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
|
/*****************************************************************************
Copyright (c) 1994, 2016, Oracle and/or its affiliates. All Rights Reserved.
Copyright (c) 2012, Facebook Inc.
Copyright (c) 2014, 2022, MariaDB Corporation.
This program is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free Software
Foundation; version 2 of the License.
This program is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with
this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1335 USA
*****************************************************************************/
/**************************************************//**
@file btr/btr0btr.cc
The B-tree
Created 6/2/1994 Heikki Tuuri
*******************************************************/
#include "btr0btr.h"
#include "page0page.h"
#include "page0zip.h"
#include "gis0rtree.h"
#include "btr0cur.h"
#include "btr0sea.h"
#include "btr0pcur.h"
#include "btr0defragment.h"
#include "rem0cmp.h"
#include "lock0lock.h"
#include "ibuf0ibuf.h"
#include "trx0trx.h"
#include "srv0mon.h"
#include "gis0geo.h"
#include "dict0boot.h"
#include "row0sel.h" /* row_search_max_autoinc() */
#include "log.h"
/**************************************************************//**
Checks if the page in the cursor can be merged with given page.
If necessary, re-organize the merge_page.
@return true if possible to merge. */
static
bool
btr_can_merge_with_page(
/*====================*/
btr_cur_t* cursor, /*!< in: cursor on the page to merge */
uint32_t page_no, /*!< in: a sibling page */
buf_block_t** merge_block, /*!< out: the merge block */
mtr_t* mtr); /*!< in: mini-transaction */
/*
Latching strategy of the InnoDB B-tree
--------------------------------------
Node pointer page latches acquisition is protected by index->lock latch.
Before MariaDB 10.2.2, all node pointer pages were protected by index->lock
either in S (shared) or X (exclusive) mode and block->lock was not acquired on
node pointer pages.
After MariaDB 10.2.2, block->lock S-latch or X-latch is used to protect
node pointer pages and obtaiment of node pointer page latches is protected by
index->lock.
(0) Definition: B-tree level.
(0.1) The leaf pages of the B-tree are at level 0.
(0.2) The parent of a page at level L has level L+1. (The level of the
root page is equal to the tree height.)
(0.3) The B-tree lock (index->lock) is the parent of the root page and
has a level = tree height + 1.
Index->lock has 3 possible locking modes:
(1) S-latch:
(1.1) All latches for pages must be obtained in descending order of tree level.
(1.2) Before obtaining the first node pointer page latch at a given B-tree
level, parent latch must be held (at level +1 ).
(1.3) If a node pointer page is already latched at the same level
we can only obtain latch to its right sibling page latch at the same level.
(1.4) Release of the node pointer page latches must be done in
child-to-parent order. (Prevents deadlocks when obtained index->lock
in SX mode).
(1.4.1) Level L node pointer page latch can be released only when
no latches at children level i.e. level < L are hold.
(1.4.2) All latches from node pointer pages must be released so
that no latches are obtained between.
(1.5) [implied by (1.1), (1.2)] Root page latch must be first node pointer
latch obtained.
(2) SX-latch:
In this case rules (1.2) and (1.3) from S-latch case are relaxed and
merged into (2.2) and rule (1.4) is removed. Thus, latch acquisition
can be skipped at some tree levels and latches can be obtained in
a less restricted order.
(2.1) [identical to (1.1)]: All latches for pages must be obtained in descending
order of tree level.
(2.2) When a node pointer latch at level L is obtained,
the left sibling page latch in the same level or some ancestor
page latch (at level > L) must be hold.
(2.3) [implied by (2.1), (2.2)] The first node pointer page latch obtained can
be any node pointer page.
(3) X-latch:
Node pointer latches can be obtained in any order.
NOTE: New rules after MariaDB 10.2.2 does not affect the latching rules of leaf pages:
index->lock S-latch is needed in read for the node pointer traversal. When the leaf
level is reached, index-lock can be released (and with the MariaDB 10.2.2 changes, all
node pointer latches). Left to right index travelsal in leaf page level can be safely done
by obtaining right sibling leaf page latch and then releasing the old page latch.
Single leaf page modifications (BTR_MODIFY_LEAF) are protected by index->lock
S-latch.
B-tree operations involving page splits or merges (BTR_MODIFY_TREE) and page
allocations are protected by index->lock X-latch.
Node pointers
-------------
Leaf pages of a B-tree contain the index records stored in the
tree. On levels n > 0 we store 'node pointers' to pages on level
n - 1. For each page there is exactly one node pointer stored:
thus the our tree is an ordinary B-tree, not a B-link tree.
A node pointer contains a prefix P of an index record. The prefix
is long enough so that it determines an index record uniquely.
The file page number of the child page is added as the last
field. To the child page we can store node pointers or index records
which are >= P in the alphabetical order, but < P1 if there is
a next node pointer on the level, and P1 is its prefix.
If a node pointer with a prefix P points to a non-leaf child,
then the leftmost record in the child must have the same
prefix P. If it points to a leaf node, the child is not required
to contain any record with a prefix equal to P. The leaf case
is decided this way to allow arbitrary deletions in a leaf node
without touching upper levels of the tree.
We have predefined a special minimum record which we
define as the smallest record in any alphabetical order.
A minimum record is denoted by setting a bit in the record
header. A minimum record acts as the prefix of a node pointer
which points to a leftmost node on any level of the tree.
File page allocation
--------------------
In the root node of a B-tree there are two file segment headers.
The leaf pages of a tree are allocated from one file segment, to
make them consecutive on disk if possible. From the other file segment
we allocate pages for the non-leaf levels of the tree.
*/
/** Check a file segment header within a B-tree root page.
@param offset file segment header offset
@param block B-tree root page
@param space tablespace
@return whether the segment header is valid */
static bool btr_root_fseg_validate(ulint offset,
const buf_block_t &block,
const fil_space_t &space)
{
ut_ad(block.page.id().space() == space.id);
const uint16_t hdr= mach_read_from_2(offset + FSEG_HDR_OFFSET +
block.page.frame);
if (FIL_PAGE_DATA <= hdr && hdr <= srv_page_size - FIL_PAGE_DATA_END &&
mach_read_from_4(block.page.frame + offset + FSEG_HDR_SPACE) == space.id)
return true;
sql_print_error("InnoDB: Index root page " UINT32PF " in %s is corrupted "
"at " ULINTPF,
block.page.id().page_no(),
UT_LIST_GET_FIRST(space.chain)->name);
return false;
}
/** Report a decryption failure. */
ATTRIBUTE_COLD void btr_decryption_failed(const dict_index_t &index)
{
ib_push_warning(static_cast<void*>(nullptr), DB_DECRYPTION_FAILED,
"Table %s is encrypted but encryption service or"
" used key_id is not available. "
" Can't continue reading table.",
index.table->name.m_name);
index.table->file_unreadable= true;
}
/** Get an index page and declare its latching order level.
@param[in] index index tree
@param[in] page page number
@param[in] mode latch mode
@param[in] merge whether change buffer merge should be attempted
@param[in,out] mtr mini-transaction
@param[out] err error code
@return block */
buf_block_t *btr_block_get(const dict_index_t &index,
uint32_t page, ulint mode, bool merge,
mtr_t *mtr, dberr_t *err)
{
dberr_t local_err;
if (!err)
err= &local_err;
buf_block_t *block=
buf_page_get_gen(page_id_t{index.table->space->id, page},
index.table->space->zip_size(), mode, nullptr, BUF_GET,
mtr, err, merge && !index.is_clust());
ut_ad(!block == (*err != DB_SUCCESS));
if (UNIV_LIKELY(block != nullptr))
{
if (!!page_is_comp(block->page.frame) != index.table->not_redundant() ||
btr_page_get_index_id(block->page.frame) != index.id ||
!fil_page_index_page_check(block->page.frame) ||
index.is_spatial() !=
(fil_page_get_type(block->page.frame) == FIL_PAGE_RTREE))
{
*err= DB_PAGE_CORRUPTED;
block= nullptr;
}
}
else if (*err == DB_DECRYPTION_FAILED)
btr_decryption_failed(index);
return block;
}
/**************************************************************//**
Gets the root node of a tree and x- or s-latches it.
@return root page, x- or s-latched */
buf_block_t*
btr_root_block_get(
/*===============*/
const dict_index_t* index, /*!< in: index tree */
rw_lock_type_t mode, /*!< in: either RW_S_LATCH
or RW_X_LATCH */
mtr_t* mtr, /*!< in: mtr */
dberr_t* err) /*!< out: error code */
{
if (!index->table || !index->table->space)
{
*err= DB_TABLESPACE_NOT_FOUND;
return nullptr;
}
buf_block_t *block = btr_block_get(*index, index->page, mode, false, mtr,
err);
if (block)
{
if (index->is_ibuf());
else if (!btr_root_fseg_validate(FIL_PAGE_DATA + PAGE_BTR_SEG_LEAF,
*block, *index->table->space) ||
!btr_root_fseg_validate(FIL_PAGE_DATA + PAGE_BTR_SEG_TOP,
*block, *index->table->space))
{
*err= DB_CORRUPTION;
block= nullptr;
}
}
return block;
}
/**************************************************************//**
Gets the root node of a tree and sx-latches it for segment access.
@return root page, sx-latched */
static
page_t*
btr_root_get(
/*=========*/
const dict_index_t* index, /*!< in: index tree */
mtr_t* mtr, /*!< in: mtr */
dberr_t* err) /*!< out: error code */
{
/* Intended to be used for accessing file segment lists.
Concurrent read of other data is allowed. */
if (buf_block_t *root= btr_root_block_get(index, RW_SX_LATCH, mtr, err))
return root->page.frame;
return nullptr;
}
/**************************************************************//**
Checks a file segment header within a B-tree root page and updates
the segment header space id.
@return TRUE if valid */
static
bool
btr_root_fseg_adjust_on_import(
/*===========================*/
fseg_header_t* seg_header, /*!< in/out: segment header */
page_zip_des_t* page_zip, /*!< in/out: compressed page,
or NULL */
ulint space) /*!< in: tablespace identifier */
{
ulint offset = mach_read_from_2(seg_header + FSEG_HDR_OFFSET);
if (offset < FIL_PAGE_DATA
|| offset > srv_page_size - FIL_PAGE_DATA_END) {
return false;
}
seg_header += FSEG_HDR_SPACE;
mach_write_to_4(seg_header, space);
if (UNIV_LIKELY_NULL(page_zip)) {
memcpy(page_zip->data + page_offset(seg_header), seg_header,
4);
}
return true;
}
/**************************************************************//**
Checks and adjusts the root node of a tree during IMPORT TABLESPACE.
@return error code, or DB_SUCCESS */
dberr_t
btr_root_adjust_on_import(
/*======================*/
const dict_index_t* index) /*!< in: index tree */
{
dberr_t err;
mtr_t mtr;
page_t* page;
page_zip_des_t* page_zip;
dict_table_t* table = index->table;
DBUG_EXECUTE_IF("ib_import_trigger_corruption_3",
return(DB_CORRUPTION););
mtr_start(&mtr);
mtr_set_log_mode(&mtr, MTR_LOG_NO_REDO);
buf_block_t* block = buf_page_get_gen(
page_id_t(table->space->id, index->page),
table->space->zip_size(), RW_X_LATCH, NULL, BUF_GET,
&mtr, &err);
if (!block) {
ut_ad(err != DB_SUCCESS);
goto func_exit;
}
page = buf_block_get_frame(block);
page_zip = buf_block_get_page_zip(block);
if (!fil_page_index_page_check(page) || page_has_siblings(page)) {
err = DB_CORRUPTION;
} else if (dict_index_is_clust(index)) {
bool page_is_compact_format;
page_is_compact_format = page_is_comp(page) > 0;
/* Check if the page format and table format agree. */
if (page_is_compact_format != dict_table_is_comp(table)) {
err = DB_CORRUPTION;
} else {
/* Check that the table flags and the tablespace
flags match. */
uint32_t tf = dict_tf_to_fsp_flags(table->flags);
uint32_t sf = table->space->flags;
sf &= ~FSP_FLAGS_MEM_MASK;
tf &= ~FSP_FLAGS_MEM_MASK;
if (fil_space_t::is_flags_equal(tf, sf)
|| fil_space_t::is_flags_equal(sf, tf)) {
mysql_mutex_lock(&fil_system.mutex);
table->space->flags = (table->space->flags
& ~FSP_FLAGS_MEM_MASK)
| (tf & FSP_FLAGS_MEM_MASK);
mysql_mutex_unlock(&fil_system.mutex);
err = DB_SUCCESS;
} else {
err = DB_CORRUPTION;
}
}
} else {
err = DB_SUCCESS;
}
/* Check and adjust the file segment headers, if all OK so far. */
if (err == DB_SUCCESS
&& (!btr_root_fseg_adjust_on_import(
FIL_PAGE_DATA + PAGE_BTR_SEG_LEAF
+ page, page_zip, table->space_id)
|| !btr_root_fseg_adjust_on_import(
FIL_PAGE_DATA + PAGE_BTR_SEG_TOP
+ page, page_zip, table->space_id))) {
err = DB_CORRUPTION;
}
func_exit:
mtr_commit(&mtr);
return(err);
}
/**************************************************************//**
Creates a new index page (not the root, and also not
used in page reorganization). @see btr_page_empty(). */
void
btr_page_create(
/*============*/
buf_block_t* block, /*!< in/out: page to be created */
page_zip_des_t* page_zip,/*!< in/out: compressed page, or NULL */
dict_index_t* index, /*!< in: index */
ulint level, /*!< in: the B-tree level of the page */
mtr_t* mtr) /*!< in: mtr */
{
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
byte *index_id= my_assume_aligned<2>(PAGE_HEADER + PAGE_INDEX_ID +
block->page.frame);
if (UNIV_LIKELY_NULL(page_zip))
{
mach_write_to_8(index_id, index->id);
page_create_zip(block, index, level, 0, mtr);
}
else
{
page_create(block, mtr, dict_table_is_comp(index->table));
if (index->is_spatial())
{
static_assert(((FIL_PAGE_INDEX & 0xff00) | byte(FIL_PAGE_RTREE)) ==
FIL_PAGE_RTREE, "compatibility");
mtr->write<1>(*block, FIL_PAGE_TYPE + 1 + block->page.frame,
byte(FIL_PAGE_RTREE));
if (mach_read_from_8(block->page.frame + FIL_RTREE_SPLIT_SEQ_NUM))
mtr->memset(block, FIL_RTREE_SPLIT_SEQ_NUM, 8, 0);
}
/* Set the level of the new index page */
mtr->write<2,mtr_t::MAYBE_NOP>(*block,
my_assume_aligned<2>(PAGE_HEADER +
PAGE_LEVEL +
block->page.frame),
level);
mtr->write<8,mtr_t::MAYBE_NOP>(*block, index_id, index->id);
}
}
/**************************************************************//**
Allocates a new file page to be used in an ibuf tree. Takes the page from
the free list of the tree, which must contain pages!
@return new allocated block, x-latched */
static
buf_block_t*
btr_page_alloc_for_ibuf(
/*====================*/
dict_index_t* index, /*!< in: index tree */
mtr_t* mtr, /*!< in: mtr */
dberr_t* err) /*!< out: error code */
{
buf_block_t *root= btr_root_block_get(index, RW_SX_LATCH, mtr, err);
if (UNIV_UNLIKELY(!root))
return root;
buf_block_t *new_block=
buf_page_get_gen(page_id_t(index->table->space_id,
mach_read_from_4(PAGE_HEADER +
PAGE_BTR_IBUF_FREE_LIST +
FLST_FIRST + FIL_ADDR_PAGE +
root->page.frame)),
index->table->space->zip_size(), RW_X_LATCH, nullptr,
BUF_GET, mtr, err);
if (new_block)
*err= flst_remove(root, PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST, new_block,
PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST_NODE, mtr);
ut_d(if (*err == DB_SUCCESS)
flst_validate(root, PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST, mtr));
return new_block;
}
/**************************************************************//**
Allocates a new file page to be used in an index tree. NOTE: we assume
that the caller has made the reservation for free extents!
@retval NULL if no page could be allocated */
static MY_ATTRIBUTE((nonnull, warn_unused_result))
buf_block_t*
btr_page_alloc_low(
/*===============*/
dict_index_t* index, /*!< in: index */
uint32_t hint_page_no, /*!< in: hint of a good page */
byte file_direction, /*!< in: direction where a possible
page split is made */
ulint level, /*!< in: level where the page is placed
in the tree */
mtr_t* mtr, /*!< in/out: mini-transaction
for the allocation */
mtr_t* init_mtr, /*!< in/out: mtr or another
mini-transaction in which the
page should be initialized. */
dberr_t* err) /*!< out: error code */
{
const auto savepoint= mtr->get_savepoint();
buf_block_t *root= btr_root_block_get(index, RW_NO_LATCH, mtr, err);
if (UNIV_UNLIKELY(!root))
return root;
if (mtr->have_u_or_x_latch(*root))
{
#ifdef BTR_CUR_HASH_ADAPT
ut_ad(!root->index || !root->index->freed());
#endif
mtr->release_block_at_savepoint(savepoint, root);
}
else
{
mtr->u_lock_register(savepoint);
root->page.lock.u_lock();
#ifdef BTR_CUR_HASH_ADAPT
btr_search_drop_page_hash_index(root, true);
#endif
}
fseg_header_t *seg_header= root->page.frame +
(level ? PAGE_HEADER + PAGE_BTR_SEG_TOP : PAGE_HEADER + PAGE_BTR_SEG_LEAF);
return fseg_alloc_free_page_general(seg_header, hint_page_no, file_direction,
true, mtr, init_mtr, err);
}
/**************************************************************//**
Allocates a new file page to be used in an index tree. NOTE: we assume
that the caller has made the reservation for free extents!
@retval NULL if no page could be allocated */
buf_block_t*
btr_page_alloc(
/*===========*/
dict_index_t* index, /*!< in: index */
uint32_t hint_page_no, /*!< in: hint of a good page */
byte file_direction, /*!< in: direction where a possible
page split is made */
ulint level, /*!< in: level where the page is placed
in the tree */
mtr_t* mtr, /*!< in/out: mini-transaction
for the allocation */
mtr_t* init_mtr, /*!< in/out: mini-transaction
for x-latching and initializing
the page */
dberr_t* err) /*!< out: error code */
{
ut_ad(level < BTR_MAX_NODE_LEVEL);
return index->is_ibuf()
? btr_page_alloc_for_ibuf(index, mtr, err)
: btr_page_alloc_low(index, hint_page_no, file_direction, level,
mtr, init_mtr, err);
}
/**************************************************************//**
Frees a page used in an ibuf tree. Puts the page to the free list of the
ibuf tree. */
static
dberr_t
btr_page_free_for_ibuf(
/*===================*/
dict_index_t* index, /*!< in: index tree */
buf_block_t* block, /*!< in: block to be freed, x-latched */
mtr_t* mtr) /*!< in: mtr */
{
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
dberr_t err;
if (buf_block_t *root= btr_root_block_get(index, RW_SX_LATCH, mtr, &err))
{
err= flst_add_first(root, PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST,
block, PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST_NODE, mtr);
ut_d(if (err == DB_SUCCESS)
flst_validate(root, PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST, mtr));
}
return err;
}
/** Free an index page.
@param[in,out] index index tree
@param[in,out] block block to be freed
@param[in,out] mtr mini-transaction
@param[in] blob whether this is freeing a BLOB page
@param[in] latched whether index->table->space->x_lock() was called
@return error code */
dberr_t btr_page_free(dict_index_t* index, buf_block_t* block, mtr_t* mtr,
bool blob, bool space_latched)
{
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
#if defined BTR_CUR_HASH_ADAPT && defined UNIV_DEBUG
if (btr_search_check_marked_free_index(block))
{
ut_ad(!blob);
ut_ad(page_is_leaf(block->page.frame));
}
#endif
const uint32_t page{block->page.id().page_no()};
ut_ad(index->table->space_id == block->page.id().space());
/* The root page is freed by btr_free_root(). */
ut_ad(page != index->page);
ut_ad(mtr->is_named_space(index->table->space));
/* The page gets invalid for optimistic searches: increment the frame
modify clock */
buf_block_modify_clock_inc(block);
/* TODO: Discard any operations for block from mtr->m_log.
The page will be freed, so previous changes to it by this
mini-transaction should not matter. */
if (index->is_ibuf())
return btr_page_free_for_ibuf(index, block, mtr);
fil_space_t *space= index->table->space;
dberr_t err;
const auto savepoint= mtr->get_savepoint();
if (buf_block_t *root= btr_root_block_get(index, RW_NO_LATCH, mtr, &err))
{
if (mtr->have_u_or_x_latch(*root))
{
#ifdef BTR_CUR_HASH_ADAPT
ut_ad(!root->index || !root->index->freed());
#endif
mtr->release_block_at_savepoint(savepoint, root);
}
else
{
mtr->u_lock_register(savepoint);
root->page.lock.u_lock();
#ifdef BTR_CUR_HASH_ADAPT
btr_search_drop_page_hash_index(root, true);
#endif
}
err= fseg_free_page(&root->page.frame[blob ||
page_is_leaf(block->page.frame)
? PAGE_HEADER + PAGE_BTR_SEG_LEAF
: PAGE_HEADER + PAGE_BTR_SEG_TOP],
space, page, mtr, space_latched);
}
if (err == DB_SUCCESS)
buf_page_free(space, page, mtr);
/* The page was marked free in the allocation bitmap, but it
should remain exclusively latched until mtr_t::commit() or until it
is explicitly freed from the mini-transaction. */
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
return err;
}
/** Set the child page number in a node pointer record.
