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
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
|
/*****************************************************************************
Copyright (c) 1995, 2017, Oracle and/or its affiliates. All Rights Reserved.
Copyright (c) 2013, 2019, MariaDB Corporation.
This program is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free Software
Foundation; version 2 of the License.
This program is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with
this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1335 USA
*****************************************************************************/
/**************************************************//**
@file include/fil0fil.h
The low-level file system
Created 10/25/1995 Heikki Tuuri
*******************************************************/
#ifndef fil0fil_h
#define fil0fil_h
#include "fsp0types.h"
#ifndef UNIV_INNOCHECKSUM
#include "log0recv.h"
#include "dict0types.h"
#ifdef UNIV_LINUX
# include <set>
#endif
// Forward declaration
extern my_bool srv_use_doublewrite_buf;
extern struct buf_dblwr_t* buf_dblwr;
class page_id_t;
/** Structure containing encryption specification */
struct fil_space_crypt_t;
/** File types */
enum fil_type_t {
/** temporary tablespace (temporary undo log or tables) */
FIL_TYPE_TEMPORARY,
/** a tablespace that is being imported (no logging until finished) */
FIL_TYPE_IMPORT,
/** persistent tablespace (for system, undo log or tables) */
FIL_TYPE_TABLESPACE,
/** redo log covering changes to files of FIL_TYPE_TABLESPACE */
FIL_TYPE_LOG
};
/** Check if fil_type is any of FIL_TYPE_TEMPORARY, FIL_TYPE_IMPORT
or FIL_TYPE_TABLESPACE.
@param[in] type variable of type fil_type_t
@return true if any of FIL_TYPE_TEMPORARY, FIL_TYPE_IMPORT
or FIL_TYPE_TABLESPACE */
inline
bool
fil_type_is_data(
fil_type_t type)
{
return(type == FIL_TYPE_TEMPORARY
|| type == FIL_TYPE_IMPORT
|| type == FIL_TYPE_TABLESPACE);
}
struct fil_node_t;
#endif
/** Tablespace or log data space */
struct fil_space_t {
#ifndef UNIV_INNOCHECKSUM
ulint id; /*!< space id */
hash_node_t hash; /*!< hash chain node */
char* name; /*!< Tablespace name */
lsn_t max_lsn;
/*!< LSN of the most recent
fil_names_write_if_was_clean().
Reset to 0 by fil_names_clear().
Protected by log_sys.mutex.
If and only if this is nonzero, the
tablespace will be in named_spaces. */
/** Log sequence number of the latest MLOG_INDEX_LOAD record
that was found while parsing the redo log */
lsn_t enable_lsn;
bool stop_new_ops;
/*!< we set this true when we start
deleting a single-table tablespace.
When this is set following new ops
are not allowed:
* read IO request
* ibuf merge
* file flush
Note that we can still possibly have
new write operations because we don't
check this flag when doing flush
batches. */
/** whether undo tablespace truncation is in progress */
bool is_being_truncated;
#ifdef UNIV_DEBUG
/** reference count for operations who want to skip redo log in the
file space in order to make modify_check() pass. */
Atomic_counter<ulint> redo_skipped_count;
#endif
fil_type_t purpose;/*!< purpose */
UT_LIST_BASE_NODE_T(fil_node_t) chain;
/*!< base node for the file chain */
ulint size; /*!< tablespace file size in pages;
0 if not known yet */
ulint size_in_header;
/* FSP_SIZE in the tablespace header;
0 if not known yet */
ulint free_len;
/*!< length of the FSP_FREE list */
ulint free_limit;
/*!< contents of FSP_FREE_LIMIT */
ulint recv_size;
/*!< recovered tablespace size in pages;
0 if no size change was read from the redo log,
or if the size change was implemented */
ulint n_reserved_extents;
/*!< number of reserved free extents for
ongoing operations like B-tree page split */
ulint n_pending_flushes; /*!< this is positive when flushing
the tablespace to disk; dropping of the
tablespace is forbidden if this is positive */
/** Number of pending buffer pool operations accessing the tablespace
without holding a table lock or dict_sys.latch S-latch
that would prevent the table (and tablespace) from being
dropped. An example is change buffer merge.
The tablespace cannot be dropped while this is nonzero,
or while fil_node_t::n_pending is nonzero.
Protected by fil_system.mutex and std::atomic. */
std::atomic<ulint> n_pending_ops;
/** Number of pending block read or write operations
(when a write is imminent or a read has recently completed).
The tablespace object cannot be freed while this is nonzero,
but it can be detached from fil_system.
Note that fil_node_t::n_pending tracks actual pending I/O requests.
Protected by fil_system.mutex and std::atomic. */
std::atomic<ulint> n_pending_ios;
rw_lock_t latch; /*!< latch protecting the file space storage
allocation */
UT_LIST_NODE_T(fil_space_t) unflushed_spaces;
/*!< list of spaces with at least one unflushed
file we have written to */
UT_LIST_NODE_T(fil_space_t) named_spaces;
/*!< list of spaces for which MLOG_FILE_NAME
records have been issued */
/** Checks that this tablespace in a list of unflushed tablespaces.
@return true if in a list */
bool is_in_unflushed_spaces() const;
UT_LIST_NODE_T(fil_space_t) space_list;
/*!< list of all spaces */
/** other tablespaces needing key rotation */
UT_LIST_NODE_T(fil_space_t) rotation_list;
/** Checks that this tablespace needs key rotation.
@return true if in a rotation list */
bool is_in_rotation_list() const;
/** MariaDB encryption data */
fil_space_crypt_t* crypt_data;
/** True if the device this filespace is on supports atomic writes */
bool atomic_write_supported;
/** True if file system storing this tablespace supports
punch hole */
bool punch_hole;
ulint magic_n;/*!< FIL_SPACE_MAGIC_N */
/** @return whether the tablespace is about to be dropped */
bool is_stopping() const { return stop_new_ops; }
/** @return whether doublewrite buffering is needed */
bool use_doublewrite() const
{
return !atomic_write_supported
&& srv_use_doublewrite_buf && buf_dblwr;
}
/** Append a file to the chain of files of a space.
@param[in] name file name of a file that is not open
@param[in] handle file handle, or OS_FILE_CLOSED
@param[in] size file size in entire database pages
@param[in] is_raw whether this is a raw device
@param[in] atomic_write true if atomic write could be enabled
@param[in] max_pages maximum number of pages in file,
or ULINT_MAX for unlimited
@return file object */
fil_node_t* add(const char* name, pfs_os_file_t handle,
ulint size, bool is_raw, bool atomic_write,
ulint max_pages = ULINT_MAX);
#ifdef UNIV_DEBUG
/** Assert that the mini-transaction is compatible with
updating an allocation bitmap page.
@param[in] mtr mini-transaction */
void modify_check(const mtr_t& mtr) const;
#endif /* UNIV_DEBUG */
/** Try to reserve free extents.
