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
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
3324
3325
3326
3327
3328
3329
3330
3331
3332
3333
3334
3335
3336
3337
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
3376
3377
3378
3379
3380
3381
3382
3383
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
3410
3411
3412
3413
3414
3415
3416
3417
3418
3419
3420
3421
3422
3423
3424
3425
3426
3427
3428
3429
3430
3431
3432
3433
3434
3435
3436
3437
3438
3439
3440
3441
3442
3443
3444
3445
3446
3447
3448
3449
3450
3451
3452
3453
3454
3455
3456
3457
3458
3459
3460
3461
3462
3463
3464
3465
3466
3467
3468
3469
3470
3471
3472
3473
3474
3475
3476
3477
3478
3479
3480
3481
3482
3483
3484
3485
3486
3487
3488
3489
3490
3491
3492
3493
3494
3495
3496
3497
3498
3499
3500
3501
3502
3503
3504
3505
3506
3507
3508
3509
3510
3511
3512
3513
3514
3515
3516
3517
3518
3519
3520
3521
3522
3523
3524
3525
3526
3527
3528
3529
3530
3531
3532
3533
3534
3535
3536
3537
3538
3539
3540
3541
3542
3543
3544
3545
3546
3547
3548
3549
3550
3551
3552
3553
3554
3555
3556
3557
3558
3559
3560
3561
3562
3563
3564
3565
3566
3567
3568
3569
3570
3571
3572
3573
3574
3575
3576
3577
3578
3579
3580
3581
3582
3583
3584
3585
3586
3587
3588
3589
3590
3591
3592
3593
3594
3595
3596
3597
3598
3599
3600
3601
3602
3603
3604
3605
3606
3607
3608
3609
3610
3611
3612
3613
3614
3615
3616
3617
3618
3619
3620
3621
3622
3623
3624
3625
3626
3627
3628
3629
3630
3631
3632
3633
3634
3635
3636
3637
3638
3639
3640
3641
3642
3643
3644
3645
3646
3647
3648
3649
3650
3651
3652
3653
3654
3655
3656
3657
3658
3659
3660
3661
3662
3663
3664
3665
3666
3667
3668
3669
3670
3671
3672
3673
3674
3675
3676
3677
3678
3679
3680
3681
3682
3683
3684
3685
3686
3687
3688
3689
3690
3691
3692
3693
3694
3695
3696
3697
3698
3699
3700
3701
3702
3703
3704
3705
3706
3707
3708
3709
3710
3711
3712
3713
3714
3715
3716
3717
3718
3719
3720
3721
3722
3723
3724
3725
3726
3727
3728
3729
3730
3731
3732
3733
3734
3735
3736
3737
3738
3739
3740
3741
3742
3743
3744
3745
3746
3747
3748
3749
3750
3751
3752
3753
3754
3755
3756
3757
3758
3759
3760
3761
3762
3763
3764
3765
3766
3767
3768
3769
3770
3771
3772
3773
3774
3775
3776
3777
3778
3779
3780
3781
3782
3783
3784
3785
3786
3787
3788
3789
3790
3791
3792
3793
3794
3795
3796
3797
3798
3799
3800
3801
3802
3803
3804
3805
3806
3807
3808
3809
3810
3811
3812
3813
3814
3815
3816
3817
3818
3819
3820
3821
3822
3823
3824
3825
3826
3827
3828
3829
3830
3831
3832
3833
3834
3835
3836
3837
3838
3839
3840
3841
3842
3843
3844
3845
3846
3847
3848
3849
3850
3851
3852
3853
3854
3855
3856
3857
3858
3859
3860
3861
3862
3863
3864
3865
3866
3867
3868
3869
3870
3871
3872
3873
3874
3875
3876
3877
3878
3879
3880
3881
3882
3883
3884
3885
3886
3887
3888
3889
3890
3891
3892
3893
3894
3895
3896
3897
3898
3899
3900
3901
3902
3903
3904
3905
3906
3907
3908
3909
3910
3911
3912
3913
3914
3915
3916
3917
3918
3919
3920
3921
3922
3923
3924
3925
3926
3927
3928
3929
3930
3931
3932
3933
3934
3935
3936
3937
3938
3939
3940
3941
3942
3943
3944
3945
3946
3947
3948
3949
3950
3951
3952
3953
3954
3955
3956
3957
3958
3959
3960
3961
3962
3963
3964
3965
3966
3967
3968
3969
3970
3971
3972
3973
3974
3975
3976
3977
3978
3979
3980
3981
3982
3983
3984
3985
3986
3987
3988
3989
3990
3991
3992
3993
3994
3995
3996
3997
3998
3999
4000
4001
4002
4003
4004
4005
4006
4007
4008
4009
4010
4011
4012
4013
4014
4015
4016
4017
4018
4019
4020
4021
4022
4023
4024
4025
4026
4027
4028
4029
4030
4031
4032
4033
4034
4035
4036
4037
4038
4039
4040
4041
4042
4043
4044
4045
4046
4047
4048
4049
4050
4051
4052
4053
4054
4055
4056
4057
4058
4059
4060
4061
4062
4063
4064
4065
4066
4067
4068
4069
4070
4071
4072
4073
4074
4075
4076
4077
4078
4079
4080
4081
4082
4083
4084
4085
4086
4087
4088
4089
4090
4091
4092
4093
4094
4095
4096
4097
4098
4099
4100
4101
4102
4103
4104
4105
4106
4107
4108
4109
4110
4111
4112
4113
4114
4115
4116
4117
4118
4119
4120
4121
4122
4123
4124
4125
4126
4127
4128
4129
4130
4131
4132
4133
4134
4135
4136
4137
4138
4139
4140
4141
4142
4143
4144
4145
4146
4147
4148
4149
4150
4151
4152
4153
4154
4155
4156
4157
4158
4159
4160
4161
4162
4163
4164
4165
4166
4167
4168
4169
4170
4171
4172
4173
4174
4175
4176
4177
4178
4179
4180
4181
4182
4183
4184
4185
4186
4187
4188
4189
4190
4191
4192
4193
4194
4195
4196
4197
4198
4199
4200
4201
4202
4203
4204
4205
4206
4207
4208
4209
4210
4211
4212
4213
4214
4215
4216
4217
4218
4219
4220
4221
4222
4223
4224
4225
4226
4227
4228
4229
4230
4231
4232
4233
4234
4235
4236
4237
4238
4239
4240
4241
4242
4243
4244
4245
4246
4247
4248
4249
4250
4251
4252
4253
4254
4255
4256
4257
4258
4259
4260
4261
4262
4263
4264
4265
4266
4267
4268
4269
4270
4271
4272
4273
4274
4275
4276
4277
4278
4279
4280
4281
4282
4283
4284
4285
4286
4287
4288
4289
4290
4291
4292
4293
4294
4295
4296
4297
4298
4299
4300
4301
4302
4303
4304
4305
4306
4307
4308
4309
4310
4311
4312
4313
4314
4315
4316
4317
4318
4319
4320
4321
4322
4323
4324
4325
4326
4327
4328
4329
4330
4331
4332
4333
4334
4335
4336
4337
4338
4339
4340
4341
4342
4343
4344
4345
4346
4347
4348
4349
4350
4351
4352
4353
4354
4355
4356
4357
4358
4359
4360
4361
4362
4363
4364
4365
4366
4367
4368
4369
4370
4371
4372
4373
4374
4375
4376
4377
4378
4379
4380
4381
4382
4383
4384
4385
4386
4387
4388
4389
4390
4391
4392
4393
4394
4395
4396
4397
4398
4399
4400
4401
4402
4403
4404
4405
4406
4407
4408
4409
4410
4411
4412
4413
4414
4415
4416
4417
4418
4419
4420
4421
4422
4423
4424
4425
4426
4427
4428
4429
4430
4431
4432
4433
4434
4435
4436
4437
4438
4439
4440
4441
4442
4443
4444
4445
4446
4447
4448
4449
4450
4451
4452
4453
4454
4455
4456
4457
4458
4459
4460
4461
4462
4463
4464
4465
4466
4467
4468
4469
4470
4471
4472
4473
4474
4475
4476
4477
4478
4479
4480
4481
4482
4483
4484
4485
4486
4487
4488
4489
4490
4491
4492
4493
4494
4495
4496
4497
4498
4499
4500
4501
4502
4503
4504
4505
4506
4507
4508
4509
4510
4511
4512
4513
4514
4515
4516
4517
4518
4519
4520
4521
4522
4523
4524
4525
4526
4527
4528
4529
4530
4531
4532
4533
4534
4535
4536
4537
4538
4539
4540
4541
4542
4543
4544
4545
4546
4547
4548
4549
4550
4551
4552
4553
4554
4555
4556
4557
4558
4559
4560
4561
4562
4563
4564
4565
4566
4567
4568
4569
4570
4571
4572
4573
4574
4575
4576
4577
4578
4579
4580
4581
4582
4583
4584
4585
4586
4587
4588
4589
4590
4591
4592
4593
4594
4595
4596
4597
4598
4599
4600
4601
4602
4603
4604
4605
4606
4607
4608
4609
4610
4611
4612
4613
4614
4615
4616
4617
4618
4619
4620
4621
4622
4623
4624
4625
4626
4627
4628
4629
4630
4631
4632
4633
4634
4635
4636
4637
4638
4639
4640
4641
4642
4643
4644
4645
4646
4647
4648
4649
4650
4651
4652
4653
4654
4655
4656
4657
4658
4659
4660
4661
4662
4663
4664
4665
4666
4667
4668
4669
4670
4671
4672
4673
4674
4675
4676
4677
4678
4679
4680
4681
4682
4683
4684
4685
4686
4687
4688
4689
4690
4691
4692
4693
4694
4695
4696
4697
4698
4699
4700
4701
4702
4703
4704
4705
4706
4707
4708
4709
4710
4711
4712
4713
4714
4715
4716
4717
4718
4719
4720
4721
4722
4723
4724
4725
4726
4727
4728
4729
4730
4731
4732
4733
4734
4735
4736
4737
4738
4739
4740
4741
4742
4743
4744
4745
4746
4747
4748
4749
4750
4751
4752
4753
4754
4755
4756
4757
4758
4759
4760
4761
4762
4763
4764
4765
4766
4767
4768
4769
4770
4771
4772
4773
4774
4775
4776
4777
4778
4779
4780
4781
4782
4783
4784
4785
4786
4787
4788
4789
4790
4791
4792
4793
4794
4795
4796
4797
4798
4799
4800
4801
4802
4803
4804
4805
4806
4807
4808
4809
4810
4811
4812
4813
4814
4815
4816
4817
4818
4819
4820
4821
4822
4823
4824
4825
4826
4827
4828
4829
4830
4831
4832
4833
4834
4835
4836
4837
4838
4839
4840
4841
4842
4843
4844
4845
4846
4847
4848
4849
4850
4851
4852
4853
4854
4855
4856
4857
4858
4859
4860
4861
4862
4863
4864
4865
4866
4867
4868
4869
4870
4871
4872
4873
4874
4875
4876
4877
4878
4879
4880
4881
4882
4883
4884
4885
4886
4887
4888
4889
4890
4891
4892
4893
4894
4895
4896
4897
4898
4899
4900
4901
4902
4903
4904
4905
4906
4907
4908
4909
4910
4911
4912
4913
4914
4915
4916
4917
4918
4919
4920
4921
4922
4923
4924
4925
4926
4927
4928
4929
4930
4931
4932
4933
4934
4935
4936
4937
4938
4939
4940
4941
4942
4943
4944
4945
4946
4947
4948
4949
4950
4951
4952
4953
4954
4955
4956
4957
4958
4959
4960
4961
4962
4963
4964
4965
4966
4967
4968
4969
4970
4971
4972
4973
4974
4975
4976
4977
4978
4979
4980
4981
4982
4983
4984
4985
4986
4987
4988
4989
4990
4991
4992
4993
4994
4995
4996
4997
4998
4999
5000
5001
5002
5003
5004
5005
5006
5007
5008
5009
5010
5011
5012
5013
5014
5015
5016
5017
5018
5019
5020
5021
5022
5023
5024
5025
5026
5027
5028
5029
5030
5031
5032
5033
5034
5035
5036
5037
5038
5039
5040
5041
5042
5043
5044
5045
5046
5047
5048
5049
5050
5051
5052
5053
5054
5055
5056
5057
5058
5059
5060
5061
5062
5063
5064
5065
5066
5067
5068
5069
5070
5071
5072
5073
5074
5075
5076
5077
5078
5079
5080
5081
5082
5083
5084
5085
5086
5087
5088
5089
5090
5091
5092
5093
5094
5095
5096
5097
5098
5099
5100
5101
5102
5103
5104
5105
5106
5107
5108
5109
5110
5111
5112
5113
5114
5115
5116
5117
5118
5119
5120
5121
5122
5123
5124
5125
5126
5127
5128
5129
5130
5131
5132
5133
5134
5135
5136
5137
5138
5139
5140
5141
5142
5143
5144
5145
5146
5147
5148
5149
5150
5151
5152
5153
5154
5155
5156
5157
5158
5159
5160
5161
5162
5163
5164
5165
5166
5167
5168
5169
5170
5171
5172
5173
5174
5175
5176
5177
5178
5179
5180
5181
5182
5183
5184
5185
5186
5187
5188
5189
5190
5191
5192
5193
5194
5195
5196
5197
5198
5199
5200
5201
5202
5203
5204
5205
5206
5207
5208
5209
5210
5211
5212
5213
5214
5215
5216
5217
5218
5219
5220
5221
5222
5223
5224
5225
5226
5227
5228
5229
5230
5231
5232
5233
5234
5235
5236
5237
5238
5239
5240
5241
5242
5243
5244
5245
5246
5247
5248
5249
5250
5251
5252
5253
5254
5255
5256
5257
5258
5259
5260
5261
5262
5263
5264
5265
5266
5267
5268
5269
5270
5271
5272
5273
5274
5275
5276
5277
5278
5279
5280
5281
5282
5283
5284
5285
5286
5287
5288
5289
5290
5291
5292
5293
5294
5295
5296
5297
5298
5299
5300
5301
5302
5303
5304
5305
5306
5307
5308
5309
5310
5311
5312
5313
5314
5315
5316
5317
5318
5319
5320
5321
5322
5323
5324
5325
5326
5327
5328
5329
5330
5331
5332
5333
5334
5335
5336
5337
5338
5339
5340
5341
5342
5343
5344
5345
5346
5347
5348
5349
5350
5351
5352
5353
5354
5355
5356
5357
5358
5359
5360
5361
5362
5363
5364
5365
5366
5367
5368
5369
5370
5371
5372
5373
5374
5375
5376
5377
5378
5379
5380
5381
5382
5383
5384
5385
5386
5387
5388
5389
5390
5391
5392
5393
5394
5395
5396
5397
5398
5399
5400
5401
5402
5403
5404
5405
5406
5407
5408
5409
5410
5411
5412
5413
5414
5415
5416
5417
5418
5419
5420
5421
5422
5423
5424
5425
5426
5427
5428
5429
5430
5431
5432
5433
5434
5435
5436
5437
5438
5439
5440
5441
5442
5443
5444
5445
5446
5447
5448
5449
5450
5451
5452
5453
5454
5455
5456
5457
5458
5459
5460
5461
5462
5463
5464
5465
5466
5467
5468
5469
5470
5471
5472
5473
5474
5475
5476
5477
5478
5479
5480
5481
5482
5483
5484
5485
5486
5487
5488
5489
5490
5491
5492
5493
5494
5495
5496
5497
5498
5499
5500
5501
5502
5503
5504
5505
5506
5507
5508
5509
5510
5511
5512
5513
5514
5515
5516
5517
5518
5519
5520
5521
5522
5523
5524
5525
5526
5527
5528
5529
5530
5531
5532
5533
5534
5535
5536
5537
5538
5539
5540
5541
5542
5543
5544
5545
5546
5547
5548
5549
5550
5551
5552
5553
5554
5555
5556
5557
5558
5559
5560
5561
5562
5563
5564
5565
5566
5567
5568
5569
5570
5571
5572
5573
5574
5575
5576
5577
5578
5579
5580
5581
5582
5583
5584
5585
5586
5587
5588
5589
5590
5591
5592
5593
5594
5595
5596
5597
5598
5599
5600
5601
5602
5603
5604
5605
5606
5607
5608
5609
5610
5611
5612
5613
5614
5615
5616
5617
5618
5619
5620
5621
5622
5623
5624
5625
5626
5627
5628
5629
5630
5631
5632
5633
5634
5635
5636
5637
5638
5639
5640
5641
5642
5643
5644
5645
5646
5647
5648
5649
5650
5651
5652
5653
5654
5655
5656
5657
5658
5659
5660
5661
5662
5663
5664
5665
5666
5667
5668
5669
5670
5671
5672
5673
5674
5675
5676
5677
5678
5679
5680
5681
5682
5683
5684
5685
5686
5687
5688
5689
5690
5691
5692
5693
5694
5695
5696
5697
5698
5699
5700
5701
5702
5703
5704
5705
5706
5707
5708
5709
5710
5711
5712
5713
5714
5715
5716
5717
5718
5719
5720
5721
5722
5723
5724
5725
5726
5727
5728
5729
5730
5731
5732
5733
5734
5735
5736
5737
5738
5739
5740
5741
5742
5743
5744
5745
5746
5747
5748
5749
5750
5751
5752
5753
5754
5755
5756
5757
5758
5759
5760
5761
5762
5763
5764
5765
5766
5767
5768
5769
5770
5771
5772
5773
5774
5775
5776
5777
5778
5779
5780
5781
5782
5783
5784
5785
5786
5787
5788
5789
5790
5791
5792
5793
5794
5795
5796
5797
5798
5799
5800
5801
5802
5803
5804
5805
5806
5807
5808
5809
5810
5811
5812
5813
5814
5815
5816
5817
5818
5819
5820
5821
5822
5823
5824
5825
5826
5827
5828
5829
5830
5831
5832
5833
5834
5835
5836
5837
5838
5839
5840
5841
5842
5843
5844
5845
5846
5847
5848
5849
5850
5851
5852
5853
5854
5855
5856
5857
5858
5859
5860
5861
5862
5863
5864
5865
5866
5867
5868
5869
5870
5871
5872
5873
5874
5875
5876
5877
5878
5879
5880
5881
5882
5883
5884
5885
5886
5887
5888
5889
5890
5891
5892
5893
5894
5895
5896
5897
5898
5899
5900
5901
5902
5903
5904
5905
5906
5907
5908
5909
5910
5911
5912
5913
5914
5915
5916
5917
5918
5919
5920
5921
5922
5923
5924
5925
5926
5927
5928
5929
5930
5931
5932
5933
5934
5935
5936
5937
5938
5939
5940
5941
5942
5943
5944
5945
5946
5947
5948
5949
5950
5951
5952
5953
5954
5955
5956
5957
5958
5959
5960
5961
5962
5963
5964
5965
5966
5967
5968
5969
5970
5971
5972
5973
5974
5975
5976
5977
5978
5979
5980
5981
5982
5983
5984
5985
5986
5987
5988
5989
5990
5991
5992
5993
5994
5995
5996
5997
5998
5999
6000
6001
6002
6003
6004
6005
6006
6007
6008
6009
6010
6011
6012
6013
6014
6015
6016
6017
6018
6019
6020
6021
6022
6023
6024
6025
6026
6027
6028
6029
6030
6031
6032
6033
6034
6035
6036
6037
6038
6039
6040
6041
6042
6043
6044
6045
6046
6047
6048
6049
6050
6051
6052
6053
6054
6055
6056
6057
6058
6059
6060
6061
6062
6063
6064
6065
6066
6067
6068
6069
6070
6071
6072
6073
6074
6075
6076
6077
6078
6079
6080
6081
6082
6083
6084
6085
6086
6087
6088
6089
6090
6091
6092
6093
6094
6095
6096
6097
6098
6099
6100
6101
6102
6103
6104
6105
6106
6107
6108
6109
6110
6111
6112
6113
6114
6115
6116
6117
6118
6119
6120
6121
6122
6123
6124
6125
6126
6127
6128
6129
6130
6131
6132
6133
6134
6135
6136
6137
6138
6139
6140
6141
6142
6143
6144
6145
6146
6147
6148
6149
6150
6151
6152
6153
6154
6155
6156
6157
6158
6159
6160
6161
6162
6163
6164
6165
6166
6167
6168
6169
6170
6171
6172
6173
6174
6175
6176
6177
6178
6179
6180
6181
6182
6183
6184
6185
6186
6187
6188
6189
6190
6191
6192
6193
6194
6195
6196
6197
6198
6199
6200
6201
6202
6203
6204
6205
6206
6207
6208
6209
6210
6211
6212
6213
6214
6215
6216
6217
6218
6219
6220
6221
6222
6223
6224
6225
6226
6227
6228
6229
6230
6231
6232
6233
6234
6235
6236
6237
6238
6239
6240
6241
6242
6243
6244
6245
6246
6247
6248
6249
6250
6251
6252
6253
6254
6255
6256
6257
6258
6259
6260
6261
6262
6263
6264
6265
6266
6267
6268
6269
6270
6271
6272
6273
6274
6275
6276
6277
6278
6279
6280
6281
6282
6283
6284
6285
6286
6287
6288
6289
6290
6291
6292
6293
6294
6295
6296
6297
6298
6299
6300
6301
6302
6303
6304
6305
6306
6307
6308
6309
6310
6311
6312
6313
6314
6315
6316
6317
6318
6319
6320
6321
6322
6323
6324
6325
6326
6327
6328
6329
6330
6331
6332
6333
6334
6335
6336
6337
6338
6339
6340
6341
6342
6343
6344
6345
6346
6347
6348
6349
6350
6351
6352
6353
6354
6355
6356
6357
6358
6359
6360
6361
6362
6363
6364
6365
6366
6367
6368
6369
6370
6371
6372
6373
6374
6375
6376
6377
6378
6379
6380
6381
6382
6383
6384
6385
6386
6387
6388
6389
6390
6391
6392
6393
6394
6395
6396
6397
6398
6399
6400
6401
6402
6403
6404
6405
6406
6407
6408
6409
6410
6411
6412
6413
6414
6415
6416
6417
6418
6419
6420
6421
6422
6423
6424
6425
6426
6427
6428
6429
6430
6431
6432
6433
6434
6435
6436
6437
6438
6439
6440
6441
6442
6443
6444
6445
6446
6447
6448
6449
6450
6451
6452
6453
6454
6455
6456
6457
6458
6459
6460
6461
6462
6463
6464
6465
6466
6467
6468
6469
6470
6471
6472
6473
6474
6475
6476
6477
6478
6479
6480
6481
6482
6483
6484
6485
6486
6487
6488
6489
6490
6491
6492
6493
6494
6495
6496
6497
6498
6499
6500
6501
6502
6503
6504
6505
6506
6507
6508
6509
6510
6511
6512
6513
6514
6515
6516
6517
6518
6519
6520
6521
6522
6523
6524
6525
6526
6527
6528
6529
6530
6531
6532
6533
6534
6535
6536
6537
6538
6539
6540
6541
6542
6543
6544
6545
6546
6547
6548
6549
6550
6551
6552
6553
6554
6555
6556
6557
6558
6559
6560
6561
6562
6563
6564
6565
6566
6567
6568
6569
6570
6571
6572
6573
6574
6575
6576
6577
6578
6579
6580
6581
6582
6583
6584
6585
6586
6587
6588
6589
6590
6591
6592
6593
6594
6595
6596
6597
6598
6599
6600
6601
6602
6603
6604
6605
6606
6607
6608
6609
6610
6611
6612
6613
6614
6615
6616
6617
6618
6619
6620
6621
6622
6623
6624
6625
6626
6627
6628
6629
6630
6631
6632
6633
6634
6635
6636
6637
6638
6639
6640
6641
6642
6643
6644
6645
6646
6647
6648
6649
6650
6651
6652
6653
6654
6655
6656
6657
6658
6659
6660
6661
6662
6663
6664
6665
6666
6667
6668
6669
6670
6671
6672
6673
6674
6675
6676
6677
6678
6679
6680
6681
6682
6683
6684
6685
6686
6687
6688
6689
6690
6691
6692
6693
6694
6695
6696
6697
6698
6699
6700
6701
6702
6703
6704
6705
6706
6707
6708
6709
6710
6711
6712
6713
6714
6715
6716
6717
6718
6719
6720
6721
6722
6723
6724
6725
6726
6727
6728
6729
6730
6731
6732
6733
6734
6735
6736
6737
6738
6739
6740
6741
6742
6743
6744
6745
6746
6747
6748
6749
6750
6751
6752
6753
6754
6755
6756
6757
6758
6759
6760
6761
6762
6763
6764
6765
6766
6767
6768
6769
6770
6771
6772
6773
6774
6775
6776
6777
6778
6779
6780
6781
6782
6783
6784
6785
6786
6787
6788
6789
6790
6791
6792
6793
6794
6795
6796
6797
6798
6799
6800
6801
6802
6803
6804
6805
6806
6807
6808
6809
6810
6811
6812
6813
6814
6815
6816
6817
6818
6819
6820
6821
6822
6823
6824
6825
6826
6827
6828
6829
6830
6831
6832
6833
6834
6835
6836
6837
6838
6839
6840
6841
6842
6843
6844
6845
6846
6847
6848
6849
6850
6851
6852
6853
6854
6855
6856
6857
6858
6859
6860
6861
6862
6863
6864
6865
6866
6867
6868
6869
6870
6871
6872
6873
6874
6875
6876
6877
6878
6879
6880
6881
6882
6883
6884
6885
6886
6887
6888
6889
6890
6891
6892
6893
6894
6895
6896
6897
6898
6899
6900
6901
6902
6903
6904
6905
6906
6907
6908
6909
6910
6911
6912
6913
6914
6915
6916
6917
6918
6919
6920
6921
6922
6923
6924
6925
6926
6927
6928
6929
6930
6931
6932
6933
6934
6935
6936
6937
6938
6939
6940
6941
6942
6943
6944
6945
6946
6947
6948
6949
6950
6951
6952
6953
6954
6955
6956
6957
6958
6959
6960
6961
6962
6963
6964
6965
6966
6967
6968
6969
6970
6971
6972
6973
6974
6975
6976
6977
6978
6979
6980
6981
6982
6983
6984
6985
6986
6987
6988
6989
6990
6991
6992
6993
6994
6995
6996
6997
6998
6999
7000
7001
7002
7003
7004
7005
7006
7007
7008
7009
7010
7011
7012
7013
7014
7015
7016
7017
7018
7019
7020
7021
7022
7023
7024
7025
7026
7027
7028
7029
7030
7031
7032
7033
7034
7035
7036
7037
7038
7039
7040
7041
7042
7043
7044
7045
7046
7047
7048
7049
7050
7051
7052
7053
7054
7055
7056
7057
7058
7059
7060
7061
7062
7063
7064
7065
7066
7067
7068
7069
7070
7071
7072
7073
7074
7075
7076
7077
7078
7079
7080
7081
7082
7083
7084
7085
7086
7087
7088
7089
7090
7091
7092
7093
7094
7095
7096
7097
7098
7099
7100
7101
7102
7103
7104
7105
7106
7107
7108
7109
7110
7111
7112
7113
7114
7115
7116
7117
7118
7119
7120
7121
7122
7123
7124
7125
7126
7127
7128
7129
7130
7131
7132
7133
7134
7135
7136
7137
7138
7139
7140
7141
7142
7143
7144
7145
7146
7147
7148
7149
7150
7151
7152
7153
7154
7155
7156
7157
7158
7159
7160
7161
7162
7163
7164
7165
7166
7167
7168
7169
7170
7171
7172
7173
7174
7175
7176
7177
7178
7179
7180
7181
7182
7183
7184
7185
7186
7187
7188
7189
7190
7191
7192
7193
7194
7195
7196
7197
7198
7199
7200
7201
7202
7203
7204
7205
7206
7207
7208
7209
7210
7211
7212
7213
7214
7215
7216
7217
7218
7219
7220
7221
7222
7223
7224
7225
7226
7227
7228
7229
7230
7231
7232
7233
7234
7235
7236
7237
7238
7239
7240
7241
7242
7243
7244
7245
7246
7247
7248
7249
7250
7251
7252
7253
7254
7255
7256
7257
7258
7259
7260
7261
7262
7263
7264
7265
7266
7267
7268
7269
7270
7271
7272
7273
7274
7275
7276
7277
7278
7279
7280
7281
7282
7283
7284
7285
7286
7287
7288
7289
7290
7291
7292
7293
7294
7295
7296
7297
7298
7299
7300
7301
7302
7303
7304
7305
7306
7307
7308
7309
7310
7311
7312
7313
7314
7315
7316
7317
7318
7319
7320
7321
7322
7323
7324
7325
7326
7327
7328
7329
7330
7331
7332
7333
7334
7335
7336
7337
7338
7339
7340
7341
7342
7343
7344
7345
7346
7347
7348
7349
7350
7351
7352
7353
7354
7355
7356
7357
7358
7359
7360
7361
7362
7363
7364
7365
7366
7367
7368
7369
7370
7371
7372
7373
7374
7375
7376
7377
7378
7379
7380
7381
7382
7383
7384
7385
7386
7387
7388
7389
7390
7391
7392
7393
7394
7395
7396
7397
7398
7399
7400
7401
7402
7403
7404
7405
7406
7407
7408
7409
7410
7411
7412
7413
7414
7415
7416
7417
7418
7419
7420
7421
7422
7423
7424
7425
7426
7427
7428
7429
7430
7431
7432
7433
7434
7435
7436
7437
7438
7439
7440
7441
7442
7443
7444
7445
7446
7447
7448
7449
7450
7451
7452
7453
7454
7455
7456
7457
7458
7459
7460
7461
7462
7463
7464
7465
7466
7467
7468
7469
7470
7471
7472
7473
7474
7475
7476
7477
|
/* Copyright (C) 2000-2005 MySQL AB & Innobase Oy
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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */
/* This file defines the InnoDB handler: the interface between MySQL and InnoDB
NOTE: You can only use noninlined InnoDB functions in this file, because we
have disabled the InnoDB inlining in this file. */
/* TODO list for the InnoDB handler in 5.0:
- Remove the flag trx->active_trans and look at the InnoDB
trx struct state field
- fix savepoint functions to use savepoint storage area
- Find out what kind of problems the OS X case-insensitivity causes to
table and database names; should we 'normalize' the names like we do
in Windows?
