-
Notifications
You must be signed in to change notification settings - Fork 115
/
Copy pathparallel_backend_tbb.h
1357 lines (1188 loc) · 45.2 KB
/
parallel_backend_tbb.h
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
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
// -*- C++ -*-
//===----------------------------------------------------------------------===//
//
// Copyright (C) Intel Corporation
//
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
// This file incorporates work covered by the following copyright and permission
// notice:
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
//
//===----------------------------------------------------------------------===//
// This header guard is used to check inclusion of TBB backend.
// Changing this macro may result in broken tests.
#ifndef _ONEDPL_PARALLEL_BACKEND_TBB_H
#define _ONEDPL_PARALLEL_BACKEND_TBB_H
#include <cassert>
#include <algorithm>
#include <type_traits>
#include "parallel_backend_utils.h"
#include "execution_impl.h"
// Bring in minimal required subset of Intel(R) Threading Building Blocks (Intel(R) TBB)
#include <tbb/blocked_range.h>
#include <tbb/parallel_for.h>
#include <tbb/parallel_for_each.h>
#include <tbb/parallel_reduce.h>
#include <tbb/parallel_scan.h>
#include <tbb/parallel_invoke.h>
#include <tbb/task_arena.h>
#include <tbb/tbb_allocator.h>
#include <tbb/enumerable_thread_specific.h>
#if TBB_INTERFACE_VERSION > 12000
# include <tbb/task.h>
#endif
#if TBB_INTERFACE_VERSION < 10000
# error Intel(R) Threading Building Blocks 2018 is required; older versions are not supported.
#endif
namespace oneapi
{
namespace dpl
{
namespace __tbb_backend
{
//! Raw memory buffer with automatic freeing and no exceptions.
/** Some of our algorithms need to start with raw memory buffer,
not an initialize array, because initialization/destruction
would make the span be at least O(N). */
// tbb::allocator can improve performance in some cases.
template <typename _ExecutionPolicy, typename _Tp>
using __buffer = oneapi::dpl::__utils::__buffer_impl<std::decay_t<_ExecutionPolicy>, _Tp, tbb::tbb_allocator>;
// Wrapper for tbb::task
inline void
__cancel_execution(oneapi::dpl::__internal::__tbb_backend_tag)
{
#if TBB_INTERFACE_VERSION <= 12000
tbb::task::self().group()->cancel_group_execution();
#else
tbb::task::current_context()->cancel_group_execution();
#endif
}
//------------------------------------------------------------------------
// parallel_for
//------------------------------------------------------------------------
template <class _Index, class _RealBody>
class __parallel_for_body
{
public:
__parallel_for_body(const _RealBody& __body) : _M_body(__body) {}
__parallel_for_body(const __parallel_for_body& __body) : _M_body(__body._M_body) {}
void
operator()(const tbb::blocked_range<_Index>& __range) const
{
_M_body(__range.begin(), __range.end());
}
private:
_RealBody _M_body;
};
//! Evaluation of brick f[i,j) for each subrange [i,j) of [first,last)
// wrapper over tbb::parallel_for
template <class _ExecutionPolicy, class _Index, class _Fp>
void
__parallel_for(oneapi::dpl::__internal::__tbb_backend_tag, _ExecutionPolicy&&, _Index __first, _Index __last, _Fp __f)
{
tbb::this_task_arena::isolate([=]() {
tbb::parallel_for(tbb::blocked_range<_Index>(__first, __last), __parallel_for_body<_Index, _Fp>(__f));
});
}
//! Evaluation of brick f[i,j) for each subrange [i,j) of [first,last)
// wrapper over tbb::parallel_reduce
template <class _ExecutionPolicy, class _Value, class _Index, typename _RealBody, typename _Reduction>
_Value
__parallel_reduce(oneapi::dpl::__internal::__tbb_backend_tag, _ExecutionPolicy&&, _Index __first, _Index __last,
const _Value& __identity, const _RealBody& __real_body, const _Reduction& __reduction)
{
return tbb::this_task_arena::isolate([__first, __last, &__identity, &__real_body, &__reduction]() -> _Value {
