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
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
|
/******************************************************
The transaction
(c) 1996 Innobase Oy
Created 3/26/1996 Heikki Tuuri
*******************************************************/
#include "trx0trx.h"
#ifdef UNIV_NONINL
#include "trx0trx.ic"
#endif
#include "trx0undo.h"
#include "trx0rseg.h"
#include "log0log.h"
#include "que0que.h"
#include "lock0lock.h"
#include "trx0roll.h"
#include "usr0sess.h"
#include "read0read.h"
#include "srv0srv.h"
#include "thr0loc.h"
#include "btr0sea.h"
#include "os0proc.h"
#include "trx0xa.h"
/* Copy of the prototype for innobase_mysql_print_thd: this
copy MUST be equal to the one in mysql/sql/ha_innodb.cc ! */
void innobase_mysql_print_thd(
FILE* f,
void* thd,
ulint max_query_len);
/* Dummy session used currently in MySQL interface */
sess_t* trx_dummy_sess = NULL;
/* Number of transactions currently allocated for MySQL: protected by
the kernel mutex */
ulint trx_n_mysql_transactions = 0;
/*****************************************************************
Starts the transaction if it is not yet started. */
void
trx_start_if_not_started_noninline(
/*===============================*/
trx_t* trx) /* in: transaction */
{
trx_start_if_not_started(trx);
}
/*****************************************************************
Set detailed error message for the transaction. */
void
trx_set_detailed_error(
/*===================*/
trx_t* trx, /* in: transaction struct */
const char* msg) /* in: detailed error message */
{
ut_strlcpy(trx->detailed_error, msg, sizeof(trx->detailed_error));
}
/*****************************************************************
Set detailed error message for the transaction from a file. Note that the
file is rewinded before reading from it. */
void
trx_set_detailed_error_from_file(
/*=============================*/
trx_t* trx, /* in: transaction struct */
FILE* file) /* in: file to read message from */
{
os_file_read_string(file, trx->detailed_error,
sizeof(trx->detailed_error));
}
/********************************************************************
Retrieves the error_info field from a trx. */
void*
trx_get_error_info(
/*===============*/
/* out: the error info */
trx_t* trx) /* in: trx object */
{
return(trx->error_info);
}
/********************************************************************
Creates and initializes a transaction object. */
trx_t*
trx_create(
/*=======*/
/* out, own: the transaction */
sess_t* sess) /* in: session or NULL */
{
trx_t* trx;
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
trx = mem_alloc(sizeof(trx_t));
trx->magic_n = TRX_MAGIC_N;
trx->op_info = "";
trx->type = TRX_USER;
trx->conc_state = TRX_NOT_STARTED;
trx->start_time = time(NULL);
trx->isolation_level = TRX_ISO_REPEATABLE_READ;
trx->id = ut_dulint_zero;
trx->no = ut_dulint_max;
trx->support_xa = TRUE;
trx->check_foreigns = TRUE;
trx->check_unique_secondary = TRUE;
trx->flush_log_later = FALSE;
trx->must_flush_log_later = FALSE;
trx->dict_operation = FALSE;
trx->mysql_thd = NULL;
trx->mysql_query_str = NULL;
trx->n_mysql_tables_in_use = 0;
trx->mysql_n_tables_locked = 0;
trx->mysql_log_file_name = NULL;
trx->mysql_log_offset = 0;
trx->mysql_master_log_file_name = "";
trx->mysql_master_log_pos = 0;
trx->repl_wait_binlog_name = NULL;
trx->repl_wait_binlog_pos = 0;
mutex_create(&trx->undo_mutex, SYNC_TRX_UNDO);
trx->rseg = NULL;
trx->undo_no = ut_dulint_zero;
trx->last_sql_stat_start.least_undo_no = ut_dulint_zero;
trx->insert_undo = NULL;
trx->update_undo = NULL;
trx->undo_no_arr = NULL;
trx->error_state = DB_SUCCESS;
trx->detailed_error[0] = '\0';
trx->sess = sess;
trx->que_state = TRX_QUE_RUNNING;
trx->n_active_thrs = 0;
trx->handling_signals = FALSE;
UT_LIST_INIT(trx->signals);
UT_LIST_INIT(trx->reply_signals);
trx->graph = NULL;
trx->wait_lock = NULL;
trx->was_chosen_as_deadlock_victim = FALSE;
UT_LIST_INIT(trx->wait_thrs);
trx->lock_heap = mem_heap_create_in_buffer(256);
UT_LIST_INIT(trx->trx_locks);
UT_LIST_INIT(trx->trx_savepoints);
trx->dict_operation_lock_mode = 0;
trx->has_search_latch = FALSE;
trx->search_latch_timeout = BTR_SEA_TIMEOUT;
trx->declared_to_be_inside_innodb = FALSE;
trx->n_tickets_to_enter_innodb = 0;
trx->auto_inc_lock = NULL;
trx->global_read_view_heap = mem_heap_create(256);
trx->global_read_view = NULL;
trx->read_view = NULL;
/* Set X/Open XA transaction identification to NULL */
memset(&trx->xid, 0, sizeof(trx->xid));
trx->xid.formatID = -1;
trx_reset_new_rec_lock_info(trx);
return(trx);
}
/************************************************************************
Creates a transaction object for MySQL. */
trx_t*
trx_allocate_for_mysql(void)
/*========================*/
/* out, own: transaction object */
{
trx_t* trx;
mutex_enter(&kernel_mutex);
/* Open a dummy session */
if (!trx_dummy_sess) {
trx_dummy_sess = sess_open();
}
trx = trx_create(trx_dummy_sess);
trx_n_mysql_transactions++;
UT_LIST_ADD_FIRST(mysql_trx_list, trx_sys->mysql_trx_list, trx);
mutex_exit(&kernel_mutex);
trx->mysql_thread_id = os_thread_get_curr_id();
trx->mysql_process_no = os_proc_get_number();
return(trx);
}
/************************************************************************
Creates a transaction object for background operations by the master thread. */
trx_t*
trx_allocate_for_background(void)
/*=============================*/
/* out, own: transaction object */
{
trx_t* trx;
mutex_enter(&kernel_mutex);
/* Open a dummy session */
if (!trx_dummy_sess) {
trx_dummy_sess = sess_open();
}
trx = trx_create(trx_dummy_sess);
mutex_exit(&kernel_mutex);
return(trx);
}
/************************************************************************
Releases the search latch if trx has reserved it. */
void
trx_search_latch_release_if_reserved(
/*=================================*/
trx_t* trx) /* in: transaction */
