TosaLegalizeCommon.cpp

//===----------------------------------------------------------------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
// Also available under a BSD-style license. See LICENSE.
//
//===----------------------------------------------------------------------===//

#include "torch-mlir/Conversion/TorchToTosa/TosaLegalizeCommon.h"
#include "mlir/Dialect/Tosa/IR/TosaOps.h" // from @llvm-project
#include "mlir/Dialect/Tosa/Utils/ConversionUtils.h"
#include "torch-mlir/Conversion/Utils/Utils.h"
#include "torch-mlir/Dialect/Torch/IR/TorchOps.h"

#include <cstdint>
#include <iterator>
#include <numeric>

#include "mlir/Dialect/Tensor/IR/Tensor.h" // from @llvm-project
#include "llvm/Support/FormatVariadic.h"

namespace mlir {
namespace tosa {

using namespace mlir::torch::Torch;

// This function is a helper for `convertTorchIndexToTfIndices`.
//
// We convert PyTorch index to TensorFlow-style indices so that we can use
// `convertGatherNdOp` and `convertScatterNdOp` functions, which lower Gather
// and Scatter operators to TOSA using TensorFlow-style indices.
// The difference between PyTorch/ONNX Gather/Scatter and TensorFlow
// Gather/Scatter ops is that PyTorch/ONNX take in the dimension that you want
// to gather/scatter elements, while in TensorFlow, the indices point directly
// to positions that you want to gather/scatter elements.
std::optional<Value>
createOneDimTfIndices(PatternRewriter &rewriter, Operation *op,
                      SmallVector<int64_t> indicesOneDimShape, int32_t dim,
                      ArrayRef<int64_t> indexShape) {
  unsigned indexRank = indexShape.size();
  SmallVector<int32_t> indicesVec;         // input vec to create tosaConstant
  SmallVector<int32_t> indicesMetaElement; // torch.meshgrid inputs

  // Create torch.meshgrid inputs
  // Example: indexShape=[1,4,2]
  // dim0: indicesMetaElement = torch.arange(0, 1) = [0]
  // dim1: indicesMetaElement = torch.arange(0, 4) = [0,1,2,3]
  // dim2: indicesMetaElement = torch.arange(0, 2) = [0,1]
  for (int i = 0; i < indexShape[dim]; i++)
    indicesMetaElement.push_back(i);

  int preDimMetaElementRepeatTimes = 1;
  int postDimMetaElementRepeatTimes = 1;

  // Compute total number of times meta element range should repeat
  // = product(indexShape[0:dim])
  // dim0: preDimMetaElementRepeatTimes = 1
  // dim1: preDimMetaElementRepeatTimes = 1
  // dim2: preDimMetaElementRepeatTimes = 1 x 4 = 4
  for (int i = 0; i < dim; i++)
    preDimMetaElementRepeatTimes *= indexShape[i];

  // Compute total number of times meta element repeat
  // = product(indexShape[dim+1:indexRank])
  // dim0: postDimMetaElementRepeatTimes = 4 x 2 = 8
  // dim1: postDimMetaElementRepeatTimes = 2
  // dim2: postDimMetaElementRepeatTimes = 1
  for (int i = dim + 1; i < static_cast<int>(indexRank); i++)
    postDimMetaElementRepeatTimes *= indexShape[i];

  // Example using dim1:
  // preDimMetaElementRepeatTimes = 1
  // postDimMetaElementRepeatTimes = 2
  // Using postDimMetaElementRepeatTimes, we get the meta element range:
  // [0 0 1 1 2 2 3 3]
  // Using preDimMetaElementRepeatTimes, we get the full one dim indices:
  // [0 0 1 1 2 2 3 3]
  //
  // Let's use a clearer example:
  // indexShape = [3, 4, 2]
  // Target dim = 1
  // => preDimMetaElementRepeatTimes = 3
  //    postDimMetaElementRepeatTimes = 2
  // Using postDimMetaElementRepeatTimes, we get the meta element range:
  // [0 0 1 1 2 2]
  // Using preDimMetaElementRepeatTimes, we get the full one dim indices:
  // [0 0 1 1 2 2 0 0 1 1 2 2 0 0 1 1 2 2]
  for (int i = 0; i < preDimMetaElementRepeatTimes; i++) {
    for (size_t elementId = 0; elementId < indicesMetaElement.size();
         elementId++) {
      for (int j = 0; j < postDimMetaElementRepeatTimes; j++) {
        indicesVec.push_back(indicesMetaElement[elementId]);
      }
    }
  }

  // Create tosa::ConstOp Tensor for indicesVec with target shape.
  // torch.unsqueeze(torch.stack(torch.meshgrid)))
  // dim0:          tensor([[   [ [0], [0] ],
  //			            	[ [0], [0] ],
  //			            	[ [0], [0] ],
  //			              	[ [0], [0] ], ]]) 1*4*2*1
  // dim1:	        tensor([[   [ [0], [0] ],
  //			             	[ [1], [1] ],
  //			            	[ [2], [2] ],
  //			             	[ [3], [3] ], ]]) 1*4*2*1
  // dim2/last dim:	tensor([[   [ [0], [1] ],
  //		                   	[ [0], [1] ],
  //			            	[ [0], [1] ],
  //		    	        	[ [0], [1] ], ]]) 1*4*2*1
  auto indicesDim = getConstTensor<int32_t>(rewriter, op,
                                            /*vec=*/indicesVec,
                                            /*shape=*/indicesOneDimShape);
  return indicesDim;
}

// Default function to create TOSA op with shift value
tosa::MulOp createMulOpAndCast(PatternRewriter &rewriter, Operation *op,
                               TensorType outType, Value lhs, Value rhs,
                               int32_t shift) {
  lhs = tosa::tosaCastTensorToType(rewriter, lhs, outType).value();
  rhs = tosa::tosaCastTensorToType(rewriter, rhs, outType).value();

  auto constShift = tosa::getTosaMulShiftConstTensor(rewriter, op, shift);

  return tosa::CreateOpAndInfer<tosa::MulOp>(rewriter, op->getLoc(), outType,
                                             lhs, rhs, constShift);
}

template <>
tosa::IntDivOp
createBinaryOpAndCast<IntDivOp>(PatternRewriter &rewriter, Operation *op,
                                TensorType outType, Value lhs, Value rhs) {
  auto lhsElemTy = cast<TensorType>(lhs.getType()).getElementType();
  auto rhsElemTy = cast<TensorType>(rhs.getType()).getElementType();
  if (isa<mlir::FloatType>(lhsElemTy) || isa<mlir::FloatType>(rhsElemTy)) {
    (void)rewriter.notifyMatchFailure(
        op, "tosa.int_div only supports integer type");
  }

  lhs = tosa::tosaCastTensorToType(rewriter, lhs, outType).value();
  rhs = tosa::tosaCastTensorToType(rewriter, rhs, outType).value();
  return tosa::CreateOpAndInfer<tosa::IntDivOp>(rewriter, op->getLoc(), outType,
                                                lhs, rhs);
}

std::optional<Value> convertTorchIndexToTfIndices(PatternRewriter &rewriter,
                                                  Operation *op,
                                                  Value paramsValue,
                                                  Value indexValue,
                                                  int32_t axis) {
  // For easy understanding of this algorithm, the following comments are with
  // an exact example: torch.aten.gather(!torch.vtensor<[1,4,3],f32>, axis=2,
  // !torch.vtensor<[1,4,2],si64>) -> !torch.vtensor<[1,4,2],f32>
  // https://gist.github.com/AmosLewis/2f18434397025211da4491735bcc6db6
  //
  // Convert Torch Index     to       TF Indices
  //    [[                         [[   d0 d1  d2  d0 d1 d2
  //        [0,0],                     [[0, 0, 0],[0, 0, 0]],
  //        [1,0],                     [[0, 1, 1],[0, 1, 0]],
  //        [2,1],                     [[0, 2, 2],[0, 2, 1]],
  //        [2,1]                      [[0, 3, 2],[0, 3, 1]]
  //    ]] 1*4*2                   ]] 1*4*2*3

  auto paramsType = dyn_cast<RankedTensorType>(paramsValue.getType());
  auto indexType = dyn_cast<RankedTensorType>(indexValue.getType());
  auto paramsShape = paramsType.getShape(); // [1 4 3]
  auto indexShape = indexType.getShape();   // [1 4 2]
  int paramsRank = paramsShape.size();      // 3
  int indexRank = indexShape.size();        // 3

