summaryrefslogtreecommitdiff
path: root/innobase/btr/btr0btr.c
blob: c27fb73ff8dd2e8c7b7077ca9d436dfa2f0d8fa8 (plain)
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
/******************************************************
The B-tree

(c) 1994-1996 Innobase Oy

Created 6/2/1994 Heikki Tuuri
*******************************************************/
 
#include "btr0btr.h"

#ifdef UNIV_NONINL
#include "btr0btr.ic"
#endif

#include "fsp0fsp.h"
#include "page0page.h"
#include "btr0cur.h"
#include "btr0sea.h"
#include "btr0pcur.h"
#include "rem0cmp.h"
#include "lock0lock.h"
#include "ibuf0ibuf.h"
#include "trx0trx.h"

/*
Latching strategy of the InnoDB B-tree
--------------------------------------
A tree latch protects all non-leaf nodes of the tree. Each node of a tree
also has a latch of its own.

A B-tree operation normally first acquires an S-latch on the tree. It
searches down the tree and releases the tree latch when it has the
leaf node latch. To save CPU time we do not acquire any latch on
non-leaf nodes of the tree during a search, those pages are only bufferfixed.

If an operation needs to restructure the tree, it acquires an X-latch on
the tree before searching to a leaf node. If it needs, for example, to
split a leaf,
(1) InnoDB decides the split point in the leaf,
(2) allocates a new page,
(3) inserts the appropriate node pointer to the first non-leaf level,
(4) releases the tree X-latch,
(5) and then moves records from the leaf to the new allocated page.

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.
*/

/******************************************************************
Creates a new index page to the tree (not the root, and also not
used in page reorganization). */
static
void
btr_page_create(
/*============*/
	page_t*		page,	/* in: page to be created */
	dict_tree_t*	tree,	/* in: index tree */
	mtr_t*		mtr);	/* in: mtr */
/****************************************************************
Returns the upper level node pointer to a page. It is assumed that
mtr holds an x-latch on the tree. */
static
rec_t*
btr_page_get_father_node_ptr(
/*=========================*/
				/* out: pointer to node pointer record */
	dict_tree_t*	tree,	/* in: index tree */
	page_t*		page,	/* in: page: must contain at least one
				user record */
	mtr_t*		mtr);	/* in: mtr */
/*****************************************************************
Empties an index page. */
static
void
btr_page_empty(
/*===========*/
	page_t*	page,	/* in: page to be emptied */
	mtr_t*	mtr);	/* in: mtr */
/*****************************************************************
Returns TRUE if the insert fits on the appropriate half-page
with the chosen split_rec. */
static
ibool
btr_page_insert_fits(
/*=================*/
					/* out: TRUE if fits */
	btr_cur_t*	cursor,		/* in: cursor at which insert
					should be made */
	rec_t*		split_rec,	/* in: suggestion for first record
					on upper half-page, or NULL if
					tuple should be first */
	const ulint*	offsets,	/* in: rec_get_offsets(
					split_rec, cursor->index) */
	dtuple_t*	tuple,		/* in: tuple to insert */
	mem_heap_t*	heap);		/* in: temporary memory heap */

/******************************************************************
Gets the root node of a tree and x-latches it. */

page_t*
btr_root_get(
/*=========*/
				/* out: root page, x-latched */
	dict_tree_t*	tree,	/* in: index tree */
	mtr_t*		mtr)	/* in: mtr */
{
	ulint	space;
	ulint	root_page_no;
	page_t*	root;
	
	space = dict_tree_get_space(tree);
	root_page_no = dict_tree_get_page(tree);

	root = btr_page_get(space, root_page_no, RW_X_LATCH, mtr);
	ut_a((ibool)!!page_is_comp(root) ==
			UT_LIST_GET_FIRST(tree->tree_indexes)->table->comp);
	
	return(root);
}

/*****************************************************************
Gets pointer to the previous user record in the tree. It is assumed that
the caller has appropriate latches on the page and its neighbor. */

rec_t*
btr_get_prev_user_rec(
/*==================*/
			/* out: previous user record, NULL if there is none */
	rec_t*	rec,	/* in: record on leaf level */
	mtr_t*	mtr)	/* in: mtr holding a latch on the page, and if
			needed, also to the previous page */
{
	page_t*	page;
	page_t*	prev_page;
	ulint	prev_page_no;
	ulint	space;

	if (!page_rec_is_infimum(rec)) {

		rec_t*	prev_rec = page_rec_get_prev(rec);

		if (!page_rec_is_infimum(prev_rec)) {

			return(prev_rec);
		}
	}
	
	page = buf_frame_align(rec);
	prev_page_no = btr_page_get_prev(page, mtr);
	space = buf_frame_get_space_id(page);
	
	if (prev_page_no != FIL_NULL) {

		prev_page = buf_page_get_with_no_latch(space, prev_page_no,
									mtr);
		/* The caller must already have a latch to the brother */
		ut_ad((mtr_memo_contains(mtr, buf_block_align(prev_page),
		      				MTR_MEMO_PAGE_S_FIX))
		      || (mtr_memo_contains(mtr, buf_block_align(prev_page),
		      				MTR_MEMO_PAGE_X_FIX)));
		ut_a(page_is_comp(prev_page) == page_is_comp(page));

		return(page_rec_get_prev(page_get_supremum_rec(prev_page)));
	}

	return(NULL);
}

/*****************************************************************
Gets pointer to the next user record in the tree. It is assumed that the
caller has appropriate latches on the page and its neighbor. */

rec_t*
btr_get_next_user_rec(
/*==================*/
			/* out: next user record, NULL if there is none */
	rec_t*	rec,	/* in: record on leaf level */
	mtr_t*	mtr)	/* in: mtr holding a latch on the page, and if
			needed, also to the next page */
{
	page_t*	page;
	page_t*	next_page;
	ulint	next_page_no;
	ulint	space;

	if (!page_rec_is_supremum(rec)) {

		rec_t*	next_rec = page_rec_get_next(rec);

		if (!page_rec_is_supremum(next_rec)) {

			return(next_rec);
		}
	}
	
	page = buf_frame_align(rec);
	next_page_no = btr_page_get_next(page, mtr);
	space = buf_frame_get_space_id(page);
	
	if (next_page_no != FIL_NULL) {

		next_page = buf_page_get_with_no_latch(space, next_page_no,
									mtr);
		/* The caller must already have a latch to the brother */
		ut_ad((mtr_memo_contains(mtr, buf_block_align(next_page),
		      				MTR_MEMO_PAGE_S_FIX))
		      || (mtr_memo_contains(mtr, buf_block_align(next_page),
		      				MTR_MEMO_PAGE_X_FIX)));

		ut_a(page_is_comp(next_page) == page_is_comp(page));
		return(page_rec_get_next(page_get_infimum_rec(next_page)));
	}

	return(NULL);
}

/******************************************************************
Creates a new index page to the tree (not the root, and also not used in
page reorganization). */
static
void
btr_page_create(
/*============*/
	page_t*		page,	/* in: page to be created */
	dict_tree_t*	tree,	/* in: index tree */
	mtr_t*		mtr)	/* in: mtr */
{
	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
			      				MTR_MEMO_PAGE_X_FIX));
	page_create(page, mtr,
			UT_LIST_GET_FIRST(tree->tree_indexes)->table->comp);
	buf_block_align(page)->check_index_page_at_flush = TRUE;
	
	btr_page_set_index_id(page, tree->id, mtr);
}

/******************************************************************
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! */
static
page_t*
btr_page_alloc_for_ibuf(
/*====================*/
				/* out: new allocated page, x-latched */
	dict_tree_t*	tree,	/* in: index tree */
	mtr_t*		mtr)	/* in: mtr */
{
	fil_addr_t	node_addr;
	page_t*		root;
	page_t*		new_page;

	root = btr_root_get(tree, mtr);
	
	node_addr = flst_get_first(root + PAGE_HEADER
					+ PAGE_BTR_IBUF_FREE_LIST, mtr);
	ut_a(node_addr.page != FIL_NULL);

	new_page = buf_page_get(dict_tree_get_space(tree), node_addr.page,
							RW_X_LATCH, mtr);
#ifdef UNIV_SYNC_DEBUG
	buf_page_dbg_add_level(new_page, SYNC_TREE_NODE_NEW);
#endif /* UNIV_SYNC_DEBUG */

	flst_remove(root + PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST,
		    new_page + PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST_NODE,
									mtr);
	ut_ad(flst_validate(root + PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST, mtr));

	return(new_page);
}

/******************************************************************
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! */

page_t*
btr_page_alloc(
/*===========*/
					/* out: new allocated page, x-latched;
					NULL if out of space */
	dict_tree_t*	tree,		/* in: index tree */
	ulint		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: mtr */
{
	fseg_header_t*	seg_header;
	page_t*		root;
	page_t*		new_page;
	ulint		new_page_no;

	if (tree->type & DICT_IBUF) {

		return(btr_page_alloc_for_ibuf(tree, mtr));
	}

	root = btr_root_get(tree, mtr);
		
	if (level == 0) {
		seg_header = root + PAGE_HEADER + PAGE_BTR_SEG_LEAF;
	} else {
		seg_header = root + PAGE_HEADER + PAGE_BTR_SEG_TOP;
	}

	/* Parameter TRUE below states that the caller has made the
	reservation for free extents, and thus we know that a page can
	be allocated: */
	
	new_page_no = fseg_alloc_free_page_general(seg_header, hint_page_no,
						file_direction, TRUE, mtr);
	if (new_page_no == FIL_NULL) {

		return(NULL);
	}

	new_page = buf_page_get(dict_tree_get_space(tree), new_page_no,
							RW_X_LATCH, mtr);
#ifdef UNIV_SYNC_DEBUG
	buf_page_dbg_add_level(new_page, SYNC_TREE_NODE_NEW);
#endif /* UNIV_SYNC_DEBUG */
							
	return(new_page);
}	

/******************************************************************
Gets the number of pages in a B-tree. */

ulint
btr_get_size(
/*=========*/
				/* out: number of pages */
	dict_index_t*	index,	/* in: index */
	ulint		flag)	/* in: BTR_N_LEAF_PAGES or BTR_TOTAL_SIZE */
{
	fseg_header_t*	seg_header;
	page_t*		root;
	ulint		n;
	ulint		dummy;
	mtr_t		mtr;

	mtr_start(&mtr);

	mtr_s_lock(dict_tree_get_lock(index->tree), &mtr);

	root = btr_root_get(index->tree, &mtr);
		
	if (flag == BTR_N_LEAF_PAGES) {
		seg_header = root + PAGE_HEADER + PAGE_BTR_SEG_LEAF;
		
		fseg_n_reserved_pages(seg_header, &n, &mtr);
		