return tbb::parallel_reduce(
tbb::blocked_range<_Index>(__first, __last), __identity,
[__real_body](const tbb::blocked_range<_Index>& __r, const _Value& __value) -> _Value {
return __real_body(__r.begin(), __r.end(), __value);
},
__reduction);
});
}
//------------------------------------------------------------------------
// parallel_transform_reduce
//
// Notation:
// r(i,j,init) returns reduction of init with reduction over [i,j)
// u(i) returns f(i,i+1,identity) for a hypothetical left identity element of r
// c(x,y) combines values x and y that were the result of r or u
//------------------------------------------------------------------------
template <class _Index, class _Up, class _Tp, class _Cp, class _Rp>
struct __par_trans_red_body
{
alignas(_Tp) char _M_sum_storage[sizeof(_Tp)]; // Holds generalized non-commutative sum when has_sum==true
_Rp _M_brick_reduce; // Most likely to have non-empty layout
_Up _M_u;
_Cp _M_combine;
bool _M_has_sum; // Put last to minimize size of class
_Tp&
sum()
{
__TBB_ASSERT(_M_has_sum, "sum expected");
return *(_Tp*)_M_sum_storage;
}
__par_trans_red_body(_Up __u, _Tp __init, _Cp __c, _Rp __r)
: _M_brick_reduce(__r), _M_u(__u), _M_combine(__c), _M_has_sum(true)
{
new (_M_sum_storage) _Tp(__init);
}
__par_trans_red_body(__par_trans_red_body& __left, tbb::split)
: _M_brick_reduce(__left._M_brick_reduce), _M_u(__left._M_u), _M_combine(__left._M_combine), _M_has_sum(false)
{
}
~__par_trans_red_body()
{
// 17.6.5.12 tells us to not worry about catching exceptions from destructors.
if (_M_has_sum)
sum().~_Tp();
}
void
join(__par_trans_red_body& __rhs)
{
sum() = _M_combine(sum(), __rhs.sum());
}
void
operator()(const tbb::blocked_range<_Index>& __range)
{
_Index __i = __range.begin();
_Index __j = __range.end();
if (!_M_has_sum)
{
__TBB_ASSERT(__range.size() > 1, "there should be at least 2 elements");
new (&_M_sum_storage)
_Tp(_M_combine(_M_u(__i), _M_u(__i + 1))); // The condition i+1 < j is provided by the grain size of 3
_M_has_sum = true;
::std::advance(__i, 2);
if (__i == __j)
return;
}
sum() = _M_brick_reduce(__i, __j, sum());
}
};
template <class _ExecutionPolicy, class _Index, class _Up, class _Tp, class _Cp, class _Rp>
_Tp
__parallel_transform_reduce(oneapi::dpl::__internal::__tbb_backend_tag, _ExecutionPolicy&&, _Index __first,
_Index __last, _Up __u, _Tp __init, _Cp __combine, _Rp __brick_reduce)
{
__tbb_backend::__par_trans_red_body<_Index, _Up, _Tp, _Cp, _Rp> __body(__u, __init, __combine, __brick_reduce);
// The grain size of 3 is used in order to provide minimum 2 elements for each body
tbb::this_task_arena::isolate(
[__first, __last, &__body]() { tbb::parallel_reduce(tbb::blocked_range<_Index>(__first, __last, 3), __body); });
return __body.sum();
}
//------------------------------------------------------------------------
// parallel_scan
//------------------------------------------------------------------------
template <class _Index, class _Up, class _Tp, class _Cp, class _Rp, class _Sp>
class __trans_scan_body
{
alignas(_Tp) char _M_sum_storage[sizeof(_Tp)]; // Holds generalized non-commutative sum when has_sum==true
_Rp _M_brick_reduce; // Most likely to have non-empty layout
_Up _M_u;
_Cp _M_combine;
_Sp _M_scan;
bool _M_has_sum; // Put last to minimize size of class
public:
__trans_scan_body(_Up __u, _Tp __init, _Cp __combine, _Rp __reduce, _Sp __scan)
: _M_brick_reduce(__reduce), _M_u(__u), _M_combine(__combine), _M_scan(__scan), _M_has_sum(true)
{
new (_M_sum_storage) _Tp(__init);
}
__trans_scan_body(__trans_scan_body& __b, tbb::split)
: _M_brick_reduce(__b._M_brick_reduce), _M_u(__b._M_u), _M_combine(__b._M_combine), _M_scan(__b._M_scan),
_M_has_sum(false)
{
}
~__trans_scan_body()
{
// 17.6.5.12 tells us to not worry about catching exceptions from destructors.