  // Initialize the final tf indices shape, and the shape of each dim that can
  // concat to this tf indices
  SmallVector<int64_t> indicesShape;       // [1 4 2 3]
  SmallVector<int64_t> indicesOneDimShape; // [1 4 2 1]
  for (auto shape : indexShape) {
    indicesShape.push_back(shape);
    indicesOneDimShape.push_back(shape);
  }
  indicesShape.push_back(paramsRank);
  indicesOneDimShape.push_back(1);

  // Get the chosen axis index
  // indexValue reshape to indicesDim: shape append 1
  // [1 4 2] -> [1 4 2 1]
  // dim2:	tensor([[   [ [0], [0] ],
  //			    [ [1], [0] ],
  //			    [ [2], [1] ],
  //			    [ [2], [1] ], ]]) 1*4*2*1
  auto indicesChosenAxis = tosa::CreateOpAndInfer<tosa::ReshapeOp>(
      rewriter, op->getLoc(),
      GetTypeFromTensorShape(indicesOneDimShape, indexType.getElementType()),
      indexValue,
      tosa::getTosaConstShape(rewriter, op->getLoc(), indicesOneDimShape));

  SmallVector<Value> concatInputs;
  for (auto dim = 0; dim < paramsRank; dim++) {
    if (dim != axis) {
      auto indices = createOneDimTfIndices(rewriter, op, indicesOneDimShape,
                                           dim, indexShape);
      concatInputs.push_back(indices.value());
    } else {
      // the chosen axis indices will be replaced by index[i][j][k]
      concatInputs.push_back(indicesChosenAxis.getResult());
    }
  }

  // detailed example explanation
  // https://gist.github.com/AmosLewis/932a8dee3ba7657dcc6d09a4da4775d4 Get TF
  // indices: 1*4*2*3
  // [[  d0 d1  d2  d0 d1 d2
  //    [[0, 0, 0],[0, 0, 0]],
  //    [[0, 1, 1],[0, 1, 0]],
  //    [[0, 2, 2],[0, 2, 1]],
  //    [[0, 3, 2],[0, 3, 1]]
  // ]]
  auto indicesTf = tosa::CreateOpAndInfer<tosa::ConcatOp>(
      rewriter, op->getLoc(),
      GetTypeFromTensorShape(indicesShape, rewriter.getIntegerType(32)),
      concatInputs, indexRank);

  return indicesTf.getResult();
}

// Lowers Gather operators to a sequence of TOSA ops.
// taken from
// https://github.com/tensorflow/tensorflow/blob/master/tensorflow/compiler/mlir/tosa/transforms/legalize_common.cc
std::optional<Value> convertGatherNdOp(PatternRewriter &rewriter, Operation *op,
                                       Type outType, Value paramsValue,
                                       Value indicesValue) {
  auto resultType = dyn_cast<ShapedType>(outType);
  auto paramsType = dyn_cast<RankedTensorType>(paramsValue.getType());
  auto indicesType = dyn_cast<RankedTensorType>(indicesValue.getType());

  if (!resultType || !paramsType || !indicesType)
    return std::nullopt;

  // N: number of batches
  // Always 1 for GatherND
  //
  // Because TOSA's GATHER operator already uses the symbol 'N' for
  // the number of batches, we will use the symbol 'ND' to specify the
  // number of dimensions that are sliced from params instead of'N' in
  // the TF MLIR documentation.
  //
  // ND: indices.shape[-1]
  //
  // W: number of indices in each batch
  // Computed as:
  // product(indices.shape[0:-1]) (all but the last dimension)
  //
  // K: range of each index
  // Computed as:
  // product(params.shape[0:ND-1])
  //
  // C: number of channels for each index
  // Computed as:
  // product(params.shape[ND:])
  //
  // The params tensor needs to be reshaped, but not transposed, to move the
  // dimensions into [N, K, C] order.
  //
  // The dimensions of the input params[] tensor are grouped in the following
  // order to begin with:
  //
  //  [ParamIndices, ParamChannels]
  //  |------------||-------------|
  //         K              C
  //
  // The reshape simply flattens the params tensor into a 2D [K, C] shape.
  //
  // Indices needs to be put in the form of [N, W], but a simple flattening
  // will not suffice, because the indices need to index into a [W]-shape
  // vector instead of the params.shape[0:ND-1] tensor that we had before.
  //
  // To flatten the coordinates, first reshape indices to a [W, ND] matrix,
  // where the matrix now represents W ND-dimensional coordinates into the
  // params tensor.
  //
  // From here, we take each of the ND dimensions and multiply it with
  // the size of the next params dimension (or 1 for the last
  // dimension), then sum all these together with a reduce_sum
  // operator.  This is exactly the same mathematics as one would use
  // flatten the indices of an N-dimensional row-major array into a
  // 1-D array in C.
  //
  // More precisely, do an element-wise multiply with [params.shape[1
  // .. ND], 1] in axis 1, then reduce_sum in axis 1 to flatten to a
  // [W]-shaped tensor, then trivially reshape to [N=1, W] to be
  // compatible with the GATHER operator's shape.
  //
  // Then perform the tosa.GATHER() operation.
  //
  // Now we have result = [N, K, C].
  //
  // Reshape with a single, simple reshape to the final output shape of:
  //  [Indices, ParamChannels]
  //
  // Where, Indices is indices.shape[0:ND-1]
  //
  // For easy understanding, all following comments take an exact value for each
  // argument Example: Take TF style indices as input
  //    func.func @torch.aten.gather(%arg0: !torch.vtensor<[1,4,3],f32>,
  //        %arg1: !torch.vtensor<[1,4,2,3],i32>) -> !torch.vtensor<[1,4,2],f32>
  // Detail algorithm visualization:
  // https://gist.github.com/AmosLewis/bb6e3a0ad9fd1705c9f9d42a2eefbb88

  int N = 1, W = 1, K = 1, C = 1, ND = 1;

  int paramsRank = paramsType.getShape().size();   // 3
  int indicesRank = indicesType.getShape().size(); // 4

  //  ND: indices.shape[-1]
  ND = indicesType.getShape()[indicesRank - 1]; // 3

  if (ND > paramsRank) {
    (void)rewriter.notifyMatchFailure(
        op, "size of last dimension of indices must be <= params rank");
    return std::nullopt;
  }

