regular_tile_access_iterator_tensor_op.h

Go to the documentation of this file.

/***************************************************************************************************
 * Copyright (c) 2017-2019, NVIDIA CORPORATION.  All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without modification, are permitted
 * provided that the following conditions are met:
 *     * Redistributions of source code must retain the above copyright notice, this list of
 *       conditions and the following disclaimer.
 *     * Redistributions in binary form must reproduce the above copyright notice, this list of
 *       conditions and the following disclaimer in the documentation and/or other materials
 *       provided with the distribution.
 *     * Neither the name of the NVIDIA CORPORATION nor the names of its contributors may be used
 *       to endorse or promote products derived from this software without specific prior written
 *       permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR
 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
 * FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL NVIDIA CORPORATION BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
 * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
 * STRICT LIABILITY, OR TOR (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 *
 **************************************************************************************************/
#pragma once

#include "cutlass/array.h"
#include "cutlass/cutlass.h"
#include "cutlass/layout/pitch_linear.h"
#include "cutlass/layout/tensor_op_multiplicand_sm75.h"
#include "cutlass/matrix_coord.h"
#include "cutlass/matrix_shape.h"
#include "cutlass/tensor_ref.h"
#include "cutlass/transform/threadblock/regular_tile_access_iterator.h"


namespace cutlass {
namespace transform {
namespace threadblock {


template <typename Shape_, typename Element_, int AdvanceRank,
          typename ThreadMap_, int Alignment>
class RegularTileAccessIterator<
    Shape_, Element_,
    layout::TensorOpMultiplicandCongruous<sizeof_bits<Element_>::value,
                                          int(128 / sizeof(Element_))>,
    AdvanceRank, ThreadMap_, Alignment> {
 public:
  static_assert(
      AdvanceRank == 0 || AdvanceRank == 1,
      "Specialization for pitch-linear iterator may along advance along the "
      "contiguous(rank=0) or strided(rank=1) dimension.");

  using Shape = Shape_;
  using Element = Element_;
  using Layout =
      layout::TensorOpMultiplicandCongruous<sizeof_bits<Element_>::value,
                                            int(128 / sizeof(Element_))>;
  static int const kAdvanceRank = AdvanceRank;
  static int const kAlignment = Alignment;

  using Index = typename Layout::Index;
  using LongIndex = typename Layout::LongIndex;

  using TensorRef = TensorRef<Element, Layout>;
  using TensorCoord = typename Layout::TensorCoord;

  using ThreadMap = ThreadMap_;

  struct Detail {
    static int const kAccessSizeInBits = 128;

    static_assert(sizeof_bits<Element_>::value *
                          ThreadMap::kElementsPerAccess ==
                      kAccessSizeInBits,
                  "This iterator requires a policy whose access size is 128bs");

    static int const kPointerCount =
        (ThreadMap::Iterations::kStrided > 1 ? 2 : 1);
  };

  using AccessType = Array<Element, Layout::kElementsPerAccess>;

 private:
  //
  // Data members
  //

  Index stride_;

  AccessType *pointer_[Detail::kPointerCount];

  Index byte_offset_;

  int iteration_contiguous_;

  int iteration_strided_;

 public:
  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator(TensorRef ref,  
                            int thread_id   
                            )
      : stride_(ref.stride(0) / Layout::kElementsPerAccess),
        byte_offset_(0) {
    layout::PitchLinearCoord thread_offset_base =
        ThreadMap::initial_offset(thread_id);

    CUTLASS_PRAGMA_UNROLL
    for (int i = 0; i < Detail::kPointerCount; ++i) {
      // This is the offset of a thread within a threadblock tile for a specific
      // pointer (units of elements)
      layout::PitchLinearCoord thread_offset_in_threadblock_tile =
          thread_offset_base +
          layout::PitchLinearCoord{
              0, ThreadMap::Detail::WarpThreadArrangement::kStrided * i};

      // initialize pointer
      pointer_[i] = reinterpret_cast<AccessType *>(
          ref.data() + ref.offset(thread_offset_in_threadblock_tile));
    }

    set_iteration_index(0);
  }

  CUTLASS_HOST_DEVICE
  void set_iteration_index(int index) {
    iteration_contiguous_ = index % ThreadMap::Iterations::kContiguous;
    iteration_strided_ = index / ThreadMap::Iterations::kContiguous;
  }

