[PATCH] fix for partmap detection on RAID/LVM

[Robert Millan]

This is my proposed approach for fixing the partmap detection problem in
LVM/RAID.

The problem: grub-probe just tests for partmap in the target drive directly
(which usually doesn't have any), and not in the physical drives that make
the RAID array (or LVM volume).

My proposed solution:  RAID implements grub_raid_parent, which returns a
linked list of "parent" (alternate naming suggestions welcome!) drives
conforming a specific array.  grub_disk_t abstracts that by providing a
parent() member (for GRUB_UTIL only) on top of it (so e.g. LVM can
implement the same).  grub-probe receives the list from this interface
and probes partmap in each member.

I had in mind something nifty with probe() recursing into itself, in order
to support bizarre combinations like lvm inside dm-raid, but on second
thoughts my current change is not intrusive and wouldn't preclude this kind
of restructuring if someone feels like it.

Comments?

-- 
Robert Millan

<GPLv2> I know my rights; I want my phone call!
<DRM> What use is a phone call… if you are unable to speak?
(as seen on /.)

raid_partmap.diff
Description: Text Data