[Mlir-commits] [mlir] [MLIR][NVGPU] Add truncf and extf Ops (PR #199700)

[Srinivasa Ravi]

================
@@ -699,6 +699,138 @@ LogicalResult RcpOp::verify() {
   return success();
 }
 
+//===----------------------------------------------------------------------===//
+// NVGPU_ConvertFPTruncOp
+//===----------------------------------------------------------------------===//
+
+static bool isShapedContainerType(Type t) {
+  return llvm::isa<VectorType, RankedTensorType, UnrankedTensorType>(t);
----------------
Wolfram70 wrote:

That makes sense. We currently cannot lower the tensor types either, so it looks fairly redundant for now. Perhaps we can add them back if it is required in the future.
Removed tensor type support from these Ops for now in the latest revision. Thanks!

https://github.com/llvm/llvm-project/pull/199700