  CUTLASS_HOST_DEVICE
  void add_pointer_offset(LongIndex pointer_offset) {
    byte_offset_ += pointer_offset * sizeof(Element);
  }

  CUTLASS_HOST_DEVICE
  AccessType *get() const {
    AccessType *access_ptr = pointer_[iteration_strided_ & 1];
    int stride_idx = (iteration_strided_ & ~1);

    int access_offset = stride_idx * ThreadMap::Delta::kStrided * stride_ +
                        iteration_contiguous_ * ThreadMap::Delta::kContiguous /
                            ThreadMap::kElementsPerAccess;

    char *access_byte_ptr =
        reinterpret_cast<char *>(access_ptr + access_offset);
    return reinterpret_cast<AccessType *>(access_byte_ptr + byte_offset_);
  }

  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator &operator++() {
    ++iteration_contiguous_;

    if (iteration_contiguous_ < ThreadMap::Iterations::kContiguous)
      return *this;

    // Enter here only if (iteration_contiguous_ ==
    // ThreadMap::Iteration::kContiguous)
    iteration_contiguous_ = 0;
    ++iteration_strided_;

    if (iteration_strided_ < ThreadMap::Iterations::kStrided) {
      return *this;
    }

    // Enter here only if (iteration_strided_ == ThreadMap::Iteration::kStrided)
    // which means we enter the next tile.
    iteration_strided_ = 0;

    return *this;
  }

  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator operator++(int) {
    RegularTileAccessIterator prev(*this);
    this->operator++();

    return prev;
  }

  CUTLASS_DEVICE
  void add_tile_offset(TensorCoord const &coord) {
    add_pointer_offset(coord.contiguous() * Shape::kContiguous +
                       coord.strided() * Shape::kStrided * stride_ *
                           Layout::kElementsPerAccess);
  }
};


template <typename Shape_, typename Element_, int AdvanceRank,
          typename ThreadMap_, int Alignment>
class RegularTileAccessIterator<
    Shape_, Element_,
    layout::ColumnMajorTensorOpMultiplicandCongruous<
        sizeof_bits<Element_>::value, int(128 / sizeof(Element_))>,
    AdvanceRank, ThreadMap_, Alignment> {
 public:
  static_assert(
      AdvanceRank == 0 || AdvanceRank == 1,
      "Specialization for column-major iterator may along advance along the "
      "columns(rank=0) or rows(rank=1) dimension.");

  using Shape = Shape_;
  using Element = Element_;
  using Layout = layout::ColumnMajorTensorOpMultiplicandCongruous<
      sizeof_bits<Element_>::value, int(128 / sizeof(Element_))>;
  static int const kAdvanceRank = AdvanceRank;
  static int const kAlignment = Alignment;

  using Index = typename Layout::Index;
  using LongIndex = typename Layout::LongIndex;

  using TensorRef = TensorRef<Element, Layout>;
  using TensorCoord = typename Layout::TensorCoord;

  using ThreadMap = ThreadMap_;

  using UnderlyingIterator = RegularTileAccessIterator<
      layout::PitchLinearShape<Shape::kRow, Shape::kColumn>, Element,
      layout::TensorOpMultiplicandCongruous<sizeof_bits<Element_>::value,
                                            int(128 / sizeof(Element_))>,
      (kAdvanceRank == 0 ? 0 : 1), ThreadMap_>;

  using AccessType = typename UnderlyingIterator::AccessType;

 private:
  UnderlyingIterator iterator_;

 public:
  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator(TensorRef ref,  
                            int thread_id   
                            )
      : iterator_({ref.data(), ref.stride()}, thread_id) {}

  CUTLASS_HOST_DEVICE
  void set_iteration_index(int index) { iterator_.set_iteration_index(index); }

  CUTLASS_HOST_DEVICE
  void add_pointer_offset(LongIndex pointer_offset) {
    iterator_.add_pointer_offset(pointer_offset);
  }

  CUTLASS_HOST_DEVICE
  AccessType *get() const {
    return reinterpret_cast<AccessType *>(iterator_.get());
  }

  CUTLASS_DEVICE
  void add_tile_offset(TensorCoord const &coord) {
    iterator_.add_tile_offset({coord.row(), coord.column()});
  }

  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator &operator++() {
    ++iterator_;
    return *this;
  }

