[Bug 97287] GL45-CTS.vertex_attrib_binding.basic-inputL-case1 fails

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=97287

            Bug ID: 97287
           Summary: GL45-CTS.vertex_attrib_binding.basic-inputL-case1
                    fails
           Product: Mesa
           Version: git
          Hardware: Other
                OS: All
            Status: NEW
          Severity: normal
          Priority: medium
         Component: Drivers/DRI/i965
          Assignee: idr at freedesktop.org
          Reporter: idr at freedesktop.org
        QA Contact: intel-3d-bugs at lists.freedesktop.org

This test configures 4 double-precision vertex arrays as inputs and uses a
trivial shader with transform feedback to write the data out.  XFB output 6
(which comes from an immediate-mode value) and output 7 (which comes from the
last vertex array) have incorrect values.

I investigated this a bit:

 * I modified the shader to use immediate values to set the outputs.  This
produced the correct results, so I believe that XFB of doubles works correctly.

 * I thought some of the offsets or strides (vertex array 7 uses a zero stride
to get the same value for both vertices) could cause problems.  Modifying those
values did not help.

 * Disabling one of the earlier vertex arrays (I commented out the
"glEnableVertexAttribArray(5);" line and the lines that set expected_data[5]
and [5+8]) causes outputs 6 and 7 to get the expected values.

GL45-CTS.vertex_attrib_binding.basic-inputL-case2 is very similar.  However, it
only uses 3 inputs, and it passes.

My current theory is that having 4+ 64-bit inputs causes problems.  I don't
know that code very well, so I'm not sure where to start looking.  I'll come
back to this problem later, but if someone more familiar with that code has any
bright ideas, feel free to take the bug.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-3d-bugs/attachments/20160811/926880f9/attachment.html>