-
Notifications
You must be signed in to change notification settings - Fork 0
/
peqc.cpp
2224 lines (1938 loc) · 80.4 KB
/
peqc.cpp
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
//
// Created by ylf9811 on 2021/7/6.
//
#include "peqc.h"
using namespace std;
extern "C" {
#include <athread.h>
#include <pthread.h>
void slave_ngspefunc();
void slave_formatpefunc();
void slave_peallfunc();
void slave_decompressfunc();
void slave_compressfunc();
}
void PrintRef(neoReference &ref) {
printf("%.*s\n", ref.lname, (char *) ref.base + ref.pname);
printf("%.*s\n", ref.lseq, (char *) ref.base + ref.pseq);
printf("%.*s\n", ref.lstrand, (char *) ref.base + ref.pstrand);
printf("%.*s\n", ref.lqual, (char *) ref.base + ref.pqual);
}
void SkipToLineEnd(char *data_, int64_t &pos_, const int64_t size_) {
// cerr << "pos: " << pos_ << " size: " << size_ << endl;
ASSERT(pos_ < size_);
while (data_[pos_] != '\n' && data_[pos_] != '\r' && pos_ < size_) ++pos_;
if (data_[pos_] == '\r' && pos_ < size_) {
if (data_[pos_ + 1] == '\n') {
//usesCrlf = true;
++pos_;
}
}
}
int64_t GetNextFastq(char *data_, int64_t pos_, const int64_t size_) {
SkipToLineEnd(data_, pos_, size_);
++pos_;
// find beginning of the next record
while (data_[pos_] != '@') {
SkipToLineEnd(data_, pos_, size_);
++pos_;
}
int64_t pos0 = pos_;
SkipToLineEnd(data_, pos_, size_);
++pos_;
if (data_[pos_] == '@')// previous one was a quality field
return pos_;
//-----[haoz:] is the following code necessary??-------------//
SkipToLineEnd(data_, pos_, size_);
++pos_;
if (data_[pos_] != '+') cout << "core dump is pos: " << pos_ << " char: " << data_[pos_] << endl;
ASSERT(data_[pos_] == '+');// pos0 was the start of tag
return pos0;
}
static bool use_in_mem = 0;
static bool use_out_mem = 0;
static bool use_swidx_file = 0;
#define use_mpi_file
char* OutMemData1;
char* OutMemData2;
/**
* @brief Construct function
* @param cmd_info1 : cmd information
*/
PeQc::PeQc(CmdInfo *cmd_info1, int my_rank_, int comm_size_) {
my_rank = my_rank_;
comm_size = comm_size_;
part_sizes = new int64_t[comm_size];
now_pos1_ = 0;
now_pos2_ = 0;
zip_now_pos1_ = 0;
zip_now_pos2_ = 0;
cmd_info_ = cmd_info1;
filter_ = new Filter(cmd_info1);
done_thread_number_ = 0;
int out_block_nums = int(1.0 * cmd_info1->in_file_size1_ / cmd_info1->out_block_size_);
out_queue_ = NULL;
p_out_queue_ = NULL;
//out_queue1_ = NULL;
//out_queue2_ = NULL;
in_is_zip_ = cmd_info1->in_file_name1_.find(".gz") != string::npos;
out_is_zip_ = cmd_info1->out_file_name1_.find(".gz") != string::npos;
if(in_is_zip_ && comm_size > 1) {
#ifdef USE_LIBDEFLATE
#else
if(my_rank == 0) fprintf(stderr, "compress file input for multi-process is not support, please use -DUSE_LIBDEFLATE in makefile.\n");
exit(0);
#endif
}
ifstream gFile;
gFile.open(cmd_info1->in_file_name1_.c_str());
gFile.seekg(0, ios_base::end);
long long real_file_size = gFile.tellg();
gFile.close();
gFile.open(cmd_info1->in_file_name2_.c_str());
gFile.seekg(0, ios_base::end);
long long real_file_size2 = gFile.tellg();
gFile.close();
if(!in_is_zip_ && real_file_size != real_file_size2 && comm_size > 1) {
fprintf(stderr, "PE in files size error %lld %lld\n", real_file_size, real_file_size2);
exit(0);
}
int64_t start_pos, end_pos;
if(in_is_zip_) {
#ifdef USE_CC_GZ
for(int i = 0; i < 64; i++) {
