Re: [version 1.6.21] Assertion (dis k != NULL) at ../../libparted/disk.c:1037 in function ped_disk_next_partition()

[Andrew Clausen]

On Sun, Mar 27, 2005 at 09:01:15AM +0200, Sven Luther wrote:
> Andrew, what do you think, for future versions of libparted, to modify the
> detection behavior.
> 
> Instead of running the probes, and stoping once you found, we would run all
> probes, put the results in a list, and then if there is a single element in
> the list, go for it, if not, we ask the user in an exception about this fact,
> with eventually a choice to clobber the badly detected info.

I independently came to this conclusion also.  (I've been speaking in private
with some SGI people who have been running into problems.)  We already
do this for filesystem probing.

> We maybe need to add a clobber-light in this case, since the full clobber may
> well break the real file-system or partition table on the disk.

I don't think this will be necessary.  AFAIK, a "clobber-light" should
always suffice where a "clobber-heavy" works.

> Also, it would be nice to have the same kind of double checking for
> filesystems which are one partitions marked as being of some other type.

I'm accepting patches.  I guess we need to add disk_ops->check...

Cheers,
Andrew