cuda-cccl 0.3.2__cp313-cp313-manylinux_2_24_aarch64.whl → 0.3.4__cp313-cp313-manylinux_2_24_aarch64.whl

cuda/cccl/headers/include/thrust/system/tbb/detail/remove.inl

@@ -1,87 +0,0 @@
-/*
- *  Copyright 2008-2013 NVIDIA Corporation
- *
- *  Licensed under the Apache License, Version 2.0 (the "License");
- *  you may not use this file except in compliance with the License.
- *  You may obtain a copy of the License at
- *
- *      http://www.apache.org/licenses/LICENSE-2.0
- *
- *  Unless required by applicable law or agreed to in writing, software
- *  distributed under the License is distributed on an "AS IS" BASIS,
- *  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- *  See the License for the specific language governing permissions and
- *  limitations under the License.
- */
-
-#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/system/detail/generic/remove.h>
-#include <thrust/system/tbb/detail/remove.h>
-
-THRUST_NAMESPACE_BEGIN
-namespace system
-{
-namespace tbb
-{
-namespace detail
-{
-
-template <typename DerivedPolicy, typename ForwardIterator, typename Predicate>
-ForwardIterator
-remove_if(execution_policy<DerivedPolicy>& exec, ForwardIterator first, ForwardIterator last, Predicate pred)
-{
-  // tbb prefers generic::remove_if to cpp::remove_if
-  return thrust::system::detail::generic::remove_if(exec, first, last, pred);
-}
-
-template <typename DerivedPolicy, typename ForwardIterator, typename InputIterator, typename Predicate>
-ForwardIterator remove_if(
-  execution_policy<DerivedPolicy>& exec,
-  ForwardIterator first,
-  ForwardIterator last,
-  InputIterator stencil,
-  Predicate pred)
-{
-  // tbb prefers generic::remove_if to cpp::remove_if
-  return thrust::system::detail::generic::remove_if(exec, first, last, stencil, pred);
-}
-
-template <typename DerivedPolicy, typename InputIterator, typename OutputIterator, typename Predicate>
-OutputIterator remove_copy_if(
-  execution_policy<DerivedPolicy>& exec, InputIterator first, InputIterator last, OutputIterator result, Predicate pred)
-{
-  // tbb prefers generic::remove_copy_if to cpp::remove_copy_if
-  return thrust::system::detail::generic::remove_copy_if(exec, first, last, result, pred);
-}
-
-template <typename DerivedPolicy,
-          typename InputIterator1,
-          typename InputIterator2,
-          typename OutputIterator,
-          typename Predicate>
-OutputIterator remove_copy_if(
-  execution_policy<DerivedPolicy>& exec,
-  InputIterator1 first,
-  InputIterator1 last,
-  InputIterator2 stencil,
-  OutputIterator result,
-  Predicate pred)
-{
-  // tbb prefers generic::remove_copy_if to cpp::remove_copy_if
-  return thrust::system::detail::generic::remove_copy_if(exec, first, last, stencil, result, pred);
-}
-
-} // end namespace detail
-} // end namespace tbb
-} // end namespace system
-THRUST_NAMESPACE_END

cuda/cccl/headers/include/thrust/system/tbb/detail/scan.inl

@@ -1,312 +0,0 @@
-/*
- *  Copyright 2008-2013 NVIDIA Corporation
- *
- *  Licensed under the Apache License, Version 2.0 (the "License");
- *  you may not use this file except in compliance with the License.
- *  You may obtain a copy of the License at
- *
- *      http://www.apache.org/licenses/LICENSE-2.0
- *
- *  Unless required by applicable law or agreed to in writing, software
- *  distributed under the License is distributed on an "AS IS" BASIS,
- *  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- *  See the License for the specific language governing permissions and
- *  limitations under the License.
