aboutsummaryrefslogblamecommitdiffstats
path: root/fs/xfs/xfs_log_recover.c
blob: 1f0016b0b4ec2a211d3d19dd15740b5058ebb894 (plain) (tree)
1
2
3
4
5
6
7
8
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
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
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
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
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
  
                                                      
  
                                                                 
                                             


                                                                   
  

                                                                       
   
                
                   
                      
                    
                    
                     
                      
                   




                           
                             
                        
                        
                       
                      

                           

                         

                             





















































































                                                                               
          




























                                                                  
                  


























































































































































































































































































































































                                                                                
          



































































                                                                                
                                                                              





































































































































































































                                                                                
                                          









































































































































































































































































































































































































































































































































































































                                                                                          
                                                              






























































































































































































































































































































































































































































































































































































                                                                                                                                                  
                                                                                













































                                                                              
                                                           

                                                                     
                                                                             
                       
                                                  

                                                                       
                                                                    

                       

                                           

                                                                     
                                               

                       
                                                       


                                                               
                                                              


                                 

                                                                    



                                                                    
                                                                               

                                       

                                                                    



                                                                    
                                                                               

                                       

                                                                    



                                                                    
                                                                               















                                                                      



                                                            




























                                                                         
                                                  





















































































                                                                         

                                                                               





















                                                                              
                                                                               






































































































































































































































































































































                                                                                                                                                         
                                          

                                                     
                                                     





































                                                                         
                                                                   
























































































































                                                                            
































                                                                           




                                                                   

                                        
























                                                                        
                                                                   







































































































































































































































































































































































                                                                               
                                                              

                                                      
 
                                                           
























































                                                                              
                                                                        

                                                                               
                                                                       

















                                                                              
                                                                    

































































                                                                               
                                                                    


























































































































































































































































































































































































































































































































































































                                                                                     
                             



                                           
                                                                


















                                                                          

                                                                               













































                                                                        

                                                                             















































                                                                            




                                                                                








                                                                            




                                                                                



























                                                                           
/*
 * Copyright (c) 2000-2003,2005 Silicon Graphics, Inc.
 * All Rights Reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License as
 * published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it would be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write the Free Software Foundation,
 * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
 */
#include "xfs.h"
#include "xfs_fs.h"
#include "xfs_types.h"
#include "xfs_bit.h"
#include "xfs_log.h"
#include "xfs_inum.h"
#include "xfs_trans.h"
#include "xfs_sb.h"
#include "xfs_ag.h"
#include "xfs_dir.h"
#include "xfs_dir2.h"
#include "xfs_dmapi.h"
#include "xfs_mount.h"
#include "xfs_error.h"
#include "xfs_bmap_btree.h"
#include "xfs_alloc_btree.h"
#include "xfs_ialloc_btree.h"
#include "xfs_dir_sf.h"
#include "xfs_dir2_sf.h"
#include "xfs_attr_sf.h"
#include "xfs_dinode.h"
#include "xfs_inode.h"
#include "xfs_inode_item.h"
#include "xfs_imap.h"
#include "xfs_alloc.h"
#include "xfs_ialloc.h"
#include "xfs_log_priv.h"
#include "xfs_buf_item.h"
#include "xfs_log_recover.h"
#include "xfs_extfree_item.h"
#include "xfs_trans_priv.h"
#include "xfs_quota.h"
#include "xfs_rw.h"

STATIC int	xlog_find_zeroed(xlog_t *, xfs_daddr_t *);
STATIC int	xlog_clear_stale_blocks(xlog_t *, xfs_lsn_t);
STATIC void	xlog_recover_insert_item_backq(xlog_recover_item_t **q,
					       xlog_recover_item_t *item);
#if defined(DEBUG)
STATIC void	xlog_recover_check_summary(xlog_t *);
STATIC void	xlog_recover_check_ail(xfs_mount_t *, xfs_log_item_t *, int);
#else
#define	xlog_recover_check_summary(log)
#define	xlog_recover_check_ail(mp, lip, gen)
#endif


/*
 * Sector aligned buffer routines for buffer create/read/write/access
 */

#define XLOG_SECTOR_ROUNDUP_BBCOUNT(log, bbs)	\
	( ((log)->l_sectbb_mask && (bbs & (log)->l_sectbb_mask)) ? \
	((bbs + (log)->l_sectbb_mask + 1) & ~(log)->l_sectbb_mask) : (bbs) )
#define XLOG_SECTOR_ROUNDDOWN_BLKNO(log, bno)	((bno) & ~(log)->l_sectbb_mask)

xfs_buf_t *
xlog_get_bp(
	xlog_t		*log,
	int		num_bblks)
{
	ASSERT(num_bblks > 0);

	if (log->l_sectbb_log) {
		if (num_bblks > 1)
			num_bblks += XLOG_SECTOR_ROUNDUP_BBCOUNT(log, 1);
		num_bblks = XLOG_SECTOR_ROUNDUP_BBCOUNT(log, num_bblks);
	}
	return xfs_buf_get_noaddr(BBTOB(num_bblks), log->l_mp->m_logdev_targp);
}

void
xlog_put_bp(
	xfs_buf_t	*bp)
{
	xfs_buf_free(bp);
}


/*
 * nbblks should be uint, but oh well.  Just want to catch that 32-bit length.
 */
int
xlog_bread(
	xlog_t		*log,
	xfs_daddr_t	blk_no,
	int		nbblks,
	xfs_buf_t	*bp)
{
	int		error;

	if (log->l_sectbb_log) {
		blk_no = XLOG_SECTOR_ROUNDDOWN_BLKNO(log, blk_no);
		nbblks = XLOG_SECTOR_ROUNDUP_BBCOUNT(log, nbblks);
	}

	ASSERT(nbblks > 0);
	ASSERT(BBTOB(nbblks) <= XFS_BUF_SIZE(bp));
	ASSERT(bp);

	XFS_BUF_SET_ADDR(bp, log->l_logBBstart + blk_no);
	XFS_BUF_READ(bp);
	XFS_BUF_BUSY(bp);
	XFS_BUF_SET_COUNT(bp, BBTOB(nbblks));
	XFS_BUF_SET_TARGET(bp, log->l_mp->m_logdev_targp);

	xfsbdstrat(log->l_mp, bp);
	if ((error = xfs_iowait(bp)))
		xfs_ioerror_alert("xlog_bread", log->l_mp,
				  bp, XFS_BUF_ADDR(bp));
	return error;
}

/*
 * Write out the buffer at the given block for the given number of blocks.
 * The buffer is kept locked across the write and is returned locked.
 * This can only be used for synchronous log writes.
 */
STATIC int
xlog_bwrite(
	xlog_t		*log,
	xfs_daddr_t	blk_no,
	int		nbblks,
	xfs_buf_t	*bp)
{
	int		error;

	if (log->l_sectbb_log) {
		blk_no = XLOG_SECTOR_ROUNDDOWN_BLKNO(log, blk_no);
		nbblks = XLOG_SECTOR_ROUNDUP_BBCOUNT(log, nbblks);
	}

	ASSERT(nbblks > 0);
	ASSERT(BBTOB(nbblks) <= XFS_BUF_SIZE(bp));

	XFS_BUF_SET_ADDR(bp, log->l_logBBstart + blk_no);
	XFS_BUF_ZEROFLAGS(bp);
	XFS_BUF_BUSY(bp);
	XFS_BUF_HOLD(bp);
	XFS_BUF_PSEMA(bp, PRIBIO);
	XFS_BUF_SET_COUNT(bp, BBTOB(nbblks));
	XFS_BUF_SET_TARGET(bp, log->l_mp->m_logdev_targp);

	if ((error = xfs_bwrite(log->l_mp, bp)))
		xfs_ioerror_alert("xlog_bwrite", log->l_mp,
				  bp, XFS_BUF_ADDR(bp));
	return error;
}

STATIC xfs_caddr_t
xlog_align(
	xlog_t		*log,
	xfs_daddr_t	blk_no,
	int		nbblks,
	xfs_buf_t	*bp)
{
	xfs_caddr_t	ptr;

	if (!log->l_sectbb_log)
		return XFS_BUF_PTR(bp);

	ptr = XFS_BUF_PTR(bp) + BBTOB((int)blk_no & log->l_sectbb_mask);
	ASSERT(XFS_BUF_SIZE(bp) >=
		BBTOB(nbblks + (blk_no & log->l_sectbb_mask)));
	return ptr;
}

#ifdef DEBUG
/*
 * dump debug superblock and log record information
 */
STATIC void
xlog_header_check_dump(
	xfs_mount_t		*mp,
	xlog_rec_header_t	*head)
{
	int			b;

	printk("%s:  SB : uuid = ", __FUNCTION__);
	for (b = 0; b < 16; b++)
		printk("%02x",((unsigned char *)&mp->m_sb.sb_uuid)[b]);
	printk(", fmt = %d\n", XLOG_FMT);
	printk("    log : uuid = ");
	for (b = 0; b < 16; b++)
		printk("%02x",((unsigned char *)&head->h_fs_uuid)[b]);
	printk(", fmt = %d\n", INT_GET(head->h_fmt, ARCH_CONVERT));
}
#else
#define xlog_header_check_dump(mp, head)
#endif

/*
 * check log record header for recovery
 */
STATIC int
xlog_header_check_recover(
	xfs_mount_t		*mp,
	xlog_rec_header_t	*head)
{
	ASSERT(INT_GET(head->h_magicno, ARCH_CONVERT) == XLOG_HEADER_MAGIC_NUM);

	/*
	 * IRIX doesn't write the h_fmt field and leaves it zeroed
	 * (XLOG_FMT_UNKNOWN). This stops us from trying to recover
	 * a dirty log created in IRIX.
	 */
	if (unlikely(INT_GET(head->h_fmt, ARCH_CONVERT) != XLOG_FMT)) {
		xlog_warn(
	"XFS: dirty log written in incompatible format - can't recover");
		xlog_header_check_dump(mp, head);
		XFS_ERROR_REPORT("xlog_header_check_recover(1)",
				 XFS_ERRLEVEL_HIGH, mp);
		return XFS_ERROR(EFSCORRUPTED);
	} else if (unlikely(!uuid_equal(&mp->m_sb.sb_uuid, &head->h_fs_uuid))) {
		xlog_warn(
	"XFS: dirty log entry has mismatched uuid - can't recover");
		xlog_header_check_dump(mp, head);
		XFS_ERROR_REPORT("xlog_header_check_recover(2)",
				 XFS_ERRLEVEL_HIGH, mp);
		return XFS_ERROR(EFSCORRUPTED);
	}
	return 0;
}

/*
 * read the head block of the log and check the header
 */
STATIC int
xlog_header_check_mount(
	xfs_mount_t		*mp,
	xlog_rec_header_t	*head)
{
	ASSERT(INT_GET(head->h_magicno, ARCH_CONVERT) == XLOG_HEADER_MAGIC_NUM);

	if (uuid_is_nil(&head->h_fs_uuid)) {
		/*
		 * IRIX doesn't write the h_fs_uuid or h_fmt fields. If
		 * h_fs_uuid is nil, we assume this log was last mounted
		 * by IRIX and continue.
		 */
		xlog_warn("XFS: nil uuid in log - IRIX style log");
	} else if (unlikely(!uuid_equal(&mp->m_sb.sb_uuid, &head->h_fs_uuid))) {
		xlog_warn("XFS: log has mismatched uuid - can't recover");
		xlog_header_check_dump(mp, head);
		XFS_ERROR_REPORT("xlog_header_check_mount",
				 XFS_ERRLEVEL_HIGH, mp);
		return XFS_ERROR(EFSCORRUPTED);
	}
	return 0;
}

STATIC void
xlog_recover_iodone(
	struct xfs_buf	*bp)
{
	xfs_mount_t	*mp;

	ASSERT(XFS_BUF_FSPRIVATE(bp, void *));

	if (XFS_BUF_GETERROR(bp)) {
		/*
		 * We're not going to bother about retrying
		 * this during recovery. One strike!
		 */
		mp = XFS_BUF_FSPRIVATE(bp, xfs_mount_t *);
		xfs_ioerror_alert("xlog_recover_iodone",
				  mp, bp, XFS_BUF_ADDR(bp));
		xfs_force_shutdown(mp, XFS_METADATA_IO_ERROR);
	}
	XFS_BUF_SET_FSPRIVATE(bp, NULL);
	XFS_BUF_CLR_IODONE_FUNC(bp);
	xfs_biodone(bp);
}

