forked from AnonymousRepo123/AlphaSparse
-
Notifications
You must be signed in to change notification settings - Fork 0
/
shared_memory_total_warp_reduce_op.cc
executable file
·1819 lines (1471 loc) · 95.6 KB
/
shared_memory_total_warp_reduce_op.cc
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
#include "shared_memory_total_warp_reduce_op.hpp"
shared_memory_total_warp_reduce_template_t *init_shared_memory_total_warp_reduce_template(code_builder_t *builder, unsigned long dense_block_id)
{
assert(builder != NULL);
assert(builder->op_manager != NULL);
assert(builder->op_manager->matrix != NULL);
sparse_struct_t *matrix = builder->op_manager->matrix;
assert(matrix->block_coor_table.item_arr.size() > dense_block_id);
// 创建对应的模板
shared_memory_total_warp_reduce_template_t *new_template = new shared_memory_total_warp_reduce_template_t();
new_template->dense_block_index = dense_block_id;
new_template->matrix = matrix;
new_template->kernal_first_row_index = matrix->block_coor_table.item_arr[dense_block_id]->min_dense_row_index;
new_template->kernal_first_col_index = matrix->block_coor_table.item_arr[dense_block_id]->min_dense_col_index;
compressed_block_t *compressed_block_view = matrix->block_coor_table.item_arr[dense_block_id]->compressed_block_ptr;
if (matrix->block_coor_table.item_arr[dense_block_id]->min_dense_col_index == 0 && matrix->block_coor_table.item_arr[dense_block_id]->max_dense_col_index == matrix->dense_col_number - 1)
{
// 稠密子块之间没有共享的行
}
else
{
new_template->is_atom_add = true;
}
// 首先处理每一线程的全局行索引，将全局行索引搞出来
// 分别遍历三个层次的索引
index_of_compress_block_t *block_level_index = compressed_block_view->read_index[2];
index_of_compress_block_t *warp_level_index = compressed_block_view->read_index[3];
index_of_compress_block_t *thread_level_index = compressed_block_view->read_index[4];
assert(block_level_index->level_of_this_index == TBLOCK_LEVEL);
assert(warp_level_index->level_of_this_index == WRAP_LEVEL);
assert(thread_level_index->level_of_this_index == THREAD_LEVEL);
// 行最大值相同
assert(matrix->block_coor_table.item_arr[dense_block_id]->max_dense_row_index <= block_level_index->max_row_index);
assert(matrix->block_coor_table.item_arr[dense_block_id]->compressed_block_ptr->read_index[0]->max_row_index == block_level_index->max_row_index);
// 有效的部分应该只有
new_template->effective_row_num = matrix->block_coor_table.item_arr[dense_block_id]->max_dense_row_index - matrix->block_coor_table.item_arr[dense_block_id]->min_dense_row_index + 1;
if (thread_level_index->row_number_of_block_arr != NULL)
{
cout << "row num in thread level block must be 1, thread level index shouldn't have this metadata" << endl;
assert(false);
}
// 检查warp粒度的块包含的行数量，每个warp粒度的块包含一行
// for (unsigned long warp_level_block_id = 0; warp_level_block_id < warp_level_index->block_num; warp_level_block_id++)
// {
// unsigned long block_row_num = read_from_array_with_data_type(warp_level_index->row_number_of_block_arr, warp_level_index->data_type_of_row_number_of_block_arr, warp_level_block_id);
// if (block_row_num != 1)
// {
// cout << "row num in warp level block must be 1" << endl;
// assert(false);
// }
// }
// 查看线程块粒度的块包含的行，查看线程块粒度的块之间有没有共享行，判断是否需要全局内存中的原子加归约
unsigned long global_min_row_index = read_from_array_with_data_type(block_level_index->index_of_the_first_row_arr, block_level_index->data_type_of_index_of_the_first_row_arr, 0);
unsigned long global_max_row_index = global_min_row_index + read_from_array_with_data_type(block_level_index->row_number_of_block_arr, block_level_index->data_type_of_row_number_of_block_arr, 0) - 1;
for (unsigned long index_of_block_level_index = 1; index_of_block_level_index < block_level_index->block_num; index_of_block_level_index++)
{
// 当前块的起始行号和包含的行的数量，
unsigned long cur_min_row_index = read_from_array_with_data_type(block_level_index->index_of_the_first_row_arr, block_level_index->data_type_of_index_of_the_first_row_arr, index_of_block_level_index);
if (cur_min_row_index <= global_max_row_index)
{
// 代表有重合的部分
new_template->is_atom_add = true;
// 这里的重合代表一行可能有多个线程块，就不支持这个模板了
cout << "several tblock level block in one row, not supported in this template" << endl;
assert(false);
break;
}
unsigned long cur_row_num = read_from_array_with_data_type(block_level_index->row_number_of_block_arr, block_level_index->data_type_of_row_number_of_block_arr, index_of_block_level_index);
if (cur_row_num + cur_min_row_index - 1 > global_max_row_index)
{
// 没有重合就修改元数据
global_max_row_index = cur_min_row_index + cur_row_num - 1;
}
}
// 当前稠密子块的行数量，这个值也是块最大行号+1
unsigned long total_row_num = compressed_block_view->read_index[0]->max_row_index - compressed_block_view->read_index[0]->min_row_index + 1;
vector<unsigned long> new_row_offset_in_warp_tmp_result_vec;
new_row_offset_in_warp_tmp_result_vec.push_back(0);
// 用一个数组记录每一行中间结果的数量，
vector<unsigned long> warp_level_result_num_of_each_row(total_row_num);
// 每个warp粒度的块的首个非零元的索引，使用CSR的索引，一共是warp块的数量+1，这一点和其他的模板不一样
vector<unsigned long> global_warp_block_first_nz_vec(warp_level_index->block_num + 1);
// 每一行的warp粒度的块的数量一开始全都初始化为0
for (unsigned long i = 0; i < total_row_num; i++)
{
warp_level_result_num_of_each_row[i] = 0;
}
// 分别遍历三个层次
// 遍历三个层次的索引，计算每一行warp结果的数量，并且计算每个warp粒度的块的非零元起始索引
for (unsigned long index_of_block_level_index = 0; index_of_block_level_index < block_level_index->block_num; index_of_block_level_index++)
{
// 当前block的首行行号
unsigned long block_first_row_index = read_from_array_with_data_type(block_level_index->index_of_the_first_row_arr, block_level_index->data_type_of_index_of_the_first_row_arr, index_of_block_level_index);
// 当前block第一个非零元索引
unsigned long block_first_nz_index = read_from_array_with_data_type(block_level_index->coo_begin_index_arr, block_level_index->data_type_of_coo_begin_index_arr, index_of_block_level_index);
// block中第一个warp号和下一个block的首warp
unsigned long this_block_first_warp_index = read_from_array_with_data_type(block_level_index->index_arr, block_level_index->index_data_type, index_of_block_level_index);
