Re: Lilo problem w/2.1.51

Werner Almesberger (almesber@lrc.di.epfl.ch)
Wed, 27 Aug 1997 01:15:55 +0200 (MET DST)


Charles W. Doolittle, N1SPX wrote:
> Device 0x0300: Invalid partition table, 1st entry
> 3D address: 1/1/0 (53)
> Linear address: 40/0/0 (39)
>
> Is what I get. It does not change, as far as I can tell.
> It will not change.
[...]
> It looks like hda is tried to be LBAd.

Unless you change the partition table or the kernel's idea of the disk
geometry/translation, this problem won't go away. If you're absolutely
sure that it's just some bad math in the Linux kernel, you can tell
LILO to ignore the inconsistency with the option IGNORE-TABLE. You may
also want to use LINEAR, because LILO's normal mapping is very likely
to fail. Comment out all options that may write things to the disk
(LOCK, CHANGE, etc.) while experimenting, just in case.

NB: there's also an option called FIX-TABLE, which will change the
partition table in a way that makes LILO and probably also fdisk happy.
If the LBA calculation is wrong (what you suspect, and what looks
likely), FIX-TABLE will do exactly what you don't want: write back
incorrect information.

I'd suggest to check the LBA settings in the BIOS. If they're okay, it
may be something in the IDE driver, so looking for recent changes may be
a good idea (Mark ?). If you're in a hurry, try IGNORE-TABLE. fdisk will
still report all the inconsistencies, so you don't need LILO's complaints
for researching the problem further.

- Werner

-- 
  _________________________________________________________________________
 / Werner Almesberger, DI-LRC,EPFL,CH   werner.almesberger@lrc.di.epfl.ch /
/_IN_R_133__Tel_+41_21_693_6621__Fax_+41_21_693_6610_____________________/