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
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
|
/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* Copyright by The HDF Group. *
* Copyright by the Board of Trustees of the University of Illinois. *
* All rights reserved. *
* *
* This file is part of HDF5. The full HDF5 copyright notice, including *
* terms governing use, modification, and redistribution, is contained in *
* the COPYING file, which can be found at the root of the source code *
* distribution tree, or in https://www.hdfgroup.org/licenses. *
* If you do not have access to either file, you may request a copy from *
* help@hdfgroup.org. *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
/*-------------------------------------------------------------------------
*
* Created: H5PB.c
*
* Purpose: Page Buffer routines.
*
*-------------------------------------------------------------------------
*/
/****************/
/* Module Setup */
/****************/
#define H5F_FRIEND /* Suppress error about including H5Fpkg */
#include "H5PBmodule.h" /* This source code file is part of the H5PB module */
/***********/
/* Headers */
/***********/
#include "H5private.h" /* Generic Functions */
#include "H5Eprivate.h" /* Error handling */
#include "H5Fpkg.h" /* Files */
#include "H5FDprivate.h" /* File drivers */
#include "H5Iprivate.h" /* IDs */
#include "H5MMprivate.h" /* Memory management */
#include "H5PBpkg.h" /* File access */
#include "H5SLprivate.h" /* Skip List */
/****************/
/* Local Macros */
/****************/
#define H5PB__PREPEND(page_ptr, head_ptr, tail_ptr, len) \
{ \
if ((head_ptr) == NULL) { \
(head_ptr) = (page_ptr); \
(tail_ptr) = (page_ptr); \
} /* end if */ \
else { \
(head_ptr)->prev = (page_ptr); \
(page_ptr)->next = (head_ptr); \
(head_ptr) = (page_ptr); \
} /* end else */ \
(len)++; \
} /* H5PB__PREPEND() */
#define H5PB__REMOVE(page_ptr, head_ptr, tail_ptr, len) \
{ \
if ((head_ptr) == (page_ptr)) { \
(head_ptr) = (page_ptr)->next; \
if ((head_ptr) != NULL) \
(head_ptr)->prev = NULL; \
} /* end if */ \
else \
(page_ptr)->prev->next = (page_ptr)->next; \
if ((tail_ptr) == (page_ptr)) { \
(tail_ptr) = (page_ptr)->prev; \
if ((tail_ptr) != NULL) \
(tail_ptr)->next = NULL; \
} /* end if */ \
else \
(page_ptr)->next->prev = (page_ptr)->prev; \
page_ptr->next = NULL; \
page_ptr->prev = NULL; \
(len)--; \
}
#define H5PB__INSERT_LRU(page_buf, page_ptr) \
{ \
HDassert(page_buf); \
HDassert(page_ptr); \
/* insert the entry at the head of the list. */ \
H5PB__PREPEND((page_ptr), (page_buf)->LRU_head_ptr, (page_buf)->LRU_tail_ptr, \
(page_buf)->LRU_list_len) \
}
#define H5PB__REMOVE_LRU(page_buf, page_ptr) \
{ \
HDassert(page_buf); \
HDassert(page_ptr); \
/* remove the entry from the list. */ \
H5PB__REMOVE((page_ptr), (page_buf)->LRU_head_ptr, (page_buf)->LRU_tail_ptr, \
(page_buf)->LRU_list_len) \
}
#define H5PB__MOVE_TO_TOP_LRU(page_buf, page_ptr) \
{ \
HDassert(page_buf); \
HDassert(page_ptr); \
/* Remove entry and insert at the head of the list. */ \
H5PB__REMOVE((page_ptr), (page_buf)->LRU_head_ptr, (page_buf)->LRU_tail_ptr, \
(page_buf)->LRU_list_len) \
H5PB__PREPEND((page_ptr), (page_buf)->LRU_head_ptr, (page_buf)->LRU_tail_ptr, \
(page_buf)->LRU_list_len) \
}
/******************/
/* Local Typedefs */
/******************/
/* Iteration context for destroying page buffer */
typedef struct {
H5PB_t *page_buf;
hbool_t actual_slist;
} H5PB_ud1_t;
/********************/
/* Package Typedefs */
/********************/
/********************/
/* Local Prototypes */
/********************/
static herr_t H5PB__insert_entry(H5PB_t *page_buf, H5PB_entry_t *page_entry);
static htri_t H5PB__make_space(H5F_shared_t *f_sh, H5PB_t *page_buf, H5FD_mem_t inserted_type);
static herr_t H5PB__write_entry(H5F_shared_t *f_sh, H5PB_entry_t *page_entry);
/*********************/
/* Package Variables */
/*********************/
/*****************************/
/* Library Private Variables */
/*****************************/
/*******************/
/* Local Variables */
/*******************/
/* Declare a free list to manage the H5PB_t struct */
H5FL_DEFINE_STATIC(H5PB_t);
/* Declare a free list to manage the H5PB_entry_t struct */
H5FL_DEFINE_STATIC(H5PB_entry_t);
/*-------------------------------------------------------------------------
* Function: H5PB_reset_stats
*
* Purpose: This function was created without documentation.
* What follows is my best understanding of Mohamad's intent.
*
* Reset statistics collected for the page buffer layer.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
herr_t
H5PB_reset_stats(H5PB_t *page_buf)
{
FUNC_ENTER_NOAPI_NOERR
/* Sanity checks */
HDassert(page_buf);
page_buf->accesses[0] = 0;
page_buf->accesses[1] = 0;
page_buf->hits[0] = 0;
page_buf->hits[1] = 0;
page_buf->misses[0] = 0;
page_buf->misses[1] = 0;
page_buf->evictions[0] = 0;
page_buf->evictions[1] = 0;
page_buf->bypasses[0] = 0;
page_buf->bypasses[1] = 0;
FUNC_LEAVE_NOAPI(SUCCEED)
} /* H5PB_reset_stats() */
/*-------------------------------------------------------------------------
* Function: H5PB_get_stats
*
* Purpose: This function was created without documentation.
* What follows is my best understanding of Mohamad's intent.
*
* Retrieve statistics collected about page accesses for the page buffer layer.
* --accesses: the number of metadata and raw data accesses to the page buffer layer
* --hits: the number of metadata and raw data hits in the page buffer layer
* --misses: the number of metadata and raw data misses in the page buffer layer
* --evictions: the number of metadata and raw data evictions from the page buffer layer
* --bypasses: the number of metadata and raw data accesses that bypass the page buffer layer
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
herr_t
H5PB_get_stats(const H5PB_t *page_buf, unsigned accesses[2], unsigned hits[2], unsigned misses[2],
unsigned evictions[2], unsigned bypasses[2])
{
FUNC_ENTER_NOAPI_NOERR
/* Sanity checks */
HDassert(page_buf);
accesses[0] = page_buf->accesses[0];
accesses[1] = page_buf->accesses[1];
hits[0] = page_buf->hits[0];
hits[1] = page_buf->hits[1];
misses[0] = page_buf->misses[0];
misses[1] = page_buf->misses[1];
evictions[0] = page_buf->evictions[0];
evictions[1] = page_buf->evictions[1];
bypasses[0] = page_buf->bypasses[0];
bypasses[1] = page_buf->bypasses[1];
FUNC_LEAVE_NOAPI(SUCCEED)
} /* H5PB_get_stats */
/*-------------------------------------------------------------------------
* Function: H5PB_print_stats()
*
* Purpose: This function was created without documentation.
