%0 = "tf_device.cluster_func"(%arg0, %arg1) {func = @_func, use_spmd_for_xla_partitioning = true, use_tpu = true, num_cores_per_replica = 2 : i64} : (tensor<2x4xf32>, tensor<4x2xf32>) -> tensor<2x2xf32>
  %1:2 = "tf.TPUPartitionedOutputV2"(%0) {device = "", partition_dims = [2, 1]} : (tensor<2x2xf32>) -> (tensor<1x2xf32>, tensor<1x2xf32>)

// CHECK-NEXT:  %[[argmax:.*]] = "tfl.arg_max"(%[[softmax]], %[[cst]]) : (tensor<128x16xf32>, tensor<i32>) -> tensor<128xi32>
  %3 = "tfl.arg_max"(%2, %cst) : (tensor<128x16xf32>, tensor<i32>) -> tensor<128xi32>
// CHECK-NEXT:  return %[[softmax]], %[[argmax]] : tensor<128x16xf32>, tensor<128xi32>
  func.return %2, %3 : tensor<128x16xf32>, tensor<128xi32>
}

// CHECK-LABEL: PruneUnusedLstm

// CHECK:}

func.func @reshape(%arg0 : tensor<1x128xi32>) -> tensor<4x32x1xi32>{
  %0 = "vhlo.reshape_v1"(%arg0) : (tensor<1x128xi32>) -> tensor<4x32x1xi32>
  func.return %0 : tensor<4x32x1xi32>
}

//CHECK:func.func private @reshape(%arg0: tensor<1x128xi32>) -> tensor<4x32x1xi32> {
//CHECK-NEXT: %0 = "vhlo.reshape_v1"(%arg0) : (tensor<1x128xi32>) -> tensor<4x32x1xi32>
//CHECK-NEXT: return %0 : tensor<4x32x1xi32>

// `tfl.concatenation`.

func.func @concatenate_float(%arg0: tensor<3x2xf32>, %arg1: tensor<1x2xf32>) -> tensor<4x2xf32> {
  %0 = "stablehlo.concatenate"(%arg0, %arg1) {dimension = 0 : i64} : (tensor<3x2xf32>, tensor<1x2xf32>) -> tensor<4x2xf32>
  return %0 : tensor<4x2xf32>
}
// CHECK-LABEL: concatenate_float
// CHECK-NOT: tfl.concatenation
// CHECK: stablehlo.concatenate

// -----

  func.func @dynamic_update_slice(%arg0: tensor<3x4xi32>, %arg1: tensor<2x2xi32>, %arg2: tensor<2xi32>) -> tensor<3x4xi32> {

    // CHECK: %[[SLICE0:.*]] = "mhlo.slice"(%[[ARG2]])
    // CHECK-DAG-SAME: start_indices = dense<0> : tensor<1xi64>
    // CHECK-DAG-SAME: limit_indices = dense<1> : tensor<1xi64>
    // CHECK-DAG-SAME: strides = dense<1> : tensor<1xi64>

  return %2 : tensor<1x3xf32>
}
func.func private @composite_dot_general_fn_1(%arg0: tensor<1x1024xf32>, %arg1: tensor<1024x3xf32>) -> tensor<1x3xf32> attributes {_from_xla_call_module} {
  %0 = stablehlo.dot_general %arg0, %arg1, contracting_dims = [1] x [0] : (tensor<1x1024xf32>, tensor<1024x3xf32>) -> tensor<1x3xf32>
  return %0 : tensor<1x3xf32>
}
// CHECK: func.func @main

      %2 = "tf.XlaCallModule"(%arg0, %1, %0) <{Sout = [#tf_type.shape<?x2>], module = "", version = 9 : i64}> {_original_entry_function = "composite_fn_1"} : (tensor<?x2xf32>, tensor<2x2xf32>, tensor<2xf32>) -> tensor<?x2xf32>
      return %2 : tensor<?x2xf32>
    }

    ```mlir
      %0 = "tf.Const"() {value = dense<[[42.0]]> : tensor<1x1xf32>} : () -> tensor<1x1xf32>
      %1 = "tf.Const"() {device = "", value = dense<[[42.0]]> : tensor<1x1xf32>} : () -> tensor<1x1xf32>
      %2 = "tf.Const"() {device = "baz", value = dense<[[42.0]]> : tensor<1x1xf32>} : () -> tensor<1x1xf32>
    ```

    then running this pass with 'default-device=foobar', we get:

    ```mlir

//     : (tensor<2xf32>, tensor<2xi64>) -> tensor<1x2xf32>
//   %inp1 = "tf.Reshape"(%arg1, %shape)
//     : (tensor<2x2x2xf32>, tensor<2xi64>) -> tensor<4x2xf32>
//   %items0 = "tf.Unpack"(%[[INP0]]) {axis = 0 : i64}
//     : (tensor<1x2xf32>) -> tensor<2xf32>
//   %items1:4 = "tf.Unpack"(%[[INP1]]) {axis = 0 : i64}
//     : (tensor<4x2xf32>) -> (tensor<2xf32>, tensor<2xf32>, tensor<2xf32>,
//     tensor<2xf32>)

func.func @dot_general_with_bias_same_shape_fn(%arg0: tensor<1x2xf32>) -> tensor<1x3xf32> {
  %0 = stablehlo.constant dense<2.000000e+00> : tensor<2x3xf32>
  %1 = stablehlo.constant dense<2.000000e+00> : tensor<1x3xf32>
  %2 = stablehlo.dot_general %arg0, %0, contracting_dims = [1] x [0], precision = [DEFAULT, DEFAULT] : (tensor<1x2xf32>, tensor<2x3xf32>) -> tensor<1x3xf32>