summaryrefslogtreecommitdiffstats
path: root/src/pager.c
blob: 37588f0b259e90ddfc2d874eb3c4785876a2d79d (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
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
3324
3325
3326
3327
3328
3329
3330
3331
3332
3333
3334
3335
3336
3337
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
3376
3377
3378
3379
3380
3381
3382
3383
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
3410
3411
3412
3413
3414
3415
3416
3417
3418
3419
3420
3421
3422
3423
3424
3425
3426
3427
3428
3429
3430
3431
3432
3433
3434
3435
3436
3437
3438
3439
3440
3441
3442
3443
3444
3445
3446
3447
3448
3449
3450
3451
3452
3453
3454
3455
3456
3457
3458
3459
3460
3461
3462
3463
3464
3465
3466
3467
3468
3469
3470
3471
3472
3473
3474
3475
3476
3477
3478
3479
3480
3481
3482
3483
3484
3485
3486
3487
3488
3489
3490
3491
3492
3493
3494
3495
3496
3497
3498
3499
3500
3501
3502
3503
3504
3505
3506
3507
3508
3509
3510
3511
3512
3513
3514
3515
3516
3517
3518
3519
3520
3521
3522
3523
3524
3525
3526
3527
3528
3529
3530
3531
3532
3533
3534
3535
3536
3537
3538
3539
3540
3541
3542
3543
3544
3545
3546
3547
3548
3549
3550
3551
3552
3553
3554
3555
3556
3557
3558
3559
3560
3561
3562
3563
3564
3565
3566
3567
3568
3569
3570
3571
3572
3573
3574
3575
3576
3577
3578
3579
3580
3581
3582
3583
3584
3585
3586
3587
3588
3589
3590
3591
3592
3593
3594
3595
3596
3597
3598
3599
3600
3601
3602
3603
3604
3605
3606
3607
3608
3609
3610
3611
3612
3613
3614
3615
3616
3617
3618
3619
3620
3621
3622
3623
3624
3625
3626
3627
3628
3629
3630
3631
3632
3633
3634
3635
3636
3637
3638
3639
3640
3641
3642
3643
3644
3645
3646
3647
3648
3649
3650
3651
3652
3653
3654
3655
3656
3657
3658
3659
3660
3661
3662
3663
3664
3665
3666
3667
3668
3669
3670
3671
3672
3673
3674
3675
3676
3677
3678
3679
3680
3681
3682
3683
3684
3685
3686
3687
3688
3689
3690
3691
3692
3693
3694
3695
3696
3697
3698
3699
3700
3701
3702
3703
3704
3705
3706
3707
3708
3709
3710
3711
3712
3713
3714
3715
3716
3717
3718
3719
3720
3721
3722
3723
3724
3725
3726
3727
3728
3729
3730
3731
3732
3733
3734
3735
3736
3737
3738
3739
3740
3741
3742
3743
3744
3745
3746
3747
3748
3749
3750
3751
3752
3753
3754
3755
3756
3757
3758
3759
3760
3761
3762
3763
3764
3765
3766
3767
3768
3769
3770
3771
3772
3773
3774
3775
3776
3777
3778
3779
3780
3781
3782
3783
3784
3785
3786
3787
3788
3789
3790
3791
3792
3793
3794
3795
3796
3797
3798
3799
3800
3801
3802
3803
3804
3805
3806
3807
3808
3809
3810
3811
3812
3813
3814
3815
3816
3817
3818
3819
3820
3821
3822
3823
3824
3825
3826
3827
3828
3829
3830
3831
3832
3833
3834
3835
3836
3837
3838
3839
3840
3841
3842
3843
3844
3845
3846
3847
3848
3849
3850
3851
3852
3853
3854
3855
3856
3857
3858
3859
3860
3861
3862
3863
3864
3865
3866
3867
3868
3869
3870
3871
3872
3873
3874
3875
3876
3877
3878
3879
3880
3881
3882
3883
3884
3885
3886
3887
3888
3889
3890
3891
3892
3893
3894
3895
3896
3897
3898
3899
3900
3901
3902
3903
3904
3905
3906
3907
3908
3909
3910
3911
3912
3913
3914
3915
3916
3917
3918
3919
3920
3921
3922
3923
3924
3925
3926
3927
3928
3929
3930
3931
3932
3933
3934
3935
3936
3937
3938
3939
3940
3941
3942
3943
3944
3945
3946
3947
3948
3949
3950
3951
3952
3953
3954
3955
3956
3957
3958
3959
3960
3961
3962
3963
3964
3965
3966
3967
3968
3969
3970
3971
3972
3973
3974
3975
3976
3977
3978
3979
3980
3981
3982
3983
3984
3985
3986
3987
3988
3989
3990
3991
3992
3993
3994
3995
3996
3997
3998
3999
4000
4001
4002
4003
4004
4005
4006
4007
4008
4009
4010
4011
4012
4013
4014
4015
4016
4017
4018
4019
4020
4021
4022
4023
4024
4025
4026
4027
4028
4029
4030
4031
4032
4033
4034
4035
4036
4037
4038
4039
4040
4041
4042
4043
4044
4045
4046
4047
4048
4049
4050
4051
4052
4053
4054
4055
4056
4057
4058
4059
4060
4061
4062
4063
4064
4065
4066
4067
4068
4069
4070
4071
4072
4073
4074
4075
4076
4077
4078
4079
4080
4081
4082
4083
4084
4085
4086
4087
4088
4089
4090
4091
4092
4093
4094
4095
4096
4097
4098
4099
4100
4101
4102
4103
4104
4105
4106
4107
4108
4109
4110
4111
4112
4113
4114
4115
4116
4117
4118
4119
4120
4121
4122
4123
4124
4125
4126
4127
4128
4129
4130
4131
4132
4133
4134
4135
4136
4137
4138
4139
4140
4141
4142
4143
4144
4145
4146
4147
4148
4149
4150
4151
4152
4153
4154
4155
4156
4157
4158
4159
4160
4161
4162
4163
4164
4165
4166
4167
4168
4169
4170
4171
4172
4173
4174
4175
4176
4177
4178
4179
4180
4181
4182
4183
4184
4185
4186
4187
4188
4189
4190
4191
4192
4193
4194
4195
4196
4197
4198
4199
4200
4201
4202
4203
4204
4205
4206
4207
4208
4209
4210
4211
4212
4213
4214
4215
4216
4217
4218
4219
4220
4221
4222
4223
4224
4225
4226
4227
4228
4229
4230
4231
4232
4233
4234
4235
4236
4237
4238
4239
4240
4241
4242
4243
4244
4245
4246
4247
4248
4249
4250
4251
4252
4253
4254
4255
4256
4257
4258
4259
4260
4261
4262
4263
4264
4265
4266
4267
4268
4269
4270
4271
4272
4273
4274
4275
4276
4277
4278
4279
4280
4281
4282
4283
4284
4285
4286
4287
4288
4289
4290
4291
4292
4293
4294
4295
4296
4297
4298
4299
4300
4301
4302
4303
4304
4305
4306
4307
4308
4309
4310
4311
4312
4313
4314
4315
4316
4317
4318
4319
4320
4321
4322
4323
4324
4325
4326
4327
4328
4329
4330
4331
4332
4333
4334
4335
4336
4337
4338
4339
4340
4341
4342
4343
4344
4345
4346
4347
4348
4349
4350
4351
4352
4353
4354
4355
4356
4357
4358
4359
4360
4361
4362
4363
4364
4365
4366
4367
4368
4369
4370
4371
4372
4373
4374
4375
4376
4377
4378
4379
4380
4381
4382
4383
4384
4385
4386
4387
4388
4389
4390
4391
4392
4393
4394
4395
4396
4397
4398
4399
4400
4401
4402
4403
4404
4405
4406
4407
4408
4409
4410
4411
4412
4413
4414
4415
4416
4417
4418
4419
4420
4421
4422
4423
4424
4425
4426
4427
4428
4429
4430
4431
4432
4433
4434
4435
4436
4437
4438
4439
4440
4441
4442
4443
4444
4445
4446
4447
4448
4449
4450
4451
4452
4453
4454
4455
4456
4457
4458
4459
4460
4461
4462
4463
4464
4465
4466
4467
4468
4469
4470
4471
4472
4473
4474
4475
4476
4477
4478
4479
4480
4481
4482
4483
4484
4485
4486
4487
4488
4489
4490
4491
4492
4493
4494
4495
4496
4497
4498
4499
4500
4501
4502
4503
4504
4505
4506
4507
4508
4509
4510
4511
4512
4513
4514
4515
4516
4517
4518
4519
4520
4521
4522
4523
4524
4525
4526
4527
4528
4529
4530
4531
4532
4533
4534
4535
4536
4537
4538
4539
4540
4541
4542
4543
4544
4545
4546
4547
4548
4549
4550
4551
4552
4553
4554
4555
4556
4557
4558
4559
4560
4561
4562
4563
4564
4565
4566
4567
4568
4569
4570
4571
4572
4573
4574
4575
4576
4577
4578
4579
4580
4581
4582
4583
4584
4585
4586
4587
4588
4589
4590
4591
4592
4593
4594
4595
4596
4597
4598
4599
4600
4601
4602
4603
4604
4605
4606
4607
4608
4609
4610
4611
4612
4613
4614
4615
4616
4617
4618
4619
4620
4621
4622
4623
4624
4625
4626
4627
4628
4629
4630
4631
4632
4633
4634
4635
4636
4637
4638
4639
4640
4641
4642
4643
4644
4645
4646
4647
4648
4649
4650
4651
4652
4653
4654
4655
4656
4657
4658
4659
4660
4661
4662
4663
4664
4665
4666
4667
4668
4669
4670
4671
4672
4673
4674
4675
4676
4677
4678
4679
4680
4681
4682
4683
4684
4685
4686
4687
4688
4689
4690
4691
4692
4693
4694
4695
4696
4697
4698
4699
4700
4701
4702
4703
4704
4705
4706
4707
4708
4709
4710
4711
4712
4713
4714
4715
4716
4717
4718
4719
4720
4721
4722
4723
4724
4725
4726
4727
4728
4729
4730
4731
4732
4733
4734
4735
4736
4737
4738
4739
4740
4741
4742
4743
4744
4745
4746
4747
4748
4749
4750
4751
4752
4753
4754
4755
4756
4757
4758
4759
4760
4761
4762
4763
4764
4765
4766
4767
4768
4769
4770
4771
4772
4773
4774
4775
4776
4777
4778
4779
4780
4781
4782
4783
4784
4785
4786
4787
4788
4789
4790
4791
4792
4793
4794
4795
4796
4797
4798
4799
4800
4801
4802
4803
4804
4805
4806
4807
4808
4809
4810
4811
4812
4813
4814
4815
4816
4817
4818
4819
4820
4821
4822
4823
4824
4825
4826
4827
4828
4829
4830
4831
4832
4833
4834
4835
4836
4837
4838
4839
4840
4841
4842
4843
4844
4845
4846
4847
4848
4849
4850
4851
4852
4853
4854
4855
4856
4857
4858
4859
4860
4861
4862
4863
4864
4865
4866
4867
4868
4869
4870
4871
4872
4873
4874
4875
4876
4877
4878
4879
4880
4881
4882
4883
4884
4885
4886
4887
4888
4889
4890
4891
4892
4893
4894
4895
4896
4897
4898
4899
4900
4901
4902
4903
4904
4905
4906
4907
4908
4909
4910
4911
4912
4913
4914
4915
4916
4917
4918
4919
4920
4921
4922
4923
4924
4925
4926
4927
4928
4929
4930
4931
4932
4933
4934
4935
4936
4937
4938
4939
4940
4941
4942
4943
4944
4945
4946
4947
4948
4949
4950
4951
4952
4953
4954
4955
4956
4957
4958
4959
4960
4961
4962
4963
4964
4965
4966
4967
4968
4969
4970
4971
4972
4973
4974
4975
4976
4977
4978
4979
4980
4981
4982
4983
4984
4985
4986
4987
4988
4989
4990
4991
4992
4993
4994
4995
4996
4997
4998
4999
5000
5001
5002
5003
5004
5005
5006
5007
5008
5009
5010
5011
5012
5013
5014
5015
5016
5017
5018
5019
5020
5021
5022
5023
5024
5025
5026
5027
5028
5029
5030
5031
5032
5033
5034
5035
5036
5037
5038
5039
5040
5041
5042
5043
5044
5045
5046
5047
5048
5049
5050
5051
5052
5053
5054
5055
5056
5057
5058
5059
5060
5061
5062
5063
5064
5065
5066
5067
5068
5069
5070
5071
5072
5073
5074
5075
5076
5077
5078
5079
5080
5081
5082
5083
5084
5085
5086
5087
5088
5089
5090
5091
5092
5093
5094
5095
5096
5097
5098
5099
5100
5101
5102
5103
5104
5105
5106
5107
5108
5109
5110
5111
5112
5113
5114
5115
5116
5117
5118
5119
5120
5121
5122
5123
5124
5125
5126
5127
5128
5129
5130
5131
5132
5133
5134
5135
5136
5137
5138
5139
5140
5141
5142
5143
5144
5145
5146
5147
5148
5149
5150
5151
5152
5153
5154
5155
5156
5157
5158
5159
5160
5161
5162
5163
5164
5165
5166
5167
5168
5169
5170
5171
5172
5173
5174
5175
5176
5177
5178
5179
5180
5181
5182
5183
5184
5185
5186
5187
5188
5189
5190
5191
5192
5193
5194
5195
5196
5197
5198
5199
5200
5201
5202
5203
5204
5205
5206
5207
5208
5209
5210
5211
5212
5213
5214
5215
5216
5217
5218
5219
5220
5221
5222
5223
5224
5225
5226
5227
5228
5229
5230
5231
5232
5233
5234
5235
5236
5237
5238
5239
5240
5241
5242
5243
5244
5245
5246
5247
5248
5249
5250
5251
5252
5253
5254
5255
5256
5257
5258
5259
5260
5261
5262
5263
5264
5265
5266
5267
5268
5269
5270
5271
5272
5273
5274
5275
5276
5277
5278
5279
5280
5281
5282
5283
5284
5285
5286
5287
5288
5289
5290
5291
5292
5293
5294
5295
5296
5297
5298
5299
5300
5301
5302
5303
5304
5305
5306
5307
5308
5309
5310
5311
5312
5313
5314
5315
5316
5317
5318
5319
5320
5321
5322
5323
5324
5325
5326
5327
5328
5329
5330
5331
5332
5333
5334
5335
5336
5337
5338
5339
5340
5341
5342
5343
5344
5345
5346
5347
5348
5349
5350
5351
5352
5353
5354
5355
5356
5357
5358
5359
5360
5361
5362
5363
5364
5365
5366
5367
5368
5369
5370
5371
5372
5373
5374
5375
5376
5377
5378
5379
5380
5381
5382
5383
5384
5385
5386
5387
5388
5389
5390
5391
5392
5393
5394
5395
5396
5397
5398
5399
5400
5401
5402
5403
5404
5405
5406
5407
5408
5409
5410
5411
5412
5413
5414
5415
5416
5417
5418
5419
5420
5421
5422
5423
5424
5425
5426
5427
5428
5429
5430
5431
5432
5433
5434
5435
5436
5437
5438
5439
5440
5441
5442
5443
5444
5445
5446
5447
5448
5449
5450
5451
5452
5453
5454
5455
5456
5457
5458
5459
5460
5461
5462
5463
5464
5465
5466
5467
5468
5469
5470
5471
5472
5473
5474
5475
5476
5477
5478
5479
5480
5481
5482
5483
5484
5485
5486
5487
5488
5489
5490
5491
5492
5493
5494
5495
5496
5497
5498
5499
5500
5501
5502
5503
5504
5505
5506
5507
5508
5509
5510
5511
5512
5513
5514
5515
5516
5517
5518
5519
5520
5521
5522
5523
5524
5525
5526
5527
5528
5529
5530
5531
5532
5533
5534
5535
5536
5537
5538
5539
5540
5541
5542
5543
5544
5545
5546
5547
5548
5549
5550
5551
5552
5553
5554
5555
5556
5557
5558
5559
5560
5561
5562
5563
5564
5565
5566
5567
5568
5569
5570
5571
5572
5573
5574
5575
5576
5577
5578
5579
5580
5581
5582
5583
5584
5585
5586
5587
5588
5589
5590
5591
5592
5593
5594
5595
5596
5597
5598
5599
5600
5601
5602
5603
5604
5605
5606
5607
5608
5609
5610
5611
5612
5613
5614
5615
5616
5617
5618
5619
5620
5621
5622
5623
5624
5625
5626
5627
5628
5629
5630
5631
5632
5633
5634
5635
5636
5637
5638
5639
5640
5641
5642
5643
5644
5645
5646
5647
5648
5649
5650
5651
5652
5653
5654
5655
5656
5657
5658
5659
5660
5661
5662
5663
5664
5665
5666
5667
5668
5669
5670
5671
5672
5673
5674
5675
5676
5677
5678
5679
5680
5681
5682
5683
5684
5685
5686
5687
5688
5689
5690
5691
5692
5693
5694
5695
5696
5697
5698
5699
5700
5701
5702
5703
5704
5705
5706
5707
5708
5709
5710
5711
5712
5713
5714
5715
5716
5717
5718
5719
5720
5721
5722
5723
5724
5725
5726
5727
5728
5729
5730
5731
5732
5733
5734
5735
5736
5737
5738
5739
5740
5741
5742
5743
5744
5745
5746
5747
5748
5749
5750
5751
5752
5753
5754
5755
5756
5757
5758
5759
5760
5761
5762
5763
5764
5765
5766
5767
5768
5769
5770
5771
5772
5773
5774
5775
5776
5777
5778
5779
5780
5781
5782
5783
5784
5785
5786
5787
5788
5789
5790
5791
5792
5793
5794
5795
5796
5797
5798
5799
5800
5801
5802
5803
5804
5805
5806
5807
5808
5809
5810
5811
5812
5813
5814
5815
5816
5817
5818
5819
5820
5821
5822
5823
5824
5825
5826
5827
5828
5829
5830
5831
5832
5833
5834
5835
5836
5837
5838
5839
5840
5841
5842
5843
5844
5845
5846
5847
5848
5849
5850
5851
5852
5853
5854
5855
5856
5857
5858
5859
5860
5861
5862
5863
5864
5865
5866
5867
5868
5869
5870
5871
5872
5873
5874
5875
5876
5877
5878
5879
5880
5881
5882
5883
5884
5885
5886
5887
5888
5889
5890
5891
5892
5893
5894
5895
5896
5897
5898
5899
5900
5901
5902
5903
5904
5905
5906
5907
5908
5909
5910
5911
5912
5913
5914
5915
5916
5917
5918
5919
5920
5921
5922
5923
5924
5925
5926
5927
5928
5929
5930
5931
5932
5933
5934
5935
5936
5937
5938
5939
5940
5941
5942
5943
5944
5945
5946
5947
5948
5949
5950
5951
5952
5953
5954
5955
5956
5957
5958
5959
5960
5961
5962
5963
5964
5965
5966
5967
5968
5969
5970
5971
5972
5973
5974
5975
5976
5977
5978
5979
5980
5981
5982
5983
5984
5985
5986
5987
5988
5989
5990
5991
5992
5993
5994
5995
5996
5997
5998
5999
6000
6001
6002
6003
6004
6005
6006
6007
6008
6009
6010
6011
6012
6013
6014
6015
6016
6017
6018
6019
6020
6021
6022
6023
6024
6025
6026
6027
6028
6029
6030
6031
6032
6033
6034
6035
6036
6037
6038
6039
6040
6041
6042
6043
6044
6045
6046
6047
6048
6049
6050
6051
6052
6053
6054
6055
6056
6057
6058
6059
6060
6061
6062
6063
6064
6065
6066
6067
6068
6069
6070
6071
6072
6073
6074
6075
6076
6077
6078
6079
6080
6081
6082
6083
6084
6085
6086
6087
6088
6089
6090
6091
6092
6093
6094
6095
6096
6097
6098
6099
6100
6101
6102
6103
6104
6105
6106
6107
6108
6109
6110
6111
6112
6113
6114
6115
6116
6117
6118
6119
6120
6121
6122
6123
6124
6125
6126
6127
6128
6129
6130
6131
6132
6133
6134
6135
6136
6137
6138
6139
6140
6141
6142
6143
6144
6145
6146
6147
6148
6149
6150
6151
6152
6153
6154
6155
6156
6157
6158
6159
6160
6161
6162
6163
6164
6165
6166
6167
6168
6169
6170
6171
6172
6173
6174
6175
6176
6177
6178
6179
6180
6181
6182
6183
6184
6185
6186
6187
6188
6189
6190
6191
6192
6193
6194
6195
6196
6197
6198
6199
6200
6201
6202
6203
6204
6205
6206
6207
6208
6209
6210
6211
6212
6213
6214
6215
6216
6217
6218
6219
6220
6221
6222
6223
6224
6225
6226
6227
6228
6229
6230
6231
6232
6233
6234
6235
6236
6237
6238
6239
6240
6241
6242
6243
6244
6245
6246
6247
6248
6249
6250
6251
6252
6253
6254
6255
6256
6257
6258
6259
6260
6261
6262
6263
6264
6265
6266
6267
6268
6269
6270
6271
6272
6273
6274
6275
6276
6277
6278
6279
6280
6281
6282
6283
6284
6285
6286
6287
6288
6289
6290
6291
6292
6293
6294
6295
6296
6297
6298
6299
6300
6301
6302
6303
6304
6305
6306
6307
6308
6309
6310
6311
6312
6313
6314
6315
6316
6317
6318
6319
6320
6321
6322
6323
6324
6325
6326
6327
6328
6329
6330
6331
6332
6333
6334
6335
6336
6337
6338
6339
6340
6341
6342
6343
6344
6345
6346
6347
6348
6349
6350
6351
6352
6353
6354
6355
6356
6357
6358
6359
6360
6361
6362
6363
6364
6365
6366
6367
6368
6369
6370
6371
6372
6373
6374
6375
6376
6377
6378
6379
6380
6381
6382
6383
6384
6385
6386
6387
6388
6389
6390
6391
6392
6393
6394
6395
6396
6397
6398
6399
6400
6401
6402
6403
6404
6405
6406
6407
6408
6409
6410
6411
6412
6413
6414
6415
6416
6417
6418
6419
6420
6421
6422
6423
6424
6425
6426
6427
6428
6429
6430
6431
6432
6433
6434
6435
6436
6437
6438
6439
6440
6441
6442
6443
6444
6445
6446
6447
6448
6449
6450
6451
6452
6453
6454
6455
6456
6457
6458
6459
6460
6461
6462
6463
6464
6465
6466
6467
6468
6469
6470
6471
6472
6473
6474
6475
6476
6477
6478
6479
6480
6481
6482
6483
6484
6485
6486
6487
6488
6489
6490
6491
6492
6493
6494
6495
6496
6497
6498
6499
6500
6501
6502
6503
6504
6505
6506
6507
6508
6509
6510
6511
6512
6513
6514
6515
6516
6517
6518
6519
6520
6521
6522
6523
6524
6525
6526
6527
6528
6529
6530
6531
6532
6533
6534
6535
6536
6537
6538
6539
6540
6541
6542
6543
6544
6545
6546
6547
6548
6549
6550
6551
6552
6553
6554
6555
6556
6557
6558
6559
6560
6561
6562
6563
6564
6565
6566
6567
6568
6569
6570
6571
6572
6573
6574
6575
6576
6577
6578
6579
6580
6581
6582
6583
6584
6585
6586
6587
6588
6589
6590
6591
6592
6593
6594
6595
6596
6597
6598
6599
6600
6601
6602
6603
6604
6605
6606
6607
6608
6609
6610
6611
6612
6613
6614
6615
6616
6617
6618
6619
6620
6621
6622
6623
6624
6625
6626
6627
6628
6629
6630
6631
6632
6633
6634
6635
6636
6637
6638
6639
6640
6641
6642
6643
6644
6645
6646
6647
6648
6649
6650
6651
6652
6653
6654
6655
6656
6657
6658
6659
6660
6661
6662
6663
6664
6665
6666
6667
6668
6669
6670
6671
6672
6673
6674
6675
6676
6677
6678
6679
6680
6681
6682
6683
6684
6685
6686
6687
6688
6689
6690
6691
6692
6693
6694
6695
6696
6697
6698
6699
6700
6701
6702
6703
6704
6705
6706
6707
6708
6709
6710
6711
6712
6713
6714
6715
6716
6717
6718
6719
6720
6721
6722
6723
6724
6725
6726
6727
6728
6729
6730
6731
6732
6733
6734
6735
6736
6737
6738
6739
6740
6741
6742
6743
6744
6745
6746
6747
6748
6749
6750
6751
6752
6753
6754
6755
6756
6757
6758
6759
6760
6761
6762
6763
6764
6765
6766
6767
6768
6769
6770
6771
6772
6773
6774
6775
6776
6777
6778
6779
6780
6781
6782
6783
6784
6785
6786
6787
6788
6789
6790
6791
6792
6793
6794
6795
6796
6797
6798
6799
6800
6801
6802
6803
6804
6805
6806
6807
6808
6809
6810
6811
6812
6813
6814
6815
6816
6817
6818
6819
6820
6821
6822
6823
6824
6825
6826
6827
6828
6829
6830
6831
6832
6833
6834
6835
6836
6837
6838
6839
6840
6841
6842
6843
6844
6845
6846
6847
6848
6849
6850
6851
6852
6853
6854
6855
6856
6857
6858
6859
6860
6861
6862
6863
6864
6865
6866
6867
6868
6869
6870
6871
6872
6873
6874
6875
6876
6877
6878
6879
6880
6881
6882
6883
6884
6885
6886
6887
6888
6889
6890
6891
6892
6893
6894
6895
6896
6897
6898
6899
6900
6901
6902
6903
6904
6905
6906
6907
6908
6909
6910
6911
6912
6913
6914
6915
6916
6917
6918
6919
6920
6921
6922
6923
6924
6925
6926
6927
6928
6929
6930
6931
6932
6933
6934
6935
6936
6937
6938
6939
6940
6941
6942
6943
6944
6945
6946
6947
6948
6949
6950
6951
6952
6953
6954
6955
6956
6957
6958
6959
6960
6961
6962
6963
6964
6965
6966
6967
6968
6969
6970
6971
6972
6973
6974
6975
6976
6977
6978
6979
6980
6981
6982
6983
6984
6985
6986
6987
6988
6989
6990
6991
6992
6993
6994
6995
6996
6997
6998
6999
7000
7001
7002
7003
7004
7005
7006
7007
7008
7009
7010
7011
7012
7013
7014
7015
7016
7017
7018
7019
7020
7021
7022
7023
7024
7025
7026
7027
7028
7029
7030
7031
7032
7033
7034
7035
7036
7037
7038
7039
7040
7041
7042
7043
7044
7045
7046
7047
7048
7049
7050
7051
7052
7053
7054
7055
7056
7057
7058
7059
7060
7061
7062
7063
7064
7065
7066
7067
7068
7069
7070
7071
7072
7073
7074
7075
7076
7077
7078
7079
7080
7081
7082
7083
7084
7085
7086
7087
7088
7089
7090
7091
7092
7093
7094
7095
7096
7097
7098
7099
7100
7101
7102
7103
7104
7105
7106
7107
7108
7109
7110
7111
7112
7113
7114
7115
7116
7117
7118
7119
7120
7121
7122
7123
7124
7125
7126
7127
7128
7129
7130
7131
7132
7133
7134
7135
7136
7137
7138
7139
7140
7141
7142
7143
7144
7145
7146
7147
7148
7149
7150
7151
7152
7153
7154
7155
7156
7157
7158
7159
7160
7161
7162
7163
7164
7165
7166
7167
7168
7169
7170
7171
7172
7173
7174
7175
7176
7177
7178
7179
7180
7181
7182
7183
7184
7185
7186
7187
7188
7189
7190
7191
7192
7193
7194
7195
7196
7197
7198
7199
7200
7201
7202
7203
7204
7205
7206
7207
7208
7209
7210
7211
7212
7213
7214
7215
7216
7217
7218
7219
7220
7221
7222
7223
7224
7225
7226
7227
7228
7229
7230
7231
7232
7233
7234
7235
7236
7237
7238
7239
7240
7241
7242
7243
7244
7245
7246
7247
7248
7249
7250
7251
7252
7253
7254
7255
7256
7257
7258
7259
7260
7261
7262
7263
7264
7265
7266
7267
7268
7269
7270
7271
7272
7273
7274
7275
7276
7277
7278
7279
7280
7281
7282
7283
7284
7285
7286
7287
7288
7289
7290
7291
7292
7293
7294
7295
7296
7297
7298
7299
7300
7301
7302
7303
7304
7305
7306
7307
7308
7309
7310
7311
7312
7313
7314
7315
7316
7317
7318
7319
7320
7321
7322
7323
7324
7325
7326
7327
7328
7329
7330
7331
7332
7333
7334
7335
7336
7337
7338
7339
7340
7341
7342
7343
7344
7345
7346
7347
7348
7349
7350
7351
7352
7353
7354
7355
7356
7357
7358
7359
7360
7361
7362
7363
7364
7365
7366
7367
7368
7369
7370
7371
7372
7373
7374
7375
7376
7377
7378
7379
7380
7381
7382
7383
7384
7385
7386
7387
7388
7389
7390
7391
7392
7393
7394
7395
7396
7397
7398
7399
7400
7401
7402
7403
7404
7405
7406
7407
7408
7409
7410
7411
7412
7413
7414
7415
7416
7417
7418
7419
7420
7421
7422
7423
7424
7425
7426
7427
7428
7429
7430
7431
7432
7433
7434
7435
7436
7437
7438
7439
7440
7441
7442
7443
7444
7445
7446
7447
7448
7449
7450
7451
7452
7453
7454
7455
7456
7457
7458
7459
7460
7461
7462
7463
7464
7465
7466
7467
7468
7469
7470
7471
7472
7473
7474
7475
7476
7477
7478
7479
7480
7481
7482
7483
7484
7485
7486
7487
7488
7489
7490
7491
7492
7493
7494
7495
7496
7497
7498
7499
7500
7501
7502
7503
7504
7505
7506
7507
7508
7509
7510
7511
7512
7513
7514
7515
7516
7517
7518
7519
7520
7521
7522
7523
7524
7525
7526
7527
7528
7529
7530
7531
7532
7533
7534
7535
7536
7537
7538
7539
7540
7541
7542
7543
7544
7545
7546
7547
7548
7549
7550
7551
7552
7553
7554
7555
7556
7557
7558
7559
7560
7561
7562
7563
7564
7565
7566
7567
7568
7569
7570
7571
7572
7573
7574
7575
7576
7577
7578
7579
7580
7581
7582
7583
7584
7585
7586
7587
7588
7589
7590
7591
7592
7593
7594
7595
7596
7597
7598
7599
7600
7601
7602
7603
7604
7605
7606
7607
7608
7609
7610
7611
7612
7613
7614
7615
7616
7617
7618
7619
7620
7621
7622
7623
7624
7625
7626
7627
7628
7629
7630
7631
7632
7633
7634
7635
7636
7637
7638
7639
7640
7641
7642
7643
7644
7645
7646
7647
7648
7649
7650
7651
7652
7653
7654
7655
7656
7657
7658
7659
7660
7661
7662
7663
7664
7665
7666
7667
7668
7669
7670
7671
7672
7673
7674
7675
7676
7677
7678
7679
7680
7681
7682
7683
7684
7685
7686
7687
7688
7689
7690
7691
7692
7693
7694
7695
7696
7697
7698
7699
7700
7701
7702
7703
7704
7705
7706
7707
7708
7709
7710
7711
7712
7713
7714
7715
7716
7717
7718
7719
7720
7721
7722
7723
7724
7725
7726
7727
7728
7729
7730
7731
7732
7733
7734
7735
7736
7737
7738
7739
7740
7741
7742
7743
7744
7745
7746
7747
7748
7749
7750
7751
7752
7753
7754
7755
7756
7757
7758
7759
7760
7761
7762
7763
7764
7765
7766
7767
7768
7769
7770
7771
7772
7773
7774
7775
7776
7777
7778
7779
7780
7781
7782
7783
7784
7785
7786
7787
7788
7789
7790
7791
7792
7793
7794
7795
7796
7797
7798
7799
7800
/*
** 2001 September 15
**
** The author disclaims copyright to this source code.  In place of
** a legal notice, here is a blessing:
**
**    May you do good and not evil.
**    May you find forgiveness for yourself and forgive others.
**    May you share freely, never taking more than you give.
**
*************************************************************************
** This is the implementation of the page cache subsystem or "pager".
**
** The pager is used to access a database disk file.  It implements
** atomic commit and rollback through the use of a journal file that
** is separate from the database file.  The pager also implements file
** locking to prevent two processes from writing the same database
** file simultaneously, or one process from reading the database while
** another is writing.
*/
#ifndef SQLITE_OMIT_DISKIO
#include "sqliteInt.h"
#include "wal.h"


/******************* NOTES ON THE DESIGN OF THE PAGER ************************
**
** This comment block describes invariants that hold when using a rollback
** journal.  These invariants do not apply for journal_mode=WAL,
** journal_mode=MEMORY, or journal_mode=OFF.
**
** Within this comment block, a page is deemed to have been synced
** automatically as soon as it is written when PRAGMA synchronous=OFF.
** Otherwise, the page is not synced until the xSync method of the VFS
** is called successfully on the file containing the page.
**
** Definition:  A page of the database file is said to be "overwriteable" if
** one or more of the following are true about the page:
**
**     (a)  The original content of the page as it was at the beginning of
**          the transaction has been written into the rollback journal and
**          synced.
**
**     (b)  The page was a freelist leaf page at the start of the transaction.
**
**     (c)  The page number is greater than the largest page that existed in
**          the database file at the start of the transaction.
**
** (1) A page of the database file is never overwritten unless one of the
**     following are true:
**
**     (a) The page and all other pages on the same sector are overwriteable.
**
**     (b) The atomic page write optimization is enabled, and the entire
**         transaction other than the update of the transaction sequence
**         number consists of a single page change.
**
** (2) The content of a page written into the rollback journal exactly matches
**     both the content in the database when the rollback journal was written
**     and the content in the database at the beginning of the current
**     transaction.
**
** (3) Writes to the database file are an integer multiple of the page size
**     in length and are aligned on a page boundary.
**
** (4) Reads from the database file are either aligned on a page boundary and
**     an integer multiple of the page size in length or are taken from the
**     first 100 bytes of the database file.
**
** (5) All writes to the database file are synced prior to the rollback journal
**     being deleted, truncated, or zeroed.
**
** (6) If a super-journal file is used, then all writes to the database file
**     are synced prior to the super-journal being deleted.
**
** Definition: Two databases (or the same database at two points it time)
** are said to be "logically equivalent" if they give the same answer to
** all queries.  Note in particular the content of freelist leaf
** pages can be changed arbitrarily without affecting the logical equivalence
** of the database.
**
** (7) At any time, if any subset, including the empty set and the total set,
**     of the unsynced changes to a rollback journal are removed and the
**     journal is rolled back, the resulting database file will be logically
**     equivalent to the database file at the beginning of the transaction.
**
** (8) When a transaction is rolled back, the xTruncate method of the VFS
**     is called to restore the database file to the same size it was at
**     the beginning of the transaction.  (In some VFSes, the xTruncate
**     method is a no-op, but that does not change the fact the SQLite will
**     invoke it.)
**
** (9) Whenever the database file is modified, at least one bit in the range
**     of bytes from 24 through 39 inclusive will be changed prior to releasing
**     the EXCLUSIVE lock, thus signaling other connections on the same
**     database to flush their caches.
**
** (10) The pattern of bits in bytes 24 through 39 shall not repeat in less
**      than one billion transactions.
**
** (11) A database file is well-formed at the beginning and at the conclusion
**      of every transaction.
**
** (12) An EXCLUSIVE lock is held on the database file when writing to
**      the database file.
**
** (13) A SHARED lock is held on the database file while reading any
**      content out of the database file.
**
******************************************************************************/

/*
** Macros for troubleshooting.  Normally turned off
*/
#if 0
int sqlite3PagerTrace=1;  /* True to enable tracing */
#define sqlite3DebugPrintf printf
#define PAGERTRACE(X)     if( sqlite3PagerTrace ){ sqlite3DebugPrintf X; }
#else
#define PAGERTRACE(X)
#endif

/*
** The following two macros are used within the PAGERTRACE() macros above
** to print out file-descriptors.
**
** PAGERID() takes a pointer to a Pager struct as its argument. The
** associated file-descriptor is returned. FILEHANDLEID() takes an sqlite3_file
** struct as its argument.
*/
#define PAGERID(p) (SQLITE_PTR_TO_INT(p->fd))
#define FILEHANDLEID(fd) (SQLITE_PTR_TO_INT(fd))

/*
** The Pager.eState variable stores the current 'state' of a pager. A
** pager may be in any one of the seven states shown in the following
** state diagram.
**
**                            OPEN <------+------+
**                              |         |      |
**                              V         |      |
**               +---------> READER-------+      |
**               |              |                |
**               |              V                |
**               |<-------WRITER_LOCKED------> ERROR
**               |              |                ^ 
**               |              V                |
**               |<------WRITER_CACHEMOD-------->|
**               |              |                |
**               |              V                |
**               |<-------WRITER_DBMOD---------->|
**               |              |                |
**               |              V                |
**               +<------WRITER_FINISHED-------->+
**
**
** List of state transitions and the C [function] that performs each:
**
**   OPEN              -> READER              [sqlite3PagerSharedLock]
**   READER            -> OPEN                [pager_unlock]
**
**   READER            -> WRITER_LOCKED       [sqlite3PagerBegin]
**   WRITER_LOCKED     -> WRITER_CACHEMOD     [pager_open_journal]
**   WRITER_CACHEMOD   -> WRITER_DBMOD        [syncJournal]
**   WRITER_DBMOD      -> WRITER_FINISHED     [sqlite3PagerCommitPhaseOne]
**   WRITER_***        -> READER              [pager_end_transaction]
**
**   WRITER_***        -> ERROR               [pager_error]
**   ERROR             -> OPEN                [pager_unlock]
**
**
**  OPEN:
**
**    The pager starts up in this state. Nothing is guaranteed in this
**    state - the file may or may not be locked and the database size is
**    unknown. The database may not be read or written.
**
**    * No read or write transaction is active.
**    * Any lock, or no lock at all, may be held on the database file.
**    * The dbSize, dbOrigSize and dbFileSize variables may not be trusted.
**
**  READER:
**
**    In this state all the requirements for reading the database in
**    rollback (non-WAL) mode are met. Unless the pager is (or recently
**    was) in exclusive-locking mode, a user-level read transaction is
**    open. The database size is known in this state.
**
**    A connection running with locking_mode=normal enters this state when
**    it opens a read-transaction on the database and returns to state
**    OPEN after the read-transaction is completed. However a connection
**    running in locking_mode=exclusive (including temp databases) remains in
**    this state even after the read-transaction is closed. The only way
**    a locking_mode=exclusive connection can transition from READER to OPEN
**    is via the ERROR state (see below).
**
**    * A read transaction may be active (but a write-transaction cannot).
**    * A SHARED or greater lock is held on the database file.
**    * The dbSize variable may be trusted (even if a user-level read
**      transaction is not active). The dbOrigSize and dbFileSize variables
**      may not be trusted at this point.
**    * If the database is a WAL database, then the WAL connection is open.
**    * Even if a read-transaction is not open, it is guaranteed that
**      there is no hot-journal in the file-system.
**
**  WRITER_LOCKED:
**
**    The pager moves to this state from READER when a write-transaction
**    is first opened on the database. In WRITER_LOCKED state, all locks
**    required to start a write-transaction are held, but no actual
**    modifications to the cache or database have taken place.
**
**    In rollback mode, a RESERVED or (if the transaction was opened with
**    BEGIN EXCLUSIVE) EXCLUSIVE lock is obtained on the database file when
**    moving to this state, but the journal file is not written to or opened
**    to in this state. If the transaction is committed or rolled back while
**    in WRITER_LOCKED state, all that is required is to unlock the database
**    file.
**
**    IN WAL mode, WalBeginWriteTransaction() is called to lock the log file.
**    If the connection is running with locking_mode=exclusive, an attempt
**    is made to obtain an EXCLUSIVE lock on the database file.
**
**    * A write transaction is active.
**    * If the connection is open in rollback-mode, a RESERVED or greater
**      lock is held on the database file.
**    * If the connection is open in WAL-mode, a WAL write transaction
**      is open (i.e. sqlite3WalBeginWriteTransaction() has been successfully
**      called).
**    * The dbSize, dbOrigSize and dbFileSize variables are all valid.
**    * The contents of the pager cache have not been modified.
**    * The journal file may or may not be open.
**    * Nothing (not even the first header) has been written to the journal.
**
**  WRITER_CACHEMOD:
**
**    A pager moves from WRITER_LOCKED state to this state when a page is
**    first modified by the upper layer. In rollback mode the journal file
**    is opened (if it is not already open) and a header written to the
**    start of it. The database file on disk has not been modified.
**
**    * A write transaction is active.
**    * A RESERVED or greater lock is held on the database file.
**    * The journal file is open and the first header has been written
**      to it, but the header has not been synced to disk.
**    * The contents of the page cache have been modified.
**
**  WRITER_DBMOD:
**
**    The pager transitions from WRITER_CACHEMOD into WRITER_DBMOD state
**    when it modifies the contents of the database file. WAL connections
**    never enter this state (since they do not modify the database file,
**    just the log file).
**
**    * A write transaction is active.
**    * An EXCLUSIVE or greater lock is held on the database file.
**    * The journal file is open and the first header has been written
**      and synced to disk.
**    * The contents of the page cache have been modified (and possibly
**      written to disk).
**
**  WRITER_FINISHED:
**
**    It is not possible for a WAL connection to enter this state.
**
**    A rollback-mode pager changes to WRITER_FINISHED state from WRITER_DBMOD
**    state after the entire transaction has been successfully written into the
**    database file. In this state the transaction may be committed simply
**    by finalizing the journal file. Once in WRITER_FINISHED state, it is
**    not possible to modify the database further. At this point, the upper
**    layer must either commit or rollback the transaction.
**
**    * A write transaction is active.
**    * An EXCLUSIVE or greater lock is held on the database file.
**    * All writing and syncing of journal and database data has finished.
**      If no error occurred, all that remains is to finalize the journal to
**      commit the transaction. If an error did occur, the caller will need
**      to rollback the transaction.
**
**  ERROR:
**
**    The ERROR state is entered when an IO or disk-full error (including
**    SQLITE_IOERR_NOMEM) occurs at a point in the code that makes it
**    difficult to be sure that the in-memory pager state (cache contents,
**    db size etc.) are consistent with the contents of the file-system.
**
**    Temporary pager files may enter the ERROR state, but in-memory pagers
**    cannot.
**
**    For example, if an IO error occurs while performing a rollback,
**    the contents of the page-cache may be left in an inconsistent state.
**    At this point it would be dangerous to change back to READER state
**    (as usually happens after a rollback). Any subsequent readers might
**    report database corruption (due to the inconsistent cache), and if
**    they upgrade to writers, they may inadvertently corrupt the database
**    file. To avoid this hazard, the pager switches into the ERROR state
**    instead of READER following such an error.
**
**    Once it has entered the ERROR state, any attempt to use the pager
**    to read or write data returns an error. Eventually, once all
**    outstanding transactions have been abandoned, the pager is able to
**    transition back to OPEN state, discarding the contents of the
**    page-cache and any other in-memory state at the same time. Everything
**    is reloaded from disk (and, if necessary, hot-journal rollback performed)
**    when a read-transaction is next opened on the pager (transitioning
**    the pager into READER state). At that point the system has recovered
**    from the error.
**
**    Specifically, the pager jumps into the ERROR state if:
**
**      1. An error occurs while attempting a rollback. This happens in
**         function sqlite3PagerRollback().
**
**      2. An error occurs while attempting to finalize a journal file
**         following a commit in function sqlite3PagerCommitPhaseTwo().
**
**      3. An error occurs while attempting to write to the journal or
**         database file in function pagerStress() in order to free up
**         memory.
**
**    In other cases, the error is returned to the b-tree layer. The b-tree
**    layer then attempts a rollback operation. If the error condition
**    persists, the pager enters the ERROR state via condition (1) above.
**
**    Condition (3) is necessary because it can be triggered by a read-only
**    statement executed within a transaction. In this case, if the error
**    code were simply returned to the user, the b-tree layer would not
**    automatically attempt a rollback, as it assumes that an error in a
**    read-only statement cannot leave the pager in an internally inconsistent
**    state.
**
**    * The Pager.errCode variable is set to something other than SQLITE_OK.
**    * There are one or more outstanding references to pages (after the
**      last reference is dropped the pager should move back to OPEN state).
**    * The pager is not an in-memory pager.
**   
**
** Notes:
**
**   * A pager is never in WRITER_DBMOD or WRITER_FINISHED state if the
**     connection is open in WAL mode. A WAL connection is always in one
**     of the first four states.
**
**   * Normally, a connection open in exclusive mode is never in PAGER_OPEN
**     state. There are two exceptions: immediately after exclusive-mode has
**     been turned on (and before any read or write transactions are
**     executed), and when the pager is leaving the "error state".
**
**   * See also: assert_pager_state().
*/
#define PAGER_OPEN                  0
#define PAGER_READER                1
#define PAGER_WRITER_LOCKED         2
#define PAGER_WRITER_CACHEMOD       3
#define PAGER_WRITER_DBMOD          4
#define PAGER_WRITER_FINISHED       5
#define PAGER_ERROR                 6