* What follows is my best understanding of Mohamad's intent.
*
* Print out statistics collected for the page buffer layer.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
herr_t
H5PB_print_stats(const H5PB_t *page_buf)
{
FUNC_ENTER_NOAPI_NOINIT_NOERR
HDassert(page_buf);
HDprintf("PAGE BUFFER STATISTICS:\n");
HDprintf("******* METADATA\n");
HDprintf("\t Total Accesses: %u\n", page_buf->accesses[0]);
HDprintf("\t Hits: %u\n", page_buf->hits[0]);
HDprintf("\t Misses: %u\n", page_buf->misses[0]);
HDprintf("\t Evictions: %u\n", page_buf->evictions[0]);
HDprintf("\t Bypasses: %u\n", page_buf->bypasses[0]);
HDprintf("\t Hit Rate = %f%%\n",
((double)page_buf->hits[0] / (page_buf->accesses[0] - page_buf->bypasses[0])) * 100);
HDprintf("*****************\n\n");
HDprintf("******* RAWDATA\n");
HDprintf("\t Total Accesses: %u\n", page_buf->accesses[1]);
HDprintf("\t Hits: %u\n", page_buf->hits[1]);
HDprintf("\t Misses: %u\n", page_buf->misses[1]);
HDprintf("\t Evictions: %u\n", page_buf->evictions[1]);
HDprintf("\t Bypasses: %u\n", page_buf->bypasses[1]);
HDprintf("\t Hit Rate = %f%%\n",
((double)page_buf->hits[1] / (page_buf->accesses[1] - page_buf->bypasses[0])) * 100);
HDprintf("*****************\n\n");
FUNC_LEAVE_NOAPI(SUCCEED)
} /* H5PB_print_stats */
/*-------------------------------------------------------------------------
* Function: H5PB_create
*
* Purpose: Create and setup the PB on the file.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
herr_t
H5PB_create(H5F_shared_t *f_sh, size_t size, unsigned page_buf_min_meta_perc, unsigned page_buf_min_raw_perc)
{
H5PB_t *page_buf = NULL;
herr_t ret_value = SUCCEED; /* Return value */
FUNC_ENTER_NOAPI(FAIL)
/* Sanity checks */
HDassert(f_sh);
/* Check args */
if (f_sh->fs_strategy != H5F_FSPACE_STRATEGY_PAGE)
HGOTO_ERROR(H5E_FILE, H5E_CANTINIT, FAIL, "Enabling Page Buffering requires PAGE file space strategy")
/* round down the size if it is larger than the page size */
else if (size > f_sh->fs_page_size) {
hsize_t temp_size;
temp_size = (size / f_sh->fs_page_size) * f_sh->fs_page_size;
H5_CHECKED_ASSIGN(size, size_t, temp_size, hsize_t);
} /* end if */
else if (0 != size % f_sh->fs_page_size)
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTINIT, FAIL, "Page Buffer size must be >= to the page size")
/* Allocate the new page buffering structure */
if (NULL == (page_buf = H5FL_CALLOC(H5PB_t)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_NOSPACE, FAIL, "memory allocation failed")
page_buf->max_size = size;
H5_CHECKED_ASSIGN(page_buf->page_size, size_t, f_sh->fs_page_size, hsize_t);
page_buf->min_meta_perc = page_buf_min_meta_perc;
page_buf->min_raw_perc = page_buf_min_raw_perc;
/* Calculate the minimum page count for metadata and raw data
* based on the fractions provided
*/
page_buf->min_meta_count = (unsigned)((size * page_buf_min_meta_perc) / (f_sh->fs_page_size * 100));
page_buf->min_raw_count = (unsigned)((size * page_buf_min_raw_perc) / (f_sh->fs_page_size * 100));
if (NULL == (page_buf->slist_ptr = H5SL_create(H5SL_TYPE_HADDR, NULL)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTCREATE, FAIL, "can't create skip list")
if (NULL == (page_buf->mf_slist_ptr = H5SL_create(H5SL_TYPE_HADDR, NULL)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTCREATE, FAIL, "can't create skip list")
if (NULL == (page_buf->page_fac = H5FL_fac_init(page_buf->page_size)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTINIT, FAIL, "can't create page factory")
f_sh->page_buf = page_buf;
done:
if (ret_value < 0) {
if (page_buf != NULL) {
if (page_buf->slist_ptr != NULL)
H5SL_close(page_buf->slist_ptr);
if (page_buf->mf_slist_ptr != NULL)
H5SL_close(page_buf->mf_slist_ptr);
if (page_buf->page_fac != NULL)
H5FL_fac_term(page_buf->page_fac);
page_buf = H5FL_FREE(H5PB_t, page_buf);
} /* end if */
} /* end if */
FUNC_LEAVE_NOAPI(ret_value)
} /* H5PB_create */
/*-------------------------------------------------------------------------
* Function: H5PB__flush_cb
*
* Purpose: Callback to flush PB skiplist entries.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
static herr_t
H5PB__flush_cb(void *item, void H5_ATTR_UNUSED *key, void *_op_data)
{
H5PB_entry_t *page_entry = (H5PB_entry_t *)item; /* Pointer to page entry node */
H5F_shared_t *f_sh = (H5F_shared_t *)_op_data;
herr_t ret_value = SUCCEED; /* Return value */
FUNC_ENTER_STATIC
/* Sanity checks */
HDassert(page_entry);
HDassert(f_sh);
/* Flush the page if it's dirty */
if (page_entry->is_dirty)
if (H5PB__write_entry(f_sh, page_entry) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_WRITEERROR, FAIL, "file write failed")
done:
FUNC_LEAVE_NOAPI(ret_value)
} /* H5PB__flush_cb() */
/*-------------------------------------------------------------------------
* Function: H5PB_flush
*
* Purpose: Flush/Free all the PB entries to the file.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
herr_t
H5PB_flush(H5F_shared_t *f_sh)
{
herr_t ret_value = SUCCEED; /* Return value */
FUNC_ENTER_NOAPI(FAIL)
/* Sanity check */
HDassert(f_sh);
/* Flush all the entries in the PB skiplist, if we have write access on the file */
if (f_sh->page_buf && (H5F_ACC_RDWR & H5F_SHARED_INTENT(f_sh))) {
H5PB_t *page_buf = f_sh->page_buf;
/* Iterate over all entries in page buffer skip list */
if (H5SL_iterate(page_buf->slist_ptr, H5PB__flush_cb, f_sh))
HGOTO_ERROR(H5E_PAGEBUF, H5E_BADITER, FAIL, "can't flush page buffer skip list")
} /* end if */
done:
FUNC_LEAVE_NOAPI(ret_value)
} /* H5PB_flush */
/*-------------------------------------------------------------------------
* Function: H5PB__dest_cb
*
* Purpose: Callback to free PB skiplist entries.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