{
if (trx->has_search_latch) {
rw_lock_s_unlock(&btr_search_latch);
trx->has_search_latch = FALSE;
}
}
/************************************************************************
Frees a transaction object. */
void
trx_free(
/*=====*/
trx_t* trx) /* in, own: trx object */
{
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
if (trx->declared_to_be_inside_innodb) {
ut_print_timestamp(stderr);
fputs(
" InnoDB: Error: Freeing a trx which is declared to be processing\n"
"InnoDB: inside InnoDB.\n", stderr);
trx_print(stderr, trx, 600);
putc('\n', stderr);
}
if (trx->n_mysql_tables_in_use != 0
|| trx->mysql_n_tables_locked != 0) {
ut_print_timestamp(stderr);
fprintf(stderr,
" InnoDB: Error: MySQL is freeing a thd\n"
"InnoDB: though trx->n_mysql_tables_in_use is %lu\n"
"InnoDB: and trx->mysql_n_tables_locked is %lu.\n",
(ulong)trx->n_mysql_tables_in_use,
(ulong)trx->mysql_n_tables_locked);
trx_print(stderr, trx, 600);
ut_print_buf(stderr, trx, sizeof(trx_t));
}
ut_a(trx->magic_n == TRX_MAGIC_N);
trx->magic_n = 11112222;
ut_a(trx->conc_state == TRX_NOT_STARTED);
mutex_free(&(trx->undo_mutex));
ut_a(trx->insert_undo == NULL);
ut_a(trx->update_undo == NULL);
if (trx->undo_no_arr) {
trx_undo_arr_free(trx->undo_no_arr);
}
if (trx->repl_wait_binlog_name != NULL) {
mem_free(trx->repl_wait_binlog_name);
}
ut_a(UT_LIST_GET_LEN(trx->signals) == 0);
ut_a(UT_LIST_GET_LEN(trx->reply_signals) == 0);
ut_a(trx->wait_lock == NULL);
ut_a(UT_LIST_GET_LEN(trx->wait_thrs) == 0);
ut_a(!trx->has_search_latch);
ut_a(!trx->auto_inc_lock);
ut_a(trx->dict_operation_lock_mode == 0);
if (trx->lock_heap) {
mem_heap_free(trx->lock_heap);
}
ut_a(UT_LIST_GET_LEN(trx->trx_locks) == 0);
if (trx->global_read_view_heap) {
mem_heap_free(trx->global_read_view_heap);
}
trx->global_read_view = NULL;
ut_a(trx->read_view == NULL);
mem_free(trx);
}
/************************************************************************
Frees a transaction object for MySQL. */
void
trx_free_for_mysql(
/*===============*/
trx_t* trx) /* in, own: trx object */
{
mutex_enter(&kernel_mutex);
UT_LIST_REMOVE(mysql_trx_list, trx_sys->mysql_trx_list, trx);
trx_free(trx);
ut_a(trx_n_mysql_transactions > 0);
trx_n_mysql_transactions--;
mutex_exit(&kernel_mutex);
}
/************************************************************************
Frees a transaction object of a background operation of the master thread. */
void
trx_free_for_background(
/*====================*/
trx_t* trx) /* in, own: trx object */
{
mutex_enter(&kernel_mutex);
trx_free(trx);
mutex_exit(&kernel_mutex);
}
/********************************************************************
Inserts the trx handle in the trx system trx list in the right position.
The list is sorted on the trx id so that the biggest id is at the list
start. This function is used at the database startup to insert incomplete
transactions to the list. */
static
void
trx_list_insert_ordered(
/*====================*/
trx_t* trx) /* in: trx handle */
{
trx_t* trx2;
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
trx2 = UT_LIST_GET_FIRST(trx_sys->trx_list);
while (trx2 != NULL) {
if (ut_dulint_cmp(trx->id, trx2->id) >= 0) {
ut_ad(ut_dulint_cmp(trx->id, trx2->id) == 1);
break;
}
trx2 = UT_LIST_GET_NEXT(trx_list, trx2);
}
if (trx2 != NULL) {
trx2 = UT_LIST_GET_PREV(trx_list, trx2);
if (trx2 == NULL) {
UT_LIST_ADD_FIRST(trx_list, trx_sys->trx_list, trx);
} else {
UT_LIST_INSERT_AFTER(trx_list, trx_sys->trx_list,
trx2, trx);
}
} else {
UT_LIST_ADD_LAST(trx_list, trx_sys->trx_list, trx);
}
}
/********************************************************************
Creates trx objects for transactions and initializes the trx list of
trx_sys at database start. Rollback segment and undo log lists must
already exist when this function is called, because the lists of
transactions to be rolled back or cleaned up are built based on the
undo log lists. */
void
trx_lists_init_at_db_start(void)
/*============================*/
{
trx_rseg_t* rseg;
trx_undo_t* undo;
trx_t* trx;
UT_LIST_INIT(trx_sys->trx_list);
/* Look from the rollback segments if there exist undo logs for
transactions */
rseg = UT_LIST_GET_FIRST(trx_sys->rseg_list);
while (rseg != NULL) {
undo = UT_LIST_GET_FIRST(rseg->insert_undo_list);
while (undo != NULL) {
trx = trx_create(NULL);
trx->id = undo->trx_id;
trx->xid = undo->xid;
trx->insert_undo = undo;
trx->rseg = rseg;
if (undo->state != TRX_UNDO_ACTIVE) {
/* Prepared transactions are left in
the prepared state waiting for a
commit or abort decision from MySQL */
if (undo->state == TRX_UNDO_PREPARED) {
fprintf(stderr,
"InnoDB: Transaction %lu %lu was in the XA prepared state.\n",
ut_dulint_get_high(trx->id),
ut_dulint_get_low(trx->id));
if (srv_force_recovery == 0) {
trx->conc_state = TRX_PREPARED;
} else {
fprintf(stderr,
"InnoDB: Since innodb_force_recovery > 0, we will rollback it anyway.\n");
trx->conc_state = TRX_ACTIVE;
}
} else {
trx->conc_state =
TRX_COMMITTED_IN_MEMORY;
}
/* We give a dummy value for the trx no;
this should have no relevance since purge
is not interested in committed transaction
numbers, unless they are in the history
list, in which case it looks the number
from the disk based undo log structure */
trx->no = trx->id;
} else {
trx->conc_state = TRX_ACTIVE;
/* A running transaction always has the number
field inited to ut_dulint_max */
trx->no = ut_dulint_max;
}
if (undo->dict_operation) {
trx->dict_operation = undo->dict_operation;
trx->table_id = undo->table_id;
}
if (!undo->empty) {
trx->undo_no = ut_dulint_add(undo->top_undo_no,
1);
}
trx_list_insert_ordered(trx);
undo = UT_LIST_GET_NEXT(undo_list, undo);
}
undo = UT_LIST_GET_FIRST(rseg->update_undo_list);
while (undo != NULL) {
trx = trx_get_on_id(undo->trx_id);
if (NULL == trx) {
trx = trx_create(NULL);
trx->id = undo->trx_id;
trx->xid = undo->xid;
if (undo->state != TRX_UNDO_ACTIVE) {