@param[in,out] block non-leaf index page
@param[in,out] rec node pointer record in the page
@param[in] offsets rec_get_offsets(rec)
@param[in] page_no child page number
@param[in,out] mtr mini-transaction
Sets the child node file address in a node pointer. */
inline void btr_node_ptr_set_child_page_no(buf_block_t *block,
rec_t *rec, const rec_offs *offsets,
ulint page_no, mtr_t *mtr)
{
ut_ad(rec_offs_validate(rec, NULL, offsets));
ut_ad(!page_rec_is_leaf(rec));
ut_ad(!rec_offs_comp(offsets) || rec_get_node_ptr_flag(rec));
const ulint offs= rec_offs_data_size(offsets);
ut_ad(rec_offs_nth_size(offsets, rec_offs_n_fields(offsets) - 1) ==
REC_NODE_PTR_SIZE);
if (UNIV_LIKELY_NULL(block->page.zip.data))
page_zip_write_node_ptr(block, rec, offs, page_no, mtr);
else
mtr->write<4>(*block, rec + offs - REC_NODE_PTR_SIZE, page_no);
}
MY_ATTRIBUTE((nonnull(1,2,3,4),warn_unused_result))
/************************************************************//**
Returns the child page of a node pointer and sx-latches it.
@return child page, sx-latched */
static
buf_block_t*
btr_node_ptr_get_child(
/*===================*/
const rec_t* node_ptr,/*!< in: node pointer */
dict_index_t* index, /*!< in: index */
const rec_offs* offsets,/*!< in: array returned by rec_get_offsets() */
mtr_t* mtr, /*!< in: mtr */
dberr_t* err = nullptr) /*!< out: error code */
{
ut_ad(rec_offs_validate(node_ptr, index, offsets));
ut_ad(index->table->space_id
== page_get_space_id(page_align(node_ptr)));
return btr_block_get(
*index, btr_node_ptr_get_child_page_no(node_ptr, offsets),
RW_SX_LATCH, btr_page_get_level(page_align(node_ptr)) == 1,
mtr, err);
}
MY_ATTRIBUTE((nonnull(2,3,5), warn_unused_result))
/************************************************************//**
Returns the upper level node pointer to a page. It is assumed that mtr holds
an sx-latch on the tree.
@return rec_get_offsets() of the node pointer record */
static
rec_offs*
btr_page_get_father_node_ptr_func(
/*==============================*/
rec_offs* offsets,/*!< in: work area for the return value */
mem_heap_t* heap, /*!< in: memory heap to use */
btr_cur_t* cursor, /*!< in: cursor pointing to user record,
out: cursor on node pointer record,
its page x-latched */
ulint latch_mode,/*!< in: BTR_CONT_MODIFY_TREE
or BTR_CONT_SEARCH_TREE */
mtr_t* mtr) /*!< in: mtr */
{
ut_ad(latch_mode == BTR_CONT_MODIFY_TREE
|| latch_mode == BTR_CONT_SEARCH_TREE);
const uint32_t page_no = btr_cur_get_block(cursor)->page.id().page_no();
dict_index_t* index = btr_cur_get_index(cursor);
ut_ad(!dict_index_is_spatial(index));
ut_ad(srv_read_only_mode
|| mtr->memo_contains_flagged(&index->lock, MTR_MEMO_X_LOCK
| MTR_MEMO_SX_LOCK));
ut_ad(dict_index_get_page(index) != page_no);
const auto level = btr_page_get_level(btr_cur_get_page(cursor));
const rec_t* user_rec = btr_cur_get_rec(cursor);
ut_a(page_rec_is_user_rec(user_rec));
if (btr_cur_search_to_nth_level(index, level + 1,
dict_index_build_node_ptr(index,
user_rec, 0,
heap, level),
PAGE_CUR_LE, latch_mode,
cursor, 0, mtr) != DB_SUCCESS) {
return nullptr;
}
const rec_t* node_ptr = btr_cur_get_rec(cursor);
offsets = rec_get_offsets(node_ptr, index, offsets, 0,
ULINT_UNDEFINED, &heap);
if (btr_node_ptr_get_child_page_no(node_ptr, offsets) != page_no) {
offsets = nullptr;
}
return(offsets);
}
#define btr_page_get_father_node_ptr(of,heap,cur,mtr) \
btr_page_get_father_node_ptr_func( \
of,heap,cur,BTR_CONT_MODIFY_TREE,mtr)
#define btr_page_get_father_node_ptr_for_validate(of,heap,cur,mtr) \
btr_page_get_father_node_ptr_func( \
of,heap,cur,BTR_CONT_SEARCH_TREE,mtr)
/************************************************************//**
Returns the upper level node pointer to a page. It is assumed that mtr holds
an x-latch on the tree.
@return rec_get_offsets() of the node pointer record */
static
rec_offs*
btr_page_get_father_block(
/*======================*/
rec_offs* offsets,/*!< in: work area for the return value */
mem_heap_t* heap, /*!< in: memory heap to use */
dict_index_t* index, /*!< in: b-tree index */
buf_block_t* block, /*!< in: child page in the index */
mtr_t* mtr, /*!< in: mtr */
btr_cur_t* cursor) /*!< out: cursor on node pointer record,
its page x-latched */
{
rec_t* rec
= page_rec_get_next(page_get_infimum_rec(buf_block_get_frame(
block)));
if (UNIV_UNLIKELY(!rec)) {
return nullptr;
}
btr_cur_position(index, rec, block, cursor);
return(btr_page_get_father_node_ptr(offsets, heap, cursor, mtr));
}
/** Seek to the parent page of a B-tree page.
@param[in,out] index b-tree
@param[in] block child page
@param[in,out] mtr mini-transaction
@param[out] cursor cursor pointing to the x-latched parent page
@return whether the cursor was successfully positioned */
bool btr_page_get_father(dict_index_t* index, buf_block_t* block, mtr_t* mtr,
btr_cur_t* cursor)
{
mem_heap_t* heap;
rec_t* rec
= page_rec_get_next(page_get_infimum_rec(buf_block_get_frame(
block)));
if (UNIV_UNLIKELY(!rec)) {
return false;
}
btr_cur_position(index, rec, block, cursor);
heap = mem_heap_create(100);
const bool got = btr_page_get_father_node_ptr(NULL, heap, cursor, mtr);
mem_heap_free(heap);
return got;
}
#ifdef UNIV_DEBUG
/** PAGE_INDEX_ID value for freed index B-trees */
constexpr index_id_t BTR_FREED_INDEX_ID = 0;
#endif
/** Free a B-tree root page. btr_free_but_not_root() must already
have been called.
@param block index root page
@param space tablespace
@param mtr mini-transaction */
static void btr_free_root(buf_block_t *block, const fil_space_t &space,
mtr_t *mtr)
{
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX |
MTR_MEMO_PAGE_SX_FIX));
ut_ad(mtr->is_named_space(&space));
btr_search_drop_page_hash_index(block);
if (btr_root_fseg_validate(PAGE_HEADER + PAGE_BTR_SEG_TOP, *block, space))
{
/* Free the entire segment in small steps. */
ut_d(mtr->freeing_tree());
while (!fseg_free_step(PAGE_HEADER + PAGE_BTR_SEG_TOP +
block->page.frame, mtr));
}
}
MY_ATTRIBUTE((warn_unused_result))
/** Prepare to free a B-tree.
@param[in] page_id page id
@param[in] zip_size ROW_FORMAT=COMPRESSED page size, or 0
@param[in] index_id PAGE_INDEX_ID contents
@param[in,out] mtr mini-transaction
@return root block, to invoke btr_free_but_not_root() and btr_free_root()
@retval NULL if the page is no longer a matching B-tree page */
static
buf_block_t *btr_free_root_check(const page_id_t page_id, ulint zip_size,
index_id_t index_id, mtr_t *mtr)
{
ut_ad(page_id.space() != SRV_TMP_SPACE_ID);
ut_ad(index_id != BTR_FREED_INDEX_ID);
buf_block_t *block= buf_page_get_gen(page_id, zip_size, RW_X_LATCH,
nullptr, BUF_GET_POSSIBLY_FREED, mtr);
if (!block);
else if (fil_page_index_page_check(block->page.frame) &&
index_id == btr_page_get_index_id(block->page.frame))
/* This should be a root page. It should not be possible to
reassign the same index_id for some other index in the
tablespace. */
ut_ad(!page_has_siblings(block->page.frame));
else
block= nullptr;
return block;
}
/** Initialize the root page of the b-tree
@param[in,out] block root block
@param[in] index_id index id
@param[in] index index of root page
@param[in,out] mtr mini-transaction */
static void btr_root_page_init(buf_block_t *block, index_id_t index_id,
dict_index_t *index, mtr_t *mtr)
{
constexpr uint16_t field= PAGE_HEADER + PAGE_INDEX_ID;
byte *page_index_id= my_assume_aligned<2>(field + block->page.frame);
/* Create a new index page on the allocated segment page */
if (UNIV_LIKELY_NULL(block->page.zip.data))
{
mach_write_to_8(page_index_id, index_id);
ut_ad(!page_has_siblings(block->page.zip.data));
page_create_zip(block, index, 0, 0, mtr);
}
else
{
page_create(block, mtr, index && index->table->not_redundant());
if (index && index->is_spatial())
{
static_assert(((FIL_PAGE_INDEX & 0xff00) | byte(FIL_PAGE_RTREE)) ==
FIL_PAGE_RTREE, "compatibility");
mtr->write<1>(*block, FIL_PAGE_TYPE + 1 + block->page.frame,
byte(FIL_PAGE_RTREE));
if (mach_read_from_8(block->page.frame + FIL_RTREE_SPLIT_SEQ_NUM))
mtr->memset(block, FIL_RTREE_SPLIT_SEQ_NUM, 8, 0);
}
/* Set the level of the new index page */
mtr->write<2,mtr_t::MAYBE_NOP>(
*block, PAGE_HEADER + PAGE_LEVEL + block->page.frame, 0U);
mtr->write<8,mtr_t::MAYBE_NOP>(*block, page_index_id, index_id);
}
}
/** Create the root node for a new index tree.
@param[in] type type of the index
@param[in] index_id index id
@param[in,out] space tablespace where created
@param[in] index index, or NULL to create a system table
@param[in,out] mtr mini-transaction
@param[out] err error code
@return page number of the created root
@retval FIL_NULL if did not succeed */
uint32_t
btr_create(
ulint type,
fil_space_t* space,
index_id_t index_id,
dict_index_t* index,
mtr_t* mtr,
dberr_t* err)
{
buf_block_t* block;
ut_ad(mtr->is_named_space(space));
ut_ad(index_id != BTR_FREED_INDEX_ID);
ut_ad(index || space == fil_system.sys_space);
/* Create the two new segments (one, in the case of an ibuf tree) for
the index tree; the segment headers are put on the allocated root page
(for an ibuf tree, not in the root, but on a separate ibuf header
page) */
if (UNIV_UNLIKELY(type & DICT_IBUF)) {
/* Allocate first the ibuf header page */
buf_block_t* ibuf_hdr_block = fseg_create(
space, IBUF_HEADER + IBUF_TREE_SEG_HEADER, mtr, err);
if (ibuf_hdr_block == NULL) {
return(FIL_NULL);
}
ut_ad(ibuf_hdr_block->page.id().page_no()
== IBUF_HEADER_PAGE_NO);
/* Allocate then the next page to the segment: it will be the
tree root page */
block = fseg_alloc_free_page_general(
buf_block_get_frame(ibuf_hdr_block)
+ IBUF_HEADER + IBUF_TREE_SEG_HEADER,
IBUF_TREE_ROOT_PAGE_NO,
FSP_UP, false, mtr, mtr, err);
if (block == NULL) {
return(FIL_NULL);
}
ut_ad(block->page.id() == page_id_t(0,IBUF_TREE_ROOT_PAGE_NO));
flst_init(block, PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST, mtr);
} else {
block = fseg_create(space, PAGE_HEADER + PAGE_BTR_SEG_TOP,
mtr, err);
if (block == NULL) {
return(FIL_NULL);
}
if (!fseg_create(space, PAGE_HEADER + PAGE_BTR_SEG_LEAF, mtr,
err, false, block)) {
/* Not enough space for new segment, free root
segment before return. */
btr_free_root(block, *space, mtr);
return(FIL_NULL);
}
}
ut_ad(!page_has_siblings(block->page.frame));
btr_root_page_init(block, index_id, index, mtr);
/* We reset the free bits for the page in a separate
mini-transaction to allow creation of several trees in the
same mtr, otherwise the latch on a bitmap page would prevent
it because of the latching order.
Note: Insert Buffering is disabled for temporary tables given that
most temporary tables are smaller in size and short-lived. */
if (!(type & DICT_CLUSTERED)
&& (!index || !index->table->is_temporary())) {
ibuf_reset_free_bits(block);
}
/* In the following assertion we test that two records of maximum
allowed size fit on the root page: this fact is needed to ensure
correctness of split algorithms */
ut_ad(page_get_max_insert_size(block->page.frame, 2)
> 2 * BTR_PAGE_MAX_REC_SIZE);
return(block->page.id().page_no());
}
/** Free a B-tree except the root page. The root page MUST be freed after
this by calling btr_free_root.
@param[in,out] block root page
@param[in] log_mode mtr logging mode */
static
void
btr_free_but_not_root(
buf_block_t* block,
mtr_log_t log_mode
#ifdef BTR_CUR_HASH_ADAPT
,bool ahi=false
#endif
)
{
mtr_t mtr;
ut_ad(fil_page_index_page_check(block->page.frame));
ut_ad(!page_has_siblings(block->page.frame));
leaf_loop:
mtr_start(&mtr);
ut_d(mtr.freeing_tree());
mtr_set_log_mode(&mtr, log_mode);
fil_space_t *space = mtr.set_named_space_id(block->page.id().space());
if (!btr_root_fseg_validate(FIL_PAGE_DATA + PAGE_BTR_SEG_LEAF,
*block, *space)
|| !btr_root_fseg_validate(FIL_PAGE_DATA + PAGE_BTR_SEG_TOP,
*block, *space)) {
mtr_commit(&mtr);
return;
}
/* NOTE: page hash indexes are dropped when a page is freed inside
fsp0fsp. */
bool finished = fseg_free_step(PAGE_HEADER + PAGE_BTR_SEG_LEAF
+ block->page.frame, &mtr
#ifdef BTR_CUR_HASH_ADAPT
, ahi
#endif /* BTR_CUR_HASH_ADAPT */
);
mtr_commit(&mtr);
if (!finished) {
goto leaf_loop;
}
top_loop:
mtr_start(&mtr);
mtr_set_log_mode(&mtr, log_mode);
space = mtr.set_named_space_id(block->page.id().space());
finished = !btr_root_fseg_validate(FIL_PAGE_DATA + PAGE_BTR_SEG_TOP,
*block, *space)
|| fseg_free_step_not_header(PAGE_HEADER + PAGE_BTR_SEG_TOP
+ block->page.frame, &mtr
#ifdef BTR_CUR_HASH_ADAPT
,ahi
#endif /* BTR_CUR_HASH_ADAPT */
);
mtr_commit(&mtr);
if (!finished) {
goto top_loop;
}
}
/** Clear the index tree and reinitialize the root page, in the
rollback of TRX_UNDO_EMPTY. The BTR_SEG_LEAF is freed and reinitialized.
@param thr query thread
@return error code */
TRANSACTIONAL_TARGET
dberr_t dict_index_t::clear(que_thr_t *thr)
{
mtr_t mtr;
mtr.start();
if (table->is_temporary())
mtr.set_log_mode(MTR_LOG_NO_REDO);
else
set_modified(mtr);
dberr_t err;
if (buf_block_t *root_block=
buf_page_get_gen(page_id_t(table->space->id, page),
table->space->zip_size(),
RW_X_LATCH, nullptr, BUF_GET, &mtr, &err))
{
btr_free_but_not_root(root_block, mtr.get_log_mode()
#ifdef BTR_CUR_HASH_ADAPT
,n_ahi_pages() != 0
#endif
);
#ifdef BTR_CUR_HASH_ADAPT
if (root_block->index)
btr_search_drop_page_hash_index(root_block);
ut_ad(n_ahi_pages() == 0);
#endif
mtr.memset(root_block, PAGE_HEADER + PAGE_BTR_SEG_LEAF,
FSEG_HEADER_SIZE, 0);
if (fseg_create(table->space, PAGE_HEADER + PAGE_BTR_SEG_LEAF, &mtr,
&err, false, root_block))
btr_root_page_init(root_block, id, this, &mtr);
}
mtr.commit();
return err;
}
/** Free a persistent index tree if it exists.
@param[in,out] space tablespce
@param[in] page root page number
@param[in] index_id PAGE_INDEX_ID contents
@param[in,out] mtr mini-transaction */
void btr_free_if_exists(fil_space_t *space, uint32_t page,
index_id_t index_id, mtr_t *mtr)
{
if (buf_block_t *root= btr_free_root_check(page_id_t(space->id, page),
space->zip_size(),
index_id, mtr))
{
btr_free_but_not_root(root, mtr->get_log_mode());
mtr->set_named_space(space);
btr_free_root(root, *space, mtr);
}
}
/** Drop a temporary table
@param table temporary table */
void btr_drop_temporary_table(const dict_table_t &table)
{
ut_ad(table.is_temporary());
ut_ad(table.space == fil_system.temp_space);
mtr_t mtr;
mtr.start();
for (const dict_index_t *index= table.indexes.start; index;
index= dict_table_get_next_index(index))
{
if (buf_block_t *block= buf_page_get_low({SRV_TMP_SPACE_ID, index->page}, 0,
RW_X_LATCH, nullptr, BUF_GET, &mtr,
nullptr, false))
{
btr_free_but_not_root(block, MTR_LOG_NO_REDO);
mtr.set_log_mode(MTR_LOG_NO_REDO);
btr_free_root(block, *fil_system.temp_space, &mtr);
mtr.commit();
mtr.start();
}
}
mtr.commit();
}
/** Read the last used AUTO_INCREMENT value from PAGE_ROOT_AUTO_INC.
@param[in,out] index clustered index
@return the last used AUTO_INCREMENT value
@retval 0 on error or if no AUTO_INCREMENT value was used yet */
ib_uint64_t
btr_read_autoinc(dict_index_t* index)
{
ut_ad(index->is_primary());
ut_ad(index->table->persistent_autoinc);
ut_ad(!index->table->is_temporary());
mtr_t mtr;
mtr.start();
ib_uint64_t autoinc;
if (buf_block_t* block = buf_page_get(
page_id_t(index->table->space_id, index->page),
index->table->space->zip_size(),
RW_S_LATCH, &mtr)) {
autoinc = page_get_autoinc(block->page.frame);
} else {
autoinc = 0;
}
mtr.commit();
return autoinc;
}
/** Read the last used AUTO_INCREMENT value from PAGE_ROOT_AUTO_INC,
or fall back to MAX(auto_increment_column).
@param[in] table table containing an AUTO_INCREMENT column
@param[in] col_no index of the AUTO_INCREMENT column
@return the AUTO_INCREMENT value
@retval 0 on error or if no AUTO_INCREMENT value was used yet */
ib_uint64_t
btr_read_autoinc_with_fallback(const dict_table_t* table, unsigned col_no)
{
ut_ad(table->persistent_autoinc);
ut_ad(!table->is_temporary());
dict_index_t* index = dict_table_get_first_index(table);
if (index == NULL) {
return 0;
}
mtr_t mtr;
mtr.start();
buf_block_t* block = buf_page_get(
page_id_t(index->table->space_id, index->page),
index->table->space->zip_size(),
RW_S_LATCH, &mtr);
ib_uint64_t autoinc = block
? page_get_autoinc(block->page.frame) : 0;
const bool retry = block && autoinc == 0
&& !page_is_empty(block->page.frame);
mtr.commit();
if (retry) {
/* This should be an old data file where
PAGE_ROOT_AUTO_INC was initialized to 0.
Fall back to reading MAX(autoinc_col).
There should be an index on it. */
const dict_col_t* autoinc_col
= dict_table_get_nth_col(table, col_no);
while (index && index->fields[0].col != autoinc_col) {
index = dict_table_get_next_index(index);
}
if (index) {
autoinc = row_search_max_autoinc(index);
}
}
return autoinc;
}
/** Write the next available AUTO_INCREMENT value to PAGE_ROOT_AUTO_INC.
@param[in,out] index clustered index
@param[in] autoinc the AUTO_INCREMENT value
@param[in] reset whether to reset the AUTO_INCREMENT
to a possibly smaller value than currently
exists in the page */
void
btr_write_autoinc(dict_index_t* index, ib_uint64_t autoinc, bool reset)
{
ut_ad(index->is_primary());
ut_ad(index->table->persistent_autoinc);
ut_ad(!index->table->is_temporary());
mtr_t mtr;
mtr.start();
fil_space_t *space= index->table->space;
if (buf_block_t *root= buf_page_get(page_id_t(space->id, index->page),
space->zip_size(), RW_SX_LATCH, &mtr))
{
mtr.set_named_space(space);
page_set_autoinc(root, autoinc, &mtr, reset);
}
mtr.commit();
}
/** Reorganize an index page.