static herr_t
H5PB__dest_cb(void *item, void H5_ATTR_UNUSED *key, void *_op_data)
{
H5PB_entry_t *page_entry = (H5PB_entry_t *)item; /* Pointer to page entry node */
H5PB_ud1_t * op_data = (H5PB_ud1_t *)_op_data;
FUNC_ENTER_STATIC_NOERR
/* Sanity checking */
HDassert(page_entry);
HDassert(op_data);
HDassert(op_data->page_buf);
/* Remove entry from LRU list */
if (op_data->actual_slist) {
H5PB__REMOVE_LRU(op_data->page_buf, page_entry)
page_entry->page_buf_ptr = H5FL_FAC_FREE(op_data->page_buf->page_fac, page_entry->page_buf_ptr);
} /* end if */
/* Free page entry */
page_entry = H5FL_FREE(H5PB_entry_t, page_entry);
FUNC_LEAVE_NOAPI(SUCCEED)
} /* H5PB__dest_cb() */
/*-------------------------------------------------------------------------
* Function: H5PB_dest
*
* Purpose: Flush and destroy the PB on the file if it exists.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
herr_t
H5PB_dest(H5F_shared_t *f_sh)
{
herr_t ret_value = SUCCEED; /* Return value */
FUNC_ENTER_NOAPI(FAIL)
/* Sanity checks */
HDassert(f_sh);
/* flush and destroy the page buffer, if it exists */
if (f_sh->page_buf) {
H5PB_t * page_buf = f_sh->page_buf;
H5PB_ud1_t op_data; /* Iteration context */
if (H5PB_flush(f_sh) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTFLUSH, FAIL, "can't flush page buffer")
/* Set up context info */
op_data.page_buf = page_buf;
/* Destroy the skip list containing all the entries in the PB */
op_data.actual_slist = TRUE;
if (H5SL_destroy(page_buf->slist_ptr, H5PB__dest_cb, &op_data))
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTCLOSEOBJ, FAIL, "can't destroy page buffer skip list")
/* Destroy the skip list containing the new entries */
op_data.actual_slist = FALSE;
if (H5SL_destroy(page_buf->mf_slist_ptr, H5PB__dest_cb, &op_data))
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTCLOSEOBJ, FAIL, "can't destroy page buffer skip list")
/* Destroy the page factory */
if (H5FL_fac_term(page_buf->page_fac) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTRELEASE, FAIL, "can't destroy page buffer page factory")
f_sh->page_buf = H5FL_FREE(H5PB_t, page_buf);
} /* end if */
done:
FUNC_LEAVE_NOAPI(ret_value)
} /* H5PB_dest */
/*-------------------------------------------------------------------------
* Function: H5PB_add_new_page
*
* Purpose: Add a new page to the new page skip list. This is called
* from the MF layer when a new page is allocated to
* indicate to the page buffer layer that a read of the page
* from the file is not necessary since it's an empty page.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
herr_t
H5PB_add_new_page(H5F_shared_t *f_sh, H5FD_mem_t type, haddr_t page_addr)
{
H5PB_t * page_buf; /* Page buffer to operate on */
H5PB_entry_t *page_entry = NULL; /* Pointer to the corresponding page entry */
herr_t ret_value = SUCCEED; /* Return value */
FUNC_ENTER_NOAPI(FAIL)
/* Sanity checks */
HDassert(f_sh);
page_buf = f_sh->page_buf;
HDassert(page_buf);
/* If there is an existing page, this means that at some point the
* file free space manager freed and re-allocated a page at the same
* address. No need to do anything here then...
*/
/* MSC - to be safe, might want to dig in the MF layer and remove
* the page when it is freed from this list if it still exists and
* remove this check
*/
if (NULL == H5SL_search(page_buf->mf_slist_ptr, &(page_addr))) {
/* Create the new PB entry */
if (NULL == (page_entry = H5FL_CALLOC(H5PB_entry_t)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_NOSPACE, FAIL, "memory allocation failed")
/* Initialize page fields */
page_entry->addr = page_addr;
page_entry->type = (H5F_mem_page_t)type;
page_entry->is_dirty = FALSE;
/* Insert entry in skip list */
if (H5SL_insert(page_buf->mf_slist_ptr, page_entry, &(page_entry->addr)) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_BADVALUE, FAIL, "Can't insert entry in skip list")
} /* end if */
done:
if (ret_value < 0)
if (page_entry)
page_entry = H5FL_FREE(H5PB_entry_t, page_entry);
FUNC_LEAVE_NOAPI(ret_value)
} /* H5PB_add_new_page */
/*-------------------------------------------------------------------------
* Function: H5PB_update_entry
*
* Purpose: In PHDF5, entries that are written by other processes and just
* marked clean by this process have to have their corresponding
* pages updated if they exist in the page buffer.
* This routine checks and update the pages.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
herr_t
H5PB_update_entry(H5PB_t *page_buf, haddr_t addr, size_t size, const void *buf)
{
H5PB_entry_t *page_entry; /* Pointer to the corresponding page entry */
haddr_t page_addr;
FUNC_ENTER_NOAPI_NOERR
/* Sanity checks */
HDassert(page_buf);
HDassert(size <= page_buf->page_size);
HDassert(buf);
/* calculate the aligned address of the first page */
page_addr = (addr / page_buf->page_size) * page_buf->page_size;
/* search for the page and update if found */
page_entry = (H5PB_entry_t *)H5SL_search(page_buf->slist_ptr, (void *)(&page_addr));
if (page_entry) {
haddr_t offset;
HDassert(addr + size <= page_addr + page_buf->page_size);
offset = addr - page_addr;
H5MM_memcpy((uint8_t *)page_entry->page_buf_ptr + offset, buf, size);
/* move to top of LRU list */
H5PB__MOVE_TO_TOP_LRU(page_buf, page_entry)
} /* end if */
FUNC_LEAVE_NOAPI(SUCCEED)
} /* H5PB_update_entry */
/*-------------------------------------------------------------------------
* Function: H5PB_remove_entry
*
* Purpose: Remove possible metadata entry with ADDR from the PB cache.
* This is in response to the data corruption bug from fheap.c
* with page buffering + page strategy.
* Note: Large metadata page bypasses the PB cache.