/*
** The Pager.eLock variable is almost always set to one of the
** following locking-states, according to the lock currently held on
** the database file: NO_LOCK, SHARED_LOCK, RESERVED_LOCK or EXCLUSIVE_LOCK.
** This variable is kept up to date as locks are taken and released by
** the pagerLockDb() and pagerUnlockDb() wrappers.
**
** If the VFS xLock() or xUnlock() returns an error other than SQLITE_BUSY
** (i.e. one of the SQLITE_IOERR subtypes), it is not clear whether or not
** the operation was successful. In these circumstances pagerLockDb() and
** pagerUnlockDb() take a conservative approach - eLock is always updated
** when unlocking the file, and only updated when locking the file if the
** VFS call is successful. This way, the Pager.eLock variable may be set
** to a less exclusive (lower) value than the lock that is actually held
** at the system level, but it is never set to a more exclusive value.
**
** This is usually safe. If an xUnlock fails or appears to fail, there may
** be a few redundant xLock() calls or a lock may be held for longer than
** required, but nothing really goes wrong.
**
** The exception is when the database file is unlocked as the pager moves
** from ERROR to OPEN state. At this point there may be a hot-journal file
** in the file-system that needs to be rolled back (as part of an OPEN->SHARED
** transition, by the same pager or any other). If the call to xUnlock()
** fails at this point and the pager is left holding an EXCLUSIVE lock, this
** can confuse the call to xCheckReservedLock() call made later as part
** of hot-journal detection.
**
** xCheckReservedLock() is defined as returning true "if there is a RESERVED
** lock held by this process or any others". So xCheckReservedLock may
** return true because the caller itself is holding an EXCLUSIVE lock (but
** doesn't know it because of a previous error in xUnlock). If this happens
** a hot-journal may be mistaken for a journal being created by an active
** transaction in another process, causing SQLite to read from the database
** without rolling it back.
**
** To work around this, if a call to xUnlock() fails when unlocking the
** database in the ERROR state, Pager.eLock is set to UNKNOWN_LOCK. It
** is only changed back to a real locking state after a successful call
** to xLock(EXCLUSIVE). Also, the code to do the OPEN->SHARED state transition
** omits the check for a hot-journal if Pager.eLock is set to UNKNOWN_LOCK
** lock. Instead, it assumes a hot-journal exists and obtains an EXCLUSIVE
** lock on the database file before attempting to roll it back. See function
** PagerSharedLock() for more detail.
**
** Pager.eLock may only be set to UNKNOWN_LOCK when the pager is in
** PAGER_OPEN state.
*/
#define UNKNOWN_LOCK                (EXCLUSIVE_LOCK+1)

/*
** The maximum allowed sector size. 64KiB. If the xSectorsize() method
** returns a value larger than this, then MAX_SECTOR_SIZE is used instead.
** This could conceivably cause corruption following a power failure on
** such a system. This is currently an undocumented limit.
*/
#define MAX_SECTOR_SIZE 0x10000


/*
** An instance of the following structure is allocated for each active
** savepoint and statement transaction in the system. All such structures
** are stored in the Pager.aSavepoint[] array, which is allocated and
** resized using sqlite3Realloc().
**
** When a savepoint is created, the PagerSavepoint.iHdrOffset field is
** set to 0. If a journal-header is written into the main journal while
** the savepoint is active, then iHdrOffset is set to the byte offset
** immediately following the last journal record written into the main
** journal before the journal-header. This is required during savepoint
** rollback (see pagerPlaybackSavepoint()).
*/
typedef struct PagerSavepoint PagerSavepoint;
struct PagerSavepoint {
  i64 iOffset;                 /* Starting offset in main journal */
  i64 iHdrOffset;              /* See above */
  Bitvec *pInSavepoint;        /* Set of pages in this savepoint */
  Pgno nOrig;                  /* Original number of pages in file */
  Pgno iSubRec;                /* Index of first record in sub-journal */
  int bTruncateOnRelease;      /* If stmt journal may be truncated on RELEASE */
#ifndef SQLITE_OMIT_WAL
  u32 aWalData[WAL_SAVEPOINT_NDATA];        /* WAL savepoint context */
#endif
};

/*
** Bits of the Pager.doNotSpill flag.  See further description below.
*/
#define SPILLFLAG_OFF         0x01 /* Never spill cache.  Set via pragma */
#define SPILLFLAG_ROLLBACK    0x02 /* Current rolling back, so do not spill */
#define SPILLFLAG_NOSYNC      0x04 /* Spill is ok, but do not sync */

/*
** An open page cache is an instance of struct Pager. A description of
** some of the more important member variables follows:
**
** eState
**
**   The current 'state' of the pager object. See the comment and state
**   diagram above for a description of the pager state.
**
** eLock
**
**   For a real on-disk database, the current lock held on the database file -
**   NO_LOCK, SHARED_LOCK, RESERVED_LOCK or EXCLUSIVE_LOCK.
**
**   For a temporary or in-memory database (neither of which require any
**   locks), this variable is always set to EXCLUSIVE_LOCK. Since such
**   databases always have Pager.exclusiveMode==1, this tricks the pager
**   logic into thinking that it already has all the locks it will ever
**   need (and no reason to release them).
**
**   In some (obscure) circumstances, this variable may also be set to
**   UNKNOWN_LOCK. See the comment above the #define of UNKNOWN_LOCK for
**   details.
**
** changeCountDone
**
**   This boolean variable is used to make sure that the change-counter
**   (the 4-byte header field at byte offset 24 of the database file) is
**   not updated more often than necessary.
**
**   It is set to true when the change-counter field is updated, which
**   can only happen if an exclusive lock is held on the database file.
**   It is cleared (set to false) whenever an exclusive lock is
**   relinquished on the database file. Each time a transaction is committed,
**   The changeCountDone flag is inspected. If it is true, the work of
**   updating the change-counter is omitted for the current transaction.
**
**   This mechanism means that when running in exclusive mode, a connection
**   need only update the change-counter once, for the first transaction
**   committed.
**
** setSuper
**
**   When PagerCommitPhaseOne() is called to commit a transaction, it may
**   (or may not) specify a super-journal name to be written into the
**   journal file before it is synced to disk.
**
**   Whether or not a journal file contains a super-journal pointer affects
**   the way in which the journal file is finalized after the transaction is
**   committed or rolled back when running in "journal_mode=PERSIST" mode.
**   If a journal file does not contain a super-journal pointer, it is
**   finalized by overwriting the first journal header with zeroes. If
**   it does contain a super-journal pointer the journal file is finalized
**   by truncating it to zero bytes, just as if the connection were
**   running in "journal_mode=truncate" mode.
**
**   Journal files that contain super-journal pointers cannot be finalized
**   simply by overwriting the first journal-header with zeroes, as the
**   super-journal pointer could interfere with hot-journal rollback of any
**   subsequently interrupted transaction that reuses the journal file.
**
**   The flag is cleared as soon as the journal file is finalized (either
**   by PagerCommitPhaseTwo or PagerRollback). If an IO error prevents the
**   journal file from being successfully finalized, the setSuper flag
**   is cleared anyway (and the pager will move to ERROR state).
**
** doNotSpill
**
**   This variables control the behavior of cache-spills  (calls made by
**   the pcache module to the pagerStress() routine to write cached data
**   to the file-system in order to free up memory).
**
**   When bits SPILLFLAG_OFF or SPILLFLAG_ROLLBACK of doNotSpill are set,
**   writing to the database from pagerStress() is disabled altogether.
**   The SPILLFLAG_ROLLBACK case is done in a very obscure case that
**   comes up during savepoint rollback that requires the pcache module
**   to allocate a new page to prevent the journal file from being written
**   while it is being traversed by code in pager_playback().  The SPILLFLAG_OFF
**   case is a user preference.
**
**   If the SPILLFLAG_NOSYNC bit is set, writing to the database from
**   pagerStress() is permitted, but syncing the journal file is not.
**   This flag is set by sqlite3PagerWrite() when the file-system sector-size
**   is larger than the database page-size in order to prevent a journal sync
**   from happening in between the journalling of two pages on the same sector.
**
** subjInMemory
**
**   This is a boolean variable. If true, then any required sub-journal
**   is opened as an in-memory journal file. If false, then in-memory
**   sub-journals are only used for in-memory pager files.
**
**   This variable is updated by the upper layer each time a new
**   write-transaction is opened.
**
** dbSize, dbOrigSize, dbFileSize
**
**   Variable dbSize is set to the number of pages in the database file.
**   It is valid in PAGER_READER and higher states (all states except for
**   OPEN and ERROR).
**
**   dbSize is set based on the size of the database file, which may be
**   larger than the size of the database (the value stored at offset
**   28 of the database header by the btree). If the size of the file
**   is not an integer multiple of the page-size, the value stored in
**   dbSize is rounded down (i.e. a 5KB file with 2K page-size has dbSize==2).
**   Except, any file that is greater than 0 bytes in size is considered
**   to have at least one page. (i.e. a 1KB file with 2K page-size leads
**   to dbSize==1).
**
**   During a write-transaction, if pages with page-numbers greater than
**   dbSize are modified in the cache, dbSize is updated accordingly.
**   Similarly, if the database is truncated using PagerTruncateImage(),
**   dbSize is updated.
**
**   Variables dbOrigSize and dbFileSize are valid in states
**   PAGER_WRITER_LOCKED and higher. dbOrigSize is a copy of the dbSize
**   variable at the start of the transaction. It is used during rollback,
**   and to determine whether or not pages need to be journalled before
**   being modified.
**
**   Throughout a write-transaction, dbFileSize contains the size of
**   the file on disk in pages. It is set to a copy of dbSize when the
**   write-transaction is first opened, and updated when VFS calls are made
**   to write or truncate the database file on disk.
**
**   The only reason the dbFileSize variable is required is to suppress
**   unnecessary calls to xTruncate() after committing a transaction. If,
**   when a transaction is committed, the dbFileSize variable indicates
**   that the database file is larger than the database image (Pager.dbSize),
**   pager_truncate() is called. The pager_truncate() call uses xFilesize()
**   to measure the database file on disk, and then truncates it if required.
**   dbFileSize is not used when rolling back a transaction. In this case
**   pager_truncate() is called unconditionally (which means there may be
**   a call to xFilesize() that is not strictly required). In either case,
**   pager_truncate() may cause the file to become smaller or larger.
**
** dbHintSize
**
**   The dbHintSize variable is used to limit the number of calls made to
**   the VFS xFileControl(FCNTL_SIZE_HINT) method.
**
**   dbHintSize is set to a copy of the dbSize variable when a
**   write-transaction is opened (at the same time as dbFileSize and
**   dbOrigSize). If the xFileControl(FCNTL_SIZE_HINT) method is called,
**   dbHintSize is increased to the number of pages that correspond to the
**   size-hint passed to the method call. See pager_write_pagelist() for
**   details.
**
** errCode
**
**   The Pager.errCode variable is only ever used in PAGER_ERROR state. It
**   is set to zero in all other states. In PAGER_ERROR state, Pager.errCode
**   is always set to SQLITE_FULL, SQLITE_IOERR or one of the SQLITE_IOERR_XXX
**   sub-codes.
**
** syncFlags, walSyncFlags
**
**   syncFlags is either SQLITE_SYNC_NORMAL (0x02) or SQLITE_SYNC_FULL (0x03).
**   syncFlags is used for rollback mode.  walSyncFlags is used for WAL mode
**   and contains the flags used to sync the checkpoint operations in the
**   lower two bits, and sync flags used for transaction commits in the WAL
**   file in bits 0x04 and 0x08.  In other words, to get the correct sync flags
**   for checkpoint operations, use (walSyncFlags&0x03) and to get the correct
**   sync flags for transaction commit, use ((walSyncFlags>>2)&0x03).  Note
**   that with synchronous=NORMAL in WAL mode, transaction commit is not synced
**   meaning that the 0x04 and 0x08 bits are both zero.
*/
struct Pager {
  sqlite3_vfs *pVfs;          /* OS functions to use for IO */
  u8 exclusiveMode;           /* Boolean. True if locking_mode==EXCLUSIVE */
  u8 journalMode;             /* One of the PAGER_JOURNALMODE_* values */
  u8 useJournal;              /* Use a rollback journal on this file */
  u8 noSync;                  /* Do not sync the journal if true */
  u8 fullSync;                /* Do extra syncs of the journal for robustness */
  u8 extraSync;               /* sync directory after journal delete */
  u8 syncFlags;               /* SYNC_NORMAL or SYNC_FULL otherwise */
  u8 walSyncFlags;            /* See description above */
  u8 tempFile;                /* zFilename is a temporary or immutable file */
  u8 noLock;                  /* Do not lock (except in WAL mode) */
  u8 readOnly;                /* True for a read-only database */
  u8 memDb;                   /* True to inhibit all file I/O */
  u8 memVfs;                  /* VFS-implemented memory database */

  /**************************************************************************
  ** The following block contains those class members that change during
  ** routine operation.  Class members not in this block are either fixed
  ** when the pager is first created or else only change when there is a
  ** significant mode change (such as changing the page_size, locking_mode,
  ** or the journal_mode).  From another view, these class members describe
  ** the "state" of the pager, while other class members describe the
  ** "configuration" of the pager.
  */
  u8 eState;                  /* Pager state (OPEN, READER, WRITER_LOCKED..) */
  u8 eLock;                   /* Current lock held on database file */
  u8 changeCountDone;         /* Set after incrementing the change-counter */
  u8 setSuper;                /* Super-jrnl name is written into jrnl */
  u8 doNotSpill;              /* Do not spill the cache when non-zero */
  u8 subjInMemory;            /* True to use in-memory sub-journals */
  u8 bUseFetch;               /* True to use xFetch() */
  u8 hasHeldSharedLock;       /* True if a shared lock has ever been held */
  Pgno dbSize;                /* Number of pages in the database */
  Pgno dbOrigSize;            /* dbSize before the current transaction */
  Pgno dbFileSize;            /* Number of pages in the database file */
  Pgno dbHintSize;            /* Value passed to FCNTL_SIZE_HINT call */
  int errCode;                /* One of several kinds of errors */
  int nRec;                   /* Pages journalled since last j-header written */
  u32 cksumInit;              /* Quasi-random value added to every checksum */
  u32 nSubRec;                /* Number of records written to sub-journal */
  Bitvec *pInJournal;         /* One bit for each page in the database file */
  sqlite3_file *fd;           /* File descriptor for database */
  sqlite3_file *jfd;          /* File descriptor for main journal */
  sqlite3_file *sjfd;         /* File descriptor for sub-journal */
  i64 journalOff;             /* Current write offset in the journal file */
  i64 journalHdr;             /* Byte offset to previous journal header */
  sqlite3_backup *pBackup;    /* Pointer to list of ongoing backup processes */
  PagerSavepoint *aSavepoint; /* Array of active savepoints */
  int nSavepoint;             /* Number of elements in aSavepoint[] */
  u32 iDataVersion;           /* Changes whenever database content changes */
  char dbFileVers[16];        /* Changes whenever database file changes */

  int nMmapOut;               /* Number of mmap pages currently outstanding */
  sqlite3_int64 szMmap;       /* Desired maximum mmap size */
  PgHdr *pMmapFreelist;       /* List of free mmap page headers (pDirty) */
  /*
  ** End of the routinely-changing class members
  ***************************************************************************/

  u16 nExtra;                 /* Add this many bytes to each in-memory page */
  i16 nReserve;               /* Number of unused bytes at end of each page */
  u32 vfsFlags;               /* Flags for sqlite3_vfs.xOpen() */
  u32 sectorSize;             /* Assumed sector size during rollback */
  Pgno mxPgno;                /* Maximum allowed size of the database */
  Pgno lckPgno;               /* Page number for the locking page */
  i64 pageSize;               /* Number of bytes in a page */
  i64 journalSizeLimit;       /* Size limit for persistent journal files */
  char *zFilename;            /* Name of the database file */
  char *zJournal;             /* Name of the journal file */
  int (*xBusyHandler)(void*); /* Function to call when busy */
  void *pBusyHandlerArg;      /* Context argument for xBusyHandler */
  u32 aStat[4];               /* Total cache hits, misses, writes, spills */
#ifdef SQLITE_TEST
  int nRead;                  /* Database pages read */
#endif
  void (*xReiniter)(DbPage*); /* Call this routine when reloading pages */
  int (*xGet)(Pager*,Pgno,DbPage**,int); /* Routine to fetch a patch */
  char *pTmpSpace;            /* Pager.pageSize bytes of space for tmp use */
  PCache *pPCache;            /* Pointer to page cache object */
#ifndef SQLITE_OMIT_WAL
  Wal *pWal;                  /* Write-ahead log used by "journal_mode=wal" */
  char *zWal;                 /* File name for write-ahead log */
#endif
};

/*
** Indexes for use with Pager.aStat[]. The Pager.aStat[] array contains
** the values accessed by passing SQLITE_DBSTATUS_CACHE_HIT, CACHE_MISS
** or CACHE_WRITE to sqlite3_db_status().
*/
#define PAGER_STAT_HIT   0
#define PAGER_STAT_MISS  1
#define PAGER_STAT_WRITE 2
#define PAGER_STAT_SPILL 3

/*
** The following global variables hold counters used for
** testing purposes only.  These variables do not exist in
** a non-testing build.  These variables are not thread-safe.
*/
#ifdef SQLITE_TEST
int sqlite3_pager_readdb_count = 0;    /* Number of full pages read from DB */
int sqlite3_pager_writedb_count = 0;   /* Number of full pages written to DB */
int sqlite3_pager_writej_count = 0;    /* Number of pages written to journal */
# define PAGER_INCR(v)  v++
#else
# define PAGER_INCR(v)
#endif



/*
** Journal files begin with the following magic string.  The data
** was obtained from /dev/random.  It is used only as a sanity check.
**
** Since version 2.8.0, the journal format contains additional sanity
** checking information.  If the power fails while the journal is being
** written, semi-random garbage data might appear in the journal
** file after power is restored.  If an attempt is then made
** to roll the journal back, the database could be corrupted.  The additional
** sanity checking data is an attempt to discover the garbage in the
** journal and ignore it.
**
** The sanity checking information for the new journal format consists
** of a 32-bit checksum on each page of data.  The checksum covers both
** the page number and the pPager->pageSize bytes of data for the page.
** This cksum is initialized to a 32-bit random value that appears in the
** journal file right after the header.  The random initializer is important,
** because garbage data that appears at the end of a journal is likely
** data that was once in other files that have now been deleted.  If the
** garbage data came from an obsolete journal file, the checksums might
** be correct.  But by initializing the checksum to random value which
** is different for every journal, we minimize that risk.
*/
static const unsigned char aJournalMagic[] = {
  0xd9, 0xd5, 0x05, 0xf9, 0x20, 0xa1, 0x63, 0xd7,
};

/*
** The size of the of each page record in the journal is given by
** the following macro.
*/
#define JOURNAL_PG_SZ(pPager)  ((pPager->pageSize) + 8)

/*
** The journal header size for this pager. This is usually the same
** size as a single disk sector. See also setSectorSize().
*/
#define JOURNAL_HDR_SZ(pPager) (pPager->sectorSize)

/*
** The macro MEMDB is true if we are dealing with an in-memory database.
** We do this as a macro so that if the SQLITE_OMIT_MEMORYDB macro is set,
** the value of MEMDB will be a constant and the compiler will optimize
** out code that would never execute.
*/
#ifdef SQLITE_OMIT_MEMORYDB
# define MEMDB 0
#else
# define MEMDB pPager->memDb
#endif

/*
** The macro USEFETCH is true if we are allowed to use the xFetch and xUnfetch
** interfaces to access the database using memory-mapped I/O.
*/
#if SQLITE_MAX_MMAP_SIZE>0
# define USEFETCH(x) ((x)->bUseFetch)
#else
# define USEFETCH(x) 0
#endif

/*
** The argument to this macro is a file descriptor (type sqlite3_file*).
** Return 0 if it is not open, or non-zero (but not 1) if it is.
**
** This is so that expressions can be written as:
**
**   if( isOpen(pPager->jfd) ){ ...
**
** instead of
**
**   if( pPager->jfd->pMethods ){ ...
*/
#define isOpen(pFd) ((pFd)->pMethods!=0)

#ifdef SQLITE_DIRECT_OVERFLOW_READ
/*
** Return true if page pgno can be read directly from the database file
** by the b-tree layer. This is the case if:
**
**   * the database file is open,
**   * there are no dirty pages in the cache, and
**   * the desired page is not currently in the wal file.
*/
int sqlite3PagerDirectReadOk(Pager *pPager, Pgno pgno){
  if( pPager->fd->pMethods==0 ) return 0;
  if( sqlite3PCacheIsDirty(pPager->pPCache) ) return 0;
#ifndef SQLITE_OMIT_WAL
  if( pPager->pWal ){
    u32 iRead = 0;
    (void)sqlite3WalFindFrame(pPager->pWal, pgno, &iRead);
    return iRead==0;
  }
#endif
  return 1;
}
#endif

#ifndef SQLITE_OMIT_WAL
# define pagerUseWal(x) ((x)->pWal!=0)
#else
# define pagerUseWal(x) 0
# define pagerRollbackWal(x) 0
# define pagerWalFrames(v,w,x,y) 0
# define pagerOpenWalIfPresent(z) SQLITE_OK
# define pagerBeginReadTransaction(z) SQLITE_OK
#endif

#ifndef NDEBUG
/*
** Usage:
**
**   assert( assert_pager_state(pPager) );
**
** This function runs many asserts to try to find inconsistencies in
** the internal state of the Pager object.
*/
static int assert_pager_state(Pager *p){
  Pager *pPager = p;

  /* State must be valid. */
  assert( p->eState==PAGER_OPEN
       || p->eState==PAGER_READER
       || p->eState==PAGER_WRITER_LOCKED
       || p->eState==PAGER_WRITER_CACHEMOD
       || p->eState==PAGER_WRITER_DBMOD
       || p->eState==PAGER_WRITER_FINISHED
       || p->eState==PAGER_ERROR
  );

  /* Regardless of the current state, a temp-file connection always behaves
  ** as if it has an exclusive lock on the database file. It never updates
  ** the change-counter field, so the changeCountDone flag is always set.
  */
  assert( p->tempFile==0 || p->eLock==EXCLUSIVE_LOCK );
  assert( p->tempFile==0 || pPager->changeCountDone );

  /* If the useJournal flag is clear, the journal-mode must be "OFF".
  ** And if the journal-mode is "OFF", the journal file must not be open.
  */
  assert( p->journalMode==PAGER_JOURNALMODE_OFF || p->useJournal );
  assert( p->journalMode!=PAGER_JOURNALMODE_OFF || !isOpen(p->jfd) );

  /* Check that MEMDB implies noSync. And an in-memory journal. Since
  ** this means an in-memory pager performs no IO at all, it cannot encounter
  ** either SQLITE_IOERR or SQLITE_FULL during rollback or while finalizing
  ** a journal file. (although the in-memory journal implementation may
  ** return SQLITE_IOERR_NOMEM while the journal file is being written). It
  ** is therefore not possible for an in-memory pager to enter the ERROR
  ** state.
  */
  if( MEMDB ){
    assert( !isOpen(p->fd) );
    assert( p->noSync );
    assert( p->journalMode==PAGER_JOURNALMODE_OFF
         || p->journalMode==PAGER_JOURNALMODE_MEMORY
    );
    assert( p->eState!=PAGER_ERROR && p->eState!=PAGER_OPEN );
    assert( pagerUseWal(p)==0 );
  }

  /* If changeCountDone is set, a RESERVED lock or greater must be held
  ** on the file.
  */
  assert( pPager->changeCountDone==0 || pPager->eLock>=RESERVED_LOCK );
  assert( p->eLock!=PENDING_LOCK );

  switch( p->eState ){
    case PAGER_OPEN:
      assert( !MEMDB );
      assert( pPager->errCode==SQLITE_OK );
      assert( sqlite3PcacheRefCount(pPager->pPCache)==0 || pPager->tempFile );
      break;

    case PAGER_READER:
      assert( pPager->errCode==SQLITE_OK );
      assert( p->eLock!=UNKNOWN_LOCK );
      assert( p->eLock>=SHARED_LOCK );
      break;

    case PAGER_WRITER_LOCKED:
      assert( p->eLock!=UNKNOWN_LOCK );
      assert( pPager->errCode==SQLITE_OK );
      if( !pagerUseWal(pPager) ){
        assert( p->eLock>=RESERVED_LOCK );
      }
      assert( pPager->dbSize==pPager->dbOrigSize );
      assert( pPager->dbOrigSize==pPager->dbFileSize );
      assert( pPager->dbOrigSize==pPager->dbHintSize );
      assert( pPager->setSuper==0 );
      break;

    case PAGER_WRITER_CACHEMOD:
      assert( p->eLock!=UNKNOWN_LOCK );
      assert( pPager->errCode==SQLITE_OK );
      if( !pagerUseWal(pPager) ){
        /* It is possible that if journal_mode=wal here that neither the
        ** journal file nor the WAL file are open. This happens during
        ** a rollback transaction that switches from journal_mode=off
        ** to journal_mode=wal.
        */
        assert( p->eLock>=RESERVED_LOCK );
        assert( isOpen(p->jfd)
             || p->journalMode==PAGER_JOURNALMODE_OFF
             || p->journalMode==PAGER_JOURNALMODE_WAL
        );
      }
      assert( pPager->dbOrigSize==pPager->dbFileSize );
      assert( pPager->dbOrigSize==pPager->dbHintSize );
      break;

    case PAGER_WRITER_DBMOD:
      assert( p->eLock==EXCLUSIVE_LOCK );
      assert( pPager->errCode==SQLITE_OK );
      assert( !pagerUseWal(pPager) );
      assert( p->eLock>=EXCLUSIVE_LOCK );
      assert( isOpen(p->jfd)
           || p->journalMode==PAGER_JOURNALMODE_OFF
           || p->journalMode==PAGER_JOURNALMODE_WAL
           || (sqlite3OsDeviceCharacteristics(p->fd)&SQLITE_IOCAP_BATCH_ATOMIC)
      );
      assert( pPager->dbOrigSize<=pPager->dbHintSize );
      break;

    case PAGER_WRITER_FINISHED:
      assert( p->eLock==EXCLUSIVE_LOCK );
      assert( pPager->errCode==SQLITE_OK );
      assert( !pagerUseWal(pPager) );
      assert( isOpen(p->jfd)
           || p->journalMode==PAGER_JOURNALMODE_OFF
           || p->journalMode==PAGER_JOURNALMODE_WAL
           || (sqlite3OsDeviceCharacteristics(p->fd)&SQLITE_IOCAP_BATCH_ATOMIC)
      );
      break;

    case PAGER_ERROR:
      /* There must be at least one outstanding reference to the pager if
      ** in ERROR state. Otherwise the pager should have already dropped
      ** back to OPEN state.
      */
      assert( pPager->errCode!=SQLITE_OK );
      assert( sqlite3PcacheRefCount(pPager->pPCache)>0 || pPager->tempFile );
      break;
  }

  return 1;
}
#endif /* ifndef NDEBUG */

#ifdef SQLITE_DEBUG
/*
** Return a pointer to a human readable string in a static buffer
** containing the state of the Pager object passed as an argument. This
** is intended to be used within debuggers. For example, as an alternative
** to "print *pPager" in gdb:
**
** (gdb) printf "%s", print_pager_state(pPager)
**
** This routine has external linkage in order to suppress compiler warnings
** about an unused function.  It is enclosed within SQLITE_DEBUG and so does
** not appear in normal builds.
*/
char *print_pager_state(Pager *p){
  static char zRet[1024];

  sqlite3_snprintf(1024, zRet,
      "Filename:      %s\n"
      "State:         %s errCode=%d\n"
      "Lock:          %s\n"
      "Locking mode:  locking_mode=%s\n"
      "Journal mode:  journal_mode=%s\n"
      "Backing store: tempFile=%d memDb=%d useJournal=%d\n"
      "Journal:       journalOff=%lld journalHdr=%lld\n"
      "Size:          dbsize=%d dbOrigSize=%d dbFileSize=%d\n"
      , p->zFilename
      , p->eState==PAGER_OPEN            ? "OPEN" :
        p->eState==PAGER_READER          ? "READER" :
        p->eState==PAGER_WRITER_LOCKED   ? "WRITER_LOCKED" :
        p->eState==PAGER_WRITER_CACHEMOD ? "WRITER_CACHEMOD" :
        p->eState==PAGER_WRITER_DBMOD    ? "WRITER_DBMOD" :
        p->eState==PAGER_WRITER_FINISHED ? "WRITER_FINISHED" :
        p->eState==PAGER_ERROR           ? "ERROR" : "?error?"
      , (int)p->errCode
      , p->eLock==NO_LOCK         ? "NO_LOCK" :
        p->eLock==RESERVED_LOCK   ? "RESERVED" :
        p->eLock==EXCLUSIVE_LOCK  ? "EXCLUSIVE" :
        p->eLock==SHARED_LOCK     ? "SHARED" :
        p->eLock==UNKNOWN_LOCK    ? "UNKNOWN" : "?error?"
      , p->exclusiveMode ? "exclusive" : "normal"
      , p->journalMode==PAGER_JOURNALMODE_MEMORY   ? "memory" :
        p->journalMode==PAGER_JOURNALMODE_OFF      ? "off" :
        p->journalMode==PAGER_JOURNALMODE_DELETE   ? "delete" :
        p->journalMode==PAGER_JOURNALMODE_PERSIST  ? "persist" :
        p->journalMode==PAGER_JOURNALMODE_TRUNCATE ? "truncate" :
        p->journalMode==PAGER_JOURNALMODE_WAL      ? "wal" : "?error?"
      , (int)p->tempFile, (int)p->memDb, (int)p->useJournal
      , p->journalOff, p->journalHdr
      , (int)p->dbSize, (int)p->dbOrigSize, (int)p->dbFileSize
  );

  return zRet;
}
#endif

/* Forward references to the various page getters */
static int getPageNormal(Pager*,Pgno,DbPage**,int);
static int getPageError(Pager*,Pgno,DbPage**,int);
#if SQLITE_MAX_MMAP_SIZE>0
static int getPageMMap(Pager*,Pgno,DbPage**,int);
#endif

/*
** Set the Pager.xGet method for the appropriate routine used to fetch
** content from the pager.
*/
static void setGetterMethod(Pager *pPager){
  if( pPager->errCode ){
    pPager->xGet = getPageError;
#if SQLITE_MAX_MMAP_SIZE>0
  }else if( USEFETCH(pPager) ){
    pPager->xGet = getPageMMap;
#endif /* SQLITE_MAX_MMAP_SIZE>0 */
  }else{
    pPager->xGet = getPageNormal;
  }
}

/*
** Return true if it is necessary to write page *pPg into the sub-journal.
** A page needs to be written into the sub-journal if there exists one
** or more open savepoints for which:
**
**   * The page-number is less than or equal to PagerSavepoint.nOrig, and
**   * The bit corresponding to the page-number is not set in
**     PagerSavepoint.pInSavepoint.
*/
static int subjRequiresPage(PgHdr *pPg){
  Pager *pPager = pPg->pPager;
  PagerSavepoint *p;
  Pgno pgno = pPg->pgno;
  int i;
  for(i=0; i<pPager->nSavepoint; i++){
    p = &pPager->aSavepoint[i];
    if( p->nOrig>=pgno && 0==sqlite3BitvecTestNotNull(p->pInSavepoint, pgno) ){
      for(i=i+1; i<pPager->nSavepoint; i++){
        pPager->aSavepoint[i].bTruncateOnRelease = 0;
      }
      return 1;
    }
  }
  return 0;
}

#ifdef SQLITE_DEBUG
/*
** Return true if the page is already in the journal file.
*/
static int pageInJournal(Pager *pPager, PgHdr *pPg){
  return sqlite3BitvecTest(pPager->pInJournal, pPg->pgno);
}
#endif

/*
** Read a 32-bit integer from the given file descriptor.  Store the integer
** that is read in *pRes.  Return SQLITE_OK if everything worked, or an
** error code is something goes wrong.
**
** All values are stored on disk as big-endian.
*/
static int read32bits(sqlite3_file *fd, i64 offset, u32 *pRes){
  unsigned char ac[4];
  int rc = sqlite3OsRead(fd, ac, sizeof(ac), offset);
  if( rc==SQLITE_OK ){
    *pRes = sqlite3Get4byte(ac);
  }
  return rc;
}

/*
** Write a 32-bit integer into a string buffer in big-endian byte order.
*/
#define put32bits(A,B)  sqlite3Put4byte((u8*)A,B)


/*
** Write a 32-bit integer into the given file descriptor.  Return SQLITE_OK
** on success or an error code is something goes wrong.
*/
static int write32bits(sqlite3_file *fd, i64 offset, u32 val){
  char ac[4];
  put32bits(ac, val);
  return sqlite3OsWrite(fd, ac, 4, offset);
}

/*
** Unlock the database file to level eLock, which must be either NO_LOCK
** or SHARED_LOCK. Regardless of whether or not the call to xUnlock()
** succeeds, set the Pager.eLock variable to match the (attempted) new lock.
**
** Except, if Pager.eLock is set to UNKNOWN_LOCK when this function is
** called, do not modify it. See the comment above the #define of
** UNKNOWN_LOCK for an explanation of this.
*/
static int pagerUnlockDb(Pager *pPager, int eLock){
  int rc = SQLITE_OK;

  assert( !pPager->exclusiveMode || pPager->eLock==eLock );
  assert( eLock==NO_LOCK || eLock==SHARED_LOCK );
  assert( eLock!=NO_LOCK || pagerUseWal(pPager)==0 );
  if( isOpen(pPager->fd) ){
    assert( pPager->eLock>=eLock );
    rc = pPager->noLock ? SQLITE_OK : sqlite3OsUnlock(pPager->fd, eLock);
    if( pPager->eLock!=UNKNOWN_LOCK ){
      pPager->eLock = (u8)eLock;
    }
    IOTRACE(("UNLOCK %p %d\n", pPager, eLock))
  }
  pPager->changeCountDone = pPager->tempFile; /* ticket fb3b3024ea238d5c */
  return rc;
}

/*
** Lock the database file to level eLock, which must be either SHARED_LOCK,
** RESERVED_LOCK or EXCLUSIVE_LOCK. If the caller is successful, set the
** Pager.eLock variable to the new locking state.
**
** Except, if Pager.eLock is set to UNKNOWN_LOCK when this function is
** called, do not modify it unless the new locking state is EXCLUSIVE_LOCK.
** See the comment above the #define of UNKNOWN_LOCK for an explanation
** of this.
*/
static int pagerLockDb(Pager *pPager, int eLock){
  int rc = SQLITE_OK;

  assert( eLock==SHARED_LOCK || eLock==RESERVED_LOCK || eLock==EXCLUSIVE_LOCK );
  if( pPager->eLock<eLock || pPager->eLock==UNKNOWN_LOCK ){
    rc = pPager->noLock ? SQLITE_OK : sqlite3OsLock(pPager->fd, eLock);
    if( rc==SQLITE_OK && (pPager->eLock!=UNKNOWN_LOCK||eLock==EXCLUSIVE_LOCK) ){
      pPager->eLock = (u8)eLock;
      IOTRACE(("LOCK %p %d\n", pPager, eLock))
    }
  }
  return rc;
}

/*
** This function determines whether or not the atomic-write or
** atomic-batch-write optimizations can be used with this pager. The
** atomic-write optimization can be used if:
**
**  (a) the value returned by OsDeviceCharacteristics() indicates that
**      a database page may be written atomically, and
**  (b) the value returned by OsSectorSize() is less than or equal
**      to the page size.
**
** If it can be used, then the value returned is the size of the journal
** file when it contains rollback data for exactly one page.
**
** The atomic-batch-write optimization can be used if OsDeviceCharacteristics()
** returns a value with the SQLITE_IOCAP_BATCH_ATOMIC bit set. -1 is
** returned in this case.
**
** If neither optimization can be used, 0 is returned.
*/
static int jrnlBufferSize(Pager *pPager){
  assert( !MEMDB );

#if defined(SQLITE_ENABLE_ATOMIC_WRITE) \
 || defined(SQLITE_ENABLE_BATCH_ATOMIC_WRITE)
  int dc;                           /* Device characteristics */

  assert( isOpen(pPager->fd) );
  dc = sqlite3OsDeviceCharacteristics(pPager->fd);
#else
  UNUSED_PARAMETER(pPager);
#endif

#ifdef SQLITE_ENABLE_BATCH_ATOMIC_WRITE
  if( pPager->dbSize>0 && (dc&SQLITE_IOCAP_BATCH_ATOMIC) ){
    return -1;
  }
#endif

#ifdef SQLITE_ENABLE_ATOMIC_WRITE
  {
    int nSector = pPager->sectorSize;
    int szPage = pPager->pageSize;

    assert(SQLITE_IOCAP_ATOMIC512==(512>>8));
    assert(SQLITE_IOCAP_ATOMIC64K==(65536>>8));
    if( 0==(dc&(SQLITE_IOCAP_ATOMIC|(szPage>>8)) || nSector>szPage) ){
      return 0;
    }
  }

  return JOURNAL_HDR_SZ(pPager) + JOURNAL_PG_SZ(pPager);
#endif

  return 0;
}

/*
** If SQLITE_CHECK_PAGES is defined then we do some sanity checking
** on the cache using a hash function.  This is used for testing
** and debugging only.
*/
#ifdef SQLITE_CHECK_PAGES
/*
** Return a 32-bit hash of the page data for pPage.
*/
static u32 pager_datahash(int nByte, unsigned char *pData){
  u32 hash = 0;
  int i;
  for(i=0; i<nByte; i++){
    hash = (hash*1039) + pData[i];
  }
  return hash;
}
static u32 pager_pagehash(PgHdr *pPage){
  return pager_datahash(pPage->pPager->pageSize, (unsigned char *)pPage->pData);
}
static void pager_set_pagehash(PgHdr *pPage){
  pPage->pageHash = pager_pagehash(pPage);
}

/*
** The CHECK_PAGE macro takes a PgHdr* as an argument. If SQLITE_CHECK_PAGES
** is defined, and NDEBUG is not defined, an assert() statement checks
** that the page is either dirty or still matches the calculated page-hash.
*/
#define CHECK_PAGE(x) checkPage(x)
static void checkPage(PgHdr *pPg){
  Pager *pPager = pPg->pPager;
  assert( pPager->eState!=PAGER_ERROR );
  assert( (pPg->flags&PGHDR_DIRTY) || pPg->pageHash==pager_pagehash(pPg) );
}

#else
#define pager_datahash(X,Y)  0
#define pager_pagehash(X)  0
#define pager_set_pagehash(X)
#define CHECK_PAGE(x)
#endif  /* SQLITE_CHECK_PAGES */

/*
** When this is called the journal file for pager pPager must be open.
** This function attempts to read a super-journal file name from the
** end of the file and, if successful, copies it into memory supplied
** by the caller. See comments above writeSuperJournal() for the format
** used to store a super-journal file name at the end of a journal file.
**
** zSuper must point to a buffer of at least nSuper bytes allocated by
** the caller. This should be sqlite3_vfs.mxPathname+1 (to ensure there is
** enough space to write the super-journal name). If the super-journal
** name in the journal is longer than nSuper bytes (including a
** nul-terminator), then this is handled as if no super-journal name
** were present in the journal.
**
** If a super-journal file name is present at the end of the journal
** file, then it is copied into the buffer pointed to by zSuper. A
** nul-terminator byte is appended to the buffer following the
** super-journal file name.
**
** If it is determined that no super-journal file name is present
** zSuper[0] is set to 0 and SQLITE_OK returned.
**
** If an error occurs while reading from the journal file, an SQLite
** error code is returned.
*/
static int readSuperJournal(sqlite3_file *pJrnl, char *zSuper, u32 nSuper){
  int rc;                    /* Return code */
  u32 len;                   /* Length in bytes of super-journal name */
  i64 szJ;                   /* Total size in bytes of journal file pJrnl */
  u32 cksum;                 /* MJ checksum value read from journal */
  u32 u;                     /* Unsigned loop counter */
  unsigned char aMagic[8];   /* A buffer to hold the magic header */
  zSuper[0] = '\0';

  if( SQLITE_OK!=(rc = sqlite3OsFileSize(pJrnl, &szJ))
   || szJ<16
   || SQLITE_OK!=(rc = read32bits(pJrnl, szJ-16, &len))
   || len>=nSuper
   || len>szJ-16
   || len==0
   || SQLITE_OK!=(rc = read32bits(pJrnl, szJ-12, &cksum))
   || SQLITE_OK!=(rc = sqlite3OsRead(pJrnl, aMagic, 8, szJ-8))
   || memcmp(aMagic, aJournalMagic, 8)
   || SQLITE_OK!=(rc = sqlite3OsRead(pJrnl, zSuper, len, szJ-16-len))
  ){
    return rc;
  }

  /* See if the checksum matches the super-journal name */
  for(u=0; u<len; u++){
    cksum -= zSuper[u];
  }
  if( cksum ){
    /* If the checksum doesn't add up, then one or more of the disk sectors
    ** containing the super-journal filename is corrupted. This means
    ** definitely roll back, so just return SQLITE_OK and report a (nul)
    ** super-journal filename.
    */
    len = 0;
  }
  zSuper[len] = '\0';
  zSuper[len+1] = '\0';
  
  return SQLITE_OK;
}

/*
** Return the offset of the sector boundary at or immediately
** following the value in pPager->journalOff, assuming a sector
** size of pPager->sectorSize bytes.
**
** i.e for a sector size of 512:
**
**   Pager.journalOff          Return value
**   ---------------------------------------
**   0                         0
**   512                       512
**   100                       512
**   2000                      2048
**
*/
static i64 journalHdrOffset(Pager *pPager){
  i64 offset = 0;
  i64 c = pPager->journalOff;
  if( c ){
    offset = ((c-1)/JOURNAL_HDR_SZ(pPager) + 1) * JOURNAL_HDR_SZ(pPager);
  }
  assert( offset%JOURNAL_HDR_SZ(pPager)==0 );
  assert( offset>=c );
  assert( (offset-c)<JOURNAL_HDR_SZ(pPager) );
  return offset;
}