cc_gz_in_buffer[i] = new char[BLOCK_SIZE];
}
#endif
vector<size_t> block_sizes;
if(use_swidx_file) {
ifstream iff_idx;
iff_idx.open(cmd_info1->in_file_name1_ + ".swidx");
size_t block_size = 0;
while (iff_idx >> block_size) {
block_sizes.push_back(block_size);
}
iff_idx.close();
} else {
ifstream iff_idx;
iff_idx.open(cmd_info1->in_file_name1_, ios::binary | ios::ate);
if (!iff_idx.is_open()) {
cerr << "Failed to open file!" << endl;
exit(0);
}
size_t file_size = iff_idx.tellg();
size_t blocknum = 0;
iff_idx.seekg(-static_cast<int>(sizeof(size_t)), ios::end);
iff_idx.read(reinterpret_cast<char*>(&blocknum), sizeof(size_t));
real_file_size -= (blocknum + 1) * sizeof(size_t);
block_sizes.reserve(blocknum);
iff_idx.seekg(-static_cast<int>((blocknum + 1) * sizeof(size_t)), ios::end);
size_t block_size = 0;
for (size_t i = 0; i < blocknum; ++i) {
iff_idx.read(reinterpret_cast<char*>(&block_size), sizeof(size_t));
block_sizes.push_back(block_size);
}
iff_idx.close();
}
if (block_sizes.empty()) {
// Handle error or empty file scenario
cerr << "Error: No data read from index file." << endl;
exit(0);
}
int pre_size = ceil(1.0 * block_sizes.size() / comm_size);
int start_line = pre_size * my_rank;
int end_line = min(start_line + pre_size, (int) block_sizes.size());
for (int i = 1; i < block_sizes.size(); i++) {
block_sizes[i] += block_sizes[i - 1];
}
start_pos = (start_line == 0) ? 0 : block_sizes[start_line - 1];
end_pos = (end_line == block_sizes.size()) ? real_file_size : block_sizes[end_line - 1];
assert(block_sizes.back() == real_file_size);
start_line_ = start_line;
end_line_ = end_line;
} else {
int64_t pre_size = ceil(1.0 * real_file_size / comm_size);
start_pos = pre_size * my_rank;
end_pos = start_pos + pre_size;
if(end_pos > real_file_size) end_pos = real_file_size;
}
if(in_is_zip_) {
fprintf(stderr, "rank%d line: [%d %d]\n", my_rank, start_line_, end_line_);
}
int64_t right_pos;
if(in_is_zip_) {
right_pos = 0;
} else {
FILE *pre_fp;
pre_fp = fopen(cmd_info1->in_file_name1_.c_str(), "rb");
fseek(pre_fp, start_pos, SEEK_SET);
char *tmp_chunk = new char[1 << 20];
int res_size = fread(tmp_chunk, sizeof(char), 1 << 20, pre_fp);
MPI_Barrier(MPI_COMM_WORLD);
if(my_rank == 0) right_pos = 0;
else right_pos = GetNextFastq(tmp_chunk, 0, res_size);
fclose(pre_fp);
}
MPI_Barrier(MPI_COMM_WORLD);
long long now_pos = right_pos + start_pos;
long long now_poss[comm_size];
now_poss[0] = now_pos;
if(my_rank) {
MPI_Send(&now_pos, 1, MPI_LONG, 0, 0, MPI_COMM_WORLD);
} else {
for(int ii = 1; ii < comm_size; ii++) {
MPI_Recv(&(now_poss[ii]), 1, MPI_LONG, ii, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE);
}
}
MPI_Barrier(MPI_COMM_WORLD);
if(my_rank == 0) {
for(int ii = 1; ii < comm_size; ii++) {
MPI_Send(now_poss, comm_size, MPI_LONG, ii, 0, MPI_COMM_WORLD);
}
} else {
MPI_Recv(now_poss, comm_size, MPI_LONG, 0, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE);
}
MPI_Barrier(MPI_COMM_WORLD);
for(int i = 0; i < comm_size; i++) {
if(i == comm_size - 1) part_sizes[i] = real_file_size - now_poss[i];
else part_sizes[i] = now_poss[i + 1] - now_poss[i];
}
start_pos_ = now_poss[my_rank];
if(my_rank == 0 && start_pos_ != 0) fprintf(stderr, "GG size division\n");
if(my_rank == comm_size - 1) end_pos_ = real_file_size;
else end_pos_ = now_poss[my_rank + 1];