if (_M_has_sum)
sum().~_Tp();
}
_Tp&
sum() const
{
__TBB_ASSERT(_M_has_sum, "sum expected");
return *const_cast<_Tp*>(reinterpret_cast<_Tp const*>(_M_sum_storage));
}
void
operator()(const tbb::blocked_range<_Index>& __range, tbb::pre_scan_tag)
{
_Index __i = __range.begin();
_Index __j = __range.end();
if (!_M_has_sum)
{
new (&_M_sum_storage) _Tp(_M_u(__i));
_M_has_sum = true;
++__i;
if (__i == __j)
return;
}
sum() = _M_brick_reduce(__i, __j, sum());
}
void
operator()(const tbb::blocked_range<_Index>& __range, tbb::final_scan_tag)
{
sum() = _M_scan(__range.begin(), __range.end(), sum());
}
void
reverse_join(__trans_scan_body& __a)
{
if (_M_has_sum)
{
sum() = _M_combine(__a.sum(), sum());
}
else
{
new (&_M_sum_storage) _Tp(__a.sum());
_M_has_sum = true;
}
}
void
assign(__trans_scan_body& __b)
{
sum() = __b.sum();
}
};
template <typename _Index>
_Index
__split(_Index __m)
{
_Index __k = 1;
while (2 * __k < __m)
__k *= 2;
return __k;
}
//------------------------------------------------------------------------
// __parallel_strict_scan
//------------------------------------------------------------------------
// TODO: Here are improvements to consider:
// 1. Affinitize leaves of upsweep and leaves of downsweep. For working sets that
// fit in cache, this might reduce memory interconnect load significantly.
// 2. Add automatic tilesize adjustment. Initial stealing during upsweep ought to provide a good hint.
// 3. Use continuation-passing style for the tasks. For downsweep, a binomial tree pattern is likely optimal.
template <typename _Index, typename _Tp, typename _Rp, typename _Cp>
void
__upsweep(_Index __i, _Index __m, _Index __tilesize, _Tp* __r, _Index __lastsize, _Rp __reduce, _Cp __combine)
{
if (__m == 1)
__r[0] = __reduce(__i * __tilesize, __lastsize);
else
{
_Index __k = __split(__m);
tbb::parallel_invoke(
[=] { __tbb_backend::__upsweep(__i, __k, __tilesize, __r, __tilesize, __reduce, __combine); },
[=] {
__tbb_backend::__upsweep(__i + __k, __m - __k, __tilesize, __r + __k, __lastsize, __reduce, __combine);
});
if (__m == 2 * __k)
__r[__m - 1] = __combine(__r[__k - 1], __r[__m - 1]);
}
}
template <typename _Index, typename _Tp, typename _Cp, typename _Sp>
void
__downsweep(_Index __i, _Index __m, _Index __tilesize, _Tp* __r, _Index __lastsize, _Tp __initial, _Cp __combine,
_Sp __scan)
{
if (__m == 1)
__scan(__i * __tilesize, __lastsize, __initial);
else
{
const _Index __k = __split(__m);
tbb::parallel_invoke(
[=] { __tbb_backend::__downsweep(__i, __k, __tilesize, __r, __tilesize, __initial, __combine, __scan); },
// Assumes that __combine never throws.
//TODO: Consider adding a requirement for user functors to be constant.
[=, &__combine] {
__tbb_backend::__downsweep(__i + __k, __m - __k, __tilesize, __r + __k, __lastsize,
__combine(__initial, __r[__k - 1]), __combine, __scan);
});
}
}
// Adapted from Intel(R) Cilk(TM) version from cilkpub.