	} else if (flag == BTR_TOTAL_SIZE) {
		seg_header = root + PAGE_HEADER + PAGE_BTR_SEG_TOP;

		n = fseg_n_reserved_pages(seg_header, &dummy, &mtr);
		
		seg_header = root + PAGE_HEADER + PAGE_BTR_SEG_LEAF;
		
		n += fseg_n_reserved_pages(seg_header, &dummy, &mtr);		
	} else {
		ut_error;
	}

	mtr_commit(&mtr);

	return(n);
}	

/******************************************************************
Frees a page used in an ibuf tree. Puts the page to the free list of the
ibuf tree. */
static
void
btr_page_free_for_ibuf(
/*===================*/
	dict_tree_t*	tree,	/* in: index tree */
	page_t*		page,	/* in: page to be freed, x-latched */	
	mtr_t*		mtr)	/* in: mtr */
{
	page_t*		root;

	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
			      				MTR_MEMO_PAGE_X_FIX));
	root = btr_root_get(tree, mtr);
	
	flst_add_first(root + PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST,
		       page + PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST_NODE, mtr);

	ut_ad(flst_validate(root + PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST,
									mtr));
}

/******************************************************************
Frees a file page used in an index tree. Can be used also to (BLOB)
external storage pages, because the page level 0 can be given as an
argument. */

void
btr_page_free_low(
/*==============*/
	dict_tree_t*	tree,	/* in: index tree */
	page_t*		page,	/* in: page to be freed, x-latched */	
	ulint		level,	/* in: page level */
	mtr_t*		mtr)	/* in: mtr */
{
	fseg_header_t*	seg_header;
	page_t*		root;
	ulint		space;
	ulint		page_no;

	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
			      				MTR_MEMO_PAGE_X_FIX));
	/* The page gets invalid for optimistic searches: increment the frame
	modify clock */

	buf_frame_modify_clock_inc(page);
	
	if (tree->type & DICT_IBUF) {

		btr_page_free_for_ibuf(tree, page, mtr);

		return;
	}

	root = btr_root_get(tree, mtr);
	
	if (level == 0) {
		seg_header = root + PAGE_HEADER + PAGE_BTR_SEG_LEAF;
	} else {
		seg_header = root + PAGE_HEADER + PAGE_BTR_SEG_TOP;
	}

	space = buf_frame_get_space_id(page);
	page_no = buf_frame_get_page_no(page);
	
	fseg_free_page(seg_header, space, page_no, mtr);
}	

/******************************************************************
Frees a file page used in an index tree. NOTE: cannot free field external
storage pages because the page must contain info on its level. */

void
btr_page_free(
/*==========*/
	dict_tree_t*	tree,	/* in: index tree */
	page_t*		page,	/* in: page to be freed, x-latched */	
	mtr_t*		mtr)	/* in: mtr */
{
	ulint		level;

	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
			      				MTR_MEMO_PAGE_X_FIX));
	level = btr_page_get_level(page, mtr);
	
	btr_page_free_low(tree, page, level, mtr);
}	

/******************************************************************
Sets the child node file address in a node pointer. */
UNIV_INLINE
void
btr_node_ptr_set_child_page_no(
/*===========================*/
	rec_t*		rec,	/* in: node pointer record */
	const ulint*	offsets,/* in: array returned by rec_get_offsets() */
	ulint		page_no,/* in: child node address */
	mtr_t*		mtr)	/* in: mtr */
{
	byte*	field;
	ulint	len;

	ut_ad(rec_offs_validate(rec, NULL, offsets));
	ut_ad(0 < btr_page_get_level(buf_frame_align(rec), mtr));
	ut_ad(!rec_offs_comp(offsets) || rec_get_node_ptr_flag(rec));

	/* The child address is in the last field */	
	field = rec_get_nth_field(rec, offsets,
					rec_offs_n_fields(offsets) - 1, &len);

	ut_ad(len == 4);
	
	mlog_write_ulint(field, page_no, MLOG_4BYTES, mtr);
}

/****************************************************************
Returns the child page of a node pointer and x-latches it. */
static
page_t*
btr_node_ptr_get_child(
/*===================*/
				/* out: child page, x-latched */
	rec_t*		node_ptr,/* in: node pointer */
	const ulint*	offsets,/* in: array returned by rec_get_offsets() */
	mtr_t*		mtr)	/* in: mtr */
{
	ulint	page_no;
	ulint	space;
	page_t*	page;

	ut_ad(rec_offs_validate(node_ptr, NULL, offsets));
	space = buf_frame_get_space_id(node_ptr);
	page_no = btr_node_ptr_get_child_page_no(node_ptr, offsets);

	page = btr_page_get(space, page_no, RW_X_LATCH, mtr);
	
	return(page);
}

/****************************************************************
Returns the upper level node pointer to a page. It is assumed that mtr holds
an x-latch on the tree. */
static
rec_t*
btr_page_get_father_for_rec(
/*========================*/
				/* out: pointer to node pointer record,
				its page x-latched */
	dict_tree_t*	tree,	/* in: index tree */
	page_t*		page,	/* in: page: must contain at least one
				user record */
	rec_t*		user_rec,/* in: user_record on page */
	mtr_t*		mtr)	/* in: mtr */
{
	mem_heap_t*	heap;
	dtuple_t*	tuple;
	btr_cur_t	cursor;
	rec_t*		node_ptr;
	dict_index_t*	index;
	ulint		offsets_[REC_OFFS_NORMAL_SIZE];
	ulint*		offsets	= offsets_;
	*offsets_ = (sizeof offsets_) / sizeof *offsets_;

	ut_ad(mtr_memo_contains(mtr, dict_tree_get_lock(tree),
							MTR_MEMO_X_LOCK));
	ut_a(page_rec_is_user_rec(user_rec));
	
	ut_ad(dict_tree_get_page(tree) != buf_frame_get_page_no(page));

	heap = mem_heap_create(100);

	tuple = dict_tree_build_node_ptr(tree, user_rec, 0, heap,
					 btr_page_get_level(page, mtr));
	index = UT_LIST_GET_FIRST(tree->tree_indexes);

	/* In the following, we choose just any index from the tree as the
	first parameter for btr_cur_search_to_nth_level. */

	btr_cur_search_to_nth_level(index,
				btr_page_get_level(page, mtr) + 1,
				tuple, PAGE_CUR_LE,
				BTR_CONT_MODIFY_TREE, &cursor, 0, mtr);

	node_ptr = btr_cur_get_rec(&cursor);
	offsets = rec_get_offsets(node_ptr, index, offsets,
						ULINT_UNDEFINED, &heap);

	if (btr_node_ptr_get_child_page_no(node_ptr, offsets) !=
                                                buf_frame_get_page_no(page)) {
		rec_t*	print_rec;
		fputs("InnoDB: Dump of the child page:\n", stderr);
		buf_page_print(buf_frame_align(page));
		fputs("InnoDB: Dump of the parent page:\n", stderr);
		buf_page_print(buf_frame_align(node_ptr));

		fputs("InnoDB: Corruption of an index tree: table ", stderr);
		ut_print_name(stderr, NULL, index->table_name);
		fputs(", index ", stderr);
		ut_print_name(stderr, NULL, index->name);
		fprintf(stderr, ",\n"
"InnoDB: father ptr page no %lu, child page no %lu\n",
			(ulong)
			btr_node_ptr_get_child_page_no(node_ptr, offsets),
			(ulong) buf_frame_get_page_no(page));
		print_rec = page_rec_get_next(page_get_infimum_rec(page));
		offsets = rec_get_offsets(print_rec, index,
				offsets, ULINT_UNDEFINED, &heap);
		page_rec_print(print_rec, offsets);
		offsets = rec_get_offsets(node_ptr, index, offsets,
					ULINT_UNDEFINED, &heap);
		page_rec_print(node_ptr, offsets);

		fputs(
"InnoDB: You should dump + drop + reimport the table to fix the\n"
"InnoDB: corruption. If the crash happens at the database startup, see\n"
"InnoDB: http://dev.mysql.com/doc/mysql/en/Forcing_recovery.html about\n"
"InnoDB: forcing recovery. Then dump + drop + reimport.\n", stderr);
	}

	ut_a(btr_node_ptr_get_child_page_no(node_ptr, offsets) ==
						buf_frame_get_page_no(page));
	mem_heap_free(heap);

	return(node_ptr);
}

/****************************************************************
Returns the upper level node pointer to a page. It is assumed that
mtr holds an x-latch on the tree. */
static
rec_t*
btr_page_get_father_node_ptr(
/*=========================*/
				/* out: pointer to node pointer record */
	dict_tree_t*	tree,	/* in: index tree */
	page_t*		page,	/* in: page: must contain at least one
				user record */
	mtr_t*		mtr)	/* in: mtr */
{
	return(btr_page_get_father_for_rec(tree, page,
		page_rec_get_next(page_get_infimum_rec(page)), mtr));
}

/****************************************************************
Creates the root node for a new index tree. */

ulint
btr_create(
/*=======*/
			/* out: page number of the created root, FIL_NULL if
			did not succeed */
	ulint	type,	/* in: type of the index */
	ulint	space,	/* in: space where created */
	dulint	index_id,/* in: index id */
	ulint	comp,	/* in: nonzero=compact page format */
	mtr_t*	mtr)	/* in: mini-transaction handle */
{
	ulint		page_no;
	buf_frame_t*	ibuf_hdr_frame;
	buf_frame_t*	frame;
	page_t*		page;

	/* 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 (type & DICT_IBUF) {
		/* Allocate first the ibuf header page */
		ibuf_hdr_frame = fseg_create(space, 0,
				IBUF_HEADER + IBUF_TREE_SEG_HEADER, mtr);

#ifdef UNIV_SYNC_DEBUG
		buf_page_dbg_add_level(ibuf_hdr_frame, SYNC_TREE_NODE_NEW);
#endif /* UNIV_SYNC_DEBUG */
		ut_ad(buf_frame_get_page_no(ibuf_hdr_frame)
 						== IBUF_HEADER_PAGE_NO);
		/* Allocate then the next page to the segment: it will be the
 		tree root page */

 		page_no = fseg_alloc_free_page(
				ibuf_hdr_frame + IBUF_HEADER
 				+ IBUF_TREE_SEG_HEADER, IBUF_TREE_ROOT_PAGE_NO,
				FSP_UP, mtr);
		ut_ad(page_no == IBUF_TREE_ROOT_PAGE_NO);

		frame = buf_page_get(space, page_no, RW_X_LATCH, mtr);
	} else {
		frame = fseg_create(space, 0, PAGE_HEADER + PAGE_BTR_SEG_TOP,
									mtr);
	}
	
	if (frame == NULL) {

		return(FIL_NULL);
	}

	page_no = buf_frame_get_page_no(frame);
	
