// RUN: tf-mlir-translate -mlir-tf-to-hlo-text %s -tf-input-shapes=8,16,16,64:64 -tf-xla-emit-use-tuple-args -tf-xla-emit-return-tuple | FileCheck %s
// RUN: tf-mlir-translate -mlir-tf-to-hlo-text-via-builder %s -tf-input-shapes=8,16,16,64:64 | FileCheck %s

module attributes {tf.versions = {producer = 179 : i32}} {

# RUN: tf-mlir-translate -graphdef-to-mlir -tf-enable-shape-inference-on-import=false %s -o - | FileCheck %s

# Verify that the _input_shapes attribute of the FunctionDef is respected.
# This also checks that the output type is correctly inferred based on
# that.
#CHECK: func private @identity_function0(%arg0: tensor<i32>) -> tensor<i32>

node {
  name: "Placeholder"
  op: "Placeholder"
  attr {
    key: "dtype"
    value {
      type: DT_BOOL

Keith Randall <******@****.***> 1627678827 -0700

// Verifies that constants generated post shape inference are propagated.
// get_shape result in this test.
module attributes {tf.versions = {producer = 179 : i32}} {

  // CHECK-LABEL: func @main
  func.func @main(%arg0: tensor<10x19xf32>, %arg1: tensor<19x10xf32> {mhlo.is_same_data_across_replicas = true}) -> tensor<?xi64> {
    %0 = "tf.Shape"(%arg0) : (tensor<10x19xf32>) -> tensor<2xi64>

    llvm::ArrayRef<int64_t> shapes) {
  return llvm::to_vector(llvm::map_range(shapes, [](int64_t shape) {
    return shape == kTFDynamicSize ? mlir::ShapedType::kDynamic : shape;
  }));
}

llvm::SmallVector<int64_t> ConvertMlirShapeToTF(
    llvm::ArrayRef<int64_t> shapes) {
  return llvm::to_vector(llvm::map_range(shapes, [](int64_t shape) {
    return mlir::ShapedType::isDynamic(shape) ? kTFDynamicSize : shape;
  }));
}

TensorSpec::TensorSpec()
    : shape_(std::initializer_list<int64_t>()), dtype_(DT_FLOAT) {}

TensorSpec::TensorSpec(PartialTensorShape shape, DataType dtype)
    : shape_(std::move(shape)), dtype_(dtype) {}

TensorSpec::TensorSpec(const TensorSpecProto& proto)
    : shape_(proto.shape()), dtype_(proto.dtype()) {}

const PartialTensorShape& TensorSpec::shape() const { return shape_; }

bool CanBeRefined(Type type);

// Returns a new arg type based on the shape and element type. If there are
// dynamic bounds attribute to the arg, update the bounds based on the shape
// as well.
Type GetNewArgType(Type old_arg_type, ArrayRef<int64_t> shape,
                   Type element_type, mlir::MLIRContext* context);

// Refines all the shapes in a module, skipping the inference for all ops
// whose type is in ops_to_skip.

      se::Stream* stream, se::DeviceMemoryBase* device_memory_base,
      const xla::Shape& shape,
      tsl::AsyncValueRef<std::unique_ptr<se::Event>>& done_event)
      : stream_(stream),
        device_memory_base_(device_memory_base),
        shape_(shape),
        done_event_(done_event) {}

  // Copies 'cpu_tensor' to `device_memory_base_` with `shape_`.
  // `device_tensor` is unused.

node: {
  name: "x"
  op: "Placeholder"
  attr: {
    key: "shape"
    value: {
      shape: {
        dim: {
          size: -1
        }
      }
    }
  }
  attr: {
    key: "dtype"
    value: {
      type: DT_INT32
    }
  }
}
node: {
  name: "y"
  op: "Placeholder"
  attr: {
    key: "shape"
    value: {
      shape: {
        dim: {
          size: -1
        }
      }

# RUN: tf-mlir-translate -graphdef-to-mlir -tf-enable-shape-inference-on-import=false %s -tf-input-arrays=input0,input1 -tf-input-shapes=10:10 -tf-output-arrays=Add -o - | FileCheck --check-prefix=NONE %s