Re: [FYI] lilo can not parse compact

[Mark van Walraven <markv@wave.co.nz>]

Hi Ross,

On Sun, Jan 02, 2000 at 12:47:49PM -0800, Ross Boylan wrote:
> I can confirm that lilo ran successfully when I changed
> 	root=
> to
> 	root=/dev/hda6
> 
> I had thought that lilo was supposed to default correctly with a blank, but
> I guess not.

The man page for lilo.conf says:

       root=root-device
	    This specifies the device that should be mounted as
	    root.   If  the  special  name current is used, the
	    root device is set to the device on which the  root
	    file  system  is currently mounted. If the root has
	    been changed with  -r , the  respective  device  is
	    used.  If  the variable `root' is omitted, the root
	    device setting contained in  the  kernel  image  is
	    used.   (And  that is set at compile time using the
	    ROOT_DEV variable in the kernel Makefile,  and  can
	    later be changed with the rdev(8) program.)

So, it appears that omitting a 'root' specification altogether is ok,
but a blank value is not.

> So this looks to be more a bug with the setup of lilo.conf at initial
> system creation (I suppose that
> it uses liloconfig) than with lilo itself.

The initial lilo.conf is generated by dbootstrap (file bootconfig.c),
which deliberately writes a blank if the root partition is not on the
first IDE or SCSI disk, or is on a SCSI disk when hda also exists.
(In both cases, a problem box, warning the installer to boot the system
using the rescue disk and configure lilo manually, is displayed.)

Since your root partition is a logical partition, so you should have been
prompted for the name of an extended partition in which to install lilo.
Which one did you choose?

I think there may be a problem in bootconfig.c.

In any event, I don't think this is a lilo problem.

Regards,

Mark.