#ifdef UNIV_SYNC_DEBUG
	buf_page_dbg_add_level(frame, SYNC_TREE_NODE_NEW);
#endif /* UNIV_SYNC_DEBUG */

	if (type & DICT_IBUF) {
		/* It is an insert buffer tree: initialize the free list */

		ut_ad(page_no == IBUF_TREE_ROOT_PAGE_NO);
		
		flst_init(frame + PAGE_HEADER + PAGE_BTR_IBUF_FREE_LIST, mtr);
	} else {	
		/* It is a non-ibuf tree: create a file segment for leaf
		pages */
		fseg_create(space, page_no, PAGE_HEADER + PAGE_BTR_SEG_LEAF,
									mtr);
		/* The fseg create acquires a second latch on the page,
		therefore we must declare it: */
#ifdef UNIV_SYNC_DEBUG
		buf_page_dbg_add_level(frame, SYNC_TREE_NODE_NEW);
#endif /* UNIV_SYNC_DEBUG */
	}
	
	/* Create a new index page on the the allocated segment page */
	page = page_create(frame, mtr, comp);
	buf_block_align(page)->check_index_page_at_flush = TRUE;

	/* Set the index id of the page */
	btr_page_set_index_id(page, index_id, mtr);

	/* Set the level of the new index page */
	btr_page_set_level(page, 0, mtr);
	
	/* Set the next node and previous node fields */
	btr_page_set_next(page, FIL_NULL, mtr);
	btr_page_set_prev(page, FIL_NULL, mtr);

	/* We reset the free bits for the page 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 */
	
	ibuf_reset_free_bits_with_type(type, page);

	/* 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(page, 2) > 2 * BTR_PAGE_MAX_REC_SIZE);

	return(page_no);
}

/****************************************************************
Frees a B-tree except the root page, which MUST be freed after this
by calling btr_free_root. */

void
btr_free_but_not_root(
/*==================*/
	ulint	space,		/* in: space where created */
	ulint	root_page_no)	/* in: root page number */
{
	ibool	finished;
	page_t*	root;
	mtr_t	mtr;

leaf_loop:	
	mtr_start(&mtr);
	
	root = btr_page_get(space, root_page_no, RW_X_LATCH, &mtr);	

	/* NOTE: page hash indexes are dropped when a page is freed inside
	fsp0fsp. */

	finished = fseg_free_step(
				root + PAGE_HEADER + PAGE_BTR_SEG_LEAF, &mtr);
	mtr_commit(&mtr);

	if (!finished) {

		goto leaf_loop;
	}
top_loop:
	mtr_start(&mtr);
	
	root = btr_page_get(space, root_page_no, RW_X_LATCH, &mtr);	

	finished = fseg_free_step_not_header(
				root + PAGE_HEADER + PAGE_BTR_SEG_TOP, &mtr);
	mtr_commit(&mtr);

	if (!finished) {

		goto top_loop;
	}	
}

/****************************************************************
Frees the B-tree root page. Other tree MUST already have been freed. */

void
btr_free_root(
/*==========*/
	ulint	space,		/* in: space where created */
	ulint	root_page_no,	/* in: root page number */
	mtr_t*	mtr)		/* in: a mini-transaction which has already
				been started */
{
	ibool	finished;
	page_t*	root;

	root = btr_page_get(space, root_page_no, RW_X_LATCH, mtr);

	btr_search_drop_page_hash_index(root);	
top_loop:	
	finished = fseg_free_step(
				root + PAGE_HEADER + PAGE_BTR_SEG_TOP, mtr);
	if (!finished) {

		goto top_loop;
	}	
}

/*****************************************************************
Reorganizes an index page. */
static
void
btr_page_reorganize_low(
/*====================*/
	ibool		recovery,/* in: TRUE if called in recovery:
				locks should not be updated, i.e.,
				there cannot exist locks on the
				page, and a hash index should not be
				dropped: it cannot exist */
	page_t*		page,	/* in: page to be reorganized */
	dict_index_t*	index,	/* in: record descriptor */
	mtr_t*		mtr)	/* in: mtr */
{
	page_t*	new_page;
	ulint	log_mode;
	ulint	data_size1;
	ulint	data_size2;
	ulint	max_ins_size1;
	ulint	max_ins_size2;

	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
			      				MTR_MEMO_PAGE_X_FIX));
	ut_ad(!!page_is_comp(page) == index->table->comp);
	data_size1 = page_get_data_size(page);
	max_ins_size1 = page_get_max_insert_size_after_reorganize(page, 1);

	/* Write the log record */
	mlog_open_and_write_index(mtr, page, index, page_is_comp(page)
			? MLOG_COMP_PAGE_REORGANIZE
			: MLOG_PAGE_REORGANIZE, 0);

	/* Turn logging off */
	log_mode = mtr_set_log_mode(mtr, MTR_LOG_NONE);

	new_page = buf_frame_alloc();

	/* Copy the old page to temporary space */
	buf_frame_copy(new_page, page);

	if (!recovery) {
		btr_search_drop_page_hash_index(page);
	}

	/* Recreate the page: note that global data on page (possible
	segment headers, next page-field, etc.) is preserved intact */

	page_create(page, mtr, page_is_comp(page));
	buf_block_align(page)->check_index_page_at_flush = TRUE;
	
	/* Copy the records from the temporary space to the recreated page;
	do not copy the lock bits yet */

	page_copy_rec_list_end_no_locks(page, new_page,
				page_get_infimum_rec(new_page), index, mtr);
	/* Copy max trx id to recreated page */
	page_set_max_trx_id(page, page_get_max_trx_id(new_page));
	
	if (!recovery) {
		/* Update the record lock bitmaps */
		lock_move_reorganize_page(page, new_page);
	}

	data_size2 = page_get_data_size(page);
	max_ins_size2 = page_get_max_insert_size_after_reorganize(page, 1);

	if (data_size1 != data_size2 || max_ins_size1 != max_ins_size2) {
		buf_page_print(page);
		buf_page_print(new_page);
	        fprintf(stderr,
"InnoDB: Error: page old data size %lu new data size %lu\n"
"InnoDB: Error: page old max ins size %lu new max ins size %lu\n"
"InnoDB: Submit a detailed bug report to http://bugs.mysql.com\n",
			(unsigned long) data_size1, (unsigned long) data_size2,
			(unsigned long) max_ins_size1,
			(unsigned long) max_ins_size2);
	}

	buf_frame_free(new_page);

	/* Restore logging mode */
	mtr_set_log_mode(mtr, log_mode);
}

/*****************************************************************
Reorganizes an index page. */

void
btr_page_reorganize(
/*================*/
	page_t*		page,	/* in: page to be reorganized */
	dict_index_t*	index,	/* in: record descriptor */
	mtr_t*		mtr)	/* in: mtr */
{
	btr_page_reorganize_low(FALSE, page, index, mtr);
}

/***************************************************************
Parses a redo log record of reorganizing a page. */

byte*
btr_parse_page_reorganize(
/*======================*/
				/* out: end of log record or NULL */
	byte*		ptr,	/* in: buffer */
	byte*		end_ptr __attribute__((unused)),
				/* in: buffer end */
	dict_index_t*	index,	/* in: record descriptor */
	page_t*		page,	/* in: page or NULL */
	mtr_t*		mtr)	/* in: mtr or NULL */
{
	ut_ad(ptr && end_ptr);

	/* The record is empty, except for the record initial part */

	if (page) {
		btr_page_reorganize_low(TRUE, page, index, mtr);
	}

	return(ptr);
}

/*****************************************************************
Empties an index page. */
static
void
btr_page_empty(
/*===========*/
	page_t*	page,	/* in: page to be emptied */
	mtr_t*	mtr)	/* in: mtr */
{
	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
			      				MTR_MEMO_PAGE_X_FIX));
	btr_search_drop_page_hash_index(page);

	/* Recreate the page: note that global data on page (possible
	segment headers, next page-field, etc.) is preserved intact */

	page_create(page, mtr, page_is_comp(page));
	buf_block_align(page)->check_index_page_at_flush = TRUE;
}

/*****************************************************************
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. */

rec_t*
btr_root_raise_and_insert(
/*======================*/
				/* out: inserted record */
	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 */
	dtuple_t*	tuple,	/* in: tuple to insert */
	mtr_t*		mtr)	/* in: mtr */
{
	dict_tree_t*	tree;
	page_t*		root;
	page_t*		new_page;
	ulint		new_page_no;
	rec_t*		rec;
	mem_heap_t*	heap;
	dtuple_t*	node_ptr;
	ulint		level;	
	rec_t*		node_ptr_rec;
	page_cur_t*	page_cursor;
	
	root = btr_cur_get_page(cursor);
	tree = btr_cur_get_tree(cursor);

	ut_ad(dict_tree_get_page(tree) == buf_frame_get_page_no(root));
	ut_ad(mtr_memo_contains(mtr, dict_tree_get_lock(tree),
							MTR_MEMO_X_LOCK));
	ut_ad(mtr_memo_contains(mtr, buf_block_align(root),
			      				MTR_MEMO_PAGE_X_FIX));
	btr_search_drop_page_hash_index(root);

	/* 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. */
	
	new_page = btr_page_alloc(tree, 0, FSP_NO_DIR,
				  btr_page_get_level(root, mtr), mtr);

	btr_page_create(new_page, tree, mtr);

	level = btr_page_get_level(root, mtr);
	
	/* Set the levels of the new index page and root page */
	btr_page_set_level(new_page, level, mtr);
	btr_page_set_level(root, level + 1, mtr);
	
	/* Set the next node and previous node fields of new page */
	btr_page_set_next(new_page, FIL_NULL, mtr);
	btr_page_set_prev(new_page, FIL_NULL, mtr);

	/* Move the records from root to the new page */

	page_move_rec_list_end(new_page, root, page_get_infimum_rec(root),
							cursor->index, mtr);
	/* 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 */
	
	lock_update_root_raise(new_page, root);

	/* Create a memory heap where the node pointer is stored */
	heap = mem_heap_create(100);

	rec = page_rec_get_next(page_get_infimum_rec(new_page));
	new_page_no = buf_frame_get_page_no(new_page);
	