@param cursor index page cursor
@param index the index that the cursor belongs to
@param mtr mini-transaction */
static dberr_t btr_page_reorganize_low(page_cur_t *cursor, dict_index_t *index,
mtr_t *mtr)
{
buf_block_t *const block= cursor->block;
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
ut_ad(!is_buf_block_get_page_zip(block));
ut_ad(fil_page_index_page_check(block->page.frame));
ut_ad(index->is_dummy ||
block->page.id().space() == index->table->space->id);
ut_ad(index->is_dummy || block->page.id().page_no() != index->page ||
!page_has_siblings(block->page.frame));
/* Save the cursor position. */
const ulint pos= page_rec_get_n_recs_before(cursor->rec);
if (UNIV_UNLIKELY(pos == ULINT_UNDEFINED))
return DB_CORRUPTION;
btr_search_drop_page_hash_index(block);
buf_block_t *old= buf_block_alloc();
/* Copy the old page to temporary space */
memcpy_aligned<UNIV_PAGE_SIZE_MIN>(old->page.frame, block->page.frame,
srv_page_size);
const mtr_log_t log_mode= mtr->set_log_mode(MTR_LOG_NO_REDO);
page_create(block, mtr, index->table->not_redundant());
if (index->is_spatial())
block->page.frame[FIL_PAGE_TYPE + 1]= byte(FIL_PAGE_RTREE);
static_assert(((FIL_PAGE_INDEX & 0xff00) | byte(FIL_PAGE_RTREE)) ==
FIL_PAGE_RTREE, "compatibility");
/* Copy the records from the temporary space to the recreated page;
do not copy the lock bits yet */
dberr_t err=
page_copy_rec_list_end_no_locks(block, old,
page_get_infimum_rec(old->page.frame),
index, mtr);
mtr->set_log_mode(log_mode);
if (UNIV_UNLIKELY(err != DB_SUCCESS))
return err;
/* Copy the PAGE_MAX_TRX_ID or PAGE_ROOT_AUTO_INC. */
ut_ad(!page_get_max_trx_id(block->page.frame));
memcpy_aligned<8>(PAGE_MAX_TRX_ID + PAGE_HEADER + block->page.frame,
PAGE_MAX_TRX_ID + PAGE_HEADER + old->page.frame, 8);
#ifdef UNIV_DEBUG
if (page_get_max_trx_id(block->page.frame))
/* PAGE_MAX_TRX_ID must be zero on non-leaf pages other than
clustered index root pages. */
ut_ad(dict_index_is_sec_or_ibuf(index)
? page_is_leaf(block->page.frame)
: block->page.id().page_no() == index->page);
else
/* PAGE_MAX_TRX_ID is unused in clustered index pages (other than
the root where it is repurposed as PAGE_ROOT_AUTO_INC), non-leaf
pages, and in temporary tables. It was always zero-initialized in
page_create(). PAGE_MAX_TRX_ID must be nonzero on
dict_index_is_sec_or_ibuf() leaf pages. */
ut_ad(index->table->is_temporary() || !page_is_leaf(block->page.frame) ||
!dict_index_is_sec_or_ibuf(index));
#endif
const uint16_t data_size1= page_get_data_size(old->page.frame);
const uint16_t data_size2= page_get_data_size(block->page.frame);
const ulint max1=
page_get_max_insert_size_after_reorganize(old->page.frame, 1);
const ulint max2=
page_get_max_insert_size_after_reorganize(block->page.frame, 1);
if (UNIV_UNLIKELY(data_size1 != data_size2 || max1 != max2))
{
sql_print_error("InnoDB: Page old data size %u new data size %u"
", page old max ins size %zu new max ins size %zu",
data_size1, data_size2, max1, max2);
return DB_CORRUPTION;
}
/* Restore the cursor position. */
if (!pos)
ut_ad(cursor->rec == page_get_infimum_rec(block->page.frame));
else if (!(cursor->rec= page_rec_get_nth(block->page.frame, pos)))
return DB_CORRUPTION;
if (block->page.id().page_no() != index->page ||
fil_page_get_type(old->page.frame) != FIL_PAGE_TYPE_INSTANT)
ut_ad(!memcmp(old->page.frame, block->page.frame, PAGE_HEADER));
else if (!index->is_instant())
{
ut_ad(!memcmp(old->page.frame, block->page.frame, FIL_PAGE_TYPE));
ut_ad(!memcmp(old->page.frame + FIL_PAGE_TYPE + 2,
block->page.frame + FIL_PAGE_TYPE + 2,
PAGE_HEADER - FIL_PAGE_TYPE - 2));
mtr->write<2,mtr_t::FORCED>(*block, FIL_PAGE_TYPE + block->page.frame,
FIL_PAGE_INDEX);
}
else
{
/* Preserve the PAGE_INSTANT information. */
memcpy_aligned<2>(FIL_PAGE_TYPE + block->page.frame,
FIL_PAGE_TYPE + old->page.frame, 2);
memcpy_aligned<2>(PAGE_HEADER + PAGE_INSTANT + block->page.frame,
PAGE_HEADER + PAGE_INSTANT + old->page.frame, 2);
if (!index->table->instant);
else if (page_is_comp(block->page.frame))
{
memcpy(PAGE_NEW_INFIMUM + block->page.frame,
PAGE_NEW_INFIMUM + old->page.frame, 8);
memcpy(PAGE_NEW_SUPREMUM + block->page.frame,
PAGE_NEW_SUPREMUM + old->page.frame, 8);
}
else
{
memcpy(PAGE_OLD_INFIMUM + block->page.frame,
PAGE_OLD_INFIMUM + old->page.frame, 8);
memcpy(PAGE_OLD_SUPREMUM + block->page.frame,
PAGE_OLD_SUPREMUM + old->page.frame, 8);
}
ut_ad(!memcmp(old->page.frame, block->page.frame, PAGE_HEADER));
}
ut_ad(!memcmp(old->page.frame + PAGE_MAX_TRX_ID + PAGE_HEADER,
block->page.frame + PAGE_MAX_TRX_ID + PAGE_HEADER,
PAGE_DATA - (PAGE_MAX_TRX_ID + PAGE_HEADER)));
if (index->has_locking())
lock_move_reorganize_page(block, old);
/* Write log for the changes, if needed. */
if (log_mode == MTR_LOG_ALL)
{
/* Check and log the changes in the page header. */
ulint a, e;
for (a= PAGE_HEADER, e= PAGE_MAX_TRX_ID + PAGE_HEADER; a < e; a++)
{
if (old->page.frame[a] == block->page.frame[a])
continue;
while (--e, old->page.frame[e] == block->page.frame[e]);
e++;
ut_ad(a < e);
/* Write log for the changed page header fields. */
mtr->memcpy(*block, a, e - a);
break;
}
const uint16_t top= page_header_get_offs(block->page.frame, PAGE_HEAP_TOP);
if (page_is_comp(block->page.frame))
{
/* info_bits=0, n_owned=1, heap_no=0, status */
ut_ad(!memcmp(PAGE_NEW_INFIMUM - REC_N_NEW_EXTRA_BYTES +
block->page.frame,
PAGE_NEW_INFIMUM - REC_N_NEW_EXTRA_BYTES +
old->page.frame, 3));
/* If the 'next' pointer of the infimum record has changed, log it. */
a= PAGE_NEW_INFIMUM - 2;
e= a + 2;
if (block->page.frame[a] == old->page.frame[a])
a++;
if (--e, block->page.frame[e] != old->page.frame[e])
e++;
if (ulint len= e - a)
mtr->memcpy(*block, a, len);
/* The infimum record itself must not change. */
ut_ad(!memcmp(PAGE_NEW_INFIMUM + block->page.frame,
PAGE_NEW_INFIMUM + old->page.frame, 8));
/* Log any change of the n_owned of the supremum record. */
a= PAGE_NEW_SUPREMUM - REC_N_NEW_EXTRA_BYTES;
if (block->page.frame[a] != old->page.frame[a])
mtr->memcpy(*block, a, 1);
/* The rest of the supremum record must not change. */
ut_ad(!memcmp(&block->page.frame[a + 1], &old->page.frame[a + 1],
PAGE_NEW_SUPREMUM_END - PAGE_NEW_SUPREMUM +
REC_N_NEW_EXTRA_BYTES - 1));
/* Log the differences in the payload. */
for (a= PAGE_NEW_SUPREMUM_END, e= top; a < e; a++)
{
if (old->page.frame[a] == block->page.frame[a])
continue;
while (--e, old->page.frame[e] == block->page.frame[e]);
e++;
ut_ad(a < e);
/* TODO: write MEMMOVE records to minimize this further! */
mtr->memcpy(*block, a, e - a);
break;
}
}
else
{
/* info_bits=0, n_owned=1, heap_no=0, number of fields, 1-byte format */
ut_ad(!memcmp(PAGE_OLD_INFIMUM - REC_N_OLD_EXTRA_BYTES +
block->page.frame,
PAGE_OLD_INFIMUM - REC_N_OLD_EXTRA_BYTES +
old->page.frame, 4));
/* If the 'next' pointer of the infimum record has changed, log it. */
a= PAGE_OLD_INFIMUM - 2;
e= a + 2;
if (block->page.frame[a] == old->page.frame[a])
a++;
if (--e, block->page.frame[e] != old->page.frame[e])
e++;
if (ulint len= e - a)
mtr->memcpy(*block, a, len);
/* The infimum record itself must not change. */
ut_ad(!memcmp(PAGE_OLD_INFIMUM + block->page.frame,
PAGE_OLD_INFIMUM + old->page.frame, 8));
/* Log any change of the n_owned of the supremum record. */
a= PAGE_OLD_SUPREMUM - REC_N_OLD_EXTRA_BYTES;
if (block->page.frame[a] != old->page.frame[a])
mtr->memcpy(*block, a, 1);
ut_ad(!memcmp(&block->page.frame[a + 1], &old->page.frame[a + 1],
PAGE_OLD_SUPREMUM_END - PAGE_OLD_SUPREMUM +
REC_N_OLD_EXTRA_BYTES - 1));
/* Log the differences in the payload. */
for (a= PAGE_OLD_SUPREMUM_END, e= top; a < e; a++)
{
if (old->page.frame[a] == block->page.frame[a])
continue;
while (--e, old->page.frame[e] == block->page.frame[e]);
e++;
ut_ad(a < e);
/* TODO: write MEMMOVE records to minimize this further! */
mtr->memcpy(*block, a, e - a);
break;
}
}
e= srv_page_size - PAGE_DIR;
a= e - PAGE_DIR_SLOT_SIZE * page_dir_get_n_slots(block->page.frame);
/* Zero out the payload area. */
mtr->memset(*block, top, a - top, 0);
/* Log changes to the page directory. */
for (; a < e; a++)
{
if (old->page.frame[a] == block->page.frame[a])
continue;
while (--e, old->page.frame[e] == block->page.frame[e]);
e++;
ut_ad(a < e);
/* Write log for the changed page directory slots. */
mtr->memcpy(*block, a, e - a);
break;
}
}
buf_block_free(old);
MONITOR_INC(MONITOR_INDEX_REORG_ATTEMPTS);
MONITOR_INC(MONITOR_INDEX_REORG_SUCCESSFUL);
return DB_SUCCESS;
}
/*************************************************************//**
Reorganizes an index page.
IMPORTANT: On success, the caller will have to update IBUF_BITMAP_FREE
if this is a compressed leaf page in a secondary index. This has to
be done either within the same mini-transaction, or by invoking
ibuf_reset_free_bits() before mtr_commit(). On uncompressed pages,
IBUF_BITMAP_FREE is unaffected by reorganization.
@return error code
@retval DB_FAIL if reorganizing a ROW_FORMAT=COMPRESSED page failed */
dberr_t
btr_page_reorganize_block(
ulint z_level,/*!< in: compression level to be used
if dealing with compressed page */
buf_block_t* block, /*!< in/out: B-tree page */
dict_index_t* index, /*!< in: the index tree of the page */
mtr_t* mtr) /*!< in/out: mini-transaction */
{
if (buf_block_get_page_zip(block))
return page_zip_reorganize(block, index, z_level, mtr, true);
page_cur_t cur;
page_cur_set_before_first(block, &cur);
return btr_page_reorganize_low(&cur, index, mtr);
}
/*************************************************************//**
Reorganizes an index page.
IMPORTANT: On success, the caller will have to update IBUF_BITMAP_FREE
if this is a compressed leaf page in a secondary index. This has to
be done either within the same mini-transaction, or by invoking
ibuf_reset_free_bits() before mtr_commit(). On uncompressed pages,
IBUF_BITMAP_FREE is unaffected by reorganization.
@return error code
@retval DB_FAIL if reorganizing a ROW_FORMAT=COMPRESSED page failed */
dberr_t
btr_page_reorganize(
/*================*/
page_cur_t* cursor, /*!< in/out: page cursor */
dict_index_t* index, /*!< in: the index tree of the page */
mtr_t* mtr) /*!< in/out: mini-transaction */
{
if (!buf_block_get_page_zip(cursor->block))
return btr_page_reorganize_low(cursor, index, mtr);
ulint pos= page_rec_get_n_recs_before(cursor->rec);
if (UNIV_UNLIKELY(pos == ULINT_UNDEFINED))
return DB_CORRUPTION;
dberr_t err= page_zip_reorganize(cursor->block, index, page_zip_level, mtr,
true);
if (err == DB_FAIL);
else if (!pos)
ut_ad(cursor->rec == page_get_infimum_rec(cursor->block->page.frame));
else if (!(cursor->rec= page_rec_get_nth(cursor->block->page.frame, pos)))
err= DB_CORRUPTION;
return err;
}
/** Empty an index page (possibly the root page). @see btr_page_create().
@param[in,out] block page to be emptied
@param[in,out] page_zip compressed page frame, or NULL
@param[in] index index of the page
@param[in] level B-tree level of the page (0=leaf)
@param[in,out] mtr mini-transaction */
void
btr_page_empty(
buf_block_t* block,
page_zip_des_t* page_zip,
dict_index_t* index,
ulint level,
mtr_t* mtr)
{
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
ut_ad(page_zip == buf_block_get_page_zip(block));
ut_ad(!index->is_dummy);
ut_ad(index->table->space->id == block->page.id().space());
#ifdef UNIV_ZIP_DEBUG
ut_a(!page_zip
|| page_zip_validate(page_zip, block->page.frame, index));
#endif /* UNIV_ZIP_DEBUG */
btr_search_drop_page_hash_index(block);
/* Recreate the page: note that global data on page (possible
segment headers, next page-field, etc.) is preserved intact */
/* Preserve PAGE_ROOT_AUTO_INC when creating a clustered index
root page. */
const ib_uint64_t autoinc
= dict_index_is_clust(index)
&& index->page == block->page.id().page_no()
? page_get_autoinc(block->page.frame)
: 0;
if (page_zip) {
page_create_zip(block, index, level, autoinc, mtr);
} else {
page_create(block, mtr, index->table->not_redundant());
if (index->is_spatial()) {
static_assert(((FIL_PAGE_INDEX & 0xff00)
| byte(FIL_PAGE_RTREE))
== FIL_PAGE_RTREE, "compatibility");
mtr->write<1>(*block, FIL_PAGE_TYPE + 1
+ block->page.frame,
byte(FIL_PAGE_RTREE));
if (mach_read_from_8(block->page.frame
+ FIL_RTREE_SPLIT_SEQ_NUM)) {
mtr->memset(block, FIL_RTREE_SPLIT_SEQ_NUM,
8, 0);
}
}
mtr->write<2,mtr_t::MAYBE_NOP>(*block, PAGE_HEADER + PAGE_LEVEL
+ block->page.frame, level);
if (autoinc) {
mtr->write<8>(*block, PAGE_HEADER + PAGE_MAX_TRX_ID
+ block->page.frame, autoinc);
}
}
}
/** Write instant ALTER TABLE metadata to a root page.
@param[in,out] root clustered index root page
@param[in] index clustered index with instant ALTER TABLE
@param[in,out] mtr mini-transaction */
void btr_set_instant(buf_block_t* root, const dict_index_t& index, mtr_t* mtr)
{
ut_ad(index.n_core_fields > 0);
ut_ad(index.n_core_fields < REC_MAX_N_FIELDS);
ut_ad(index.is_instant());
ut_ad(fil_page_get_type(root->page.frame) == FIL_PAGE_TYPE_INSTANT
|| fil_page_get_type(root->page.frame) == FIL_PAGE_INDEX);
ut_ad(!page_has_siblings(root->page.frame));
ut_ad(root->page.id().page_no() == index.page);
rec_t* infimum = page_get_infimum_rec(root->page.frame);
rec_t* supremum = page_get_supremum_rec(root->page.frame);
byte* page_type = root->page.frame + FIL_PAGE_TYPE;
uint16_t i = page_header_get_field(root->page.frame, PAGE_INSTANT);
switch (mach_read_from_2(page_type)) {
case FIL_PAGE_TYPE_INSTANT:
ut_ad(page_get_instant(root->page.frame)
== index.n_core_fields);
if (memcmp(infimum, "infimum", 8)
|| memcmp(supremum, "supremum", 8)) {
ut_ad(index.table->instant);
ut_ad(!memcmp(infimum, field_ref_zero, 8));
ut_ad(!memcmp(supremum, field_ref_zero, 7));
/* The n_core_null_bytes only matters for
ROW_FORMAT=COMPACT and ROW_FORMAT=DYNAMIC tables. */
ut_ad(supremum[7] == index.n_core_null_bytes
|| !index.table->not_redundant());
return;
}
break;
default:
ut_ad("wrong page type" == 0);
/* fall through */
case FIL_PAGE_INDEX:
ut_ad(!page_is_comp(root->page.frame)
|| !page_get_instant(root->page.frame));
ut_ad(!memcmp(infimum, "infimum", 8));
ut_ad(!memcmp(supremum, "supremum", 8));
mtr->write<2>(*root, page_type, FIL_PAGE_TYPE_INSTANT);
ut_ad(i <= PAGE_NO_DIRECTION);
i |= static_cast<uint16_t>(index.n_core_fields << 3);
mtr->write<2>(*root, PAGE_HEADER + PAGE_INSTANT
+ root->page.frame, i);
break;
}
if (index.table->instant) {
mtr->memset(root, infimum - root->page.frame, 8, 0);
mtr->memset(root, supremum - root->page.frame, 7, 0);
mtr->write<1,mtr_t::MAYBE_NOP>(*root, &supremum[7],
index.n_core_null_bytes);
}
}
/** Reset the table to the canonical format on ROLLBACK of instant ALTER TABLE.
@param[in] index clustered index with instant ALTER TABLE
@param[in] all whether to reset FIL_PAGE_TYPE as well
@param[in,out] mtr mini-transaction
@return error code */
ATTRIBUTE_COLD
dberr_t btr_reset_instant(const dict_index_t &index, bool all, mtr_t *mtr)
{
ut_ad(!index.table->is_temporary());
ut_ad(index.is_primary());
dberr_t err;
if (buf_block_t *root= btr_root_block_get(&index, RW_SX_LATCH, mtr, &err))
{
byte *page_type= root->page.frame + FIL_PAGE_TYPE;
if (all)
{
ut_ad(mach_read_from_2(page_type) == FIL_PAGE_TYPE_INSTANT ||
mach_read_from_2(page_type) == FIL_PAGE_INDEX);
mtr->write<2,mtr_t::MAYBE_NOP>(*root, page_type, FIL_PAGE_INDEX);
byte *instant= PAGE_INSTANT + PAGE_HEADER + root->page.frame;
mtr->write<2,mtr_t::MAYBE_NOP>(*root, instant,
page_ptr_get_direction(instant + 1));
}
else
ut_ad(mach_read_from_2(page_type) == FIL_PAGE_TYPE_INSTANT);
static const byte supremuminfimum[8 + 8] = "supremuminfimum";
uint16_t infimum, supremum;
if (page_is_comp(root->page.frame))
{
infimum= PAGE_NEW_INFIMUM;
supremum= PAGE_NEW_SUPREMUM;
}
else
{
infimum= PAGE_OLD_INFIMUM;
supremum= PAGE_OLD_SUPREMUM;
}
ut_ad(!memcmp(&root->page.frame[infimum], supremuminfimum + 8, 8) ==
!memcmp(&root->page.frame[supremum], supremuminfimum, 8));
mtr->memcpy<mtr_t::MAYBE_NOP>(*root, &root->page.frame[infimum],
supremuminfimum + 8, 8);
mtr->memcpy<mtr_t::MAYBE_NOP>(*root, &root->page.frame[supremum],
supremuminfimum, 8);
}
return err;
}
/*************************************************************//**
Makes tree one level higher by splitting the root, and inserts
the tuple. It is assumed that mtr contains an x-latch on the tree.
NOTE that the operation of this function must always succeed,
we cannot reverse it: therefore enough free disk space must be
guaranteed to be available before this function is called.