/*
** The journal file must be open when this function is called.
**
** This function is a no-op if the journal file has not been written to
** within the current transaction (i.e. if Pager.journalOff==0).
**
** If doTruncate is non-zero or the Pager.journalSizeLimit variable is
** set to 0, then truncate the journal file to zero bytes in size. Otherwise,
** zero the 28-byte header at the start of the journal file. In either case,
** if the pager is not in no-sync mode, sync the journal file immediately
** after writing or truncating it.
**
** If Pager.journalSizeLimit is set to a positive, non-zero value, and
** following the truncation or zeroing described above the size of the
** journal file in bytes is larger than this value, then truncate the
** journal file to Pager.journalSizeLimit bytes. The journal file does
** not need to be synced following this operation.
**
** If an IO error occurs, abandon processing and return the IO error code.
** Otherwise, return SQLITE_OK.
*/
static int zeroJournalHdr(Pager *pPager, int doTruncate){
  int rc = SQLITE_OK;                               /* Return code */
  assert( isOpen(pPager->jfd) );
  assert( !sqlite3JournalIsInMemory(pPager->jfd) );
  if( pPager->journalOff ){
    const i64 iLimit = pPager->journalSizeLimit;    /* Local cache of jsl */

    IOTRACE(("JZEROHDR %p\n", pPager))
    if( doTruncate || iLimit==0 ){
      rc = sqlite3OsTruncate(pPager->jfd, 0);
    }else{
      static const char zeroHdr[28] = {0};
      rc = sqlite3OsWrite(pPager->jfd, zeroHdr, sizeof(zeroHdr), 0);
    }
    if( rc==SQLITE_OK && !pPager->noSync ){
      rc = sqlite3OsSync(pPager->jfd, SQLITE_SYNC_DATAONLY|pPager->syncFlags);
    }

    /* At this point the transaction is committed but the write lock
    ** is still held on the file. If there is a size limit configured for
    ** the persistent journal and the journal file currently consumes more
    ** space than that limit allows for, truncate it now. There is no need
    ** to sync the file following this operation.
    */
    if( rc==SQLITE_OK && iLimit>0 ){
      i64 sz;
      rc = sqlite3OsFileSize(pPager->jfd, &sz);
      if( rc==SQLITE_OK && sz>iLimit ){
        rc = sqlite3OsTruncate(pPager->jfd, iLimit);
      }
    }
  }
  return rc;
}

/*
** The journal file must be open when this routine is called. A journal
** header (JOURNAL_HDR_SZ bytes) is written into the journal file at the
** current location.
**
** The format for the journal header is as follows:
** - 8 bytes: Magic identifying journal format.
** - 4 bytes: Number of records in journal, or -1 no-sync mode is on.
** - 4 bytes: Random number used for page hash.
** - 4 bytes: Initial database page count.
** - 4 bytes: Sector size used by the process that wrote this journal.
** - 4 bytes: Database page size.
**
** Followed by (JOURNAL_HDR_SZ - 28) bytes of unused space.
*/
static int writeJournalHdr(Pager *pPager){
  int rc = SQLITE_OK;                 /* Return code */
  char *zHeader = pPager->pTmpSpace;  /* Temporary space used to build header */
  u32 nHeader = (u32)pPager->pageSize;/* Size of buffer pointed to by zHeader */
  u32 nWrite;                         /* Bytes of header sector written */
  int ii;                             /* Loop counter */

  assert( isOpen(pPager->jfd) );      /* Journal file must be open. */

  if( nHeader>JOURNAL_HDR_SZ(pPager) ){
    nHeader = JOURNAL_HDR_SZ(pPager);
  }

  /* If there are active savepoints and any of them were created
  ** since the most recent journal header was written, update the
  ** PagerSavepoint.iHdrOffset fields now.
  */
  for(ii=0; ii<pPager->nSavepoint; ii++){
    if( pPager->aSavepoint[ii].iHdrOffset==0 ){
      pPager->aSavepoint[ii].iHdrOffset = pPager->journalOff;
    }
  }

  pPager->journalHdr = pPager->journalOff = journalHdrOffset(pPager);

  /*
  ** Write the nRec Field - the number of page records that follow this
  ** journal header. Normally, zero is written to this value at this time.
  ** After the records are added to the journal (and the journal synced,
  ** if in full-sync mode), the zero is overwritten with the true number
  ** of records (see syncJournal()).
  **
  ** A faster alternative is to write 0xFFFFFFFF to the nRec field. When
  ** reading the journal this value tells SQLite to assume that the
  ** rest of the journal file contains valid page records. This assumption
  ** is dangerous, as if a failure occurred whilst writing to the journal
  ** file it may contain some garbage data. There are two scenarios
  ** where this risk can be ignored:
  **
  **   * When the pager is in no-sync mode. Corruption can follow a
  **     power failure in this case anyway.
  **
  **   * When the SQLITE_IOCAP_SAFE_APPEND flag is set. This guarantees
  **     that garbage data is never appended to the journal file.
  */
  assert( isOpen(pPager->fd) || pPager->noSync );
  if( pPager->noSync || (pPager->journalMode==PAGER_JOURNALMODE_MEMORY)
   || (sqlite3OsDeviceCharacteristics(pPager->fd)&SQLITE_IOCAP_SAFE_APPEND)
  ){
    memcpy(zHeader, aJournalMagic, sizeof(aJournalMagic));
    put32bits(&zHeader[sizeof(aJournalMagic)], 0xffffffff);
  }else{
    memset(zHeader, 0, sizeof(aJournalMagic)+4);
  }



  /* The random check-hash initializer */
  if( pPager->journalMode!=PAGER_JOURNALMODE_MEMORY ){
    sqlite3_randomness(sizeof(pPager->cksumInit), &pPager->cksumInit);
  }
#ifdef SQLITE_DEBUG
  else{
    /* The Pager.cksumInit variable is usually randomized above to protect
    ** against there being existing records in the journal file. This is
    ** dangerous, as following a crash they may be mistaken for records
    ** written by the current transaction and rolled back into the database
    ** file, causing corruption. The following assert statements verify
    ** that this is not required in "journal_mode=memory" mode, as in that
    ** case the journal file is always 0 bytes in size at this point. 
    ** It is advantageous to avoid the sqlite3_randomness() call if possible 
    ** as it takes the global PRNG mutex.  */
    i64 sz = 0;
    sqlite3OsFileSize(pPager->jfd, &sz);
    assert( sz==0 );
    assert( pPager->journalOff==journalHdrOffset(pPager) );
    assert( sqlite3JournalIsInMemory(pPager->jfd) );
  }
#endif
  put32bits(&zHeader[sizeof(aJournalMagic)+4], pPager->cksumInit);

  /* The initial database size */
  put32bits(&zHeader[sizeof(aJournalMagic)+8], pPager->dbOrigSize);
  /* The assumed sector size for this process */
  put32bits(&zHeader[sizeof(aJournalMagic)+12], pPager->sectorSize);

  /* The page size */
  put32bits(&zHeader[sizeof(aJournalMagic)+16], pPager->pageSize);

  /* Initializing the tail of the buffer is not necessary.  Everything
  ** works find if the following memset() is omitted.  But initializing
  ** the memory prevents valgrind from complaining, so we are willing to
  ** take the performance hit.
  */
  memset(&zHeader[sizeof(aJournalMagic)+20], 0,
         nHeader-(sizeof(aJournalMagic)+20));

  /* In theory, it is only necessary to write the 28 bytes that the
  ** journal header consumes to the journal file here. Then increment the
  ** Pager.journalOff variable by JOURNAL_HDR_SZ so that the next
  ** record is written to the following sector (leaving a gap in the file
  ** that will be implicitly filled in by the OS).
  **
  ** However it has been discovered that on some systems this pattern can
  ** be significantly slower than contiguously writing data to the file,
  ** even if that means explicitly writing data to the block of
  ** (JOURNAL_HDR_SZ - 28) bytes that will not be used. So that is what
  ** is done.
  **
  ** The loop is required here in case the sector-size is larger than the
  ** database page size. Since the zHeader buffer is only Pager.pageSize
  ** bytes in size, more than one call to sqlite3OsWrite() may be required
  ** to populate the entire journal header sector.
  */
  for(nWrite=0; rc==SQLITE_OK&&nWrite<JOURNAL_HDR_SZ(pPager); nWrite+=nHeader){
    IOTRACE(("JHDR %p %lld %d\n", pPager, pPager->journalHdr, nHeader))
    rc = sqlite3OsWrite(pPager->jfd, zHeader, nHeader, pPager->journalOff);
    assert( pPager->journalHdr <= pPager->journalOff );
    pPager->journalOff += nHeader;
  }

  return rc;
}

/*
** The journal file must be open when this is called. A journal header file
** (JOURNAL_HDR_SZ bytes) is read from the current location in the journal
** file. The current location in the journal file is given by
** pPager->journalOff. See comments above function writeJournalHdr() for
** a description of the journal header format.
**
** If the header is read successfully, *pNRec is set to the number of
** page records following this header and *pDbSize is set to the size of the
** database before the transaction began, in pages. Also, pPager->cksumInit
** is set to the value read from the journal header. SQLITE_OK is returned
** in this case.
**
** If the journal header file appears to be corrupted, SQLITE_DONE is
** returned and *pNRec and *PDbSize are undefined.  If JOURNAL_HDR_SZ bytes
** cannot be read from the journal file an error code is returned.
*/
static int readJournalHdr(
  Pager *pPager,               /* Pager object */
  int isHot,
  i64 journalSize,             /* Size of the open journal file in bytes */
  u32 *pNRec,                  /* OUT: Value read from the nRec field */
  u32 *pDbSize                 /* OUT: Value of original database size field */
){
  int rc;                      /* Return code */
  unsigned char aMagic[8];     /* A buffer to hold the magic header */
  i64 iHdrOff;                 /* Offset of journal header being read */

  assert( isOpen(pPager->jfd) );      /* Journal file must be open. */

  /* Advance Pager.journalOff to the start of the next sector. If the
  ** journal file is too small for there to be a header stored at this
  ** point, return SQLITE_DONE.
  */
  pPager->journalOff = journalHdrOffset(pPager);
  if( pPager->journalOff+JOURNAL_HDR_SZ(pPager) > journalSize ){
    return SQLITE_DONE;
  }
  iHdrOff = pPager->journalOff;

  /* Read in the first 8 bytes of the journal header. If they do not match
  ** the  magic string found at the start of each journal header, return
  ** SQLITE_DONE. If an IO error occurs, return an error code. Otherwise,
  ** proceed.
  */
  if( isHot || iHdrOff!=pPager->journalHdr ){
    rc = sqlite3OsRead(pPager->jfd, aMagic, sizeof(aMagic), iHdrOff);
    if( rc ){
      return rc;
    }
    if( memcmp(aMagic, aJournalMagic, sizeof(aMagic))!=0 ){
      return SQLITE_DONE;
    }
  }

  /* Read the first three 32-bit fields of the journal header: The nRec
  ** field, the checksum-initializer and the database size at the start
  ** of the transaction. Return an error code if anything goes wrong.
  */
  if( SQLITE_OK!=(rc = read32bits(pPager->jfd, iHdrOff+8, pNRec))
   || SQLITE_OK!=(rc = read32bits(pPager->jfd, iHdrOff+12, &pPager->cksumInit))
   || SQLITE_OK!=(rc = read32bits(pPager->jfd, iHdrOff+16, pDbSize))
  ){
    return rc;
  }

  if( pPager->journalOff==0 ){
    u32 iPageSize;               /* Page-size field of journal header */
    u32 iSectorSize;             /* Sector-size field of journal header */

    /* Read the page-size and sector-size journal header fields. */
    if( SQLITE_OK!=(rc = read32bits(pPager->jfd, iHdrOff+20, &iSectorSize))
     || SQLITE_OK!=(rc = read32bits(pPager->jfd, iHdrOff+24, &iPageSize))
    ){
      return rc;
    }

    /* Versions of SQLite prior to 3.5.8 set the page-size field of the
    ** journal header to zero. In this case, assume that the Pager.pageSize
    ** variable is already set to the correct page size.
    */
    if( iPageSize==0 ){
      iPageSize = pPager->pageSize;
    }

    /* Check that the values read from the page-size and sector-size fields
    ** are within range. To be 'in range', both values need to be a power
    ** of two greater than or equal to 512 or 32, and not greater than their
    ** respective compile time maximum limits.
    */
    if( iPageSize<512                  || iSectorSize<32
     || iPageSize>SQLITE_MAX_PAGE_SIZE || iSectorSize>MAX_SECTOR_SIZE
     || ((iPageSize-1)&iPageSize)!=0   || ((iSectorSize-1)&iSectorSize)!=0
    ){
      /* If the either the page-size or sector-size in the journal-header is
      ** invalid, then the process that wrote the journal-header must have
      ** crashed before the header was synced. In this case stop reading
      ** the journal file here.
      */
      return SQLITE_DONE;
    }

    /* Update the page-size to match the value read from the journal.
    ** Use a testcase() macro to make sure that malloc failure within
    ** PagerSetPagesize() is tested.
    */
    rc = sqlite3PagerSetPagesize(pPager, &iPageSize, -1);
    testcase( rc!=SQLITE_OK );

    /* Update the assumed sector-size to match the value used by
    ** the process that created this journal. If this journal was
    ** created by a process other than this one, then this routine
    ** is being called from within pager_playback(). The local value
    ** of Pager.sectorSize is restored at the end of that routine.
    */
    pPager->sectorSize = iSectorSize;
  }

  pPager->journalOff += JOURNAL_HDR_SZ(pPager);
  return rc;
}


/*
** Write the supplied super-journal name into the journal file for pager
** pPager at the current location. The super-journal name must be the last
** thing written to a journal file. If the pager is in full-sync mode, the
** journal file descriptor is advanced to the next sector boundary before
** anything is written. The format is:
**
**   + 4 bytes: PAGER_SJ_PGNO.
**   + N bytes: super-journal filename in utf-8.
**   + 4 bytes: N (length of super-journal name in bytes, no nul-terminator).
**   + 4 bytes: super-journal name checksum.
**   + 8 bytes: aJournalMagic[].
**
** The super-journal page checksum is the sum of the bytes in the super-journal
** name, where each byte is interpreted as a signed 8-bit integer.
**
** If zSuper is a NULL pointer (occurs for a single database transaction),
** this call is a no-op.
*/
static int writeSuperJournal(Pager *pPager, const char *zSuper){
  int rc;                          /* Return code */
  int nSuper;                      /* Length of string zSuper */
  i64 iHdrOff;                     /* Offset of header in journal file */
  i64 jrnlSize;                    /* Size of journal file on disk */
  u32 cksum = 0;                   /* Checksum of string zSuper */

  assert( pPager->setSuper==0 );
  assert( !pagerUseWal(pPager) );

  if( !zSuper
   || pPager->journalMode==PAGER_JOURNALMODE_MEMORY
   || !isOpen(pPager->jfd)
  ){
    return SQLITE_OK;
  }
  pPager->setSuper = 1;
  assert( pPager->journalHdr <= pPager->journalOff );

  /* Calculate the length in bytes and the checksum of zSuper */
  for(nSuper=0; zSuper[nSuper]; nSuper++){
    cksum += zSuper[nSuper];
  }

  /* If in full-sync mode, advance to the next disk sector before writing
  ** the super-journal name. This is in case the previous page written to
  ** the journal has already been synced.
  */
  if( pPager->fullSync ){
    pPager->journalOff = journalHdrOffset(pPager);
  }
  iHdrOff = pPager->journalOff;

  /* Write the super-journal data to the end of the journal file. If
  ** an error occurs, return the error code to the caller.
  */
  if( (0 != (rc = write32bits(pPager->jfd, iHdrOff, PAGER_SJ_PGNO(pPager))))
   || (0 != (rc = sqlite3OsWrite(pPager->jfd, zSuper, nSuper, iHdrOff+4)))
   || (0 != (rc = write32bits(pPager->jfd, iHdrOff+4+nSuper, nSuper)))
   || (0 != (rc = write32bits(pPager->jfd, iHdrOff+4+nSuper+4, cksum)))
   || (0 != (rc = sqlite3OsWrite(pPager->jfd, aJournalMagic, 8,
                                 iHdrOff+4+nSuper+8)))
  ){
    return rc;
  }
  pPager->journalOff += (nSuper+20);

  /* If the pager is in persistent-journal mode, then the physical
  ** journal-file may extend past the end of the super-journal name
  ** and 8 bytes of magic data just written to the file. This is
  ** dangerous because the code to rollback a hot-journal file
  ** will not be able to find the super-journal name to determine
  ** whether or not the journal is hot.
  **
  ** Easiest thing to do in this scenario is to truncate the journal
  ** file to the required size.
  */
  if( SQLITE_OK==(rc = sqlite3OsFileSize(pPager->jfd, &jrnlSize))
   && jrnlSize>pPager->journalOff
  ){
    rc = sqlite3OsTruncate(pPager->jfd, pPager->journalOff);
  }
  return rc;
}

/*
** Discard the entire contents of the in-memory page-cache.
*/
static void pager_reset(Pager *pPager){
  pPager->iDataVersion++;
  sqlite3BackupRestart(pPager->pBackup);
  sqlite3PcacheClear(pPager->pPCache);
}

/*
** Return the pPager->iDataVersion value
*/
u32 sqlite3PagerDataVersion(Pager *pPager){
  return pPager->iDataVersion;
}

/*
** Free all structures in the Pager.aSavepoint[] array and set both
** Pager.aSavepoint and Pager.nSavepoint to zero. Close the sub-journal
** if it is open and the pager is not in exclusive mode.
*/
static void releaseAllSavepoints(Pager *pPager){
  int ii;               /* Iterator for looping through Pager.aSavepoint */
  for(ii=0; ii<pPager->nSavepoint; ii++){
    sqlite3BitvecDestroy(pPager->aSavepoint[ii].pInSavepoint);
  }
  if( !pPager->exclusiveMode || sqlite3JournalIsInMemory(pPager->sjfd) ){
    sqlite3OsClose(pPager->sjfd);
  }
  sqlite3_free(pPager->aSavepoint);
  pPager->aSavepoint = 0;
  pPager->nSavepoint = 0;
  pPager->nSubRec = 0;
}

/*
** Set the bit number pgno in the PagerSavepoint.pInSavepoint
** bitvecs of all open savepoints. Return SQLITE_OK if successful
** or SQLITE_NOMEM if a malloc failure occurs.
*/
static int addToSavepointBitvecs(Pager *pPager, Pgno pgno){
  int ii;                   /* Loop counter */
  int rc = SQLITE_OK;       /* Result code */

  for(ii=0; ii<pPager->nSavepoint; ii++){
    PagerSavepoint *p = &pPager->aSavepoint[ii];
    if( pgno<=p->nOrig ){
      rc |= sqlite3BitvecSet(p->pInSavepoint, pgno);
      testcase( rc==SQLITE_NOMEM );
      assert( rc==SQLITE_OK || rc==SQLITE_NOMEM );
    }
  }
  return rc;
}

/*
** This function is a no-op if the pager is in exclusive mode and not
** in the ERROR state. Otherwise, it switches the pager to PAGER_OPEN
** state.
**
** If the pager is not in exclusive-access mode, the database file is
** completely unlocked. If the file is unlocked and the file-system does
** not exhibit the UNDELETABLE_WHEN_OPEN property, the journal file is
** closed (if it is open).
**
** If the pager is in ERROR state when this function is called, the
** contents of the pager cache are discarded before switching back to
** the OPEN state. Regardless of whether the pager is in exclusive-mode
** or not, any journal file left in the file-system will be treated
** as a hot-journal and rolled back the next time a read-transaction
** is opened (by this or by any other connection).
*/
static void pager_unlock(Pager *pPager){

  assert( pPager->eState==PAGER_READER
       || pPager->eState==PAGER_OPEN
       || pPager->eState==PAGER_ERROR
  );

  sqlite3BitvecDestroy(pPager->pInJournal);
  pPager->pInJournal = 0;
  releaseAllSavepoints(pPager);

  if( pagerUseWal(pPager) ){
    assert( !isOpen(pPager->jfd) );
    sqlite3WalEndReadTransaction(pPager->pWal);
    pPager->eState = PAGER_OPEN;
  }else if( !pPager->exclusiveMode ){
    int rc;                       /* Error code returned by pagerUnlockDb() */
    int iDc = isOpen(pPager->fd)?sqlite3OsDeviceCharacteristics(pPager->fd):0;

    /* If the operating system support deletion of open files, then
    ** close the journal file when dropping the database lock.  Otherwise
    ** another connection with journal_mode=delete might delete the file
    ** out from under us.
    */
    assert( (PAGER_JOURNALMODE_MEMORY   & 5)!=1 );
    assert( (PAGER_JOURNALMODE_OFF      & 5)!=1 );
    assert( (PAGER_JOURNALMODE_WAL      & 5)!=1 );
    assert( (PAGER_JOURNALMODE_DELETE   & 5)!=1 );
    assert( (PAGER_JOURNALMODE_TRUNCATE & 5)==1 );
    assert( (PAGER_JOURNALMODE_PERSIST  & 5)==1 );
    if( 0==(iDc & SQLITE_IOCAP_UNDELETABLE_WHEN_OPEN)
     || 1!=(pPager->journalMode & 5)
    ){
      sqlite3OsClose(pPager->jfd);
    }

    /* If the pager is in the ERROR state and the call to unlock the database
    ** file fails, set the current lock to UNKNOWN_LOCK. See the comment
    ** above the #define for UNKNOWN_LOCK for an explanation of why this
    ** is necessary.
    */
    rc = pagerUnlockDb(pPager, NO_LOCK);
    if( rc!=SQLITE_OK && pPager->eState==PAGER_ERROR ){
      pPager->eLock = UNKNOWN_LOCK;
    }

    /* The pager state may be changed from PAGER_ERROR to PAGER_OPEN here
    ** without clearing the error code. This is intentional - the error
    ** code is cleared and the cache reset in the block below.
    */
    assert( pPager->errCode || pPager->eState!=PAGER_ERROR );
    pPager->eState = PAGER_OPEN;
  }

  /* If Pager.errCode is set, the contents of the pager cache cannot be
  ** trusted. Now that there are no outstanding references to the pager,
  ** it can safely move back to PAGER_OPEN state. This happens in both
  ** normal and exclusive-locking mode.
  */
  assert( pPager->errCode==SQLITE_OK || !MEMDB );
  if( pPager->errCode ){
    if( pPager->tempFile==0 ){
      pager_reset(pPager);
      pPager->changeCountDone = 0;
      pPager->eState = PAGER_OPEN;
    }else{
      pPager->eState = (isOpen(pPager->jfd) ? PAGER_OPEN : PAGER_READER);
    }
    if( USEFETCH(pPager) ) sqlite3OsUnfetch(pPager->fd, 0, 0);
    pPager->errCode = SQLITE_OK;
    setGetterMethod(pPager);
  }

  pPager->journalOff = 0;
  pPager->journalHdr = 0;
  pPager->setSuper = 0;
}

/*
** This function is called whenever an IOERR or FULL error that requires
** the pager to transition into the ERROR state may have occurred.
** The first argument is a pointer to the pager structure, the second
** the error-code about to be returned by a pager API function. The
** value returned is a copy of the second argument to this function.
**
** If the second argument is SQLITE_FULL, SQLITE_IOERR or one of the
** IOERR sub-codes, the pager enters the ERROR state and the error code
** is stored in Pager.errCode. While the pager remains in the ERROR state,
** all major API calls on the Pager will immediately return Pager.errCode.
**
** The ERROR state indicates that the contents of the pager-cache
** cannot be trusted. This state can be cleared by completely discarding
** the contents of the pager-cache. If a transaction was active when
** the persistent error occurred, then the rollback journal may need
** to be replayed to restore the contents of the database file (as if
** it were a hot-journal).
*/
static int pager_error(Pager *pPager, int rc){
  int rc2 = rc & 0xff;
  assert( rc==SQLITE_OK || !MEMDB );
  assert(
       pPager->errCode==SQLITE_FULL ||
       pPager->errCode==SQLITE_OK ||
       (pPager->errCode & 0xff)==SQLITE_IOERR
  );
  if( rc2==SQLITE_FULL || rc2==SQLITE_IOERR ){
    pPager->errCode = rc;
    pPager->eState = PAGER_ERROR;
    setGetterMethod(pPager);
  }
  return rc;
}

static int pager_truncate(Pager *pPager, Pgno nPage);

/*
** The write transaction open on pPager is being committed (bCommit==1)
** or rolled back (bCommit==0).
**
** Return TRUE if and only if all dirty pages should be flushed to disk.
**
** Rules:
**
**   *  For non-TEMP databases, always sync to disk.  This is necessary
**      for transactions to be durable.
**
**   *  Sync TEMP database only on a COMMIT (not a ROLLBACK) when the backing
**      file has been created already (via a spill on pagerStress()) and
**      when the number of dirty pages in memory exceeds 25% of the total
**      cache size.
*/
static int pagerFlushOnCommit(Pager *pPager, int bCommit){
  if( pPager->tempFile==0 ) return 1;
  if( !bCommit ) return 0;
  if( !isOpen(pPager->fd) ) return 0;
  return (sqlite3PCachePercentDirty(pPager->pPCache)>=25);
}

/*
** This routine ends a transaction. A transaction is usually ended by
** either a COMMIT or a ROLLBACK operation. This routine may be called
** after rollback of a hot-journal, or if an error occurs while opening
** the journal file or writing the very first journal-header of a
** database transaction.
**
** This routine is never called in PAGER_ERROR state. If it is called
** in PAGER_NONE or PAGER_SHARED state and the lock held is less
** exclusive than a RESERVED lock, it is a no-op.
**
** Otherwise, any active savepoints are released.
**
** If the journal file is open, then it is "finalized". Once a journal
** file has been finalized it is not possible to use it to roll back a
** transaction. Nor will it be considered to be a hot-journal by this
** or any other database connection. Exactly how a journal is finalized
** depends on whether or not the pager is running in exclusive mode and
** the current journal-mode (Pager.journalMode value), as follows:
**
**   journalMode==MEMORY
**     Journal file descriptor is simply closed. This destroys an
**     in-memory journal.
**
**   journalMode==TRUNCATE
**     Journal file is truncated to zero bytes in size.
**
**   journalMode==PERSIST
**     The first 28 bytes of the journal file are zeroed. This invalidates
**     the first journal header in the file, and hence the entire journal
**     file. An invalid journal file cannot be rolled back.
**
**   journalMode==DELETE
**     The journal file is closed and deleted using sqlite3OsDelete().
**
**     If the pager is running in exclusive mode, this method of finalizing
**     the journal file is never used. Instead, if the journalMode is
**     DELETE and the pager is in exclusive mode, the method described under
**     journalMode==PERSIST is used instead.
**
** After the journal is finalized, the pager moves to PAGER_READER state.
** If running in non-exclusive rollback mode, the lock on the file is
** downgraded to a SHARED_LOCK.
**
** SQLITE_OK is returned if no error occurs. If an error occurs during
** any of the IO operations to finalize the journal file or unlock the
** database then the IO error code is returned to the user. If the
** operation to finalize the journal file fails, then the code still
** tries to unlock the database file if not in exclusive mode. If the
** unlock operation fails as well, then the first error code related
** to the first error encountered (the journal finalization one) is
** returned.
*/
static int pager_end_transaction(Pager *pPager, int hasSuper, int bCommit){
  int rc = SQLITE_OK;      /* Error code from journal finalization operation */
  int rc2 = SQLITE_OK;     /* Error code from db file unlock operation */

  /* Do nothing if the pager does not have an open write transaction
  ** or at least a RESERVED lock. This function may be called when there
  ** is no write-transaction active but a RESERVED or greater lock is
  ** held under two circumstances:
  **
  **   1. After a successful hot-journal rollback, it is called with
  **      eState==PAGER_NONE and eLock==EXCLUSIVE_LOCK.
  **
  **   2. If a connection with locking_mode=exclusive holding an EXCLUSIVE
  **      lock switches back to locking_mode=normal and then executes a
  **      read-transaction, this function is called with eState==PAGER_READER
  **      and eLock==EXCLUSIVE_LOCK when the read-transaction is closed.
  */
  assert( assert_pager_state(pPager) );
  assert( pPager->eState!=PAGER_ERROR );
  if( pPager->eState<PAGER_WRITER_LOCKED && pPager->eLock<RESERVED_LOCK ){
    return SQLITE_OK;
  }

  releaseAllSavepoints(pPager);
  assert( isOpen(pPager->jfd) || pPager->pInJournal==0
      || (sqlite3OsDeviceCharacteristics(pPager->fd)&SQLITE_IOCAP_BATCH_ATOMIC)
  );
  if( isOpen(pPager->jfd) ){
    assert( !pagerUseWal(pPager) );

    /* Finalize the journal file. */
    if( sqlite3JournalIsInMemory(pPager->jfd) ){
      /* assert( pPager->journalMode==PAGER_JOURNALMODE_MEMORY ); */
      sqlite3OsClose(pPager->jfd);
    }else if( pPager->journalMode==PAGER_JOURNALMODE_TRUNCATE ){
      if( pPager->journalOff==0 ){
        rc = SQLITE_OK;
      }else{
        rc = sqlite3OsTruncate(pPager->jfd, 0);
        if( rc==SQLITE_OK && pPager->fullSync ){
          /* Make sure the new file size is written into the inode right away.
          ** Otherwise the journal might resurrect following a power loss and
          ** cause the last transaction to roll back.  See
          ** https://bugzilla.mozilla.org/show_bug.cgi?id=1072773
          */
          rc = sqlite3OsSync(pPager->jfd, pPager->syncFlags);
        }
      }
      pPager->journalOff = 0;
    }else if( pPager->journalMode==PAGER_JOURNALMODE_PERSIST
      || (pPager->exclusiveMode && pPager->journalMode!=PAGER_JOURNALMODE_WAL)
    ){
      rc = zeroJournalHdr(pPager, hasSuper||pPager->tempFile);
      pPager->journalOff = 0;
    }else{
      /* This branch may be executed with Pager.journalMode==MEMORY if
      ** a hot-journal was just rolled back. In this case the journal
      ** file should be closed and deleted. If this connection writes to
      ** the database file, it will do so using an in-memory journal.
      */
      int bDelete = !pPager->tempFile;
      assert( sqlite3JournalIsInMemory(pPager->jfd)==0 );
      assert( pPager->journalMode==PAGER_JOURNALMODE_DELETE
           || pPager->journalMode==PAGER_JOURNALMODE_MEMORY
           || pPager->journalMode==PAGER_JOURNALMODE_WAL
      );
      sqlite3OsClose(pPager->jfd);
      if( bDelete ){
        rc = sqlite3OsDelete(pPager->pVfs, pPager->zJournal, pPager->extraSync);
      }
    }
  }

#ifdef SQLITE_CHECK_PAGES
  sqlite3PcacheIterateDirty(pPager->pPCache, pager_set_pagehash);
  if( pPager->dbSize==0 && sqlite3PcacheRefCount(pPager->pPCache)>0 ){
    PgHdr *p = sqlite3PagerLookup(pPager, 1);
    if( p ){
      p->pageHash = 0;
      sqlite3PagerUnrefNotNull(p);
    }
  }
#endif

  sqlite3BitvecDestroy(pPager->pInJournal);
  pPager->pInJournal = 0;
  pPager->nRec = 0;
  if( rc==SQLITE_OK ){
    if( MEMDB || pagerFlushOnCommit(pPager, bCommit) ){
      sqlite3PcacheCleanAll(pPager->pPCache);
    }else{
      sqlite3PcacheClearWritable(pPager->pPCache);
    }
    sqlite3PcacheTruncate(pPager->pPCache, pPager->dbSize);
  }

  if( pagerUseWal(pPager) ){
    /* Drop the WAL write-lock, if any. Also, if the connection was in
    ** locking_mode=exclusive mode but is no longer, drop the EXCLUSIVE
    ** lock held on the database file.
    */
    rc2 = sqlite3WalEndWriteTransaction(pPager->pWal);
    assert( rc2==SQLITE_OK );
  }else if( rc==SQLITE_OK && bCommit && pPager->dbFileSize>pPager->dbSize ){
    /* This branch is taken when committing a transaction in rollback-journal
    ** mode if the database file on disk is larger than the database image.
    ** At this point the journal has been finalized and the transaction
    ** successfully committed, but the EXCLUSIVE lock is still held on the
    ** file. So it is safe to truncate the database file to its minimum
    ** required size.  */
    assert( pPager->eLock==EXCLUSIVE_LOCK );
    rc = pager_truncate(pPager, pPager->dbSize);
  }

  if( rc==SQLITE_OK && bCommit ){
    rc = sqlite3OsFileControl(pPager->fd, SQLITE_FCNTL_COMMIT_PHASETWO, 0);
    if( rc==SQLITE_NOTFOUND ) rc = SQLITE_OK;
  }

  if( !pPager->exclusiveMode
   && (!pagerUseWal(pPager) || sqlite3WalExclusiveMode(pPager->pWal, 0))
  ){
    rc2 = pagerUnlockDb(pPager, SHARED_LOCK);
  }
  pPager->eState = PAGER_READER;
  pPager->setSuper = 0;

  return (rc==SQLITE_OK?rc2:rc);
}

/* Forward reference */
static int pager_playback(Pager *pPager, int isHot);

/*
** Execute a rollback if a transaction is active and unlock the
** database file.
**
** If the pager has already entered the ERROR state, do not attempt
** the rollback at this time. Instead, pager_unlock() is called. The
** call to pager_unlock() will discard all in-memory pages, unlock
** the database file and move the pager back to OPEN state. If this
** means that there is a hot-journal left in the file-system, the next
** connection to obtain a shared lock on the pager (which may be this one)
** will roll it back.
**
** If the pager has not already entered the ERROR state, but an IO or
** malloc error occurs during a rollback, then this will itself cause
** the pager to enter the ERROR state. Which will be cleared by the
** call to pager_unlock(), as described above.
*/
static void pagerUnlockAndRollback(Pager *pPager){
  if( pPager->eState!=PAGER_ERROR && pPager->eState!=PAGER_OPEN ){
    assert( assert_pager_state(pPager) );
    if( pPager->eState>=PAGER_WRITER_LOCKED ){
      sqlite3BeginBenignMalloc();
      sqlite3PagerRollback(pPager);
      sqlite3EndBenignMalloc();
    }else if( !pPager->exclusiveMode ){
      assert( pPager->eState==PAGER_READER );
      pager_end_transaction(pPager, 0, 0);
    }
  }else if( pPager->eState==PAGER_ERROR
         && pPager->journalMode==PAGER_JOURNALMODE_MEMORY
         && isOpen(pPager->jfd)
  ){
    /* Special case for a ROLLBACK due to I/O error with an in-memory
    ** journal:  We have to rollback immediately, before the journal is
    ** closed, because once it is closed, all content is forgotten. */
    int errCode = pPager->errCode;
    u8 eLock = pPager->eLock;
    pPager->eState = PAGER_OPEN;
    pPager->errCode = SQLITE_OK;
    pPager->eLock = EXCLUSIVE_LOCK;
    pager_playback(pPager, 1);
    pPager->errCode = errCode;
    pPager->eLock = eLock;
  }
  pager_unlock(pPager);
}

/*
** Parameter aData must point to a buffer of pPager->pageSize bytes
** of data. Compute and return a checksum based on the contents of the
** page of data and the current value of pPager->cksumInit.
**
** This is not a real checksum. It is really just the sum of the
** random initial value (pPager->cksumInit) and every 200th byte
** of the page data, starting with byte offset (pPager->pageSize%200).
** Each byte is interpreted as an 8-bit unsigned integer.
**
** Changing the formula used to compute this checksum results in an
** incompatible journal file format.
**
** If journal corruption occurs due to a power failure, the most likely
** scenario is that one end or the other of the record will be changed.
** It is much less likely that the two ends of the journal record will be
** correct and the middle be corrupt.  Thus, this "checksum" scheme,
** though fast and simple, catches the mostly likely kind of corruption.
*/
static u32 pager_cksum(Pager *pPager, const u8 *aData){
  u32 cksum = pPager->cksumInit;         /* Checksum value to return */
  int i = pPager->pageSize-200;          /* Loop counter */
  while( i>0 ){
    cksum += aData[i];
    i -= 200;
  }
  return cksum;
}

/*
** Read a single page from either the journal file (if isMainJrnl==1) or
** from the sub-journal (if isMainJrnl==0) and playback that page.
** The page begins at offset *pOffset into the file. The *pOffset
** value is increased to the start of the next page in the journal.
**
** The main rollback journal uses checksums - the statement journal does
** not.
**
** If the page number of the page record read from the (sub-)journal file
** is greater than the current value of Pager.dbSize, then playback is
** skipped and SQLITE_OK is returned.
**
** If pDone is not NULL, then it is a record of pages that have already
** been played back.  If the page at *pOffset has already been played back
** (if the corresponding pDone bit is set) then skip the playback.
** Make sure the pDone bit corresponding to the *pOffset page is set
** prior to returning.
**
** If the page record is successfully read from the (sub-)journal file
** and played back, then SQLITE_OK is returned. If an IO error occurs
** while reading the record from the (sub-)journal file or while writing
** to the database file, then the IO error code is returned. If data
** is successfully read from the (sub-)journal file but appears to be
** corrupted, SQLITE_DONE is returned. Data is considered corrupted in
** two circumstances:
**
**   * If the record page-number is illegal (0 or PAGER_SJ_PGNO), or
**   * If the record is being rolled back from the main journal file
**     and the checksum field does not match the record content.
**
** Neither of these two scenarios are possible during a savepoint rollback.
**
** If this is a savepoint rollback, then memory may have to be dynamically
** allocated by this function. If this is the case and an allocation fails,
** SQLITE_NOMEM is returned.
*/
static int pager_playback_one_page(
  Pager *pPager,                /* The pager being played back */
  i64 *pOffset,                 /* Offset of record to playback */
  Bitvec *pDone,                /* Bitvec of pages already played back */
  int isMainJrnl,               /* 1 -> main journal. 0 -> sub-journal. */
  int isSavepnt                 /* True for a savepoint rollback */
){
  int rc;
  PgHdr *pPg;                   /* An existing page in the cache */
  Pgno pgno;                    /* The page number of a page in journal */
  u32 cksum;                    /* Checksum used for sanity checking */
  char *aData;                  /* Temporary storage for the page */
  sqlite3_file *jfd;            /* The file descriptor for the journal file */
  int isSynced;                 /* True if journal page is synced */

  assert( (isMainJrnl&~1)==0 );      /* isMainJrnl is 0 or 1 */
  assert( (isSavepnt&~1)==0 );       /* isSavepnt is 0 or 1 */
  assert( isMainJrnl || pDone );     /* pDone always used on sub-journals */
  assert( isSavepnt || pDone==0 );   /* pDone never used on non-savepoint */

  aData = pPager->pTmpSpace;
  assert( aData );         /* Temp storage must have already been allocated */
  assert( pagerUseWal(pPager)==0 || (!isMainJrnl && isSavepnt) );

  /* Either the state is greater than PAGER_WRITER_CACHEMOD (a transaction
  ** or savepoint rollback done at the request of the caller) or this is
  ** a hot-journal rollback. If it is a hot-journal rollback, the pager
  ** is in state OPEN and holds an EXCLUSIVE lock. Hot-journal rollback
  ** only reads from the main journal, not the sub-journal.
  */
  assert( pPager->eState>=PAGER_WRITER_CACHEMOD
       || (pPager->eState==PAGER_OPEN && pPager->eLock==EXCLUSIVE_LOCK)
  );
  assert( pPager->eState>=PAGER_WRITER_CACHEMOD || isMainJrnl );

  /* Read the page number and page data from the journal or sub-journal
  ** file. Return an error code to the caller if an IO error occurs.
  */
  jfd = isMainJrnl ? pPager->jfd : pPager->sjfd;
  rc = read32bits(jfd, *pOffset, &pgno);
  if( rc!=SQLITE_OK ) return rc;
  rc = sqlite3OsRead(jfd, (u8*)aData, pPager->pageSize, (*pOffset)+4);
  if( rc!=SQLITE_OK ) return rc;
  *pOffset += pPager->pageSize + 4 + isMainJrnl*4;

  /* Sanity checking on the page.  This is more important that I originally
  ** thought.  If a power failure occurs while the journal is being written,
  ** it could cause invalid data to be written into the journal.  We need to
  ** detect this invalid data (with high probability) and ignore it.
  */
  if( pgno==0 || pgno==PAGER_SJ_PGNO(pPager) ){
    assert( !isSavepnt );
    return SQLITE_DONE;
  }
  if( pgno>(Pgno)pPager->dbSize || sqlite3BitvecTest(pDone, pgno) ){
    return SQLITE_OK;
  }
  if( isMainJrnl ){
    rc = read32bits(jfd, (*pOffset)-4, &cksum);
    if( rc ) return rc;
    if( !isSavepnt && pager_cksum(pPager, (u8*)aData)!=cksum ){
      return SQLITE_DONE;
    }
  }

  /* If this page has already been played back before during the current
  ** rollback, then don't bother to play it back again.
  */
  if( pDone && (rc = sqlite3BitvecSet(pDone, pgno))!=SQLITE_OK ){
    return rc;
  }

  /* When playing back page 1, restore the nReserve setting
  */
  if( pgno==1 && pPager->nReserve!=((u8*)aData)[20] ){
    pPager->nReserve = ((u8*)aData)[20];
  }

  /* If the pager is in CACHEMOD state, then there must be a copy of this
  ** page in the pager cache. In this case just update the pager cache,
  ** not the database file. The page is left marked dirty in this case.
  **
  ** An exception to the above rule: If the database is in no-sync mode
  ** and a page is moved during an incremental vacuum then the page may
  ** not be in the pager cache. Later: if a malloc() or IO error occurs
  ** during a Movepage() call, then the page may not be in the cache
  ** either. So the condition described in the above paragraph is not
  ** assert()able.
  **
  ** If in WRITER_DBMOD, WRITER_FINISHED or OPEN state, then we update the
  ** pager cache if it exists and the main file. The page is then marked
  ** not dirty. Since this code is only executed in PAGER_OPEN state for
  ** a hot-journal rollback, it is guaranteed that the page-cache is empty
  ** if the pager is in OPEN state.
  **
  ** Ticket #1171:  The statement journal might contain page content that is
  ** different from the page content at the start of the transaction.
  ** This occurs when a page is changed prior to the start of a statement
  ** then changed again within the statement.  When rolling back such a
  ** statement we must not write to the original database unless we know
  ** for certain that original page contents are synced into the main rollback
  ** journal.  Otherwise, a power loss might leave modified data in the
  ** database file without an entry in the rollback journal that can
  ** restore the database to its original form.  Two conditions must be
  ** met before writing to the database files. (1) the database must be
  ** locked.  (2) we know that the original page content is fully synced
  ** in the main journal either because the page is not in cache or else
  ** the page is marked as needSync==0.
  **
  ** 2008-04-14:  When attempting to vacuum a corrupt database file, it
  ** is possible to fail a statement on a database that does not yet exist.
  ** Do not attempt to write if database file has never been opened.
  */
  if( pagerUseWal(pPager) ){
    pPg = 0;
  }else{
    pPg = sqlite3PagerLookup(pPager, pgno);
  }
  assert( pPg || !MEMDB );
  assert( pPager->eState!=PAGER_OPEN || pPg==0 || pPager->tempFile );
  PAGERTRACE(("PLAYBACK %d page %d hash(%08x) %s\n",
           PAGERID(pPager), pgno, pager_datahash(pPager->pageSize, (u8*)aData),
           (isMainJrnl?"main-journal":"sub-journal")
  ));
  if( isMainJrnl ){
    isSynced = pPager->noSync || (*pOffset <= pPager->journalHdr);
  }else{
    isSynced = (pPg==0 || 0==(pPg->flags & PGHDR_NEED_SYNC));
  }
  if( isOpen(pPager->fd)
   && (pPager->eState>=PAGER_WRITER_DBMOD || pPager->eState==PAGER_OPEN)
   && isSynced
  ){
    i64 ofst = (pgno-1)*(i64)pPager->pageSize;
    testcase( !isSavepnt && pPg!=0 && (pPg->flags&PGHDR_NEED_SYNC)!=0 );
    assert( !pagerUseWal(pPager) );