unsigned long next_block_first_warp_index = read_from_array_with_data_type(block_level_index->index_arr, block_level_index->index_data_type, index_of_block_level_index + 1);
// 当前线程块粒度的块的第一个warp粒度的块的线程粒度的块的大小
assert(this_block_first_warp_index < warp_level_index->block_num && thread_level_index->coo_block_size_arr != NULL);
// 遍历所有warp粒度的索引
for (unsigned long index_of_warp_level_index = this_block_first_warp_index; index_of_warp_level_index < next_block_first_warp_index; index_of_warp_level_index++)
{
assert(index_of_warp_level_index < warp_level_index->block_num);
// 首先查看一个warp的相对行索引
unsigned long warp_first_row_index = read_from_array_with_data_type(warp_level_index->index_of_the_first_row_arr, warp_level_index->data_type_of_index_of_the_first_row_arr, index_of_warp_level_index);
unsigned long warp_first_nz_index = read_from_array_with_data_type(warp_level_index->coo_begin_index_arr, warp_level_index->data_type_of_coo_begin_index_arr, index_of_warp_level_index);
unsigned long block_row_num = read_from_array_with_data_type(warp_level_index->row_number_of_block_arr, warp_level_index->data_type_of_row_number_of_block_arr, index_of_warp_level_index);
if (block_row_num != 1)
{
cout << "row num in warp level block must be 1" << endl;
assert(false);
}
// 没必要进行thread层次的遍历，计算当前全局行号
unsigned long global_warp_row_index = block_first_row_index + warp_first_row_index;
unsigned long global_warp_first_nz_index = warp_first_nz_index + block_first_nz_index;
warp_level_result_num_of_each_row[global_warp_row_index]++;
global_warp_block_first_nz_vec[index_of_warp_level_index] = global_warp_first_nz_index;
// 检查非零元的数量是32的倍数
if (index_of_warp_level_index > 0)
{
assert((global_warp_block_first_nz_vec[index_of_warp_level_index] - global_warp_block_first_nz_vec[index_of_warp_level_index - 1]) % 32 == 0);
}
}
}
// global_warp_block_first_nz_vec最后一位填的是nnz
global_warp_block_first_nz_vec[global_warp_block_first_nz_vec.size() - 1] = matrix->block_coor_table.item_arr[dense_block_id]->compressed_block_ptr->padding_arr_size;
assert((global_warp_block_first_nz_vec[global_warp_block_first_nz_vec.size() - 1] - global_warp_block_first_nz_vec[global_warp_block_first_nz_vec.size() - 2]) % 32 == 0);
// 如果每一行的中间结果只有一个，就不需要shared memory层次的归约
// 遍历每一行结果的数量，从而得出每一行在中间结果中的偏移量
for (unsigned long row_index = 0; row_index < warp_level_result_num_of_each_row.size(); row_index++)
{
new_row_offset_in_warp_tmp_result_vec.push_back(new_row_offset_in_warp_tmp_result_vec[new_row_offset_in_warp_tmp_result_vec.size() - 1] + warp_level_result_num_of_each_row[row_index]);
}
assert(new_row_offset_in_warp_tmp_result_vec.size() == total_row_num + 1);
// 交错存储是没必要的，因为warp内的内容同属于一行
new_template->data_type_of_row_offset_in_warp_tmp_result = find_most_suitable_data_type(new_row_offset_in_warp_tmp_result_vec[new_row_offset_in_warp_tmp_result_vec.size() - 1]);
new_template->size_of_row_offset_in_warp_tmp_result = new_row_offset_in_warp_tmp_result_vec.size();
// 申请数组存储归约信息
new_template->row_offset_in_warp_tmp_result = malloc_arr(new_template->size_of_row_offset_in_warp_tmp_result, new_template->data_type_of_row_offset_in_warp_tmp_result);
copy_unsigned_long_arr_to_others(&(new_row_offset_in_warp_tmp_result_vec[0]), new_template->row_offset_in_warp_tmp_result, new_template->data_type_of_row_offset_in_warp_tmp_result, new_template->size_of_row_offset_in_warp_tmp_result);
// 考虑到空行，最多也是等于最后一个数据
assert(total_row_num >= read_from_array_with_data_type(block_level_index->index_of_the_first_row_arr, block_level_index->data_type_of_index_of_the_first_row_arr, block_level_index->block_num - 1));
// 记录每个block的起始行号，在归约的时候使用，直接就有现成的，拷贝进来即可，但是最后一个需要是所有的行数量，使用的是CSR类型的索引
new_template->data_type_of_block_first_row_index = find_most_suitable_data_type(total_row_num);
new_template->size_of_block_first_row_index = block_level_index->length;
new_template->block_first_row_index = malloc_arr(new_template->size_of_block_first_row_index, new_template->data_type_of_block_first_row_index);
// 不同类型数组之间的拷贝，从block_level_index中拷贝到new_template中
for (unsigned long block_id = 0; block_id < block_level_index->block_num; block_id++)
{
unsigned long source_arr_content = read_from_array_with_data_type(block_level_index->index_of_the_first_row_arr, block_level_index->data_type_of_index_of_the_first_row_arr, block_id);
// 写数据
write_to_array_with_data_type(new_template->block_first_row_index, new_template->data_type_of_block_first_row_index, block_id, source_arr_content);
}
// 最后写一个数据，写的的是行的数量
write_to_array_with_data_type(new_template->block_first_row_index, new_template->data_type_of_block_first_row_index, block_level_index->block_num, total_row_num);
// 每个block的warp粒度的块的数量
new_template->data_type_of_block_begin_warp_index_offset = block_level_index->index_data_type;
new_template->size_of_block_begin_warp_index_offset = block_level_index->length;
new_template->block_begin_warp_index_offset = block_level_index->index_arr;
// 排序相关
// 一些排序的数据
// 最后给出排序索引类型和具体的数组
if (compressed_block_view->y_write_index.size() > 0)
{
// 在子块内排序了
assert(compressed_block_view->is_sorted == true && builder->sub_block_sort_type_vec[dense_block_id] == SUB_BLOCK_SORT && matrix->is_sorted == false);
new_template->global_sort_index = false;
new_template->local_sort_index = true;
// 拷贝
new_template->data_type_of_row_index_before_sort = compressed_block_view->y_write_index[0]->index_data_type;
new_template->row_index_before_sort = compressed_block_view->y_write_index[0]->index_arr;
new_template->size_of_row_index_before_sort = compressed_block_view->y_write_index[0]->length;
}
else if (matrix->sorted_row_index != NULL)
{
cout << "have global sort" << endl;
// 在全局范围内有排序
assert(compressed_block_view->is_sorted == false && matrix->is_sorted == true && builder->sub_block_sort_type_vec[dense_block_id] == GLOBAL_SORT);
new_template->global_sort_index = true;