*/
#ifdef USE_PRAGMA_IMPLEMENTATION
#pragma implementation // gcc: Class implementation
#endif
#include "mysql_priv.h"
#include "slave.h"
#ifdef HAVE_INNOBASE_DB
#include <m_ctype.h>
#include <hash.h>
#include <myisampack.h>
#include <mysys_err.h>
#include <my_sys.h>
#define MAX_ULONG_BIT ((ulong) 1 << (sizeof(ulong)*8-1))
#include "ha_innodb.h"
pthread_mutex_t innobase_share_mutex, /* to protect innobase_open_files */
prepare_commit_mutex; /* to force correct commit order in
binlog */
ulong commit_threads= 0;
pthread_mutex_t commit_threads_m;
pthread_cond_t commit_cond;
pthread_mutex_t commit_cond_m;
bool innodb_inited= 0;
/*-----------------------------------------------------------------*/
/* These variables are used to implement (semi-)synchronous MySQL binlog
replication for InnoDB tables. */
pthread_cond_t innobase_repl_cond; /* Posix cond variable;
this variable is signaled
when enough binlog has been
sent to slave, so that a
waiting trx can return the
'ok' message to the client
for a commit */
pthread_mutex_t innobase_repl_cond_mutex; /* Posix cond variable mutex
that also protects the next
innobase_repl_... variables */
uint innobase_repl_state; /* 1 if synchronous replication
is switched on and is working
ok; else 0 */
uint innobase_repl_file_name_inited = 0; /* This is set to 1 when
innobase_repl_file_name
contains meaningful data */
char* innobase_repl_file_name; /* The binlog name up to which
we have sent some binlog to
the slave */
my_off_t innobase_repl_pos; /* The position in that file
up to which we have sent the
binlog to the slave */
uint innobase_repl_n_wait_threads = 0; /* This tells how many
transactions currently are
waiting for the binlog to be
sent to the client */
uint innobase_repl_wait_file_name_inited = 0; /* This is set to 1
when we know the 'smallest'
wait position */
char* innobase_repl_wait_file_name; /* NULL, or the 'smallest'
innobase_repl_file_name that
a transaction is waiting for */
my_off_t innobase_repl_wait_pos; /* The smallest position in
that file that a trx is
waiting for: the trx can
proceed and send an 'ok' to
the client when MySQL has sent
the binlog up to this position
to the slave */
/*-----------------------------------------------------------------*/
/* Store MySQL definition of 'byte': in Linux it is char while InnoDB
uses unsigned char; the header univ.i which we include next defines
'byte' as a macro which expands to 'unsigned char' */
typedef byte mysql_byte;
#define INSIDE_HA_INNOBASE_CC
/* Include necessary InnoDB headers */
extern "C" {
#include "../innobase/include/univ.i"
#include "../innobase/include/os0file.h"
#include "../innobase/include/os0thread.h"
#include "../innobase/include/srv0start.h"
#include "../innobase/include/srv0srv.h"
#include "../innobase/include/trx0roll.h"
#include "../innobase/include/trx0trx.h"
#include "../innobase/include/trx0sys.h"
#include "../innobase/include/mtr0mtr.h"
#include "../innobase/include/row0ins.h"
#include "../innobase/include/row0mysql.h"
#include "../innobase/include/row0sel.h"
#include "../innobase/include/row0upd.h"
#include "../innobase/include/log0log.h"
#include "../innobase/include/lock0lock.h"
#include "../innobase/include/dict0crea.h"
#include "../innobase/include/btr0cur.h"
#include "../innobase/include/btr0btr.h"
#include "../innobase/include/fsp0fsp.h"
#include "../innobase/include/sync0sync.h"
#include "../innobase/include/fil0fil.h"
#include "../innobase/include/trx0xa.h"
}
#define HA_INNOBASE_ROWS_IN_TABLE 10000 /* to get optimization right */
#define HA_INNOBASE_RANGE_COUNT 100
ulong innobase_large_page_size = 0;
/* The default values for the following, type long or longlong, start-up
parameters are declared in mysqld.cc: */
long innobase_mirrored_log_groups, innobase_log_files_in_group,
innobase_log_buffer_size, innobase_buffer_pool_awe_mem_mb,
innobase_additional_mem_pool_size, innobase_file_io_threads,
innobase_lock_wait_timeout, innobase_force_recovery,
innobase_open_files;
longlong innobase_buffer_pool_size, innobase_log_file_size;
/* The default values for the following char* start-up parameters
are determined in innobase_init below: */
char* innobase_data_home_dir = NULL;
char* innobase_data_file_path = NULL;
char* innobase_log_group_home_dir = NULL;
char* innobase_log_arch_dir = NULL;/* unused */
/* The following has a misleading name: starting from 4.0.5, this also
affects Windows: */
char* innobase_unix_file_flush_method = NULL;
/* Below we have boolean-valued start-up parameters, and their default
values */
ulong innobase_fast_shutdown = 1;
my_bool innobase_log_archive = FALSE;/* unused */
my_bool innobase_use_doublewrite = TRUE;
my_bool innobase_use_checksums = TRUE;
my_bool innobase_use_large_pages = FALSE;
my_bool innobase_use_native_aio = FALSE;
my_bool innobase_file_per_table = FALSE;
my_bool innobase_locks_unsafe_for_binlog = FALSE;
my_bool innobase_rollback_on_timeout = FALSE;
my_bool innobase_create_status_file = FALSE;
my_bool innobase_adaptive_hash_index = TRUE;
static char *internal_innobase_data_file_path = NULL;
/* The following counter is used to convey information to InnoDB
about server activity: in selects it is not sensible to call
srv_active_wake_master_thread after each fetch or search, we only do
it every INNOBASE_WAKE_INTERVAL'th step. */
#define INNOBASE_WAKE_INTERVAL 32
ulong innobase_active_counter = 0;
static HASH innobase_open_tables;
#ifdef __NETWARE__ /* some special cleanup for NetWare */
bool nw_panic = FALSE;
#endif
static mysql_byte* innobase_get_key(INNOBASE_SHARE *share,uint *length,
my_bool not_used __attribute__((unused)));
static INNOBASE_SHARE *get_share(const char *table_name);
static void free_share(INNOBASE_SHARE *share);
static int innobase_close_connection(THD* thd);
static int innobase_commit(THD* thd, bool all);
static int innobase_rollback(THD* thd, bool all);
static int innobase_rollback_to_savepoint(THD* thd, void *savepoint);
static int innobase_savepoint(THD* thd, void *savepoint);
static int innobase_release_savepoint(THD* thd, void *savepoint);
handlerton innobase_hton = {
"InnoDB",
SHOW_OPTION_YES,
"Supports transactions, row-level locking, and foreign keys",
DB_TYPE_INNODB,
innobase_init,
0, /* slot */
sizeof(trx_named_savept_t), /* savepoint size. TODO: use it */
innobase_close_connection,
innobase_savepoint,
innobase_rollback_to_savepoint,
innobase_release_savepoint,
innobase_commit, /* commit */
innobase_rollback, /* rollback */
innobase_xa_prepare, /* prepare */
innobase_xa_recover, /* recover */
innobase_commit_by_xid, /* commit_by_xid */
innobase_rollback_by_xid, /* rollback_by_xid */
innobase_create_cursor_view,
innobase_set_cursor_view,
innobase_close_cursor_view,
HTON_NO_FLAGS
};
/***********************************************************************
This function checks each index name for a table against reserved
system default primary index name 'GEN_CLUST_INDEX'. If a name matches,
this function pushes an error message to the client, and returns true. */
static
bool
innobase_index_name_is_reserved(
/*============================*/
/* out: true if index name matches a
reserved name */
const trx_t* trx, /* in: InnoDB transaction handle */
const TABLE* form, /* in: information on table
columns and indexes */
const char* norm_name); /* in: table name */
/* "GEN_CLUST_INDEX" is the name reserved for Innodb default
system primary index. */
static const char innobase_index_reserve_name[]= "GEN_CLUST_INDEX";
/*********************************************************************
Commits a transaction in an InnoDB database. */
void
innobase_commit_low(
/*================*/
trx_t* trx); /* in: transaction handle */
struct show_var_st innodb_status_variables[]= {
{"buffer_pool_pages_data",
(char*) &export_vars.innodb_buffer_pool_pages_data, SHOW_LONG},
{"buffer_pool_pages_dirty",
(char*) &export_vars.innodb_buffer_pool_pages_dirty, SHOW_LONG},
{"buffer_pool_pages_flushed",
(char*) &export_vars.innodb_buffer_pool_pages_flushed, SHOW_LONG},
{"buffer_pool_pages_free",
(char*) &export_vars.innodb_buffer_pool_pages_free, SHOW_LONG},
#ifdef UNIV_DEBUG
{"buffer_pool_pages_latched",
(char*) &export_vars.innodb_buffer_pool_pages_latched, SHOW_LONG},
#endif /* UNIV_DEBUG */
{"buffer_pool_pages_misc",
(char*) &export_vars.innodb_buffer_pool_pages_misc, SHOW_LONG},
{"buffer_pool_pages_total",
(char*) &export_vars.innodb_buffer_pool_pages_total, SHOW_LONG},
{"buffer_pool_read_ahead_rnd",
(char*) &export_vars.innodb_buffer_pool_read_ahead_rnd, SHOW_LONG},
{"buffer_pool_read_ahead_seq",
(char*) &export_vars.innodb_buffer_pool_read_ahead_seq, SHOW_LONG},
{"buffer_pool_read_requests",
(char*) &export_vars.innodb_buffer_pool_read_requests, SHOW_LONG},
{"buffer_pool_reads",
(char*) &export_vars.innodb_buffer_pool_reads, SHOW_LONG},
{"buffer_pool_wait_free",
(char*) &export_vars.innodb_buffer_pool_wait_free, SHOW_LONG},
{"buffer_pool_write_requests",
(char*) &export_vars.innodb_buffer_pool_write_requests, SHOW_LONG},
{"data_fsyncs",
(char*) &export_vars.innodb_data_fsyncs, SHOW_LONG},
{"data_pending_fsyncs",
(char*) &export_vars.innodb_data_pending_fsyncs, SHOW_LONG},
{"data_pending_reads",
(char*) &export_vars.innodb_data_pending_reads, SHOW_LONG},
{"data_pending_writes",
(char*) &export_vars.innodb_data_pending_writes, SHOW_LONG},
{"data_read",
(char*) &export_vars.innodb_data_read, SHOW_LONG},
{"data_reads",
(char*) &export_vars.innodb_data_reads, SHOW_LONG},
{"data_writes",
(char*) &export_vars.innodb_data_writes, SHOW_LONG},
{"data_written",
(char*) &export_vars.innodb_data_written, SHOW_LONG},
{"dblwr_pages_written",
(char*) &export_vars.innodb_dblwr_pages_written, SHOW_LONG},
{"dblwr_writes",
(char*) &export_vars.innodb_dblwr_writes, SHOW_LONG},
{"log_waits",
(char*) &export_vars.innodb_log_waits, SHOW_LONG},
{"log_write_requests",
(char*) &export_vars.innodb_log_write_requests, SHOW_LONG},
{"log_writes",
(char*) &export_vars.innodb_log_writes, SHOW_LONG},
{"os_log_fsyncs",
(char*) &export_vars.innodb_os_log_fsyncs, SHOW_LONG},
{"os_log_pending_fsyncs",
(char*) &export_vars.innodb_os_log_pending_fsyncs, SHOW_LONG},
{"os_log_pending_writes",
(char*) &export_vars.innodb_os_log_pending_writes, SHOW_LONG},
{"os_log_written",
(char*) &export_vars.innodb_os_log_written, SHOW_LONG},
{"page_size",
(char*) &export_vars.innodb_page_size, SHOW_LONG},
{"pages_created",
(char*) &export_vars.innodb_pages_created, SHOW_LONG},
{"pages_read",
(char*) &export_vars.innodb_pages_read, SHOW_LONG},
{"pages_written",
(char*) &export_vars.innodb_pages_written, SHOW_LONG},
{"row_lock_current_waits",
(char*) &export_vars.innodb_row_lock_current_waits, SHOW_LONG},
{"row_lock_time",
(char*) &export_vars.innodb_row_lock_time, SHOW_LONGLONG},
{"row_lock_time_avg",
(char*) &export_vars.innodb_row_lock_time_avg, SHOW_LONG},
{"row_lock_time_max",
(char*) &export_vars.innodb_row_lock_time_max, SHOW_LONG},
{"row_lock_waits",
(char*) &export_vars.innodb_row_lock_waits, SHOW_LONG},
{"rows_deleted",
(char*) &export_vars.innodb_rows_deleted, SHOW_LONG},
{"rows_inserted",
(char*) &export_vars.innodb_rows_inserted, SHOW_LONG},
{"rows_read",
(char*) &export_vars.innodb_rows_read, SHOW_LONG},
{"rows_updated",
(char*) &export_vars.innodb_rows_updated, SHOW_LONG},
{NullS, NullS, SHOW_LONG}};
/* General functions */
/**********************************************************************
Save some CPU by testing the value of srv_thread_concurrency in inline
functions. */
inline
void
innodb_srv_conc_enter_innodb(
/*=========================*/
trx_t* trx) /* in: transaction handle */
{
if (UNIV_LIKELY(!srv_thread_concurrency)) {
return;
}
srv_conc_enter_innodb(trx);
}
/**********************************************************************
Save some CPU by testing the value of srv_thread_concurrency in inline
functions. */
inline
void
innodb_srv_conc_exit_innodb(
/*========================*/
trx_t* trx) /* in: transaction handle */
{
if (UNIV_LIKELY(!srv_thread_concurrency)) {
return;
}
srv_conc_exit_innodb(trx);
}
/**********************************************************************
Releases possible search latch and InnoDB thread FIFO ticket. These should
be released at each SQL statement end, and also when mysqld passes the
control to the client. It does no harm to release these also in the middle
of an SQL statement. */
inline
void
innobase_release_stat_resources(
/*============================*/
trx_t* trx) /* in: transaction object */
{
if (trx->has_search_latch) {
trx_search_latch_release_if_reserved(trx);
}
if (trx->declared_to_be_inside_innodb) {
/* Release our possible ticket in the FIFO */
srv_conc_force_exit_innodb(trx);
}
}
/************************************************************************
Call this function when mysqld passes control to the client. That is to
avoid deadlocks on the adaptive hash S-latch possibly held by thd. For more
documentation, see handler.cc. */
void
innobase_release_temporary_latches(
/*===============================*/
THD *thd)
{
trx_t* trx;
if (!innodb_inited) {
return;
}
trx = (trx_t*) thd->ha_data[innobase_hton.slot];
if (trx) {
innobase_release_stat_resources(trx);
}
}
/************************************************************************
Increments innobase_active_counter and every INNOBASE_WAKE_INTERVALth
time calls srv_active_wake_master_thread. This function should be used
when a single database operation may introduce a small need for
server utility activity, like checkpointing. */
inline
void
innobase_active_small(void)
/*=======================*/
{
innobase_active_counter++;
if ((innobase_active_counter % INNOBASE_WAKE_INTERVAL) == 0) {
srv_active_wake_master_thread();
}
}
/************************************************************************
Converts an InnoDB error code to a MySQL error code and also tells to MySQL
about a possible transaction rollback inside InnoDB caused by a lock wait
timeout or a deadlock. */
static
int
convert_error_code_to_mysql(
/*========================*/
/* out: MySQL error code */
int error, /* in: InnoDB error code */
THD* thd) /* in: user thread handle or NULL */
{
if (error == DB_SUCCESS) {
return(0);
} else if (error == (int) DB_DUPLICATE_KEY) {
return(HA_ERR_FOUND_DUPP_KEY);
} else if (error == (int) DB_RECORD_NOT_FOUND) {
return(HA_ERR_NO_ACTIVE_RECORD);
} else if (error == (int) DB_ERROR) {
return(-1); /* unspecified error */
} else if (error == (int) DB_DEADLOCK) {
/* Since we rolled back the whole transaction, we must
tell it also to MySQL so that MySQL knows to empty the
cached binlog for this transaction */
mark_transaction_to_rollback(thd, TRUE);
return(HA_ERR_LOCK_DEADLOCK);
} else if (error == (int) DB_LOCK_WAIT_TIMEOUT) {
/* Starting from 5.0.13, we let MySQL just roll back the
latest SQL statement in a lock wait timeout. Previously, we
rolled back the whole transaction. */
mark_transaction_to_rollback(thd,
(bool)row_rollback_on_timeout);
return(HA_ERR_LOCK_WAIT_TIMEOUT);
} else if (error == (int) DB_NO_REFERENCED_ROW) {
return(HA_ERR_NO_REFERENCED_ROW);
} else if (error == (int) DB_ROW_IS_REFERENCED) {
return(HA_ERR_ROW_IS_REFERENCED);
} else if (error == (int) DB_CANNOT_ADD_CONSTRAINT) {
return(HA_ERR_CANNOT_ADD_FOREIGN);
} else if (error == (int) DB_CANNOT_DROP_CONSTRAINT) {
return(HA_ERR_ROW_IS_REFERENCED); /* TODO: This is a bit
misleading, a new MySQL error
code should be introduced */
} else if (error == (int) DB_COL_APPEARS_TWICE_IN_INDEX) {
return(HA_ERR_CRASHED);
} else if (error == (int) DB_OUT_OF_FILE_SPACE) {
return(HA_ERR_RECORD_FILE_FULL);
} else if (error == (int) DB_TABLE_IS_BEING_USED) {
return(HA_ERR_WRONG_COMMAND);
} else if (error == (int) DB_TABLE_NOT_FOUND) {
return(HA_ERR_NO_SUCH_TABLE);
} else if (error == (int) DB_TOO_BIG_RECORD) {
return(HA_ERR_TO_BIG_ROW);
} else if (error == (int) DB_CORRUPTION) {
return(HA_ERR_CRASHED);
} else if (error == (int) DB_NO_SAVEPOINT) {
return(HA_ERR_NO_SAVEPOINT);
} else if (error == (int) DB_LOCK_TABLE_FULL) {
/* Since we rolled back the whole transaction, we must
tell it also to MySQL so that MySQL knows to empty the
cached binlog for this transaction */
mark_transaction_to_rollback(thd, TRUE);
return(HA_ERR_LOCK_TABLE_FULL);
} else if (error == DB_TOO_MANY_CONCURRENT_TRXS) {
return(HA_ERR_TOO_MANY_CONCURRENT_TRXS);
} else if (error == DB_UNSUPPORTED) {
return(HA_ERR_UNSUPPORTED);
} else {
return(-1); // Unknown error
}
}
/*****************************************************************
Prints info of a THD object (== user session thread) to the given file.
NOTE that /mysql/innobase/trx/trx0trx.c must contain the prototype for
this function! */
extern "C"
void
innobase_mysql_print_thd(
/*=====================*/
FILE* f, /* in: output stream */
void* input_thd, /* in: pointer to a MySQL THD object */
uint max_query_len) /* in: max query length to print, or 0 to
use the default max length */
{
THD* thd;
const Security_context *sctx;
const char* s;
thd = (THD*) input_thd;
/* We probably want to have original user as part of debug output. */
sctx = &thd->main_security_ctx;
fprintf(f, "MySQL thread id %lu, query id %lu",
thd->thread_id, (ulong) thd->query_id);
if (sctx->host) {
putc(' ', f);
fputs(sctx->host, f);
}
if (sctx->ip) {
putc(' ', f);
fputs(sctx->ip, f);
}
if (sctx->user) {
putc(' ', f);
fputs(sctx->user, f);
}
if ((s = thd->proc_info)) {
putc(' ', f);
fputs(s, f);
}
/* We have to quarantine an access to thd->query and
thd->query_length with thd->LOCK_thd_data mutex. */
VOID(pthread_mutex_lock(&thd->LOCK_thd_data));
if ((s = thd->query)) {
/* 3100 is chosen because currently 3000 is the maximum
max_query_len we ever give this. */
char buf[3100];
uint len;
/* If buf is too small, we dynamically allocate storage
in this. */
char* dyn_str = NULL;
/* Points to buf or dyn_str. */
char* str = buf;
if (max_query_len == 0)
{
/* ADDITIONAL SAFETY: the default is to print at
most 300 chars to reduce the probability of a
seg fault if there is a race in
thd->query_length in MySQL; after May 14, 2004
probably no race any more, but better be
safe */
max_query_len = 300;
}
len = min(thd->query_length, max_query_len);
if (len > (sizeof(buf) - 1))
{
dyn_str = my_malloc(len + 1, MYF(0));
str = dyn_str;
}
/* Use strmake to reduce the timeframe for a race,
compared to fwrite() */
len = (uint) (strmake(str, s, len) - str);
putc('\n', f);
fwrite(str, 1, len, f);
if (dyn_str)
{
my_free(dyn_str, MYF(0));
}
}
VOID(pthread_mutex_unlock(&thd->LOCK_thd_data));
putc('\n', f);
}
/**********************************************************************
Get the variable length bounds of the given character set.
NOTE that the exact prototype of this function has to be in
/innobase/data/data0type.ic! */
extern "C"
void
innobase_get_cset_width(
/*====================*/
ulint cset, /* in: MySQL charset-collation code */
ulint* mbminlen, /* out: minimum length of a char (in bytes) */
ulint* mbmaxlen) /* out: maximum length of a char (in bytes) */
{
CHARSET_INFO* cs;
ut_ad(cset < 256);
ut_ad(mbminlen);
ut_ad(mbmaxlen);
cs = all_charsets[cset];
if (cs) {
*mbminlen = cs->mbminlen;
*mbmaxlen = cs->mbmaxlen;
} else {
ut_a(cset == 0);
*mbminlen = *mbmaxlen = 0;
}
}
/**********************************************************************
Compares NUL-terminated UTF-8 strings case insensitively.
NOTE that the exact prototype of this function has to be in
/innobase/dict/dict0dict.c! */
extern "C"
int
innobase_strcasecmp(
/*================*/
/* out: 0 if a=b, <0 if a<b, >1 if a>b */
const char* a, /* in: first string to compare */
const char* b) /* in: second string to compare */
{
return(my_strcasecmp(system_charset_info, a, b));
}
/**********************************************************************
Makes all characters in a NUL-terminated UTF-8 string lower case.
NOTE that the exact prototype of this function has to be in
/innobase/dict/dict0dict.c! */
extern "C"
void
innobase_casedn_str(
/*================*/
char* a) /* in/out: string to put in lower case */
{
my_casedn_str(system_charset_info, a);
}
/*************************************************************************
Creates a temporary file. */
extern "C"
int
innobase_mysql_tmpfile(void)
/*========================*/
/* out: temporary file descriptor, or < 0 on error */
{
char filename[FN_REFLEN];
int fd2 = -1;
File fd = create_temp_file(filename, mysql_tmpdir, "ib",
#ifdef __WIN__
O_BINARY | O_TRUNC | O_SEQUENTIAL |
O_TEMPORARY | O_SHORT_LIVED |
#endif /* __WIN__ */
O_CREAT | O_EXCL | O_RDWR,
MYF(MY_WME));
if (fd >= 0) {
#ifndef __WIN__
/* On Windows, open files cannot be removed, but files can be
created with the O_TEMPORARY flag to the same effect
("delete on close"). */
unlink(filename);
#endif /* !__WIN__ */
/* Copy the file descriptor, so that the additional resources
allocated by create_temp_file() can be freed by invoking
my_close().
Because the file descriptor returned by this function
will be passed to fdopen(), it will be closed by invoking
fclose(), which in turn will invoke close() instead of
my_close(). */
fd2 = dup(fd);
if (fd2 < 0) {
DBUG_PRINT("error",("Got error %d on dup",fd2));
my_errno=errno;
my_error(EE_OUT_OF_FILERESOURCES,
MYF(ME_BELL+ME_WAITTANG),
filename, my_errno);
}
my_close(fd, MYF(MY_WME));
}
return(fd2);
}
/*************************************************************************
Gets the InnoDB transaction handle for a MySQL handler object, creates
an InnoDB transaction struct if the corresponding MySQL thread struct still
lacks one. */
static
trx_t*
check_trx_exists(
/*=============*/
/* out: InnoDB transaction handle */
THD* thd) /* in: user thread handle */
{
trx_t* trx;
ut_ad(thd == current_thd);
trx = (trx_t*) thd->ha_data[innobase_hton.slot];
if (trx == NULL) {
DBUG_ASSERT(thd != NULL);
trx = trx_allocate_for_mysql();
trx->mysql_thd = thd;
trx->mysql_query_str = &(thd->query);
trx->active_trans = 0;
/* Update the info whether we should skip XA steps that eat
CPU time */
trx->support_xa = (ibool)(thd->variables.innodb_support_xa);
thd->ha_data[innobase_hton.slot] = trx;
} else {
if (trx->magic_n != TRX_MAGIC_N) {
mem_analyze_corruption((byte*)trx);
ut_a(0);
}
}
if (thd->options & OPTION_NO_FOREIGN_KEY_CHECKS) {
trx->check_foreigns = FALSE;
} else {
trx->check_foreigns = TRUE;
}
if (thd->options & OPTION_RELAXED_UNIQUE_CHECKS) {
trx->check_unique_secondary = FALSE;
} else {
trx->check_unique_secondary = TRUE;
}
return(trx);
}
/*************************************************************************
Construct ha_innobase handler. */
ha_innobase::ha_innobase(TABLE *table_arg)
:handler(&innobase_hton, table_arg),
int_table_flags(HA_REC_NOT_IN_SEQ |
HA_NULL_IN_KEY |
HA_CAN_INDEX_BLOBS |
HA_CAN_SQL_HANDLER |
HA_NOT_EXACT_COUNT |
HA_PRIMARY_KEY_IN_READ_INDEX |
HA_CAN_GEOMETRY |
HA_TABLE_SCAN_ON_INDEX),
start_of_scan(0),
num_write_row(0)
{}
/*************************************************************************
Updates the user_thd field in a handle and also allocates a new InnoDB
transaction handle if needed, and updates the transaction fields in the
prebuilt struct. */
inline
int
ha_innobase::update_thd(
/*====================*/
/* out: 0 or error code */
THD* thd) /* in: thd to use the handle */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
trx_t* trx;
trx = check_trx_exists(thd);
if (prebuilt->trx != trx) {
row_update_prebuilt_trx(prebuilt, trx);
}
user_thd = thd;
return(0);
}
/*************************************************************************
Registers that InnoDB takes part in an SQL statement, so that MySQL knows to
roll back the statement if the statement results in an error. This MUST be
called for every SQL statement that may be rolled back by MySQL. Calling this
several times to register the same statement is allowed, too. */
inline
void
innobase_register_stmt(
/*===================*/
THD* thd) /* in: MySQL thd (connection) object */
{
/* Register the statement */
trans_register_ha(thd, FALSE, &innobase_hton);
}
/*************************************************************************
Registers an InnoDB transaction in MySQL, so that the MySQL XA code knows
to call the InnoDB prepare and commit, or rollback for the transaction. This
MUST be called for every transaction for which the user may call commit or
rollback. Calling this several times to register the same transaction is
allowed, too.
This function also registers the current SQL statement. */
inline
void
innobase_register_trx_and_stmt(
/*===========================*/
THD* thd) /* in: MySQL thd (connection) object */
{
/* NOTE that actually innobase_register_stmt() registers also
the transaction in the AUTOCOMMIT=1 mode. */
innobase_register_stmt(thd);
if (thd->options & (OPTION_NOT_AUTOCOMMIT | OPTION_BEGIN)) {
/* No autocommit mode, register for a transaction */
trans_register_ha(thd, TRUE, &innobase_hton);
}
}
/* BACKGROUND INFO: HOW THE MYSQL QUERY CACHE WORKS WITH INNODB
------------------------------------------------------------
1) The use of the query cache for TBL is disabled when there is an
uncommitted change to TBL.
2) When a change to TBL commits, InnoDB stores the current value of
its global trx id counter, let us denote it by INV_TRX_ID, to the table object
in the InnoDB data dictionary, and does only allow such transactions whose
id <= INV_TRX_ID to use the query cache.
3) When InnoDB does an INSERT/DELETE/UPDATE to a table TBL, or an implicit
modification because an ON DELETE CASCADE, we invalidate the MySQL query cache
of TBL immediately.
How this is implemented inside InnoDB:
1) Since every modification always sets an IX type table lock on the InnoDB
table, it is easy to check if there can be uncommitted modifications for a
table: just check if there are locks in the lock list of the table.
2) When a transaction inside InnoDB commits, it reads the global trx id
counter and stores the value INV_TRX_ID to the tables on which it had a lock.
3) If there is an implicit table change from ON DELETE CASCADE or SET NULL,
InnoDB calls an invalidate method for the MySQL query cache for that table.
How this is implemented inside sql_cache.cc:
1) The query cache for an InnoDB table TBL is invalidated immediately at an
INSERT/UPDATE/DELETE, just like in the case of MyISAM. No need to delay
invalidation to the transaction commit.
2) To store or retrieve a value from the query cache of an InnoDB table TBL,
any query must first ask InnoDB's permission. We must pass the thd as a
parameter because InnoDB will look at the trx id, if any, associated with
that thd.
3) Use of the query cache for InnoDB tables is now allowed also when
AUTOCOMMIT==0 or we are inside BEGIN ... COMMIT. Thus transactions no longer
put restrictions on the use of the query cache.
*/
/**********************************************************************
The MySQL query cache uses this to check from InnoDB if the query cache at
the moment is allowed to operate on an InnoDB table. The SQL query must
be a non-locking SELECT.
The query cache is allowed to operate on certain query only if this function
returns TRUE for all tables in the query.
If thd is not in the autocommit state, this function also starts a new
transaction for thd if there is no active trx yet, and assigns a consistent
read view to it if there is no read view yet.
Why a deadlock of threads is not possible: the query cache calls this function
at the start of a SELECT processing. Then the calling thread cannot be
holding any InnoDB semaphores. The calling thread is holding the
query cache mutex, and this function will reserver the InnoDB kernel mutex.
Thus, the 'rank' in sync0sync.h of the MySQL query cache mutex is above
the InnoDB kernel mutex. */
my_bool
innobase_query_caching_of_table_permitted(
/*======================================*/
/* out: TRUE if permitted, FALSE if not;
note that the value FALSE does not mean
we should invalidate the query cache:
invalidation is called explicitly */
THD* thd, /* in: thd of the user who is trying to
store a result to the query cache or
retrieve it */
char* full_name, /* in: concatenation of database name,
the null character '\0', and the table
name */
uint full_name_len, /* in: length of the full name, i.e.
len(dbname) + len(tablename) + 1 */
ulonglong *unused) /* unused for this engine */
{
ibool is_autocommit;
trx_t* trx;
char norm_name[1000];
ut_a(full_name_len < 999);
if (thd->variables.tx_isolation == ISO_SERIALIZABLE) {
/* In the SERIALIZABLE mode we add LOCK IN SHARE MODE to every
plain SELECT if AUTOCOMMIT is not on. */
return((my_bool)FALSE);
}
trx = check_trx_exists(thd);
if (trx->has_search_latch) {
ut_print_timestamp(stderr);
sql_print_error("The calling thread is holding the adaptive "
"search, latch though calling "
"innobase_query_caching_of_table_permitted.");
mutex_enter_noninline(&kernel_mutex);
trx_print(stderr, trx, 1024);
mutex_exit_noninline(&kernel_mutex);
}
innobase_release_stat_resources(trx);
if (!(thd->options & (OPTION_NOT_AUTOCOMMIT | OPTION_BEGIN))) {
is_autocommit = TRUE;
} else {
is_autocommit = FALSE;
}
if (is_autocommit && trx->n_mysql_tables_in_use == 0) {
/* We are going to retrieve the query result from the query
cache. This cannot be a store operation to the query cache
because then MySQL would have locks on tables already.
