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
93 lines
3.5 KiB
C++
93 lines
3.5 KiB
C++
//===----------------------------------------------------------------------===//
|
|
//
|
|
// Part of libcu++, the C++ Standard Library for your entire system,
|
|
// 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
|
|
// SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES.
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
#ifndef _CUDA___FWD_MDSPAN_H
|
|
#define _CUDA___FWD_MDSPAN_H
|
|
|
|
#include <cuda/std/detail/__config>
|
|
|
|
#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 <cuda/std/__cstddef/types.h>
|
|
#include <cuda/std/__limits/numeric_limits.h>
|
|
#include <cuda/std/__type_traits/make_signed.h>
|
|
#include <cuda/std/__utility/integer_sequence.h>
|
|
|
|
#include <cuda/std/__cccl/prologue.h>
|
|
|
|
_CCCL_BEGIN_NAMESPACE_CUDA
|
|
|
|
//! @brief Class to describe the strides of a multi-dimensional array layout.
|
|
//!
|
|
//! Similar to extents, but for strides. Supports both static (compile-time known)
|
|
//! and dynamic (runtime) stride values. Uses dynamic_stride as the tag for dynamic values.
|
|
//!
|
|
//! @tparam _OffsetType The signed integer type for stride values (supports negative strides)
|
|
//! @tparam _Strides... The stride values, where dynamic_stride indicates a runtime value
|
|
template <class _OffsetType, ::cuda::std::ptrdiff_t... _Strides>
|
|
class strides;
|
|
|
|
//! @brief Tag value indicating a dynamic stride (similar to dynamic_extent for extents)
|
|
inline constexpr ::cuda::std::ptrdiff_t dynamic_stride = (::cuda::std::numeric_limits<::cuda::std::ptrdiff_t>::min)();
|
|
|
|
namespace __strides_detail
|
|
{
|
|
template <class _OffsetType, class _Seq>
|
|
struct __make_dstrides_impl;
|
|
|
|
template <class _OffsetType, ::cuda::std::ptrdiff_t... _Idx>
|
|
struct __make_dstrides_impl<_OffsetType, ::cuda::std::integer_sequence<::cuda::std::ptrdiff_t, _Idx...>>
|
|
{
|
|
using type = strides<_OffsetType, ((void) _Idx, dynamic_stride)...>;
|
|
};
|
|
} // namespace __strides_detail
|
|
|
|
//! @brief Alias template for strides with all dynamic stride values
|
|
template <class _OffsetType, ::cuda::std::size_t _Rank>
|
|
using dstrides = typename __strides_detail::
|
|
__make_dstrides_impl<_OffsetType, ::cuda::std::make_integer_sequence<::cuda::std::ptrdiff_t, _Rank>>::type;
|
|
|
|
template <::cuda::std::size_t _Rank, class _OffsetType = ::cuda::std::ptrdiff_t>
|
|
using steps = dstrides<_OffsetType, _Rank>;
|
|
|
|
template <class _Tp>
|
|
inline constexpr bool __is_cuda_strides_v = false;
|
|
|
|
template <class _OffsetType, ::cuda::std::ptrdiff_t... _Strides>
|
|
inline constexpr bool __is_cuda_strides_v<strides<_OffsetType, _Strides...>> = true;
|
|
|
|
//! @brief Layout policy with relaxed stride mapping that supports negative strides and offsets.
|
|
//!
|
|
//! Unlike `layout_stride`, this layout allows:
|
|
//! - Negative strides (for reverse iteration)
|
|
//! - Zero strides (for broadcasting)
|
|
//! - A base offset (to accommodate negative strides)
|
|
//!
|
|
//! @note This layout is NOT always unique, exhaustive, or strided in the standard sense.
|
|
struct layout_stride_relaxed
|
|
{
|
|
template <class _Extents,
|
|
class _Strides = dstrides<::cuda::std::make_signed_t<typename _Extents::index_type>, _Extents::rank()>,
|
|
class _OffsetType = ::cuda::std::ptrdiff_t>
|
|
class mapping;
|
|
};
|
|
|
|
_CCCL_END_NAMESPACE_CUDA
|
|
|
|
#include <cuda/std/__cccl/epilogue.h>
|
|
|
|
#endif // _CUDA___FWD_MDSPAN_H
|