new_template->local_sort_index = false;
// 拷贝
new_template->data_type_of_row_index_before_sort = matrix->data_type_of_sorted_row_index;
new_template->row_index_before_sort = matrix->sorted_row_index;
new_template->size_of_row_index_before_sort = matrix->dense_row_number;
}
// 每个warp粒度的块的第一个非零元的索引。不需要block索引，采用的CSR的索引方法
new_template->size_of_global_warp_block_first_nz = global_warp_block_first_nz_vec.size();
new_template->data_type_of_global_warp_block_first_nz = find_most_suitable_data_type(global_warp_block_first_nz_vec[global_warp_block_first_nz_vec.size() - 1]);
new_template->global_warp_block_first_nz = malloc_arr(new_template->size_of_global_warp_block_first_nz, new_template->data_type_of_global_warp_block_first_nz);
// 将数据拷贝进来，从global_warp_block_first_nz_vec中拷贝进来，数量比warp粒度的块的数量多一个。
for (unsigned long i = 0; i < global_warp_block_first_nz_vec.size(); i++)
{
unsigned long nz_index = global_warp_block_first_nz_vec[i];
write_to_array_with_data_type(new_template->global_warp_block_first_nz, new_template->data_type_of_global_warp_block_first_nz, i, nz_index);
}
// 列索引的值直接做一个拷贝，可以拷贝交错存储的版本
// 值
new_template->data_type_of_val_arr = compressed_block_view->val_data_type;
new_template->val_arr = compressed_block_view->staggered_padding_val_arr;
new_template->size_of_val_arr = compressed_block_view->staggered_padding_val_arr_size;
// 这两个数组的大小和warpnz的最后一个非零元大小相同
assert(new_template->size_of_val_arr == read_from_array_with_data_type(new_template->global_warp_block_first_nz, new_template->data_type_of_global_warp_block_first_nz, new_template->size_of_global_warp_block_first_nz - 1));
// 列
new_template->data_type_of_col_index_arr = compressed_block_view->read_index[6]->index_data_type;
new_template->col_index_arr = compressed_block_view->read_index[6]->index_arr;
new_template->size_of_col_index_arr = compressed_block_view->read_index[6]->length;
assert(new_template->size_of_val_arr == new_template->size_of_col_index_arr);
return new_template;
}
bool is_supported_by_shared_memory_total_warp_reduce_template(sparse_struct_t *matrix, unsigned long dense_block_id)
{
assert(dense_block_id < matrix->block_coor_table.item_arr.size());
assert(matrix->block_coor_table.item_arr[dense_block_id] != NULL);
assert(matrix->block_coor_table.item_arr[dense_block_id]->compressed_block_ptr != NULL);
assert(matrix->block_coor_table.item_arr[dense_block_id]->compressed_block_ptr->read_index.size() == 7);
compressed_block_t *compressed_block_view = matrix->block_coor_table.item_arr[dense_block_id]->compressed_block_ptr;
index_of_compress_block_t *block_level_index = compressed_block_view->read_index[2];
index_of_compress_block_t *warp_level_index = compressed_block_view->read_index[3];
index_of_compress_block_t *thread_level_index = compressed_block_view->read_index[4];
assert(block_level_index->level_of_this_index == TBLOCK_LEVEL);
assert(warp_level_index->level_of_this_index == WRAP_LEVEL);
assert(thread_level_index->level_of_this_index == THREAD_LEVEL);
if (thread_level_index->row_number_of_block_arr != NULL)
{
return false;
}
// 查看线程块粒度的块包含的行，查看线程块粒度的块之间有没有共享行，判断是否需要全局内存中的原子加归约
unsigned long global_min_row_index = read_from_array_with_data_type(block_level_index->index_of_the_first_row_arr, block_level_index->data_type_of_index_of_the_first_row_arr, 0);
unsigned long global_max_row_index = global_min_row_index + read_from_array_with_data_type(block_level_index->row_number_of_block_arr, block_level_index->data_type_of_row_number_of_block_arr, 0) - 1;
for (unsigned long index_of_block_level_index = 1; index_of_block_level_index < block_level_index->block_num; index_of_block_level_index++)
{
// 当前块的起始行号和包含的行的数量，
unsigned long cur_min_row_index = read_from_array_with_data_type(block_level_index->index_of_the_first_row_arr, block_level_index->data_type_of_index_of_the_first_row_arr, index_of_block_level_index);
if (cur_min_row_index <= global_max_row_index)
{
// 这里的重合代表一行可能有多个线程块，就不支持这个模板了
return false;
}
unsigned long cur_row_num = read_from_array_with_data_type(block_level_index->row_number_of_block_arr, block_level_index->data_type_of_row_number_of_block_arr, index_of_block_level_index);
if (cur_row_num + cur_min_row_index - 1 > global_max_row_index)
{
// 没有重合就修改元数据
global_max_row_index = cur_min_row_index + cur_row_num - 1;
}
}
// 用一个bool判断是不是每个warp对应一行，
bool many_warp_one_row = false;
// 判断当前WLB是不是第一个WLB
bool is_first_WLB = true;
unsigned long last_WLB_first_global_row_index = 0;
// 遍历三个层次
// 遍历三个层次的索引，计算每一行warp结果的数量，并且计算每个warp粒度的块的非零元起始索引
for (unsigned long index_of_block_level_index = 0; index_of_block_level_index < block_level_index->block_num; index_of_block_level_index++)
{
// 当前block的首行行号
unsigned long block_first_row_index = read_from_array_with_data_type(block_level_index->index_of_the_first_row_arr, block_level_index->data_type_of_index_of_the_first_row_arr, index_of_block_level_index);
// 当前block第一个非零元索引
unsigned long block_first_nz_index = read_from_array_with_data_type(block_level_index->coo_begin_index_arr, block_level_index->data_type_of_coo_begin_index_arr, index_of_block_level_index);
// block中第一个warp号和下一个block的首warp
unsigned long this_block_first_warp_index = read_from_array_with_data_type(block_level_index->index_arr, block_level_index->index_data_type, index_of_block_level_index);
unsigned long next_block_first_warp_index = read_from_array_with_data_type(block_level_index->index_arr, block_level_index->index_data_type, index_of_block_level_index + 1);
// 当前线程块粒度的块的第一个warp粒度的块的线程粒度的块的大小
assert(this_block_first_warp_index < warp_level_index->block_num && thread_level_index->coo_block_size_arr != NULL);
// 遍历所有warp粒度的索引
for (unsigned long index_of_warp_level_index = this_block_first_warp_index; index_of_warp_level_index < next_block_first_warp_index; index_of_warp_level_index++)
{
assert(index_of_warp_level_index < warp_level_index->block_num);
// 首先查看一个warp的相对行索引