TODO: if the user has used LOCK TABLES to lock the table,
then we open a transaction in the call of row_.. below.
That trx can stay open until UNLOCK TABLES. The same problem
exists even if we do not use the query cache. MySQL should be
modified so that it ALWAYS calls some cleanup function when
the processing of a query ends!
We can imagine we instantaneously serialize this consistent
read trx to the current trx id counter. If trx2 would have
changed the tables of a query result stored in the cache, and
trx2 would have already committed, making the result obsolete,
then trx2 would have already invalidated the cache. Thus we
can trust the result in the cache is ok for this query. */
return((my_bool)TRUE);
}
/* Normalize the table name to InnoDB format */
memcpy(norm_name, full_name, full_name_len);
norm_name[strlen(norm_name)] = '/'; /* InnoDB uses '/' as the
separator between db and table */
norm_name[full_name_len] = '\0';
#ifdef __WIN__
innobase_casedn_str(norm_name);
#endif
/* The call of row_search_.. will start a new transaction if it is
not yet started */
if (trx->active_trans == 0) {
innobase_register_trx_and_stmt(thd);
trx->active_trans = 1;
}
if (row_search_check_if_query_cache_permitted(trx, norm_name)) {
/* printf("Query cache for %s permitted\n", norm_name); */
return((my_bool)TRUE);
}
/* printf("Query cache for %s NOT permitted\n", norm_name); */
return((my_bool)FALSE);
}
/*********************************************************************
Invalidates the MySQL query cache for the table.
NOTE that the exact prototype of this function has to be in
/innobase/row/row0ins.c! */
extern "C"
void
innobase_invalidate_query_cache(
/*============================*/
trx_t* trx, /* in: transaction which modifies the table */
char* full_name, /* in: concatenation of database name, null
char '\0', table name, null char'\0';
NOTE that in Windows this is always
in LOWER CASE! */
ulint full_name_len) /* in: full name length where also the null
chars count */
{
/* Note that the sync0sync.h rank of the query cache mutex is just
above the InnoDB kernel mutex. The caller of this function must not
have latches of a lower rank. */
/* Argument TRUE below means we are using transactions */
#ifdef HAVE_QUERY_CACHE
query_cache.invalidate((THD*)(trx->mysql_thd),
(const char*)full_name,
(uint32)full_name_len,
TRUE);
#endif
}
/*********************************************************************
Get the quote character to be used in SQL identifiers.
This definition must match the one in innobase/ut/ut0ut.c! */
extern "C"
int
mysql_get_identifier_quote_char(
/*============================*/
/* out: quote character to be
used in SQL identifiers; EOF if none */
trx_t* trx, /* in: transaction */
const char* name, /* in: name to print */
ulint namelen)/* in: length of name */
{
if (!trx || !trx->mysql_thd) {
return(EOF);
}
return(get_quote_char_for_identifier((THD*) trx->mysql_thd,
name, (int) namelen));
}
/**************************************************************************
Determines if the currently running transaction has been interrupted. */
extern "C"
ibool
trx_is_interrupted(
/*===============*/
/* out: TRUE if interrupted */
trx_t* trx) /* in: transaction */
{
return(trx && trx->mysql_thd && ((THD*) trx->mysql_thd)->killed);
}
/**************************************************************************
Obtain a pointer to the MySQL THD object, as in current_thd(). This
definition must match the one in sql/ha_innodb.cc! */
extern "C"
void*
innobase_current_thd(void)
/*======================*/
/* out: MySQL THD object */
{
return(current_thd);
}
/*********************************************************************
Call this when you have opened a new table handle in HANDLER, before you
call index_read_idx() etc. Actually, we can let the cursor stay open even
over a transaction commit! Then you should call this before every operation,
fetch next etc. This function inits the necessary things even after a
transaction commit. */
void
ha_innobase::init_table_handle_for_HANDLER(void)
/*============================================*/
{
row_prebuilt_t* prebuilt;
/* If current thd does not yet have a trx struct, create one.
If the current handle does not yet have a prebuilt struct, create
one. Update the trx pointers in the prebuilt struct. Normally
this operation is done in external_lock. */
update_thd(current_thd);
/* Initialize the prebuilt struct much like it would be inited in
external_lock */
prebuilt = (row_prebuilt_t*)innobase_prebuilt;
innobase_release_stat_resources(prebuilt->trx);
/* If the transaction is not started yet, start it */
trx_start_if_not_started_noninline(prebuilt->trx);
/* Assign a read view if the transaction does not have it yet */
trx_assign_read_view(prebuilt->trx);
/* Set the MySQL flag to mark that there is an active transaction */
if (prebuilt->trx->active_trans == 0) {
innobase_register_trx_and_stmt(current_thd);
prebuilt->trx->active_trans = 1;
}
/* We did the necessary inits in this function, no need to repeat them
in row_search_for_mysql */
prebuilt->sql_stat_start = FALSE;
/* We let HANDLER always to do the reads as consistent reads, even
if the trx isolation level would have been specified as SERIALIZABLE */
prebuilt->select_lock_type = LOCK_NONE;
prebuilt->stored_select_lock_type = LOCK_NONE;
/* Always fetch all columns in the index record */
prebuilt->hint_need_to_fetch_extra_cols = ROW_RETRIEVE_ALL_COLS;
/* We want always to fetch all columns in the whole row? Or do
we???? */
prebuilt->read_just_key = FALSE;
prebuilt->used_in_HANDLER = TRUE;
prebuilt->keep_other_fields_on_keyread = FALSE;
}
/*************************************************************************
Opens an InnoDB database. */
bool
innobase_init(void)
/*===============*/
/* out: &innobase_hton, or NULL on error */
{
static char current_dir[3]; /* Set if using current lib */
int err;
bool ret;
char *default_path;
DBUG_ENTER("innobase_init");
if (have_innodb != SHOW_OPTION_YES)
goto error;
ut_a(DATA_MYSQL_TRUE_VARCHAR == (ulint)MYSQL_TYPE_VARCHAR);
/* Check that values don't overflow on 32-bit systems. */
if (sizeof(ulint) == 4) {
if (innobase_buffer_pool_size > UINT_MAX32) {
sql_print_error(
"innobase_buffer_pool_size can't be over 4GB"
" on 32-bit systems");
goto error;
}
if (innobase_log_file_size > UINT_MAX32) {
sql_print_error(
"innobase_log_file_size can't be over 4GB"
" on 32-bit systems");
goto error;
}
}
os_innodb_umask = (ulint)my_umask;
/* First calculate the default path for innodb_data_home_dir etc.,
in case the user has not given any value.
Note that when using the embedded server, the datadirectory is not
necessarily the current directory of this program. */
if (mysqld_embedded) {
default_path = mysql_real_data_home;
fil_path_to_mysql_datadir = mysql_real_data_home;
} else {
/* It's better to use current lib, to keep paths short */
current_dir[0] = FN_CURLIB;
current_dir[1] = FN_LIBCHAR;
current_dir[2] = 0;
default_path = current_dir;
}
ut_a(default_path);
if (specialflag & SPECIAL_NO_PRIOR) {
srv_set_thread_priorities = FALSE;
} else {
srv_set_thread_priorities = TRUE;
srv_query_thread_priority = QUERY_PRIOR;
}
/* Set InnoDB initialization parameters according to the values
read from MySQL .cnf file */
/*--------------- Data files -------------------------*/
/* The default dir for data files is the datadir of MySQL */
srv_data_home = (innobase_data_home_dir ? innobase_data_home_dir :
default_path);
/* Set default InnoDB data file size to 10 MB and let it be
auto-extending. Thus users can use InnoDB in >= 4.0 without having
to specify any startup options. */
if (!innobase_data_file_path) {
innobase_data_file_path = (char*) "ibdata1:10M:autoextend";
}
/* Since InnoDB edits the argument in the next call, we make another
copy of it: */
internal_innobase_data_file_path = my_strdup(innobase_data_file_path,
MYF(MY_FAE));
ret = (bool) srv_parse_data_file_paths_and_sizes(
internal_innobase_data_file_path,
&srv_data_file_names,
&srv_data_file_sizes,
&srv_data_file_is_raw_partition,
&srv_n_data_files,
&srv_auto_extend_last_data_file,
&srv_last_file_size_max);
if (ret == FALSE) {
sql_print_error(
"InnoDB: syntax error in innodb_data_file_path");
my_free(internal_innobase_data_file_path,
MYF(MY_ALLOW_ZERO_PTR));
goto error;
}
/* -------------- Log files ---------------------------*/
/* The default dir for log files is the datadir of MySQL */
if (!innobase_log_group_home_dir) {
innobase_log_group_home_dir = default_path;
}
#ifdef UNIV_LOG_ARCHIVE
/* Since innodb_log_arch_dir has no relevance under MySQL,
starting from 4.0.6 we always set it the same as
innodb_log_group_home_dir: */
innobase_log_arch_dir = innobase_log_group_home_dir;
srv_arch_dir = innobase_log_arch_dir;
#endif /* UNIG_LOG_ARCHIVE */
ret = (bool)
srv_parse_log_group_home_dirs(innobase_log_group_home_dir,
&srv_log_group_home_dirs);
if (ret == FALSE || innobase_mirrored_log_groups != 1) {
sql_print_error("syntax error in innodb_log_group_home_dir, or a "
"wrong number of mirrored log groups");
my_free(internal_innobase_data_file_path,
MYF(MY_ALLOW_ZERO_PTR));
goto error;
}
/* --------------------------------------------------*/
srv_file_flush_method_str = innobase_unix_file_flush_method;
srv_n_log_groups = (ulint) innobase_mirrored_log_groups;
srv_n_log_files = (ulint) innobase_log_files_in_group;
srv_log_file_size = (ulint) innobase_log_file_size;
#ifdef UNIV_LOG_ARCHIVE
srv_log_archive_on = (ulint) innobase_log_archive;
#endif /* UNIV_LOG_ARCHIVE */
srv_log_buffer_size = (ulint) innobase_log_buffer_size;
/* We set srv_pool_size here in units of 1 kB. InnoDB internally
changes the value so that it becomes the number of database pages. */
if (innobase_buffer_pool_awe_mem_mb == 0) {
/* Careful here: we first convert the signed long int to ulint
and only after that divide */
srv_pool_size = ((ulint) innobase_buffer_pool_size) / 1024;
} else {
srv_use_awe = TRUE;
srv_pool_size = (ulint)
(1024 * innobase_buffer_pool_awe_mem_mb);
srv_awe_window_size = (ulint) innobase_buffer_pool_size;
/* Note that what the user specified as
innodb_buffer_pool_size is actually the AWE memory window
size in this case, and the real buffer pool size is
determined by .._awe_mem_mb. */
}
srv_mem_pool_size = (ulint) innobase_additional_mem_pool_size;
srv_n_file_io_threads = (ulint) innobase_file_io_threads;
srv_lock_wait_timeout = (ulint) innobase_lock_wait_timeout;
srv_force_recovery = (ulint) innobase_force_recovery;
srv_use_doublewrite_buf = (ibool) innobase_use_doublewrite;
srv_use_checksums = (ibool) innobase_use_checksums;
srv_use_adaptive_hash_indexes = (ibool) innobase_adaptive_hash_index;
os_use_large_pages = (ibool) innobase_use_large_pages;
os_large_page_size = (ulint) innobase_large_page_size;
row_rollback_on_timeout = (ibool) innobase_rollback_on_timeout;
srv_file_per_table = (ibool) innobase_file_per_table;
srv_locks_unsafe_for_binlog = (ibool) innobase_locks_unsafe_for_binlog;
srv_max_n_open_files = (ulint) innobase_open_files;
srv_innodb_status = (ibool) innobase_create_status_file;
srv_print_verbose_log = mysqld_embedded ? 0 : 1;
/* Store the default charset-collation number of this MySQL
installation */
data_mysql_default_charset_coll = (ulint)default_charset_info->number;
ut_a(DATA_MYSQL_LATIN1_SWEDISH_CHARSET_COLL ==
my_charset_latin1.number);
ut_a(DATA_MYSQL_BINARY_CHARSET_COLL == my_charset_bin.number);
/* Store the latin1_swedish_ci character ordering table to InnoDB. For
non-latin1_swedish_ci charsets we use the MySQL comparison functions,
and consequently we do not need to know the ordering internally in
InnoDB. */
ut_a(0 == strcmp((char*)my_charset_latin1.name,
(char*)"latin1_swedish_ci"));
memcpy(srv_latin1_ordering, my_charset_latin1.sort_order, 256);
/* Since we in this module access directly the fields of a trx
struct, and due to different headers and flags it might happen that
mutex_t has a different size in this module and in InnoDB
modules, we check at run time that the size is the same in
these compilation modules. */
srv_sizeof_trx_t_in_ha_innodb_cc = sizeof(trx_t);
err = innobase_start_or_create_for_mysql();
if (err != DB_SUCCESS) {
my_free(internal_innobase_data_file_path,
MYF(MY_ALLOW_ZERO_PTR));
goto error;
}
(void) hash_init(&innobase_open_tables,system_charset_info, 32, 0, 0,
(hash_get_key) innobase_get_key, 0, 0);
pthread_mutex_init(&innobase_share_mutex, MY_MUTEX_INIT_FAST);
pthread_mutex_init(&prepare_commit_mutex, MY_MUTEX_INIT_FAST);
pthread_mutex_init(&commit_threads_m, MY_MUTEX_INIT_FAST);
pthread_mutex_init(&commit_cond_m, MY_MUTEX_INIT_FAST);
pthread_cond_init(&commit_cond, NULL);
innodb_inited= 1;
/* If this is a replication slave and we needed to do a crash recovery,
set the master binlog position to what InnoDB internally knew about
how far we got transactions durable inside InnoDB. There is a
problem here: if the user used also MyISAM tables, InnoDB might not
know the right position for them.
THIS DOES NOT WORK CURRENTLY because replication seems to initialize
glob_mi also after innobase_init. */
/* if (trx_sys_mysql_master_log_pos != -1) {
ut_memcpy(glob_mi.log_file_name, trx_sys_mysql_master_log_name,
1 + ut_strlen(trx_sys_mysql_master_log_name));
glob_mi.pos = trx_sys_mysql_master_log_pos;
}
*/
DBUG_RETURN(FALSE);
error:
have_innodb= SHOW_OPTION_DISABLED; // If we couldn't use handler
DBUG_RETURN(TRUE);
}
/***********************************************************************
Closes an InnoDB database. */
bool
innobase_end(void)
/*==============*/
/* out: TRUE if error */
{
int err= 0;
DBUG_ENTER("innobase_end");
#ifdef __NETWARE__ /* some special cleanup for NetWare */
if (nw_panic) {
set_panic_flag_for_netware();
}
#endif
if (innodb_inited) {
srv_fast_shutdown = (ulint) innobase_fast_shutdown;
innodb_inited = 0;
if (innobase_shutdown_for_mysql() != DB_SUCCESS) {
err = 1;
}
hash_free(&innobase_open_tables);
my_free(internal_innobase_data_file_path,
MYF(MY_ALLOW_ZERO_PTR));
pthread_mutex_destroy(&innobase_share_mutex);
pthread_mutex_destroy(&prepare_commit_mutex);
pthread_mutex_destroy(&commit_threads_m);
pthread_mutex_destroy(&commit_cond_m);
pthread_cond_destroy(&commit_cond);
}
DBUG_RETURN(err);
}
/********************************************************************
Flushes InnoDB logs to disk and makes a checkpoint. Really, a commit flushes
the logs, and the name of this function should be innobase_checkpoint. */
bool
innobase_flush_logs(void)
/*=====================*/
/* out: TRUE if error */
{
bool result = 0;
DBUG_ENTER("innobase_flush_logs");
log_buffer_flush_to_disk();
DBUG_RETURN(result);
}
/*********************************************************************
Commits a transaction in an InnoDB database. */
void
innobase_commit_low(
/*================*/
trx_t* trx) /* in: transaction handle */
{
if (trx->conc_state == TRX_NOT_STARTED) {
return;
}
#ifdef HAVE_REPLICATION
THD *thd=current_thd;
if (thd && thd->slave_thread) {
/* Update the replication position info inside InnoDB */
trx->mysql_master_log_file_name
= active_mi->rli.group_master_log_name;
trx->mysql_master_log_pos = ((ib_longlong)
active_mi->rli.future_group_master_log_pos);
}
#endif /* HAVE_REPLICATION */
trx_commit_for_mysql(trx);
}
/*********************************************************************
Creates an InnoDB transaction struct for the thd if it does not yet have one.
Starts a new InnoDB transaction if a transaction is not yet started. And
assigns a new snapshot for a consistent read if the transaction does not yet
have one. */
int
innobase_start_trx_and_assign_read_view(
/*====================================*/
/* out: 0 */
THD* thd) /* in: MySQL thread handle of the user for whom
the transaction should be committed */
{
trx_t* trx;
DBUG_ENTER("innobase_start_trx_and_assign_read_view");
/* Create a new trx struct for thd, if it does not yet have one */
trx = check_trx_exists(thd);
/* This is just to play safe: release a possible FIFO ticket and
search latch. Since we will reserve the kernel mutex, we have to
release the search system latch first to obey the latching order. */
innobase_release_stat_resources(trx);
/* If the transaction is not started yet, start it */
trx_start_if_not_started_noninline(trx);
/* Assign a read view if the transaction does not have it yet */
trx_assign_read_view(trx);
/* Set the MySQL flag to mark that there is an active transaction */
if (trx->active_trans == 0) {
innobase_register_trx_and_stmt(current_thd);
trx->active_trans = 1;
}
DBUG_RETURN(0);
}
/*********************************************************************
Commits a transaction in an InnoDB database or marks an SQL statement
ended. */
static
int
innobase_commit(
/*============*/
/* out: 0 */
THD* thd, /* in: MySQL thread handle of the user for whom
the transaction should be committed */
bool all) /* in: TRUE - commit transaction
FALSE - the current SQL statement ended */
{
trx_t* trx;
DBUG_ENTER("innobase_commit");
DBUG_PRINT("trans", ("ending transaction"));
trx = check_trx_exists(thd);
/* Update the info whether we should skip XA steps that eat CPU time */
trx->support_xa = (ibool)(thd->variables.innodb_support_xa);
/* Release a possible FIFO ticket and search latch. Since we will
reserve the kernel mutex, we have to release the search system latch
first to obey the latching order. */
if (trx->has_search_latch) {
trx_search_latch_release_if_reserved(trx);
}
/* The flag trx->active_trans is set to 1 in
1. ::external_lock(),
2. ::start_stmt(),
3. innobase_query_caching_of_table_permitted(),
4. innobase_savepoint(),
5. ::init_table_handle_for_HANDLER(),
6. innobase_start_trx_and_assign_read_view(),
7. ::transactional_table_lock()
and it is only set to 0 in a commit or a rollback. If it is 0 we know
there cannot be resources to be freed and we could return immediately.
For the time being, we play safe and do the cleanup though there should
be nothing to clean up. */
if (trx->active_trans == 0
&& trx->conc_state != TRX_NOT_STARTED) {
sql_print_error("trx->active_trans == 0, but trx->conc_state != "
"TRX_NOT_STARTED");
}
if (all
|| (!(thd->options & (OPTION_NOT_AUTOCOMMIT | OPTION_BEGIN)))) {
/* We were instructed to commit the whole transaction, or
this is an SQL statement end and autocommit is on */
/* We need current binlog position for ibbackup to work.
Note, the position is current because of prepare_commit_mutex */
retry:
if (srv_commit_concurrency > 0)
{
pthread_mutex_lock(&commit_cond_m);
commit_threads++;
if (commit_threads > srv_commit_concurrency)
{
commit_threads--;
pthread_cond_wait(&commit_cond, &commit_cond_m);
pthread_mutex_unlock(&commit_cond_m);
goto retry;
}
else
pthread_mutex_unlock(&commit_cond_m);
}
trx->mysql_log_file_name = mysql_bin_log.get_log_fname();
trx->mysql_log_offset =
(ib_longlong)mysql_bin_log.get_log_file()->pos_in_file;
innobase_commit_low(trx);
if (srv_commit_concurrency > 0)
{
pthread_mutex_lock(&commit_cond_m);
commit_threads--;
pthread_cond_signal(&commit_cond);
pthread_mutex_unlock(&commit_cond_m);
}
if (trx->active_trans == 2) {
pthread_mutex_unlock(&prepare_commit_mutex);
}
trx->active_trans = 0;
} else {
/* We just mark the SQL statement ended and do not do a
transaction commit */
if (trx->auto_inc_lock) {
/* If we had reserved the auto-inc lock for some
table in this SQL statement we release it now */
row_unlock_table_autoinc_for_mysql(trx);
}
/* Store the current undo_no of the transaction so that we
know where to roll back if we have to roll back the next
SQL statement */
trx_mark_sql_stat_end(trx);
}
/* Tell the InnoDB server that there might be work for utility
threads: */
if (trx->declared_to_be_inside_innodb) {
/* Release our possible ticket in the FIFO */
srv_conc_force_exit_innodb(trx);
}
srv_active_wake_master_thread();
DBUG_RETURN(0);
}
/* TODO: put the
MySQL-4.1 functionality back to 5.0. This is needed to get InnoDB Hot Backup
to work. */
/*********************************************************************
This is called when MySQL writes the binlog entry for the current
transaction. Writes to the InnoDB tablespace info which tells where the
MySQL binlog entry for the current transaction ended. Also commits the
transaction inside InnoDB but does NOT flush InnoDB log files to disk.
To flush you have to call innobase_commit_complete(). We have separated
flushing to eliminate the bottleneck of LOCK_log in log.cc which disabled
InnoDB's group commit capability. */
int
innobase_report_binlog_offset_and_commit(
/*=====================================*/
/* out: 0 */
THD* thd, /* in: user thread */
void* trx_handle, /* in: InnoDB trx handle */
char* log_file_name, /* in: latest binlog file name */
my_off_t end_offset) /* in: the offset in the binlog file
up to which we wrote */
{
trx_t* trx;
trx = (trx_t*)trx_handle;
ut_a(trx != NULL);
trx->mysql_log_file_name = log_file_name;
trx->mysql_log_offset = (ib_longlong)end_offset;
trx->flush_log_later = TRUE;
innobase_commit(thd, TRUE);
trx->flush_log_later = FALSE;
return(0);
}
#if 0
/***********************************************************************
This function stores the binlog offset and flushes logs. */
void
innobase_store_binlog_offset_and_flush_log(
/*=======================================*/
char *binlog_name, /* in: binlog name */
longlong offset) /* in: binlog offset */
{
mtr_t mtr;
assert(binlog_name != NULL);
/* Start a mini-transaction */
mtr_start_noninline(&mtr);
/* Update the latest MySQL binlog name and offset info
in trx sys header */
trx_sys_update_mysql_binlog_offset(
binlog_name,
offset,
TRX_SYS_MYSQL_LOG_INFO, &mtr);
/* Commits the mini-transaction */
mtr_commit(&mtr);
/* Synchronous flush of the log buffer to disk */
log_buffer_flush_to_disk();
}
#endif
/*********************************************************************
This is called after MySQL has written the binlog entry for the current
transaction. Flushes the InnoDB log files to disk if required. */
int
innobase_commit_complete(
/*=====================*/
/* out: 0 */
THD* thd) /* in: user thread */
{
trx_t* trx;
trx = (trx_t*) thd->ha_data[innobase_hton.slot];
if (trx && trx->active_trans) {
trx->active_trans = 0;
if (UNIV_UNLIKELY(srv_flush_log_at_trx_commit == 0)) {
return(0);
}
trx_commit_complete_for_mysql(trx);
}
return(0);
}
/*********************************************************************
Rolls back a transaction or the latest SQL statement. */
static int
innobase_rollback(
/*==============*/
/* out: 0 or error number */
THD* thd, /* in: handle to the MySQL thread of the user
whose transaction should be rolled back */
bool all) /* in: TRUE - commit transaction
FALSE - the current SQL statement ended */
{
int error = 0;
trx_t* trx;
DBUG_ENTER("innobase_rollback");
DBUG_PRINT("trans", ("aborting transaction"));
trx = check_trx_exists(thd);
/* Update the info whether we should skip XA steps that eat CPU time */
trx->support_xa = (ibool)(thd->variables.innodb_support_xa);
/* Release a possible FIFO ticket and search latch. Since we will
reserve the kernel mutex, we have to release the search system latch
first to obey the latching order. */
innobase_release_stat_resources(trx);
if (trx->auto_inc_lock) {
/* If we had reserved the auto-inc lock for some table (if
we come here to roll back the latest SQL statement) we
release it now before a possibly lengthy rollback */
row_unlock_table_autoinc_for_mysql(trx);
}
if (all
|| (!(thd->options & (OPTION_NOT_AUTOCOMMIT | OPTION_BEGIN)))) {
error = trx_rollback_for_mysql(trx);
trx->active_trans = 0;
} else {
error = trx_rollback_last_sql_stat_for_mysql(trx);
}
DBUG_RETURN(convert_error_code_to_mysql(error, NULL));
}
/*********************************************************************
Rolls back a transaction */
int
innobase_rollback_trx(
/*==================*/
/* out: 0 or error number */
trx_t* trx) /* in: transaction */
{
int error = 0;
DBUG_ENTER("innobase_rollback_trx");
DBUG_PRINT("trans", ("aborting transaction"));
/* Release a possible FIFO ticket and search latch. Since we will
reserve the kernel mutex, we have to release the search system latch
first to obey the latching order. */
innobase_release_stat_resources(trx);
if (trx->auto_inc_lock) {
/* If we had reserved the auto-inc lock for some table (if
we come here to roll back the latest SQL statement) we
release it now before a possibly lengthy rollback */
row_unlock_table_autoinc_for_mysql(trx);
}
error = trx_rollback_for_mysql(trx);
DBUG_RETURN(convert_error_code_to_mysql(error, NULL));
}
/*********************************************************************
Rolls back a transaction to a savepoint. */
static int
innobase_rollback_to_savepoint(
/*===========================*/
/* out: 0 if success, HA_ERR_NO_SAVEPOINT if
no savepoint with the given name */
THD* thd, /* in: handle to the MySQL thread of the user
whose transaction should be rolled back */
void *savepoint) /* in: savepoint data */
{
ib_longlong mysql_binlog_cache_pos;
int error = 0;
trx_t* trx;
char name[64];
DBUG_ENTER("innobase_rollback_to_savepoint");
trx = check_trx_exists(thd);
/* Release a possible FIFO ticket and search latch. Since we will
reserve the kernel mutex, we have to release the search system latch
first to obey the latching order. */
innobase_release_stat_resources(trx);
/* TODO: use provided savepoint data area to store savepoint data */
longlong2str((ulint)savepoint, name, 36);
error = (int) trx_rollback_to_savepoint_for_mysql(trx, name,
&mysql_binlog_cache_pos);
DBUG_RETURN(convert_error_code_to_mysql(error, NULL));
}
/*********************************************************************
Release transaction savepoint name. */
static
int
innobase_release_savepoint(
/*=======================*/
/* out: 0 if success, HA_ERR_NO_SAVEPOINT if
no savepoint with the given name */
THD* thd, /* in: handle to the MySQL thread of the user
whose transaction should be rolled back */
void* savepoint) /* in: savepoint data */
{
int error = 0;
trx_t* trx;
char name[64];
DBUG_ENTER("innobase_release_savepoint");
trx = check_trx_exists(thd);
/* TODO: use provided savepoint data area to store savepoint data */
longlong2str((ulint)savepoint, name, 36);
error = (int) trx_release_savepoint_for_mysql(trx, name);
DBUG_RETURN(convert_error_code_to_mysql(error, NULL));
}
/*********************************************************************
Sets a transaction savepoint. */
static
int
innobase_savepoint(
/*===============*/
/* out: always 0, that is, always succeeds */
THD* thd, /* in: handle to the MySQL thread */
void* savepoint) /* in: savepoint data */
{
int error = 0;
trx_t* trx;
DBUG_ENTER("innobase_savepoint");
/*
In the autocommit mode there is no sense to set a savepoint
(unless we are in sub-statement), so SQL layer ensures that
this method is never called in such situation.
