[Bug 86469] Unreal Engine demo doesn't run

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=86469

--- Comment #15 from Iago Toral <itoral at igalia.com> ---
(In reply to Matt Turner from comment #10)
> (In reply to Tapani Pälli from comment #8)
> > I dont have access right now but the last clause is something like
> > "regardless of the actual hw implementation thread should not assume
> > registers above 16 can be used" but yep, one can experiment
> 
> I still don't think you're understanding what I'm saying. The statement
> you're referring to says
> 
> > Regardless of actual hardware implementation, the thread should not assume that MRF addresses above m15 wrap to legal MRF registers.
> 
> My point is that you cannot draw any conclusions about MRF > 15 from a page
> that claims they do not exist.
(In reply to Matt Turner from comment #7)
> (In reply to Tapani Pälli from comment #6)
> > SNB bspec, EU Overview -> Messages -> Message Register File (MRF), last
> > paragraph.
> 
> That page doesn't say "it's not safe". If anything it says MRFs > 15 don't
> exist ("There are 16 MRF registers").
> 
> It wouldn't surprise me if the registers don't exist and the "24" is a
> mistake, but it also wouldn't surprise me if SNB actually has 24 MRFs and
> they just didn't update a bunch of documentation. Experimentation is the
> only way to be sure.


FWIW, I've sent a RFC series to the mailing list that implements this test:
http://lists.freedesktop.org/archives/mesa-dev/2015-September/094584.html

Spoiler alert: SNB supports 24 MRF registers

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/intel-3d-bugs/attachments/20150916/18bc69e2/attachment.html>