// CHECK: module attributes {tf.versions = {bad_consumers = [], min_consumer = 12 : i32, producer = 919 : i32}}
module  {
  // CHECK: tf_executor.graph
  tfg.graph #tf_type.version<producer = 919, min_consumer = 12> {

func.func @controls_per_replica() {
  tf_executor.graph {
    %1 = tf_executor.ControlTrigger {}
    %2 = tf_executor.ControlTrigger {}
    %3 = tf_executor.island(%1, %2) {
      tf_device.replicate {n = 2 : i32} {
        tf_device.return
      }
      tf_executor.yield
    }
    tf_executor.fetch
  }
  func.return
}

// CHECK: %[[CT_0:.*]] = tf_executor.ControlTrigger

// CHECK-DAG: %[[CTL_2:.*]] = tf_executor.island(%[[CTL_0]]) wraps "tf.NoOp"()
// CHECK-DAG: %[[CTL_3:.*]] = tf_executor.island(%[[CTL_1]]) wraps "tf.NoOp"()

// CHECK: tf_executor.fetch
// CHECK-SAME: !tf_executor.control, !tf_executor.control
// CHECK-NEXT: }
// CHECK-NEXT: return

// Checks that the location for the init op is properly set.
// CHECK-LOC-LABEL: func.func @main

# `tf._resource_arg_unique_id` argument attribute of the function
# @test_func_name0.

# CHECK:  func @main
# CHECK:    tf_executor.graph
# CHECK:      "tf.VarHandleOp"()
# CHECK:      "tf.LegacyCall"
# CHECK-SAME:   <{_disable_call_shape_inference = true, f = @test_func_name0}> {device = ""}
# CHECK:      tf_executor.fetch
# CHECK:    return
# CHECK:  func private @test_func_name0
# CHECK-SAME:   tf._resource_arg_unique_id = 0

```
### `-tf-executor-to-functional-conversion`

_Lifts tf_executor.island inner ops from a tf_executor.graph_

This pass converts tf_executor.graphs consisting of only tf_executor.islands and
a tf_executor.fetch into a sea of nodes consisting of TensorFlow Dialect ops by
lifting such ops out of a tf_executor.graph's tf_executor.islands. If V1 control
flow ops are present in a tf_executor.graph, an error will be returned.

func.func @nested_entry_functions() attributes {tf.entry_function = {}} {
  tf_executor.graph {
     %control = tf_executor.island wraps "tf.StatefulPartitionedCall"() {config = "", config_proto = "", device = "/device:CPU:0", executor_type = "", f = @func} : () -> ()
     tf_executor.fetch
  }
  func.return
}

func.func @func() attributes {tf.entry_function = {}} {
  func.return
}

// tf_executor.island {_parallel_execution_ids=r0:0,p0:0} wraps "tf.OpC"
// tf_executor.island {_parallel_execution_ids=r0:0,p0:1} wraps "tf.OpD"
// tf_executor.island {_parallel_execution_ids=r0:1} wraps "tf.OpB"
// tf_executor.island {_parallel_execution_ids=r0:1,p0:0} wraps "tf.OpC"
// tf_executor.island {_parallel_execution_ids=r0:1,p0:1} wraps "tf.OpD"
//
// "tf.OpA" will not have `_parallel_execution_ids` attr,

    %outputs_4, %control_4 = tf_executor.island wraps "tf.Identity"(%outputs_3) {T = "tfdtype$DT_FLOAT", device = "", name = "output_0_shard_0"} : (tensor<0xf32>) -> tensor<0xf32>
    %control_5 = tf_executor.island(%control, %control_4) wraps "tf.NoOp"() : () -> ()
    tf_executor.fetch %control_5 : !tf_executor.control
  }
  func.return
}

// -----

// CHECK-LABEL: func.func @convert_xla_must_compile

        %0 = tf_executor.graph {
          %outputs, %control = tf_executor.island wraps "tf.XlaSharding"(%arg0) {_XlaSharding = "\08\03\1A\02\02\01\22\02\00\01", sharding = "\08\03\1A\02\02\01\22\02\00\01"} : (tensor<2x2xf32>) -> tensor<2x2xf32>

      tf_executor.yield
    }
    tf_executor.fetch %island : !tf_executor.control
  }
  func.return
}

// -----
// Tests handling of Cast, Identity, and IdentityN

!tf_res = tensor<*x!tf_type.resource<tensor<32xf32>>>
// CHECK-LABEL: func @aliasing_identities
func.func @aliasing_identities(%arg0: tensor<32xf32>) -> () {
  tf_executor.graph {
    // CHECK: tf_executor.island