- */
-
-#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/advance.h>
-#include <thrust/detail/function.h>
-#include <thrust/detail/type_traits.h>
-#include <thrust/distance.h>
-#include <thrust/iterator/iterator_traits.h>
-#include <thrust/system/tbb/detail/scan.h>
-
-#include <cuda/std/__functional/invoke.h>
-
-#include <tbb/blocked_range.h>
-#include <tbb/parallel_scan.h>
-
-THRUST_NAMESPACE_BEGIN
-namespace system
-{
-namespace tbb
-{
-namespace detail
-{
-namespace scan_detail
-{
-
-template <typename InputIterator, typename OutputIterator, typename BinaryFunction, typename ValueType, bool HasInit>
-struct inclusive_body
-{
-  InputIterator input;
-  OutputIterator output;
-  thrust::detail::wrapped_function<BinaryFunction, ValueType> binary_op;
-  ValueType sum;
-  bool first_call;
-
-  inclusive_body(InputIterator input, OutputIterator output, BinaryFunction binary_op, ValueType init)
-      : input(input)
-      , output(output)
-      , binary_op{binary_op}
-      , sum(init)
-      , first_call(true)
-  {}
-
-  inclusive_body(inclusive_body& b, ::tbb::split)
-      : input(b.input)
-      , output(b.output)
-      , binary_op{b.binary_op}
-      , sum(b.sum)
-      , first_call(true)
-  {}
-
-  template <typename Size>
-  void operator()(const ::tbb::blocked_range<Size>& r, ::tbb::pre_scan_tag)
-  {
-    InputIterator iter = input + r.begin();
-
-    ValueType temp = *iter;
-
-    ++iter;
-
-    for (Size i = r.begin() + 1; i != r.end(); ++i, ++iter)
-    {
-      temp = binary_op(temp, *iter);
-    }
-
-    if (first_call)
-    {
-      sum = temp;
-    }
-    else
-    {
-      sum = binary_op(sum, temp);
-    }
-
-    first_call = false;
-  }
-
-  template <typename Size>
-  void operator()(const ::tbb::blocked_range<Size>& r, ::tbb::final_scan_tag)
-  {
-    InputIterator iter1  = input + r.begin();
-    OutputIterator iter2 = output + r.begin();
-
-    if (first_call)
-    {
-      if constexpr (HasInit)
-      {
-        *iter2 = sum = binary_op(sum, *iter1);
-      }
-      else
-      {
-        *iter2 = sum = *iter1;
-      }
-      ++iter1;
-      ++iter2;
-      for (Size i = r.begin() + 1; i != r.end(); ++i, ++iter1, ++iter2)
-      {
-        *iter2 = sum = binary_op(sum, *iter1);
-      }
-    }
-    else
-    {
-      for (Size i = r.begin(); i != r.end(); ++i, ++iter1, ++iter2)
-      {
-        *iter2 = sum = binary_op(sum, *iter1);
-      }
-    }
-
-    first_call = false;
-  }
-
-  void reverse_join(inclusive_body& b)
-  {
-    // Only accumulate this functor's partial sum if this functor has been
-    // called at least once -- otherwise we'll over-count the initial value.
-    if (!first_call)
-    {
-      sum = binary_op(b.sum, sum);
-    }
-  }
-
-  void assign(inclusive_body& b)
-  {
-    sum = b.sum;
-  }
-};
-
-template <typename InputIterator, typename OutputIterator, typename BinaryFunction, typename ValueType>
-struct exclusive_body
-{
-  InputIterator input;
-  OutputIterator output;
-  thrust::detail::wrapped_function<BinaryFunction, ValueType> binary_op;
-  ValueType sum;
-  bool first_call;
-
-  exclusive_body(InputIterator input, OutputIterator output, BinaryFunction binary_op, ValueType init)
-      : input(input)
-      , output(output)
-      , binary_op{binary_op}
-      , sum(init)
-      , first_call(true)
-  {}
-
-  exclusive_body(exclusive_body& b, ::tbb::split)
-      : input(b.input)
-      , output(b.output)
-      , binary_op{b.binary_op}
-      , sum(b.sum)
-      , first_call(true)
-  {}
-
-  template <typename Size>
-  void operator()(const ::tbb::blocked_range<Size>& r, ::tbb::pre_scan_tag)
-  {
-    InputIterator iter = input + r.begin();
-
-    ValueType temp = *iter;
-
-    ++iter;
-
-    for (Size i = r.begin() + 1; i != r.end(); ++i, ++iter)
-    {
-      temp = binary_op(temp, *iter);
-    }
-
-    if (first_call && r.begin() > 0)
-    {
-      sum = temp;
-    }
-    else
-    {
-      sum = binary_op(sum, temp);
-    }
-
-    first_call = false;
-  }
-
-  template <typename Size>
-  void operator()(const ::tbb::blocked_range<Size>& r, ::tbb::final_scan_tag)
-  {
-    InputIterator iter1  = input + r.begin();
-    OutputIterator iter2 = output + r.begin();
-
-    for (Size i = r.begin(); i != r.end(); ++i, ++iter1, ++iter2)
-    {
-      ValueType temp = binary_op(sum, *iter1);
-      *iter2         = sum;
-      sum            = temp;
-    }
-
-    first_call = false;
-  }
-
-  void reverse_join(exclusive_body& b)
-  {
-    // Only accumulate this functor's partial sum if this functor has been
-    // called at least once -- otherwise we'll over-count the initial value.
