/home/runner/work/cccl/cccl/cub/cub/device/device_merge_sort.cuh
File members: /home/runner/work/cccl/cccl/cub/cub/device/device_merge_sort.cuh
/******************************************************************************
* Copyright (c) 2011-2022, 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/detail/choose_offset.cuh>
#include <cub/detail/nvtx.cuh>
#include <cub/device/dispatch/dispatch_merge_sort.cuh>
#include <cub/util_deprecated.cuh>
#include <cub/util_namespace.cuh>
CUB_NAMESPACE_BEGIN
struct DeviceMergeSort
{
private:
// Name reported for NVTX ranges
_CCCL_HOST_DEVICE static constexpr auto GetName() -> const char*
{
return "cub::DeviceMergeSort";
}
// Internal version without NVTX range
template <typename KeyIteratorT, typename ValueIteratorT, typename OffsetT, typename CompareOpT>
CUB_RUNTIME_FUNCTION static cudaError_t SortPairsNoNVTX(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyIteratorT d_keys,
ValueIteratorT d_items,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream = 0)
{
using PromotedOffsetT = detail::promote_small_offset_t<OffsetT>;
using DispatchMergeSortT =
DispatchMergeSort<KeyIteratorT, ValueIteratorT, KeyIteratorT, ValueIteratorT, PromotedOffsetT, CompareOpT>;
return DispatchMergeSortT::Dispatch(
d_temp_storage, temp_storage_bytes, d_keys, d_items, d_keys, d_items, num_items, compare_op, stream);
}
public:
template <typename KeyIteratorT, typename ValueIteratorT, typename OffsetT, typename CompareOpT>
CUB_RUNTIME_FUNCTION static cudaError_t SortPairs(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyIteratorT d_keys,
ValueIteratorT d_items,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream = 0)
{
CUB_DETAIL_NVTX_RANGE_SCOPE_IF(d_temp_storage, GetName());
return SortPairsNoNVTX(d_temp_storage, temp_storage_bytes, d_keys, d_items, num_items, compare_op, stream);
}
template <typename KeyIteratorT, typename ValueIteratorT, typename OffsetT, typename CompareOpT>
CUB_DETAIL_RUNTIME_DEBUG_SYNC_IS_NOT_SUPPORTED CUB_RUNTIME_FUNCTION static cudaError_t SortPairs(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyIteratorT d_keys,
ValueIteratorT d_items,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream,
bool debug_synchronous)
{
CUB_DETAIL_RUNTIME_DEBUG_SYNC_USAGE_LOG
return SortPairs<KeyIteratorT, ValueIteratorT, OffsetT, CompareOpT>(
d_temp_storage, temp_storage_bytes, d_keys, d_items, num_items, compare_op, stream);
}
template <typename KeyInputIteratorT,
typename ValueInputIteratorT,
typename KeyIteratorT,
typename ValueIteratorT,
typename OffsetT,
typename CompareOpT>
CUB_RUNTIME_FUNCTION static cudaError_t SortPairsCopy(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyInputIteratorT d_input_keys,
ValueInputIteratorT d_input_items,
KeyIteratorT d_output_keys,
ValueIteratorT d_output_items,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream = 0)
{
CUB_DETAIL_NVTX_RANGE_SCOPE_IF(d_temp_storage, GetName());
using PromotedOffsetT = detail::promote_small_offset_t<OffsetT>;
using DispatchMergeSortT =
DispatchMergeSort<KeyInputIteratorT, ValueInputIteratorT, KeyIteratorT, ValueIteratorT, PromotedOffsetT, CompareOpT>;
return DispatchMergeSortT::Dispatch(
d_temp_storage,
temp_storage_bytes,
d_input_keys,
d_input_items,
d_output_keys,
d_output_items,
num_items,
compare_op,
stream);
}
template <typename KeyInputIteratorT,
typename ValueInputIteratorT,
typename KeyIteratorT,
typename ValueIteratorT,
typename OffsetT,
typename CompareOpT>