  // Calculate N, K, W, C.  (N is always 1)
  // number of indices in each batch. product(indices.shape[0:-1]) (all but the
  // last dimension) W = 1*4*2 = 8
  for (int i = 0; i < (indicesRank - 1); i++) {
    W *= indicesType.getShape()[i];
  }

  // K: range of each index, total number of inputs(chould be gather) after
  // flattened k = 1*1*4*3 = 12
  for (int i = 0; i < ND; i++) {
    K *= paramsType.getShape()[i];
  }

  // C: number of channels for each index : numbers of values inside each
  // input(chould be gather) C = product(params.shape[ND:] ND = 3, paramsRank,
  // C = 1
  for (int i = ND; i < paramsRank; i++) {
    C *= paramsType.getShape()[i];
  }

  // int N = 1, W = 8, K = 12, C = 1, ND = 3;
  SmallVector<int64_t, 3> tosaValuesShape({N, K, C});  // {1,12,1}
  SmallVector<int64_t, 2> tosaIndicesShape({N, W});    // {1,8}
  SmallVector<int64_t, 2> indicesMatrixShape({W, ND}); // {8,3}
  SmallVector<int64_t, 2> indicesMatrixReducesumShape(
      {W, 1}); // {8,1} This is different from tf tosa code
  SmallVector<int64_t, 3> tosaGatherResultShape({N, W, C}); // {1,8,1}

  // %2 = "tosa.reshape"(%0) {new_shape = [1, 12, 1]} : (tensor<1x4x3xf32>) ->
  // tensor<1x12x1xf32>
  auto tosaValuesReshapeOp = tosa::CreateOpAndInfer<tosa::ReshapeOp>(
      rewriter, op->getLoc(),
      GetTypeFromTensorShape(tosaValuesShape, paramsType.getElementType()),
      paramsValue,
      tosa::getTosaConstShape(rewriter, op->getLoc(), tosaValuesShape));

  // %3 = "tosa.reshape"(%1) {new_shape = [8, 3]} : (tensor<1x4x2x3xi32>) ->
  // tensor<8x3xi32> Flatten the input indices tensor to an [W, ND] matrix.
  Value indicesMatrixReshapeOp = tosa::CreateOpAndInfer<tosa::ReshapeOp>(
      rewriter, op->getLoc(),
      GetTypeFromTensorShape(indicesMatrixShape, indicesType.getElementType()),
      indicesValue,
      tosa::getTosaConstShape(rewriter, op->getLoc(), indicesMatrixShape));

  SmallVector<int32_t> flattenedCoeffVec; //  [12,3,1]
  // flattenedCoeffVec = [4,3,1]
  for (int i = 1; i < ND; i++) {
    flattenedCoeffVec.push_back(paramsType.getShape()[i]);
  }
  flattenedCoeffVec.push_back(1);

  // flattenedCoeffVec = [12,3,1]
  for (int i = ND - 1; i > 0; i--) {
    flattenedCoeffVec[i - 1] *= flattenedCoeffVec[i];
  }

  // Create the tosaConstTensor for the flattenedCoeffVec
  // %4 = "tosa.const"() {value = dense<[12, 3, 1]> : tensor<3xi32>} : () ->
  // tensor<3xi32>
  auto flattenedCoeffValue =
      getConstTensor<int32_t>(rewriter, op, flattenedCoeffVec,
                              {static_cast<int64_t>(flattenedCoeffVec.size())});

  if (!flattenedCoeffValue)
    return std::nullopt;

  if (mlir::tosa::EqualizeRanks(rewriter, op->getLoc(), indicesMatrixReshapeOp,
                                flattenedCoeffValue.value())
          .failed())
    return std::nullopt;

  // Multiply the coefficients by the coordinates
  // %5 = "tosa.mul"(%3, %4) {shift = 0 : i32} : (tensor<8x3xi32>,
  // tensor<3xi32>) -> tensor<8x3xi32>
  auto flattenedIndicesMulOp = tosa::createMulOpAndCast(
      rewriter, op,
      GetTypeFromTensorShape(indicesMatrixShape, indicesType.getElementType()),
      indicesMatrixReshapeOp, flattenedCoeffValue.value(), 0);

  // Sum up the products of the coefficients and coordinates
  // %6 = "tosa.reduce_sum"(%5) {axis = 1 : i64} : (tensor<8x3xi32>) ->
  // tensor<8x1xi32>
  auto flattenedIndicesReduceOp = tosa::CreateOpAndInfer<tosa::ReduceSumOp>(
      rewriter, op->getLoc(),
      GetTypeFromTensorShape(indicesMatrixReducesumShape,
                             indicesType.getElementType()),
      flattenedIndicesMulOp.getResult(), rewriter.getI32IntegerAttr(1));

  // And reshape to [N, W]
  // %7 = "tosa.reshape"(%6) {new_shape = [1, 8]} : (tensor<8x1xi32>) ->
  // tensor<1x8xi32>
  auto tosaIndicesReshapeOp = tosa::CreateOpAndInfer<tosa::ReshapeOp>(
      rewriter, op->getLoc(),
      GetTypeFromTensorShape(tosaIndicesShape, indicesType.getElementType()),
      flattenedIndicesReduceOp.getResult(),
      tosa::getTosaConstShape(rewriter, op->getLoc(), tosaIndicesShape));

  // Now the gather op itself
  // %9 = "tosa.gather"(%2, %7) : (tensor<1x12x1xf32>, tensor<1x8xi32>) ->
  // tensor<1x8x1xf32>
  auto resultElemTy = resultType.getElementType();
  Value valuesForGather = tosaValuesReshapeOp.getResult();
  Type gatherElemTy = resultElemTy;
  if (auto intTy = dyn_cast<IntegerType>(resultElemTy);
      intTy && intTy.getWidth() == 1) {
    auto i8Ty = rewriter.getI8Type();
    valuesForGather = tosa::tosaCastTensorToType(
                          rewriter, valuesForGather,
                          GetTypeFromTensorShape(tosaValuesShape, i8Ty))
                          .value();
    gatherElemTy = i8Ty;
  }

  auto tosaGatherOp = tosa::CreateOpAndInfer<tosa::GatherOp>(
      rewriter, op->getLoc(),
      GetTypeFromTensorShape(tosaGatherResultShape, gatherElemTy),
      valuesForGather, tosaIndicesReshapeOp.getResult());

  // Finally, reshape back to the original output shape of [Indices,
  // ParamChannels]. %10 = "tosa.reshape"(%9) {new_shape = [1, 4, 2]} :
  // (tensor<1x8x1xf32>) -> tensor<1x4x2xf32> %11 = torch_c.from_builtin_tensor
  // %10 : tensor<1x4x2xf32> -> !torch.vtensor<[1,4,2],f32>
  Value gatherResult = tosaGatherOp.getResult();
  if (auto intTy = dyn_cast<IntegerType>(resultElemTy);
      intTy && intTy.getWidth() == 1) {
    gatherResult =
        tosa::tosaCastTensorToType(
            rewriter, gatherResult,
            GetTypeFromTensorShape(tosaGatherResultShape, resultElemTy))
            .value();
  }

  return tosa::CreateOpAndInfer<tosa::ReshapeOp>(
             rewriter, op->getLoc(), resultType, gatherResult,
             tosa::getTosaConstShape(rewriter, op->getLoc(),
                                     resultType.getShape()))
      .getResult();
}