  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator operator++(int) {
    RegularTileAccessIterator prev(*this);
    ++iterator_;

    return prev;
  }
};


template <typename Shape_, typename Element_, int AdvanceRank,
          typename ThreadMap_, int Alignment>
class RegularTileAccessIterator<
    Shape_, Element_,
    layout::RowMajorTensorOpMultiplicandCongruous<sizeof_bits<Element_>::value,
                                                  int(128 / sizeof(Element_))>,
    AdvanceRank, ThreadMap_, Alignment> {
 public:
  static_assert(
      AdvanceRank == 0 || AdvanceRank == 1,
      "Specialization for row-major iterator may along advance along the "
      "columns(rank=0) or rows(rank=1) dimension.");

  using Shape = Shape_;
  using Element = Element_;
  using Layout = layout::RowMajorTensorOpMultiplicandCongruous<
      sizeof_bits<Element_>::value, int(128 / sizeof(Element_))>;
  static int const kAdvanceRank = AdvanceRank;
  static int const kAlignment = Alignment;

  using Index = typename Layout::Index;
  using LongIndex = typename Layout::LongIndex;

  using TensorRef = TensorRef<Element, Layout>;
  using TensorCoord = typename Layout::TensorCoord;

  using ThreadMap = ThreadMap_;

  using UnderlyingIterator = RegularTileAccessIterator<
      layout::PitchLinearShape<Shape::kColumn, Shape::kRow>, Element,
      layout::TensorOpMultiplicandCongruous<sizeof_bits<Element_>::value,
                                            int(128 / sizeof(Element_))>,
      (kAdvanceRank == 0 ? 1 : 0), ThreadMap_>;

  using AccessType = typename UnderlyingIterator::AccessType;

 private:
  UnderlyingIterator iterator_;

 public:
  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator(TensorRef ref,  
                            int thread_id   
                            )
      : iterator_({ref.data(), ref.stride()}, thread_id) {}

  CUTLASS_HOST_DEVICE
  void set_iteration_index(int index) { iterator_.set_iteration_index(index); }

  CUTLASS_HOST_DEVICE
  void add_pointer_offset(LongIndex pointer_offset) {
    iterator_.add_pointer_offset(pointer_offset);
  }

  CUTLASS_HOST_DEVICE
  AccessType *get() const {
    return reinterpret_cast<AccessType *>(iterator_.get());
  }

  CUTLASS_DEVICE
  void add_tile_offset(TensorCoord const &coord) {
    iterator_.add_tile_offset({coord.column(), coord.row()});
  }

  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator &operator++() {
    ++iterator_;
    return *this;
  }

  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator operator++(int) {
    RegularTileAccessIterator prev(*this);
    ++iterator_;

    return prev;
  }
};


template <typename Shape_, typename Element_, int AdvanceRank,
          typename ThreadMap_, int Alignment, int Crosswise>
class RegularTileAccessIterator<Shape_, Element_,
                                layout::TensorOpMultiplicandCrosswise<
                                    sizeof_bits<Element_>::value, Crosswise>,
                                AdvanceRank, ThreadMap_, Alignment> {
 public:
  static_assert(
      AdvanceRank == 0 || AdvanceRank == 1,
      "Specialization for pitch-linear iterator may along advance along the "
      "contiguous(rank=0) or strided(rank=1) dimension.");

  using Shape = Shape_;
  using Element = Element_;
  using Layout =
      layout::TensorOpMultiplicandCrosswise<sizeof_bits<Element_>::value,
                                            Crosswise>;
  static int const kAdvanceRank = AdvanceRank;
  static int const kAlignment = Alignment;
  static int const kCrosswise = Crosswise;

  using Index = typename Layout::Index;
  using LongIndex = typename Layout::LongIndex;

  using TensorRef = TensorRef<Element, Layout>;
  using TensorCoord = typename Layout::TensorCoord;

  using ThreadMap = ThreadMap_;

  struct Detail {
    static int const kAccessSizeInBits = 128;

    static_assert(sizeof_bits<Element_>::value *
                          ThreadMap::kElementsPerAccess ==
                      kAccessSizeInBits,
                  "This iterator requires a policy whose access size is 128bs");

    static int const kPointerCount =
        (ThreadMap::Iterations::kStrided > 1 ? 2 : 1);
  };

  using AccessType = Array<Element, Layout::kElementsPerAccess>;

 private:
  //
  // Data members
  //

  int sections_;

  int sections_per_stage_;

  Index stride_;

  AccessType *pointer_[Detail::kPointerCount];

