Two partition tables on one hard drive?

[Folkert Rienstra]

Come to think of it, couldn't the problem be that the partition table disk
geometry actually *matches* (part of) the (published) BIOS setting?

It does, although one have to interpret the register contents.

No, it doesn't: "DH = maximum head number"
It reports the max head number which already is the (number of heads)-1

The interrupt reply will vary depending on BIOS settings.

Or in plain english: the bios setting will only be in effect
when the drive has not been partitioned yet.

Or in other words:
The bios setting is of no effect when the drive is partitioned.
Which is what I always thought it was, an advisory setting
for Fdisk to use when partitioning a drive.

No,

So Yes.

but one have to know the clean interrupt 13h reporting to be able
to see if there is a partition table match.

That doesn't make sense when it doesn't report the table entry(s)
but reports the bios setting instead. It doesn't make sense when that
bios setting isn't in effect when the drive has been partitioned.
It's the one *or* it is the other.
The bios will use the MBR entry for accessing the partitioned drive
and will use the bios entry in case the drive hasn't been partitioned yet.

How difficult can that be?

Then copy the MBR to a file on floppy, zero the MBR, reboot, make the
BIOS setting again, run interrupt 13h, write down the numbers on a
piece of paper. Then write back the MBR. How difficult can it be?

Far too difficult for something that is of no importance, when there is
no conflict.

What you describe only makes sense for when the partion tables original
values have been overwritten and you want to set them back without the
use of external software and use the setup utility for that instead.

And it is only necessary when you intend to use it with utilities that only
use Int13, not Int13 ext.
It will also only make a difference when the drive was previously written
to using Int13 and the original CHS and you need to read them back again,
using that original CHS.

Does not apply.

Exactly.
So either a CHS LBA mismatch is reported or the CHS values
are completely ignored and the LBA values are used entirely.

Guess not.

Set it so interrupt 13h function 8h returns 255 heads when no
partitions are present.

Ah. Any prayers to say to make that happen?

There is reason enough for a "may".

As much as in 'hell "may" freeze over first', sure.

Yes.

I would like to explain my interpretation of these matters, but it
seems as our basic understanding of disk geometry is so different,
that it may not be possible.

Well, as long as you keep insisting that something basic as values greater
than 1023 can be placed in a 10-bits register I will agree that it is
virtually impossible for you to explain your basic understanding of disk
geometry to me.

Or anyone else.

And I don't even think it has anything to do with diskgeometry 'ansich'
either. Just understanding how the bios (setup) and BIOS works.

 

[Folkert Rienstra]

I believe it is your turn again.
Now that we have established that my understanding of drive geometry
isn't all that flawed as you make it out to be, I think you owe me.