Pardon the silly question, but what's preventing us (i.e. someone capable!) from compiling and distributing a Xnu kernel that runs on the Xeon Prestonia (the one we're all having trouble with)?

I should imagine it's just a matter of telling the kernel to recognize that CPU's subtype, as opposed to the Nocona which it already supports.

The only real differences are that the Nocona supports a quad-cpu architecture and comes with more cache. They're both single-core hyperthreading P4 Xeons at heart.

Is this possible or am I off the mark on something? I'm happy to test on a zillion different machines if anyone would like, I've got access to a *lot* of cpus and boards.

Cheers,
mindslip

probably we might have more luck with the Xeon Willamente.. I dont think Perstonia supports SSE3..Nocona is the only Xeon that has SSE3. So Willamente has also no SSE3, only SSE2 but works..

Did this get figured out?

Seriously! It seems like this would be a breeze for hackers of the world. I would definately pay a good $10 (PayPal) to anybody who can help me out. And I'm sure you can find another 4 people like me and buy yourself a nice black DVD burner to match your underground hacking chamber. I shouldn't have said that.

I'm sick of Windows XP, and Vista is--to be politically correct--really stupid.

And yeah, ok, Eskurza, I'll downgrade my system.

Not all of the XNU kernel is opensourced, enough of it is not released that the compiled Darwin kernel will not boot OS X, hence why we need to patch the Kernel for SSE2