@param[in] n_free_now current number of free extents
@param[in] n_to_reserve number of extents to reserve
@return whether the reservation succeeded */
bool reserve_free_extents(ulint n_free_now, ulint n_to_reserve)
{
ut_ad(rw_lock_own(&latch, RW_LOCK_X));
if (n_reserved_extents + n_to_reserve > n_free_now) {
return false;
}
n_reserved_extents += n_to_reserve;
return true;
}
/** Release the reserved free extents.
@param[in] n_reserved number of reserved extents */
void release_free_extents(ulint n_reserved)
{
if (!n_reserved) return;
ut_ad(rw_lock_own(&latch, RW_LOCK_X));
ut_a(n_reserved_extents >= n_reserved);
n_reserved_extents -= n_reserved;
}
/** Rename a file.
@param[in] name table name after renaming
@param[in] path tablespace file name after renaming
@param[in] log whether to write redo log
@param[in] replace whether to ignore the existence of path
@return error code
@retval DB_SUCCESS on success */
dberr_t rename(const char* name, const char* path, bool log,
bool replace = false);
/** Note that the tablespace has been imported.
Initially, purpose=FIL_TYPE_IMPORT so that no redo log is
written while the space ID is being updated in each page. */
inline void set_imported();
/** @return whether the storage device is rotational (HDD, not SSD) */
inline bool is_rotational() const;
/** Open each file. Only invoked on fil_system.temp_space.
@return whether all files were opened */
bool open();
/** Close each file. Only invoked on fil_system.temp_space. */
void close();
/** Acquire a tablespace reference. */
void acquire() { n_pending_ops++; }
/** Release a tablespace reference. */
void release() { ut_ad(referenced()); n_pending_ops--; }
/** @return whether references are being held */
bool referenced() const { return n_pending_ops; }
/** Acquire a tablespace reference for I/O. */
void acquire_for_io() { n_pending_ios++; }
/** Release a tablespace reference for I/O. */
void release_for_io() { ut_ad(pending_io()); n_pending_ios--; }
/** @return whether I/O is pending */
bool pending_io() const { return n_pending_ios; }
#endif
/** FSP_SPACE_FLAGS and FSP_FLAGS_MEM_ flags;
check fsp0types.h to more info about flags. */
ulint flags;
/** Determine if full_crc32 is used for a data file
@param[in] flags tablespace flags (FSP_FLAGS)
@return whether the full_crc32 algorithm is active */
static bool full_crc32(ulint flags) {
return flags & FSP_FLAGS_FCRC32_MASK_MARKER;
}
/** @return whether innodb_checksum_algorithm=full_crc32 is active */
bool full_crc32() const { return full_crc32(flags); }
/** Determine the logical page size.
@param flags tablespace flags (FSP_FLAGS)
@return the logical page size
@retval 0 if the flags are invalid */
static unsigned logical_size(ulint flags) {
ulint page_ssize = 0;
if (full_crc32(flags)) {
page_ssize = FSP_FLAGS_FCRC32_GET_PAGE_SSIZE(flags);
} else {
page_ssize = FSP_FLAGS_GET_PAGE_SSIZE(flags);
}
switch (page_ssize) {
case 3: return 4096;
case 4: return 8192;
case 5:
{ ut_ad(full_crc32(flags)); return 16384; }
case 0:
{ ut_ad(!full_crc32(flags)); return 16384; }
case 6: return 32768;
case 7: return 65536;
default: return 0;
}
}
/** Determine the ROW_FORMAT=COMPRESSED page size.
@param flags tablespace flags (FSP_FLAGS)
@return the ROW_FORMAT=COMPRESSED page size
@retval 0 if ROW_FORMAT=COMPRESSED is not used */
static unsigned zip_size(ulint flags) {
if (full_crc32(flags)) {
return 0;
}
ulint zip_ssize = FSP_FLAGS_GET_ZIP_SSIZE(flags);
return zip_ssize
? (UNIV_ZIP_SIZE_MIN >> 1) << zip_ssize : 0;
}
/** Determine the physical page size.
@param flags tablespace flags (FSP_FLAGS)
@return the physical page size */
static unsigned physical_size(ulint flags) {
if (full_crc32(flags)) {
return logical_size(flags);
}
ulint zip_ssize = FSP_FLAGS_GET_ZIP_SSIZE(flags);
return zip_ssize
? (UNIV_ZIP_SIZE_MIN >> 1) << zip_ssize
: unsigned(srv_page_size);
}
/** @return the ROW_FORMAT=COMPRESSED page size
@retval 0 if ROW_FORMAT=COMPRESSED is not used */
unsigned zip_size() const { return zip_size(flags); }
/** @return the physical page size */
unsigned physical_size() const { return physical_size(flags); }
/** Check whether the compression enabled in tablespace.
@param[in] flags tablespace flags */
static bool is_compressed(ulint flags) {
if (full_crc32(flags)) {
ulint algo = FSP_FLAGS_FCRC32_GET_COMPRESSED_ALGO(
flags);
DBUG_ASSERT(algo <= PAGE_ALGORITHM_LAST);
return algo > 0;
}
return FSP_FLAGS_HAS_PAGE_COMPRESSION(flags);
}
/** @return whether the compression enabled for the tablespace. */
bool is_compressed() const { return is_compressed(flags); }
/** Get the compression algorithm for full crc32 format.
@param[in] flags tablespace flags
@return algorithm type of tablespace */
static ulint get_compression_algo(ulint flags)
{
return full_crc32(flags)
? FSP_FLAGS_FCRC32_GET_COMPRESSED_ALGO(flags)
: 0;
}
/** @return the page_compressed algorithm
@retval 0 if not page_compressed */
ulint get_compression_algo() const {
return fil_space_t::get_compression_algo(flags);
}
/** Determine if the page_compressed page contains an extra byte
for exact compressed stream length
@param[in] flags tablespace flags
@return whether the extra byte is needed */
static bool full_crc32_page_compressed_len(ulint flags)
{
DBUG_ASSERT(full_crc32(flags));
switch (get_compression_algo(flags)) {
case PAGE_LZ4_ALGORITHM:
case PAGE_LZO_ALGORITHM:
case PAGE_SNAPPY_ALGORITHM:
return true;
}
return false;
}
/** Whether the full checksum matches with non full checksum flags.
@param[in] flags flags present
@param[in] expected expected flags
@return true if it is equivalent */
static bool is_flags_full_crc32_equal(ulint flags, ulint expected)
{
ut_ad(full_crc32(flags));
if (full_crc32(expected)) {
return get_compression_algo(flags)
== get_compression_algo(expected);
}
ulint page_ssize = FSP_FLAGS_FCRC32_GET_PAGE_SSIZE(flags);
ulint space_page_ssize = FSP_FLAGS_GET_PAGE_SSIZE(expected);
if (page_ssize == 5) {
if (space_page_ssize) {
return false;
}
} else if (space_page_ssize != page_ssize) {
return false;
}
return is_compressed(expected) == is_compressed(flags);
}
/** Whether old tablespace flags match full_crc32 flags.