// Lower indexput op to tosa::scatter op
// Mostly take from the up function convertGatherNdOp()
std::optional<Value> convertScatterNdOp(PatternRewriter &rewriter,
                                        Operation *op, Type outType,
                                        Value paramsValue, Value indicesValue,
                                        Value fillValues) {
  auto resultType = dyn_cast<ShapedType>(outType);
  auto paramsType = dyn_cast<RankedTensorType>(paramsValue.getType());
  auto indicesType = dyn_cast<RankedTensorType>(indicesValue.getType());
  auto fillValuesType = dyn_cast<RankedTensorType>(fillValues.getType());

  if (!resultType || !paramsType || !indicesType)
    return std::nullopt;

  // N: number of batches
  // Always 1 for ScatterOp
  //
  // Because TOSA's Scatter operator already uses the symbol 'N' for
  // the number of batches, we will use the symbol 'ND' to specify the
  // number of dimensions that are sliced from params instead of'N' in
  // the TF MLIR documentation.
  //
  // ND: indices.shape[-1]
  //
  // W: number of indices in each batch
  // Computed as:
  // product(indices.shape[0:-1]) (all but the last dimension)
  //
  // K: range of each index
  // Computed as:
  // product(params.shape[0:ND-1])
  //
  // C: number of channels for each index
  // Computed as:
  // product(params.shape[ND:])
  //
  // The params tensor needs to be reshaped, but not transposed, to move the
  // dimensions into [N, K, C] order.
  //
  // The dimensions of the input params[] tensor are grouped in the following
  // order to begin with:
  //
  //  [ParamIndices, ParamChannels]
  //  |------------||-------------|
  //         K              C
  //
  // The reshape simply flattens the params tensor into a 2D [K, C] shape.
  //
  // Indices needs to be put in the form of [N, W], but a simple flattening
  // will not suffice, because the indices need to index into a [W]-shape
  // vector instead of the params.shape[0:ND-1] tensor that we had before.
  //
  // To flatten the coordinates, first reshape indices to a [W, ND] matrix,
  // where the matrix now represents W ND-dimensional coordinates into the
  // params tensor.
  //
  // From here, we take each of the ND dimensions and multiply it with
  // the size of the next params dimension (or 1 for the last
  // dimension), then sum all these together with a reduce_sum
  // operator.  This is exactly the same mathematics as one would use
  // flatten the indices of an N-dimensional row-major array into a
  // 1-D array in C.
  //
  // More precisely, do an element-wise multiply with [params.shape[1
  // .. ND], 1] in axis 1, then reduce_sum in axis 1 to flatten to a
  // [W]-shaped tensor, then trivially reshape to [N=1, W] to be
  // compatible with the scatter operator's shape.
  //
  // Then perform the tosa.scatter() operation.
  //
  // Now we have result = [N, K, C].
  //
  // Reshape with a single, simple reshape to the final output shape of:
  //  [Indices, ParamChannels]
  //
  // Where, Indices is indices.shape[0:ND-1]
  //
  // For easy understanding, all following comments take an exact value for each
  // argument Example: Take TF style indices as input
  //  torch.aten._index_put_impl %input, %indices, %fillValue, %false, %false :
  //    !torch.vtensor<[1,4],si64>, !torch.vtensor<[3,2],si64>,
  //    !torch.vtensor<[1,3],si64>, !torch.bool, !torch.bool ->
  //    !torch.vtensor<[1,4],si64>
  // Detail algorithm visualization:

  int N = 1, W = 1, K = 1, fillK = 1, C = 1, ND = 1;

  int paramsRank = paramsType.getShape().size();   // 2
  int indicesRank = indicesType.getShape().size(); // 2

  //  ND: indices.shape[-1]
  ND = indicesType.getShape()[indicesRank - 1]; // 2 depth of input

  if (ND > paramsRank) {
    (void)rewriter.notifyMatchFailure(
        op, "size of last dimension of indices must be <= params rank");
    return std::nullopt;
  }

  // Calculate N, K, W, C.  (N is always 1)
  // number of indices/selected value in each batch product(indices.shape[0:-1])
  // (all but the last dimension) W = 1*3 = 3
  for (int i = 0; i < (indicesRank - 1); i++) {
    W *= indicesType.getShape()[i];
  }

  // K: range of each index, total number of inputs(chould be scatter) after
  // flattened k = 1*1*4 = 4
  for (int i = 0; i < ND; i++) {
    K *= paramsType.getShape()[i];
  }

  // C: number of channels for each index : numbers of values inside each
  // input(chould be scatter) C = product(params.shape[ND:] ND = 2, paramsRank,
  // C = 1
  for (int i = ND; i < paramsRank; i++) {
    C *= paramsType.getShape()[i];
  }

  // int N = 1, W = 3, K = 4, fillk = 3, C = 1, ND = 2;
  SmallVector<int64_t, 3> tosaInputValuesShape({N, K, C});     // {1,4,1}
  SmallVector<int64_t, 2> tosaIndicesShape({N, W});            // {1,3}
  SmallVector<int64_t, 2> indicesMatrixShape({W, ND});         // {3,2}
  SmallVector<int64_t, 2> indicesMatrixReducesumShape({W, 1}); // {3,1}

  // Preprocess fill value.
  // There are 2 cases of fillValues,
  // 1. !torch.vtensor<[1,3],si64>
  //  [[0,0,0]] -> [[[0], [0], [0]]]
  // 2. !torch.vtensor<[],si64>
  //    reshape(1) tile(3)  reshape(1,3)   reshape(1,3,1)
  //  [] -> [0] -> [0,0,0] -> [[0,0,0]] -> [[[0], [0], [0]]]
  //    reshape to [1] and then tile to same number of indicesValue.shape[0],
  //    [1,1,1]
  if (fillValuesType.getRank() == 0) {
    // [] -> [0]
    SmallVector<int64_t, 1> oneShape({1}); // {3,1}
    auto tosaFillValuesOneReshapeOp = tosa::CreateOpAndInfer<tosa::ReshapeOp>(
        rewriter, op->getLoc(),
        GetTypeFromTensorShape(oneShape, fillValuesType.getElementType()),
        fillValues, tosa::getTosaConstShape(rewriter, op->getLoc(), oneShape));

    // [0] -> [0,0,0]
    SmallVector<int64_t, 1> tileShape({W}); // {3}
    auto tileOpMultiples =
        tosa::getTosaConstShape(rewriter, op->getLoc(), tileShape);
    auto tosaFillValuesTileOp = tosa::CreateOpAndInfer<tosa::TileOp>(
        rewriter, op->getLoc(),
        GetTypeFromTensorShape(tileShape, fillValuesType.getElementType()),
        tosaFillValuesOneReshapeOp.getResult(), tileOpMultiples);

    // [0,0,0] -> [[0,0,0]]
    SmallVector<int64_t, 2> newTosaFillValuesShape({N, W}); // {1,3}
    auto newTosaFillValuesReshapeOp = tosa::CreateOpAndInfer<tosa::ReshapeOp>(
        rewriter, op->getLoc(),
        GetTypeFromTensorShape(newTosaFillValuesShape,
                               fillValuesType.getElementType()),
        tosaFillValuesTileOp.getResult(),
        tosa::getTosaConstShape(rewriter, op->getLoc(),
                                newTosaFillValuesShape));
    fillValues = newTosaFillValuesReshapeOp.getResult();
    fillValuesType = dyn_cast<RankedTensorType>(fillValues.getType());
  }

