this is a patch including both the previous fixes for the Intel PCI->ISA
bridge and a preliminary support for the ALi 1533 bridge. It has had some
preliminary testing and seems to work, though required parameters may
vary (as with smcinit).
I've verified in every way I can that it does the same thing as
tosh1800-smcinit so I would really like someone who can successfully use
tosh1800-smcinit on their Toshibas to test just using this patch onto
Millers tree. Any arguments previously supplied to smcinit would need to
be supplied to the kernel module instead.
Samuel, will you send this upstream?
If anyone has the datasheet or reverse-engineered register map for the ALi
1533 then send it to me. (There are some drivers in the kernel written by
authors who obviously has this, but ALi does not answer my requests.)