radv: Use nir_fold_16bit_image_load_store_conversions.

Totals from 10 (0.01% of 134913) affected shaders:
CodeSize: 53316 -> 53168 (-0.28%)
Instrs: 9219 -> 9117 (-1.11%)
Latency: 41744 -> 41595 (-0.36%)
InvThroughput: 16616 -> 16412 (-1.23%)

Signed-off-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15179>
This commit is contained in:
Georg Lehmann 2022-02-25 18:55:42 +01:00 committed by Marge Bot
parent 7a6dbe0c77
commit 9bca149353

View file

@ -4514,6 +4514,7 @@ radv_create_shaders(struct radv_pipeline *pipeline, struct radv_pipeline_layout
/* Skip if there are potentially conflicting rounding modes */
if (!nir_has_any_rounding_mode_enabled(stages[i].nir->info.float_controls_execution_mode))
NIR_PASS_V(stages[i].nir, nir_fold_16bit_sampler_conversions, 0, sampler_dims);
NIR_PASS_V(stages[i].nir, nir_fold_16bit_image_load_store_conversions);
NIR_PASS_V(stages[i].nir, nir_opt_vectorize, opt_vectorize_callback, NULL);
}