func.func @empty_island_multiple_data_results(%arg0: tensor<*xf32>, %arg1: tensor<*xi32>) {
  tf_executor.graph {
    %0:3 = tf_executor.island {
      // CHECK: %[[IDENTITY_N:.*]]:2 = "tf.IdentityN"
      // CHECK-SAME: (%[[ARG_0]], %[[ARG_1]])
      // CHECK: tf_executor.yield %[[IDENTITY_N]]#0, %[[IDENTITY_N]]#1
      tf_executor.yield %arg0, %arg1 : tensor<*xf32>, tensor<*xi32>
    }
    tf_executor.fetch
  }
  func.return
}

Aditya Manthramurthy <******@****.***> 1714599073 -0700

Laura Kassovic <******@****.***> 1701107622 -0800

  %val1 = "tf.Identity"(%val0) {_class = ["loc:@v"]} : (tensor<96x!tf_type.f32ref>) -> tensor<96xf32>
  func.return
}

// -----

// Test case: Two ReadVariable ops.

func.func @f() {
  // CHECK: "tf.VarHandleOp"

  // During lowering to resource variables, this pass will preserve the
  // locations of the ReadVariableOps as Identity ops to keep the original graph
  // composition and order.

     */
    String getName(T element);

    /**
     * Returns the original identity name of the given element which
     * may be used to uniquely identify the element if the name returned
     * by getName() is not unique in the hierarchy.
     *
     * @param element the element, cannot be null
     * @return the identity name of the element, never null
     */
    String getIdentityName(T element);

    /**

//
// input   min cst       max cst          input   min cst       max cst
//  \       |             |                \       |             |
//   \  (tf.Identity) (tf.Identity)   =>    \  (tf.Identity) (tf.Identity)
//    \     |             |                  \     |             |
//       tf.FakeQuantWithMinMaxVars       tf.FakeQuantWithMinMaxVars
//                   |                                 |

  auto ctrl_dep =
      NodeWith(Op("Identity"), Inputs(Out(0, switch_on_compilation_pred)));
  // Check that we pipe int32 inputs through an IdentityN to avoid extra D2H
  // copies.
  EXPECT_THAT(
      stateful_partitioned_call_op,
      NodeWith(Inputs(Out(NodeWith(Op("IdentityN"), CtrlDeps(ctrl_dep))))));
}
#endif

}  // namespace

    call_op_returns.append(call_op.getResults().begin(),
                           call_op.getResults().end());
    SetFunctionPrivate(func_op);
  }

  // Creates Identity/IdentityN ops for returing values. This allows us to
  // restore the same output tensor names in python.
  int32_t output_count = 0;
  // Map from node name to the list of the OutputInfos of its outputs that are

  //tensorflow/c/experimental/ops/gen:generate_cpp -- \
  --output_dir="${current_dir}" \
  --api_dirs="${api_dir}" \
  --source_dir=third_party/tensorflow"

${generate} \
  --category=array \
  Identity \
  IdentityN \
  ZerosLike \
  Shape \
  ExpandDims \
  OnesLike

${generate} \
  --category=math \
  Mul \
  Conj \
  AddV2 \
  MatMul \
  Neg \
  Sum \
  Sub \
  Div \
  DivNoNan \

    ```mlir
    func @my_fn(%arg0: tensor<i32>, %arg1: tensor<i32>) -> (tensor<i32>, tensor<i32>) {
      %identity = "tf.Identity"(%arg0) : (tensor<i32>) -> tensor<i32>
      %identity_n:2 = "tf.IdentityN"(%arg1, %identity) : (tensor<i32>, tensor<i32>) -> (tensor<i32>, tensor<i32>)
      return %identity, %identity_n#0 : tensor<i32>, tensor<i32>
    }
    ```
  }];