*/
DBUG_ASSERT(thd->options & (OPTION_NOT_AUTOCOMMIT | OPTION_BEGIN) ||
thd->in_sub_stmt);
trx = check_trx_exists(thd);
/* Release a possible FIFO ticket and search latch. Since we will
reserve the kernel mutex, we have to release the search system latch
first to obey the latching order. */
innobase_release_stat_resources(trx);
/* cannot happen outside of transaction */
DBUG_ASSERT(trx->active_trans);
/* TODO: use provided savepoint data area to store savepoint data */
char name[64];
longlong2str((ulint)savepoint,name,36);
error = (int) trx_savepoint_for_mysql(trx, name, (ib_longlong)0);
DBUG_RETURN(convert_error_code_to_mysql(error, NULL));
}
/*********************************************************************
Frees a possible InnoDB trx object associated with the current THD. */
static
int
innobase_close_connection(
/*======================*/
/* out: 0 or error number */
THD* thd) /* in: handle to the MySQL thread of the user
whose resources should be free'd */
{
trx_t* trx;
trx = (trx_t*)thd->ha_data[innobase_hton.slot];
ut_a(trx);
if (trx->active_trans == 0
&& trx->conc_state != TRX_NOT_STARTED) {
sql_print_error("trx->active_trans == 0, but trx->conc_state != "
"TRX_NOT_STARTED");
}
if (trx->conc_state != TRX_NOT_STARTED &&
global_system_variables.log_warnings)
sql_print_warning("MySQL is closing a connection that has an active "
"InnoDB transaction. %lu row modifications will "
"roll back.",
(ulong)trx->undo_no.low);
innobase_rollback_trx(trx);
trx_free_for_mysql(trx);
return(0);
}
/*****************************************************************************
** InnoDB database tables
*****************************************************************************/
/********************************************************************
Get the record format from the data dictionary. */
enum row_type
ha_innobase::get_row_type() const
/*=============================*/
/* out: ROW_TYPE_REDUNDANT or ROW_TYPE_COMPACT */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
if (prebuilt && prebuilt->table) {
if (prebuilt->table->comp) {
return(ROW_TYPE_COMPACT);
} else {
return(ROW_TYPE_REDUNDANT);
}
}
ut_ad(0);
return(ROW_TYPE_NOT_USED);
}
/********************************************************************
Gives the file extension of an InnoDB single-table tablespace. */
static const char* ha_innobase_exts[] = {
".ibd",
NullS
};
const char**
ha_innobase::bas_ext() const
/*========================*/
/* out: file extension string */
{
return ha_innobase_exts;
}
/*********************************************************************
Normalizes a table name string. A normalized name consists of the
database name catenated to '/' and table name. An example:
test/mytable. On Windows normalization puts both the database name and the
table name always to lower case. */
static
void
normalize_table_name(
/*=================*/
char* norm_name, /* out: normalized name as a
null-terminated string */
const char* name) /* in: table name string */
{
char* name_ptr;
char* db_ptr;
char* ptr;
/* Scan name from the end */
ptr = strend(name)-1;
while (ptr >= name && *ptr != '\\' && *ptr != '/') {
ptr--;
}
name_ptr = ptr + 1;
DBUG_ASSERT(ptr > name);
ptr--;
while (ptr >= name && *ptr != '\\' && *ptr != '/') {
ptr--;
}
db_ptr = ptr + 1;
memcpy(norm_name, db_ptr, strlen(name) + 1 - (db_ptr - name));
norm_name[name_ptr - db_ptr - 1] = '/';
#ifdef __WIN__
innobase_casedn_str(norm_name);
#endif
}
/*********************************************************************
Creates and opens a handle to a table which already exists in an InnoDB
database. */
int
ha_innobase::open(
/*==============*/
/* out: 1 if error, 0 if success */
const char* name, /* in: table name */
int mode, /* in: not used */
uint test_if_locked) /* in: not used */
{
dict_table_t* ib_table;
char norm_name[1000];
THD* thd;
DBUG_ENTER("ha_innobase::open");
UT_NOT_USED(mode);
UT_NOT_USED(test_if_locked);
thd = current_thd;
/* Under some cases MySQL seems to call this function while
holding btr_search_latch. This breaks the latching order as
we acquire dict_sys->mutex below and leads to a deadlock. */
if (thd != NULL) {
innobase_release_temporary_latches(thd);
}
normalize_table_name(norm_name, name);
user_thd = NULL;
last_query_id = (ulong)-1;
if (!(share=get_share(name))) {
DBUG_RETURN(1);
}
/* Create buffers for packing the fields of a record. Why
table->reclength did not work here? Obviously, because char
fields when packed actually became 1 byte longer, when we also
stored the string length as the first byte. */
upd_and_key_val_buff_len =
table->s->reclength + table->s->max_key_length
+ MAX_REF_PARTS * 3;
if (!(mysql_byte*) my_multi_malloc(MYF(MY_WME),
&upd_buff, upd_and_key_val_buff_len,
&key_val_buff, upd_and_key_val_buff_len,
NullS)) {
free_share(share);
DBUG_RETURN(1);
}
/* Get pointer to a table object in InnoDB dictionary cache */
ib_table = dict_table_get_and_increment_handle_count(
norm_name, NULL);
if (NULL == ib_table) {
ut_print_timestamp(stderr);
sql_print_error("Cannot find table %s from the internal data "
"dictionary\nof InnoDB though the .frm file "
"for the table exists. Maybe you\nhave "
"deleted and recreated InnoDB data files but "
"have forgotten\nto delete the corresponding "
".frm files of InnoDB tables, or you\n"
"have moved .frm files to another database?\n"
"See http://dev.mysql.com/doc/refman/5.0/en/innodb-troubleshooting.html\n"
"how you can resolve the problem.\n",
norm_name);
free_share(share);
my_free((gptr) upd_buff, MYF(0));
my_errno = ENOENT;
DBUG_RETURN(HA_ERR_NO_SUCH_TABLE);
}
if (ib_table->ibd_file_missing && !thd->tablespace_op) {
ut_print_timestamp(stderr);
sql_print_error("MySQL is trying to open a table handle but "
"the .ibd file for\ntable %s does not exist.\n"
"Have you deleted the .ibd file from the "
"database directory under\nthe MySQL datadir, "
"or have you used DISCARD TABLESPACE?\n"
"See http://dev.mysql.com/doc/refman/5.0/en/innodb-troubleshooting.html\n"
"how you can resolve the problem.\n",
norm_name);
free_share(share);
my_free((gptr) upd_buff, MYF(0));
my_errno = ENOENT;
dict_table_decrement_handle_count(ib_table);
DBUG_RETURN(HA_ERR_NO_SUCH_TABLE);
}
innobase_prebuilt = row_create_prebuilt(ib_table);
((row_prebuilt_t*)innobase_prebuilt)->mysql_row_len =
table->s->reclength;
((row_prebuilt_t*)innobase_prebuilt)->default_rec
= table->s->default_values;
ut_ad(table->s->default_values);
/* Looks like MySQL-3.23 sometimes has primary key number != 0 */
primary_key = table->s->primary_key;
key_used_on_scan = primary_key;
/* Allocate a buffer for a 'row reference'. A row reference is
a string of bytes of length ref_length which uniquely specifies
a row in our table. Note that MySQL may also compare two row
references for equality by doing a simple memcmp on the strings
of length ref_length! */
if (!row_table_got_default_clust_index(ib_table)) {
if (primary_key >= MAX_KEY) {
sql_print_error("Table %s has a primary key in InnoDB data "
"dictionary, but not in MySQL!", name);
}
((row_prebuilt_t*)innobase_prebuilt)
->clust_index_was_generated = FALSE;
/* MySQL allocates the buffer for ref. key_info->key_length
includes space for all key columns + one byte for each column
that may be NULL. ref_length must be as exact as possible to
save space, because all row reference buffers are allocated
based on ref_length. */
ref_length = table->key_info[primary_key].key_length;
} else {
if (primary_key != MAX_KEY) {
sql_print_error("Table %s has no primary key in InnoDB data "
"dictionary, but has one in MySQL! If you "
"created the table with a MySQL version < "
"3.23.54 and did not define a primary key, "
"but defined a unique key with all non-NULL "
"columns, then MySQL internally treats that "
"key as the primary key. You can fix this "
"error by dump + DROP + CREATE + reimport "
"of the table.", name);
}
((row_prebuilt_t*)innobase_prebuilt)
->clust_index_was_generated = TRUE;
ref_length = DATA_ROW_ID_LEN;
/* If we automatically created the clustered index, then
MySQL does not know about it, and MySQL must NOT be aware
of the index used on scan, to make it avoid checking if we
update the column of the index. That is why we assert below
that key_used_on_scan is the undefined value MAX_KEY.
The column is the row id in the automatical generation case,
and it will never be updated anyway. */
if (key_used_on_scan != MAX_KEY) {
sql_print_warning("Table %s key_used_on_scan is %lu even "
"though there is no primary key inside "
"InnoDB.", name, (ulong) key_used_on_scan);
}
}
block_size = 16 * 1024; /* Index block size in InnoDB: used by MySQL
in query optimization */
/* Init table lock structure */
thr_lock_data_init(&share->lock,&lock,(void*) 0);
info(HA_STATUS_NO_LOCK | HA_STATUS_VARIABLE | HA_STATUS_CONST);
DBUG_RETURN(0);
}
uint
ha_innobase::max_supported_key_part_length() const
{
return(DICT_MAX_INDEX_COL_LEN - 1);
}
/**********************************************************************
Closes a handle to an InnoDB table. */
int
ha_innobase::close(void)
/*====================*/
/* out: 0 */
{
THD* thd;
DBUG_ENTER("ha_innobase::close");
thd = current_thd; // avoid calling current_thd twice, it may be slow
if (thd != NULL) {
innobase_release_temporary_latches(thd);
}
row_prebuilt_free((row_prebuilt_t*) innobase_prebuilt);
my_free((gptr) upd_buff, MYF(0));
free_share(share);
/* Tell InnoDB server that there might be work for
utility threads: */
srv_active_wake_master_thread();
DBUG_RETURN(0);
}
/* The following accessor functions should really be inside MySQL code! */
/******************************************************************
Gets field offset for a field in a table. */
inline
uint
get_field_offset(
/*=============*/
/* out: offset */
TABLE* table, /* in: MySQL table object */
Field* field) /* in: MySQL field object */
{
return((uint) (field->ptr - (char*) table->record[0]));
}
/******************************************************************
Checks if a field in a record is SQL NULL. Uses the record format
information in table to track the null bit in record. */
inline
uint
field_in_record_is_null(
/*====================*/
/* out: 1 if NULL, 0 otherwise */
TABLE* table, /* in: MySQL table object */
Field* field, /* in: MySQL field object */
char* record) /* in: a row in MySQL format */
{
int null_offset;
if (!field->null_ptr) {
return(0);
}
null_offset = (uint) ((char*) field->null_ptr
- (char*) table->record[0]);
if (record[null_offset] & field->null_bit) {
return(1);
}
return(0);
}
/******************************************************************
Sets a field in a record to SQL NULL. Uses the record format
information in table to track the null bit in record. */
inline
void
set_field_in_record_to_null(
/*========================*/
TABLE* table, /* in: MySQL table object */
Field* field, /* in: MySQL field object */
char* record) /* in: a row in MySQL format */
{
int null_offset;
null_offset = (uint) ((char*) field->null_ptr
- (char*) table->record[0]);
record[null_offset] = record[null_offset] | field->null_bit;
}
extern "C" {
/*****************************************************************
InnoDB uses this function to compare two data fields for which the data type
is such that we must use MySQL code to compare them. NOTE that the prototype
of this function is in rem0cmp.c in InnoDB source code! If you change this
function, remember to update the prototype there! */
int
innobase_mysql_cmp(
/*===============*/
/* out: 1, 0, -1, if a is greater,
equal, less than b, respectively */
int mysql_type, /* in: MySQL type */
uint charset_number, /* in: number of the charset */
unsigned char* a, /* in: data field */
unsigned int a_length, /* in: data field length,
not UNIV_SQL_NULL */
unsigned char* b, /* in: data field */
unsigned int b_length) /* in: data field length,
not UNIV_SQL_NULL */
{
CHARSET_INFO* charset;
enum_field_types mysql_tp;
int ret;
DBUG_ASSERT(a_length != UNIV_SQL_NULL);
DBUG_ASSERT(b_length != UNIV_SQL_NULL);
mysql_tp = (enum_field_types) mysql_type;
switch (mysql_tp) {
case MYSQL_TYPE_BIT:
case MYSQL_TYPE_STRING:
case MYSQL_TYPE_VAR_STRING:
case FIELD_TYPE_TINY_BLOB:
case FIELD_TYPE_MEDIUM_BLOB:
case FIELD_TYPE_BLOB:
case FIELD_TYPE_LONG_BLOB:
case MYSQL_TYPE_VARCHAR:
/* Use the charset number to pick the right charset struct for
the comparison. Since the MySQL function get_charset may be
slow before Bar removes the mutex operation there, we first
look at 2 common charsets directly. */
if (charset_number == default_charset_info->number) {
charset = default_charset_info;
} else if (charset_number == my_charset_latin1.number) {
charset = &my_charset_latin1;
} else {
charset = get_charset(charset_number, MYF(MY_WME));
if (charset == NULL) {
sql_print_error("InnoDB needs charset %lu for doing "
"a comparison, but MySQL cannot "
"find that charset.",
(ulong) charset_number);
ut_a(0);
}
}
/* Starting from 4.1.3, we use strnncollsp() in comparisons of
non-latin1_swedish_ci strings. NOTE that the collation order
changes then: 'b\0\0...' is ordered BEFORE 'b ...'. Users
having indexes on such data need to rebuild their tables! */
ret = charset->coll->strnncollsp(charset,
a, a_length,
b, b_length, 0);
if (ret < 0) {
return(-1);
} else if (ret > 0) {
return(1);
} else {
return(0);
}
default:
assert(0);
}
return(0);
}
}
/******************************************************************
Converts a MySQL type to an InnoDB type. Note that this function returns
the 'mtype' of InnoDB. InnoDB differentiates between MySQL's old <= 4.1
VARCHAR and the new true VARCHAR in >= 5.0.3 by the 'prtype'. */
inline
ulint
get_innobase_type_from_mysql_type(
/*==============================*/
/* out: DATA_BINARY, DATA_VARCHAR, ... */
ulint* unsigned_flag, /* out: DATA_UNSIGNED if an 'unsigned type';
at least ENUM and SET, and unsigned integer
types are 'unsigned types' */
Field* field) /* in: MySQL field */
{
/* The following asserts try to check that the MySQL type code fits in
8 bits: this is used in ibuf and also when DATA_NOT_NULL is ORed to
the type */
DBUG_ASSERT((ulint)FIELD_TYPE_STRING < 256);
DBUG_ASSERT((ulint)FIELD_TYPE_VAR_STRING < 256);
DBUG_ASSERT((ulint)FIELD_TYPE_DOUBLE < 256);
DBUG_ASSERT((ulint)FIELD_TYPE_FLOAT < 256);
DBUG_ASSERT((ulint)FIELD_TYPE_DECIMAL < 256);
if (field->flags & UNSIGNED_FLAG) {
*unsigned_flag = DATA_UNSIGNED;
} else {
*unsigned_flag = 0;
}
if (field->real_type() == FIELD_TYPE_ENUM
|| field->real_type() == FIELD_TYPE_SET) {
/* MySQL has field->type() a string type for these, but the
data is actually internally stored as an unsigned integer
code! */
*unsigned_flag = DATA_UNSIGNED; /* MySQL has its own unsigned
flag set to zero, even though
internally this is an unsigned
integer type */
return(DATA_INT);
}
switch (field->type()) {
/* NOTE that we only allow string types in DATA_MYSQL
and DATA_VARMYSQL */
case MYSQL_TYPE_VAR_STRING: /* old <= 4.1 VARCHAR */
case MYSQL_TYPE_VARCHAR: /* new >= 5.0.3 true VARCHAR */
if (field->binary()) {
return(DATA_BINARY);
} else if (strcmp(
field->charset()->name,
"latin1_swedish_ci") == 0) {
return(DATA_VARCHAR);
} else {
return(DATA_VARMYSQL);
}
case MYSQL_TYPE_BIT:
case MYSQL_TYPE_STRING: if (field->binary()) {
return(DATA_FIXBINARY);
} else if (strcmp(
field->charset()->name,
"latin1_swedish_ci") == 0) {
return(DATA_CHAR);
} else {
return(DATA_MYSQL);
}
case FIELD_TYPE_NEWDECIMAL:
return(DATA_FIXBINARY);
case FIELD_TYPE_LONG:
case FIELD_TYPE_LONGLONG:
case FIELD_TYPE_TINY:
case FIELD_TYPE_SHORT:
case FIELD_TYPE_INT24:
case FIELD_TYPE_DATE:
case FIELD_TYPE_DATETIME:
case FIELD_TYPE_YEAR:
case FIELD_TYPE_NEWDATE:
case FIELD_TYPE_TIME:
case FIELD_TYPE_TIMESTAMP:
return(DATA_INT);
case FIELD_TYPE_FLOAT:
return(DATA_FLOAT);
case FIELD_TYPE_DOUBLE:
return(DATA_DOUBLE);
case FIELD_TYPE_DECIMAL:
return(DATA_DECIMAL);
case FIELD_TYPE_GEOMETRY:
case FIELD_TYPE_TINY_BLOB:
case FIELD_TYPE_MEDIUM_BLOB:
case FIELD_TYPE_BLOB:
case FIELD_TYPE_LONG_BLOB:
return(DATA_BLOB);
default:
assert(0);
}
return(0);
}
/***********************************************************************
Writes an unsigned integer value < 64k to 2 bytes, in the little-endian
storage format. */
inline
void
innobase_write_to_2_little_endian(
/*==============================*/
byte* buf, /* in: where to store */
ulint val) /* in: value to write, must be < 64k */
{
ut_a(val < 256 * 256);
buf[0] = (byte)(val & 0xFF);
buf[1] = (byte)(val / 256);
}
/***********************************************************************
Reads an unsigned integer value < 64k from 2 bytes, in the little-endian
storage format. */
inline
uint
innobase_read_from_2_little_endian(
/*===============================*/
/* out: value */
const mysql_byte* buf) /* in: from where to read */
{
return (uint) ((ulint)(buf[0]) + 256 * ((ulint)(buf[1])));
}
/***********************************************************************
Stores a key value for a row to a buffer. */
uint
ha_innobase::store_key_val_for_row(
/*===============================*/
/* out: key value length as stored in buff */
uint keynr, /* in: key number */
char* buff, /* in/out: buffer for the key value (in MySQL
format) */
uint buff_len,/* in: buffer length */
const mysql_byte* record)/* in: row in MySQL format */
{
KEY* key_info = table->key_info + keynr;
KEY_PART_INFO* key_part = key_info->key_part;
KEY_PART_INFO* end = key_part + key_info->key_parts;
char* buff_start = buff;
enum_field_types mysql_type;
Field* field;
ibool is_null;
DBUG_ENTER("store_key_val_for_row");
/* The format for storing a key field in MySQL is the following:
1. If the column can be NULL, then in the first byte we put 1 if the
field value is NULL, 0 otherwise.
2. If the column is of a BLOB type (it must be a column prefix field
in this case), then we put the length of the data in the field to the
next 2 bytes, in the little-endian format. If the field is SQL NULL,
then these 2 bytes are set to 0. Note that the length of data in the
field is <= column prefix length.
3. In a column prefix field, prefix_len next bytes are reserved for
data. In a normal field the max field length next bytes are reserved
for data. For a VARCHAR(n) the max field length is n. If the stored
value is the SQL NULL then these data bytes are set to 0.
4. We always use a 2 byte length for a true >= 5.0.3 VARCHAR. Note that
in the MySQL row format, the length is stored in 1 or 2 bytes,
depending on the maximum allowed length. But in the MySQL key value
format, the length always takes 2 bytes.
We have to zero-fill the buffer so that MySQL is able to use a
simple memcmp to compare two key values to determine if they are
equal. MySQL does this to compare contents of two 'ref' values. */
bzero(buff, buff_len);
for (; key_part != end; key_part++) {
is_null = FALSE;
if (key_part->null_bit) {
if (record[key_part->null_offset]
& key_part->null_bit) {
*buff = 1;
is_null = TRUE;
} else {
*buff = 0;
}
buff++;
}
field = key_part->field;
mysql_type = field->type();
if (mysql_type == MYSQL_TYPE_VARCHAR) {
/* >= 5.0.3 true VARCHAR */
ulint lenlen;
ulint len;
byte* data;
ulint key_len;
ulint true_len;
CHARSET_INFO* cs;
int error=0;
key_len = key_part->length;
if (is_null) {
buff += key_len + 2;
continue;
}
cs = field->charset();
lenlen = (ulint)
(((Field_varstring*)field)->length_bytes);
data = row_mysql_read_true_varchar(&len,
(byte*) (record
+ (ulint)get_field_offset(table, field)),
lenlen);
true_len = len;
/* For multi byte character sets we need to calculate
the true length of the key */
if (len > 0 && cs->mbmaxlen > 1) {
true_len = (ulint) cs->cset->well_formed_len(cs,
(const char *) data,
(const char *) data + len,
(uint) (key_len /
cs->mbmaxlen),
&error);
}
/* In a column prefix index, we may need to truncate
the stored value: */
if (true_len > key_len) {
true_len = key_len;
}
/* The length in a key value is always stored in 2
bytes */
row_mysql_store_true_var_len((byte*)buff, true_len, 2);
buff += 2;
memcpy(buff, data, true_len);
/* Note that we always reserve the maximum possible
length of the true VARCHAR in the key value, though
only len first bytes after the 2 length bytes contain
actual data. The rest of the space was reset to zero
in the bzero() call above. */
buff += key_len;
} else if (mysql_type == FIELD_TYPE_TINY_BLOB
|| mysql_type == FIELD_TYPE_MEDIUM_BLOB
|| mysql_type == FIELD_TYPE_BLOB
|| mysql_type == FIELD_TYPE_LONG_BLOB
/* MYSQL_TYPE_GEOMETRY data is treated
as BLOB data in innodb. */
|| mysql_type == FIELD_TYPE_GEOMETRY) {
CHARSET_INFO* cs;
ulint key_len;
ulint true_len;
int error=0;
ulint blob_len;
byte* blob_data;
ut_a(key_part->key_part_flag & HA_PART_KEY_SEG);
key_len = key_part->length;
if (is_null) {
buff += key_len + 2;
continue;
}
cs = field->charset();
blob_data = row_mysql_read_blob_ref(&blob_len,
(byte*) (record
+ (ulint)get_field_offset(table, field)),
(ulint) field->pack_length());
true_len = blob_len;
ut_a(get_field_offset(table, field)
== key_part->offset);
/* For multi byte character sets we need to calculate
the true length of the key */
if (blob_len > 0 && cs->mbmaxlen > 1) {
true_len = (ulint) cs->cset->well_formed_len(cs,
(const char *) blob_data,
(const char *) blob_data
+ blob_len,
(uint) (key_len /
cs->mbmaxlen),
&error);
}
/* All indexes on BLOB and TEXT are column prefix
indexes, and we may need to truncate the data to be
stored in the key value: */
if (true_len > key_len) {
true_len = key_len;
}
/* MySQL reserves 2 bytes for the length and the
storage of the number is little-endian */
innobase_write_to_2_little_endian(
(byte*)buff, true_len);
buff += 2;
memcpy(buff, blob_data, true_len);
/* Note that we always reserve the maximum possible
length of the BLOB prefix in the key value. */
buff += key_len;
} else {
/* Here we handle all other data types except the
true VARCHAR, BLOB and TEXT. Note that the column
value we store may be also in a column prefix
index. */
CHARSET_INFO* cs;
ulint true_len;
ulint key_len;
const mysql_byte* src_start;
int error=0;
enum_field_types real_type;
key_len = key_part->length;
if (is_null) {
buff += key_len;
continue;
}
src_start = record + key_part->offset;
real_type = field->real_type();
true_len = key_len;
/* Character set for the field is defined only
to fields whose type is string and real field
type is not enum or set. For these fields check
if character set is multi byte. */
if (real_type != FIELD_TYPE_ENUM
&& real_type != FIELD_TYPE_SET
&& ( mysql_type == MYSQL_TYPE_VAR_STRING
|| mysql_type == MYSQL_TYPE_STRING)) {
cs = field->charset();
/* For multi byte character sets we need to
calculate the true length of the key */
if (key_len > 0 && cs->mbmaxlen > 1) {
true_len = (ulint)
cs->cset->well_formed_len(cs,
(const char *)src_start,
(const char *)src_start
+ key_len,
(uint) (key_len /
cs->mbmaxlen),
&error);
}
}
memcpy(buff, src_start, true_len);
buff += true_len;
/* Pad the unused space with spaces. Note that no
padding is ever needed for UCS-2 because in MySQL,
all UCS2 characters are 2 bytes, as MySQL does not
support surrogate pairs, which are needed to represent
characters in the range U+10000 to U+10FFFF. */
if (true_len < key_len) {
ulint pad_len = key_len - true_len;
memset(buff, ' ', pad_len);
buff += pad_len;
}
}
}
ut_a(buff <= buff_start + buff_len);
DBUG_RETURN((uint)(buff - buff_start));
}
/******************************************************************
Builds a 'template' to the prebuilt struct. The template is used in fast
retrieval of just those column values MySQL needs in its processing. */
static
void
build_template(
/*===========*/
row_prebuilt_t* prebuilt, /* in: prebuilt struct */
THD* thd, /* in: current user thread, used
only if templ_type is
ROW_MYSQL_REC_FIELDS */
TABLE* table, /* in: MySQL table */
ulint templ_type) /* in: ROW_MYSQL_WHOLE_ROW or
ROW_MYSQL_REC_FIELDS */
{
dict_index_t* index;
dict_index_t* clust_index;
mysql_row_templ_t* templ;
Field* field;
ulint n_fields;
ulint n_requested_fields = 0;
ibool fetch_all_in_key = FALSE;
ibool fetch_primary_key_cols = FALSE;
ulint i;
/* byte offset of the end of last requested column */
ulint mysql_prefix_len = 0;
if (prebuilt->select_lock_type == LOCK_X) {
/* We always retrieve the whole clustered index record if we
use exclusive row level locks, for example, if the read is
done in an UPDATE statement. */
templ_type = ROW_MYSQL_WHOLE_ROW;
}
if (templ_type == ROW_MYSQL_REC_FIELDS) {
if (prebuilt->hint_need_to_fetch_extra_cols
== ROW_RETRIEVE_ALL_COLS) {
/* We know we must at least fetch all columns in the key, or
all columns in the table */
if (prebuilt->read_just_key) {
/* MySQL has instructed us that it is enough to
fetch the columns in the key; looks like MySQL
can set this flag also when there is only a
prefix of the column in the key: in that case we
retrieve the whole column from the clustered
index */
fetch_all_in_key = TRUE;
} else {
templ_type = ROW_MYSQL_WHOLE_ROW;
}
} else if (prebuilt->hint_need_to_fetch_extra_cols
== ROW_RETRIEVE_PRIMARY_KEY) {
/* We must at least fetch all primary key cols. Note that if
the clustered index was internally generated by InnoDB on the
row id (no primary key was defined), then
row_search_for_mysql() will always retrieve the row id to a
special buffer in the prebuilt struct. */
fetch_primary_key_cols = TRUE;
}
}
clust_index = dict_table_get_first_index_noninline(prebuilt->table);
if (templ_type == ROW_MYSQL_REC_FIELDS) {
index = prebuilt->index;
} else {
index = clust_index;
}
if (index == clust_index) {
prebuilt->need_to_access_clustered = TRUE;
} else {
prebuilt->need_to_access_clustered = FALSE;
/* Below we check column by column if we need to access
the clustered index */
}
n_fields = (ulint)table->s->fields; /* number of columns */
if (!prebuilt->mysql_template) {
prebuilt->mysql_template = (mysql_row_templ_t*)
mem_alloc_noninline(
n_fields * sizeof(mysql_row_templ_t));
}
prebuilt->template_type = templ_type;
prebuilt->null_bitmap_len = table->s->null_bytes;
prebuilt->templ_contains_blob = FALSE;
/* Note that in InnoDB, i is the column number. MySQL calls columns
'fields'. */
for (i = 0; i < n_fields; i++) {
templ = prebuilt->mysql_template + n_requested_fields;
field = table->field[i];
if (UNIV_LIKELY(templ_type == ROW_MYSQL_REC_FIELDS)) {
/* Decide which columns we should fetch
and which we can skip. */
register const ibool index_contains_field =
dict_index_contains_col_or_prefix(index, i);
if (!index_contains_field && prebuilt->read_just_key) {
/* If this is a 'key read', we do not need
columns that are not in the key */
goto skip_field;
}
if (index_contains_field && fetch_all_in_key) {
/* This field is needed in the query */
goto include_field;
}
if (thd->query_id == field->query_id) {
/* This field is needed in the query */
goto include_field;
}
if (fetch_primary_key_cols
&& dict_table_col_in_clustered_key(index->table,
i)) {
/* This field is needed in the query */
goto include_field;
}
/* This field is not needed in the query, skip it */
goto skip_field;
}
include_field:
n_requested_fields++;
templ->col_no = i;
if (index == clust_index) {
templ->rec_field_no = (index->table->cols + i)
->clust_pos;
} else {
templ->rec_field_no = dict_index_get_nth_col_pos(
index, i);
}
if (templ->rec_field_no == ULINT_UNDEFINED) {
prebuilt->need_to_access_clustered = TRUE;
}
if (field->null_ptr) {
templ->mysql_null_byte_offset =
(ulint) ((char*) field->null_ptr
- (char*) table->record[0]);
templ->mysql_null_bit_mask = (ulint) field->null_bit;
} else {
templ->mysql_null_bit_mask = 0;
}
templ->mysql_col_offset = (ulint)
get_field_offset(table, field);
templ->mysql_col_len = (ulint) field->pack_length();
if (mysql_prefix_len < templ->mysql_col_offset
+ templ->mysql_col_len) {
mysql_prefix_len = templ->mysql_col_offset
+ templ->mysql_col_len;
}
templ->type = index->table->cols[i].type.mtype;
templ->mysql_type = (ulint)field->type();
if (templ->mysql_type == DATA_MYSQL_TRUE_VARCHAR) {
templ->mysql_length_bytes = (ulint)
(((Field_varstring*)field)->length_bytes);
}
templ->charset = dtype_get_charset_coll_noninline(
index->table->cols[i].type.prtype);
templ->mbminlen = index->table->cols[i].type.mbminlen;
templ->mbmaxlen = index->table->cols[i].type.mbmaxlen;
templ->is_unsigned = index->table->cols[i].type.prtype
& DATA_UNSIGNED;
if (templ->type == DATA_BLOB) {
prebuilt->templ_contains_blob = TRUE;
}
skip_field:
;
}
prebuilt->n_template = n_requested_fields;
prebuilt->mysql_prefix_len = mysql_prefix_len;
if (index != clust_index && prebuilt->need_to_access_clustered) {
/* Change rec_field_no's to correspond to the clustered index
record */
for (i = 0; i < n_requested_fields; i++) {
templ = prebuilt->mysql_template + i;
templ->rec_field_no =
(index->table->cols + templ->col_no)->clust_pos;
}
}
}
/************************************************************************
Stores a row in an InnoDB database, to the table specified in this
handle. */
int
ha_innobase::write_row(
/*===================*/
/* out: error code */
mysql_byte* record) /* in: a row in MySQL format */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*)innobase_prebuilt;
int error;
longlong auto_inc;
longlong dummy;
ibool auto_inc_used= FALSE;
DBUG_ENTER("ha_innobase::write_row");
if (prebuilt->trx !=
(trx_t*) current_thd->ha_data[innobase_hton.slot]) {
sql_print_error("The transaction object for the table handle is at "
"%p, but for the current thread it is at %p",
prebuilt->trx,
(trx_t*) current_thd->ha_data[innobase_hton.slot]);
fputs("InnoDB: Dump of 200 bytes around prebuilt: ", stderr);
ut_print_buf(stderr, ((const byte*)prebuilt) - 100, 200);
fputs("\n"
"InnoDB: Dump of 200 bytes around transaction.all: ",
stderr);
ut_print_buf(stderr,
((byte*)(&(current_thd->ha_data[innobase_hton.slot]))) - 100,
200);
putc('\n', stderr);
ut_error;
}
statistic_increment(current_thd->status_var.ha_write_count,
&LOCK_status);
if (table->timestamp_field_type & TIMESTAMP_AUTO_SET_ON_INSERT)
table->timestamp_field->set_time();
if ((user_thd->lex->sql_command == SQLCOM_ALTER_TABLE
|| user_thd->lex->sql_command == SQLCOM_OPTIMIZE
|| user_thd->lex->sql_command == SQLCOM_CREATE_INDEX
|| user_thd->lex->sql_command == SQLCOM_DROP_INDEX)
&& num_write_row >= 10000) {
/* ALTER TABLE is COMMITted at every 10000 copied rows.