// Let i:len denote a counted interval of length n starting at i. s denotes a generalized-sum value.
// Expected actions of the functors are:
// reduce(i,len) -> s -- return reduction value of i:len.
// combine(s1,s2) -> s -- return merged sum
// apex(s) -- do any processing necessary between reduce and scan.
// scan(i,len,initial) -- perform scan over i:len starting with initial.
// The initial range 0:n is partitioned into consecutive subranges.
// reduce and scan are each called exactly once per subrange.
// Thus callers can rely upon side effects in reduce.
// combine must not throw an exception.
// apex is called exactly once, after all calls to reduce and before all calls to scan.
// For example, it's useful for allocating a __buffer used by scan but whose size is the sum of all reduction values.
// T must have a trivial constructor and destructor.
template <class _ExecutionPolicy, typename _Index, typename _Tp, typename _Rp, typename _Cp, typename _Sp, typename _Ap>
void
__parallel_strict_scan(oneapi::dpl::__internal::__tbb_backend_tag, _ExecutionPolicy&& __exec, _Index __n, _Tp __initial,
_Rp __reduce, _Cp __combine, _Sp __scan, _Ap __apex)
{
tbb::this_task_arena::isolate([=, &__combine]() {
if (__n > 1)
{
_Index __p = tbb::this_task_arena::max_concurrency();
const _Index __slack = 4;
_Index __tilesize = (__n - 1) / (__slack * __p) + 1;
_Index __m = (__n - 1) / __tilesize;
__tbb_backend::__buffer<_ExecutionPolicy, _Tp> __buf(__exec, __m + 1);
_Tp* __r = __buf.get();
__tbb_backend::__upsweep(_Index(0), _Index(__m + 1), __tilesize, __r, __n - __m * __tilesize, __reduce,
__combine);
// When __apex is a no-op and __combine has no side effects, a good optimizer
// should be able to eliminate all code between here and __apex.
// Alternatively, provide a default value for __apex that can be
// recognized by metaprogramming that conditionlly executes the following.
size_t __k = __m + 1;
_Tp __t = __r[__k - 1];
while ((__k &= __k - 1))
__t = __combine(__r[__k - 1], __t);
__apex(__combine(__initial, __t));
__tbb_backend::__downsweep(_Index(0), _Index(__m + 1), __tilesize, __r, __n - __m * __tilesize, __initial,
__combine, __scan);
return;
}
// Fewer than 2 elements in sequence, or out of memory. Handle has single block.
_Tp __sum = __initial;
if (__n)
__sum = __combine(__sum, __reduce(_Index(0), __n));
__apex(__sum);
if (__n)
__scan(_Index(0), __n, __initial);
});
}
template <class _ExecutionPolicy, class _Index, class _Up, class _Tp, class _Cp, class _Rp, class _Sp>
_Tp
__parallel_transform_scan(oneapi::dpl::__internal::__tbb_backend_tag, _ExecutionPolicy&&, _Index __n, _Up __u,
_Tp __init, _Cp __combine, _Rp __brick_reduce, _Sp __scan)
{
__trans_scan_body<_Index, _Up, _Tp, _Cp, _Rp, _Sp> __body(__u, __init, __combine, __brick_reduce, __scan);
auto __range = tbb::blocked_range<_Index>(0, __n);
tbb::this_task_arena::isolate([__range, &__body]() { tbb::parallel_scan(__range, __body); });
return __body.sum();
}
//------------------------------------------------------------------------
// parallel_stable_sort
//------------------------------------------------------------------------
//------------------------------------------------------------------------
// stable_sort utilities
//
// These are used by parallel implementations but do not depend on them.