/*
 * This routine finds (to an approximation) the first block in the physical
 * log which contains the given cycle.  It uses a binary search algorithm.
 * Note that the algorithm can not be perfect because the disk will not
 * necessarily be perfect.
 */
int
xlog_find_cycle_start(
	xlog_t		*log,
	xfs_buf_t	*bp,
	xfs_daddr_t	first_blk,
	xfs_daddr_t	*last_blk,
	uint		cycle)
{
	xfs_caddr_t	offset;
	xfs_daddr_t	mid_blk;
	uint		mid_cycle;
	int		error;

	mid_blk = BLK_AVG(first_blk, *last_blk);
	while (mid_blk != first_blk && mid_blk != *last_blk) {
		if ((error = xlog_bread(log, mid_blk, 1, bp)))
			return error;
		offset = xlog_align(log, mid_blk, 1, bp);
		mid_cycle = GET_CYCLE(offset, ARCH_CONVERT);
		if (mid_cycle == cycle) {
			*last_blk = mid_blk;
			/* last_half_cycle == mid_cycle */
		} else {
			first_blk = mid_blk;
			/* first_half_cycle == mid_cycle */
		}
		mid_blk = BLK_AVG(first_blk, *last_blk);
	}
	ASSERT((mid_blk == first_blk && mid_blk+1 == *last_blk) ||
	       (mid_blk == *last_blk && mid_blk-1 == first_blk));

	return 0;
}

/*
 * Check that the range of blocks does not contain the cycle number
 * given.  The scan needs to occur from front to back and the ptr into the
 * region must be updated since a later routine will need to perform another
 * test.  If the region is completely good, we end up returning the same
 * last block number.
 *
 * Set blkno to -1 if we encounter no errors.  This is an invalid block number
 * since we don't ever expect logs to get this large.
 */
STATIC int
xlog_find_verify_cycle(
	xlog_t		*log,
	xfs_daddr_t	start_blk,
	int		nbblks,
	uint		stop_on_cycle_no,
	xfs_daddr_t	*new_blk)
{
	xfs_daddr_t	i, j;
	uint		cycle;
	xfs_buf_t	*bp;
	xfs_daddr_t	bufblks;
	xfs_caddr_t	buf = NULL;
	int		error = 0;

	bufblks = 1 << ffs(nbblks);

	while (!(bp = xlog_get_bp(log, bufblks))) {
		/* can't get enough memory to do everything in one big buffer */
		bufblks >>= 1;
		if (bufblks <= log->l_sectbb_log)
			return ENOMEM;
	}

	for (i = start_blk; i < start_blk + nbblks; i += bufblks) {
		int	bcount;

		bcount = min(bufblks, (start_blk + nbblks - i));

		if ((error = xlog_bread(log, i, bcount, bp)))
			goto out;

		buf = xlog_align(log, i, bcount, bp);
		for (j = 0; j < bcount; j++) {
			cycle = GET_CYCLE(buf, ARCH_CONVERT);
			if (cycle == stop_on_cycle_no) {
				*new_blk = i+j;
				goto out;
			}

			buf += BBSIZE;
		}
	}

	*new_blk = -1;

out:
	xlog_put_bp(bp);
	return error;
}

/*
 * Potentially backup over partial log record write.
 *
 * In the typical case, last_blk is the number of the block directly after
 * a good log record.  Therefore, we subtract one to get the block number
 * of the last block in the given buffer.  extra_bblks contains the number
 * of blocks we would have read on a previous read.  This happens when the
 * last log record is split over the end of the physical log.
 *
 * extra_bblks is the number of blocks potentially verified on a previous
 * call to this routine.
 */
STATIC int
xlog_find_verify_log_record(
	xlog_t			*log,
	xfs_daddr_t		start_blk,
	xfs_daddr_t		*last_blk,
	int			extra_bblks)
{
	xfs_daddr_t		i;
	xfs_buf_t		*bp;
	xfs_caddr_t		offset = NULL;
	xlog_rec_header_t	*head = NULL;
	int			error = 0;
	int			smallmem = 0;
	int			num_blks = *last_blk - start_blk;
	int			xhdrs;

	ASSERT(start_blk != 0 || *last_blk != start_blk);

	if (!(bp = xlog_get_bp(log, num_blks))) {
		if (!(bp = xlog_get_bp(log, 1)))
			return ENOMEM;
		smallmem = 1;
	} else {
		if ((error = xlog_bread(log, start_blk, num_blks, bp)))
			goto out;
		offset = xlog_align(log, start_blk, num_blks, bp);
		offset += ((num_blks - 1) << BBSHIFT);
	}

	for (i = (*last_blk) - 1; i >= 0; i--) {
		if (i < start_blk) {
			/* valid log record not found */
			xlog_warn(
		"XFS: Log inconsistent (didn't find previous header)");
			ASSERT(0);
			error = XFS_ERROR(EIO);
			goto out;
		}

		if (smallmem) {
			if ((error = xlog_bread(log, i, 1, bp)))
				goto out;
			offset = xlog_align(log, i, 1, bp);
		}

		head = (xlog_rec_header_t *)offset;

		if (XLOG_HEADER_MAGIC_NUM ==
		    INT_GET(head->h_magicno, ARCH_CONVERT))
			break;

		if (!smallmem)
			offset -= BBSIZE;
	}

	/*
	 * We hit the beginning of the physical log & still no header.  Return
	 * to caller.  If caller can handle a return of -1, then this routine
	 * will be called again for the end of the physical log.
	 */
	if (i == -1) {
		error = -1;
		goto out;
	}

	/*
	 * We have the final block of the good log (the first block
	 * of the log record _before_ the head. So we check the uuid.
	 */
	if ((error = xlog_header_check_mount(log->l_mp, head)))
		goto out;

	/*
	 * We may have found a log record header before we expected one.
	 * last_blk will be the 1st block # with a given cycle #.  We may end
	 * up reading an entire log record.  In this case, we don't want to
	 * reset last_blk.  Only when last_blk points in the middle of a log
	 * record do we update last_blk.
	 */
	if (XFS_SB_VERSION_HASLOGV2(&log->l_mp->m_sb)) {
		uint	h_size = INT_GET(head->h_size, ARCH_CONVERT);

		xhdrs = h_size / XLOG_HEADER_CYCLE_SIZE;
		if (h_size % XLOG_HEADER_CYCLE_SIZE)
			xhdrs++;
	} else {
		xhdrs = 1;
	}

	if (*last_blk - i + extra_bblks
			!= BTOBB(INT_GET(head->h_len, ARCH_CONVERT)) + xhdrs)
		*last_blk = i;

out:
	xlog_put_bp(bp);
	return error;
}

/*
 * Head is defined to be the point of the log where the next log write
 * write could go.  This means that incomplete LR writes at the end are
 * eliminated when calculating the head.  We aren't guaranteed that previous
 * LR have complete transactions.  We only know that a cycle number of
 * current cycle number -1 won't be present in the log if we start writing
 * from our current block number.
 *
 * last_blk contains the block number of the first block with a given
 * cycle number.
 *
 * Return: zero if normal, non-zero if error.
 */
STATIC int
xlog_find_head(
	xlog_t 		*log,
	xfs_daddr_t	*return_head_blk)
{
	xfs_buf_t	*bp;
	xfs_caddr_t	offset;
	xfs_daddr_t	new_blk, first_blk, start_blk, last_blk, head_blk;
	int		num_scan_bblks;
	uint		first_half_cycle, last_half_cycle;
	uint		stop_on_cycle;
	int		error, log_bbnum = log->l_logBBsize;

	/* Is the end of the log device zeroed? */
	if ((error = xlog_find_zeroed(log, &first_blk)) == -1) {
		*return_head_blk = first_blk;

		/* Is the whole lot zeroed? */
		if (!first_blk) {
			/* Linux XFS shouldn't generate totally zeroed logs -
			 * mkfs etc write a dummy unmount record to a fresh
			 * log so we can store the uuid in there
			 */
			xlog_warn("XFS: totally zeroed log");
		}

		return 0;
	} else if (error) {
		xlog_warn("XFS: empty log check failed");
		return error;
	}

	first_blk = 0;			/* get cycle # of 1st block */
	bp = xlog_get_bp(log, 1);
	if (!bp)
		return ENOMEM;
	if ((error = xlog_bread(log, 0, 1, bp)))
		goto bp_err;
	offset = xlog_align(log, 0, 1, bp);
	first_half_cycle = GET_CYCLE(offset, ARCH_CONVERT);

	last_blk = head_blk = log_bbnum - 1;	/* get cycle # of last block */
	if ((error = xlog_bread(log, last_blk, 1, bp)))
		goto bp_err;
	offset = xlog_align(log, last_blk, 1, bp);
	last_half_cycle = GET_CYCLE(offset, ARCH_CONVERT);
	ASSERT(last_half_cycle != 0);

	/*
	 * If the 1st half cycle number is equal to the last half cycle number,
	 * then the entire log is stamped with the same cycle number.  In this
	 * case, head_blk can't be set to zero (which makes sense).  The below
	 * math doesn't work out properly with head_blk equal to zero.  Instead,
	 * we set it to log_bbnum which is an invalid block number, but this
	 * value makes the math correct.  If head_blk doesn't changed through
	 * all the tests below, *head_blk is set to zero at the very end rather
	 * than log_bbnum.  In a sense, log_bbnum and zero are the same block
	 * in a circular file.
	 */
	if (first_half_cycle == last_half_cycle) {
		/*
		 * In this case we believe that the entire log should have
		 * cycle number last_half_cycle.  We need to scan backwards
		 * from the end verifying that there are no holes still
		 * containing last_half_cycle - 1.  If we find such a hole,
		 * then the start of that hole will be the new head.  The
		 * simple case looks like
		 *        x | x ... | x - 1 | x
		 * Another case that fits this picture would be
		 *        x | x + 1 | x ... | x
		 * In this case the head really is somewhere at the end of the
		 * log, as one of the latest writes at the beginning was
		 * incomplete.
		 * One more case is
		 *        x | x + 1 | x ... | x - 1 | x
		 * This is really the combination of the above two cases, and
		 * the head has to end up at the start of the x-1 hole at the
		 * end of the log.
		 *
		 * In the 256k log case, we will read from the beginning to the
		 * end of the log and search for cycle numbers equal to x-1.
		 * We don't worry about the x+1 blocks that we encounter,
		 * because we know that they cannot be the head since the log
		 * started with x.
		 */
		head_blk = log_bbnum;
		stop_on_cycle = last_half_cycle - 1;
	} else {
		/*
		 * In this case we want to find the first block with cycle
		 * number matching last_half_cycle.  We expect the log to be
		 * some variation on
		 *        x + 1 ... | x ...
		 * The first block with cycle number x (last_half_cycle) will
		 * be where the new head belongs.  First we do a binary search
		 * for the first occurrence of last_half_cycle.  The binary
		 * search may not be totally accurate, so then we scan back
		 * from there looking for occurrences of last_half_cycle before
		 * us.  If that backwards scan wraps around the beginning of
		 * the log, then we look for occurrences of last_half_cycle - 1
		 * at the end of the log.  The cases we're looking for look
		 * like
		 *        x + 1 ... | x | x + 1 | x ...
		 *                               ^ binary search stopped here
		 * or
		 *        x + 1 ... | x ... | x - 1 | x
		 *        <---------> less than scan distance
		 */
		stop_on_cycle = last_half_cycle;
		if ((error = xlog_find_cycle_start(log, bp, first_blk,
						&head_blk, last_half_cycle)))
			goto bp_err;
	}

	/*
	 * Now validate the answer.  Scan back some number of maximum possible
	 * blocks and make sure each one has the expected cycle number.  The
	 * maximum is determined by the total possible amount of buffering
	 * in the in-core log.  The following number can be made tighter if
	 * we actually look at the block size of the filesystem.
	 */
	num_scan_bblks = XLOG_TOTAL_REC_SHIFT(log);
	if (head_blk >= num_scan_bblks) {
		/*
		 * We are guaranteed that the entire check can be performed
		 * in one buffer.
		 */
		start_blk = head_blk - num_scan_bblks;
		if ((error = xlog_find_verify_cycle(log,
						start_blk, num_scan_bblks,
						stop_on_cycle, &new_blk)))
			goto bp_err;
		if (new_blk != -1)
			head_blk = new_blk;
	} else {		/* need to read 2 parts of log */
		/*
		 * We are going to scan backwards in the log in two parts.
		 * First we scan the physical end of the log.  In this part
		 * of the log, we are looking for blocks with cycle number
		 * last_half_cycle - 1.
		 * If we find one, then we know that the log starts there, as
		 * we've found a hole that didn't get written in going around
		 * the end of the physical log.  The simple case for this is
		 *        x + 1 ... | x ... | x - 1 | x
		 *        <---------> less than scan distance
		 * If all of the blocks at the end of the log have cycle number
		 * last_half_cycle, then we check the blocks at the start of
		 * the log looking for occurrences of last_half_cycle.  If we
		 * find one, then our current estimate for the location of the
		 * first occurrence of last_half_cycle is wrong and we move
		 * back to the hole we've found.  This case looks like
		 *        x + 1 ... | x | x + 1 | x ...
		 *                               ^ binary search stopped here
		 * Another case we need to handle that only occurs in 256k
		 * logs is
		 *        x + 1 ... | x ... | x+1 | x ...
		 *                   ^ binary search stops here
		 * In a 256k log, the scan at the end of the log will see the
		 * x + 1 blocks.  We need to skip past those since that is
		 * certainly not the head of the log.  By searching for
		 * last_half_cycle-1 we accomplish that.
		 */
		start_blk = log_bbnum - num_scan_bblks + head_blk;
		ASSERT(head_blk <= INT_MAX &&
			(xfs_daddr_t) num_scan_bblks - head_blk >= 0);
		if ((error = xlog_find_verify_cycle(log, start_blk,
					num_scan_bblks - (int)head_blk,
					(stop_on_cycle - 1), &new_blk)))
			goto bp_err;
		if (new_blk != -1) {
			head_blk = new_blk;
			goto bad_blk;
		}

		/*
		 * Scan beginning of log now.  The last part of the physical
		 * log is good.  This scan needs to verify that it doesn't find
		 * the last_half_cycle.
		 */
		start_blk = 0;
		ASSERT(head_blk <= INT_MAX);
		if ((error = xlog_find_verify_cycle(log,
					start_blk, (int)head_blk,
					stop_on_cycle, &new_blk)))
			goto bp_err;
		if (new_blk != -1)
			head_blk = new_blk;
	}

 bad_blk:
	/*
	 * Now we need to make sure head_blk is not pointing to a block in
	 * the middle of a log record.
	 */
	num_scan_bblks = XLOG_REC_SHIFT(log);
	if (head_blk >= num_scan_bblks) {
		start_blk = head_blk - num_scan_bblks; /* don't read head_blk */

