diff options
| author | Jason Ekstrand <[email protected]> | 2017-08-28 17:33:33 -0700 |
|---|---|---|
| committer | Jason Ekstrand <[email protected]> | 2018-03-07 12:13:47 -0800 |
| commit | 44681e47955687d5590779fdc44373d3fb204fdc (patch) | |
| tree | 147add39a1abf8411c643b306e510fb71f14624c /src/amd | |
| parent | 9812fce60b6ffbcd136b66bfb609143449ad3f7c (diff) | |
nir: Generalize nir_intrinsic_vote_eq
The SPIR-V extension wants us to be able to do an AllEqual on any vector
or scalar type. This has two implications:
1) We need to be able to handle vectors so we switch the vote_eq
intrinsics to be vectorized intrinsics.
2) We need to handle floats which have different behavior with respect
to +-0, NaN, etc. than the integer variant so we need two variants.
Reviewed-by: Lionel Landwerlin <[email protected]>
Diffstat (limited to 'src/amd')
| -rw-r--r-- | src/amd/common/ac_nir_to_llvm.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/src/amd/common/ac_nir_to_llvm.c b/src/amd/common/ac_nir_to_llvm.c index cca796de71b..4b3b253c439 100644 --- a/src/amd/common/ac_nir_to_llvm.c +++ b/src/amd/common/ac_nir_to_llvm.c @@ -4563,7 +4563,7 @@ static void visit_intrinsic(struct ac_nir_context *ctx, result = LLVMBuildSExt(ctx->ac.builder, tmp, ctx->ac.i32, ""); break; } - case nir_intrinsic_vote_eq: { + case nir_intrinsic_vote_ieq: { LLVMValueRef tmp = ac_build_vote_eq(&ctx->ac, get_src(ctx, instr->src[0])); result = LLVMBuildSExt(ctx->ac.builder, tmp, ctx->ac.i32, ""); break; |