56fd68e7dd
CCCL (CUDA C++ Core Libraries) provides: - CUB: device/block/warp-level GPU primitives (reduce, scan, sort, topk) - Thrust: high-level parallel algorithms (transform_reduce, sort, scan) - libcudacxx: CUDA C++ standard library (atomics, barriers, memory) - cudax: experimental features (memory resources, allocators) - Tuning policies: per-SM hardware-specific algorithm parameters Competition optimization vectors mapped to CCCL: - Output TPS (83% weight): warp_reduce, block_reduce, device_topk - Input TPS (14% weight): device_scan, block_load, prefetch - Cache TPS (3% weight): prefix caching strategy patterns - Memory (0.9 util): pooled/cached/buddy allocators Source: https://github.com/NVIDIA/cccl (shallow clone, HEAD only) License: Apache-2.0
134 lines
5.2 KiB
C++
134 lines
5.2 KiB
C++
// SPDX-FileCopyrightText: Copyright (c) 2008-2013, NVIDIA Corporation. All rights reserved.
|
|
// SPDX-License-Identifier: Apache-2.0
|
|
|
|
#pragma once
|
|
|
|
#include <thrust/detail/config.h>
|
|
|
|
#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 <thrust/detail/copy.h>
|
|
#include <thrust/detail/nvtx_policy.h>
|
|
#include <thrust/system/detail/generic/select_system.h>
|
|
|
|
// Include all active backend system implementations (generic, sequential, host and device)
|
|
#include <thrust/system/detail/generic/copy.h>
|
|
#include <thrust/system/detail/sequential/copy.h>
|
|
#include __THRUST_HOST_SYSTEM_ALGORITH_DETAIL_HEADER_INCLUDE(copy.h)
|
|
#include __THRUST_DEVICE_SYSTEM_ALGORITH_DETAIL_HEADER_INCLUDE(copy.h)
|
|
|
|
// Some build systems need a hint to know which files we could include
|
|
#if 0
|
|
# include <thrust/system/cpp/detail/copy.h>
|
|
# include <thrust/system/cuda/detail/copy.h>
|
|
# include <thrust/system/omp/detail/copy.h>
|
|
# include <thrust/system/tbb/detail/copy.h>
|
|
#endif
|
|
|
|
THRUST_NAMESPACE_BEGIN
|
|
|
|
_CCCL_EXEC_CHECK_DISABLE
|
|
template <typename DerivedPolicy, typename InputIterator, typename OutputIterator>
|
|
_CCCL_HOST_DEVICE OutputIterator
|
|
copy(const thrust::detail::execution_policy_base<DerivedPolicy>& exec,
|
|
InputIterator first,
|
|
InputIterator last,
|
|
OutputIterator result)
|
|
{
|
|
_CCCL_NVTX_RANGE_SCOPE_IF(detail::should_enable_nvtx_for_policy<DerivedPolicy>(), "thrust::copy");
|
|
using thrust::system::detail::generic::copy;
|
|
return copy(thrust::detail::derived_cast(thrust::detail::strip_const(exec)), first, last, result);
|
|
} // end copy()
|
|
|
|
_CCCL_EXEC_CHECK_DISABLE
|
|
template <typename DerivedPolicy, typename InputIterator, typename Size, typename OutputIterator>
|
|
_CCCL_HOST_DEVICE OutputIterator copy_n(
|
|
const thrust::detail::execution_policy_base<DerivedPolicy>& exec, InputIterator first, Size n, OutputIterator result)
|
|
{
|
|
_CCCL_NVTX_RANGE_SCOPE_IF(detail::should_enable_nvtx_for_policy<DerivedPolicy>(), "thrust::copy_n");
|
|
using thrust::system::detail::generic::copy_n;
|
|
return copy_n(thrust::detail::derived_cast(thrust::detail::strip_const(exec)), first, n, result);
|
|
} // end copy_n()
|
|
|
|
namespace detail
|
|
{
|
|