		/* start ptr at last block ptr before head_blk */
		if ((error = xlog_find_verify_log_record(log, start_blk,
							&head_blk, 0)) == -1) {
			error = XFS_ERROR(EIO);
			goto bp_err;
		} else if (error)
			goto bp_err;
	} else {
		start_blk = 0;
		ASSERT(head_blk <= INT_MAX);
		if ((error = xlog_find_verify_log_record(log, start_blk,
							&head_blk, 0)) == -1) {
			/* We hit the beginning of the log during our search */
			start_blk = log_bbnum - num_scan_bblks + head_blk;
			new_blk = log_bbnum;
			ASSERT(start_blk <= INT_MAX &&
				(xfs_daddr_t) log_bbnum-start_blk >= 0);
			ASSERT(head_blk <= INT_MAX);
			if ((error = xlog_find_verify_log_record(log,
							start_blk, &new_blk,
							(int)head_blk)) == -1) {
				error = XFS_ERROR(EIO);
				goto bp_err;
			} else if (error)
				goto bp_err;
			if (new_blk != log_bbnum)
				head_blk = new_blk;
		} else if (error)
			goto bp_err;
	}

	xlog_put_bp(bp);
	if (head_blk == log_bbnum)
		*return_head_blk = 0;
	else
		*return_head_blk = head_blk;
	/*
	 * When returning here, we have a good block number.  Bad block
	 * means that during a previous crash, we didn't have a clean break
	 * from cycle number N to cycle number N-1.  In this case, we need
	 * to find the first block with cycle number N-1.
	 */
	return 0;

 bp_err:
	xlog_put_bp(bp);

	if (error)
	    xlog_warn("XFS: failed to find log head");
	return error;
}

/*
 * Find the sync block number or the tail of the log.
 *
 * This will be the block number of the last record to have its
 * associated buffers synced to disk.  Every log record header has
 * a sync lsn embedded in it.  LSNs hold block numbers, so it is easy
 * to get a sync block number.  The only concern is to figure out which
 * log record header to believe.
 *
 * The following algorithm uses the log record header with the largest
 * lsn.  The entire log record does not need to be valid.  We only care
 * that the header is valid.
 *
 * We could speed up search by using current head_blk buffer, but it is not
 * available.
 */
int
xlog_find_tail(
	xlog_t			*log,
	xfs_daddr_t		*head_blk,
	xfs_daddr_t		*tail_blk)
{
	xlog_rec_header_t	*rhead;
	xlog_op_header_t	*op_head;
	xfs_caddr_t		offset = NULL;
	xfs_buf_t		*bp;
	int			error, i, found;
	xfs_daddr_t		umount_data_blk;
	xfs_daddr_t		after_umount_blk;
	xfs_lsn_t		tail_lsn;
	int			hblks;

	found = 0;

	/*
	 * Find previous log record
	 */
	if ((error = xlog_find_head(log, head_blk)))
		return error;

	bp = xlog_get_bp(log, 1);
	if (!bp)
		return ENOMEM;
	if (*head_blk == 0) {				/* special case */
		if ((error = xlog_bread(log, 0, 1, bp)))
			goto bread_err;
		offset = xlog_align(log, 0, 1, bp);
		if (GET_CYCLE(offset, ARCH_CONVERT) == 0) {
			*tail_blk = 0;
			/* leave all other log inited values alone */
			goto exit;
		}
	}

	/*
	 * Search backwards looking for log record header block
	 */
	ASSERT(*head_blk < INT_MAX);
	for (i = (int)(*head_blk) - 1; i >= 0; i--) {
		if ((error = xlog_bread(log, i, 1, bp)))
			goto bread_err;
		offset = xlog_align(log, i, 1, bp);
		if (XLOG_HEADER_MAGIC_NUM ==
		    INT_GET(*(uint *)offset, ARCH_CONVERT)) {
			found = 1;
			break;
		}
	}
	/*
	 * If we haven't found the log record header block, start looking
	 * again from the end of the physical log.  XXXmiken: There should be
	 * a check here to make sure we didn't search more than N blocks in
	 * the previous code.
	 */
	if (!found) {
		for (i = log->l_logBBsize - 1; i >= (int)(*head_blk); i--) {
			if ((error = xlog_bread(log, i, 1, bp)))
				goto bread_err;
			offset = xlog_align(log, i, 1, bp);
			if (XLOG_HEADER_MAGIC_NUM ==
			    INT_GET(*(uint*)offset, ARCH_CONVERT)) {
				found = 2;
				break;
			}
		}
	}
	if (!found) {
		xlog_warn("XFS: xlog_find_tail: couldn't find sync record");
		ASSERT(0);
		return XFS_ERROR(EIO);
	}

	/* find blk_no of tail of log */
	rhead = (xlog_rec_header_t *)offset;
	*tail_blk = BLOCK_LSN(INT_GET(rhead->h_tail_lsn, ARCH_CONVERT));

	/*
	 * Reset log values according to the state of the log when we
	 * crashed.  In the case where head_blk == 0, we bump curr_cycle
	 * one because the next write starts a new cycle rather than
	 * continuing the cycle of the last good log record.  At this
	 * point we have guaranteed that all partial log records have been
	 * accounted for.  Therefore, we know that the last good log record
	 * written was complete and ended exactly on the end boundary
	 * of the physical log.
	 */
	log->l_prev_block = i;
	log->l_curr_block = (int)*head_blk;
	log->l_curr_cycle = INT_GET(rhead->h_cycle, ARCH_CONVERT);
	if (found == 2)
		log->l_curr_cycle++;
	log->l_tail_lsn = INT_GET(rhead->h_tail_lsn, ARCH_CONVERT);
	log->l_last_sync_lsn = INT_GET(rhead->h_lsn, ARCH_CONVERT);
	log->l_grant_reserve_cycle = log->l_curr_cycle;
	log->l_grant_reserve_bytes = BBTOB(log->l_curr_block);
	log->l_grant_write_cycle = log->l_curr_cycle;
	log->l_grant_write_bytes = BBTOB(log->l_curr_block);

	/*
	 * Look for unmount record.  If we find it, then we know there
	 * was a clean unmount.  Since 'i' could be the last block in
	 * the physical log, we convert to a log block before comparing
	 * to the head_blk.
	 *
	 * Save the current tail lsn to use to pass to
	 * xlog_clear_stale_blocks() below.  We won't want to clear the
	 * unmount record if there is one, so we pass the lsn of the
	 * unmount record rather than the block after it.
	 */
	if (XFS_SB_VERSION_HASLOGV2(&log->l_mp->m_sb)) {
		int	h_size = INT_GET(rhead->h_size, ARCH_CONVERT);
		int	h_version = INT_GET(rhead->h_version, ARCH_CONVERT);

		if ((h_version & XLOG_VERSION_2) &&
		    (h_size > XLOG_HEADER_CYCLE_SIZE)) {
			hblks = h_size / XLOG_HEADER_CYCLE_SIZE;
			if (h_size % XLOG_HEADER_CYCLE_SIZE)
				hblks++;
		} else {
			hblks = 1;
		}
	} else {
		hblks = 1;
	}
	after_umount_blk = (i + hblks + (int)
		BTOBB(INT_GET(rhead->h_len, ARCH_CONVERT))) % log->l_logBBsize;
	tail_lsn = log->l_tail_lsn;
	if (*head_blk == after_umount_blk &&
	    INT_GET(rhead->h_num_logops, ARCH_CONVERT) == 1) {
		umount_data_blk = (i + hblks) % log->l_logBBsize;
		if ((error = xlog_bread(log, umount_data_blk, 1, bp))) {
			goto bread_err;
		}
		offset = xlog_align(log, umount_data_blk, 1, bp);
		op_head = (xlog_op_header_t *)offset;
		if (op_head->oh_flags & XLOG_UNMOUNT_TRANS) {
			/*
			 * Set tail and last sync so that newly written
			 * log records will point recovery to after the
			 * current unmount record.
			 */
			ASSIGN_ANY_LSN_HOST(log->l_tail_lsn, log->l_curr_cycle,
					after_umount_blk);
			ASSIGN_ANY_LSN_HOST(log->l_last_sync_lsn, log->l_curr_cycle,
					after_umount_blk);
			*tail_blk = after_umount_blk;
		}
	}

	/*
	 * Make sure that there are no blocks in front of the head
	 * with the same cycle number as the head.  This can happen
	 * because we allow multiple outstanding log writes concurrently,
	 * and the later writes might make it out before earlier ones.
	 *
	 * We use the lsn from before modifying it so that we'll never
	 * overwrite the unmount record after a clean unmount.
	 *
	 * Do this only if we are going to recover the filesystem
	 *
	 * NOTE: This used to say "if (!readonly)"
	 * However on Linux, we can & do recover a read-only filesystem.
	 * We only skip recovery if NORECOVERY is specified on mount,
	 * in which case we would not be here.
	 *
	 * But... if the -device- itself is readonly, just skip this.
	 * We can't recover this device anyway, so it won't matter.
	 */
	if (!xfs_readonly_buftarg(log->l_mp->m_logdev_targp)) {
		error = xlog_clear_stale_blocks(log, tail_lsn);
	}

bread_err:
exit:
	xlog_put_bp(bp);

	if (error)
		xlog_warn("XFS: failed to locate log tail");
	return error;
}

/*
 * Is the log zeroed at all?
 *
 * The last binary search should be changed to perform an X block read
 * once X becomes small enough.  You can then search linearly through
 * the X blocks.  This will cut down on the number of reads we need to do.
 *
 * If the log is partially zeroed, this routine will pass back the blkno
 * of the first block with cycle number 0.  It won't have a complete LR
 * preceding it.
 *
 * Return:
 *	0  => the log is completely written to
 *	-1 => use *blk_no as the first block of the log
 *	>0 => error has occurred
 */
int
xlog_find_zeroed(
	xlog_t		*log,
	xfs_daddr_t	*blk_no)
{
	xfs_buf_t	*bp;
	xfs_caddr_t	offset;
	uint	        first_cycle, last_cycle;
	xfs_daddr_t	new_blk, last_blk, start_blk;
	xfs_daddr_t     num_scan_bblks;
	int	        error, log_bbnum = log->l_logBBsize;

	/* check totally zeroed log */
	bp = xlog_get_bp(log, 1);
	if (!bp)
		return ENOMEM;
	if ((error = xlog_bread(log, 0, 1, bp)))
		goto bp_err;
	offset = xlog_align(log, 0, 1, bp);
	first_cycle = GET_CYCLE(offset, ARCH_CONVERT);
	if (first_cycle == 0) {		/* completely zeroed log */
		*blk_no = 0;
		xlog_put_bp(bp);
		return -1;
	}

	/* check partially zeroed log */
	if ((error = xlog_bread(log, log_bbnum-1, 1, bp)))
		goto bp_err;
	offset = xlog_align(log, log_bbnum-1, 1, bp);
	last_cycle = GET_CYCLE(offset, ARCH_CONVERT);
	if (last_cycle != 0) {		/* log completely written to */
		xlog_put_bp(bp);
		return 0;
	} else if (first_cycle != 1) {
		/*
		 * If the cycle of the last block is zero, the cycle of
		 * the first block must be 1. If it's not, maybe we're
		 * not looking at a log... Bail out.
		 */
		xlog_warn("XFS: Log inconsistent or not a log (last==0, first!=1)");
		return XFS_ERROR(EINVAL);
	}

	/* we have a partially zeroed log */
	last_blk = log_bbnum-1;
	if ((error = xlog_find_cycle_start(log, bp, 0, &last_blk, 0)))
		goto bp_err;

	/*
	 * Validate the answer.  Because there is no way to guarantee that
	 * the entire log is made up of log records which are the same size,
	 * we scan over the defined maximum blocks.  At this point, the maximum
	 * is not chosen to mean anything special.   XXXmiken
	 */
	num_scan_bblks = XLOG_TOTAL_REC_SHIFT(log);
	ASSERT(num_scan_bblks <= INT_MAX);

	if (last_blk < num_scan_bblks)
		num_scan_bblks = last_blk;
	start_blk = last_blk - num_scan_bblks;

	/*
	 * We search for any instances of cycle number 0 that occur before
	 * our current estimate of the head.  What we're trying to detect is
	 *        1 ... | 0 | 1 | 0...
	 *                       ^ binary search ends here
	 */
	if ((error = xlog_find_verify_cycle(log, start_blk,
					 (int)num_scan_bblks, 0, &new_blk)))
		goto bp_err;
	if (new_blk != -1)
		last_blk = new_blk;

	/*
	 * Potentially backup over partial log record write.  We don't need
	 * to search the end of the log because we know it is zero.
	 */
	if ((error = xlog_find_verify_log_record(log, start_blk,
				&last_blk, 0)) == -1) {
	    error = XFS_ERROR(EIO);
	    goto bp_err;
	} else if (error)
	    goto bp_err;

	*blk_no = last_blk;
bp_err:
	xlog_put_bp(bp);
	if (error)
		return error;
	return -1;
}

/*
 * These are simple subroutines used by xlog_clear_stale_blocks() below
 * to initialize a buffer full of empty log record headers and write
 * them into the log.
 */
STATIC void
xlog_add_record(
	xlog_t			*log,
	xfs_caddr_t		buf,
	int			cycle,
	int			block,
	int			tail_cycle,
	int			tail_block)
{
	xlog_rec_header_t	*recp = (xlog_rec_header_t *)buf;

	memset(buf, 0, BBSIZE);
	INT_SET(recp->h_magicno, ARCH_CONVERT, XLOG_HEADER_MAGIC_NUM);
	INT_SET(recp->h_cycle, ARCH_CONVERT, cycle);
	INT_SET(recp->h_version, ARCH_CONVERT,
			XFS_SB_VERSION_HASLOGV2(&log->l_mp->m_sb) ? 2 : 1);
	ASSIGN_ANY_LSN_DISK(recp->h_lsn, cycle, block);
	ASSIGN_ANY_LSN_DISK(recp->h_tail_lsn, tail_cycle, tail_block);
	INT_SET(recp->h_fmt, ARCH_CONVERT, XLOG_FMT);
	memcpy(&recp->h_fs_uuid, &log->l_mp->m_sb.sb_uuid, sizeof(uuid_t));
}

