Mesa (main): aco: skip s_barrier if TCS patches are within subgroup

[GitLab Mirror]

Module: Mesa
Branch: main
Commit: 152092b8eaddd46a11f1a08514742e72ba1c195e
URL:    http://cgit.freedesktop.org/mesa/mesa/commit/?id=152092b8eaddd46a11f1a08514742e72ba1c195e

Author: Rhys Perry <pendingchaos02 at gmail.com>
Date:   Wed May  4 13:03:52 2022 +0100

aco: skip s_barrier if TCS patches are within subgroup

fossil-db (Sienna Cichlid):
Totals from 518 (0.32% of 162293) affected shaders:
Instrs: 124943 -> 123908 (-0.83%)
CodeSize: 708764 -> 704624 (-0.58%)
Latency: 618380 -> 618279 (-0.02%)
InvThroughput: 214061 -> 214051 (-0.00%)

Signed-off-by: Rhys Perry <pendingchaos02 at gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof at gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/16356>

---

 src/amd/compiler/aco_instruction_selection.cpp | 18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)

diff --git a/src/amd/compiler/aco_instruction_selection.cpp b/src/amd/compiler/aco_instruction_selection.cpp
index 4bd7ca1cea6..8996d4835fd 100644
--- a/src/amd/compiler/aco_instruction_selection.cpp
+++ b/src/amd/compiler/aco_instruction_selection.cpp
@@ -11004,13 +11004,6 @@ create_fs_exports(isel_context* ctx)
    ctx->block->kind |= block_kind_export_end;
 }
 
-static void
-create_workgroup_barrier(Builder& bld)
-{
-   bld.barrier(aco_opcode::p_barrier,
-               memory_sync_info(storage_shared, semantic_acqrel, scope_workgroup), scope_workgroup);
-}
-
 static void
 emit_stream_output(isel_context* ctx, Temp const* so_buffers, Temp const* so_write_offset,
                    const struct radv_stream_output* output)
@@ -11522,8 +11515,15 @@ select_program(Program* program, unsigned shader_count, struct nir_shader* const
          bool tcs_skip_barrier = ctx.stage == vertex_tess_control_hs &&
                                  ctx.tcs_temp_only_inputs == nir->info.inputs_read;
 
-         if (!ngg_gs && !tcs_skip_barrier)
-            create_workgroup_barrier(bld);
+         if (!ngg_gs && !tcs_skip_barrier) {
+            sync_scope scope =
+               ctx.stage == vertex_tess_control_hs &&
+                     program->wave_size % ctx.options->key.tcs.tess_input_vertices == 0
+                  ? scope_subgroup
+                  : scope_workgroup;
+            bld.barrier(aco_opcode::p_barrier,
+                        memory_sync_info(storage_shared, semantic_acqrel, scope), scope);
+         }
 
          if (ctx.stage == vertex_geometry_gs || ctx.stage == tess_eval_geometry_gs) {
             ctx.gs_wave_id = bld.pseudo(aco_opcode::p_extract, bld.def(s1, m0), bld.def(s1, scc),