	/* Build the node pointer (= node key and page address) for the
	child */

	node_ptr = dict_tree_build_node_ptr(tree, rec, new_page_no, heap,
					                          level);
	/* Reorganize the root to get free space */
	btr_page_reorganize(root, cursor->index, mtr);

	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,
						cursor->index, mtr);

	ut_ad(node_ptr_rec);

	/* 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: */

	btr_set_min_rec_mark(node_ptr_rec, page_is_comp(root), mtr);
		
	/* Free the memory heap */
	mem_heap_free(heap);

	/* We play safe and reset the free bits for the new page */

/*	fprintf(stderr, "Root raise new page no %lu\n",
					buf_frame_get_page_no(new_page)); */

	ibuf_reset_free_bits(UT_LIST_GET_FIRST(tree->tree_indexes),
								new_page);
	/* Reposition the cursor to the child node */
	page_cur_search(new_page, cursor->index, tuple,
				PAGE_CUR_LE, page_cursor);
	
	/* Split the child and insert tuple */
	return(btr_page_split_and_insert(cursor, tuple, mtr));
}	

/*****************************************************************
Decides if the page should be split at the convergence point of inserts
converging to the left. */

ibool
btr_page_get_split_rec_to_left(
/*===========================*/
				/* out: TRUE if split recommended */
	btr_cur_t*	cursor,	/* in: cursor at which to insert */
	rec_t**		split_rec) /* out: if split recommended,
				the first record on upper half page,
				or NULL if tuple to be inserted should
				be first */
{
	page_t*	page;
	rec_t*	insert_point;
	rec_t*	infimum;

	page = btr_cur_get_page(cursor);
	insert_point = btr_cur_get_rec(cursor);

	if (page_header_get_ptr(page, PAGE_LAST_INSERT)
	    == page_rec_get_next(insert_point)) {

	     	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 (infimum != insert_point
		    && page_rec_get_next(infimum) != insert_point) {

			*split_rec = insert_point;
		} else {
	     		*split_rec = page_rec_get_next(insert_point);
	     	}

		return(TRUE);
	}

	return(FALSE);
}

/*****************************************************************
Decides if the page should be split at the convergence point of inserts
converging to the right. */

ibool
btr_page_get_split_rec_to_right(
/*============================*/
				/* out: TRUE if split recommended */
	btr_cur_t*	cursor,	/* in: cursor at which to insert */
	rec_t**		split_rec) /* out: if split recommended,
				the first record on upper half page,
				or NULL if tuple to be inserted should
				be first */
{
	page_t*	page;
	rec_t*	insert_point;

	page = btr_cur_get_page(cursor);
	insert_point = btr_cur_get_rec(cursor);

	/* 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 (UNIV_LIKELY(page_header_get_ptr(page, PAGE_LAST_INSERT)
				== insert_point)) {

		rec_t*	next_rec;

		next_rec = page_rec_get_next(insert_point);

		if (page_rec_is_supremum(next_rec)) {
split_at_new:
			/* Split at the new record to insert */
	     		*split_rec = NULL;
		} else {
			rec_t*	next_next_rec = page_rec_get_next(next_rec);
			if (page_rec_is_supremum(next_next_rec)) {

				goto split_at_new;
			}

			/* 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 = next_next_rec;
		}

		return(TRUE);
	}

	return(FALSE);
}

/*****************************************************************
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. */
static
rec_t*
btr_page_get_sure_split_rec(
/*========================*/
					/* out: split record, or NULL if
					tuple will be the first record on
					upper half-page */
	btr_cur_t*	cursor,		/* in: cursor at which insert
					should be made */
	dtuple_t*	tuple)		/* in: tuple to insert */	
{
	page_t*	page;
	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;
	ulint*	offsets;

	page = btr_cur_get_page(cursor);
	
	insert_size = rec_get_converted_size(cursor->index, tuple);
	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 = 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 */

	for (;;) {
		/* 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, ULINT_UNDEFINED, &heap);
			incl_data += rec_offs_size(offsets);
		}

		n++;
		
		if (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 (UNIV_LIKELY_NULL(heap)) {
				mem_heap_free(heap);
			}
			return(rec);
		}
	}
}		

/*****************************************************************
Returns TRUE if the insert fits on the appropriate half-page with the
chosen split_rec. */
static
ibool
btr_page_insert_fits(
/*=================*/
					/* out: TRUE if fits */
	btr_cur_t*	cursor,		/* in: cursor at which insert
					should be made */
	rec_t*		split_rec,	/* in: suggestion for first record
					on upper half-page, or NULL if
					tuple to be inserted should be first */
	const ulint*	offsets,	/* in: rec_get_offsets(
					split_rec, cursor->index) */
	dtuple_t*	tuple,		/* in: tuple to insert */
	mem_heap_t*	heap)		/* in: temporary memory heap */
{
	page_t*	page;
	ulint	insert_size;
	ulint	free_space;
	ulint	total_data;
	ulint	total_n_recs;
	rec_t*	rec;
	rec_t*	end_rec;
	ulint*	offs;
	
	page = btr_cur_get_page(cursor);

	ut_ad(!split_rec == !offsets);
	ut_ad(!offsets
		|| !page_is_comp(page) == !rec_offs_comp(offsets));
	ut_ad(!offsets
		|| rec_offs_validate(split_rec, cursor->index, offsets));

	insert_size = rec_get_converted_size(cursor->index, tuple);
	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 = 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 (split_rec == NULL) {
		rec = page_rec_get_next(page_get_infimum_rec(page));
		end_rec = page_rec_get_next(btr_cur_get_rec(cursor));

	} else if (cmp_dtuple_rec(tuple, split_rec, offsets) >= 0) {

		rec = page_rec_get_next(page_get_infimum_rec(page));
 		end_rec = split_rec;
	} else {
		rec = split_rec;
		end_rec = page_get_supremum_rec(page);
	}

	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);
	}

	offs = NULL;

	while (rec != end_rec) {
		/* In this loop we calculate the amount of reserved
		space after rec is removed from page. */

		offs = rec_get_offsets(rec, cursor->index, offs,
						ULINT_UNDEFINED, &heap);

		total_data -= rec_offs_size(offs);
		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);
		}

		rec = page_rec_get_next(rec);
	}

	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. */

void
btr_insert_on_non_leaf_level(
/*=========================*/
	dict_tree_t*	tree,	/* in: tree */
	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;		
	ulint		err;
	rec_t*		rec;

	ut_ad(level > 0);
	
	/* In the following, choose just any index from the tree as the
	first parameter for btr_cur_search_to_nth_level. */

	btr_cur_search_to_nth_level(UT_LIST_GET_FIRST(tree->tree_indexes),
				    level, tuple, PAGE_CUR_LE,
				    BTR_CONT_MODIFY_TREE,
				    &cursor, 0, mtr);

	err = btr_cur_pessimistic_insert(BTR_NO_LOCKING_FLAG
					| BTR_KEEP_SYS_FLAG
					| BTR_NO_UNDO_LOG_FLAG,
					&cursor, tuple,
					&rec, &dummy_big_rec, NULL, mtr);
	ut_a(err == DB_SUCCESS);
}

/******************************************************************
Attaches the halves of an index page on the appropriate level in an
index tree. */
static
void
btr_attach_half_pages(
/*==================*/
	dict_tree_t*	tree,		/* in: the index tree */
	page_t*		page,		/* in: page to be split */
	rec_t*		split_rec,	/* in: first record on upper
					half page */
	page_t*		new_page,	/* in: the new half page */
	ulint		direction,	/* in: FSP_UP or FSP_DOWN */
	mtr_t*		mtr)		/* in: mtr */
{
	ulint		space;
	rec_t*		node_ptr;
	page_t*		prev_page;
	page_t*		next_page;
	ulint		prev_page_no;
	ulint		next_page_no;
	ulint		level;
	page_t*		lower_page;
	page_t*		upper_page;
	ulint		lower_page_no;
	ulint		upper_page_no;
	dtuple_t*	node_ptr_upper;
	mem_heap_t* 	heap;

	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
			      				MTR_MEMO_PAGE_X_FIX));
	ut_ad(mtr_memo_contains(mtr, buf_block_align(new_page),
			      				MTR_MEMO_PAGE_X_FIX));
	ut_a(page_is_comp(page) == page_is_comp(new_page));

	/* 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) {

		lower_page_no = buf_frame_get_page_no(new_page);
		upper_page_no = buf_frame_get_page_no(page);
		lower_page = new_page;
		upper_page = page;

		/* Look from the tree for the node pointer to page */
		node_ptr = btr_page_get_father_node_ptr(tree, page, mtr);

		/* Replace the address of the old child node (= page) with the 
		address of the new lower half */

		btr_node_ptr_set_child_page_no(node_ptr,
			rec_get_offsets(node_ptr,
					UT_LIST_GET_FIRST(tree->tree_indexes),
					NULL, ULINT_UNDEFINED, &heap),
			lower_page_no, mtr);
		mem_heap_empty(heap);
	} else {
		lower_page_no = buf_frame_get_page_no(page);
		upper_page_no = buf_frame_get_page_no(new_page);
		lower_page = page;
		upper_page = new_page;
	}
				   
	/* Get the level of the split pages */
	level = btr_page_get_level(page, mtr);

	/* Build the node pointer (= node key and page address) for the upper
	half */

	node_ptr_upper = dict_tree_build_node_ptr(tree, split_rec,
					     upper_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. */

	btr_insert_on_non_leaf_level(tree, level + 1, node_ptr_upper, mtr);
		
	/* Free the memory heap */
	mem_heap_free(heap);

	/* Get the previous and next pages of page */

	prev_page_no = btr_page_get_prev(page, mtr);
	next_page_no = btr_page_get_next(page, mtr);
	space = buf_frame_get_space_id(page);
	
	/* Update page links of the level */
	
	if (prev_page_no != FIL_NULL) {

		prev_page = btr_page_get(space, prev_page_no, RW_X_LATCH, mtr);
		ut_a(page_is_comp(prev_page) == page_is_comp(page));

		btr_page_set_next(prev_page, lower_page_no, mtr);
	}

	if (next_page_no != FIL_NULL) {

		next_page = btr_page_get(space, next_page_no, RW_X_LATCH, mtr);
		ut_a(page_is_comp(next_page) == page_is_comp(page));

		btr_page_set_prev(next_page, upper_page_no, mtr);
	}
	
	btr_page_set_prev(lower_page, prev_page_no, mtr);
	btr_page_set_next(lower_page, upper_page_no, mtr);
	btr_page_set_level(lower_page, level, mtr);

	btr_page_set_prev(upper_page, lower_page_no, mtr);
	btr_page_set_next(upper_page, next_page_no, mtr);
	btr_page_set_level(upper_page, level, mtr);
}