STATIC int
xlog_write_log_records(
	xlog_t		*log,
	int		cycle,
	int		start_block,
	int		blocks,
	int		tail_cycle,
	int		tail_block)
{
	xfs_caddr_t	offset;
	xfs_buf_t	*bp;
	int		balign, ealign;
	int		sectbb = XLOG_SECTOR_ROUNDUP_BBCOUNT(log, 1);
	int		end_block = start_block + blocks;
	int		bufblks;
	int		error = 0;
	int		i, j = 0;

	bufblks = 1 << ffs(blocks);
	while (!(bp = xlog_get_bp(log, bufblks))) {
		bufblks >>= 1;
		if (bufblks <= log->l_sectbb_log)
			return ENOMEM;
	}

	/* We may need to do a read at the start to fill in part of
	 * the buffer in the starting sector not covered by the first
	 * write below.
	 */
	balign = XLOG_SECTOR_ROUNDDOWN_BLKNO(log, start_block);
	if (balign != start_block) {
		if ((error = xlog_bread(log, start_block, 1, bp))) {
			xlog_put_bp(bp);
			return error;
		}
		j = start_block - balign;
	}

	for (i = start_block; i < end_block; i += bufblks) {
		int		bcount, endcount;

		bcount = min(bufblks, end_block - start_block);
		endcount = bcount - j;

		/* We may need to do a read at the end to fill in part of
		 * the buffer in the final sector not covered by the write.
		 * If this is the same sector as the above read, skip it.
		 */
		ealign = XLOG_SECTOR_ROUNDDOWN_BLKNO(log, end_block);
		if (j == 0 && (start_block + endcount > ealign)) {
			offset = XFS_BUF_PTR(bp);
			balign = BBTOB(ealign - start_block);
			XFS_BUF_SET_PTR(bp, offset + balign, BBTOB(sectbb));
			if ((error = xlog_bread(log, ealign, sectbb, bp)))
				break;
			XFS_BUF_SET_PTR(bp, offset, bufblks);
		}

		offset = xlog_align(log, start_block, endcount, bp);
		for (; j < endcount; j++) {
			xlog_add_record(log, offset, cycle, i+j,
					tail_cycle, tail_block);
			offset += BBSIZE;
		}
		error = xlog_bwrite(log, start_block, endcount, bp);
		if (error)
			break;
		start_block += endcount;
		j = 0;
	}
	xlog_put_bp(bp);
	return error;
}

/*
 * This routine is called to blow away any incomplete log writes out
 * in front of the log head.  We do this so that we won't become confused
 * if we come up, write only a little bit more, and then crash again.
 * If we leave the partial log records out there, this situation could
 * cause us to think those partial writes are valid blocks since they
 * have the current cycle number.  We get rid of them by overwriting them
 * with empty log records with the old cycle number rather than the
 * current one.
 *
 * The tail lsn is passed in rather than taken from
 * the log so that we will not write over the unmount record after a
 * clean unmount in a 512 block log.  Doing so would leave the log without
 * any valid log records in it until a new one was written.  If we crashed
 * during that time we would not be able to recover.
 */
STATIC int
xlog_clear_stale_blocks(
	xlog_t		*log,
	xfs_lsn_t	tail_lsn)
{
	int		tail_cycle, head_cycle;
	int		tail_block, head_block;
	int		tail_distance, max_distance;
	int		distance;
	int		error;

	tail_cycle = CYCLE_LSN(tail_lsn);
	tail_block = BLOCK_LSN(tail_lsn);
	head_cycle = log->l_curr_cycle;
	head_block = log->l_curr_block;

	/*
	 * Figure out the distance between the new head of the log
	 * and the tail.  We want to write over any blocks beyond the
	 * head that we may have written just before the crash, but
	 * we don't want to overwrite the tail of the log.
	 */
	if (head_cycle == tail_cycle) {
		/*
		 * The tail is behind the head in the physical log,
		 * so the distance from the head to the tail is the
		 * distance from the head to the end of the log plus
		 * the distance from the beginning of the log to the
		 * tail.
		 */
		if (unlikely(head_block < tail_block || head_block >= log->l_logBBsize)) {
			XFS_ERROR_REPORT("xlog_clear_stale_blocks(1)",
					 XFS_ERRLEVEL_LOW, log->l_mp);
			return XFS_ERROR(EFSCORRUPTED);
		}
		tail_distance = tail_block + (log->l_logBBsize - head_block);
	} else {
		/*
		 * The head is behind the tail in the physical log,
		 * so the distance from the head to the tail is just
		 * the tail block minus the head block.
		 */
		if (unlikely(head_block >= tail_block || head_cycle != (tail_cycle + 1))){
			XFS_ERROR_REPORT("xlog_clear_stale_blocks(2)",
					 XFS_ERRLEVEL_LOW, log->l_mp);
			return XFS_ERROR(EFSCORRUPTED);
		}
		tail_distance = tail_block - head_block;
	}

	/*
	 * If the head is right up against the tail, we can't clear
	 * anything.
	 */
	if (tail_distance <= 0) {
		ASSERT(tail_distance == 0);
		return 0;
	}

	max_distance = XLOG_TOTAL_REC_SHIFT(log);
	/*
	 * Take the smaller of the maximum amount of outstanding I/O
	 * we could have and the distance to the tail to clear out.
	 * We take the smaller so that we don't overwrite the tail and
	 * we don't waste all day writing from the head to the tail
	 * for no reason.
	 */
	max_distance = MIN(max_distance, tail_distance);

	if ((head_block + max_distance) <= log->l_logBBsize) {
		/*
		 * We can stomp all the blocks we need to without
		 * wrapping around the end of the log.  Just do it
		 * in a single write.  Use the cycle number of the
		 * current cycle minus one so that the log will look like:
		 *     n ... | n - 1 ...
		 */
		error = xlog_write_log_records(log, (head_cycle - 1),
				head_block, max_distance, tail_cycle,
				tail_block);
		if (error)
			return error;
	} else {
		/*
		 * We need to wrap around the end of the physical log in
		 * order to clear all the blocks.  Do it in two separate
		 * I/Os.  The first write should be from the head to the
		 * end of the physical log, and it should use the current
		 * cycle number minus one just like above.
		 */
		distance = log->l_logBBsize - head_block;
		error = xlog_write_log_records(log, (head_cycle - 1),
				head_block, distance, tail_cycle,
				tail_block);

		if (error)
			return error;

		/*
		 * Now write the blocks at the start of the physical log.
		 * This writes the remainder of the blocks we want to clear.
		 * It uses the current cycle number since we're now on the
		 * same cycle as the head so that we get:
		 *    n ... n ... | n - 1 ...
		 *    ^^^^^ blocks we're writing
		 */
		distance = max_distance - (log->l_logBBsize - head_block);
		error = xlog_write_log_records(log, head_cycle, 0, distance,
				tail_cycle, tail_block);
		if (error)
			return error;
	}

	return 0;
}

/******************************************************************************
 *
 *		Log recover routines
 *
 ******************************************************************************
 */

STATIC xlog_recover_t *
xlog_recover_find_tid(
	xlog_recover_t		*q,
	xlog_tid_t		tid)
{
	xlog_recover_t		*p = q;

	while (p != NULL) {
		if (p->r_log_tid == tid)
		    break;
		p = p->r_next;
	}
	return p;
}

STATIC void
xlog_recover_put_hashq(
	xlog_recover_t		**q,
	xlog_recover_t		*trans)
{
	trans->r_next = *q;
	*q = trans;
}

STATIC void
xlog_recover_add_item(
	xlog_recover_item_t	**itemq)
{
	xlog_recover_item_t	*item;

	item = kmem_zalloc(sizeof(xlog_recover_item_t), KM_SLEEP);
	xlog_recover_insert_item_backq(itemq, item);
}

STATIC int
xlog_recover_add_to_cont_trans(
	xlog_recover_t		*trans,
	xfs_caddr_t		dp,
	int			len)
{
	xlog_recover_item_t	*item;
	xfs_caddr_t		ptr, old_ptr;
	int			old_len;

	item = trans->r_itemq;
	if (item == 0) {
		/* finish copying rest of trans header */
		xlog_recover_add_item(&trans->r_itemq);
		ptr = (xfs_caddr_t) &trans->r_theader +
				sizeof(xfs_trans_header_t) - len;
		memcpy(ptr, dp, len); /* d, s, l */
		return 0;
	}
	item = item->ri_prev;

	old_ptr = item->ri_buf[item->ri_cnt-1].i_addr;
	old_len = item->ri_buf[item->ri_cnt-1].i_len;

	ptr = kmem_realloc(old_ptr, len+old_len, old_len, 0u);
	memcpy(&ptr[old_len], dp, len); /* d, s, l */
	item->ri_buf[item->ri_cnt-1].i_len += len;
	item->ri_buf[item->ri_cnt-1].i_addr = ptr;
	return 0;
}

/*
 * The next region to add is the start of a new region.  It could be
 * a whole region or it could be the first part of a new region.  Because
 * of this, the assumption here is that the type and size fields of all
 * format structures fit into the first 32 bits of the structure.
 *
 * This works because all regions must be 32 bit aligned.  Therefore, we
 * either have both fields or we have neither field.  In the case we have
 * neither field, the data part of the region is zero length.  We only have
 * a log_op_header and can throw away the header since a new one will appear
 * later.  If we have at least 4 bytes, then we can determine how many regions
 * will appear in the current log item.
 */
STATIC int
xlog_recover_add_to_trans(
	xlog_recover_t		*trans,
	xfs_caddr_t		dp,
	int			len)
{
	xfs_inode_log_format_t	*in_f;			/* any will do */
	xlog_recover_item_t	*item;
	xfs_caddr_t		ptr;

	if (!len)
		return 0;
	item = trans->r_itemq;
	if (item == 0) {
		ASSERT(*(uint *)dp == XFS_TRANS_HEADER_MAGIC);
		if (len == sizeof(xfs_trans_header_t))
			xlog_recover_add_item(&trans->r_itemq);
		memcpy(&trans->r_theader, dp, len); /* d, s, l */
		return 0;
	}

	ptr = kmem_alloc(len, KM_SLEEP);
	memcpy(ptr, dp, len);
	in_f = (xfs_inode_log_format_t *)ptr;

	if (item->ri_prev->ri_total != 0 &&
	     item->ri_prev->ri_total == item->ri_prev->ri_cnt) {
		xlog_recover_add_item(&trans->r_itemq);
	}
	item = trans->r_itemq;
	item = item->ri_prev;

	if (item->ri_total == 0) {		/* first region to be added */
		item->ri_total	= in_f->ilf_size;
		ASSERT(item->ri_total <= XLOG_MAX_REGIONS_IN_ITEM);
		item->ri_buf = kmem_zalloc((item->ri_total *
					    sizeof(xfs_log_iovec_t)), KM_SLEEP);
	}
	ASSERT(item->ri_total > item->ri_cnt);
	/* Description region is ri_buf[0] */
	item->ri_buf[item->ri_cnt].i_addr = ptr;
	item->ri_buf[item->ri_cnt].i_len  = len;
	item->ri_cnt++;
	return 0;
}

STATIC void
xlog_recover_new_tid(
	xlog_recover_t		**q,
	xlog_tid_t		tid,
	xfs_lsn_t		lsn)
{
	xlog_recover_t		*trans;

	trans = kmem_zalloc(sizeof(xlog_recover_t), KM_SLEEP);
	trans->r_log_tid   = tid;
	trans->r_lsn	   = lsn;
	xlog_recover_put_hashq(q, trans);
}

STATIC int
xlog_recover_unlink_tid(
	xlog_recover_t		**q,
	xlog_recover_t		*trans)
{
	xlog_recover_t		*tp;
	int			found = 0;

	ASSERT(trans != 0);
	if (trans == *q) {
		*q = (*q)->r_next;
	} else {
		tp = *q;
		while (tp != 0) {
			if (tp->r_next == trans) {
				found = 1;
				break;
			}
			tp = tp->r_next;
		}
		if (!found) {
			xlog_warn(
			     "XFS: xlog_recover_unlink_tid: trans not found");
			ASSERT(0);
			return XFS_ERROR(EIO);
		}
		tp->r_next = tp->r_next->r_next;
	}
	return 0;
}

STATIC void
xlog_recover_insert_item_backq(
	xlog_recover_item_t	**q,
	xlog_recover_item_t	*item)
{
	if (*q == 0) {
		item->ri_prev = item->ri_next = item;
		*q = item;
	} else {
		item->ri_next		= *q;
		item->ri_prev		= (*q)->ri_prev;
		(*q)->ri_prev		= item;
		item->ri_prev->ri_next	= item;
	}
}

STATIC void
xlog_recover_insert_item_frontq(
	xlog_recover_item_t	**q,
	xlog_recover_item_t	*item)
{
	xlog_recover_insert_item_backq(q, item);
	*q = item;
}

