Semantics of the 'dumb' interface

[Alan Cox]

> I think the intel driver forces a reset to the system scanout on release. I've
> actually never found a test to indicate it was completely necessary on
> other GPUs

I need to force this on the GMA500 because we want some kind of sane
scanout left and the nice guys at Dell decided the Mini 10 didn't
actually need a sysrq key so sysrq-V isn't going to help.

On the bright side GEM and the dumb kms stuff now seem to work with a new
device id added to libkms and a daft bug in libkms dumb.c fixed (if you
report a 32bit fb type, you probably want to allocate a 32bit fb not a
16bit one)

Alan