unsigned long warp_first_row_index = read_from_array_with_data_type(warp_level_index->index_of_the_first_row_arr, warp_level_index->data_type_of_index_of_the_first_row_arr, index_of_warp_level_index);
unsigned long warp_first_nz_index = read_from_array_with_data_type(warp_level_index->coo_begin_index_arr, warp_level_index->data_type_of_coo_begin_index_arr, index_of_warp_level_index);
unsigned long block_row_num = read_from_array_with_data_type(warp_level_index->row_number_of_block_arr, warp_level_index->data_type_of_row_number_of_block_arr, index_of_warp_level_index);
// cout << "block_row_num:" << block_row_num << endl;
if (block_row_num != 1)
{
// 一个warp只能处理一行
return false;
}
// 没必要进行thread层次的遍历，计算当前全局行号
unsigned long global_warp_row_index = block_first_row_index + warp_first_row_index;
unsigned long global_warp_first_nz_index = warp_first_nz_index + block_first_nz_index;
//判断行号是否重合
if (is_first_WLB == true)
{
last_WLB_first_global_row_index = global_warp_row_index;
is_first_WLB = false;
}
else
{
if (last_WLB_first_global_row_index == global_warp_row_index)
{
many_warp_one_row = true;
}
last_WLB_first_global_row_index == global_warp_row_index;
}
}
}
// 如果不是多个warp处理一行，那么检查不通过
if (many_warp_one_row == false)
{
return false;
}
// 这里做一个比较粗的过滤，只要每一个BLB内的WLB数量超过阈值，代表shared memory爆炸了
for (unsigned long BLB_id = 0; BLB_id < block_level_index->block_num; BLB_id++)
{
// 减一下，得出当前BLB中WLB的数量
unsigned long first_WLB_index_of_cur_BLB = read_from_array_with_data_type(block_level_index->index_arr, block_level_index->index_data_type, BLB_id);
unsigned long first_WLB_index_of_next_BLB = read_from_array_with_data_type(block_level_index->index_arr, block_level_index->index_data_type, BLB_id);
assert(first_WLB_index_of_next_BLB > first_WLB_index_of_cur_BLB);
// BLB中WLB的数量
unsigned long WLB_num_of_cur_BLB = first_WLB_index_of_next_BLB - first_WLB_index_of_cur_BLB;
// WLB_num_of_cur_BLB是共享内存的大小，
if (WLB_num_of_cur_BLB > get_config()["SHARED_MEM_TOTAL_SIZE"].as_integer() - 5)
{
return false;
}
}
return true;
}
// 每个warp最多只能负责一行、不能一个warp只负责一行、一行不能被两个线程块处理。
bool is_supported_by_shared_memory_total_warp_reduce_template(code_builder_t *builder, unsigned long dense_block_id)
{
assert(builder != NULL);
assert(builder->op_manager != NULL);
assert(builder->op_manager->matrix != NULL);
sparse_struct_t *matrix = builder->op_manager->matrix;
return is_supported_by_shared_memory_total_warp_reduce_template(matrix, dense_block_id);
}
void store_template_data(shared_memory_total_warp_reduce_template_t *output_template, string output_dir, bool force_not_share_global_sort_index)
{
assert(output_template != NULL);
srand(time(0));
unsigned long matrix_id = rand() + time(0) % 1000;
output_dir = output_dir + "/" + to_string(matrix_id) + "_" + to_string(get_config()["DEFAULT_DEVICE_ID"].as_integer());
// 创建这个文件夹
system(("mkdir " + output_dir).c_str());
// 只要不压缩，就持久化
if (output_template->row_offset_in_warp_tmp_result_compress == NONE_COMPRESS)
{
assert(output_template->row_offset_in_warp_tmp_result != NULL);
print_arr_to_file_with_data_type(output_template->row_offset_in_warp_tmp_result, output_template->data_type_of_row_offset_in_warp_tmp_result, output_template->size_of_row_offset_in_warp_tmp_result, output_dir + "/row_offset_in_warp_tmp_result");
}
if (output_template->block_first_row_index_compress == NONE_COMPRESS)
{
assert(output_template->block_first_row_index != NULL);
print_arr_to_file_with_data_type(output_template->block_first_row_index, output_template->data_type_of_block_first_row_index, output_template->size_of_block_first_row_index, output_dir + "/block_first_row_index");
}
if (output_template->block_begin_warp_index_offset_compress == NONE_COMPRESS)
{
assert(output_template->block_begin_warp_index_offset != NULL);
print_arr_to_file_with_data_type(output_template->block_begin_warp_index_offset, output_template->data_type_of_block_begin_warp_index_offset, output_template->size_of_block_begin_warp_index_offset, output_dir + "/block_begin_warp_index_offset");
}
if (output_template->row_index_before_sort_compress == NONE_COMPRESS && output_template->row_index_before_sort != NULL)
{
assert(output_template->row_index_before_sort != NULL);
// 如果是全局排序，只有第一个才需要存排序之后的行索引
if (output_template->local_sort_index == true)
{
assert(output_template->global_sort_index == false);
print_arr_to_file_with_data_type(output_template->row_index_before_sort, output_template->data_type_of_row_index_before_sort, output_template->size_of_row_index_before_sort, output_dir + "/row_index_before_sort");
}
else if (output_template->global_sort_index == true && (output_template->dense_block_index == 0 || force_not_share_global_sort_index == true))
{
assert(output_template->local_sort_index == false);
print_arr_to_file_with_data_type(output_template->row_index_before_sort, output_template->data_type_of_row_index_before_sort, output_template->size_of_row_index_before_sort, output_dir + "/row_index_before_sort");
}
}
if (output_template->global_warp_block_first_nz_compress == NONE_COMPRESS)
{
assert(output_template->global_warp_block_first_nz != NULL);
print_arr_to_file_with_data_type(output_template->global_warp_block_first_nz, output_template->data_type_of_global_warp_block_first_nz, output_template->size_of_global_warp_block_first_nz, output_dir + "/global_warp_block_first_nz");
}
// 值
assert(output_template->val_arr != NULL);
print_arr_to_file_with_data_type(output_template->val_arr, output_template->data_type_of_val_arr, output_template->size_of_val_arr, output_dir + "/val_arr");
// 列
assert(output_template->col_index_arr != NULL);
print_arr_to_file_with_data_type(output_template->col_index_arr, output_template->data_type_of_col_index_arr, output_template->size_of_col_index_arr, output_dir + "/col_index_arr");
output_template->hash_of_this_template = matrix_id;
}
string code_of_template_data_struct(shared_memory_total_warp_reduce_template_t *output_template, unsigned long dense_block_id)
