%cst = stablehlo.constant dense<1.0> : tensor<1x3xf32>
      %0 = stablehlo.dot_general %arg0, %arg1, contracting_dims = [1] x [0], precision = [] : (tensor<1x1024xf32>, tensor<1024x3xf32>) -> tensor<1x3xf32>
      %1 = stablehlo.subtract %cst, %0 : tensor<1x3xf32>
      %2 = stablehlo.add %0, %cst : tensor<1x3xf32>
      return %2 : tensor<1x3xf32>
    }
  }
)mlir";

// 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>

      %1 = "quantfork.qcast"(%0) {volatile} : (tensor<1x3xf32>) -> tensor<1x3x!quant.uniform<i8:f32, 0.13170163023705575:-1>>
      %2 = "quantfork.dcast"(%1) : (tensor<1x3x!quant.uniform<i8:f32, 0.13170163023705575:-1>>) -> tensor<1x3xf32>
      %3 = stablehlo.reshape %2 : (tensor<1x3xf32>) -> tensor<3x1xf32>

  // 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>
}

func.func @no_change_to_dag(%arg0: tensor<!tf_type.resource<tensor<10x3xf32>>>, %arg1: tensor<!tf_type.resource<tensor<10x3xf32>>>, %arg2: tensor<!tf_type.resource<tensor<10x3xf32>>>, %arg3: tensor<!tf_type.resource<tensor<10x3xf32>>>) -> (tensor<!tf_type.resource<tensor<10x3xf32>>>, tensor<!tf_type.resource<tensor<10x3xf32>>>, tensor<!tf_type.resource...

    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);

  %8 = "tfl.concatenation"(%2, %0) {axis = -1 : i32, fused_activation_function = "NONE"} : (tensor<1x1xf32>, tensor<1x1xf32>) -> tensor<1x2xf32>
  %9 = "quantfork.stats"(%8) {layerStats = dense<[-0.488159984, 0.189515018]> : tensor<2xf32>} : (tensor<1x2xf32>) -> tensor<1x2xf32>
  %10 = "tfl.concatenation"(%9, %7) {axis = -1 : i32, fused_activation_function = "NONE"} : (tensor<1x2xf32>, tensor<1x2xf32>) -> tensor<1x4xf32>

// CHECK-SAME: ([[ARG0:%.*]]: tensor<!tf_type.resource<tensor<10x3xf32>>>, [[ARG1:%.*]]: tensor<!tf_type.resource<tensor<10x3xf32>>>, [[ARG2:%.*]]: tensor<!tf_type.resource<tensor<10x3xf32>>>, [[ARG3:%.*]]: tensor<!tf_type.resource<tensor<10x3xf32>>>)

func.func @pushTposeAfterAddSimpleWithFold(%arg0: tensor<2x3xi32>) -> tensor<3x2xi32> {
  %perm = arith.constant dense<[1, 0]> : tensor<2xi32>
  %0 = "tfl.transpose"(%arg0, %perm) : (tensor<2x3xi32>, tensor<2xi32>) -> tensor<3x2xi32>
  %cst = arith.constant dense<[[1, 2], [3, 4], [5, 6]]> : tensor<3x2xi32>
  %1 = tfl.add %0, %cst { fused_activation_function = "NONE" } : tensor<3x2xi32>
  func.return %1 : tensor<3x2xi32>
}