Re: Assertion (metadata_length > 0) at disk_dos.c:1790

[Jonathan Purcell]

Andrew Clausen wrote:
> Hi Jonathan,
>
> On Tue, Feb 01, 2005 at 02:39:41PM +0000, Jonathan Purcell wrote:
>
> > I have seen this assertion where I specify two extended partitions 
> > exactly next to each other;
>
>
> I assume you mean logical partitions; a partition table can have only
> one extended partition.

Yes apologies for incorrect terminology.

> In most cases, you can't have logical partitions right next to each
> other.  Each logical partition must have its own partition table, just
> before the partition starts.
>
>
> > particularly where I use a constraint 
> > generated with ped_constraint_exact(). Should the function 
> > add_logical_part_metadata just return success where the metadata length 
> > is zero?
>
>
> No... in the cases libparted is complaining, you actually need a
> non-zero gap.

Ah ok so this metadata partition is space for the partition table before 
each logical partition. In which case the assertion I saw was as 
expected. Thanks for the info.

Regards

Jonathan