/*****************************************************************
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 must be guaranteed to be available before
this function is called. */

rec_t*
btr_page_split_and_insert(
/*======================*/
				/* out: inserted record; NOTE: the tree
				x-latch is released! NOTE: 2 free disk
				pages must be available! */
	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 */
	dtuple_t*	tuple,	/* in: tuple to insert */
	mtr_t*		mtr)	/* in: mtr */
{
	dict_tree_t*	tree;
	page_t*		page;
	ulint		page_no;
	byte		direction;
	ulint		hint_page_no;
	page_t*		new_page;
	rec_t*		split_rec;
	page_t*		left_page;
	page_t*		right_page;
	page_t*		insert_page;
	page_cur_t*	page_cursor;
	rec_t*		first_rec;
	byte*		buf = 0; /* remove warning */
	rec_t*		move_limit;
	ibool		insert_will_fit;
	ulint		n_iterations = 0;
	rec_t*		rec;
	mem_heap_t*	heap;
	ulint		n_uniq;
	ulint*		offsets;

	heap = mem_heap_create(1024);
	n_uniq = dict_index_get_n_unique_in_tree(cursor->index);
func_start:
	mem_heap_empty(heap);
	offsets = NULL;
	tree = btr_cur_get_tree(cursor);
	
	ut_ad(mtr_memo_contains(mtr, dict_tree_get_lock(tree),
							MTR_MEMO_X_LOCK));
#ifdef UNIV_SYNC_DEBUG
	ut_ad(rw_lock_own(dict_tree_get_lock(tree), RW_LOCK_EX));
#endif /* UNIV_SYNC_DEBUG */

	page = btr_cur_get_page(cursor);

	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
			      				MTR_MEMO_PAGE_X_FIX));
	ut_ad(page_get_n_recs(page) >= 2);
	
	page_no = buf_frame_get_page_no(page);

	/* 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 */

	if (n_iterations > 0) {
		direction = FSP_UP;
		hint_page_no = page_no + 1;
		split_rec = btr_page_get_sure_split_rec(cursor, tuple);
		
	} else if (btr_page_get_split_rec_to_right(cursor, &split_rec)) {
		direction = FSP_UP;
		hint_page_no = page_no + 1;

	} else if (btr_page_get_split_rec_to_left(cursor, &split_rec)) {
		direction = FSP_DOWN;
		hint_page_no = page_no - 1;
	} else {
		direction = FSP_UP;
		hint_page_no = page_no + 1;
		split_rec = page_get_middle_rec(page);
	}

	/* 2. Allocate a new page to the tree */
	new_page = btr_page_alloc(tree, hint_page_no, direction,
					btr_page_get_level(page, mtr), mtr);
	btr_page_create(new_page, tree, mtr);
	
	/* 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 != NULL) {
		first_rec = split_rec;
		move_limit = split_rec;
	} else {
		buf = mem_alloc(rec_get_converted_size(cursor->index, tuple));

		first_rec = rec_convert_dtuple_to_rec(buf,
							cursor->index, tuple);
		move_limit = page_rec_get_next(btr_cur_get_rec(cursor));
	}
	
	/* 4. Do first the modifications in the tree structure */

	btr_attach_half_pages(tree, page, first_rec, new_page, direction, mtr);

	if (split_rec == NULL) {
		mem_free(buf);
	}

	/* 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. */

	if (split_rec) {
		offsets = rec_get_offsets(split_rec, cursor->index, offsets,
							n_uniq, &heap);

		insert_will_fit = btr_page_insert_fits(cursor,
					split_rec, offsets, tuple, heap);
	} else {
		insert_will_fit = btr_page_insert_fits(cursor,
					NULL, NULL, tuple, heap);
	}
	
	if (insert_will_fit && (btr_page_get_level(page, mtr) == 0)) {

		mtr_memo_release(mtr, dict_tree_get_lock(tree),
							MTR_MEMO_X_LOCK);
	}

	/* 5. Move then the records to the new page */
	if (direction == FSP_DOWN) {
/*		fputs("Split left\n", stderr); */

		page_move_rec_list_start(new_page, page, move_limit,
							cursor->index, mtr);
		left_page = new_page;
		right_page = page;

		lock_update_split_left(right_page, left_page);
	} else {
/*		fputs("Split right\n", stderr); */

		page_move_rec_list_end(new_page, page, move_limit,
							cursor->index, mtr);
		left_page = page;
		right_page = new_page;

		lock_update_split_right(right_page, left_page);
	}

	/* 6. The split and the tree modification is now completed. Decide the
	page where the tuple should be inserted */

	if (split_rec == NULL) {
		insert_page = right_page;

	} else {
		offsets = rec_get_offsets(first_rec, cursor->index,
						offsets, n_uniq, &heap);

		if (cmp_dtuple_rec(tuple, first_rec, offsets) >= 0) {

			insert_page = right_page;
		} else {
			insert_page = left_page;
		}
	}

	/* 7. Reposition the cursor for insert and try insertion */
	page_cursor = btr_cur_get_page_cur(cursor);

	page_cur_search(insert_page, cursor->index, tuple,
						PAGE_CUR_LE, page_cursor);

	rec = page_cur_tuple_insert(page_cursor, tuple, cursor->index, mtr);

	if (rec != NULL) {
		/* Insert fit on the page: update the free bits for the
		left and right pages in the same mtr */

		ibuf_update_free_bits_for_two_pages_low(cursor->index,
							left_page,
							right_page, mtr);
		/* fprintf(stderr, "Split and insert done %lu %lu\n",
				buf_frame_get_page_no(left_page),
				buf_frame_get_page_no(right_page)); */
		mem_heap_free(heap);
		return(rec);
	}
	
	/* 8. If insert did not fit, try page reorganization */

	btr_page_reorganize(insert_page, cursor->index, mtr);

	page_cur_search(insert_page, cursor->index, tuple,
						PAGE_CUR_LE, page_cursor);
	rec = page_cur_tuple_insert(page_cursor, tuple, cursor->index, mtr);

	if (rec == NULL) {
		/* The insert did not fit on the page: loop back to the
		start of the function for a new split */
		
		/* We play safe and reset the free bits for new_page */
		ibuf_reset_free_bits(cursor->index, new_page);

		/* fprintf(stderr, "Split second round %lu\n",
					buf_frame_get_page_no(page)); */
		n_iterations++;
		ut_ad(n_iterations < 2);
		ut_ad(!insert_will_fit);

		goto func_start;
	}

	/* Insert fit on the page: update the free bits for the
	left and right pages in the same mtr */

	ibuf_update_free_bits_for_two_pages_low(cursor->index, left_page,
							right_page, mtr);
	/* fprintf(stderr, "Split and insert done %lu %lu\n",
				buf_frame_get_page_no(left_page),
				buf_frame_get_page_no(right_page)); */

	ut_ad(page_validate(left_page, UT_LIST_GET_FIRST(tree->tree_indexes)));
	ut_ad(page_validate(right_page, UT_LIST_GET_FIRST(tree->tree_indexes)));

	mem_heap_free(heap);
	return(rec);
}

/*****************************************************************
Removes a page from the level list of pages. */
static
void
btr_level_list_remove(
/*==================*/
	dict_tree_t*	tree __attribute__((unused)), /* in: index tree */
	page_t*		page,	/* in: page to remove */
	mtr_t*		mtr)	/* in: mtr */
{	
	ulint	space;
	ulint	prev_page_no;
	page_t*	prev_page;
	ulint	next_page_no;
	page_t*	next_page;
	
	ut_ad(tree && page && mtr);
	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
			      				MTR_MEMO_PAGE_X_FIX));
	/* Get the previous and next page numbers of page */

	prev_page_no = btr_page_get_prev(page, mtr);
	next_page_no = btr_page_get_next(page, mtr);
	space = buf_frame_get_space_id(page);
	
	/* Update page links of the level */
	
	if (prev_page_no != FIL_NULL) {

		prev_page = btr_page_get(space, prev_page_no, RW_X_LATCH, mtr);
		ut_a(page_is_comp(prev_page) == page_is_comp(page));

		btr_page_set_next(prev_page, next_page_no, mtr);
	}

	if (next_page_no != FIL_NULL) {

		next_page = btr_page_get(space, next_page_no, RW_X_LATCH, mtr);
		ut_a(page_is_comp(next_page) == page_is_comp(page));

		btr_page_set_prev(next_page, prev_page_no, mtr);
	}
}
	
/********************************************************************
Writes the redo log record for setting an index record as the predefined
minimum record. */
UNIV_INLINE
void
btr_set_min_rec_mark_log(
/*=====================*/
	rec_t*	rec,	/* in: record */
	ulint	comp,	/* nonzero=compact record format */
	mtr_t*	mtr)	/* in: mtr */
{
	mlog_write_initial_log_record(rec,
		comp ? MLOG_COMP_REC_MIN_MARK : MLOG_REC_MIN_MARK, mtr);

	/* Write rec offset as a 2-byte ulint */
	mlog_catenate_ulint(mtr, ut_align_offset(rec, UNIV_PAGE_SIZE),
								MLOG_2BYTES);
}

/********************************************************************
Parses the redo log record for setting an index record as the predefined
minimum record. */

byte*
btr_parse_set_min_rec_mark(
/*=======================*/
			/* out: end of log record or NULL */
	byte*	ptr,	/* in: buffer */
	byte*	end_ptr,/* in: buffer end */
	ulint	comp,	/* in: nonzero=compact page format */
	page_t*	page,	/* in: page or NULL */
	mtr_t*	mtr)	/* in: mtr or NULL */
{
	rec_t*	rec;

	if (end_ptr < ptr + 2) {

		return(NULL);
	}

	if (page) {
		ut_a(!page_is_comp(page) == !comp);

		rec = page + mach_read_from_2(ptr);

		btr_set_min_rec_mark(rec, comp, mtr);
	}

	return(ptr + 2);
}

/********************************************************************
Sets a record as the predefined minimum record. */

void
btr_set_min_rec_mark(
/*=================*/
	rec_t*	rec,	/* in: record */
	ulint	comp,	/* in: nonzero=compact page format */
	mtr_t*	mtr)	/* in: mtr */
{
	ulint	info_bits;

	info_bits = rec_get_info_bits(rec, comp);

	rec_set_info_bits(rec, comp, info_bits | REC_INFO_MIN_REC_FLAG);

	btr_set_min_rec_mark_log(rec, comp, mtr);
}

/*****************************************************************
Deletes on the upper level the node pointer to a page. */

void
btr_node_ptr_delete(
/*================*/
	dict_tree_t*	tree,	/* in: index tree */
	page_t*		page,	/* in: page whose node pointer is deleted */
	mtr_t*		mtr)	/* in: mtr */
{
	rec_t*		node_ptr;
	btr_cur_t	cursor;
	ibool		compressed;
	ulint		err;
	
