auto resource_type =
          mlir::TF::ResourceType::get({resource_subtype}, builder.getContext());

      auto tensor_type = mlir::cast<mlir::TensorType>(mlir_arg.getType());
      if (tensor_type.hasRank()) {
        mlir_arg.setType(
            GetTypeFromTFTensorShape(tensor_type.getShape(), resource_type));
      } else {
        mlir_arg.setType(mlir::UnrankedTensorType::get(resource_type));
      }

    if (matchPattern(op_state.operands[1], m_Constant(&shape_attr))) {
      auto shape_ty =
          op_state.operands[1].getType().dyn_cast<RankedTensorType>();
      if (shape_ty != nullptr && shape_ty.hasRank() && shape_ty.getRank() > 1) {
        llvm::SmallVector<mlir::Attribute, 4> shape;
        int32_t dim_size = 0;
        for (const auto& dim :
             llvm::enumerate(shape_attr.getValues<llvm::APInt>())) {

    // Here we assume that the element_shape won't be changed before calling
    // the first `TensorListSetItemOp`.
    if (auto shaped_type = element_shape.getType().dyn_cast<ShapedType>()) {
      if (shaped_type.hasRank() && shaped_type.getRank() == 0) {
        bool element_shape_acquired = false;
        auto uses = op.getResult().getUses();
        for (auto &use : llvm::make_early_inc_range(uses)) {

			t.lastHeading = dt.heading
			t.out(dt.heading)
		}
		if vflag > 0 {
			fmt.Printf("# go tool dist test -run=^%s$\n", dt.name)
		}
	}
}

func (t *tester) hasBash() bool {
	switch gohostos {
	case "windows", "plan9":
		return false
	}
	return true
}

// hasParallelism is a copy of the function