-    if (!first_call)
-    {
-      sum = binary_op(b.sum, sum);
-    }
-  }
-
-  void assign(exclusive_body& b)
-  {
-    sum = b.sum;
-  }
-};
-
-} // namespace scan_detail
-
-template <typename InputIterator, typename OutputIterator, typename BinaryFunction>
-OutputIterator
-inclusive_scan(tag, InputIterator first, InputIterator last, OutputIterator result, BinaryFunction binary_op)
-{
-  using namespace thrust::detail;
-
-  // Use the input iterator's value type per https://wg21.link/P0571
-  using ValueType = thrust::detail::it_value_t<InputIterator>;
-
-  using Size = thrust::detail::it_difference_t<InputIterator>;
-  Size n     = ::cuda::std::distance(first, last);
-
-  if (n != 0)
-  {
-    using Body = typename scan_detail::inclusive_body<InputIterator, OutputIterator, BinaryFunction, ValueType, false>;
-    Body scan_body(first, result, binary_op, *first);
-    ::tbb::parallel_scan(::tbb::blocked_range<Size>(0, n), scan_body);
-  }
-
-  ::cuda::std::advance(result, n);
-
-  return result;
-}
-
-template <typename InputIterator, typename OutputIterator, typename InitialValueType, typename BinaryFunction>
-OutputIterator inclusive_scan(
-  tag, InputIterator first, InputIterator last, OutputIterator result, InitialValueType init, BinaryFunction binary_op)
-{
-  using namespace thrust::detail;
-
-  // Use the input iterator's value type and the initial value type per wg21.link/p2322
-  using ValueType =
-    typename ::cuda::std::__accumulator_t<BinaryFunction, thrust::detail::it_value_t<InputIterator>, InitialValueType>;
-
-  using Size = thrust::detail::it_difference_t<InputIterator>;
-  Size n     = ::cuda::std::distance(first, last);
-
-  if (n != 0)
-  {
-    using Body = typename scan_detail::inclusive_body<InputIterator, OutputIterator, BinaryFunction, ValueType, true>;
-    Body scan_body(first, result, binary_op, init);
-    ::tbb::parallel_scan(::tbb::blocked_range<Size>(0, n), scan_body);
-  }
-
-  ::cuda::std::advance(result, n);
-
-  return result;
-}
-
-template <typename InputIterator, typename OutputIterator, typename InitialValueType, typename BinaryFunction>
-OutputIterator exclusive_scan(
-  tag, InputIterator first, InputIterator last, OutputIterator result, InitialValueType init, BinaryFunction binary_op)
-{
-  using namespace thrust::detail;
-
-  // Use the initial value type per https://wg21.link/P0571
-  using ValueType = InitialValueType;
-
-  using Size = thrust::detail::it_difference_t<InputIterator>;
-  Size n     = ::cuda::std::distance(first, last);
-
-  if (n != 0)
-  {
-    using Body = typename scan_detail::exclusive_body<InputIterator, OutputIterator, BinaryFunction, ValueType>;
-    Body scan_body(first, result, binary_op, init);
-    ::tbb::parallel_scan(::tbb::blocked_range<Size>(0, n), scan_body);
-  }
-
-  ::cuda::std::advance(result, n);
-
-  return result;
-}
-
-} // end namespace detail
-} // end namespace tbb
-} // end namespace system
-THRUST_NAMESPACE_END

cuda/cccl/headers/include/thrust/system/tbb/detail/sort.inl

@@ -1,295 +0,0 @@
-/*
- *  Copyright 2008-2013 NVIDIA Corporation
- *
- *  Licensed under the Apache License, Version 2.0 (the "License");
- *  you may not use this file except in compliance with the License.
- *  You may obtain a copy of the License at
- *
- *      http://www.apache.org/licenses/LICENSE-2.0
- *
- *  Unless required by applicable law or agreed to in writing, software
- *  distributed under the License is distributed on an "AS IS" BASIS,
- *  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- *  See the License for the specific language governing permissions and
- *  limitations under the License.