STATIC int
xlog_recover_reorder_trans(
	xlog_t			*log,
	xlog_recover_t		*trans)
{
	xlog_recover_item_t	*first_item, *itemq, *itemq_next;
	xfs_buf_log_format_t	*buf_f;
	xfs_buf_log_format_v1_t	*obuf_f;
	ushort			flags = 0;

	first_item = itemq = trans->r_itemq;
	trans->r_itemq = NULL;
	do {
		itemq_next = itemq->ri_next;
		buf_f = (xfs_buf_log_format_t *)itemq->ri_buf[0].i_addr;
		switch (ITEM_TYPE(itemq)) {
		case XFS_LI_BUF:
			flags = buf_f->blf_flags;
			break;
		case XFS_LI_6_1_BUF:
		case XFS_LI_5_3_BUF:
			obuf_f = (xfs_buf_log_format_v1_t*)buf_f;
			flags = obuf_f->blf_flags;
			break;
		}

		switch (ITEM_TYPE(itemq)) {
		case XFS_LI_BUF:
		case XFS_LI_6_1_BUF:
		case XFS_LI_5_3_BUF:
			if (!(flags & XFS_BLI_CANCEL)) {
				xlog_recover_insert_item_frontq(&trans->r_itemq,
								itemq);
				break;
			}
		case XFS_LI_INODE:
		case XFS_LI_6_1_INODE:
		case XFS_LI_5_3_INODE:
		case XFS_LI_DQUOT:
		case XFS_LI_QUOTAOFF:
		case XFS_LI_EFD:
		case XFS_LI_EFI:
			xlog_recover_insert_item_backq(&trans->r_itemq, itemq);
			break;
		default:
			xlog_warn(
	"XFS: xlog_recover_reorder_trans: unrecognized type of log operation");
			ASSERT(0);
			return XFS_ERROR(EIO);
		}
		itemq = itemq_next;
	} while (first_item != itemq);
	return 0;
}

/*
 * Build up the table of buf cancel records so that we don't replay
 * cancelled data in the second pass.  For buffer records that are
 * not cancel records, there is nothing to do here so we just return.
 *
 * If we get a cancel record which is already in the table, this indicates
 * that the buffer was cancelled multiple times.  In order to ensure
 * that during pass 2 we keep the record in the table until we reach its
 * last occurrence in the log, we keep a reference count in the cancel
 * record in the table to tell us how many times we expect to see this
 * record during the second pass.
 */
STATIC void
xlog_recover_do_buffer_pass1(
	xlog_t			*log,
	xfs_buf_log_format_t	*buf_f)
{
	xfs_buf_cancel_t	*bcp;
	xfs_buf_cancel_t	*nextp;
	xfs_buf_cancel_t	*prevp;
	xfs_buf_cancel_t	**bucket;
	xfs_buf_log_format_v1_t	*obuf_f;
	xfs_daddr_t		blkno = 0;
	uint			len = 0;
	ushort			flags = 0;

	switch (buf_f->blf_type) {
	case XFS_LI_BUF:
		blkno = buf_f->blf_blkno;
		len = buf_f->blf_len;
		flags = buf_f->blf_flags;
		break;
	case XFS_LI_6_1_BUF:
	case XFS_LI_5_3_BUF:
		obuf_f = (xfs_buf_log_format_v1_t*)buf_f;
		blkno = (xfs_daddr_t) obuf_f->blf_blkno;
		len = obuf_f->blf_len;
		flags = obuf_f->blf_flags;
		break;
	}

	/*
	 * If this isn't a cancel buffer item, then just return.
	 */
	if (!(flags & XFS_BLI_CANCEL))
		return;

	/*
	 * Insert an xfs_buf_cancel record into the hash table of
	 * them.  If there is already an identical record, bump
	 * its reference count.
	 */
	bucket = &log->l_buf_cancel_table[(__uint64_t)blkno %
					  XLOG_BC_TABLE_SIZE];
	/*
	 * If the hash bucket is empty then just insert a new record into
	 * the bucket.
	 */
	if (*bucket == NULL) {
		bcp = (xfs_buf_cancel_t *)kmem_alloc(sizeof(xfs_buf_cancel_t),
						     KM_SLEEP);
		bcp->bc_blkno = blkno;
		bcp->bc_len = len;
		bcp->bc_refcount = 1;
		bcp->bc_next = NULL;
		*bucket = bcp;
		return;
	}

	/*
	 * The hash bucket is not empty, so search for duplicates of our
	 * record.  If we find one them just bump its refcount.  If not
	 * then add us at the end of the list.
	 */
	prevp = NULL;
	nextp = *bucket;
	while (nextp != NULL) {
		if (nextp->bc_blkno == blkno && nextp->bc_len == len) {
			nextp->bc_refcount++;
			return;
		}
		prevp = nextp;
		nextp = nextp->bc_next;
	}
	ASSERT(prevp != NULL);
	bcp = (xfs_buf_cancel_t *)kmem_alloc(sizeof(xfs_buf_cancel_t),
					     KM_SLEEP);
	bcp->bc_blkno = blkno;
	bcp->bc_len = len;
	bcp->bc_refcount = 1;
	bcp->bc_next = NULL;
	prevp->bc_next = bcp;
}

/*
 * Check to see whether the buffer being recovered has a corresponding
 * entry in the buffer cancel record table.  If it does then return 1
 * so that it will be cancelled, otherwise return 0.  If the buffer is
 * actually a buffer cancel item (XFS_BLI_CANCEL is set), then decrement
 * the refcount on the entry in the table and remove it from the table
 * if this is the last reference.
 *
 * We remove the cancel record from the table when we encounter its
 * last occurrence in the log so that if the same buffer is re-used
 * again after its last cancellation we actually replay the changes
 * made at that point.
 */
STATIC int
xlog_check_buffer_cancelled(
	xlog_t			*log,
	xfs_daddr_t		blkno,
	uint			len,
	ushort			flags)
{
	xfs_buf_cancel_t	*bcp;
	xfs_buf_cancel_t	*prevp;
	xfs_buf_cancel_t	**bucket;

	if (log->l_buf_cancel_table == NULL) {
		/*
		 * There is nothing in the table built in pass one,
		 * so this buffer must not be cancelled.
		 */
		ASSERT(!(flags & XFS_BLI_CANCEL));
		return 0;
	}

	bucket = &log->l_buf_cancel_table[(__uint64_t)blkno %
					  XLOG_BC_TABLE_SIZE];
	bcp = *bucket;
	if (bcp == NULL) {
		/*
		 * There is no corresponding entry in the table built
		 * in pass one, so this buffer has not been cancelled.
		 */
		ASSERT(!(flags & XFS_BLI_CANCEL));
		return 0;
	}

	/*
	 * Search for an entry in the buffer cancel table that
	 * matches our buffer.
	 */
	prevp = NULL;
	while (bcp != NULL) {
		if (bcp->bc_blkno == blkno && bcp->bc_len == len) {
			/*
			 * We've go a match, so return 1 so that the
			 * recovery of this buffer is cancelled.
			 * If this buffer is actually a buffer cancel
			 * log item, then decrement the refcount on the
			 * one in the table and remove it if this is the
			 * last reference.
			 */
			if (flags & XFS_BLI_CANCEL) {
				bcp->bc_refcount--;
				if (bcp->bc_refcount == 0) {
					if (prevp == NULL) {
						*bucket = bcp->bc_next;
					} else {
						prevp->bc_next = bcp->bc_next;
					}
					kmem_free(bcp,
						  sizeof(xfs_buf_cancel_t));
				}
			}
			return 1;
		}
		prevp = bcp;
		bcp = bcp->bc_next;
	}
	/*
	 * We didn't find a corresponding entry in the table, so
	 * return 0 so that the buffer is NOT cancelled.
	 */
	ASSERT(!(flags & XFS_BLI_CANCEL));
	return 0;
}

STATIC int
xlog_recover_do_buffer_pass2(
	xlog_t			*log,
	xfs_buf_log_format_t	*buf_f)
{
	xfs_buf_log_format_v1_t	*obuf_f;
	xfs_daddr_t		blkno = 0;
	ushort			flags = 0;
	uint			len = 0;

	switch (buf_f->blf_type) {
	case XFS_LI_BUF:
		blkno = buf_f->blf_blkno;
		flags = buf_f->blf_flags;
		len = buf_f->blf_len;
		break;
	case XFS_LI_6_1_BUF:
	case XFS_LI_5_3_BUF:
		obuf_f = (xfs_buf_log_format_v1_t*)buf_f;
		blkno = (xfs_daddr_t) obuf_f->blf_blkno;
		flags = obuf_f->blf_flags;
		len = (xfs_daddr_t) obuf_f->blf_len;
		break;
	}

	return xlog_check_buffer_cancelled(log, blkno, len, flags);
}

/*
 * Perform recovery for a buffer full of inodes.  In these buffers,
 * the only data which should be recovered is that which corresponds
 * to the di_next_unlinked pointers in the on disk inode structures.
 * The rest of the data for the inodes is always logged through the
 * inodes themselves rather than the inode buffer and is recovered
 * in xlog_recover_do_inode_trans().
 *
 * The only time when buffers full of inodes are fully recovered is
 * when the buffer is full of newly allocated inodes.  In this case
 * the buffer will not be marked as an inode buffer and so will be
 * sent to xlog_recover_do_reg_buffer() below during recovery.
 */
STATIC int
xlog_recover_do_inode_buffer(
	xfs_mount_t		*mp,
	xlog_recover_item_t	*item,
	xfs_buf_t		*bp,
	xfs_buf_log_format_t	*buf_f)
{
	int			i;
	int			item_index;
	int			bit;
	int			nbits;
	int			reg_buf_offset;
	int			reg_buf_bytes;
	int			next_unlinked_offset;
	int			inodes_per_buf;
	xfs_agino_t		*logged_nextp;
	xfs_agino_t		*buffer_nextp;
	xfs_buf_log_format_v1_t	*obuf_f;
	unsigned int		*data_map = NULL;
	unsigned int		map_size = 0;

	switch (buf_f->blf_type) {
	case XFS_LI_BUF:
		data_map = buf_f->blf_data_map;
		map_size = buf_f->blf_map_size;
		break;
	case XFS_LI_6_1_BUF:
	case XFS_LI_5_3_BUF:
		obuf_f = (xfs_buf_log_format_v1_t*)buf_f;
		data_map = obuf_f->blf_data_map;
		map_size = obuf_f->blf_map_size;
		break;
	}
	/*
	 * Set the variables corresponding to the current region to
	 * 0 so that we'll initialize them on the first pass through
	 * the loop.
	 */
	reg_buf_offset = 0;
	reg_buf_bytes = 0;
	bit = 0;
	nbits = 0;
	item_index = 0;
	inodes_per_buf = XFS_BUF_COUNT(bp) >> mp->m_sb.sb_inodelog;
	for (i = 0; i < inodes_per_buf; i++) {
		next_unlinked_offset = (i * mp->m_sb.sb_inodesize) +
			offsetof(xfs_dinode_t, di_next_unlinked);

		while (next_unlinked_offset >=
		       (reg_buf_offset + reg_buf_bytes)) {
			/*
			 * The next di_next_unlinked field is beyond
			 * the current logged region.  Find the next
			 * logged region that contains or is beyond
			 * the current di_next_unlinked field.
			 */
			bit += nbits;
			bit = xfs_next_bit(data_map, map_size, bit);

			/*
			 * If there are no more logged regions in the
			 * buffer, then we're done.
			 */
			if (bit == -1) {
				return 0;
			}

			nbits = xfs_contig_bits(data_map, map_size,
							 bit);
			ASSERT(nbits > 0);
			reg_buf_offset = bit << XFS_BLI_SHIFT;
			reg_buf_bytes = nbits << XFS_BLI_SHIFT;
			item_index++;
		}

		/*
		 * If the current logged region starts after the current
		 * di_next_unlinked field, then move on to the next
		 * di_next_unlinked field.
		 */
		if (next_unlinked_offset < reg_buf_offset) {
			continue;
		}

		ASSERT(item->ri_buf[item_index].i_addr != NULL);
		ASSERT((item->ri_buf[item_index].i_len % XFS_BLI_CHUNK) == 0);
		ASSERT((reg_buf_offset + reg_buf_bytes) <= XFS_BUF_COUNT(bp));

		/*
		 * The current logged region contains a copy of the
		 * current di_next_unlinked field.  Extract its value
		 * and copy it to the buffer copy.
		 */
		logged_nextp = (xfs_agino_t *)
			       ((char *)(item->ri_buf[item_index].i_addr) +
				(next_unlinked_offset - reg_buf_offset));
		if (unlikely(*logged_nextp == 0)) {
			xfs_fs_cmn_err(CE_ALERT, mp,
				"bad inode buffer log record (ptr = 0x%p, bp = 0x%p).  XFS trying to replay bad (0) inode di_next_unlinked field",
				item, bp);
			XFS_ERROR_REPORT("xlog_recover_do_inode_buf",
					 XFS_ERRLEVEL_LOW, mp);
			return XFS_ERROR(EFSCORRUPTED);
		}

		buffer_nextp = (xfs_agino_t *)xfs_buf_offset(bp,
					      next_unlinked_offset);
		INT_SET(*buffer_nextp, ARCH_CONVERT, *logged_nextp);
	}

	return 0;
}

/*
 * Perform a 'normal' buffer recovery.  Each logged region of the
 * buffer should be copied over the corresponding region in the
 * given buffer.  The bitmap in the buf log format structure indicates
 * where to place the logged data.
 */
/*ARGSUSED*/
STATIC void
xlog_recover_do_reg_buffer(
	xfs_mount_t		*mp,
	xlog_recover_item_t	*item,
	xfs_buf_t		*bp,
	xfs_buf_log_format_t	*buf_f)
{
	int			i;
	int			bit;
	int			nbits;
	xfs_buf_log_format_v1_t	*obuf_f;
	unsigned int		*data_map = NULL;
	unsigned int		map_size = 0;
	int                     error;

	switch (buf_f->blf_type) {
	case XFS_LI_BUF:
		data_map = buf_f->blf_data_map;
		map_size = buf_f->blf_map_size;
		break;
	case XFS_LI_6_1_BUF:
	case XFS_LI_5_3_BUF:
		obuf_f = (xfs_buf_log_format_v1_t*)buf_f;
		data_map = obuf_f->blf_data_map;
		map_size = obuf_f->blf_map_size;
		break;
	}
	bit = 0;
	i = 1;  /* 0 is the buf format structure */
	while (1) {
		bit = xfs_next_bit(data_map, map_size, bit);
		if (bit == -1)
			break;
		nbits = xfs_contig_bits(data_map, map_size, bit);
		ASSERT(nbits > 0);
		ASSERT(item->ri_buf[i].i_addr != 0);
		ASSERT(item->ri_buf[i].i_len % XFS_BLI_CHUNK == 0);
		ASSERT(XFS_BUF_COUNT(bp) >=
		       ((uint)bit << XFS_BLI_SHIFT)+(nbits<<XFS_BLI_SHIFT));

