attr->contains("DstT") &&
      attr->at("SrcT").type() == attr->at("DstT").type() &&
      attr->contains("Truncate") && !attr->at("Truncate").b()) {
    node_def->set_op("Identity");
    attr->insert({{"T", attr->at("SrcT")}});
    attr->erase("SrcT");
    attr->erase("DstT");
    attr->erase("Truncate");
  }
}

}  // namespace

Status GetAttrValuesFromOperation(

	shouldPanic("SetCap", func() { vs.SetCap(10) })
	shouldPanic("SetLen", func() { vs.SetLen(-1) })
	shouldPanic("SetCap", func() { vs.SetCap(-1) })
	shouldPanic("SetCap", func() { vs.SetCap(6) }) // smaller than len
	vs.SetLen(5)
	if len(xs) != 5 || cap(xs) != 8 {
		t.Errorf("after SetLen(5), len, cap = %d, %d, want 5, 8", len(xs), cap(xs))
	}
	vs.SetCap(6)
	if len(xs) != 5 || cap(xs) != 6 {

        TensorShapeProto shape_proto;
        context->ShapeHandleToProto(handle, &shape_proto);
        if (!shape_proto.unknown_rank()) {
          NodeDef const_def;
          const_def.set_op("Const");
          Node* var_node;
          TF_RETURN_IF_ERROR(n->input_node(0, &var_node));
          const_def.set_name(
              graph->NewName(absl::StrCat("var_shape_", var_node->name())));

          if (TF::TensorListSetItemOp set_op =
                  llvm::dyn_cast<TF::TensorListSetItemOp>(use.getOwner())) {
            element_shape = rewriter.create<TF::ShapeOp>(
                op.getLoc(),
                tensorflow::GetTypeFromTFTensorShape({-1}, shape_dtype),
                set_op.getItem());
            element_shape_acquired = true;

  if (!name.empty())
    node_def->set_name(std::string(ParseTensorName(name.str()).node()));
  else
    node_def->set_name(
        std::string(op_to_name_.GetUniqueName(func.getName().str())));

  node_def->set_op(FunctionLibraryDefinition::kArgOp);

  mlir::TensorType arg_type = mlir::cast<mlir::TensorType>(arg.getType());
  if (auto resource_type =
          mlir::dyn_cast<mlir::TF::ResourceType>(arg_type.getElementType())) {

  if (!name.empty())
    node_def->set_name(std::string(ParseTensorName(name.str()).node()));
  else
    node_def->set_name(
        std::string(op_to_name_.GetUniqueName(func.getName().str())));

  node_def->set_op(FunctionLibraryDefinition::kArgOp);

  mlir::TensorType arg_type = mlir::cast<mlir::TensorType>(arg.getType());
  if (auto resource_type =
          mlir::dyn_cast<mlir::TF::ResourceType>(arg_type.getElementType())) {

# This is the image used in a multi-stage build to apply capabilities to Docker-wrapped binaries.
readonly KUBE_BUILD_SETCAP_IMAGE="${KUBE_BUILD_SETCAP_IMAGE:-$KUBE_BASE_IMAGE_REGISTRY/setcap:$__default_setcap_version}"

# Get the set of master binaries that run in Docker (on Linux)
# Entry format is "<binary-name>,<base-image>".
# Binaries are placed in /usr/local/bin inside the image.

func (s *ProxyServer) setupConntrack(ctx context.Context) error {
	ct := &realConntracker{}

	max, err := getConntrackMax(ctx, s.Config.Conntrack)
	if err != nil {
		return err
	}
	if max > 0 {
		err := ct.SetMax(ctx, max)
		if err != nil {
			if err != errReadOnlySysFS {
				return err
			}
			// errReadOnlySysFS is caused by a known docker issue (https://github.com/docker/docker/issues/24000),

    TF_ASSIGN_OR_RETURN(auto tf_name,
                        GetTensorFlowOpName(inst->getName().getStringRef()));
    op_name.append(tf_name);
  }

  node_def->set_name(name.str());
  node_def->set_op(std::string(op_name.str()));

  // Update NodeDef constructed out of an MLIR Case/If/While op to map it to
  // either TensorFlow StatelessX or X op depending on the additional attribute.

  const int num_inputs = input_shapes->num_items;
  NodeDef node_def;
  tensorflow::ImmediateExecutionOperation* op = tensorflow::unwrap(tfe_op);
  node_def.set_name(op->Name());
  node_def.set_op(op->Name());
  for (int i = 0; i < num_inputs; ++i) {
    node_def.add_input("dummy_input");
  }
  OperationFromInterface(op)->Attrs().FillAttrValueMap(node_def.mutable_attr());