	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
							MTR_MEMO_PAGE_X_FIX));
	/* Delete node pointer on father page */

	node_ptr = btr_page_get_father_node_ptr(tree, page, mtr);

	btr_cur_position(UT_LIST_GET_FIRST(tree->tree_indexes), node_ptr,
								&cursor);
	compressed = btr_cur_pessimistic_delete(&err, TRUE, &cursor, FALSE,
									mtr);
	ut_a(err == DB_SUCCESS);

	if (!compressed) {
		btr_cur_compress_if_useful(&cursor, mtr);
	}
}

/*****************************************************************
If page is the only on its level, this function moves its records to the
father page, thus reducing the tree height. */
static
void
btr_lift_page_up(
/*=============*/
	dict_tree_t*	tree,	/* in: index tree */
	page_t*		page,	/* 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: mtr */
{
	page_t*		father_page;
	ulint		page_level;
	dict_index_t*	index;

	ut_ad(btr_page_get_prev(page, mtr) == FIL_NULL);
	ut_ad(btr_page_get_next(page, mtr) == FIL_NULL);
	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
			      				MTR_MEMO_PAGE_X_FIX));
	father_page = buf_frame_align(
			btr_page_get_father_node_ptr(tree, page, mtr));
	
	page_level = btr_page_get_level(page, mtr);
	index = UT_LIST_GET_FIRST(tree->tree_indexes);

	btr_search_drop_page_hash_index(page);
	
	/* Make the father empty */
	btr_page_empty(father_page, mtr);

	/* Move records to the father */
 	page_copy_rec_list_end(father_page, page, page_get_infimum_rec(page),
								index, mtr);
	lock_update_copy_and_discard(father_page, page);

	btr_page_set_level(father_page, page_level, mtr);

	/* Free the file page */
	btr_page_free(tree, page, mtr);		

	/* We play safe and reset the free bits for the father */
	ibuf_reset_free_bits(index, father_page);
	ut_ad(page_validate(father_page, index));
	ut_ad(btr_check_node_ptr(tree, father_page, mtr));
}	

/*****************************************************************
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. NOTE: it is assumed that the caller has reserved
enough free extents so that the compression will always succeed if done! */

void
btr_compress(
/*=========*/
	btr_cur_t*	cursor,	/* in: cursor on the page to merge or lift;
				the page must not be empty: in record delete
				use btr_discard_page if the page would become
				empty */
	mtr_t*		mtr)	/* in: mtr */
{
	dict_tree_t*	tree;
	ulint		space;
	ulint		left_page_no;
	ulint		right_page_no;
	page_t*		merge_page;
	page_t*		father_page;
	ibool		is_left;
	page_t*		page;
	rec_t*		orig_pred;
	rec_t*		orig_succ;
	rec_t*		node_ptr;
	ulint		data_size;
	ulint		n_recs;
	ulint		max_ins_size;
	ulint		max_ins_size_reorg;
	ulint		level;
	ulint		comp;

	page = btr_cur_get_page(cursor);
	tree = btr_cur_get_tree(cursor);
	comp = page_is_comp(page);
	ut_a((ibool)!!comp == cursor->index->table->comp);

	ut_ad(mtr_memo_contains(mtr, dict_tree_get_lock(tree),
							MTR_MEMO_X_LOCK));
	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
							MTR_MEMO_PAGE_X_FIX));
	level = btr_page_get_level(page, mtr);
	space = dict_tree_get_space(tree);

	left_page_no = btr_page_get_prev(page, mtr);
	right_page_no = btr_page_get_next(page, mtr);

/*	fprintf(stderr, "Merge left page %lu right %lu \n", left_page_no,
							right_page_no); */

	node_ptr = btr_page_get_father_node_ptr(tree, page, mtr);
	ut_ad(!comp || rec_get_status(node_ptr) == REC_STATUS_NODE_PTR);
	father_page = buf_frame_align(node_ptr);
	ut_a(comp == page_is_comp(father_page));

	/* Decide the page to which we try to merge and which will inherit
	the locks */

	if (left_page_no != FIL_NULL) {

		is_left = TRUE;
		merge_page = btr_page_get(space, left_page_no, RW_X_LATCH,
									mtr);
	} else if (right_page_no != FIL_NULL) {

		is_left = FALSE;
		merge_page = btr_page_get(space, right_page_no, RW_X_LATCH,
									mtr);
	} else {
		/* The page is the only one on the level, lift the records
		to the father */
		btr_lift_page_up(tree, page, mtr);

		return;
	}
	
	n_recs = page_get_n_recs(page);
	data_size = page_get_data_size(page);
	ut_a(page_is_comp(merge_page) == comp);

	max_ins_size_reorg = page_get_max_insert_size_after_reorganize(
							merge_page, n_recs);
	if (data_size > max_ins_size_reorg) {

		/* No space for merge */

		return;
	}

	ut_ad(page_validate(merge_page, cursor->index));

	max_ins_size = page_get_max_insert_size(merge_page, n_recs);
	
	if (data_size > max_ins_size) {

		/* We have to reorganize merge_page */

		btr_page_reorganize(merge_page, cursor->index, mtr);

		max_ins_size = page_get_max_insert_size(merge_page, n_recs);

		ut_ad(page_validate(merge_page, cursor->index));
		ut_ad(page_get_max_insert_size(merge_page, n_recs)
							== max_ins_size_reorg);
	}

	if (data_size > max_ins_size) {

		/* Add fault tolerance, though this should never happen */

		return;
	}

	btr_search_drop_page_hash_index(page);

	/* Remove the page from the level list */
	btr_level_list_remove(tree, page, mtr);

	if (is_left) {
		btr_node_ptr_delete(tree, page, mtr);
	} else {
		mem_heap_t*	heap		= NULL;
		ulint		offsets_[REC_OFFS_NORMAL_SIZE];
		*offsets_ = (sizeof offsets_) / sizeof *offsets_;
		/* 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(node_ptr,
				rec_get_offsets(node_ptr, cursor->index,
				offsets_, ULINT_UNDEFINED, &heap),
				right_page_no, mtr);
		if (UNIV_LIKELY_NULL(heap)) {
			mem_heap_free(heap);
		}
		btr_node_ptr_delete(tree, merge_page, mtr);
	}
	
	/* Move records to the merge page */
	if (is_left) {
		orig_pred = page_rec_get_prev(
					page_get_supremum_rec(merge_page));
		page_copy_rec_list_start(merge_page, page,
			page_get_supremum_rec(page), cursor->index, mtr);

		lock_update_merge_left(merge_page, orig_pred, page);
	} else {
		orig_succ = page_rec_get_next(
					page_get_infimum_rec(merge_page));
		page_copy_rec_list_end(merge_page, page,
			page_get_infimum_rec(page), cursor->index, mtr);

		lock_update_merge_right(orig_succ, page);
	}

	/* We have added new records to merge_page: update its free bits */
	ibuf_update_free_bits_if_full(cursor->index, merge_page,
					UNIV_PAGE_SIZE, ULINT_UNDEFINED);
					
	ut_ad(page_validate(merge_page, cursor->index));

	/* Free the file page */
	btr_page_free(tree, page, mtr);		

	ut_ad(btr_check_node_ptr(tree, merge_page, mtr));
}	

/*****************************************************************
Discards a page that is the only page on its level. */
static
void
btr_discard_only_page_on_level(
/*===========================*/
	dict_tree_t*	tree,	/* in: index tree */
	page_t*		page,	/* in: page which is the only on its level */
	mtr_t*		mtr)	/* in: mtr */
{
	rec_t*	node_ptr;
	page_t*	father_page;
	ulint	page_level;
	
	ut_ad(btr_page_get_prev(page, mtr) == FIL_NULL);
	ut_ad(btr_page_get_next(page, mtr) == FIL_NULL);
	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
							MTR_MEMO_PAGE_X_FIX));
	btr_search_drop_page_hash_index(page);

	node_ptr = btr_page_get_father_node_ptr(tree, page, mtr);
	father_page = buf_frame_align(node_ptr);

	page_level = btr_page_get_level(page, mtr);

	lock_update_discard(page_get_supremum_rec(father_page), page);

	btr_page_set_level(father_page, page_level, mtr);

	/* Free the file page */
	btr_page_free(tree, page, mtr);		

	if (buf_frame_get_page_no(father_page) == dict_tree_get_page(tree)) {
		/* The father is the root page */

		btr_page_empty(father_page, mtr);

		/* We play safe and reset the free bits for the father */
		ibuf_reset_free_bits(UT_LIST_GET_FIRST(tree->tree_indexes),
								father_page);
	} else {
		ut_ad(page_get_n_recs(father_page) == 1);

		btr_discard_only_page_on_level(tree, father_page, 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. */

void
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_tree_t*	tree;
	ulint		space;
	ulint		left_page_no;
	ulint		right_page_no;
	page_t*		merge_page;
	ibool		is_left;
	page_t*		page;
	rec_t*		node_ptr;
	
	page = btr_cur_get_page(cursor);
	tree = btr_cur_get_tree(cursor);

	ut_ad(dict_tree_get_page(tree) != buf_frame_get_page_no(page));
	ut_ad(mtr_memo_contains(mtr, dict_tree_get_lock(tree),
							MTR_MEMO_X_LOCK));
	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
							MTR_MEMO_PAGE_X_FIX));
	space = dict_tree_get_space(tree);
	
	/* Decide the page which will inherit the locks */

	left_page_no = btr_page_get_prev(page, mtr);
	right_page_no = btr_page_get_next(page, mtr);

	if (left_page_no != FIL_NULL) {
		is_left = TRUE;
		merge_page = btr_page_get(space, left_page_no, RW_X_LATCH,
									mtr);
	} else if (right_page_no != FIL_NULL) {
		is_left = FALSE;
		merge_page = btr_page_get(space, right_page_no, RW_X_LATCH,
									mtr);
	} else {
		btr_discard_only_page_on_level(tree, page, mtr);

		return;
	}

	ut_a(page_is_comp(merge_page) == page_is_comp(page));
	btr_search_drop_page_hash_index(page);
	
	if (left_page_no == FIL_NULL && btr_page_get_level(page, mtr) > 0) {

		/* We have to mark the leftmost node pointer on the right
		side page as the predefined minimum record */

		node_ptr = page_rec_get_next(page_get_infimum_rec(merge_page));

		ut_ad(page_rec_is_user_rec(node_ptr));

		btr_set_min_rec_mark(node_ptr, page_is_comp(merge_page), mtr);
	}	
	
	btr_node_ptr_delete(tree, page, mtr);