    /* Write the data read from the journal back into the database file.
    ** This is usually safe even for an encrypted database - as the data
    ** was encrypted before it was written to the journal file. The exception
    ** is if the data was just read from an in-memory sub-journal. In that
    ** case it must be encrypted here before it is copied into the database
    ** file.  */
    rc = sqlite3OsWrite(pPager->fd, (u8 *)aData, pPager->pageSize, ofst);

    if( pgno>pPager->dbFileSize ){
      pPager->dbFileSize = pgno;
    }
    if( pPager->pBackup ){
      sqlite3BackupUpdate(pPager->pBackup, pgno, (u8*)aData);
    }
  }else if( !isMainJrnl && pPg==0 ){
    /* If this is a rollback of a savepoint and data was not written to
    ** the database and the page is not in-memory, there is a potential
    ** problem. When the page is next fetched by the b-tree layer, it
    ** will be read from the database file, which may or may not be
    ** current.
    **
    ** There are a couple of different ways this can happen. All are quite
    ** obscure. When running in synchronous mode, this can only happen
    ** if the page is on the free-list at the start of the transaction, then
    ** populated, then moved using sqlite3PagerMovepage().
    **
    ** The solution is to add an in-memory page to the cache containing
    ** the data just read from the sub-journal. Mark the page as dirty
    ** and if the pager requires a journal-sync, then mark the page as
    ** requiring a journal-sync before it is written.
    */
    assert( isSavepnt );
    assert( (pPager->doNotSpill & SPILLFLAG_ROLLBACK)==0 );
    pPager->doNotSpill |= SPILLFLAG_ROLLBACK;
    rc = sqlite3PagerGet(pPager, pgno, &pPg, 1);
    assert( (pPager->doNotSpill & SPILLFLAG_ROLLBACK)!=0 );
    pPager->doNotSpill &= ~SPILLFLAG_ROLLBACK;
    if( rc!=SQLITE_OK ) return rc;
    sqlite3PcacheMakeDirty(pPg);
  }
  if( pPg ){
    /* No page should ever be explicitly rolled back that is in use, except
    ** for page 1 which is held in use in order to keep the lock on the
    ** database active. However such a page may be rolled back as a result
    ** of an internal error resulting in an automatic call to
    ** sqlite3PagerRollback().
    */
    void *pData;
    pData = pPg->pData;
    memcpy(pData, (u8*)aData, pPager->pageSize);
    pPager->xReiniter(pPg);
    /* It used to be that sqlite3PcacheMakeClean(pPg) was called here.  But
    ** that call was dangerous and had no detectable benefit since the cache
    ** is normally cleaned by sqlite3PcacheCleanAll() after rollback and so
    ** has been removed. */
    pager_set_pagehash(pPg);

    /* If this was page 1, then restore the value of Pager.dbFileVers.
    ** Do this before any decoding. */
    if( pgno==1 ){
      memcpy(&pPager->dbFileVers, &((u8*)pData)[24],sizeof(pPager->dbFileVers));
    }
    sqlite3PcacheRelease(pPg);
  }
  return rc;
}

/*
** Parameter zSuper is the name of a super-journal file. A single journal
** file that referred to the super-journal file has just been rolled back.
** This routine checks if it is possible to delete the super-journal file,
** and does so if it is.
**
** Argument zSuper may point to Pager.pTmpSpace. So that buffer is not
** available for use within this function.
**
** When a super-journal file is created, it is populated with the names
** of all of its child journals, one after another, formatted as utf-8
** encoded text. The end of each child journal file is marked with a
** nul-terminator byte (0x00). i.e. the entire contents of a super-journal
** file for a transaction involving two databases might be:
**
**   "/home/bill/a.db-journal\x00/home/bill/b.db-journal\x00"
**
** A super-journal file may only be deleted once all of its child
** journals have been rolled back.
**
** This function reads the contents of the super-journal file into
** memory and loops through each of the child journal names. For
** each child journal, it checks if:
**
**   * if the child journal exists, and if so
**   * if the child journal contains a reference to super-journal
**     file zSuper
**
** If a child journal can be found that matches both of the criteria
** above, this function returns without doing anything. Otherwise, if
** no such child journal can be found, file zSuper is deleted from
** the file-system using sqlite3OsDelete().
**
** If an IO error within this function, an error code is returned. This
** function allocates memory by calling sqlite3Malloc(). If an allocation
** fails, SQLITE_NOMEM is returned. Otherwise, if no IO or malloc errors
** occur, SQLITE_OK is returned.
**
** TODO: This function allocates a single block of memory to load
** the entire contents of the super-journal file. This could be
** a couple of kilobytes or so - potentially larger than the page
** size.
*/
static int pager_delsuper(Pager *pPager, const char *zSuper){
  sqlite3_vfs *pVfs = pPager->pVfs;
  int rc;                   /* Return code */
  sqlite3_file *pSuper;     /* Malloc'd super-journal file descriptor */
  sqlite3_file *pJournal;   /* Malloc'd child-journal file descriptor */
  char *zSuperJournal = 0;  /* Contents of super-journal file */
  i64 nSuperJournal;        /* Size of super-journal file */
  char *zJournal;           /* Pointer to one journal within MJ file */
  char *zSuperPtr;          /* Space to hold super-journal filename */
  char *zFree = 0;          /* Free this buffer */
  int nSuperPtr;            /* Amount of space allocated to zSuperPtr[] */

  /* Allocate space for both the pJournal and pSuper file descriptors.
  ** If successful, open the super-journal file for reading.
  */
  pSuper = (sqlite3_file *)sqlite3MallocZero(pVfs->szOsFile * 2);
  if( !pSuper ){
    rc = SQLITE_NOMEM_BKPT;
    pJournal = 0;
  }else{
    const int flags = (SQLITE_OPEN_READONLY|SQLITE_OPEN_SUPER_JOURNAL);
    rc = sqlite3OsOpen(pVfs, zSuper, pSuper, flags, 0);
    pJournal = (sqlite3_file *)(((u8 *)pSuper) + pVfs->szOsFile);
  }
  if( rc!=SQLITE_OK ) goto delsuper_out;

  /* Load the entire super-journal file into space obtained from
  ** sqlite3_malloc() and pointed to by zSuperJournal.   Also obtain
  ** sufficient space (in zSuperPtr) to hold the names of super-journal
  ** files extracted from regular rollback-journals.
  */
  rc = sqlite3OsFileSize(pSuper, &nSuperJournal);
  if( rc!=SQLITE_OK ) goto delsuper_out;
  nSuperPtr = pVfs->mxPathname+1;
  zFree = sqlite3Malloc(4 + nSuperJournal + nSuperPtr + 2);
  if( !zFree ){
    rc = SQLITE_NOMEM_BKPT;
    goto delsuper_out;
  }
  zFree[0] = zFree[1] = zFree[2] = zFree[3] = 0;
  zSuperJournal = &zFree[4];
  zSuperPtr = &zSuperJournal[nSuperJournal+2];
  rc = sqlite3OsRead(pSuper, zSuperJournal, (int)nSuperJournal, 0);
  if( rc!=SQLITE_OK ) goto delsuper_out;
  zSuperJournal[nSuperJournal] = 0;
  zSuperJournal[nSuperJournal+1] = 0;

  zJournal = zSuperJournal;
  while( (zJournal-zSuperJournal)<nSuperJournal ){
    int exists;
    rc = sqlite3OsAccess(pVfs, zJournal, SQLITE_ACCESS_EXISTS, &exists);
    if( rc!=SQLITE_OK ){
      goto delsuper_out;
    }
    if( exists ){
      /* One of the journals pointed to by the super-journal exists.
      ** Open it and check if it points at the super-journal. If
      ** so, return without deleting the super-journal file.
      ** NB:  zJournal is really a MAIN_JOURNAL.  But call it a
      ** SUPER_JOURNAL here so that the VFS will not send the zJournal
      ** name into sqlite3_database_file_object().
      */
      int c;
      int flags = (SQLITE_OPEN_READONLY|SQLITE_OPEN_SUPER_JOURNAL);
      rc = sqlite3OsOpen(pVfs, zJournal, pJournal, flags, 0);
      if( rc!=SQLITE_OK ){
        goto delsuper_out;
      }

      rc = readSuperJournal(pJournal, zSuperPtr, nSuperPtr);
      sqlite3OsClose(pJournal);
      if( rc!=SQLITE_OK ){
        goto delsuper_out;
      }

      c = zSuperPtr[0]!=0 && strcmp(zSuperPtr, zSuper)==0;
      if( c ){
        /* We have a match. Do not delete the super-journal file. */
        goto delsuper_out;
      }
    }
    zJournal += (sqlite3Strlen30(zJournal)+1);
  }

  sqlite3OsClose(pSuper);
  rc = sqlite3OsDelete(pVfs, zSuper, 0);

delsuper_out:
  sqlite3_free(zFree);
  if( pSuper ){
    sqlite3OsClose(pSuper);
    assert( !isOpen(pJournal) );
    sqlite3_free(pSuper);
  }
  return rc;
}


/*
** This function is used to change the actual size of the database
** file in the file-system. This only happens when committing a transaction,
** or rolling back a transaction (including rolling back a hot-journal).
**
** If the main database file is not open, or the pager is not in either
** DBMOD or OPEN state, this function is a no-op. Otherwise, the size
** of the file is changed to nPage pages (nPage*pPager->pageSize bytes).
** If the file on disk is currently larger than nPage pages, then use the VFS
** xTruncate() method to truncate it.
**
** Or, it might be the case that the file on disk is smaller than
** nPage pages. Some operating system implementations can get confused if
** you try to truncate a file to some size that is larger than it
** currently is, so detect this case and write a single zero byte to
** the end of the new file instead.
**
** If successful, return SQLITE_OK. If an IO error occurs while modifying
** the database file, return the error code to the caller.
*/
static int pager_truncate(Pager *pPager, Pgno nPage){
  int rc = SQLITE_OK;
  assert( pPager->eState!=PAGER_ERROR );
  assert( pPager->eState!=PAGER_READER );
  PAGERTRACE(("Truncate %d npage %u\n", PAGERID(pPager), nPage));

 
  if( isOpen(pPager->fd)
   && (pPager->eState>=PAGER_WRITER_DBMOD || pPager->eState==PAGER_OPEN)
  ){
    i64 currentSize, newSize;
    int szPage = pPager->pageSize;
    assert( pPager->eLock==EXCLUSIVE_LOCK );
    /* TODO: Is it safe to use Pager.dbFileSize here? */
    rc = sqlite3OsFileSize(pPager->fd, &currentSize);
    newSize = szPage*(i64)nPage;
    if( rc==SQLITE_OK && currentSize!=newSize ){
      if( currentSize>newSize ){
        rc = sqlite3OsTruncate(pPager->fd, newSize);
      }else if( (currentSize+szPage)<=newSize ){
        char *pTmp = pPager->pTmpSpace;
        memset(pTmp, 0, szPage);
        testcase( (newSize-szPage) == currentSize );
        testcase( (newSize-szPage) >  currentSize );
        sqlite3OsFileControlHint(pPager->fd, SQLITE_FCNTL_SIZE_HINT, &newSize);
        rc = sqlite3OsWrite(pPager->fd, pTmp, szPage, newSize-szPage);
      }
      if( rc==SQLITE_OK ){
        pPager->dbFileSize = nPage;
      }
    }
  }
  return rc;
}

/*
** Return a sanitized version of the sector-size of OS file pFile. The
** return value is guaranteed to lie between 32 and MAX_SECTOR_SIZE.
*/
int sqlite3SectorSize(sqlite3_file *pFile){
  int iRet = sqlite3OsSectorSize(pFile);
  if( iRet<32 ){
    iRet = 512;
  }else if( iRet>MAX_SECTOR_SIZE ){
    assert( MAX_SECTOR_SIZE>=512 );
    iRet = MAX_SECTOR_SIZE;
  }
  return iRet;
}

/*
** Set the value of the Pager.sectorSize variable for the given
** pager based on the value returned by the xSectorSize method
** of the open database file. The sector size will be used
** to determine the size and alignment of journal header and
** super-journal pointers within created journal files.
**
** For temporary files the effective sector size is always 512 bytes.
**
** Otherwise, for non-temporary files, the effective sector size is
** the value returned by the xSectorSize() method rounded up to 32 if
** it is less than 32, or rounded down to MAX_SECTOR_SIZE if it
** is greater than MAX_SECTOR_SIZE.
**
** If the file has the SQLITE_IOCAP_POWERSAFE_OVERWRITE property, then set
** the effective sector size to its minimum value (512).  The purpose of
** pPager->sectorSize is to define the "blast radius" of bytes that
** might change if a crash occurs while writing to a single byte in
** that range.  But with POWERSAFE_OVERWRITE, the blast radius is zero
** (that is what POWERSAFE_OVERWRITE means), so we minimize the sector
** size.  For backwards compatibility of the rollback journal file format,
** we cannot reduce the effective sector size below 512.
*/
static void setSectorSize(Pager *pPager){
  assert( isOpen(pPager->fd) || pPager->tempFile );

  if( pPager->tempFile
   || (sqlite3OsDeviceCharacteristics(pPager->fd) &
              SQLITE_IOCAP_POWERSAFE_OVERWRITE)!=0
  ){
    /* Sector size doesn't matter for temporary files. Also, the file
    ** may not have been opened yet, in which case the OsSectorSize()
    ** call will segfault. */
    pPager->sectorSize = 512;
  }else{
    pPager->sectorSize = sqlite3SectorSize(pPager->fd);
  }
}

/*
** Playback the journal and thus restore the database file to
** the state it was in before we started making changes. 
**
** The journal file format is as follows:
**
**  (1)  8 byte prefix.  A copy of aJournalMagic[].
**  (2)  4 byte big-endian integer which is the number of valid page records
**       in the journal.  If this value is 0xffffffff, then compute the
**       number of page records from the journal size.
**  (3)  4 byte big-endian integer which is the initial value for the
**       sanity checksum.
**  (4)  4 byte integer which is the number of pages to truncate the
**       database to during a rollback.
**  (5)  4 byte big-endian integer which is the sector size.  The header
**       is this many bytes in size.
**  (6)  4 byte big-endian integer which is the page size.
**  (7)  zero padding out to the next sector size.
**  (8)  Zero or more pages instances, each as follows:
**        +  4 byte page number.
**        +  pPager->pageSize bytes of data.
**        +  4 byte checksum
**
** When we speak of the journal header, we mean the first 7 items above.
** Each entry in the journal is an instance of the 8th item.
**
** Call the value from the second bullet "nRec".  nRec is the number of
** valid page entries in the journal.  In most cases, you can compute the
** value of nRec from the size of the journal file.  But if a power
** failure occurred while the journal was being written, it could be the
** case that the size of the journal file had already been increased but
** the extra entries had not yet made it safely to disk.  In such a case,
** the value of nRec computed from the file size would be too large.  For
** that reason, we always use the nRec value in the header.
**
** If the nRec value is 0xffffffff it means that nRec should be computed
** from the file size.  This value is used when the user selects the
** no-sync option for the journal.  A power failure could lead to corruption
** in this case.  But for things like temporary table (which will be
** deleted when the power is restored) we don't care. 
**
** If the file opened as the journal file is not a well-formed
** journal file then all pages up to the first corrupted page are rolled
** back (or no pages if the journal header is corrupted). The journal file
** is then deleted and SQLITE_OK returned, just as if no corruption had
** been encountered.
**
** If an I/O or malloc() error occurs, the journal-file is not deleted
** and an error code is returned.
**
** The isHot parameter indicates that we are trying to rollback a journal
** that might be a hot journal.  Or, it could be that the journal is
** preserved because of JOURNALMODE_PERSIST or JOURNALMODE_TRUNCATE.
** If the journal really is hot, reset the pager cache prior rolling
** back any content.  If the journal is merely persistent, no reset is
** needed.
*/
static int pager_playback(Pager *pPager, int isHot){
  sqlite3_vfs *pVfs = pPager->pVfs;
  i64 szJ;                 /* Size of the journal file in bytes */
  u32 nRec;                /* Number of Records in the journal */
  u32 u;                   /* Unsigned loop counter */
  Pgno mxPg = 0;           /* Size of the original file in pages */
  int rc;                  /* Result code of a subroutine */
  int res = 1;             /* Value returned by sqlite3OsAccess() */
  char *zSuper = 0;        /* Name of super-journal file if any */
  int needPagerReset;      /* True to reset page prior to first page rollback */
  int nPlayback = 0;       /* Total number of pages restored from journal */
  u32 savedPageSize = pPager->pageSize;

  /* Figure out how many records are in the journal.  Abort early if
  ** the journal is empty.
  */
  assert( isOpen(pPager->jfd) );
  rc = sqlite3OsFileSize(pPager->jfd, &szJ);
  if( rc!=SQLITE_OK ){
    goto end_playback;
  }

  /* Read the super-journal name from the journal, if it is present.
  ** If a super-journal file name is specified, but the file is not
  ** present on disk, then the journal is not hot and does not need to be
  ** played back.
  **
  ** TODO: Technically the following is an error because it assumes that
  ** buffer Pager.pTmpSpace is (mxPathname+1) bytes or larger. i.e. that
  ** (pPager->pageSize >= pPager->pVfs->mxPathname+1). Using os_unix.c,
  ** mxPathname is 512, which is the same as the minimum allowable value
  ** for pageSize.
  */
  zSuper = pPager->pTmpSpace;
  rc = readSuperJournal(pPager->jfd, zSuper, pPager->pVfs->mxPathname+1);
  if( rc==SQLITE_OK && zSuper[0] ){
    rc = sqlite3OsAccess(pVfs, zSuper, SQLITE_ACCESS_EXISTS, &res);
  }
  zSuper = 0;
  if( rc!=SQLITE_OK || !res ){
    goto end_playback;
  }
  pPager->journalOff = 0;
  needPagerReset = isHot;

  /* This loop terminates either when a readJournalHdr() or
  ** pager_playback_one_page() call returns SQLITE_DONE or an IO error
  ** occurs.
  */
  while( 1 ){
    /* Read the next journal header from the journal file.  If there are
    ** not enough bytes left in the journal file for a complete header, or
    ** it is corrupted, then a process must have failed while writing it.
    ** This indicates nothing more needs to be rolled back.
    */
    rc = readJournalHdr(pPager, isHot, szJ, &nRec, &mxPg);
    if( rc!=SQLITE_OK ){
      if( rc==SQLITE_DONE ){
        rc = SQLITE_OK;
      }
      goto end_playback;
    }

    /* If nRec is 0xffffffff, then this journal was created by a process
    ** working in no-sync mode. This means that the rest of the journal
    ** file consists of pages, there are no more journal headers. Compute
    ** the value of nRec based on this assumption.
    */
    if( nRec==0xffffffff ){
      assert( pPager->journalOff==JOURNAL_HDR_SZ(pPager) );
      nRec = (int)((szJ - JOURNAL_HDR_SZ(pPager))/JOURNAL_PG_SZ(pPager));
    }

    /* If nRec is 0 and this rollback is of a transaction created by this
    ** process and if this is the final header in the journal, then it means
    ** that this part of the journal was being filled but has not yet been
    ** synced to disk.  Compute the number of pages based on the remaining
    ** size of the file.
    **
    ** The third term of the test was added to fix ticket #2565.
    ** When rolling back a hot journal, nRec==0 always means that the next
    ** chunk of the journal contains zero pages to be rolled back.  But
    ** when doing a ROLLBACK and the nRec==0 chunk is the last chunk in
    ** the journal, it means that the journal might contain additional
    ** pages that need to be rolled back and that the number of pages
    ** should be computed based on the journal file size.
    */
    if( nRec==0 && !isHot &&
        pPager->journalHdr+JOURNAL_HDR_SZ(pPager)==pPager->journalOff ){
      nRec = (int)((szJ - pPager->journalOff) / JOURNAL_PG_SZ(pPager));
    }

    /* If this is the first header read from the journal, truncate the
    ** database file back to its original size.
    */
    if( pPager->journalOff==JOURNAL_HDR_SZ(pPager) ){
      rc = pager_truncate(pPager, mxPg);
      if( rc!=SQLITE_OK ){
        goto end_playback;
      }
      pPager->dbSize = mxPg;
      if( pPager->mxPgno<mxPg ){
        pPager->mxPgno = mxPg;
      }
    }

    /* Copy original pages out of the journal and back into the
    ** database file and/or page cache.
    */
    for(u=0; u<nRec; u++){
      if( needPagerReset ){
        pager_reset(pPager);
        needPagerReset = 0;
      }
      rc = pager_playback_one_page(pPager,&pPager->journalOff,0,1,0);
      if( rc==SQLITE_OK ){
        nPlayback++;
      }else{
        if( rc==SQLITE_DONE ){
          pPager->journalOff = szJ;
          break;
        }else if( rc==SQLITE_IOERR_SHORT_READ ){
          /* If the journal has been truncated, simply stop reading and
          ** processing the journal. This might happen if the journal was
          ** not completely written and synced prior to a crash.  In that
          ** case, the database should have never been written in the
          ** first place so it is OK to simply abandon the rollback. */
          rc = SQLITE_OK;
          goto end_playback;
        }else{
          /* If we are unable to rollback, quit and return the error
          ** code.  This will cause the pager to enter the error state
          ** so that no further harm will be done.  Perhaps the next
          ** process to come along will be able to rollback the database.
          */
          goto end_playback;
        }
      }
    }
  }
  /*NOTREACHED*/
  assert( 0 );

end_playback:
  if( rc==SQLITE_OK ){
    rc = sqlite3PagerSetPagesize(pPager, &savedPageSize, -1);
  }
  /* Following a rollback, the database file should be back in its original
  ** state prior to the start of the transaction, so invoke the
  ** SQLITE_FCNTL_DB_UNCHANGED file-control method to disable the
  ** assertion that the transaction counter was modified.
  */
#ifdef SQLITE_DEBUG
  sqlite3OsFileControlHint(pPager->fd,SQLITE_FCNTL_DB_UNCHANGED,0);
#endif

  /* If this playback is happening automatically as a result of an IO or
  ** malloc error that occurred after the change-counter was updated but
  ** before the transaction was committed, then the change-counter
  ** modification may just have been reverted. If this happens in exclusive
  ** mode, then subsequent transactions performed by the connection will not
  ** update the change-counter at all. This may lead to cache inconsistency
  ** problems for other processes at some point in the future. So, just
  ** in case this has happened, clear the changeCountDone flag now.
  */
  pPager->changeCountDone = pPager->tempFile;

  if( rc==SQLITE_OK ){
    /* Leave 4 bytes of space before the super-journal filename in memory.
    ** This is because it may end up being passed to sqlite3OsOpen(), in
    ** which case it requires 4 0x00 bytes in memory immediately before
    ** the filename. */
    zSuper = &pPager->pTmpSpace[4];
    rc = readSuperJournal(pPager->jfd, zSuper, pPager->pVfs->mxPathname+1);
    testcase( rc!=SQLITE_OK );
  }
  if( rc==SQLITE_OK
   && (pPager->eState>=PAGER_WRITER_DBMOD || pPager->eState==PAGER_OPEN)
  ){
    rc = sqlite3PagerSync(pPager, 0);
  }
  if( rc==SQLITE_OK ){
    rc = pager_end_transaction(pPager, zSuper[0]!='\0', 0);
    testcase( rc!=SQLITE_OK );
  }
  if( rc==SQLITE_OK && zSuper[0] && res ){
    /* If there was a super-journal and this routine will return success,
    ** see if it is possible to delete the super-journal.
    */
    assert( zSuper==&pPager->pTmpSpace[4] );
    memset(pPager->pTmpSpace, 0, 4);
    rc = pager_delsuper(pPager, zSuper);
    testcase( rc!=SQLITE_OK );
  }
  if( isHot && nPlayback ){
    sqlite3_log(SQLITE_NOTICE_RECOVER_ROLLBACK, "recovered %d pages from %s",
                nPlayback, pPager->zJournal);
  }

  /* The Pager.sectorSize variable may have been updated while rolling
  ** back a journal created by a process with a different sector size
  ** value. Reset it to the correct value for this process.
  */
  setSectorSize(pPager);
  return rc;
}


/*
** Read the content for page pPg out of the database file (or out of
** the WAL if that is where the most recent copy if found) into
** pPg->pData. A shared lock or greater must be held on the database
** file before this function is called.
**
** If page 1 is read, then the value of Pager.dbFileVers[] is set to
** the value read from the database file.
**
** If an IO error occurs, then the IO error is returned to the caller.
** Otherwise, SQLITE_OK is returned.
*/
static int readDbPage(PgHdr *pPg){
  Pager *pPager = pPg->pPager; /* Pager object associated with page pPg */
  int rc = SQLITE_OK;          /* Return code */

#ifndef SQLITE_OMIT_WAL
  u32 iFrame = 0;              /* Frame of WAL containing pgno */

  assert( pPager->eState>=PAGER_READER && !MEMDB );
  assert( isOpen(pPager->fd) );

  if( pagerUseWal(pPager) ){
    rc = sqlite3WalFindFrame(pPager->pWal, pPg->pgno, &iFrame);
    if( rc ) return rc;
  }
  if( iFrame ){
    rc = sqlite3WalReadFrame(pPager->pWal, iFrame,pPager->pageSize,pPg->pData);
  }else
#endif
  {
    i64 iOffset = (pPg->pgno-1)*(i64)pPager->pageSize;
    rc = sqlite3OsRead(pPager->fd, pPg->pData, pPager->pageSize, iOffset);
    if( rc==SQLITE_IOERR_SHORT_READ ){
      rc = SQLITE_OK;
    }
  }

  if( pPg->pgno==1 ){
    if( rc ){
      /* If the read is unsuccessful, set the dbFileVers[] to something
      ** that will never be a valid file version.  dbFileVers[] is a copy
      ** of bytes 24..39 of the database.  Bytes 28..31 should always be
      ** zero or the size of the database in page. Bytes 32..35 and 35..39
      ** should be page numbers which are never 0xffffffff.  So filling
      ** pPager->dbFileVers[] with all 0xff bytes should suffice.
      **
      ** For an encrypted database, the situation is more complex:  bytes
      ** 24..39 of the database are white noise.  But the probability of
      ** white noise equaling 16 bytes of 0xff is vanishingly small so
      ** we should still be ok.
      */
      memset(pPager->dbFileVers, 0xff, sizeof(pPager->dbFileVers));
    }else{
      u8 *dbFileVers = &((u8*)pPg->pData)[24];
      memcpy(&pPager->dbFileVers, dbFileVers, sizeof(pPager->dbFileVers));
    }
  }
  PAGER_INCR(sqlite3_pager_readdb_count);
  PAGER_INCR(pPager->nRead);
  IOTRACE(("PGIN %p %d\n", pPager, pPg->pgno));
  PAGERTRACE(("FETCH %d page %d hash(%08x)\n",
               PAGERID(pPager), pPg->pgno, pager_pagehash(pPg)));

  return rc;
}

/*
** Update the value of the change-counter at offsets 24 and 92 in
** the header and the sqlite version number at offset 96.
**
** This is an unconditional update.  See also the pager_incr_changecounter()
** routine which only updates the change-counter if the update is actually
** needed, as determined by the pPager->changeCountDone state variable.
*/
static void pager_write_changecounter(PgHdr *pPg){
  u32 change_counter;
  if( NEVER(pPg==0) ) return;

  /* Increment the value just read and write it back to byte 24. */
  change_counter = sqlite3Get4byte((u8*)pPg->pPager->dbFileVers)+1;
  put32bits(((char*)pPg->pData)+24, change_counter);

  /* Also store the SQLite version number in bytes 96..99 and in
  ** bytes 92..95 store the change counter for which the version number
  ** is valid. */
  put32bits(((char*)pPg->pData)+92, change_counter);
  put32bits(((char*)pPg->pData)+96, SQLITE_VERSION_NUMBER);
}

#ifndef SQLITE_OMIT_WAL
/*
** This function is invoked once for each page that has already been
** written into the log file when a WAL transaction is rolled back.
** Parameter iPg is the page number of said page. The pCtx argument
** is actually a pointer to the Pager structure.
**
** If page iPg is present in the cache, and has no outstanding references,
** it is discarded. Otherwise, if there are one or more outstanding
** references, the page content is reloaded from the database. If the
** attempt to reload content from the database is required and fails,
** return an SQLite error code. Otherwise, SQLITE_OK.
*/
static int pagerUndoCallback(void *pCtx, Pgno iPg){
  int rc = SQLITE_OK;
  Pager *pPager = (Pager *)pCtx;
  PgHdr *pPg;

  assert( pagerUseWal(pPager) );
  pPg = sqlite3PagerLookup(pPager, iPg);
  if( pPg ){
    if( sqlite3PcachePageRefcount(pPg)==1 ){
      sqlite3PcacheDrop(pPg);
    }else{
      rc = readDbPage(pPg);
      if( rc==SQLITE_OK ){
        pPager->xReiniter(pPg);
      }
      sqlite3PagerUnrefNotNull(pPg);
    }
  }

  /* Normally, if a transaction is rolled back, any backup processes are
  ** updated as data is copied out of the rollback journal and into the
  ** database. This is not generally possible with a WAL database, as
  ** rollback involves simply truncating the log file. Therefore, if one
  ** or more frames have already been written to the log (and therefore
  ** also copied into the backup databases) as part of this transaction,
  ** the backups must be restarted.
  */
  sqlite3BackupRestart(pPager->pBackup);

  return rc;
}

/*
** This function is called to rollback a transaction on a WAL database.
*/
static int pagerRollbackWal(Pager *pPager){
  int rc;                         /* Return Code */
  PgHdr *pList;                   /* List of dirty pages to revert */

  /* For all pages in the cache that are currently dirty or have already
  ** been written (but not committed) to the log file, do one of the
  ** following:
  **
  **   + Discard the cached page (if refcount==0), or
  **   + Reload page content from the database (if refcount>0).
  */
  pPager->dbSize = pPager->dbOrigSize;
  rc = sqlite3WalUndo(pPager->pWal, pagerUndoCallback, (void *)pPager);
  pList = sqlite3PcacheDirtyList(pPager->pPCache);
  while( pList && rc==SQLITE_OK ){
    PgHdr *pNext = pList->pDirty;
    rc = pagerUndoCallback((void *)pPager, pList->pgno);
    pList = pNext;
  }

  return rc;
}

/*
** This function is a wrapper around sqlite3WalFrames(). As well as logging
** the contents of the list of pages headed by pList (connected by pDirty),
** this function notifies any active backup processes that the pages have
** changed.
**
** The list of pages passed into this routine is always sorted by page number.
** Hence, if page 1 appears anywhere on the list, it will be the first page.
*/
static int pagerWalFrames(
  Pager *pPager,                  /* Pager object */
  PgHdr *pList,                   /* List of frames to log */
  Pgno nTruncate,                 /* Database size after this commit */
  int isCommit                    /* True if this is a commit */
){
  int rc;                         /* Return code */
  int nList;                      /* Number of pages in pList */
  PgHdr *p;                       /* For looping over pages */

  assert( pPager->pWal );
  assert( pList );
#ifdef SQLITE_DEBUG
  /* Verify that the page list is in ascending order */
  for(p=pList; p && p->pDirty; p=p->pDirty){
    assert( p->pgno < p->pDirty->pgno );
  }
#endif

  assert( pList->pDirty==0 || isCommit );
  if( isCommit ){
    /* If a WAL transaction is being committed, there is no point in writing
    ** any pages with page numbers greater than nTruncate into the WAL file.
    ** They will never be read by any client. So remove them from the pDirty
    ** list here. */
    PgHdr **ppNext = &pList;
    nList = 0;
    for(p=pList; (*ppNext = p)!=0; p=p->pDirty){
      if( p->pgno<=nTruncate ){
        ppNext = &p->pDirty;
        nList++;
      }
    }
    assert( pList );
  }else{
    nList = 1;
  }
  pPager->aStat[PAGER_STAT_WRITE] += nList;

  if( pList->pgno==1 ) pager_write_changecounter(pList);
  rc = sqlite3WalFrames(pPager->pWal,
      pPager->pageSize, pList, nTruncate, isCommit, pPager->walSyncFlags
  );
  if( rc==SQLITE_OK && pPager->pBackup ){
    for(p=pList; p; p=p->pDirty){
      sqlite3BackupUpdate(pPager->pBackup, p->pgno, (u8 *)p->pData);
    }
  }

#ifdef SQLITE_CHECK_PAGES
  pList = sqlite3PcacheDirtyList(pPager->pPCache);
  for(p=pList; p; p=p->pDirty){
    pager_set_pagehash(p);
  }
#endif

  return rc;
}

/*
** Begin a read transaction on the WAL.
**
** This routine used to be called "pagerOpenSnapshot()" because it essentially
** makes a snapshot of the database at the current point in time and preserves
** that snapshot for use by the reader in spite of concurrently changes by
** other writers or checkpointers.
*/
static int pagerBeginReadTransaction(Pager *pPager){
  int rc;                         /* Return code */
  int changed = 0;                /* True if cache must be reset */

  assert( pagerUseWal(pPager) );
  assert( pPager->eState==PAGER_OPEN || pPager->eState==PAGER_READER );

  /* sqlite3WalEndReadTransaction() was not called for the previous
  ** transaction in locking_mode=EXCLUSIVE.  So call it now.  If we
  ** are in locking_mode=NORMAL and EndRead() was previously called,
  ** the duplicate call is harmless.
  */
  sqlite3WalEndReadTransaction(pPager->pWal);

  rc = sqlite3WalBeginReadTransaction(pPager->pWal, &changed);
  if( rc!=SQLITE_OK || changed ){
    pager_reset(pPager);
    if( USEFETCH(pPager) ) sqlite3OsUnfetch(pPager->fd, 0, 0);
  }

  return rc;
}
#endif

/*
** This function is called as part of the transition from PAGER_OPEN
** to PAGER_READER state to determine the size of the database file
** in pages (assuming the page size currently stored in Pager.pageSize).
**
** If no error occurs, SQLITE_OK is returned and the size of the database
** in pages is stored in *pnPage. Otherwise, an error code (perhaps
** SQLITE_IOERR_FSTAT) is returned and *pnPage is left unmodified.
*/
static int pagerPagecount(Pager *pPager, Pgno *pnPage){
  Pgno nPage;                     /* Value to return via *pnPage */

  /* Query the WAL sub-system for the database size. The WalDbsize()
  ** function returns zero if the WAL is not open (i.e. Pager.pWal==0), or
  ** if the database size is not available. The database size is not
  ** available from the WAL sub-system if the log file is empty or
  ** contains no valid committed transactions.
  */
  assert( pPager->eState==PAGER_OPEN );
  assert( pPager->eLock>=SHARED_LOCK );
  assert( isOpen(pPager->fd) );
  assert( pPager->tempFile==0 );
  nPage = sqlite3WalDbsize(pPager->pWal);

  /* If the number of pages in the database is not available from the
  ** WAL sub-system, determine the page count based on the size of
  ** the database file.  If the size of the database file is not an
  ** integer multiple of the page-size, round up the result.
  */
  if( nPage==0 && ALWAYS(isOpen(pPager->fd)) ){
    i64 n = 0;                    /* Size of db file in bytes */
    int rc = sqlite3OsFileSize(pPager->fd, &n);
    if( rc!=SQLITE_OK ){
      return rc;
    }
    nPage = (Pgno)((n+pPager->pageSize-1) / pPager->pageSize);
  }

  /* If the current number of pages in the file is greater than the
  ** configured maximum pager number, increase the allowed limit so
  ** that the file can be read.
  */
  if( nPage>pPager->mxPgno ){
    pPager->mxPgno = (Pgno)nPage;
  }

  *pnPage = nPage;
  return SQLITE_OK;
}

#ifndef SQLITE_OMIT_WAL
/*
** Check if the *-wal file that corresponds to the database opened by pPager
** exists if the database is not empty, or verify that the *-wal file does
** not exist (by deleting it) if the database file is empty.
**
** If the database is not empty and the *-wal file exists, open the pager
** in WAL mode.  If the database is empty or if no *-wal file exists and
** if no error occurs, make sure Pager.journalMode is not set to
** PAGER_JOURNALMODE_WAL.
**
** Return SQLITE_OK or an error code.
**
** The caller must hold a SHARED lock on the database file to call this
** function. Because an EXCLUSIVE lock on the db file is required to delete
** a WAL on a none-empty database, this ensures there is no race condition
** between the xAccess() below and an xDelete() being executed by some
** other connection.
*/
static int pagerOpenWalIfPresent(Pager *pPager){
  int rc = SQLITE_OK;
  assert( pPager->eState==PAGER_OPEN );
  assert( pPager->eLock>=SHARED_LOCK );

  if( !pPager->tempFile ){
    int isWal;                    /* True if WAL file exists */
    rc = sqlite3OsAccess(
        pPager->pVfs, pPager->zWal, SQLITE_ACCESS_EXISTS, &isWal
    );
    if( rc==SQLITE_OK ){
      if( isWal ){
        Pgno nPage;                   /* Size of the database file */

        rc = pagerPagecount(pPager, &nPage);
        if( rc ) return rc;
        if( nPage==0 ){
          rc = sqlite3OsDelete(pPager->pVfs, pPager->zWal, 0);
        }else{
          testcase( sqlite3PcachePagecount(pPager->pPCache)==0 );
          rc = sqlite3PagerOpenWal(pPager, 0);
        }
      }else if( pPager->journalMode==PAGER_JOURNALMODE_WAL ){
        pPager->journalMode = PAGER_JOURNALMODE_DELETE;
      }
    }
  }
  return rc;
}
#endif

/*
** Playback savepoint pSavepoint. Or, if pSavepoint==NULL, then playback
** the entire super-journal file. The case pSavepoint==NULL occurs when
** a ROLLBACK TO command is invoked on a SAVEPOINT that is a transaction
** savepoint.
**
** When pSavepoint is not NULL (meaning a non-transaction savepoint is
** being rolled back), then the rollback consists of up to three stages,
** performed in the order specified:
**
**   * Pages are played back from the main journal starting at byte
**     offset PagerSavepoint.iOffset and continuing to
**     PagerSavepoint.iHdrOffset, or to the end of the main journal
**     file if PagerSavepoint.iHdrOffset is zero.
**
**   * If PagerSavepoint.iHdrOffset is not zero, then pages are played
**     back starting from the journal header immediately following
**     PagerSavepoint.iHdrOffset to the end of the main journal file.
**
**   * Pages are then played back from the sub-journal file, starting
**     with the PagerSavepoint.iSubRec and continuing to the end of
**     the journal file.
**
** Throughout the rollback process, each time a page is rolled back, the
** corresponding bit is set in a bitvec structure (variable pDone in the
** implementation below). This is used to ensure that a page is only
** rolled back the first time it is encountered in either journal.
**
** If pSavepoint is NULL, then pages are only played back from the main
** journal file. There is no need for a bitvec in this case.
**
** In either case, before playback commences the Pager.dbSize variable
** is reset to the value that it held at the start of the savepoint
** (or transaction). No page with a page-number greater than this value
** is played back. If one is encountered it is simply skipped.
*/
static int pagerPlaybackSavepoint(Pager *pPager, PagerSavepoint *pSavepoint){
  i64 szJ;                 /* Effective size of the main journal */
  i64 iHdrOff;             /* End of first segment of main-journal records */
  int rc = SQLITE_OK;      /* Return code */
  Bitvec *pDone = 0;       /* Bitvec to ensure pages played back only once */

  assert( pPager->eState!=PAGER_ERROR );
  assert( pPager->eState>=PAGER_WRITER_LOCKED );

  /* Allocate a bitvec to use to store the set of pages rolled back */
  if( pSavepoint ){
    pDone = sqlite3BitvecCreate(pSavepoint->nOrig);
    if( !pDone ){
      return SQLITE_NOMEM_BKPT;
    }
  }

  /* Set the database size back to the value it was before the savepoint
  ** being reverted was opened.
  */
  pPager->dbSize = pSavepoint ? pSavepoint->nOrig : pPager->dbOrigSize;
  pPager->changeCountDone = pPager->tempFile;

  if( !pSavepoint && pagerUseWal(pPager) ){
    return pagerRollbackWal(pPager);
  }

  /* Use pPager->journalOff as the effective size of the main rollback
  ** journal.  The actual file might be larger than this in
  ** PAGER_JOURNALMODE_TRUNCATE or PAGER_JOURNALMODE_PERSIST.  But anything
  ** past pPager->journalOff is off-limits to us.
  */
  szJ = pPager->journalOff;
  assert( pagerUseWal(pPager)==0 || szJ==0 );

  /* Begin by rolling back records from the main journal starting at
  ** PagerSavepoint.iOffset and continuing to the next journal header.
  ** There might be records in the main journal that have a page number
  ** greater than the current database size (pPager->dbSize) but those
  ** will be skipped automatically.  Pages are added to pDone as they
  ** are played back.
  */
  if( pSavepoint && !pagerUseWal(pPager) ){
    iHdrOff = pSavepoint->iHdrOffset ? pSavepoint->iHdrOffset : szJ;
    pPager->journalOff = pSavepoint->iOffset;
    while( rc==SQLITE_OK && pPager->journalOff<iHdrOff ){
      rc = pager_playback_one_page(pPager, &pPager->journalOff, pDone, 1, 1);
    }
    assert( rc!=SQLITE_DONE );
  }else{
    pPager->journalOff = 0;
  }

  /* Continue rolling back records out of the main journal starting at
  ** the first journal header seen and continuing until the effective end
  ** of the main journal file.  Continue to skip out-of-range pages and
  ** continue adding pages rolled back to pDone.
  */
  while( rc==SQLITE_OK && pPager->journalOff<szJ ){
    u32 ii;            /* Loop counter */
    u32 nJRec = 0;     /* Number of Journal Records */
    u32 dummy;
    rc = readJournalHdr(pPager, 0, szJ, &nJRec, &dummy);
    assert( rc!=SQLITE_DONE );

    /*
    ** The "pPager->journalHdr+JOURNAL_HDR_SZ(pPager)==pPager->journalOff"
    ** test is related to ticket #2565.  See the discussion in the
    ** pager_playback() function for additional information.
    */
    if( nJRec==0
     && pPager->journalHdr+JOURNAL_HDR_SZ(pPager)==pPager->journalOff
    ){
      nJRec = (u32)((szJ - pPager->journalOff)/JOURNAL_PG_SZ(pPager));
    }
    for(ii=0; rc==SQLITE_OK && ii<nJRec && pPager->journalOff<szJ; ii++){
      rc = pager_playback_one_page(pPager, &pPager->journalOff, pDone, 1, 1);
    }
    assert( rc!=SQLITE_DONE );
  }
  assert( rc!=SQLITE_OK || pPager->journalOff>=szJ );