@return inserted record */
rec_t*
btr_root_raise_and_insert(
/*======================*/
ulint flags, /*!< in: undo logging and locking flags */
btr_cur_t* cursor, /*!< in: cursor at which to insert: must be
on the root page; when the function returns,
the cursor is positioned on the predecessor
of the inserted record */
rec_offs** offsets,/*!< out: offsets on inserted record */
mem_heap_t** heap, /*!< in/out: pointer to memory heap, or NULL */
const dtuple_t* tuple, /*!< in: tuple to insert */
ulint n_ext, /*!< in: number of externally stored columns */
mtr_t* mtr, /*!< in: mtr */
dberr_t* err) /*!< out: error code */
{
dict_index_t* index;
rec_t* rec;
dtuple_t* node_ptr;
ulint level;
rec_t* node_ptr_rec;
page_cur_t* page_cursor;
page_zip_des_t* root_page_zip;
page_zip_des_t* new_page_zip;
buf_block_t* root;
buf_block_t* new_block;
root = btr_cur_get_block(cursor);
root_page_zip = buf_block_get_page_zip(root);
ut_ad(!page_is_empty(root->page.frame));
index = btr_cur_get_index(cursor);
ut_ad(index->n_core_null_bytes <= UT_BITS_IN_BYTES(index->n_nullable));
#ifdef UNIV_ZIP_DEBUG
ut_a(!root_page_zip
|| page_zip_validate(root_page_zip, root->page.frame, index));
#endif /* UNIV_ZIP_DEBUG */
const page_id_t root_id{root->page.id()};
ut_ad(mtr->memo_contains_flagged(&index->lock, MTR_MEMO_X_LOCK
| MTR_MEMO_SX_LOCK));
ut_ad(mtr->memo_contains_flagged(root, MTR_MEMO_PAGE_X_FIX));
if (index->page != root_id.page_no()) {
ut_ad("corrupted root page number" == 0);
return nullptr;
}
if (index->is_ibuf()) {
} else if (!btr_root_fseg_validate(FIL_PAGE_DATA + PAGE_BTR_SEG_LEAF,
*root, *index->table->space)
|| !btr_root_fseg_validate(FIL_PAGE_DATA + PAGE_BTR_SEG_TOP,
*root, *index->table->space)) {
return nullptr;
}
/* Allocate a new page to the tree. Root splitting is done by first
moving the root records to the new page, emptying the root, putting
a node pointer to the new page, and then splitting the new page. */
level = btr_page_get_level(root->page.frame);
new_block = btr_page_alloc(index, 0, FSP_NO_DIR, level, mtr, mtr, err);
if (!new_block) {
return nullptr;
}
new_page_zip = buf_block_get_page_zip(new_block);
ut_a(!new_page_zip == !root_page_zip);
ut_a(!new_page_zip
|| page_zip_get_size(new_page_zip)
== page_zip_get_size(root_page_zip));
btr_page_create(new_block, new_page_zip, index, level, mtr);
if (page_has_siblings(new_block->page.frame)) {
compile_time_assert(FIL_PAGE_NEXT == FIL_PAGE_PREV + 4);
compile_time_assert(FIL_NULL == 0xffffffff);
static_assert(FIL_PAGE_PREV % 8 == 0, "alignment");
memset_aligned<8>(new_block->page.frame + FIL_PAGE_PREV,
0xff, 8);
mtr->memset(new_block, FIL_PAGE_PREV, 8, 0xff);
if (UNIV_LIKELY_NULL(new_page_zip)) {
memset_aligned<8>(new_page_zip->data + FIL_PAGE_PREV,
0xff, 8);
}
}
/* Copy the records from root to the new page one by one. */
dberr_t e;
if (!err) {
err = &e;
}
if (0
#ifdef UNIV_ZIP_COPY
|| new_page_zip
#endif /* UNIV_ZIP_COPY */
|| !page_copy_rec_list_end(new_block, root,
page_get_infimum_rec(root->page.frame),
index, mtr, err)) {
switch (*err) {
case DB_SUCCESS:
break;
case DB_FAIL:
*err = DB_SUCCESS;
break;
default:
return nullptr;
}
ut_a(new_page_zip);
/* Copy the page byte for byte. */
page_zip_copy_recs(new_block, root_page_zip,
root->page.frame, index, mtr);
/* Update the lock table and possible hash index. */
if (index->has_locking()) {
lock_move_rec_list_end(
new_block, root,
page_get_infimum_rec(root->page.frame));
}
/* Move any existing predicate locks */
if (dict_index_is_spatial(index)) {
lock_prdt_rec_move(new_block, root_id);
} else {
btr_search_move_or_delete_hash_entries(
new_block, root);
}
}
constexpr uint16_t max_trx_id = PAGE_HEADER + PAGE_MAX_TRX_ID;
if (dict_index_is_sec_or_ibuf(index)) {
/* In secondary indexes and the change buffer,
PAGE_MAX_TRX_ID can be reset on the root page, because
the field only matters on leaf pages, and the root no
longer is a leaf page. (Older versions of InnoDB did
set PAGE_MAX_TRX_ID on all secondary index pages.) */
byte* p = my_assume_aligned<8>(
PAGE_HEADER + PAGE_MAX_TRX_ID + root->page.frame);
if (mach_read_from_8(p)) {
mtr->memset(root, max_trx_id, 8, 0);
if (UNIV_LIKELY_NULL(root->page.zip.data)) {
memset_aligned<8>(max_trx_id
+ root->page.zip.data, 0, 8);
}
}
} else {
/* PAGE_ROOT_AUTO_INC is only present in the clustered index
root page; on other clustered index pages, we want to reserve
the field PAGE_MAX_TRX_ID for future use. */
byte* p = my_assume_aligned<8>(
PAGE_HEADER + PAGE_MAX_TRX_ID + new_block->page.frame);
if (mach_read_from_8(p)) {
mtr->memset(new_block, max_trx_id, 8, 0);
if (UNIV_LIKELY_NULL(new_block->page.zip.data)) {
memset_aligned<8>(max_trx_id
+ new_block->page.zip.data,
0, 8);
}
}
}
/* If this is a pessimistic insert which is actually done to
perform a pessimistic update then we have stored the lock
information of the record to be inserted on the infimum of the
root page: we cannot discard the lock structs on the root page */
if (index->has_locking()) {
lock_update_root_raise(*new_block, root_id);
}
/* Create a memory heap where the node pointer is stored */
if (!*heap) {
*heap = mem_heap_create(1000);
}
const uint32_t new_page_no = new_block->page.id().page_no();
rec = page_rec_get_next(page_get_infimum_rec(new_block->page.frame));
ut_ad(rec); /* We just created the page. */
/* Build the node pointer (= node key and page address) for the
child */
if (dict_index_is_spatial(index)) {
rtr_mbr_t new_mbr;
rtr_page_cal_mbr(index, new_block, &new_mbr, *heap);
node_ptr = rtr_index_build_node_ptr(
index, &new_mbr, rec, new_page_no, *heap);
} else {
node_ptr = dict_index_build_node_ptr(
index, rec, new_page_no, *heap, level);
}
/* The node pointer must be marked as the predefined minimum record,
as there is no lower alphabetical limit to records in the leftmost
node of a level: */
dtuple_set_info_bits(node_ptr,
dtuple_get_info_bits(node_ptr)
| REC_INFO_MIN_REC_FLAG);
/* Rebuild the root page to get free space */
btr_page_empty(root, root_page_zip, index, level + 1, mtr);
/* btr_page_empty() is supposed to zero-initialize the field. */
ut_ad(!page_get_instant(root->page.frame));
if (index->is_instant()) {
ut_ad(!root_page_zip);
btr_set_instant(root, *index, mtr);
}
ut_ad(!page_has_siblings(root->page.frame));
page_cursor = btr_cur_get_page_cur(cursor);
/* Insert node pointer to the root */
page_cur_set_before_first(root, page_cursor);
node_ptr_rec = page_cur_tuple_insert(page_cursor, node_ptr,
index, offsets, heap, 0, mtr);
/* The root page should only contain the node pointer
to new_block at this point. Thus, the data should fit. */
ut_a(node_ptr_rec);
/* We play safe and reset the free bits for the new page */
if (!dict_index_is_clust(index)
&& !index->table->is_temporary()) {
ibuf_reset_free_bits(new_block);
}
if (tuple) {
ut_ad(dtuple_check_typed(tuple));
/* Reposition the cursor to the child node */
ulint low_match = 0, up_match = 0;
if (page_cur_search_with_match(new_block, index, tuple,
PAGE_CUR_LE,
&up_match, &low_match,
page_cursor, nullptr)) {
if (err) {
*err = DB_CORRUPTION;
}
return nullptr;
}
} else {
/* Set cursor to first record on child node */
page_cur_set_before_first(new_block, page_cursor);
}
/* Split the child and insert tuple */
return btr_page_split_and_insert(flags, cursor, offsets, heap,
tuple, n_ext, mtr, err);
}
/** Decide if the page should be split at the convergence point of inserts
converging to the left.
@param[in] cursor insert position
@return the first record to be moved to the right half page
@retval NULL if no split is recommended */
rec_t* btr_page_get_split_rec_to_left(const btr_cur_t* cursor)
{
rec_t* split_rec = btr_cur_get_rec(cursor);
const page_t* page = page_align(split_rec);
if (page_header_get_ptr(page, PAGE_LAST_INSERT)
!= page_rec_get_next(split_rec)) {
return NULL;
}
/* The metadata record must be present in the leftmost leaf page
of the clustered index, if and only if index->is_instant().
However, during innobase_instant_try(), index->is_instant()
would already hold when row_ins_clust_index_entry_low()
is being invoked to insert the the metadata record.
So, we can only assert that when the metadata record exists,
index->is_instant() must hold. */
ut_ad(!page_is_leaf(page) || page_has_prev(page)
|| cursor->index->is_instant()
|| !(rec_get_info_bits(page_rec_get_next_const(
page_get_infimum_rec(page)),
cursor->index->table->not_redundant())
& REC_INFO_MIN_REC_FLAG));
const rec_t* infimum = page_get_infimum_rec(page);
/* If the convergence is in the middle of a page, include also
the record immediately before the new insert to the upper
page. Otherwise, we could repeatedly move from page to page
lots of records smaller than the convergence point. */
if (split_rec == infimum
|| split_rec == page_rec_get_next_const(infimum)) {
split_rec = page_rec_get_next(split_rec);
}
return split_rec;
}
/** Decide if the page should be split at the convergence point of inserts
converging to the right.
@param[in] cursor insert position
@param[out] split_rec if split recommended, the first record
on the right half page, or
NULL if the to-be-inserted record
should be first
@return whether split is recommended */
bool
btr_page_get_split_rec_to_right(const btr_cur_t* cursor, rec_t** split_rec)
{
rec_t* insert_point = btr_cur_get_rec(cursor);
const page_t* page = page_align(insert_point);
/* We use eager heuristics: if the new insert would be right after
the previous insert on the same page, we assume that there is a
pattern of sequential inserts here. */
if (page_header_get_ptr(page, PAGE_LAST_INSERT) != insert_point) {
return false;
}
insert_point = page_rec_get_next(insert_point);
if (!insert_point || page_rec_is_supremum(insert_point)) {
insert_point = NULL;
} else {
insert_point = page_rec_get_next(insert_point);
if (page_rec_is_supremum(insert_point)) {
insert_point = NULL;
}
/* If there are >= 2 user records up from the insert
point, split all but 1 off. We want to keep one because
then sequential inserts can use the adaptive hash
index, as they can do the necessary checks of the right
search position just by looking at the records on this
page. */
}
*split_rec = insert_point;
return true;
}
/*************************************************************//**
Calculates a split record such that the tuple will certainly fit on
its half-page when the split is performed. We assume in this function
only that the cursor page has at least one user record.
@return split record, or NULL if tuple will be the first record on
the lower or upper half-page (determined by btr_page_tuple_smaller()) */
static
rec_t*
btr_page_get_split_rec(
/*===================*/
btr_cur_t* cursor, /*!< in: cursor at which insert should be made */
const dtuple_t* tuple, /*!< in: tuple to insert */
ulint n_ext) /*!< in: number of externally stored columns */
{
page_t* page;
page_zip_des_t* page_zip;
ulint insert_size;
ulint free_space;
ulint total_data;
ulint total_n_recs;
ulint total_space;
ulint incl_data;
rec_t* ins_rec;
rec_t* rec;
rec_t* next_rec;
ulint n;
mem_heap_t* heap;
rec_offs* offsets;
page = btr_cur_get_page(cursor);
insert_size = rec_get_converted_size(cursor->index, tuple, n_ext);
free_space = page_get_free_space_of_empty(page_is_comp(page));
page_zip = btr_cur_get_page_zip(cursor);
if (page_zip) {
/* Estimate the free space of an empty compressed page. */
ulint free_space_zip = page_zip_empty_size(
cursor->index->n_fields,
page_zip_get_size(page_zip));
if (free_space > (ulint) free_space_zip) {
free_space = (ulint) free_space_zip;
}
}
/* free_space is now the free space of a created new page */
total_data = page_get_data_size(page) + insert_size;
total_n_recs = ulint(page_get_n_recs(page)) + 1;
ut_ad(total_n_recs >= 2);
total_space = total_data + page_dir_calc_reserved_space(total_n_recs);
n = 0;
incl_data = 0;
ins_rec = btr_cur_get_rec(cursor);
rec = page_get_infimum_rec(page);
heap = NULL;
offsets = NULL;
/* We start to include records to the left half, and when the
space reserved by them exceeds half of total_space, then if
the included records fit on the left page, they will be put there
if something was left over also for the right page,
otherwise the last included record will be the first on the right
half page */
do {
/* Decide the next record to include */
if (rec == ins_rec) {
rec = NULL; /* NULL denotes that tuple is
now included */
} else if (rec == NULL) {
rec = page_rec_get_next(ins_rec);
} else {
rec = page_rec_get_next(rec);
}
if (rec == NULL) {
/* Include tuple */
incl_data += insert_size;
} else {
offsets = rec_get_offsets(rec, cursor->index, offsets,
page_is_leaf(page)
? cursor->index->n_core_fields
: 0,
ULINT_UNDEFINED, &heap);
incl_data += rec_offs_size(offsets);
}
n++;
} while (incl_data + page_dir_calc_reserved_space(n)
< total_space / 2);
if (incl_data + page_dir_calc_reserved_space(n) <= free_space) {
/* The next record will be the first on
the right half page if it is not the
supremum record of page */
if (rec == ins_rec) {
rec = NULL;
goto func_exit;
} else if (rec == NULL) {
next_rec = page_rec_get_next(ins_rec);
} else {
next_rec = page_rec_get_next(rec);
}
ut_ad(next_rec);
if (!page_rec_is_supremum(next_rec)) {
rec = next_rec;
}
}
func_exit:
if (heap) {
mem_heap_free(heap);
}
return(rec);
}
/*************************************************************//**
Returns TRUE if the insert fits on the appropriate half-page with the
chosen split_rec.
@return true if fits */
static MY_ATTRIBUTE((nonnull(1,3,4,6), warn_unused_result))
bool
btr_page_insert_fits(
/*=================*/
btr_cur_t* cursor, /*!< in: cursor at which insert
should be made */
const rec_t* split_rec,/*!< in: suggestion for first record
on upper half-page, or NULL if
tuple to be inserted should be first */
rec_offs** offsets,/*!< in: rec_get_offsets(
split_rec, cursor->index); out: garbage */
const dtuple_t* tuple, /*!< in: tuple to insert */
ulint n_ext, /*!< in: number of externally stored columns */
mem_heap_t** heap) /*!< in: temporary memory heap */
{
page_t* page;
ulint insert_size;
ulint free_space;
ulint total_data;
ulint total_n_recs;
const rec_t* rec;
const rec_t* end_rec;
page = btr_cur_get_page(cursor);
ut_ad(!split_rec
|| !page_is_comp(page) == !rec_offs_comp(*offsets));
ut_ad(!split_rec
|| rec_offs_validate(split_rec, cursor->index, *offsets));
insert_size = rec_get_converted_size(cursor->index, tuple, n_ext);
free_space = page_get_free_space_of_empty(page_is_comp(page));
/* free_space is now the free space of a created new page */
total_data = page_get_data_size(page) + insert_size;
total_n_recs = ulint(page_get_n_recs(page)) + 1;
/* We determine which records (from rec to end_rec, not including
end_rec) will end up on the other half page from tuple when it is
inserted. */
if (!(end_rec = split_rec)) {
end_rec = page_rec_get_next(btr_cur_get_rec(cursor));
} else if (cmp_dtuple_rec(tuple, split_rec, cursor->index,
*offsets) < 0) {
rec = split_rec;
end_rec = page_get_supremum_rec(page);
goto got_rec;
}
if (!(rec = page_rec_get_next(page_get_infimum_rec(page)))) {
return false;
}
got_rec:
if (total_data + page_dir_calc_reserved_space(total_n_recs)
<= free_space) {
/* Ok, there will be enough available space on the
half page where the tuple is inserted */
return(true);
}
while (rec != end_rec) {
/* In this loop we calculate the amount of reserved
space after rec is removed from page. */
*offsets = rec_get_offsets(rec, cursor->index, *offsets,
page_is_leaf(page)
? cursor->index->n_core_fields
: 0,
ULINT_UNDEFINED, heap);
total_data -= rec_offs_size(*offsets);
total_n_recs--;
if (total_data + page_dir_calc_reserved_space(total_n_recs)
<= free_space) {
/* Ok, there will be enough available space on the
half page where the tuple is inserted */
return(true);
}
if (!(rec = page_rec_get_next_const(rec))) {
break;
}
}
return(false);
}
/*******************************************************//**
Inserts a data tuple to a tree on a non-leaf level. It is assumed
that mtr holds an x-latch on the tree. */
dberr_t
btr_insert_on_non_leaf_level(
ulint flags, /*!< in: undo logging and locking flags */
dict_index_t* index, /*!< in: index */
ulint level, /*!< in: level, must be > 0 */
dtuple_t* tuple, /*!< in: the record to be inserted */
mtr_t* mtr) /*!< in: mtr */
{
big_rec_t* dummy_big_rec;
btr_cur_t cursor;
rec_t* rec;
mem_heap_t* heap = NULL;
rec_offs offsets_[REC_OFFS_NORMAL_SIZE];
rec_offs* offsets = offsets_;
rec_offs_init(offsets_);
rtr_info_t rtr_info;
ut_ad(level > 0);
auto mode = PAGE_CUR_LE;
if (index->is_spatial()) {
mode = PAGE_CUR_RTREE_INSERT;
/* For spatial index, initialize structures to track
its parents etc. */
rtr_init_rtr_info(&rtr_info, false, &cursor, index, false);
rtr_info_update_btr(&cursor, &rtr_info);
}
flags |= BTR_NO_LOCKING_FLAG | BTR_KEEP_SYS_FLAG
| BTR_NO_UNDO_LOG_FLAG;
dberr_t err = btr_cur_search_to_nth_level(index, level, tuple, mode,
BTR_CONT_MODIFY_TREE,
&cursor, 0, mtr);
ut_ad(cursor.flag == BTR_CUR_BINARY);
if (UNIV_LIKELY(err == DB_SUCCESS)) {
err = btr_cur_optimistic_insert(flags,
&cursor, &offsets, &heap,
tuple, &rec,
&dummy_big_rec, 0, NULL, mtr);
}
if (err == DB_FAIL) {
err = btr_cur_pessimistic_insert(flags,
&cursor, &offsets, &heap,
tuple, &rec,
&dummy_big_rec, 0, NULL, mtr);
}
if (UNIV_LIKELY_NULL(heap)) {
mem_heap_free(heap);
}
if (index->is_spatial()) {
ut_ad(cursor.rtr_info);
rtr_clean_rtr_info(&rtr_info, true);
}
return err;
}
static_assert(FIL_PAGE_OFFSET % 4 == 0, "alignment");
static_assert(FIL_PAGE_PREV % 4 == 0, "alignment");
static_assert(FIL_PAGE_NEXT % 4 == 0, "alignment");
MY_ATTRIBUTE((nonnull,warn_unused_result))
/**************************************************************//**
Attaches the halves of an index page on the appropriate level in an
index tree. */
static
dberr_t
btr_attach_half_pages(
/*==================*/
ulint flags, /*!< in: undo logging and
locking flags */
dict_index_t* index, /*!< in: the index tree */
buf_block_t* block, /*!< in/out: page to be split */
const rec_t* split_rec, /*!< in: first record on upper
half page */
buf_block_t* new_block, /*!< in/out: the new half page */
ulint direction, /*!< in: FSP_UP or FSP_DOWN */
mtr_t* mtr) /*!< in: mtr */
{
dtuple_t* node_ptr_upper;
mem_heap_t* heap;
buf_block_t* prev_block = NULL;
buf_block_t* next_block = NULL;
buf_block_t* lower_block;
buf_block_t* upper_block;
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
ut_ad(mtr->memo_contains_flagged(new_block, MTR_MEMO_PAGE_X_FIX));
/* Create a memory heap where the data tuple is stored */
heap = mem_heap_create(1024);
/* Based on split direction, decide upper and lower pages */
if (direction == FSP_DOWN) {
btr_cur_t cursor;
rec_offs* offsets;
lower_block = new_block;
upper_block = block;
/* Look up the index for the node pointer to page */
offsets = btr_page_get_father_block(NULL, heap, index,
block, mtr, &cursor);
/* Replace the address of the old child node (= page) with the
address of the new lower half */
btr_node_ptr_set_child_page_no(
btr_cur_get_block(&cursor),
btr_cur_get_rec(&cursor),
offsets, lower_block->page.id().page_no(), mtr);
mem_heap_empty(heap);
} else {
lower_block = block;
upper_block = new_block;
}
/* Get the level of the split pages */
const ulint level = btr_page_get_level(block->page.frame);
ut_ad(level == btr_page_get_level(new_block->page.frame));
/* Get the previous and next pages of page */
const uint32_t prev_page_no = btr_page_get_prev(block->page.frame);
const uint32_t next_page_no = btr_page_get_next(block->page.frame);
/* for consistency, both blocks should be locked, before change */
if (prev_page_no != FIL_NULL && direction == FSP_DOWN) {
prev_block = btr_block_get(*index, prev_page_no, RW_X_LATCH,
!level, mtr);
}
if (next_page_no != FIL_NULL && direction != FSP_DOWN) {
next_block = btr_block_get(*index, next_page_no, RW_X_LATCH,
!level, mtr);
}
/* Build the node pointer (= node key and page address) for the upper
half */
node_ptr_upper = dict_index_build_node_ptr(
index, split_rec, upper_block->page.id().page_no(),
heap, level);
/* Insert it next to the pointer to the lower half. Note that this
may generate recursion leading to a split on the higher level. */
dberr_t err = btr_insert_on_non_leaf_level(
flags, index, level + 1, node_ptr_upper, mtr);
/* Free the memory heap */
mem_heap_free(heap);
if (UNIV_UNLIKELY(err != DB_SUCCESS)) {
return err;
}
/* Update page links of the level */
if (prev_block) {
if (UNIV_UNLIKELY(memcmp_aligned<4>(prev_block->page.frame
+ FIL_PAGE_NEXT,
block->page.frame
+ FIL_PAGE_OFFSET,
4))) {
return DB_CORRUPTION;
}
btr_page_set_next(prev_block, lower_block->page.id().page_no(),
mtr);
}
if (next_block) {
if (UNIV_UNLIKELY(memcmp_aligned<4>(next_block->page.frame
+ FIL_PAGE_PREV,
block->page.frame
+ FIL_PAGE_OFFSET,
4))) {
return DB_CORRUPTION;
}
btr_page_set_prev(next_block, upper_block->page.id().page_no(),
mtr);
}
if (direction == FSP_DOWN) {
ut_ad(lower_block == new_block);
ut_ad(btr_page_get_next(upper_block->page.frame)
== next_page_no);
btr_page_set_prev(lower_block, prev_page_no, mtr);
} else {
ut_ad(upper_block == new_block);
ut_ad(btr_page_get_prev(lower_block->page.frame)
== prev_page_no);
btr_page_set_next(upper_block, next_page_no, mtr);
}
btr_page_set_prev(upper_block, lower_block->page.id().page_no(), mtr);
btr_page_set_next(lower_block, upper_block->page.id().page_no(), mtr);
return DB_SUCCESS;
}
/*************************************************************//**
Determine if a tuple is smaller than any record on the page.