//------------------------------------------------------------------------
#define _ONEDPL_MERGE_CUT_OFF 2000
template <typename _Func>
class __func_task;
template <typename _Func>
class __root_task;
#if TBB_INTERFACE_VERSION <= 12000
class __task : public tbb::task
{
public:
template <typename _Fn>
__task*
make_continuation(_Fn&& __f)
{
return new (allocate_continuation()) __func_task<::std::decay_t<_Fn>>(::std::forward<_Fn>(__f));
}
template <typename _Fn>
__task*
make_child_of(__task* parent, _Fn&& __f)
{
return new (parent->allocate_child()) __func_task<::std::decay_t<_Fn>>(::std::forward<_Fn>(__f));
}
template <typename _Fn>
__task*
make_additional_child_of(tbb::task* parent, _Fn&& __f)
{
return new (tbb::task::allocate_additional_child_of(*parent))
__func_task<::std::decay_t<_Fn>>(::std::forward<_Fn>(__f));
}
inline void
recycle_as_continuation()
{
tbb::task::recycle_as_continuation();
}
inline void
recycle_as_child_of(__task* parent)
{
tbb::task::recycle_as_child_of(*parent);
}
inline void
spawn(__task* __t)
{
tbb::task::spawn(*__t);
}
template <typename _Fn>
static inline void
spawn_root_and_wait(__root_task<_Fn>& __root)
{
tbb::task::spawn_root_and_wait(*__root._M_task);
}
};
template <typename _Func>
class __func_task : public __task
{
_Func _M_func;
tbb::task*
execute()
{
return _M_func(this);
};
public:
template <typename _Fn>
__func_task(_Fn&& __f) : _M_func{::std::forward<_Fn>(__f)}
{
}
_Func&
body()
{
return _M_func;
}
};
template <typename _Func>
class __root_task
{
tbb::task* _M_task;
public:
template <typename... Args>
__root_task(Args&&... args)
: _M_task{new (tbb::task::allocate_root()) __func_task<_Func>{_Func(::std::forward<Args>(args)...)}}
{
}
friend class __task;
friend class __func_task<_Func>;
};
#else // TBB_INTERFACE_VERSION > 12000
class __task : public tbb::detail::d1::task
{
protected:
tbb::detail::d1::small_object_allocator _M_allocator{};
tbb::detail::d1::execution_data* _M_execute_data{};
__task* _M_parent{};
::std::atomic<int> _M_refcount{};
bool _M_recycle{};
template <typename _Fn>
__task*
allocate_func_task(_Fn&& __f)
{
assert(_M_execute_data != nullptr);
tbb::detail::d1::small_object_allocator __alloc{};
auto __t = __alloc.new_object<__func_task<::std::decay_t<_Fn>>>(*_M_execute_data, ::std::forward<_Fn>(__f));
__t->_M_allocator = __alloc;
return __t;
}
public:
__task*
parent()
{
return _M_parent;
}
void
set_ref_count(int __n)
{
_M_refcount.store(__n, ::std::memory_order_release);
}
template <typename _Fn>
__task*
make_continuation(_Fn&& __f)
{
auto __t = allocate_func_task(::std::forward<_Fn&&>(__f));
__t->_M_parent = _M_parent;
_M_parent = nullptr;
return __t;
}
template <typename _Fn>
__task*
make_child_of(__task* __parent, _Fn&& __f)
{
auto __t = allocate_func_task(::std::forward<_Fn&&>(__f));
__t->_M_parent = __parent;
return __t;
}
template <typename _Fn>
__task*
make_additional_child_of(__task* __parent, _Fn&& __f)
{
auto __t = make_child_of(__parent, ::std::forward<_Fn>(__f));
assert(__parent->_M_refcount.load(::std::memory_order_relaxed) > 0);
++__parent->_M_refcount;
return __t;
}
inline void
recycle_as_continuation()
{
_M_recycle = true;
}
inline void
recycle_as_child_of(__task* parent)
{
_M_recycle = true;
_M_parent = parent;
}
inline void
spawn(__task* __t)
{
assert(_M_execute_data != nullptr);
tbb::detail::d1::spawn(*__t, *_M_execute_data->context);
}
template <typename _Fn>
static inline void
spawn_root_and_wait(__root_task<_Fn>& __root)
{
tbb::detail::d1::execute_and_wait(*__root._M_func_task, __root._M_context, __root._M_wait_object,
__root._M_context);
}
template <typename _Func>
friend class __func_task;
};
template <typename _Func>
class __func_task : public __task
{
_Func _M_func;
__task*
execute(tbb::detail::d1::execution_data& __ed) override
{
_M_execute_data = &__ed;
_M_recycle = false;
__task* __next = _M_func(this);
return finalize(__next);
};
__task*
cancel(tbb::detail::d1::execution_data&) override
{
return finalize(nullptr);
}
__task*
finalize(__task* __next)
{
bool __recycle = _M_recycle;
_M_recycle = false;
if (__recycle)
{
return __next;
}
auto __parent = _M_parent;
auto __alloc = _M_allocator;
auto __ed = _M_execute_data;
this->~__func_task();
assert(__parent != nullptr);
assert(__parent->_M_refcount.load(::std::memory_order_relaxed) > 0);
auto __refcount = --__parent->_M_refcount;
// Placing the deallocation after the refcount decrement allows another thread to proceed with tree
// folding concurrently with this task cleanup.