@param[in] flags flags present
@param[in] expected expected flags
@return true if it is equivalent */
static bool is_flags_non_full_crc32_equal(ulint flags, ulint expected)
{
ut_ad(!full_crc32(flags));
if (!full_crc32(expected)) {
return false;
}
ulint page_ssize = FSP_FLAGS_GET_PAGE_SSIZE(flags);
ulint space_page_ssize = FSP_FLAGS_FCRC32_GET_PAGE_SSIZE(
expected);
if (page_ssize) {
if (space_page_ssize != 5) {
return false;
}
} else if (space_page_ssize != page_ssize) {
return false;
}
return is_compressed(expected) == is_compressed(flags);
}
/** Whether both fsp flags are equivalent */
static bool is_flags_equal(ulint flags, ulint expected)
{
if (!((flags ^ expected) & ~(1U << FSP_FLAGS_POS_RESERVED))) {
return true;
}
return full_crc32(flags)
? is_flags_full_crc32_equal(flags, expected)
: is_flags_non_full_crc32_equal(flags, expected);
}
/** Validate the tablespace flags for full crc32 format.
@param[in] flags the content of FSP_SPACE_FLAGS
@return whether the flags are correct in full crc32 format */
static bool is_fcrc32_valid_flags(ulint flags)
{
ut_ad(flags & FSP_FLAGS_FCRC32_MASK_MARKER);
const ulint page_ssize = physical_size(flags);
if (page_ssize < 3 || page_ssize & 8) {
return false;
}
flags >>= FSP_FLAGS_FCRC32_POS_COMPRESSED_ALGO;
return flags <= PAGE_ALGORITHM_LAST;
}
/** Validate the tablespace flags.
@param[in] flags content of FSP_SPACE_FLAGS
@param[in] is_ibd whether this is an .ibd file
(not system tablespace)
@return whether the flags are correct. */
static bool is_valid_flags(ulint flags, bool is_ibd)
{
DBUG_EXECUTE_IF("fsp_flags_is_valid_failure",
return false;);
if (full_crc32(flags)) {
return is_fcrc32_valid_flags(flags);
}
if (flags == 0) {
return true;
}
if (flags & ~FSP_FLAGS_MASK) {
return false;
}
if ((flags & (FSP_FLAGS_MASK_POST_ANTELOPE
| FSP_FLAGS_MASK_ATOMIC_BLOBS))
== FSP_FLAGS_MASK_ATOMIC_BLOBS) {
/* If the "atomic blobs" flag (indicating
ROW_FORMAT=DYNAMIC or ROW_FORMAT=COMPRESSED) flag
is set, then the "post Antelope"
(ROW_FORMAT!=REDUNDANT) flag must also be set. */
return false;
}
/* Bits 10..14 should be 0b0000d where d is the DATA_DIR flag
of MySQL 5.6 and MariaDB 10.0, which we ignore.
In the buggy FSP_SPACE_FLAGS written by MariaDB 10.1.0 to 10.1.20,
bits 10..14 would be nonzero 0bsssaa where sss is
nonzero PAGE_SSIZE (3, 4, 6, or 7)
and aa is ATOMIC_WRITES (not 0b11). */
if (FSP_FLAGS_GET_RESERVED(flags) & ~1U) {
return false;
}
const ulint ssize = FSP_FLAGS_GET_PAGE_SSIZE(flags);
if (ssize == 1 || ssize == 2 || ssize == 5 || ssize & 8) {
/* the page_size is not between 4k and 64k;
16k should be encoded as 0, not 5 */
return false;
}
const ulint zssize = FSP_FLAGS_GET_ZIP_SSIZE(flags);
if (zssize == 0) {
/* not ROW_FORMAT=COMPRESSED */
} else if (zssize > (ssize ? ssize : 5)) {
/* Invalid KEY_BLOCK_SIZE */
return false;
} else if (~flags & (FSP_FLAGS_MASK_POST_ANTELOPE
| FSP_FLAGS_MASK_ATOMIC_BLOBS)) {
/* both these flags should be set for
ROW_FORMAT=COMPRESSED */
return false;
}
/* The flags do look valid. But, avoid misinterpreting
buggy MariaDB 10.1 format flags for
PAGE_COMPRESSED=1 PAGE_COMPRESSION_LEVEL={0,2,3}
as valid-looking PAGE_SSIZE if this is known to be
an .ibd file and we are using the default innodb_page_size=16k. */
return(ssize == 0 || !is_ibd
|| srv_page_size != UNIV_PAGE_SIZE_ORIG);
}
};
#ifndef UNIV_INNOCHECKSUM
/** Value of fil_space_t::magic_n */
#define FIL_SPACE_MAGIC_N 89472
/** File node of a tablespace or the log data space */
struct fil_node_t {
/** tablespace containing this file */
fil_space_t* space;
/** file name; protected by fil_system.mutex and log_sys.mutex. */
char* name;
/** file handle (valid if is_open) */
pfs_os_file_t handle;
/** whether the file actually is a raw device or disk partition */
bool is_raw_disk;
/** whether the file is on non-rotational media (SSD) */
bool on_ssd;
/** size of the file in database pages (0 if not known yet);
the possible last incomplete megabyte may be ignored
if space->id == 0 */
ulint size;
/** initial size of the file in database pages;
FIL_IBD_FILE_INITIAL_SIZE by default */
ulint init_size;
/** maximum size of the file in database pages (0 if unlimited) */
ulint max_size;
/** count of pending i/o's; is_open must be true if nonzero */
ulint n_pending;
/** count of pending flushes; is_open must be true if nonzero */
ulint n_pending_flushes;
/** whether the file is currently being extended */
bool being_extended;
/** whether this file had writes after lasy fsync() */
bool needs_flush;
/** link to other files in this tablespace */
UT_LIST_NODE_T(fil_node_t) chain;
/** link to the fil_system.LRU list (keeping track of open files) */
UT_LIST_NODE_T(fil_node_t) LRU;
/** whether this file could use atomic write (data file) */
bool atomic_write;
/** Filesystem block size */
ulint block_size;
/** FIL_NODE_MAGIC_N */
ulint magic_n;
/** @return whether this file is open */
bool is_open() const
{
return(handle != OS_FILE_CLOSED);
}
/** Read the first page of a data file.
@param[in] first whether this is the very first read
@return whether the page was found valid */
bool read_page0(bool first);
/** Determine some file metadata when creating or reading the file.
