forked from uxlfoundation/oneDPL
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathCHANGES
More file actions
226 lines (165 loc) · 7.96 KB
/
CHANGES
File metadata and controls
226 lines (165 loc) · 7.96 KB
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
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
------------------------------------------------------------------------
The list of most significant changes made over time in Parallel STL.
Parallel STL 20190321 release
PSTL_VERSION == 205
Features / APIs:
- Improved performance in stable_sort and sort algorithms.
------------------------------------------------------------------------
Parallel STL release within Intel(R) Parallel Studio XE 2019 Update 3
PSTL_VERSION == 204
Bugs fixed:
- Fixed compilation error for usage of PSTL_USE_PARALLEL_POLICIES macro.
- Fixed compilation error for reordering standard included files.
Examples:
- Added CMake support for convex_hull and dot_product samples.
------------------------------------------------------------------------
Parallel STL 20181204 release
PSTL_VERSION == 203
Features / APIs:
- More algorithms support parallel execution policies:
set_union, set_intersection, set_difference, set_symmetric_difference.
- Improved performance in reverse, reverse_copy algorithm with vector
execution policies.
- Added FindTBB CMake module - Preview Feature.
Bugs fixed:
- Fixed several algorithms which use scan and reduction patterns with
a user defined binary operation with vector execution policies.
Examples:
- CMake support for gamma_correction sample was added.
------------------------------------------------------------------------
Parallel STL release within Intel(R) Parallel Studio XE 2019 Update 1
PSTL_VERSION == 202
Features / APIs:
- Fixed compilation error for reduce and transform_reduce algorithms
with Microsoft* Visual Studio 2017 15.8.
------------------------------------------------------------------------
Parallel STL 20181004 release
PSTL_VERSION == 201
Features / APIs:
- Fixed a number of algorithms to work with mutable function objects.
- Fixed algorithms to throw std::bad_alloc when additional memory cannot
be obtained.
------------------------------------------------------------------------
Parallel STL release within Intel(R) Parallel Studio XE 2019
PSTL_VERSION == 200
Features / APIs:
- More algorithms support parallel and vector execution policies:
rotate.
- More algorithms support parallel execution policies:
includes, nth_element, remove, remove_if, unique.
- More algorithms support vector execution policies:
exclusive_scan, inclusive_scan, max_element, minmax_element,
min_element, reduce, transform_exclusive_scan, transform_inclusive_scan,
transform_reduce.
- Arranged types and variables naming: single- and double-underscore.
(by Thomas Rodgers).
------------------------------------------------------------------------
Parallel STL 20180619 release
PSTL_VERSION == 106
Features / APIs:
- More algorithms support parallel and vector execution policies:
adjacent_difference, partition, reverse, reverse_copy, rotate_copy,
stable_partition.
- More algorithms support parallel execution policies:
inplace_merge, partial_sort_copy.
- Split algorithm declarations and implementation by files.
(by Thomas Rodgers).
- CMake support - Preview Feature
(by Amit Prakash Ambasta and Henry Schreiner)
------------------------------------------------------------------------
Parallel STL release within Intel(R) Parallel Studio XE 2018 Update 3
PSTL_VERSION == 105
Features / APIs:
- More algorithms support parallel and vector execution policies:
is_partitioned, lexicographical_compare, swap_ranges.
- Added specialization of parallel_transform_scan pattern for better
performance with floating point types.
- Arranged iterator types in public interface according to the standard
(by Thomas Rodgers).
Bugs fixed:
- Fixed an error in parallel version of partial_sort algorithm.
- Fixed an error in equal algorithm connected with values comparison.
------------------------------------------------------------------------
Parallel STL 20180329 release
PSTL_VERSION == 104
Features / APIs:
- More algorithms support parallel and vector execution policies:
find_first_of, is_heap, is_heap_until, replace, replace_if.
- More algorithms support vector execution policies:
remove, remove_if.
- More algorithms support parallel execution policies:
partial_sort.
------------------------------------------------------------------------
Parallel STL release within Intel(R) Parallel Studio XE 2018 Update 2
PSTL_VERSION == 103
Features / APIs:
- More algorithms support parallel and/or vector execution policies:
find_end, merge, search, search_n
------------------------------------------------------------------------
Parallel STL 20171127 release
PSTL_VERSION == 102
Features / APIs:
- Added Parallel STL version macros:
PSTL_VERSION, PSTL_VERSION_MAJOR, PSTL_VERSION_MINOR.
- More algorithms support parallel and vector execution policies:
move, partition_copy, mismatch.
- More algorithms support parallel execution policies:
min_element, max_element, minmax_element.
------------------------------------------------------------------------
Parallel STL release within Intel(R) Parallel Studio XE 2018 Update 1
Features / APIs:
- More algorithms support parallel and vector execution policies:
destroy, destroy_n, uninitialized_copy, uninitialized_copy_n,
uninitialized_default_construct, uninitialized_default_construct_n,
uninitialized_fill, uninitialized_fill_n, uninitialized_move,
uninitialized_move_n, uninitialized_value_construct,
uninitialized_value_construct_n.
- Improved performance in find_end and search algorithms.
- Added macro PSTL_USE_NONTEMPORAL_STORES that can improve performance
of copy, copy_n, fill, fill_n, generate, generate_n algorithms with
unseq and par_unseq policies; by default the macro is not defined.
Bugs fixed:
- Fixed transform_inclusive_scan to correctly process the first element.
- Fixed compile time error in sort algorithm when used with zip_iterator
and some other custom iterator types.
- Fixed several algorithms to allow use of non-const functors.
------------------------------------------------------------------------
Parallel STL release within Intel(R) Parallel Studio XE 2018
Features / APIs:
- Aligned the implementation with the draft N4659 of the C++ standard.
In particular, inner_product no longer supports execution policies.
- reduce and transform_reduce support unseq and par_unseq execution
policies if std::plus<> is used for reduction.
- Added counting_iterator and zip_iterator to support advanced use cases.
To use, include pstl/iterators.h header file.
- Added macro PSTL_USE_PARALLEL_POLICIES to control usage of parallel
execution policies (par, par_unseq); by default these are enabled.
Examples:
- Added dot_product and convex_hull samples to demonstrate Parallel STL
usage.
------------------------------------------------------------------------
Parallel STL beta update release
Features / APIs:
- sequenced_policy is supported by all C++17 algorithms.
- Added other execution policies support for inner_product algorithm.
------------------------------------------------------------------------
Parallel STL beta initial release
Features / APIs:
- Implemented C++17 execution policies: sequenced_policy,
parallel_policy, parallel_unsequenced_policy.
- Implemented unsequenced_policy based on the ISO C++ working group
paper P0076R3.
- sequenced_policy is supported by all except a few C++17 algorithms.
- Other execution policies are supported by the following algorithms:
adjacent_find, all_of, any_of, copy, copy_if, copy_n, count,
count_if, equal, exclusive_scan, fill, fill_n, find, find_if,
find_if_not, for_each, for_each_n, generate, generate_n,
inclusive_scan, none_of, is_sorted, is_sorted_until, reduce,
remove_copy, remove_copy_if, sort, stable_sort, transform,
transform_exclusive_scan, transform_inclusive_scan,
transform_reduce, unique_copy.
Documentation:
- Initial set of documents: Getting Started Guide, Release Notes.
Examples:
- Added gamma_correction sample to demonstrate Parallel STL usage.
------------------------------------------------------------------------