{
string return_str = "typedef struct compressed_dense_block_" + to_string(dense_block_id) + "\n{\n";
// 对应的位置分别存储行号和块号
if (output_template->row_offset_in_warp_tmp_result_compress == NONE_COMPRESS)
{
assert(output_template->row_offset_in_warp_tmp_result != NULL);
return_str = return_str + code_line_of_pointer_define(output_template->data_type_of_row_offset_in_warp_tmp_result, code_of_arr_var_name(dense_block_id, -1, "row_offset_in_warp_tmp_result"));
return_str = return_str + code_of_data_type(UNSIGNED_LONG) + " size_of_" + code_of_arr_var_name(dense_block_id, -1, "row_offset_in_warp_tmp_result") + " = " + to_string(output_template->size_of_row_offset_in_warp_tmp_result) + ";\n";
}
return_str = return_str + "\n";
if (output_template->block_first_row_index_compress == NONE_COMPRESS)
{
assert(output_template->block_first_row_index != NULL);
return_str = return_str + code_line_of_pointer_define(output_template->data_type_of_block_first_row_index, code_of_arr_var_name(dense_block_id, -1, "block_first_row_index"));
return_str = return_str + code_of_data_type(UNSIGNED_LONG) + " size_of_" + code_of_arr_var_name(dense_block_id, -1, "block_first_row_index") + " = " + to_string(output_template->size_of_block_first_row_index) + ";\n";
}
return_str = return_str + "\n";
if (output_template->block_begin_warp_index_offset_compress == NONE_COMPRESS)
{
assert(output_template->block_begin_warp_index_offset != NULL);
return_str = return_str + code_line_of_pointer_define(output_template->data_type_of_block_begin_warp_index_offset, code_of_arr_var_name(dense_block_id, -1, "block_begin_warp_index_offset"));
return_str = return_str + code_of_data_type(UNSIGNED_LONG) + " size_of_" + code_of_arr_var_name(dense_block_id, -1, "block_begin_warp_index_offset") + " = " + to_string(output_template->size_of_block_begin_warp_index_offset) + ";\n";
}
return_str = return_str + "\n";
if (output_template->row_index_before_sort_compress == NONE_COMPRESS && output_template->row_index_before_sort != NULL)
{
assert(output_template->row_index_before_sort != NULL);
return_str = return_str + code_line_of_pointer_define(output_template->data_type_of_row_index_before_sort, code_of_arr_var_name(dense_block_id, -1, "row_index_before_sort"));
return_str = return_str + code_of_data_type(UNSIGNED_LONG) + " size_of_" + code_of_arr_var_name(dense_block_id, -1, "row_index_before_sort") + " = " + to_string(output_template->size_of_row_index_before_sort) + ";\n";
}
return_str = return_str + "\n";
if (output_template->global_warp_block_first_nz_compress == NONE_COMPRESS)
{
assert(output_template->global_warp_block_first_nz != NULL);
return_str = return_str + code_line_of_pointer_define(output_template->data_type_of_global_warp_block_first_nz, code_of_arr_var_name(dense_block_id, -1, "global_warp_block_first_nz"));
return_str = return_str + code_of_data_type(UNSIGNED_LONG) + " size_of_" + code_of_arr_var_name(dense_block_id, -1, "global_warp_block_first_nz") + " = " + to_string(output_template->size_of_global_warp_block_first_nz) + ";\n";
}
return_str = return_str + "\n";
return_str = return_str + "\n";
assert(output_template->val_arr != NULL);
return_str = return_str + code_line_of_pointer_define(output_template->data_type_of_val_arr, code_of_arr_var_name(dense_block_id, -1, "val_arr"));
return_str = return_str + code_of_data_type(UNSIGNED_LONG) + " size_of_" + code_of_arr_var_name(dense_block_id, -1, "val_arr") + " = " + to_string(output_template->size_of_val_arr) + ";\n";
return_str = return_str + "\n";
assert(output_template->col_index_arr != NULL);
return_str = return_str + code_line_of_pointer_define(output_template->data_type_of_col_index_arr, code_of_arr_var_name(dense_block_id, -1, "col_index_arr"));
return_str = return_str + code_of_data_type(UNSIGNED_LONG) + " size_of_" + code_of_arr_var_name(dense_block_id, -1, "col_index_arr") + " = " + to_string(output_template->size_of_col_index_arr) + ";\n";
return_str = return_str + "}";
return_str = return_str + "compressed_dense_block_" + to_string(dense_block_id) + "_t;\n";
return return_str;
}
string code_of_read_template_data_from_file_func_define(shared_memory_total_warp_reduce_template_t *output_template, unsigned long dense_block_id, bool force_not_share_global_sort_index)
{
string return_str = "compressed_dense_block_" + to_string(dense_block_id) + "_t* read_dense_block_" + to_string(dense_block_id) + "_from_file(string file_name_prefix)\n{\n";
return_str = return_str + "compressed_dense_block_" + to_string(dense_block_id) + "_t *template_data = new " + "compressed_dense_block_" + to_string(dense_block_id) + "_t();\n";
// 对应的位置分别存储行号和块号
if (output_template->row_offset_in_warp_tmp_result_compress == NONE_COMPRESS)
{
assert(output_template->row_offset_in_warp_tmp_result != NULL);
return_str = return_str + "template_data->" + code_of_arr_var_name(dense_block_id, -1, "row_offset_in_warp_tmp_result") + " = (" + code_of_data_type(output_template->data_type_of_row_offset_in_warp_tmp_result) + " *)";
return_str = return_str + "read_arr_from_file_with_data_type(template_data->size_of_" + code_of_arr_var_name(dense_block_id, -1, "row_offset_in_warp_tmp_result") + ", " + convert_data_type_to_string(output_template->data_type_of_row_offset_in_warp_tmp_result) + ", ";
// 要读的文件名
return_str = return_str + "file_name_prefix + \"/row_offset_in_warp_tmp_result\");\n";
}
return_str = return_str + "\n";
if (output_template->block_first_row_index_compress == NONE_COMPRESS)
{
assert(output_template->block_first_row_index != NULL);
return_str = return_str + "template_data->" + code_of_arr_var_name(dense_block_id, -1, "block_first_row_index") + " = (" + code_of_data_type(output_template->data_type_of_block_first_row_index) + " *)";
return_str = return_str + "read_arr_from_file_with_data_type(template_data->size_of_" + code_of_arr_var_name(dense_block_id, -1, "block_first_row_index") + ", " + convert_data_type_to_string(output_template->data_type_of_block_first_row_index) + ", ";