@param file the file that is being created, or OS_FILE_CLOSED */
void find_metadata(os_file_t file = OS_FILE_CLOSED
#ifdef UNIV_LINUX
, struct stat* statbuf = NULL
#endif
);
/** Close the file handle. */
void close();
};
/** Value of fil_node_t::magic_n */
#define FIL_NODE_MAGIC_N 89389
inline void fil_space_t::set_imported()
{
ut_ad(purpose == FIL_TYPE_IMPORT);
purpose = FIL_TYPE_TABLESPACE;
UT_LIST_GET_FIRST(chain)->find_metadata();
}
inline bool fil_space_t::is_rotational() const
{
for (const fil_node_t* node = UT_LIST_GET_FIRST(chain);
node != NULL; node = UT_LIST_GET_NEXT(chain, node)) {
if (!node->on_ssd) {
return true;
}
}
return false;
}
/** Common InnoDB file extentions */
enum ib_extention {
NO_EXT = 0,
IBD = 1,
ISL = 2,
CFG = 3
};
extern const char* dot_ext[];
#define DOT_IBD dot_ext[IBD]
#define DOT_ISL dot_ext[ISL]
#define DOT_CFG dot_ext[CFG]
/** When mysqld is run, the default directory "." is the mysqld datadir,
but in the MySQL Embedded Server Library and mysqlbackup it is not the default
directory, and we must set the base file path explicitly */
extern const char* fil_path_to_mysql_datadir;
/* Space address data type; this is intended to be used when
addresses accurate to a byte are stored in file pages. If the page part
of the address is FIL_NULL, the address is considered undefined. */
typedef byte fil_faddr_t; /*!< 'type' definition in C: an address
stored in a file page is a string of bytes */
#else
# include "univ.i"
#endif /* !UNIV_INNOCHECKSUM */
/** Initial size of a single-table tablespace in pages */
#define FIL_IBD_FILE_INITIAL_SIZE 4U
/** 'null' (undefined) page offset in the context of file spaces */
#define FIL_NULL ULINT32_UNDEFINED
#define FIL_ADDR_PAGE 0U /* first in address is the page offset */
#define FIL_ADDR_BYTE 4U /* then comes 2-byte byte offset within page*/
#define FIL_ADDR_SIZE 6U /* address size is 6 bytes */
/** File space address */
struct fil_addr_t {
ulint page; /*!< page number within a space */
ulint boffset; /*!< byte offset within the page */
};
/** The byte offsets on a file page for various variables @{ */
#define FIL_PAGE_SPACE_OR_CHKSUM 0 /*!< in < MySQL-4.0.14 space id the
page belongs to (== 0) but in later
versions the 'new' checksum of the
page */
#define FIL_PAGE_OFFSET 4U /*!< page offset inside space */
#define FIL_PAGE_PREV 8U /*!< if there is a 'natural'
predecessor of the page, its
offset. Otherwise FIL_NULL.
This field is not set on BLOB
pages, which are stored as a
singly-linked list. See also
FIL_PAGE_NEXT. */
#define FIL_PAGE_NEXT 12U /*!< if there is a 'natural' successor
of the page, its offset.
Otherwise FIL_NULL.
B-tree index pages
(FIL_PAGE_TYPE contains FIL_PAGE_INDEX)
on the same PAGE_LEVEL are maintained
as a doubly linked list via
FIL_PAGE_PREV and FIL_PAGE_NEXT
in the collation order of the
smallest user record on each page. */
#define FIL_PAGE_LSN 16U /*!< lsn of the end of the newest
modification log record to the page */
#define FIL_PAGE_TYPE 24U /*!< file page type: FIL_PAGE_INDEX,...,
2 bytes.
The contents of this field can only
be trusted in the following case:
if the page is an uncompressed
B-tree index page, then it is
guaranteed that the value is
FIL_PAGE_INDEX.
The opposite does not hold.
In tablespaces created by
MySQL/InnoDB 5.1.7 or later, the
contents of this field is valid
for all uncompressed pages. */
/** For the first page in a system tablespace data file(ibdata*, not *.ibd):
the file has been flushed to disk at least up to this lsn
For other pages: 32-bit key version used to encrypt the page + 32-bit checksum
or 64 bites of zero if no encryption */
#define FIL_PAGE_FILE_FLUSH_LSN_OR_KEY_VERSION 26U
/** This overloads FIL_PAGE_FILE_FLUSH_LSN for RTREE Split Sequence Number */
#define FIL_RTREE_SPLIT_SEQ_NUM FIL_PAGE_FILE_FLUSH_LSN_OR_KEY_VERSION
/** Start of the page_compressed content */
#define FIL_PAGE_COMP_ALGO FIL_PAGE_FILE_FLUSH_LSN_OR_KEY_VERSION
/** starting from 4.1.x this contains the space id of the page */
#define FIL_PAGE_ARCH_LOG_NO_OR_SPACE_ID 34U
#define FIL_PAGE_SPACE_ID FIL_PAGE_ARCH_LOG_NO_OR_SPACE_ID
#define FIL_PAGE_DATA 38U /*!< start of the data on the page */
/** 32-bit key version used to encrypt the page in full_crc32 format.
For non-encrypted page, it contains 0. */
#define FIL_PAGE_FCRC32_KEY_VERSION 0
/** page_compressed without innodb_checksum_algorithm=full_crc32 @{ */
/** Number of bytes used to store actual payload data size on
page_compressed pages when not using full_crc32. */
#define FIL_PAGE_COMP_SIZE 0
/** Number of bytes for FIL_PAGE_COMP_SIZE */
#define FIL_PAGE_COMP_METADATA_LEN 2
/** Number of bytes used to store actual compression method
for encrypted tables when not using full_crc32. */
#define FIL_PAGE_ENCRYPT_COMP_ALGO 2
/** Extra header size for encrypted page_compressed pages when
not using full_crc32 */
#define FIL_PAGE_ENCRYPT_COMP_METADATA_LEN 4
/* @} */
/** File page trailer @{ */
#define FIL_PAGE_END_LSN_OLD_CHKSUM 8 /*!< the low 4 bytes of this are used
to store the page checksum, the
last 4 bytes should be identical
to the last 4 bytes of FIL_PAGE_LSN */
#define FIL_PAGE_DATA_END 8 /*!< size of the page trailer */
/** Store the last 4 bytes of FIL_PAGE_LSN */
#define FIL_PAGE_FCRC32_END_LSN 8
/** Store crc32 checksum at the end of the page */
#define FIL_PAGE_FCRC32_CHECKSUM 4
/* @} */
/** File page types (values of FIL_PAGE_TYPE) @{ */
/** page_compressed, encrypted=YES (not used for full_crc32) */
#define FIL_PAGE_PAGE_COMPRESSED_ENCRYPTED 37401
/** page_compressed (not used for full_crc32) */
#define FIL_PAGE_PAGE_COMPRESSED 34354 /*!< page compressed page */
#define FIL_PAGE_INDEX 17855 /*!< B-tree node */
#define FIL_PAGE_RTREE 17854 /*!< R-tree node (SPATIAL INDEX) */
#define FIL_PAGE_UNDO_LOG 2 /*!< Undo log page */
#define FIL_PAGE_INODE 3 /*!< Index node */
#define FIL_PAGE_IBUF_FREE_LIST 4 /*!< Insert buffer free list */
/* File page types introduced in MySQL/InnoDB 5.1.7 */
#define FIL_PAGE_TYPE_ALLOCATED 0 /*!< Freshly allocated page */
#define FIL_PAGE_IBUF_BITMAP 5 /*!< Insert buffer bitmap */
#define FIL_PAGE_TYPE_SYS 6 /*!< System page */
#define FIL_PAGE_TYPE_TRX_SYS 7 /*!< Transaction system data */
#define FIL_PAGE_TYPE_FSP_HDR 8 /*!< File space header */
#define FIL_PAGE_TYPE_XDES 9 /*!< Extent descriptor page */
#define FIL_PAGE_TYPE_BLOB 10 /*!< Uncompressed BLOB page */
#define FIL_PAGE_TYPE_ZBLOB 11 /*!< First compressed BLOB page */
#define FIL_PAGE_TYPE_ZBLOB2 12 /*!< Subsequent compressed BLOB page */
#define FIL_PAGE_TYPE_UNKNOWN 13 /*!< In old tablespaces, garbage
in FIL_PAGE_TYPE is replaced with this
value when flushing pages. */
/* File page types introduced in MySQL 5.7, not supported in MariaDB */
//#define FIL_PAGE_COMPRESSED 14
//#define FIL_PAGE_ENCRYPTED 15
//#define FIL_PAGE_COMPRESSED_AND_ENCRYPTED 16
//#define FIL_PAGE_ENCRYPTED_RTREE 17
/** Clustered index root page after instant ADD COLUMN */
#define FIL_PAGE_TYPE_INSTANT 18
/** Used by i_s.cc to index into the text description.