		/*
		 * Do a sanity check if this is a dquot buffer. Just checking
		 * the first dquot in the buffer should do. XXXThis is
		 * probably a good thing to do for other buf types also.
		 */
		error = 0;
		if (buf_f->blf_flags &
		   (XFS_BLI_UDQUOT_BUF|XFS_BLI_PDQUOT_BUF|XFS_BLI_GDQUOT_BUF)) {
			error = xfs_qm_dqcheck((xfs_disk_dquot_t *)
					       item->ri_buf[i].i_addr,
					       -1, 0, XFS_QMOPT_DOWARN,
					       "dquot_buf_recover");
		}
		if (!error)
			memcpy(xfs_buf_offset(bp,
				(uint)bit << XFS_BLI_SHIFT),	/* dest */
				item->ri_buf[i].i_addr,		/* source */
				nbits<<XFS_BLI_SHIFT);		/* length */
		i++;
		bit += nbits;
	}

	/* Shouldn't be any more regions */
	ASSERT(i == item->ri_total);
}

/*
 * Do some primitive error checking on ondisk dquot data structures.
 */
int
xfs_qm_dqcheck(
	xfs_disk_dquot_t *ddq,
	xfs_dqid_t	 id,
	uint		 type,	  /* used only when IO_dorepair is true */
	uint		 flags,
	char		 *str)
{
	xfs_dqblk_t	 *d = (xfs_dqblk_t *)ddq;
	int		errs = 0;

	/*
	 * We can encounter an uninitialized dquot buffer for 2 reasons:
	 * 1. If we crash while deleting the quotainode(s), and those blks got
	 *    used for user data. This is because we take the path of regular
	 *    file deletion; however, the size field of quotainodes is never
	 *    updated, so all the tricks that we play in itruncate_finish
	 *    don't quite matter.
	 *
	 * 2. We don't play the quota buffers when there's a quotaoff logitem.
	 *    But the allocation will be replayed so we'll end up with an
	 *    uninitialized quota block.
	 *
	 * This is all fine; things are still consistent, and we haven't lost
	 * any quota information. Just don't complain about bad dquot blks.
	 */
	if (be16_to_cpu(ddq->d_magic) != XFS_DQUOT_MAGIC) {
		if (flags & XFS_QMOPT_DOWARN)
			cmn_err(CE_ALERT,
			"%s : XFS dquot ID 0x%x, magic 0x%x != 0x%x",
			str, id, be16_to_cpu(ddq->d_magic), XFS_DQUOT_MAGIC);
		errs++;
	}
	if (ddq->d_version != XFS_DQUOT_VERSION) {
		if (flags & XFS_QMOPT_DOWARN)
			cmn_err(CE_ALERT,
			"%s : XFS dquot ID 0x%x, version 0x%x != 0x%x",
			str, id, ddq->d_version, XFS_DQUOT_VERSION);
		errs++;
	}

	if (ddq->d_flags != XFS_DQ_USER &&
	    ddq->d_flags != XFS_DQ_PROJ &&
	    ddq->d_flags != XFS_DQ_GROUP) {
		if (flags & XFS_QMOPT_DOWARN)
			cmn_err(CE_ALERT,
			"%s : XFS dquot ID 0x%x, unknown flags 0x%x",
			str, id, ddq->d_flags);
		errs++;
	}

	if (id != -1 && id != be32_to_cpu(ddq->d_id)) {
		if (flags & XFS_QMOPT_DOWARN)
			cmn_err(CE_ALERT,
			"%s : ondisk-dquot 0x%p, ID mismatch: "
			"0x%x expected, found id 0x%x",
			str, ddq, id, be32_to_cpu(ddq->d_id));
		errs++;
	}

	if (!errs && ddq->d_id) {
		if (ddq->d_blk_softlimit &&
		    be64_to_cpu(ddq->d_bcount) >=
				be64_to_cpu(ddq->d_blk_softlimit)) {
			if (!ddq->d_btimer) {
				if (flags & XFS_QMOPT_DOWARN)
					cmn_err(CE_ALERT,
					"%s : Dquot ID 0x%x (0x%p) "
					"BLK TIMER NOT STARTED",
					str, (int)be32_to_cpu(ddq->d_id), ddq);
				errs++;
			}
		}
		if (ddq->d_ino_softlimit &&
		    be64_to_cpu(ddq->d_icount) >=
				be64_to_cpu(ddq->d_ino_softlimit)) {
			if (!ddq->d_itimer) {
				if (flags & XFS_QMOPT_DOWARN)
					cmn_err(CE_ALERT,
					"%s : Dquot ID 0x%x (0x%p) "
					"INODE TIMER NOT STARTED",
					str, (int)be32_to_cpu(ddq->d_id), ddq);
				errs++;
			}
		}
		if (ddq->d_rtb_softlimit &&
		    be64_to_cpu(ddq->d_rtbcount) >=
				be64_to_cpu(ddq->d_rtb_softlimit)) {
			if (!ddq->d_rtbtimer) {
				if (flags & XFS_QMOPT_DOWARN)
					cmn_err(CE_ALERT,
					"%s : Dquot ID 0x%x (0x%p) "
					"RTBLK TIMER NOT STARTED",
					str, (int)be32_to_cpu(ddq->d_id), ddq);
				errs++;
			}
		}
	}

	if (!errs || !(flags & XFS_QMOPT_DQREPAIR))
		return errs;

	if (flags & XFS_QMOPT_DOWARN)
		cmn_err(CE_NOTE, "Re-initializing dquot ID 0x%x", id);

	/*
	 * Typically, a repair is only requested by quotacheck.
	 */
	ASSERT(id != -1);
	ASSERT(flags & XFS_QMOPT_DQREPAIR);
	memset(d, 0, sizeof(xfs_dqblk_t));

	d->dd_diskdq.d_magic = cpu_to_be16(XFS_DQUOT_MAGIC);
	d->dd_diskdq.d_version = XFS_DQUOT_VERSION;
	d->dd_diskdq.d_flags = type;
	d->dd_diskdq.d_id = cpu_to_be32(id);

	return errs;
}

/*
 * Perform a dquot buffer recovery.
 * Simple algorithm: if we have found a QUOTAOFF logitem of the same type
 * (ie. USR or GRP), then just toss this buffer away; don't recover it.
 * Else, treat it as a regular buffer and do recovery.
 */
STATIC void
xlog_recover_do_dquot_buffer(
	xfs_mount_t		*mp,
	xlog_t			*log,
	xlog_recover_item_t	*item,
	xfs_buf_t		*bp,
	xfs_buf_log_format_t	*buf_f)
{
	uint			type;

	/*
	 * Filesystems are required to send in quota flags at mount time.
	 */
	if (mp->m_qflags == 0) {
		return;
	}

	type = 0;
	if (buf_f->blf_flags & XFS_BLI_UDQUOT_BUF)
		type |= XFS_DQ_USER;
	if (buf_f->blf_flags & XFS_BLI_PDQUOT_BUF)
		type |= XFS_DQ_PROJ;
	if (buf_f->blf_flags & XFS_BLI_GDQUOT_BUF)
		type |= XFS_DQ_GROUP;
	/*
	 * This type of quotas was turned off, so ignore this buffer
	 */
	if (log->l_quotaoffs_flag & type)
		return;

	xlog_recover_do_reg_buffer(mp, item, bp, buf_f);
}

/*
 * This routine replays a modification made to a buffer at runtime.
 * There are actually two types of buffer, regular and inode, which
 * are handled differently.  Inode buffers are handled differently
 * in that we only recover a specific set of data from them, namely
 * the inode di_next_unlinked fields.  This is because all other inode
 * data is actually logged via inode records and any data we replay
 * here which overlaps that may be stale.
 *
 * When meta-data buffers are freed at run time we log a buffer item
 * with the XFS_BLI_CANCEL bit set to indicate that previous copies
 * of the buffer in the log should not be replayed at recovery time.
 * This is so that if the blocks covered by the buffer are reused for
 * file data before we crash we don't end up replaying old, freed
 * meta-data into a user's file.
 *
 * To handle the cancellation of buffer log items, we make two passes
 * over the log during recovery.  During the first we build a table of
 * those buffers which have been cancelled, and during the second we
 * only replay those buffers which do not have corresponding cancel
 * records in the table.  See xlog_recover_do_buffer_pass[1,2] above
 * for more details on the implementation of the table of cancel records.
 */
STATIC int
xlog_recover_do_buffer_trans(
	xlog_t			*log,
	xlog_recover_item_t	*item,
	int			pass)
{
	xfs_buf_log_format_t	*buf_f;
	xfs_buf_log_format_v1_t	*obuf_f;
	xfs_mount_t		*mp;
	xfs_buf_t		*bp;
	int			error;
	int			cancel;
	xfs_daddr_t		blkno;
	int			len;
	ushort			flags;

	buf_f = (xfs_buf_log_format_t *)item->ri_buf[0].i_addr;

	if (pass == XLOG_RECOVER_PASS1) {
		/*
		 * In this pass we're only looking for buf items
		 * with the XFS_BLI_CANCEL bit set.
		 */
		xlog_recover_do_buffer_pass1(log, buf_f);
		return 0;
	} else {
		/*
		 * In this pass we want to recover all the buffers
		 * which have not been cancelled and are not
		 * cancellation buffers themselves.  The routine
		 * we call here will tell us whether or not to
		 * continue with the replay of this buffer.
		 */
		cancel = xlog_recover_do_buffer_pass2(log, buf_f);
		if (cancel) {
			return 0;
		}
	}
	switch (buf_f->blf_type) {
	case XFS_LI_BUF:
		blkno = buf_f->blf_blkno;
		len = buf_f->blf_len;
		flags = buf_f->blf_flags;
		break;
	case XFS_LI_6_1_BUF:
	case XFS_LI_5_3_BUF:
		obuf_f = (xfs_buf_log_format_v1_t*)buf_f;
		blkno = obuf_f->blf_blkno;
		len = obuf_f->blf_len;
		flags = obuf_f->blf_flags;
		break;
	default:
		xfs_fs_cmn_err(CE_ALERT, log->l_mp,
			"xfs_log_recover: unknown buffer type 0x%x, logdev %s",
			buf_f->blf_type, log->l_mp->m_logname ?
			log->l_mp->m_logname : "internal");
		XFS_ERROR_REPORT("xlog_recover_do_buffer_trans",
				 XFS_ERRLEVEL_LOW, log->l_mp);
		return XFS_ERROR(EFSCORRUPTED);
	}

	mp = log->l_mp;
	if (flags & XFS_BLI_INODE_BUF) {
		bp = xfs_buf_read_flags(mp->m_ddev_targp, blkno, len,
								XFS_BUF_LOCK);
	} else {
		bp = xfs_buf_read(mp->m_ddev_targp, blkno, len, 0);
	}
	if (XFS_BUF_ISERROR(bp)) {
		xfs_ioerror_alert("xlog_recover_do..(read#1)", log->l_mp,
				  bp, blkno);
		error = XFS_BUF_GETERROR(bp);
		xfs_buf_relse(bp);
		return error;
	}

	error = 0;
	if (flags & XFS_BLI_INODE_BUF) {
		error = xlog_recover_do_inode_buffer(mp, item, bp, buf_f);
	} else if (flags &
		  (XFS_BLI_UDQUOT_BUF|XFS_BLI_PDQUOT_BUF|XFS_BLI_GDQUOT_BUF)) {
		xlog_recover_do_dquot_buffer(mp, log, item, bp, buf_f);
	} else {
		xlog_recover_do_reg_buffer(mp, item, bp, buf_f);
	}
	if (error)
		return XFS_ERROR(error);