@return TRUE if smaller */
static MY_ATTRIBUTE((nonnull, warn_unused_result))
bool
btr_page_tuple_smaller(
/*===================*/
btr_cur_t* cursor, /*!< in: b-tree cursor */
const dtuple_t* tuple, /*!< in: tuple to consider */
rec_offs** offsets,/*!< in/out: temporary storage */
ulint n_uniq, /*!< in: number of unique fields
in the index page records */
mem_heap_t** heap) /*!< in/out: heap for offsets */
{
buf_block_t* block;
const rec_t* first_rec;
page_cur_t pcur;
/* Read the first user record in the page. */
block = btr_cur_get_block(cursor);
page_cur_set_before_first(block, &pcur);
if (UNIV_UNLIKELY(!(first_rec = page_cur_move_to_next(&pcur)))) {
ut_ad("corrupted page" == 0);
return false;
}
*offsets = rec_get_offsets(first_rec, cursor->index, *offsets,
page_is_leaf(block->page.frame)
? cursor->index->n_core_fields : 0,
n_uniq, heap);
return cmp_dtuple_rec(tuple, first_rec, cursor->index, *offsets) < 0;
}
/** Insert the tuple into the right sibling page, if the cursor is at the end
of a page.
@param[in] flags undo logging and locking flags
@param[in,out] cursor cursor at which to insert; when the function succeeds,
the cursor is positioned before the insert point.
@param[out] offsets offsets on inserted record
@param[in,out] heap memory heap for allocating offsets
@param[in] tuple tuple to insert
@param[in] n_ext number of externally stored columns
@param[in,out] mtr mini-transaction
@return inserted record (first record on the right sibling page);
the cursor will be positioned on the page infimum
@retval NULL if the operation was not performed */
static
rec_t*
btr_insert_into_right_sibling(
ulint flags,
btr_cur_t* cursor,
rec_offs** offsets,
mem_heap_t* heap,
const dtuple_t* tuple,
ulint n_ext,
mtr_t* mtr)
{
buf_block_t* block = btr_cur_get_block(cursor);
page_t* page = buf_block_get_frame(block);
const uint32_t next_page_no = btr_page_get_next(page);
ut_ad(mtr->memo_contains_flagged(&cursor->index->lock,
MTR_MEMO_X_LOCK | MTR_MEMO_SX_LOCK));
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
ut_ad(heap);
ut_ad(dtuple_check_typed(tuple));
if (next_page_no == FIL_NULL || !page_rec_is_supremum(
page_rec_get_next(btr_cur_get_rec(cursor)))) {
return nullptr;
}
page_cur_t next_page_cursor;
buf_block_t* next_block;
page_t* next_page;
btr_cur_t next_father_cursor;
rec_t* rec = nullptr;
ulint max_size;
next_block = btr_block_get(*cursor->index, next_page_no, RW_X_LATCH,
page_is_leaf(page), mtr);
if (UNIV_UNLIKELY(!next_block)) {
return nullptr;
}
next_page = buf_block_get_frame(next_block);
const bool is_leaf = page_is_leaf(next_page);
if (!btr_page_get_father(cursor->index, next_block, mtr,
&next_father_cursor)) {
return nullptr;
}
ulint up_match = 0, low_match = 0;
if (page_cur_search_with_match(next_block, cursor->index, tuple,
PAGE_CUR_LE, &up_match, &low_match,
&next_page_cursor, nullptr)) {
return nullptr;
}
max_size = page_get_max_insert_size_after_reorganize(next_page, 1);
/* Extends gap lock for the next page */
if (is_leaf && cursor->index->has_locking()) {
lock_update_node_pointer(block, next_block);
}
rec = page_cur_tuple_insert(
&next_page_cursor, tuple, cursor->index, offsets, &heap,
n_ext, mtr);
if (!rec) {
if (is_leaf
&& next_block->page.zip.ssize
&& !dict_index_is_clust(cursor->index)
&& !cursor->index->table->is_temporary()) {
/* Reset the IBUF_BITMAP_FREE bits, because
page_cur_tuple_insert() will have attempted page
reorganize before failing. */
ibuf_reset_free_bits(next_block);
}
return nullptr;
}
ibool compressed;
dberr_t err;
ulint level = btr_page_get_level(next_page);
/* adjust cursor position */
*btr_cur_get_page_cur(cursor) = next_page_cursor;
ut_ad(btr_cur_get_rec(cursor) == page_get_infimum_rec(next_page));
ut_ad(page_rec_get_next(page_get_infimum_rec(next_page)) == rec);
/* We have to change the parent node pointer */
compressed = btr_cur_pessimistic_delete(
&err, TRUE, &next_father_cursor,
BTR_CREATE_FLAG, false, mtr);
if (err != DB_SUCCESS) {
return nullptr;
}
if (!compressed) {
btr_cur_compress_if_useful(&next_father_cursor, false, mtr);
}
dtuple_t* node_ptr = dict_index_build_node_ptr(
cursor->index, rec, next_block->page.id().page_no(),
heap, level);
if (btr_insert_on_non_leaf_level(flags, cursor->index, level + 1,
node_ptr, mtr) != DB_SUCCESS) {
return nullptr;
}
ut_ad(rec_offs_validate(rec, cursor->index, *offsets));
if (is_leaf
&& !dict_index_is_clust(cursor->index)
&& !cursor->index->table->is_temporary()) {
/* Update the free bits of the B-tree page in the
insert buffer bitmap. */
if (next_block->page.zip.ssize) {
ibuf_update_free_bits_zip(next_block, mtr);
} else {
ibuf_update_free_bits_if_full(
next_block, max_size,
rec_offs_size(*offsets) + PAGE_DIR_SLOT_SIZE);
}
}
return(rec);
}
/*************************************************************//**
Moves record list end to another page. Moved records include
split_rec.
IMPORTANT: The caller will have to update IBUF_BITMAP_FREE
if new_block is a compressed leaf page in a secondary index.
This has to be done either within the same mini-transaction,
or by invoking ibuf_reset_free_bits() before mtr_commit().
@return error code */
static
dberr_t
page_move_rec_list_end(
/*===================*/
buf_block_t* new_block, /*!< in/out: index page where to move */
buf_block_t* block, /*!< in: index page from where to move */
rec_t* split_rec, /*!< in: first record to move */
dict_index_t* index, /*!< in: record descriptor */
mtr_t* mtr) /*!< in: mtr */
{
page_t* new_page = buf_block_get_frame(new_block);
ulint old_data_size;
ulint new_data_size;
ulint old_n_recs;
ulint new_n_recs;
ut_ad(!dict_index_is_spatial(index));
old_data_size = page_get_data_size(new_page);
old_n_recs = page_get_n_recs(new_page);
#ifdef UNIV_ZIP_DEBUG
{
page_zip_des_t* new_page_zip
= buf_block_get_page_zip(new_block);
page_zip_des_t* page_zip
= buf_block_get_page_zip(block);
ut_a(!new_page_zip == !page_zip);
ut_a(!new_page_zip
|| page_zip_validate(new_page_zip, new_page, index));
ut_a(!page_zip
|| page_zip_validate(page_zip, page_align(split_rec),
index));
}
#endif /* UNIV_ZIP_DEBUG */
dberr_t err;
if (!page_copy_rec_list_end(new_block, block,
split_rec, index, mtr, &err)) {
return err;
}
new_data_size = page_get_data_size(new_page);
new_n_recs = page_get_n_recs(new_page);
ut_ad(new_data_size >= old_data_size);
return page_delete_rec_list_end(split_rec, block, index,
new_n_recs - old_n_recs,
new_data_size - old_data_size, mtr);
}
/*************************************************************//**
Moves record list start to another page. Moved records do not include
split_rec.
IMPORTANT: The caller will have to update IBUF_BITMAP_FREE
if new_block is a compressed leaf page in a secondary index.
This has to be done either within the same mini-transaction,
or by invoking ibuf_reset_free_bits() before mtr_commit().
@return error code */
static
dberr_t
page_move_rec_list_start(
/*=====================*/
buf_block_t* new_block, /*!< in/out: index page where to move */
buf_block_t* block, /*!< in/out: page containing split_rec */
rec_t* split_rec, /*!< in: first record not to move */
dict_index_t* index, /*!< in: record descriptor */
mtr_t* mtr) /*!< in: mtr */
{
dberr_t err;
if (page_copy_rec_list_start(new_block, block, split_rec, index, mtr, &err))
page_delete_rec_list_start(split_rec, block, index, mtr);
return err;
}
/*************************************************************//**
Splits an index page to halves and inserts the tuple. It is assumed
that mtr holds an x-latch to the index tree. NOTE: the tree x-latch is
released within this function! NOTE that the operation of this
function must always succeed, we cannot reverse it: therefore enough
free disk space (2 pages) must be guaranteed to be available before
this function is called.
@return inserted record or NULL if run out of space */
rec_t*
btr_page_split_and_insert(
/*======================*/
ulint flags, /*!< in: undo logging and locking flags */
btr_cur_t* cursor, /*!< in: cursor at which to insert; when the
function returns, the cursor is positioned
on the predecessor of the inserted record */
rec_offs** offsets,/*!< out: offsets on inserted record */
mem_heap_t** heap, /*!< in/out: pointer to memory heap, or NULL */
const dtuple_t* tuple, /*!< in: tuple to insert */
ulint n_ext, /*!< in: number of externally stored columns */
mtr_t* mtr, /*!< in: mtr */
dberr_t* err) /*!< out: error code */
{
buf_block_t* block;
page_t* page;
page_zip_des_t* page_zip;
buf_block_t* new_block;
page_t* new_page;
page_zip_des_t* new_page_zip;
rec_t* split_rec;
buf_block_t* left_block;
buf_block_t* right_block;
page_cur_t* page_cursor;
rec_t* first_rec;
byte* buf = 0; /* remove warning */
rec_t* move_limit;
ulint n_iterations = 0;
ulint n_uniq;
ut_ad(*err == DB_SUCCESS);
ut_ad(dtuple_check_typed(tuple));
if (cursor->index->is_spatial()) {
/* Split rtree page and update parent */
return rtr_page_split_and_insert(flags, cursor, offsets, heap,
tuple, n_ext, mtr, err);
}
if (!*heap) {
*heap = mem_heap_create(1024);
}
n_uniq = dict_index_get_n_unique_in_tree(cursor->index);
func_start:
mem_heap_empty(*heap);
*offsets = NULL;
ut_ad(mtr->memo_contains_flagged(&cursor->index->lock, MTR_MEMO_X_LOCK
| MTR_MEMO_SX_LOCK));
ut_ad(!dict_index_is_online_ddl(cursor->index)
|| (flags & BTR_CREATE_FLAG)
|| dict_index_is_clust(cursor->index));
ut_ad(cursor->index->lock.have_u_or_x());
block = btr_cur_get_block(cursor);
page = buf_block_get_frame(block);
page_zip = buf_block_get_page_zip(block);
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
ut_ad(!page_is_empty(page));
/* try to insert to the next page if possible before split */
if (rec_t* rec = btr_insert_into_right_sibling(
flags, cursor, offsets, *heap, tuple, n_ext, mtr)) {
return(rec);
}
/* 1. Decide the split record; split_rec == NULL means that the
tuple to be inserted should be the first record on the upper
half-page */
bool insert_left = false;
uint32_t hint_page_no = block->page.id().page_no() + 1;
byte direction = FSP_UP;
if (n_iterations > 0) {
split_rec = btr_page_get_split_rec(cursor, tuple, n_ext);
if (split_rec == NULL) {
insert_left = btr_page_tuple_smaller(
cursor, tuple, offsets, n_uniq, heap);
}
} else if (btr_page_get_split_rec_to_right(cursor, &split_rec)) {
} else if ((split_rec = btr_page_get_split_rec_to_left(cursor))) {
direction = FSP_DOWN;
hint_page_no -= 2;
} else {
/* If there is only one record in the index page, we
can't split the node in the middle by default. We need
to determine whether the new record will be inserted
to the left or right. */
if (page_get_n_recs(page) > 1) {
split_rec = page_get_middle_rec(page);
} else if (btr_page_tuple_smaller(cursor, tuple,
offsets, n_uniq, heap)) {
split_rec = page_rec_get_next(
page_get_infimum_rec(page));
} else {
split_rec = NULL;
goto got_split_rec;
}
if (UNIV_UNLIKELY(!split_rec)) {
*err = DB_CORRUPTION;
return nullptr;
}
}
got_split_rec:
/* 2. Allocate a new page to the index */
const uint16_t page_level = btr_page_get_level(page);
new_block = btr_page_alloc(cursor->index, hint_page_no, direction,
page_level, mtr, mtr, err);
if (!new_block) {
return nullptr;
}
new_page = buf_block_get_frame(new_block);
new_page_zip = buf_block_get_page_zip(new_block);
if (page_level && UNIV_LIKELY_NULL(new_page_zip)) {
/* ROW_FORMAT=COMPRESSED non-leaf pages are not expected
to contain FIL_NULL in FIL_PAGE_PREV at this stage. */
memset_aligned<4>(new_page + FIL_PAGE_PREV, 0, 4);
}
btr_page_create(new_block, new_page_zip, cursor->index,
page_level, mtr);
/* Only record the leaf level page splits. */
if (!page_level) {
cursor->index->stat_defrag_n_page_split ++;
cursor->index->stat_defrag_modified_counter ++;
btr_defragment_save_defrag_stats_if_needed(cursor->index);
}
/* 3. Calculate the first record on the upper half-page, and the
first record (move_limit) on original page which ends up on the
upper half */
if (split_rec) {
first_rec = move_limit = split_rec;
*offsets = rec_get_offsets(split_rec, cursor->index, *offsets,
page_is_leaf(page)
? cursor->index->n_core_fields : 0,
n_uniq, heap);
insert_left = cmp_dtuple_rec(tuple, split_rec, cursor->index,
*offsets) < 0;
if (!insert_left && new_page_zip && n_iterations > 0) {
/* If a compressed page has already been split,
avoid further splits by inserting the record
to an empty page. */
split_rec = NULL;
goto insert_empty;
}
} else if (insert_left) {
if (UNIV_UNLIKELY(!n_iterations)) {
corrupted:
*err = DB_CORRUPTION;
return nullptr;
}
first_rec = page_rec_get_next(page_get_infimum_rec(page));
insert_move_limit:
move_limit = page_rec_get_next(btr_cur_get_rec(cursor));
if (UNIV_UNLIKELY(!first_rec || !move_limit)) {
goto corrupted;
}
} else {
insert_empty:
ut_ad(!split_rec);
ut_ad(!insert_left);
buf = UT_NEW_ARRAY_NOKEY(
byte,
rec_get_converted_size(cursor->index, tuple, n_ext));
first_rec = rec_convert_dtuple_to_rec(buf, cursor->index,
tuple, n_ext);
goto insert_move_limit;
}
/* 4. Do first the modifications in the tree structure */
/* FIXME: write FIL_PAGE_PREV,FIL_PAGE_NEXT in new_block earlier! */
*err = btr_attach_half_pages(flags, cursor->index, block,
first_rec, new_block, direction, mtr);
if (UNIV_UNLIKELY(*err != DB_SUCCESS)) {
return nullptr;
}
/* If the split is made on the leaf level and the insert will fit
on the appropriate half-page, we may release the tree x-latch.
We can then move the records after releasing the tree latch,
thus reducing the tree latch contention. */
const bool insert_will_fit = !new_page_zip
&& btr_page_insert_fits(cursor, split_rec, offsets, tuple,
n_ext, heap);
if (!split_rec && !insert_left) {
UT_DELETE_ARRAY(buf);
buf = NULL;
}
if (!srv_read_only_mode
&& insert_will_fit
&& page_is_leaf(page)
&& !dict_index_is_online_ddl(cursor->index)) {
mtr->memo_release(&cursor->index->lock,
MTR_MEMO_X_LOCK | MTR_MEMO_SX_LOCK);
/* NOTE: We cannot release root block latch here, because it
has segment header and already modified in most of cases.*/
}
/* 5. Move then the records to the new page */
if (direction == FSP_DOWN) {
/* fputs("Split left\n", stderr); */
if (0
#ifdef UNIV_ZIP_COPY
|| page_zip
#endif /* UNIV_ZIP_COPY */
|| (*err = page_move_rec_list_start(new_block, block,
move_limit,
cursor->index, mtr))) {
if (*err != DB_FAIL) {
return nullptr;
}
/* For some reason, compressing new_block failed,
even though it should contain fewer records than
the original page. Copy the page byte for byte
and then delete the records from both pages
as appropriate. Deleting will always succeed. */
ut_a(new_page_zip);
page_zip_copy_recs(new_block,
page_zip, page, cursor->index, mtr);
*err = page_delete_rec_list_end(move_limit
- page + new_page,
new_block,
cursor->index,
ULINT_UNDEFINED,
ULINT_UNDEFINED, mtr);
if (*err != DB_SUCCESS) {
return nullptr;
}
/* Update the lock table and possible hash index. */
if (cursor->index->has_locking()) {
lock_move_rec_list_start(
new_block, block, move_limit,
new_page + PAGE_NEW_INFIMUM);
}
btr_search_move_or_delete_hash_entries(
new_block, block);
/* Delete the records from the source page. */
page_delete_rec_list_start(move_limit, block,
cursor->index, mtr);
}
left_block = new_block;
right_block = block;
if (cursor->index->has_locking()) {
lock_update_split_left(right_block, left_block);
}
} else {
/* fputs("Split right\n", stderr); */
if (0
#ifdef UNIV_ZIP_COPY
|| page_zip
#endif /* UNIV_ZIP_COPY */
|| (*err = page_move_rec_list_end(new_block, block,
move_limit,
cursor->index, mtr))) {
if (*err != DB_FAIL) {
return nullptr;
}
/* For some reason, compressing new_page failed,
even though it should contain fewer records than
the original page. Copy the page byte for byte
and then delete the records from both pages
as appropriate. Deleting will always succeed. */
ut_a(new_page_zip);
page_zip_copy_recs(new_block,
page_zip, page, cursor->index, mtr);
page_delete_rec_list_start(move_limit - page
+ new_page, new_block,
cursor->index, mtr);
/* Update the lock table and possible hash index. */
if (cursor->index->has_locking()) {
lock_move_rec_list_end(new_block, block,
move_limit);
}
btr_search_move_or_delete_hash_entries(
new_block, block);
/* Delete the records from the source page. */
*err = page_delete_rec_list_end(move_limit, block,
cursor->index,
ULINT_UNDEFINED,
ULINT_UNDEFINED, mtr);
if (*err != DB_SUCCESS) {
return nullptr;
}
}
left_block = block;
right_block = new_block;
if (cursor->index->has_locking()) {
lock_update_split_right(right_block, left_block);
}
}
#ifdef UNIV_ZIP_DEBUG
if (page_zip) {
ut_a(page_zip_validate(page_zip, page, cursor->index));
ut_a(page_zip_validate(new_page_zip, new_page, cursor->index));
}
#endif /* UNIV_ZIP_DEBUG */
/* At this point, split_rec, move_limit and first_rec may point
to garbage on the old page. */
/* 6. The split and the tree modification is now completed. Decide the
page where the tuple should be inserted */
rec_t* rec;
buf_block_t* const insert_block = insert_left
? left_block : right_block;
/* 7. Reposition the cursor for insert and try insertion */
page_cursor = btr_cur_get_page_cur(cursor);
ulint up_match = 0, low_match = 0;
if (page_cur_search_with_match(insert_block, cursor->index, tuple,
PAGE_CUR_LE, &up_match, &low_match,
page_cursor, nullptr)) {
*err = DB_CORRUPTION;
return nullptr;
}
rec = page_cur_tuple_insert(page_cursor, tuple, cursor->index,
offsets, heap, n_ext, mtr);
#ifdef UNIV_ZIP_DEBUG
{
page_t* insert_page
= buf_block_get_frame(insert_block);
page_zip_des_t* insert_page_zip
= buf_block_get_page_zip(insert_block);
ut_a(!insert_page_zip
|| page_zip_validate(insert_page_zip, insert_page,
cursor->index));
}
#endif /* UNIV_ZIP_DEBUG */
if (rec != NULL) {
goto func_exit;
}
/* 8. If insert did not fit, try page reorganization.