The IX table lock for the original table has to be re-issued.
As this method will be called on a temporary table where the
contents of the original table is being copied to, it is
a bit tricky to determine the source table. The cursor
position in the source table need not be adjusted after the
intermediate COMMIT, since writes by other transactions are
being blocked by a MySQL table lock TL_WRITE_ALLOW_READ. */
dict_table_t* src_table;
ulint mode;
num_write_row = 0;
/* Commit the transaction. This will release the table
locks, so they have to be acquired again. */
/* Altering an InnoDB table */
/* Get the source table. */
src_table = lock_get_src_table(
prebuilt->trx, prebuilt->table, &mode);
if (!src_table) {
no_commit:
/* Unknown situation: do not commit */
/*
ut_print_timestamp(stderr);
fprintf(stderr,
" InnoDB error: ALTER TABLE is holding lock"
" on %lu tables!\n",
prebuilt->trx->mysql_n_tables_locked);
*/
;
} else if (src_table == prebuilt->table) {
/* Source table is not in InnoDB format:
no need to re-acquire locks on it. */
/* Altering to InnoDB format */
innobase_commit(user_thd, 1);
/* Note that this transaction is still active. */
prebuilt->trx->active_trans = 1;
/* We will need an IX lock on the destination table. */
prebuilt->sql_stat_start = TRUE;
} else {
/* Ensure that there are no other table locks than
LOCK_IX and LOCK_AUTO_INC on the destination table. */
if (!lock_is_table_exclusive(prebuilt->table,
prebuilt->trx)) {
goto no_commit;
}
/* Commit the transaction. This will release the table
locks, so they have to be acquired again. */
innobase_commit(user_thd, 1);
/* Note that this transaction is still active. */
prebuilt->trx->active_trans = 1;
/* Re-acquire the table lock on the source table. */
row_lock_table_for_mysql(prebuilt, src_table, mode);
/* We will need an IX lock on the destination table. */
prebuilt->sql_stat_start = TRUE;
}
}
num_write_row++;
if (last_query_id != user_thd->query_id) {
prebuilt->sql_stat_start = TRUE;
last_query_id = user_thd->query_id;
innobase_release_stat_resources(prebuilt->trx);
}
if (table->next_number_field && record == table->record[0]) {
/* This is the case where the table has an
auto-increment column */
/* Initialize the auto-inc counter if it has not been
initialized yet */
if (0 == dict_table_autoinc_peek(prebuilt->table)) {
/* This call initializes the counter */
error = innobase_read_and_init_auto_inc(&dummy);
if (error) {
/* Deadlock or lock wait timeout */
goto func_exit;
}
/* We have to set sql_stat_start to TRUE because
the above call probably has called a select, and
has reset that flag; row_insert_for_mysql has to
know to set the IX intention lock on the table,
something it only does at the start of each
statement */
prebuilt->sql_stat_start = TRUE;
}
/* We have to use the transactional lock mechanism on the
auto-inc counter of the table to ensure that replication and
roll-forward of the binlog exactly imitates also the given
auto-inc values. The lock is released at each SQL statement's
end. This lock also prevents a race where two threads would
call ::get_auto_increment() simultaneously. */
error = row_lock_table_autoinc_for_mysql(prebuilt);
if (error != DB_SUCCESS) {
/* Deadlock or lock wait timeout */
error = convert_error_code_to_mysql(error, user_thd);
goto func_exit;
}
/* We must use the handler code to update the auto-increment
value to be sure that we increment it correctly. */
if ((error= update_auto_increment()))
goto func_exit;
auto_inc_used = 1;
}
if (prebuilt->mysql_template == NULL
|| prebuilt->template_type != ROW_MYSQL_WHOLE_ROW) {
/* Build the template used in converting quickly between
the two database formats */
build_template(prebuilt, NULL, table, ROW_MYSQL_WHOLE_ROW);
}
innodb_srv_conc_enter_innodb(prebuilt->trx);
error = row_insert_for_mysql((byte*) record, prebuilt);
if (error == DB_SUCCESS && auto_inc_used) {
/* Fetch the value that was set in the autoincrement field */
auto_inc = table->next_number_field->val_int();
if (auto_inc != 0) {
/* This call will update the counter according to the
value that was inserted in the table */
dict_table_autoinc_update(prebuilt->table, auto_inc);
}
}
/* A REPLACE command and LOAD DATA INFILE REPLACE handle a duplicate
key error themselves, and we must update the autoinc counter if we are
performing those statements. */
if (error == DB_DUPLICATE_KEY && auto_inc_used
&& (user_thd->lex->sql_command == SQLCOM_REPLACE
|| user_thd->lex->sql_command == SQLCOM_REPLACE_SELECT
|| (user_thd->lex->sql_command == SQLCOM_LOAD
&& user_thd->lex->duplicates == DUP_REPLACE))) {
auto_inc = table->next_number_field->val_int();
if (auto_inc != 0) {
dict_table_autoinc_update(prebuilt->table, auto_inc);
}
}
innodb_srv_conc_exit_innodb(prebuilt->trx);
error = convert_error_code_to_mysql(error, user_thd);
/* Tell InnoDB server that there might be work for
utility threads: */
func_exit:
innobase_active_small();
DBUG_RETURN(error);
}
/**************************************************************************
Checks which fields have changed in a row and stores information
of them to an update vector. */
static
int
calc_row_difference(
/*================*/
/* out: error number or 0 */
upd_t* uvect, /* in/out: update vector */
mysql_byte* old_row, /* in: old row in MySQL format */
mysql_byte* new_row, /* in: new row in MySQL format */
struct st_table* table, /* in: table in MySQL data
dictionary */
mysql_byte* upd_buff, /* in: buffer to use */
ulint buff_len, /* in: buffer length */
row_prebuilt_t* prebuilt, /* in: InnoDB prebuilt struct */
THD* thd) /* in: user thread */
{
mysql_byte* original_upd_buff = upd_buff;
Field* field;
enum_field_types field_mysql_type;
uint n_fields;
ulint o_len;
ulint n_len;
ulint col_pack_len;
byte* new_mysql_row_col;
byte* o_ptr;
byte* n_ptr;
byte* buf;
upd_field_t* ufield;
ulint col_type;
ulint n_changed = 0;
dfield_t dfield;
uint i;
n_fields = table->s->fields;
/* We use upd_buff to convert changed fields */
buf = (byte*) upd_buff;
for (i = 0; i < n_fields; i++) {
field = table->field[i];
/* if (thd->query_id != field->query_id) { */
/* TODO: check that these fields cannot have
changed! */
/* goto skip_field;
}*/
o_ptr = (byte*) old_row + get_field_offset(table, field);
n_ptr = (byte*) new_row + get_field_offset(table, field);
/* Use new_mysql_row_col and col_pack_len save the values */
new_mysql_row_col = n_ptr;
col_pack_len = field->pack_length();
o_len = col_pack_len;
n_len = col_pack_len;
/* We use o_ptr and n_ptr to dig up the actual data for
comparison. */
field_mysql_type = field->type();
col_type = prebuilt->table->cols[i].type.mtype;
switch (col_type) {
case DATA_BLOB:
o_ptr = row_mysql_read_blob_ref(&o_len, o_ptr, o_len);
n_ptr = row_mysql_read_blob_ref(&n_len, n_ptr, n_len);
break;
case DATA_VARCHAR:
case DATA_BINARY:
case DATA_VARMYSQL:
if (field_mysql_type == MYSQL_TYPE_VARCHAR) {
/* This is a >= 5.0.3 type true VARCHAR where
the real payload data length is stored in
1 or 2 bytes */
o_ptr = row_mysql_read_true_varchar(
&o_len, o_ptr,
(ulint)
(((Field_varstring*)field)->length_bytes));
n_ptr = row_mysql_read_true_varchar(
&n_len, n_ptr,
(ulint)
(((Field_varstring*)field)->length_bytes));
}
break;
default:
;
}
if (field->null_ptr) {
if (field_in_record_is_null(table, field,
(char*) old_row)) {
o_len = UNIV_SQL_NULL;
}
if (field_in_record_is_null(table, field,
(char*) new_row)) {
n_len = UNIV_SQL_NULL;
}
}
if (o_len != n_len || (o_len != UNIV_SQL_NULL &&
0 != memcmp(o_ptr, n_ptr, o_len))) {
/* The field has changed */
ufield = uvect->fields + n_changed;
/* Let us use a dummy dfield to make the conversion
from the MySQL column format to the InnoDB format */
dfield.type = (prebuilt->table->cols + i)->type;
if (n_len != UNIV_SQL_NULL) {
buf = row_mysql_store_col_in_innobase_format(
&dfield,
(byte*)buf,
TRUE,
new_mysql_row_col,
col_pack_len,
prebuilt->table->comp);
ufield->new_val.data = dfield.data;
ufield->new_val.len = dfield.len;
} else {
ufield->new_val.data = NULL;
ufield->new_val.len = UNIV_SQL_NULL;
}
ufield->exp = NULL;
ufield->field_no = prebuilt->table->cols[i].clust_pos;
n_changed++;
}
}
uvect->n_fields = n_changed;
uvect->info_bits = 0;
ut_a(buf <= (byte*)original_upd_buff + buff_len);
return(0);
}
/**************************************************************************
Updates a row given as a parameter to a new value. Note that we are given
whole rows, not just the fields which are updated: this incurs some
overhead for CPU when we check which fields are actually updated.
TODO: currently InnoDB does not prevent the 'Halloween problem':
in a searched update a single row can get updated several times
if its index columns are updated! */
int
ha_innobase::update_row(
/*====================*/
/* out: error number or 0 */
const mysql_byte* old_row,/* in: old row in MySQL format */
mysql_byte* new_row)/* in: new row in MySQL format */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
upd_t* uvect;
int error = 0;
DBUG_ENTER("ha_innobase::update_row");
ut_ad(prebuilt->trx ==
(trx_t*) current_thd->ha_data[innobase_hton.slot]);
if (table->timestamp_field_type & TIMESTAMP_AUTO_SET_ON_UPDATE)
table->timestamp_field->set_time();
if (last_query_id != user_thd->query_id) {
prebuilt->sql_stat_start = TRUE;
last_query_id = user_thd->query_id;
innobase_release_stat_resources(prebuilt->trx);
}
if (prebuilt->upd_node) {
uvect = prebuilt->upd_node->update;
} else {
uvect = row_get_prebuilt_update_vector(prebuilt);
}
/* Build an update vector from the modified fields in the rows
(uses upd_buff of the handle) */
calc_row_difference(uvect, (mysql_byte*) old_row, new_row, table,
upd_buff, (ulint)upd_and_key_val_buff_len,
prebuilt, user_thd);
/* This is not a delete */
prebuilt->upd_node->is_delete = FALSE;
assert(prebuilt->template_type == ROW_MYSQL_WHOLE_ROW);
innodb_srv_conc_enter_innodb(prebuilt->trx);
error = row_update_for_mysql((byte*) old_row, prebuilt);
/* We need to do some special AUTOINC handling for the following case:
INSERT INTO t (c1,c2) VALUES(x,y) ON DUPLICATE KEY UPDATE ...
We need to use the AUTOINC counter that was actually used by
MySQL in the UPDATE statement, which can be different from the
value used in the INSERT statement.*/
if (error == DB_SUCCESS
&& table->next_number_field && new_row == table->record[0]
&& user_thd->lex->sql_command == SQLCOM_INSERT
&& user_thd->lex->duplicates == DUP_UPDATE) {
longlong auto_inc;
auto_inc = table->next_number_field->val_int();
if (auto_inc != 0) {
dict_table_autoinc_update(prebuilt->table, auto_inc);
}
}
innodb_srv_conc_exit_innodb(prebuilt->trx);
error = convert_error_code_to_mysql(error, user_thd);
/* Tell InnoDB server that there might be work for
utility threads: */
innobase_active_small();
DBUG_RETURN(error);
}
/**************************************************************************
Deletes a row given as the parameter. */
int
ha_innobase::delete_row(
/*====================*/
/* out: error number or 0 */
const mysql_byte* record) /* in: a row in MySQL format */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
int error = 0;
DBUG_ENTER("ha_innobase::delete_row");
ut_ad(prebuilt->trx ==
(trx_t*) current_thd->ha_data[innobase_hton.slot]);
if (last_query_id != user_thd->query_id) {
prebuilt->sql_stat_start = TRUE;
last_query_id = user_thd->query_id;
innobase_release_stat_resources(prebuilt->trx);
}
if (!prebuilt->upd_node) {
row_get_prebuilt_update_vector(prebuilt);
}
/* This is a delete */
prebuilt->upd_node->is_delete = TRUE;
innodb_srv_conc_enter_innodb(prebuilt->trx);
error = row_update_for_mysql((byte*) record, prebuilt);
innodb_srv_conc_exit_innodb(prebuilt->trx);
error = convert_error_code_to_mysql(error, user_thd);
/* Tell the InnoDB server that there might be work for
utility threads: */
innobase_active_small();
DBUG_RETURN(error);
}
/**************************************************************************
Removes a new lock set on a row. This method does nothing unless the
option innodb_locks_unsafe_for_binlog is set.*/
void
ha_innobase::unlock_row(void)
/*=========================*/
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
DBUG_ENTER("ha_innobase::unlock_row");
if (last_query_id != user_thd->query_id) {
ut_print_timestamp(stderr);
sql_print_error("last_query_id is %lu != user_thd_query_id is "
"%lu", (ulong) last_query_id,
(ulong) user_thd->query_id);
mem_analyze_corruption((byte *) prebuilt->trx);
ut_error;
}
/* Consistent read does not take any locks, thus there is
nothing to unlock. */
if (prebuilt->select_lock_type == LOCK_NONE) {
DBUG_VOID_RETURN;
}
if (srv_locks_unsafe_for_binlog) {
row_unlock_for_mysql(prebuilt, FALSE);
}
DBUG_VOID_RETURN;
}
/**********************************************************************
Initializes a handle to use an index. */
int
ha_innobase::index_init(
/*====================*/
/* out: 0 or error number */
uint keynr) /* in: key (index) number */
{
int error = 0;
DBUG_ENTER("index_init");
error = change_active_index(keynr);
DBUG_RETURN(error);
}
/**********************************************************************
Currently does nothing. */
int
ha_innobase::index_end(void)
/*========================*/
{
int error = 0;
DBUG_ENTER("index_end");
active_index=MAX_KEY;
DBUG_RETURN(error);
}
/*************************************************************************
Converts a search mode flag understood by MySQL to a flag understood
by InnoDB. */
inline
ulint
convert_search_mode_to_innobase(
/*============================*/
enum ha_rkey_function find_flag)
{
switch (find_flag) {
case HA_READ_KEY_EXACT: return(PAGE_CUR_GE);
/* the above does not require the index to be UNIQUE */
case HA_READ_KEY_OR_NEXT: return(PAGE_CUR_GE);
case HA_READ_KEY_OR_PREV: return(PAGE_CUR_LE);
case HA_READ_AFTER_KEY: return(PAGE_CUR_G);
case HA_READ_BEFORE_KEY: return(PAGE_CUR_L);
case HA_READ_PREFIX: return(PAGE_CUR_GE);
case HA_READ_PREFIX_LAST: return(PAGE_CUR_LE);
case HA_READ_PREFIX_LAST_OR_PREV:return(PAGE_CUR_LE);
/* In MySQL-4.0 HA_READ_PREFIX and HA_READ_PREFIX_LAST always
pass a complete-field prefix of a key value as the search
tuple. I.e., it is not allowed that the last field would
just contain n first bytes of the full field value.
MySQL uses a 'padding' trick to convert LIKE 'abc%'
type queries so that it can use as a search tuple
a complete-field-prefix of a key value. Thus, the InnoDB
search mode PAGE_CUR_LE_OR_EXTENDS is never used.
TODO: when/if MySQL starts to use also partial-field
prefixes, we have to deal with stripping of spaces
and comparison of non-latin1 char type fields in
innobase_mysql_cmp() to get PAGE_CUR_LE_OR_EXTENDS to
work correctly. */
case HA_READ_MBR_CONTAIN:
case HA_READ_MBR_INTERSECT:
case HA_READ_MBR_WITHIN:
case HA_READ_MBR_DISJOINT:
case HA_READ_MBR_EQUAL:
return(PAGE_CUR_UNSUPP);
/* do not use "default:" in order to produce a gcc warning:
enumeration value '...' not handled in switch
(if -Wswitch or -Wall is used)
*/
}
my_error(ER_CHECK_NOT_IMPLEMENTED, MYF(0), "this functionality");
return(PAGE_CUR_UNSUPP);
}
/*
BACKGROUND INFO: HOW A SELECT SQL QUERY IS EXECUTED
---------------------------------------------------
The following does not cover all the details, but explains how we determine
the start of a new SQL statement, and what is associated with it.
For each table in the database the MySQL interpreter may have several
table handle instances in use, also in a single SQL query. For each table
handle instance there is an InnoDB 'prebuilt' struct which contains most
of the InnoDB data associated with this table handle instance.
A) if the user has not explicitly set any MySQL table level locks:
1) MySQL calls ::external_lock to set an 'intention' table level lock on
the table of the handle instance. There we set
prebuilt->sql_stat_start = TRUE. The flag sql_stat_start should be set
true if we are taking this table handle instance to use in a new SQL
statement issued by the user. We also increment trx->n_mysql_tables_in_use.
2) If prebuilt->sql_stat_start == TRUE we 'pre-compile' the MySQL search
instructions to prebuilt->template of the table handle instance in
::index_read. The template is used to save CPU time in large joins.
3) In row_search_for_mysql, if prebuilt->sql_stat_start is true, we
allocate a new consistent read view for the trx if it does not yet have one,
or in the case of a locking read, set an InnoDB 'intention' table level
lock on the table.
4) We do the SELECT. MySQL may repeatedly call ::index_read for the
same table handle instance, if it is a join.
5) When the SELECT ends, MySQL removes its intention table level locks
in ::external_lock. When trx->n_mysql_tables_in_use drops to zero,
(a) we execute a COMMIT there if the autocommit is on,
(b) we also release possible 'SQL statement level resources' InnoDB may
have for this SQL statement. The MySQL interpreter does NOT execute
autocommit for pure read transactions, though it should. That is why the
table handler in that case has to execute the COMMIT in ::external_lock.
B) If the user has explicitly set MySQL table level locks, then MySQL
does NOT call ::external_lock at the start of the statement. To determine
when we are at the start of a new SQL statement we at the start of
::index_read also compare the query id to the latest query id where the
table handle instance was used. If it has changed, we know we are at the
start of a new SQL statement. Since the query id can theoretically
overwrap, we use this test only as a secondary way of determining the
start of a new SQL statement. */
/**************************************************************************
Positions an index cursor to the index specified in the handle. Fetches the
row if any. */
int
ha_innobase::index_read(
/*====================*/
/* out: 0, HA_ERR_KEY_NOT_FOUND,
or error number */
mysql_byte* buf, /* in/out: buffer for the returned
row */
const mysql_byte* key_ptr,/* in: key value; if this is NULL
we position the cursor at the
start or end of index; this can
also contain an InnoDB row id, in
which case key_len is the InnoDB
row id length; the key value can
also be a prefix of a full key value,
and the last column can be a prefix
of a full column */
uint key_len,/* in: key value length */
enum ha_rkey_function find_flag)/* in: search flags from my_base.h */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
ulint mode;
dict_index_t* index;
ulint match_mode = 0;
int error;
ulint ret;
DBUG_ENTER("index_read");
ut_ad(prebuilt->trx ==
(trx_t*) current_thd->ha_data[innobase_hton.slot]);
statistic_increment(current_thd->status_var.ha_read_key_count,
&LOCK_status);
if (last_query_id != user_thd->query_id) {
prebuilt->sql_stat_start = TRUE;
last_query_id = user_thd->query_id;
innobase_release_stat_resources(prebuilt->trx);
}
index = prebuilt->index;
/* Note that if the index for which the search template is built is not
necessarily prebuilt->index, but can also be the clustered index */
if (prebuilt->sql_stat_start) {
build_template(prebuilt, user_thd, table,
ROW_MYSQL_REC_FIELDS);
}
if (key_ptr) {
/* Convert the search key value to InnoDB format into
prebuilt->search_tuple */
row_sel_convert_mysql_key_to_innobase(prebuilt->search_tuple,
(byte*) key_val_buff,
(ulint)upd_and_key_val_buff_len,
index,
(byte*) key_ptr,
(ulint) key_len, prebuilt->trx);
} else {
/* We position the cursor to the last or the first entry
in the index */
dtuple_set_n_fields(prebuilt->search_tuple, 0);
}
mode = convert_search_mode_to_innobase(find_flag);
match_mode = 0;
if (find_flag == HA_READ_KEY_EXACT) {
match_mode = ROW_SEL_EXACT;
} else if (find_flag == HA_READ_PREFIX
|| find_flag == HA_READ_PREFIX_LAST) {
match_mode = ROW_SEL_EXACT_PREFIX;
}
last_match_mode = (uint) match_mode;
if (mode != PAGE_CUR_UNSUPP) {
innodb_srv_conc_enter_innodb(prebuilt->trx);
ret = row_search_for_mysql((byte*) buf, mode, prebuilt,
match_mode, 0);
innodb_srv_conc_exit_innodb(prebuilt->trx);
} else {
ret = DB_UNSUPPORTED;
}
if (ret == DB_SUCCESS) {
error = 0;
table->status = 0;
} else if (ret == DB_RECORD_NOT_FOUND) {
error = HA_ERR_KEY_NOT_FOUND;
table->status = STATUS_NOT_FOUND;
} else if (ret == DB_END_OF_INDEX) {
error = HA_ERR_KEY_NOT_FOUND;
table->status = STATUS_NOT_FOUND;
} else {
error = convert_error_code_to_mysql((int) ret, user_thd);
table->status = STATUS_NOT_FOUND;
}
DBUG_RETURN(error);
}
/***********************************************************************
The following functions works like index_read, but it find the last
row with the current key value or prefix. */
int
ha_innobase::index_read_last(
/*=========================*/
/* out: 0, HA_ERR_KEY_NOT_FOUND, or an
error code */
mysql_byte* buf, /* out: fetched row */
const mysql_byte* key_ptr, /* in: key value, or a prefix of a full
key value */
uint key_len) /* in: length of the key val or prefix
in bytes */
{
return(index_read(buf, key_ptr, key_len, HA_READ_PREFIX_LAST));
}
/************************************************************************
Changes the active index of a handle. */
int
ha_innobase::change_active_index(
/*=============================*/
/* out: 0 or error code */
uint keynr) /* in: use this index; MAX_KEY means always clustered
index, even if it was internally generated by
InnoDB */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
KEY* key=0;
statistic_increment(current_thd->status_var.ha_read_key_count,
&LOCK_status);
DBUG_ENTER("change_active_index");
ut_ad(user_thd == current_thd);
ut_ad(prebuilt->trx ==
(trx_t*) current_thd->ha_data[innobase_hton.slot]);
active_index = keynr;
if (keynr != MAX_KEY && table->s->keys > 0) {
key = table->key_info + active_index;
prebuilt->index = dict_table_get_index_noninline(
prebuilt->table,
key->name);
} else {
prebuilt->index = dict_table_get_first_index_noninline(
prebuilt->table);
}
if (!prebuilt->index) {
sql_print_error("Innodb could not find key n:o %u with name %s "
"from dict cache for table %s",
keynr, key ? key->name : "NULL",
prebuilt->table->name);
DBUG_RETURN(1);
}
assert(prebuilt->search_tuple != 0);
dtuple_set_n_fields(prebuilt->search_tuple, prebuilt->index->n_fields);
dict_index_copy_types(prebuilt->search_tuple, prebuilt->index,
prebuilt->index->n_fields);
/* MySQL changes the active index for a handle also during some
queries, for example SELECT MAX(a), SUM(a) first retrieves the MAX()
and then calculates the sum. Previously we played safe and used
the flag ROW_MYSQL_WHOLE_ROW below, but that caused unnecessary
copying. Starting from MySQL-4.1 we use a more efficient flag here. */
build_template(prebuilt, user_thd, table, ROW_MYSQL_REC_FIELDS);
DBUG_RETURN(0);
}
/**************************************************************************
Positions an index cursor to the index specified in keynr. Fetches the
row if any. */
/* ??? This is only used to read whole keys ??? */
int
ha_innobase::index_read_idx(
/*========================*/
/* out: error number or 0 */
mysql_byte* buf, /* in/out: buffer for the returned
row */
uint keynr, /* in: use this index */
const mysql_byte* key, /* in: key value; if this is NULL
we position the cursor at the
start or end of index */
uint key_len, /* in: key value length */
enum ha_rkey_function find_flag)/* in: search flags from my_base.h */
{
if (change_active_index(keynr)) {
return(1);
}
return(index_read(buf, key, key_len, find_flag));
}
/***************************************************************************
Reads the next or previous row from a cursor, which must have previously been
positioned using index_read. */
int
ha_innobase::general_fetch(
/*=======================*/
/* out: 0, HA_ERR_END_OF_FILE, or error
number */
mysql_byte* buf, /* in/out: buffer for next row in MySQL
format */
uint direction, /* in: ROW_SEL_NEXT or ROW_SEL_PREV */
uint match_mode) /* in: 0, ROW_SEL_EXACT, or
ROW_SEL_EXACT_PREFIX */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
ulint ret;
int error = 0;
DBUG_ENTER("general_fetch");
ut_ad(prebuilt->trx ==
(trx_t*) current_thd->ha_data[innobase_hton.slot]);
innodb_srv_conc_enter_innodb(prebuilt->trx);
ret = row_search_for_mysql((byte*)buf, 0, prebuilt, match_mode,
direction);
innodb_srv_conc_exit_innodb(prebuilt->trx);
if (ret == DB_SUCCESS) {
error = 0;
table->status = 0;
} else if (ret == DB_RECORD_NOT_FOUND) {
error = HA_ERR_END_OF_FILE;
table->status = STATUS_NOT_FOUND;
} else if (ret == DB_END_OF_INDEX) {
error = HA_ERR_END_OF_FILE;
table->status = STATUS_NOT_FOUND;
} else {
error = convert_error_code_to_mysql((int) ret, user_thd);
table->status = STATUS_NOT_FOUND;
}
DBUG_RETURN(error);
}
/***************************************************************************
Reads the next row from a cursor, which must have previously been
positioned using index_read. */
int
ha_innobase::index_next(
/*====================*/
/* out: 0, HA_ERR_END_OF_FILE, or error
number */
mysql_byte* buf) /* in/out: buffer for next row in MySQL
format */
{
statistic_increment(current_thd->status_var.ha_read_next_count,
&LOCK_status);
return(general_fetch(buf, ROW_SEL_NEXT, 0));
}
/***********************************************************************
Reads the next row matching to the key value given as the parameter. */
int
ha_innobase::index_next_same(
/*=========================*/
/* out: 0, HA_ERR_END_OF_FILE, or error
number */
mysql_byte* buf, /* in/out: buffer for the row */
const mysql_byte* key, /* in: key value */
uint keylen) /* in: key value length */
{
statistic_increment(current_thd->status_var.ha_read_next_count,
&LOCK_status);
return(general_fetch(buf, ROW_SEL_NEXT, last_match_mode));
}
/***************************************************************************
Reads the previous row from a cursor, which must have previously been
positioned using index_read. */
int
ha_innobase::index_prev(
/*====================*/
/* out: 0, HA_ERR_END_OF_FILE, or error
number */
mysql_byte* buf) /* in/out: buffer for previous row in MySQL
format */
{
statistic_increment(current_thd->status_var.ha_read_prev_count,
&LOCK_status);
return(general_fetch(buf, ROW_SEL_PREV, 0));
}
/************************************************************************
Positions a cursor on the first record in an index and reads the
corresponding row to buf. */
int
ha_innobase::index_first(
/*=====================*/
/* out: 0, HA_ERR_END_OF_FILE,
or error code */
mysql_byte* buf) /* in/out: buffer for the row */
{
int error;
DBUG_ENTER("index_first");
statistic_increment(current_thd->status_var.ha_read_first_count,
&LOCK_status);
error = index_read(buf, NULL, 0, HA_READ_AFTER_KEY);
/* MySQL does not seem to allow this to return HA_ERR_KEY_NOT_FOUND */
if (error == HA_ERR_KEY_NOT_FOUND) {
error = HA_ERR_END_OF_FILE;
}
DBUG_RETURN(error);
}
/************************************************************************
Positions a cursor on the last record in an index and reads the
corresponding row to buf. */
int
ha_innobase::index_last(
/*====================*/
/* out: 0, HA_ERR_END_OF_FILE, or error code */
mysql_byte* buf) /* in/out: buffer for the row */
{
int error;
DBUG_ENTER("index_last");
statistic_increment(current_thd->status_var.ha_read_last_count,
&LOCK_status);
error = index_read(buf, NULL, 0, HA_READ_BEFORE_KEY);
/* MySQL does not seem to allow this to return HA_ERR_KEY_NOT_FOUND */
if (error == HA_ERR_KEY_NOT_FOUND) {
error = HA_ERR_END_OF_FILE;
}
DBUG_RETURN(error);
}
/********************************************************************
Initialize a table scan. */
int
ha_innobase::rnd_init(
/*==================*/
/* out: 0 or error number */
bool scan) /* in: ???????? */
{
int err;
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
/* Store the active index value so that we can restore the original
value after a scan */
if (prebuilt->clust_index_was_generated) {
err = change_active_index(MAX_KEY);
} else {
err = change_active_index(primary_key);
}
start_of_scan = 1;
return(err);
}
/*********************************************************************
Ends a table scan. */
int
ha_innobase::rnd_end(void)
/*======================*/
/* out: 0 or error number */
{
return(index_end());
}
/*********************************************************************
Reads the next row in a table scan (also used to read the FIRST row
in a table scan). */
int
ha_innobase::rnd_next(
/*==================*/
/* out: 0, HA_ERR_END_OF_FILE, or error number */
mysql_byte* buf)/* in/out: returns the row in this buffer,
in MySQL format */
{
int error;
DBUG_ENTER("rnd_next");
statistic_increment(current_thd->status_var.ha_read_rnd_next_count,
&LOCK_status);
if (start_of_scan) {
error = index_first(buf);
if (error == HA_ERR_KEY_NOT_FOUND) {
error = HA_ERR_END_OF_FILE;
}
start_of_scan = 0;
} else {
error = general_fetch(buf, ROW_SEL_NEXT, 0);
}
DBUG_RETURN(error);
}
/**************************************************************************
Fetches a row from the table based on a row reference. */
int
ha_innobase::rnd_pos(
/*=================*/
/* out: 0, HA_ERR_KEY_NOT_FOUND,
or error code */
mysql_byte* buf, /* in/out: buffer for the row */
mysql_byte* pos) /* in: primary key value of the row in the
MySQL format, or the row id if the clustered
index was internally generated by InnoDB;
the length of data in pos has to be
ref_length */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
int error;
uint keynr = active_index;