/* Prepared transactions are left in
the prepared state waiting for a
commit or abort decision from MySQL */
if (undo->state == TRX_UNDO_PREPARED) {
fprintf(stderr,
"InnoDB: Transaction %lu %lu was in the XA prepared state.\n",
ut_dulint_get_high(trx->id),
ut_dulint_get_low(trx->id));
if (srv_force_recovery == 0) {
trx->conc_state =
TRX_PREPARED;
} else {
fprintf(stderr,
"InnoDB: Since innodb_force_recovery > 0, we will rollback it anyway.\n");
trx->conc_state =
TRX_ACTIVE;
}
} else {
trx->conc_state =
TRX_COMMITTED_IN_MEMORY;
}
/* We give a dummy value for the trx
number */
trx->no = trx->id;
} else {
trx->conc_state = TRX_ACTIVE;
/* A running transaction always has
the number field inited to
ut_dulint_max */
trx->no = ut_dulint_max;
}
trx->rseg = rseg;
trx_list_insert_ordered(trx);
if (undo->dict_operation) {
trx->dict_operation =
undo->dict_operation;
trx->table_id = undo->table_id;
}
}
trx->update_undo = undo;
if ((!undo->empty)
&& (ut_dulint_cmp(undo->top_undo_no,
trx->undo_no) >= 0)) {
trx->undo_no = ut_dulint_add(undo->top_undo_no,
1);
}
undo = UT_LIST_GET_NEXT(undo_list, undo);
}
rseg = UT_LIST_GET_NEXT(rseg_list, rseg);
}
}
/**********************************************************************
Assigns a rollback segment to a transaction in a round-robin fashion.
Skips the SYSTEM rollback segment if another is available. */
UNIV_INLINE
ulint
trx_assign_rseg(void)
/*=================*/
/* out: assigned rollback segment id */
{
trx_rseg_t* rseg = trx_sys->latest_rseg;
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
loop:
/* Get next rseg in a round-robin fashion */
rseg = UT_LIST_GET_NEXT(rseg_list, rseg);
if (rseg == NULL) {
rseg = UT_LIST_GET_FIRST(trx_sys->rseg_list);
}
/* If it is the SYSTEM rollback segment, and there exist others, skip
it */
if ((rseg->id == TRX_SYS_SYSTEM_RSEG_ID)
&& (UT_LIST_GET_LEN(trx_sys->rseg_list) > 1)) {
goto loop;
}
trx_sys->latest_rseg = rseg;
return(rseg->id);
}
/********************************************************************
Starts a new transaction. */
ibool
trx_start_low(
/*==========*/
/* out: TRUE */
trx_t* trx, /* in: transaction */
ulint rseg_id)/* in: rollback segment id; if ULINT_UNDEFINED
is passed, the system chooses the rollback segment
automatically in a round-robin fashion */
{
trx_rseg_t* rseg;
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
ut_ad(trx->rseg == NULL);
if (trx->type == TRX_PURGE) {
trx->id = ut_dulint_zero;
trx->conc_state = TRX_ACTIVE;
trx->start_time = time(NULL);
return(TRUE);
}
ut_ad(trx->conc_state != TRX_ACTIVE);
if (rseg_id == ULINT_UNDEFINED) {
rseg_id = trx_assign_rseg();
}
rseg = trx_sys_get_nth_rseg(trx_sys, rseg_id);
trx->id = trx_sys_get_new_trx_id();
/* The initial value for trx->no: ut_dulint_max is used in
read_view_open_now: */
trx->no = ut_dulint_max;
trx->rseg = rseg;
trx->conc_state = TRX_ACTIVE;
trx->start_time = time(NULL);
UT_LIST_ADD_FIRST(trx_list, trx_sys->trx_list, trx);
return(TRUE);
}
/********************************************************************
Starts a new transaction. */
ibool
trx_start(
/*======*/
/* out: TRUE */
trx_t* trx, /* in: transaction */
ulint rseg_id)/* in: rollback segment id; if ULINT_UNDEFINED
is passed, the system chooses the rollback segment
automatically in a round-robin fashion */
{
ibool ret;
mutex_enter(&kernel_mutex);
ret = trx_start_low(trx, rseg_id);
mutex_exit(&kernel_mutex);
return(ret);
}
/********************************************************************
Commits a transaction. */
void
trx_commit_off_kernel(
/*==================*/
trx_t* trx) /* in: transaction */
{
page_t* update_hdr_page;
dulint lsn;
trx_rseg_t* rseg;
trx_undo_t* undo;
ibool must_flush_log = FALSE;
mtr_t mtr;
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
trx->must_flush_log_later = FALSE;
rseg = trx->rseg;
if (trx->insert_undo != NULL || trx->update_undo != NULL) {
mutex_exit(&kernel_mutex);
mtr_start(&mtr);
must_flush_log = TRUE;
/* Change the undo log segment states from TRX_UNDO_ACTIVE
to some other state: these modifications to the file data
structure define the transaction as committed in the file
based world, at the serialization point of the log sequence
number lsn obtained below. */
mutex_enter(&(rseg->mutex));
if (trx->insert_undo != NULL) {
trx_undo_set_state_at_finish(trx, trx->insert_undo,
&mtr);
}
undo = trx->update_undo;
if (undo) {
mutex_enter(&kernel_mutex);
trx->no = trx_sys_get_new_trx_no();
mutex_exit(&kernel_mutex);
/* It is not necessary to obtain trx->undo_mutex here
because only a single OS thread is allowed to do the
transaction commit for this transaction. */
update_hdr_page = trx_undo_set_state_at_finish(trx,
undo, &mtr);
/* We have to do the cleanup for the update log while
holding the rseg mutex because update log headers
have to be put to the history list in the order of
the trx number. */
trx_undo_update_cleanup(trx, update_hdr_page, &mtr);
}
mutex_exit(&(rseg->mutex));
/* Update the latest MySQL binlog name and offset info
in trx sys header if MySQL binlogging is on or the database
server is a MySQL replication slave */
if (trx->mysql_log_file_name
&& trx->mysql_log_file_name[0] != '\0') {
trx_sys_update_mysql_binlog_offset(
trx->mysql_log_file_name,
trx->mysql_log_offset,
TRX_SYS_MYSQL_LOG_INFO, &mtr);
trx->mysql_log_file_name = NULL;
}
if (trx->mysql_master_log_file_name[0] != '\0') {
/* This database server is a MySQL replication slave */
trx_sys_update_mysql_binlog_offset(
trx->mysql_master_log_file_name,
trx->mysql_master_log_pos,
TRX_SYS_MYSQL_MASTER_LOG_INFO, &mtr);
}
/* The following call commits the mini-transaction, making the
whole transaction committed in the file-based world, at this
log sequence number. The transaction becomes 'durable' when
we write the log to disk, but in the logical sense the commit
in the file-based data structures (undo logs etc.) happens
here.