p_out_queue_ = new vector<rabbit::fq::FastqDataPairChunk *>[1 << 20];
p_queueP1 = 0;
p_queueP2 = 0;
p_queueNumNow = 0;
p_queueSizeLim = Q_lim_pe;
if (cmd_info1->write_data_) {
out_queue_ = new pair<CIPair, CIPair>[1 << 20];
queueP1 = 0;
queueP2 = 0;
queueNumNow = 0;
queueSizeLim = Q_lim_pe * 64;
if(use_out_mem) {
OutMemData1 = new char[128 << 20];
OutMemData2 = new char[128 << 20];
}
if (out_is_zip_) {
if (cmd_info1->use_pigz_) {
fprintf(stderr, "use pigz TODO...\n");
exit(0);
} else {
#ifdef Verbose
printf("open gzip stream1 %s\n", cmd_info1->out_file_name1_.c_str());
printf("open gzip stream2 %s\n", cmd_info1->out_file_name2_.c_str());
#endif
#ifdef USE_LIBDEFLATE
ofstream file1(cmd_info1->out_file_name1_.c_str());
if (file1.is_open()) {
file1 << "Hello, file1!" << endl;
file1.close();
} else {
cerr << "Failed to open file1: " << cmd_info1->out_file_name1_ << endl;
}
int err1 = MPI_File_open(MPI_COMM_WORLD, cmd_info1->out_file_name1_.c_str(), MPI_MODE_CREATE | MPI_MODE_WRONLY, MPI_INFO_NULL, &fh1);
if (err1 != MPI_SUCCESS) {
char error_string[BUFSIZ];
int length_of_error_string, error_class;
MPI_Error_class(err1, &error_class);
MPI_Error_string(error_class, error_string, &length_of_error_string);
fprintf(stderr, "%3d: Couldn't open file1, MPI Error: %s\n", my_rank, error_string);
MPI_Abort(MPI_COMM_WORLD, err1);
exit(0);
}
ofstream file2(cmd_info1->out_file_name2_.c_str());
if (file2.is_open()) {
file2 << "Hello, file2!" << endl;
file2.close();
} else {
cerr << "Failed to open file2: " << cmd_info1->out_file_name2_ << endl;
}
int err2 = MPI_File_open(MPI_COMM_WORLD, cmd_info1->out_file_name2_.c_str(), MPI_MODE_CREATE | MPI_MODE_WRONLY, MPI_INFO_NULL, &fh2);
if (err2 != MPI_SUCCESS) {
char error_string[BUFSIZ];
int length_of_error_string, error_class;
MPI_Error_class(err2, &error_class);
MPI_Error_string(error_class, error_string, &length_of_error_string);
fprintf(stderr, "%3d: Couldn't open file, MPI Error: %s\n", my_rank, error_string);
MPI_Abort(MPI_COMM_WORLD, err2);
exit(0);
}
#else
zip_out_stream1 = gzopen(cmd_info1->out_file_name1_.c_str(), "w");
gzsetparams(zip_out_stream1, cmd_info1->compression_level_, Z_DEFAULT_STRATEGY);
gzbuffer(zip_out_stream1, 1024 * 1024);
zip_out_stream2 = gzopen(cmd_info1->out_file_name2_.c_str(), "w");
gzsetparams(zip_out_stream2, cmd_info1->compression_level_, Z_DEFAULT_STRATEGY);
gzbuffer(zip_out_stream2, 1024 * 1024);
#endif
}
} else {
#ifdef Verbose
printf("open stream1 %s\n", cmd_info1->out_file_name1_.c_str());
if (cmd_info_->interleaved_out_ == 0)
printf("open stream2 %s\n", cmd_info1->out_file_name2_.c_str());
#endif
#ifdef use_mpi_file
ofstream file1(cmd_info1->out_file_name1_.c_str());
if (file1.is_open()) {
file1 << "Hello, file!" << endl;
file1.close();
} else {
cerr << "Failed to open file1: " << cmd_info1->out_file_name1_ << endl;
}
int err1 = MPI_File_open(MPI_COMM_WORLD, cmd_info1->out_file_name1_.c_str(), MPI_MODE_CREATE | MPI_MODE_WRONLY, MPI_INFO_NULL, &fh1);
if (err1 != MPI_SUCCESS) {
char error_string[BUFSIZ];
int length_of_error_string, error_class;
MPI_Error_class(err1, &error_class);
MPI_Error_string(error_class, error_string, &length_of_error_string);
fprintf(stderr, "%3d: Couldn't open file, MPI Error: %s\n", my_rank, error_string);
MPI_Abort(MPI_COMM_WORLD, err1);
exit(0);