// 要读的文件名
return_str = return_str + "file_name_prefix + \"/block_first_row_index\");\n";
}
return_str = return_str + "\n";
if (output_template->block_begin_warp_index_offset_compress == NONE_COMPRESS)
{
assert(output_template->block_begin_warp_index_offset != NULL);
return_str = return_str + "template_data->" + code_of_arr_var_name(dense_block_id, -1, "block_begin_warp_index_offset") + " = (" + code_of_data_type(output_template->data_type_of_block_begin_warp_index_offset) + " *)";
return_str = return_str + "read_arr_from_file_with_data_type(template_data->size_of_" + code_of_arr_var_name(dense_block_id, -1, "block_begin_warp_index_offset") + ", " + convert_data_type_to_string(output_template->data_type_of_block_begin_warp_index_offset) + ", ";
// 要读的文件名
return_str = return_str + "file_name_prefix + \"/block_begin_warp_index_offset\");\n";
}
return_str = return_str + "\n";
if (output_template->row_index_before_sort_compress == NONE_COMPRESS && output_template->row_index_before_sort != NULL)
{
// 如果有全局的排序索引，只有0号块需要存储
if (output_template->global_sort_index == true)
{
if (dense_block_id == 0 || force_not_share_global_sort_index == true)
{
// 存一个全局的排序
assert(output_template->row_index_before_sort != NULL);
return_str = return_str + "template_data->" + code_of_arr_var_name(dense_block_id, -1, "row_index_before_sort") + " = (" + code_of_data_type(output_template->data_type_of_row_index_before_sort) + " *)";
return_str = return_str + "read_arr_from_file_with_data_type(template_data->size_of_" + code_of_arr_var_name(dense_block_id, -1, "row_index_before_sort") + ", " + convert_data_type_to_string(output_template->data_type_of_row_index_before_sort) + ", ";
// 要读的文件名
return_str = return_str + "file_name_prefix + \"/row_index_before_sort\");\n";
}
else
{
// 如果已经有了就直接拷贝全局的排序
return_str = return_str + "template_data->" + code_of_arr_var_name(dense_block_id, -1, "row_index_before_sort") + " = NULL;\n";
}
}
else if (output_template->local_sort_index == true)
{
assert(output_template->row_index_before_sort != NULL);
return_str = return_str + "template_data->" + code_of_arr_var_name(dense_block_id, -1, "row_index_before_sort") + " = (" + code_of_data_type(output_template->data_type_of_row_index_before_sort) + " *)";
return_str = return_str + "read_arr_from_file_with_data_type(template_data->size_of_" + code_of_arr_var_name(dense_block_id, -1, "row_index_before_sort") + ", " + convert_data_type_to_string(output_template->data_type_of_row_index_before_sort) + ", ";
// 要读的文件名
return_str = return_str + "file_name_prefix + \"/row_index_before_sort\");\n";
}
else
{
cout << "error" << endl;
assert(false);
}
}
return_str = return_str + "\n";
// block和warp的收个非零元索引
if (output_template->global_warp_block_first_nz_compress == NONE_COMPRESS)
{
assert(output_template->global_warp_block_first_nz != NULL);
return_str = return_str + "template_data->" + code_of_arr_var_name(dense_block_id, -1, "global_warp_block_first_nz") + " = (" + code_of_data_type(output_template->data_type_of_global_warp_block_first_nz) + " *)";
return_str = return_str + "read_arr_from_file_with_data_type(template_data->size_of_" + code_of_arr_var_name(dense_block_id, -1, "global_warp_block_first_nz") + ", " + convert_data_type_to_string(output_template->data_type_of_global_warp_block_first_nz) + ", ";
// 要读的文件名
return_str = return_str + "file_name_prefix + \"/global_warp_block_first_nz\");\n";
}
return_str = return_str + "\n";
return_str = return_str + "\n";
assert(output_template->val_arr != NULL);
return_str = return_str + "template_data->" + code_of_arr_var_name(dense_block_id, -1, "val_arr") + " = (" + code_of_data_type(output_template->data_type_of_val_arr) + " *)";
return_str = return_str + "read_arr_from_file_with_data_type(template_data->size_of_" + code_of_arr_var_name(dense_block_id, -1, "val_arr") + ", " + convert_data_type_to_string(output_template->data_type_of_val_arr) + ", ";
// 要读的文件名
return_str = return_str + "file_name_prefix + \"/val_arr\");\n";
return_str = return_str + "\n";
assert(output_template->col_index_arr != NULL);
return_str = return_str + "template_data->" + code_of_arr_var_name(dense_block_id, -1, "col_index_arr") + " = (" + code_of_data_type(output_template->data_type_of_col_index_arr) + " *)";
return_str = return_str + "read_arr_from_file_with_data_type(template_data->size_of_" + code_of_arr_var_name(dense_block_id, -1, "col_index_arr") + ", " + convert_data_type_to_string(output_template->data_type_of_col_index_arr) + ", ";
// 要读的文件名
return_str = return_str + "file_name_prefix + \"/col_index_arr\");\n";
return_str = return_str + "return template_data;\n";
return_str = return_str + "}\n";
return return_str;
}
// kernal，注意归约
string code_of_template_kernal(shared_memory_total_warp_reduce_template_t *output_template, unsigned long dense_block_id)
{
if (output_template->thread_num_of_row_reduce != get_config()["HALF_MAX_ROW_REDUCE_THREAD"].as_integer() && output_template->thread_num_of_row_reduce != get_config()["MAX_ROW_REDUCE_THREAD"].as_integer())
{
assert(output_template->thread_num_in_block % output_template->thread_num_of_row_reduce == 0);
}
assert(output_template->thread_num_in_block % 32 == 0);
// 内核函数的声明
string return_str = "__global__ void spmv_" + to_string(dense_block_id) + "(";
// 用一个变量表明当前形参是不是第一个，如果是第一个就不用点逗号
bool is_first_param = true;
// 这里加入形参的声明
if (output_template->row_offset_in_warp_tmp_result_compress == NONE_COMPRESS)
{
assert(output_template->row_offset_in_warp_tmp_result != NULL);
return_str = return_str + code_of_a_formal_param_declare(output_template->data_type_of_row_offset_in_warp_tmp_result, "* row_offset_in_warp_tmp_result");
is_first_param = false;
}
if (output_template->block_first_row_index_compress == NONE_COMPRESS)
{
if (is_first_param == false)
{
return_str = return_str + ", ";
}
else
{
is_first_param = false;
}