Note: FIL_PAGE_TYPE_INSTANT maps to the same as FIL_PAGE_INDEX. */
#define FIL_PAGE_TYPE_LAST FIL_PAGE_TYPE_UNKNOWN
/*!< Last page type */
/** Set in FIL_PAGE_TYPE if for full_crc32 pages in page_compressed format.
If the flag is set, then the following holds for the remaining bits
of FIL_PAGE_TYPE:
Bits 0..7 will contain the compressed page size in bytes.
Bits 8..14 are reserved and must be 0. */
#define FIL_PAGE_COMPRESS_FCRC32_MARKER 15
/* @} */
/** @return whether the page type is B-tree or R-tree index */
inline bool fil_page_type_is_index(ulint page_type)
{
switch (page_type) {
case FIL_PAGE_TYPE_INSTANT:
case FIL_PAGE_INDEX:
case FIL_PAGE_RTREE:
return(true);
}
return(false);
}
/** Check whether the page is index page (either regular Btree index or Rtree
index */
#define fil_page_index_page_check(page) \
fil_page_type_is_index(fil_page_get_type(page))
/** Enum values for encryption table option */
enum fil_encryption_t {
/** Encrypted if innodb_encrypt_tables=ON (srv_encrypt_tables) */
FIL_ENCRYPTION_DEFAULT,
/** Encrypted */
FIL_ENCRYPTION_ON,
/** Not encrypted */
FIL_ENCRYPTION_OFF
};
#ifndef UNIV_INNOCHECKSUM
/** The number of fsyncs done to the log */
extern ulint fil_n_log_flushes;
/** Number of pending redo log flushes */
extern ulint fil_n_pending_log_flushes;
/** Number of pending tablespace flushes */
extern ulint fil_n_pending_tablespace_flushes;
/** Look up a tablespace.
The caller should hold an InnoDB table lock or a MDL that prevents
the tablespace from being dropped during the operation,
or the caller should be in single-threaded crash recovery mode
(no user connections that could drop tablespaces).
If this is not the case, fil_space_acquire() and fil_space_t::release()
should be used instead.
@param[in] id tablespace ID
@return tablespace, or NULL if not found */
fil_space_t*
fil_space_get(
ulint id)
MY_ATTRIBUTE((warn_unused_result));
/** The tablespace memory cache; also the totality of logs (the log
data space) is stored here; below we talk about tablespaces, but also
the ib_logfiles form a 'space' and it is handled here */
struct fil_system_t {
/**
Constructor.
Some members may require late initialisation, thus we just mark object as
uninitialised. Real initialisation happens in create().
*/
fil_system_t(): m_initialised(false)
{
UT_LIST_INIT(LRU, &fil_node_t::LRU);
UT_LIST_INIT(space_list, &fil_space_t::space_list);
UT_LIST_INIT(rotation_list, &fil_space_t::rotation_list);
UT_LIST_INIT(unflushed_spaces, &fil_space_t::unflushed_spaces);
UT_LIST_INIT(named_spaces, &fil_space_t::named_spaces);
}
bool is_initialised() const { return m_initialised; }
/**
Create the file system interface at database start.
@param[in] hash_size hash table size
*/
void create(ulint hash_size);
/** Close the file system interface at shutdown */
void close();
private:
bool m_initialised;
#ifdef UNIV_LINUX
/** available block devices that reside on non-rotational storage */
std::vector<dev_t> ssd;
public:
/** @return whether a file system device is on non-rotational storage */
bool is_ssd(dev_t dev) const
{
/* Linux seems to allow up to 15 partitions per block device.
If the detected ssd carries "partition number 0" (it is the whole device),
compare the candidate file system number without the partition number. */
for (const auto s : ssd)
if (dev == s || (dev & ~15U) == s)
return true;
return false;
}
#endif
public:
ib_mutex_t mutex; /*!< The mutex protecting the cache */
fil_space_t* sys_space; /*!< The innodb_system tablespace */
fil_space_t* temp_space; /*!< The innodb_temporary tablespace */
hash_table_t* spaces; /*!< The hash table of spaces in the
system; they are hashed on the space
id */
UT_LIST_BASE_NODE_T(fil_node_t) LRU;
/*!< base node for the LRU list of the
most recently used open files with no
pending i/o's; if we start an i/o on
the file, we first remove it from this
list, and return it to the start of
the list when the i/o ends;
log files and the system tablespace are
not put to this list: they are opened
after the startup, and kept open until
shutdown */
UT_LIST_BASE_NODE_T(fil_space_t) unflushed_spaces;
/*!< base node for the list of those
tablespaces whose files contain
unflushed writes; those spaces have
at least one file node where
needs_flush == true */
ulint n_open; /*!< number of files currently open */
ulint max_assigned_id;/*!< maximum space id in the existing
tables, or assigned during the time
mysqld has been up; at an InnoDB
startup we scan the data dictionary
and set here the maximum of the
space id's of the tables there */
UT_LIST_BASE_NODE_T(fil_space_t) space_list;
/*!< list of all file spaces */
UT_LIST_BASE_NODE_T(fil_space_t) named_spaces;
/*!< list of all file spaces
for which a MLOG_FILE_NAME
record has been written since
the latest redo log checkpoint.
Protected only by log_sys.mutex. */
UT_LIST_BASE_NODE_T(fil_space_t) rotation_list;
/*!< list of all file spaces needing
key rotation.*/
bool space_id_reuse_warned;
/*!< whether fil_space_create()
has issued a warning about
potential space_id reuse */
/** Trigger a call to fil_node_t::read_page0()
@param[in] id tablespace identifier
@return tablespace
@retval NULL if the tablespace does not exist or cannot be read */
fil_space_t* read_page0(ulint id);
/** Return the next fil_space_t from key rotation list.
Once started, the caller must keep calling this until it returns NULL.
fil_space_acquire() and fil_space_release() are invoked here which
blocks a concurrent operation from dropping the tablespace.
@param[in] prev_space Previous tablespace or NULL to start
from beginning of fil_system->rotation
list
@param[in] recheck recheck of the tablespace is needed or
still encryption thread does write page0
for it
@param[in] key_version key version of the key state thread
If NULL, use the first fil_space_t on fil_system->space_list.
