radv: lower ffma in nir.

So it appears the Vulkan SPIR-V fma opcode can be equivalent to a
mad operation, and the fma hw opcode on AMD hw is issued like a double
opcode so is slower. Also the radeonsi stack does this.

This appears to improve performance on a number of games from Feral,
and thanks to Feral for noticing the problem.

I'm reposting this one as Marek indicated he thinks this is what
we should be doing on AMD hw.

Reviewed-by: Nicolai Hähnle <[email protected]>
Reviewed-by: Marek Olšák <[email protected]>
Cc: "17.2" <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
(cherry picked from commit 2c61594d84911f486aa2edb4b8e561e780139d20)
[Emil Velikov: use correct file radv_shader.c -> radv_pipeline.c]
Signed-off-by: Emil Velikov <[email protected]>

Conflicts:
	src/amd/vulkan/radv_shader.c

1 files changed, 1 insertions, 0 deletions

diff --git a/src/amd/vulkan/radv_pipeline.c b/src/amd/vulkan/radv_pipeline.c
index bbcaeaa4e1f..ddd146424e8 100644
--- a/src/amd/vulkan/radv_pipeline.c
+++ b/src/amd/vulkan/radv_pipeline.c
@@ -65,6 +65,7 @@ static const struct nir_shader_compiler_options nir_options = {
 	.lower_unpack_unorm_4x8 = true,
 	.lower_extract_byte = true,
 	.lower_extract_word = true,
+	.lower_ffma = true,
 	.max_unroll_iterations = 32
 };