  Index byte_offset_;

  int iteration_contiguous_;

  int iteration_strided_;

 public:
  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator(TensorRef ref,  
                            int thread_id   
                            )
      : sections_(ref.stride(0) / kCrosswise),
        sections_per_stage_(Shape::kContiguous / kCrosswise),
        // stride_ = kCrosswise x sections_ x kFactor
        stride_(ref.stride(0) * Layout::kFactor / Layout::kElementsPerAccess),
        byte_offset_(0) {
    layout::PitchLinearCoord thread_offset_base =
        ThreadMap::initial_offset(thread_id);

    CUTLASS_PRAGMA_UNROLL
    for (int i = 0; i < Detail::kPointerCount; ++i) {
      // This is the offset of a thread within a threadblock tile for a specific
      // pointer (units of elements)
      layout::PitchLinearCoord thread_offset_in_threadblock_tile =
          thread_offset_base +
          layout::PitchLinearCoord{
              0, ThreadMap::Detail::WarpThreadArrangement::kStrided * i};
      // initialize pointer
      pointer_[i] = reinterpret_cast<AccessType *>(ref.data()) +
                    ref.offset(thread_offset_in_threadblock_tile) /
                        Layout::kElementsPerAccess;
    }

    set_iteration_index(0);
  }

  CUTLASS_HOST_DEVICE
  void set_iteration_index(int index) {
    iteration_contiguous_ = index % ThreadMap::Iterations::kContiguous;
    iteration_strided_ = index / ThreadMap::Iterations::kContiguous;
  }

  CUTLASS_HOST_DEVICE
  void add_pointer_offset(LongIndex pointer_offset) {
    byte_offset_ += pointer_offset * sizeof_bits<Element>::value / 8;
  }

  CUTLASS_HOST_DEVICE
  AccessType *get() const {
    AccessType *access_ptr = pointer_[iteration_strided_ & 1];
    int stride_idx = (iteration_strided_ & ~1);

    int access_offset =
        stride_idx * ThreadMap::Delta::kStrided * stride_ / Layout::kFactor +
        iteration_contiguous_ * ThreadMap::Delta::kContiguous /
            ThreadMap::kElementsPerAccess;
    char *access_byte_ptr =
        reinterpret_cast<char *>(access_ptr + access_offset);
    return reinterpret_cast<AccessType *>(access_byte_ptr + byte_offset_);
  }

  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator &operator++() {
    ++iteration_contiguous_;

    if (iteration_contiguous_ < ThreadMap::Iterations::kContiguous)
      return *this;

    // Enter here only if (iteration_contiguous_ ==
    // ThreadMap::Iteration::kContiguous)
    iteration_contiguous_ = 0;
    ++iteration_strided_;

    if (iteration_strided_ < ThreadMap::Iterations::kStrided) {
      return *this;
    }

    // Enter here only if (iteration_strided_ == ThreadMap::Iteration::kStrided)
    // which means we enter the next section.
    iteration_strided_ = 0;

    return *this;
  }

  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator operator++(int) {
    RegularTileAccessIterator prev(*this);
    this->operator++();

    return prev;
  }

  CUTLASS_DEVICE
  void add_tile_offset(TensorCoord const &coord) {
    add_pointer_offset(coord.contiguous() * sections_per_stage_ * stride_ *
                           ThreadMap::kElementsPerAccess / sections_ +
                       coord.strided() * Shape::kStrided * stride_ *
                           Layout::kElementsPerAccess);
  }
};


template <typename Shape_, typename Element_, int AdvanceRank,
          typename ThreadMap_, int Alignment, int Crosswise>
class RegularTileAccessIterator<
    Shape_, Element_,
    layout::ColumnMajorTensorOpMultiplicandCrosswise<
        sizeof_bits<Element_>::value, Crosswise>,
    AdvanceRank, ThreadMap_, Alignment> {
 public:
  static_assert(
      AdvanceRank == 0 || AdvanceRank == 1,
      "Specialization for column-major iterator may along advance along the "
      "columns(rank=0) or rows(rank=1) dimension.");

  using Shape = Shape_;
  using Element = Element_;
  using Layout = layout::ColumnMajorTensorOpMultiplicandCrosswise<
      sizeof_bits<Element_>::value, Crosswise>;
  static int const kAdvanceRank = AdvanceRank;
  static int const kAlignment = Alignment;