  // fillK: range of each index, total number of fillInput(could be scatter)
  // after flattened k = 1*1*3 = 3
  for (int i = 0; i < ND; i++) {
    fillK *= fillValuesType.getShape()[i];
  }
  SmallVector<int64_t, 3> tosaFillValuesShape({N, fillK, C}); // {1,3,1}

  // Reshape/Flatten fillValues to 3d tensor
  // [[0,0,0]] -> [[[0], [0], [0]]]
  // %10 = "tosa.reshape"(%1) {new_shape = array<i64: 1, 3, 1>} :
  // (tensor<1x3xi64>) -> tensor<1x3x1xi64>
  auto tosaFillValuesReshapeOp = tosa::CreateOpAndInfer<tosa::ReshapeOp>(
      rewriter, op->getLoc(),
      GetTypeFromTensorShape(tosaFillValuesShape,
                             fillValuesType.getElementType()),
      fillValues,
      tosa::getTosaConstShape(rewriter, op->getLoc(), tosaFillValuesShape));

  // Reshape/Flatten input to 3d tensor
  // [[1, 2, 3, 4]] -> [[[1], [2], [3], [4]]]
  // %9 = "tosa.reshape"(%0) {new_shape = array<i64: 1, 4, 1>} :
  // (tensor<1x4xi64>) -> tensor<1x4x1xi64>
  auto tosaValuesReshapeOp = tosa::CreateOpAndInfer<tosa::ReshapeOp>(
      rewriter, op->getLoc(),
      GetTypeFromTensorShape(tosaInputValuesShape, paramsType.getElementType()),
      paramsValue,
      tosa::getTosaConstShape(rewriter, op->getLoc(), tosaInputValuesShape));

  // Reshape/Flatten the input indices tensor to a 2d  [W, ND] matrix.
  // [[0, 1], [0, 2],  [0, 3]] -> [[0, 1], [0, 2],  [0, 3]]
  // %11 = "tosa.reshape"(%8) {new_shape = array<i64: 3, 2>} : (tensor<3x2xi32>)
  // -> tensor<3x2xi32>
  Value indicesMatrixReshapeOp = tosa::CreateOpAndInfer<tosa::ReshapeOp>(
      rewriter, op->getLoc(),
      GetTypeFromTensorShape(indicesMatrixShape, indicesType.getElementType()),
      indicesValue,
      tosa::getTosaConstShape(rewriter, op->getLoc(), indicesMatrixShape));

  SmallVector<int32_t> flattenedCoeffVec; //  [4,1]
  // flattenedCoeffVec = [4,1]
  for (int i = 1; i < ND; i++) {
    flattenedCoeffVec.push_back(paramsType.getShape()[i]);
  }
  flattenedCoeffVec.push_back(1);

  // flattenedCoeffVec = [4,1]
  for (int i = ND - 1; i > 0; i--) {
    flattenedCoeffVec[i - 1] *= flattenedCoeffVec[i];
  }

  // Create the tosaConstTensor for the flattenedCoeffVec.
  // %12 = "tosa.const"() {value = dense<[4, 1]> : tensor<2xi32>} : () ->
  // tensor<2xi32>
  auto flattenedCoeffValue =
      getConstTensor<int32_t>(rewriter, op, flattenedCoeffVec,
                              {static_cast<int64_t>(flattenedCoeffVec.size())});

  if (!flattenedCoeffValue)
    return std::nullopt;

  if (mlir::tosa::EqualizeRanks(rewriter, op->getLoc(), indicesMatrixReshapeOp,
                                flattenedCoeffValue.value())
          .failed())
    return std::nullopt;

  // Multiply the coefficients by the coordinates.
  // [[0, 1], [0, 2],  [0, 3]] X [4, 1] -> [[4*0, 1*1], [4*0, 1*2], [4*0, 1*3]]
  // %13 = "tosa.mul"(%11, %12) {shift = 0 : i32} : (tensor<3x2xi32>,
  // tensor<2xi32>) -> tensor<3x2xi32>
  auto flattenedIndicesMulOp = tosa::createMulOpAndCast(
      rewriter, op,
      GetTypeFromTensorShape(indicesMatrixShape, indicesType.getElementType()),
      indicesMatrixReshapeOp, flattenedCoeffValue.value(), 0);

  // Sum up the products of the coefficients and coordinates
  // [[4*0 + 1*1], [4*0 + 1*2], [4*0 + 1*3]] = [[1],[2],[3]]
  // %14 = "tosa.reduce_sum"(%13) {axis = 1 : i64} : (tensor<3x2xi32>) ->
  // tensor<3x1xi32>
  auto flattenedIndicesReduceOp = tosa::CreateOpAndInfer<tosa::ReduceSumOp>(
      rewriter, op->getLoc(),
      GetTypeFromTensorShape(indicesMatrixReducesumShape,
                             indicesType.getElementType()),
      flattenedIndicesMulOp.getResult(), rewriter.getI32IntegerAttr(1));

  // And reshape to [N, W]
  // [[1],[2],[3]] -> [[1,2,3]]
  // %15 = "tosa.reshape"(%14) {new_shape = array<i64: 1, 3>} :
  // (tensor<3x1xi32>) -> tensor<1x3xi32>
  auto tosaIndicesReshapeOp = tosa::CreateOpAndInfer<tosa::ReshapeOp>(
      rewriter, op->getLoc(),
      GetTypeFromTensorShape(tosaIndicesShape, indicesType.getElementType()),
      flattenedIndicesReduceOp.getResult(),
      tosa::getTosaConstShape(rewriter, op->getLoc(), tosaIndicesShape));

  // Now the Scatter op itself
  // %16 = "tosa.scatter"(%9, %15, %10) : (tensor<1x4x1xi64>, tensor<1x3xi32>,
  // tensor<1x3x1xi64>) -> tensor<1x4x1xi64> input = [[[1], [2], [3], [4]]],
  // indices = [[1,2,3]], fillValues= [[[0], [0], [0]]] result = [[[1], [0],
  // [0], [0]]]
  auto tosaScatterOp = tosa::CreateOpAndInfer<tosa::ScatterOp>(
      rewriter, op->getLoc(),
      GetTypeFromTensorShape(tosaInputValuesShape, resultType.getElementType()),
      tosaValuesReshapeOp.getResult(), tosaIndicesReshapeOp.getResult(),
      tosaFillValuesReshapeOp.getResult());

  // Finally, reshape back to the original output shape of [Indices,
  // ParamChannels].
  // [[1, 0, 0, 0]]
  // %17 = "tosa.reshape"(%16) {new_shape = array<i64: 1, 4>} :
  // (tensor<1x4x1xi64>) -> tensor<1x4xi64>
  return tosa::CreateOpAndInfer<tosa::ReshapeOp>(
             rewriter, op->getLoc(), resultType, tosaScatterOp.getResult(),
             tosa::getTosaConstShape(rewriter, op->getLoc(),
                                     resultType.getShape()))
      .getResult();
}

// Common function for lowering reduce operations to TOSA ops.
template <typename T>
std::optional<Value> convertReduceOpCommon(
    PatternRewriter &rewriter, Operation *op, RankedTensorType output_type,
    Value input_value, ElementsAttr axes_elems, bool keep_dims,
    Type reduce_element_type, bool is_quantized, double input_scale,
    int64_t input_zp, double output_scale, int64_t output_zp) {
  RankedTensorType input_type =
      dyn_cast<RankedTensorType>(input_value.getType());
  if (!input_type)
    return std::nullopt;

