TCLUG Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [TCLUG:2449] Boot from hard drive using NT bootloader?



I am no good at reading fdisk output, but I think that NTFS lies in the
first 8128 cylinders.. And fdisk is unable to actully handle the size
which is
why it prints 4096480+ and so is confused with the disk setup.

Now all of lilo documentation says that you must have lilo data in the
first
1024 Cylinders of the disk. In your case when you copy the lilo file
to NT somewhere on your C: drive it could get copied to any of the
8128 cylinders.

If the exact location on the disk where this gets copied happens to be
beyond 1024 cylinder then you might be in trouble.

One work around I see immediately is the possibilty of having creating a
small partition disk before the NT partion in the disk that  can be
accessed by NT
as drive D: or something and then installing lilo files here.

Another is to use "loadlin". Loadlin allows you to boot linux from
within dos.
It can be windows95 in the safe mode too.

Create DOS/FAT partition and install dos on it anywhere on the disk and
copy
loadlin and the kernel file onto there. Look at the instruction on how
to 
use loadlin. It is pretty straight forward.

Then in the autoexec.bat of that partion execute loadlin with the
appropriate parameters.

Mark this drive to be bootable and voila you should in theory be able to
use
the NT loader to boot this drive which when it boots will run the
autoexec
which will boot linux without the use of lilo.

Let me know if this works out.

--
regards,
sandipan

> clue:
> 
> fdisk /dev/hdb prints this:
> 
>    Device Boot   Begin    Start      End   Blocks   Id  System
> /dev/hdb1   ?        1        1     8128  4096480+   7  OS/2 HPFS
> Partition 1 does not end on cylinder boundary:
>      phys=(507, 255, 63) should be (507, 15, 63)
> /dev/hdb2         7677     8129    10544  1217664   83  Linux native
> Partition 2 does not end on cylinder boundary:
>      phys=(658, 255, 63) should be (658, 15, 63)
> 
> This is an 8.4GB drive.  Could it have something to do with LILO's
> 1024-cylinder limit? Is there a way to not use LILO?
> 
> Is LILO being fixed to use larger disks?
> 
> What else could it be?
> 
> Chris Schumann <whizkid@dwave.net>