b[unzig[zig]] *= qt[zig]
	}
	idct(b)
	dst, stride := []byte(nil), 0
	if d.nComp == 1 {
		dst, stride = d.img1.Pix[8*(by*d.img1.Stride+bx):], d.img1.Stride
	} else {
		switch compIndex {
		case 0:
			dst, stride = d.img3.Y[8*(by*d.img3.YStride+bx):], d.img3.YStride
		case 1:
			dst, stride = d.img3.Cb[8*(by*d.img3.CStride+bx):], d.img3.CStride
		case 2:
			dst, stride = d.img3.Cr[8*(by*d.img3.CStride+bx):], d.img3.CStride

  %rst = "tf.Conv2D"(%arg, %fq) {T = "tfdtype$DT_FLOAT", data_format = "NHWC", dilations = [1, 2, 3, 1], padding = "SAME", strides = [1, 4, 5, 1]} : (tensor<256x32x32x3xf32>, tensor<3x3x3x16xf32>) -> tensor<256x8x7x16xf32>
  func.return %rst : tensor<256x8x7x16xf32>

  %cst2 = arith.constant dense<[1.0, 2.0]> : tensor<2xf32>
  %0 = "tf.Conv2D"(%arg0, %cst0) {T = "tfdtype$DT_FLOAT", data_format = "NHWC", dilations = [1, 2, 3, 1], padding = "SAME", strides = [1, 4, 5, 1]} : (tensor<1x112x112x3xf32>, tensor<1x3x3x2xf32>) -> tensor<1x28x23x2xf32>
  %1 = "tf.Mul"(%0, %cst2) : (tensor<1x28x23x2xf32>, tensor<2xf32>) -> tensor<1x28x23x2xf32>

  func.return %1 : tensor<1x28x23x2xf32>

  %cst_2 = "tf.Const"() {value = dense<[0.3, 0.4]> : tensor<2xf32>} : () -> tensor<2xf32>
  %0 = "tf.Conv2D"(%arg_0, %cst_0) {data_format = "NHWC", dilations = [1, 1, 2, 1], explicit_paddings = [], padding = "SAME", strides = [1, 1, 2, 1], use_cudnn_on_gpu = true} : (tensor<1x3x4x3xf32>, tensor<2x3x3x2xf32>) -> tensor<1x3x2x2xf32>

  fields.emplace_back(pool_size_id, pool_size);

  auto strides_id = ::mlir::StringAttr::get(context, "strides");
  fields.emplace_back(strides_id, strides);

  DictionaryAttr dict = DictionaryAttr::get(context, fields);
  return TF::FuncAttr::get(context, "MaxUnpooling2D", dict);
}

}  // namespace

class PerceptionUtilsTest : public ::testing::Test {
 protected:
  PerceptionUtilsTest() {}

    emitError(loc,
              "zero point is expected to be a constant with a single value");
    return {};
  }
  if (strides.size() != num_dims || dilations.size() != num_dims) {
    emitError(loc,
              absl::StrFormat(
                  "strides and dilations are expected to be %d-element arrays",
                  num_dims));
    return {};
  }

  xla::ConvolutionDimensionNumbers dnums;

		}
	}
}

func TestDumper(t *testing.T) {
	var in [40]byte
	for i := range in {
		in[i] = byte(i + 30)
	}

	for stride := 1; stride < len(in); stride++ {
		var out bytes.Buffer
		dumper := Dumper(&out)
		done := 0
		for done < len(in) {
			todo := done + stride
			if todo > len(in) {
				todo = len(in)
			}
			dumper.Write(in[done:todo])
			done = todo
		}

		dumper.Close()

module {
func.func @main(%arg0: tensor<4x68x68x3xf32>, %arg1: tensor<5x5x3x8xf32>) -> tensor<4x64x64x8xf32> {
  %0 = "tf.Conv2D"(%arg0, %arg1) {padding = "VALID", strides = [1, 1, 1, 1]} : (tensor<4x68x68x3xf32>, tensor<5x5x3x8xf32>) -> tensor<4x64x64x8xf32>
  func.return %0 : tensor<4x64x64x8xf32>
}

    %2 = "tf.Conv2D"(%1, %0) {attr_map = "0:strides,1:use_cudnn_on_gpu,2:padding,3:explicit_paddings,4:dilations", data_format = "NHWC", device = "", dilations = [1, 1, 1, 1], explicit_paddings = [], padding = "SAME", strides = [1, 1, 2, 1], use_cudnn_on_gpu = true} : (tensor<1x3x4x3xf32>, tensor<2x3x3x2xf32>) -> tensor<*xf32>

  %bias = arith.constant dense<3.0> : tensor<16xf32>
  %value = arith.constant dense<4.0> : tensor<16xf32>
  %0 = "tf.Conv2D"(%arg, %filter) {T = "tfdtype$DT_FLOAT", data_format = "NHWC", dilations = [1, 2, 3, 1], padding = "SAME", strides = [1, 4, 5, 1]} : (tensor<256x32x32x3xf32>, tensor<3x3x3x16xf32>) -> tensor<256x8x7x16xf32>
  %1 = "tf.BiasAdd"(%0, %bias) {T = "tfdtype$DT_FLOAT", data_format = "NHWC"}: (tensor<256x8x7x16xf32>, tensor<16xf32>) -> tensor<256x8x7x16xf32>