  ArrayRef<int64_t> input_shape = input_type.getShape();
  ArrayRef<int64_t> output_shape = output_type.getShape();
  auto input_rank = input_shape.size();
  Value val = input_value;

  if (axes_elems.getNumElements() == 0) {
    // No axes means return the original tensor.
    auto identity_op = CreateOpAndInfer<tosa::IdentityOp>(
        rewriter, op->getLoc(), output_type, val);
    val = identity_op.getResult();
  } else {
    // Reduce along each axis
    SmallVector<int64_t> shape_vec(input_shape.begin(), input_shape.end());

    if (is_quantized) {
      val = buildRescaleToInt32(rewriter, op, val, input_scale, input_zp);
    }

    for (int i = 0; i < axes_elems.getNumElements(); i++) {
      int64_t axis_val = axes_elems.getValues<IntegerAttr>()[i].getInt();
      if (axis_val < 0)
        axis_val += input_rank;
      auto axis_attr = rewriter.getI32IntegerAttr(axis_val);

      shape_vec[axis_val] = 1;
      RankedTensorType reduce_type =
          RankedTensorType::get(shape_vec, reduce_element_type);

      Value reduce_op;
      if constexpr (std::is_same<T, tosa::ReduceMinOp>() ||
                    std::is_same<T, tosa::ReduceMaxOp>()) {
        // Use default NaN Propagation mode "PROPAGATE" for tosa.reduce_min
        // and tosa.reduce_max
        reduce_op = CreateOpAndInfer<T>(
            rewriter, op->getLoc(), reduce_type, val, axis_attr,
            /*nan_mode=*/
            tosa::NanPropagationModeAttr::get(
                rewriter.getContext(), tosa::NanPropagationMode::PROPAGATE));
      } else {
        reduce_op = CreateOpAndInfer<T>(rewriter, op->getLoc(), reduce_type,
                                        val, axis_attr);
      }

      val = reduce_op;
    }

    if (is_quantized) {
      RankedTensorType output_rescale_type =
          RankedTensorType::get(shape_vec, output_type.getElementType());
      val = buildRescale(rewriter, op, output_rescale_type, val, output_scale,
                         0, output_zp, tosa::RoundingMode::SINGLE_ROUND, true);
    }

    // Optionally squeeze out the reduced axes.
    if (!keep_dims) {
      auto squeezedType =
          RankedTensorType::get(output_shape, reduce_element_type);
      auto reshape_op = CreateOpAndInfer<tosa::ReshapeOp>(
          rewriter, op->getLoc(), squeezedType, val,
          tosa::getTosaConstShape(rewriter, op->getLoc(), output_shape));
      val = reshape_op.getResult();
    }
  }

  // Ensure the result element type matches the expected output type.
  if (val.getType() != output_type) {
    auto casted = tosa::tosaCastTensorToType(rewriter, val, output_type);
    if (!casted)
      return std::nullopt;
    val = casted.value();
  }

  return val;
}

// Lowers ReduceAll to a sequence of TOSA ops.
std::optional<Value>
convertReduceAllOp(PatternRewriter &rewriter, Operation *op,
                   RankedTensorType output_type, Value input_value,
                   ElementsAttr axes_elems, bool keep_dims) {
  RankedTensorType input_type =
      dyn_cast<RankedTensorType>(input_value.getType());
  if (!input_type)
    return std::nullopt;

  return convertReduceOpCommon<tosa::ReduceAllOp>(
      rewriter, op, output_type, input_value, axes_elems, keep_dims,
      output_type.getElementType(), false, 1.0f, 0, 1.0f, 0);
}

// Lowers ReduceAny to a sequence of TOSA ops.
std::optional<Value>
convertReduceAnyOp(PatternRewriter &rewriter, Operation *op,
                   RankedTensorType output_type, Value input_value,
                   ElementsAttr axes_elems, bool keep_dims) {
  RankedTensorType input_type =
      dyn_cast<RankedTensorType>(input_value.getType());
  if (!input_type)
    return std::nullopt;

  return convertReduceOpCommon<tosa::ReduceAnyOp>(
      rewriter, op, output_type, input_value, axes_elems, keep_dims,
      output_type.getElementType(), false, 1.0f, 0, 1.0f, 0);
}

// Lowers ReduceMin to a sequence of TOSA ops.
std::optional<Value>
convertReduceMinOp(PatternRewriter &rewriter, Operation *op,
                   RankedTensorType output_type, Value input_value,
                   ElementsAttr axes_elems, bool keep_dims) {
  RankedTensorType input_type =
      dyn_cast<RankedTensorType>(input_value.getType());
  if (!input_type)
    return std::nullopt;

  return convertReduceOpCommon<tosa::ReduceMinOp>(
      rewriter, op, output_type, input_value, axes_elems, keep_dims,
      output_type.getElementType(), false, 1.0f, 0, 1.0f, 0);
}

// Lowers ReduceMax to a sequence of TOSA ops.
std::optional<Value>
convertReduceMaxOp(PatternRewriter &rewriter, Operation *op,
                   RankedTensorType output_type, Value input_value,
                   ElementsAttr axes_elems, bool keep_dims) {
  RankedTensorType input_type =
      dyn_cast<RankedTensorType>(input_value.getType());
  if (!input_type)
    return std::nullopt;

  return convertReduceOpCommon<tosa::ReduceMaxOp>(
      rewriter, op, output_type, input_value, axes_elems, keep_dims,
      output_type.getElementType(), false, 1.0f, 0, 1.0f, 0);
}

// Lowers ReduceProd to a sequence of TOSA ops.
std::optional<Value>
convertReduceProdOp(PatternRewriter &rewriter, Operation *op,
                    RankedTensorType output_type, Value input_value,
                    ElementsAttr axes_elems, bool keep_dims) {
  RankedTensorType input_type =
      dyn_cast<RankedTensorType>(input_value.getType());
  if (!input_type)
    return std::nullopt;

  bool input_is_qtype =
      isa<mlir::quant::UniformQuantizedType>(input_type.getElementType());
  bool output_is_qtype =
      isa<mlir::quant::UniformQuantizedType>(output_type.getElementType());

  if (input_is_qtype || output_is_qtype) {
    op->emitOpError("ConvertReduceProdOp: input/output tensor should "
                    "be all floating-point.");
    return std::nullopt;
  }

  return convertReduceOpCommon<tosa::ReduceProductOp>(
      rewriter, op, output_type, input_value, axes_elems, keep_dims,
      output_type.getElementType(), false, 1.0f, 0, 1.0f, 0);
}

// Lowers ReduceSum to a sequence of TOSA ops.
std::optional<Value>
convertReduceSumOp(PatternRewriter &rewriter, Operation *op,
                   RankedTensorType output_type, Value input_value,
                   ElementsAttr axes_elems, bool keep_dims) {
  RankedTensorType input_type =
      dyn_cast<RankedTensorType>(input_value.getType());
  if (!input_type)
    return std::nullopt;

  bool input_is_qtype =
      isa<mlir::quant::UniformQuantizedType>(input_type.getElementType());
  bool output_is_qtype =
      isa<mlir::quant::UniformQuantizedType>(output_type.getElementType());

  if (input_is_qtype != output_is_qtype) {
    op->emitOpError("ConvertReduceSumOp: input/output tensor should "
                    "be all quantized or all floating-point.");
    return std::nullopt;
  }

  double input_scale = 1.0f;
  double output_scale = 1.0f;
  int64_t input_zp = 0;
  int64_t output_zp = 0;
  Type reduce_element_type = input_type.getElementType();

  if (input_is_qtype) {
    auto input_qtype =
        cast<mlir::quant::UniformQuantizedType>(input_type.getElementType());
    auto output_qtype =
        cast<mlir::quant::UniformQuantizedType>(output_type.getElementType());

    int32_t input_shift = 20;

    input_scale =
        static_cast<double>(1 << input_shift) * input_qtype.getScale();
    output_scale =
        1.0 / (output_qtype.getScale() * static_cast<double>(1 << input_shift));

    input_zp = input_qtype.getZeroPoint();
    output_zp = output_qtype.getZeroPoint();
    reduce_element_type = rewriter.getI32Type();
  }

  return convertReduceOpCommon<tosa::ReduceSumOp>(
      rewriter, op, output_type, input_value, axes_elems, keep_dims,
      reduce_element_type, input_is_qtype, input_scale, input_zp, output_scale,
      output_zp);
}

