ronis-REM0VE@ronispc-d0t-chem-d0t-mc...QUEBEC-IS.invalid

wrote:
> I'm setting up a Dell that came with windows XP installed on its hard
> drive. I've added another drive, configured to be the master, and
> changed the original drive to be the slave. I installed 10.1 on the
> new drive and had lilo self-install on the MBR of that drive. The
> install script recognized the windows partition on the original drive
> and added an entry for it (on /dev/hdb2 with table=/dev/hdb). While
> booting into Linux works exactly as expected (the original drive is
> also accessible), if I try to boot into the windows partition the
> system hangs. One exception to this behavior is when I set the boot
> flag on windows partition; now the system boots directly into windows,
> and I never see the LILO menu.
>
> Any suggestions?
>
> David
>
>
Hi!
NOT a lilo problem!
The MS-Glassware retains its install configuration and remembers that it
is "egoistic" only resident of your box residing on primary master drive
and on primary partition.
So in the usual Linux way lets teach it that WE tell the software who is
the boss.
Now to work:
You swapped the drives, so something has to tell XP what happened.
Lilo can do it, so can the BIOS. You can instruct a modern BIOS to boot
from various drives but it is "last resort".
The usual (at least for me) way is to utilise the flexibility that is in
lilo and cheat the software to believe that it is on primary master
drive. It is called mapping the drives. Assignments are:
Primary master = IDE0(hda) = 0x80
Primary slave = IDE1(hdb) = 0x81
Secondary master=IDE2(hdc) = 0x82
Secondary slave =IDE3(hdd) = 0x83

So in the stanza that calls your XP install you add
map-drive=0x81 to=0x80
map-drive=0x80 to=0x81

Assuming that the master slave change was on primary IDE.

And rerun lilo so this change is recorded in booting choices.

Have fun

Stanislaw
Slack user from Ulladulla.