[Bug 109216] 4-27% performance drop in Vulkan benchmarks

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=109216

--- Comment #18 from Eero Tamminen <eero.t.tamminen at intel.com> ---
(In reply to Ian Romanick from comment #17)
> I put up a WIP: merge request.  This series fixes the issue in the Sascha
> Willems demos for me.  I get the same number of loops with this series as
> there was before 8fb8ebfbb05d.
> 
> There's a one instruction regression in the deferredmultisampling demo
> compared to 8fb8ebfbb05d~1, but I can live with that.
> 
> https://gitlab.freedesktop.org/mesa/mesa/merge_requests/107
> 
> I can't get a successful CI run even with Mesa master right now, so this is
> definitely not fully tested.

Your branch includes bug 109304 so it won't work with Aztec Ruins (which also
regressed).

If you can provide patch series I can apply on top of latest Mesa, I could do
full test with it.

-- 
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/20190114/b6a6abcc/attachment-0001.html>