Re: reiser4 filesystem detection

[Andrew Clausen]

Hi Bart,

On Tue, Dec 28, 2004 at 04:57:05PM +0100, B.Hakvoort wrote:
> Why do the specific probefunctions (e.g. ntfs_probe in fs_ntfs/ntfs.c)
> return a PedGeometry? I've skimmed through the probe-sections in
> filesys.c and figured it has something to do with the heuristics, but i
> can't see why a simple boolean wouldn't suffice. 'true' if geom contains
> fs signature, 'false' if not.
> Or maybe an integer to assign 'scores' to a filesystem...

The PedGeometry returned is the detected file system start/end.
This is useful for:
 * trying to find "lost" partitions.  (i.e. the rescue command)
 * if there are two different file system signatures on the same place,
then the file system that matches the partition closer is more likely
to be the "real" file system.

Cheers,
Andrew