NOTE that transaction numbers, which are assigned only to
transactions with an update undo log, do not necessarily come
in exactly the same order as commit lsn's, if the transactions
have different rollback segments. To get exactly the same
order we should hold the kernel mutex up to this point,
adding to to the contention of the kernel mutex. However, if
a transaction T2 is able to see modifications made by
a transaction T1, T2 will always get a bigger transaction
number and a bigger commit lsn than T1. */
/*--------------*/
mtr_commit(&mtr);
/*--------------*/
lsn = mtr.end_lsn;
mutex_enter(&kernel_mutex);
}
ut_ad(trx->conc_state == TRX_ACTIVE
|| trx->conc_state == TRX_PREPARED);
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
/* The following assignment makes the transaction committed in memory
and makes its changes to data visible to other transactions.
NOTE that there is a small discrepancy from the strict formal
visibility rules here: a human user of the database can see
modifications made by another transaction T even before the necessary
log segment has been flushed to the disk. If the database happens to
crash before the flush, the user has seen modifications from T which
will never be a committed transaction. However, any transaction T2
which sees the modifications of the committing transaction T, and
which also itself makes modifications to the database, will get an lsn
larger than the committing transaction T. In the case where the log
flush fails, and T never gets committed, also T2 will never get
committed. */
/*--------------------------------------*/
trx->conc_state = TRX_COMMITTED_IN_MEMORY;
/*--------------------------------------*/
lock_release_off_kernel(trx);
if (trx->global_read_view) {
read_view_close(trx->global_read_view);
mem_heap_empty(trx->global_read_view_heap);
trx->global_read_view = NULL;
}
trx->read_view = NULL;
if (must_flush_log) {
mutex_exit(&kernel_mutex);
if (trx->insert_undo != NULL) {
trx_undo_insert_cleanup(trx);
}
/* NOTE that we could possibly make a group commit more
efficient here: call os_thread_yield here to allow also other
trxs to come to commit! */
/*-------------------------------------*/
/* Depending on the my.cnf options, we may now write the log
buffer to the log files, making the transaction durable if
the OS does not crash. We may also flush the log files to
disk, making the transaction durable also at an OS crash or a
power outage.
The idea in InnoDB's group commit is that a group of
transactions gather behind a trx doing a physical disk write
to log files, and when that physical write has been completed,
one of those transactions does a write which commits the whole
group. Note that this group commit will only bring benefit if
there are > 2 users in the database. Then at least 2 users can
gather behind one doing the physical log write to disk.
If we are calling trx_commit() under MySQL's binlog mutex, we
will delay possible log write and flush to a separate function
trx_commit_complete_for_mysql(), which is only called when the
thread has released the binlog mutex. This is to make the
group commit algorithm to work. Otherwise, the MySQL binlog
mutex would serialize all commits and prevent a group of
transactions from gathering. */
if (trx->flush_log_later) {
/* Do nothing yet */
trx->must_flush_log_later = TRUE;
} else if (srv_flush_log_at_trx_commit == 0) {
/* Do nothing */
} else if (srv_flush_log_at_trx_commit == 1) {
if (srv_unix_file_flush_method == SRV_UNIX_NOSYNC) {
/* Write the log but do not flush it to disk */
log_write_up_to(lsn, LOG_WAIT_ONE_GROUP,
FALSE);
} else {
/* Write the log to the log files AND flush
them to disk */
log_write_up_to(lsn, LOG_WAIT_ONE_GROUP, TRUE);
}
} else if (srv_flush_log_at_trx_commit == 2) {
/* Write the log but do not flush it to disk */
log_write_up_to(lsn, LOG_WAIT_ONE_GROUP, FALSE);
} else {
ut_error;
}
trx->commit_lsn = lsn;
/*-------------------------------------*/
mutex_enter(&kernel_mutex);
}
/* Free savepoints */
trx_roll_savepoints_free(trx, NULL);
trx->conc_state = TRX_NOT_STARTED;
trx->rseg = NULL;
trx->undo_no = ut_dulint_zero;
trx->last_sql_stat_start.least_undo_no = ut_dulint_zero;
ut_ad(UT_LIST_GET_LEN(trx->wait_thrs) == 0);
ut_ad(UT_LIST_GET_LEN(trx->trx_locks) == 0);
UT_LIST_REMOVE(trx_list, trx_sys->trx_list, trx);
}
/********************************************************************
Cleans up a transaction at database startup. The cleanup is needed if
the transaction already got to the middle of a commit when the database
crashed, andf we cannot roll it back. */
void
trx_cleanup_at_db_startup(
/*======================*/
trx_t* trx) /* in: transaction */
{
if (trx->insert_undo != NULL) {
trx_undo_insert_cleanup(trx);
}
trx->conc_state = TRX_NOT_STARTED;
trx->rseg = NULL;
trx->undo_no = ut_dulint_zero;
trx->last_sql_stat_start.least_undo_no = ut_dulint_zero;
UT_LIST_REMOVE(trx_list, trx_sys->trx_list, trx);
}
/************************************************************************
Assigns a read view for a consistent read query. All the consistent reads
within the same transaction will get the same read view, which is created
when this function is first called for a new started transaction. */
read_view_t*
trx_assign_read_view(
/*=================*/
/* out: consistent read view */
trx_t* trx) /* in: active transaction */
{
ut_ad(trx->conc_state == TRX_ACTIVE);
if (trx->read_view) {
return(trx->read_view);
}
mutex_enter(&kernel_mutex);
if (!trx->read_view) {
trx->read_view = read_view_open_now(trx->id,
trx->global_read_view_heap);
trx->global_read_view = trx->read_view;
}
mutex_exit(&kernel_mutex);
return(trx->read_view);
}
/********************************************************************
Commits a transaction. NOTE that the kernel mutex is temporarily released. */
static
void
trx_handle_commit_sig_off_kernel(
/*=============================*/
trx_t* trx, /* in: transaction */
que_thr_t** next_thr) /* in/out: next query thread to run;
if the value which is passed in is