* Note: Update of raw data page (large or small sized) is handled by the PB cache.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Vailin Choi; Feb 2017
*
*-------------------------------------------------------------------------
*/
herr_t
H5PB_remove_entry(const H5F_shared_t *f_sh, haddr_t addr)
{
H5PB_t * page_buf; /* Page buffer to operate on */
H5PB_entry_t *page_entry = NULL; /* Pointer to the page entry being searched */
herr_t ret_value = SUCCEED; /* Return value */
FUNC_ENTER_NOAPI(FAIL)
/* Sanity checks */
HDassert(f_sh);
page_buf = f_sh->page_buf;
HDassert(page_buf);
/* Search for address in the skip list */
page_entry = (H5PB_entry_t *)H5SL_search(page_buf->slist_ptr, (void *)(&addr));
/* If found, remove the entry from the PB cache */
if (page_entry) {
HDassert(page_entry->type != H5F_MEM_PAGE_DRAW);
if (NULL == H5SL_remove(page_buf->slist_ptr, &(page_entry->addr)))
HGOTO_ERROR(H5E_CACHE, H5E_BADVALUE, FAIL, "Page Entry is not in skip list")
/* Remove from LRU list */
H5PB__REMOVE_LRU(page_buf, page_entry)
HDassert(H5SL_count(page_buf->slist_ptr) == page_buf->LRU_list_len);
page_buf->meta_count--;
page_entry->page_buf_ptr = H5FL_FAC_FREE(page_buf->page_fac, page_entry->page_buf_ptr);
page_entry = H5FL_FREE(H5PB_entry_t, page_entry);
} /* end if */
done:
FUNC_LEAVE_NOAPI(ret_value)
} /* H5PB_remove_entry */
/*-------------------------------------------------------------------------
* Function: H5PB_read
*
* Purpose: Reads in the data from the page containing it if it exists
* in the PB cache; otherwise reads in the page through the VFD.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
herr_t
H5PB_read(H5F_shared_t *f_sh, H5FD_mem_t type, haddr_t addr, size_t size, void *buf /*out*/)
{
H5PB_t * page_buf; /* Page buffering info for this file */
H5PB_entry_t *page_entry; /* Pointer to the corresponding page entry */
H5FD_t * file; /* File driver pointer */
haddr_t first_page_addr, last_page_addr; /* Addresses of the first and last pages covered by I/O */
haddr_t offset;
haddr_t search_addr; /* Address of current page */
hsize_t num_touched_pages; /* Number of pages accessed */
size_t access_size = 0;
hbool_t bypass_pb = FALSE; /* Whether to bypass page buffering */
hsize_t i; /* Local index variable */
herr_t ret_value = SUCCEED; /* Return value */
FUNC_ENTER_NOAPI(FAIL)
/* Sanity checks */
HDassert(f_sh);
HDassert(type != H5FD_MEM_GHEAP);
/* Get pointer to page buffer info for this file */
page_buf = f_sh->page_buf;
#ifdef H5_HAVE_PARALLEL
if (H5F_SHARED_HAS_FEATURE(f_sh, H5FD_FEAT_HAS_MPI)) {
#if 1
bypass_pb = TRUE;
#else
/* MSC - why this stopped working ? */
int mpi_size;
if ((mpi_size = H5F_shared_mpi_get_size(f_sh)) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTGET, FAIL, "can't retrieve MPI communicator size")
if (1 != mpi_size)
bypass_pb = TRUE;
#endif
} /* end if */
#endif
/* If page buffering is disabled, or the I/O size is larger than that of a
* single page, or if this is a parallel raw data access, bypass page
* buffering.
*/
if (NULL == page_buf || size >= page_buf->page_size || (bypass_pb && H5FD_MEM_DRAW == type)) {
if (H5F__accum_read(f_sh, type, addr, size, buf) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_READERROR, FAIL, "read through metadata accumulator failed")
/* Update statistics */
if (page_buf) {
if (type == H5FD_MEM_DRAW)
page_buf->bypasses[1]++;
else
page_buf->bypasses[0]++;
} /* end if */
/* If page buffering is disabled, or if this is a large metadata access,
* or if this is parallel raw data access, we are done here
*/
if (NULL == page_buf || (size >= page_buf->page_size && H5FD_MEM_DRAW != type) ||
(bypass_pb && H5FD_MEM_DRAW == type))
HGOTO_DONE(SUCCEED)
} /* end if */
/* Update statistics */
if (page_buf) {
if (type == H5FD_MEM_DRAW)
page_buf->accesses[1]++;
else
page_buf->accesses[0]++;
} /* end if */
/* Calculate the aligned address of the first page */
first_page_addr = (addr / page_buf->page_size) * page_buf->page_size;
/* For Raw data calculate the aligned address of the last page and
* the number of pages accessed if more than 1 page is accessed
*/
if (H5FD_MEM_DRAW == type) {
last_page_addr = ((addr + size - 1) / page_buf->page_size) * page_buf->page_size;
/* How many pages does this write span */
num_touched_pages =
(last_page_addr / page_buf->page_size + 1) - (first_page_addr / page_buf->page_size);
if (first_page_addr == last_page_addr) {
HDassert(1 == num_touched_pages);
last_page_addr = HADDR_UNDEF;
} /* end if */
} /* end if */
/* Otherwise set last page addr to HADDR_UNDEF */
else {
num_touched_pages = 1;
last_page_addr = HADDR_UNDEF;
} /* end else */
/* Translate to file driver I/O info object */
file = f_sh->lf;
/* Copy raw data from dirty pages into the read buffer if the read
request spans pages in the page buffer*/
if (H5FD_MEM_DRAW == type && size >= page_buf->page_size) {
H5SL_node_t *node;
/* For each touched page in the page buffer, check if it
* exists in the page Buffer and is dirty. If it does, we
* update the buffer with what's in the page so we get the up
* to date data into the buffer after the big read from the file.