  using Index = typename Layout::Index;
  using LongIndex = typename Layout::LongIndex;

  using TensorRef = TensorRef<Element, Layout>;
  using TensorCoord = typename Layout::TensorCoord;

  using ThreadMap = ThreadMap_;

  using UnderlyingIterator = RegularTileAccessIterator<
      layout::PitchLinearShape<Shape::kRow, Shape::kColumn>, Element,
      layout::TensorOpMultiplicandCrosswise<sizeof_bits<Element_>::value,
                                            Crosswise>,
      (kAdvanceRank == 0 ? 0 : 1), ThreadMap_>;

  using AccessType = typename UnderlyingIterator::AccessType;

 private:
  UnderlyingIterator iterator_;

 public:
  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator(TensorRef ref,  
                            int thread_id   
                            )
      : iterator_({ref.data(), ref.stride()}, thread_id) {}

  CUTLASS_HOST_DEVICE
  void set_iteration_index(int index) { iterator_.set_iteration_index(index); }

  CUTLASS_HOST_DEVICE
  void add_pointer_offset(LongIndex pointer_offset) {
    iterator_.add_pointer_offset(pointer_offset);
  }

  CUTLASS_HOST_DEVICE
  AccessType *get() const {
    return reinterpret_cast<AccessType *>(iterator_.get());
  }

  CUTLASS_DEVICE
  void add_tile_offset(TensorCoord const &coord) {
    iterator_.add_tile_offset({coord.row(), coord.column()});
  }

  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator &operator++() {
    ++iterator_;
    return *this;
  }

  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator operator++(int) {
    RegularTileAccessIterator prev(*this);
    ++iterator_;

    return prev;
  }
};


template <typename Shape_, typename Element_, int AdvanceRank,
          typename ThreadMap_, int Alignment, int Crosswise>
class RegularTileAccessIterator<Shape_, Element_,
                                layout::RowMajorTensorOpMultiplicandCrosswise<
                                    sizeof_bits<Element_>::value, Crosswise>,
                                AdvanceRank, ThreadMap_, Alignment> {
 public:
  static_assert(
      AdvanceRank == 0 || AdvanceRank == 1,
      "Specialization for row-major iterator may along advance along the "
      "columns(rank=0) or rows(rank=1) dimension.");

  using Shape = Shape_;
  using Element = Element_;
  using Layout = layout::RowMajorTensorOpMultiplicandCrosswise<
      sizeof_bits<Element_>::value, Crosswise>;
  static int const kAdvanceRank = AdvanceRank;
  static int const kAlignment = Alignment;

  using Index = typename Layout::Index;
  using LongIndex = typename Layout::LongIndex;

  using TensorRef = TensorRef<Element, Layout>;
  using TensorCoord = typename Layout::TensorCoord;

  using ThreadMap = ThreadMap_;

  using UnderlyingIterator = RegularTileAccessIterator<
      layout::PitchLinearShape<Shape::kColumn, Shape::kRow>, Element,
      layout::TensorOpMultiplicandCrosswise<sizeof_bits<Element_>::value,
                                            Crosswise>,
      (kAdvanceRank == 0 ? 1 : 0), ThreadMap_>;

  using AccessType = typename UnderlyingIterator::AccessType;

 private:
  UnderlyingIterator iterator_;

 public:
  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator(TensorRef ref,  
                            int thread_id   
                            )
      : iterator_({ref.data(), ref.stride()}, thread_id) {}

  CUTLASS_HOST_DEVICE
  void set_iteration_index(int index) { iterator_.set_iteration_index(index); }

  CUTLASS_HOST_DEVICE
  void add_pointer_offset(LongIndex pointer_offset) {
    iterator_.add_pointer_offset(pointer_offset);
  }

  CUTLASS_HOST_DEVICE
  AccessType *get() const {
    return reinterpret_cast<AccessType *>(iterator_.get());
  }

  CUTLASS_DEVICE
  void add_tile_offset(TensorCoord const &coord) {
    iterator_.add_tile_offset({coord.column(), coord.row()});
  }

  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator &operator++() {
    ++iterator_;
    return *this;
  }

  CUTLASS_HOST_DEVICE
  RegularTileAccessIterator operator++(int) {
    RegularTileAccessIterator prev(*this);
    ++iterator_;

    return prev;
  }
};

}  // namespace threadblock
}  // namespace transform
}  // namespace cutlass