OpenHarmony/third_party_mesa3d
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

aco: handle clustered uniform reductions correctly

Browse Source 
Alternatively we could just trust divergence analysis to do the right thing.

Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30235>
This commit is contained in:
Georg Lehmann
2024-07-19 08:39:32 +02:00
committed by Marge Bot
parent 65acc81e9d
commit efb9258814
1 changed files with 6 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
src/amd/compiler/aco_instruction_selection.cpp
View File

@@ -8835,16 +8835,18 @@ visit_intrinsic(isel_context* ctx, nir_intrinsic_instr* instr)
const unsigned bit_size = instr->src[0].ssa->bit_size;
assert(bit_size != 1);
if (!nir_src_is_divergent(instr->src[0]) && cluster_size == ctx->program->wave_size) {
if (!nir_src_is_divergent(instr->src[0])) {
/* We use divergence analysis to assign the regclass, so check if it's
* working as expected */
ASSERTED bool expected_divergent = instr->intrinsic == nir_intrinsic_exclusive_scan;
if (instr->intrinsic == nir_intrinsic_inclusive_scan)
expected_divergent = op == nir_op_iadd || op == nir_op_fadd || op == nir_op_ixor;
if (instr->intrinsic == nir_intrinsic_inclusive_scan ||
cluster_size != ctx->program->wave_size)
expected_divergent = op == nir_op_iadd || op == nir_op_fadd || op == nir_op_ixor ||
op == nir_op_imul || op == nir_op_fmul;
assert(instr->def.divergent == expected_divergent);
if (instr->intrinsic == nir_intrinsic_reduce) {
if (emit_uniform_reduce(ctx, instr))
if (!instr->def.divergent && emit_uniform_reduce(ctx, instr))
break;
} else if (emit_uniform_scan(ctx, instr)) {
break;