On 12/17/2010 06:19 AM, ehem+grub@m5p.com wrote:
> I've got a system with an oddball, but valid setup that is outside the
> bounds of what GRUB supports. As such I'm starting to work on
> familiarizing myself with GRUB2's internals on the way to solving the
> problem.
>
> I've found two issues worth fixing so far and I'm attaching patches.
> First, what is refered to as "a magic number used by Windows NT" is
> apparently used as a volume identifier to identify disks. 
It's windows NT Volume ID. I don't see how it is a misnomer.
> Current
> versions of LILO in fact use it to locate the disk holding its second
> stage. Second, perhaps minor, but I'm surprised someone chose to break
> the checksum verification code in grub-core/partmap/sun.c and didn't
> didn't keep the magic number check with it.
This seems to be just moving the code around. There are many different
coding styles. Changing from one of them to another is waste of effort
unless it's done to uniformise with rest of codebase. I don't have
strong preference to one or another style so I prefer not to touch anything.
>  I hereby release the two
> attached patches under GPL version 3 and subsequent varients.
>
>   

> I'm not very familiar with Bazaar, shouldn't be a problem. Does seem git
> is taking over from Subversion.
>
>
>   
>
>
> _______________________________________________
> Grub-devel mailing list
> Grub-devel@gnu.org
> http://lists.gnu.org/mailman/listinfo/grub-devel
>   


-- 
Regards
Vladimir 'φ-coder/phcoder' Serbinenko