// CHECK-NEXT: %[[ADDN2:.*]] = "tf.AddN"(%[[ADDN]], %[[ZEROS_LIKE]]) : (tensor<10xf32>, tensor<10xf32>) -> tensor<10xf32>
  %addn2 = "tf.AddN"(%addn, %zeros-like) : (tensor<!tf_type.variant<tensor<f32>>>, tensor<!tf_type.variant<tensor<f32>>>) -> tensor<!tf_type.variant<tensor<f32>>>
  %stack = "tf.TensorListStack"(%addn2, %elem_shape) : (tensor<!tf_type.variant<tensor<f32>>>, tensor<0xi32>) -> tensor<10xf32>

// CHECK:    %[[ADD3:.*]], %[[ADD3_control:.*]] = tf_executor.island(%[[ISLAND1]], %[[ADD2_control]]) wraps "tf.Add"(%[[ADD2]], %[[ADD2]])
// CHECK:    %[[PRINT2:.*]], %[[PRINT2_control:.*]] = tf_executor.island wraps "tf.Print"(%[[ADD3]]) <{message = "add result"}>
// CHECK:    tf_executor.fetch %[[ADD2]], %[[MUL]], %[[PRINT1_control]], %[[PRINT2_control:.*]] :
// CHECK:  }

  // CHECK-NOT: "tf.ReadVariableOp"
  // CHECK: %[[CONST:.*]] = "tf.Const"() <{value = dense<4.200000e+01> : tensor<f32>}>
  // CHECK: %[[ADD1:[0-9]*]] = "tf.AddV2"(%arg1, %[[CONST]])
  // CHECK: %[[ADD2:[0-9]*]] = "tf.AddV2"(%[[ADD1]], %arg1)
  // CHECK: %[[PACK:[0-9]*]] = "tf.Pack"(%[[CONST]], %[[ADD2]])
  // CHECK: return %[[PACK]]

  %0 = "tf.Const"() {value = dense<4.200000e+01> : tensor<f32>} : () -> tensor<f32>

// CHECK:       tf.WhileRegion
// CHECK:       ^bb0
// CHECK:       tf.OpA
// CHECK:       ^bb0([[ARG_2:%[a-zA-Z0-9_]+]]
// CHECK-SAME:  [[ARG_3:%[a-zA-Z0-9_]+]]: tensor<i32>)
// CHECK-NEXT:  [[RES_3:%.*]] = "tf.AddV2"([[ARG_2]], [[RES_1]])
// CHECK-NEXT:  [[RES_4:%.*]] = "tf.Div"([[ARG_3]], [[RES_2]])
// CHECK:       "tf.Yield"([[RES_3]], [[RES_4]])

  // CHECK-NEXT: tf_mlrt.executeop([[y]]
  %w = "tf.AddV2"(%y, %z) {__op_key = 2: i32}: (tensor<i32>, tensor<i32>) -> tensor<i32>
  // CHECK-NEXT: tf_mlrt.executeop([[y]]
  %u = "tf.AddV2"(%y, %z) {__op_key = 3: i32} : (tensor<i32>, tensor<i32>) -> tensor<i32>
  // CHECK-NEXT: tf_mlrt.executeop([[y]]
  %v = "tf.AddV2"(%y, %z) {__op_key = 4: i32}: (tensor<i32>, tensor<i32>) -> tensor<i32>

  // CHECK: [[handle:%.*]] = "tf.VarHandleOp"()
  // CHECK: "tf.ReadVariableOp"([[handle]])
  %0 = "tf.ReadVariableOp"(%arg1) {device = "/device:CPU:0"} : (tensor<*x!tf_type.resource>) -> tensor<1x3xf32>
  %1 = "tf.AddV2"(%arg0, %0) {device = "/device:CPU:0"} : (tensor<1x3xf32>, tensor<1x3xf32>) -> tensor<1x3xf32>
  %2 = "tf.Identity"(%1) {device = "/device:CPU:0"} : (tensor<1x3xf32>) -> tensor<1x3xf32>
  func.return %2 : tensor<1x3xf32>
}

		map1, map2 = fake, fake
	}
	return
}

func TestConvertCPUProfile(t *testing.T) {
	addr1, addr2, map1, map2 := testPCs(t)

	b := []uint64{
		3, 0, 500, // hz = 500
		5, 0, 10, uint64(addr1 + 1), uint64(addr1 + 2), // 10 samples in addr1
		5, 0, 40, uint64(addr2 + 1), uint64(addr2 + 2), // 40 samples in addr2
		5, 0, 10, uint64(addr1 + 1), uint64(addr1 + 2), // 10 samples in addr1
	}

  // CHECK-NEXT: "tf.AddV2"({{.*}}, [[v]]) {device = "/CPU:0"} : (tensor<i32>, tensor<i32>) -> tensor<i32>
  // CHECK-NEXT: return
  %const = "tf.Const"() {device = "/CPU:0", value = dense<0> : tensor<i32>} : () -> tensor<i32>
  %x = "tf.AddV2"(%const, %const) {device = "/CPU:0"} : (tensor<i32>, tensor<i32>) -> tensor<i32>
  %r = "tf.AddV2"(%arg, %x) {device = "/CPU:0"} : (tensor<i32>, tensor<i32>) -> tensor<i32>

// CHECK: %[[mul:.*]] = "tf.Mul"(%arg0, %[[CONST_0]]) : (tensor<*xf32>, tensor<2xf32>) -> tensor<*xf32>
// CHECK: %[[add:.*]] = "tf.AddV2"(%[[mul]], %[[CONST]]) : (tensor<*xf32>, tensor<2xf32>) -> tensor<*xf32>
// CHECK-NEXT: return %[[add]] : tensor<*xf32>

// -----

func.func @not_decompose_batch_norm(%arg0: tensor<*xf32>) -> (tensor<*xf32>) {

// CHECK-LABEL: func @otherVariantWhileBody
// CHECK:       [[CST:%.*]] = "tf.Const"()
// CHECK-NEXT:  [[ADD:%.*]] = "tf.AddV2"(%arg2, [[CST]])
// CHECK-NEXT:  [[TENSOR_MAP_INSERT_RESULT:%.*]] = "tf.TensorMapInsert"(%arg3, %arg2, %arg2)
// CHECK-NEXT:  [[ADD_2:%.*]] = "tf.AddV2"(%arg0, [[CST]])
// CHECK-NEXT:  return [[ADD_2]], %arg1, [[ADD]], [[TENSOR_MAP_INSERT_RESULT]]