  /* Finally,  rollback pages from the sub-journal.  Page that were
  ** previously rolled back out of the main journal (and are hence in pDone)
  ** will be skipped.  Out-of-range pages are also skipped.
  */
  if( pSavepoint ){
    u32 ii;            /* Loop counter */
    i64 offset = (i64)pSavepoint->iSubRec*(4+pPager->pageSize);

    if( pagerUseWal(pPager) ){
      rc = sqlite3WalSavepointUndo(pPager->pWal, pSavepoint->aWalData);
    }
    for(ii=pSavepoint->iSubRec; rc==SQLITE_OK && ii<pPager->nSubRec; ii++){
      assert( offset==(i64)ii*(4+pPager->pageSize) );
      rc = pager_playback_one_page(pPager, &offset, pDone, 0, 1);
    }
    assert( rc!=SQLITE_DONE );
  }

  sqlite3BitvecDestroy(pDone);
  if( rc==SQLITE_OK ){
    pPager->journalOff = szJ;
  }

  return rc;
}

/*
** Change the maximum number of in-memory pages that are allowed
** before attempting to recycle clean and unused pages.
*/
void sqlite3PagerSetCachesize(Pager *pPager, int mxPage){
  sqlite3PcacheSetCachesize(pPager->pPCache, mxPage);
}

/*
** Change the maximum number of in-memory pages that are allowed
** before attempting to spill pages to journal.
*/
int sqlite3PagerSetSpillsize(Pager *pPager, int mxPage){
  return sqlite3PcacheSetSpillsize(pPager->pPCache, mxPage);
}

/*
** Invoke SQLITE_FCNTL_MMAP_SIZE based on the current value of szMmap.
*/
static void pagerFixMaplimit(Pager *pPager){
#if SQLITE_MAX_MMAP_SIZE>0
  sqlite3_file *fd = pPager->fd;
  if( isOpen(fd) && fd->pMethods->iVersion>=3 ){
    sqlite3_int64 sz;
    sz = pPager->szMmap;
    pPager->bUseFetch = (sz>0);
    setGetterMethod(pPager);
    sqlite3OsFileControlHint(pPager->fd, SQLITE_FCNTL_MMAP_SIZE, &sz);
  }
#endif
}

/*
** Change the maximum size of any memory mapping made of the database file.
*/
void sqlite3PagerSetMmapLimit(Pager *pPager, sqlite3_int64 szMmap){
  pPager->szMmap = szMmap;
  pagerFixMaplimit(pPager);
}

/*
** Free as much memory as possible from the pager.
*/
void sqlite3PagerShrink(Pager *pPager){
  sqlite3PcacheShrink(pPager->pPCache);
}

/*
** Adjust settings of the pager to those specified in the pgFlags parameter.
**
** The "level" in pgFlags & PAGER_SYNCHRONOUS_MASK sets the robustness
** of the database to damage due to OS crashes or power failures by
** changing the number of syncs()s when writing the journals.
** There are four levels:
**
**    OFF       sqlite3OsSync() is never called.  This is the default
**              for temporary and transient files.
**
**    NORMAL    The journal is synced once before writes begin on the
**              database.  This is normally adequate protection, but
**              it is theoretically possible, though very unlikely,
**              that an inopertune power failure could leave the journal
**              in a state which would cause damage to the database
**              when it is rolled back.
**
**    FULL      The journal is synced twice before writes begin on the
**              database (with some additional information - the nRec field
**              of the journal header - being written in between the two
**              syncs).  If we assume that writing a
**              single disk sector is atomic, then this mode provides
**              assurance that the journal will not be corrupted to the
**              point of causing damage to the database during rollback.
**
**    EXTRA     This is like FULL except that is also syncs the directory
**              that contains the rollback journal after the rollback
**              journal is unlinked.
**
** The above is for a rollback-journal mode.  For WAL mode, OFF continues
** to mean that no syncs ever occur.  NORMAL means that the WAL is synced
** prior to the start of checkpoint and that the database file is synced
** at the conclusion of the checkpoint if the entire content of the WAL
** was written back into the database.  But no sync operations occur for
** an ordinary commit in NORMAL mode with WAL.  FULL means that the WAL
** file is synced following each commit operation, in addition to the
** syncs associated with NORMAL.  There is no difference between FULL
** and EXTRA for WAL mode.
**
** Do not confuse synchronous=FULL with SQLITE_SYNC_FULL.  The
** SQLITE_SYNC_FULL macro means to use the MacOSX-style full-fsync
** using fcntl(F_FULLFSYNC).  SQLITE_SYNC_NORMAL means to do an
** ordinary fsync() call.  There is no difference between SQLITE_SYNC_FULL
** and SQLITE_SYNC_NORMAL on platforms other than MacOSX.  But the
** synchronous=FULL versus synchronous=NORMAL setting determines when
** the xSync primitive is called and is relevant to all platforms.
**
** Numeric values associated with these states are OFF==1, NORMAL=2,
** and FULL=3.
*/
void sqlite3PagerSetFlags(
  Pager *pPager,        /* The pager to set safety level for */
  unsigned pgFlags      /* Various flags */
){
  unsigned level = pgFlags & PAGER_SYNCHRONOUS_MASK;
  if( pPager->tempFile ){
    pPager->noSync = 1;
    pPager->fullSync = 0;
    pPager->extraSync = 0;
  }else{
    pPager->noSync =  level==PAGER_SYNCHRONOUS_OFF ?1:0;
    pPager->fullSync = level>=PAGER_SYNCHRONOUS_FULL ?1:0;
    pPager->extraSync = level==PAGER_SYNCHRONOUS_EXTRA ?1:0;
  }
  if( pPager->noSync ){
    pPager->syncFlags = 0;
  }else if( pgFlags & PAGER_FULLFSYNC ){
    pPager->syncFlags = SQLITE_SYNC_FULL;
  }else{
    pPager->syncFlags = SQLITE_SYNC_NORMAL;
  }
  pPager->walSyncFlags = (pPager->syncFlags<<2);
  if( pPager->fullSync ){
    pPager->walSyncFlags |= pPager->syncFlags;
  }
  if( (pgFlags & PAGER_CKPT_FULLFSYNC) && !pPager->noSync ){
    pPager->walSyncFlags |= (SQLITE_SYNC_FULL<<2);
  }
  if( pgFlags & PAGER_CACHESPILL ){
    pPager->doNotSpill &= ~SPILLFLAG_OFF;
  }else{
    pPager->doNotSpill |= SPILLFLAG_OFF;
  }
}

/*
** The following global variable is incremented whenever the library
** attempts to open a temporary file.  This information is used for
** testing and analysis only. 
*/
#ifdef SQLITE_TEST
int sqlite3_opentemp_count = 0;
#endif

/*
** Open a temporary file.
**
** Write the file descriptor into *pFile. Return SQLITE_OK on success
** or some other error code if we fail. The OS will automatically
** delete the temporary file when it is closed.
**
** The flags passed to the VFS layer xOpen() call are those specified
** by parameter vfsFlags ORed with the following:
**
**     SQLITE_OPEN_READWRITE
**     SQLITE_OPEN_CREATE
**     SQLITE_OPEN_EXCLUSIVE
**     SQLITE_OPEN_DELETEONCLOSE
*/
static int pagerOpentemp(
  Pager *pPager,        /* The pager object */
  sqlite3_file *pFile,  /* Write the file descriptor here */
  int vfsFlags          /* Flags passed through to the VFS */
){
  int rc;               /* Return code */

#ifdef SQLITE_TEST
  sqlite3_opentemp_count++;  /* Used for testing and analysis only */
#endif

  vfsFlags |=  SQLITE_OPEN_READWRITE | SQLITE_OPEN_CREATE |
            SQLITE_OPEN_EXCLUSIVE | SQLITE_OPEN_DELETEONCLOSE;
  rc = sqlite3OsOpen(pPager->pVfs, 0, pFile, vfsFlags, 0);
  assert( rc!=SQLITE_OK || isOpen(pFile) );
  return rc;
}

/*
** Set the busy handler function.
**
** The pager invokes the busy-handler if sqlite3OsLock() returns
** SQLITE_BUSY when trying to upgrade from no-lock to a SHARED lock,
** or when trying to upgrade from a RESERVED lock to an EXCLUSIVE
** lock. It does *not* invoke the busy handler when upgrading from
** SHARED to RESERVED, or when upgrading from SHARED to EXCLUSIVE
** (which occurs during hot-journal rollback). Summary:
**
**   Transition                        | Invokes xBusyHandler
**   --------------------------------------------------------
**   NO_LOCK       -> SHARED_LOCK      | Yes
**   SHARED_LOCK   -> RESERVED_LOCK    | No
**   SHARED_LOCK   -> EXCLUSIVE_LOCK   | No
**   RESERVED_LOCK -> EXCLUSIVE_LOCK   | Yes
**
** If the busy-handler callback returns non-zero, the lock is
** retried. If it returns zero, then the SQLITE_BUSY error is
** returned to the caller of the pager API function.
*/
void sqlite3PagerSetBusyHandler(
  Pager *pPager,                       /* Pager object */
  int (*xBusyHandler)(void *),         /* Pointer to busy-handler function */
  void *pBusyHandlerArg                /* Argument to pass to xBusyHandler */
){
  void **ap;
  pPager->xBusyHandler = xBusyHandler;
  pPager->pBusyHandlerArg = pBusyHandlerArg;
  ap = (void **)&pPager->xBusyHandler;
  assert( ((int(*)(void *))(ap[0]))==xBusyHandler );
  assert( ap[1]==pBusyHandlerArg );
  sqlite3OsFileControlHint(pPager->fd, SQLITE_FCNTL_BUSYHANDLER, (void *)ap);
}

/*
** Change the page size used by the Pager object. The new page size
** is passed in *pPageSize.
**
** If the pager is in the error state when this function is called, it
** is a no-op. The value returned is the error state error code (i.e.
** one of SQLITE_IOERR, an SQLITE_IOERR_xxx sub-code or SQLITE_FULL).
**
** Otherwise, if all of the following are true:
**
**   * the new page size (value of *pPageSize) is valid (a power
**     of two between 512 and SQLITE_MAX_PAGE_SIZE, inclusive), and
**
**   * there are no outstanding page references, and
**
**   * the database is either not an in-memory database or it is
**     an in-memory database that currently consists of zero pages.
**
** then the pager object page size is set to *pPageSize.
**
** If the page size is changed, then this function uses sqlite3PagerMalloc()
** to obtain a new Pager.pTmpSpace buffer. If this allocation attempt
** fails, SQLITE_NOMEM is returned and the page size remains unchanged.
** In all other cases, SQLITE_OK is returned.
**
** If the page size is not changed, either because one of the enumerated
** conditions above is not true, the pager was in error state when this
** function was called, or because the memory allocation attempt failed,
** then *pPageSize is set to the old, retained page size before returning.
*/
int sqlite3PagerSetPagesize(Pager *pPager, u32 *pPageSize, int nReserve){
  int rc = SQLITE_OK;

  /* It is not possible to do a full assert_pager_state() here, as this
  ** function may be called from within PagerOpen(), before the state
  ** of the Pager object is internally consistent.
  **
  ** At one point this function returned an error if the pager was in
  ** PAGER_ERROR state. But since PAGER_ERROR state guarantees that
  ** there is at least one outstanding page reference, this function
  ** is a no-op for that case anyhow.
  */

  u32 pageSize = *pPageSize;
  assert( pageSize==0 || (pageSize>=512 && pageSize<=SQLITE_MAX_PAGE_SIZE) );
  if( (pPager->memDb==0 || pPager->dbSize==0)
   && sqlite3PcacheRefCount(pPager->pPCache)==0
   && pageSize && pageSize!=(u32)pPager->pageSize
  ){
    char *pNew = NULL;             /* New temp space */
    i64 nByte = 0;

    if( pPager->eState>PAGER_OPEN && isOpen(pPager->fd) ){
      rc = sqlite3OsFileSize(pPager->fd, &nByte);
    }
    if( rc==SQLITE_OK ){
      /* 8 bytes of zeroed overrun space is sufficient so that the b-tree
      * cell header parser will never run off the end of the allocation */
      pNew = (char *)sqlite3PageMalloc(pageSize+8);
      if( !pNew ){
        rc = SQLITE_NOMEM_BKPT;
      }else{
        memset(pNew+pageSize, 0, 8);
      }
    }

    if( rc==SQLITE_OK ){
      pager_reset(pPager);
      rc = sqlite3PcacheSetPageSize(pPager->pPCache, pageSize);
    }
    if( rc==SQLITE_OK ){
      sqlite3PageFree(pPager->pTmpSpace);
      pPager->pTmpSpace = pNew;
      pPager->dbSize = (Pgno)((nByte+pageSize-1)/pageSize);
      pPager->pageSize = pageSize;
      pPager->lckPgno = (Pgno)(PENDING_BYTE/pageSize) + 1;
    }else{
      sqlite3PageFree(pNew);
    }
  }

  *pPageSize = pPager->pageSize;
  if( rc==SQLITE_OK ){
    if( nReserve<0 ) nReserve = pPager->nReserve;
    assert( nReserve>=0 && nReserve<1000 );
    pPager->nReserve = (i16)nReserve;
    pagerFixMaplimit(pPager);
  }
  return rc;
}

/*
** Return a pointer to the "temporary page" buffer held internally
** by the pager.  This is a buffer that is big enough to hold the
** entire content of a database page.  This buffer is used internally
** during rollback and will be overwritten whenever a rollback
** occurs.  But other modules are free to use it too, as long as
** no rollbacks are happening.
*/
void *sqlite3PagerTempSpace(Pager *pPager){
  return pPager->pTmpSpace;
}

/*
** Attempt to set the maximum database page count if mxPage is positive.
** Make no changes if mxPage is zero or negative.  And never reduce the
** maximum page count below the current size of the database.
**
** Regardless of mxPage, return the current maximum page count.
*/
Pgno sqlite3PagerMaxPageCount(Pager *pPager, Pgno mxPage){
  if( mxPage>0 ){
    pPager->mxPgno = mxPage;
  }
  assert( pPager->eState!=PAGER_OPEN );      /* Called only by OP_MaxPgcnt */
  /* assert( pPager->mxPgno>=pPager->dbSize ); */
  /* OP_MaxPgcnt ensures that the parameter passed to this function is not
  ** less than the total number of valid pages in the database. But this
  ** may be less than Pager.dbSize, and so the assert() above is not valid */
  return pPager->mxPgno;
}

/*
** The following set of routines are used to disable the simulated
** I/O error mechanism.  These routines are used to avoid simulated
** errors in places where we do not care about errors.
**
** Unless -DSQLITE_TEST=1 is used, these routines are all no-ops
** and generate no code.
*/
#ifdef SQLITE_TEST
extern int sqlite3_io_error_pending;
extern int sqlite3_io_error_hit;
static int saved_cnt;
void disable_simulated_io_errors(void){
  saved_cnt = sqlite3_io_error_pending;
  sqlite3_io_error_pending = -1;
}
void enable_simulated_io_errors(void){
  sqlite3_io_error_pending = saved_cnt;
}
#else
# define disable_simulated_io_errors()
# define enable_simulated_io_errors()
#endif

/*
** Read the first N bytes from the beginning of the file into memory
** that pDest points to.
**
** If the pager was opened on a transient file (zFilename==""), or
** opened on a file less than N bytes in size, the output buffer is
** zeroed and SQLITE_OK returned. The rationale for this is that this
** function is used to read database headers, and a new transient or
** zero sized database has a header than consists entirely of zeroes.
**
** If any IO error apart from SQLITE_IOERR_SHORT_READ is encountered,
** the error code is returned to the caller and the contents of the
** output buffer undefined.
*/
int sqlite3PagerReadFileheader(Pager *pPager, int N, unsigned char *pDest){
  int rc = SQLITE_OK;
  memset(pDest, 0, N);
  assert( isOpen(pPager->fd) || pPager->tempFile );

  /* This routine is only called by btree immediately after creating
  ** the Pager object.  There has not been an opportunity to transition
  ** to WAL mode yet.
  */
  assert( !pagerUseWal(pPager) );

  if( isOpen(pPager->fd) ){
    IOTRACE(("DBHDR %p 0 %d\n", pPager, N))
    rc = sqlite3OsRead(pPager->fd, pDest, N, 0);
    if( rc==SQLITE_IOERR_SHORT_READ ){
      rc = SQLITE_OK;
    }
  }
  return rc;
}

/*
** This function may only be called when a read-transaction is open on
** the pager. It returns the total number of pages in the database.
**
** However, if the file is between 1 and <page-size> bytes in size, then
** this is considered a 1 page file.
*/
void sqlite3PagerPagecount(Pager *pPager, int *pnPage){
  assert( pPager->eState>=PAGER_READER );
  assert( pPager->eState!=PAGER_WRITER_FINISHED );
  *pnPage = (int)pPager->dbSize;
}


/*
** Try to obtain a lock of type locktype on the database file. If
** a similar or greater lock is already held, this function is a no-op
** (returning SQLITE_OK immediately).
**
** Otherwise, attempt to obtain the lock using sqlite3OsLock(). Invoke
** the busy callback if the lock is currently not available. Repeat
** until the busy callback returns false or until the attempt to
** obtain the lock succeeds.
**
** Return SQLITE_OK on success and an error code if we cannot obtain
** the lock. If the lock is obtained successfully, set the Pager.state
** variable to locktype before returning.
*/
static int pager_wait_on_lock(Pager *pPager, int locktype){
  int rc;                              /* Return code */

  /* Check that this is either a no-op (because the requested lock is
  ** already held), or one of the transitions that the busy-handler
  ** may be invoked during, according to the comment above
  ** sqlite3PagerSetBusyhandler().
  */
  assert( (pPager->eLock>=locktype)
       || (pPager->eLock==NO_LOCK && locktype==SHARED_LOCK)
       || (pPager->eLock==RESERVED_LOCK && locktype==EXCLUSIVE_LOCK)
  );

  do {
    rc = pagerLockDb(pPager, locktype);
  }while( rc==SQLITE_BUSY && pPager->xBusyHandler(pPager->pBusyHandlerArg) );
  return rc;
}

/*
** Function assertTruncateConstraint(pPager) checks that one of the
** following is true for all dirty pages currently in the page-cache:
**
**   a) The page number is less than or equal to the size of the
**      current database image, in pages, OR
**
**   b) if the page content were written at this time, it would not
**      be necessary to write the current content out to the sub-journal.
**
** If the condition asserted by this function were not true, and the
** dirty page were to be discarded from the cache via the pagerStress()
** routine, pagerStress() would not write the current page content to
** the database file. If a savepoint transaction were rolled back after
** this happened, the correct behavior would be to restore the current
** content of the page. However, since this content is not present in either
** the database file or the portion of the rollback journal and
** sub-journal rolled back the content could not be restored and the
** database image would become corrupt. It is therefore fortunate that
** this circumstance cannot arise.
*/
#if defined(SQLITE_DEBUG)
static void assertTruncateConstraintCb(PgHdr *pPg){
  Pager *pPager = pPg->pPager;
  assert( pPg->flags&PGHDR_DIRTY );
  if( pPg->pgno>pPager->dbSize ){      /* if (a) is false */
    Pgno pgno = pPg->pgno;
    int i;
    for(i=0; i<pPg->pPager->nSavepoint; i++){
      PagerSavepoint *p = &pPager->aSavepoint[i];
      assert( p->nOrig<pgno || sqlite3BitvecTestNotNull(p->pInSavepoint,pgno) );
    }
  }
}
static void assertTruncateConstraint(Pager *pPager){
  sqlite3PcacheIterateDirty(pPager->pPCache, assertTruncateConstraintCb);
}
#else
# define assertTruncateConstraint(pPager)
#endif

/*
** Truncate the in-memory database file image to nPage pages. This
** function does not actually modify the database file on disk. It
** just sets the internal state of the pager object so that the
** truncation will be done when the current transaction is committed.
**
** This function is only called right before committing a transaction.
** Once this function has been called, the transaction must either be
** rolled back or committed. It is not safe to call this function and
** then continue writing to the database.
*/
void sqlite3PagerTruncateImage(Pager *pPager, Pgno nPage){
  assert( pPager->dbSize>=nPage || CORRUPT_DB );
  assert( pPager->eState>=PAGER_WRITER_CACHEMOD );
  pPager->dbSize = nPage;

  /* At one point the code here called assertTruncateConstraint() to
  ** ensure that all pages being truncated away by this operation are,
  ** if one or more savepoints are open, present in the savepoint
  ** journal so that they can be restored if the savepoint is rolled
  ** back. This is no longer necessary as this function is now only
  ** called right before committing a transaction. So although the
  ** Pager object may still have open savepoints (Pager.nSavepoint!=0),
  ** they cannot be rolled back. So the assertTruncateConstraint() call
  ** is no longer correct. */
}


/*
** This function is called before attempting a hot-journal rollback. It
** syncs the journal file to disk, then sets pPager->journalHdr to the
** size of the journal file so that the pager_playback() routine knows
** that the entire journal file has been synced.
**
** Syncing a hot-journal to disk before attempting to roll it back ensures
** that if a power-failure occurs during the rollback, the process that
** attempts rollback following system recovery sees the same journal
** content as this process.
**
** If everything goes as planned, SQLITE_OK is returned. Otherwise,
** an SQLite error code.
*/
static int pagerSyncHotJournal(Pager *pPager){
  int rc = SQLITE_OK;
  if( !pPager->noSync ){
    rc = sqlite3OsSync(pPager->jfd, SQLITE_SYNC_NORMAL);
  }
  if( rc==SQLITE_OK ){
    rc = sqlite3OsFileSize(pPager->jfd, &pPager->journalHdr);
  }
  return rc;
}

#if SQLITE_MAX_MMAP_SIZE>0
/*
** Obtain a reference to a memory mapped page object for page number pgno.
** The new object will use the pointer pData, obtained from xFetch().
** If successful, set *ppPage to point to the new page reference
** and return SQLITE_OK. Otherwise, return an SQLite error code and set
** *ppPage to zero.
**
** Page references obtained by calling this function should be released
** by calling pagerReleaseMapPage().
*/
static int pagerAcquireMapPage(
  Pager *pPager,                  /* Pager object */
  Pgno pgno,                      /* Page number */
  void *pData,                    /* xFetch()'d data for this page */
  PgHdr **ppPage                  /* OUT: Acquired page object */
){
  PgHdr *p;                       /* Memory mapped page to return */
 
  if( pPager->pMmapFreelist ){
    *ppPage = p = pPager->pMmapFreelist;
    pPager->pMmapFreelist = p->pDirty;
    p->pDirty = 0;
    assert( pPager->nExtra>=8 );
    memset(p->pExtra, 0, 8);
  }else{
    *ppPage = p = (PgHdr *)sqlite3MallocZero(sizeof(PgHdr) + pPager->nExtra);
    if( p==0 ){
      sqlite3OsUnfetch(pPager->fd, (i64)(pgno-1) * pPager->pageSize, pData);
      return SQLITE_NOMEM_BKPT;
    }
    p->pExtra = (void *)&p[1];
    p->flags = PGHDR_MMAP;
    p->nRef = 1;
    p->pPager = pPager;
  }

  assert( p->pExtra==(void *)&p[1] );
  assert( p->pPage==0 );
  assert( p->flags==PGHDR_MMAP );
  assert( p->pPager==pPager );
  assert( p->nRef==1 );

  p->pgno = pgno;
  p->pData = pData;
  pPager->nMmapOut++;

  return SQLITE_OK;
}
#endif

/*
** Release a reference to page pPg. pPg must have been returned by an
** earlier call to pagerAcquireMapPage().
*/
static void pagerReleaseMapPage(PgHdr *pPg){
  Pager *pPager = pPg->pPager;
  pPager->nMmapOut--;
  pPg->pDirty = pPager->pMmapFreelist;
  pPager->pMmapFreelist = pPg;

  assert( pPager->fd->pMethods->iVersion>=3 );
  sqlite3OsUnfetch(pPager->fd, (i64)(pPg->pgno-1)*pPager->pageSize, pPg->pData);
}

/*
** Free all PgHdr objects stored in the Pager.pMmapFreelist list.
*/
static void pagerFreeMapHdrs(Pager *pPager){
  PgHdr *p;
  PgHdr *pNext;
  for(p=pPager->pMmapFreelist; p; p=pNext){
    pNext = p->pDirty;
    sqlite3_free(p);
  }
}

/* Verify that the database file has not be deleted or renamed out from
** under the pager.  Return SQLITE_OK if the database is still where it ought
** to be on disk.  Return non-zero (SQLITE_READONLY_DBMOVED or some other error
** code from sqlite3OsAccess()) if the database has gone missing.
*/
static int databaseIsUnmoved(Pager *pPager){
  int bHasMoved = 0;
  int rc;

  if( pPager->tempFile ) return SQLITE_OK;
  if( pPager->dbSize==0 ) return SQLITE_OK;
  assert( pPager->zFilename && pPager->zFilename[0] );
  rc = sqlite3OsFileControl(pPager->fd, SQLITE_FCNTL_HAS_MOVED, &bHasMoved);
  if( rc==SQLITE_NOTFOUND ){
    /* If the HAS_MOVED file-control is unimplemented, assume that the file
    ** has not been moved.  That is the historical behavior of SQLite: prior to
    ** version 3.8.3, it never checked */
    rc = SQLITE_OK;
  }else if( rc==SQLITE_OK && bHasMoved ){
    rc = SQLITE_READONLY_DBMOVED;
  }
  return rc;
}


/*
** Shutdown the page cache.  Free all memory and close all files.
**
** If a transaction was in progress when this routine is called, that
** transaction is rolled back.  All outstanding pages are invalidated
** and their memory is freed.  Any attempt to use a page associated
** with this page cache after this function returns will likely
** result in a coredump.
**
** This function always succeeds. If a transaction is active an attempt
** is made to roll it back. If an error occurs during the rollback
** a hot journal may be left in the filesystem but no error is returned
** to the caller.
*/
int sqlite3PagerClose(Pager *pPager, sqlite3 *db){
  u8 *pTmp = (u8*)pPager->pTmpSpace;
  assert( db || pagerUseWal(pPager)==0 );
  assert( assert_pager_state(pPager) );
  disable_simulated_io_errors();
  sqlite3BeginBenignMalloc();
  pagerFreeMapHdrs(pPager);
  /* pPager->errCode = 0; */
  pPager->exclusiveMode = 0;
#ifndef SQLITE_OMIT_WAL
  {
    u8 *a = 0;
    assert( db || pPager->pWal==0 );
    if( db && 0==(db->flags & SQLITE_NoCkptOnClose)
     && SQLITE_OK==databaseIsUnmoved(pPager)
    ){
      a = pTmp;
    }
    sqlite3WalClose(pPager->pWal, db, pPager->walSyncFlags, pPager->pageSize,a);
    pPager->pWal = 0;
  }
#endif
  pager_reset(pPager);
  if( MEMDB ){
    pager_unlock(pPager);
  }else{
    /* If it is open, sync the journal file before calling UnlockAndRollback.
    ** If this is not done, then an unsynced portion of the open journal
    ** file may be played back into the database. If a power failure occurs
    ** while this is happening, the database could become corrupt.
    **
    ** If an error occurs while trying to sync the journal, shift the pager
    ** into the ERROR state. This causes UnlockAndRollback to unlock the
    ** database and close the journal file without attempting to roll it
    ** back or finalize it. The next database user will have to do hot-journal
    ** rollback before accessing the database file.
    */
    if( isOpen(pPager->jfd) ){
      pager_error(pPager, pagerSyncHotJournal(pPager));
    }
    pagerUnlockAndRollback(pPager);
  }
  sqlite3EndBenignMalloc();
  enable_simulated_io_errors();
  PAGERTRACE(("CLOSE %d\n", PAGERID(pPager)));
  IOTRACE(("CLOSE %p\n", pPager))
  sqlite3OsClose(pPager->jfd);
  sqlite3OsClose(pPager->fd);
  sqlite3PageFree(pTmp);
  sqlite3PcacheClose(pPager->pPCache);
  assert( !pPager->aSavepoint && !pPager->pInJournal );
  assert( !isOpen(pPager->jfd) && !isOpen(pPager->sjfd) );

  sqlite3_free(pPager);
  return SQLITE_OK;
}

#if !defined(NDEBUG) || defined(SQLITE_TEST)
/*
** Return the page number for page pPg.
*/
Pgno sqlite3PagerPagenumber(DbPage *pPg){
  return pPg->pgno;
}
#endif

/*
** Increment the reference count for page pPg.
*/
void sqlite3PagerRef(DbPage *pPg){
  sqlite3PcacheRef(pPg);
}

/*
** Sync the journal. In other words, make sure all the pages that have
** been written to the journal have actually reached the surface of the
** disk and can be restored in the event of a hot-journal rollback.
**
** If the Pager.noSync flag is set, then this function is a no-op.
** Otherwise, the actions required depend on the journal-mode and the
** device characteristics of the file-system, as follows:
**
**   * If the journal file is an in-memory journal file, no action need
**     be taken.
**
**   * Otherwise, if the device does not support the SAFE_APPEND property,
**     then the nRec field of the most recently written journal header
**     is updated to contain the number of journal records that have
**     been written following it. If the pager is operating in full-sync
**     mode, then the journal file is synced before this field is updated.
**
**   * If the device does not support the SEQUENTIAL property, then
**     journal file is synced.
**
** Or, in pseudo-code:
**
**   if( NOT <in-memory journal> ){
**     if( NOT SAFE_APPEND ){
**       if( <full-sync mode> ) xSync(<journal file>);
**       <update nRec field>
**     }
**     if( NOT SEQUENTIAL ) xSync(<journal file>);
**   }
**
** If successful, this routine clears the PGHDR_NEED_SYNC flag of every
** page currently held in memory before returning SQLITE_OK. If an IO
** error is encountered, then the IO error code is returned to the caller.
*/
static int syncJournal(Pager *pPager, int newHdr){
  int rc;                         /* Return code */

  assert( pPager->eState==PAGER_WRITER_CACHEMOD
       || pPager->eState==PAGER_WRITER_DBMOD
  );
  assert( assert_pager_state(pPager) );
  assert( !pagerUseWal(pPager) );

  rc = sqlite3PagerExclusiveLock(pPager);
  if( rc!=SQLITE_OK ) return rc;

  if( !pPager->noSync ){
    assert( !pPager->tempFile );
    if( isOpen(pPager->jfd) && pPager->journalMode!=PAGER_JOURNALMODE_MEMORY ){
      const int iDc = sqlite3OsDeviceCharacteristics(pPager->fd);
      assert( isOpen(pPager->jfd) );

      if( 0==(iDc&SQLITE_IOCAP_SAFE_APPEND) ){
        /* This block deals with an obscure problem. If the last connection
        ** that wrote to this database was operating in persistent-journal
        ** mode, then the journal file may at this point actually be larger
        ** than Pager.journalOff bytes. If the next thing in the journal
        ** file happens to be a journal-header (written as part of the
        ** previous connection's transaction), and a crash or power-failure
        ** occurs after nRec is updated but before this connection writes
        ** anything else to the journal file (or commits/rolls back its
        ** transaction), then SQLite may become confused when doing the
        ** hot-journal rollback following recovery. It may roll back all
        ** of this connections data, then proceed to rolling back the old,
        ** out-of-date data that follows it. Database corruption.
        **
        ** To work around this, if the journal file does appear to contain
        ** a valid header following Pager.journalOff, then write a 0x00
        ** byte to the start of it to prevent it from being recognized.
        **
        ** Variable iNextHdrOffset is set to the offset at which this
        ** problematic header will occur, if it exists. aMagic is used
        ** as a temporary buffer to inspect the first couple of bytes of
        ** the potential journal header.
        */
        i64 iNextHdrOffset;
        u8 aMagic[8];
        u8 zHeader[sizeof(aJournalMagic)+4];

        memcpy(zHeader, aJournalMagic, sizeof(aJournalMagic));
        put32bits(&zHeader[sizeof(aJournalMagic)], pPager->nRec);

        iNextHdrOffset = journalHdrOffset(pPager);
        rc = sqlite3OsRead(pPager->jfd, aMagic, 8, iNextHdrOffset);
        if( rc==SQLITE_OK && 0==memcmp(aMagic, aJournalMagic, 8) ){
          static const u8 zerobyte = 0;
          rc = sqlite3OsWrite(pPager->jfd, &zerobyte, 1, iNextHdrOffset);
        }
        if( rc!=SQLITE_OK && rc!=SQLITE_IOERR_SHORT_READ ){
          return rc;
        }

        /* Write the nRec value into the journal file header. If in
        ** full-synchronous mode, sync the journal first. This ensures that
        ** all data has really hit the disk before nRec is updated to mark
        ** it as a candidate for rollback.
        **
        ** This is not required if the persistent media supports the
        ** SAFE_APPEND property. Because in this case it is not possible
        ** for garbage data to be appended to the file, the nRec field
        ** is populated with 0xFFFFFFFF when the journal header is written
        ** and never needs to be updated.
        */
        if( pPager->fullSync && 0==(iDc&SQLITE_IOCAP_SEQUENTIAL) ){
          PAGERTRACE(("SYNC journal of %d\n", PAGERID(pPager)));
          IOTRACE(("JSYNC %p\n", pPager))
          rc = sqlite3OsSync(pPager->jfd, pPager->syncFlags);
          if( rc!=SQLITE_OK ) return rc;
        }
        IOTRACE(("JHDR %p %lld\n", pPager, pPager->journalHdr));
        rc = sqlite3OsWrite(
            pPager->jfd, zHeader, sizeof(zHeader), pPager->journalHdr
        );
        if( rc!=SQLITE_OK ) return rc;
      }
      if( 0==(iDc&SQLITE_IOCAP_SEQUENTIAL) ){
        PAGERTRACE(("SYNC journal of %d\n", PAGERID(pPager)));
        IOTRACE(("JSYNC %p\n", pPager))
        rc = sqlite3OsSync(pPager->jfd, pPager->syncFlags|
          (pPager->syncFlags==SQLITE_SYNC_FULL?SQLITE_SYNC_DATAONLY:0)
        );
        if( rc!=SQLITE_OK ) return rc;
      }

      pPager->journalHdr = pPager->journalOff;
      if( newHdr && 0==(iDc&SQLITE_IOCAP_SAFE_APPEND) ){
        pPager->nRec = 0;
        rc = writeJournalHdr(pPager);
        if( rc!=SQLITE_OK ) return rc;
      }
    }else{
      pPager->journalHdr = pPager->journalOff;
    }
  }

  /* Unless the pager is in noSync mode, the journal file was just
  ** successfully synced. Either way, clear the PGHDR_NEED_SYNC flag on
  ** all pages.
  */
  sqlite3PcacheClearSyncFlags(pPager->pPCache);
  pPager->eState = PAGER_WRITER_DBMOD;
  assert( assert_pager_state(pPager) );
  return SQLITE_OK;
}

/*
** The argument is the first in a linked list of dirty pages connected
** by the PgHdr.pDirty pointer. This function writes each one of the
** in-memory pages in the list to the database file. The argument may
** be NULL, representing an empty list. In this case this function is
** a no-op.
**
** The pager must hold at least a RESERVED lock when this function
** is called. Before writing anything to the database file, this lock
** is upgraded to an EXCLUSIVE lock. If the lock cannot be obtained,
** SQLITE_BUSY is returned and no data is written to the database file.
**
** If the pager is a temp-file pager and the actual file-system file
** is not yet open, it is created and opened before any data is
** written out.
**
** Once the lock has been upgraded and, if necessary, the file opened,
** the pages are written out to the database file in list order. Writing
** a page is skipped if it meets either of the following criteria:
**
**   * The page number is greater than Pager.dbSize, or
**   * The PGHDR_DONT_WRITE flag is set on the page.
**
** If writing out a page causes the database file to grow, Pager.dbFileSize
** is updated accordingly. If page 1 is written out, then the value cached
** in Pager.dbFileVers[] is updated to match the new value stored in
** the database file.
**
** If everything is successful, SQLITE_OK is returned. If an IO error
** occurs, an IO error code is returned. Or, if the EXCLUSIVE lock cannot
** be obtained, SQLITE_BUSY is returned.
*/
static int pager_write_pagelist(Pager *pPager, PgHdr *pList){
  int rc = SQLITE_OK;                  /* Return code */

  /* This function is only called for rollback pagers in WRITER_DBMOD state. */
  assert( !pagerUseWal(pPager) );
  assert( pPager->tempFile || pPager->eState==PAGER_WRITER_DBMOD );
  assert( pPager->eLock==EXCLUSIVE_LOCK );
  assert( isOpen(pPager->fd) || pList->pDirty==0 );

  /* If the file is a temp-file has not yet been opened, open it now. It
  ** is not possible for rc to be other than SQLITE_OK if this branch
  ** is taken, as pager_wait_on_lock() is a no-op for temp-files.
  */
  if( !isOpen(pPager->fd) ){
    assert( pPager->tempFile && rc==SQLITE_OK );
    rc = pagerOpentemp(pPager, pPager->fd, pPager->vfsFlags);
  }

  /* Before the first write, give the VFS a hint of what the final
  ** file size will be.
  */
  assert( rc!=SQLITE_OK || isOpen(pPager->fd) );
  if( rc==SQLITE_OK
   && pPager->dbHintSize<pPager->dbSize
   && (pList->pDirty || pList->pgno>pPager->dbHintSize)
  ){
    sqlite3_int64 szFile = pPager->pageSize * (sqlite3_int64)pPager->dbSize;
    sqlite3OsFileControlHint(pPager->fd, SQLITE_FCNTL_SIZE_HINT, &szFile);
    pPager->dbHintSize = pPager->dbSize;
  }

  while( rc==SQLITE_OK && pList ){
    Pgno pgno = pList->pgno;

    /* If there are dirty pages in the page cache with page numbers greater
    ** than Pager.dbSize, this means sqlite3PagerTruncateImage() was called to
    ** make the file smaller (presumably by auto-vacuum code). Do not write
    ** any such pages to the file.
    **
    ** Also, do not write out any page that has the PGHDR_DONT_WRITE flag
    ** set (set by sqlite3PagerDontWrite()).
    */
    if( pgno<=pPager->dbSize && 0==(pList->flags&PGHDR_DONT_WRITE) ){
      i64 offset = (pgno-1)*(i64)pPager->pageSize;   /* Offset to write */
      char *pData;                                   /* Data to write */   

      assert( (pList->flags&PGHDR_NEED_SYNC)==0 );
      if( pList->pgno==1 ) pager_write_changecounter(pList);

      pData = pList->pData;

      /* Write out the page data. */
      rc = sqlite3OsWrite(pPager->fd, pData, pPager->pageSize, offset);

      /* If page 1 was just written, update Pager.dbFileVers to match
      ** the value now stored in the database file. If writing this
      ** page caused the database file to grow, update dbFileSize.
      */
      if( pgno==1 ){
        memcpy(&pPager->dbFileVers, &pData[24], sizeof(pPager->dbFileVers));
      }
      if( pgno>pPager->dbFileSize ){
        pPager->dbFileSize = pgno;
      }
      pPager->aStat[PAGER_STAT_WRITE]++;

      /* Update any backup objects copying the contents of this pager. */
      sqlite3BackupUpdate(pPager->pBackup, pgno, (u8*)pList->pData);

      PAGERTRACE(("STORE %d page %d hash(%08x)\n",
                   PAGERID(pPager), pgno, pager_pagehash(pList)));
      IOTRACE(("PGOUT %p %d\n", pPager, pgno));
      PAGER_INCR(sqlite3_pager_writedb_count);
    }else{
      PAGERTRACE(("NOSTORE %d page %d\n", PAGERID(pPager), pgno));
    }
    pager_set_pagehash(pList);
    pList = pList->pDirty;
  }

  return rc;
}

/*
** Ensure that the sub-journal file is open. If it is already open, this
** function is a no-op.
**
** SQLITE_OK is returned if everything goes according to plan. An
** SQLITE_IOERR_XXX error code is returned if a call to sqlite3OsOpen()
** fails.
*/
static int openSubJournal(Pager *pPager){
  int rc = SQLITE_OK;
  if( !isOpen(pPager->sjfd) ){
    const int flags =  SQLITE_OPEN_SUBJOURNAL | SQLITE_OPEN_READWRITE
      | SQLITE_OPEN_CREATE | SQLITE_OPEN_EXCLUSIVE
      | SQLITE_OPEN_DELETEONCLOSE;
    int nStmtSpill = sqlite3Config.nStmtSpill;
    if( pPager->journalMode==PAGER_JOURNALMODE_MEMORY || pPager->subjInMemory ){
      nStmtSpill = -1;
    }
    rc = sqlite3JournalOpen(pPager->pVfs, 0, pPager->sjfd, flags, nStmtSpill);
  }
  return rc;
}

/*
** Append a record of the current state of page pPg to the sub-journal.
**
** If successful, set the bit corresponding to pPg->pgno in the bitvecs
** for all open savepoints before returning.
**
** This function returns SQLITE_OK if everything is successful, an IO
** error code if the attempt to write to the sub-journal fails, or
** SQLITE_NOMEM if a malloc fails while setting a bit in a savepoint
** bitvec.
*/
static int subjournalPage(PgHdr *pPg){
  int rc = SQLITE_OK;
  Pager *pPager = pPg->pPager;
  if( pPager->journalMode!=PAGER_JOURNALMODE_OFF ){

    /* Open the sub-journal, if it has not already been opened */
    assert( pPager->useJournal );
    assert( isOpen(pPager->jfd) || pagerUseWal(pPager) );
    assert( isOpen(pPager->sjfd) || pPager->nSubRec==0 );
    assert( pagerUseWal(pPager)
         || pageInJournal(pPager, pPg)
         || pPg->pgno>pPager->dbOrigSize
    );
    rc = openSubJournal(pPager);

    /* If the sub-journal was opened successfully (or was already open),
    ** write the journal record into the file.  */
    if( rc==SQLITE_OK ){
      void *pData = pPg->pData;
      i64 offset = (i64)pPager->nSubRec*(4+pPager->pageSize);
      char *pData2;
      pData2 = pData;
      PAGERTRACE(("STMT-JOURNAL %d page %d\n", PAGERID(pPager), pPg->pgno));
      rc = write32bits(pPager->sjfd, offset, pPg->pgno);
      if( rc==SQLITE_OK ){
        rc = sqlite3OsWrite(pPager->sjfd, pData2, pPager->pageSize, offset+4);
      }
    }
  }
  if( rc==SQLITE_OK ){
    pPager->nSubRec++;
    assert( pPager->nSavepoint>0 );
    rc = addToSavepointBitvecs(pPager, pPg->pgno);
  }
  return rc;
}
static int subjournalPageIfRequired(PgHdr *pPg){
  if( subjRequiresPage(pPg) ){
    return subjournalPage(pPg);
  }else{
    return SQLITE_OK;
  }
}

/*
** This function is called by the pcache layer when it has reached some
** soft memory limit. The first argument is a pointer to a Pager object
** (cast as a void*). The pager is always 'purgeable' (not an in-memory
** database). The second argument is a reference to a page that is
** currently dirty but has no outstanding references. The page
** is always associated with the Pager object passed as the first
** argument.
**
** The job of this function is to make pPg clean by writing its contents
** out to the database file, if possible. This may involve syncing the
** journal file.
**
** If successful, sqlite3PcacheMakeClean() is called on the page and
** SQLITE_OK returned. If an IO error occurs while trying to make the
** page clean, the IO error code is returned. If the page cannot be
** made clean for some other reason, but no error occurs, then SQLITE_OK
** is returned by sqlite3PcacheMakeClean() is not called.
*/
static int pagerStress(void *p, PgHdr *pPg){
  Pager *pPager = (Pager *)p;
  int rc = SQLITE_OK;

  assert( pPg->pPager==pPager );
  assert( pPg->flags&PGHDR_DIRTY );