*/
node = H5SL_find(page_buf->slist_ptr, (void *)(&first_page_addr));
for (i = 0; i < num_touched_pages; i++) {
search_addr = i * page_buf->page_size + first_page_addr;
/* if we still haven't located a starting page, search again */
if (!node && i != 0)
node = H5SL_find(page_buf->slist_ptr, (void *)(&search_addr));
/* if the current page is in the Page Buffer, do the updates */
if (node) {
page_entry = (H5PB_entry_t *)H5SL_item(node);
HDassert(page_entry);
/* If the current page address falls out of the access
block, then there are no more pages to go over */
if (page_entry->addr >= addr + size)
break;
HDassert(page_entry->addr == search_addr);
if (page_entry->is_dirty) {
/* special handling for the first page if it is not a full page access */
if (i == 0 && first_page_addr != addr) {
offset = addr - first_page_addr;
HDassert(page_buf->page_size > offset);
H5MM_memcpy(buf, (uint8_t *)page_entry->page_buf_ptr + offset,
page_buf->page_size - (size_t)offset);
/* move to top of LRU list */
H5PB__MOVE_TO_TOP_LRU(page_buf, page_entry)
} /* end if */
/* special handling for the last page if it is not a full page access */
else if (num_touched_pages > 1 && i == num_touched_pages - 1 &&
search_addr < addr + size) {
offset = (num_touched_pages - 2) * page_buf->page_size +
(page_buf->page_size - (addr - first_page_addr));
H5MM_memcpy((uint8_t *)buf + offset, page_entry->page_buf_ptr,
(size_t)((addr + size) - last_page_addr));
/* move to top of LRU list */
H5PB__MOVE_TO_TOP_LRU(page_buf, page_entry)
} /* end else-if */
/* copy the entire fully accessed pages */
else {
offset = i * page_buf->page_size;
H5MM_memcpy((uint8_t *)buf + (i * page_buf->page_size), page_entry->page_buf_ptr,
page_buf->page_size);
} /* end else */
} /* end if */
node = H5SL_next(node);
} /* end if */
} /* end for */
} /* end if */
else {
/* A raw data access could span 1 or 2 PB entries at this point so
we need to handle that */
HDassert(1 == num_touched_pages || 2 == num_touched_pages);
for (i = 0; i < num_touched_pages; i++) {
haddr_t buf_offset;
/* Calculate the aligned address of the page to search for it in the skip list */
search_addr = (0 == i ? first_page_addr : last_page_addr);
/* Calculate the access size if the access spans more than 1 page */
if (1 == num_touched_pages)
access_size = size;
else
access_size = (0 == i ? (size_t)((first_page_addr + page_buf->page_size) - addr)
: (size - access_size));
/* Lookup the page in the skip list */
page_entry = (H5PB_entry_t *)H5SL_search(page_buf->slist_ptr, (void *)(&search_addr));
/* if found */
if (page_entry) {
offset = (0 == i ? addr - page_entry->addr : 0);
buf_offset = (0 == i ? 0 : size - access_size);
/* copy the requested data from the page into the input buffer */
H5MM_memcpy((uint8_t *)buf + buf_offset, (uint8_t *)page_entry->page_buf_ptr + offset,
access_size);
/* Update LRU */
H5PB__MOVE_TO_TOP_LRU(page_buf, page_entry)
/* Update statistics */
if (type == H5FD_MEM_DRAW)
page_buf->hits[1]++;
else
page_buf->hits[0]++;
} /* end if */
/* if not found */
else {
void * new_page_buf = NULL;
size_t page_size = page_buf->page_size;
haddr_t eoa;
/* make space for new entry */
if ((H5SL_count(page_buf->slist_ptr) * page_buf->page_size) >= page_buf->max_size) {
htri_t can_make_space;
/* check if we can make space in page buffer */
if ((can_make_space = H5PB__make_space(f_sh, page_buf, type)) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_NOSPACE, FAIL, "make space in Page buffer Failed")
/* if make_space returns 0, then we can't use the page
buffer for this I/O and we need to bypass */
if (0 == can_make_space) {
/* make space can't return FALSE on second touched page since the first is of the same
* type */
HDassert(0 == i);
/* read entire block from VFD and return */
if (H5FD_read(file, type, addr, size, buf) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_READERROR, FAIL, "driver read request failed")
/* Break out of loop */
break;
} /* end if */
} /* end if */
/* Read page from VFD */
if (NULL == (new_page_buf = H5FL_FAC_MALLOC(page_buf->page_fac)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTALLOC, FAIL,
"memory allocation failed for page buffer entry")
/* Read page through the VFD layer, but make sure we don't read past the EOA. */
/* Retrieve the 'eoa' for the file */
if (HADDR_UNDEF == (eoa = H5F_shared_get_eoa(f_sh, type)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTGET, FAIL, "driver get_eoa request failed")
/* If the entire page falls outside the EOA, then fail */
if (search_addr > eoa)
HGOTO_ERROR(H5E_PAGEBUF, H5E_BADVALUE, FAIL,
"reading an entire page that is outside the file EOA")
/* Adjust the read size to not go beyond the EOA */
if (search_addr + page_size > eoa)
page_size = (size_t)(eoa - search_addr);
/* Read page from VFD */
if (H5FD_read(file, type, search_addr, page_size, new_page_buf) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_READERROR, FAIL, "driver read request failed")
/* Copy the requested data from the page into the input buffer */
offset = (0 == i ? addr - search_addr : 0);
buf_offset = (0 == i ? 0 : size - access_size);
H5MM_memcpy((uint8_t *)buf + buf_offset, (uint8_t *)new_page_buf + offset, access_size);
/* Create the new PB entry */
if (NULL == (page_entry = H5FL_CALLOC(H5PB_entry_t)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_NOSPACE, FAIL, "memory allocation failed")
page_entry->page_buf_ptr = new_page_buf;
page_entry->addr = search_addr;
page_entry->type = (H5F_mem_page_t)type;
page_entry->is_dirty = FALSE;
/* Insert page into PB */
if (H5PB__insert_entry(page_buf, page_entry) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTSET, FAIL, "error inserting new page in page buffer")
/* Update statistics */
if (type == H5FD_MEM_DRAW)
page_buf->misses[1]++;
else
page_buf->misses[0]++;
} /* end else */
} /* end for */
} /* end else */
done:
FUNC_LEAVE_NOAPI(ret_value)
} /* end H5PB_read() */
/*-------------------------------------------------------------------------
* Function: H5PB_write
*
* Purpose: Write data into the Page Buffer. If the page exists in the
* cache, update it; otherwise read it from disk, update it, and
* insert into cache.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
herr_t
H5PB_write(H5F_shared_t *f_sh, H5FD_mem_t type, haddr_t addr, size_t size, const void *buf)
{
H5PB_t * page_buf; /* Page buffering info for this file */
H5PB_entry_t *page_entry; /* Pointer to the corresponding page entry */
H5FD_t * file; /* File driver pointer */
haddr_t first_page_addr, last_page_addr; /* Addresses of the first and last pages covered by I/O */
haddr_t offset;
haddr_t search_addr; /* Address of current page */
hsize_t num_touched_pages; /* Number of pages accessed */
size_t access_size = 0;
hbool_t bypass_pb = FALSE; /* Whether to bypass page buffering */
hsize_t i; /* Local index variable */
herr_t ret_value = SUCCEED; /* Return value */
FUNC_ENTER_NOAPI(FAIL)
/* Sanity checks */
HDassert(f_sh);
/* Get pointer to page buffer info for this file */
page_buf = f_sh->page_buf;
#ifdef H5_HAVE_PARALLEL
if (H5F_SHARED_HAS_FEATURE(f_sh, H5FD_FEAT_HAS_MPI)) {
#if 1
bypass_pb = TRUE;
#else
/* MSC - why this stopped working ? */
int mpi_size;
if ((mpi_size = H5F_shared_mpi_get_size(f_sh)) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTGET, FAIL, "can't retrieve MPI communicator size")
if (1 != mpi_size)
bypass_pb = TRUE;
#endif
} /* end if */
#endif
/* If page buffering is disabled, or the I/O size is larger than that of a
* single page, or if this is a parallel raw data access, bypass page
* buffering.