- */
-
-#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/seq.h>
-#include <thrust/detail/temporary_array.h>
-#include <thrust/distance.h>
-#include <thrust/iterator/iterator_traits.h>
-#include <thrust/merge.h>
-#include <thrust/sort.h>
-
-#include <tbb/parallel_invoke.h>
-
-THRUST_NAMESPACE_BEGIN
-namespace system
-{
-namespace tbb
-{
-namespace detail
-{
-namespace sort_detail
-{
-
-// TODO tune this based on data type and comp
-const static int threshold = 128 * 1024;
-
-template <typename DerivedPolicy, typename Iterator1, typename Iterator2, typename StrictWeakOrdering>
-void merge_sort(execution_policy<DerivedPolicy>& exec,
-                Iterator1 first1,
-                Iterator1 last1,
-                Iterator2 first2,
-                StrictWeakOrdering comp,
-                bool inplace);
-
-template <typename DerivedPolicy, typename Iterator1, typename Iterator2, typename StrictWeakOrdering>
-struct merge_sort_closure
-{
-  execution_policy<DerivedPolicy>& exec;
-  Iterator1 first1, last1;
-  Iterator2 first2;
-  StrictWeakOrdering comp;
-  bool inplace;
-
-  merge_sort_closure(
-    execution_policy<DerivedPolicy>& exec,
-    Iterator1 first1,
-    Iterator1 last1,
-    Iterator2 first2,
-    StrictWeakOrdering comp,
-    bool inplace)
-      : exec(exec)
-      , first1(first1)
-      , last1(last1)
-      , first2(first2)
-      , comp(comp)
-      , inplace(inplace)
-  {}
-
-  void operator()(void) const
-  {
-    merge_sort(exec, first1, last1, first2, comp, inplace);
-  }
-};
-
-template <typename DerivedPolicy, typename Iterator1, typename Iterator2, typename StrictWeakOrdering>
-void merge_sort(execution_policy<DerivedPolicy>& exec,
-                Iterator1 first1,
-                Iterator1 last1,
-                Iterator2 first2,
-                StrictWeakOrdering comp,
-                bool inplace)
-{
-  using difference_type = thrust::detail::it_difference_t<Iterator1>;
-
-  difference_type n = ::cuda::std::distance(first1, last1);
-
-  if (n < threshold)
-  {
-    thrust::stable_sort(thrust::seq, first1, last1, comp);
-
-    if (!inplace)
-    {
-      thrust::copy(thrust::seq, first1, last1, first2);
-    }
-
-    return;
-  }
-
-  Iterator1 mid1  = first1 + (n / 2);
-  Iterator2 mid2  = first2 + (n / 2);
-  Iterator2 last2 = first2 + n;
-
-  using Closure = merge_sort_closure<DerivedPolicy, Iterator1, Iterator2, StrictWeakOrdering>;
-
-  Closure left(exec, first1, mid1, first2, comp, !inplace);
-  Closure right(exec, mid1, last1, mid2, comp, !inplace);
-
-  ::tbb::parallel_invoke(left, right);
-
-  if (inplace)
-  {
-    thrust::merge(exec, first2, mid2, mid2, last2, first1, comp);
-  }
-  else
-  {
-    thrust::merge(exec, first1, mid1, mid1, last1, first2, comp);
-  }
-}
-
-} // end namespace sort_detail
-
-namespace sort_by_key_detail
-{
-
-// TODO tune this based on data type and comp
-const static int threshold = 128 * 1024;
-
-template <typename DerivedPolicy,
-          typename Iterator1,
-          typename Iterator2,
-          typename Iterator3,
-          typename Iterator4,
-          typename StrictWeakOrdering>
-void merge_sort_by_key(
-  execution_policy<DerivedPolicy>& exec,
-  Iterator1 first1,
-  Iterator1 last1,
-  Iterator2 first2,
-  Iterator3 first3,
-  Iterator4 first4,
-  StrictWeakOrdering comp,
-  bool inplace);
-
-template <typename DerivedPolicy,
-          typename Iterator1,
-          typename Iterator2,
-          typename Iterator3,
-          typename Iterator4,
-          typename StrictWeakOrdering>
-struct merge_sort_by_key_closure
-{
-  execution_policy<DerivedPolicy>& exec;
-  Iterator1 first1, last1;
-  Iterator2 first2;
-  Iterator3 first3;
-  Iterator4 first4;