For compressed pages, page_cur_tuple_insert() will have
attempted this already. */
if (page_cur_get_page_zip(page_cursor)) {
goto insert_failed;
}
*err = btr_page_reorganize(page_cursor, cursor->index, mtr);
if (*err != DB_SUCCESS) {
return nullptr;
}
rec = page_cur_tuple_insert(page_cursor, tuple, cursor->index,
offsets, heap, n_ext, mtr);
if (rec == NULL) {
/* The insert did not fit on the page: loop back to the
start of the function for a new split */
insert_failed:
/* We play safe and reset the free bits for new_page */
if (!dict_index_is_clust(cursor->index)
&& !cursor->index->table->is_temporary()) {
ibuf_reset_free_bits(new_block);
ibuf_reset_free_bits(block);
}
n_iterations++;
ut_ad(n_iterations < 2
|| buf_block_get_page_zip(insert_block));
ut_ad(!insert_will_fit);
goto func_start;
}
func_exit:
/* Insert fit on the page: update the free bits for the
left and right pages in the same mtr */
if (!dict_index_is_clust(cursor->index)
&& !cursor->index->table->is_temporary()
&& page_is_leaf(page)) {
ibuf_update_free_bits_for_two_pages_low(
left_block, right_block, mtr);
}
MONITOR_INC(MONITOR_INDEX_SPLIT);
ut_ad(page_validate(buf_block_get_frame(left_block), cursor->index));
ut_ad(page_validate(buf_block_get_frame(right_block), cursor->index));
ut_ad(!rec || rec_offs_validate(rec, cursor->index, *offsets));
return(rec);
}
/** Remove a page from the level list of pages.
@param[in] block page to remove
@param[in] index index tree
@param[in,out] mtr mini-transaction */
dberr_t btr_level_list_remove(const buf_block_t& block,
const dict_index_t& index, mtr_t* mtr)
{
ut_ad(mtr->memo_contains_flagged(&block, MTR_MEMO_PAGE_X_FIX));
ut_ad(block.zip_size() == index.table->space->zip_size());
ut_ad(index.table->space->id == block.page.id().space());
/* Get the previous and next page numbers of page */
const page_t* page = block.page.frame;
const uint32_t prev_page_no = btr_page_get_prev(page);
const uint32_t next_page_no = btr_page_get_next(page);
/* Update page links of the level */
dberr_t err;
if (prev_page_no != FIL_NULL) {
buf_block_t* prev_block = btr_block_get(
index, prev_page_no, RW_X_LATCH, page_is_leaf(page),
mtr, &err);
if (UNIV_UNLIKELY(!prev_block)) {
return err;
}
if (UNIV_UNLIKELY(memcmp_aligned<4>(prev_block->page.frame
+ FIL_PAGE_NEXT,
page + FIL_PAGE_OFFSET,
4))) {
return DB_CORRUPTION;
}
btr_page_set_next(prev_block, next_page_no, mtr);
}
if (next_page_no != FIL_NULL) {
buf_block_t* next_block = btr_block_get(
index, next_page_no, RW_X_LATCH, page_is_leaf(page),
mtr, &err);
if (UNIV_UNLIKELY(!next_block)) {
return err;
}
if (UNIV_UNLIKELY(memcmp_aligned<4>(next_block->page.frame
+ FIL_PAGE_PREV,
page + FIL_PAGE_OFFSET,
4))) {
return DB_CORRUPTION;
}
btr_page_set_prev(next_block, prev_page_no, mtr);
}
return DB_SUCCESS;
}
/*************************************************************//**
If page is the only on its level, this function moves its records to the
father page, thus reducing the tree height.
@return father block */
buf_block_t*
btr_lift_page_up(
dict_index_t* index, /*!< in: index tree */
buf_block_t* block, /*!< in: page which is the only on its level;
must not be empty: use
btr_discard_only_page_on_level if the last
record from the page should be removed */
mtr_t* mtr, /*!< in/out: mini-transaction */
dberr_t* err) /*!< out: error code */
{
buf_block_t* father_block;
ulint page_level;
page_zip_des_t* father_page_zip;
page_t* page = buf_block_get_frame(block);
ulint root_page_no;
buf_block_t* blocks[BTR_MAX_LEVELS];
ulint n_blocks; /*!< last used index in blocks[] */
ulint i;
bool lift_father_up;
buf_block_t* block_orig = block;
ut_ad(!page_has_siblings(page));
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
page_level = btr_page_get_level(page);
root_page_no = dict_index_get_page(index);
{
btr_cur_t cursor;
rec_offs* offsets = NULL;
mem_heap_t* heap = mem_heap_create(
sizeof(*offsets)
* (REC_OFFS_HEADER_SIZE + 1 + 1
+ unsigned(index->n_fields)));
buf_block_t* b;
if (dict_index_is_spatial(index)) {
offsets = rtr_page_get_father_block(
NULL, heap, index, block, mtr,
NULL, &cursor);
} else {
offsets = btr_page_get_father_block(offsets, heap,
index, block,
mtr, &cursor);
}
father_block = btr_cur_get_block(&cursor);
father_page_zip = buf_block_get_page_zip(father_block);
n_blocks = 0;
/* Store all ancestor pages so we can reset their
levels later on. We have to do all the searches on
the tree now because later on, after we've replaced
the first level, the tree is in an inconsistent state
and can not be searched. */
for (b = father_block;
b->page.id().page_no() != root_page_no; ) {
ut_a(n_blocks < BTR_MAX_LEVELS);
if (dict_index_is_spatial(index)) {
offsets = rtr_page_get_father_block(
NULL, heap, index, b, mtr,
NULL, &cursor);
} else {
offsets = btr_page_get_father_block(offsets,
heap,
index, b,
mtr,
&cursor);
}
blocks[n_blocks++] = b = btr_cur_get_block(&cursor);
}
lift_father_up = (n_blocks && page_level == 0);
if (lift_father_up) {
/* The father page also should be the only on its level (not
root). We should lift up the father page at first.
Because the leaf page should be lifted up only for root page.
The freeing page is based on page_level (==0 or !=0)
to choose segment. If the page_level is changed ==0 from !=0,
later freeing of the page doesn't find the page allocation
to be freed.*/
block = father_block;
page = buf_block_get_frame(block);
page_level = btr_page_get_level(page);
ut_ad(!page_has_siblings(page));
ut_ad(mtr->memo_contains_flagged(block,
MTR_MEMO_PAGE_X_FIX));
father_block = blocks[0];
father_page_zip = buf_block_get_page_zip(father_block);
}
mem_heap_free(heap);
}
btr_search_drop_page_hash_index(block);
/* Make the father empty */
btr_page_empty(father_block, father_page_zip, index, page_level, mtr);
/* btr_page_empty() is supposed to zero-initialize the field. */
ut_ad(!page_get_instant(father_block->page.frame));
if (index->is_instant()
&& father_block->page.id().page_no() == root_page_no) {
ut_ad(!father_page_zip);
btr_set_instant(father_block, *index, mtr);
}
page_level++;
/* Copy the records to the father page one by one. */
if (0
#ifdef UNIV_ZIP_COPY
|| father_page_zip
#endif /* UNIV_ZIP_COPY */
|| !page_copy_rec_list_end(father_block, block,
page_get_infimum_rec(page),
index, mtr, err)) {
switch (*err) {
case DB_SUCCESS:
break;
case DB_FAIL:
*err = DB_SUCCESS;
break;
default:
return nullptr;
}
const page_zip_des_t* page_zip
= buf_block_get_page_zip(block);
ut_a(father_page_zip);
ut_a(page_zip);
/* Copy the page byte for byte. */
page_zip_copy_recs(father_block,
page_zip, page, index, mtr);
/* Update the lock table and possible hash index. */
if (index->has_locking()) {
lock_move_rec_list_end(father_block, block,
page_get_infimum_rec(page));
}
/* Also update the predicate locks */
if (dict_index_is_spatial(index)) {
lock_prdt_rec_move(father_block, block->page.id());
} else {
btr_search_move_or_delete_hash_entries(
father_block, block);
}
}
if (index->has_locking()) {
const page_id_t id{block->page.id()};
/* Free predicate page locks on the block */
if (index->is_spatial()) {
lock_sys.prdt_page_free_from_discard(id);
} else {
lock_update_copy_and_discard(*father_block, id);
}
}
/* Go upward to root page, decrementing levels by one. */
for (i = lift_father_up ? 1 : 0; i < n_blocks; i++, page_level++) {
ut_ad(btr_page_get_level(blocks[i]->page.frame)
== page_level + 1);
btr_page_set_level(blocks[i], page_level, mtr);
}
if (dict_index_is_spatial(index)) {
rtr_check_discard_page(index, NULL, block);
}
/* Free the file page */
btr_page_free(index, block, mtr);
/* We play it safe and reset the free bits for the father */
if (!dict_index_is_clust(index)
&& !index->table->is_temporary()) {
ibuf_reset_free_bits(father_block);
}
ut_ad(page_validate(father_block->page.frame, index));
ut_ad(btr_check_node_ptr(index, father_block, mtr));
return(lift_father_up ? block_orig : father_block);
}
/*************************************************************//**
Tries to merge the page first to the left immediate brother if such a
brother exists, and the node pointers to the current page and to the brother
reside on the same page. If the left brother does not satisfy these
conditions, looks at the right brother. If the page is the only one on that
level lifts the records of the page to the father page, thus reducing the
tree height. It is assumed that mtr holds an x-latch on the tree and on the
page. If cursor is on the leaf level, mtr must also hold x-latches to the
brothers, if they exist.
@return error code */
dberr_t
btr_compress(
/*=========*/
btr_cur_t* cursor, /*!< in/out: cursor on the page to merge
or lift; the page must not be empty:
when deleting records, use btr_discard_page()
if the page would become empty */
bool adjust, /*!< in: whether the cursor position should be
adjusted even when compression occurs */
mtr_t* mtr) /*!< in/out: mini-transaction */
{
dict_index_t* index;
buf_block_t* merge_block = nullptr;
page_t* merge_page = nullptr;
page_zip_des_t* merge_page_zip;
ibool is_left;
buf_block_t* block;
page_t* page;
btr_cur_t father_cursor;
mem_heap_t* heap;
rec_offs* offsets;
ulint nth_rec = 0; /* remove bogus warning */
bool mbr_changed = false;
#ifdef UNIV_DEBUG
bool leftmost_child;
#endif
DBUG_ENTER("btr_compress");
block = btr_cur_get_block(cursor);
page = btr_cur_get_page(cursor);
index = btr_cur_get_index(cursor);
ut_ad(mtr->memo_contains_flagged(&index->lock, MTR_MEMO_X_LOCK
| MTR_MEMO_SX_LOCK));
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
MONITOR_INC(MONITOR_INDEX_MERGE_ATTEMPTS);
const uint32_t left_page_no = btr_page_get_prev(page);
const uint32_t right_page_no = btr_page_get_next(page);
dberr_t err = DB_SUCCESS;
ut_ad(page_is_leaf(page) || left_page_no != FIL_NULL
|| (REC_INFO_MIN_REC_FLAG & rec_get_info_bits(
page_rec_get_next(page_get_infimum_rec(page)),
page_is_comp(page))));
heap = mem_heap_create(100);
if (dict_index_is_spatial(index)) {
offsets = rtr_page_get_father_block(
NULL, heap, index, block, mtr, cursor, &father_cursor);
ut_ad(cursor->page_cur.block->page.id() == block->page.id());
rec_t* my_rec = father_cursor.page_cur.rec;
ulint page_no = btr_node_ptr_get_child_page_no(my_rec, offsets);
if (page_no != block->page.id().page_no()) {
ib::info() << "father positioned on page "
<< page_no << "instead of "
<< block->page.id().page_no();
offsets = btr_page_get_father_block(
NULL, heap, index, block, mtr, &father_cursor);
}
} else {
offsets = btr_page_get_father_block(
NULL, heap, index, block, mtr, &father_cursor);
}
if (adjust) {
nth_rec = page_rec_get_n_recs_before(btr_cur_get_rec(cursor));
if (UNIV_UNLIKELY(!nth_rec || nth_rec == ULINT_UNDEFINED)) {
err = DB_CORRUPTION;
goto err_exit;
}
}
if (left_page_no == FIL_NULL && right_page_no == FIL_NULL) {
/* The page is the only one on the level, lift the records
to the father */
merge_block = btr_lift_page_up(index, block, mtr, &err);
goto func_exit;
}
ut_d(leftmost_child =
left_page_no != FIL_NULL
&& (page_rec_get_next(
page_get_infimum_rec(
btr_cur_get_page(&father_cursor)))
== btr_cur_get_rec(&father_cursor)));
/* Decide the page to which we try to merge and which will inherit
the locks */
is_left = btr_can_merge_with_page(cursor, left_page_no,
&merge_block, mtr);
DBUG_EXECUTE_IF("ib_always_merge_right", is_left = FALSE;);
retry:
if (!is_left
&& !btr_can_merge_with_page(cursor, right_page_no, &merge_block,
mtr)) {
if (!merge_block) {
merge_page = NULL;
}
cannot_merge:
err = DB_FAIL;
goto err_exit;
}
merge_page = buf_block_get_frame(merge_block);
if (UNIV_UNLIKELY(memcmp_aligned<4>(merge_page + (is_left
? FIL_PAGE_NEXT
: FIL_PAGE_PREV),
block->page.frame
+ FIL_PAGE_OFFSET, 4))) {
err = DB_CORRUPTION;
goto err_exit;
}
ut_ad(page_validate(merge_page, index));
merge_page_zip = buf_block_get_page_zip(merge_block);
#ifdef UNIV_ZIP_DEBUG
if (merge_page_zip) {
const page_zip_des_t* page_zip
= buf_block_get_page_zip(block);
ut_a(page_zip);
ut_a(page_zip_validate(merge_page_zip, merge_page, index));
ut_a(page_zip_validate(page_zip, page, index));
}
#endif /* UNIV_ZIP_DEBUG */
/* Move records to the merge page */
if (is_left) {
btr_cur_t cursor2;
rtr_mbr_t new_mbr;
rec_offs* offsets2 = NULL;
/* For rtree, we need to update father's mbr. */
if (index->is_spatial()) {
/* We only support merge pages with the same parent
page */
if (!rtr_check_same_block(
index, &cursor2,
btr_cur_get_block(&father_cursor),
merge_block, heap)) {
is_left = false;
goto retry;
}
/* Set rtr_info for cursor2, since it is
necessary in recursive page merge. */
cursor2.rtr_info = cursor->rtr_info;
cursor2.tree_height = cursor->tree_height;
offsets2 = rec_get_offsets(
btr_cur_get_rec(&cursor2), index, NULL,
page_is_leaf(btr_cur_get_page(&cursor2))
? index->n_fields : 0,
ULINT_UNDEFINED, &heap);
/* Check if parent entry needs to be updated */
mbr_changed = rtr_merge_mbr_changed(
&cursor2, &father_cursor,
offsets2, offsets, &new_mbr);
}
rec_t* orig_pred = page_copy_rec_list_start(
merge_block, block, page_get_supremum_rec(page),
index, mtr, &err);
if (!orig_pred) {
goto err_exit;
}
btr_search_drop_page_hash_index(block);
/* Remove the page from the level list */
err = btr_level_list_remove(*block, *index, mtr);
if (UNIV_UNLIKELY(err != DB_SUCCESS)) {
goto err_exit;
}
const page_id_t id{block->page.id()};
if (dict_index_is_spatial(index)) {
rec_t* my_rec = father_cursor.page_cur.rec;
ulint page_no = btr_node_ptr_get_child_page_no(
my_rec, offsets);
if (page_no != block->page.id().page_no()) {
ib::fatal() << "father positioned on "
<< page_no << " instead of "
<< block->page.id().page_no();
}
if (mbr_changed) {
rtr_update_mbr_field(
&cursor2, offsets2, &father_cursor,
merge_page, &new_mbr, NULL, mtr);
} else {
rtr_node_ptr_delete(&father_cursor, mtr);
}
/* No GAP lock needs to be worrying about */
lock_sys.prdt_page_free_from_discard(id);
} else {
err = btr_cur_node_ptr_delete(&father_cursor, mtr);
if (UNIV_UNLIKELY(err != DB_SUCCESS)) {
goto err_exit;
}
if (index->has_locking()) {
lock_update_merge_left(
*merge_block, orig_pred, id);
}
}
if (adjust) {
ulint n = page_rec_get_n_recs_before(orig_pred);
if (UNIV_UNLIKELY(!n || n == ULINT_UNDEFINED)) {
err = DB_CORRUPTION;
goto err_exit;
}
nth_rec += n;
}
} else {
rec_t* orig_succ;
ibool compressed;
dberr_t err;
btr_cur_t cursor2;
/* father cursor pointing to node ptr
of the right sibling */
byte fil_page_prev[4];
if (dict_index_is_spatial(index)) {
cursor2.rtr_info = NULL;
/* For spatial index, we disallow merge of blocks
with different parents, since the merge would need
to update entry (for MBR and Primary key) in the
parent of block being merged */
if (!rtr_check_same_block(
index, &cursor2,
btr_cur_get_block(&father_cursor),
merge_block, heap)) {
goto cannot_merge;
}
/* Set rtr_info for cursor2, since it is
necessary in recursive page merge. */
cursor2.rtr_info = cursor->rtr_info;
cursor2.tree_height = cursor->tree_height;
} else if (!btr_page_get_father(index, merge_block, mtr,
&cursor2)) {
goto cannot_merge;
}
if (merge_page_zip && left_page_no == FIL_NULL) {
/* The function page_zip_compress(), which will be
invoked by page_copy_rec_list_end() below,
requires that FIL_PAGE_PREV be FIL_NULL.
Clear the field, but prepare to restore it. */
static_assert(FIL_PAGE_PREV % 8 == 0, "alignment");
memcpy(fil_page_prev, merge_page + FIL_PAGE_PREV, 4);
compile_time_assert(FIL_NULL == 0xffffffffU);
memset_aligned<4>(merge_page + FIL_PAGE_PREV, 0xff, 4);
}
orig_succ = page_copy_rec_list_end(merge_block, block,
page_get_infimum_rec(page),
cursor->index, mtr, &err);
if (!orig_succ) {
ut_a(merge_page_zip);
if (left_page_no == FIL_NULL) {
/* FIL_PAGE_PREV was restored from
merge_page_zip. */
ut_ad(!memcmp(fil_page_prev,
merge_page + FIL_PAGE_PREV, 4));
}
goto err_exit;
}
btr_search_drop_page_hash_index(block);
if (merge_page_zip && left_page_no == FIL_NULL) {
/* Restore FIL_PAGE_PREV in order to avoid an assertion
failure in btr_level_list_remove(), which will set
the field again to FIL_NULL. Even though this makes
merge_page and merge_page_zip inconsistent for a
split second, it is harmless, because the pages
are X-latched. */
memcpy(merge_page + FIL_PAGE_PREV, fil_page_prev, 4);
}
/* Remove the page from the level list */
err = btr_level_list_remove(*block, *index, mtr);
if (UNIV_UNLIKELY(err != DB_SUCCESS)) {
goto err_exit;
}
ut_ad(btr_node_ptr_get_child_page_no(
btr_cur_get_rec(&father_cursor), offsets)
== block->page.id().page_no());
/* Replace the address of the old child node (= page) with the
address of the merge page to the right */
btr_node_ptr_set_child_page_no(
btr_cur_get_block(&father_cursor),
btr_cur_get_rec(&father_cursor),
offsets, right_page_no, mtr);
#ifdef UNIV_DEBUG
if (!page_is_leaf(page) && left_page_no == FIL_NULL) {
ut_ad(REC_INFO_MIN_REC_FLAG & rec_get_info_bits(
page_rec_get_next(page_get_infimum_rec(
buf_block_get_frame(merge_block))),
page_is_comp(page)));
}
#endif /* UNIV_DEBUG */
/* For rtree, we need to update father's mbr. */
if (index->is_spatial()) {
rec_offs* offsets2;
ulint rec_info;
offsets2 = rec_get_offsets(
btr_cur_get_rec(&cursor2), index, NULL,
page_is_leaf(btr_cur_get_page(&cursor2))
? index->n_fields : 0,
ULINT_UNDEFINED, &heap);
ut_ad(btr_node_ptr_get_child_page_no(
btr_cur_get_rec(&cursor2), offsets2)
== right_page_no);
rec_info = rec_get_info_bits(
btr_cur_get_rec(&father_cursor),
rec_offs_comp(offsets));
if (rec_info & REC_INFO_MIN_REC_FLAG) {
/* When the father node ptr is minimal rec,
we will keep it and delete the node ptr of
merge page. */
rtr_merge_and_update_mbr(&father_cursor,
&cursor2,
offsets, offsets2,
merge_page, mtr);
} else {
/* Otherwise, we will keep the node ptr of
merge page and delete the father node ptr.
