// CHECK-NEXT:   value {
// CHECK-NEXT:     list {
// CHECK-NEXT:     }
// CHECK-NEXT:   }
// CHECK-NEXT: }

module attributes {tf.versions = {bad_consumers = [], min_consumer = 0 : i32, producer = 413 : i32}} {
  func.func @main() -> tensor<*xi64> attributes {tf.entry_function = {control_outputs = "", inputs = "", outputs = "result"}} {
    %0 = tf_executor.graph {

// RUN: tf-mlir-translate -mlir-to-graphdef %s -o - | FileCheck %s

module attributes {tf.versions = {producer = 946 : i32}, tf_saved_model.semantics, tfl.description = "MLIR Converted.", tfl.schema_version = 3 : i32}  {

module attributes {tf.versions = {bad_consumers = [], min_consumer = 0 : i32, producer = 268 : i32}} {
  func.func @main() -> () {
    func.return
  }

  %2 = "tfl.add"(%arg0, %arg3) {fused_activation_function = "RELU6"} : (tensor<1xf32>, tensor<1xf32>) -> tensor<1xf32>
  %3 = "tfl.pack"(%1, %2) {axis = 0 : i32, values_count = 2 : i32} : (tensor<1xf32>, tensor<1xf32>) -> tensor<2x1xf32>
  func.return %3 : tensor<2x1xf32>
}

// CHECK: %[[CST:.*]] = arith.constant dense<1> : tensor<4xi32>

  %cst = "tf.Const"() {value = dense<3> : tensor<i32>} : () -> tensor<i32>
  %cst_1 = "tf.Const"() {value = dense<[2, 0, 1]> : tensor<3xi32>} : () -> tensor<3xi32>
  %cst_2 = "tf.Const"() {value = dense<[512, 64]> : tensor<2xi32>} : () -> tensor<2xi32>
  %0 = "tf.ExpandDims"(%arg0, %cst) {device = ""} : (tensor<8x64xf32>, tensor<i32>) -> tensor<8x64x1xf32>

};

TEST_F(VerifyClusteringPassTest, OnlyTfFunctionalPasses) {
  static constexpr char kMlirModuleStr[] = R"(
  module attributes {tf.versions = {bad_consumers = [], min_consumer = 0 : i32, producer = 268 : i32}} {
    func.func @main() -> tensor<1xi32> {
      %0 = "tf.Const"() {value = dense<1000> : tensor<1xi32>} : () -> tensor<1xi32>
      return %0 : tensor<1xi32>
    }
  })";
  CreateModule(kMlirModuleStr);

module attributes {tf.versions = {bad_consumers = [], min_consumer = 0 : i32, producer = 268 : i32}} {
  func.func @main() {
    tf_executor.graph {
      %control = tf_executor.island {
        tf_executor.yield
      }
      tf_executor.fetch %control : !tf_executor.control
    }
    tf_executor.graph {
      %control = tf_executor.island {
        tf_executor.yield
      }
      tf_executor.fetch %control : !tf_executor.control
    }

    return %4: tensor<*xi32>
  }
  func.func @__initializer(%arg0: tensor<*x!tf_type.string>) -> tensor<i32> {
    %0 = "tf.Const"() <{value = dense<1> : tensor<i32>}> : () -> tensor<i32>
    return %0 : tensor<i32>
  }
}


// -----
//  Decomposable Resource Ops usage
//
// CHECK-LABEL:  func.func @serving_default
// CHECK:      "tf.VarHandleOp"

};

TEST_F(InferenceMetricsPassTest, RecordsTrueForTPUPartitionedCallOp) {
  static constexpr char kMlirModuleStr[] = R"(
  module attributes {tf.versions = {bad_consumers = [], min_consumer = 0 : i32, producer = 268 : i32}} {
    func.func @tpu_partitioned_call_func(%arg0: tensor<?xi32>) -> (tensor<?xi32>) {
      func.return %arg0 : tensor<?xi32>
    }

module attributes {tf.versions = {bad_consumers = [], min_consumer = 0 : i32, producer = 268 : i32}} {
  func.func @main() {
    tf_executor.graph {
      %control = tf_executor.island {
        tf_executor.yield
      }
      tf_executor.fetch %control : !tf_executor.control
    }
    tf_executor.graph {
      %control = tf_executor.island {
        tf_executor.yield
      }
      tf_executor.fetch %control : !tf_executor.control
    }