// TFLite variables legalization patterns
include "mlir/IR/PatternBase.td"
include "tensorflow/compiler/mlir/lite/ir/tfl_ops.td"
include "tensorflow/compiler/mlir/tensorflow/ir/tf_ops.td"

def HasSupportedElementType :
  Constraint<CPred<"HasSupportedElementType($0.getDefiningOp())">>;

def IsSupportedElementType :
  Constraint<CPred<"IsSupportedElementType($0.getType().cast<ShapedType>())">>;

include "mlir/Dialect/Func/IR/FuncOps.td"
include "mlir/Dialect/Arith/IR/ArithOps.td"
include "tensorflow/compiler/mlir/quantization/common/attrs_and_constraints.td"
include "tensorflow/compiler/mlir/tensorflow/ir/tf_ops.td"

// Converts tf.Const to arith.constant for statically shaped, non-opaque constants.
// Needed for QuantizationDriver to recognize constants.
def ConvertTfConstToArithConst : Pat<

from tensorflow.python.framework import dtypes
from tensorflow.python.framework import ops
from tensorflow.python.framework import tensor_spec
from tensorflow.python.module import module
from tensorflow.python.ops import io_ops
from tensorflow.python.ops import lookup_ops
from tensorflow.python.ops import variables
from tensorflow.python.platform import test
from tensorflow.python.saved_model import save_options

        name = name + "_ops_cc",
        srcs = [name + "_ops.cc"],
        deps = [
            "//tensorflow/core:framework",
            "//tensorflow/core:lib",
            "//tensorflow/core:protos_all_cc",
        ],
        alwayslink = 1,
    )

    tf_custom_op_library(
        name = name + "_ops.so",
        srcs = [name + "_ops.cc"],
    )

    tf_gen_op_wrapper_py(

#ifndef TENSORFLOW_COMPILER_MLIR_QUANTIZATION_TENSORFLOW_CC_CONST_OP_SIZE_H_
#define TENSORFLOW_COMPILER_MLIR_QUANTIZATION_TENSORFLOW_CC_CONST_OP_SIZE_H_

#include <cstdint>

#include "tensorflow/compiler/mlir/tensorflow/ir/tf_ops.h"

namespace mlir {
namespace quant {

// Returns the size in bytes of the underlying data of `const_op`. If the
// underlying type's size cannot be determined, it assumes 4 bytes per element.

#include "mlir/Support/LLVM.h"  // from @llvm-project
#include "tensorflow/compiler/mlir/quantization/common/quantization_lib/quantization_utils.h"
#include "tensorflow/compiler/mlir/tensorflow/ir/tf_ops.h"

namespace mlir::quant {

std::unique_ptr<OpQuantSpec> GetUniformOpQuantSpec(Operation* op) {
  auto spec = std::make_unique<OpQuantSpec>();
  if (isa<TF::UniformQuantizedConvolutionHybridOp>(op) ||

include "mlir/IR/PatternBase.td"
include "mlir/Dialect/Arith/IR/ArithOps.td"
include "mlir/Dialect/Func/IR/FuncOps.td"
include "tensorflow/compiler/mlir/lite/ir/tfl_ops.td"
include "tensorflow/compiler/mlir/tensorflow/ir/tf_ops.td"

// Basically x - y => x + y * -1
def SubToAdd : Pat<(TFL_SubOp $lhs, $rhs, $act),
  (TFL_AddOp $lhs, (TFL_MulOp $rhs,
                  (Arith_ConstantOp ConstantAttr<RankedF32ElementsAttr<[]>,

See the License for the specific language governing permissions and
limitations under the License.
==============================================================================*/
include "tensorflow/compiler/mlir/tensorflow/ir/tf_ops.td"
include "mlir/IR/OpBase.td"
include "mlir/IR/PatternBase.td"
include "tensorflow/compiler/mlir/quantization/common/attrs_and_constraints.td"

// Combines the two variadic arguments ($in_tensors and $captured_tensors).

limitations under the License.
==============================================================================*/

include "mlir/IR/OpBase.td"
include "mlir/IR/PatternBase.td"
include "tensorflow/compiler/mlir/tensorflow/ir/tf_ops.td"

// TODO(jpienaar): Move this somewhere general.
class GetI64ScalarElementsAttr<int value> :
  NativeCodeCall<"DenseElementsAttr::get<int64_t>(RankedTensorType::get({}, $_builder.getIntegerType(64)), " # value # ")">;

limitations under the License.
==============================================================================*/
include "mlir/IR/OpBase.td"
include "mlir/IR/PatternBase.td"
include "tensorflow/compiler/mlir/tensorflow/ir/tf_ops.td"

// Note that this may not be a valid transformation for general tensorflow
// program, but from the perspective of mlir-reduce, it's worth to give it a try
// because the validity will be checked separately by the tester.