}
if (cmd_info_->interleaved_out_ == 0) {
ofstream file2(cmd_info1->out_file_name2_.c_str());
if (file2.is_open()) {
file2 << "Hello, file2!" << endl;
file2.close();
} else {
cerr << "Failed to open file2: " << cmd_info1->out_file_name2_ << endl;
}
int err2 = MPI_File_open(MPI_COMM_WORLD, cmd_info1->out_file_name2_.c_str(), MPI_MODE_CREATE | MPI_MODE_WRONLY, MPI_INFO_NULL, &fh2);
if (err2 != MPI_SUCCESS) {
char error_string[BUFSIZ];
int length_of_error_string, error_class;
MPI_Error_class(err2, &error_class);
MPI_Error_string(error_class, error_string, &length_of_error_string);
fprintf(stderr, "%3d: Couldn't open file, MPI Error: %s\n", my_rank, error_string);
MPI_Abort(MPI_COMM_WORLD, err2);
exit(0);
}
}
#else
if(my_rank == 0) {
printf("pre real file1 size %lld\n", real_file_size);
int fd = open(cmd_info1->out_file_name1_.c_str(), O_CREAT | O_TRUNC | O_RDWR | O_EXCL, 0644);
ftruncate(fd, sizeof(char) * real_file_size);
out_stream1_ = fdopen(fd, "w");
} else {
int found = 0;
do {
if(-1 == access(cmd_info1->out_file_name1_.c_str(), F_OK)) {
if(ENOENT == errno) {
//cerr << "waiting file be created..." << endl;
usleep(10000);
} else {
//cerr << "file open GG" << endl;
usleep(10000);
//exit(0);
}
} else {
out_stream1_ = fopen(cmd_info1->out_file_name1_.c_str(), "r+b");
found = 1;
}
} while(found == 0);
}
if (cmd_info_->interleaved_out_ == 0){
if(my_rank == 0) {
printf("pre real file2 size %lld\n", real_file_size2);
int fd = open(cmd_info1->out_file_name2_.c_str(), O_CREAT | O_TRUNC | O_RDWR | O_EXCL, 0644);
ftruncate(fd, sizeof(char) * real_file_size2);
out_stream2_ = fdopen(fd, "w");
} else {
int found = 0;
do {
if(-1 == access(cmd_info1->out_file_name2_.c_str(), F_OK)) {
if(ENOENT == errno) {
//cerr << "waiting file be created..." << endl;
usleep(10000);
} else {
//cerr << "file open GG" << endl;
usleep(10000);
//exit(0);
}
} else {
out_stream2_ = fopen(cmd_info1->out_file_name2_.c_str(), "r+b");
found = 1;
}
} while(found == 0);
}
//MPI_Barrier(MPI_COMM_WORLD);
}
#endif
}
}
duplicate_ = NULL;
if (cmd_info1->state_duplicate_) {
duplicate_ = new Duplicate(cmd_info1);
}
umier_ = NULL;
if (cmd_info1->add_umi_) {
umier_ = new Umier(cmd_info1);
}
if (cmd_info1->use_pugz_) {
fprintf(stderr, "use pugz TODO...\n");
exit(0);
}
if (cmd_info1->use_pigz_) {
fprintf(stderr, "use pigz TODO...\n");
exit(0);
}
pugzDone1 = 0;
pugzDone2 = 0;
writerDone1 = 0;
writerDone2 = 0;
}
PeQc::~PeQc() {
delete[] part_sizes;
delete filter_;
delete[] p_out_queue_;
if (cmd_info_->write_data_) {
delete []out_queue_;
}
if(in_is_zip_) {
#ifdef USE_CC_GZ
for(int i = 0; i < 64; i++) {
delete []cc_gz_in_buffer[i];
}
#endif
}
if (cmd_info_->state_duplicate_) {
delete duplicate_;
}
if (cmd_info_->add_umi_) {
delete umier_;
}
}
string PeQc::Read2String(neoReference &ref) {
return string((char *) ref.base + ref.pname, ref.lname) + "\n" +
string((char *) ref.base + ref.pseq, ref.lseq) + "\n" +
string((char *) ref.base + ref.pstrand, ref.lstrand) + "\n" +
string((char *) ref.base + ref.pqual, ref.lqual) + "\n";
}
void PeQc::Read2Chars(neoReference &ref, char *out_data, int &pos) {
//PrintRef(ref);
memcpy(out_data + pos, ref.base + ref.pname, ref.lname);
pos += ref.lname;
out_data[pos++] = '\n';
memcpy(out_data + pos, ref.base + ref.pseq, ref.lseq);
pos += ref.lseq;
out_data[pos++] = '\n';