*/
if (NULL == page_buf || size >= page_buf->page_size || bypass_pb) {
if (H5F__accum_write(f_sh, type, addr, size, buf) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_WRITEERROR, FAIL, "write through metadata accumulator failed")
/* Update statistics */
if (page_buf) {
if (type == H5FD_MEM_DRAW || type == H5FD_MEM_GHEAP)
page_buf->bypasses[1]++;
else
page_buf->bypasses[0]++;
} /* end if */
/* If page buffering is disabled, or if this is a large metadata access,
* or if this is a parallel raw data access, we are done here
*/
if (NULL == page_buf || (size >= page_buf->page_size && H5FD_MEM_DRAW != type) ||
(bypass_pb && H5FD_MEM_DRAW == type))
HGOTO_DONE(SUCCEED)
#ifdef H5_HAVE_PARALLEL
if (bypass_pb) {
if (H5PB_update_entry(page_buf, addr, size, buf) > 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTUPDATE, FAIL, "failed to update PB with metadata cache")
HGOTO_DONE(SUCCEED)
} /* end if */
#endif
} /* end if */
/* Update statistics */
if (page_buf) {
if (type == H5FD_MEM_DRAW || type == H5FD_MEM_GHEAP)
page_buf->accesses[1]++;
else
page_buf->accesses[0]++;
} /* end if */
/* Calculate the aligned address of the first page */
first_page_addr = (addr / page_buf->page_size) * page_buf->page_size;
/* For raw data calculate the aligned address of the last page and
* the number of pages accessed if more than 1 page is accessed
*/
if (H5FD_MEM_DRAW == type) {
last_page_addr = (addr + size - 1) / page_buf->page_size * page_buf->page_size;
/* how many pages does this write span */
num_touched_pages =
(last_page_addr / page_buf->page_size + 1) - (first_page_addr / page_buf->page_size);
if (first_page_addr == last_page_addr) {
HDassert(1 == num_touched_pages);
last_page_addr = HADDR_UNDEF;
} /* end if */
} /* end if */
/* Otherwise set last page addr to HADDR_UNDEF */
else {
num_touched_pages = 1;
last_page_addr = HADDR_UNDEF;
} /* end else */
/* Translate to file driver I/O info object */
file = f_sh->lf;
/* Check if existing pages for raw data need to be updated since raw data access is not atomic */
if (H5FD_MEM_DRAW == type && size >= page_buf->page_size) {
/* For each touched page, check if it exists in the page buffer, and
* update it with the data in the buffer to keep it up to date
*/
for (i = 0; i < num_touched_pages; i++) {
search_addr = i * page_buf->page_size + first_page_addr;
/* Special handling for the first page if it is not a full page update */
if (i == 0 && first_page_addr != addr) {
/* Lookup the page in the skip list */
page_entry = (H5PB_entry_t *)H5SL_search(page_buf->slist_ptr, (void *)(&search_addr));
if (page_entry) {
offset = addr - first_page_addr;
HDassert(page_buf->page_size > offset);
/* Update page's data */
H5MM_memcpy((uint8_t *)page_entry->page_buf_ptr + offset, buf,
page_buf->page_size - (size_t)offset);
/* Mark page dirty and push to top of LRU */
page_entry->is_dirty = TRUE;
H5PB__MOVE_TO_TOP_LRU(page_buf, page_entry)
} /* end if */
} /* end if */
/* Special handling for the last page if it is not a full page update */
else if (num_touched_pages > 1 && i == (num_touched_pages - 1) &&
(search_addr + page_buf->page_size) != (addr + size)) {
HDassert(search_addr + page_buf->page_size > addr + size);
/* Lookup the page in the skip list */
page_entry = (H5PB_entry_t *)H5SL_search(page_buf->slist_ptr, (void *)(&search_addr));
if (page_entry) {
offset = (num_touched_pages - 2) * page_buf->page_size +
(page_buf->page_size - (addr - first_page_addr));
/* Update page's data */
H5MM_memcpy(page_entry->page_buf_ptr, (const uint8_t *)buf + offset,
(size_t)((addr + size) - last_page_addr));
/* Mark page dirty and push to top of LRU */
page_entry->is_dirty = TRUE;
H5PB__MOVE_TO_TOP_LRU(page_buf, page_entry)
} /* end if */
} /* end else-if */
/* Discard all fully written pages from the page buffer */
else {
page_entry = (H5PB_entry_t *)H5SL_remove(page_buf->slist_ptr, (void *)(&search_addr));
if (page_entry) {
/* Remove from LRU list */
H5PB__REMOVE_LRU(page_buf, page_entry)
/* Decrement page count of appropriate type */
if (H5F_MEM_PAGE_DRAW == page_entry->type || H5F_MEM_PAGE_GHEAP == page_entry->type)
page_buf->raw_count--;
else
page_buf->meta_count--;
/* Free page info */
page_entry->page_buf_ptr = H5FL_FAC_FREE(page_buf->page_fac, page_entry->page_buf_ptr);
page_entry = H5FL_FREE(H5PB_entry_t, page_entry);
} /* end if */
} /* end else */
} /* end for */
} /* end if */
else {
/* An access could span 1 or 2 PBs at this point so we need to handle that */
HDassert(1 == num_touched_pages || 2 == num_touched_pages);
for (i = 0; i < num_touched_pages; i++) {
haddr_t buf_offset;
/* Calculate the aligned address of the page to search for it in the skip list */
search_addr = (0 == i ? first_page_addr : last_page_addr);
/* Calculate the access size if the access spans more than 1 page */
if (1 == num_touched_pages)
access_size = size;
else
access_size =
(0 == i ? (size_t)(first_page_addr + page_buf->page_size - addr) : (size - access_size));
/* Lookup the page in the skip list */
page_entry = (H5PB_entry_t *)H5SL_search(page_buf->slist_ptr, (void *)(&search_addr));
/* If found */
if (page_entry) {
offset = (0 == i ? addr - page_entry->addr : 0);
buf_offset = (0 == i ? 0 : size - access_size);
/* Copy the requested data from the input buffer into the page */
H5MM_memcpy((uint8_t *)page_entry->page_buf_ptr + offset, (const uint8_t *)buf + buf_offset,
access_size);
/* Mark page dirty and push to top of LRU */
page_entry->is_dirty = TRUE;
H5PB__MOVE_TO_TOP_LRU(page_buf, page_entry)
/* Update statistics */
if (type == H5FD_MEM_DRAW || type == H5FD_MEM_GHEAP)
page_buf->hits[1]++;
else
page_buf->hits[0]++;
} /* end if */
/* If not found */
else {
void * new_page_buf;
size_t page_size = page_buf->page_size;
/* Make space for new entry */
if ((H5SL_count(page_buf->slist_ptr) * page_buf->page_size) >= page_buf->max_size) {
htri_t can_make_space;
/* Check if we can make space in page buffer */
if ((can_make_space = H5PB__make_space(f_sh, page_buf, type)) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_NOSPACE, FAIL, "make space in Page buffer Failed")
