Boot Suse 10.1 from Lilo help!
I already had slackware installed so I install Suse 10.1. Somehow I messed up grub and didn't install it on the MRB. I booted slackware from a floppy and ran Lilo. I have tryed to add suse to the Lilo but it stops after it says "can't mount root, Kernel panic".
Here is my Lilo.conf: Quote:
Quote:
|
I have an idea what the problem might be, but I am not totally sure I have it all figured out. (The worst that could happen is one more failed boot, right?)
I think the problem has to do with the difference between the way LILO and GRUB work. GRUB, as I understand it, basically interprets a configuration file at boot time. LILO, in effect, "compiles" the info from lilo.conf and stores absolute disk locations of the relevant items. So the interpretation of most or all of the paths in lilo.conf is in terms of the system running when you run lilo. I.e. in your case, when slackware is running. So I think when you try to boot Suse, even though you tell it to use /dev/hda3 as /root, you are still telling it to use the kernel and initrd from your slackware install. So after booting slackware from your floppy, I think you need to mount /dev/hda3 somewhere, for example, at /mnt/hda3. And then change ... Code:
image = /boot/vmlinuz Code:
image = /mnt/hda3/boot/vmlinuz I have experimented with something like this on one of my systems, but w/o an initrd. It seems to work like I think it should, except the boot command line reported in the log is a little crazy with regard to the path of the kernel image. |
Thanks for the Reply!
I tryed what you said and it quit while it was loading the image. Lilo is strange. I think I'm going to have to reinstall Suse and try to get grub on the MRB. |
I finally figured it out. My solution was very similar to yours.
I found out what vmlinuz and initrd linked to and copyed them to /boot/suse (I guess the symbolic link doesn't work). I haven't tried out safemode yet but it should work. Here is my new lilo.conf Quote:
|
I am glad you got it working. If I understood you correctly, when you tried my suggestion it started to load the kernel but didn't finish. This sounds like it may have been a problem with disk addressing. Originally, LILO had 1024 cylinder limit. Everything that is normally in the the /boot directory had to be in the first 1024 cylinders of the disk, which I think generally works out to about 8.4 MB. Newer versions of LILO could overcome this limitation, but you had to specify the lba32 global option to LILO. Since /dev/hda3 has higher disk addesses than /dev/hda2, this sounds like a plausible explanation.
Interpreting symlinks is a function of a running kernel, so I wouldn't expect them to work at boot time. Again, congratulations. |
Thanks
It may take me a while, but I am learning! |
All times are GMT -5. The time now is 09:20 AM. |