[Mesa-dev] [PATCH v2 2/3] nir: Add a discard optimization pass
eero.t.tamminen at intel.com
Wed Jul 11 10:05:27 UTC 2018
On 06.07.2018 00:28, Jason Ekstrand wrote:
> On Thu, Jul 5, 2018 at 2:18 PM, Jason Ekstrand <jason at jlekstrand.net
> <mailto:jason at jlekstrand.net>> wrote:
> >> Optimizing for the latter case is an essentially
> >> heuristic assumption that needs to be verified experimentally. Have you
> >> tested the effect of this pass on non-DX workloads extensively?
> > Yes, it is a trade-off. No, I have not done particularly extensive
> > testing. We do, however, know of non-DXVK workloads that would benefit
> > from this. I believe Manhattan is one such example though I have not yet
> > benchmarked it.
> You should grab some numbers then to make sure there are no
> I'm working on that. Unfortunately the perf system is giving me
> trouble so I don't have the numbers yet.
> But keep in mind that the i965 scheduler is already
> performing a similar optimization (locally, but with cycle-count
> information). This will only help over the existing
> optimization if the
> shaders that represent a bottleneck in Manhattan have sufficient
> flow for the basic block boundaries to represent a problem to the
> (local) scheduler.
> I'm not sure about the manhattan shader but the Skyrim shader does
> have control flow which the discard has to get moved above.
> I have results from the perf system now and somehow this pass makes
> manhattan noticeably worse. I'll look into that.
Note: All the more complex GfxBench tests use discard:
Egypt, T-Rex, Manhattan, CarChase, AztecRuins...
(Most of the other benchmarks we're running, don't use them.)
More information about the mesa-dev