DBUG_ENTER("rnd_pos");
DBUG_DUMP("key", (uchar*) pos, ref_length);
statistic_increment(current_thd->status_var.ha_read_rnd_count,
&LOCK_status);
ut_ad(prebuilt->trx ==
(trx_t*) current_thd->ha_data[innobase_hton.slot]);
if (prebuilt->clust_index_was_generated) {
/* No primary key was defined for the table and we
generated the clustered index from the row id: the
row reference is the row id, not any key value
that MySQL knows of */
error = change_active_index(MAX_KEY);
} else {
error = change_active_index(primary_key);
}
if (error) {
DBUG_PRINT("error", ("Got error: %d", error));
DBUG_RETURN(error);
}
/* Note that we assume the length of the row reference is fixed
for the table, and it is == ref_length */
error = index_read(buf, pos, ref_length, HA_READ_KEY_EXACT);
if (error) {
DBUG_PRINT("error", ("Got error: %d", error));
}
change_active_index(keynr);
DBUG_RETURN(error);
}
/*************************************************************************
Stores a reference to the current row to 'ref' field of the handle. Note
that in the case where we have generated the clustered index for the
table, the function parameter is illogical: we MUST ASSUME that 'record'
is the current 'position' of the handle, because if row ref is actually
the row id internally generated in InnoDB, then 'record' does not contain
it. We just guess that the row id must be for the record where the handle
was positioned the last time. */
void
ha_innobase::position(
/*==================*/
const mysql_byte* record) /* in: row in MySQL format */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
uint len;
ut_ad(prebuilt->trx ==
(trx_t*) current_thd->ha_data[innobase_hton.slot]);
if (prebuilt->clust_index_was_generated) {
/* No primary key was defined for the table and we
generated the clustered index from row id: the
row reference will be the row id, not any key value
that MySQL knows of */
len = DATA_ROW_ID_LEN;
memcpy(ref, prebuilt->row_id, len);
} else {
len = store_key_val_for_row(primary_key, (char*)ref,
ref_length, record);
}
/* We assume that the 'ref' value len is always fixed for the same
table. */
if (len != ref_length) {
sql_print_error("Stored ref len is %lu, but table ref len is %lu",
(ulong) len, (ulong) ref_length);
}
}
/*********************************************************************
Creates a table definition to an InnoDB database. */
static
int
create_table_def(
/*=============*/
trx_t* trx, /* in: InnoDB transaction handle */
TABLE* form, /* in: information on table
columns and indexes */
const char* table_name, /* in: table name */
const char* path_of_temp_table,/* in: if this is a table explicitly
created by the user with the
TEMPORARY keyword, then this
parameter is the dir path where the
table should be placed if we create
an .ibd file for it (no .ibd extension
in the path, though); otherwise this
is NULL */
ibool comp) /* in: TRUE=compact record format */
{
Field* field;
dict_table_t* table;
ulint n_cols;
int error;
ulint col_type;
ulint col_len;
ulint nulls_allowed;
ulint unsigned_type;
ulint binary_type;
ulint long_true_varchar;
ulint charset_no;
ulint i;
DBUG_ENTER("create_table_def");
DBUG_PRINT("enter", ("table_name: %s", table_name));
n_cols = form->s->fields;
/* We pass 0 as the space id, and determine at a lower level the space
id where to store the table */
table = dict_mem_table_create(table_name, 0, n_cols, comp);
if (path_of_temp_table) {
table->dir_path_of_temp_table =
mem_heap_strdup(table->heap, path_of_temp_table);
}
for (i = 0; i < n_cols; i++) {
field = form->field[i];
col_type = get_innobase_type_from_mysql_type(&unsigned_type,
field);
if (field->null_ptr) {
nulls_allowed = 0;
} else {
nulls_allowed = DATA_NOT_NULL;
}
if (field->binary()) {
binary_type = DATA_BINARY_TYPE;
} else {
binary_type = 0;
}
charset_no = 0;
if (dtype_is_string_type(col_type)) {
charset_no = (ulint)field->charset()->number;
ut_a(charset_no < 256); /* in data0type.h we assume
that the number fits in one
byte */
}
ut_a(field->type() < 256); /* we assume in dtype_form_prtype()
that this fits in one byte */
col_len = field->pack_length();
/* The MySQL pack length contains 1 or 2 bytes length field
for a true VARCHAR. Let us subtract that, so that the InnoDB
column length in the InnoDB data dictionary is the real
maximum byte length of the actual data. */
long_true_varchar = 0;
if (field->type() == MYSQL_TYPE_VARCHAR) {
col_len -= ((Field_varstring*)field)->length_bytes;
if (((Field_varstring*)field)->length_bytes == 2) {
long_true_varchar = DATA_LONG_TRUE_VARCHAR;
}
}
dict_mem_table_add_col(table,
(char*) field->field_name,
col_type,
dtype_form_prtype(
(ulint)field->type()
| nulls_allowed | unsigned_type
| binary_type | long_true_varchar,
charset_no),
col_len,
0);
}
error = row_create_table_for_mysql(table, trx);
error = convert_error_code_to_mysql(error, NULL);
DBUG_RETURN(error);
}
/*********************************************************************
Creates an index in an InnoDB database. */
static
int
create_index(
/*=========*/
trx_t* trx, /* in: InnoDB transaction handle */
TABLE* form, /* in: information on table
columns and indexes */
const char* table_name, /* in: table name */
uint key_num) /* in: index number */
{
Field* field;
dict_index_t* index;
int error;
ulint n_fields;
KEY* key;
KEY_PART_INFO* key_part;
ulint ind_type;
ulint col_type;
ulint prefix_len;
ulint is_unsigned;
ulint i;
ulint j;
ulint* field_lengths;
DBUG_ENTER("create_index");
key = form->key_info + key_num;
n_fields = key->key_parts;
/* Assert that "GEN_CLUST_INDEX" cannot be used as non-primary index */
ut_a(innobase_strcasecmp(key->name, innobase_index_reserve_name) != 0);
ind_type = 0;
if (key_num == form->s->primary_key) {
ind_type = ind_type | DICT_CLUSTERED;
}
if (key->flags & HA_NOSAME ) {
ind_type = ind_type | DICT_UNIQUE;
}
/* We pass 0 as the space id, and determine at a lower level the space
id where to store the table */
index = dict_mem_index_create((char*) table_name, key->name, 0,
ind_type, n_fields);
field_lengths = (ulint*) my_malloc(sizeof(ulint) * n_fields,
MYF(MY_FAE));
for (i = 0; i < n_fields; i++) {
key_part = key->key_part + i;
/* (The flag HA_PART_KEY_SEG denotes in MySQL a column prefix
field in an index: we only store a specified number of first
bytes of the column to the index field.) The flag does not
seem to be properly set by MySQL. Let us fall back on testing
the length of the key part versus the column. */
field = NULL;
for (j = 0; j < form->s->fields; j++) {
field = form->field[j];
if (0 == innobase_strcasecmp(
field->field_name,
key_part->field->field_name)) {
/* Found the corresponding column */
break;
}
}
ut_a(j < form->s->fields);
col_type = get_innobase_type_from_mysql_type(
&is_unsigned, key_part->field);
if (DATA_BLOB == col_type
|| (key_part->length < field->pack_length()
&& field->type() != MYSQL_TYPE_VARCHAR)
|| (field->type() == MYSQL_TYPE_VARCHAR
&& key_part->length < field->pack_length()
- ((Field_varstring*)field)->length_bytes)) {
prefix_len = key_part->length;
if (col_type == DATA_INT
|| col_type == DATA_FLOAT
|| col_type == DATA_DOUBLE
|| col_type == DATA_DECIMAL) {
sql_print_error("MySQL is trying to create a column "
"prefix index field, on an "
"inappropriate data type. Table "
"name %s, column name %s.",
table_name,
key_part->field->field_name);
prefix_len = 0;
}
} else {
prefix_len = 0;
}
field_lengths[i] = key_part->length;
/* We assume all fields should be sorted in ascending
order, hence the '0': */
dict_mem_index_add_field(index,
(char*) key_part->field->field_name,
0, prefix_len);
}
/* Even though we've defined max_supported_key_part_length, we
still do our own checking using field_lengths to be absolutely
sure we don't create too long indexes. */
error = row_create_index_for_mysql(index, trx, field_lengths);
error = convert_error_code_to_mysql(error, NULL);
my_free((gptr) field_lengths, MYF(0));
DBUG_RETURN(error);
}
/*********************************************************************
Creates an index to an InnoDB table when the user has defined no
primary index. */
static
int
create_clustered_index_when_no_primary(
/*===================================*/
trx_t* trx, /* in: InnoDB transaction handle */
const char* table_name) /* in: table name */
{
dict_index_t* index;
int error;
/* We pass 0 as the space id, and determine at a lower level the space
id where to store the table */
index = dict_mem_index_create(table_name,
innobase_index_reserve_name,
0, DICT_CLUSTERED, 0);
error = row_create_index_for_mysql(index, trx, NULL);
error = convert_error_code_to_mysql(error, NULL);
return(error);
}
/*********************************************************************
Update create_info. Used in SHOW CREATE TABLE et al. */
void
ha_innobase::update_create_info(
/*============================*/
HA_CREATE_INFO* create_info) /* in/out: create info */
{
if (!(create_info->used_fields & HA_CREATE_USED_AUTO)) {
ha_innobase::info(HA_STATUS_AUTO);
create_info->auto_increment_value = auto_increment_value;
}
}
/*********************************************************************
Creates a new table to an InnoDB database. */
int
ha_innobase::create(
/*================*/
/* out: error number */
const char* name, /* in: table name */
TABLE* form, /* in: information on table
columns and indexes */
HA_CREATE_INFO* create_info) /* in: more information of the
created table, contains also the
create statement string */
{
int error;
dict_table_t* innobase_table;
trx_t* parent_trx;
trx_t* trx;
int primary_key_no;
uint i;
char name2[FN_REFLEN];
char norm_name[FN_REFLEN];
THD *thd= current_thd;
ib_longlong auto_inc_value;
DBUG_ENTER("ha_innobase::create");
DBUG_ASSERT(thd != NULL);
if (form->s->fields > 1000) {
/* The limit probably should be REC_MAX_N_FIELDS - 3 = 1020,
but we play safe here */
DBUG_RETURN(HA_ERR_TO_BIG_ROW);
}
/* Get the transaction associated with the current thd, or create one
if not yet created */
parent_trx = check_trx_exists(current_thd);
/* In case MySQL calls this in the middle of a SELECT query, release
possible adaptive hash latch to avoid deadlocks of threads */
trx_search_latch_release_if_reserved(parent_trx);
trx = trx_allocate_for_mysql();
trx->mysql_thd = thd;
trx->mysql_query_str = &((*thd).query);
if (thd->options & OPTION_NO_FOREIGN_KEY_CHECKS) {
trx->check_foreigns = FALSE;
}
if (thd->options & OPTION_RELAXED_UNIQUE_CHECKS) {
trx->check_unique_secondary = FALSE;
}
if (lower_case_table_names) {
srv_lower_case_table_names = TRUE;
} else {
srv_lower_case_table_names = FALSE;
}
fn_format(name2, name, "", "", 2); // Remove the .frm extension
normalize_table_name(norm_name, name2);
/* Latch the InnoDB data dictionary exclusively so that no deadlocks
or lock waits can happen in it during a table create operation.
Drop table etc. do this latching in row0mysql.c. */
row_mysql_lock_data_dictionary(trx);
/* Look for a primary key */
primary_key_no= (table->s->primary_key != MAX_KEY ?
(int) table->s->primary_key :
-1);
/* Our function row_get_mysql_key_number_for_index assumes
the primary key is always number 0, if it exists */
DBUG_ASSERT(primary_key_no == -1 || primary_key_no == 0);
/* Check for name conflicts (with reserved name) for
any user indices to be created. */
if (innobase_index_name_is_reserved(trx, form, norm_name)) {
error = -1;
goto cleanup;
}
/* Create the table definition in InnoDB */
error = create_table_def(trx, form, norm_name,
create_info->options & HA_LEX_CREATE_TMP_TABLE ? name2 : NULL,
form->s->row_type != ROW_TYPE_REDUNDANT);
if (error) {
goto cleanup;
}
/* Create the keys */
if (form->s->keys == 0 || primary_key_no == -1) {
/* Create an index which is used as the clustered index;
order the rows by their row id which is internally generated
by InnoDB */
error = create_clustered_index_when_no_primary(trx,
norm_name);
if (error) {
goto cleanup;
}
}
if (primary_key_no != -1) {
/* In InnoDB the clustered index must always be created
first */
if ((error = create_index(trx, form, norm_name,
(uint) primary_key_no))) {
goto cleanup;
}
}
for (i = 0; i < form->s->keys; i++) {
if (i != (uint) primary_key_no) {
if ((error = create_index(trx, form, norm_name, i))) {
goto cleanup;
}
}
}
if (current_thd->query != NULL) {
LEX_STRING q;
if (thd->convert_string(&q, system_charset_info,
current_thd->query,
current_thd->query_length,
current_thd->charset())) {
error = HA_ERR_OUT_OF_MEM;
goto cleanup;
}
error = row_table_add_foreign_constraints(trx,
q.str, norm_name,
create_info->options & HA_LEX_CREATE_TMP_TABLE);
error = convert_error_code_to_mysql(error, NULL);
if (error) {
goto cleanup;
}
}
innobase_commit_low(trx);
row_mysql_unlock_data_dictionary(trx);
/* Flush the log to reduce probability that the .frm files and
the InnoDB data dictionary get out-of-sync if the user runs
with innodb_flush_log_at_trx_commit = 0 */
log_buffer_flush_to_disk();
innobase_table = dict_table_get(norm_name, NULL);
DBUG_ASSERT(innobase_table != 0);
if ((create_info->used_fields & HA_CREATE_USED_AUTO) &&
(create_info->auto_increment_value != 0)) {
/* Query was ALTER TABLE...AUTO_INCREMENT = x; or
CREATE TABLE ...AUTO_INCREMENT = x; Find out a table
definition from the dictionary and get the current value
of the auto increment field. Set a new value to the
auto increment field if the value is greater than the
maximum value in the column. */
auto_inc_value = create_info->auto_increment_value;
dict_table_autoinc_initialize(innobase_table, auto_inc_value);
}
/* Tell the InnoDB server that there might be work for
utility threads: */
srv_active_wake_master_thread();
trx_free_for_mysql(trx);
DBUG_RETURN(0);
cleanup:
innobase_commit_low(trx);
row_mysql_unlock_data_dictionary(trx);
trx_free_for_mysql(trx);
DBUG_RETURN(error);
}
/*********************************************************************
Discards or imports an InnoDB tablespace. */
int
ha_innobase::discard_or_import_tablespace(
/*======================================*/
/* out: 0 == success, -1 == error */
my_bool discard) /* in: TRUE if discard, else import */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
dict_table_t* dict_table;
trx_t* trx;
int err;
DBUG_ENTER("ha_innobase::discard_or_import_tablespace");
ut_a(prebuilt->trx && prebuilt->trx->magic_n == TRX_MAGIC_N);
ut_a(prebuilt->trx ==
(trx_t*) current_thd->ha_data[innobase_hton.slot]);
dict_table = prebuilt->table;
trx = prebuilt->trx;
if (discard) {
err = row_discard_tablespace_for_mysql(dict_table->name, trx);
} else {
err = row_import_tablespace_for_mysql(dict_table->name, trx);
}
err = convert_error_code_to_mysql(err, NULL);
DBUG_RETURN(err);
}
/*********************************************************************
Deletes all rows of an InnoDB table. */
int
ha_innobase::delete_all_rows(void)
/*==============================*/
/* out: error number */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*)innobase_prebuilt;
int error;
trx_t* trx;
THD* thd = current_thd;
DBUG_ENTER("ha_innobase::delete_all_rows");
if (thd->lex->sql_command != SQLCOM_TRUNCATE) {
fallback:
/* We only handle TRUNCATE TABLE t as a special case.
DELETE FROM t will have to use ha_innobase::delete_row(). */
DBUG_RETURN(my_errno=HA_ERR_WRONG_COMMAND);
}
/* Get the transaction associated with the current thd, or create one
if not yet created */
trx = check_trx_exists(thd);
/* Truncate the table in InnoDB */
error = row_truncate_table_for_mysql(prebuilt->table, trx);
if (error == DB_ERROR) {
/* Cannot truncate; resort to ha_innobase::delete_row() */
goto fallback;
}
error = convert_error_code_to_mysql(error, NULL);
DBUG_RETURN(error);
}
/*********************************************************************
Drops a table from an InnoDB database. Before calling this function,
MySQL calls innobase_commit to commit the transaction of the current user.
Then the current user cannot have locks set on the table. Drop table
operation inside InnoDB will remove all locks any user has on the table
inside InnoDB. */
int
ha_innobase::delete_table(
/*======================*/
/* out: error number */
const char* name) /* in: table name */
{
ulint name_len;
int error;
trx_t* parent_trx;
trx_t* trx;
THD *thd= current_thd;
char norm_name[1000];
DBUG_ENTER("ha_innobase::delete_table");
/* Get the transaction associated with the current thd, or create one
if not yet created */
parent_trx = check_trx_exists(current_thd);
/* In case MySQL calls this in the middle of a SELECT query, release
possible adaptive hash latch to avoid deadlocks of threads */
trx_search_latch_release_if_reserved(parent_trx);
if (lower_case_table_names) {
srv_lower_case_table_names = TRUE;
} else {
srv_lower_case_table_names = FALSE;
}
trx = trx_allocate_for_mysql();
trx->mysql_thd = current_thd;
trx->mysql_query_str = &((*current_thd).query);
if (thd->options & OPTION_NO_FOREIGN_KEY_CHECKS) {
trx->check_foreigns = FALSE;
}
if (thd->options & OPTION_RELAXED_UNIQUE_CHECKS) {
trx->check_unique_secondary = FALSE;
}
name_len = strlen(name);
assert(name_len < 1000);
/* Strangely, MySQL passes the table name without the '.frm'
extension, in contrast to ::create */
normalize_table_name(norm_name, name);
/* Drop the table in InnoDB */
error = row_drop_table_for_mysql(norm_name, trx,
thd->lex->sql_command == SQLCOM_DROP_DB);
/* Flush the log to reduce probability that the .frm files and
the InnoDB data dictionary get out-of-sync if the user runs
with innodb_flush_log_at_trx_commit = 0 */
log_buffer_flush_to_disk();
/* Tell the InnoDB server that there might be work for
utility threads: */
srv_active_wake_master_thread();
innobase_commit_low(trx);
trx_free_for_mysql(trx);
error = convert_error_code_to_mysql(error, NULL);
DBUG_RETURN(error);
}
/*********************************************************************
Removes all tables in the named database inside InnoDB. */
int
innobase_drop_database(
/*===================*/
/* out: error number */
char* path) /* in: database path; inside InnoDB the name
of the last directory in the path is used as
the database name: for example, in 'mysql/data/test'
the database name is 'test' */
{
ulint len = 0;
trx_t* parent_trx;
trx_t* trx;
char* ptr;
int error;
char* namebuf;
/* Get the transaction associated with the current thd, or create one
if not yet created */
parent_trx = check_trx_exists(current_thd);
/* In case MySQL calls this in the middle of a SELECT query, release
possible adaptive hash latch to avoid deadlocks of threads */
trx_search_latch_release_if_reserved(parent_trx);
ptr = strend(path) - 2;
while (ptr >= path && *ptr != '\\' && *ptr != '/') {
ptr--;
len++;
}
ptr++;
namebuf = my_malloc((uint) len + 2, MYF(0));
memcpy(namebuf, ptr, len);
namebuf[len] = '/';
namebuf[len + 1] = '\0';
#ifdef __WIN__
innobase_casedn_str(namebuf);
#endif
trx = trx_allocate_for_mysql();
trx->mysql_thd = current_thd;
trx->mysql_query_str = &((*current_thd).query);
if (current_thd->options & OPTION_NO_FOREIGN_KEY_CHECKS) {
trx->check_foreigns = FALSE;
}
error = row_drop_database_for_mysql(namebuf, trx);
my_free(namebuf, MYF(0));
/* Flush the log to reduce probability that the .frm files and
the InnoDB data dictionary get out-of-sync if the user runs
with innodb_flush_log_at_trx_commit = 0 */
log_buffer_flush_to_disk();
/* Tell the InnoDB server that there might be work for
utility threads: */
srv_active_wake_master_thread();
innobase_commit_low(trx);
trx_free_for_mysql(trx);
error = convert_error_code_to_mysql(error, NULL);
return(error);
}
/*************************************************************************
Renames an InnoDB table. */
int
ha_innobase::rename_table(
/*======================*/
/* out: 0 or error code */
const char* from, /* in: old name of the table */
const char* to) /* in: new name of the table */
{
ulint name_len1;
ulint name_len2;
int error;
trx_t* parent_trx;
trx_t* trx;
char norm_from[1000];
char norm_to[1000];
DBUG_ENTER("ha_innobase::rename_table");
/* Get the transaction associated with the current thd, or create one
if not yet created */
parent_trx = check_trx_exists(current_thd);
/* In case MySQL calls this in the middle of a SELECT query, release
possible adaptive hash latch to avoid deadlocks of threads */
trx_search_latch_release_if_reserved(parent_trx);
if (lower_case_table_names) {
srv_lower_case_table_names = TRUE;
} else {
srv_lower_case_table_names = FALSE;
}
trx = trx_allocate_for_mysql();
trx->mysql_thd = current_thd;
trx->mysql_query_str = &((*current_thd).query);
if (current_thd->options & OPTION_NO_FOREIGN_KEY_CHECKS) {
trx->check_foreigns = FALSE;
}
name_len1 = strlen(from);
name_len2 = strlen(to);
assert(name_len1 < 1000);
assert(name_len2 < 1000);
normalize_table_name(norm_from, from);
normalize_table_name(norm_to, to);
/* Rename the table in InnoDB */
error = row_rename_table_for_mysql(norm_from, norm_to, trx);
/* Flush the log to reduce probability that the .frm files and
the InnoDB data dictionary get out-of-sync if the user runs
with innodb_flush_log_at_trx_commit = 0 */
log_buffer_flush_to_disk();
/* Tell the InnoDB server that there might be work for
utility threads: */
srv_active_wake_master_thread();
innobase_commit_low(trx);
trx_free_for_mysql(trx);
error = convert_error_code_to_mysql(error, NULL);
DBUG_RETURN(error);
}
/*************************************************************************
Estimates the number of index records in a range. */
ha_rows
ha_innobase::records_in_range(
/*==========================*/
/* out: estimated number of
rows */
uint keynr, /* in: index number */
key_range *min_key, /* in: start key value of the
range, may also be 0 */
key_range *max_key) /* in: range end key val, may
also be 0 */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
KEY* key;
dict_index_t* index;
mysql_byte* key_val_buff2 = (mysql_byte*) my_malloc(
table->s->reclength
+ table->s->max_key_length + 100,
MYF(MY_FAE));
ulint buff2_len = table->s->reclength
+ table->s->max_key_length + 100;
dtuple_t* range_start;
dtuple_t* range_end;
ib_longlong n_rows;
ulint mode1;
ulint mode2;
void* heap1;
void* heap2;
DBUG_ENTER("records_in_range");
prebuilt->trx->op_info = (char*)"estimating records in index range";
/* In case MySQL calls this in the middle of a SELECT query, release
possible adaptive hash latch to avoid deadlocks of threads */
trx_search_latch_release_if_reserved(prebuilt->trx);
active_index = keynr;
key = table->key_info + active_index;
index = dict_table_get_index_noninline(prebuilt->table, key->name);
range_start = dtuple_create_for_mysql(&heap1, key->key_parts);
dict_index_copy_types(range_start, index, key->key_parts);
range_end = dtuple_create_for_mysql(&heap2, key->key_parts);
dict_index_copy_types(range_end, index, key->key_parts);
row_sel_convert_mysql_key_to_innobase(
range_start, (byte*) key_val_buff,
(ulint)upd_and_key_val_buff_len,
index,
(byte*) (min_key ? min_key->key :
(const mysql_byte*) 0),
(ulint) (min_key ? min_key->length : 0),
prebuilt->trx);
row_sel_convert_mysql_key_to_innobase(
range_end, (byte*) key_val_buff2,
buff2_len, index,
(byte*) (max_key ? max_key->key :
(const mysql_byte*) 0),
(ulint) (max_key ? max_key->length : 0),
prebuilt->trx);
mode1 = convert_search_mode_to_innobase(min_key ? min_key->flag :
HA_READ_KEY_EXACT);
mode2 = convert_search_mode_to_innobase(max_key ? max_key->flag :
HA_READ_KEY_EXACT);
if (mode1 != PAGE_CUR_UNSUPP && mode2 != PAGE_CUR_UNSUPP) {
n_rows = btr_estimate_n_rows_in_range(index, range_start,
mode1, range_end,
mode2);
} else {
n_rows = HA_POS_ERROR;
}
dtuple_free_for_mysql(heap1);
dtuple_free_for_mysql(heap2);
my_free((gptr) key_val_buff2, MYF(0));
prebuilt->trx->op_info = (char*)"";
/* The MySQL optimizer seems to believe an estimate of 0 rows is
always accurate and may return the result 'Empty set' based on that.
The accuracy is not guaranteed, and even if it were, for a locking
read we should anyway perform the search to set the next-key lock.
Add 1 to the value to make sure MySQL does not make the assumption! */
if (n_rows == 0) {
n_rows = 1;
}
DBUG_RETURN((ha_rows) n_rows);
}
/*************************************************************************
Gives an UPPER BOUND to the number of rows in a table. This is used in
filesort.cc. */
ha_rows
ha_innobase::estimate_rows_upper_bound(void)
/*======================================*/
/* out: upper bound of rows */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
dict_index_t* index;
ulonglong estimate;
ulonglong local_data_file_length;
DBUG_ENTER("estimate_rows_upper_bound");
/* We do not know if MySQL can call this function before calling
external_lock(). To be safe, update the thd of the current table
handle. */
update_thd(current_thd);
prebuilt->trx->op_info = (char*)
"calculating upper bound for table rows";
/* In case MySQL calls this in the middle of a SELECT query, release
possible adaptive hash latch to avoid deadlocks of threads */
trx_search_latch_release_if_reserved(prebuilt->trx);
index = dict_table_get_first_index_noninline(prebuilt->table);
local_data_file_length = ((ulonglong) index->stat_n_leaf_pages)
* UNIV_PAGE_SIZE;
/* Calculate a minimum length for a clustered index record and from
that an upper bound for the number of rows. Since we only calculate
new statistics in row0mysql.c when a table has grown by a threshold
factor, we must add a safety factor 2 in front of the formula below. */
estimate = 2 * local_data_file_length /
dict_index_calc_min_rec_len(index);
prebuilt->trx->op_info = (char*)"";
DBUG_RETURN((ha_rows) estimate);
}
/*************************************************************************
How many seeks it will take to read through the table. This is to be
comparable to the number returned by records_in_range so that we can
decide if we should scan the table or use keys. */
double
ha_innobase::scan_time()
/*====================*/
/* out: estimated time measured in disk seeks */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
/* Since MySQL seems to favor table scans too much over index
searches, we pretend that a sequential read takes the same time
as a random disk read, that is, we do not divide the following
by 10, which would be physically realistic. */
return((double) (prebuilt->table->stat_clustered_index_size));
}
/**********************************************************************
Calculate the time it takes to read a set of ranges through an index
This enables us to optimise reads for clustered indexes. */
double
ha_innobase::read_time(
/*===================*/
/* out: estimated time measured in disk seeks */
uint index, /* in: key number */
uint ranges, /* in: how many ranges */
ha_rows rows) /* in: estimated number of rows in the ranges */
{
ha_rows total_rows;
double time_for_scan;
if (index != table->s->primary_key) {
/* Not clustered */
return(handler::read_time(index, ranges, rows));
}
if (rows <= 2) {
return((double) rows);
}
/* Assume that the read time is proportional to the scan time for all
rows + at most one seek per range. */
time_for_scan = scan_time();
if ((total_rows = estimate_rows_upper_bound()) < rows) {
return(time_for_scan);
}
return(ranges + (double) rows / (double) total_rows * time_for_scan);
}
/*************************************************************************
Returns statistics information of the table to the MySQL interpreter,
in various fields of the handle object. */
int
ha_innobase::info(
/*==============*/
uint flag) /* in: what information MySQL requests */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
dict_table_t* ib_table;
dict_index_t* index;
ha_rows rec_per_key;
ib_longlong n_rows;
ulong j;
ulong i;
char path[FN_REFLEN];
os_file_stat_t stat_info;
DBUG_ENTER("info");
/* If we are forcing recovery at a high level, we will suppress
statistics calculation on tables, because that may crash the
server if an index is badly corrupted. */
if (srv_force_recovery >= SRV_FORCE_NO_IBUF_MERGE) {
/* We return success (0) instead of HA_ERR_CRASHED,
because we want MySQL to process this query and not
stop, like it would do if it received the error code
HA_ERR_CRASHED. */
DBUG_RETURN(0);
}
/* We do not know if MySQL can call this function before calling
external_lock(). To be safe, update the thd of the current table
handle. */
update_thd(current_thd);
/* In case MySQL calls this in the middle of a SELECT query, release
possible adaptive hash latch to avoid deadlocks of threads */
prebuilt->trx->op_info = (char*)"returning various info to MySQL";
trx_search_latch_release_if_reserved(prebuilt->trx);
ib_table = prebuilt->table;
if (flag & HA_STATUS_TIME) {
/* In sql_show we call with this flag: update then statistics
so that they are up-to-date */
prebuilt->trx->op_info = (char*)"updating table statistics";
dict_update_statistics(ib_table);
prebuilt->trx->op_info = (char*)
"returning various info to MySQL";
my_snprintf(path, sizeof(path), "%s/%s%s",
mysql_data_home, ib_table->name,
reg_ext);
unpack_filename(path,path);
/* Note that we do not know the access time of the table,
nor the CHECK TABLE time, nor the UPDATE or INSERT time. */
if (os_file_get_status(path,&stat_info)) {
create_time = stat_info.ctime;
}
}
if (flag & HA_STATUS_VARIABLE) {
n_rows = ib_table->stat_n_rows;
/* Because we do not protect stat_n_rows by any mutex in a
delete, it is theoretically possible that the value can be
smaller than zero! TODO: fix this race.