	/* Remove the page from the level list */
	btr_level_list_remove(tree, page, mtr);

	if (is_left) {
		lock_update_discard(page_get_supremum_rec(merge_page), page);
	} else {
		lock_update_discard(page_rec_get_next(
				    page_get_infimum_rec(merge_page)), page);
	}

	/* Free the file page */
	btr_page_free(tree, page, mtr);		

	ut_ad(btr_check_node_ptr(tree, merge_page, mtr));
}	

#ifdef UNIV_BTR_PRINT
/*****************************************************************
Prints size info of a B-tree. */

void
btr_print_size(
/*===========*/
	dict_tree_t*	tree)	/* in: index tree */
{
	page_t*		root;
	fseg_header_t*	seg;
	mtr_t		mtr;

	if (tree->type & DICT_IBUF) {
		fputs(
	"Sorry, cannot print info of an ibuf tree: use ibuf functions\n",
			stderr);

		return;
	}

	mtr_start(&mtr);
	
	root = btr_root_get(tree, &mtr);

	seg = root + PAGE_HEADER + PAGE_BTR_SEG_TOP;

	fputs("INFO OF THE NON-LEAF PAGE SEGMENT\n", stderr);
	fseg_print(seg, &mtr);

	if (!(tree->type & DICT_UNIVERSAL)) {

		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_tree_t*	tree,	/* in: index tree */
	page_t*		page,	/* 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() */
	ulint**		offsets,/* in/out: buffer for rec_get_offsets() */
	mtr_t*		mtr)	/* in: mtr */
{
	page_cur_t	cursor;
	ulint		n_recs;
	ulint		i	= 0;
	mtr_t		mtr2;
	rec_t*		node_ptr;
	page_t*		child;
	dict_index_t*	index;

	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
							MTR_MEMO_PAGE_X_FIX));
	fprintf(stderr, "NODE ON LEVEL %lu page number %lu\n",
	       (ulong) btr_page_get_level(page, mtr),
	       (ulong) buf_frame_get_page_no(page));
	
	index = UT_LIST_GET_FIRST(tree->tree_indexes);
	page_print(page, index, width, width);
	
	n_recs = page_get_n_recs(page);
	
	page_cur_set_before_first(page, &cursor);
	page_cur_move_to_next(&cursor);

	while (!page_cur_is_after_last(&cursor)) {

		if (0 == btr_page_get_level(page, mtr)) {

			/* If this is the leaf level, do nothing */

		} else if ((i <= width) || (i >= n_recs - width)) {

			mtr_start(&mtr2);

			node_ptr = page_cur_get_rec(&cursor);

			*offsets = rec_get_offsets(node_ptr, index, *offsets,
					ULINT_UNDEFINED, heap);
			child = btr_node_ptr_get_child(node_ptr,
					*offsets, &mtr2);
			btr_print_recursive(tree, 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_tree(
/*===========*/
	dict_tree_t*	tree,	/* in: tree */
	ulint		width)	/* in: print this many entries from start
				and end */
{
	mtr_t		mtr;
	page_t*		root;
	mem_heap_t*	heap	= NULL;
	ulint		offsets_[REC_OFFS_NORMAL_SIZE];
	ulint*		offsets	= offsets_;
	*offsets_ = (sizeof offsets_) / sizeof *offsets_;

	fputs("--------------------------\n"
		"INDEX TREE PRINT\n", stderr);

	mtr_start(&mtr);

	root = btr_root_get(tree, &mtr);

	btr_print_recursive(tree, root, width, &heap, &offsets, &mtr);
	if (UNIV_LIKELY_NULL(heap)) {
		mem_heap_free(heap);
	}

	mtr_commit(&mtr);

	btr_validate_tree(tree, NULL);
}
#endif /* UNIV_BTR_PRINT */

/****************************************************************
Checks that the node pointer to a page is appropriate. */

ibool
btr_check_node_ptr(
/*===============*/
				/* out: TRUE */
	dict_tree_t*	tree,	/* in: index tree */
	page_t*		page,	/* in: index page */
	mtr_t*		mtr)	/* in: mtr */
{
	mem_heap_t*	heap;
	rec_t*		node_ptr;
	dtuple_t*	node_ptr_tuple;

	ut_ad(mtr_memo_contains(mtr, buf_block_align(page),
							MTR_MEMO_PAGE_X_FIX));
	if (dict_tree_get_page(tree) == buf_frame_get_page_no(page)) {

		return(TRUE);
	}

	node_ptr = btr_page_get_father_node_ptr(tree, page, mtr);
 
	if (btr_page_get_level(page, mtr) == 0) {

		return(TRUE);
	}
	
	heap = mem_heap_create(256);
		
	node_ptr_tuple = dict_tree_build_node_ptr(
				tree,
				page_rec_get_next(page_get_infimum_rec(page)),
				0, heap, btr_page_get_level(page, mtr));
				
	ut_a(cmp_dtuple_rec(node_ptr_tuple, node_ptr,
			rec_get_offsets(node_ptr,
			dict_tree_find_index(tree, node_ptr),
			NULL, ULINT_UNDEFINED, &heap)) == 0);

	mem_heap_free(heap);

	return(TRUE);
}

/****************************************************************
Display identification information for a record. */
static
void
btr_index_rec_validate_report(
/*==========================*/
	page_t*		page,	/* in: index page */
	rec_t*		rec,	/* in: index record */
	dict_index_t*	index)	/* in: index */
{
	fputs("InnoDB: Record in ", stderr);
	dict_index_name_print(stderr, NULL, index);
	fprintf(stderr, ", page %lu, at offset %lu\n",
		buf_frame_get_page_no(page), (ulint)(rec - page));
}

/****************************************************************
Checks the size and number of fields in a record based on the definition of
the index. */

ibool
btr_index_rec_validate(
/*====================*/
					/* out: TRUE if ok */
	rec_t*		rec,		/* in: index record */
	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;
	ulint		n;
	ulint		i;
	page_t*		page;
	mem_heap_t*	heap	= NULL;
	ulint		offsets_[REC_OFFS_NORMAL_SIZE];
	ulint*		offsets	= offsets_;
	*offsets_ = (sizeof offsets_) / sizeof *offsets_;

	page = buf_frame_align(rec);

	if (UNIV_UNLIKELY(index->type & DICT_UNIVERSAL)) {
	        /* The insert buffer index tree can contain records from any
	        other index: we cannot check the number of fields or
	        their length */

	        return(TRUE);
	}

	if (UNIV_UNLIKELY((ibool)!!page_is_comp(page) != index->table->comp)) {
		btr_index_rec_validate_report(page, rec, index);
		fprintf(stderr, "InnoDB: compact flag=%lu, should be %lu\n",
			(ulong) !!page_is_comp(page),
			(ulong) index->table->comp);
		return(FALSE);
	}

	n = dict_index_get_n_fields(index);

	if (!page_is_comp(page)
			&& UNIV_UNLIKELY(rec_get_n_fields_old(rec) != n)) {
		btr_index_rec_validate_report(page, rec, index);
		fprintf(stderr, "InnoDB: has %lu fields, should have %lu\n",
			(ulong) rec_get_n_fields_old(rec), (ulong) n);

		if (dump_on_error) {
			buf_page_print(page);

			fputs("InnoDB: corrupt record ", stderr);
			rec_print_old(stderr, rec);
			putc('\n', stderr);
		}
		return(FALSE);
	}

	offsets = rec_get_offsets(rec, index, offsets, ULINT_UNDEFINED, &heap);

	for (i = 0; i < n; i++) {
		dtype_t*	type = dict_index_get_nth_type(index, i);
		ulint		fixed_size = dtype_get_fixed_size(type);

		rec_get_nth_field(rec, offsets, i, &len);

		/* Note that prefix indexes are not fixed size even when
		their type is CHAR. */

		if ((dict_index_get_nth_field(index, i)->prefix_len == 0
		    && len != UNIV_SQL_NULL && fixed_size
		    && len != fixed_size)
		   ||
		   (dict_index_get_nth_field(index, i)->prefix_len > 0
		    && len != UNIV_SQL_NULL
		    && len >
			   dict_index_get_nth_field(index, i)->prefix_len)) {

			btr_index_rec_validate_report(page, rec, index);
			fprintf(stderr,
"InnoDB: field %lu len is %lu, should be %lu\n",
				(ulong) i, (ulong) len, (ulong) dtype_get_fixed_size(type));

			if (dump_on_error) {
				buf_page_print(page);

				fputs("InnoDB: corrupt record ", stderr);
				rec_print_new(stderr, rec, offsets);
				putc('\n', stderr);
			}
			if (UNIV_LIKELY_NULL(heap)) {
				mem_heap_free(heap);
			}
			return(FALSE);
		}
	}

	if (UNIV_LIKELY_NULL(heap)) {
		mem_heap_free(heap);
	}
	return(TRUE);			
}

/****************************************************************
Checks the size and number of fields in records based on the definition of
the index. */
static
ibool
btr_index_page_validate(
/*====================*/
				/* out: TRUE if ok */
	page_t*		page,	/* in: index page */
	dict_index_t*	index)	/* in: index */
{
	page_cur_t 	cur;
	ibool		ret	= TRUE;
	
	page_cur_set_before_first(page, &cur);
	page_cur_move_to_next(&cur);

	for (;;) {
		if (page_cur_is_after_last(&cur)) {

			break;
		}

		if (!btr_index_rec_validate(cur.rec, index, TRUE)) {

			return(FALSE);
		}

		page_cur_move_to_next(&cur);
	}

	return(ret);	
}

/****************************************************************
Report an error on one page of an index tree. */
static
void
btr_validate_report1(
				/* out: TRUE if ok */
	dict_index_t*	index,	/* in: index */
	ulint		level,	/* in: B-tree level */
	page_t*		page)	/* in: index page */
{
	fprintf(stderr, "InnoDB: Error in page %lu of ",
		buf_frame_get_page_no(page));
	dict_index_name_print(stderr, NULL, index);
	if (level) {
		fprintf(stderr, ", index tree level %lu", level);
	}
	putc('\n', stderr);
}