memcpy(out_data + pos, ref.base + ref.pstrand, ref.lstrand);
pos += ref.lstrand;
out_data[pos++] = '\n';
memcpy(out_data + pos, ref.base + ref.pqual, ref.lqual);
pos += ref.lqual;
out_data[pos++] = '\n';
}
static rabbit::fq::FastqFileReader *fqFileReader;
void PeQc::ProducerPeFastqTask64(string file, string file2, rabbit::fq::FastqDataPool *fastqPool) {
double t0 = GetTime();
rabbit::uint32 tmpSize = SWAP1_SIZE;
if (cmd_info_->seq_len_ <= 200) tmpSize = SWAP2_SIZE;
int64_t n_chunks = 0;
double t_sum1 = 0;
double t_sum2 = 0;
bool is_first = 1;
int64_t tot_size = 0;
vector<rabbit::fq::FastqDataPairChunk *> tmp_chunks;
while (true) {
double tt0 = GetTime();
rabbit::fq::FastqDataPairChunk *fqdatachunk;
int64_t offset;
if(is_first) {
offset = 0;
for(int i = 0; i < my_rank; i++)
offset += part_sizes[i];
} else {
offset = -1;
}
is_first = 0;
fqdatachunk = fqFileReader->readNextPairChunk();
t_sum1 += GetTime() - tt0;
tt0 = GetTime();
if ((fqdatachunk == NULL) || (fqdatachunk != NULL && fqdatachunk->left_part->size == 1ll << 32)) {
if(tmp_chunks.size()) {
// fprintf(stderr, "producer %d put last %d\n", my_rank, tmp_chunks.size());
while (p_queueNumNow >= p_queueSizeLim) {
usleep(1000);
//fprintf(stderr, "producer %d waiting2\n", my_rank);
}
p_out_queue_[p_queueP2++] = tmp_chunks;
p_queueNumNow++;
tmp_chunks.clear();
n_chunks++;
}
// fprintf(stderr, "producer %d stop, tot chunk size %d\n", my_rank, n_chunks);
tmp_chunks.clear();
while (p_queueNumNow >= p_queueSizeLim) {
usleep(100);
}
p_out_queue_[p_queueP2++] = tmp_chunks;
p_queueNumNow++;
break;
}
tot_size += fqdatachunk->left_part->size;
tmp_chunks.push_back(fqdatachunk);
// fprintf(stderr, "producer %d read a chunk\n", my_rank);
if(tmp_chunks.size() == 64) {
// fprintf(stderr, "producer %d put 64\n", my_rank);
while (p_queueNumNow >= p_queueSizeLim) {
usleep(1000);
}
p_out_queue_[p_queueP2++] = tmp_chunks;
p_queueNumNow++;
tmp_chunks.clear();
n_chunks++;
}
t_sum2 += GetTime() - tt0;
}
//printf("totsize %lld\n", tot_size);
fprintf(stderr, "producer%d sum1 cost %lf\n", my_rank, t_sum1);
fprintf(stderr, "producer%d sum2 cost %lf\n", my_rank, t_sum2);
fprintf(stderr, "producer%d cost %lf\n", my_rank, GetTime() - t0);
fprintf(stderr, "producer %d in func done\n", my_rank);
//dq.SetCompleted();
}
struct formatpe_data {
vector <neoReference> *data1;
vector <neoReference> *data2;
rabbit::fq::FastqDataPairChunk *fqdatachunk;
uint64_t *res;
int fqSize;
int my_rank;
};
struct PeMerge_data {
char* out_data1[64];
char* out_data2[64];
int out_lens1[64] = {0};
int out_lens2[64] = {0};
std::vector <neoReference> *pass_data1[64];
std::vector <neoReference> *pass_data2[64];
int pass_data_size1[64];
int pass_data_size2[64];
};
struct PeAll_data{
formatpe_data para1[64];
qc_data *para2;
PeMerge_data *para3;
int out_len_slave1[64];
int out_len_slave2[64];
bool write_data;
};
double tsum1 = 0;
double tsum2 = 0;
double tsum3 = 0;
double tsum4 = 0;
double tsum4_1 = 0;
double tsum4_1_1 = 0;
double tsum4_1_2 = 0;
double tsum4_1_3 = 0;
double tsum4_2 = 0;
double tsum4_3 = 0;
double tsum4_4 = 0;
double tsum5 = 0;
double tsum6 = 0;
double tsum7 = 0;
double t_wait_producer = 0;
double t_format = 0;
double t_resize = 0;
double t_decom = 0;
double t_ngsfunc = 0;
double t_slave_gz = 0;
double t_slave_gz2 = 0;
