[Bug 92196] [BDW] ES3-CTS.shaders.uniform_block.random.nested_structs_arrays.9 Fails

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=92196

--- Comment #2 from Jason Ekstrand <jason at jlekstrand.net> ---
I did a little digging on this this afternoon.  I'm going to write down what I
found here before I head home.  I don't know when I'll have time to poke at
this again, so I want to leave notes for the next person.

I believe Connor's patch is a complete red herring; it only changes things
because it causes register allocation to be subtly different on SIMD8.  Running
the test, I noticed that if you run it with INTEL_DEBUG=no16, it fails and,
before the window closes, it flashes yellow.  Looking at the shader it
generates we see

> dEQP_FragColor = vec4(1.0, v_vtxResult, result, 1.0);

So a color of yellow means that v_vtxResult is fine, but result is 0.  This
means that the failure is in one of the comparisons in the fragment shader.  In
particular, the fault is not the vertex shader or nor is it in the
interpolation of the vertex result.  I see two potential ways forward:

 1) Play with the dEQP test case and see if we can get a smaller failing case

 2) Step through the fragment shader in the simulator and see if we can see
where the false comparison is coming from.

My first attempt would probably be number 2 but, as I said above, I don't know
when I'll have time to look at it again so if someone else wants to give it a
try, feel free.

-- 
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/20151008/893f23b4/attachment.html>