This is for keeping the rec order in upper
level. */
rtr_merge_and_update_mbr(&cursor2,
&father_cursor,
offsets2, offsets,
merge_page, mtr);
}
const page_id_t id{block->page.id()};
lock_sys.prdt_page_free_from_discard(id);
} else {
compressed = btr_cur_pessimistic_delete(&err, TRUE,
&cursor2,
BTR_CREATE_FLAG,
false, mtr);
ut_a(err == DB_SUCCESS);
if (!compressed) {
btr_cur_compress_if_useful(&cursor2, false,
mtr);
}
if (index->has_locking()) {
lock_update_merge_right(
merge_block, orig_succ, block);
}
}
}
if (!dict_index_is_clust(index)
&& !index->table->is_temporary()
&& page_is_leaf(merge_page)) {
/* Update the free bits of the B-tree page in the
insert buffer bitmap. This has to be done in a
separate mini-transaction that is committed before the
main mini-transaction. We cannot update the insert
buffer bitmap in this mini-transaction, because
btr_compress() can be invoked recursively without
committing the mini-transaction in between. Since
insert buffer bitmap pages have a lower rank than
B-tree pages, we must not access other pages in the
same mini-transaction after accessing an insert buffer
bitmap page. */
/* The free bits in the insert buffer bitmap must
never exceed the free space on a page. It is safe to
decrement or reset the bits in the bitmap in a
mini-transaction that is committed before the
mini-transaction that affects the free space. */
/* It is unsafe to increment the bits in a separately
committed mini-transaction, because in crash recovery,
the free bits could momentarily be set too high. */
if (merge_block->zip_size()) {
/* Because the free bits may be incremented
and we cannot update the insert buffer bitmap
in the same mini-transaction, the only safe
thing we can do here is the pessimistic
approach: reset the free bits. */
ibuf_reset_free_bits(merge_block);
} else {
/* On uncompressed pages, the free bits will
never increase here. Thus, it is safe to
write the bits accurately in a separate
mini-transaction. */
ibuf_update_free_bits_if_full(merge_block,
srv_page_size,
ULINT_UNDEFINED);
}
}
ut_ad(page_validate(merge_page, index));
#ifdef UNIV_ZIP_DEBUG
ut_a(!merge_page_zip || page_zip_validate(merge_page_zip, merge_page,
index));
#endif /* UNIV_ZIP_DEBUG */
if (dict_index_is_spatial(index)) {
rtr_check_discard_page(index, NULL, block);
}
/* Free the file page */
err = btr_page_free(index, block, mtr);
if (err == DB_SUCCESS) {
ut_ad(leftmost_child
|| btr_check_node_ptr(index, merge_block, mtr));
func_exit:
if (adjust) {
ut_ad(nth_rec > 0);
if (rec_t* nth
= page_rec_get_nth(merge_block->page.frame,
nth_rec)) {
btr_cur_position(index, nth,
merge_block, cursor);
} else {
err = DB_CORRUPTION;
goto err_exit;
}
}
MONITOR_INC(MONITOR_INDEX_MERGE_SUCCESSFUL);
} else {
err_exit:
/* We play it safe and reset the free bits. */
if (merge_block && merge_block->zip_size()
&& page_is_leaf(merge_block->page.frame)
&& !index->is_clust()) {
ibuf_reset_free_bits(merge_block);
}
}
mem_heap_free(heap);
DBUG_RETURN(err);
}
/*************************************************************//**
Discards a page that is the only page on its level. This will empty
the whole B-tree, leaving just an empty root page. This function
should almost never be reached, because btr_compress(), which is invoked in
delete operations, calls btr_lift_page_up() to flatten the B-tree. */
ATTRIBUTE_COLD
static
void
btr_discard_only_page_on_level(
/*===========================*/
dict_index_t* index, /*!< in: index tree */
buf_block_t* block, /*!< in: page which is the only on its level */
mtr_t* mtr) /*!< in: mtr */
{
ulint page_level = 0;
ut_ad(!index->is_dummy);
/* Save the PAGE_MAX_TRX_ID from the leaf page. */
const trx_id_t max_trx_id = page_get_max_trx_id(block->page.frame);
const rec_t* r = page_rec_get_next(
page_get_infimum_rec(block->page.frame));
/* In the caller we checked that a valid key exists in the page,
because we were able to look up a parent page. */
ut_ad(r);
ut_ad(rec_is_metadata(r, *index) == index->is_instant());
while (block->page.id().page_no() != dict_index_get_page(index)) {
btr_cur_t cursor;
buf_block_t* father;
const page_t* page = buf_block_get_frame(block);
ut_a(page_get_n_recs(page) == 1);
ut_a(page_level == btr_page_get_level(page));
ut_a(!page_has_siblings(page));
ut_ad(fil_page_index_page_check(page));
ut_ad(block->page.id().space() == index->table->space->id);
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
btr_search_drop_page_hash_index(block);
if (index->is_spatial()) {
/* Check any concurrent search having this page */
rtr_check_discard_page(index, NULL, block);
if (!rtr_page_get_father(index, block, mtr, nullptr,
&cursor)) {
return;
}
} else {
if (!btr_page_get_father(index, block, mtr, &cursor)) {
return;
}
}
father = btr_cur_get_block(&cursor);
if (index->has_locking()) {
lock_update_discard(
father, PAGE_HEAP_NO_SUPREMUM, block);
}
/* Free the file page */
if (btr_page_free(index, block, mtr) != DB_SUCCESS) {
return;
}
block = father;
page_level++;
}
/* block is the root page, which must be empty, except
for the node pointer to the (now discarded) block(s). */
ut_ad(!page_has_siblings(block->page.frame));
mem_heap_t* heap = nullptr;
const rec_t* rec = nullptr;
rec_offs* offsets = nullptr;
if (index->table->instant || index->must_avoid_clear_instant_add()) {
if (!rec_is_metadata(r, *index)) {
} else if (!index->table->instant
|| rec_is_alter_metadata(r, *index)) {
heap = mem_heap_create(srv_page_size);
offsets = rec_get_offsets(r, index, nullptr,
index->n_core_fields,
ULINT_UNDEFINED, &heap);
rec = rec_copy(mem_heap_alloc(heap,
rec_offs_size(offsets)),
r, offsets);
rec_offs_make_valid(rec, index, true, offsets);
}
}
btr_page_empty(block, buf_block_get_page_zip(block), index, 0, mtr);
ut_ad(page_is_leaf(buf_block_get_frame(block)));
/* btr_page_empty() is supposed to zero-initialize the field. */
ut_ad(!page_get_instant(block->page.frame));
if (index->is_primary()) {
if (rec) {
page_cur_t cur;
page_cur_set_before_first(block, &cur);
DBUG_ASSERT(index->table->instant);
DBUG_ASSERT(rec_is_alter_metadata(rec, *index));
btr_set_instant(block, *index, mtr);
rec = page_cur_insert_rec_low(&cur, index, rec,
offsets, mtr);
ut_ad(rec);
mem_heap_free(heap);
} else if (index->is_instant()) {
index->clear_instant_add();
}
} else if (!index->table->is_temporary()) {
/* We play it safe and reset the free bits for the root */
ibuf_reset_free_bits(block);
ut_a(max_trx_id);
page_set_max_trx_id(block,
buf_block_get_page_zip(block),
max_trx_id, mtr);
}
}
/*************************************************************//**
Discards a page from a B-tree. This is used to remove the last record from
a B-tree page: the whole page must be removed at the same time. This cannot
be used for the root page, which is allowed to be empty. */
dberr_t
btr_discard_page(
/*=============*/
btr_cur_t* cursor, /*!< in: cursor on the page to discard: not on
the root page */
mtr_t* mtr) /*!< in: mtr */
{
dict_index_t* index;
buf_block_t* merge_block;
buf_block_t* block;
btr_cur_t parent_cursor;
block = btr_cur_get_block(cursor);
index = btr_cur_get_index(cursor);
ut_ad(dict_index_get_page(index) != block->page.id().page_no());
ut_ad(mtr->memo_contains_flagged(&index->lock, MTR_MEMO_X_LOCK
| MTR_MEMO_SX_LOCK));
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
MONITOR_INC(MONITOR_INDEX_DISCARD);
if (index->is_spatial()
? !rtr_page_get_father(index, block, mtr, cursor, &parent_cursor)
: !btr_page_get_father(index, block, mtr, &parent_cursor)) {
return DB_CORRUPTION;
}
/* Decide the page which will inherit the locks */
const uint32_t left_page_no = btr_page_get_prev(block->page.frame);
const uint32_t right_page_no = btr_page_get_next(block->page.frame);
ut_d(bool parent_is_different = false);
if (left_page_no != FIL_NULL) {
dberr_t err;
merge_block = btr_block_get(*index, left_page_no, RW_X_LATCH,
true, mtr, &err);
if (UNIV_UNLIKELY(!merge_block)) {
return err;
}
if (UNIV_UNLIKELY(memcmp_aligned<4>(merge_block->page.frame
+ FIL_PAGE_NEXT,
block->page.frame
+ FIL_PAGE_OFFSET, 4))) {
return DB_CORRUPTION;
}
ut_d(parent_is_different =
(page_rec_get_next(
page_get_infimum_rec(
btr_cur_get_page(
&parent_cursor)))
== btr_cur_get_rec(&parent_cursor)));
} else if (right_page_no != FIL_NULL) {
dberr_t err;
merge_block = btr_block_get(*index, right_page_no, RW_X_LATCH,
true, mtr, &err);
if (UNIV_UNLIKELY(!merge_block)) {
return err;
}
if (UNIV_UNLIKELY(memcmp_aligned<4>(merge_block->page.frame
+ FIL_PAGE_PREV,
block->page.frame
+ FIL_PAGE_OFFSET, 4))) {
return DB_CORRUPTION;
}
ut_d(parent_is_different = page_rec_is_supremum(
page_rec_get_next(btr_cur_get_rec(&parent_cursor))));
if (page_is_leaf(merge_block->page.frame)) {
} else if (rec_t* node_ptr =
page_rec_get_next(page_get_infimum_rec(
merge_block->page.frame))) {
ut_ad(page_rec_is_user_rec(node_ptr));
/* We have to mark the leftmost node pointer as the
predefined minimum record. */
btr_set_min_rec_mark<true>(node_ptr, *merge_block,
mtr);
} else {
return DB_CORRUPTION;
}
} else {
btr_discard_only_page_on_level(index, block, mtr);
return DB_SUCCESS;
}
if (UNIV_UNLIKELY(memcmp_aligned<2>(&merge_block->page.frame
[PAGE_HEADER + PAGE_LEVEL],
&block->page.frame
[PAGE_HEADER + PAGE_LEVEL], 2))) {
return DB_CORRUPTION;
}
btr_search_drop_page_hash_index(block);
if (dict_index_is_spatial(index)) {
rtr_node_ptr_delete(&parent_cursor, mtr);
} else if (dberr_t err =
btr_cur_node_ptr_delete(&parent_cursor, mtr)) {
return err;
}
/* Remove the page from the level list */
if (dberr_t err = btr_level_list_remove(*block, *index, mtr)) {
return err;
}
#ifdef UNIV_ZIP_DEBUG
{
page_zip_des_t* merge_page_zip
= buf_block_get_page_zip(merge_block);
ut_a(!merge_page_zip
|| page_zip_validate(merge_page_zip,
merge_block->page.frame, index));
}
#endif /* UNIV_ZIP_DEBUG */
if (index->has_locking()) {
if (left_page_no != FIL_NULL) {
lock_update_discard(merge_block, PAGE_HEAP_NO_SUPREMUM,
block);
} else {
lock_update_discard(merge_block,
lock_get_min_heap_no(merge_block),
block);
}
if (index->is_spatial()) {
rtr_check_discard_page(index, cursor, block);
}
}
/* Free the file page */
dberr_t err = btr_page_free(index, block, mtr);
if (err == DB_SUCCESS) {
/* btr_check_node_ptr() needs parent block latched.
If the merge_block's parent block is not same,
we cannot use btr_check_node_ptr() */
ut_ad(parent_is_different
|| btr_check_node_ptr(index, merge_block, mtr));
if (btr_cur_get_block(&parent_cursor)->page.id().page_no()
== index->page
&& !page_has_siblings(btr_cur_get_page(&parent_cursor))
&& page_get_n_recs(btr_cur_get_page(&parent_cursor))
== 1) {
btr_lift_page_up(index, merge_block, mtr, &err);
}
}
return err;
}
#ifdef UNIV_BTR_PRINT
/*************************************************************//**
Prints size info of a B-tree. */
void
btr_print_size(
/*===========*/
dict_index_t* index) /*!< in: index tree */
{
page_t* root;
fseg_header_t* seg;
mtr_t mtr;
if (dict_index_is_ibuf(index)) {
fputs("Sorry, cannot print info of an ibuf tree:"
" use ibuf functions\n", stderr);
return;
}
mtr_start(&mtr);
root = btr_root_get(index, &mtr);
seg = root + PAGE_HEADER + PAGE_BTR_SEG_TOP;
fputs("INFO OF THE NON-LEAF PAGE SEGMENT\n", stderr);
fseg_print(seg, &mtr);
if (!dict_index_is_ibuf(index)) {
seg = root + PAGE_HEADER + PAGE_BTR_SEG_LEAF;
fputs("INFO OF THE LEAF PAGE SEGMENT\n", stderr);
fseg_print(seg, &mtr);
}
mtr_commit(&mtr);
}
/************************************************************//**
Prints recursively index tree pages. */
static
void
btr_print_recursive(
/*================*/
dict_index_t* index, /*!< in: index tree */
buf_block_t* block, /*!< in: index page */
ulint width, /*!< in: print this many entries from start
and end */
mem_heap_t** heap, /*!< in/out: heap for rec_get_offsets() */
rec_offs** offsets,/*!< in/out: buffer for rec_get_offsets() */
mtr_t* mtr) /*!< in: mtr */
{
const page_t* page = buf_block_get_frame(block);
page_cur_t cursor;
ulint n_recs;
ulint i = 0;
mtr_t mtr2;
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_SX_FIX));
ib::info() << "NODE ON LEVEL " << btr_page_get_level(page)
<< " page " << block->page.id;
page_print(block, index, width, width);
n_recs = page_get_n_recs(page);
page_cur_set_before_first(block, &cursor);
page_cur_move_to_next(&cursor);
while (!page_cur_is_after_last(&cursor)) {
if (page_is_leaf(page)) {
/* If this is the leaf level, do nothing */
} else if ((i <= width) || (i >= n_recs - width)) {
const rec_t* node_ptr;
mtr_start(&mtr2);
node_ptr = page_cur_get_rec(&cursor);
*offsets = rec_get_offsets(
node_ptr, index, *offsets, 0,
ULINT_UNDEFINED, heap);
if (buf_block_t *child =
btr_node_ptr_get_child(node_ptr, index, *offsets,
&mtr2)) {
btr_print_recursive(index, child, width, heap,
offsets, &mtr2);
}
mtr_commit(&mtr2);
}
page_cur_move_to_next(&cursor);
i++;
}
}
/**************************************************************//**
Prints directories and other info of all nodes in the tree. */
void
btr_print_index(
/*============*/
dict_index_t* index, /*!< in: index */
ulint width) /*!< in: print this many entries from start
and end */
{
mtr_t mtr;
buf_block_t* root;
mem_heap_t* heap = NULL;
rec_offs offsets_[REC_OFFS_NORMAL_SIZE];
rec_offs* offsets = offsets_;
rec_offs_init(offsets_);
fputs("--------------------------\n"
"INDEX TREE PRINT\n", stderr);
mtr_start(&mtr);
root = btr_root_block_get(index, RW_SX_LATCH, &mtr);
btr_print_recursive(index, root, width, &heap, &offsets, &mtr);
if (heap) {
mem_heap_free(heap);
}
mtr_commit(&mtr);
ut_ad(btr_validate_index(index, 0));
}
#endif /* UNIV_BTR_PRINT */
#ifdef UNIV_DEBUG
/************************************************************//**
Checks that the node pointer to a page is appropriate.
@return TRUE */
ibool
btr_check_node_ptr(
/*===============*/
dict_index_t* index, /*!< in: index tree */
buf_block_t* block, /*!< in: index page */
mtr_t* mtr) /*!< in: mtr */
{
mem_heap_t* heap;
dtuple_t* tuple;
rec_offs* offsets;
btr_cur_t cursor;
page_t* page = buf_block_get_frame(block);
ut_ad(mtr->memo_contains_flagged(block, MTR_MEMO_PAGE_X_FIX));
if (dict_index_get_page(index) == block->page.id().page_no()) {
return(TRUE);
}
heap = mem_heap_create(256);
if (dict_index_is_spatial(index)) {
offsets = rtr_page_get_father_block(NULL, heap, index, block, mtr,
NULL, &cursor);
} else {
offsets = btr_page_get_father_block(NULL, heap, index, block, mtr,
&cursor);
}
if (page_is_leaf(page)) {
goto func_exit;
}
tuple = dict_index_build_node_ptr(
index, page_rec_get_next(page_get_infimum_rec(page)), 0, heap,
btr_page_get_level(page));
/* For spatial index, the MBR in the parent rec could be different
with that of first rec of child, their relationship should be
"WITHIN" relationship */
if (dict_index_is_spatial(index)) {
ut_a(!cmp_dtuple_rec_with_gis(
tuple, btr_cur_get_rec(&cursor),
PAGE_CUR_WITHIN));
} else {
ut_a(!cmp_dtuple_rec(tuple, btr_cur_get_rec(&cursor), index,
offsets));
}
func_exit:
mem_heap_free(heap);
return(TRUE);
}
#endif /* UNIV_DEBUG */
/************************************************************//**
Display identification information for a record. */
static
void
btr_index_rec_validate_report(
/*==========================*/
const page_t* page, /*!< in: index page */
const rec_t* rec, /*!< in: index record */
const dict_index_t* index) /*!< in: index */
{
ib::info() << "Record in index " << index->name
<< " of table " << index->table->name
<< ", page " << page_id_t(page_get_space_id(page),
page_get_page_no(page))
<< ", at offset " << page_offset(rec);
}
/************************************************************//**
Checks the size and number of fields in a record based on the definition of
the index.
@return TRUE if ok */
ibool
btr_index_rec_validate(
/*===================*/
const rec_t* rec, /*!< in: index record */
const dict_index_t* index, /*!< in: index */
ibool dump_on_error) /*!< in: TRUE if the function
should print hex dump of record
and page on error */
{
ulint len;
const page_t* page;
mem_heap_t* heap = NULL;
rec_offs offsets_[REC_OFFS_NORMAL_SIZE];
rec_offs* offsets = offsets_;
rec_offs_init(offsets_);
page = page_align(rec);
ut_ad(index->n_core_fields);
if (index->is_ibuf()) {
/* The insert buffer index tree can contain records from any
other index: we cannot check the number of fields or
their length */
return(TRUE);
}
#ifdef VIRTUAL_INDEX_DEBUG
if (dict_index_has_virtual(index)) {
fprintf(stderr, "index name is %s\n", index->name());
}
#endif
if ((ibool)!!page_is_comp(page) != dict_table_is_comp(index->table)) {
btr_index_rec_validate_report(page, rec, index);
ib::error() << "Compact flag=" << !!page_is_comp(page)
<< ", should be " << dict_table_is_comp(index->table);
return(FALSE);
}
const bool is_alter_metadata = page_is_leaf(page)
&& !page_has_prev(page)
&& index->is_primary() && index->table->instant
&& rec == page_rec_get_next_const(page_get_infimum_rec(page));
if (is_alter_metadata
&& !rec_is_alter_metadata(rec, page_is_comp(page))) {
btr_index_rec_validate_report(page, rec, index);
ib::error() << "First record is not ALTER TABLE metadata";
return FALSE;
}
if (!page_is_comp(page)) {
const ulint n_rec_fields = rec_get_n_fields_old(rec);
if (n_rec_fields == DICT_FLD__SYS_INDEXES__MERGE_THRESHOLD
&& index->id == DICT_INDEXES_ID) {
/* A record for older SYS_INDEXES table
(missing merge_threshold column) is acceptable. */
} else if (is_alter_metadata) {
if (n_rec_fields != ulint(index->n_fields) + 1) {
goto n_field_mismatch;
}
} else if (n_rec_fields < index->n_core_fields
|| n_rec_fields > index->n_fields) {
n_field_mismatch:
btr_index_rec_validate_report(page, rec, index);
ib::error() << "Has " << rec_get_n_fields_old(rec)
<< " fields, should have "
<< index->n_core_fields << ".."
<< index->n_fields;
if (dump_on_error) {
fputs("InnoDB: corrupt record ", stderr);
rec_print_old(stderr, rec);
putc('\n', stderr);
}
return(FALSE);
}
}
offsets = rec_get_offsets(rec, index, offsets, page_is_leaf(page)
? index->n_core_fields : 0,
ULINT_UNDEFINED, &heap);
const dict_field_t* field = index->fields;
ut_ad(rec_offs_n_fields(offsets)
== ulint(index->n_fields) + is_alter_metadata);
for (unsigned i = 0; i < rec_offs_n_fields(offsets); i++) {
rec_get_nth_field_offs(offsets, i, &len);
ulint fixed_size;
if (is_alter_metadata && i == index->first_user_field()) {
fixed_size = FIELD_REF_SIZE;
if (len != FIELD_REF_SIZE
|| !rec_offs_nth_extern(offsets, i)) {
goto len_mismatch;
}
continue;
} else {
fixed_size = dict_col_get_fixed_size(
field->col, page_is_comp(page));
if (rec_offs_nth_extern(offsets, i)) {
const byte* data = rec_get_nth_field(
rec, offsets, i, &len);
len -= BTR_EXTERN_FIELD_REF_SIZE;
ulint extern_len = mach_read_from_4(
data + len + BTR_EXTERN_LEN + 4);
if (fixed_size == extern_len) {
goto next_field;
}
}
}
/* Note that if fixed_size != 0, it equals the
length of a fixed-size column in the clustered index.