__alloc.deallocate(this, *__ed);
if (__refcount == 0)
{
assert(__next == nullptr);
return __parent;
}
return __next;
}
friend class __root_task<_Func>;
public:
template <typename _Fn>
__func_task(_Fn&& __f) : _M_func(::std::forward<_Fn>(__f))
{
}
_Func&
body()
{
return _M_func;
}
};
template <typename _Func>
class __root_task : public __task
{
__task*
execute(tbb::detail::d1::execution_data&) override
{
_M_wait_object.release();
return nullptr;
};
__task*
cancel(tbb::detail::d1::execution_data&) override
{
_M_wait_object.release();
return nullptr;
}
__func_task<_Func>* _M_func_task{};
tbb::detail::d1::wait_context _M_wait_object{0};
tbb::task_group_context _M_context{};
public:
template <typename... Args>
__root_task(Args&&... args) : _M_wait_object{1}
{
tbb::detail::d1::small_object_allocator __alloc{};
_M_func_task = __alloc.new_object<__func_task<_Func>>(_Func(::std::forward<Args>(args)...));
_M_func_task->_M_allocator = __alloc;
_M_func_task->_M_parent = this;
_M_refcount.store(1, ::std::memory_order_relaxed);
}
friend class __task;
};
#endif // TBB_INTERFACE_VERSION <= 12000
template <typename _RandomAccessIterator1, typename _RandomAccessIterator2, typename _Compare, typename _Cleanup,
typename _LeafMerge>
class __merge_func
{
typedef typename ::std::iterator_traits<_RandomAccessIterator1>::difference_type _DifferenceType1;
typedef typename ::std::iterator_traits<_RandomAccessIterator2>::difference_type _DifferenceType2;
typedef typename ::std::common_type_t<_DifferenceType1, _DifferenceType2> _SizeType;
typedef typename ::std::iterator_traits<_RandomAccessIterator1>::value_type _ValueType;
_RandomAccessIterator1 _M_x_beg;
_RandomAccessIterator2 _M_z_beg;
_SizeType _M_xs, _M_xe;
_SizeType _M_ys, _M_ye;
_SizeType _M_zs;
_Compare _M_comp;
_LeafMerge _M_leaf_merge;
_SizeType _M_nsort; //number of elements to be sorted for partial_sort alforithm
static const _SizeType __merge_cut_off = _ONEDPL_MERGE_CUT_OFF;
bool _root; //means a task is merging root task
bool _x_orig; //"true" means X(or left ) subrange is in the original container; false - in the buffer
bool _y_orig; //"true" means Y(or right) subrange is in the original container; false - in the buffer
bool _split; //"true" means a merge task is a split task for parallel merging, the execution logic differs
struct __move_value
{
template <typename Iterator1, typename Iterator2>
void
operator()(Iterator1 __x, Iterator2 __z)
{
*__z = ::std::move(*__x);
}
};
struct __move_value_construct
{
template <typename Iterator1, typename Iterator2>
void
operator()(Iterator1 __x, Iterator2 __z)
{
::new (::std::addressof(*__z)) _ValueType(::std::move(*__x));
}
};
struct __move_range
{
template <typename Iterator1, typename Iterator2>
Iterator2
operator()(Iterator1 __first1, Iterator1 __last1, Iterator2 __first2)
{
if (__last1 - __first1 < __merge_cut_off)
return ::std::move(__first1, __last1, __first2);
auto __n = __last1 - __first1;
