[Mlir-commits] [mlir] [MLIR][NVVM] Add nvvm.fadd and nvvm.fsub Ops (PR #179162)

[Guray Ozen]

================
@@ -6176,10 +6176,9 @@ def NVVM_Tcgen05MMAWsSparseOp : NVVM_Op<"tcgen05.mma.ws.sp",
   }];
 }
 
-def NVVMFloatType : AnyTypeOf<[F16, BF16, F32, F64, VectorOfLengthAndType<[2], [F16, BF16]>]>;
+def SIMTFloatType : AnyTypeOf<[F16, BF16, F32, F64, VectorOfLengthAndType<[2], [F16, BF16, F32, F64]>]>;
 
-def NVVM_FloatAdditionOp : 
-  NVVM_SingleResultIntrinsicOp<"fadd", [Pure, Commutative]> {
+def NVVM_FAddOp : NVVM_Op<"fadd", [Pure, Commutative]> {
----------------
grypp wrote:

Maybe you can also create `Arith_FloatBinaryOp` helper class. I feel like arithmetic ops will grow in nvvm dialect. 

https://github.com/llvm/llvm-project/pull/179162