double t_push_q = 0;
double t_copy_data = 0;
struct ConsumerWriteInfoPE{
char* buffer;
char* buffer2;
int buffer_len;
int buffer_len2;
long long file_offset;
long long file_offset2;
};
vector<ConsumerWriteInfoPE> writeInfosPE;
static int pe_pre_out_len_slave1[64];
static int pe_pre_out_len_slave2[64];
static int pe_pre_pass_data_size1[64];
static int pe_pre_pass_data_size2[64];
void PeQc::ProcessFormatQCWrite(bool &allIsNull, vector <neoReference> *data1, vector <neoReference> *data2,
vector <neoReference> *pass_data1, vector <neoReference> *pass_data2,
vector <neoReference> *pre_pass_data1, vector <neoReference> *pre_pass_data2,
vector <rabbit::fq::FastqDataPairChunk *> fqdatachunks, vector <rabbit::fq::FastqDataPairChunk *> pre_fqdatachunks,
qc_data *para, rabbit::fq::FastqDataPool *fastq_data_pool) {
double tt0 = GetTime();
formatpe_data para2[64];
uint64_t counts[64] = {0};
int fq_size = fqdatachunks.size();
for(int i = 0; i < 64; i++) {
para2[i].fqSize = fq_size;
para2[i].my_rank = my_rank;
}
for(int i = 0; i < fq_size; i++) {
para2[i].data1 = &(data1[i]);
para2[i].data2 = &(data2[i]);
para2[i].fqdatachunk = fqdatachunks[i];
para2[i].res = counts;
para2[i].fqSize = fq_size;
}
for(int i = 0; i < 64; i++) {
para->data1_[i] = &data1[i];
para->data2_[i] = &data2[i];
para->pass_data1_[i] = &pass_data1[i];
para->pass_data2_[i] = &pass_data2[i];
}
tsum1 += GetTime() - tt0;
tt0 = GetTime();
PeMerge_data pe_merge_data;
PeAll_data pe_all_data;
pe_all_data.write_data = cmd_info_->write_data_;
int out_lens1[64] = {0};
int out_lens2[64] = {0};
char *out_data1[64];
char *out_data2[64];
for(int i = 0; i < 64; i++) {
pe_all_data.para1[i] = para2[i];
pe_all_data.out_len_slave1[i] = 0;
pe_all_data.out_len_slave2[i] = 0;
if(cmd_info_->write_data_) {
out_lens1[i] = pe_pre_out_len_slave1[i];
if(out_lens1[i]) out_data1[i] = new char[out_lens1[i]];
else out_data1[i] = (char *)(NULL);
out_lens2[i] = pe_pre_out_len_slave2[i];
if(out_lens2[i]) out_data2[i] = new char[out_lens2[i]];
else out_data2[i] = (char *)(NULL);
}
}
for(int i = 0; i < 64; i++) {
pe_merge_data.out_data1[i] = out_data1[i];
pe_merge_data.out_lens1[i] = out_lens1[i];
pe_merge_data.pass_data1[i] = &(pre_pass_data1[i]);
pe_merge_data.pass_data_size1[i] = pe_pre_pass_data_size1[i];
pe_merge_data.out_data2[i] = out_data2[i];
pe_merge_data.out_lens2[i] = out_lens2[i];
pe_merge_data.pass_data2[i] = &(pre_pass_data2[i]);
pe_merge_data.pass_data_size2[i] = pe_pre_pass_data_size2[i];
}
pe_all_data.para2 = para;
pe_all_data.para3 = &pe_merge_data;
tsum2 += GetTime() - tt0;
tt0 = GetTime();
__real_athread_spawn((void *)slave_peallfunc, &pe_all_data, 1);
athread_join();
tsum3 += GetTime() - tt0;
tt0 = GetTime();
if (cmd_info_->write_data_) {
long long now_pos_base1 = 0;
long long now_pos_base2 = 0;
bool use_consumer_libdeflate = 0;
#ifdef CONSUMER_USE_LIBDEFLATE
use_consumer_libdeflate = 1;
#endif
if(!out_is_zip_ || use_consumer_libdeflate == 0) {
int out_len1 = 0;
for(int i = 0; i < 64; i++) {
out_len1 += out_lens1[i];
}
if(cmd_info_->splitWrite_ == 0) {
int now_size1 = out_len1;
int now_sizes1[comm_size];
now_sizes1[0] = now_size1;
MPI_Barrier(MPI_COMM_WORLD);
if(my_rank) {
MPI_Send(&now_size1, 1, MPI_INT, 0, 0, MPI_COMM_WORLD);
} else {
for(int ii = 1; ii < comm_size; ii++) {