tbb::parallel_for(tbb::blocked_range<_SizeType>(0, __n, __merge_cut_off),
[__first1, __first2](const tbb::blocked_range<_SizeType>& __range) {
::std::move(__first1 + __range.begin(), __first1 + __range.end(),
__first2 + __range.begin());
});
return __first2 + __n;
}
};
struct __move_range_construct
{
template <typename Iterator1, typename Iterator2>
Iterator2
operator()(Iterator1 __first1, Iterator1 __last1, Iterator2 __first2)
{
if (__last1 - __first1 < __merge_cut_off)
{
for (; __first1 != __last1; ++__first1, ++__first2)
__move_value_construct()(__first1, __first2);
return __first2;
}
auto __n = __last1 - __first1;
tbb::parallel_for(tbb::blocked_range<_SizeType>(0, __n, __merge_cut_off),
[__first1, __first2](const tbb::blocked_range<_SizeType>& __range) {
for (auto i = __range.begin(); i != __range.end(); ++i)
__move_value_construct()(__first1 + i, __first2 + i);
});
return __first2 + __n;
}
};
struct __cleanup_range
{
template <typename Iterator>
void
operator()(Iterator __first, Iterator __last)
{
if (__last - __first < __merge_cut_off)
_Cleanup()(__first, __last);
else
{
auto __n = __last - __first;
tbb::parallel_for(tbb::blocked_range<_SizeType>(0, __n, __merge_cut_off),
[__first](const tbb::blocked_range<_SizeType>& __range) {
_Cleanup()(__first + __range.begin(), __first + __range.end());
});
}
}
};
public:
__merge_func(_SizeType __xs, _SizeType __xe, _SizeType __ys, _SizeType __ye, _SizeType __zs, _Compare __comp,
_Cleanup, _LeafMerge __leaf_merge, _SizeType __nsort, _RandomAccessIterator1 __x_beg,
_RandomAccessIterator2 __z_beg, bool __x_orig, bool __y_orig, bool __root)
: _M_x_beg(__x_beg), _M_z_beg(__z_beg), _M_xs(__xs), _M_xe(__xe), _M_ys(__ys), _M_ye(__ye), _M_zs(__zs),
_M_comp(__comp), _M_leaf_merge(__leaf_merge), _M_nsort(__nsort), _root(__root), _x_orig(__x_orig),
_y_orig(__y_orig), _split(false)
{
}
bool
is_left(_SizeType __idx) const
{
return _M_xs == __idx;
}
template <typename IndexType>
void
set_odd(IndexType __idx, bool __on_off)
{
if (is_left(__idx))
_x_orig = __on_off;
else
_y_orig = __on_off;
}
__task*
operator()(__task* __self);
private:
__merge_func*
parent_merge(__task* __self) const
{
return _root ? nullptr : &static_cast<__func_task<__merge_func>*>(__self->parent())->body();
}
bool
x_less_y()
{
auto __nx = (_M_xe - _M_xs);
auto __ny = (_M_ye - _M_ys);
assert(__nx > 0 && __ny > 0);
assert(_M_nsort > 0);
auto __kx = ::std::min(_M_nsort, __nx);
auto __ky = ::std::min(_M_nsort, __ny);
assert(_x_orig == _y_orig);
if (_x_orig)
{
assert(::std::is_sorted(_M_x_beg + _M_xs, _M_x_beg + _M_xs + __kx, _M_comp));
assert(::std::is_sorted(_M_x_beg + _M_ys, _M_x_beg + _M_ys + __ky, _M_comp));
return !_M_comp(*(_M_x_beg + _M_ys), *(_M_x_beg + _M_xs + __kx - 1));
}
assert(::std::is_sorted(_M_z_beg + _M_xs, _M_z_beg + _M_xs + __kx, _M_comp));
assert(::std::is_sorted(_M_z_beg + _M_ys, _M_z_beg + _M_ys + __ky, _M_comp));
return !_M_comp(*(_M_z_beg + _M_zs + __nx), *(_M_z_beg + _M_zs + __kx - 1));
}
void
move_x_range()
{
