%0 = "tfl.var_handle"() {container = "", shared_name = "states"} : () -> tensor<!tf_type.resource<tensor<1x2x3xf32>>>
  %1 = "tfl.read_variable"(%0) : (tensor<!tf_type.resource<tensor<1x2x3xf32>>>) -> tensor<1x2x3xf32>
  %2 = "tfl.quantize"(%1) {qtype = tensor<1x2x3x!quant.uniform<u8<1:255>:f32, 0.047244094488188976:128>>} : (tensor<1x2x3xf32>) -> tensor<1x2x3x!quant.uniform<u8<1:255>:f32, 0.047244094488188976:128>>

}

func.func @matmulNoTransposeAOrB(%arg0: tensor<1x1280xf32>, %arg1: tensor<1280x1000xf32>) -> tensor<1x1000xf32> {
  %166 = "tf.MatMul"(%arg0, %arg1) {T = "tfdtype$DT_FLOAT", _output_shapes = ["tfshape$dim { size = 1} dim { size = 1000}"], device = "", name = "matmul", transpose_a = false, transpose_b = false} : (tensor<1x1280xf32>, tensor<1280x1000xf32>) -> tensor<1x1000xf32>
  func.return %166 : tensor<1x1000xf32>

// CHECK: return [[MUL1]]
}

func.func @batchmatmul2fullyconnected(%arg0: tensor<4x128x2xf32>) -> (tensor<4x128x1xf32>) {
  %0 = "tf.Const"() {value = dense<[[1.0], [2.0]]> : tensor<2x1xf32>} : () -> tensor<2x1xf32>
  %1 = "tf.BatchMatMulV2"(%arg0, %0) : (tensor<4x128x2xf32>, tensor<2x1xf32>) -> tensor<4x128x1xf32>
  func.return %1 : tensor<4x128x1xf32>

  // CHECK-LABEL: batchmatmul2fullyconnected

// CHECK-NEXT:    return %2 : tensor<3x2xf32>
}

func.func @transpose_3d(%arg0: tensor<1x2x3xf32>) -> tensor<3x2x1xf32> {
  %0 = "mhlo.transpose"(%arg0) <{permutation = dense<[2, 1, 0]> : tensor<3xi64>}> : (tensor<1x2x3xf32>) -> tensor<3x2x1xf32>
  func.return %0 : tensor<3x2x1xf32>

// CHECK-LABEL:   transpose_3d

// result shapes.
// CHECK-LABEL: func @testShapeInvariantWhile
func.func @testShapeInvariantWhile(%arg0: tensor<1x2x3xf32>) -> tensor<1x8x3xf32> {
  %0 = "tf.While"(%arg0) {cond = @cond, body = @body, is_stateless = false, shape_invariant} : (tensor<1x2x3xf32>) -> tensor<1x8x3xf32>
  func.return %0 : tensor<1x8x3xf32>
}

// -----
// WhileRegion tests

// Simple While region
// CHECK-LABEL: testValidWhileRegion

// CHECK-LABEL: testConvertPackToReshapeAxis0
func.func @testConvertPackToReshapeAxis0(%arg0: tensor<2x3xf32>) -> tensor<1x2x3xf32> {
  %0 = "tf.Pack"(%arg0) {axis = 0 : i64, _xla_outside_compilation = "1", device = "/job:localhost/replica:0/task:0/device:GPU:0"} : (tensor<2x3xf32>) -> tensor<1x2x3xf32>
  func.return %0 : tensor<1x2x3xf32>
  // CHECK: %[[SHAPE:.*]] = "tf.Const"() <{value = dense<[1, 2, 3]> : tensor<3xi32>}> : () -> tensor<3xi32>

}

// -----

// CHECK-LABEL: @transpose_3d_int32
func.func @transpose_3d_int32(%arg0: tensor<1x2x3xf32>) -> tensor<3x2x1xf32> {
  %permutation = "tf.Const"() {value = dense<[2, 1, 0]> : tensor<3xi32>} : () -> (tensor<3xi32>)
  // CHECK: "mhlo.transpose"
  %0 = "tf.Transpose"(%arg0, %permutation) : (tensor<1x2x3xf32>, tensor<3xi32>) -> tensor<3x2x1xf32>
  func.return %0 : tensor<3x2x1xf32>
}

// -----

}

// -----

func.func @testPReluInvalidBroadcast(%arg0: tensor<15x14x2x14xf32>, %arg1: tensor<1x1x3xf32>) -> tensor<15x14x2x14xf32> {
  // expected-error @+1 {{'tfl.prelu' op operands don't have broadcast-compatible shapes}}
  %0 = "tfl.prelu"(%arg0, %arg1) : (tensor<15x14x2x14xf32>, tensor<1x1x3xf32>) -> tensor<15x14x2x14xf32>
  func.return %0 : tensor<15x14x2x14xf32>
}

// -----