MPI_Recv(&(now_sizes1[ii]), 1, MPI_INT, ii, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE);
}
}
MPI_Barrier(MPI_COMM_WORLD);
if(my_rank == 0) {
for(int ii = 1; ii < comm_size; ii++) {
MPI_Send(now_sizes1, comm_size, MPI_INT, ii, 0, MPI_COMM_WORLD);
}
} else {
MPI_Recv(now_sizes1, comm_size, MPI_INT, 0, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE);
}
MPI_Barrier(MPI_COMM_WORLD);
int pre_sizes1[comm_size];
pre_sizes1[0] = 0;
for(int ii = 1; ii < comm_size; ii++) {
pre_sizes1[ii] = pre_sizes1[ii - 1] + now_sizes1[ii - 1];
}
now_pos_base1 = now_pos1_ + pre_sizes1[my_rank];
for(int ii = 0; ii < comm_size; ii++) {
now_pos1_ += now_sizes1[ii];
}
int out_len2 = 0;
for(int i = 0; i < 64; i++) {
out_len2 += out_lens2[i];
}
int now_size2 = out_len2;
int now_sizes2[comm_size];
now_sizes2[0] = now_size2;
MPI_Barrier(MPI_COMM_WORLD);
if(my_rank) {
MPI_Send(&now_size2, 1, MPI_INT, 0, 0, MPI_COMM_WORLD);
} else {
for(int ii = 1; ii < comm_size; ii++) {
MPI_Recv(&(now_sizes2[ii]), 1, MPI_INT, ii, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE);
}
}
MPI_Barrier(MPI_COMM_WORLD);
if(my_rank == 0) {
for(int ii = 1; ii < comm_size; ii++) {
MPI_Send(now_sizes2, comm_size, MPI_INT, ii, 0, MPI_COMM_WORLD);
}
} else {
MPI_Recv(now_sizes2, comm_size, MPI_INT, 0, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE);
}
MPI_Barrier(MPI_COMM_WORLD);
int pre_sizes2[comm_size];
pre_sizes2[0] = 0;
for(int ii = 1; ii < comm_size; ii++) {
pre_sizes2[ii] = pre_sizes2[ii - 1] + now_sizes2[ii - 1];
}
now_pos_base2 = now_pos2_ + pre_sizes2[my_rank];
for(int ii = 0; ii < comm_size; ii++) {
now_pos2_ += now_sizes2[ii];
}
}
}
writeInfosPE.clear();
for(int i = 0; i < 64; i++) {
writeInfosPE.push_back({out_data1[i], out_data2[i], out_lens1[i], out_lens2[i], now_pos_base1, now_pos_base2});
}
}
for(int i = 0; i < 64; i++) {
pe_pre_out_len_slave1[i] = pe_all_data.out_len_slave1[i];
pe_pre_out_len_slave2[i] = pe_all_data.out_len_slave2[i];
pe_pre_pass_data_size1[i] = pe_merge_data.pass_data_size1[i];
pe_pre_pass_data_size2[i] = pe_merge_data.pass_data_size2[i];
}
tsum4 += GetTime() - tt0;
tt0 = GetTime();
int isNull = 1;
int all_stop = 1;
for(int i = 0; i < 64; i++) {
if(counts[i]) isNull = 0;
}
if(cmd_info_->splitWrite_ == 0) {
int isNulls[comm_size];
isNulls[0] = isNull;
MPI_Barrier(MPI_COMM_WORLD);
if(my_rank) {
MPI_Send(&isNull, 1, MPI_INT, 0, 0, MPI_COMM_WORLD);
} else {
for(int ii = 1; ii < comm_size; ii++) {
MPI_Recv(&(isNulls[ii]), 1, MPI_INT, ii, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE);
}
}
MPI_Barrier(MPI_COMM_WORLD);
if(my_rank == 0) {
for(int ii = 1; ii < comm_size; ii++) {
MPI_Send(isNulls, comm_size, MPI_INT, ii, 0, MPI_COMM_WORLD);
}
} else {
MPI_Recv(isNulls, comm_size, MPI_INT, 0, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE);
}
for(int ii = 0; ii < comm_size; ii++) {
if(isNulls[ii] == 0) all_stop = 0;
}
if(all_stop) allIsNull = 1;
} else {
allIsNull = isNull;
}
tsum5 += GetTime() - tt0;
tt0 = GetTime();
if(cmd_info_->splitWrite_ == 0 && !all_stop && isNull) {
fprintf(stderr, "consumer%d push null\n", my_rank);
vector<rabbit::fq::FastqDataPairChunk *> tmp_chunks;
tmp_chunks.clear();
while (p_queueNumNow >= p_queueSizeLim) {
usleep(100);
}
p_out_queue_[p_queueP2++] = tmp_chunks;
p_queueNumNow++;
}
tsum6 += GetTime() - tt0;
tt0 = GetTime();