const auto __nx = (_M_xe - _M_xs);
const auto __ny = (_M_ye - _M_ys);
assert(__nx > 0 && __ny > 0);
if (_x_orig)
__move_range_construct()(_M_x_beg + _M_xs, _M_x_beg + _M_xe, _M_z_beg + _M_zs);
else
{
__move_range()(_M_z_beg + _M_zs, _M_z_beg + _M_zs + __nx, _M_x_beg + _M_xs);
if constexpr (!::std::is_trivially_destructible_v<_ValueType>)
__cleanup_range()(_M_z_beg + _M_zs, _M_z_beg + _M_zs + __nx);
}
_x_orig = !_x_orig;
}
void
move_y_range()
{
const auto __nx = (_M_xe - _M_xs);
const auto __ny = (_M_ye - _M_ys);
if (_y_orig)
__move_range_construct()(_M_x_beg + _M_ys, _M_x_beg + _M_ye, _M_z_beg + _M_zs + __nx);
else
{
__move_range()(_M_z_beg + _M_zs + __nx, _M_z_beg + _M_zs + __nx + __ny, _M_x_beg + _M_ys);
if constexpr (!::std::is_trivially_destructible_v<_ValueType>)
__cleanup_range()(_M_z_beg + _M_zs + __nx, _M_z_beg + _M_zs + __nx + __ny);
}
_y_orig = !_y_orig;
}
__task*
merge_ranges(__task* __self)
{
assert(_M_nsort > 0);
assert(_x_orig == _y_orig); //two merged subrange must be lie into the same buffer
const auto __nx = (_M_xe - _M_xs);
const auto __ny = (_M_ye - _M_ys);
const auto __n = __nx + __ny;
// need to merge {x} and {y}
if (__n > __merge_cut_off)
return split_merging(__self);
//merge to buffer
if (_x_orig)
{
_M_leaf_merge(_M_x_beg + _M_xs, _M_x_beg + _M_xe, _M_x_beg + _M_ys, _M_x_beg + _M_ye, _M_z_beg + _M_zs,
_M_comp, __move_value_construct(), __move_value_construct(), __move_range_construct(),
__move_range_construct());
assert(parent_merge(__self)); //not root merging task
}
//merge to "origin"
else
{
assert(_x_orig == _y_orig);
_M_leaf_merge(_M_z_beg + _M_xs, _M_z_beg + _M_xe, _M_z_beg + _M_ys, _M_z_beg + _M_ye, _M_x_beg + _M_zs,
_M_comp, __move_value(), __move_value(), __move_range(), __move_range());
if constexpr (!::std::is_trivially_destructible_v<_ValueType>)
{
__cleanup_range()(_M_z_beg + _M_xs, _M_z_beg + _M_xe);
__cleanup_range()(_M_z_beg + _M_ys, _M_z_beg + _M_ye);
}
}
return nullptr;
}
__task*
process_ranges(__task* __self)
{
assert(_x_orig == _y_orig);
assert(!_split);
auto p = parent_merge(__self);
if (!p)
{ //root merging task
//optimization, just for sort algorithm, //{x} <= {y}
if (x_less_y()) //we have a solution
{
if (!_x_orig)
{ //we have to move the solution to the origin
move_x_range(); //parallel moving
move_y_range(); //parallel moving
}
return nullptr;
}
//else: if we have data in the origin,
//we have to move data to the buffer for final merging into the origin.
if (_x_orig)
{
move_x_range(); //parallel moving
move_y_range(); //parallel moving
}
// need to merge {x} and {y}.
return merge_ranges(__self);
}
//else: not root merging task (parent_merge() == NULL)
//optimization, just for sort algorithm, //{x} <= {y}
if (x_less_y())
{
const auto id_range = _M_zs;
p->set_odd(id_range, _x_orig);
return nullptr;
}
//else: we have to revert "_x(y)_orig" flag of the parent merging task
const auto id_range = _M_zs;
p->set_odd(id_range, !_x_orig);
return merge_ranges(__self);
}