And by "something like this", I obviously _really_ meant "something
almost, but not entirely, unlike this".
The "insert_resource_expand_to_fit()" calling convention is much simpler,
and it should be just
insert_resource_expand_to_fit(&iomem_resource, res);
which has the downside (?) that it will actually _modify_ the resource as
it expands it to fit. Of course, it's not like we don't do that in other
places too (ie the "sanitize" phase), so I don't think it's much of a real
downside.
Anyway, I'm not appending a fixed patch, becuase the fix is so trivial as
to be more easily done by hand.
Linus
--