a pointer to a NULL pointer, then the
calling function can start running
a new query thread */
{
trx_sig_t* sig;
trx_sig_t* next_sig;
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
trx->que_state = TRX_QUE_COMMITTING;
trx_commit_off_kernel(trx);
ut_ad(UT_LIST_GET_LEN(trx->wait_thrs) == 0);
/* Remove all TRX_SIG_COMMIT signals from the signal queue and send
reply messages to them */
sig = UT_LIST_GET_FIRST(trx->signals);
while (sig != NULL) {
next_sig = UT_LIST_GET_NEXT(signals, sig);
if (sig->type == TRX_SIG_COMMIT) {
trx_sig_reply(sig, next_thr);
trx_sig_remove(trx, sig);
}
sig = next_sig;
}
trx->que_state = TRX_QUE_RUNNING;
}
/***************************************************************
The transaction must be in the TRX_QUE_LOCK_WAIT state. Puts it to
the TRX_QUE_RUNNING state and releases query threads which were
waiting for a lock in the wait_thrs list. */
void
trx_end_lock_wait(
/*==============*/
trx_t* trx) /* in: transaction */
{
que_thr_t* thr;
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
ut_ad(trx->que_state == TRX_QUE_LOCK_WAIT);
thr = UT_LIST_GET_FIRST(trx->wait_thrs);
while (thr != NULL) {
que_thr_end_wait_no_next_thr(thr);
UT_LIST_REMOVE(trx_thrs, trx->wait_thrs, thr);
thr = UT_LIST_GET_FIRST(trx->wait_thrs);
}
trx->que_state = TRX_QUE_RUNNING;
}
/***************************************************************
Moves the query threads in the lock wait list to the SUSPENDED state and puts
the transaction to the TRX_QUE_RUNNING state. */
static
void
trx_lock_wait_to_suspended(
/*=======================*/
trx_t* trx) /* in: transaction in the TRX_QUE_LOCK_WAIT state */
{
que_thr_t* thr;
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
ut_ad(trx->que_state == TRX_QUE_LOCK_WAIT);
thr = UT_LIST_GET_FIRST(trx->wait_thrs);
while (thr != NULL) {
thr->state = QUE_THR_SUSPENDED;
UT_LIST_REMOVE(trx_thrs, trx->wait_thrs, thr);
thr = UT_LIST_GET_FIRST(trx->wait_thrs);
}
trx->que_state = TRX_QUE_RUNNING;
}
/***************************************************************
Moves the query threads in the sig reply wait list of trx to the SUSPENDED
state. */
static
void
trx_sig_reply_wait_to_suspended(
/*============================*/
trx_t* trx) /* in: transaction */
{
trx_sig_t* sig;
que_thr_t* thr;
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
sig = UT_LIST_GET_FIRST(trx->reply_signals);
while (sig != NULL) {
thr = sig->receiver;
ut_ad(thr->state == QUE_THR_SIG_REPLY_WAIT);
thr->state = QUE_THR_SUSPENDED;
sig->receiver = NULL;
UT_LIST_REMOVE(reply_signals, trx->reply_signals, sig);
sig = UT_LIST_GET_FIRST(trx->reply_signals);
}
}
/*********************************************************************
Checks the compatibility of a new signal with the other signals in the
queue. */
static
ibool
trx_sig_is_compatible(
/*==================*/
/* out: TRUE if the signal can be queued */
trx_t* trx, /* in: trx handle */
ulint type, /* in: signal type */
ulint sender) /* in: TRX_SIG_SELF or TRX_SIG_OTHER_SESS */
{
trx_sig_t* sig;
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
if (UT_LIST_GET_LEN(trx->signals) == 0) {
return(TRUE);
}
if (sender == TRX_SIG_SELF) {
if (type == TRX_SIG_ERROR_OCCURRED) {
return(TRUE);
} else if (type == TRX_SIG_BREAK_EXECUTION) {
return(TRUE);
} else {
return(FALSE);
}
}
ut_ad(sender == TRX_SIG_OTHER_SESS);
sig = UT_LIST_GET_FIRST(trx->signals);
if (type == TRX_SIG_COMMIT) {
while (sig != NULL) {
if (sig->type == TRX_SIG_TOTAL_ROLLBACK) {
return(FALSE);
}
sig = UT_LIST_GET_NEXT(signals, sig);
}
return(TRUE);
} else if (type == TRX_SIG_TOTAL_ROLLBACK) {
while (sig != NULL) {
if (sig->type == TRX_SIG_COMMIT) {
return(FALSE);
}
sig = UT_LIST_GET_NEXT(signals, sig);
}
return(TRUE);
} else if (type == TRX_SIG_BREAK_EXECUTION) {
return(TRUE);
} else {
ut_error;
return(FALSE);
}
}
/********************************************************************
Sends a signal to a trx object. */
void
trx_sig_send(
/*=========*/
trx_t* trx, /* in: trx handle */
ulint type, /* in: signal type */
ulint sender, /* in: TRX_SIG_SELF or
TRX_SIG_OTHER_SESS */
que_thr_t* receiver_thr, /* in: query thread which wants the
reply, or NULL; if type is
TRX_SIG_END_WAIT, this must be NULL */
trx_savept_t* savept, /* in: possible rollback savepoint, or
NULL */
que_thr_t** next_thr) /* in/out: next query thread to run;
if the value which is passed in is
a pointer to a NULL pointer, then the
calling function can start running
a new query thread; if the parameter
is NULL, it is ignored */
{
trx_sig_t* sig;
trx_t* receiver_trx;
ut_ad(trx);
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
if (!trx_sig_is_compatible(trx, type, sender)) {
/* The signal is not compatible with the other signals in
the queue: die */
ut_error;
}
/* Queue the signal object */
if (UT_LIST_GET_LEN(trx->signals) == 0) {
/* The signal list is empty: the 'sig' slot must be unused
(we improve performance a bit by avoiding mem_alloc) */
sig = &(trx->sig);
} else {
/* It might be that the 'sig' slot is unused also in this
case, but we choose the easy way of using mem_alloc */
sig = mem_alloc(sizeof(trx_sig_t));
}
UT_LIST_ADD_LAST(signals, trx->signals, sig);
sig->type = type;
sig->state = TRX_SIG_WAITING;
sig->sender = sender;
sig->receiver = receiver_thr;
if (savept) {
sig->savept = *savept;
}
if (receiver_thr) {
receiver_trx = thr_get_trx(receiver_thr);
UT_LIST_ADD_LAST(reply_signals, receiver_trx->reply_signals,
sig);
}
if (trx->sess->state == SESS_ERROR) {
trx_sig_reply_wait_to_suspended(trx);
}
if ((sender != TRX_SIG_SELF) || (type == TRX_SIG_BREAK_EXECUTION)) {
ut_error;
}
/* If there were no other signals ahead in the queue, try to start
handling of the signal */
if (UT_LIST_GET_FIRST(trx->signals) == sig) {
trx_sig_start_handle(trx, next_thr);
}
}
/********************************************************************
Ends signal handling. If the session is in the error state, and
trx->graph_before_signal_handling != NULL, then returns control to the error
