}) {is_stateless = false} : (tensor<i32>, tensor<1xf32>) -> (tensor<i32>, tensor<1xf32>)
  func.return %0#1 : tensor<1xf32>
}

func.func @cond(%arg0: tensor<*xi32>, %arg1: tensor<*xf32>) -> tensor<i1> {
  %cst = arith.constant dense<0> : tensor<i32> loc("Const")
  %0 = "tfl.greater"(%arg0, %cst) : (tensor<*xi32>, tensor<i32>) -> tensor<i1>
  func.return %0 : tensor<i1>
}

  %3 = "tfl.depthwise_conv_2d"(%arg0, %2, %0) {depth_multiplier = 4 : i32, dilation_h_factor = 2 : i32, dilation_w_factor = 2 : i32, fused_activation_function = "NONE", padding = "VALID", stride_h = 4 : i32, stride_w = 5 : i32} : (tensor<1x224x224x3xf32>, tensor<32x3x3x3xf32>, tensor<32xf32>) -> tensor<1x112x112x32xf32>

  func.return %3 : tensor<1x112x112x32xf32>

  // CHECK:      %[[READ:.*]] = "tf.ReadVariableOp"
  // CHECK-NEXT: tf_device.replicate
  // CHECK-NEXT:   "tf.OpA"(%[[READ]])
  tf_device.replicate {n = 2 : i32} {
    %0 = "tf.ReadVariableOp"(%arg0) : (tensor<*x!tf_type.resource<tensor<f32>>>) -> tensor<f32>
    "tf.OpA"(%0) : (tensor<f32>) -> ()
  }
  func.return
}

// -----

const maxTries = 10000 // Up to 100ms per test.

func main() {
	var i32 int32
	var i64 int64
	var b bool
	var s string

	var sync = make(chan bool)

	for buffer := 0; buffer < 2; buffer++ {
		c32 := make(chan int32, buffer)
		c64 := make(chan int64, buffer)
		cb := make(chan bool, buffer)
		cs := make(chan string, buffer)

		select {
		case i32 = <-c32:
			panic("blocked i32sender")
		default:
		}

module attributes {tf.versions = {bad_consumers = [], min_consumer = 0 : i32, producer = 268 : i32}} {
  func.func @main() -> () {
     %s0 = "tf.Const"() {value = dense<[4, 1]> : tensor<2xi32>} : () -> tensor<2xi32>
    func.return
  }

func.func @serving_default(%arg0: tensor<1x3xf32>) -> (tensor<1x3xf32>) {
  %outputs  =  "tf.TPUCompilationResult"() {_tpu_compilation_status = "cluster", device = ""} : () -> tensor<!tf_type.string>

// RUN: stablehlo-quant-opt %s -split-input-file -stablehlo-restore-function-name | FileCheck %s

module attributes {tf.versions = {bad_consumers = [], min_consumer = 12 : i32, producer = 1646 : i32}, tf_saved_model.semantics} {
  // CHECK-LABEL: @serving_default
  // CHECK-SAME: %[[ARG0:[^:[:space:]]+]]
  // CHECK-SAME: %[[ARG1:[^:[:space:]]+]]
  func.func private @serving_default(%arg0: tensor<1x4xf32>, %arg1: tensor<4x3xf32>) -> tensor<1x3xf32> {

module attributes {tf.versions = {producer = 888 : i32}} {
  // CHECK-LABEL: func @annotate_broadcast_values
  func.func @annotate_broadcast_values(%arg0: tensor<?xi32>) -> tensor<?xi32> {
    %0 = "tf._A"(%arg0) : (tensor<?xi32>) -> tensor<?xi32>
    %1 = "tf._B"(%arg0) : (tensor<?xi32>) -> tensor<?xi32>
    %5:2 = tf_device.replicate([%0, %arg0] as %ri_0: tensor<?xi32>) {n = 2 : i32} {
      %2 = "tf._F"(%arg0) : (tensor<?xi32>) -> tensor<?xi32>

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

// RUN: flatbuffer_translate -mlir-to-tflite-flatbuffer %s -o - | flatbuffer_to_string - | FileCheck %s

func.func @main(tensor<8xi32>, tensor<8xi32>, tensor<i32>) -> tensor<8xi32> {
^bb0(%arg0: tensor<8xi32>, %arg1: tensor<8xi32>, %arg2: tensor<i32>):
// CHECK: {
// CHECK-NEXT:   version: 3,
// CHECK-NEXT:   operator_codes: [ {
// CHECK-NEXT:     deprecated_builtin_code: 127,
// CHECK-NEXT:     version: 1,
// CHECK-NEXT:     builtin_code: UNSORTED_SEGMENT_PROD