// Lowers ReduceMean to a sequence of TOSA ops.
std::optional<Value>
convertReduceMeanOp(PatternRewriter &rewriter, Operation *op,
                    RankedTensorType output_type, Value input_value,
                    ElementsAttr axes_elems, bool keep_dims) {
  // reduce_mean is lowered as followed:
  // op1 = reduce_sum(input)
  // op2 = mul(op1, 1.0 / num_elements_on_reduced_axis)

  RankedTensorType input_type =
      dyn_cast<RankedTensorType>(input_value.getType());
  if (!input_type)
    return std::nullopt;

  bool input_is_qtype =
      isa<mlir::quant::UniformQuantizedType>(input_type.getElementType());
  bool output_is_qtype =
      isa<mlir::quant::UniformQuantizedType>(output_type.getElementType());

  if (input_is_qtype != output_is_qtype) {
    op->emitOpError("ConvertReduceSumOp: input/output tensor should "
                    "be all quantized or all floating-point.");
    return std::nullopt;
  }

  // Only supports float type mean() if it's non-quantized
  if (!input_is_qtype && !isa<mlir::FloatType>(output_type.getElementType())) {
    op->emitWarning(
        "Failed convertReduceMean: input unquantized type but output element "
        "not FloatType!");
    return std::nullopt;
  }

  int64_t input_rank = input_type.getRank();
  ArrayRef<int64_t> inputShape = input_type.getShape();
  int64_t num_elems_on_reduced_axis = 1;
  for (int i = 0; i < axes_elems.getNumElements(); i++) {
    int64_t axis_val = axes_elems.getValues<IntegerAttr>()[i].getInt();
    if (axis_val < 0)
      axis_val += input_rank;
    if (inputShape[axis_val] < 0)
      op->emitOpError("Failed convertReduceMean: support for dynamic input "
                      "shape not implemented");
    num_elems_on_reduced_axis *= inputShape[axis_val];
  }
  double div_scale = 1.0 / static_cast<double>(num_elems_on_reduced_axis);

  double input_scale = 1.0f;
  double output_scale = 1.0f;
  int64_t input_zp = 0;
  int64_t output_zp = 0;
  Type reduce_element_type = input_type.getElementType();

  if (input_is_qtype) {
    auto input_qtype =
        cast<mlir::quant::UniformQuantizedType>(input_type.getElementType());
    auto output_qtype =
        cast<mlir::quant::UniformQuantizedType>(output_type.getElementType());

    // Combine 'div_scale' as part of output rescale
    output_scale = div_scale * input_qtype.getScale() / output_qtype.getScale();

    input_zp = input_qtype.getZeroPoint();
    output_zp = output_qtype.getZeroPoint();
    reduce_element_type = rewriter.getI32Type();
  }

  auto val = convertReduceOpCommon<tosa::ReduceSumOp>(
      rewriter, op, output_type, input_value, axes_elems, keep_dims,
      reduce_element_type, input_is_qtype, input_scale, input_zp, output_scale,
      output_zp);

  if (!val.has_value())
    return std::nullopt;

  if (!input_is_qtype) {
    Value div_const = getTosaConstTensorSingleF32(rewriter, op, div_scale);

    if (mlir::tosa::EqualizeRanks(rewriter, op->getLoc(), val.value(),
                                  div_const)
            .failed())
      return std::nullopt;

    return tosa::createMulOpAndCast(rewriter, op, output_type, val.value(),
                                    div_const, 0)
        .getResult();
  }

  return val;
}

// Lowers LinalgVectorNorm to a sequence of TOSA ops.
std::optional<Value>
convertLinalgVectorNormOp(PatternRewriter &rewriter, Operation *op,
                          RankedTensorType output_type, Value input_value,
                          ElementsAttr axes_elems, bool keep_dims) {
  RankedTensorType input_type =
      dyn_cast<RankedTensorType>(input_value.getType());
  if (!input_type)
    return std::nullopt;

  Type elemType = output_type.getElementType();
  if (!isa<mlir::FloatType>(elemType)) {
    op->emitOpError("Only floating-point datatype legalization supported for "
                    "AtenLinalgVectorNorm op");
    return std::nullopt;
  }

  auto linalgVectorNormOp = cast<AtenLinalgVectorNormOp>(op);
  // TODO: Add support for ord = {0, +inf, -inf}.
  auto epsilon = 1e-5;
  double ordLiteralFloat = 1.0;
  int64_t ordLiteralInt = 1;
  Value ordVal;
  if (matchPattern(linalgVectorNormOp.getOrd(),
                   torch::Torch::m_TorchConstantFloat(&ordLiteralFloat))) {
    ordVal = tosa::getConstTensor<float>(rewriter, op,
                                         {static_cast<float>(ordLiteralFloat)},
                                         {}, elemType)
                 .value();
  } else if (matchPattern(linalgVectorNormOp.getOrd(),
                          torch::Torch::m_TorchConstantInt(&ordLiteralInt))) {
    ordVal = tosa::getConstTensor<float>(rewriter, op,
                                         {static_cast<float>(ordLiteralInt)},
                                         {}, elemType)
                 .value();
  } else {
    op->emitOpError("only support FP or INT type ord parameter");
    return std::nullopt;
  }

  Value ordValRank0 = ordVal;
  if (mlir::tosa::EqualizeRanks(rewriter, op->getLoc(), input_value, ordVal)
          .failed())
    return std::nullopt;

  if (fabs(ordLiteralFloat) < epsilon ||
      fabs(static_cast<double>(ordLiteralInt)) < epsilon) {
    op->emitOpError("unimplemented: L0 norm");
    return std::nullopt;
  }

  if (std::isinf(ordLiteralFloat) ||
      std::isinf(static_cast<double>(ordLiteralInt))) {
    op->emitOpError("unimplemented: ord = +/- inf");
    return std::nullopt;
  }

  auto input_value_casted =
      tosa::tosaCastTensorToType(rewriter, input_value, output_type).value();
  auto absVal = CreateOpAndInfer<tosa::AbsOp>(
                    rewriter, op->getLoc(),
                    RankedTensorType::get(input_type.getShape(), elemType),
                    input_value_casted)
                    .getResult();
  auto powVal = CreateOpAndInfer<tosa::PowOp>(
                    rewriter, op->getLoc(),
                    RankedTensorType::get(input_type.getShape(), elemType),
                    absVal, ordVal)
                    .getResult();
  std::optional<Value> result = convertReduceSumOp(
      rewriter, op, output_type, powVal, axes_elems, keep_dims);
  if (!result)
    return std::nullopt;

