// CHECK-LABEL: func @fold_identity_test
func.func @fold_identity_test(%arg0: tensor<2x2xf32>, %arg1: tensor<2x2xf32>) -> tensor<2x2xf32> {
  // CHECK: tf.MatMul
  %outputs = "tf.MatMul"(%arg0, %arg1) {device = "/device:CPU:0", transpose_a = false, transpose_b = false} : (tensor<2x2xf32>, tensor<2x2xf32>) -> tensor<2x2xf32>
  // CHECK-NOT: tf.Identity

    %1 = "tf.ReadVariableOp"(%0) {device = ""} : (tensor<!tf_type.resource<tensor<100x50xf32>>>) -> tensor<100x50xf32>
    %2 = "tf.VarHandleOp"() {_class = ["loc:@dense/bias"], allowed_devices = [], container = "", device = "", shared_name = "dense/bias"} : () -> tensor<!tf_type.resource<tensor<50xf32>>>

      %outputs_11, %control_12 = tf_executor.island wraps "tf.Placeholder"() {device = "", dtype = "tfdtype$DT_FLOAT", name = "y", shape = "tfshape$dim { }"} : () -> tensor<0xf32>

  }

  // Returns a device-to-device stream, in round-robin fashion.
  se::Stream* GetDeviceToDeviceStream();

  Status ThenExecute(Device* device, stream_executor::Stream* stream,
                     std::function<void()> func) override;

 private:
  bool UseMultipleStreams() const { return stream_ != host_to_device_stream_; }

  // The main compute stream of the device, used to synchronize the transfer

    %1 = "tf.Relu"(%0) {device = "tpu0"} : (tensor<?xi32>) -> tensor<?xi32>
    %2 = "tf.Add"(%0, %1) {device = "tpu0"} : (tensor<?xi32>, tensor<?xi32>) -> tensor<?xi32>
    %3 = "tf.Relu"(%arg1) {device = "tpu1"} : (tensor<?xi32>) -> tensor<?xi32>
    %4 = "tf.Add"(%3, %arg1) {device = "tpu1"} : (tensor<?xi32>, tensor<?xi32>) -> tensor<?xi32>
    %5 = "tf.Relu"(%4) {device = "tpu0"} : (tensor<?xi32>) -> tensor<?xi32>

                 SP_Event event, TF_Status* const status) {}
void WaitForEvent(const SP_Device* const device, SP_Stream stream,
                  SP_Event event, TF_Status* const status) {}
void CreateTimer(const SP_Device* const device, SP_Timer* timer,
                 TF_Status* const status) {}
void DestroyTimer(const SP_Device* const device, SP_Timer timer) {}

func.func @conv2d_unmatching_loc_pattern(%arg0: tensor<1x3x4x3xf32>) -> (tensor<1x3x2x2xf32>) {
  %cst = "tf.Const"() {device = "", value = dense_resource<__elided__> : tensor<2x3x3x2xbf16>} : () -> tensor<2x3x3x2xbf16>
  %0 = "tf.Cast"(%arg0) {Truncate = false, device = ""} : (tensor<1x3x4x3xf32>) -> tensor<1x3x4x3xbf16>

  // CHECK: {{%.*}} = tfrt_fallback_async.executeop {{.*}} device("/device:CPU:0") "tf.MatMul"
  // CHECK-SAME: {T = f32, transpose_a = false, transpose_b = false}

    %2:3 = "tf.UnpackHyp"(%arg1) {_tpu_replicate = "cluster_Fn", _xla_outside_compilation = "0", device = "", max_seq_length = 16 : i64} : (tensor<*x!tf_type.string>) -> (tensor<*xi32>, tensor<*xi32>, tensor<*xf32>)
    %3 = "tf.Reshape"(%2#2, %1) {_tpu_replicate = "cluster_Fn", _xla_outside_compilation = "0", device = ""} : (tensor<*xf32>, tensor<*xi32>) -> tensor<*xf32>

  %dense_default_1 = "tf.Const"() {device = "/device:CPU:0", dtype = f32, value = dense<[]> : tensor<0xf32>} : () -> tensor<0xf32>
  %dense_keys = "tf.Const"() {device = "/device:CPU:0", dtype = !tf_type.string, value = dense<""> : tensor<2x!tf_type.string>} : () -> tensor<2x!tf_type.string>