cuda-cccl 0.1.3.2.0.dev438__cp312-cp312-manylinux_2_24_aarch64.whl → 0.3.1__cp312-cp312-manylinux_2_24_aarch64.whl

cuda/cccl/cooperative/experimental/warp/__init__.py

@@ -1,9 +0,0 @@
-# Copyright (c) 2024, NVIDIA CORPORATION & AFFILIATES. ALL RIGHTS RESERVED.
-#
-# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
-
-from cuda.cccl.cooperative.experimental.warp._warp_merge_sort import merge_sort_keys
-from cuda.cccl.cooperative.experimental.warp._warp_reduce import reduce, sum
-from cuda.cccl.cooperative.experimental.warp._warp_scan import exclusive_sum
-
-__all__ = ["exclusive_sum", "reduce", "sum", "merge_sort_keys"]

cuda/cccl/headers/include/cub/device/dispatch/dispatch_advance_iterators.cuh

@@ -1,111 +0,0 @@
-// SPDX-FileCopyrightText: Copyright (c) 2025, NVIDIA CORPORATION. All rights reserved.
-// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
-
-#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 <cuda/std/__type_traits/integral_constant.h>
-#include <cuda/std/__type_traits/void_t.h>
-#include <cuda/std/__utility/declval.h>
-
-CUB_NAMESPACE_BEGIN
-
-namespace detail
-{
-template <typename T, typename U, typename = void>
-struct has_plus_operator : ::cuda::std::false_type
-{};
-
-template <typename T, typename U>
-struct has_plus_operator<T, U, ::cuda::std::void_t<decltype(::cuda::std::declval<T>() + ::cuda::std::declval<U>())>>
-    : ::cuda::std::true_type
-{};
-
-template <typename T, typename U>
-constexpr bool has_plus_operator_v = has_plus_operator<T, U>::value;
-
-// Helper function that advances a given iterator only if it supports being advanced by the given offset
-template <typename IteratorT, typename OffsetT>
-CUB_RUNTIME_FUNCTION _CCCL_VISIBILITY_HIDDEN _CCCL_FORCEINLINE IteratorT
-advance_iterators_if_supported(IteratorT iter, [[maybe_unused]] OffsetT offset)
-{
-  if constexpr (has_plus_operator_v<IteratorT, OffsetT>)
-  {
-    // If operator+ is valid, advance the iterator.
-    return iter + offset;
-  }
-  else
-  {
-    // Otherwise, return iter unmodified.
-    return iter;
-  }
-}
-
-template <typename T, typename U, typename = void>
-struct has_add_assign_operator : ::cuda::std::false_type
-{};
-
-template <typename T, typename U>
-struct has_add_assign_operator<T,
-                               U,
-                               ::cuda::std::void_t<decltype(::cuda::std::declval<T&>() += ::cuda::std::declval<U>())>>
-    : ::cuda::std::true_type
-{};
-
-template <typename T, typename U>
-constexpr bool has_add_assign_operator_v = has_add_assign_operator<T, U>::value;
-
-// Helper function that advances a given iterator only if it supports being advanced by the given offset
-template <typename IteratorT, typename OffsetT>
-CUB_RUNTIME_FUNCTION _CCCL_VISIBILITY_HIDDEN _CCCL_FORCEINLINE void
-advance_iterators_inplace_if_supported(IteratorT& iter, [[maybe_unused]] OffsetT offset)
-{
-  if constexpr (has_add_assign_operator_v<IteratorT, OffsetT>)
-  {
-    // If operator+ is valid, advance the iterator.
-    iter += offset;
-  }
-}
-
-// Helper function that checks whether all of the given iterators support the + operator with the given offset
-template <typename OffsetT, typename... Iterators>
-CUB_RUNTIME_FUNCTION _CCCL_VISIBILITY_HIDDEN _CCCL_FORCEINLINE bool
-all_iterators_support_plus_operator(OffsetT /*offset*/, Iterators... /*iters*/)
-{
-  if constexpr ((has_plus_operator_v<Iterators, OffsetT> && ...))
-  {
-    return true;
-  }
-  else
-  {
-    return false;
-  }
-}
-
-// Helper function that checks whether all of the given iterators support the + operator with the given offset
-template <typename OffsetT, typename... Iterators>
-CUB_RUNTIME_FUNCTION _CCCL_VISIBILITY_HIDDEN _CCCL_FORCEINLINE bool
-all_iterators_support_add_assign_operator(OffsetT /*offset*/, Iterators... /*iters*/)
-{
-  if constexpr ((has_add_assign_operator_v<Iterators, OffsetT> && ...))
-  {
-    return true;
-  }
-  else
-  {
-    return false;
-  }
-}
-
-} // namespace detail
-
-CUB_NAMESPACE_END

cuda/cccl/headers/include/cuda/std/__cuda/ensure_current_device.h

@@ -1,72 +0,0 @@
-//===----------------------------------------------------------------------===//
-//
-// 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) 2024 NVIDIA CORPORATION & AFFILIATES.
-//
-//===----------------------------------------------------------------------===//
-
-#ifndef _CUDA__STD__CUDA_ENSURE_CURRENT_DEVICE_H
-#define _CUDA__STD__CUDA_ENSURE_CURRENT_DEVICE_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
-
-#if _CCCL_HAS_CTK()
-
-#  include <cuda/std/__cuda/api_wrapper.h>
-
-#  include <cuda/std/__cccl/prologue.h>
-
-_CCCL_BEGIN_NAMESPACE_CUDA
-
-#  if !_CCCL_COMPILER(NVRTC)
-
-//! @brief `__ensure_current_device` is a simple helper that the current device is set to the right one.
-//! Only changes the current device if the target device is not the current one
-struct __ensure_current_device
-{
-  int __target_device_   = 0;
-  int __original_device_ = 0;
-
-  //! @brief Queries the current device and if that is different than \p __target_device sets the current device to
-  //! \p __target_device
-  __ensure_current_device(const int __target_device)
-      : __target_device_(__target_device)
-  {
-    _CCCL_TRY_CUDA_API(::cudaGetDevice, "Failed to query current device", &__original_device_);
-    if (__original_device_ != __target_device_)
-    {
-      _CCCL_TRY_CUDA_API(::cudaSetDevice, "Failed to set device", __target_device_);
-    }
-  }
-
-  //! @brief If the \p __original_device was not equal to \p __target_device sets the current device back to
-  //! \p __original_device
-  ~__ensure_current_device()
-  {
-    if (__original_device_ != __target_device_)
-    {
-      _CCCL_LOG_CUDA_API(::cudaSetDevice, "Failed to set device", __original_device_);
-    }
-  }
-};
-
-#  endif // !_CCCL_COMPILER(NVRTC)
-
-_CCCL_END_NAMESPACE_CUDA
-
-#  include <cuda/std/__cccl/epilogue.h>
-
-#endif // _CCCL_HAS_CTK()
-
-#endif //_CUDA__STD__CUDA_ENSURE_CURRENT_DEVICE_H

cuda/cccl/parallel/experimental/.gitignore

@@ -1,4 +0,0 @@
-# these are generated at build time depending on the CUDA
-# version we're building for:
-cu12/
-cu13/