handling routine of the graph (currently just returns the control to the
graph root which then will send an error message to the client). */
void
trx_end_signal_handling(
/*====================*/
trx_t* trx) /* in: trx */
{
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
ut_ad(trx->handling_signals == TRUE);
trx->handling_signals = FALSE;
trx->graph = trx->graph_before_signal_handling;
if (trx->graph && (trx->sess->state == SESS_ERROR)) {
que_fork_error_handle(trx, trx->graph);
}
}
/********************************************************************
Starts handling of a trx signal. */
void
trx_sig_start_handle(
/*=================*/
trx_t* trx, /* in: trx handle */
que_thr_t** next_thr) /* in/out: next query thread to run;
if the value which is passed in is
a pointer to a NULL pointer, then the
calling function can start running
a new query thread; if the parameter
is NULL, it is ignored */
{
trx_sig_t* sig;
ulint type;
loop:
/* We loop in this function body as long as there are queued signals
we can process immediately */
ut_ad(trx);
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
if (trx->handling_signals && (UT_LIST_GET_LEN(trx->signals) == 0)) {
trx_end_signal_handling(trx);
return;
}
if (trx->conc_state == TRX_NOT_STARTED) {
trx_start_low(trx, ULINT_UNDEFINED);
}
/* If the trx is in a lock wait state, moves the waiting query threads
to the suspended state */
if (trx->que_state == TRX_QUE_LOCK_WAIT) {
trx_lock_wait_to_suspended(trx);
}
/* If the session is in the error state and this trx has threads
waiting for reply from signals, moves these threads to the suspended
state, canceling wait reservations; note that if the transaction has
sent a commit or rollback signal to itself, and its session is not in
the error state, then nothing is done here. */
if (trx->sess->state == SESS_ERROR) {
trx_sig_reply_wait_to_suspended(trx);
}
/* If there are no running query threads, we can start processing of a
signal, otherwise we have to wait until all query threads of this
transaction are aware of the arrival of the signal. */
if (trx->n_active_thrs > 0) {
return;
}
if (trx->handling_signals == FALSE) {
trx->graph_before_signal_handling = trx->graph;
trx->handling_signals = TRUE;
}
sig = UT_LIST_GET_FIRST(trx->signals);
type = sig->type;
if (type == TRX_SIG_COMMIT) {
trx_handle_commit_sig_off_kernel(trx, next_thr);
} else if ((type == TRX_SIG_TOTAL_ROLLBACK)
|| (type == TRX_SIG_ROLLBACK_TO_SAVEPT)) {
trx_rollback(trx, sig, next_thr);
/* No further signals can be handled until the rollback
completes, therefore we return */
return;
} else if (type == TRX_SIG_ERROR_OCCURRED) {
trx_rollback(trx, sig, next_thr);
/* No further signals can be handled until the rollback
completes, therefore we return */
return;
} else if (type == TRX_SIG_BREAK_EXECUTION) {
trx_sig_reply(sig, next_thr);
trx_sig_remove(trx, sig);
} else {
ut_error;
}
goto loop;
}
/********************************************************************
Send the reply message when a signal in the queue of the trx has been
handled. */
void
trx_sig_reply(
/*==========*/
trx_sig_t* sig, /* in: signal */
que_thr_t** next_thr) /* in/out: next query thread to run;
if the value which is passed in is
a pointer to a NULL pointer, then the
calling function can start running
a new query thread */
{
trx_t* receiver_trx;
ut_ad(sig);
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
if (sig->receiver != NULL) {
ut_ad((sig->receiver)->state == QUE_THR_SIG_REPLY_WAIT);
receiver_trx = thr_get_trx(sig->receiver);
UT_LIST_REMOVE(reply_signals, receiver_trx->reply_signals,
sig);
ut_ad(receiver_trx->sess->state != SESS_ERROR);
que_thr_end_wait(sig->receiver, next_thr);
sig->receiver = NULL;
}
}
/********************************************************************
Removes a signal object from the trx signal queue. */
void
trx_sig_remove(
/*===========*/
trx_t* trx, /* in: trx handle */
trx_sig_t* sig) /* in, own: signal */
{
ut_ad(trx && sig);
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
ut_ad(sig->receiver == NULL);
UT_LIST_REMOVE(signals, trx->signals, sig);
sig->type = 0; /* reset the field to catch possible bugs */
if (sig != &(trx->sig)) {
mem_free(sig);
}
}
/*************************************************************************
Creates a commit command node struct. */
commit_node_t*
commit_node_create(
/*===============*/
/* out, own: commit node struct */
mem_heap_t* heap) /* in: mem heap where created */
{
commit_node_t* node;
node = mem_heap_alloc(heap, sizeof(commit_node_t));
node->common.type = QUE_NODE_COMMIT;
node->state = COMMIT_NODE_SEND;
return(node);
}
/***************************************************************
Performs an execution step for a commit type node in a query graph. */
que_thr_t*
trx_commit_step(
/*============*/
/* out: query thread to run next, or NULL */
que_thr_t* thr) /* in: query thread */
{
commit_node_t* node;
que_thr_t* next_thr;
node = thr->run_node;
ut_ad(que_node_get_type(node) == QUE_NODE_COMMIT);
if (thr->prev_node == que_node_get_parent(node)) {
node->state = COMMIT_NODE_SEND;
}
if (node->state == COMMIT_NODE_SEND) {
mutex_enter(&kernel_mutex);
node->state = COMMIT_NODE_WAIT;
next_thr = NULL;
thr->state = QUE_THR_SIG_REPLY_WAIT;
/* Send the commit signal to the transaction */
trx_sig_send(thr_get_trx(thr), TRX_SIG_COMMIT, TRX_SIG_SELF,
thr, NULL, &next_thr);
mutex_exit(&kernel_mutex);
return(next_thr);
}
ut_ad(node->state == COMMIT_NODE_WAIT);
node->state = COMMIT_NODE_SEND;
thr->run_node = que_node_get_parent(node);
return(thr);
}
/**************************************************************************
Does the transaction commit for MySQL. */
ulint
trx_commit_for_mysql(
/*=================*/
/* out: 0 or error number */
trx_t* trx) /* in: trx handle */
{
/* Because we do not do the commit by sending an Innobase
sig to the transaction, we must here make sure that trx has been
started. */
ut_a(trx);
trx->op_info = "committing";
trx_start_if_not_started(trx);
mutex_enter(&kernel_mutex);