CUB_DETAIL_RUNTIME_DEBUG_SYNC_IS_NOT_SUPPORTED CUB_RUNTIME_FUNCTION static cudaError_t SortPairsCopy(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyInputIteratorT d_input_keys,
ValueInputIteratorT d_input_items,
KeyIteratorT d_output_keys,
ValueIteratorT d_output_items,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream,
bool debug_synchronous)
{
CUB_DETAIL_RUNTIME_DEBUG_SYNC_USAGE_LOG
return SortPairsCopy<KeyInputIteratorT, ValueInputIteratorT, KeyIteratorT, ValueIteratorT, OffsetT, CompareOpT>(
d_temp_storage,
temp_storage_bytes,
d_input_keys,
d_input_items,
d_output_keys,
d_output_items,
num_items,
compare_op,
stream);
}
private:
// Internal version without NVTX range
template <typename KeyIteratorT, typename OffsetT, typename CompareOpT>
CUB_RUNTIME_FUNCTION static cudaError_t SortKeysNoNVTX(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyIteratorT d_keys,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream = 0)
{
using PromotedOffsetT = detail::promote_small_offset_t<OffsetT>;
using DispatchMergeSortT =
DispatchMergeSort<KeyIteratorT, NullType*, KeyIteratorT, NullType*, PromotedOffsetT, CompareOpT>;
return DispatchMergeSortT::Dispatch(
d_temp_storage,
temp_storage_bytes,
d_keys,
static_cast<NullType*>(nullptr),
d_keys,
static_cast<NullType*>(nullptr),
num_items,
compare_op,
stream);
}
public:
template <typename KeyIteratorT, typename OffsetT, typename CompareOpT>
CUB_RUNTIME_FUNCTION static cudaError_t SortKeys(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyIteratorT d_keys,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream = 0)
{
CUB_DETAIL_NVTX_RANGE_SCOPE_IF(d_temp_storage, GetName());
return SortKeysNoNVTX(d_temp_storage, temp_storage_bytes, d_keys, num_items, compare_op, stream);
}
template <typename KeyIteratorT, typename OffsetT, typename CompareOpT>
CUB_DETAIL_RUNTIME_DEBUG_SYNC_IS_NOT_SUPPORTED CUB_RUNTIME_FUNCTION static cudaError_t SortKeys(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyIteratorT d_keys,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream,
bool debug_synchronous)
{
CUB_DETAIL_RUNTIME_DEBUG_SYNC_USAGE_LOG
return SortKeys<KeyIteratorT, OffsetT, CompareOpT>(
d_temp_storage, temp_storage_bytes, d_keys, num_items, compare_op, stream);
}
private:
// Internal version without NVTX range
template <typename KeyInputIteratorT, typename KeyIteratorT, typename OffsetT, typename CompareOpT>
CUB_RUNTIME_FUNCTION static cudaError_t SortKeysCopyNoNVTX(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyInputIteratorT d_input_keys,
KeyIteratorT d_output_keys,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream = 0)
{
using PromotedOffsetT = detail::promote_small_offset_t<OffsetT>;
using DispatchMergeSortT =
DispatchMergeSort<KeyInputIteratorT, NullType*, KeyIteratorT, NullType*, PromotedOffsetT, CompareOpT>;
return DispatchMergeSortT::Dispatch(
d_temp_storage,
temp_storage_bytes,
d_input_keys,
static_cast<NullType*>(nullptr),
d_output_keys,
static_cast<NullType*>(nullptr),
num_items,
compare_op,
stream);
}
public:
template <typename KeyInputIteratorT, typename KeyIteratorT, typename OffsetT, typename CompareOpT>
CUB_RUNTIME_FUNCTION static cudaError_t SortKeysCopy(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyInputIteratorT d_input_keys,
KeyIteratorT d_output_keys,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream = 0)
{
CUB_DETAIL_NVTX_RANGE_SCOPE_IF(d_temp_storage, GetName());
return SortKeysCopyNoNVTX(
d_temp_storage, temp_storage_bytes, d_input_keys, d_output_keys, num_items, compare_op, stream);
}