assert(output_template->block_first_row_index != NULL);
return_str = return_str + code_of_a_formal_param_declare(output_template->data_type_of_block_first_row_index, "* block_first_row_index");
}
if (output_template->block_begin_warp_index_offset_compress == NONE_COMPRESS)
{
if (is_first_param == false)
{
return_str = return_str + ", ";
}
else
{
is_first_param = false;
}
assert(output_template->block_begin_warp_index_offset != NULL);
return_str = return_str + code_of_a_formal_param_declare(output_template->data_type_of_block_begin_warp_index_offset, "* block_begin_warp_index_offset");
}
if (output_template->row_index_before_sort_compress == NONE_COMPRESS && output_template->row_index_before_sort != NULL)
{
// 这里代表有排序过
if (is_first_param == false)
{
return_str = return_str + ", ";
}
else
{
is_first_param = false;
}
return_str = return_str + code_of_a_formal_param_declare(output_template->data_type_of_row_index_before_sort, "* row_index_before_sort");
}
if (output_template->global_warp_block_first_nz_compress == NONE_COMPRESS)
{
if (is_first_param == false)
{
return_str = return_str + ", ";
}
else
{
is_first_param = false;
}
assert(output_template->global_warp_block_first_nz != NULL);
return_str = return_str + code_of_a_formal_param_declare(output_template->data_type_of_global_warp_block_first_nz, "* global_warp_block_first_nz");
}
if (is_first_param == false)
{
return_str = return_str + ", ";
}
else
{
is_first_param = false;
}
assert(output_template->val_arr != NULL);
return_str = return_str + code_of_a_formal_param_declare(output_template->data_type_of_val_arr, "* val_arr");
return_str = return_str + ", ";
assert(output_template->col_index_arr != NULL);
return_str = return_str + code_of_a_formal_param_declare(output_template->data_type_of_col_index_arr, "* col_index_arr");
// x的值
return_str = return_str + ", ";
return_str = return_str + code_of_a_formal_param_declare(output_template->data_type_of_val_arr, "* device_x_arr");
// y的值
return_str = return_str + ", ";
return_str = return_str + code_of_a_formal_param_declare(output_template->data_type_of_val_arr, "* device_y_arr");
return_str = return_str + ")\n{\n";
// 这里是kernal的本体，首先判断哪些变量是需要的，哪些是不需要的
// 下一个BLB的第一个WLB的索引，减少一次显存读，当每个BLB中WLB的数量一样，并且等于线程块中warp数量的时候，就不需要下一个BLB的第一个WLB的索引
bool need_next_block_first_warp_index = true;
if (output_template->block_begin_warp_index_offset_compress == LINEAR_COMPRESS)
{
assert(output_template->block_begin_warp_index_offset_compress_meta != NULL);
linear_compress_t *compressor = (linear_compress_t *)output_template->block_begin_warp_index_offset_compress_meta;
if (compressor->coefficient == output_template->thread_num_in_block / 32)
{
// 节省一次读显存
need_next_block_first_warp_index = false;
}
}
// 如果一个WLB中只有32个非零元，就不需要next_warp_block_first_nz用来作为WLB遍历自己内容的下界
bool need_next_warp_block_first_nz = true;
if (output_template->global_warp_block_first_nz_compress == LINEAR_COMPRESS)
{
assert(output_template->global_warp_block_first_nz_compress_meta != NULL);
linear_compress_t *compressor = (linear_compress_t *)output_template->global_warp_block_first_nz_compress_meta;
if (compressor->coefficient == 32)
{
need_next_warp_block_first_nz = false;
}
}
// 线程索引
return_str = return_str + "int bid = blockIdx.x;\n";
return_str = return_str + "int tid_in_block = threadIdx.x;\n";
return_str = return_str + "int wid_in_block = threadIdx.x / 32;\n";
return_str = return_str + "int tid_in_warp = threadIdx.x % 32;\n";
if (need_next_block_first_warp_index == true)
{
return_str = return_str + "int warp_num_in_block = blockDim.x / 32;\n";
}
// 查看是否需要首行和首列索引
if (output_template->kernal_first_row_index != 0)
{
return_str = return_str + code_of_data_type(find_most_suitable_data_type(output_template->kernal_first_row_index + 1)) + " kernal_first_row_index = " + to_string(output_template->kernal_first_row_index) + ";\n";
}
if (output_template->kernal_first_col_index != 0)
{
return_str = return_str + code_of_data_type(find_most_suitable_data_type(output_template->kernal_first_col_index + 1)) + " kernal_first_col_index = " + to_string(output_template->kernal_first_col_index) + ";\n";
}
// 当前shared memory的占用
unsigned long shared_memory_used_size = 0;
// 最大的WLB数量
unsigned long max_WLB_num_of_BLB = 0;
// 遍历所有的BLB，找出WLB数量最多的BLB中WLB的数量
for (unsigned long block_level_block_id = 0; block_level_block_id < output_template->size_of_block_begin_warp_index_offset - 1; block_level_block_id++)
{
// 当前BLB的WLB数量
unsigned long WLB_index_of_this_BLB = read_from_array_with_data_type(output_template->block_begin_warp_index_offset, output_template->data_type_of_block_begin_warp_index_offset, block_level_block_id);
unsigned long WLB_index_of_next_BLB = read_from_array_with_data_type(output_template->block_begin_warp_index_offset, output_template->data_type_of_block_begin_warp_index_offset, block_level_block_id + 1);
unsigned long WLB_num_of_cur_BLB = WLB_index_of_next_BLB - WLB_index_of_this_BLB;
if (WLB_num_of_cur_BLB > max_WLB_num_of_BLB)
{
max_WLB_num_of_BLB = WLB_num_of_cur_BLB;
}
}
return_str = return_str + "\n";
// 申请一个共享内存存储所有warp的中间结果
return_str = return_str + "__shared__ " + code_of_data_type(output_template->data_type_of_val_arr) + " warp_tmp_result_inner_block[" + to_string(max_WLB_num_of_BLB) + "];\n";
shared_memory_used_size = shared_memory_used_size + max_WLB_num_of_BLB;
return_str = return_str + "\n";
// 只要WLB的首行行号没有被压缩，就需要使用共享内存执行广播
if (output_template->block_first_row_index_compress == NONE_COMPRESS)
{
assert(output_template->block_first_row_index_compress_meta == NULL);
return_str = return_str + "__shared__ " + code_of_data_type(output_template->data_type_of_block_first_row_index) + " this_block_first_row_index_shared[1];\n";
return_str = return_str + "__shared__ " + code_of_data_type(output_template->data_type_of_block_first_row_index) + " next_block_first_row_index_shared[1];\n";
shared_memory_used_size = shared_memory_used_size + 2;
}
return_str = return_str + "\n";