trx_commit_off_kernel(trx);
mutex_exit(&kernel_mutex);
trx->op_info = "";
return(0);
}
/**************************************************************************
If required, flushes the log to disk if we called trx_commit_for_mysql()
with trx->flush_log_later == TRUE. */
ulint
trx_commit_complete_for_mysql(
/*==========================*/
/* out: 0 or error number */
trx_t* trx) /* in: trx handle */
{
dulint lsn = trx->commit_lsn;
ut_a(trx);
trx->op_info = "flushing log";
if (!trx->must_flush_log_later) {
/* Do nothing */
} else if (srv_flush_log_at_trx_commit == 0) {
/* Do nothing */
} else if (srv_flush_log_at_trx_commit == 1) {
if (srv_unix_file_flush_method == SRV_UNIX_NOSYNC) {
/* Write the log but do not flush it to disk */
log_write_up_to(lsn, LOG_WAIT_ONE_GROUP, FALSE);
} else {
/* Write the log to the log files AND flush them to
disk */
log_write_up_to(lsn, LOG_WAIT_ONE_GROUP, TRUE);
}
} else if (srv_flush_log_at_trx_commit == 2) {
/* Write the log but do not flush it to disk */
log_write_up_to(lsn, LOG_WAIT_ONE_GROUP, FALSE);
} else {
ut_error;
}
trx->must_flush_log_later = FALSE;
trx->op_info = "";
return(0);
}
/**************************************************************************
Marks the latest SQL statement ended. */
void
trx_mark_sql_stat_end(
/*==================*/
trx_t* trx) /* in: trx handle */
{
ut_a(trx);
if (trx->conc_state == TRX_NOT_STARTED) {
trx->undo_no = ut_dulint_zero;
}
trx->last_sql_stat_start.least_undo_no = trx->undo_no;
}
/**************************************************************************
Prints info about a transaction to the given file. The caller must own the
kernel mutex and must have called
innobase_mysql_prepare_print_arbitrary_thd(), unless he knows that MySQL
or InnoDB cannot meanwhile change the info printed here. */
void
trx_print(
/*======*/
FILE* f, /* in: output stream */
trx_t* trx, /* in: transaction */
ulint max_query_len) /* in: max query length to print, or 0 to
use the default max length */
{
ibool newline;
fprintf(f, "TRANSACTION %lu %lu",
(ulong) ut_dulint_get_high(trx->id),
(ulong) ut_dulint_get_low(trx->id));
switch (trx->conc_state) {
case TRX_NOT_STARTED:
fputs(", not started", f);
break;
case TRX_ACTIVE:
fprintf(f, ", ACTIVE %lu sec",
(ulong)difftime(time(NULL), trx->start_time));
break;
case TRX_PREPARED:
fprintf(f, ", ACTIVE (PREPARED) %lu sec",
(ulong)difftime(time(NULL), trx->start_time));
break;
case TRX_COMMITTED_IN_MEMORY:
fputs(", COMMITTED IN MEMORY", f);
break;
default:
fprintf(f, " state %lu", (ulong) trx->conc_state);
}
#ifdef UNIV_LINUX
fprintf(f, ", process no %lu", trx->mysql_process_no);
#endif
fprintf(f, ", OS thread id %lu",
(ulong) os_thread_pf(trx->mysql_thread_id));
if (*trx->op_info) {
putc(' ', f);
fputs(trx->op_info, f);
}
if (trx->type != TRX_USER) {
fputs(" purge trx", f);
}
if (trx->declared_to_be_inside_innodb) {
fprintf(f, ", thread declared inside InnoDB %lu",
(ulong) trx->n_tickets_to_enter_innodb);
}
putc('\n', f);
if (trx->n_mysql_tables_in_use > 0 || trx->mysql_n_tables_locked > 0) {
fprintf(f, "mysql tables in use %lu, locked %lu\n",
(ulong) trx->n_mysql_tables_in_use,
(ulong) trx->mysql_n_tables_locked);
}
newline = TRUE;
switch (trx->que_state) {
case TRX_QUE_RUNNING:
newline = FALSE; break;
case TRX_QUE_LOCK_WAIT:
fputs("LOCK WAIT ", f); break;
case TRX_QUE_ROLLING_BACK:
fputs("ROLLING BACK ", f); break;
case TRX_QUE_COMMITTING:
fputs("COMMITTING ", f); break;
default:
fprintf(f, "que state %lu ", (ulong) trx->que_state);
}
if (0 < UT_LIST_GET_LEN(trx->trx_locks) ||
mem_heap_get_size(trx->lock_heap) > 400) {
newline = TRUE;
fprintf(f, "%lu lock struct(s), heap size %lu",
(ulong) UT_LIST_GET_LEN(trx->trx_locks),
(ulong) mem_heap_get_size(trx->lock_heap));
fprintf(f, "%lu row lock(s)",
(ulong) lock_number_of_rows_locked(trx));
}
if (trx->has_search_latch) {
newline = TRUE;
fputs(", holds adaptive hash latch", f);
}
if (ut_dulint_cmp(trx->undo_no, ut_dulint_zero) != 0) {
newline = TRUE;
fprintf(f, ", undo log entries %lu",
(ulong) ut_dulint_get_low(trx->undo_no));
}
if (newline) {
putc('\n', f);
}
if (trx->mysql_thd != NULL) {
innobase_mysql_print_thd(f, trx->mysql_thd, max_query_len);
}
}
/********************************************************************
Prepares a transaction. */
void
trx_prepare_off_kernel(
/*===================*/
trx_t* trx) /* in: transaction */
{
page_t* update_hdr_page;
trx_rseg_t* rseg;
ibool must_flush_log = FALSE;
dulint lsn;
mtr_t mtr;
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
rseg = trx->rseg;
if (trx->insert_undo != NULL || trx->update_undo != NULL) {
mutex_exit(&kernel_mutex);
mtr_start(&mtr);
must_flush_log = TRUE;
/* Change the undo log segment states from TRX_UNDO_ACTIVE
to TRX_UNDO_PREPARED: these modifications to the file data
structure define the transaction as prepared in the
file-based world, at the serialization point of lsn. */
mutex_enter(&(rseg->mutex));
if (trx->insert_undo != NULL) {
/* It is not necessary to obtain trx->undo_mutex here
because only a single OS thread is allowed to do the
transaction prepare for this transaction. */
trx_undo_set_state_at_prepare(trx, trx->insert_undo,
&mtr);
}
if (trx->update_undo) {
update_hdr_page = trx_undo_set_state_at_prepare(trx,
trx->update_undo, &mtr);
}
mutex_exit(&(rseg->mutex));
/*--------------*/
mtr_commit(&mtr); /* This mtr commit makes the
transaction prepared in the file-based
world */
/*--------------*/
lsn = mtr.end_lsn;
mutex_enter(&kernel_mutex);
}
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&kernel_mutex));
#endif /* UNIV_SYNC_DEBUG */
/*--------------------------------------*/
trx->conc_state = TRX_PREPARED;
/*--------------------------------------*/
if (must_flush_log) {
/* Depending on the my.cnf options, we may now write the log
buffer to the log files, making the prepared state of the
transaction durable if the OS does not crash. We may also
flush the log files to disk, making the prepared state of the
transaction durable also at an OS crash or a power outage.
