[Bug 107510] [GEN8+] up to 10% perf drop on several 3D benchmarks

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=107510

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |VERIFIED

--- Comment #25 from Eero Tamminen <eero.t.tamminen at intel.com> ---
(In reply to Eero Tamminen from comment #0)
> On SKL GT2, the regressions are following:
> - 10% GfxBench Tessellation (onscreen & offscreen)
> - 5% SynMark2 GSCloth
> - 2-3% GpuTest Julia FP64
> - 1-2% Unigine Heaven (high quality with tessellation, no AA)
> - 1% GfxBench ALU2 (onscreen & offscreen), SynMark PSPom

(In reply to Jason Ekstrand from comment #24)
> This is fixed by the following commit now in master:
...
>     intel/nir: Vectorize all IO

All the other tests improved by the amount regressed, except for these two
which were not affected by the fix:
- GpuTest Julia FP64
- SynMark PSPom

I assume PSPom drop was fixed already in October, with the "nir: Copy
propagation between blocks" commit (but it regressed last month, I'll file
separate bug for that).  FP64 case isn't really relevant.

VERIFIED.

-- 
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/20190313/68ea871b/attachment.html>