  /* The doNotSpill NOSYNC bit is set during times when doing a sync of
  ** journal (and adding a new header) is not allowed.  This occurs
  ** during calls to sqlite3PagerWrite() while trying to journal multiple
  ** pages belonging to the same sector.
  **
  ** The doNotSpill ROLLBACK and OFF bits inhibits all cache spilling
  ** regardless of whether or not a sync is required.  This is set during
  ** a rollback or by user request, respectively.
  **
  ** Spilling is also prohibited when in an error state since that could
  ** lead to database corruption.   In the current implementation it
  ** is impossible for sqlite3PcacheFetch() to be called with createFlag==3
  ** while in the error state, hence it is impossible for this routine to
  ** be called in the error state.  Nevertheless, we include a NEVER()
  ** test for the error state as a safeguard against future changes.
  */
  if( NEVER(pPager->errCode) ) return SQLITE_OK;
  testcase( pPager->doNotSpill & SPILLFLAG_ROLLBACK );
  testcase( pPager->doNotSpill & SPILLFLAG_OFF );
  testcase( pPager->doNotSpill & SPILLFLAG_NOSYNC );
  if( pPager->doNotSpill
   && ((pPager->doNotSpill & (SPILLFLAG_ROLLBACK|SPILLFLAG_OFF))!=0
      || (pPg->flags & PGHDR_NEED_SYNC)!=0)
  ){
    return SQLITE_OK;
  }

  pPager->aStat[PAGER_STAT_SPILL]++;
  pPg->pDirty = 0;
  if( pagerUseWal(pPager) ){
    /* Write a single frame for this page to the log. */
    rc = subjournalPageIfRequired(pPg);
    if( rc==SQLITE_OK ){
      rc = pagerWalFrames(pPager, pPg, 0, 0);
    }
  }else{
   
#ifdef SQLITE_ENABLE_BATCH_ATOMIC_WRITE
    if( pPager->tempFile==0 ){
      rc = sqlite3JournalCreate(pPager->jfd);
      if( rc!=SQLITE_OK ) return pager_error(pPager, rc);
    }
#endif
 
    /* Sync the journal file if required. */
    if( pPg->flags&PGHDR_NEED_SYNC
     || pPager->eState==PAGER_WRITER_CACHEMOD
    ){
      rc = syncJournal(pPager, 1);
    }
 
    /* Write the contents of the page out to the database file. */
    if( rc==SQLITE_OK ){
      assert( (pPg->flags&PGHDR_NEED_SYNC)==0 );
      rc = pager_write_pagelist(pPager, pPg);
    }
  }

  /* Mark the page as clean. */
  if( rc==SQLITE_OK ){
    PAGERTRACE(("STRESS %d page %d\n", PAGERID(pPager), pPg->pgno));
    sqlite3PcacheMakeClean(pPg);
  }

  return pager_error(pPager, rc);
}

/*
** Flush all unreferenced dirty pages to disk.
*/
int sqlite3PagerFlush(Pager *pPager){
  int rc = pPager->errCode;
  if( !MEMDB ){
    PgHdr *pList = sqlite3PcacheDirtyList(pPager->pPCache);
    assert( assert_pager_state(pPager) );
    while( rc==SQLITE_OK && pList ){
      PgHdr *pNext = pList->pDirty;
      if( pList->nRef==0 ){
        rc = pagerStress((void*)pPager, pList);
      }
      pList = pNext;
    }
  }

  return rc;
}

/*
** Allocate and initialize a new Pager object and put a pointer to it
** in *ppPager. The pager should eventually be freed by passing it
** to sqlite3PagerClose().
**
** The zFilename argument is the path to the database file to open.
** If zFilename is NULL then a randomly-named temporary file is created
** and used as the file to be cached. Temporary files are be deleted
** automatically when they are closed. If zFilename is ":memory:" then
** all information is held in cache. It is never written to disk.
** This can be used to implement an in-memory database.
**
** The nExtra parameter specifies the number of bytes of space allocated
** along with each page reference. This space is available to the user
** via the sqlite3PagerGetExtra() API.  When a new page is allocated, the
** first 8 bytes of this space are zeroed but the remainder is uninitialized.
** (The extra space is used by btree as the MemPage object.)
**
** The flags argument is used to specify properties that affect the
** operation of the pager. It should be passed some bitwise combination
** of the PAGER_* flags.
**
** The vfsFlags parameter is a bitmask to pass to the flags parameter
** of the xOpen() method of the supplied VFS when opening files.
**
** If the pager object is allocated and the specified file opened
** successfully, SQLITE_OK is returned and *ppPager set to point to
** the new pager object. If an error occurs, *ppPager is set to NULL
** and error code returned. This function may return SQLITE_NOMEM
** (sqlite3Malloc() is used to allocate memory), SQLITE_CANTOPEN or
** various SQLITE_IO_XXX errors.
*/
int sqlite3PagerOpen(
  sqlite3_vfs *pVfs,       /* The virtual file system to use */
  Pager **ppPager,         /* OUT: Return the Pager structure here */
  const char *zFilename,   /* Name of the database file to open */
  int nExtra,              /* Extra bytes append to each in-memory page */
  int flags,               /* flags controlling this file */
  int vfsFlags,            /* flags passed through to sqlite3_vfs.xOpen() */
  void (*xReinit)(DbPage*) /* Function to reinitialize pages */
){
  u8 *pPtr;
  Pager *pPager = 0;       /* Pager object to allocate and return */
  int rc = SQLITE_OK;      /* Return code */
  int tempFile = 0;        /* True for temp files (incl. in-memory files) */
  int memDb = 0;           /* True if this is an in-memory file */
  int memJM = 0;           /* Memory journal mode */
  int readOnly = 0;        /* True if this is a read-only file */
  int journalFileSize;     /* Bytes to allocate for each journal fd */
  char *zPathname = 0;     /* Full path to database file */
  int nPathname = 0;       /* Number of bytes in zPathname */
  int useJournal = (flags & PAGER_OMIT_JOURNAL)==0; /* False to omit journal */
  int pcacheSize = sqlite3PcacheSize();       /* Bytes to allocate for PCache */
  u32 szPageDflt = SQLITE_DEFAULT_PAGE_SIZE;  /* Default page size */
  const char *zUri = 0;    /* URI args to copy */
  int nUriByte = 1;        /* Number of bytes of URI args at *zUri */

  /* Figure out how much space is required for each journal file-handle
  ** (there are two of them, the main journal and the sub-journal).  */
  journalFileSize = ROUND8(sqlite3JournalSize(pVfs));

  /* Set the output variable to NULL in case an error occurs. */
  *ppPager = 0;

#ifndef SQLITE_OMIT_MEMORYDB
  if( flags & PAGER_MEMORY ){
    memDb = 1;
    if( zFilename && zFilename[0] ){
      zPathname = sqlite3DbStrDup(0, zFilename);
      if( zPathname==0  ) return SQLITE_NOMEM_BKPT;
      nPathname = sqlite3Strlen30(zPathname);
      zFilename = 0;
    }
  }
#endif

  /* Compute and store the full pathname in an allocated buffer pointed
  ** to by zPathname, length nPathname. Or, if this is a temporary file,
  ** leave both nPathname and zPathname set to 0.
  */
  if( zFilename && zFilename[0] ){
    const char *z;
    nPathname = pVfs->mxPathname+1;
    zPathname = sqlite3DbMallocRaw(0, nPathname*2);
    if( zPathname==0 ){
      return SQLITE_NOMEM_BKPT;
    }
    zPathname[0] = 0; /* Make sure initialized even if FullPathname() fails */
    rc = sqlite3OsFullPathname(pVfs, zFilename, nPathname, zPathname);
    if( rc!=SQLITE_OK ){
      if( rc==SQLITE_OK_SYMLINK ){
        if( vfsFlags & SQLITE_OPEN_NOFOLLOW ){
          rc = SQLITE_CANTOPEN_SYMLINK;
        }else{
          rc = SQLITE_OK;
        }
      }
    }
    nPathname = sqlite3Strlen30(zPathname);
    z = zUri = &zFilename[sqlite3Strlen30(zFilename)+1];
    while( *z ){
      z += strlen(z)+1;
      z += strlen(z)+1;
    }
    nUriByte = (int)(&z[1] - zUri);
    assert( nUriByte>=1 );
    if( rc==SQLITE_OK && nPathname+8>pVfs->mxPathname ){
      /* This branch is taken when the journal path required by
      ** the database being opened will be more than pVfs->mxPathname
      ** bytes in length. This means the database cannot be opened,
      ** as it will not be possible to open the journal file or even
      ** check for a hot-journal before reading.
      */
      rc = SQLITE_CANTOPEN_BKPT;
    }
    if( rc!=SQLITE_OK ){
      sqlite3DbFree(0, zPathname);
      return rc;
    }
  }

  /* Allocate memory for the Pager structure, PCache object, the
  ** three file descriptors, the database file name and the journal
  ** file name. The layout in memory is as follows:
  **
  **     Pager object                    (sizeof(Pager) bytes)
  **     PCache object                   (sqlite3PcacheSize() bytes)
  **     Database file handle            (pVfs->szOsFile bytes)
  **     Sub-journal file handle         (journalFileSize bytes)
  **     Main journal file handle        (journalFileSize bytes)
  **     Ptr back to the Pager           (sizeof(Pager*) bytes)
  **     \0\0\0\0 database prefix        (4 bytes)
  **     Database file name              (nPathname+1 bytes)
  **     URI query parameters            (nUriByte bytes)
  **     Journal filename                (nPathname+8+1 bytes)
  **     WAL filename                    (nPathname+4+1 bytes)
  **     \0\0\0 terminator               (3 bytes)
  **
  ** Some 3rd-party software, over which we have no control, depends on
  ** the specific order of the filenames and the \0 separators between them
  ** so that it can (for example) find the database filename given the WAL
  ** filename without using the sqlite3_filename_database() API.  This is a
  ** misuse of SQLite and a bug in the 3rd-party software, but the 3rd-party
  ** software is in widespread use, so we try to avoid changing the filename
  ** order and formatting if possible.  In particular, the details of the
  ** filename format expected by 3rd-party software should be as follows:
  **
  **   - Main Database Path
  **   - \0
  **   - Multiple URI components consisting of:
  **     - Key
  **     - \0
  **     - Value
  **     - \0
  **   - \0
  **   - Journal Path
  **   - \0
  **   - WAL Path (zWALName)
  **   - \0
  **
  ** The sqlite3_create_filename() interface and the databaseFilename() utility
  ** that is used by sqlite3_filename_database() and kin also depend on the
  ** specific formatting and order of the various filenames, so if the format
  ** changes here, be sure to change it there as well.
  */
  assert( SQLITE_PTRSIZE==sizeof(Pager*) );
  pPtr = (u8 *)sqlite3MallocZero(
    ROUND8(sizeof(*pPager)) +            /* Pager structure */
    ROUND8(pcacheSize) +                 /* PCache object */
    ROUND8(pVfs->szOsFile) +             /* The main db file */
    journalFileSize * 2 +                /* The two journal files */
    SQLITE_PTRSIZE +                     /* Space to hold a pointer */
    4 +                                  /* Database prefix */
    nPathname + 1 +                      /* database filename */
    nUriByte +                           /* query parameters */
    nPathname + 8 + 1 +                  /* Journal filename */
#ifndef SQLITE_OMIT_WAL
    nPathname + 4 + 1 +                  /* WAL filename */
#endif
    3                                    /* Terminator */
  );
  assert( EIGHT_BYTE_ALIGNMENT(SQLITE_INT_TO_PTR(journalFileSize)) );
  if( !pPtr ){
    sqlite3DbFree(0, zPathname);
    return SQLITE_NOMEM_BKPT;
  }
  pPager = (Pager*)pPtr;                  pPtr += ROUND8(sizeof(*pPager));
  pPager->pPCache = (PCache*)pPtr;        pPtr += ROUND8(pcacheSize);
  pPager->fd = (sqlite3_file*)pPtr;       pPtr += ROUND8(pVfs->szOsFile);
  pPager->sjfd = (sqlite3_file*)pPtr;     pPtr += journalFileSize;
  pPager->jfd =  (sqlite3_file*)pPtr;     pPtr += journalFileSize;
  assert( EIGHT_BYTE_ALIGNMENT(pPager->jfd) );
  memcpy(pPtr, &pPager, SQLITE_PTRSIZE);  pPtr += SQLITE_PTRSIZE;

  /* Fill in the Pager.zFilename and pPager.zQueryParam fields */
                                          pPtr += 4;  /* Skip zero prefix */
  pPager->zFilename = (char*)pPtr;
  if( nPathname>0 ){
    memcpy(pPtr, zPathname, nPathname);   pPtr += nPathname + 1;
    if( zUri ){
      memcpy(pPtr, zUri, nUriByte);       pPtr += nUriByte;
    }else{
                                          pPtr++;
    }
  }


  /* Fill in Pager.zJournal */
  if( nPathname>0 ){
    pPager->zJournal = (char*)pPtr;
    memcpy(pPtr, zPathname, nPathname);   pPtr += nPathname;
    memcpy(pPtr, "-journal",8);           pPtr += 8 + 1;
#ifdef SQLITE_ENABLE_8_3_NAMES
    sqlite3FileSuffix3(zFilename,pPager->zJournal);
    pPtr = (u8*)(pPager->zJournal + sqlite3Strlen30(pPager->zJournal)+1);
#endif
  }else{
    pPager->zJournal = 0;
  }

#ifndef SQLITE_OMIT_WAL
  /* Fill in Pager.zWal */
  if( nPathname>0 ){
    pPager->zWal = (char*)pPtr;
    memcpy(pPtr, zPathname, nPathname);   pPtr += nPathname;
    memcpy(pPtr, "-wal", 4);              pPtr += 4 + 1;
#ifdef SQLITE_ENABLE_8_3_NAMES
    sqlite3FileSuffix3(zFilename, pPager->zWal);
    pPtr = (u8*)(pPager->zWal + sqlite3Strlen30(pPager->zWal)+1);
#endif
  }else{
    pPager->zWal = 0;
  }
#endif
  (void)pPtr;  /* Suppress warning about unused pPtr value */

  if( nPathname ) sqlite3DbFree(0, zPathname);
  pPager->pVfs = pVfs;
  pPager->vfsFlags = vfsFlags;

  /* Open the pager file.
  */
  if( zFilename && zFilename[0] ){
    int fout = 0;                    /* VFS flags returned by xOpen() */
    rc = sqlite3OsOpen(pVfs, pPager->zFilename, pPager->fd, vfsFlags, &fout);
    assert( !memDb );
    pPager->memVfs = memJM = (fout&SQLITE_OPEN_MEMORY)!=0;
    readOnly = (fout&SQLITE_OPEN_READONLY)!=0;

    /* If the file was successfully opened for read/write access,
    ** choose a default page size in case we have to create the
    ** database file. The default page size is the maximum of:
    **
    **    + SQLITE_DEFAULT_PAGE_SIZE,
    **    + The value returned by sqlite3OsSectorSize()
    **    + The largest page size that can be written atomically.
    */
    if( rc==SQLITE_OK ){
      int iDc = sqlite3OsDeviceCharacteristics(pPager->fd);
      if( !readOnly ){
        setSectorSize(pPager);
        assert(SQLITE_DEFAULT_PAGE_SIZE<=SQLITE_MAX_DEFAULT_PAGE_SIZE);
        if( szPageDflt<pPager->sectorSize ){
          if( pPager->sectorSize>SQLITE_MAX_DEFAULT_PAGE_SIZE ){
            szPageDflt = SQLITE_MAX_DEFAULT_PAGE_SIZE;
          }else{
            szPageDflt = (u32)pPager->sectorSize;
          }
        }
#ifdef SQLITE_ENABLE_ATOMIC_WRITE
        {
          int ii;
          assert(SQLITE_IOCAP_ATOMIC512==(512>>8));
          assert(SQLITE_IOCAP_ATOMIC64K==(65536>>8));
          assert(SQLITE_MAX_DEFAULT_PAGE_SIZE<=65536);
          for(ii=szPageDflt; ii<=SQLITE_MAX_DEFAULT_PAGE_SIZE; ii=ii*2){
            if( iDc&(SQLITE_IOCAP_ATOMIC|(ii>>8)) ){
              szPageDflt = ii;
            }
          }
        }
#endif
      }
      pPager->noLock = sqlite3_uri_boolean(pPager->zFilename, "nolock", 0);
      if( (iDc & SQLITE_IOCAP_IMMUTABLE)!=0
       || sqlite3_uri_boolean(pPager->zFilename, "immutable", 0) ){
          vfsFlags |= SQLITE_OPEN_READONLY;
          goto act_like_temp_file;
      }
    }
  }else{
    /* If a temporary file is requested, it is not opened immediately.
    ** In this case we accept the default page size and delay actually
    ** opening the file until the first call to OsWrite().
    **
    ** This branch is also run for an in-memory database. An in-memory
    ** database is the same as a temp-file that is never written out to
    ** disk and uses an in-memory rollback journal.
    **
    ** This branch also runs for files marked as immutable.
    */
act_like_temp_file:
    tempFile = 1;
    pPager->eState = PAGER_READER;     /* Pretend we already have a lock */
    pPager->eLock = EXCLUSIVE_LOCK;    /* Pretend we are in EXCLUSIVE mode */
    pPager->noLock = 1;                /* Do no locking */
    readOnly = (vfsFlags&SQLITE_OPEN_READONLY);
  }

  /* The following call to PagerSetPagesize() serves to set the value of
  ** Pager.pageSize and to allocate the Pager.pTmpSpace buffer.
  */
  if( rc==SQLITE_OK ){
    assert( pPager->memDb==0 );
    rc = sqlite3PagerSetPagesize(pPager, &szPageDflt, -1);
    testcase( rc!=SQLITE_OK );
  }

  /* Initialize the PCache object. */
  if( rc==SQLITE_OK ){
    nExtra = ROUND8(nExtra);
    assert( nExtra>=8 && nExtra<1000 );
    rc = sqlite3PcacheOpen(szPageDflt, nExtra, !memDb,
                       !memDb?pagerStress:0, (void *)pPager, pPager->pPCache);
  }

  /* If an error occurred above, free the  Pager structure and close the file.
  */
  if( rc!=SQLITE_OK ){
    sqlite3OsClose(pPager->fd);
    sqlite3PageFree(pPager->pTmpSpace);
    sqlite3_free(pPager);
    return rc;
  }

  PAGERTRACE(("OPEN %d %s\n", FILEHANDLEID(pPager->fd), pPager->zFilename));
  IOTRACE(("OPEN %p %s\n", pPager, pPager->zFilename))

  pPager->useJournal = (u8)useJournal;
  /* pPager->stmtOpen = 0; */
  /* pPager->stmtInUse = 0; */
  /* pPager->nRef = 0; */
  /* pPager->stmtSize = 0; */
  /* pPager->stmtJSize = 0; */
  /* pPager->nPage = 0; */
  pPager->mxPgno = SQLITE_MAX_PAGE_COUNT;
  /* pPager->state = PAGER_UNLOCK; */
  /* pPager->errMask = 0; */
  pPager->tempFile = (u8)tempFile;
  assert( tempFile==PAGER_LOCKINGMODE_NORMAL
          || tempFile==PAGER_LOCKINGMODE_EXCLUSIVE );
  assert( PAGER_LOCKINGMODE_EXCLUSIVE==1 );
  pPager->exclusiveMode = (u8)tempFile;
  pPager->changeCountDone = pPager->tempFile;
  pPager->memDb = (u8)memDb;
  pPager->readOnly = (u8)readOnly;
  assert( useJournal || pPager->tempFile );
  sqlite3PagerSetFlags(pPager, (SQLITE_DEFAULT_SYNCHRONOUS+1)|PAGER_CACHESPILL);
  /* pPager->pFirst = 0; */
  /* pPager->pFirstSynced = 0; */
  /* pPager->pLast = 0; */
  pPager->nExtra = (u16)nExtra;
  pPager->journalSizeLimit = SQLITE_DEFAULT_JOURNAL_SIZE_LIMIT;
  assert( isOpen(pPager->fd) || tempFile );
  setSectorSize(pPager);
  if( !useJournal ){
    pPager->journalMode = PAGER_JOURNALMODE_OFF;
  }else if( memDb || memJM ){
    pPager->journalMode = PAGER_JOURNALMODE_MEMORY;
  }
  /* pPager->xBusyHandler = 0; */
  /* pPager->pBusyHandlerArg = 0; */
  pPager->xReiniter = xReinit;
  setGetterMethod(pPager);
  /* memset(pPager->aHash, 0, sizeof(pPager->aHash)); */
  /* pPager->szMmap = SQLITE_DEFAULT_MMAP_SIZE // will be set by btree.c */

  *ppPager = pPager;
  return SQLITE_OK;
}

/*
** Return the sqlite3_file for the main database given the name
** of the corresponding WAL or Journal name as passed into
** xOpen.
*/
sqlite3_file *sqlite3_database_file_object(const char *zName){
  Pager *pPager;
  const char *p;
  while( zName[-1]!=0 || zName[-2]!=0 || zName[-3]!=0 || zName[-4]!=0 ){
    zName--;
  }
  p = zName - 4 - sizeof(Pager*);
  assert( EIGHT_BYTE_ALIGNMENT(p) );
  pPager = *(Pager**)p;
  return pPager->fd;
}


/*
** This function is called after transitioning from PAGER_UNLOCK to
** PAGER_SHARED state. It tests if there is a hot journal present in
** the file-system for the given pager. A hot journal is one that
** needs to be played back. According to this function, a hot-journal
** file exists if the following criteria are met:
**
**   * The journal file exists in the file system, and
**   * No process holds a RESERVED or greater lock on the database file, and
**   * The database file itself is greater than 0 bytes in size, and
**   * The first byte of the journal file exists and is not 0x00.
**
** If the current size of the database file is 0 but a journal file
** exists, that is probably an old journal left over from a prior
** database with the same name. In this case the journal file is
** just deleted using OsDelete, *pExists is set to 0 and SQLITE_OK
** is returned.
**
** This routine does not check if there is a super-journal filename
** at the end of the file. If there is, and that super-journal file
** does not exist, then the journal file is not really hot. In this
** case this routine will return a false-positive. The pager_playback()
** routine will discover that the journal file is not really hot and
** will not roll it back.
**
** If a hot-journal file is found to exist, *pExists is set to 1 and
** SQLITE_OK returned. If no hot-journal file is present, *pExists is
** set to 0 and SQLITE_OK returned. If an IO error occurs while trying
** to determine whether or not a hot-journal file exists, the IO error
** code is returned and the value of *pExists is undefined.
*/
static int hasHotJournal(Pager *pPager, int *pExists){
  sqlite3_vfs * const pVfs = pPager->pVfs;
  int rc = SQLITE_OK;           /* Return code */
  int exists = 1;               /* True if a journal file is present */
  int jrnlOpen = !!isOpen(pPager->jfd);

  assert( pPager->useJournal );
  assert( isOpen(pPager->fd) );
  assert( pPager->eState==PAGER_OPEN );

  assert( jrnlOpen==0 || ( sqlite3OsDeviceCharacteristics(pPager->jfd) &
    SQLITE_IOCAP_UNDELETABLE_WHEN_OPEN
  ));

  *pExists = 0;
  if( !jrnlOpen ){
    rc = sqlite3OsAccess(pVfs, pPager->zJournal, SQLITE_ACCESS_EXISTS, &exists);
  }
  if( rc==SQLITE_OK && exists ){
    int locked = 0;             /* True if some process holds a RESERVED lock */

    /* Race condition here:  Another process might have been holding the
    ** the RESERVED lock and have a journal open at the sqlite3OsAccess()
    ** call above, but then delete the journal and drop the lock before
    ** we get to the following sqlite3OsCheckReservedLock() call.  If that
    ** is the case, this routine might think there is a hot journal when
    ** in fact there is none.  This results in a false-positive which will
    ** be dealt with by the playback routine.  Ticket #3883.
    */
    rc = sqlite3OsCheckReservedLock(pPager->fd, &locked);
    if( rc==SQLITE_OK && !locked ){
      Pgno nPage;                 /* Number of pages in database file */

      assert( pPager->tempFile==0 );
      rc = pagerPagecount(pPager, &nPage);
      if( rc==SQLITE_OK ){
        /* If the database is zero pages in size, that means that either (1) the
        ** journal is a remnant from a prior database with the same name where
        ** the database file but not the journal was deleted, or (2) the initial
        ** transaction that populates a new database is being rolled back.
        ** In either case, the journal file can be deleted.  However, take care
        ** not to delete the journal file if it is already open due to
        ** journal_mode=PERSIST.
        */
        if( nPage==0 && !jrnlOpen ){
          sqlite3BeginBenignMalloc();
          if( pagerLockDb(pPager, RESERVED_LOCK)==SQLITE_OK ){
            sqlite3OsDelete(pVfs, pPager->zJournal, 0);
            if( !pPager->exclusiveMode ) pagerUnlockDb(pPager, SHARED_LOCK);
          }
          sqlite3EndBenignMalloc();
        }else{
          /* The journal file exists and no other connection has a reserved
          ** or greater lock on the database file. Now check that there is
          ** at least one non-zero bytes at the start of the journal file.
          ** If there is, then we consider this journal to be hot. If not,
          ** it can be ignored.
          */
          if( !jrnlOpen ){
            int f = SQLITE_OPEN_READONLY|SQLITE_OPEN_MAIN_JOURNAL;
            rc = sqlite3OsOpen(pVfs, pPager->zJournal, pPager->jfd, f, &f);
          }
          if( rc==SQLITE_OK ){
            u8 first = 0;
            rc = sqlite3OsRead(pPager->jfd, (void *)&first, 1, 0);
            if( rc==SQLITE_IOERR_SHORT_READ ){
              rc = SQLITE_OK;
            }
            if( !jrnlOpen ){
              sqlite3OsClose(pPager->jfd);
            }
            *pExists = (first!=0);
          }else if( rc==SQLITE_CANTOPEN ){
            /* If we cannot open the rollback journal file in order to see if
            ** it has a zero header, that might be due to an I/O error, or
            ** it might be due to the race condition described above and in
            ** ticket #3883.  Either way, assume that the journal is hot.
            ** This might be a false positive.  But if it is, then the
            ** automatic journal playback and recovery mechanism will deal
            ** with it under an EXCLUSIVE lock where we do not need to
            ** worry so much with race conditions.
            */
            *pExists = 1;
            rc = SQLITE_OK;
          }
        }
      }
    }
  }

  return rc;
}

/*
** This function is called to obtain a shared lock on the database file.
** It is illegal to call sqlite3PagerGet() until after this function
** has been successfully called. If a shared-lock is already held when
** this function is called, it is a no-op.
**
** The following operations are also performed by this function.
**
**   1) If the pager is currently in PAGER_OPEN state (no lock held
**      on the database file), then an attempt is made to obtain a
**      SHARED lock on the database file. Immediately after obtaining
**      the SHARED lock, the file-system is checked for a hot-journal,
**      which is played back if present. Following any hot-journal
**      rollback, the contents of the cache are validated by checking
**      the 'change-counter' field of the database file header and
**      discarded if they are found to be invalid.
**
**   2) If the pager is running in exclusive-mode, and there are currently
**      no outstanding references to any pages, and is in the error state,
**      then an attempt is made to clear the error state by discarding
**      the contents of the page cache and rolling back any open journal
**      file.
**
** If everything is successful, SQLITE_OK is returned. If an IO error
** occurs while locking the database, checking for a hot-journal file or
** rolling back a journal file, the IO error code is returned.
*/
int sqlite3PagerSharedLock(Pager *pPager){
  int rc = SQLITE_OK;                /* Return code */

  /* This routine is only called from b-tree and only when there are no
  ** outstanding pages. This implies that the pager state should either
  ** be OPEN or READER. READER is only possible if the pager is or was in
  ** exclusive access mode.  */
  assert( sqlite3PcacheRefCount(pPager->pPCache)==0 );
  assert( assert_pager_state(pPager) );
  assert( pPager->eState==PAGER_OPEN || pPager->eState==PAGER_READER );
  assert( pPager->errCode==SQLITE_OK );

  if( !pagerUseWal(pPager) && pPager->eState==PAGER_OPEN ){
    int bHotJournal = 1;          /* True if there exists a hot journal-file */

    assert( !MEMDB );
    assert( pPager->tempFile==0 || pPager->eLock==EXCLUSIVE_LOCK );

    rc = pager_wait_on_lock(pPager, SHARED_LOCK);
    if( rc!=SQLITE_OK ){
      assert( pPager->eLock==NO_LOCK || pPager->eLock==UNKNOWN_LOCK );
      goto failed;
    }

    /* If a journal file exists, and there is no RESERVED lock on the
    ** database file, then it either needs to be played back or deleted.
    */
    if( pPager->eLock<=SHARED_LOCK ){
      rc = hasHotJournal(pPager, &bHotJournal);
    }
    if( rc!=SQLITE_OK ){
      goto failed;
    }
    if( bHotJournal ){
      if( pPager->readOnly ){
        rc = SQLITE_READONLY_ROLLBACK;
        goto failed;
      }

      /* Get an EXCLUSIVE lock on the database file. At this point it is
      ** important that a RESERVED lock is not obtained on the way to the
      ** EXCLUSIVE lock. If it were, another process might open the
      ** database file, detect the RESERVED lock, and conclude that the
      ** database is safe to read while this process is still rolling the
      ** hot-journal back.
      **
      ** Because the intermediate RESERVED lock is not requested, any
      ** other process attempting to access the database file will get to
      ** this point in the code and fail to obtain its own EXCLUSIVE lock
      ** on the database file.
      **
      ** Unless the pager is in locking_mode=exclusive mode, the lock is
      ** downgraded to SHARED_LOCK before this function returns.
      */
      rc = pagerLockDb(pPager, EXCLUSIVE_LOCK);
      if( rc!=SQLITE_OK ){
        goto failed;
      }

      /* If it is not already open and the file exists on disk, open the
      ** journal for read/write access. Write access is required because
      ** in exclusive-access mode the file descriptor will be kept open
      ** and possibly used for a transaction later on. Also, write-access
      ** is usually required to finalize the journal in journal_mode=persist
      ** mode (and also for journal_mode=truncate on some systems).
      **
      ** If the journal does not exist, it usually means that some
      ** other connection managed to get in and roll it back before
      ** this connection obtained the exclusive lock above. Or, it
      ** may mean that the pager was in the error-state when this
      ** function was called and the journal file does not exist.
      */
      if( !isOpen(pPager->jfd) && pPager->journalMode!=PAGER_JOURNALMODE_OFF ){
        sqlite3_vfs * const pVfs = pPager->pVfs;
        int bExists;              /* True if journal file exists */
        rc = sqlite3OsAccess(
            pVfs, pPager->zJournal, SQLITE_ACCESS_EXISTS, &bExists);
        if( rc==SQLITE_OK && bExists ){
          int fout = 0;
          int f = SQLITE_OPEN_READWRITE|SQLITE_OPEN_MAIN_JOURNAL;
          assert( !pPager->tempFile );
          rc = sqlite3OsOpen(pVfs, pPager->zJournal, pPager->jfd, f, &fout);
          assert( rc!=SQLITE_OK || isOpen(pPager->jfd) );
          if( rc==SQLITE_OK && fout&SQLITE_OPEN_READONLY ){
            rc = SQLITE_CANTOPEN_BKPT;
            sqlite3OsClose(pPager->jfd);
          }
        }
      }

      /* Playback and delete the journal.  Drop the database write
      ** lock and reacquire the read lock. Purge the cache before
      ** playing back the hot-journal so that we don't end up with
      ** an inconsistent cache.  Sync the hot journal before playing
      ** it back since the process that crashed and left the hot journal
      ** probably did not sync it and we are required to always sync
      ** the journal before playing it back.
      */
      if( isOpen(pPager->jfd) ){
        assert( rc==SQLITE_OK );
        rc = pagerSyncHotJournal(pPager);
        if( rc==SQLITE_OK ){
          rc = pager_playback(pPager, !pPager->tempFile);
          pPager->eState = PAGER_OPEN;
        }
      }else if( !pPager->exclusiveMode ){
        pagerUnlockDb(pPager, SHARED_LOCK);
      }

      if( rc!=SQLITE_OK ){
        /* This branch is taken if an error occurs while trying to open
        ** or roll back a hot-journal while holding an EXCLUSIVE lock. The
        ** pager_unlock() routine will be called before returning to unlock
        ** the file. If the unlock attempt fails, then Pager.eLock must be
        ** set to UNKNOWN_LOCK (see the comment above the #define for
        ** UNKNOWN_LOCK above for an explanation).
        **
        ** In order to get pager_unlock() to do this, set Pager.eState to
        ** PAGER_ERROR now. This is not actually counted as a transition
        ** to ERROR state in the state diagram at the top of this file,
        ** since we know that the same call to pager_unlock() will very
        ** shortly transition the pager object to the OPEN state. Calling
        ** assert_pager_state() would fail now, as it should not be possible
        ** to be in ERROR state when there are zero outstanding page
        ** references.
        */
        pager_error(pPager, rc);
        goto failed;
      }

      assert( pPager->eState==PAGER_OPEN );
      assert( (pPager->eLock==SHARED_LOCK)
           || (pPager->exclusiveMode && pPager->eLock>SHARED_LOCK)
      );
    }

    if( !pPager->tempFile && pPager->hasHeldSharedLock ){
      /* The shared-lock has just been acquired then check to
      ** see if the database has been modified.  If the database has changed,
      ** flush the cache.  The hasHeldSharedLock flag prevents this from
      ** occurring on the very first access to a file, in order to save a
      ** single unnecessary sqlite3OsRead() call at the start-up.
      **
      ** Database changes are detected by looking at 15 bytes beginning
      ** at offset 24 into the file.  The first 4 of these 16 bytes are
      ** a 32-bit counter that is incremented with each change.  The
      ** other bytes change randomly with each file change when
      ** a codec is in use.
      **
      ** There is a vanishingly small chance that a change will not be
      ** detected.  The chance of an undetected change is so small that
      ** it can be neglected.
      */
      char dbFileVers[sizeof(pPager->dbFileVers)];

      IOTRACE(("CKVERS %p %d\n", pPager, sizeof(dbFileVers)));
      rc = sqlite3OsRead(pPager->fd, &dbFileVers, sizeof(dbFileVers), 24);
      if( rc!=SQLITE_OK ){
        if( rc!=SQLITE_IOERR_SHORT_READ ){
          goto failed;
        }
        memset(dbFileVers, 0, sizeof(dbFileVers));
      }

      if( memcmp(pPager->dbFileVers, dbFileVers, sizeof(dbFileVers))!=0 ){
        pager_reset(pPager);

        /* Unmap the database file. It is possible that external processes
        ** may have truncated the database file and then extended it back
        ** to its original size while this process was not holding a lock.
        ** In this case there may exist a Pager.pMap mapping that appears
        ** to be the right size but is not actually valid. Avoid this
        ** possibility by unmapping the db here. */
        if( USEFETCH(pPager) ){
          sqlite3OsUnfetch(pPager->fd, 0, 0);
        }
      }
    }

    /* If there is a WAL file in the file-system, open this database in WAL
    ** mode. Otherwise, the following function call is a no-op.
    */
    rc = pagerOpenWalIfPresent(pPager);
#ifndef SQLITE_OMIT_WAL
    assert( pPager->pWal==0 || rc==SQLITE_OK );
#endif
  }

  if( pagerUseWal(pPager) ){
    assert( rc==SQLITE_OK );
    rc = pagerBeginReadTransaction(pPager);
  }

  if( pPager->tempFile==0 && pPager->eState==PAGER_OPEN && rc==SQLITE_OK ){
    rc = pagerPagecount(pPager, &pPager->dbSize);
  }

 failed:
  if( rc!=SQLITE_OK ){
    assert( !MEMDB );
    pager_unlock(pPager);
    assert( pPager->eState==PAGER_OPEN );
  }else{
    pPager->eState = PAGER_READER;
    pPager->hasHeldSharedLock = 1;
  }
  return rc;
}

/*
** If the reference count has reached zero, rollback any active
** transaction and unlock the pager.
**
** Except, in locking_mode=EXCLUSIVE when there is nothing to in
** the rollback journal, the unlock is not performed and there is
** nothing to rollback, so this routine is a no-op.
*/
static void pagerUnlockIfUnused(Pager *pPager){
  if( sqlite3PcacheRefCount(pPager->pPCache)==0 ){
    assert( pPager->nMmapOut==0 ); /* because page1 is never memory mapped */
    pagerUnlockAndRollback(pPager);
  }
}

/*
** The page getter methods each try to acquire a reference to a
** page with page number pgno. If the requested reference is
** successfully obtained, it is copied to *ppPage and SQLITE_OK returned.
**
** There are different implementations of the getter method depending
** on the current state of the pager.
**
**     getPageNormal()         --  The normal getter
**     getPageError()          --  Used if the pager is in an error state
**     getPageMmap()           --  Used if memory-mapped I/O is enabled
**
** If the requested page is already in the cache, it is returned.
** Otherwise, a new page object is allocated and populated with data
** read from the database file. In some cases, the pcache module may
** choose not to allocate a new page object and may reuse an existing
** object with no outstanding references.
**
** The extra data appended to a page is always initialized to zeros the
** first time a page is loaded into memory. If the page requested is
** already in the cache when this function is called, then the extra
** data is left as it was when the page object was last used.
**
** If the database image is smaller than the requested page or if
** the flags parameter contains the PAGER_GET_NOCONTENT bit and the
** requested page is not already stored in the cache, then no
** actual disk read occurs. In this case the memory image of the
** page is initialized to all zeros.
**
** If PAGER_GET_NOCONTENT is true, it means that we do not care about
** the contents of the page. This occurs in two scenarios:
**
**   a) When reading a free-list leaf page from the database, and
**
**   b) When a savepoint is being rolled back and we need to load
**      a new page into the cache to be filled with the data read
**      from the savepoint journal.
**
** If PAGER_GET_NOCONTENT is true, then the data returned is zeroed instead
** of being read from the database. Additionally, the bits corresponding
** to pgno in Pager.pInJournal (bitvec of pages already written to the
** journal file) and the PagerSavepoint.pInSavepoint bitvecs of any open
** savepoints are set. This means if the page is made writable at any
** point in the future, using a call to sqlite3PagerWrite(), its contents
** will not be journaled. This saves IO.
**
** The acquisition might fail for several reasons.  In all cases,
** an appropriate error code is returned and *ppPage is set to NULL.
**
** See also sqlite3PagerLookup().  Both this routine and Lookup() attempt
** to find a page in the in-memory cache first.  If the page is not already
** in memory, this routine goes to disk to read it in whereas Lookup()
** just returns 0.  This routine acquires a read-lock the first time it
** has to go to disk, and could also playback an old journal if necessary.
** Since Lookup() never goes to disk, it never has to deal with locks
** or journal files.
*/
static int getPageNormal(
  Pager *pPager,      /* The pager open on the database file */
  Pgno pgno,          /* Page number to fetch */
  DbPage **ppPage,    /* Write a pointer to the page here */
  int flags           /* PAGER_GET_XXX flags */
){
  int rc = SQLITE_OK;
  PgHdr *pPg;
  u8 noContent;                   /* True if PAGER_GET_NOCONTENT is set */
  sqlite3_pcache_page *pBase;

  assert( pPager->errCode==SQLITE_OK );
  assert( pPager->eState>=PAGER_READER );
  assert( assert_pager_state(pPager) );
  assert( pPager->hasHeldSharedLock==1 );

  if( pgno==0 ) return SQLITE_CORRUPT_BKPT;
  pBase = sqlite3PcacheFetch(pPager->pPCache, pgno, 3);
  if( pBase==0 ){
    pPg = 0;
    rc = sqlite3PcacheFetchStress(pPager->pPCache, pgno, &pBase);
    if( rc!=SQLITE_OK ) goto pager_acquire_err;
    if( pBase==0 ){
      rc = SQLITE_NOMEM_BKPT;
      goto pager_acquire_err;
    }
  }
  pPg = *ppPage = sqlite3PcacheFetchFinish(pPager->pPCache, pgno, pBase);
  assert( pPg==(*ppPage) );
  assert( pPg->pgno==pgno );
  assert( pPg->pPager==pPager || pPg->pPager==0 );

  noContent = (flags & PAGER_GET_NOCONTENT)!=0;
  if( pPg->pPager && !noContent ){
    /* In this case the pcache already contains an initialized copy of
    ** the page. Return without further ado.  */
    assert( pgno!=PAGER_SJ_PGNO(pPager) );
    pPager->aStat[PAGER_STAT_HIT]++;
    return SQLITE_OK;

  }else{
    /* The pager cache has created a new page. Its content needs to
    ** be initialized. But first some error checks:
    **
    ** (*) obsolete.  Was: maximum page number is 2^31
    ** (2) Never try to fetch the locking page
    */
    if( pgno==PAGER_SJ_PGNO(pPager) ){
      rc = SQLITE_CORRUPT_BKPT;
      goto pager_acquire_err;
    }

    pPg->pPager = pPager;

    assert( !isOpen(pPager->fd) || !MEMDB );
    if( !isOpen(pPager->fd) || pPager->dbSize<pgno || noContent ){
      if( pgno>pPager->mxPgno ){
        rc = SQLITE_FULL;
        if( pgno<=pPager->dbSize ){
          sqlite3PcacheRelease(pPg);
          pPg = 0;
        }
        goto pager_acquire_err;
      }
      if( noContent ){
        /* Failure to set the bits in the InJournal bit-vectors is benign.
        ** It merely means that we might do some extra work to journal a
        ** page that does not need to be journaled.  Nevertheless, be sure
        ** to test the case where a malloc error occurs while trying to set
        ** a bit in a bit vector.
        */
        sqlite3BeginBenignMalloc();
        if( pgno<=pPager->dbOrigSize ){
          TESTONLY( rc = ) sqlite3BitvecSet(pPager->pInJournal, pgno);
          testcase( rc==SQLITE_NOMEM );
        }
        TESTONLY( rc = ) addToSavepointBitvecs(pPager, pgno);
        testcase( rc==SQLITE_NOMEM );
        sqlite3EndBenignMalloc();
      }
      memset(pPg->pData, 0, pPager->pageSize);
      IOTRACE(("ZERO %p %d\n", pPager, pgno));
    }else{
      assert( pPg->pPager==pPager );
      pPager->aStat[PAGER_STAT_MISS]++;
      rc = readDbPage(pPg);
      if( rc!=SQLITE_OK ){
        goto pager_acquire_err;
      }
    }
    pager_set_pagehash(pPg);
  }
  return SQLITE_OK;

pager_acquire_err:
  assert( rc!=SQLITE_OK );
  if( pPg ){
    sqlite3PcacheDrop(pPg);
  }
  pagerUnlockIfUnused(pPager);
  *ppPage = 0;
  return rc;
}

#if SQLITE_MAX_MMAP_SIZE>0
/* The page getter for when memory-mapped I/O is enabled */
static int getPageMMap(
  Pager *pPager,      /* The pager open on the database file */
  Pgno pgno,          /* Page number to fetch */
  DbPage **ppPage,    /* Write a pointer to the page here */
  int flags           /* PAGER_GET_XXX flags */
){
  int rc = SQLITE_OK;
  PgHdr *pPg = 0;
  u32 iFrame = 0;                 /* Frame to read from WAL file */

