[Bug 103824] [SNB] SynMark GSCloth test renders wrong

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=103824

            Bug ID: 103824
           Summary: [SNB] SynMark GSCloth test renders wrong
           Product: Mesa
           Version: git
          Hardware: Other
                OS: All
            Status: NEW
          Severity: normal
          Priority: medium
         Component: Drivers/DRI/i965
          Assignee: intel-3d-bugs at lists.freedesktop.org
          Reporter: eero.t.tamminen at intel.com
        QA Contact: intel-3d-bugs at lists.freedesktop.org

Created attachment 135603
  --> https://bugs.freedesktop.org/attachment.cgi?id=135603&action=edit
Incorrect rendering screenshot

Mesa fails SynMark render validation for OglGSCloth test on SNB:
   ./synmark2 OglGSCoth

Rendering is obviously incorrect (see the the attachment).  Francesco's recent
GEN6-7 Transform feedback optimization patch series seems to make this somewhat
worse.

I don't know whether Mesa has ever rendered it correctly on SNB, the issue has
been there at least a year.  Rendering is fine on GEN7+ (also with Francesco's
patch series).

Unlike most of the other tests we run regularly (which work fine), GSCloth uses
geometry shaders, streamout and transform feedback.  GfxBench Manhattan uses
also transform feedback, but that at least didn't have obvious rendering issues
on SNB like this test.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-3d-bugs/attachments/20171120/d083cbef/attachment-0001.html>