We should adjust it here.
A prefix index of the column is of fixed, but different
length. When fixed_size == 0, prefix_len is the maximum
length of the prefix index column. */
if (len_is_stored(len)
&& (field->prefix_len
? len > field->prefix_len
: (fixed_size && len != fixed_size))) {
len_mismatch:
btr_index_rec_validate_report(page, rec, index);
ib::error error;
error << "Field " << i << " len is " << len
<< ", should be " << fixed_size;
if (dump_on_error) {
error << "; ";
rec_print(error.m_oss, rec,
rec_get_info_bits(
rec, rec_offs_comp(offsets)),
offsets);
}
if (heap) {
mem_heap_free(heap);
}
return(FALSE);
}
next_field:
field++;
}
#ifdef VIRTUAL_INDEX_DEBUG
if (dict_index_has_virtual(index)) {
rec_print_new(stderr, rec, offsets);
}
#endif
if (heap) {
mem_heap_free(heap);
}
return(TRUE);
}
/************************************************************//**
Checks the size and number of fields in records based on the definition of
the index.
@return true if ok */
static
bool
btr_index_page_validate(
/*====================*/
buf_block_t* block, /*!< in: index page */
dict_index_t* index) /*!< in: index */
{
page_cur_t cur;
#ifndef DBUG_OFF
ulint nth = 1;
#endif /* !DBUG_OFF */
page_cur_set_before_first(block, &cur);
/* Directory slot 0 should only contain the infimum record. */
DBUG_EXECUTE_IF("check_table_rec_next",
ut_a(page_rec_get_nth_const(
page_cur_get_page(&cur), 0)
== cur.rec);
ut_a(page_dir_slot_get_n_owned(
page_dir_get_nth_slot(
page_cur_get_page(&cur), 0))
== 1););
while (page_cur_move_to_next(&cur)) {
if (page_cur_is_after_last(&cur)) {
return true;
}
if (!btr_index_rec_validate(cur.rec, index, TRUE)) {
break;
}
/* Verify that page_rec_get_nth_const() is correctly
retrieving each record. */
DBUG_EXECUTE_IF("check_table_rec_next",
ut_a(cur.rec == page_rec_get_nth_const(
page_cur_get_page(&cur),
page_rec_get_n_recs_before(
cur.rec)));
ut_a(nth++ == page_rec_get_n_recs_before(
cur.rec)););
}
return false;
}
/************************************************************//**
Report an error on one page of an index tree. */
static
void
btr_validate_report1(
/*=================*/
dict_index_t* index, /*!< in: index */
ulint level, /*!< in: B-tree level */
const buf_block_t* block) /*!< in: index page */
{
ib::error error;
error << "In page " << block->page.id().page_no()
<< " of index " << index->name
<< " of table " << index->table->name;
if (level > 0) {
error << ", index tree level " << level;
}
}
/************************************************************//**
Report an error on two pages of an index tree. */
static
void
btr_validate_report2(
/*=================*/
const dict_index_t* index, /*!< in: index */
ulint level, /*!< in: B-tree level */
const buf_block_t* block1, /*!< in: first index page */
const buf_block_t* block2) /*!< in: second index page */
{
ib::error error;
error << "In pages " << block1->page.id()
<< " and " << block2->page.id() << " of index " << index->name
<< " of table " << index->table->name;
if (level)
error << ", index tree level " << level;
}
/** Validate an index tree level. */
static
dberr_t
btr_validate_level(
/*===============*/
dict_index_t* index, /*!< in: index tree */
const trx_t* trx, /*!< in: transaction or NULL */
ulint level, /*!< in: level number */
bool lockout)/*!< in: true if X-latch index is intended */
{
buf_block_t* block;
page_t* page;
buf_block_t* right_block = 0; /* remove warning */
page_t* right_page = 0; /* remove warning */
page_t* father_page;
btr_cur_t node_cur;
btr_cur_t right_node_cur;
rec_t* rec;
page_cur_t cursor;
dtuple_t* node_ptr_tuple;
mtr_t mtr;
mem_heap_t* heap = mem_heap_create(256);
rec_offs* offsets = NULL;
rec_offs* offsets2= NULL;
#ifdef UNIV_ZIP_DEBUG
page_zip_des_t* page_zip;
#endif /* UNIV_ZIP_DEBUG */
ulint savepoint = 0;
ulint savepoint2 = 0;
uint32_t parent_page_no = FIL_NULL;
uint32_t parent_right_page_no = FIL_NULL;
bool rightmost_child = false;
mtr.start();
if (!srv_read_only_mode) {
if (lockout) {
mtr_x_lock_index(index, &mtr);
} else {
mtr_sx_lock_index(index, &mtr);
}
}
dberr_t err;
block = btr_root_block_get(index, RW_SX_LATCH, &mtr, &err);
if (!block) {
mtr.commit();
return err;
}
page = buf_block_get_frame(block);
fil_space_t* space = index->table->space;
while (level != btr_page_get_level(page)) {
const rec_t* node_ptr;
switch (dberr_t e =
fseg_page_is_allocated(space,
block->page.id().page_no())) {
case DB_SUCCESS_LOCKED_REC:
break;
case DB_SUCCESS:
btr_validate_report1(index, level, block);
ib::warn() << "Page is free";
e = DB_CORRUPTION;
/* fall through */
default:
err = e;
}
ut_ad(index->table->space_id == block->page.id().space());
ut_ad(block->page.id().space() == page_get_space_id(page));
#ifdef UNIV_ZIP_DEBUG
page_zip = buf_block_get_page_zip(block);
ut_a(!page_zip || page_zip_validate(page_zip, page, index));
#endif /* UNIV_ZIP_DEBUG */
if (page_is_leaf(page)) {
corrupted:
err = DB_CORRUPTION;
goto invalid_page;
}
page_cur_set_before_first(block, &cursor);
if (!(node_ptr = page_cur_move_to_next(&cursor))) {
goto corrupted;
}
offsets = rec_get_offsets(node_ptr, index, offsets, 0,
ULINT_UNDEFINED, &heap);
savepoint2 = mtr_set_savepoint(&mtr);
block = btr_node_ptr_get_child(node_ptr, index, offsets, &mtr,
&err);
if (!block) {
break;
}
page = buf_block_get_frame(block);
/* For R-Tree, since record order might not be the same as
linked index page in the lower level, we need to travers
backwards to get the first page rec in this level.
This is only used for index validation. Spatial index
does not use such scan for any of its DML or query
operations */
if (dict_index_is_spatial(index)) {
uint32_t left_page_no = btr_page_get_prev(page);
while (left_page_no != FIL_NULL) {
/* To obey latch order of tree blocks,
we should release the right_block once to
obtain lock of the uncle block. */
mtr_release_block_at_savepoint(
&mtr, savepoint2, block);
savepoint2 = mtr_set_savepoint(&mtr);
block = btr_block_get(*index, left_page_no,
RW_SX_LATCH, false,
&mtr, &err);
if (!block) {
goto invalid_page;
}
page = buf_block_get_frame(block);
left_page_no = btr_page_get_prev(page);
}
}
}
/* Now we are on the desired level. Loop through the pages on that
level. */
loop:
if (!block) {
invalid_page:
mtr.commit();
func_exit:
mem_heap_free(heap);
return err;
}
mem_heap_empty(heap);
offsets = offsets2 = NULL;
if (!srv_read_only_mode) {
if (lockout) {
mtr_x_lock_index(index, &mtr);
} else {
mtr_sx_lock_index(index, &mtr);
}
}
page = block->page.frame;
#ifdef UNIV_ZIP_DEBUG
page_zip = buf_block_get_page_zip(block);
ut_a(!page_zip || page_zip_validate(page_zip, page, index));
#endif /* UNIV_ZIP_DEBUG */
if (DB_SUCCESS_LOCKED_REC
!= fseg_page_is_allocated(space, block->page.id().page_no())) {
btr_validate_report1(index, level, block);
ib::warn() << "Page is marked as free";
err = DB_CORRUPTION;
} else if (btr_page_get_index_id(page) != index->id) {
ib::error() << "Page index id " << btr_page_get_index_id(page)
<< " != data dictionary index id " << index->id;
err = DB_CORRUPTION;
} else if (!page_validate(page, index)) {
btr_validate_report1(index, level, block);
err = DB_CORRUPTION;
} else if (btr_page_get_level(page) != level) {
btr_validate_report1(index, level, block);
ib::error() << "Page level is not " << level;
err = DB_CORRUPTION;
} else if (level == 0 && !btr_index_page_validate(block, index)) {
/* We are on level 0. Check that the records have the right
number of fields, and field lengths are right. */
err = DB_CORRUPTION;
} else if (!page_is_empty(page)) {
} else if (level) {
btr_validate_report1(index, level, block);
ib::error() << "Non-leaf page is empty";
} else if (block->page.id().page_no() != index->page) {
btr_validate_report1(index, level, block);
ib::error() << "Empty leaf page is not index root";
}
uint32_t right_page_no = btr_page_get_next(page);
uint32_t left_page_no = btr_page_get_prev(page);
if (right_page_no != FIL_NULL) {
const rec_t* right_rec;
savepoint = mtr_set_savepoint(&mtr);
right_block = btr_block_get(*index, right_page_no, RW_SX_LATCH,
!level, &mtr, &err);
if (!right_block) {
btr_validate_report1(index, level, block);
fputs("InnoDB: broken FIL_PAGE_NEXT link\n", stderr);
goto invalid_page;
}
right_page = buf_block_get_frame(right_block);
if (btr_page_get_prev(right_page) != page_get_page_no(page)) {
btr_validate_report2(index, level, block, right_block);
fputs("InnoDB: broken FIL_PAGE_NEXT"
" or FIL_PAGE_PREV links\n", stderr);
err = DB_CORRUPTION;
}
if (!(rec = page_rec_get_prev(page_get_supremum_rec(page)))) {
broken_links:
btr_validate_report1(index, level, block);
fputs("InnoDB: broken record links\n", stderr);
goto invalid_page;
}
if (!(right_rec =
page_rec_get_next(page_get_infimum_rec(right_page)))) {
goto broken_links;
}
offsets = rec_get_offsets(rec, index, offsets,
page_is_leaf(page)
? index->n_core_fields : 0,
ULINT_UNDEFINED, &heap);
offsets2 = rec_get_offsets(right_rec, index, offsets2,
page_is_leaf(right_page)
? index->n_core_fields : 0,
ULINT_UNDEFINED, &heap);
/* For spatial index, we cannot guarantee the key ordering
across pages, so skip the record compare verification for
now. Will enhanced in special R-Tree index validation scheme */
if (index->is_btree()
&& cmp_rec_rec(rec, right_rec,
offsets, offsets2, index) >= 0) {
btr_validate_report2(index, level, block, right_block);
fputs("InnoDB: records in wrong order"
" on adjacent pages\n", stderr);
rec = page_rec_get_prev(page_get_supremum_rec(page));
if (rec) {
fputs("InnoDB: record ", stderr);
rec_print(stderr, rec, index);
putc('\n', stderr);
}
fputs("InnoDB: record ", stderr);
rec = page_rec_get_next(
page_get_infimum_rec(right_page));
if (rec) {
rec_print(stderr, rec, index);
}
putc('\n', stderr);
err = DB_CORRUPTION;
}
}
if (!level || left_page_no != FIL_NULL) {
} else if (const rec_t* first =
page_rec_get_next_const(page_get_infimum_rec(page))) {
if (!(REC_INFO_MIN_REC_FLAG
& rec_get_info_bits(first, page_is_comp(page)))) {
btr_validate_report1(index, level, block);
ib::error() << "Missing REC_INFO_MIN_REC_FLAG";
err = DB_CORRUPTION;
}
} else {
err = DB_CORRUPTION;
goto node_ptr_fails;
}
/* Similarly skip the father node check for spatial index for now,
for a couple of reasons:
1) As mentioned, there is no ordering relationship between records
in parent level and linked pages in the child level.
2) Search parent from root is very costly for R-tree.
We will add special validation mechanism for R-tree later (WL #7520) */
if (index->is_btree() && block->page.id().page_no() != index->page) {
/* Check father node pointers */
rec_t* node_ptr
= page_rec_get_next(page_get_infimum_rec(page));
if (!node_ptr) {
err = DB_CORRUPTION;
goto node_ptr_fails;
}
btr_cur_position(index, node_ptr, block, &node_cur);
offsets = btr_page_get_father_node_ptr_for_validate(
offsets, heap, &node_cur, &mtr);
father_page = btr_cur_get_page(&node_cur);
node_ptr = btr_cur_get_rec(&node_cur);
parent_page_no = page_get_page_no(father_page);
parent_right_page_no = btr_page_get_next(father_page);
rightmost_child = page_rec_is_supremum(
page_rec_get_next(node_ptr));
rec = page_rec_get_prev(page_get_supremum_rec(page));
if (rec) {
btr_cur_position(index, rec, block, &node_cur);
offsets = btr_page_get_father_node_ptr_for_validate(
offsets, heap, &node_cur, &mtr);
} else {
offsets = nullptr;
}
if (!offsets || node_ptr != btr_cur_get_rec(&node_cur)
|| btr_node_ptr_get_child_page_no(node_ptr, offsets)
!= block->page.id().page_no()) {
btr_validate_report1(index, level, block);
fputs("InnoDB: node pointer to the page is wrong\n",
stderr);
fputs("InnoDB: node ptr ", stderr);
rec_print(stderr, node_ptr, index);
if (offsets) {
rec = btr_cur_get_rec(&node_cur);
fprintf(stderr, "\n"
"InnoDB: node ptr child page n:o %u\n",
btr_node_ptr_get_child_page_no(
rec, offsets));
fputs("InnoDB: record on page ", stderr);
rec_print_new(stderr, rec, offsets);
putc('\n', stderr);
}
err = DB_CORRUPTION;
goto node_ptr_fails;
}
if (page_is_leaf(page)) {
} else if (const rec_t* first_rec =
page_rec_get_next(page_get_infimum_rec(page))) {
node_ptr_tuple = dict_index_build_node_ptr(
index, first_rec,
0, heap, btr_page_get_level(page));
if (cmp_dtuple_rec(node_ptr_tuple, node_ptr, index,
offsets)) {
btr_validate_report1(index, level, block);
ib::error() << "Node ptrs differ on levels > 0";
fputs("InnoDB: node ptr ",stderr);
rec_print_new(stderr, node_ptr, offsets);
fputs("InnoDB: first rec ", stderr);
rec_print(stderr, first_rec, index);
putc('\n', stderr);
err = DB_CORRUPTION;
goto node_ptr_fails;
}
} else {
err = DB_CORRUPTION;
goto node_ptr_fails;
}
if (left_page_no == FIL_NULL) {
if (page_has_prev(father_page)
|| node_ptr != page_rec_get_next(
page_get_infimum_rec(father_page))) {
err = DB_CORRUPTION;
goto node_ptr_fails;
}
}
if (right_page_no == FIL_NULL) {
if (page_has_next(father_page)
|| node_ptr != page_rec_get_prev(
page_get_supremum_rec(father_page))) {
err = DB_CORRUPTION;
goto node_ptr_fails;
}
} else if (const rec_t* right_node_ptr
= page_rec_get_next(node_ptr)) {
if (!lockout && rightmost_child) {
/* To obey latch order of tree blocks,
we should release the right_block once to
obtain lock of the uncle block. */
mtr_release_block_at_savepoint(
&mtr, savepoint, right_block);
if (parent_right_page_no != FIL_NULL) {
btr_block_get(*index,
parent_right_page_no,
RW_SX_LATCH, false,
&mtr);
}
right_block = btr_block_get(*index,
right_page_no,
RW_SX_LATCH,
!level, &mtr,
&err);
if (!right_block) {
btr_validate_report1(index, level,
block);
fputs("InnoDB: broken FIL_PAGE_NEXT"
" link\n", stderr);
goto invalid_page;
}
}
btr_cur_position(
index,
page_get_infimum_rec(right_block->page.frame),
right_block, &right_node_cur);
if (!page_cur_move_to_next(&right_node_cur.page_cur)) {
goto node_pointer_corrupted;
}
offsets = btr_page_get_father_node_ptr_for_validate(
offsets, heap, &right_node_cur, &mtr);
if (right_node_ptr
!= page_get_supremum_rec(father_page)) {
if (btr_cur_get_rec(&right_node_cur)
!= right_node_ptr) {
node_pointer_corrupted:
err = DB_CORRUPTION;
fputs("InnoDB: node pointer to"
" the right page is wrong\n",
stderr);
btr_validate_report1(index, level,
block);
}
} else {
page_t* right_father_page
= btr_cur_get_page(&right_node_cur);
if (btr_cur_get_rec(&right_node_cur)
!= page_rec_get_next(
page_get_infimum_rec(
right_father_page))) {
err = DB_CORRUPTION;
fputs("InnoDB: node pointer 2 to"
" the right page is wrong\n",
stderr);
btr_validate_report1(index, level,
block);
}
if (page_get_page_no(right_father_page)
!= btr_page_get_next(father_page)) {
err = DB_CORRUPTION;
fputs("InnoDB: node pointer 3 to"
" the right page is wrong\n",
stderr);
btr_validate_report1(index, level,
block);
}
}
} else {
err = DB_CORRUPTION;
}
}
node_ptr_fails:
/* Commit the mini-transaction to release the latch on 'page'.
Re-acquire the latch on right_page, which will become 'page'
on the next loop. The page has already been checked. */
mtr.commit();
if (trx_is_interrupted(trx)) {
/* On interrupt, return the current status. */
} else if (right_page_no != FIL_NULL) {
mtr.start();
if (!lockout) {
if (rightmost_child) {
if (parent_right_page_no != FIL_NULL) {
btr_block_get(*index,
parent_right_page_no,
RW_SX_LATCH, false,
&mtr);
}
} else if (parent_page_no != FIL_NULL) {
btr_block_get(*index, parent_page_no,
RW_SX_LATCH, false, &mtr);
}
}
block = btr_block_get(*index, right_page_no, RW_SX_LATCH,
!level, &mtr, &err);
goto loop;
}
goto func_exit;
}
/**************************************************************//**
Checks the consistency of an index tree.
@return DB_SUCCESS if ok, error code if not */
dberr_t
btr_validate_index(
/*===============*/
dict_index_t* index, /*!< in: index */
const trx_t* trx) /*!< in: transaction or NULL */
{
/* Full Text index are implemented by auxiliary tables, not the B-tree */
if (index->online_status != ONLINE_INDEX_COMPLETE ||
(index->type & (DICT_FTS | DICT_CORRUPT)))
return DB_SUCCESS;
const bool lockout= index->is_spatial();
mtr_t mtr;
mtr.start();
if (lockout)
mtr_x_lock_index(index, &mtr);
else
mtr_sx_lock_index(index, &mtr);
dberr_t err;
if (page_t *root= btr_root_get(index, &mtr, &err))
for (auto level= btr_page_get_level(root);; level--)
{
if (dberr_t err_level= btr_validate_level(index, trx, level, lockout))
err= err_level;
if (!level)
break;
}
mtr.commit();
return err;
}
/**************************************************************//**
Checks if the page in the cursor can be merged with given page.
If necessary, re-organize the merge_page.
@return true if possible to merge. */
static
bool
btr_can_merge_with_page(
/*====================*/
btr_cur_t* cursor, /*!< in: cursor on the page to merge */
uint32_t page_no, /*!< in: a sibling page */
buf_block_t** merge_block, /*!< out: the merge block */
mtr_t* mtr) /*!< in: mini-transaction */
{
dict_index_t* index;
page_t* page;
ulint n_recs;
ulint data_size;
ulint max_ins_size_reorg;
ulint max_ins_size;
buf_block_t* mblock;
page_t* mpage;
DBUG_ENTER("btr_can_merge_with_page");
if (page_no == FIL_NULL) {
error:
*merge_block = NULL;
DBUG_RETURN(false);
}
index = btr_cur_get_index(cursor);
page = btr_cur_get_page(cursor);
mblock = btr_block_get(*index, page_no, RW_X_LATCH, page_is_leaf(page),
mtr);
if (!mblock) {
goto error;
}
mpage = buf_block_get_frame(mblock);
n_recs = page_get_n_recs(page);
data_size = page_get_data_size(page);
max_ins_size_reorg = page_get_max_insert_size_after_reorganize(
mpage, n_recs);
if (data_size > max_ins_size_reorg) {
goto error;
}
/* If compression padding tells us that merging will result in
too packed up page i.e.: which is likely to cause compression
failure then don't merge the pages. */
if (mblock->page.zip.data && page_is_leaf(mpage)
&& (page_get_data_size(mpage) + data_size
>= dict_index_zip_pad_optimal_page_size(index))) {
goto error;
}
max_ins_size = page_get_max_insert_size(mpage, n_recs);
if (data_size > max_ins_size) {
/* We have to reorganize mpage */
if (btr_page_reorganize_block(page_zip_level, mblock, index,
mtr) != DB_SUCCESS) {
goto error;
}
max_ins_size = page_get_max_insert_size(mpage, n_recs);
ut_ad(page_validate(mpage, index));
ut_ad(max_ins_size == max_ins_size_reorg);
if (data_size > max_ins_size) {
/* Add fault tolerance, though this should
never happen */
goto error;
}
}
*merge_block = mblock;
DBUG_RETURN(true);
}
|