I've tried to use lilo-21.5 in a root raid1 partition with a 2.2.16
kernel with the latest raid patch (A0).
I've installed lilo with the lilo.conf suggested by the linux-raid docs:
boot = /dev/md0 # my root raid1 partition composed by /dev/hda1 & /dev/hdc1
linear # tested also without it and also tested with lba32
image = /vmlinuz-2.2.16-RAID
root = /dev/md0
label = linux-raid
read-only
After the reboot lilo keep on printing an error followed by a 0x80 and by
a floppy seek. (Why does it seek the floppy?)
If I insert a floppy (don't ask me why I've done it!) the number
printed changes from 0x80 to 0x09.
After having read the mailing list archives of linux-raid at mail-archive.com
I've found that at least another one had a similar problem (0x80 error).
After various tries I've succeded in installing lilo in the following way:
changed lilo.conf boot option to /dev/hda, run lilo
changed lilo.conf boot option to /dev/hdc, run lilo.
With both hda & hdc disk connected this works, but if I boot the linux box
without the hdc lilo writes
L followed by ones and zeros.
Does this mean that /dev/hda is lookin for the kernel image in /dev/hdc?
Is this a LILO bug or is my motherboard BIOS buggy?
Hoping not to have wasted your time..
Ciao.
Diego Liziero ([EMAIL PROTECTED])
P.S.: I've also tried the Red Hat patched version lilo.21-15 and the lilo-21
with the raid1 patch with the same result.