So I've got an initio card, 68-pin SCSI UW hooked up to a 10k rpm uhmm... some IBM drive, 9.1Gb. I'm getting the ubiquitous LILO error of LI and then straight 99's, which of course means can't load second stage because the drive geometry is misreported or gnomes have eaten the kernel or a million other things. Before I go shotgun debugging lilo with a million different options, does anyone know an option to pass lilo that typically makes scsi happy? Its doing the typical: lba32 bit addressing assumed message when I re-run LILO, is that bad?

The lilo.conf is the typical slackware kind:

# LILO configuration file
# generated by 'liloconfig'
#
# Start LILO global section
boot = /dev/sdb
message = /boot/boot_message.txt
prompt
timeout = 200
default = GoSpeedRacer
# Override dangerous defaults that rewrite the partition table:
change-rules
reset
# VESA framebuffer console @ 1024x768x256
# vga = 773
# Normal VGA console
vga = normal


Right now I'm cheat booting with a different scsi drive on a different drive. Both drives seem fine, I've badblocked and beaten the heck out of them in the past few days.

Cheers,

Finegan

this is what I had to do to make my IDE and SCSI drives like each other.
/dev/hda=Win Xp
/dev/sda=linux /
/dev/hdc=/usr/local and /home

I switched between linux and windoze by switching the boot order in BIOS. To trick LILO into letting me boot from a scsi drive first rather than an ide drive, I added the following to the first section of my lilo.conf:

#mapping for using a scsi drive
disk = /dev/sda
bios = 0x80
disk = /dev/hda
bios = 0x81
disk = /dev/hdc
bios = 0x82