cub/warp/warp_exchange.cuh

File members: cub/warp/warp_exchange.cuh

/******************************************************************************
 * Copyright (c) 2011-2021, 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 TORT
 * (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 <cub/config.cuh>

#if defined(_CCCL_IMPLICIT_SYSTEM_HEADER_GCC)
#  pragma GCC system_header
#elif defined(_CCCL_IMPLICIT_SYSTEM_HEADER_CLANG)
#  pragma clang system_header
#elif defined(_CCCL_IMPLICIT_SYSTEM_HEADER_MSVC)
#  pragma system_header
#endif // no system header

#include <cub/util_ptx.cuh>
#include <cub/util_type.cuh>
#include <cub/warp/specializations/warp_exchange_shfl.cuh>
#include <cub/warp/specializations/warp_exchange_smem.cuh>

#include <cuda/std/type_traits>

CUB_NAMESPACE_BEGIN

enum WarpExchangeAlgorithm
{
  WARP_EXCHANGE_SMEM,
  WARP_EXCHANGE_SHUFFLE,
};

namespace detail
{
template <typename InputT, int ITEMS_PER_THREAD, int LOGICAL_WARP_THREADS, WarpExchangeAlgorithm WARP_EXCHANGE_ALGORITHM>
using InternalWarpExchangeImpl =
  ::cuda::std::_If<WARP_EXCHANGE_ALGORITHM == WARP_EXCHANGE_SMEM,
                   WarpExchangeSmem<InputT, ITEMS_PER_THREAD, LOGICAL_WARP_THREADS>,
                   WarpExchangeShfl<InputT, ITEMS_PER_THREAD, LOGICAL_WARP_THREADS>>;
} // namespace detail

template <typename InputT,
          int ITEMS_PER_THREAD,
          int LOGICAL_WARP_THREADS                      = CUB_PTX_WARP_THREADS,
          int LEGACY_PTX_ARCH                           = 0,
          WarpExchangeAlgorithm WARP_EXCHANGE_ALGORITHM = WARP_EXCHANGE_SMEM>
class WarpExchange
    : private detail::InternalWarpExchangeImpl<InputT, ITEMS_PER_THREAD, LOGICAL_WARP_THREADS, WARP_EXCHANGE_ALGORITHM>
{
  using InternalWarpExchange =
    detail::InternalWarpExchangeImpl<InputT, ITEMS_PER_THREAD, LOGICAL_WARP_THREADS, WARP_EXCHANGE_ALGORITHM>;

public:
  using TempStorage = typename InternalWarpExchange::TempStorage;

  WarpExchange() = delete;

  explicit _CCCL_DEVICE _CCCL_FORCEINLINE WarpExchange(TempStorage& temp_storage)
      : InternalWarpExchange(temp_storage)
  {}

  template <typename OutputT>
  _CCCL_DEVICE _CCCL_FORCEINLINE void
  BlockedToStriped(const InputT (&input_items)[ITEMS_PER_THREAD], OutputT (&output_items)[ITEMS_PER_THREAD])
  {
    InternalWarpExchange::BlockedToStriped(input_items, output_items);
  }

  template <typename OutputT>
  _CCCL_DEVICE _CCCL_FORCEINLINE void
  StripedToBlocked(const InputT (&input_items)[ITEMS_PER_THREAD], OutputT (&output_items)[ITEMS_PER_THREAD])
  {
    InternalWarpExchange::StripedToBlocked(input_items, output_items);
  }

  template <typename OffsetT>
  _CCCL_DEVICE _CCCL_FORCEINLINE void
  ScatterToStriped(InputT (&items)[ITEMS_PER_THREAD], OffsetT (&ranks)[ITEMS_PER_THREAD])
  {
    InternalWarpExchange::ScatterToStriped(items, ranks);
  }

  template <typename OutputT, typename OffsetT>
  _CCCL_DEVICE _CCCL_FORCEINLINE void ScatterToStriped(
    const InputT (&input_items)[ITEMS_PER_THREAD],
    OutputT (&output_items)[ITEMS_PER_THREAD],
    OffsetT (&ranks)[ITEMS_PER_THREAD])
  {
    InternalWarpExchange::ScatterToStriped(input_items, output_items, ranks);
  }

};

CUB_NAMESPACE_END