template <typename KeyInputIteratorT, typename KeyIteratorT, typename OffsetT, typename CompareOpT>
CUB_DETAIL_RUNTIME_DEBUG_SYNC_IS_NOT_SUPPORTED CUB_RUNTIME_FUNCTION static cudaError_t SortKeysCopy(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyInputIteratorT d_input_keys,
KeyIteratorT d_output_keys,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream,
bool debug_synchronous)
{
CUB_DETAIL_RUNTIME_DEBUG_SYNC_USAGE_LOG
return SortKeysCopy<KeyInputIteratorT, KeyIteratorT, OffsetT, CompareOpT>(
d_temp_storage, temp_storage_bytes, d_input_keys, d_output_keys, num_items, compare_op, stream);
}
template <typename KeyIteratorT, typename ValueIteratorT, typename OffsetT, typename CompareOpT>
CUB_RUNTIME_FUNCTION static cudaError_t StableSortPairs(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyIteratorT d_keys,
ValueIteratorT d_items,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream = 0)
{
CUB_DETAIL_NVTX_RANGE_SCOPE_IF(d_temp_storage, GetName());
using PromotedOffsetT = detail::promote_small_offset_t<OffsetT>;
return SortPairsNoNVTX<KeyIteratorT, ValueIteratorT, PromotedOffsetT, CompareOpT>(
d_temp_storage, temp_storage_bytes, d_keys, d_items, num_items, compare_op, stream);
}
template <typename KeyIteratorT, typename ValueIteratorT, typename OffsetT, typename CompareOpT>
CUB_DETAIL_RUNTIME_DEBUG_SYNC_IS_NOT_SUPPORTED CUB_RUNTIME_FUNCTION static cudaError_t StableSortPairs(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyIteratorT d_keys,
ValueIteratorT d_items,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream,
bool debug_synchronous)
{
CUB_DETAIL_RUNTIME_DEBUG_SYNC_USAGE_LOG
return StableSortPairs<KeyIteratorT, ValueIteratorT, OffsetT, CompareOpT>(
d_temp_storage, temp_storage_bytes, d_keys, d_items, num_items, compare_op, stream);
}
template <typename KeyIteratorT, typename OffsetT, typename CompareOpT>
CUB_RUNTIME_FUNCTION static cudaError_t StableSortKeys(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyIteratorT d_keys,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream = 0)
{
CUB_DETAIL_NVTX_RANGE_SCOPE_IF(d_temp_storage, GetName());
using PromotedOffsetT = detail::promote_small_offset_t<OffsetT>;
return SortKeysNoNVTX<KeyIteratorT, PromotedOffsetT, CompareOpT>(
d_temp_storage, temp_storage_bytes, d_keys, num_items, compare_op, stream);
}
template <typename KeyIteratorT, typename OffsetT, typename CompareOpT>
CUB_DETAIL_RUNTIME_DEBUG_SYNC_IS_NOT_SUPPORTED CUB_RUNTIME_FUNCTION static cudaError_t StableSortKeys(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyIteratorT d_keys,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream,
bool debug_synchronous)
{
CUB_DETAIL_RUNTIME_DEBUG_SYNC_USAGE_LOG
return StableSortKeys<KeyIteratorT, OffsetT, CompareOpT>(
d_temp_storage, temp_storage_bytes, d_keys, num_items, compare_op, stream);
}
template <typename KeyInputIteratorT, typename KeyIteratorT, typename OffsetT, typename CompareOpT>
CUB_RUNTIME_FUNCTION static cudaError_t StableSortKeysCopy(
void* d_temp_storage,
std::size_t& temp_storage_bytes,
KeyInputIteratorT d_input_keys,
KeyIteratorT d_output_keys,
OffsetT num_items,
CompareOpT compare_op,
cudaStream_t stream = 0)
{
CUB_DETAIL_NVTX_RANGE_SCOPE_IF(d_temp_storage, GetName());
using PromotedOffsetT = detail::promote_small_offset_t<OffsetT>;
return SortKeysCopyNoNVTX<KeyInputIteratorT, KeyIteratorT, PromotedOffsetT, CompareOpT>(
d_temp_storage, temp_storage_bytes, d_input_keys, d_output_keys, num_items, compare_op, stream);
}
};
CUB_NAMESPACE_END