On Mon, 28 Jun 1999 03:06:53 -0400, [EMAIL PROTECTED] wrote:
>This is not a lilo message, I think. You are trying to make a raw
>kernel boot floppy? This looks very much like a defective floppy.
>DX:0000 means to me that it is trying to read (AX:0212) 18 sectors from
>the floppy and getting _no_ joy. Try running badblocks on floppies
>until you find one with _solid_ no errors. A raw kernel boot floppy has
>no tolerance for badblocks. If you can't find one, we can make a
>filesystem on it and make a lilo boot floppy, but that is beyond what I
>can do without sleeping first. rdev /dev/fd0 /dev/hdc1 will put the
>right number in the right word of the floppy, for all it isn't in the
>doco yet. I don't even have a /dev/hdc, and it works for me. You are
>tangling with a defective floppy.
Yes, this is a raw kernel boot floppy.
Does badblocks care which of these switches I use?
badblocks -b 512 /dev/fd0 2880 or,
badblocks /dev/fd0 1440
I used the second version, though the first seemed to work.
I have zipslack 2.0.35 on hda using umsdos, and slackware 4, 2.2.6 on
hdc using the ext2 filesystem. After several attempts and no success
with 2.2.6 using badblocks, fdformat, rdev, and cp /vmlinuz I tried
this (all with the same floppy):
1. Booted zipslack, formatted the floppy, ran badblocks, copied the
2.0.35 version of vmlinuz on hda to the floppy and rebooted with the
floppy in the drive. It worked.
2. Same as 1, but I booted 2.2.6 on hdc and copied the 2.2.6 version of
vmlinuz to the floppy. The reboot failed with the same error from my
previous post.
3. Repeated 1 with zipslack. It worked.
4. Repeated 2. It didn't work.
5. On 2.2.6, I ran badblocks with the -write option, got no error
messages, repeated 2, and it still failed.
I'm happy using loadlin, but this puzzles me. I haven't read enough
about lilo to ask questions here, so feel free to get some sleep if you
need it. :)
Thanks,
Kurt Kehler