  Value reciprocalVal =
      CreateOpAndInfer<tosa::ReciprocalOp>(rewriter, op->getLoc(),
                                           ordValRank0.getType(), ordValRank0)
          .getResult();

  if (mlir::tosa::EqualizeRanks(rewriter, op->getLoc(), result.value(),
                                reciprocalVal)
          .failed())
    return std::nullopt;

  return CreateOpAndInfer<tosa::PowOp>(rewriter, op->getLoc(), output_type,
                                       result.value(), reciprocalVal)
      .getResult();
}

LogicalResult getIntegerClampAttrs(ConversionPatternRewriter &rewriter,
                                   Operation *op, Type elemTy,
                                   std::optional<int64_t> minInt,
                                   std::optional<int64_t> maxInt,
                                   IntegerAttr &minAttr, IntegerAttr &maxAttr) {

  if (!elemTy.isInteger()) {
    return rewriter.notifyMatchFailure(
        op, "getIntegerClampAttrs expects integer type");
  }

  int64_t finalMin, finalMax;
  unsigned bitwidth = elemTy.getIntOrFloatBitWidth();

  switch (bitwidth) {
  case 8:
    finalMin = minInt.value_or(std::numeric_limits<int8_t>::min());
    finalMax = maxInt.value_or(std::numeric_limits<int8_t>::max());
    break;
  case 16:
    finalMin = minInt.value_or(std::numeric_limits<int16_t>::min());
    finalMax = maxInt.value_or(std::numeric_limits<int16_t>::max());
    break;
  case 32:
    finalMin = minInt.value_or(std::numeric_limits<int32_t>::min());
    finalMax = maxInt.value_or(std::numeric_limits<int32_t>::max());
    break;
  case 64:
    finalMin = minInt.value_or(std::numeric_limits<int64_t>::min());
    finalMax = maxInt.value_or(std::numeric_limits<int64_t>::max());
    break;
  default:
    return rewriter.notifyMatchFailure(
        op, "Unsupported integer bitwidth for clamp");
  }

  minAttr = rewriter.getIntegerAttr(elemTy, finalMin);
  maxAttr = rewriter.getIntegerAttr(elemTy, finalMax);

  return success();
}

LogicalResult getFloatClampAttrs(ConversionPatternRewriter &rewriter,
                                 Operation *op, Type elemTy,
                                 std::optional<double> minFloat,
                                 std::optional<double> maxFloat,
                                 FloatAttr &minAttr, FloatAttr &maxAttr) {

  if (elemTy.isF16()) {
    minAttr = rewriter.getF16FloatAttr(
        minFloat.value_or(torch::Torch::Float16Lowest));
    maxAttr =
        rewriter.getF16FloatAttr(maxFloat.value_or(torch::Torch::Float16Max));
  } else if (elemTy.isBF16()) {
    minAttr = rewriter.getFloatAttr(
        elemTy, minFloat.value_or(torch::Torch::BFloat16Lowest));
    maxAttr = rewriter.getFloatAttr(
        elemTy, maxFloat.value_or(torch::Torch::BFloat16Max));
  } else if (elemTy.isF32()) {
    minAttr = rewriter.getF32FloatAttr(
        minFloat.value_or(std::numeric_limits<float>::lowest()));
    maxAttr = rewriter.getF32FloatAttr(
        maxFloat.value_or(std::numeric_limits<float>::max()));
  } else if (elemTy.isF64()) {
    minAttr = rewriter.getF64FloatAttr(
        minFloat.value_or(std::numeric_limits<double>::lowest()));
    maxAttr = rewriter.getF64FloatAttr(
        maxFloat.value_or(std::numeric_limits<double>::max()));
  } else {
    return rewriter.notifyMatchFailure(
        op, "Unsupported floating-point type for clamp");
  }

  return success();
}

std::optional<Value> createRoundHalfToEven(ConversionPatternRewriter &rewriter,
                                           Operation *op, Value input,
                                           RankedTensorType resultTy) {
  // To round to the nearest integer, we will consider the fractional part of
  // the input element (= input element - integer part of element). If the
  // fractional part is smaller than 0.5, round the number down. If the
  // fractional part is 0.5, apply "round half to even" rule. If the fractional
  // part is greater than 0.5, round up.
  //
  // if (frac < 0.5 || (frac == 0.5 && floor(input) % 2 == 0)):
  //   res = floor(input)
  // else:
  //   res = ceil(input)
  auto boolTy =
      RankedTensorType::get(resultTy.getShape(), rewriter.getIntegerType(1));
  auto resultElemTy = resultTy.getElementType();

  auto oneHalf =
      tosa::getConstTensor<float>(rewriter, op, 0.5, {}, resultElemTy).value();
  auto two =
      tosa::getConstTensor<float>(rewriter, op, 2, {}, resultElemTy).value();

  if (mlir::tosa::EqualizeRanks(rewriter, op->getLoc(), input, oneHalf)
          .failed() ||
      mlir::tosa::EqualizeRanks(rewriter, op->getLoc(), input, two).failed()) {
    op->emitError("Failed to equalize ranks among operands and result");
    return std::nullopt;
  }

  auto floorInput =
      tosa::FloorOp::create(rewriter, op->getLoc(), resultTy, input);

  // input - floor(input)
  auto fractionalPart = tosa::SubOp::create(rewriter, op->getLoc(), resultTy,
                                            input, floorInput.getResult());

  auto ceilInput =
      tosa::CeilOp::create(rewriter, op->getLoc(), resultTy, input);

  auto floorInputDivByTwo = tosa::createMulOpAndCast(
      rewriter, op, resultTy, floorInput.getResult(), oneHalf, /*shift=*/0);

  auto floorDivResult = tosa::FloorOp::create(rewriter, op->getLoc(), resultTy,
                                              floorInputDivByTwo.getResult());

  // (floor(input) // 2) * 2
  auto evenComparison = tosa::createMulOpAndCast(
      rewriter, op, resultTy, floorDivResult.getResult(), two, /*shift=*/0);

  // floor(input) // 2) * 2 == input <=> floor(input) % 2 == 0
  auto floorInputEven =
      tosa::EqualOp::create(rewriter, op->getLoc(), boolTy,
                            floorInput.getResult(), evenComparison.getResult());

  auto fracEqualOneHalf = tosa::EqualOp::create(
      rewriter, op->getLoc(), boolTy, fractionalPart.getResult(), oneHalf);

  auto fracLtOneHalf = tosa::GreaterOp::create(
      rewriter, op->getLoc(), boolTy, oneHalf, fractionalPart.getResult());

  // (frac == 0.5) && (floor(input) % 2 == 0)
  auto fracEqualOneHalfCond = tosa::LogicalAndOp::create(
      rewriter, op->getLoc(), boolTy, fracEqualOneHalf.getResult(),
      floorInputEven.getResult());

  // (frac < 0.5) || ((frac == 0.5) && (floor(input) % 2 == 0))
  auto floorResultCond = tosa::LogicalOrOp::create(
      rewriter, op->getLoc(), boolTy, fracLtOneHalf.getResult(),
      fracEqualOneHalfCond.getResult());

  auto selectOp = tosa::SelectOp::create(
      rewriter, op->getLoc(), resultTy, floorResultCond.getResult(),
      floorInput.getResult(), ceilInput.getResult());

  return selectOp.getResult();
}

} // namespace tosa
} // namespace mlir