Re: Big Endian fix patch

[Lennart Sorensen]

On Wed, Jul 28, 2010 at 02:01:46PM -0400, Doug Nazar wrote:
> I was just looking at that. In disk/raid.c:129 we set the flag that  
> tells the system to try looking for partitions. I've been debating  
> whether to add code to disk/raid.c to detect if there actually is a  
> partition table (by probing for all known partition types) or fix the  
> tab completion code to offer a ')' also instead of assuming you want a  
> partition.
>
> I'm not sure of the GRUB design philosophy, but the first option might  
> be considered a layering violation. Maybe grub_device_open() should  
> check and then clear the flag if it doesn't have a table.

Certainly is seems USB keys and similar may hit the same problem.
Sometimes they have partitions and sometimes they don't.  So not really
the raid code's fault.

-- 
Len Sorensen