/* If make_space returns 0, then we can't use the page
* buffer for this I/O and we need to bypass
*/
if (0 == can_make_space) {
HDassert(0 == i);
/* Write to VFD and return */
if (H5FD_write(file, type, addr, size, buf) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_WRITEERROR, FAIL, "driver write request failed")
/* Break out of loop */
break;
} /* end if */
} /* end if */
/* Don't bother searching if there is no write access */
if (H5F_ACC_RDWR & H5F_SHARED_INTENT(f_sh))
/* Lookup & remove the page from the new skip list page if
* it exists to see if this is a new page from the MF layer
*/
page_entry = (H5PB_entry_t *)H5SL_remove(page_buf->mf_slist_ptr, (void *)(&search_addr));
/* Calculate offset into the buffer of the page and the user buffer */
offset = (0 == i ? addr - search_addr : 0);
buf_offset = (0 == i ? 0 : size - access_size);
/* If found, then just update the buffer pointer to the newly allocate buffer */
if (page_entry) {
/* Allocate space for the page buffer */
if (NULL == (new_page_buf = H5FL_FAC_MALLOC(page_buf->page_fac)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTALLOC, FAIL,
"memory allocation failed for page buffer entry")
HDmemset(new_page_buf, 0, (size_t)offset);
HDmemset((uint8_t *)new_page_buf + offset + access_size, 0,
page_size - ((size_t)offset + access_size));
page_entry->page_buf_ptr = new_page_buf;
/* Update statistics */
if (type == H5FD_MEM_DRAW || type == H5FD_MEM_GHEAP)
page_buf->hits[1]++;
else
page_buf->hits[0]++;
} /* end if */
/* Otherwise read page through the VFD layer, but make sure we don't read past the EOA. */
else {
haddr_t eoa, eof = HADDR_UNDEF;
/* Allocate space for the page buffer */
if (NULL == (new_page_buf = H5FL_FAC_CALLOC(page_buf->page_fac)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTALLOC, FAIL,
"memory allocation failed for page buffer entry")
/* Create the new loaded PB entry */
if (NULL == (page_entry = H5FL_CALLOC(H5PB_entry_t)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTALLOC, FAIL, "memory allocation failed")
page_entry->page_buf_ptr = new_page_buf;
page_entry->addr = search_addr;
page_entry->type = (H5F_mem_page_t)type;
/* Retrieve the 'eoa' for the file */
if (HADDR_UNDEF == (eoa = H5F_shared_get_eoa(f_sh, type)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTGET, FAIL, "driver get_eoa request failed")
/* If the entire page falls outside the EOA, then fail */
if (search_addr > eoa)
HGOTO_ERROR(H5E_PAGEBUF, H5E_BADVALUE, FAIL,
"writing to a page that is outside the file EOA")
/* Retrieve the 'eof' for the file - The MPI-VFD EOF
* returned will most likely be HADDR_UNDEF, so skip
* that check.
*/
if (!H5F_SHARED_HAS_FEATURE(f_sh, H5FD_FEAT_HAS_MPI))
if (HADDR_UNDEF == (eof = H5FD_get_eof(f_sh->lf, H5FD_MEM_DEFAULT)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTGET, FAIL, "driver get_eof request failed")
/* Adjust the read size to not go beyond the EOA */
if (search_addr + page_size > eoa)
page_size = (size_t)(eoa - search_addr);
if (search_addr < eof) {
if (H5FD_read(file, type, search_addr, page_size, new_page_buf) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_READERROR, FAIL, "driver read request failed")
/* Update statistics */
if (type == H5FD_MEM_DRAW || type == H5FD_MEM_GHEAP)
page_buf->misses[1]++;
else
page_buf->misses[0]++;
} /* end if */
} /* end else */
/* Copy the requested data from the page into the input buffer */
H5MM_memcpy((uint8_t *)new_page_buf + offset, (const uint8_t *)buf + buf_offset, access_size);
/* Page is dirty now */
page_entry->is_dirty = TRUE;
/* Insert page into PB, evicting other pages as necessary */
if (H5PB__insert_entry(page_buf, page_entry) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTSET, FAIL, "error inserting new page in page buffer")
} /* end else */
} /* end for */
} /* end else */
done:
FUNC_LEAVE_NOAPI(ret_value)
} /* end H5PB_write() */
/*-------------------------------------------------------------------------
* Function: H5PB_enabled
*
* Purpose: Check if the page buffer may be enabled for the specified
* file and data access type.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Neil Fortner
*
*-------------------------------------------------------------------------
*/
herr_t
H5PB_enabled(H5F_shared_t *f_sh, H5FD_mem_t type, hbool_t *enabled)
{
H5PB_t *page_buf; /* Page buffering info for this file */
hbool_t bypass_pb = FALSE; /* Whether to bypass page buffering */
herr_t ret_value = SUCCEED; /* Return value */
FUNC_ENTER_NOAPI_NOERR
/* Sanity checks */
HDassert(f_sh);
/* Get pointer to page buffer info for this file */
page_buf = f_sh->page_buf;
#ifdef H5_HAVE_PARALLEL
if (H5F_SHARED_HAS_FEATURE(f_sh, H5FD_FEAT_HAS_MPI)) {
#if 1
bypass_pb = TRUE;
#else
/* MSC - why this stopped working ? */
int mpi_size;
if ((mpi_size = H5F_shared_mpi_get_size(f_sh)) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTGET, FAIL, "can't retrieve MPI communicator size")
if (1 != mpi_size)
bypass_pb = TRUE;
#endif
} /* end if */
#endif
/* If page buffering is disabled, or if this is a parallel raw data access,
* bypass page buffering. Note that page buffering may still be disabled for
* large metadata access or large non-parallel raw data access, but this
* function doesn't take I/O size into account so if it returns TRUE the
* page buffer may still be disabled for some I/O. If it returns FALSE it is
* always disabled for this access type.
*/
if (NULL == page_buf || (bypass_pb && H5FD_MEM_DRAW == type)) {
/* Update statistics, since wherever this function is called, if it
* returns FALSE, the calling function performs I/O avoiding the page
* buffer layer */
if (page_buf) {
HDassert(type == H5FD_MEM_DRAW);
page_buf->bypasses[1]++;
} /* end if */
/* Page buffer is disabled, at least for this data access type */
*enabled = FALSE;
} /* end if */
else
/* Page buffer may be enabled */
*enabled = TRUE;
FUNC_LEAVE_NOAPI(ret_value)
} /* end H5PB_enabled() */
/*-------------------------------------------------------------------------
* Function: H5PB__insert_entry()
*
* Purpose: This function was created without documentation.