-  StrictWeakOrdering comp;
-  bool inplace;
-
-  merge_sort_by_key_closure(
-    execution_policy<DerivedPolicy>& exec,
-    Iterator1 first1,
-    Iterator1 last1,
-    Iterator2 first2,
-    Iterator3 first3,
-    Iterator4 first4,
-    StrictWeakOrdering comp,
-    bool inplace)
-      : exec(exec)
-      , first1(first1)
-      , last1(last1)
-      , first2(first2)
-      , first3(first3)
-      , first4(first4)
-      , comp(comp)
-      , inplace(inplace)
-  {}
-
-  void operator()(void) const
-  {
-    merge_sort_by_key(exec, first1, last1, first2, first3, first4, comp, inplace);
-  }
-};
-
-template <typename DerivedPolicy,
-          typename Iterator1,
-          typename Iterator2,
-          typename Iterator3,
-          typename Iterator4,
-          typename StrictWeakOrdering>
-void merge_sort_by_key(
-  execution_policy<DerivedPolicy>& exec,
-  Iterator1 first1,
-  Iterator1 last1,
-  Iterator2 first2,
-  Iterator3 first3,
-  Iterator4 first4,
-  StrictWeakOrdering comp,
-  bool inplace)
-{
-  using difference_type = thrust::detail::it_difference_t<Iterator1>;
-
-  difference_type n = ::cuda::std::distance(first1, last1);
-
-  Iterator1 mid1  = first1 + (n / 2);
-  Iterator2 mid2  = first2 + (n / 2);
-  Iterator3 mid3  = first3 + (n / 2);
-  Iterator4 mid4  = first4 + (n / 2);
-  Iterator2 last2 = first2 + n;
-  Iterator3 last3 = first3 + n;
-
-  if (n < threshold)
-  {
-    thrust::stable_sort_by_key(thrust::seq, first1, last1, first2, comp);
-
-    if (!inplace)
-    {
-      thrust::copy(thrust::seq, first1, last1, first3);
-      thrust::copy(thrust::seq, first2, last2, first4);
-    }
-
-    return;
-  }
-
-  using Closure =
-    merge_sort_by_key_closure<DerivedPolicy, Iterator1, Iterator2, Iterator3, Iterator4, StrictWeakOrdering>;
-
-  Closure left(exec, first1, mid1, first2, first3, first4, comp, !inplace);
-  Closure right(exec, mid1, last1, mid2, mid3, mid4, comp, !inplace);
-
-  ::tbb::parallel_invoke(left, right);
-
-  if (inplace)
-  {
-    thrust::merge_by_key(exec, first3, mid3, mid3, last3, first4, mid4, first1, first2, comp);
-  }
-  else
-  {
-    thrust::merge_by_key(exec, first1, mid1, mid1, last1, first2, mid2, first3, first4, comp);
-  }
-}
-
-} // namespace sort_by_key_detail
-
-template <typename DerivedPolicy, typename RandomAccessIterator, typename StrictWeakOrdering>
-void stable_sort(
-  execution_policy<DerivedPolicy>& exec, RandomAccessIterator first, RandomAccessIterator last, StrictWeakOrdering comp)
-{
-  using key_type = thrust::detail::it_value_t<RandomAccessIterator>;
-
-  thrust::detail::temporary_array<key_type, DerivedPolicy> temp(exec, first, last);
-
-  sort_detail::merge_sort(exec, first, last, temp.begin(), comp, true);
-}
-
-template <typename DerivedPolicy,
-          typename RandomAccessIterator1,
-          typename RandomAccessIterator2,
-          typename StrictWeakOrdering>
-void stable_sort_by_key(
-  execution_policy<DerivedPolicy>& exec,
-  RandomAccessIterator1 first1,
-  RandomAccessIterator1 last1,
-  RandomAccessIterator2 first2,
-  StrictWeakOrdering comp)
-{
-  using key_type = thrust::detail::it_value_t<RandomAccessIterator1>;
-  using val_type = thrust::detail::it_value_t<RandomAccessIterator2>;
-
-  RandomAccessIterator2 last2 = first2 + ::cuda::std::distance(first1, last1);
-
-  thrust::detail::temporary_array<key_type, DerivedPolicy> temp1(exec, first1, last1);
-  thrust::detail::temporary_array<val_type, DerivedPolicy> temp2(exec, first2, last2);
-
-  sort_by_key_detail::merge_sort_by_key(exec, first1, last1, first2, temp1.begin(), temp2.begin(), comp, true);
-}
-
-} // end namespace detail
-} // end namespace tbb
-} // end namespace system
-THRUST_NAMESPACE_END