The idea in InnoDB's group prepare is that a group of
transactions gather behind a trx doing a physical disk write
to log files, and when that physical write has been completed,
one of those transactions does a write which prepares the whole
group. Note that this group prepare will only bring benefit if
there are > 2 users in the database. Then at least 2 users can
gather behind one doing the physical log write to disk.
TODO: find out if MySQL holds some mutex when calling this.
That would spoil our group prepare algorithm. */
mutex_exit(&kernel_mutex);
if (srv_flush_log_at_trx_commit == 0) {
/* Do nothing */
} else if (srv_flush_log_at_trx_commit == 1) {
if (srv_unix_file_flush_method == SRV_UNIX_NOSYNC) {
/* Write the log but do not flush it to disk */
log_write_up_to(lsn, LOG_WAIT_ONE_GROUP,
FALSE);
} else {
/* Write the log to the log files AND flush
them to disk */
log_write_up_to(lsn, LOG_WAIT_ONE_GROUP, TRUE);
}
} else if (srv_flush_log_at_trx_commit == 2) {
/* Write the log but do not flush it to disk */
log_write_up_to(lsn, LOG_WAIT_ONE_GROUP, FALSE);
} else {
ut_error;
}
mutex_enter(&kernel_mutex);
}
}
/**************************************************************************
Does the transaction prepare for MySQL. */
ulint
trx_prepare_for_mysql(
/*====-=============*/
/* out: 0 or error number */
trx_t* trx) /* in: trx handle */
{
/* Because we do not do the prepare by sending an Innobase
sig to the transaction, we must here make sure that trx has been
started. */
ut_a(trx);
trx->op_info = "preparing";
trx_start_if_not_started(trx);
mutex_enter(&kernel_mutex);
trx_prepare_off_kernel(trx);
mutex_exit(&kernel_mutex);
trx->op_info = "";
return(0);
}
/**************************************************************************
This function is used to find number of prepared transactions and
their transaction objects for a recovery. */
int
trx_recover_for_mysql(
/*==================*/
/* out: number of prepared transactions
stored in xid_list */
XID* xid_list, /* in/out: prepared transactions */
ulint len) /* in: number of slots in xid_list */
{
trx_t* trx;
ulint count = 0;
ut_ad(xid_list);
ut_ad(len);
/* We should set those transactions which are in the prepared state
to the xid_list */
mutex_enter(&kernel_mutex);
trx = UT_LIST_GET_FIRST(trx_sys->trx_list);
while (trx) {
if (trx->conc_state == TRX_PREPARED) {
xid_list[count] = trx->xid;
if (count == 0) {
ut_print_timestamp(stderr);
fprintf(stderr,
" InnoDB: Starting recovery for XA transactions...\n");
}
ut_print_timestamp(stderr);
fprintf(stderr,
" InnoDB: Transaction %lu %lu in prepared state after recovery\n",
(ulong) ut_dulint_get_high(trx->id),
(ulong) ut_dulint_get_low(trx->id));
ut_print_timestamp(stderr);
fprintf(stderr,
" InnoDB: Transaction contains changes to %lu rows\n",
(ulong)ut_conv_dulint_to_longlong(trx->undo_no));
count++;
if (count == len) {
break;
}
}
trx = UT_LIST_GET_NEXT(trx_list, trx);
}
mutex_exit(&kernel_mutex);
if (count > 0){
ut_print_timestamp(stderr);
fprintf(stderr,
" InnoDB: %lu transactions in prepared state after recovery\n",
(ulong) count);
}
return (count);
}
/***********************************************************************
This function is used to find one X/Open XA distributed transaction
which is in the prepared state */
trx_t*
trx_get_trx_by_xid(
/*===============*/
/* out: trx or NULL */
XID* xid) /* in: X/Open XA transaction identification */
{
trx_t* trx;
if (xid == NULL) {
return (NULL);
}
mutex_enter(&kernel_mutex);
trx = UT_LIST_GET_FIRST(trx_sys->trx_list);
while (trx) {
/* Compare two X/Open XA transaction id's: their
length should be the same and binary comparison
of gtrid_lenght+bqual_length bytes should be
the same */
if (xid->gtrid_length == trx->xid.gtrid_length &&
xid->bqual_length == trx->xid.bqual_length &&
memcmp(xid->data, trx->xid.data,
xid->gtrid_length +
xid->bqual_length) == 0) {
break;
}
trx = UT_LIST_GET_NEXT(trx_list, trx);
}
mutex_exit(&kernel_mutex);
if (trx) {
if (trx->conc_state != TRX_PREPARED) {
return(NULL);
}
return(trx);
} else {
return(NULL);
}
}
|