netns1 := p.UpsertPodCacheWithNetns(string(pod.UID), wl)
	if !reflect.DeepEqual(netns1, ns) {
		t.Fatalf("Expected the same Netns for the same uid, got %v and %v", netns1, ns)
	}

	ns2 := newFakeNsInode(inc(), 1)
	wl2 := WorkloadInfo{
		Workload: podToWorkload(pod),
		Netns:    ns2,
	}
	// when using same uid, the original netns should be returned
	netns2 := p.UpsertPodCacheWithNetns(string(pod.UID), wl2)
	if netns2 != ns {

// CHECK: return %[[CALL]]

// CHECK: quantized_dot_general_fn
// CHECK-SAME: (%[[ARG1:.+]]: tensor<1x2xf32>,  %[[ARG2:.+]]: tensor<2x3x!quant.uniform<i8<-127:127>:f32, 0.0023622048182750312>>) -> tensor<1x3xf32>
// CHECK: %[[DOT:.+]] = stablehlo.dot_general %[[ARG1]], %[[ARG2]]
// CHECK-SAME: (tensor<1x2xf32>, tensor<2x3x!quant.uniform<i8<-127:127>:f32, 0.0023622048182750312>>) -> tensor<1x3xf32>

// CHECK: func.func @init_func_restore_op
// CHECK-SAME: tf_saved_model.exported_names = ["tf_saved_model.session_initializer_restore_op"]
// CHECK-SAME: tf_saved_model.initializer_type = "restore_op"

// Test that a new save function that wraps the SaveV2 op is created.
// CHECK: func.func private @tf_quant__save(%[[ARG:.*]]: tensor<!tf_type.string>)
// CHECK: %[[VAR_HANDLE_0:.*]] = "tf.VarHandleOp"()
// CHECK-SAME: {{.*shared_name = "var_0".*}}

func.func @does_reordering_for_nested_compiles() -> () {
   // CHECK: _TPUCompileMlir
   // CHECK-SAME: Z
   // CHECK: tf_device.launch
   // CHECK-NEXT: _TPUCompileMlir
   // CHECK-SAME: X
   // CHECK: tf_device.launch
   // CHECK-NEXT: _TPUCompileMlir
   // CHECK-SAME: Y
   // CHECK: OpA
   // CHECK: OpB
   // CHECK: OpC
   "tf.OpA"() : () -> ()
   "tf_device.launch"() ({

  // CHECK: "tf.XlaHostCompute"(%arg0, %arg1)
  // CHECK-SAME: ancestors = []
  // CHECK-SAME: cost_estimate_ns = 1000000 : i64
  // CHECK-SAME: key = ""
  // CHECK-SAME: recv_key = "host_compute_channel_recv"
  // CHECK-SAME: send_key = "host_compute_channel_send"
  // CHECK-SAME: shape_inference_graph = @host_func
  // CHECK-SAME: shapes = [#tf_type.shape<*>, #tf_type.shape<3x?>]
  // CHECK-SAME: tpu_core = 0 : i64
  // CHECK: func @host_func

  // CHECK-SAME: ([[serialized]]
  // CHECK-NOT: device
  // CHECK-SAME: Tdense = [f32, f32]
  // CHECK-SAME: dense_shapes = [#corert.shape<>, #corert.shape<>]
  // CHECK-SAME: num_sparse = 2 : i64
  // CHECK-SAME: ragged_split_types = []
  // CHECK-SAME: ragged_value_types = []
  // CHECK-SAME: sparse_types = [!corert.string, i64]

  func.return %0 : tensor<i32>
}

// CHECK-LABEL: func private @merge_stateless_merged_if_0_0_then
// CHECK-SAME: ([[x:%.*]]: tensor<i32>, [[y:%.*]]: tensor<i32>)
// CHECK: [[r0:%.*]] = "tf.AddV2"([[x]], [[y]])
// CHECK: return [[r0]], [[r0]], [[r0]]

// CHECK-LABEL: func private @merge_stateless_merged_if_0_0_else
// CHECK-SAME: ([[x:%.*]]: tensor<i32>, [[y:%.*]]: tensor<i32>)

  // CHECK: %[[CONST:.*]] = mhlo.constant()
  // CHECK-SAME{LITERAL} value = dense<127> : tensor<2x3x3x2xi8>
  // CHECK-SAME: tensor<2x3x3x2x!quant.uniform<i8:f32, 1.000000e+00:3>>
  // CHECK: mhlo.convolution(%arg0, %[[CONST]])
  // CHECK-SAME{LITERAL}: dim_numbers = [b, 0, 1, f]x[0, 1, i, o]->[b, 0, 1, f]
  // CHECK-SAME{LITERAL}: window = {stride = [1, 2], pad = [[0, 0], [0, 0]], lhs_dilate = [1, 1], rhs_dilate = [2, 2]}

  // CHECK: "tf_saved_model.session_initializer"()
  // CHECK-SAME: initializers = [@init_func_restore_op]

  func.func @init_func_restore_op() -> () attributes {
      tf_saved_model.initializer_type = "restore_op",
      tf_saved_model.exported_names = ["tf_saved_model.session_initializer_restore_op"]} {
    return
  }
  // CHECK: @init_func_restore_op
  // CHECK-SAME: tf_saved_model.initializer_type = "restore_op"

  %0 = "tf_device.cluster"() ({
    // CHECK: "tf.Const"() <{value = dense<1> : tensor<i32>}>
    // CHECK-NOT: _xla_outside_compilation
    // CHECK: "tf.Const"
    // CHECK-SAME: _xla_outside_compilation
    // CHECK-SAME: tf_type.string
    // CHECK: "tf.Identity"
    // CHECK-NOT: _xla_outside_compilation
    %1 = "tf.Const"() {value = dense<1> : tensor<i32>} : () -> tensor<i32>