* What follows is my best understanding of Mohamad's intent.
*
* Insert the supplied page into the page buffer, both the
* skip list and the LRU.
*
* As best I can tell, this function imposes no limit on the
* number of entries in the page buffer beyond an assertion
* failure it the page count exceeds the limit.
*
* JRM -- 12/22/16
*
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
static herr_t
H5PB__insert_entry(H5PB_t *page_buf, H5PB_entry_t *page_entry)
{
herr_t ret_value = SUCCEED; /* Return value */
FUNC_ENTER_STATIC
/* Insert entry in skip list */
if (H5SL_insert(page_buf->slist_ptr, page_entry, &(page_entry->addr)) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTINSERT, FAIL, "can't insert entry in skip list")
HDassert(H5SL_count(page_buf->slist_ptr) * page_buf->page_size <= page_buf->max_size);
/* Increment appropriate page count */
if (H5F_MEM_PAGE_DRAW == page_entry->type || H5F_MEM_PAGE_GHEAP == page_entry->type)
page_buf->raw_count++;
else
page_buf->meta_count++;
/* Insert entry in LRU */
H5PB__INSERT_LRU(page_buf, page_entry)
done:
FUNC_LEAVE_NOAPI(ret_value)
} /* end H5PB__insert_entry() */
/*-------------------------------------------------------------------------
* Function: H5PB__make_space()
*
* Purpose: This function was created without documentation.
* What follows is my best understanding of Mohamad's intent.
*
* If necessary and if possible, evict a page from the page
* buffer to make space for the supplied page. Depending on
* the page buffer configuration and contents, and the page
* supplied this may or may not be possible.
*
* JRM -- 12/22/16
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
static htri_t
H5PB__make_space(H5F_shared_t *f_sh, H5PB_t *page_buf, H5FD_mem_t inserted_type)
{
H5PB_entry_t *page_entry; /* Pointer to page eviction candidate */
htri_t ret_value = TRUE; /* Return value */
FUNC_ENTER_STATIC
/* Sanity check */
HDassert(f_sh);
HDassert(page_buf);
/* Get oldest entry */
page_entry = page_buf->LRU_tail_ptr;
if (H5FD_MEM_DRAW == inserted_type) {
/* If threshould is 100% metadata and page buffer is full of
metadata, then we can't make space for raw data */
if (0 == page_buf->raw_count && page_buf->min_meta_count == page_buf->meta_count) {
HDassert(page_buf->meta_count * page_buf->page_size == page_buf->max_size);
HGOTO_DONE(FALSE)
} /* end if */
/* check the metadata threshold before evicting metadata items */
while (1) {
if (page_entry->prev && H5F_MEM_PAGE_META == page_entry->type &&
page_buf->min_meta_count >= page_buf->meta_count)
page_entry = page_entry->prev;
else
break;
} /* end while */
} /* end if */
else {
/* If threshould is 100% raw data and page buffer is full of
raw data, then we can't make space for meta data */
if (0 == page_buf->meta_count && page_buf->min_raw_count == page_buf->raw_count) {
HDassert(page_buf->raw_count * page_buf->page_size == page_buf->max_size);
HGOTO_DONE(FALSE)
} /* end if */
/* check the raw data threshold before evicting raw data items */
while (1) {
if (page_entry->prev &&
(H5F_MEM_PAGE_DRAW == page_entry->type || H5F_MEM_PAGE_GHEAP == page_entry->type) &&
page_buf->min_raw_count >= page_buf->raw_count)
page_entry = page_entry->prev;
else
break;
} /* end while */
} /* end else */
/* Remove from page index */
if (NULL == H5SL_remove(page_buf->slist_ptr, &(page_entry->addr)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_BADVALUE, FAIL, "Tail Page Entry is not in skip list")
/* Remove entry from LRU list */
H5PB__REMOVE_LRU(page_buf, page_entry)
HDassert(H5SL_count(page_buf->slist_ptr) == page_buf->LRU_list_len);
/* Decrement appropriate page type counter */
if (H5F_MEM_PAGE_DRAW == page_entry->type || H5F_MEM_PAGE_GHEAP == page_entry->type)
page_buf->raw_count--;
else
page_buf->meta_count--;
/* Flush page if dirty */
if (page_entry->is_dirty)
if (H5PB__write_entry(f_sh, page_entry) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_WRITEERROR, FAIL, "file write failed")
/* Update statistics */
if (page_entry->type == H5F_MEM_PAGE_DRAW || H5F_MEM_PAGE_GHEAP == page_entry->type)
page_buf->evictions[1]++;
else
page_buf->evictions[0]++;
/* Release page */
page_entry->page_buf_ptr = H5FL_FAC_FREE(page_buf->page_fac, page_entry->page_buf_ptr);
page_entry = H5FL_FREE(H5PB_entry_t, page_entry);
done:
FUNC_LEAVE_NOAPI(ret_value)
} /* end H5PB__make_space() */
/*-------------------------------------------------------------------------
* Function: H5PB__write_entry()
*
* Purpose: ???
*
* This function was created without documentation.
* What follows is my best understanding of Mohamad's intent.
*
* Return: Non-negative on success/Negative on failure
*
* Programmer: Mohamad Chaarawi
*
*-------------------------------------------------------------------------
*/
static herr_t
H5PB__write_entry(H5F_shared_t *f_sh, H5PB_entry_t *page_entry)
{
haddr_t eoa; /* Current EOA for the file */
herr_t ret_value = SUCCEED; /* Return value */
FUNC_ENTER_STATIC
/* Sanity check */
HDassert(f_sh);
HDassert(page_entry);
/* Retrieve the 'eoa' for the file */
if (HADDR_UNDEF == (eoa = H5F_shared_get_eoa(f_sh, (H5FD_mem_t)page_entry->type)))
HGOTO_ERROR(H5E_PAGEBUF, H5E_CANTGET, FAIL, "driver get_eoa request failed")
/* If the starting address of the page is larger than
* the EOA, then the entire page is discarded without writing.
*/
if (page_entry->addr <= eoa) {
H5FD_t *file; /* File driver I/O info */
size_t page_size = f_sh->page_buf->page_size;
/* Adjust the page length if it exceeds the EOA */
if ((page_entry->addr + page_size) > eoa)
page_size = (size_t)(eoa - page_entry->addr);
/* Translate to file driver I/O info object */
file = f_sh->lf;
if (H5FD_write(file, (H5FD_mem_t)page_entry->type, page_entry->addr, page_size,
page_entry->page_buf_ptr) < 0)
HGOTO_ERROR(H5E_PAGEBUF, H5E_WRITEERROR, FAIL, "file write failed")
} /* end if */
page_entry->is_dirty = FALSE;
done:
FUNC_LEAVE_NOAPI(ret_value)
} /* end H5PB__write_entry() */
|