[Bug 97532] Regression: GLB 2.7 & Glmark-2 segfault due to linker precision error (259fc505) on dead variable

bugzilla-daemon at freedesktop.org bugzilla-daemon at freedesktop.org
Wed Sep 7 13:32:49 UTC 2016


https://bugs.freedesktop.org/show_bug.cgi?id=97532

Eero Tamminen <eero.t.tamminen at intel.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|Regression: GLB 2.7         |Regression: GLB 2.7 &
                   |segfaults due to bogus      |Glmark-2 segfault due to
                   |linker precision error      |linker precision error
                   |(259fc505) on dead variable |(259fc505) on dead variable

--- Comment #15 from Eero Tamminen <eero.t.tamminen at intel.com> ---
Glmark2 GLES version in latest Ubuntu (16.04) also segfaults to this in
following subtests:
- ideas (modelview uniform declared also in FS but not used)
- jellyfish (uCurrentTime uniform used both in VS & FS)
- terrain (viewMatrix uniform declared also in FS but not used)

For example:
-------------------------
$ apt-get install glmark2-es2
$ glmark2-es2 -b terrain
...
Error: Failed to link program created from files None and None:  error:
declarations for uniform `viewMatrix` have mismatching precision qualifiers
Segmentation fault (core dumped)
-------------------------

Like GLB, it's GLES v2 program.

Ideas & terrain can be run by fixing Tapani's patch to check whether uniform is
unused in either of the stages (not just VS), but jellyfish test is using the
uniform in both stages with different precision, so it has a real bug (which
might show on backends that implement FP16 support).


Glmark2 is open source so it can be (eventually) fixed:
  https://github.com/glmark2/glmark2

But I don't think this and GLB are the only GLES 2.x programs with the problem.

-- 
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/20160907/2612f8ca/attachment-0001.html>


More information about the intel-3d-bugs mailing list