	/*
	 * Perform delayed write on the buffer.  Asynchronous writes will be
	 * slower when taking into account all the buffers to be flushed.
	 *
	 * Also make sure that only inode buffers with good sizes stay in
	 * the buffer cache.  The kernel moves inodes in buffers of 1 block
	 * or XFS_INODE_CLUSTER_SIZE bytes, whichever is bigger.  The inode
	 * buffers in the log can be a different size if the log was generated
	 * by an older kernel using unclustered inode buffers or a newer kernel
	 * running with a different inode cluster size.  Regardless, if the
	 * the inode buffer size isn't MAX(blocksize, XFS_INODE_CLUSTER_SIZE)
	 * for *our* value of XFS_INODE_CLUSTER_SIZE, then we need to keep
	 * the buffer out of the buffer cache so that the buffer won't
	 * overlap with future reads of those inodes.
	 */
	if (XFS_DINODE_MAGIC ==
	    INT_GET(*((__uint16_t *)(xfs_buf_offset(bp, 0))), ARCH_CONVERT) &&
	    (XFS_BUF_COUNT(bp) != MAX(log->l_mp->m_sb.sb_blocksize,
			(__uint32_t)XFS_INODE_CLUSTER_SIZE(log->l_mp)))) {
		XFS_BUF_STALE(bp);
		error = xfs_bwrite(mp, bp);
	} else {
		ASSERT(XFS_BUF_FSPRIVATE(bp, void *) == NULL ||
		       XFS_BUF_FSPRIVATE(bp, xfs_mount_t *) == mp);
		XFS_BUF_SET_FSPRIVATE(bp, mp);
		XFS_BUF_SET_IODONE_FUNC(bp, xlog_recover_iodone);
		xfs_bdwrite(mp, bp);
	}

	return (error);
}

STATIC int
xlog_recover_do_inode_trans(
	xlog_t			*log,
	xlog_recover_item_t	*item,
	int			pass)
{
	xfs_inode_log_format_t	*in_f;
	xfs_mount_t		*mp;
	xfs_buf_t		*bp;
	xfs_imap_t		imap;
	xfs_dinode_t		*dip;
	xfs_ino_t		ino;
	int			len;
	xfs_caddr_t		src;
	xfs_caddr_t		dest;
	int			error;
	int			attr_index;
	uint			fields;
	xfs_dinode_core_t	*dicp;

	if (pass == XLOG_RECOVER_PASS1) {
		return 0;
	}

	in_f = (xfs_inode_log_format_t *)item->ri_buf[0].i_addr;
	ino = in_f->ilf_ino;
	mp = log->l_mp;
	if (ITEM_TYPE(item) == XFS_LI_INODE) {
		imap.im_blkno = (xfs_daddr_t)in_f->ilf_blkno;
		imap.im_len = in_f->ilf_len;
		imap.im_boffset = in_f->ilf_boffset;
	} else {
		/*
		 * It's an old inode format record.  We don't know where
		 * its cluster is located on disk, and we can't allow
		 * xfs_imap() to figure it out because the inode btrees
		 * are not ready to be used.  Therefore do not pass the
		 * XFS_IMAP_LOOKUP flag to xfs_imap().  This will give
		 * us only the single block in which the inode lives
		 * rather than its cluster, so we must make sure to
		 * invalidate the buffer when we write it out below.
		 */
		imap.im_blkno = 0;
		xfs_imap(log->l_mp, NULL, ino, &imap, 0);
	}

	/*
	 * Inode buffers can be freed, look out for it,
	 * and do not replay the inode.
	 */
	if (xlog_check_buffer_cancelled(log, imap.im_blkno, imap.im_len, 0))
		return 0;

	bp = xfs_buf_read_flags(mp->m_ddev_targp, imap.im_blkno, imap.im_len,
								XFS_BUF_LOCK);
	if (XFS_BUF_ISERROR(bp)) {
		xfs_ioerror_alert("xlog_recover_do..(read#2)", mp,
				  bp, imap.im_blkno);
		error = XFS_BUF_GETERROR(bp);
		xfs_buf_relse(bp);
		return error;
	}
	error = 0;
	ASSERT(in_f->ilf_fields & XFS_ILOG_CORE);
	dip = (xfs_dinode_t *)xfs_buf_offset(bp, imap.im_boffset);

	/*
	 * Make sure the place we're flushing out to really looks
	 * like an inode!
	 */
	if (unlikely(INT_GET(dip->di_core.di_magic, ARCH_CONVERT) != XFS_DINODE_MAGIC)) {
		xfs_buf_relse(bp);
		xfs_fs_cmn_err(CE_ALERT, mp,
			"xfs_inode_recover: Bad inode magic number, dino ptr = 0x%p, dino bp = 0x%p, ino = %Ld",
			dip, bp, ino);
		XFS_ERROR_REPORT("xlog_recover_do_inode_trans(1)",
				 XFS_ERRLEVEL_LOW, mp);
		return XFS_ERROR(EFSCORRUPTED);
	}
	dicp = (xfs_dinode_core_t*)(item->ri_buf[1].i_addr);
	if (unlikely(dicp->di_magic != XFS_DINODE_MAGIC)) {
		xfs_buf_relse(bp);
		xfs_fs_cmn_err(CE_ALERT, mp,
			"xfs_inode_recover: Bad inode log record, rec ptr 0x%p, ino %Ld",
			item, ino);
		XFS_ERROR_REPORT("xlog_recover_do_inode_trans(2)",
				 XFS_ERRLEVEL_LOW, mp);
		return XFS_ERROR(EFSCORRUPTED);
	}

	/* Skip replay when the on disk inode is newer than the log one */
	if (dicp->di_flushiter <
	    INT_GET(dip->di_core.di_flushiter, ARCH_CONVERT)) {
		/*
		 * Deal with the wrap case, DI_MAX_FLUSH is less
		 * than smaller numbers
		 */
		if ((INT_GET(dip->di_core.di_flushiter, ARCH_CONVERT)
							== DI_MAX_FLUSH) &&
		    (dicp->di_flushiter < (DI_MAX_FLUSH>>1))) {
			/* do nothing */
		} else {
			xfs_buf_relse(bp);
			return 0;
		}
	}
	/* Take the opportunity to reset the flush iteration count */
	dicp->di_flushiter = 0;

	if (unlikely((dicp->di_mode & S_IFMT) == S_IFREG)) {
		if ((dicp->di_format != XFS_DINODE_FMT_EXTENTS) &&
		    (dicp->di_format != XFS_DINODE_FMT_BTREE)) {
			XFS_CORRUPTION_ERROR("xlog_recover_do_inode_trans(3)",
					 XFS_ERRLEVEL_LOW, mp, dicp);
			xfs_buf_relse(bp);
			xfs_fs_cmn_err(CE_ALERT, mp,
				"xfs_inode_recover: Bad regular inode log record, rec ptr 0x%p, ino ptr = 0x%p, ino bp = 0x%p, ino %Ld",
				item, dip, bp, ino);
			return XFS_ERROR(EFSCORRUPTED);
		}
	} else if (unlikely((dicp->di_mode & S_IFMT) == S_IFDIR)) {
		if ((dicp->di_format != XFS_DINODE_FMT_EXTENTS) &&
		    (dicp->di_format != XFS_DINODE_FMT_BTREE) &&
		    (dicp->di_format != XFS_DINODE_FMT_LOCAL)) {
			XFS_CORRUPTION_ERROR("xlog_recover_do_inode_trans(4)",
					     XFS_ERRLEVEL_LOW, mp, dicp);
			xfs_buf_relse(bp);
			xfs_fs_cmn_err(CE_ALERT, mp,
				"xfs_inode_recover: Bad dir inode log record, rec ptr 0x%p, ino ptr = 0x%p, ino bp = 0x%p, ino %Ld",
				item, dip, bp, ino);
			return XFS_ERROR(EFSCORRUPTED);
		}
	}
	if (unlikely(dicp->di_nextents + dicp->di_anextents > dicp->di_nblocks)){
		XFS_CORRUPTION_ERROR("xlog_recover_do_inode_trans(5)",
				     XFS_ERRLEVEL_LOW, mp, dicp);
		xfs_buf_relse(bp);
		xfs_fs_cmn_err(CE_ALERT, mp,
			"xfs_inode_recover: Bad inode log record, rec ptr 0x%p, dino ptr 0x%p, dino bp 0x%p, ino %Ld, total extents = %d, nblocks = %Ld",
			item, dip, bp, ino,
			dicp->di_nextents + dicp->di_anextents,
			dicp->di_nblocks);
		return XFS_ERROR(EFSCORRUPTED);
	}
	if (unlikely(dicp->di_forkoff > mp->m_sb.sb_inodesize)) {
		XFS_CORRUPTION_ERROR("xlog_recover_do_inode_trans(6)",
				     XFS_ERRLEVEL_LOW, mp, dicp);
		xfs_buf_relse(bp);
		xfs_fs_cmn_err(CE_ALERT, mp,
			"xfs_inode_recover: Bad inode log rec ptr 0x%p, dino ptr 0x%p, dino bp 0x%p, ino %Ld, forkoff 0x%x",
			item, dip, bp, ino, dicp->di_forkoff);
		return XFS_ERROR(EFSCORRUPTED);
	}
	if (unlikely(item->ri_buf[1].i_len > sizeof(xfs_dinode_core_t))) {
		XFS_CORRUPTION_ERROR("xlog_recover_do_inode_trans(7)",
				     XFS_ERRLEVEL_LOW, mp, dicp);
		xfs_buf_relse(bp);
		xfs_fs_cmn_err(CE_ALERT, mp,
			"xfs_inode_recover: Bad inode log record length %d, rec ptr 0x%p",
			item->ri_buf[1].i_len, item);
		return XFS_ERROR(EFSCORRUPTED);
	}

	/* The core is in in-core format */
	xfs_xlate_dinode_core((xfs_caddr_t)&dip->di_core,
			      (xfs_dinode_core_t*)item->ri_buf[1].i_addr, -1);

	/* the rest is in on-disk format */
	if (item->ri_buf[1].i_len > sizeof(xfs_dinode_core_t)) {
		memcpy((xfs_caddr_t) dip + sizeof(xfs_dinode_core_t),
			item->ri_buf[1].i_addr + sizeof(xfs_dinode_core_t),
			item->ri_buf[1].i_len  - sizeof(xfs_dinode_core_t));
	}

	fields = in_f->ilf_fields;
	switch (fields & (XFS_ILOG_DEV | XFS_ILOG_UUID)) {
	case XFS_ILOG_DEV:
		INT_SET(dip->di_u.di_dev, ARCH_CONVERT, in_f->ilf_u.ilfu_rdev);

		break;
	case XFS_ILOG_UUID:
		dip->di_u.di_muuid = in_f->ilf_u.ilfu_uuid;
		break;
	}

	if (in_f->ilf_size == 2)
		goto write_inode_buffer;
	len = item->ri_buf[2].i_len;
	src = item->ri_buf[2].i_addr;
	ASSERT(in_f->ilf_size <= 4);
	ASSERT((in_f->ilf_size == 3) || (fields & XFS_ILOG_AFORK));
	ASSERT(!(fields & XFS_ILOG_DFORK) ||
	       (len == in_f->ilf_dsize));

	switch (fields & XFS_ILOG_DFORK) {
	case XFS_ILOG_DDATA:
	case XFS_ILOG_DEXT:
		memcpy(&dip->di_u, src, len);
		break;

	case XFS_ILOG_DBROOT:
		xfs_bmbt_to_bmdr((xfs_bmbt_block_t *)src, len,
				 &(dip->di_u.di_bmbt),
				 XFS_DFORK_DSIZE(dip, mp));
		break;

	default:
		/*
		 * There are no data fork flags set.
		 */
		ASSERT((fields & XFS_ILOG_DFORK) == 0);
		break;
	}

	/*
	 * If we logged any attribute data, recover it.  There may or
	 * may not have been any other non-core data logged in this
	 * transaction.
	 */
	if (in_f->ilf_fields & XFS_ILOG_AFORK) {
		if (in_f->ilf_fields & XFS_ILOG_DFORK) {
			attr_index = 3;
		} else {
			attr_index = 2;
		}
		len = item->ri_buf[attr_index].i_len;
		src = item->ri_buf[attr_index].i_addr;
		ASSERT(len == in_f->ilf_asize);

		switch (in_f->ilf_fields & XFS_ILOG_AFORK) {
		case XFS_ILOG_ADATA:
		case XFS_ILOG_AEXT:
			dest = XFS_DFORK_APTR(dip);
			ASSERT(len <= XFS_DFORK_ASIZE(dip, mp));
			memcpy(dest, src, len);
			break;

		case XFS_ILOG_ABROOT:
			dest = XFS_DFORK_APTR(dip);
			xfs_bmbt_to_bmdr((xfs_bmbt_block_t *)src, len,
					 (xfs_bmdr_block_t*)dest,
					 XFS_DFORK_ASIZE(dip, mp));
			break;

		default:
			xlog_warn("XFS: xlog_recover_do_inode_trans: Invalid flag");
			ASSERT(0);
			xfs_buf_relse(bp);
			return XFS_ERROR(EIO);
		}
	}

write_inode_buffer:
	if (ITEM_TYPE(item) == XFS_LI_INODE) {
		ASSERT(XFS_BUF_FSPRIVATE(bp, void *) == NULL ||
		       XFS_BUF_FSPRIVATE(bp, xfs_mount_t *) == mp);
		XFS_BUF_SET_FSPRIVATE(bp, mp);
		XFS_BUF_SET_IODONE_FUNC(bp, xlog_recover_iodone);
		xfs_bdwrite(mp, bp);
	} else {
		XFS_BUF_STALE(bp);
		error = xfs_bwrite(mp, bp);
	}

	return (error);
}

/*
 * Recover QUOTAOFF records. We simply make a note of it in the xlog_t
 * structure, so that we know not to do any dquot item or dquot buffer recovery,
 * of that type.
 */
STATIC int
xlog_recover_do_quotaoff_trans(
	xlog_t			*log,
	xlog_recover_item_t	*item,
	int			pass)
{
	xfs_qoff_logformat_t	*qoff_f;

	if (pass == XLOG_RECOVER_PASS2) {
		return (0);
	}

	qoff_f = (xfs_qoff_logformat_t *)item->ri_buf[0].i_addr;
	ASSERT(qoff_f);

