ever hits this problem.

And one final thing, I also did this prior to the chroot, did the chroot /mnt and then ran lilo. I'm not sure it was actually needed to run properly, but thats what I had.

Cheers everyone !
Steven

Hi, sorry I didn't see this question - it isn't a VM and yes, it's still 13.37 (I just wanted to patch the system).

Steven

Not a problem, you solved it yourself.

My machine has an Adaptec AIC-7892A U160/m SCSI controller. The huge kernel would initialize that controller first, then the on-board SATA controller. The generic kernel wouldn't have the adaptec kernel module integrated, so it would initialize the on-board SATA controller first then the initrd would load the adaptec module.

End result was the drives being renamed depending depending which kernel I used to boot. Perhaps this (or something similar) is what happened to you.