@return pointer to the next fil_space_t.
@retval NULL if this was the last */
fil_space_t* keyrotate_next(
fil_space_t* prev_space,
bool remove,
uint key_version);
};
/** The tablespace memory cache. */
extern fil_system_t fil_system;
#include "fil0crypt.h"
/** Returns the latch of a file space.
@param[in] id space id
@param[out] flags tablespace flags
@return latch protecting storage allocation */
rw_lock_t*
fil_space_get_latch(
ulint id,
ulint* flags);
/** Create a space memory object and put it to the fil_system hash table.
Error messages are issued to the server log.
@param[in] name tablespace name
@param[in] id tablespace identifier
@param[in] flags tablespace flags
@param[in] purpose tablespace purpose
@param[in,out] crypt_data encryption information
@param[in] mode encryption mode
@return pointer to created tablespace, to be filled in with fil_space_t::add()
@retval NULL on failure (such as when the same tablespace exists) */
fil_space_t*
fil_space_create(
const char* name,
ulint id,
ulint flags,
fil_type_t purpose,
fil_space_crypt_t* crypt_data,
fil_encryption_t mode = FIL_ENCRYPTION_DEFAULT)
MY_ATTRIBUTE((warn_unused_result));
/*******************************************************************//**
Assigns a new space id for a new single-table tablespace. This works simply by
incrementing the global counter. If 4 billion id's is not enough, we may need
to recycle id's.
@return true if assigned, false if not */
bool
fil_assign_new_space_id(
/*====================*/
ulint* space_id); /*!< in/out: space id */
/** Frees a space object from the tablespace memory cache.
Closes the files in the chain but does not delete them.
There must not be any pending i/o's or flushes on the files.
@param[in] id tablespace identifier
@param[in] x_latched whether the caller holds X-mode space->latch
@return true if success */
bool
fil_space_free(
ulint id,
bool x_latched);
/** Set the recovered size of a tablespace in pages.
@param id tablespace ID
@param size recovered size in pages */
UNIV_INTERN
void
fil_space_set_recv_size(ulint id, ulint size);
/*******************************************************************//**
Returns the size of the space in pages. The tablespace must be cached in the
memory cache.
@return space size, 0 if space not found */
ulint
fil_space_get_size(
/*===============*/
ulint id); /*!< in: space id */
/*******************************************************************//**
Returns the flags of the space. The tablespace must be cached
in the memory cache.
@return flags, ULINT_UNDEFINED if space not found */
ulint
fil_space_get_flags(
/*================*/
ulint id); /*!< in: space id */
/*******************************************************************//**
Opens all log files and system tablespace data files. They stay open until the
database server shutdown. This should be called at a server startup after the
space objects for the log and the system tablespace have been created. The
purpose of this operation is to make sure we never run out of file descriptors
if we need to read from the insert buffer or to write to the log. */
void
fil_open_log_and_system_tablespace_files(void);
/*==========================================*/
/*******************************************************************//**
Closes all open files. There must not be any pending i/o's or not flushed
modifications in the files. */
void
fil_close_all_files(void);
/*=====================*/
/*******************************************************************//**
Closes the redo log files. There must not be any pending i/o's or not
flushed modifications in the files. */
void
fil_close_log_files(
/*================*/
bool free); /*!< in: whether to free the memory object */
/*******************************************************************//**
Sets the max tablespace id counter if the given number is bigger than the
previous value. */
void
fil_set_max_space_id_if_bigger(
/*===========================*/
ulint max_id);/*!< in: maximum known id */
/** Write the flushed LSN to the page header of the first page in the
system tablespace.
@param[in] lsn flushed LSN
@return DB_SUCCESS or error number */
dberr_t
fil_write_flushed_lsn(
lsn_t lsn)
MY_ATTRIBUTE((warn_unused_result));
/** Acquire a tablespace when it could be dropped concurrently.
Used by background threads that do not necessarily hold proper locks
for concurrency control.
@param[in] id tablespace ID
@param[in] silent whether to silently ignore missing tablespaces
@return the tablespace
@retval NULL if missing or being deleted or truncated */
UNIV_INTERN
fil_space_t*
fil_space_acquire_low(ulint id, bool silent)
MY_ATTRIBUTE((warn_unused_result));
/** Acquire a tablespace when it could be dropped concurrently.
Used by background threads that do not necessarily hold proper locks
for concurrency control.
@param[in] id tablespace ID
@return the tablespace
@retval NULL if missing or being deleted or truncated */
inline
fil_space_t*
fil_space_acquire(ulint id)
{
return (fil_space_acquire_low(id, false));
}
/** Acquire a tablespace that may not exist.
Used by background threads that do not necessarily hold proper locks
for concurrency control.
@param[in] id tablespace ID
@return the tablespace
@retval NULL if missing or being deleted */
inline
fil_space_t*
fil_space_acquire_silent(ulint id)
{
return (fil_space_acquire_low(id, true));
}
/** Acquire a tablespace for reading or writing a block,
when it could be dropped concurrently.
@param[in] id tablespace ID
@return the tablespace
@retval NULL if missing */
fil_space_t*
fil_space_acquire_for_io(ulint id);
/** Return the next fil_space_t.
Once started, the caller must keep calling this until it returns NULL.
fil_space_acquire() and fil_space_t::release() are invoked here which
blocks a concurrent operation from dropping the tablespace.
@param[in,out] prev_space Pointer to the previous fil_space_t.
If NULL, use the first fil_space_t on fil_system.space_list.
@return pointer to the next fil_space_t.
@retval NULL if this was the last */
fil_space_t*
fil_space_next(
fil_space_t* prev_space)
MY_ATTRIBUTE((warn_unused_result));
/** Return the next fil_space_t from key rotation list.
Once started, the caller must keep calling this until it returns NULL.
fil_space_acquire() and fil_space_t::release() are invoked here which
blocks a concurrent operation from dropping the tablespace.
@param[in,out] prev_space Pointer to the previous fil_space_t.
If NULL, use the first fil_space_t on fil_system.space_list.
@param[in] remove Whether to remove the previous tablespace from
the rotation list
@return pointer to the next fil_space_t.
@retval NULL if this was the last*/
fil_space_t*
fil_space_keyrotate_next(fil_space_t* prev_space, bool remove)
MY_ATTRIBUTE((warn_unused_result));
/********************************************************//**
Creates the database directory for a table if it does not exist yet. */
void
fil_create_directory_for_tablename(
/*===============================*/
const char* name); /*!< in: name in the standard
'databasename/tablename' format */
/** Replay a file rename operation if possible.
@param[in] space_id tablespace identifier
@param[in] first_page_no first page number in the file
@param[in] name old file name
@param[in] new_name new file name
@return whether the operation was successfully applied
(the name did not exist, or new_name did not exist and
name was successfully renamed to new_name) */
bool
fil_op_replay_rename(
ulint space_id,
ulint first_page_no,
const char* name,
const char* new_name)
MY_ATTRIBUTE((warn_unused_result));
/** Determine whether a table can be accessed in operations that are
not (necessarily) protected by meta-data locks.
