[Mlir-commits] [mlir] [MLIR][NVVM] Add new narrow FP convert Ops (PR #184291)

[Durgadoss R]

================
@@ -347,6 +347,32 @@ LogicalResult ConvertF32x2ToF6x2Op::verify() {
   return success();
 }
 
+LogicalResult ConvertF16x2ToF6x2Op::verify() {
+  mlir::MLIRContext *ctx = getContext();
+
+  if (!llvm::isa<mlir::Float6E2M3FNType, mlir::Float6E3M2FNType>(getDstTy())) {
----------------
durga4github wrote:

Can we not constrain this in the Op itself, like how we added the EnumAttr NoneOf/AnyOf constraints recently? This way, we do not need some of these Verifiers at all.

https://github.com/llvm/llvm-project/pull/184291