Uwe Dippel wrote:
> On Mon, 02 Jan 2006 14:06:52 +0100, M. Schatzl wrote:
>
>
>>Now that I switched to a 60G disk (cloned the other 2 partitions and the
>>Windows bootsector, then installed OpenBSD anew from the same
>>floppy/mirror as before), OpenBSD won't boot any more, except when I run
>>the boot-floopy and boot explicitely with wd0a:/bsd
OK, I finally found the error. This is the the partition table:
Disk: wd0 geometry: 5168/240/63 [78140160 Sectors]
Offset: 0 Signature: 0xAA55
Starting Ending LBA Info:
#: id C H S - C H S [ start: size ]
------------------------------------------------------------------------
0: 07 0 1 1 - 948 239 63 [ 63: 14348817 ] HPFS/QNX/AUX
1: 12 7189 0 1 - 7751 239 63 [ 108697680: 8512560 ] Compaq Diag.
*2: A6 949 0 1 - 7000 239 63 [ 14348880: 91506240 ] OpenBSD
3: 00 0 0 0 - 0 0 0 [ 0: 0 ] unused
This is the layout:
NTFS -> OpenBSD -> unused -> Compaq
Partitions #0 and #1 as well as the MBR were already transferred before
the installation of OpenBSD (via the installer). At the disk-setup
stage, I let drop myself into fdisk. Now I created the A6 partition on
unused space and proceeded,
the first time leaving the boot flag on #0:
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
The NTFS and Compaq(FAT16) partitions showed up as i and j in the
disklabel-editor.
I had to calculate the new offsets for my partitions myself because it
always remained on the initial value (which was correct for wd0a).
Installing biosboot via "/usr/mdec/installboot -v /boot
/usr/mdec/biosboot wd0" was of exactly no use. When I set the boot flag
to #2 later on, it still didn't work. The BIOS always responded with "No
OS found"-messages, though I could start OpenBSD then with a bootfloppy.
Appearently there was no damage done to the MBR, because Windows booted
without problems.
the second time with setting the flag to #2:
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
There are no partitions visible except wd0a and wd0c.
Offsets cumulated themselves every "a X"
Installing the BIOS was successful and it also booted the kernel after a
reboot.
In the second case it behaved as expected.
Obviously, the fact of setting the active partition affects the
installation. But shouldn't the flag be just meaningful on boot, to tell
the BIOS where to hook in? Maybe my assumptions/expectations are wrong,
but I suspected the installer only to honor the partition ID.
I don't know if this is an installer bug; I'm certainly not really
hellbent to reproduce it, but I could as long as I got that old disk (1
more week).
So let me know.
All the best,
/Markus