_CCCL_EXEC_CHECK_DISABLE // because we might call e.g. std::ostream_iterator's constructor
|
|
template <typename System1, typename System2, typename InputIterator, typename OutputIterator>
|
|
_CCCL_HOST_DEVICE OutputIterator two_system_copy(
|
|
const thrust::execution_policy<System1>& system1,
|
|
const thrust::execution_policy<System2>& system2,
|
|
InputIterator first,
|
|
InputIterator last,
|
|
OutputIterator result)
|
|
{
|
|
_CCCL_NVTX_RANGE_SCOPE_IF(should_enable_nvtx_for_policy<System1>() || should_enable_nvtx_for_policy<System2>(),
|
|
"thrust::two_system_copy");
|
|
using thrust::system::detail::generic::select_system;
|
|
|
|
return thrust::copy(
|
|
select_system(thrust::detail::derived_cast(thrust::detail::strip_const(system1)),
|
|
thrust::detail::derived_cast(thrust::detail::strip_const(system2))),
|
|
first,
|
|
last,
|
|
result);
|
|
} // end two_system_copy()
|
|
|
|
_CCCL_EXEC_CHECK_DISABLE // because we might call e.g. std::ostream_iterator's constructor
|
|
template <typename System1, typename System2, typename InputIterator, typename Size, typename OutputIterator>
|
|
_CCCL_HOST_DEVICE OutputIterator two_system_copy_n(
|
|
const thrust::execution_policy<System1>& system1,
|
|
const thrust::execution_policy<System2>& system2,
|
|
InputIterator first,
|
|
Size n,
|
|
OutputIterator result)
|
|
{
|
|
_CCCL_NVTX_RANGE_SCOPE_IF(should_enable_nvtx_for_policy<System1>() || should_enable_nvtx_for_policy<System2>(),
|
|
"thrust::two_system_copy_n");
|
|
using thrust::system::detail::generic::select_system;
|
|
|
|
return thrust::copy_n(
|
|
select_system(thrust::detail::derived_cast(thrust::detail::strip_const(system1)),
|
|
thrust::detail::derived_cast(thrust::detail::strip_const(system2))),
|
|
first,
|
|
n,
|
|
result);
|
|
} // end two_system_copy_n()
|
|
} // namespace detail
|
|
|
|
template <typename InputIterator, typename OutputIterator>
|
|
OutputIterator copy(InputIterator first, InputIterator last, OutputIterator result)
|
|
{
|
|
using System1 = typename thrust::iterator_system<InputIterator>::type;
|
|
using System2 = typename thrust::iterator_system<OutputIterator>::type;
|
|
_CCCL_NVTX_RANGE_SCOPE_IF(
|
|
detail::should_enable_nvtx_for_policy<System1>() || detail::should_enable_nvtx_for_policy<System2>(),
|
|
"thrust::copy");
|
|
|
|
System1 system1;
|
|
System2 system2;
|
|
|
|
return thrust::detail::two_system_copy(system1, system2, first, last, result);
|
|
} // end copy()
|
|
|
|
template <typename InputIterator, typename Size, typename OutputIterator>
|
|
OutputIterator copy_n(InputIterator first, Size n, OutputIterator result)
|
|
{
|
|
using System1 = typename thrust::iterator_system<InputIterator>::type;
|
|
using System2 = typename thrust::iterator_system<OutputIterator>::type;
|
|
_CCCL_NVTX_RANGE_SCOPE_IF(
|
|
detail::should_enable_nvtx_for_policy<System1>() || detail::should_enable_nvtx_for_policy<System2>(),
|
|
"thrust::copy_n");
|
|
|
|
System1 system1;
|
|
System2 system2;
|
|
|
|
return thrust::detail::two_system_copy_n(system1, system2, first, n, result);
|
|
} // end copy_n()
|
|
|
|
THRUST_NAMESPACE_END
|