// BLB的首个WLB索引没有被压缩，那就要用共享内存来广播
if (output_template->block_begin_warp_index_offset_compress == NONE_COMPRESS)
{
assert(output_template->block_begin_warp_index_offset_compress_meta == NULL);
return_str = return_str + "__shared__ " + code_of_data_type(output_template->data_type_of_block_begin_warp_index_offset) + " this_block_first_warp_index_shared[1];\n";
shared_memory_used_size++;
if (need_next_block_first_warp_index == true)
{
return_str = return_str + "__shared__ " + code_of_data_type(output_template->data_type_of_block_begin_warp_index_offset) + " next_block_first_warp_index_shared[1];\n";
shared_memory_used_size++;
}
}
return_str = return_str + "\n";
// 只要warp中间结果的行偏移索引没有压缩，就需要用shared mem来广播
if (output_template->row_offset_in_warp_tmp_result_compress == NONE_COMPRESS)
{
assert(output_template->row_offset_in_warp_tmp_result_compress_meta == NULL);
return_str = return_str + "__shared__ " + code_of_data_type(output_template->data_type_of_row_offset_in_warp_tmp_result) + " this_block_first_tmp_result_shared[1];\n";
shared_memory_used_size++;
}
// 实际共享内存的占用不能超标
// 尝试不去处理共享内存的超标问题，
if (shared_memory_used_size >= get_config()["SHARED_MEM_TOTAL_SIZE"].as_integer())
{
cout << "shared memory overflow" << endl;
cout << "shared_memory_used_size:" << shared_memory_used_size << endl;
// assert(false);
}
return_str = return_str + "\n";
// 根据block的数量和BLB的数量是否一致，决定要不要，加入for循环
if (output_template->tblock_num == output_template->size_of_block_begin_warp_index_offset - 1)
{
return_str = return_str + "{\n";
// BLB的索引
return_str = return_str + "unsigned int block_level_block_id = bid;\n";
}
else
{
return_str = return_str + "for (unsigned int block_level_block_id = bid; block_level_block_id < " + to_string(output_template->size_of_block_begin_warp_index_offset - 1) + "; block_level_block_id = block_level_block_id + gridDim.x)\n{\n";
}
// 所有block级别元数据的声明
return_str = return_str + code_of_data_type(output_template->data_type_of_block_begin_warp_index_offset) + " this_block_first_warp_index;\n";
if (need_next_block_first_warp_index == true)
{
return_str = return_str + code_of_data_type(output_template->data_type_of_block_begin_warp_index_offset) + " next_block_first_warp_index;\n";
}
return_str = return_str + code_of_data_type(output_template->data_type_of_block_first_row_index) + " this_block_first_row_index;\n";
return_str = return_str + code_of_data_type(output_template->data_type_of_block_first_row_index) + " next_block_first_row_index;\n";
return_str = return_str + "\n";
return_str = return_str + code_of_data_type(output_template->data_type_of_row_offset_in_warp_tmp_result) + " this_block_first_tmp_result;\n";
return_str = return_str + "\n";
return_str = return_str + "__syncthreads();\n\n";
// 块级别的几个元数据的获取
// 只有有东西没被压缩的时候，才需要获取元数据
if (output_template->block_first_row_index_compress == NONE_COMPRESS || output_template->block_begin_warp_index_offset_compress == NONE_COMPRESS)
{
// 当前线程块的第一个线程取数
return_str = return_str + "if (tid_in_block == 0)\n{\n";
// warp的索引如果没有压缩，就需要从全局内存中取
if (output_template->block_begin_warp_index_offset_compress == NONE_COMPRESS)
{
return_str = return_str + "this_block_first_warp_index_shared[0] = block_begin_warp_index_offset[block_level_block_id];\n";
if (need_next_block_first_warp_index == true)
{
return_str = return_str + "next_block_first_warp_index_shared[0] = block_begin_warp_index_offset[block_level_block_id + 1];\n";
}
}
if (output_template->block_first_row_index_compress == NONE_COMPRESS)
{
// 块的首行索引
return_str = return_str + "this_block_first_row_index_shared[0] = block_first_row_index[block_level_block_id];\n";
return_str = return_str + "next_block_first_row_index_shared[0] = block_first_row_index[block_level_block_id + 1];\n";
}
return_str = return_str + "}\n";
return_str = return_str + "__syncthreads();\n\n";
}
// 同步获得行索引和warp索引
// warp索引，压缩和不压缩
if (output_template->block_begin_warp_index_offset_compress == NONE_COMPRESS)
{
return_str = return_str + "this_block_first_warp_index = this_block_first_warp_index_shared[0];\n";
if (need_next_block_first_warp_index == true)
{
return_str = return_str + "next_block_first_warp_index = next_block_first_warp_index_shared[0];\n";
}
}
else if (output_template->block_begin_warp_index_offset_compress == LINEAR_COMPRESS)
{
linear_compress_t *compressor = (linear_compress_t *)output_template->block_begin_warp_index_offset_compress_meta;
assert(compressor != NULL);
return_str = return_str + code_of_arr_read(compressor, "this_block_first_warp_index", "block_level_block_id") + ";\n";
if (need_next_block_first_warp_index == true)
{
// 直接累加斜率，性能更高
return_str = return_str + "next_block_first_warp_index = this_block_first_warp_index + " + to_string(compressor->coefficient) + ";\n";
// return_str = return_str + code_of_arr_read(compressor, "next_block_first_warp_index", "(block_level_block_id + 1)") + ";\n";
}
}
else
{
cout << "compress type is not supported" << endl;
assert(false);
}
// 根据块首行索引的压缩情况执行行索引的计算
if (output_template->block_first_row_index_compress == NONE_COMPRESS)
{
return_str = return_str + "this_block_first_row_index = this_block_first_row_index_shared[0];\n";
return_str = return_str + "next_block_first_row_index = next_block_first_row_index_shared[0];\n";
}
else if (output_template->block_first_row_index_compress == LINEAR_COMPRESS)
{
linear_compress_t *compressor = (linear_compress_t *)output_template->block_first_row_index_compress_meta;
assert(compressor != NULL);
return_str = return_str + code_of_arr_read(compressor, "this_block_first_row_index", "block_level_block_id") + ";\n";
// 线性压缩的第二个直接用累加的方式
return_str = return_str + "next_block_first_row_index = this_block_first_row_index + " + to_string(compressor->coefficient) + ";\n";
// return_str = return_str + code_of_arr_read(compressor, "this_block_first_row_index", "(block_level_block_id + 1)") + ";\n";
}
else
{
cout << "compress type is not supported" << endl;