> I remember reading about the issue with UltraSparc documentation. Derek
> Warren has raised teh point in a blog comment on John Clingan's blog here.
>
> http://blogs.sun.com/roller/page/jcl...nk_of#comments

We don't need processor docs. That stuff is trivial.

We need *chipset docs*.

And those are still not available under Sun's new program.

So it is STILL a complete matter of reverse engineering.

Same as with Apple, of course: powerpc is trivial. But the chipsets
on every Apple machine are subtly different, full of bugs, and reading
Darwin is a terrific waste of time since the code is so horrid, and of
course, it is all just a maze of workarounds.

Verilog for ultrasparc? Give me a break. We don't care.