[snip]
Now that ppc64 is basically working I continued playing with Mark's v1. I noticed a few glitches, some of which I've started fixing. ppc64 would no longer reach the prompt with v1 applied; I noticed that, e.g., the claim_phys functions did not yet return a phys_addr_t despite using it internally. With some fixes like these it is now working.
ppc64 required some more work for the MMU translations - I needed to tweak some sizes and squeeze in some additional cells for the properties. Still need to add some more .properties pretty-printing to confirm whether the values are actually right now.
http://repo.or.cz/w/openbios/afaerber.git/shortlog/refs/heads/ppc64-ofmem
What I didn't yet look into is that the add_range() etc. functions still operate on ucell despite the range_t using a phys_addr_t now. ppc's ea_to_phys() also still has some assumptions about memory layout that need fixing.
Andreas