/****************************************************************
Report an error on two pages of an index tree. */
static
void
btr_validate_report2(
				/* out: TRUE if ok */
	dict_index_t*	index,	/* in: index */
	ulint		level,	/* in: B-tree level */
	page_t*		page1,	/* in: first index page */
	page_t*		page2)	/* in: second index page */
{
	fprintf(stderr, "InnoDB: Error in pages %lu and %lu of ",
		buf_frame_get_page_no(page1),
		buf_frame_get_page_no(page2));
	dict_index_name_print(stderr, NULL, index);
	if (level) {
		fprintf(stderr, ", index tree level %lu", level);
	}
	putc('\n', stderr);
}

/****************************************************************
Validates index tree level. */
static
ibool
btr_validate_level(
/*===============*/
				/* out: TRUE if ok */
	dict_tree_t*	tree,	/* in: index tree */
	trx_t*		trx,	/* in: transaction or NULL */
	ulint		level)	/* in: level number */
{
	ulint		space;
	page_t*		page;
	page_t*		right_page = 0; /* remove warning */
	page_t*		father_page;
	page_t*		right_father_page;
	rec_t*		node_ptr;
	rec_t*		right_node_ptr;
	rec_t*		rec;
	ulint		right_page_no;
	ulint		left_page_no;
	page_cur_t	cursor;
	dtuple_t*	node_ptr_tuple;
	ibool		ret	= TRUE;
	dict_index_t*	index;
	mtr_t		mtr;
	mem_heap_t*	heap	= mem_heap_create(256);
	ulint*		offsets	= NULL;
	ulint*		offsets2= NULL;

	mtr_start(&mtr);

	mtr_x_lock(dict_tree_get_lock(tree), &mtr);
	
	page = btr_root_get(tree, &mtr);

	space = buf_frame_get_space_id(page);

	index = UT_LIST_GET_FIRST(tree->tree_indexes);

	while (level != btr_page_get_level(page, &mtr)) {

		ut_a(btr_page_get_level(page, &mtr) > 0);

		page_cur_set_before_first(page, &cursor);
		page_cur_move_to_next(&cursor);

		node_ptr = page_cur_get_rec(&cursor);
		offsets = rec_get_offsets(node_ptr, index, offsets,
					ULINT_UNDEFINED, &heap);
		page = btr_node_ptr_get_child(node_ptr, offsets, &mtr);
	}

	/* Now we are on the desired level. Loop through the pages on that
	level. */
loop:
	if (trx_is_interrupted(trx)) {
		mtr_commit(&mtr);
		mem_heap_free(heap);
		return(ret);
	}
	mem_heap_empty(heap);
	offsets = offsets2 = NULL;
	mtr_x_lock(dict_tree_get_lock(tree), &mtr);

	/* Check ordering etc. of records */

	if (!page_validate(page, index)) {
		btr_validate_report1(index, level, page);

		ret = FALSE;
	} else if (level == 0) {
		/* We are on level 0. Check that the records have the right
		number of fields, and field lengths are right. */

		if (!btr_index_page_validate(page, index)) {

			ret = FALSE;
		}
	}
	
	ut_a(btr_page_get_level(page, &mtr) == level);

	right_page_no = btr_page_get_next(page, &mtr);
	left_page_no = btr_page_get_prev(page, &mtr);

	ut_a((page_get_n_recs(page) > 0)
	     || ((level == 0) &&
		  (buf_frame_get_page_no(page) == dict_tree_get_page(tree))));

	if (right_page_no != FIL_NULL) {
		rec_t*	right_rec;
		right_page = btr_page_get(space, right_page_no, RW_X_LATCH,
									&mtr);
		ut_a(page_is_comp(right_page) == page_is_comp(page));
		rec = page_rec_get_prev(page_get_supremum_rec(page));
		right_rec = page_rec_get_next(
					page_get_infimum_rec(right_page));
		offsets = rec_get_offsets(rec, index,
					offsets, ULINT_UNDEFINED, &heap);
		offsets2 = rec_get_offsets(right_rec, index,
					offsets2, ULINT_UNDEFINED, &heap);
		if (cmp_rec_rec(rec, right_rec, offsets, offsets2, index)
				>= 0) {

			btr_validate_report2(index, level, page, right_page);

			fputs("InnoDB: records in wrong order"
				" on adjacent pages\n", stderr);

			buf_page_print(page);
			buf_page_print(right_page);

			fputs("InnoDB: record ", stderr);
			rec = page_rec_get_prev(page_get_supremum_rec(page));
			rec_print(stderr, rec, index);
			putc('\n', stderr);
			fputs("InnoDB: record ", stderr);
			rec = page_rec_get_next(page_get_infimum_rec(
						right_page));
			rec_print(stderr, rec, index);
			putc('\n', stderr);

	  		ret = FALSE;
	  	}
	}
	
	if (level > 0 && left_page_no == FIL_NULL) {
		ut_a(REC_INFO_MIN_REC_FLAG & rec_get_info_bits(
			page_rec_get_next(page_get_infimum_rec(page)),
				page_is_comp(page)));
	}

	if (buf_frame_get_page_no(page) != dict_tree_get_page(tree)) {

		/* Check father node pointers */
	
		node_ptr = btr_page_get_father_node_ptr(tree, page, &mtr);
		father_page = buf_frame_align(node_ptr);
		offsets	= rec_get_offsets(node_ptr, index,
					offsets, ULINT_UNDEFINED, &heap);

		if (btr_node_ptr_get_child_page_no(node_ptr, offsets) !=
						buf_frame_get_page_no(page)
		   || node_ptr != btr_page_get_father_for_rec(tree, page,
			page_rec_get_prev(page_get_supremum_rec(page)),
							&mtr)) {
			btr_validate_report1(index, level, page);

			fputs("InnoDB: node pointer to the page is wrong\n",
				stderr);

			buf_page_print(father_page);
			buf_page_print(page);

			fputs("InnoDB: node ptr ", stderr);
			rec_print_new(stderr, node_ptr, offsets);

			fprintf(stderr, "\n"
				"InnoDB: node ptr child page n:o %lu\n",
				(unsigned long) btr_node_ptr_get_child_page_no(
						node_ptr, offsets));

			fputs("InnoDB: record on page ", stderr);
			rec = btr_page_get_father_for_rec(tree, page,
				page_rec_get_prev(page_get_supremum_rec(page)),
				&mtr);
			rec_print(stderr, rec, index);
			putc('\n', stderr);
		   	ret = FALSE;

		   	goto node_ptr_fails;
		}

		if (btr_page_get_level(page, &mtr) > 0) {
			offsets	= rec_get_offsets(node_ptr, index,
					offsets, ULINT_UNDEFINED, &heap);
		
			node_ptr_tuple = dict_tree_build_node_ptr(
					tree,
					page_rec_get_next(
						page_get_infimum_rec(page)),
						0, heap,
       					btr_page_get_level(page, &mtr));

			if (cmp_dtuple_rec(node_ptr_tuple, node_ptr,
								offsets)) {
			  	rec_t*	first_rec	= page_rec_get_next(
					page_get_infimum_rec(page));

				btr_validate_report1(index, level, page);

				buf_page_print(father_page);
				buf_page_print(page);

				fputs("InnoDB: Error: node ptrs differ"
					" on levels > 0\n"
					"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);
		   		ret = FALSE;

		   		goto node_ptr_fails;
			}
		}

		if (left_page_no == FIL_NULL) {
			ut_a(node_ptr == page_rec_get_next(
					page_get_infimum_rec(father_page)));
			ut_a(btr_page_get_prev(father_page, &mtr) == FIL_NULL);
		}

		if (right_page_no == FIL_NULL) {
			ut_a(node_ptr == page_rec_get_prev(
				page_get_supremum_rec(father_page)));
			ut_a(btr_page_get_next(father_page, &mtr) == FIL_NULL);
		}

		if (right_page_no != FIL_NULL) {

			right_node_ptr = btr_page_get_father_node_ptr(tree,
							right_page, &mtr);
			if (page_rec_get_next(node_ptr) !=
					page_get_supremum_rec(father_page)) {

				if (right_node_ptr !=
						page_rec_get_next(node_ptr)) {
					ret = FALSE;
					fputs(
			"InnoDB: node pointer to the right page is wrong\n",
					stderr);

					btr_validate_report1(index, level,
						page);

					buf_page_print(father_page);
					buf_page_print(page);
					buf_page_print(right_page);
				}
			} else {
				right_father_page = buf_frame_align(
							right_node_ptr);
							
				if (right_node_ptr != page_rec_get_next(
					   		page_get_infimum_rec(
							right_father_page))) {
					ret = FALSE;
					fputs(
			"InnoDB: node pointer 2 to the right page is wrong\n",
					stderr);

					btr_validate_report1(index, level,
						page);

					buf_page_print(father_page);
					buf_page_print(right_father_page);
					buf_page_print(page);
					buf_page_print(right_page);
				}

				if (buf_frame_get_page_no(right_father_page)
				   != btr_page_get_next(father_page, &mtr)) {

					ret = FALSE;
					fputs(
			"InnoDB: node pointer 3 to the right page is wrong\n",
					stderr);

					btr_validate_report1(index, level,
						page);

					buf_page_print(father_page);
					buf_page_print(right_father_page);
					buf_page_print(page);
					buf_page_print(right_page);
				}
			}					
		}
	}

node_ptr_fails:
	mtr_commit(&mtr);

	if (right_page_no != FIL_NULL) {
		ulint	comp = page_is_comp(page);
		mtr_start(&mtr);
	
		page = btr_page_get(space, right_page_no, RW_X_LATCH, &mtr);
		ut_a(page_is_comp(page) == comp);

		goto loop;
	}

	mem_heap_free(heap);
	return(ret);
}

/******************************************************************
Checks the consistency of an index tree. */

ibool
btr_validate_tree(
/*==============*/
				/* out: TRUE if ok */
	dict_tree_t*	tree,	/* in: tree */
	trx_t*		trx)	/* in: transaction or NULL */
{
	mtr_t	mtr;
	page_t*	root;
	ulint	i;
	ulint	n;

	mtr_start(&mtr);
	mtr_x_lock(dict_tree_get_lock(tree), &mtr);

	root = btr_root_get(tree, &mtr);
	n = btr_page_get_level(root, &mtr);

	for (i = 0; i <= n && !trx_is_interrupted(trx); i++) {
		if (!btr_validate_level(tree, trx, n - i)) {

			mtr_commit(&mtr);

			return(FALSE);
		}
	}

	mtr_commit(&mtr);

	return(TRUE);
}