The MySQL optimizer seems to assume in a left join that n_rows
is an accurate estimate if it is zero. Of course, it is not,
since we do not have any locks on the rows yet at this phase.
Since SHOW TABLE STATUS seems to call this function with the
HA_STATUS_TIME flag set, while the left join optimizer does not
set that flag, we add one to a zero value if the flag is not
set. That way SHOW TABLE STATUS will show the best estimate,
while the optimizer never sees the table empty. */
if (n_rows < 0) {
n_rows = 0;
}
if (n_rows == 0 && !(flag & HA_STATUS_TIME)) {
n_rows++;
}
records = (ha_rows)n_rows;
deleted = 0;
data_file_length = ((ulonglong)
ib_table->stat_clustered_index_size)
* UNIV_PAGE_SIZE;
index_file_length = ((ulonglong)
ib_table->stat_sum_of_other_index_sizes)
* UNIV_PAGE_SIZE;
delete_length = 0;
check_time = 0;
if (records == 0) {
mean_rec_length = 0;
} else {
mean_rec_length = (ulong) (data_file_length / records);
}
}
if (flag & HA_STATUS_CONST) {
index = dict_table_get_first_index_noninline(ib_table);
if (prebuilt->clust_index_was_generated) {
index = dict_table_get_next_index_noninline(index);
}
for (i = 0; i < table->s->keys; i++) {
if (index == NULL) {
ut_print_timestamp(stderr);
sql_print_error("Table %s contains fewer "
"indexes inside InnoDB than "
"are defined in the MySQL "
".frm file. Have you mixed up "
".frm files from different "
"installations? See "
"http://dev.mysql.com/doc/refman/5.0/en/innodb-troubleshooting.html\n",
ib_table->name);
break;
}
for (j = 0; j < table->key_info[i].key_parts; j++) {
if (j + 1 > index->n_uniq) {
ut_print_timestamp(stderr);
sql_print_error(
"Index %s of %s has %lu columns unique inside InnoDB, but MySQL is asking "
"statistics for %lu columns. Have you mixed up .frm files from different "
"installations? "
"See http://dev.mysql.com/doc/refman/5.0/en/innodb-troubleshooting.html\n",
index->name,
ib_table->name,
(unsigned long)
index->n_uniq, j + 1);
break;
}
if (index->stat_n_diff_key_vals[j + 1] == 0) {
rec_per_key = records;
} else {
rec_per_key = (ha_rows)(records /
index->stat_n_diff_key_vals[j + 1]);
}
/* Since MySQL seems to favor table scans
too much over index searches, we pretend
index selectivity is 2 times better than
our estimate: */
rec_per_key = rec_per_key / 2;
if (rec_per_key == 0) {
rec_per_key = 1;
}
table->key_info[i].rec_per_key[j]=
rec_per_key >= ~(ulong) 0 ? ~(ulong) 0 :
(ulong) rec_per_key;
}
index = dict_table_get_next_index_noninline(index);
}
}
if (flag & HA_STATUS_ERRKEY) {
ut_a(prebuilt->trx && prebuilt->trx->magic_n == TRX_MAGIC_N);
errkey = (unsigned int) row_get_mysql_key_number_for_index(
(dict_index_t*)
trx_get_error_info(prebuilt->trx));
}
if (flag & HA_STATUS_AUTO && table->found_next_number_field) {
longlong auto_inc;
int ret;
/* The following function call can the first time fail in
a lock wait timeout error because it reserves the auto-inc
lock on the table. If it fails, then someone is already initing
the auto-inc counter, and the second call is guaranteed to
succeed. */
ret = innobase_read_and_init_auto_inc(&auto_inc);
if (ret != 0) {
ret = innobase_read_and_init_auto_inc(&auto_inc);
if (ret != 0) {
ut_print_timestamp(stderr);
sql_print_error("Cannot get table %s auto-inc"
"counter value in ::info\n",
ib_table->name);
auto_inc = 0;
}
}
auto_increment_value = auto_inc;
}
prebuilt->trx->op_info = (char*)"";
DBUG_RETURN(0);
}
/**************************************************************************
Updates index cardinalities of the table, based on 8 random dives into
each index tree. This does NOT calculate exact statistics on the table. */
int
ha_innobase::analyze(
/*=================*/
/* out: returns always 0 (success) */
THD* thd, /* in: connection thread handle */
HA_CHECK_OPT* check_opt) /* in: currently ignored */
{
/* Simply call ::info() with all the flags */
info(HA_STATUS_TIME | HA_STATUS_CONST | HA_STATUS_VARIABLE);
return(0);
}
/**************************************************************************
This is mapped to "ALTER TABLE tablename TYPE=InnoDB", which rebuilds
the table in MySQL. */
int
ha_innobase::optimize(
/*==================*/
THD* thd, /* in: connection thread handle */
HA_CHECK_OPT* check_opt) /* in: currently ignored */
{
return(HA_ADMIN_TRY_ALTER);
}
/***********************************************************************
Tries to check that an InnoDB table is not corrupted. If corruption is
noticed, prints to stderr information about it. In case of corruption
may also assert a failure and crash the server. */
int
ha_innobase::check(
/*===============*/
/* out: HA_ADMIN_CORRUPT or
HA_ADMIN_OK */
THD* thd, /* in: user thread handle */
HA_CHECK_OPT* check_opt) /* in: check options, currently
ignored */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
ulint ret;
ut_a(prebuilt->trx && prebuilt->trx->magic_n == TRX_MAGIC_N);
ut_a(prebuilt->trx ==
(trx_t*) current_thd->ha_data[innobase_hton.slot]);
if (prebuilt->mysql_template == NULL) {
/* Build the template; we will use a dummy template
in index scans done in checking */
build_template(prebuilt, NULL, table, ROW_MYSQL_WHOLE_ROW);
}
ret = row_check_table_for_mysql(prebuilt);
if (ret == DB_SUCCESS) {
return(HA_ADMIN_OK);
}
return(HA_ADMIN_CORRUPT);
}
/*****************************************************************
Adds information about free space in the InnoDB tablespace to a table comment
which is printed out when a user calls SHOW TABLE STATUS. Adds also info on
foreign keys. */
char*
ha_innobase::update_table_comment(
/*==============================*/
/* out: table comment + InnoDB free space +
info on foreign keys */
const char* comment)/* in: table comment defined by user */
{
uint length = (uint) strlen(comment);
char* str;
row_prebuilt_t* prebuilt = (row_prebuilt_t*)innobase_prebuilt;
long flen;
/* We do not know if MySQL can call this function before calling
external_lock(). To be safe, update the thd of the current table
handle. */
if (length > 64000 - 3) {
return((char*)comment); /* string too long */
}
update_thd(current_thd);
prebuilt->trx->op_info = (char*)"returning table comment";
/* In case MySQL calls this in the middle of a SELECT query, release
possible adaptive hash latch to avoid deadlocks of threads */
trx_search_latch_release_if_reserved(prebuilt->trx);
str = NULL;
/* output the data to a temporary file */
mutex_enter_noninline(&srv_dict_tmpfile_mutex);
rewind(srv_dict_tmpfile);
fprintf(srv_dict_tmpfile, "InnoDB free: %llu kB",
fsp_get_available_space_in_free_extents(
prebuilt->table->space));
dict_print_info_on_foreign_keys(FALSE, srv_dict_tmpfile,
prebuilt->trx, prebuilt->table);
flen = ftell(srv_dict_tmpfile);
if (flen < 0) {
flen = 0;
} else if (length + flen + 3 > 64000) {
flen = 64000 - 3 - length;
}
/* allocate buffer for the full string, and
read the contents of the temporary file */
str = my_malloc(length + flen + 3, MYF(0));
if (str) {
char* pos = str + length;
if (length) {
memcpy(str, comment, length);
*pos++ = ';';
*pos++ = ' ';
}
rewind(srv_dict_tmpfile);
flen = (uint) fread(pos, 1, flen, srv_dict_tmpfile);
pos[flen] = 0;
}
mutex_exit_noninline(&srv_dict_tmpfile_mutex);
prebuilt->trx->op_info = (char*)"";
return(str ? str : (char*) comment);
}
/***********************************************************************
Gets the foreign key create info for a table stored in InnoDB. */
char*
ha_innobase::get_foreign_key_create_info(void)
/*==========================================*/
/* out, own: character string in the form which
can be inserted to the CREATE TABLE statement,
MUST be freed with ::free_foreign_key_create_info */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*)innobase_prebuilt;
char* str = 0;
long flen;
ut_a(prebuilt != NULL);
/* We do not know if MySQL can call this function before calling
external_lock(). To be safe, update the thd of the current table
handle. */
update_thd(current_thd);
prebuilt->trx->op_info = (char*)"getting info on foreign keys";
/* In case MySQL calls this in the middle of a SELECT query,
release possible adaptive hash latch to avoid
deadlocks of threads */
trx_search_latch_release_if_reserved(prebuilt->trx);
mutex_enter_noninline(&srv_dict_tmpfile_mutex);
rewind(srv_dict_tmpfile);
/* output the data to a temporary file */
dict_print_info_on_foreign_keys(TRUE, srv_dict_tmpfile,
prebuilt->trx, prebuilt->table);
prebuilt->trx->op_info = (char*)"";
flen = ftell(srv_dict_tmpfile);
if (flen < 0) {
flen = 0;
} else if (flen > 64000 - 1) {
flen = 64000 - 1;
}
/* allocate buffer for the string, and
read the contents of the temporary file */
str = my_malloc(flen + 1, MYF(0));
if (str) {
rewind(srv_dict_tmpfile);
flen = (uint) fread(str, 1, flen, srv_dict_tmpfile);
str[flen] = 0;
}
mutex_exit_noninline(&srv_dict_tmpfile_mutex);
return(str);
}
int
ha_innobase::get_foreign_key_list(THD *thd, List<FOREIGN_KEY_INFO> *f_key_list)
{
dict_foreign_t* foreign;
DBUG_ENTER("get_foreign_key_list");
row_prebuilt_t* prebuilt = (row_prebuilt_t*)innobase_prebuilt;
ut_a(prebuilt != NULL);
update_thd(current_thd);
prebuilt->trx->op_info = (char*)"getting list of foreign keys";
trx_search_latch_release_if_reserved(prebuilt->trx);
mutex_enter_noninline(&(dict_sys->mutex));
foreign = UT_LIST_GET_FIRST(prebuilt->table->foreign_list);
while (foreign != NULL)
{
uint i;
FOREIGN_KEY_INFO f_key_info;
LEX_STRING *name= 0;
const char *tmp_buff;
tmp_buff= foreign->id;
i= 0;
while (tmp_buff[i] != '/')
i++;
tmp_buff+= i + 1;
f_key_info.forein_id= make_lex_string(thd, 0, tmp_buff,
(uint) strlen(tmp_buff), 1);
tmp_buff= foreign->referenced_table_name;
i= 0;
while (tmp_buff[i] != '/')
i++;
f_key_info.referenced_db= make_lex_string(thd, 0,
tmp_buff, i, 1);
tmp_buff+= i + 1;
f_key_info.referenced_table= make_lex_string(thd, 0, tmp_buff,
(uint) strlen(tmp_buff), 1);
for (i= 0;;)
{
tmp_buff= foreign->foreign_col_names[i];
name= make_lex_string(thd, name, tmp_buff, (uint) strlen(tmp_buff), 1);
f_key_info.foreign_fields.push_back(name);
tmp_buff= foreign->referenced_col_names[i];
name= make_lex_string(thd, name, tmp_buff, (uint) strlen(tmp_buff), 1);
f_key_info.referenced_fields.push_back(name);
if (++i >= foreign->n_fields)
break;
}
ulong length= 0;
if (foreign->type == DICT_FOREIGN_ON_DELETE_CASCADE)
{
length=17;
tmp_buff= "ON DELETE CASCADE";
}
else if (foreign->type == DICT_FOREIGN_ON_DELETE_SET_NULL)
{
length=18;
tmp_buff= "ON DELETE SET NULL";
}
else if (foreign->type == DICT_FOREIGN_ON_DELETE_NO_ACTION)
{
length=19;
tmp_buff= "ON DELETE NO ACTION";
}
else if (foreign->type == DICT_FOREIGN_ON_UPDATE_CASCADE)
{
length=17;
tmp_buff= "ON UPDATE CASCADE";
}
else if (foreign->type == DICT_FOREIGN_ON_UPDATE_SET_NULL)
{
length=18;
tmp_buff= "ON UPDATE SET NULL";
}
else if (foreign->type == DICT_FOREIGN_ON_UPDATE_NO_ACTION)
{
length=19;
tmp_buff= "ON UPDATE NO ACTION";
}
f_key_info.constraint_method= make_lex_string(thd,
f_key_info.constraint_method,
tmp_buff, length, 1);
FOREIGN_KEY_INFO *pf_key_info= ((FOREIGN_KEY_INFO *)
thd->memdup((gptr) &f_key_info,
sizeof(FOREIGN_KEY_INFO)));
f_key_list->push_back(pf_key_info);
foreign = UT_LIST_GET_NEXT(foreign_list, foreign);
}
mutex_exit_noninline(&(dict_sys->mutex));
prebuilt->trx->op_info = (char*)"";
DBUG_RETURN(0);
}
/*********************************************************************
Checks if ALTER TABLE may change the storage engine of the table.
Changing storage engines is not allowed for tables for which there
are foreign key constraints (parent or child tables). */
bool
ha_innobase::can_switch_engines(void)
/*=================================*/
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
bool can_switch;
DBUG_ENTER("ha_innobase::can_switch_engines");
prebuilt->trx->op_info =
"determining if there are foreign key constraints";
row_mysql_lock_data_dictionary(prebuilt->trx);
can_switch = !UT_LIST_GET_FIRST(prebuilt->table->referenced_list)
&& !UT_LIST_GET_FIRST(prebuilt->table->foreign_list);
row_mysql_unlock_data_dictionary(prebuilt->trx);
prebuilt->trx->op_info = "";
DBUG_RETURN(can_switch);
}
/***********************************************************************
Checks if a table is referenced by a foreign key. The MySQL manual states that
a REPLACE is either equivalent to an INSERT, or DELETE(s) + INSERT. Only a
delete is then allowed internally to resolve a duplicate key conflict in
REPLACE, not an update. */
uint
ha_innobase::referenced_by_foreign_key(void)
/*========================================*/
/* out: > 0 if referenced by a FOREIGN KEY */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*)innobase_prebuilt;
if (dict_table_referenced_by_foreign_key(prebuilt->table)) {
return(1);
}
return(0);
}
/***********************************************************************
Frees the foreign key create info for a table stored in InnoDB, if it is
non-NULL. */
void
ha_innobase::free_foreign_key_create_info(
/*======================================*/
char* str) /* in, own: create info string to free */
{
if (str) {
my_free(str, MYF(0));
}
}
/***********************************************************************
Tells something additional to the handler about how to do things. */
int
ha_innobase::extra(
/*===============*/
/* out: 0 or error number */
enum ha_extra_function operation)
/* in: HA_EXTRA_RETRIEVE_ALL_COLS or some
other flag */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
/* Warning: since it is not sure that MySQL calls external_lock
before calling this function, the trx field in prebuilt can be
obsolete! */
switch (operation) {
case HA_EXTRA_FLUSH:
if (prebuilt->blob_heap) {
row_mysql_prebuilt_free_blob_heap(prebuilt);
}
break;
case HA_EXTRA_RESET:
if (prebuilt->blob_heap) {
row_mysql_prebuilt_free_blob_heap(prebuilt);
}
prebuilt->keep_other_fields_on_keyread = 0;
prebuilt->read_just_key = 0;
break;
case HA_EXTRA_RESET_STATE:
prebuilt->keep_other_fields_on_keyread = 0;
prebuilt->read_just_key = 0;
break;
case HA_EXTRA_NO_KEYREAD:
prebuilt->read_just_key = 0;
break;
case HA_EXTRA_RETRIEVE_ALL_COLS:
prebuilt->hint_need_to_fetch_extra_cols
= ROW_RETRIEVE_ALL_COLS;
break;
case HA_EXTRA_RETRIEVE_PRIMARY_KEY:
if (prebuilt->hint_need_to_fetch_extra_cols == 0) {
prebuilt->hint_need_to_fetch_extra_cols
= ROW_RETRIEVE_PRIMARY_KEY;
}
break;
case HA_EXTRA_KEYREAD:
prebuilt->read_just_key = 1;
break;
case HA_EXTRA_KEYREAD_PRESERVE_FIELDS:
prebuilt->keep_other_fields_on_keyread = 1;
break;
default:/* Do nothing */
;
}
return(0);
}
/**********************************************************************
MySQL calls this function at the start of each SQL statement inside LOCK
TABLES. Inside LOCK TABLES the ::external_lock method does not work to
mark SQL statement borders. Note also a special case: if a temporary table
is created inside LOCK TABLES, MySQL has not called external_lock() at all
on that table.
MySQL-5.0 also calls this before each statement in an execution of a stored
procedure. To make the execution more deterministic for binlogging, MySQL-5.0
locks all tables involved in a stored procedure with full explicit table
locks (thd->in_lock_tables is true in ::store_lock()) before executing the
procedure. */
int
ha_innobase::start_stmt(
/*====================*/
/* out: 0 or error code */
THD* thd, /* in: handle to the user thread */
thr_lock_type lock_type)
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
trx_t* trx;
update_thd(thd);
trx = prebuilt->trx;
/* Here we release the search latch and the InnoDB thread FIFO ticket
if they were reserved. They should have been released already at the
end of the previous statement, but because inside LOCK TABLES the
lock count method does not work to mark the end of a SELECT statement,
that may not be the case. We MUST release the search latch before an
INSERT, for example. */
innobase_release_stat_resources(trx);
prebuilt->sql_stat_start = TRUE;
prebuilt->hint_need_to_fetch_extra_cols = 0;
prebuilt->read_just_key = 0;
prebuilt->keep_other_fields_on_keyread = FALSE;
if (!prebuilt->mysql_has_locked) {
/* This handle is for a temporary table created inside
this same LOCK TABLES; since MySQL does NOT call external_lock
in this case, we must use x-row locks inside InnoDB to be
prepared for an update of a row */
prebuilt->select_lock_type = LOCK_X;
} else {
if (trx->isolation_level != TRX_ISO_SERIALIZABLE
&& thd->lex->sql_command == SQLCOM_SELECT
&& lock_type == TL_READ) {
/* For other than temporary tables, we obtain
no lock for consistent read (plain SELECT). */
prebuilt->select_lock_type = LOCK_NONE;
} else {
/* Not a consistent read: restore the
select_lock_type value. The value of
stored_select_lock_type was decided in:
1) ::store_lock(),
2) ::external_lock(),
3) ::init_table_handle_for_HANDLER(), and
4) :.transactional_table_lock(). */
prebuilt->select_lock_type =
prebuilt->stored_select_lock_type;
}
}
trx->detailed_error[0] = '\0';
/* Set the MySQL flag to mark that there is an active transaction */
if (trx->active_trans == 0) {
innobase_register_trx_and_stmt(thd);
trx->active_trans = 1;
} else {
innobase_register_stmt(thd);
}
return(0);
}
/**********************************************************************
Maps a MySQL trx isolation level code to the InnoDB isolation level code */
inline
ulint
innobase_map_isolation_level(
/*=========================*/
/* out: InnoDB isolation level */
enum_tx_isolation iso) /* in: MySQL isolation level code */
{
switch(iso) {
case ISO_REPEATABLE_READ: return(TRX_ISO_REPEATABLE_READ);
case ISO_READ_COMMITTED: return(TRX_ISO_READ_COMMITTED);
case ISO_SERIALIZABLE: return(TRX_ISO_SERIALIZABLE);
case ISO_READ_UNCOMMITTED: return(TRX_ISO_READ_UNCOMMITTED);
default: ut_a(0); return(0);
}
}
/**********************************************************************
As MySQL will execute an external lock for every new table it uses when it
starts to process an SQL statement (an exception is when MySQL calls
start_stmt for the handle) we can use this function to store the pointer to
the THD in the handle. We will also use this function to communicate
to InnoDB that a new SQL statement has started and that we must store a
savepoint to our transaction handle, so that we are able to roll back
the SQL statement in case of an error. */
int
ha_innobase::external_lock(
/*=======================*/
/* out: 0 */
THD* thd, /* in: handle to the user thread */
int lock_type) /* in: lock type */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
trx_t* trx;
DBUG_ENTER("ha_innobase::external_lock");
DBUG_PRINT("enter",("lock_type: %d", lock_type));
update_thd(thd);
trx = prebuilt->trx;
prebuilt->sql_stat_start = TRUE;
prebuilt->hint_need_to_fetch_extra_cols = 0;
prebuilt->read_just_key = 0;
prebuilt->keep_other_fields_on_keyread = FALSE;
if (lock_type == F_WRLCK) {
/* If this is a SELECT, then it is in UPDATE TABLE ...
or SELECT ... FOR UPDATE */
prebuilt->select_lock_type = LOCK_X;
prebuilt->stored_select_lock_type = LOCK_X;
}
if (lock_type != F_UNLCK) {
/* MySQL is setting a new table lock */
trx->detailed_error[0] = '\0';
/* Set the MySQL flag to mark that there is an active
transaction */
if (trx->active_trans == 0) {
innobase_register_trx_and_stmt(thd);
trx->active_trans = 1;
} else if (trx->n_mysql_tables_in_use == 0) {
innobase_register_stmt(thd);
}
trx->n_mysql_tables_in_use++;
prebuilt->mysql_has_locked = TRUE;
if (trx->n_mysql_tables_in_use == 1) {
trx->isolation_level = innobase_map_isolation_level(
(enum_tx_isolation)
thd->variables.tx_isolation);
if (trx->isolation_level <= TRX_ISO_READ_COMMITTED
&& trx->global_read_view) {
/* At low transaction isolation levels we let
each consistent read set its own snapshot */
read_view_close_for_mysql(trx);
}
}
if (trx->isolation_level == TRX_ISO_SERIALIZABLE
&& prebuilt->select_lock_type == LOCK_NONE
&& (thd->options
& (OPTION_NOT_AUTOCOMMIT | OPTION_BEGIN))) {
/* To get serializable execution, we let InnoDB
conceptually add 'LOCK IN SHARE MODE' to all SELECTs
which otherwise would have been consistent reads. An
exception is consistent reads in the AUTOCOMMIT=1 mode:
we know that they are read-only transactions, and they
can be serialized also if performed as consistent
reads. */
prebuilt->select_lock_type = LOCK_S;
prebuilt->stored_select_lock_type = LOCK_S;
}
/* Starting from 4.1.9, no InnoDB table lock is taken in LOCK
TABLES if AUTOCOMMIT=1. It does not make much sense to acquire
an InnoDB table lock if it is released immediately at the end
of LOCK TABLES, and InnoDB's table locks in that case cause
VERY easily deadlocks.
We do not set InnoDB table locks if user has not explicitly
requested a table lock. Note that thd->in_lock_tables
can be TRUE on some cases e.g. at the start of a stored
procedure call (SQLCOM_CALL). */
if (prebuilt->select_lock_type != LOCK_NONE) {
if (thd->in_lock_tables &&
thd->lex->sql_command == SQLCOM_LOCK_TABLES &&
thd->variables.innodb_table_locks &&
(thd->options & OPTION_NOT_AUTOCOMMIT)) {
ulint error;
error = row_lock_table_for_mysql(prebuilt,
NULL, 0);
if (error != DB_SUCCESS) {
error = convert_error_code_to_mysql(
(int) error, user_thd);
DBUG_RETURN((int) error);
}
}
trx->mysql_n_tables_locked++;
}
DBUG_RETURN(0);
}
/* MySQL is releasing a table lock */
trx->n_mysql_tables_in_use--;
prebuilt->mysql_has_locked = FALSE;
/* Release a possible FIFO ticket and search latch. Since we
may reserve the kernel mutex, we have to release the search
system latch first to obey the latching order. */
innobase_release_stat_resources(trx);
/* If the MySQL lock count drops to zero we know that the current SQL
statement has ended */
if (trx->n_mysql_tables_in_use == 0) {
trx->mysql_n_tables_locked = 0;
prebuilt->used_in_HANDLER = FALSE;
if (!(thd->options & (OPTION_NOT_AUTOCOMMIT | OPTION_BEGIN))) {
if (trx->active_trans != 0) {
innobase_commit(thd, TRUE);
}
} else {
if (trx->isolation_level <= TRX_ISO_READ_COMMITTED
&& trx->global_read_view) {
/* At low transaction isolation levels we let
each consistent read set its own snapshot */
read_view_close_for_mysql(trx);
}
}
}
DBUG_RETURN(0);
}
/**********************************************************************
With this function MySQL request a transactional lock to a table when
user issued query LOCK TABLES..WHERE ENGINE = InnoDB. */
int
ha_innobase::transactional_table_lock(
/*==================================*/
/* out: error code */
THD* thd, /* in: handle to the user thread */
int lock_type) /* in: lock type */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
trx_t* trx;
DBUG_ENTER("ha_innobase::transactional_table_lock");
DBUG_PRINT("enter",("lock_type: %d", lock_type));
/* We do not know if MySQL can call this function before calling
external_lock(). To be safe, update the thd of the current table
handle. */
update_thd(thd);
if (prebuilt->table->ibd_file_missing && !current_thd->tablespace_op) {
ut_print_timestamp(stderr);
fprintf(stderr, " InnoDB error:\n"
"MySQL is trying to use a table handle but the .ibd file for\n"
"table %s does not exist.\n"
"Have you deleted the .ibd file from the database directory under\n"
"the MySQL datadir?"
