%2 = "quantfork.stats"(%1) {layerStats = dense<[5.00000000e-6, 7.00000000e-1]> : tensor<2xf32>} : (tensor<1x3xf32>) -> tensor<1x3xf32>
    return %2 : tensor<1x3xf32>
  }
// CHECK: func.func private @quantize_dot_general_with_bias_same_shape_fn(%[[ARG_0:.+]]: tensor<1x2xf32>) -> tensor<1x3xf32> attributes {tf._original_func_name = "main_0"}

// CHECK-LABEL: func @basic
func.func @basic(
    %arg0: tensor<3x1xf32>,
    %arg1: tensor<!tf_type.resource<tensor<1x3xf32>>>) -> (tensor<3x3xf32>) {
  %1 = "tf.ReadVariableOp"(%arg1) {_output_shapes = ["tfshape$dim { size: 1 } dim { size: 3 }"], device = "/device:CPU:0", dtype = f32} : (tensor<!tf_type.resource<tensor<1x3xf32>>>) -> tensor<1x3xf32>

  // CHECK: {{%.*}} = tfrt_fallback_async.executeop {{.*}} device("/device:CPU:0") "tf.MatMul"

    %arg1: tensor<1x3xf32> {tf_saved_model.index_path = [0]})
      -> (tensor<3x3xf32> {tf_saved_model.index_path = []}) {
  // CHECK: {{%.*}} = corert.get_op_handler %arg0 "/device:GPU:0"
  %2 = "tf.MatMul"(%arg0, %arg1) {T = f32, _output_shapes = ["tfshape$dim { size: 3 } dim { size: 3 }"], device = "/device:GPU:0", transpose_a = false, transpose_b = false} : (tensor<3x1xf32>, tensor<1x3xf32>) -> tensor<3x3xf32>

  // CHECK-NEXT: [[out_ch:%.*]] = tfrt.merge.chains [[ch]], [[ch1]] : !tfrt.chain, !tfrt.chain
  // CHECK-NEXT: tfrt.return [[out_ch]], [[r2]] : !tfrt.chain, !tfrt_fallback.tf_tensor
  func.return %6#0 : tensor<3x3xf32>
}

    module attributes {} {
      func.func @main(%arg0: tensor<1x4xf32>) -> tensor<1x3xf32> attributes {} {
        %0 = stablehlo.constant dense<1.0> : tensor<4x3xf32>
        %1 = stablehlo.dot_general %arg0, %0, contracting_dims = [1] x [0], precision = [DEFAULT, DEFAULT] : (tensor<1x4xf32>, tensor<4x3xf32>) -> tensor<1x3xf32>
        return %1 : tensor<1x3xf32>
      }
    }
  )mlir");
  ASSERT_TRUE(module_op);

  func.return %0 : tensor<1024xf32>

  // 1.0 * 2.0 * 512 + 4.0 = 1028.0
  // CHECK: %[[CST:.*]] = arith.constant dense<1.028000e+03> : tensor<1024xf32>
  // CHECK:  return %[[CST]]
}

// CHECK-LABEL: @ConstantFoldFullyConnectedNoBias
func.func @ConstantFoldFullyConnectedNoBias() -> tensor<1024xf32> {

  // CHECK: return %[[RESULT]]
  func.return %0: tensor<10x2xf32>
}

// -----

func.func @select_first(%arg0: tensor<10x2xf32>, %arg1: tensor<10x2xf32>) -> tensor<10x2xf32> {
  func.return %arg0: tensor<10x2xf32>
}

// CHECK-LABEL: testMultiInputLegacyCallOp
func.func @testMultiInputLegacyCallOp(%arg0: tensor<10x2xf32>, %arg1: tensor<10x2xf32>) -> tensor<10x2xf32> {

    %1 = "tf.PartitionedCall"(%arg0, %cst_0) {_tfl_quant_trait = "fully_quantizable", config = "", config_proto = "", executor_type = "", f = @composite_matmul_fn_1} : (tensor<2x12xf32>, tensor<12x2xf32>) -> tensor<*xf32>
    func.return %1: tensor<*xf32>
  }

func.func @float_matmul(
  %arg0: tensor<1x10xf32>, %arg1: tensor<10x10xf32>) -> (tensor<*xf32>, tensor<*xf32>, tensor<*xf32>) {
  %cst = "tf.Const"() {value = dense<0.000000e+00> : tensor<10xf32>} : () -> tensor<10xf32>
  %0 = "tf.MatMul"(%arg0, %arg1) {
    transpose_a = false, transpose_b = false
  } : (tensor<1x10xf32>, tensor<10x10xf32>) -> tensor<*xf32>

  %prelu = "tfl.prelu"(%arg0, %cst) : (tensor<1x10x10x3xf32>, tensor<1x1x3xf32>) -> tensor<1x10x10x3xf32>
  func.return %prelu : tensor<1x10x10x3xf32>

// CHECK: %[[cst:.*]] = arith.constant dense<[{{\[}}[1.66394591, 3.61694336, 2.0382936]]]> : tensor<1x1x3xf32>