// CHECK: TPUCompile
// CHECK: TPUExecute
// CHECK-NOT: func @_func
    tf_executor.graph {
      %outputs, %control = tf_executor.island wraps "arith.constant"() {value = dense<2.000000e+00> : tensor<f32>} : () -> tensor<f32>
      %outputs_0, %control_1 = tf_executor.island wraps "arith.constant"() {value = dense<3.000000e+00> : tensor<f32>} : () -> tensor<f32>

func.func @multiple_func_body_ops() {
  tf_executor.graph {
    tf_executor.island {
      "tf.NoOp"() : () -> ()
      tf_executor.yield
    }
    tf_executor.fetch
  }
  tf_executor.graph {
    tf_executor.island {
      "tf.NoOp"() : () -> ()
      tf_executor.yield
    }
    tf_executor.fetch
  }
  func.return
}

// -----

}

// -----

// Check that tf_executor.graph can't be nested directly in a tf_executor.graph.
func.func @nested_graph() {
  tf_executor.graph {
    tf_executor.graph {}
// expected-error@-1 {{'tf_executor.graph' op unallowed directly inside another tf_executor.graph}}
  }
  func.return
}

// -----

// Check that a tf_executor.fetch is terminating a tf_executor.graph (custom parser)

      %3 = "tf_device.launch_func"(%2) {device = "/device:test_device:0", func = @_func} : (tensor<?xf32>) -> tensor<?xf32>

      // CHECK: tf_executor.yield %[[CALL_OUTPUT]]
      tf_executor.yield %3 : tensor<?xf32>
    }
    tf_executor.fetch %1#0 : tensor<?xf32>
  }
  func.return %0 : tensor<?xf32>
}

func.func @_func(%arg0: tensor<?xf32>) -> tensor<?xf32> {
  func.return %arg0 : tensor<?xf32>
}

#include "tensorflow/compiler/mlir/tensorflow/ir/tf_executor.h"

#define DEBUG_TYPE "tf-functional-to-executor"

namespace mlir {

namespace {
// This pass converts mlir functions consisting of mlir ops into a tf_executor
// dialect as a single island.
// Result like so:
//   func @my_fn(%argi...) -> (result_t) {
//     %results:[[n_args]] = tf_executor.graph {

// CHECK-LABEL: func @constant_in_island
func.func @constant_in_island(%arg0 : tensor<i1>) -> tensor<f32> {
  %0 = tf_executor.graph {
// CHECK: tf_executor.island
// CHECK: tf.Const{{.*}}2.0
    %1:2 = tf_executor.island {
      %0 = "tf.Const"() {value = dense<2.000000e+00> : tensor<f32>} : () -> tensor<f32>
      tf_executor.yield %0 : tensor<f32>
    }
// Uses two islands for no other reason than preventing canonicalization from

        tf_device.return %4 : tensor<?xi32>
      }) {device = "/device:test_device:0"} : () -> tensor<?xi32>

      // CHECK: tf_executor.yield %[[LAUNCH_OUTPUT]]
      tf_executor.yield %3 : tensor<?xi32>
    }
    tf_executor.fetch %1#0 : tensor<?xi32>
  }
  func.return %0 : tensor<?xi32>
}

// CHECK: func private @[[LAUNCH]]

// CHECK:          }
// CHECK-NEXT:   }
func.func @bar() {
  tf_executor.graph {
    %0:2 = tf_executor.island wraps "tf.Const"() {dtype = "tfdtype$DT_INT32", name = "tf.Const", value = dense<1> : tensor<i32>} : () -> tensor<i32>
    %1:2 = tf_executor.island wraps "tf.Empty"(%0#0) {dtype = "tfdtype$DT_FLOAT", name = "tf.Empty"} : (tensor<i32>) -> tensor<*xf32>
    tf_executor.fetch
  }
  func.return
}

// CHECK:        function {

    %1:2 = tf_executor.island {
      %3 = "tf.opA"(%arg0) : (tensor<i1>) -> tensor<i1>
      tf_executor.yield %3 : tensor<i1>
    }
    %2:2 = tf_executor.island(%1#1) {
      %4 = "tf.opB"() : () -> tensor<f32>
      tf_executor.yield %4 : tensor<f32>
    }

// CHECK: "tf.opA"
// CHECK: tf_executor.island
// CHECK: "tf.opB"

    tf_executor.fetch %2#0 : tensor<f32>
  }
  func.return %0 : tensor<f32>
}

    %0 = tf_executor.graph {
      %1, %c0 = tf_executor.island wraps "tf.Const"() {value = dense<1> : tensor<i1>} : () -> tensor<i1>
      tf_executor.fetch %1 : tensor<i1>
    }
    return %0 : tensor<i1>
  }

  func.func private @while_body(%arg0: tensor<i32>, %arg1: tensor<i32>) -> (tensor<i32>, tensor<i32>) {
    tf_executor.graph {