  /* It is acceptable to use a read-only (mmap) page for any page except
  ** page 1 if there is no write-transaction open or the ACQUIRE_READONLY
  ** flag was specified by the caller. And so long as the db is not a
  ** temporary or in-memory database.  */
  const int bMmapOk = (pgno>1
   && (pPager->eState==PAGER_READER || (flags & PAGER_GET_READONLY))
  );

  assert( USEFETCH(pPager) );

  /* Optimization note:  Adding the "pgno<=1" term before "pgno==0" here
  ** allows the compiler optimizer to reuse the results of the "pgno>1"
  ** test in the previous statement, and avoid testing pgno==0 in the
  ** common case where pgno is large. */
  if( pgno<=1 && pgno==0 ){
    return SQLITE_CORRUPT_BKPT;
  }
  assert( pPager->eState>=PAGER_READER );
  assert( assert_pager_state(pPager) );
  assert( pPager->hasHeldSharedLock==1 );
  assert( pPager->errCode==SQLITE_OK );

  if( bMmapOk && pagerUseWal(pPager) ){
    rc = sqlite3WalFindFrame(pPager->pWal, pgno, &iFrame);
    if( rc!=SQLITE_OK ){
      *ppPage = 0;
      return rc;
    }
  }
  if( bMmapOk && iFrame==0 ){
    void *pData = 0;
    rc = sqlite3OsFetch(pPager->fd,
        (i64)(pgno-1) * pPager->pageSize, pPager->pageSize, &pData
    );
    if( rc==SQLITE_OK && pData ){
      if( pPager->eState>PAGER_READER || pPager->tempFile ){
        pPg = sqlite3PagerLookup(pPager, pgno);
      }
      if( pPg==0 ){
        rc = pagerAcquireMapPage(pPager, pgno, pData, &pPg);
      }else{
        sqlite3OsUnfetch(pPager->fd, (i64)(pgno-1)*pPager->pageSize, pData);
      }
      if( pPg ){
        assert( rc==SQLITE_OK );
        *ppPage = pPg;
        return SQLITE_OK;
      }
    }
    if( rc!=SQLITE_OK ){
      *ppPage = 0;
      return rc;
    }
  }
  return getPageNormal(pPager, pgno, ppPage, flags);
}
#endif /* SQLITE_MAX_MMAP_SIZE>0 */

/* The page getter method for when the pager is an error state */
static int getPageError(
  Pager *pPager,      /* The pager open on the database file */
  Pgno pgno,          /* Page number to fetch */
  DbPage **ppPage,    /* Write a pointer to the page here */
  int flags           /* PAGER_GET_XXX flags */
){
  UNUSED_PARAMETER(pgno);
  UNUSED_PARAMETER(flags);
  assert( pPager->errCode!=SQLITE_OK );
  *ppPage = 0;
  return pPager->errCode;
}


/* Dispatch all page fetch requests to the appropriate getter method.
*/
int sqlite3PagerGet(
  Pager *pPager,      /* The pager open on the database file */
  Pgno pgno,          /* Page number to fetch */
  DbPage **ppPage,    /* Write a pointer to the page here */
  int flags           /* PAGER_GET_XXX flags */
){
#if 0   /* Trace page fetch by setting to 1 */
  int rc;
  printf("PAGE %u\n", pgno);
  fflush(stdout);
  rc = pPager->xGet(pPager, pgno, ppPage, flags);
  if( rc ){ 
    printf("PAGE %u failed with 0x%02x\n", pgno, rc);
    fflush(stdout);
  }
  return rc;
#else
  /* Normal, high-speed version of sqlite3PagerGet() */
  return pPager->xGet(pPager, pgno, ppPage, flags);
#endif
}

/*
** Acquire a page if it is already in the in-memory cache.  Do
** not read the page from disk.  Return a pointer to the page,
** or 0 if the page is not in cache.
**
** See also sqlite3PagerGet().  The difference between this routine
** and sqlite3PagerGet() is that _get() will go to the disk and read
** in the page if the page is not already in cache.  This routine
** returns NULL if the page is not in cache or if a disk I/O error
** has ever happened.
*/
DbPage *sqlite3PagerLookup(Pager *pPager, Pgno pgno){
  sqlite3_pcache_page *pPage;
  assert( pPager!=0 );
  assert( pgno!=0 );
  assert( pPager->pPCache!=0 );
  pPage = sqlite3PcacheFetch(pPager->pPCache, pgno, 0);
  assert( pPage==0 || pPager->hasHeldSharedLock );
  if( pPage==0 ) return 0;
  return sqlite3PcacheFetchFinish(pPager->pPCache, pgno, pPage);
}

/*
** Release a page reference.
**
** The sqlite3PagerUnref() and sqlite3PagerUnrefNotNull() may only be used
** if we know that the page being released is not the last reference to page1.
** The btree layer always holds page1 open until the end, so these first
** two routines can be used to release any page other than BtShared.pPage1.
** The assert() at tag-20230419-2 proves that this constraint is always
** honored.
**
** Use sqlite3PagerUnrefPageOne() to release page1.  This latter routine
** checks the total number of outstanding pages and if the number of
** pages reaches zero it drops the database lock.
*/
void sqlite3PagerUnrefNotNull(DbPage *pPg){
  TESTONLY( Pager *pPager = pPg->pPager; )
  assert( pPg!=0 );
  if( pPg->flags & PGHDR_MMAP ){
    assert( pPg->pgno!=1 );  /* Page1 is never memory mapped */
    pagerReleaseMapPage(pPg);
  }else{
    sqlite3PcacheRelease(pPg);
  }
  /* Do not use this routine to release the last reference to page1 */
  assert( sqlite3PcacheRefCount(pPager->pPCache)>0 ); /* tag-20230419-2 */
}
void sqlite3PagerUnref(DbPage *pPg){
  if( pPg ) sqlite3PagerUnrefNotNull(pPg);
}
void sqlite3PagerUnrefPageOne(DbPage *pPg){
  Pager *pPager;
  assert( pPg!=0 );
  assert( pPg->pgno==1 );
  assert( (pPg->flags & PGHDR_MMAP)==0 ); /* Page1 is never memory mapped */
  pPager = pPg->pPager;
  sqlite3PcacheRelease(pPg);
  pagerUnlockIfUnused(pPager);
}

/*
** This function is called at the start of every write transaction.
** There must already be a RESERVED or EXCLUSIVE lock on the database
** file when this routine is called.
**
** Open the journal file for pager pPager and write a journal header
** to the start of it. If there are active savepoints, open the sub-journal
** as well. This function is only used when the journal file is being
** opened to write a rollback log for a transaction. It is not used
** when opening a hot journal file to roll it back.
**
** If the journal file is already open (as it may be in exclusive mode),
** then this function just writes a journal header to the start of the
** already open file.
**
** Whether or not the journal file is opened by this function, the
** Pager.pInJournal bitvec structure is allocated.
**
** Return SQLITE_OK if everything is successful. Otherwise, return
** SQLITE_NOMEM if the attempt to allocate Pager.pInJournal fails, or
** an IO error code if opening or writing the journal file fails.
*/
static int pager_open_journal(Pager *pPager){
  int rc = SQLITE_OK;                        /* Return code */
  sqlite3_vfs * const pVfs = pPager->pVfs;   /* Local cache of vfs pointer */

  assert( pPager->eState==PAGER_WRITER_LOCKED );
  assert( assert_pager_state(pPager) );
  assert( pPager->pInJournal==0 );
 
  /* If already in the error state, this function is a no-op.  But on
  ** the other hand, this routine is never called if we are already in
  ** an error state. */
  if( NEVER(pPager->errCode) ) return pPager->errCode;

  if( !pagerUseWal(pPager) && pPager->journalMode!=PAGER_JOURNALMODE_OFF ){
    pPager->pInJournal = sqlite3BitvecCreate(pPager->dbSize);
    if( pPager->pInJournal==0 ){
      return SQLITE_NOMEM_BKPT;
    }
 
    /* Open the journal file if it is not already open. */
    if( !isOpen(pPager->jfd) ){
      if( pPager->journalMode==PAGER_JOURNALMODE_MEMORY ){
        sqlite3MemJournalOpen(pPager->jfd);
      }else{
        int flags = SQLITE_OPEN_READWRITE|SQLITE_OPEN_CREATE;
        int nSpill;

        if( pPager->tempFile ){
          flags |= (SQLITE_OPEN_DELETEONCLOSE|SQLITE_OPEN_TEMP_JOURNAL);
          flags |= SQLITE_OPEN_EXCLUSIVE;
          nSpill = sqlite3Config.nStmtSpill;
        }else{
          flags |= SQLITE_OPEN_MAIN_JOURNAL;
          nSpill = jrnlBufferSize(pPager);
        }
         
        /* Verify that the database still has the same name as it did when
        ** it was originally opened. */
        rc = databaseIsUnmoved(pPager);
        if( rc==SQLITE_OK ){
          rc = sqlite3JournalOpen (
              pVfs, pPager->zJournal, pPager->jfd, flags, nSpill
          );
        }
      }
      assert( rc!=SQLITE_OK || isOpen(pPager->jfd) );
    }
 
 
    /* Write the first journal header to the journal file and open
    ** the sub-journal if necessary.
    */
    if( rc==SQLITE_OK ){
      /* TODO: Check if all of these are really required. */
      pPager->nRec = 0;
      pPager->journalOff = 0;
      pPager->setSuper = 0;
      pPager->journalHdr = 0;
      rc = writeJournalHdr(pPager);
    }
  }

  if( rc!=SQLITE_OK ){
    sqlite3BitvecDestroy(pPager->pInJournal);
    pPager->pInJournal = 0;
    pPager->journalOff = 0;
  }else{
    assert( pPager->eState==PAGER_WRITER_LOCKED );
    pPager->eState = PAGER_WRITER_CACHEMOD;
  }

  return rc;
}

/*
** Begin a write-transaction on the specified pager object. If a
** write-transaction has already been opened, this function is a no-op.
**
** If the exFlag argument is false, then acquire at least a RESERVED
** lock on the database file. If exFlag is true, then acquire at least
** an EXCLUSIVE lock. If such a lock is already held, no locking
** functions need be called.
**
** If the subjInMemory argument is non-zero, then any sub-journal opened
** within this transaction will be opened as an in-memory file. This
** has no effect if the sub-journal is already opened (as it may be when
** running in exclusive mode) or if the transaction does not require a
** sub-journal. If the subjInMemory argument is zero, then any required
** sub-journal is implemented in-memory if pPager is an in-memory database,
** or using a temporary file otherwise.
*/
int sqlite3PagerBegin(Pager *pPager, int exFlag, int subjInMemory){
  int rc = SQLITE_OK;

  if( pPager->errCode ) return pPager->errCode;
  assert( pPager->eState>=PAGER_READER && pPager->eState<PAGER_ERROR );
  pPager->subjInMemory = (u8)subjInMemory;

  if( pPager->eState==PAGER_READER ){
    assert( pPager->pInJournal==0 );

    if( pagerUseWal(pPager) ){
      /* If the pager is configured to use locking_mode=exclusive, and an
      ** exclusive lock on the database is not already held, obtain it now.
      */
      if( pPager->exclusiveMode && sqlite3WalExclusiveMode(pPager->pWal, -1) ){
        rc = pagerLockDb(pPager, EXCLUSIVE_LOCK);
        if( rc!=SQLITE_OK ){
          return rc;
        }
        (void)sqlite3WalExclusiveMode(pPager->pWal, 1);
      }

      /* Grab the write lock on the log file. If successful, upgrade to
      ** PAGER_RESERVED state. Otherwise, return an error code to the caller.
      ** The busy-handler is not invoked if another connection already
      ** holds the write-lock. If possible, the upper layer will call it.
      */
      rc = sqlite3WalBeginWriteTransaction(pPager->pWal);
    }else{
      /* Obtain a RESERVED lock on the database file. If the exFlag parameter
      ** is true, then immediately upgrade this to an EXCLUSIVE lock. The
      ** busy-handler callback can be used when upgrading to the EXCLUSIVE
      ** lock, but not when obtaining the RESERVED lock.
      */
      rc = pagerLockDb(pPager, RESERVED_LOCK);
      if( rc==SQLITE_OK && exFlag ){
        rc = pager_wait_on_lock(pPager, EXCLUSIVE_LOCK);
      }
    }

    if( rc==SQLITE_OK ){
      /* Change to WRITER_LOCKED state.
      **
      ** WAL mode sets Pager.eState to PAGER_WRITER_LOCKED or CACHEMOD
      ** when it has an open transaction, but never to DBMOD or FINISHED.
      ** This is because in those states the code to roll back savepoint
      ** transactions may copy data from the sub-journal into the database
      ** file as well as into the page cache. Which would be incorrect in
      ** WAL mode.
      */
      pPager->eState = PAGER_WRITER_LOCKED;
      pPager->dbHintSize = pPager->dbSize;
      pPager->dbFileSize = pPager->dbSize;
      pPager->dbOrigSize = pPager->dbSize;
      pPager->journalOff = 0;
    }

    assert( rc==SQLITE_OK || pPager->eState==PAGER_READER );
    assert( rc!=SQLITE_OK || pPager->eState==PAGER_WRITER_LOCKED );
    assert( assert_pager_state(pPager) );
  }

  PAGERTRACE(("TRANSACTION %d\n", PAGERID(pPager)));
  return rc;
}

/*
** Write page pPg onto the end of the rollback journal.
*/
static SQLITE_NOINLINE int pagerAddPageToRollbackJournal(PgHdr *pPg){
  Pager *pPager = pPg->pPager;
  int rc;
  u32 cksum;
  char *pData2;
  i64 iOff = pPager->journalOff;

  /* We should never write to the journal file the page that
  ** contains the database locks.  The following assert verifies
  ** that we do not. */
  assert( pPg->pgno!=PAGER_SJ_PGNO(pPager) );

  assert( pPager->journalHdr<=pPager->journalOff );
  pData2 = pPg->pData;
  cksum = pager_cksum(pPager, (u8*)pData2);

  /* Even if an IO or diskfull error occurs while journalling the
  ** page in the block above, set the need-sync flag for the page.
  ** Otherwise, when the transaction is rolled back, the logic in
  ** playback_one_page() will think that the page needs to be restored
  ** in the database file. And if an IO error occurs while doing so,
  ** then corruption may follow.
  */
  pPg->flags |= PGHDR_NEED_SYNC;

  rc = write32bits(pPager->jfd, iOff, pPg->pgno);
  if( rc!=SQLITE_OK ) return rc;
  rc = sqlite3OsWrite(pPager->jfd, pData2, pPager->pageSize, iOff+4);
  if( rc!=SQLITE_OK ) return rc;
  rc = write32bits(pPager->jfd, iOff+pPager->pageSize+4, cksum);
  if( rc!=SQLITE_OK ) return rc;

  IOTRACE(("JOUT %p %d %lld %d\n", pPager, pPg->pgno,
           pPager->journalOff, pPager->pageSize));
  PAGER_INCR(sqlite3_pager_writej_count);
  PAGERTRACE(("JOURNAL %d page %d needSync=%d hash(%08x)\n",
       PAGERID(pPager), pPg->pgno,
       ((pPg->flags&PGHDR_NEED_SYNC)?1:0), pager_pagehash(pPg)));

  pPager->journalOff += 8 + pPager->pageSize;
  pPager->nRec++;
  assert( pPager->pInJournal!=0 );
  rc = sqlite3BitvecSet(pPager->pInJournal, pPg->pgno);
  testcase( rc==SQLITE_NOMEM );
  assert( rc==SQLITE_OK || rc==SQLITE_NOMEM );
  rc |= addToSavepointBitvecs(pPager, pPg->pgno);
  assert( rc==SQLITE_OK || rc==SQLITE_NOMEM );
  return rc;
}

/*
** Mark a single data page as writeable. The page is written into the
** main journal or sub-journal as required. If the page is written into
** one of the journals, the corresponding bit is set in the
** Pager.pInJournal bitvec and the PagerSavepoint.pInSavepoint bitvecs
** of any open savepoints as appropriate.
*/
static int pager_write(PgHdr *pPg){
  Pager *pPager = pPg->pPager;
  int rc = SQLITE_OK;

  /* This routine is not called unless a write-transaction has already
  ** been started. The journal file may or may not be open at this point.
  ** It is never called in the ERROR state.
  */
  assert( pPager->eState==PAGER_WRITER_LOCKED
       || pPager->eState==PAGER_WRITER_CACHEMOD
       || pPager->eState==PAGER_WRITER_DBMOD
  );
  assert( assert_pager_state(pPager) );
  assert( pPager->errCode==0 );
  assert( pPager->readOnly==0 );
  CHECK_PAGE(pPg);

  /* The journal file needs to be opened. Higher level routines have already
  ** obtained the necessary locks to begin the write-transaction, but the
  ** rollback journal might not yet be open. Open it now if this is the case.
  **
  ** This is done before calling sqlite3PcacheMakeDirty() on the page.
  ** Otherwise, if it were done after calling sqlite3PcacheMakeDirty(), then
  ** an error might occur and the pager would end up in WRITER_LOCKED state
  ** with pages marked as dirty in the cache.
  */
  if( pPager->eState==PAGER_WRITER_LOCKED ){
    rc = pager_open_journal(pPager);
    if( rc!=SQLITE_OK ) return rc;
  }
  assert( pPager->eState>=PAGER_WRITER_CACHEMOD );
  assert( assert_pager_state(pPager) );

  /* Mark the page that is about to be modified as dirty. */
  sqlite3PcacheMakeDirty(pPg);

  /* If a rollback journal is in use, them make sure the page that is about
  ** to change is in the rollback journal, or if the page is a new page off
  ** then end of the file, make sure it is marked as PGHDR_NEED_SYNC.
  */
  assert( (pPager->pInJournal!=0) == isOpen(pPager->jfd) );
  if( pPager->pInJournal!=0
   && sqlite3BitvecTestNotNull(pPager->pInJournal, pPg->pgno)==0
  ){
    assert( pagerUseWal(pPager)==0 );
    if( pPg->pgno<=pPager->dbOrigSize ){
      rc = pagerAddPageToRollbackJournal(pPg);
      if( rc!=SQLITE_OK ){
        return rc;
      }
    }else{
      if( pPager->eState!=PAGER_WRITER_DBMOD ){
        pPg->flags |= PGHDR_NEED_SYNC;
      }
      PAGERTRACE(("APPEND %d page %d needSync=%d\n",
              PAGERID(pPager), pPg->pgno,
             ((pPg->flags&PGHDR_NEED_SYNC)?1:0)));
    }
  }

  /* The PGHDR_DIRTY bit is set above when the page was added to the dirty-list
  ** and before writing the page into the rollback journal.  Wait until now,
  ** after the page has been successfully journalled, before setting the
  ** PGHDR_WRITEABLE bit that indicates that the page can be safely modified.
  */
  pPg->flags |= PGHDR_WRITEABLE;
 
  /* If the statement journal is open and the page is not in it,
  ** then write the page into the statement journal.
  */
  if( pPager->nSavepoint>0 ){
    rc = subjournalPageIfRequired(pPg);
  }

  /* Update the database size and return. */
  if( pPager->dbSize<pPg->pgno ){
    pPager->dbSize = pPg->pgno;
  }
  return rc;
}

/*
** This is a variant of sqlite3PagerWrite() that runs when the sector size
** is larger than the page size.  SQLite makes the (reasonable) assumption that
** all bytes of a sector are written together by hardware.  Hence, all bytes of
** a sector need to be journalled in case of a power loss in the middle of
** a write.
**
** Usually, the sector size is less than or equal to the page size, in which
** case pages can be individually written.  This routine only runs in the
** exceptional case where the page size is smaller than the sector size.
*/
static SQLITE_NOINLINE int pagerWriteLargeSector(PgHdr *pPg){
  int rc = SQLITE_OK;          /* Return code */
  Pgno nPageCount;             /* Total number of pages in database file */
  Pgno pg1;                    /* First page of the sector pPg is located on. */
  int nPage = 0;               /* Number of pages starting at pg1 to journal */
  int ii;                      /* Loop counter */
  int needSync = 0;            /* True if any page has PGHDR_NEED_SYNC */
  Pager *pPager = pPg->pPager; /* The pager that owns pPg */
  Pgno nPagePerSector = (pPager->sectorSize/pPager->pageSize);

  /* Set the doNotSpill NOSYNC bit to 1. This is because we cannot allow
  ** a journal header to be written between the pages journaled by
  ** this function.
  */
  assert( !MEMDB );
  assert( (pPager->doNotSpill & SPILLFLAG_NOSYNC)==0 );
  pPager->doNotSpill |= SPILLFLAG_NOSYNC;

  /* This trick assumes that both the page-size and sector-size are
  ** an integer power of 2. It sets variable pg1 to the identifier
  ** of the first page of the sector pPg is located on.
  */
  pg1 = ((pPg->pgno-1) & ~(nPagePerSector-1)) + 1;

  nPageCount = pPager->dbSize;
  if( pPg->pgno>nPageCount ){
    nPage = (pPg->pgno - pg1)+1;
  }else if( (pg1+nPagePerSector-1)>nPageCount ){
    nPage = nPageCount+1-pg1;
  }else{
    nPage = nPagePerSector;
  }
  assert(nPage>0);
  assert(pg1<=pPg->pgno);
  assert((pg1+nPage)>pPg->pgno);

  for(ii=0; ii<nPage && rc==SQLITE_OK; ii++){
    Pgno pg = pg1+ii;
    PgHdr *pPage;
    if( pg==pPg->pgno || !sqlite3BitvecTest(pPager->pInJournal, pg) ){
      if( pg!=PAGER_SJ_PGNO(pPager) ){
        rc = sqlite3PagerGet(pPager, pg, &pPage, 0);
        if( rc==SQLITE_OK ){
          rc = pager_write(pPage);
          if( pPage->flags&PGHDR_NEED_SYNC ){
            needSync = 1;
          }
          sqlite3PagerUnrefNotNull(pPage);
        }
      }
    }else if( (pPage = sqlite3PagerLookup(pPager, pg))!=0 ){
      if( pPage->flags&PGHDR_NEED_SYNC ){
        needSync = 1;
      }
      sqlite3PagerUnrefNotNull(pPage);
    }
  }

  /* If the PGHDR_NEED_SYNC flag is set for any of the nPage pages
  ** starting at pg1, then it needs to be set for all of them. Because
  ** writing to any of these nPage pages may damage the others, the
  ** journal file must contain sync()ed copies of all of them
  ** before any of them can be written out to the database file.
  */
  if( rc==SQLITE_OK && needSync ){
    assert( !MEMDB );
    for(ii=0; ii<nPage; ii++){
      PgHdr *pPage = sqlite3PagerLookup(pPager, pg1+ii);
      if( pPage ){
        pPage->flags |= PGHDR_NEED_SYNC;
        sqlite3PagerUnrefNotNull(pPage);
      }
    }
  }

  assert( (pPager->doNotSpill & SPILLFLAG_NOSYNC)!=0 );
  pPager->doNotSpill &= ~SPILLFLAG_NOSYNC;
  return rc;
}

/*
** Mark a data page as writeable. This routine must be called before
** making changes to a page. The caller must check the return value
** of this function and be careful not to change any page data unless
** this routine returns SQLITE_OK.
**
** The difference between this function and pager_write() is that this
** function also deals with the special case where 2 or more pages
** fit on a single disk sector. In this case all co-resident pages
** must have been written to the journal file before returning.
**
** If an error occurs, SQLITE_NOMEM or an IO error code is returned
** as appropriate. Otherwise, SQLITE_OK.
*/
int sqlite3PagerWrite(PgHdr *pPg){
  Pager *pPager = pPg->pPager;
  assert( (pPg->flags & PGHDR_MMAP)==0 );
  assert( pPager->eState>=PAGER_WRITER_LOCKED );
  assert( assert_pager_state(pPager) );
  if( (pPg->flags & PGHDR_WRITEABLE)!=0 && pPager->dbSize>=pPg->pgno ){
    if( pPager->nSavepoint ) return subjournalPageIfRequired(pPg);
    return SQLITE_OK;
  }else if( pPager->errCode ){
    return pPager->errCode;
  }else if( pPager->sectorSize > (u32)pPager->pageSize ){
    assert( pPager->tempFile==0 );
    return pagerWriteLargeSector(pPg);
  }else{
    return pager_write(pPg);
  }
}

/*
** Return TRUE if the page given in the argument was previously passed
** to sqlite3PagerWrite().  In other words, return TRUE if it is ok
** to change the content of the page.
*/
#ifndef NDEBUG
int sqlite3PagerIswriteable(DbPage *pPg){
  return pPg->flags & PGHDR_WRITEABLE;
}
#endif

/*
** A call to this routine tells the pager that it is not necessary to
** write the information on page pPg back to the disk, even though
** that page might be marked as dirty.  This happens, for example, when
** the page has been added as a leaf of the freelist and so its
** content no longer matters.
**
** The overlying software layer calls this routine when all of the data
** on the given page is unused. The pager marks the page as clean so
** that it does not get written to disk.
**
** Tests show that this optimization can quadruple the speed of large
** DELETE operations.
**
** This optimization cannot be used with a temp-file, as the page may
** have been dirty at the start of the transaction. In that case, if
** memory pressure forces page pPg out of the cache, the data does need
** to be written out to disk so that it may be read back in if the
** current transaction is rolled back.
*/
void sqlite3PagerDontWrite(PgHdr *pPg){
  Pager *pPager = pPg->pPager;
  if( !pPager->tempFile && (pPg->flags&PGHDR_DIRTY) && pPager->nSavepoint==0 ){
    PAGERTRACE(("DONT_WRITE page %d of %d\n", pPg->pgno, PAGERID(pPager)));
    IOTRACE(("CLEAN %p %d\n", pPager, pPg->pgno))
    pPg->flags |= PGHDR_DONT_WRITE;
    pPg->flags &= ~PGHDR_WRITEABLE;
    testcase( pPg->flags & PGHDR_NEED_SYNC );
    pager_set_pagehash(pPg);
  }
}

/*
** This routine is called to increment the value of the database file
** change-counter, stored as a 4-byte big-endian integer starting at
** byte offset 24 of the pager file.  The secondary change counter at
** 92 is also updated, as is the SQLite version number at offset 96.
**
** But this only happens if the pPager->changeCountDone flag is false.
** To avoid excess churning of page 1, the update only happens once.
** See also the pager_write_changecounter() routine that does an
** unconditional update of the change counters.
**
** If the isDirectMode flag is zero, then this is done by calling
** sqlite3PagerWrite() on page 1, then modifying the contents of the
** page data. In this case the file will be updated when the current
** transaction is committed.
**
** The isDirectMode flag may only be non-zero if the library was compiled
** with the SQLITE_ENABLE_ATOMIC_WRITE macro defined. In this case,
** if isDirect is non-zero, then the database file is updated directly
** by writing an updated version of page 1 using a call to the
** sqlite3OsWrite() function.
*/
static int pager_incr_changecounter(Pager *pPager, int isDirectMode){
  int rc = SQLITE_OK;

  assert( pPager->eState==PAGER_WRITER_CACHEMOD
       || pPager->eState==PAGER_WRITER_DBMOD
  );
  assert( assert_pager_state(pPager) );

  /* Declare and initialize constant integer 'isDirect'. If the
  ** atomic-write optimization is enabled in this build, then isDirect
  ** is initialized to the value passed as the isDirectMode parameter
  ** to this function. Otherwise, it is always set to zero.
  **
  ** The idea is that if the atomic-write optimization is not
  ** enabled at compile time, the compiler can omit the tests of
  ** 'isDirect' below, as well as the block enclosed in the
  ** "if( isDirect )" condition.
  */
#ifndef SQLITE_ENABLE_ATOMIC_WRITE
# define DIRECT_MODE 0
  assert( isDirectMode==0 );
  UNUSED_PARAMETER(isDirectMode);
#else
# define DIRECT_MODE isDirectMode
#endif

  if( !pPager->changeCountDone && pPager->dbSize>0 ){
    PgHdr *pPgHdr;                /* Reference to page 1 */

    assert( !pPager->tempFile && isOpen(pPager->fd) );

    /* Open page 1 of the file for writing. */
    rc = sqlite3PagerGet(pPager, 1, &pPgHdr, 0);
    assert( pPgHdr==0 || rc==SQLITE_OK );

    /* If page one was fetched successfully, and this function is not
    ** operating in direct-mode, make page 1 writable.  When not in
    ** direct mode, page 1 is always held in cache and hence the PagerGet()
    ** above is always successful - hence the ALWAYS on rc==SQLITE_OK.
    */
    if( !DIRECT_MODE && ALWAYS(rc==SQLITE_OK) ){
      rc = sqlite3PagerWrite(pPgHdr);
    }

    if( rc==SQLITE_OK ){
      /* Actually do the update of the change counter */
      pager_write_changecounter(pPgHdr);

      /* If running in direct mode, write the contents of page 1 to the file. */
      if( DIRECT_MODE ){
        const void *zBuf;
        assert( pPager->dbFileSize>0 );
        zBuf = pPgHdr->pData;
        if( rc==SQLITE_OK ){
          rc = sqlite3OsWrite(pPager->fd, zBuf, pPager->pageSize, 0);
          pPager->aStat[PAGER_STAT_WRITE]++;
        }
        if( rc==SQLITE_OK ){
          /* Update the pager's copy of the change-counter. Otherwise, the
          ** next time a read transaction is opened the cache will be
          ** flushed (as the change-counter values will not match).  */
          const void *pCopy = (const void *)&((const char *)zBuf)[24];
          memcpy(&pPager->dbFileVers, pCopy, sizeof(pPager->dbFileVers));
          pPager->changeCountDone = 1;
        }
      }else{
        pPager->changeCountDone = 1;
      }
    }

    /* Release the page reference. */
    sqlite3PagerUnref(pPgHdr);
  }
  return rc;
}

/*
** Sync the database file to disk. This is a no-op for in-memory databases
** or pages with the Pager.noSync flag set.
**
** If successful, or if called on a pager for which it is a no-op, this
** function returns SQLITE_OK. Otherwise, an IO error code is returned.
*/
int sqlite3PagerSync(Pager *pPager, const char *zSuper){
  int rc = SQLITE_OK;
  void *pArg = (void*)zSuper;
  rc = sqlite3OsFileControl(pPager->fd, SQLITE_FCNTL_SYNC, pArg);
  if( rc==SQLITE_NOTFOUND ) rc = SQLITE_OK;
  if( rc==SQLITE_OK && !pPager->noSync ){
    assert( !MEMDB );
    rc = sqlite3OsSync(pPager->fd, pPager->syncFlags);
  }
  return rc;
}

/*
** This function may only be called while a write-transaction is active in
** rollback. If the connection is in WAL mode, this call is a no-op.
** Otherwise, if the connection does not already have an EXCLUSIVE lock on
** the database file, an attempt is made to obtain one.
**
** If the EXCLUSIVE lock is already held or the attempt to obtain it is
** successful, or the connection is in WAL mode, SQLITE_OK is returned.
** Otherwise, either SQLITE_BUSY or an SQLITE_IOERR_XXX error code is
** returned.
*/
int sqlite3PagerExclusiveLock(Pager *pPager){
  int rc = pPager->errCode;
  assert( assert_pager_state(pPager) );
  if( rc==SQLITE_OK ){
    assert( pPager->eState==PAGER_WRITER_CACHEMOD
         || pPager->eState==PAGER_WRITER_DBMOD
         || pPager->eState==PAGER_WRITER_LOCKED
    );
    assert( assert_pager_state(pPager) );
    if( 0==pagerUseWal(pPager) ){
      rc = pager_wait_on_lock(pPager, EXCLUSIVE_LOCK);
    }
  }
  return rc;
}

/*
** Sync the database file for the pager pPager. zSuper points to the name
** of a super-journal file that should be written into the individual
** journal file. zSuper may be NULL, which is interpreted as no
** super-journal (a single database transaction).
**
** This routine ensures that:
**
**   * The database file change-counter is updated,
**   * the journal is synced (unless the atomic-write optimization is used),
**   * all dirty pages are written to the database file,
**   * the database file is truncated (if required), and
**   * the database file synced.
**
** The only thing that remains to commit the transaction is to finalize
** (delete, truncate or zero the first part of) the journal file (or
** delete the super-journal file if specified).
**
** Note that if zSuper==NULL, this does not overwrite a previous value
** passed to an sqlite3PagerCommitPhaseOne() call.
**
** If the final parameter - noSync - is true, then the database file itself
** is not synced. The caller must call sqlite3PagerSync() directly to
** sync the database file before calling CommitPhaseTwo() to delete the
** journal file in this case.
*/
int sqlite3PagerCommitPhaseOne(
  Pager *pPager,                  /* Pager object */
  const char *zSuper,            /* If not NULL, the super-journal name */
  int noSync                      /* True to omit the xSync on the db file */
){
  int rc = SQLITE_OK;             /* Return code */

  assert( pPager->eState==PAGER_WRITER_LOCKED
       || pPager->eState==PAGER_WRITER_CACHEMOD
       || pPager->eState==PAGER_WRITER_DBMOD
       || pPager->eState==PAGER_ERROR
  );
  assert( assert_pager_state(pPager) );

  /* If a prior error occurred, report that error again. */
  if( NEVER(pPager->errCode) ) return pPager->errCode;

  /* Provide the ability to easily simulate an I/O error during testing */
  if( sqlite3FaultSim(400) ) return SQLITE_IOERR;

  PAGERTRACE(("DATABASE SYNC: File=%s zSuper=%s nSize=%d\n",
      pPager->zFilename, zSuper, pPager->dbSize));

  /* If no database changes have been made, return early. */
  if( pPager->eState<PAGER_WRITER_CACHEMOD ) return SQLITE_OK;

  assert( MEMDB==0 || pPager->tempFile );
  assert( isOpen(pPager->fd) || pPager->tempFile );
  if( 0==pagerFlushOnCommit(pPager, 1) ){
    /* If this is an in-memory db, or no pages have been written to, or this
    ** function has already been called, it is mostly a no-op.  However, any
    ** backup in progress needs to be restarted.  */
    sqlite3BackupRestart(pPager->pBackup);
  }else{
    PgHdr *pList;
    if( pagerUseWal(pPager) ){
      PgHdr *pPageOne = 0;
      pList = sqlite3PcacheDirtyList(pPager->pPCache);
      if( pList==0 ){
        /* Must have at least one page for the WAL commit flag.
        ** Ticket [2d1a5c67dfc2363e44f29d9bbd57f] 2011-05-18 */
        rc = sqlite3PagerGet(pPager, 1, &pPageOne, 0);
        pList = pPageOne;
        pList->pDirty = 0;
      }
      assert( rc==SQLITE_OK );
      if( ALWAYS(pList) ){
        rc = pagerWalFrames(pPager, pList, pPager->dbSize, 1);
      }
      sqlite3PagerUnref(pPageOne);
      if( rc==SQLITE_OK ){
        sqlite3PcacheCleanAll(pPager->pPCache);
      }
    }else{
      /* The bBatch boolean is true if the batch-atomic-write commit method
      ** should be used.  No rollback journal is created if batch-atomic-write
      ** is enabled.
      */
#ifdef SQLITE_ENABLE_BATCH_ATOMIC_WRITE
      sqlite3_file *fd = pPager->fd;
      int bBatch = zSuper==0    /* An SQLITE_IOCAP_BATCH_ATOMIC commit */
        && (sqlite3OsDeviceCharacteristics(fd) & SQLITE_IOCAP_BATCH_ATOMIC)
        && !pPager->noSync
        && sqlite3JournalIsInMemory(pPager->jfd);
#else
#     define bBatch 0
#endif

#ifdef SQLITE_ENABLE_ATOMIC_WRITE
      /* The following block updates the change-counter. Exactly how it
      ** does this depends on whether or not the atomic-update optimization
      ** was enabled at compile time, and if this transaction meets the
      ** runtime criteria to use the operation:
      **
      **    * The file-system supports the atomic-write property for
      **      blocks of size page-size, and
      **    * This commit is not part of a multi-file transaction, and
      **    * Exactly one page has been modified and store in the journal file.
      **
      ** If the optimization was not enabled at compile time, then the
      ** pager_incr_changecounter() function is called to update the change
      ** counter in 'indirect-mode'. If the optimization is compiled in but
      ** is not applicable to this transaction, call sqlite3JournalCreate()
      ** to make sure the journal file has actually been created, then call
      ** pager_incr_changecounter() to update the change-counter in indirect
      ** mode.
      **
      ** Otherwise, if the optimization is both enabled and applicable,
      ** then call pager_incr_changecounter() to update the change-counter
      ** in 'direct' mode. In this case the journal file will never be
      ** created for this transaction.
      */
      if( bBatch==0 ){
        PgHdr *pPg;
        assert( isOpen(pPager->jfd)
            || pPager->journalMode==PAGER_JOURNALMODE_OFF
            || pPager->journalMode==PAGER_JOURNALMODE_WAL
            );
        if( !zSuper && isOpen(pPager->jfd)
         && pPager->journalOff==jrnlBufferSize(pPager)
         && pPager->dbSize>=pPager->dbOrigSize
         && (!(pPg = sqlite3PcacheDirtyList(pPager->pPCache)) || 0==pPg->pDirty)
        ){
          /* Update the db file change counter via the direct-write method. The
          ** following call will modify the in-memory representation of page 1
          ** to include the updated change counter and then write page 1
          ** directly to the database file. Because of the atomic-write
          ** property of the host file-system, this is safe.
          */
          rc = pager_incr_changecounter(pPager, 1);
        }else{
          rc = sqlite3JournalCreate(pPager->jfd);
          if( rc==SQLITE_OK ){
            rc = pager_incr_changecounter(pPager, 0);
          }
        }
      }
#else  /* SQLITE_ENABLE_ATOMIC_WRITE */
#ifdef SQLITE_ENABLE_BATCH_ATOMIC_WRITE
      if( zSuper ){
        rc = sqlite3JournalCreate(pPager->jfd);
        if( rc!=SQLITE_OK ) goto commit_phase_one_exit;
        assert( bBatch==0 );
      }
#endif
      rc = pager_incr_changecounter(pPager, 0);
#endif /* !SQLITE_ENABLE_ATOMIC_WRITE */
      if( rc!=SQLITE_OK ) goto commit_phase_one_exit;
 
      /* Write the super-journal name into the journal file. If a
      ** super-journal file name has already been written to the journal file,
      ** or if zSuper is NULL (no super-journal), then this call is a no-op.
      */
      rc = writeSuperJournal(pPager, zSuper);
      if( rc!=SQLITE_OK ) goto commit_phase_one_exit;
 
      /* Sync the journal file and write all dirty pages to the database.
      ** If the atomic-update optimization is being used, this sync will not
      ** create the journal file or perform any real IO.
      **
      ** Because the change-counter page was just modified, unless the
      ** atomic-update optimization is used it is almost certain that the
      ** journal requires a sync here. However, in locking_mode=exclusive
      ** on a system under memory pressure it is just possible that this is
      ** not the case. In this case it is likely enough that the redundant
      ** xSync() call will be changed to a no-op by the OS anyhow.
      */
      rc = syncJournal(pPager, 0);
      if( rc!=SQLITE_OK ) goto commit_phase_one_exit;

      pList = sqlite3PcacheDirtyList(pPager->pPCache);
#ifdef SQLITE_ENABLE_BATCH_ATOMIC_WRITE
      if( bBatch ){
        rc = sqlite3OsFileControl(fd, SQLITE_FCNTL_BEGIN_ATOMIC_WRITE, 0);
        if( rc==SQLITE_OK ){
          rc = pager_write_pagelist(pPager, pList);
          if( rc==SQLITE_OK && pPager->dbSize>pPager->dbFileSize ){
            char *pTmp = pPager->pTmpSpace;
            int szPage = (int)pPager->pageSize;
            memset(pTmp, 0, szPage);
            rc = sqlite3OsWrite(pPager->fd, pTmp, szPage,
                      ((i64)pPager->dbSize*pPager->pageSize)-szPage);
          }
          if( rc==SQLITE_OK ){
            rc = sqlite3OsFileControl(fd, SQLITE_FCNTL_COMMIT_ATOMIC_WRITE, 0);
          }
          if( rc!=SQLITE_OK ){
            sqlite3OsFileControlHint(fd, SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE, 0);
          }
        }

        if( (rc&0xFF)==SQLITE_IOERR && rc!=SQLITE_IOERR_NOMEM ){
          rc = sqlite3JournalCreate(pPager->jfd);
          if( rc!=SQLITE_OK ){
            sqlite3OsClose(pPager->jfd);
            goto commit_phase_one_exit;
          }
          bBatch = 0;
        }else{
          sqlite3OsClose(pPager->jfd);
        }
      }
#endif /* SQLITE_ENABLE_BATCH_ATOMIC_WRITE */

      if( bBatch==0 ){
        rc = pager_write_pagelist(pPager, pList);
      }
      if( rc!=SQLITE_OK ){
        assert( rc!=SQLITE_IOERR_BLOCKED );
        goto commit_phase_one_exit;
      }
      sqlite3PcacheCleanAll(pPager->pPCache);

      /* If the file on disk is smaller than the database image, use
      ** pager_truncate to grow the file here. This can happen if the database
      ** image was extended as part of the current transaction and then the
      ** last page in the db image moved to the free-list. In this case the
      ** last page is never written out to disk, leaving the database file
      ** undersized. Fix this now if it is the case.  */
      if( pPager->dbSize>pPager->dbFileSize ){
        Pgno nNew = pPager->dbSize - (pPager->dbSize==PAGER_SJ_PGNO(pPager));
        assert( pPager->eState==PAGER_WRITER_DBMOD );
        rc = pager_truncate(pPager, nNew);
        if( rc!=SQLITE_OK ) goto commit_phase_one_exit;
      }
 
      /* Finally, sync the database file. */
      if( !noSync ){
        rc = sqlite3PagerSync(pPager, zSuper);
      }
      IOTRACE(("DBSYNC %p\n", pPager))
    }
  }

commit_phase_one_exit:
  if( rc==SQLITE_OK && !pagerUseWal(pPager) ){
    pPager->eState = PAGER_WRITER_FINISHED;
  }
  return rc;
}


/*
** When this function is called, the database file has been completely
** updated to reflect the changes made by the current transaction and
** synced to disk. The journal file still exists in the file-system
** though, and if a failure occurs at this point it will eventually
** be used as a hot-journal and the current transaction rolled back.
**
** This function finalizes the journal file, either by deleting,
** truncating or partially zeroing it, so that it cannot be used
** for hot-journal rollback. Once this is done the transaction is
** irrevocably committed.
**
** If an error occurs, an IO error code is returned and the pager
** moves into the error state. Otherwise, SQLITE_OK is returned.
*/
int sqlite3PagerCommitPhaseTwo(Pager *pPager){
  int rc = SQLITE_OK;                  /* Return code */

  /* This routine should not be called if a prior error has occurred.
  ** But if (due to a coding error elsewhere in the system) it does get
  ** called, just return the same error code without doing anything. */
  if( NEVER(pPager->errCode) ) return pPager->errCode;
  pPager->iDataVersion++;

  assert( pPager->eState==PAGER_WRITER_LOCKED
       || pPager->eState==PAGER_WRITER_FINISHED
       || (pagerUseWal(pPager) && pPager->eState==PAGER_WRITER_CACHEMOD)
  );
  assert( assert_pager_state(pPager) );

  /* An optimization. If the database was not actually modified during
  ** this transaction, the pager is running in exclusive-mode and is
  ** using persistent journals, then this function is a no-op.
  **
  ** The start of the journal file currently contains a single journal
  ** header with the nRec field set to 0. If such a journal is used as
  ** a hot-journal during hot-journal rollback, 0 changes will be made
  ** to the database file. So there is no need to zero the journal
  ** header. Since the pager is in exclusive mode, there is no need
  ** to drop any locks either.
  */
  if( pPager->eState==PAGER_WRITER_LOCKED
   && pPager->exclusiveMode
   && pPager->journalMode==PAGER_JOURNALMODE_PERSIST
  ){
    assert( pPager->journalOff==JOURNAL_HDR_SZ(pPager) || !pPager->journalOff );
    pPager->eState = PAGER_READER;
    return SQLITE_OK;
  }

