Bug 27388 - [i965] piglit glsl-vs-arrays failure
[i965] piglit glsl-vs-arrays failure
Product: Mesa
Classification: Unclassified
Component: Drivers/DRI/i965
Other All
: medium normal
Assigned To: Eric Anholt
Depends on:
Blocks: 29044
  Show dependency treegraph
Reported: 2010-03-31 02:18 UTC by Gordon Jin
Modified: 2010-07-30 21:36 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Note You need to log in before you can comment on or make changes to this bug.
Description Gordon Jin 2010-03-31 02:18:30 UTC
piglit shader/glsl-vs-arrays fails with below output:

Probe at (15,15)
  Expected: 0.000000 1.000000 0.000000
  Observed: 0.000000 0.098039 0.000000
PIGLIT: {'result': 'fail' }

It may be not regression, at least not recent regression.

It passes with software rendering.

It passes (with i965) if I change glsl-vs-arrays.vert as below:
-    temp[index2] = temp[index1];
+    temp[2] = temp[3];

So it indicates the assumption of "index1=3, index2=2" are broken and the uniform vars are not passed through.  
Comment 1 Gordon Jin 2010-03-31 19:30:43 UTC
piglit shader/glsl-vs-mov-after-deref fails with the same problem:

Probe at (15,15)
  Expected: 0.000000 1.000000 0.000000
  Observed: 1.000000 1.000000 1.000000
PIGLIT: {'result': 'fail' }
Comment 2 Eric Anholt 2010-04-12 13:11:59 UTC
The test was created to show this failure.  It's not the uniforms that are the problem, but array access.
Comment 3 Eric Anholt 2010-07-21 22:21:20 UTC
glsl-vs-mov-after-deref failed with half of the problems that glsl-vs-arrays has, and is now fixed.

commit c686ee0fa7e2298408259f5533b739c7d05c78b8
Author: Eric Anholt <eric@anholt.net>
Date:   Wed Jul 21 21:45:34 2010 -0700

    i965: In the VS, multiply the address reg by the appropriate register size.
Comment 4 Eric Anholt 2010-07-23 14:05:24 UTC
commit 35bbbf47425244188334a89163191d9f00bdeced
Author: Eric Anholt <eric@anholt.net>
Date:   Fri Jul 23 10:34:29 2010 -0700

    i965: Add support for VS relative addressing of temporary arrays.
    Fixes glsl-vs-arrays.  Bug #27388.
Comment 5 zhao jian 2010-07-30 21:36:18 UTC
It passed now. so verified.