for(int i = 0; i < 64; i++) {
if(pre_fqdatachunks[i] != NULL) {
fastq_data_pool->Release(pre_fqdatachunks[i]->left_part);
fastq_data_pool->Release(pre_fqdatachunks[i]->right_part);
}
}
tsum7 += GetTime() - tt0;
}
inline void gather_and_sort_vectors_pe(int rank, int comm_size, int out_round, vector<pair<int, size_t>>& local_vec, int root) {
int local_size = local_vec.size();
vector<int> send_data(local_size * 2);
for (int i = 0; i < local_size; ++i) {
send_data[2 * i] = local_vec[i].first;
send_data[2 * i + 1] = static_cast<int>(local_vec[i].second); // Assuming size_t can be safely cast to int
}
// Root process initializes a vector to hold all received data
vector<int> recv_data;
if (rank == root) {
recv_data.reserve(local_size * 2 * comm_size); // Reserve enough space
}
// Root process receives data from all other processes
if (rank == root) {
vector<int> buffer(local_size * 2);
for (int i = 0; i < comm_size; ++i) {
if (i == root) {
// Include the root's own data
recv_data.insert(recv_data.end(), send_data.begin(), send_data.end());
} else {
MPI_Recv(buffer.data(), local_size * 2, MPI_INT, i, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE);
recv_data.insert(recv_data.end(), buffer.begin(), buffer.end());
}
}
} else {
// Other processes send their data to the root
MPI_Send(send_data.data(), local_size * 2, MPI_INT, root, 0, MPI_COMM_WORLD);
}
// Root process sorts all received data
if (rank == root) {
vector<pair<int, size_t>> all_data;
for (int i = 0; i < recv_data.size() / 2; ++i) {
all_data.emplace_back(recv_data[2 * i], static_cast<size_t>(recv_data[2 * i + 1]));
}
sort(all_data.begin(), all_data.end());
//// Print sorted data for verification
//for (const auto& p : all_data) {
// cout << "(" << p.first << ", " << p.second << ") ";
//}
//cout << endl;
// Optionally clear and repopulate local_vec
local_vec.clear();
for (const auto& item : all_data) {
local_vec.push_back(item);
}
}
}
void PeQc::ConsumerPeFastqTask64(ThreadInfo **thread_infos, rabbit::fq::FastqDataPool *fastqPool) {
rabbit::int64 id = 0;
rabbit::fq::FastqDataPairChunk *fqdatachunk;
qc_data para;
para.cmd_info_ = cmd_info_;
para.thread_info_ = thread_infos;
para.bit_len = 0;
para.duplicate_ = duplicate_;
bool allIsNull = 0;
if(cmd_info_->state_duplicate_) {
para.bit_len = duplicate_->key_len_base_;
}
athread_init();
double t0 = GetTime();
vector<rabbit::fq::FastqDataPairChunk *> fqdatachunks;
vector<rabbit::fq::FastqDataPairChunk *> pre_fqdatachunks;
int out_round = 0;
vector<neoReference> data1[64];
vector<neoReference> data2[64];
vector<neoReference> pass_data1[64];
vector<neoReference> pass_data2[64];
vector<neoReference> pre_pass_data1[64];
vector<neoReference> pre_pass_data2[64];
int max_data_size = BLOCK_SIZE / (cmd_info_->seq_len_ * 2);
for(int i = 0; i < 64; i++) {
data1[i].clear();
data2[i].clear();
pass_data1[i].clear();
pass_data2[i].clear();
pre_pass_data1[i].clear();
pre_pass_data2[i].clear();
pre_fqdatachunks.push_back(NULL);
data1[i].resize(max_data_size);
data2[i].resize(max_data_size);
if(cmd_info_->write_data_) {
pass_data1[i].resize(max_data_size);
pass_data2[i].resize(max_data_size);
pre_pass_data1[i].resize(max_data_size);
pre_pass_data2[i].resize(max_data_size);
}
}
while(true) {
double tt0 = GetTime();
while (p_queueNumNow == 0) {
usleep(1000);