  PAGERTRACE(("COMMIT %d\n", PAGERID(pPager)));
  rc = pager_end_transaction(pPager, pPager->setSuper, 1);
  return pager_error(pPager, rc);
}

/*
** If a write transaction is open, then all changes made within the
** transaction are reverted and the current write-transaction is closed.
** The pager falls back to PAGER_READER state if successful, or PAGER_ERROR
** state if an error occurs.
**
** If the pager is already in PAGER_ERROR state when this function is called,
** it returns Pager.errCode immediately. No work is performed in this case.
**
** Otherwise, in rollback mode, this function performs two functions:
**
**   1) It rolls back the journal file, restoring all database file and
**      in-memory cache pages to the state they were in when the transaction
**      was opened, and
**
**   2) It finalizes the journal file, so that it is not used for hot
**      rollback at any point in the future.
**
** Finalization of the journal file (task 2) is only performed if the
** rollback is successful.
**
** In WAL mode, all cache-entries containing data modified within the
** current transaction are either expelled from the cache or reverted to
** their pre-transaction state by re-reading data from the database or
** WAL files. The WAL transaction is then closed.
*/
int sqlite3PagerRollback(Pager *pPager){
  int rc = SQLITE_OK;                  /* Return code */
  PAGERTRACE(("ROLLBACK %d\n", PAGERID(pPager)));

  /* PagerRollback() is a no-op if called in READER or OPEN state. If
  ** the pager is already in the ERROR state, the rollback is not
  ** attempted here. Instead, the error code is returned to the caller.
  */
  assert( assert_pager_state(pPager) );
  if( pPager->eState==PAGER_ERROR ) return pPager->errCode;
  if( pPager->eState<=PAGER_READER ) return SQLITE_OK;

  if( pagerUseWal(pPager) ){
    int rc2;
    rc = sqlite3PagerSavepoint(pPager, SAVEPOINT_ROLLBACK, -1);
    rc2 = pager_end_transaction(pPager, pPager->setSuper, 0);
    if( rc==SQLITE_OK ) rc = rc2;
  }else if( !isOpen(pPager->jfd) || pPager->eState==PAGER_WRITER_LOCKED ){
    int eState = pPager->eState;
    rc = pager_end_transaction(pPager, 0, 0);
    if( !MEMDB && eState>PAGER_WRITER_LOCKED ){
      /* This can happen using journal_mode=off. Move the pager to the error
      ** state to indicate that the contents of the cache may not be trusted.
      ** Any active readers will get SQLITE_ABORT.
      */
      pPager->errCode = SQLITE_ABORT;
      pPager->eState = PAGER_ERROR;
      setGetterMethod(pPager);
      return rc;
    }
  }else{
    rc = pager_playback(pPager, 0);
  }

  assert( pPager->eState==PAGER_READER || rc!=SQLITE_OK );
  assert( rc==SQLITE_OK || rc==SQLITE_FULL || rc==SQLITE_CORRUPT
          || rc==SQLITE_NOMEM || (rc&0xFF)==SQLITE_IOERR
          || rc==SQLITE_CANTOPEN
  );

  /* If an error occurs during a ROLLBACK, we can no longer trust the pager
  ** cache. So call pager_error() on the way out to make any error persistent.
  */
  return pager_error(pPager, rc);
}

/*
** Return TRUE if the database file is opened read-only.  Return FALSE
** if the database is (in theory) writable.
*/
u8 sqlite3PagerIsreadonly(Pager *pPager){
  return pPager->readOnly;
}

#ifdef SQLITE_DEBUG
/*
** Return the sum of the reference counts for all pages held by pPager.
*/
int sqlite3PagerRefcount(Pager *pPager){
  return sqlite3PcacheRefCount(pPager->pPCache);
}
#endif

/*
** Return the approximate number of bytes of memory currently
** used by the pager and its associated cache.
*/
int sqlite3PagerMemUsed(Pager *pPager){
  int perPageSize = pPager->pageSize + pPager->nExtra
    + (int)(sizeof(PgHdr) + 5*sizeof(void*));
  return perPageSize*sqlite3PcachePagecount(pPager->pPCache)
           + sqlite3MallocSize(pPager)
           + pPager->pageSize;
}

/*
** Return the number of references to the specified page.
*/
int sqlite3PagerPageRefcount(DbPage *pPage){
  return sqlite3PcachePageRefcount(pPage);
}

#ifdef SQLITE_TEST
/*
** This routine is used for testing and analysis only.
*/
int *sqlite3PagerStats(Pager *pPager){
  static int a[11];
  a[0] = sqlite3PcacheRefCount(pPager->pPCache);
  a[1] = sqlite3PcachePagecount(pPager->pPCache);
  a[2] = sqlite3PcacheGetCachesize(pPager->pPCache);
  a[3] = pPager->eState==PAGER_OPEN ? -1 : (int) pPager->dbSize;
  a[4] = pPager->eState;
  a[5] = pPager->errCode;
  a[6] = (int)pPager->aStat[PAGER_STAT_HIT] & 0x7fffffff;
  a[7] = (int)pPager->aStat[PAGER_STAT_MISS] & 0x7fffffff;
  a[8] = 0;  /* Used to be pPager->nOvfl */
  a[9] = pPager->nRead;
  a[10] = (int)pPager->aStat[PAGER_STAT_WRITE] & 0x7fffffff;
  return a;
}
#endif

/*
** Parameter eStat must be one of SQLITE_DBSTATUS_CACHE_HIT, _MISS, _WRITE,
** or _WRITE+1.  The SQLITE_DBSTATUS_CACHE_WRITE+1 case is a translation
** of SQLITE_DBSTATUS_CACHE_SPILL.  The _SPILL case is not contiguous because
** it was added later.
**
** Before returning, *pnVal is incremented by the
** current cache hit or miss count, according to the value of eStat. If the
** reset parameter is non-zero, the cache hit or miss count is zeroed before
** returning.
*/
void sqlite3PagerCacheStat(Pager *pPager, int eStat, int reset, u64 *pnVal){

  assert( eStat==SQLITE_DBSTATUS_CACHE_HIT
       || eStat==SQLITE_DBSTATUS_CACHE_MISS
       || eStat==SQLITE_DBSTATUS_CACHE_WRITE
       || eStat==SQLITE_DBSTATUS_CACHE_WRITE+1
  );

  assert( SQLITE_DBSTATUS_CACHE_HIT+1==SQLITE_DBSTATUS_CACHE_MISS );
  assert( SQLITE_DBSTATUS_CACHE_HIT+2==SQLITE_DBSTATUS_CACHE_WRITE );
  assert( PAGER_STAT_HIT==0 && PAGER_STAT_MISS==1
           && PAGER_STAT_WRITE==2 && PAGER_STAT_SPILL==3 );

  eStat -= SQLITE_DBSTATUS_CACHE_HIT;
  *pnVal += pPager->aStat[eStat];
  if( reset ){
    pPager->aStat[eStat] = 0;
  }
}

/*
** Return true if this is an in-memory or temp-file backed pager.
*/
int sqlite3PagerIsMemdb(Pager *pPager){
  return pPager->tempFile || pPager->memVfs;
}

/*
** Check that there are at least nSavepoint savepoints open. If there are
** currently less than nSavepoints open, then open one or more savepoints
** to make up the difference. If the number of savepoints is already
** equal to nSavepoint, then this function is a no-op.
**
** If a memory allocation fails, SQLITE_NOMEM is returned. If an error
** occurs while opening the sub-journal file, then an IO error code is
** returned. Otherwise, SQLITE_OK.
*/
static SQLITE_NOINLINE int pagerOpenSavepoint(Pager *pPager, int nSavepoint){
  int rc = SQLITE_OK;                       /* Return code */
  int nCurrent = pPager->nSavepoint;        /* Current number of savepoints */
  int ii;                                   /* Iterator variable */
  PagerSavepoint *aNew;                     /* New Pager.aSavepoint array */

  assert( pPager->eState>=PAGER_WRITER_LOCKED );
  assert( assert_pager_state(pPager) );
  assert( nSavepoint>nCurrent && pPager->useJournal );

  /* Grow the Pager.aSavepoint array using realloc(). Return SQLITE_NOMEM
  ** if the allocation fails. Otherwise, zero the new portion in case a
  ** malloc failure occurs while populating it in the for(...) loop below.
  */
  aNew = (PagerSavepoint *)sqlite3Realloc(
      pPager->aSavepoint, sizeof(PagerSavepoint)*nSavepoint
  );
  if( !aNew ){
    return SQLITE_NOMEM_BKPT;
  }
  memset(&aNew[nCurrent], 0, (nSavepoint-nCurrent) * sizeof(PagerSavepoint));
  pPager->aSavepoint = aNew;

  /* Populate the PagerSavepoint structures just allocated. */
  for(ii=nCurrent; ii<nSavepoint; ii++){
    aNew[ii].nOrig = pPager->dbSize;
    if( isOpen(pPager->jfd) && pPager->journalOff>0 ){
      aNew[ii].iOffset = pPager->journalOff;
    }else{
      aNew[ii].iOffset = JOURNAL_HDR_SZ(pPager);
    }
    aNew[ii].iSubRec = pPager->nSubRec;
    aNew[ii].pInSavepoint = sqlite3BitvecCreate(pPager->dbSize);
    aNew[ii].bTruncateOnRelease = 1;
    if( !aNew[ii].pInSavepoint ){
      return SQLITE_NOMEM_BKPT;
    }
    if( pagerUseWal(pPager) ){
      sqlite3WalSavepoint(pPager->pWal, aNew[ii].aWalData);
    }
    pPager->nSavepoint = ii+1;
  }
  assert( pPager->nSavepoint==nSavepoint );
  assertTruncateConstraint(pPager);
  return rc;
}
int sqlite3PagerOpenSavepoint(Pager *pPager, int nSavepoint){
  assert( pPager->eState>=PAGER_WRITER_LOCKED );
  assert( assert_pager_state(pPager) );

  if( nSavepoint>pPager->nSavepoint && pPager->useJournal ){
    return pagerOpenSavepoint(pPager, nSavepoint);
  }else{
    return SQLITE_OK;
  }
}


/*
** This function is called to rollback or release (commit) a savepoint.
** The savepoint to release or rollback need not be the most recently
** created savepoint.
**
** Parameter op is always either SAVEPOINT_ROLLBACK or SAVEPOINT_RELEASE.
** If it is SAVEPOINT_RELEASE, then release and destroy the savepoint with
** index iSavepoint. If it is SAVEPOINT_ROLLBACK, then rollback all changes
** that have occurred since the specified savepoint was created.
**
** The savepoint to rollback or release is identified by parameter
** iSavepoint. A value of 0 means to operate on the outermost savepoint
** (the first created). A value of (Pager.nSavepoint-1) means operate
** on the most recently created savepoint. If iSavepoint is greater than
** (Pager.nSavepoint-1), then this function is a no-op.
**
** If a negative value is passed to this function, then the current
** transaction is rolled back. This is different to calling
** sqlite3PagerRollback() because this function does not terminate
** the transaction or unlock the database, it just restores the
** contents of the database to its original state.
**
** In any case, all savepoints with an index greater than iSavepoint
** are destroyed. If this is a release operation (op==SAVEPOINT_RELEASE),
** then savepoint iSavepoint is also destroyed.
**
** This function may return SQLITE_NOMEM if a memory allocation fails,
** or an IO error code if an IO error occurs while rolling back a
** savepoint. If no errors occur, SQLITE_OK is returned.
*/
int sqlite3PagerSavepoint(Pager *pPager, int op, int iSavepoint){
  int rc = pPager->errCode;
 
#ifdef SQLITE_ENABLE_ZIPVFS
  if( op==SAVEPOINT_RELEASE ) rc = SQLITE_OK;
#endif

  assert( op==SAVEPOINT_RELEASE || op==SAVEPOINT_ROLLBACK );
  assert( iSavepoint>=0 || op==SAVEPOINT_ROLLBACK );

  if( rc==SQLITE_OK && iSavepoint<pPager->nSavepoint ){
    int ii;            /* Iterator variable */
    int nNew;          /* Number of remaining savepoints after this op. */

    /* Figure out how many savepoints will still be active after this
    ** operation. Store this value in nNew. Then free resources associated
    ** with any savepoints that are destroyed by this operation.
    */
    nNew = iSavepoint + (( op==SAVEPOINT_RELEASE ) ? 0 : 1);
    for(ii=nNew; ii<pPager->nSavepoint; ii++){
      sqlite3BitvecDestroy(pPager->aSavepoint[ii].pInSavepoint);
    }
    pPager->nSavepoint = nNew;

    /* Truncate the sub-journal so that it only includes the parts
    ** that are still in use. */
    if( op==SAVEPOINT_RELEASE ){
      PagerSavepoint *pRel = &pPager->aSavepoint[nNew];
      if( pRel->bTruncateOnRelease && isOpen(pPager->sjfd) ){
        /* Only truncate if it is an in-memory sub-journal. */
        if( sqlite3JournalIsInMemory(pPager->sjfd) ){
          i64 sz = (pPager->pageSize+4)*(i64)pRel->iSubRec;
          rc = sqlite3OsTruncate(pPager->sjfd, sz);
          assert( rc==SQLITE_OK );
        }
        pPager->nSubRec = pRel->iSubRec;
      }
    }
    /* Else this is a rollback operation, playback the specified savepoint.
    ** If this is a temp-file, it is possible that the journal file has
    ** not yet been opened. In this case there have been no changes to
    ** the database file, so the playback operation can be skipped.
    */
    else if( pagerUseWal(pPager) || isOpen(pPager->jfd) ){
      PagerSavepoint *pSavepoint = (nNew==0)?0:&pPager->aSavepoint[nNew-1];
      rc = pagerPlaybackSavepoint(pPager, pSavepoint);
      assert(rc!=SQLITE_DONE);
    }
   
#ifdef SQLITE_ENABLE_ZIPVFS
    /* If the cache has been modified but the savepoint cannot be rolled
    ** back journal_mode=off, put the pager in the error state. This way,
    ** if the VFS used by this pager includes ZipVFS, the entire transaction
    ** can be rolled back at the ZipVFS level.  */
    else if(
        pPager->journalMode==PAGER_JOURNALMODE_OFF
     && pPager->eState>=PAGER_WRITER_CACHEMOD
    ){
      pPager->errCode = SQLITE_ABORT;
      pPager->eState = PAGER_ERROR;
      setGetterMethod(pPager);
    }
#endif
  }

  return rc;
}

/*
** Return the full pathname of the database file.
**
** Except, if the pager is in-memory only, then return an empty string if
** nullIfMemDb is true.  This routine is called with nullIfMemDb==1 when
** used to report the filename to the user, for compatibility with legacy
** behavior.  But when the Btree needs to know the filename for matching to
** shared cache, it uses nullIfMemDb==0 so that in-memory databases can
** participate in shared-cache.
**
** The return value to this routine is always safe to use with
** sqlite3_uri_parameter() and sqlite3_filename_database() and friends.
*/
const char *sqlite3PagerFilename(const Pager *pPager, int nullIfMemDb){
  static const char zFake[8] = { 0, 0, 0, 0, 0, 0, 0, 0 };
  if( nullIfMemDb && (pPager->memDb || sqlite3IsMemdb(pPager->pVfs)) ){
    return &zFake[4];
  }else{
    return pPager->zFilename;
  }
}

/*
** Return the VFS structure for the pager.
*/
sqlite3_vfs *sqlite3PagerVfs(Pager *pPager){
  return pPager->pVfs;
}

/*
** Return the file handle for the database file associated
** with the pager.  This might return NULL if the file has
** not yet been opened.
*/
sqlite3_file *sqlite3PagerFile(Pager *pPager){
  return pPager->fd;
}

/*
** Return the file handle for the journal file (if it exists).
** This will be either the rollback journal or the WAL file.
*/
sqlite3_file *sqlite3PagerJrnlFile(Pager *pPager){
#if SQLITE_OMIT_WAL
  return pPager->jfd;
#else
  return pPager->pWal ? sqlite3WalFile(pPager->pWal) : pPager->jfd;
#endif
}

/*
** Return the full pathname of the journal file.
*/
const char *sqlite3PagerJournalname(Pager *pPager){
  return pPager->zJournal;
}

#ifndef SQLITE_OMIT_AUTOVACUUM
/*
** Move the page pPg to location pgno in the file.
**
** There must be no references to the page previously located at
** pgno (which we call pPgOld) though that page is allowed to be
** in cache.  If the page previously located at pgno is not already
** in the rollback journal, it is not put there by by this routine.
**
** References to the page pPg remain valid. Updating any
** meta-data associated with pPg (i.e. data stored in the nExtra bytes
** allocated along with the page) is the responsibility of the caller.
**
** A transaction must be active when this routine is called. It used to be
** required that a statement transaction was not active, but this restriction
** has been removed (CREATE INDEX needs to move a page when a statement
** transaction is active).
**
** If the fourth argument, isCommit, is non-zero, then this page is being
** moved as part of a database reorganization just before the transaction
** is being committed. In this case, it is guaranteed that the database page
** pPg refers to will not be written to again within this transaction.
**
** This function may return SQLITE_NOMEM or an IO error code if an error
** occurs. Otherwise, it returns SQLITE_OK.
*/
int sqlite3PagerMovepage(Pager *pPager, DbPage *pPg, Pgno pgno, int isCommit){
  PgHdr *pPgOld;               /* The page being overwritten. */
  Pgno needSyncPgno = 0;       /* Old value of pPg->pgno, if sync is required */
  int rc;                      /* Return code */
  Pgno origPgno;               /* The original page number */

  assert( pPg->nRef>0 );
  assert( pPager->eState==PAGER_WRITER_CACHEMOD
       || pPager->eState==PAGER_WRITER_DBMOD
  );
  assert( assert_pager_state(pPager) );

  /* In order to be able to rollback, an in-memory database must journal
  ** the page we are moving from.
  */
  assert( pPager->tempFile || !MEMDB );
  if( pPager->tempFile ){
    rc = sqlite3PagerWrite(pPg);
    if( rc ) return rc;
  }

  /* If the page being moved is dirty and has not been saved by the latest
  ** savepoint, then save the current contents of the page into the
  ** sub-journal now. This is required to handle the following scenario:
  **
  **   BEGIN;
  **     <journal page X, then modify it in memory>
  **     SAVEPOINT one;
  **       <Move page X to location Y>
  **     ROLLBACK TO one;
  **
  ** If page X were not written to the sub-journal here, it would not
  ** be possible to restore its contents when the "ROLLBACK TO one"
  ** statement were is processed.
  **
  ** subjournalPage() may need to allocate space to store pPg->pgno into
  ** one or more savepoint bitvecs. This is the reason this function
  ** may return SQLITE_NOMEM.
  */
  if( (pPg->flags & PGHDR_DIRTY)!=0
   && SQLITE_OK!=(rc = subjournalPageIfRequired(pPg))
  ){
    return rc;
  }

  PAGERTRACE(("MOVE %d page %d (needSync=%d) moves to %d\n",
      PAGERID(pPager), pPg->pgno, (pPg->flags&PGHDR_NEED_SYNC)?1:0, pgno));
  IOTRACE(("MOVE %p %d %d\n", pPager, pPg->pgno, pgno))

  /* If the journal needs to be sync()ed before page pPg->pgno can
  ** be written to, store pPg->pgno in local variable needSyncPgno.
  **
  ** If the isCommit flag is set, there is no need to remember that
  ** the journal needs to be sync()ed before database page pPg->pgno
  ** can be written to. The caller has already promised not to write to it.
  */
  if( (pPg->flags&PGHDR_NEED_SYNC) && !isCommit ){
    needSyncPgno = pPg->pgno;
    assert( pPager->journalMode==PAGER_JOURNALMODE_OFF ||
            pageInJournal(pPager, pPg) || pPg->pgno>pPager->dbOrigSize );
    assert( pPg->flags&PGHDR_DIRTY );
  }

  /* If the cache contains a page with page-number pgno, remove it
  ** from its hash chain. Also, if the PGHDR_NEED_SYNC flag was set for
  ** page pgno before the 'move' operation, it needs to be retained
  ** for the page moved there.
  */
  pPg->flags &= ~PGHDR_NEED_SYNC;
  pPgOld = sqlite3PagerLookup(pPager, pgno);
  assert( !pPgOld || pPgOld->nRef==1 || CORRUPT_DB );
  if( pPgOld ){
    if( NEVER(pPgOld->nRef>1) ){
      sqlite3PagerUnrefNotNull(pPgOld);
      return SQLITE_CORRUPT_BKPT;
    }
    pPg->flags |= (pPgOld->flags&PGHDR_NEED_SYNC);
    if( pPager->tempFile ){
      /* Do not discard pages from an in-memory database since we might
      ** need to rollback later.  Just move the page out of the way. */
      sqlite3PcacheMove(pPgOld, pPager->dbSize+1);
    }else{
      sqlite3PcacheDrop(pPgOld);
    }
  }

  origPgno = pPg->pgno;
  sqlite3PcacheMove(pPg, pgno);
  sqlite3PcacheMakeDirty(pPg);

  /* For an in-memory database, make sure the original page continues
  ** to exist, in case the transaction needs to roll back.  Use pPgOld
  ** as the original page since it has already been allocated.
  */
  if( pPager->tempFile && pPgOld ){
    sqlite3PcacheMove(pPgOld, origPgno);
    sqlite3PagerUnrefNotNull(pPgOld);
  }

  if( needSyncPgno ){
    /* If needSyncPgno is non-zero, then the journal file needs to be
    ** sync()ed before any data is written to database file page needSyncPgno.
    ** Currently, no such page exists in the page-cache and the
    ** "is journaled" bitvec flag has been set. This needs to be remedied by
    ** loading the page into the pager-cache and setting the PGHDR_NEED_SYNC
    ** flag.
    **
    ** If the attempt to load the page into the page-cache fails, (due
    ** to a malloc() or IO failure), clear the bit in the pInJournal[]
    ** array. Otherwise, if the page is loaded and written again in
    ** this transaction, it may be written to the database file before
    ** it is synced into the journal file. This way, it may end up in
    ** the journal file twice, but that is not a problem.
    */
    PgHdr *pPgHdr;
    rc = sqlite3PagerGet(pPager, needSyncPgno, &pPgHdr, 0);
    if( rc!=SQLITE_OK ){
      if( needSyncPgno<=pPager->dbOrigSize ){
        assert( pPager->pTmpSpace!=0 );
        sqlite3BitvecClear(pPager->pInJournal, needSyncPgno, pPager->pTmpSpace);
      }
      return rc;
    }
    pPgHdr->flags |= PGHDR_NEED_SYNC;
    sqlite3PcacheMakeDirty(pPgHdr);
    sqlite3PagerUnrefNotNull(pPgHdr);
  }

  return SQLITE_OK;
}
#endif

/*
** The page handle passed as the first argument refers to a dirty page
** with a page number other than iNew. This function changes the page's
** page number to iNew and sets the value of the PgHdr.flags field to
** the value passed as the third parameter.
*/
void sqlite3PagerRekey(DbPage *pPg, Pgno iNew, u16 flags){
  assert( pPg->pgno!=iNew );
  pPg->flags = flags;
  sqlite3PcacheMove(pPg, iNew);
}

/*
** Return a pointer to the data for the specified page.
*/
void *sqlite3PagerGetData(DbPage *pPg){
  assert( pPg->nRef>0 || pPg->pPager->memDb );
  return pPg->pData;
}

/*
** Return a pointer to the Pager.nExtra bytes of "extra" space
** allocated along with the specified page.
*/
void *sqlite3PagerGetExtra(DbPage *pPg){
  return pPg->pExtra;
}

/*
** Get/set the locking-mode for this pager. Parameter eMode must be one
** of PAGER_LOCKINGMODE_QUERY, PAGER_LOCKINGMODE_NORMAL or
** PAGER_LOCKINGMODE_EXCLUSIVE. If the parameter is not _QUERY, then
** the locking-mode is set to the value specified.
**
** The returned value is either PAGER_LOCKINGMODE_NORMAL or
** PAGER_LOCKINGMODE_EXCLUSIVE, indicating the current (possibly updated)
** locking-mode.
*/
int sqlite3PagerLockingMode(Pager *pPager, int eMode){
  assert( eMode==PAGER_LOCKINGMODE_QUERY
            || eMode==PAGER_LOCKINGMODE_NORMAL
            || eMode==PAGER_LOCKINGMODE_EXCLUSIVE );
  assert( PAGER_LOCKINGMODE_QUERY<0 );
  assert( PAGER_LOCKINGMODE_NORMAL>=0 && PAGER_LOCKINGMODE_EXCLUSIVE>=0 );
  assert( pPager->exclusiveMode || 0==sqlite3WalHeapMemory(pPager->pWal) );
  if( eMode>=0 && !pPager->tempFile && !sqlite3WalHeapMemory(pPager->pWal) ){
    pPager->exclusiveMode = (u8)eMode;
  }
  return (int)pPager->exclusiveMode;
}

/*
** Set the journal-mode for this pager. Parameter eMode must be one of:
**
**    PAGER_JOURNALMODE_DELETE
**    PAGER_JOURNALMODE_TRUNCATE
**    PAGER_JOURNALMODE_PERSIST
**    PAGER_JOURNALMODE_OFF
**    PAGER_JOURNALMODE_MEMORY
**    PAGER_JOURNALMODE_WAL
**
** The journalmode is set to the value specified if the change is allowed.
** The change may be disallowed for the following reasons:
**
**   *  An in-memory database can only have its journal_mode set to _OFF
**      or _MEMORY.
**
**   *  Temporary databases cannot have _WAL journalmode.
**
** The returned indicate the current (possibly updated) journal-mode.
*/
int sqlite3PagerSetJournalMode(Pager *pPager, int eMode){
  u8 eOld = pPager->journalMode;    /* Prior journalmode */

  /* The eMode parameter is always valid */
  assert(      eMode==PAGER_JOURNALMODE_DELETE    /* 0 */
            || eMode==PAGER_JOURNALMODE_PERSIST   /* 1 */
            || eMode==PAGER_JOURNALMODE_OFF       /* 2 */
            || eMode==PAGER_JOURNALMODE_TRUNCATE  /* 3 */
            || eMode==PAGER_JOURNALMODE_MEMORY    /* 4 */
            || eMode==PAGER_JOURNALMODE_WAL       /* 5 */ );

  /* This routine is only called from the OP_JournalMode opcode, and
  ** the logic there will never allow a temporary file to be changed
  ** to WAL mode.
  */
  assert( pPager->tempFile==0 || eMode!=PAGER_JOURNALMODE_WAL );

  /* Do allow the journalmode of an in-memory database to be set to
  ** anything other than MEMORY or OFF
  */
  if( MEMDB ){
    assert( eOld==PAGER_JOURNALMODE_MEMORY || eOld==PAGER_JOURNALMODE_OFF );
    if( eMode!=PAGER_JOURNALMODE_MEMORY && eMode!=PAGER_JOURNALMODE_OFF ){
      eMode = eOld;
    }
  }

  if( eMode!=eOld ){

    /* Change the journal mode. */
    assert( pPager->eState!=PAGER_ERROR );
    pPager->journalMode = (u8)eMode;

    /* When transitioning from TRUNCATE or PERSIST to any other journal
    ** mode except WAL, unless the pager is in locking_mode=exclusive mode,
    ** delete the journal file.
    */
    assert( (PAGER_JOURNALMODE_TRUNCATE & 5)==1 );
    assert( (PAGER_JOURNALMODE_PERSIST & 5)==1 );
    assert( (PAGER_JOURNALMODE_DELETE & 5)==0 );
    assert( (PAGER_JOURNALMODE_MEMORY & 5)==4 );
    assert( (PAGER_JOURNALMODE_OFF & 5)==0 );
    assert( (PAGER_JOURNALMODE_WAL & 5)==5 );

    assert( isOpen(pPager->fd) || pPager->exclusiveMode );
    if( !pPager->exclusiveMode && (eOld & 5)==1 && (eMode & 1)==0 ){
      /* In this case we would like to delete the journal file. If it is
      ** not possible, then that is not a problem. Deleting the journal file
      ** here is an optimization only.
      **
      ** Before deleting the journal file, obtain a RESERVED lock on the
      ** database file. This ensures that the journal file is not deleted
      ** while it is in use by some other client.
      */
      sqlite3OsClose(pPager->jfd);
      if( pPager->eLock>=RESERVED_LOCK ){
        sqlite3OsDelete(pPager->pVfs, pPager->zJournal, 0);
      }else{
        int rc = SQLITE_OK;
        int state = pPager->eState;
        assert( state==PAGER_OPEN || state==PAGER_READER );
        if( state==PAGER_OPEN ){
          rc = sqlite3PagerSharedLock(pPager);
        }
        if( pPager->eState==PAGER_READER ){
          assert( rc==SQLITE_OK );
          rc = pagerLockDb(pPager, RESERVED_LOCK);
        }
        if( rc==SQLITE_OK ){
          sqlite3OsDelete(pPager->pVfs, pPager->zJournal, 0);
        }
        if( rc==SQLITE_OK && state==PAGER_READER ){
          pagerUnlockDb(pPager, SHARED_LOCK);
        }else if( state==PAGER_OPEN ){
          pager_unlock(pPager);
        }
        assert( state==pPager->eState );
      }
    }else if( eMode==PAGER_JOURNALMODE_OFF || eMode==PAGER_JOURNALMODE_MEMORY ){
      sqlite3OsClose(pPager->jfd);
    }
  }

  /* Return the new journal mode */
  return (int)pPager->journalMode;
}

/*
** Return the current journal mode.
*/
int sqlite3PagerGetJournalMode(Pager *pPager){
  return (int)pPager->journalMode;
}

/*
** Return TRUE if the pager is in a state where it is OK to change the
** journalmode.  Journalmode changes can only happen when the database
** is unmodified.
*/
int sqlite3PagerOkToChangeJournalMode(Pager *pPager){
  assert( assert_pager_state(pPager) );
  if( pPager->eState>=PAGER_WRITER_CACHEMOD ) return 0;
  if( NEVER(isOpen(pPager->jfd) && pPager->journalOff>0) ) return 0;
  return 1;
}

/*
** Get/set the size-limit used for persistent journal files.
**
** Setting the size limit to -1 means no limit is enforced.
** An attempt to set a limit smaller than -1 is a no-op.
*/
i64 sqlite3PagerJournalSizeLimit(Pager *pPager, i64 iLimit){
  if( iLimit>=-1 ){
    pPager->journalSizeLimit = iLimit;
    sqlite3WalLimit(pPager->pWal, iLimit);
  }
  return pPager->journalSizeLimit;
}

/*
** Return a pointer to the pPager->pBackup variable. The backup module
** in backup.c maintains the content of this variable. This module
** uses it opaquely as an argument to sqlite3BackupRestart() and
** sqlite3BackupUpdate() only.
*/
sqlite3_backup **sqlite3PagerBackupPtr(Pager *pPager){
  return &pPager->pBackup;
}

#ifndef SQLITE_OMIT_VACUUM
/*
** Unless this is an in-memory or temporary database, clear the pager cache.
*/
void sqlite3PagerClearCache(Pager *pPager){
  assert( MEMDB==0 || pPager->tempFile );
  if( pPager->tempFile==0 ) pager_reset(pPager);
}
#endif


#ifndef SQLITE_OMIT_WAL
/*
** This function is called when the user invokes "PRAGMA wal_checkpoint",
** "PRAGMA wal_blocking_checkpoint" or calls the sqlite3_wal_checkpoint()
** or wal_blocking_checkpoint() API functions.
**
** Parameter eMode is one of SQLITE_CHECKPOINT_PASSIVE, FULL or RESTART.
*/
int sqlite3PagerCheckpoint(
  Pager *pPager,                  /* Checkpoint on this pager */
  sqlite3 *db,                    /* Db handle used to check for interrupts */
  int eMode,                      /* Type of checkpoint */
  int *pnLog,                     /* OUT: Final number of frames in log */
  int *pnCkpt                     /* OUT: Final number of checkpointed frames */
){
  int rc = SQLITE_OK;
  if( pPager->pWal==0 && pPager->journalMode==PAGER_JOURNALMODE_WAL ){
    /* This only happens when a database file is zero bytes in size opened and
    ** then "PRAGMA journal_mode=WAL" is run and then sqlite3_wal_checkpoint()
    ** is invoked without any intervening transactions.  We need to start
    ** a transaction to initialize pWal.  The PRAGMA table_list statement is
    ** used for this since it starts transactions on every database file,
    ** including all ATTACHed databases.  This seems expensive for a single
    ** sqlite3_wal_checkpoint() call, but it happens very rarely.
    ** https://sqlite.org/forum/forumpost/fd0f19d229156939
    */
    sqlite3_exec(db, "PRAGMA table_list",0,0,0);
  }
  if( pPager->pWal ){
    rc = sqlite3WalCheckpoint(pPager->pWal, db, eMode,
        (eMode==SQLITE_CHECKPOINT_PASSIVE ? 0 : pPager->xBusyHandler),
        pPager->pBusyHandlerArg,
        pPager->walSyncFlags, pPager->pageSize, (u8 *)pPager->pTmpSpace,
        pnLog, pnCkpt
    );
  }
  return rc;
}

int sqlite3PagerWalCallback(Pager *pPager){
  return sqlite3WalCallback(pPager->pWal);
}

/*
** Return true if the underlying VFS for the given pager supports the
** primitives necessary for write-ahead logging.
*/
int sqlite3PagerWalSupported(Pager *pPager){
  const sqlite3_io_methods *pMethods = pPager->fd->pMethods;
  if( pPager->noLock ) return 0;
  return pPager->exclusiveMode || (pMethods->iVersion>=2 && pMethods->xShmMap);
}

/*
** Attempt to take an exclusive lock on the database file. If a PENDING lock
** is obtained instead, immediately release it.
*/
static int pagerExclusiveLock(Pager *pPager){
  int rc;                         /* Return code */
  u8 eOrigLock;                   /* Original lock */

  assert( pPager->eLock>=SHARED_LOCK );
  eOrigLock = pPager->eLock;
  rc = pagerLockDb(pPager, EXCLUSIVE_LOCK);
  if( rc!=SQLITE_OK ){
    /* If the attempt to grab the exclusive lock failed, release the
    ** pending lock that may have been obtained instead.  */
    pagerUnlockDb(pPager, eOrigLock);
  }

  return rc;
}

/*
** Call sqlite3WalOpen() to open the WAL handle. If the pager is in
** exclusive-locking mode when this function is called, take an EXCLUSIVE
** lock on the database file and use heap-memory to store the wal-index
** in. Otherwise, use the normal shared-memory.
*/
static int pagerOpenWal(Pager *pPager){
  int rc = SQLITE_OK;

  assert( pPager->pWal==0 && pPager->tempFile==0 );
  assert( pPager->eLock==SHARED_LOCK || pPager->eLock==EXCLUSIVE_LOCK );

  /* If the pager is already in exclusive-mode, the WAL module will use
  ** heap-memory for the wal-index instead of the VFS shared-memory
  ** implementation. Take the exclusive lock now, before opening the WAL
  ** file, to make sure this is safe.
  */
  if( pPager->exclusiveMode ){
    rc = pagerExclusiveLock(pPager);
  }

  /* Open the connection to the log file. If this operation fails,
  ** (e.g. due to malloc() failure), return an error code.
  */
  if( rc==SQLITE_OK ){
    rc = sqlite3WalOpen(pPager->pVfs,
        pPager->fd, pPager->zWal, pPager->exclusiveMode,
        pPager->journalSizeLimit, &pPager->pWal
    );
  }
  pagerFixMaplimit(pPager);

  return rc;
}


/*
** The caller must be holding a SHARED lock on the database file to call
** this function.
**
** If the pager passed as the first argument is open on a real database
** file (not a temp file or an in-memory database), and the WAL file
** is not already open, make an attempt to open it now. If successful,
** return SQLITE_OK. If an error occurs or the VFS used by the pager does
** not support the xShmXXX() methods, return an error code. *pbOpen is
** not modified in either case.
**
** If the pager is open on a temp-file (or in-memory database), or if
** the WAL file is already open, set *pbOpen to 1 and return SQLITE_OK
** without doing anything.
*/
int sqlite3PagerOpenWal(
  Pager *pPager,                  /* Pager object */
  int *pbOpen                     /* OUT: Set to true if call is a no-op */
){
  int rc = SQLITE_OK;             /* Return code */

  assert( assert_pager_state(pPager) );
  assert( pPager->eState==PAGER_OPEN   || pbOpen );
  assert( pPager->eState==PAGER_READER || !pbOpen );
  assert( pbOpen==0 || *pbOpen==0 );
  assert( pbOpen!=0 || (!pPager->tempFile && !pPager->pWal) );

  if( !pPager->tempFile && !pPager->pWal ){
    if( !sqlite3PagerWalSupported(pPager) ) return SQLITE_CANTOPEN;

    /* Close any rollback journal previously open */
    sqlite3OsClose(pPager->jfd);

    rc = pagerOpenWal(pPager);
    if( rc==SQLITE_OK ){
      pPager->journalMode = PAGER_JOURNALMODE_WAL;
      pPager->eState = PAGER_OPEN;
    }
  }else{
    *pbOpen = 1;
  }

  return rc;
}

/*
** This function is called to close the connection to the log file prior
** to switching from WAL to rollback mode.
**
** Before closing the log file, this function attempts to take an
** EXCLUSIVE lock on the database file. If this cannot be obtained, an
** error (SQLITE_BUSY) is returned and the log connection is not closed.
** If successful, the EXCLUSIVE lock is not released before returning.
*/
int sqlite3PagerCloseWal(Pager *pPager, sqlite3 *db){
  int rc = SQLITE_OK;

  assert( pPager->journalMode==PAGER_JOURNALMODE_WAL );

  /* If the log file is not already open, but does exist in the file-system,
  ** it may need to be checkpointed before the connection can switch to
  ** rollback mode. Open it now so this can happen.
  */
  if( !pPager->pWal ){
    int logexists = 0;
    rc = pagerLockDb(pPager, SHARED_LOCK);
    if( rc==SQLITE_OK ){
      rc = sqlite3OsAccess(
          pPager->pVfs, pPager->zWal, SQLITE_ACCESS_EXISTS, &logexists
      );
    }
    if( rc==SQLITE_OK && logexists ){
      rc = pagerOpenWal(pPager);
    }
  }
   
  /* Checkpoint and close the log. Because an EXCLUSIVE lock is held on
  ** the database file, the log and log-summary files will be deleted.
  */
  if( rc==SQLITE_OK && pPager->pWal ){
    rc = pagerExclusiveLock(pPager);
    if( rc==SQLITE_OK ){
      rc = sqlite3WalClose(pPager->pWal, db, pPager->walSyncFlags,
                           pPager->pageSize, (u8*)pPager->pTmpSpace);
      pPager->pWal = 0;
      pagerFixMaplimit(pPager);
      if( rc && !pPager->exclusiveMode ) pagerUnlockDb(pPager, SHARED_LOCK);
    }
  }
  return rc;
}

#ifdef SQLITE_ENABLE_SETLK_TIMEOUT
/*
** If pager pPager is a wal-mode database not in exclusive locking mode,
** invoke the sqlite3WalWriteLock() function on the associated Wal object
** with the same db and bLock parameters as were passed to this function.
** Return an SQLite error code if an error occurs, or SQLITE_OK otherwise.
*/
int sqlite3PagerWalWriteLock(Pager *pPager, int bLock){
  int rc = SQLITE_OK;
  if( pagerUseWal(pPager) && pPager->exclusiveMode==0 ){
    rc = sqlite3WalWriteLock(pPager->pWal, bLock);
  }
  return rc;
}

/*
** Set the database handle used by the wal layer to determine if
** blocking locks are required.
*/
void sqlite3PagerWalDb(Pager *pPager, sqlite3 *db){
  if( pagerUseWal(pPager) ){
    sqlite3WalDb(pPager->pWal, db);
  }
}
#endif

#ifdef SQLITE_ENABLE_SNAPSHOT
/*
** If this is a WAL database, obtain a snapshot handle for the snapshot
** currently open. Otherwise, return an error.
*/
int sqlite3PagerSnapshotGet(Pager *pPager, sqlite3_snapshot **ppSnapshot){
  int rc = SQLITE_ERROR;
  if( pPager->pWal ){
    rc = sqlite3WalSnapshotGet(pPager->pWal, ppSnapshot);
  }
  return rc;
}

/*
** If this is a WAL database, store a pointer to pSnapshot. Next time a
** read transaction is opened, attempt to read from the snapshot it
** identifies. If this is not a WAL database, return an error.
*/
int sqlite3PagerSnapshotOpen(
  Pager *pPager,
  sqlite3_snapshot *pSnapshot
){
  int rc = SQLITE_OK;
  if( pPager->pWal ){
    sqlite3WalSnapshotOpen(pPager->pWal, pSnapshot);
  }else{
    rc = SQLITE_ERROR;
  }
  return rc;
}

/*
** If this is a WAL database, call sqlite3WalSnapshotRecover(). If this
** is not a WAL database, return an error.
*/
int sqlite3PagerSnapshotRecover(Pager *pPager){
  int rc;
  if( pPager->pWal ){
    rc = sqlite3WalSnapshotRecover(pPager->pWal);
  }else{
    rc = SQLITE_ERROR;
  }
  return rc;
}

/*
** The caller currently has a read transaction open on the database.
** If this is not a WAL database, SQLITE_ERROR is returned. Otherwise,
** this function takes a SHARED lock on the CHECKPOINTER slot and then
** checks if the snapshot passed as the second argument is still
** available. If so, SQLITE_OK is returned.
**
** If the snapshot is not available, SQLITE_ERROR is returned. Or, if
** the CHECKPOINTER lock cannot be obtained, SQLITE_BUSY. If any error
** occurs (any value other than SQLITE_OK is returned), the CHECKPOINTER
** lock is released before returning.
*/
int sqlite3PagerSnapshotCheck(Pager *pPager, sqlite3_snapshot *pSnapshot){
  int rc;
  if( pPager->pWal ){
    rc = sqlite3WalSnapshotCheck(pPager->pWal, pSnapshot);
  }else{
    rc = SQLITE_ERROR;
  }
  return rc;
}

/*
** Release a lock obtained by an earlier successful call to
** sqlite3PagerSnapshotCheck().
*/
void sqlite3PagerSnapshotUnlock(Pager *pPager){
  assert( pPager->pWal );
  sqlite3WalSnapshotUnlock(pPager->pWal);
}

#endif /* SQLITE_ENABLE_SNAPSHOT */
#endif /* !SQLITE_OMIT_WAL */

#ifdef SQLITE_ENABLE_ZIPVFS
/*
** A read-lock must be held on the pager when this function is called. If
** the pager is in WAL mode and the WAL file currently contains one or more
** frames, return the size in bytes of the page images stored within the
** WAL frames. Otherwise, if this is not a WAL database or the WAL file
** is empty, return 0.
*/
int sqlite3PagerWalFramesize(Pager *pPager){
  assert( pPager->eState>=PAGER_READER );
  return sqlite3WalFramesize(pPager->pWal);
}
#endif

#if defined(SQLITE_USE_SEH) && !defined(SQLITE_OMIT_WAL)
int sqlite3PagerWalSystemErrno(Pager *pPager){
  return sqlite3WalSystemErrno(pPager->pWal);
}
#endif

#endif /* SQLITE_OMIT_DISKIO */