(Rollback would generally be protected, but rollback of
FOREIGN KEY CASCADE/SET NULL is not protected by meta-data locks
but only by InnoDB table locks, which may be broken by
lock_remove_all_on_table().)
@param[in] table persistent table
checked @return whether the table is accessible */
bool fil_table_accessible(const dict_table_t* table)
MY_ATTRIBUTE((warn_unused_result, nonnull));
/** Delete a tablespace and associated .ibd file.
@param[in] id tablespace identifier
@return DB_SUCCESS or error */
dberr_t
fil_delete_tablespace(
ulint id
#ifdef BTR_CUR_HASH_ADAPT
, bool drop_ahi = false /*!< whether to drop the adaptive hash index */
#endif /* BTR_CUR_HASH_ADAPT */
);
/** Prepare to truncate an undo tablespace.
@param[in] space_id undo tablespace id
@return the tablespace
@retval NULL if the tablespace does not exist */
fil_space_t* fil_truncate_prepare(ulint space_id);
/** Write log about an undo tablespace truncate operation. */
void fil_truncate_log(fil_space_t* space, ulint size, mtr_t* mtr)
MY_ATTRIBUTE((nonnull));
/*******************************************************************//**
Closes a single-table tablespace. The tablespace must be cached in the
memory cache. Free all pages used by the tablespace.
@return DB_SUCCESS or error */
dberr_t
fil_close_tablespace(
/*=================*/
trx_t* trx, /*!< in/out: Transaction covering the close */
ulint id); /*!< in: space id */
/*******************************************************************//**
Allocates and builds a file name from a path, a table or tablespace name
and a suffix. The string must be freed by caller with ut_free().
@param[in] path NULL or the direcory path or the full path and filename.
@param[in] name NULL if path is full, or Table/Tablespace name
@param[in] suffix NULL or the file extention to use.
@return own: file name */
char*
fil_make_filepath(
const char* path,
const char* name,
ib_extention suffix,
bool strip_name);
/** Create a tablespace file.
@param[in] space_id Tablespace ID
@param[in] name Tablespace name in dbname/tablename format.
@param[in] path Path and filename of the datafile to create.
@param[in] flags Tablespace flags
@param[in] size Initial size of the tablespace file in pages,
must be >= FIL_IBD_FILE_INITIAL_SIZE
@param[in] mode MariaDB encryption mode
@param[in] key_id MariaDB encryption key_id
@param[out] err DB_SUCCESS or error code
@return the created tablespace
@retval NULL on error */
fil_space_t*
fil_ibd_create(
ulint space_id,
const char* name,
const char* path,
ulint flags,
ulint size,
fil_encryption_t mode,
uint32_t key_id,
dberr_t* err)
MY_ATTRIBUTE((nonnull(2,8), warn_unused_result));
/** Try to adjust FSP_SPACE_FLAGS if they differ from the expectations.
(Typically when upgrading from MariaDB 10.1.0..10.1.20.)
@param[in,out] space tablespace
@param[in] flags desired tablespace flags */
void fsp_flags_try_adjust(fil_space_t* space, ulint flags);
/********************************************************************//**
Tries to open a single-table tablespace and optionally checks the space id is
right in it. If does not succeed, prints an error message to the .err log. This
function is used to open a tablespace when we start up mysqld, and also in
IMPORT TABLESPACE.
NOTE that we assume this operation is used either at the database startup
or under the protection of the dictionary mutex, so that two users cannot
race here. This operation does not leave the file associated with the
tablespace open, but closes it after we have looked at the space id in it.
If the validate boolean is set, we read the first page of the file and
check that the space id in the file is what we expect. We assume that
this function runs much faster if no check is made, since accessing the
file inode probably is much faster (the OS caches them) than accessing
the first page of the file. This boolean may be initially false, but if
a remote tablespace is found it will be changed to true.
If the fix_dict boolean is set, then it is safe to use an internal SQL
statement to update the dictionary tables if they are incorrect.
@param[in] validate true if we should validate the tablespace
@param[in] fix_dict true if the dictionary is available to be fixed
@param[in] purpose FIL_TYPE_TABLESPACE or FIL_TYPE_TEMPORARY
@param[in] id tablespace ID
@param[in] flags expected FSP_SPACE_FLAGS
@param[in] tablename table name
If file-per-table, it is the table name in the databasename/tablename format
@param[in] path_in expected filepath, usually read from dictionary
@param[out] err DB_SUCCESS or error code
@return tablespace
@retval NULL if the tablespace could not be opened */
fil_space_t*
fil_ibd_open(
bool validate,
bool fix_dict,
fil_type_t purpose,
ulint id,
ulint flags,
const table_name_t& tablename,
const char* path_in,
dberr_t* err = NULL)
MY_ATTRIBUTE((warn_unused_result));
enum fil_load_status {
/** The tablespace file(s) were found and valid. */
FIL_LOAD_OK,
/** The name no longer matches space_id */
FIL_LOAD_ID_CHANGED,
/** The file(s) were not found */
FIL_LOAD_NOT_FOUND,
/** The file(s) were not valid */
FIL_LOAD_INVALID
};
/** Open a single-file tablespace and add it to the InnoDB data structures.
@param[in] space_id tablespace ID
@param[in] filename path/to/databasename/tablename.ibd
@param[out] space the tablespace, or NULL on error
@return status of the operation */
enum fil_load_status
fil_ibd_load(
ulint space_id,
const char* filename,
fil_space_t*& space)
MY_ATTRIBUTE((warn_unused_result));
/***********************************************************************//**
A fault-tolerant function that tries to read the next file name in the
directory. We retry 100 times if os_file_readdir_next_file() returns -1. The
idea is to read as much good data as we can and jump over bad data.
@return 0 if ok, -1 if error even after the retries, 1 if at the end
of the directory */
int
fil_file_readdir_next_file(
/*=======================*/
dberr_t* err, /*!< out: this is set to DB_ERROR if an error
was encountered, otherwise not changed */
const char* dirname,/*!< in: directory name or path */
os_file_dir_t dir, /*!< in: directory stream */
os_file_stat_t* info); /*!< in/out: buffer where the
info is returned */
/** Determine if a matching tablespace exists in the InnoDB tablespace
memory cache. Note that if we have not done a crash recovery at the database
startup, there may be many tablespaces which are not yet in the memory cache.
@param[in] id Tablespace ID
@param[in] name Tablespace name used in fil_space_create().
@param[in] table_flags table flags
@return the tablespace
@retval NULL if no matching tablespace exists in the memory cache */
fil_space_t*
fil_space_for_table_exists_in_mem(
ulint id,
const char* name,
ulint table_flags);
/** Try to extend a tablespace if it is smaller than the specified size.
@param[in,out] space tablespace
@param[in] size desired size in pages
@return whether the tablespace is at least as big as requested */
bool
fil_space_extend(
fil_space_t* space,
ulint size);
/** Reads or writes data. This operation could be asynchronous (aio).
