File: sort.inl

package info (click to toggle)
libthrust 1.17.2-2
  • links: PTS, VCS
  • area: main
  • in suites: bookworm
  • size: 10,900 kB
  • sloc: ansic: 29,519; cpp: 23,989; python: 1,421; sh: 811; perl: 460; makefile: 112
file content (206 lines) | stat: -rw-r--r-- 6,440 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
/*
 *  Copyright 2008-2021 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>

#include <thrust/reverse.h>
#include <thrust/detail/type_traits.h>
#include <thrust/iterator/iterator_traits.h>
#include <thrust/system/detail/sequential/stable_merge_sort.h>
#include <thrust/system/detail/sequential/stable_primitive_sort.h>

THRUST_NAMESPACE_BEGIN
namespace system
{
namespace detail
{
namespace sequential
{
namespace sort_detail
{


////////////////////
// Primitive Sort //
////////////////////


template<typename KeyType, typename Compare>
struct needs_reverse
  : thrust::detail::integral_constant<
      bool,
      thrust::detail::is_same<Compare, typename thrust::greater<KeyType> >::value
    >
{};


template<typename DerivedPolicy,
         typename RandomAccessIterator,
         typename StrictWeakOrdering>
__host__ __device__
void stable_sort(sequential::execution_policy<DerivedPolicy> &exec,
                 RandomAccessIterator first,
                 RandomAccessIterator last,
                 StrictWeakOrdering,
                 thrust::detail::true_type)
{
  thrust::system::detail::sequential::stable_primitive_sort(exec, first, last);

  // if comp is greater<T> then reverse the keys
  typedef typename thrust::iterator_traits<RandomAccessIterator>::value_type KeyType;

  if(needs_reverse<KeyType,StrictWeakOrdering>::value)
  {
    thrust::reverse(exec, first, last);
  }
}


template<typename DerivedPolicy,
         typename RandomAccessIterator1,
         typename RandomAccessIterator2,
         typename StrictWeakOrdering>
__host__ __device__
void stable_sort_by_key(sequential::execution_policy<DerivedPolicy> &exec,
                        RandomAccessIterator1 first1,
                        RandomAccessIterator1 last1,
                        RandomAccessIterator2 first2,
                        StrictWeakOrdering,
                        thrust::detail::true_type)
{
  // if comp is greater<T> then reverse the keys and values
  typedef typename thrust::iterator_traits<RandomAccessIterator1>::value_type KeyType;

  // note, we also have to reverse the (unordered) input to preserve stability
  if(needs_reverse<KeyType,StrictWeakOrdering>::value)
  {
    thrust::reverse(exec, first1,  last1);
    thrust::reverse(exec, first2, first2 + (last1 - first1));
  }

  thrust::system::detail::sequential::stable_primitive_sort_by_key(exec, first1, last1, first2);

  if(needs_reverse<KeyType,StrictWeakOrdering>::value)
  {
    thrust::reverse(exec, first1,  last1);
    thrust::reverse(exec, first2, first2 + (last1 - first1));
  }
}


////////////////
// Merge Sort //
////////////////


template<typename DerivedPolicy,
         typename RandomAccessIterator,
         typename StrictWeakOrdering>
__host__ __device__
void stable_sort(sequential::execution_policy<DerivedPolicy> &exec,
                 RandomAccessIterator first,
                 RandomAccessIterator last,
                 StrictWeakOrdering comp,
                 thrust::detail::false_type)
{
  thrust::system::detail::sequential::stable_merge_sort(exec, first, last, comp);
}


template<typename DerivedPolicy,
         typename RandomAccessIterator1,
         typename RandomAccessIterator2,
         typename StrictWeakOrdering>
__host__ __device__
void stable_sort_by_key(sequential::execution_policy<DerivedPolicy> &exec,
                        RandomAccessIterator1 first1,
                        RandomAccessIterator1 last1,
                        RandomAccessIterator2 first2,
                        StrictWeakOrdering comp,
                        thrust::detail::false_type)
{
  thrust::system::detail::sequential::stable_merge_sort_by_key(exec, first1, last1, first2, comp);
}


template<typename KeyType, typename Compare>
struct use_primitive_sort
  : thrust::detail::and_<
      thrust::detail::is_arithmetic<KeyType>,
      thrust::detail::or_<
        thrust::detail::is_same<Compare, thrust::less<KeyType> >,
        thrust::detail::is_same<Compare, thrust::greater<KeyType> >
      >
    >
{};


} // end namespace sort_detail


template<typename DerivedPolicy,
         typename RandomAccessIterator,
         typename StrictWeakOrdering>
__host__ __device__
void stable_sort(sequential::execution_policy<DerivedPolicy> &exec,
                 RandomAccessIterator first,
                 RandomAccessIterator last,
                 StrictWeakOrdering comp)
{

  // the compilation time of stable_primitive_sort is too expensive to use within a single CUDA thread
#ifndef __CUDA_ARCH__
  typedef typename thrust::iterator_traits<RandomAccessIterator>::value_type KeyType;
  sort_detail::use_primitive_sort<KeyType,StrictWeakOrdering> use_primitive_sort;
#else
  thrust::detail::false_type use_primitive_sort;
#endif

  sort_detail::stable_sort(exec, first, last, comp, use_primitive_sort);
}


template<typename DerivedPolicy,
         typename RandomAccessIterator1,
         typename RandomAccessIterator2,
         typename StrictWeakOrdering>
__host__ __device__
void stable_sort_by_key(sequential::execution_policy<DerivedPolicy> &exec,
                        RandomAccessIterator1 first1,
                        RandomAccessIterator1 last1,
                        RandomAccessIterator2 first2,
                        StrictWeakOrdering comp)
{

  // the compilation time of stable_primitive_sort_by_key is too expensive to use within a single CUDA thread
#ifndef __CUDA_ARCH__
  typedef typename thrust::iterator_traits<RandomAccessIterator1>::value_type KeyType;
  sort_detail::use_primitive_sort<KeyType,StrictWeakOrdering> use_primitive_sort;
#else
  thrust::detail::false_type use_primitive_sort;
#endif

  sort_detail::stable_sort_by_key(exec, first1, last1, first2, comp, use_primitive_sort);
}


} // end namespace sequential
} // end namespace detail
} // end namespace system
THRUST_NAMESPACE_END