iris: Ask st to vectorize our IO.

(Technically this is common code, but it doesn't affect i965 or anv.)

Improves performance of GFXBench5/gl_tess_off on Skylake GT4e at 1080p
by 9.3933% +/- 0.0305157% by eliminating all spilling in the GS.

Improves performance of GFXBench5/gl_4_off (Car Chase) on Skylake GT4e
at 1080p by 0.325208% +/- 0.0842233% (n=18).

Reviewed-by: Marek Olšák <marek.olsak@amd.com>

diff --git a/src/intel/compiler/brw_compiler.c b/src/intel/compiler/brw_compiler.c
index bded536c9485379640cafd1ef967d81ba904399b..536dc86ab78954ba07f43ce86b853fd5e73586f9 100644 (file)
--- a/src/intel/compiler/brw_compiler.c
+++ b/src/intel/compiler/brw_compiler.c
@@ -46,6 +46,7 @@
    .lower_isign = true,                                                       \
    .lower_ldexp = true,                                                       \
    .lower_device_index_to_zero = true,                                        \
+   .vectorize_io = true,                                                      \
    .use_interpolated_input_intrinsics = true,                                 \
    .vertex_id_zero_based = true,                                              \
    .lower_base_vertex = true