@param[in] type IO context
@param[in] sync true if synchronous aio is desired
@param[in] page_id page id
@param[in] zip_size ROW_FORMAT=COMPRESSED page size, or 0
@param[in] byte_offset remainder of offset in bytes; in aio this
must be divisible by the OS block size
@param[in] len how many bytes to read or write; this must
not cross a file boundary; in aio this must
be a block size multiple
@param[in,out] buf buffer where to store read data or from where
to write; in aio this must be appropriately
aligned
@param[in] message message for aio handler if non-sync aio
used, else ignored
@param[in] ignore_missing_space true=ignore missing space during read
@return DB_SUCCESS, or DB_TABLESPACE_DELETED
if we are trying to do i/o on a tablespace which does not exist */
dberr_t
fil_io(
const IORequest& type,
bool sync,
const page_id_t page_id,
ulint zip_size,
ulint byte_offset,
ulint len,
void* buf,
void* message,
bool ignore_missing_space = false);
/**********************************************************************//**
Waits for an aio operation to complete. This function is used to write the
handler for completed requests. The aio array of pending requests is divided
into segments (see os0file.cc for more info). The thread specifies which
segment it wants to wait for. */
void
fil_aio_wait(
/*=========*/
ulint segment); /*!< in: the number of the segment in the aio
array to wait for */
/**********************************************************************//**
Flushes to disk possible writes cached by the OS. If the space does not exist
or is being dropped, does not do anything. */
void
fil_flush(
/*======*/
ulint space_id); /*!< in: file space id (this can be a group of
log files or a tablespace of the database) */
/** Flush a tablespace.
@param[in,out] space tablespace to flush */
void
fil_flush(fil_space_t* space);
/** Flush to disk the writes in file spaces of the given type
possibly cached by the OS.
@param[in] purpose FIL_TYPE_TABLESPACE or FIL_TYPE_LOG */
void
fil_flush_file_spaces(
fil_type_t purpose);
/******************************************************************//**
Checks the consistency of the tablespace cache.
@return true if ok */
bool
fil_validate(void);
/*==============*/
/********************************************************************//**
Returns true if file address is undefined.
@return true if undefined */
bool
fil_addr_is_null(
/*=============*/
fil_addr_t addr); /*!< in: address */
/********************************************************************//**
Get the predecessor of a file page.
@return FIL_PAGE_PREV */
ulint
fil_page_get_prev(
/*==============*/
const byte* page); /*!< in: file page */
/********************************************************************//**
Get the successor of a file page.
@return FIL_PAGE_NEXT */
ulint
fil_page_get_next(
/*==============*/
const byte* page); /*!< in: file page */
/*********************************************************************//**
Sets the file page type. */
void
fil_page_set_type(
/*==============*/
byte* page, /*!< in/out: file page */
ulint type); /*!< in: type */
/********************************************************************//**
Delete the tablespace file and any related files like .cfg.
This should not be called for temporary tables. */
void
fil_delete_file(
/*============*/
const char* path); /*!< in: filepath of the ibd tablespace */
/********************************************************************//**
Looks for a pre-existing fil_space_t with the given tablespace ID
and, if found, returns the name and filepath in newly allocated buffers that the caller must free.
@param[in] space_id The tablespace ID to search for.
@param[out] name Name of the tablespace found.
@param[out] fileapth The filepath of the first datafile for thtablespace found.
@return true if tablespace is found, false if not. */
bool
fil_space_read_name_and_filepath(
ulint space_id,
char** name,
char** filepath);
/** Convert a file name to a tablespace name.
@param[in] filename directory/databasename/tablename.ibd
@return database/tablename string, to be freed with ut_free() */
char*
fil_path_to_space_name(
const char* filename);
/** Generate redo log for swapping two .ibd files
@param[in] old_table old table
@param[in] new_table new table
@param[in] tmp_name temporary table name
@param[in,out] mtr mini-transaction
@return innodb error code */
dberr_t
fil_mtr_rename_log(
const dict_table_t* old_table,
const dict_table_t* new_table,
const char* tmp_name,
mtr_t* mtr)
MY_ATTRIBUTE((nonnull, warn_unused_result));
/** Acquire the fil_system mutex. */
#define fil_system_enter() mutex_enter(&fil_system.mutex)
/** Release the fil_system mutex. */
#define fil_system_exit() mutex_exit(&fil_system.mutex)
/*******************************************************************//**
Returns the table space by a given id, NULL if not found. */
fil_space_t*
fil_space_get_by_id(
/*================*/
ulint id); /*!< in: space id */
/** Note that a non-predefined persistent tablespace has been modified
by redo log.
@param[in,out] space tablespace */
void
fil_names_dirty(
fil_space_t* space);
/** Write MLOG_FILE_NAME records when a non-predefined persistent
tablespace was modified for the first time since the latest
fil_names_clear().
@param[in,out] space tablespace
@param[in,out] mtr mini-transaction */
void
fil_names_dirty_and_write(
fil_space_t* space,
mtr_t* mtr);
/** Write MLOG_FILE_NAME records if a persistent tablespace was modified
for the first time since the latest fil_names_clear().
@param[in,out] space tablespace
@param[in,out] mtr mini-transaction
@return whether any MLOG_FILE_NAME record was written */
inline MY_ATTRIBUTE((warn_unused_result))
bool
fil_names_write_if_was_clean(
fil_space_t* space,
mtr_t* mtr)
{
ut_ad(log_mutex_own());
if (space == NULL) {
return(false);
}
const bool was_clean = space->max_lsn == 0;
ut_ad(space->max_lsn <= log_sys.lsn);
space->max_lsn = log_sys.lsn;
if (was_clean) {
fil_names_dirty_and_write(space, mtr);
}
return(was_clean);
}
/** During crash recovery, open a tablespace if it had not been opened
yet, to get valid size and flags.
@param[in,out] space tablespace */
inline void fil_space_open_if_needed(fil_space_t* space)
{
ut_d(extern volatile bool recv_recovery_on);
ut_ad(recv_recovery_on);
if (space->size == 0) {
/* Initially, size and flags will be set to 0,
until the files are opened for the first time.
fil_space_get_size() will open the file
and adjust the size and flags. */
ut_d(ulint size =) fil_space_get_size(space->id);
ut_ad(size == space->size);
}
}
/** On a log checkpoint, reset fil_names_dirty_and_write() flags
and write out MLOG_FILE_NAME and MLOG_CHECKPOINT if needed.
@param[in] lsn checkpoint LSN
@param[in] do_write whether to always write MLOG_CHECKPOINT
@return whether anything was written to the redo log
@retval false if no flags were set and nothing written
@retval true if anything was written to the redo log */
bool
fil_names_clear(
lsn_t lsn,
bool do_write);
#ifdef UNIV_ENABLE_UNIT_TEST_MAKE_FILEPATH
void test_make_filepath();
#endif /* UNIV_ENABLE_UNIT_TEST_MAKE_FILEPATH */
/** Determine the block size of the data file.
@param[in] space tablespace
@param[in] offset page number
@return block size */
UNIV_INTERN
ulint
fil_space_get_block_size(const fil_space_t* space, unsigned offset);
#include "fil0fil.ic"
#endif /* UNIV_INNOCHECKSUM */
#endif /* fil0fil_h */
|