Re: workaround for BIOS / CHS stuff

[K . G .]

On Mon, 28 Jun 2004 11:07:13 +1000
Andrew Clausen <address@hidden> wrote:

[...]

> So, is the following a good solution?
> 
> (1) when resizing a partition, attempt to deduce the geometry from
> that partition alone.  (i.e. by inspecting its partition table entry,
> and inspecting its filesystem geometry).  Also, if the start or end
> are staying fixed, then just keep their CHS values the same.

I think this would be good.
Maybe if the partition is to small at the begining of the disk, or in
other cases where the geometry can't be guessed, fallback to method 2.
I still don't really know what to do if the geometry for that partition
is inconsistent (ie guess(startCHS) != guess(endCHS) ||
guess(FS) != guess(startCHS)). Maybe one source of information should
have the priority.
 
> (2) when creating partitions, make a global guess, similar to what
> we're doing now.

Yes, but regarding to informations reported by Szakacsits Szabolcs,
not using kernel information anymore about the drive geometry would
be a good idea.
What should be done in case of inconsistencies ? (different geom for
different partitions). Maybe using the geometry of the boot partition
if there is one would be interresting ? Does Linux use the CHS values ?
If not maybe consider only geometry of DOS partitions for the global
guess.
Or just refuse to start saying "you have a bad partition table"...
Anyway I hope inconsistent partition tables are not too current.

> When creating new partition tables, should we assume that the user is
> using LBA?

Well without any source of information it would be difficult to do
anything else :)
So either use LBA, or if the user has the right to force a geometry and
do it, use the user supplied geometry ?

Cheers,
Guillaume.