"See http://dev.mysql.com/doc/refman/5.0/en/innodb-troubleshooting.html\n"
"how you can resolve the problem.\n",
prebuilt->table->name);
DBUG_RETURN(HA_ERR_CRASHED);
}
trx = prebuilt->trx;
prebuilt->sql_stat_start = TRUE;
prebuilt->hint_need_to_fetch_extra_cols = 0;
prebuilt->read_just_key = 0;
prebuilt->keep_other_fields_on_keyread = FALSE;
if (lock_type == F_WRLCK) {
prebuilt->select_lock_type = LOCK_X;
prebuilt->stored_select_lock_type = LOCK_X;
} else if (lock_type == F_RDLCK) {
prebuilt->select_lock_type = LOCK_S;
prebuilt->stored_select_lock_type = LOCK_S;
} else {
ut_print_timestamp(stderr);
fprintf(stderr, " InnoDB error:\n"
"MySQL is trying to set transactional table lock with corrupted lock type\n"
"to table %s, lock type %d does not exist.\n",
prebuilt->table->name, lock_type);
DBUG_RETURN(HA_ERR_CRASHED);
}
/* MySQL is setting a new transactional table lock */
/* Set the MySQL flag to mark that there is an active transaction */
if (trx->active_trans == 0) {
innobase_register_trx_and_stmt(thd);
trx->active_trans = 1;
}
if (thd->in_lock_tables && thd->variables.innodb_table_locks) {
ulint error = DB_SUCCESS;
error = row_lock_table_for_mysql(prebuilt, NULL, 0);
if (error != DB_SUCCESS) {
error = convert_error_code_to_mysql((int) error, user_thd);
DBUG_RETURN((int) error);
}
if (thd->options & (OPTION_NOT_AUTOCOMMIT | OPTION_BEGIN)) {
/* Store the current undo_no of the transaction
so that we know where to roll back if we have
to roll back the next SQL statement */
trx_mark_sql_stat_end(trx);
}
}
DBUG_RETURN(0);
}
/****************************************************************************
Here we export InnoDB status variables to MySQL. */
void
innodb_export_status(void)
/*======================*/
{
if (innodb_inited) {
srv_export_innodb_status();
}
}
/****************************************************************************
Implements the SHOW INNODB STATUS command. Sends the output of the InnoDB
Monitor to the client. */
bool
innodb_show_status(
/*===============*/
THD* thd) /* in: the MySQL query thread of the caller */
{
Protocol* protocol = thd->protocol;
trx_t* trx;
static const char truncated_msg[] = "... truncated...\n";
const long MAX_STATUS_SIZE = 64000;
ulint trx_list_start = ULINT_UNDEFINED;
ulint trx_list_end = ULINT_UNDEFINED;
DBUG_ENTER("innodb_show_status");
if (have_innodb != SHOW_OPTION_YES) {
my_message(ER_NOT_SUPPORTED_YET,
"Cannot call SHOW INNODB STATUS because skip-innodb is defined",
MYF(0));
DBUG_RETURN(TRUE);
}
trx = check_trx_exists(thd);
innobase_release_stat_resources(trx);
/* We let the InnoDB Monitor to output at most MAX_STATUS_SIZE
bytes of text. */
long flen, usable_len;
char* str;
mutex_enter_noninline(&srv_monitor_file_mutex);
rewind(srv_monitor_file);
srv_printf_innodb_monitor(srv_monitor_file,
&trx_list_start, &trx_list_end);
flen = ftell(srv_monitor_file);
os_file_set_eof(srv_monitor_file);
if (flen < 0) {
flen = 0;
}
if (flen > MAX_STATUS_SIZE) {
usable_len = MAX_STATUS_SIZE;
} else {
usable_len = flen;
}
/* allocate buffer for the string, and
read the contents of the temporary file */
if (!(str = my_malloc(usable_len + 1, MYF(0))))
{
mutex_exit_noninline(&srv_monitor_file_mutex);
DBUG_RETURN(TRUE);
}
rewind(srv_monitor_file);
if (flen < MAX_STATUS_SIZE) {
/* Display the entire output. */
flen = (long) fread(str, 1, flen, srv_monitor_file);
} else if (trx_list_end < (ulint) flen
&& trx_list_start < trx_list_end
&& trx_list_start + (flen - trx_list_end)
< MAX_STATUS_SIZE - sizeof truncated_msg - 1) {
/* Omit the beginning of the list of active transactions. */
long len = (long) fread(str, 1, trx_list_start, srv_monitor_file);
memcpy(str + len, truncated_msg, sizeof truncated_msg - 1);
len += sizeof truncated_msg - 1;
usable_len = (MAX_STATUS_SIZE - 1) - len;
fseek(srv_monitor_file, flen - usable_len, SEEK_SET);
len += (long) fread(str + len, 1, usable_len, srv_monitor_file);
flen = len;
} else {
/* Omit the end of the output. */
flen = (long) fread(str, 1, MAX_STATUS_SIZE - 1, srv_monitor_file);
}
mutex_exit_noninline(&srv_monitor_file_mutex);
List<Item> field_list;
field_list.push_back(new Item_empty_string("Status", flen));
if (protocol->send_fields(&field_list, Protocol::SEND_NUM_ROWS |
Protocol::SEND_EOF)) {
my_free(str, MYF(0));
DBUG_RETURN(TRUE);
}
protocol->prepare_for_resend();
protocol->store(str, flen, system_charset_info);
my_free(str, MYF(0));
if (protocol->write()) {
DBUG_RETURN(TRUE);
}
send_eof(thd);
DBUG_RETURN(FALSE);
}
/****************************************************************************
Implements the SHOW MUTEX STATUS command. . */
bool
innodb_mutex_show_status(
/*===============*/
THD* thd) /* in: the MySQL query thread of the caller */
{
Protocol *protocol= thd->protocol;
List<Item> field_list;
mutex_t* mutex;
#ifdef UNIV_DEBUG
ulint rw_lock_count= 0;
ulint rw_lock_count_spin_loop= 0;
ulint rw_lock_count_spin_rounds= 0;
ulint rw_lock_count_os_wait= 0;
ulint rw_lock_count_os_yield= 0;
ulonglong rw_lock_wait_time= 0;
#endif /* UNIV_DEBUG */
DBUG_ENTER("innodb_mutex_show_status");
#ifdef UNIV_DEBUG
field_list.push_back(new Item_empty_string("Mutex", FN_REFLEN));
field_list.push_back(new Item_empty_string("Module", FN_REFLEN));
field_list.push_back(new Item_uint("Count", MY_INT64_NUM_DECIMAL_DIGITS));
field_list.push_back(new Item_uint("Spin_waits", MY_INT64_NUM_DECIMAL_DIGITS));
field_list.push_back(new Item_uint("Spin_rounds", MY_INT64_NUM_DECIMAL_DIGITS));
field_list.push_back(new Item_uint("OS_waits", MY_INT64_NUM_DECIMAL_DIGITS));
field_list.push_back(new Item_uint("OS_yields", MY_INT64_NUM_DECIMAL_DIGITS));
field_list.push_back(new Item_uint("OS_waits_time", MY_INT64_NUM_DECIMAL_DIGITS));
#else /* UNIV_DEBUG */
field_list.push_back(new Item_empty_string("File", FN_REFLEN));
field_list.push_back(new Item_uint("Line", MY_INT64_NUM_DECIMAL_DIGITS));
field_list.push_back(new Item_uint("OS_waits", MY_INT64_NUM_DECIMAL_DIGITS));
#endif /* UNIV_DEBUG */
if (protocol->send_fields(&field_list,
Protocol::SEND_NUM_ROWS | Protocol::SEND_EOF))
DBUG_RETURN(TRUE);
mutex_enter_noninline(&mutex_list_mutex);
mutex = UT_LIST_GET_FIRST(mutex_list);
while ( mutex != NULL )
{
#ifdef UNIV_DEBUG
if (mutex->mutex_type != 1)
{
if (mutex->count_using > 0)
{
protocol->prepare_for_resend();
protocol->store(mutex->cmutex_name, system_charset_info);
protocol->store(mutex->cfile_name, system_charset_info);
protocol->store((ulonglong)mutex->count_using);
protocol->store((ulonglong)mutex->count_spin_loop);
protocol->store((ulonglong)mutex->count_spin_rounds);
protocol->store((ulonglong)mutex->count_os_wait);
protocol->store((ulonglong)mutex->count_os_yield);
protocol->store((ulonglong)mutex->lspent_time/1000);
if (protocol->write())
{
mutex_exit_noninline(&mutex_list_mutex);
DBUG_RETURN(1);
}
}
}
else
{
rw_lock_count += mutex->count_using;
rw_lock_count_spin_loop += mutex->count_spin_loop;
rw_lock_count_spin_rounds += mutex->count_spin_rounds;
rw_lock_count_os_wait += mutex->count_os_wait;
rw_lock_count_os_yield += mutex->count_os_yield;
rw_lock_wait_time += mutex->lspent_time;
}
#else /* UNIV_DEBUG */
protocol->prepare_for_resend();
protocol->store(mutex->cfile_name, system_charset_info);
protocol->store((ulonglong)mutex->cline);
protocol->store((ulonglong)mutex->count_os_wait);
if (protocol->write())
{
mutex_exit_noninline(&mutex_list_mutex);
DBUG_RETURN(1);
}
#endif /* UNIV_DEBUG */
mutex = UT_LIST_GET_NEXT(list, mutex);
}
mutex_exit_noninline(&mutex_list_mutex);
#ifdef UNIV_DEBUG
protocol->prepare_for_resend();
protocol->store("rw_lock_mutexes", system_charset_info);
protocol->store("", system_charset_info);
protocol->store((ulonglong)rw_lock_count);
protocol->store((ulonglong)rw_lock_count_spin_loop);
protocol->store((ulonglong)rw_lock_count_spin_rounds);
protocol->store((ulonglong)rw_lock_count_os_wait);
protocol->store((ulonglong)rw_lock_count_os_yield);
protocol->store((ulonglong)rw_lock_wait_time/1000);
if (protocol->write())
{
DBUG_RETURN(1);
}
#endif /* UNIV_DEBUG */
send_eof(thd);
DBUG_RETURN(FALSE);
}
/****************************************************************************
Handling the shared INNOBASE_SHARE structure that is needed to provide table
locking.
****************************************************************************/
static mysql_byte* innobase_get_key(INNOBASE_SHARE *share,uint *length,
my_bool not_used __attribute__((unused)))
{
*length=share->table_name_length;
return (mysql_byte*) share->table_name;
}
static INNOBASE_SHARE *get_share(const char *table_name)
{
INNOBASE_SHARE *share;
pthread_mutex_lock(&innobase_share_mutex);
uint length=(uint) strlen(table_name);
if (!(share=(INNOBASE_SHARE*) hash_search(&innobase_open_tables,
(mysql_byte*) table_name,
length))) {
share = (INNOBASE_SHARE *) my_malloc(sizeof(*share)+length+1,
MYF(MY_FAE | MY_ZEROFILL));
share->table_name_length=length;
share->table_name=(char*) (share+1);
strmov(share->table_name,table_name);
if (my_hash_insert(&innobase_open_tables,
(mysql_byte*) share)) {
pthread_mutex_unlock(&innobase_share_mutex);
my_free((gptr) share,0);
return 0;
}
thr_lock_init(&share->lock);
pthread_mutex_init(&share->mutex,MY_MUTEX_INIT_FAST);
}
share->use_count++;
pthread_mutex_unlock(&innobase_share_mutex);
return share;
}
static void free_share(INNOBASE_SHARE *share)
{
pthread_mutex_lock(&innobase_share_mutex);
if (!--share->use_count)
{
hash_delete(&innobase_open_tables, (mysql_byte*) share);
thr_lock_delete(&share->lock);
pthread_mutex_destroy(&share->mutex);
my_free((gptr) share, MYF(0));
}
pthread_mutex_unlock(&innobase_share_mutex);
}
/*********************************************************************
Converts a MySQL table lock stored in the 'lock' field of the handle to
a proper type before storing pointer to the lock into an array of pointers.
MySQL also calls this if it wants to reset some table locks to a not-locked
state during the processing of an SQL query. An example is that during a
SELECT the read lock is released early on the 'const' tables where we only
fetch one row. MySQL does not call this when it releases all locks at the
end of an SQL statement. */
THR_LOCK_DATA**
ha_innobase::store_lock(
/*====================*/
/* out: pointer to the next
element in the 'to' array */
THD* thd, /* in: user thread handle */
THR_LOCK_DATA** to, /* in: pointer to an array
of pointers to lock structs;
pointer to the 'lock' field
of current handle is stored
next to this array */
enum thr_lock_type lock_type) /* in: lock type to store in
'lock'; this may also be
TL_IGNORE */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
trx_t* trx;
/* Note that trx in this function is NOT necessarily prebuilt->trx
because we call update_thd() later, in ::external_lock()! Failure to
understand this caused a serious memory corruption bug in 5.1.11. */
trx = check_trx_exists(thd);
/* NOTE: MySQL can call this function with lock 'type' TL_IGNORE!
Be careful to ignore TL_IGNORE if we are going to do something with
only 'real' locks! */
if ((lock_type == TL_READ && thd->in_lock_tables) ||
(lock_type == TL_READ_HIGH_PRIORITY && thd->in_lock_tables) ||
lock_type == TL_READ_WITH_SHARED_LOCKS ||
lock_type == TL_READ_NO_INSERT ||
(thd->lex->sql_command != SQLCOM_SELECT
&& lock_type != TL_IGNORE)) {
/* The OR cases above are in this order:
1) MySQL is doing LOCK TABLES ... READ LOCAL, or we
are processing a stored procedure or function, or
2) (we do not know when TL_READ_HIGH_PRIORITY is used), or
3) this is a SELECT ... IN SHARE MODE, or
4) we are doing a complex SQL statement like
INSERT INTO ... SELECT ... and the logical logging (MySQL
binlog) requires the use of a locking read, or
MySQL is doing LOCK TABLES ... READ.
5) we let InnoDB do locking reads for all SQL statements that
are not simple SELECTs; note that select_lock_type in this
case may get strengthened in ::external_lock() to LOCK_X.
Note that we MUST use a locking read in all data modifying
SQL statements, because otherwise the execution would not be
serializable, and also the results from the update could be
unexpected if an obsolete consistent read view would be
used. */
if (srv_locks_unsafe_for_binlog &&
trx->isolation_level != TRX_ISO_SERIALIZABLE &&
(lock_type == TL_READ || lock_type == TL_READ_NO_INSERT) &&
(thd->lex->sql_command == SQLCOM_INSERT_SELECT ||
thd->lex->sql_command == SQLCOM_UPDATE ||
thd->lex->sql_command == SQLCOM_CREATE_TABLE)) {
/* In case we have innobase_locks_unsafe_for_binlog
option set and isolation level of the transaction
is not set to serializable and MySQL is doing
INSERT INTO...SELECT or UPDATE ... = (SELECT ...) or
CREATE ... SELECT... without FOR UPDATE or
IN SHARE MODE in select, then we use consistent
read for select. */
prebuilt->select_lock_type = LOCK_NONE;
prebuilt->stored_select_lock_type = LOCK_NONE;
} else if (thd->lex->sql_command == SQLCOM_CHECKSUM) {
/* Use consistent read for checksum table */
prebuilt->select_lock_type = LOCK_NONE;
prebuilt->stored_select_lock_type = LOCK_NONE;
} else {
prebuilt->select_lock_type = LOCK_S;
prebuilt->stored_select_lock_type = LOCK_S;
}
} else if (lock_type != TL_IGNORE) {
/* We set possible LOCK_X value in external_lock, not yet
here even if this would be SELECT ... FOR UPDATE */
prebuilt->select_lock_type = LOCK_NONE;
prebuilt->stored_select_lock_type = LOCK_NONE;
}
if (lock_type != TL_IGNORE && lock.type == TL_UNLOCK) {
/* Starting from 5.0.7, we weaken also the table locks
set at the start of a MySQL stored procedure call, just like
we weaken the locks set at the start of an SQL statement.
MySQL does set thd->in_lock_tables TRUE there, but in reality
we do not need table locks to make the execution of a
single transaction stored procedure call deterministic
(if it does not use a consistent read). */
if (lock_type == TL_READ
&& thd->lex->sql_command == SQLCOM_LOCK_TABLES) {
/* We come here if MySQL is processing LOCK TABLES
... READ LOCAL. MyISAM under that table lock type
reads the table as it was at the time the lock was
granted (new inserts are allowed, but not seen by the
reader). To get a similar effect on an InnoDB table,
we must use LOCK TABLES ... READ. We convert the lock
type here, so that for InnoDB, READ LOCAL is
equivalent to READ. This will change the InnoDB
behavior in mysqldump, so that dumps of InnoDB tables
are consistent with dumps of MyISAM tables. */
lock_type = TL_READ_NO_INSERT;
}
/* If we are not doing a LOCK TABLE, DISCARD/IMPORT
TABLESPACE or TRUNCATE TABLE then allow multiple
writers. Note that ALTER TABLE uses a TL_WRITE_ALLOW_READ
< TL_WRITE_CONCURRENT_INSERT.
We especially allow multiple writers if MySQL is at the
start of a stored procedure call (SQLCOM_CALL) or a
stored function call (MySQL does have thd->in_lock_tables
TRUE there). */
if ((lock_type >= TL_WRITE_CONCURRENT_INSERT
&& lock_type <= TL_WRITE)
&& !(thd->in_lock_tables
&& thd->lex->sql_command == SQLCOM_LOCK_TABLES)
&& !thd->tablespace_op
&& thd->lex->sql_command != SQLCOM_TRUNCATE
&& thd->lex->sql_command != SQLCOM_OPTIMIZE
&& thd->lex->sql_command != SQLCOM_CREATE_TABLE) {
lock_type = TL_WRITE_ALLOW_WRITE;
}
/* In queries of type INSERT INTO t1 SELECT ... FROM t2 ...
MySQL would use the lock TL_READ_NO_INSERT on t2, and that
would conflict with TL_WRITE_ALLOW_WRITE, blocking all inserts
to t2. Convert the lock to a normal read lock to allow
concurrent inserts to t2.
We especially allow concurrent inserts if MySQL is at the
start of a stored procedure call (SQLCOM_CALL)
(MySQL does have thd->in_lock_tables TRUE there). */
if (lock_type == TL_READ_NO_INSERT
&& thd->lex->sql_command != SQLCOM_LOCK_TABLES) {
lock_type = TL_READ;
}
lock.type = lock_type;
}
*to++= &lock;
return(to);
}
/***********************************************************************
This function initializes the auto-inc counter if it has not been
initialized yet. This function does not change the value of the auto-inc
counter if it already has been initialized. In parameter ret returns
the value of the auto-inc counter. */
int
ha_innobase::innobase_read_and_init_auto_inc(
/*=========================================*/
/* out: 0 or error code: deadlock or lock wait
timeout */
longlong* ret) /* out: auto-inc value */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
longlong auto_inc;
ulint old_select_lock_type;
ibool trx_was_not_started = FALSE;
int error;
ut_a(prebuilt);
ut_a(prebuilt->trx ==
(trx_t*) current_thd->ha_data[innobase_hton.slot]);
ut_a(prebuilt->table);
if (prebuilt->trx->conc_state == TRX_NOT_STARTED) {
trx_was_not_started = TRUE;
}
/* In case MySQL calls this in the middle of a SELECT query, release
possible adaptive hash latch to avoid deadlocks of threads */
trx_search_latch_release_if_reserved(prebuilt->trx);
auto_inc = dict_table_autoinc_read(prebuilt->table);
if (auto_inc != 0) {
/* Already initialized */
*ret = auto_inc;
error = 0;
goto func_exit_early;
}
error = row_lock_table_autoinc_for_mysql(prebuilt);
if (error != DB_SUCCESS) {
error = convert_error_code_to_mysql(error, user_thd);
goto func_exit_early;
}
/* Check again if someone has initialized the counter meanwhile */
auto_inc = dict_table_autoinc_read(prebuilt->table);
if (auto_inc != 0) {
*ret = auto_inc;
error = 0;
goto func_exit_early;
}
(void) extra(HA_EXTRA_KEYREAD);
index_init(table->s->next_number_index);
/* Starting from 5.0.9, we use a consistent read to read the auto-inc
column maximum value. This eliminates the spurious deadlocks caused
by the row X-lock that we previously used. Note the following flaw
in our algorithm: if some other user meanwhile UPDATEs the auto-inc
column, our consistent read will not return the largest value. We
accept this flaw, since the deadlocks were a bigger trouble. */
/* Fetch all the columns in the key */
prebuilt->hint_need_to_fetch_extra_cols = ROW_RETRIEVE_ALL_COLS;
old_select_lock_type = prebuilt->select_lock_type;
prebuilt->select_lock_type = LOCK_NONE;
/* Eliminate an InnoDB error print that happens when we try to SELECT
from a table when no table has been locked in ::external_lock(). */
prebuilt->trx->n_mysql_tables_in_use++;
/* Since we will perform a MySQL SELECT query to determine the
auto-inc value, set prebuilt->sql_stat_start = TRUE so that it
is performed like any normal SELECT, regardless of the context
we come here. */
prebuilt->sql_stat_start = TRUE;
error = index_last(table->record[1]);
prebuilt->trx->n_mysql_tables_in_use--;
prebuilt->select_lock_type = old_select_lock_type;
if (error) {
if (error == HA_ERR_END_OF_FILE) {
/* The table was empty, initialize to 1 */
auto_inc = 1;
error = 0;
} else {
/* This should not happen in a consistent read */
sql_print_error("Consistent read of auto-inc column "
"returned %lu", (ulong) error);
auto_inc = -1;
goto func_exit;
}
} else {
/* Initialize to max(col) + 1; we use
'found_next_number_field' below because MySQL in SHOW TABLE
STATUS does not seem to set 'next_number_field'. The comment
in table.h says that 'next_number_field' is set when it is
'active'. */
auto_inc = (longlong) table->found_next_number_field->
val_int_offset(table->s->rec_buff_length) + 1;
}
dict_table_autoinc_initialize(prebuilt->table, auto_inc);
func_exit:
(void) extra(HA_EXTRA_NO_KEYREAD);
index_end();
*ret = auto_inc;
func_exit_early:
/* Since MySQL does not seem to call autocommit after SHOW TABLE
STATUS (even if we would register the trx here), we commit our
transaction here if it was started here. This is to eliminate a
dangling transaction. If the user had AUTOCOMMIT=0, then SHOW
TABLE STATUS does leave a dangling transaction if the user does not
himself call COMMIT. */
if (trx_was_not_started) {
innobase_commit_low(prebuilt->trx);
}
return(error);
}
/***********************************************************************
This function initializes the auto-inc counter if it has not been
initialized yet. This function does not change the value of the auto-inc
counter if it already has been initialized. Returns the value of the
auto-inc counter. */
ulonglong
ha_innobase::get_auto_increment()
/*=============================*/
/* out: auto-increment column value, -1 if error
(deadlock or lock wait timeout) */
{
longlong nr;
int error;
error = innobase_read_and_init_auto_inc(&nr);
if (error) {
/* This should never happen in the current (5.0.6) code, since
we call this function only after the counter has been
initialized. */
ut_print_timestamp(stderr);
sql_print_error("Error %lu in ::get_auto_increment()",
(ulong) error);
return(~(ulonglong) 0);
}
return((ulonglong) nr);
}
/* See comment in handler.h */
int
ha_innobase::reset_auto_increment(ulonglong value)
{
DBUG_ENTER("ha_innobase::reset_auto_increment");
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
int error;
error = row_lock_table_autoinc_for_mysql(prebuilt);
if (error != DB_SUCCESS) {
error = convert_error_code_to_mysql(error, user_thd);
DBUG_RETURN(error);
}
dict_table_autoinc_initialize(prebuilt->table, value);
DBUG_RETURN(0);
}
/* See comment in handler.cc */
bool
ha_innobase::get_error_message(int error, String *buf)
{
trx_t* trx = check_trx_exists(current_thd);
buf->copy(trx->detailed_error, (uint) strlen(trx->detailed_error),
system_charset_info);
return FALSE;
}
/***********************************************************************
Compares two 'refs'. A 'ref' is the (internal) primary key value of the row.
If there is no explicitly declared non-null unique key or a primary key, then
InnoDB internally uses the row id as the primary key. */
int
ha_innobase::cmp_ref(
/*=================*/
/* out: < 0 if ref1 < ref2, 0 if equal, else
> 0 */
const mysql_byte* ref1, /* in: an (internal) primary key value in the
MySQL key value format */
const mysql_byte* ref2) /* in: an (internal) primary key value in the
MySQL key value format */
{
row_prebuilt_t* prebuilt = (row_prebuilt_t*) innobase_prebuilt;
enum_field_types mysql_type;
Field* field;
KEY_PART_INFO* key_part;
KEY_PART_INFO* key_part_end;
uint len1;
uint len2;
int result;
if (prebuilt->clust_index_was_generated) {
/* The 'ref' is an InnoDB row id */
return(memcmp(ref1, ref2, DATA_ROW_ID_LEN));
}
/* Do a type-aware comparison of primary key fields. PK fields
are always NOT NULL, so no checks for NULL are performed. */
key_part = table->key_info[table->s->primary_key].key_part;
key_part_end = key_part
+ table->key_info[table->s->primary_key].key_parts;
for (; key_part != key_part_end; ++key_part) {
field = key_part->field;
mysql_type = field->type();
if (mysql_type == FIELD_TYPE_TINY_BLOB
|| mysql_type == FIELD_TYPE_MEDIUM_BLOB
|| mysql_type == FIELD_TYPE_BLOB
|| mysql_type == FIELD_TYPE_LONG_BLOB) {
/* In the MySQL key value format, a column prefix of
a BLOB is preceded by a 2-byte length field */
len1 = innobase_read_from_2_little_endian(ref1);
len2 = innobase_read_from_2_little_endian(ref2);
ref1 += 2;
ref2 += 2;
result = ((Field_blob*)field)->cmp(
(const char*)ref1, len1,
(const char*)ref2, len2);
} else {
result = field->key_cmp(ref1, ref2);
}
if (result) {
return(result);
}
ref1 += key_part->store_length;
ref2 += key_part->store_length;
}
return(0);
}
char*
ha_innobase::get_mysql_bin_log_name()
{
return(trx_sys_mysql_bin_log_name);
}
ulonglong
ha_innobase::get_mysql_bin_log_pos()
{
/* trx... is ib_longlong, which is a typedef for a 64-bit integer
(__int64 or longlong) so it's ok to cast it to ulonglong. */
return(trx_sys_mysql_bin_log_pos);
}
extern "C" {
/**********************************************************************
This function is used to find the storage length in bytes of the first n
characters for prefix indexes using a multibyte character set. The function
finds charset information and returns length of prefix_len characters in the
index field in bytes.
NOTE: the prototype of this function is copied to data0type.c! If you change
this function, you MUST change also data0type.c! */
ulint
innobase_get_at_most_n_mbchars(
/*===========================*/
/* out: number of bytes occupied by the first
n characters */
ulint charset_id, /* in: character set id */
ulint prefix_len, /* in: prefix length in bytes of the index
(this has to be divided by mbmaxlen to get the
number of CHARACTERS n in the prefix) */
ulint data_len, /* in: length of the string in bytes */
const char* str) /* in: character string */
{
ulint char_length; /* character length in bytes */
ulint n_chars; /* number of characters in prefix */
CHARSET_INFO* charset; /* charset used in the field */
charset = get_charset((uint) charset_id, MYF(MY_WME));
ut_ad(charset);
ut_ad(charset->mbmaxlen);
/* Calculate how many characters at most the prefix index contains */
n_chars = prefix_len / charset->mbmaxlen;
/* If the charset is multi-byte, then we must find the length of the
first at most n chars in the string. If the string contains less
characters than n, then we return the length to the end of the last
character. */
if (charset->mbmaxlen > 1) {
/* my_charpos() returns the byte length of the first n_chars
characters, or a value bigger than the length of str, if
there were not enough full characters in str.
Why does the code below work:
Suppose that we are looking for n UTF-8 characters.
1) If the string is long enough, then the prefix contains at
least n complete UTF-8 characters + maybe some extra
characters + an incomplete UTF-8 character. No problem in
this case. The function returns the pointer to the
end of the nth character.
2) If the string is not long enough, then the string contains
the complete value of a column, that is, only complete UTF-8
characters, and we can store in the column prefix index the
whole string. */
char_length = my_charpos(charset, str,
str + data_len, (int) n_chars);
if (char_length > data_len) {
char_length = data_len;
}
} else {
if (data_len < prefix_len) {
char_length = data_len;
} else {
char_length = prefix_len;
}
}
return(char_length);
}
}
extern "C" {
/**********************************************************************
This function returns true if
1) SQL-query in the current thread
is either REPLACE or LOAD DATA INFILE REPLACE.
2) SQL-query in the current thread
is INSERT ON DUPLICATE KEY UPDATE.
NOTE that /mysql/innobase/row/row0ins.c must contain the
prototype for this function ! */
ibool
innobase_query_is_update(void)
/*==========================*/
{
THD* thd;
thd = (THD *)innobase_current_thd();
if (thd->lex->sql_command == SQLCOM_REPLACE ||
thd->lex->sql_command == SQLCOM_REPLACE_SELECT ||
(thd->lex->sql_command == SQLCOM_LOAD &&
thd->lex->duplicates == DUP_REPLACE)) {
return(1);
}
if (thd->lex->sql_command == SQLCOM_INSERT &&
thd->lex->duplicates == DUP_UPDATE) {
return(1);
}
return(0);
}
}
/***********************************************************************
This function is used to prepare X/Open XA distributed transaction */
int
innobase_xa_prepare(
/*================*/
/* out: 0 or error number */
THD* thd, /* in: handle to the MySQL thread of the user
whose XA transaction should be prepared */
bool all) /* in: TRUE - commit transaction
FALSE - the current SQL statement ended */
{
int error = 0;
trx_t* trx = check_trx_exists(thd);
if (thd->lex->sql_command != SQLCOM_XA_PREPARE) {
/* For ibbackup to work the order of transactions in binlog
and InnoDB must be the same. Consider the situation
thread1> prepare; write to binlog; ...
<context switch>
thread2> prepare; write to binlog; commit
thread1> ... commit
To ensure this will not happen we're taking the mutex on
prepare, and releasing it on commit.
Note: only do it for normal commits, done via ha_commit_trans.
If 2pc protocol is executed by external transaction
coordinator, it will be just a regular MySQL client
executing XA PREPARE and XA COMMIT commands.
In this case we cannot know how many minutes or hours
will be between XA PREPARE and XA COMMIT, and we don't want
to block for undefined period of time.
*/
pthread_mutex_lock(&prepare_commit_mutex);
trx->active_trans = 2;
}
if (!thd->variables.innodb_support_xa) {
return(0);
}
trx->xid=thd->transaction.xid_state.xid;
/* Release a possible FIFO ticket and search latch. Since we will
reserve the kernel mutex, we have to release the search system latch
first to obey the latching order. */
innobase_release_stat_resources(trx);
if (trx->active_trans == 0 && trx->conc_state != TRX_NOT_STARTED) {
sql_print_error("trx->active_trans == 0, but trx->conc_state != "
"TRX_NOT_STARTED");
}
if (all
|| (!(thd->options & (OPTION_NOT_AUTOCOMMIT | OPTION_BEGIN)))) {
/* We were instructed to prepare the whole transaction, or
this is an SQL statement end and autocommit is on */
ut_ad(trx->active_trans);
error = (int) trx_prepare_for_mysql(trx);
} else {
/* We just mark the SQL statement ended and do not do a
transaction prepare */
if (trx->auto_inc_lock) {
/* If we had reserved the auto-inc lock for some
table in this SQL statement we release it now */
row_unlock_table_autoinc_for_mysql(trx);
}
/* Store the current undo_no of the transaction so that we
know where to roll back if we have to roll back the next
SQL statement */
trx_mark_sql_stat_end(trx);
}
/* Tell the InnoDB server that there might be work for utility
threads: */
srv_active_wake_master_thread();
return error;
}
/***********************************************************************
This function is used to recover X/Open XA distributed transactions */
int
innobase_xa_recover(
/*================*/
/* out: number of prepared transactions
stored in xid_list */
XID* xid_list, /* in/out: prepared transactions */
uint len) /* in: number of slots in xid_list */
{
if (len == 0 || xid_list == NULL) {
return(0);
}
return(trx_recover_for_mysql(xid_list, len));
}
/***********************************************************************
This function is used to commit one X/Open XA distributed transaction
which is in the prepared state */
int
innobase_commit_by_xid(
/*===================*/
/* out: 0 or error number */
XID* xid) /* in: X/Open XA transaction identification */
{
trx_t* trx;
trx = trx_get_trx_by_xid(xid);
if (trx) {
innobase_commit_low(trx);
return(XA_OK);
} else {
return(XAER_NOTA);
}
}
/***********************************************************************
This function is used to rollback one X/Open XA distributed transaction
which is in the prepared state */
int
innobase_rollback_by_xid(
/*=====================*/
/* out: 0 or error number */
XID *xid) /* in: X/Open XA transaction identification */
{
trx_t* trx;
trx = trx_get_trx_by_xid(xid);
if (trx) {
return(innobase_rollback_trx(trx));
} else {
return(XAER_NOTA);
}
}
/***********************************************************************
Create a consistent view for a cursor based on current transaction
which is created if the corresponding MySQL thread still lacks one.
This consistent view is then used inside of MySQL when accessing records
using a cursor. */
void*
innobase_create_cursor_view(void)
/*=============================*/
/* out: Pointer to cursor view or NULL */
{
return(read_cursor_view_create_for_mysql(
check_trx_exists(current_thd)));
}
/***********************************************************************
Close the given consistent cursor view of a transaction and restore
global read view to a transaction read view. Transaction is created if the
corresponding MySQL thread still lacks one. */
void
innobase_close_cursor_view(
/*=======================*/
void* curview)/* in: Consistent read view to be closed */
{
read_cursor_view_close_for_mysql(check_trx_exists(current_thd),
(cursor_view_t*) curview);
}
/***********************************************************************
Set the given consistent cursor view to a transaction which is created
if the corresponding MySQL thread still lacks one. If the given
consistent cursor view is NULL global read view of a transaction is
restored to a transaction read view. */
void
innobase_set_cursor_view(
/*=====================*/
void* curview)/* in: Consistent cursor view to be set */
{
read_cursor_set_for_mysql(check_trx_exists(current_thd),
(cursor_view_t*) curview);
}
/***********************************************************************
This function checks each index name for a table against reserved
system default primary index name 'GEN_CLUST_INDEX'. If a name matches,
this function pushes an error message to the client, and returns true. */
static
bool
innobase_index_name_is_reserved(
/*============================*/
/* out: true if an index name
matches the reserved name */
const trx_t* trx, /* in: InnoDB transaction handle */
const TABLE* form, /* in: information on table
columns and indexes */
const char* norm_name) /* in: table name */
{
KEY* key;
uint key_num; /* index number */
for (key_num = 0; key_num < form->s->keys; key_num++) {
key = form->key_info + key_num;
if (innobase_strcasecmp(key->name,
innobase_index_reserve_name) == 0) {
/* Push warning to mysql */
push_warning_printf((THD*) trx->mysql_thd,
MYSQL_ERROR::WARN_LEVEL_WARN,
ER_CANT_CREATE_TABLE,
"Cannot Create Index with name "
"'%s'. The name is reserved "
"for the system default primary "
"index.",
innobase_index_reserve_name);
return(true);
}
}
return(false);
}
#endif /* HAVE_INNOBASE_DB */
|