	/*
	 * The logitem format's flag tells us if this was user quotaoff,
	 * group/project quotaoff or both.
	 */
	if (qoff_f->qf_flags & XFS_UQUOTA_ACCT)
		log->l_quotaoffs_flag |= XFS_DQ_USER;
	if (qoff_f->qf_flags & XFS_PQUOTA_ACCT)
		log->l_quotaoffs_flag |= XFS_DQ_PROJ;
	if (qoff_f->qf_flags & XFS_GQUOTA_ACCT)
		log->l_quotaoffs_flag |= XFS_DQ_GROUP;

	return (0);
}

/*
 * Recover a dquot record
 */
STATIC int
xlog_recover_do_dquot_trans(
	xlog_t			*log,
	xlog_recover_item_t	*item,
	int			pass)
{
	xfs_mount_t		*mp;
	xfs_buf_t		*bp;
	struct xfs_disk_dquot	*ddq, *recddq;
	int			error;
	xfs_dq_logformat_t	*dq_f;
	uint			type;

	if (pass == XLOG_RECOVER_PASS1) {
		return 0;
	}
	mp = log->l_mp;

	/*
	 * Filesystems are required to send in quota flags at mount time.
	 */
	if (mp->m_qflags == 0)
		return (0);

	recddq = (xfs_disk_dquot_t *)item->ri_buf[1].i_addr;
	ASSERT(recddq);
	/*
	 * This type of quotas was turned off, so ignore this record.
	 */
	type = INT_GET(recddq->d_flags, ARCH_CONVERT) &
			(XFS_DQ_USER | XFS_DQ_PROJ | XFS_DQ_GROUP);
	ASSERT(type);
	if (log->l_quotaoffs_flag & type)
		return (0);

	/*
	 * At this point we know that quota was _not_ turned off.
	 * Since the mount flags are not indicating to us otherwise, this
	 * must mean that quota is on, and the dquot needs to be replayed.
	 * Remember that we may not have fully recovered the superblock yet,
	 * so we can't do the usual trick of looking at the SB quota bits.
	 *
	 * The other possibility, of course, is that the quota subsystem was
	 * removed since the last mount - ENOSYS.
	 */
	dq_f = (xfs_dq_logformat_t *)item->ri_buf[0].i_addr;
	ASSERT(dq_f);
	if ((error = xfs_qm_dqcheck(recddq,
			   dq_f->qlf_id,
			   0, XFS_QMOPT_DOWARN,
			   "xlog_recover_do_dquot_trans (log copy)"))) {
		return XFS_ERROR(EIO);
	}
	ASSERT(dq_f->qlf_len == 1);

	error = xfs_read_buf(mp, mp->m_ddev_targp,
			     dq_f->qlf_blkno,
			     XFS_FSB_TO_BB(mp, dq_f->qlf_len),
			     0, &bp);
	if (error) {
		xfs_ioerror_alert("xlog_recover_do..(read#3)", mp,
				  bp, dq_f->qlf_blkno);
		return error;
	}
	ASSERT(bp);
	ddq = (xfs_disk_dquot_t *)xfs_buf_offset(bp, dq_f->qlf_boffset);

	/*
	 * At least the magic num portion should be on disk because this
	 * was among a chunk of dquots created earlier, and we did some
	 * minimal initialization then.
	 */
	if (xfs_qm_dqcheck(ddq, dq_f->qlf_id, 0, XFS_QMOPT_DOWARN,
			   "xlog_recover_do_dquot_trans")) {
		xfs_buf_relse(bp);
		return XFS_ERROR(EIO);
	}

	memcpy(ddq, recddq, item->ri_buf[1].i_len);

	ASSERT(dq_f->qlf_size == 2);
	ASSERT(XFS_BUF_FSPRIVATE(bp, void *) == NULL ||
	       XFS_BUF_FSPRIVATE(bp, xfs_mount_t *) == mp);
	XFS_BUF_SET_FSPRIVATE(bp, mp);
	XFS_BUF_SET_IODONE_FUNC(bp, xlog_recover_iodone);
	xfs_bdwrite(mp, bp);

	return (0);
}

/*
 * This routine is called to create an in-core extent free intent
 * item from the efi format structure which was logged on disk.
 * It allocates an in-core efi, copies the extents from the format
 * structure into it, and adds the efi to the AIL with the given
 * LSN.
 */
STATIC void
xlog_recover_do_efi_trans(
	xlog_t			*log,
	xlog_recover_item_t	*item,
	xfs_lsn_t		lsn,
	int			pass)
{
	xfs_mount_t		*mp;
	xfs_efi_log_item_t	*efip;
	xfs_efi_log_format_t	*efi_formatp;
	SPLDECL(s);

	if (pass == XLOG_RECOVER_PASS1) {
		return;
	}

	efi_formatp = (xfs_efi_log_format_t *)item->ri_buf[0].i_addr;
	ASSERT(item->ri_buf[0].i_len ==
	       (sizeof(xfs_efi_log_format_t) +
		((efi_formatp->efi_nextents - 1) * sizeof(xfs_extent_t))));

	mp = log->l_mp;
	efip = xfs_efi_init(mp, efi_formatp->efi_nextents);
	memcpy((char *)&(efip->efi_format), (char *)efi_formatp,
	      sizeof(xfs_efi_log_format_t) +
	      ((efi_formatp->efi_nextents - 1) * sizeof(xfs_extent_t)));
	efip->efi_next_extent = efi_formatp->efi_nextents;
	efip->efi_flags |= XFS_EFI_COMMITTED;

	AIL_LOCK(mp,s);
	/*
	 * xfs_trans_update_ail() drops the AIL lock.
	 */
	xfs_trans_update_ail(mp, (xfs_log_item_t *)efip, lsn, s);
}


/*
 * This routine is called when an efd format structure is found in
 * a committed transaction in the log.  It's purpose is to cancel
 * the corresponding efi if it was still in the log.  To do this
 * it searches the AIL for the efi with an id equal to that in the
 * efd format structure.  If we find it, we remove the efi from the
 * AIL and free it.
 */
STATIC void
xlog_recover_do_efd_trans(
	xlog_t			*log,
	xlog_recover_item_t	*item,
	int			pass)
{
	xfs_mount_t		*mp;
	xfs_efd_log_format_t	*efd_formatp;
	xfs_efi_log_item_t	*efip = NULL;
	xfs_log_item_t		*lip;
	int			gen;
	__uint64_t		efi_id;
	SPLDECL(s);

	if (pass == XLOG_RECOVER_PASS1) {
		return;
	}

	efd_formatp = (xfs_efd_log_format_t *)item->ri_buf[0].i_addr;
	ASSERT(item->ri_buf[0].i_len ==
	       (sizeof(xfs_efd_log_format_t) +
		((efd_formatp->efd_nextents - 1) * sizeof(xfs_extent_t))));
	efi_id = efd_formatp->efd_efi_id;

	/*
	 * Search for the efi with the id in the efd format structure
	 * in the AIL.
	 */
	mp = log->l_mp;
	AIL_LOCK(mp,s);
	lip = xfs_trans_first_ail(mp, &gen);
	while (lip != NULL) {
		if (lip->li_type == XFS_LI_EFI) {
			efip = (xfs_efi_log_item_t *)lip;
			if (efip->efi_format.efi_id == efi_id) {
				/*
				 * xfs_trans_delete_ail() drops the
				 * AIL lock.
				 */
				xfs_trans_delete_ail(mp, lip, s);
				break;
			}
		}
		lip = xfs_trans_next_ail(mp, lip, &gen, NULL);
	}

	/*
	 * If we found it, then free it up.  If it wasn't there, it
	 * must have been overwritten in the log.  Oh well.
	 */
	if (lip != NULL) {
		xfs_efi_item_free(efip);
	} else {
		AIL_UNLOCK(mp, s);
	}
}

/*
 * Perform the transaction
 *
 * If the transaction modifies a buffer or inode, do it now.  Otherwise,
 * EFIs and EFDs get queued up by adding entries into the AIL for them.
 */
STATIC int
xlog_recover_do_trans(
	xlog_t			*log,
	xlog_recover_t		*trans,
	int			pass)
{
	int			error = 0;
	xlog_recover_item_t	*item, *first_item;

	if ((error = xlog_recover_reorder_trans(log, trans)))
		return error;
	first_item = item = trans->r_itemq;
	do {
		/*
		 * we don't need to worry about the block number being
		 * truncated in > 1 TB buffers because in user-land,
		 * we're now n32 or 64-bit so xfs_daddr_t is 64-bits so
		 * the blknos will get through the user-mode buffer
		 * cache properly.  The only bad case is o32 kernels
		 * where xfs_daddr_t is 32-bits but mount will warn us
		 * off a > 1 TB filesystem before we get here.
		 */
		if ((ITEM_TYPE(item) == XFS_LI_BUF) ||
		    (ITEM_TYPE(item) == XFS_LI_6_1_BUF) ||
		    (ITEM_TYPE(item) == XFS_LI_5_3_BUF)) {
			if  ((error = xlog_recover_do_buffer_trans(log, item,
								 pass)))
				break;
		} else if ((ITEM_TYPE(item) == XFS_LI_INODE) ||
			   (ITEM_TYPE(item) == XFS_LI_6_1_INODE) ||
			   (ITEM_TYPE(item) == XFS_LI_5_3_INODE)) {
			if ((error = xlog_recover_do_inode_trans(log, item,
								pass)))
				break;
		} else if (ITEM_TYPE(item) == XFS_LI_EFI) {
			xlog_recover_do_efi_trans(log, item, trans->r_lsn,
						  pass);
		} else if (ITEM_TYPE(item) == XFS_LI_EFD) {
			xlog_recover_do_efd_trans(log, item, pass);
		} else if (ITEM_TYPE(item) == XFS_LI_DQUOT) {
			if ((error = xlog_recover_do_dquot_trans(log, item,
								   pass)))
					break;
		} else if ((ITEM_TYPE(item) == XFS_LI_QUOTAOFF)) {
			if ((error = xlog_recover_do_quotaoff_trans(log, item,
								   pass)))
					break;
		} else {
			xlog_warn("XFS: xlog_recover_do_trans");
			ASSERT(0);
			error = XFS_ERROR(EIO);
			break;
		}
		item = item->ri_next;
	} while (first_item != item);

	return error;
}

/*
 * Free up any resources allocated by the transaction
 *
 * Remember that EFIs, EFDs, and IUNLINKs are handled later.
 */
STATIC void
xlog_recover_free_trans(
	xlog_recover_t		*trans)
{
	xlog_recover_item_t	*first_item, *item, *free_item;
	int			i;

	item = first_item = trans->r_itemq;
	do {
		free_item = item;
		item = item->ri_next;
		 /* Free the regions in the item. */
		for (i = 0; i < free_item->ri_cnt; i++) {
			kmem_free(free_item->ri_buf[i].i_addr,
				  free_item->ri_buf[i].i_len);
		}
		/* Free the item itself */
		kmem_free(free_item->ri_buf,
			  (free_item->ri_total * sizeof(xfs_log_iovec_t)));
		kmem_free(free_item, sizeof(xlog_recover_item_t));
	} while (first_item != item);
	/* Free the transaction recover structure */
	kmem_free(trans, sizeof(xlog_recover_t));
}

STATIC int
xlog_recover_commit_trans(
	xlog_t			*log,
	xlog_recover_t		**q,
	xlog_recover_t		*trans,
	int			pass)
{
	int			error;

	if ((error = xlog_recover_unlink_tid(q, trans)))
		return error;
	if ((error = xlog_recover_do_trans(log, trans, pass)))
		return error;
	xlog_recover_free_trans(trans);			/* no error */
	return 0;
}

STATIC int
xlog_recover_unmount_trans(
	xlog_recover_t		*trans)
{
	/* Do nothing now */
	xlog_warn("XFS: xlog_recover_unmount_trans: Unmount LR");
	return 0;
}

/*
 * There are two valid states of the r_state field.  0 indicates that the
 * transaction structure is in a normal state.  We have either seen the
 * start of the transaction or the last operation we added was not a partial
 * operation.  If the last operation we added to the transaction was a
 * partial operation, we need to mark r_state with XLOG_WAS_CONT_TRANS.
 *
 * NOTE: skip LRs with 0 data length.
 */
STATIC int
xlog_recover_process_data(
	xlog_t			*log,
	xlog_recover_t		*rhash[],
	xlog_rec_header_t	*rhead,
	xfs_caddr_t		dp,
	int			pass)
{
	xfs_caddr_t		lp;
	int			num_logops;
	xlog_op_header_t	*ohead;
	xlog_recover_t		*trans;
	xlog_tid_t		tid;
	int			error;
	unsigned long		hash;
	uint			flags;

	lp = dp + INT_GET(rhead->h_len, ARCH_CONVERT);
	num_logops = INT_GET(rhead->h_num_logops, ARCH_CONVERT);

	/* check the log format matches our own - else we can't recover */
	if (xlog_header_check_recover(log->l_mp, rhead))
		return (XFS_ERROR(EIO));

	while ((dp < lp) && num_logops) {
		ASSERT(dp + sizeof(xlog_op_header_t) <= lp);
		ohead = (xlog_op_header_t *)dp;
		dp += sizeof(xlog_op_header_t);
		if (ohead->oh_clientid != XFS_TRANSACTION &&
		    ohead->oh_clientid != XFS_LOG) {
			xlog_warn(
		"XFS: xlog_recover_process_data: bad clientid");
			ASSERT(0);
			return (XFS_ERROR(EIO));
		}
		tid = INT_GET(ohead->oh_tid, ARCH_CONVERT);
		hash = XLOG_RHASH(tid);
		trans = xlog_recover_find_tid(rhash[hash], tid);
		if (trans == NULL) {		   /* not found; add new tid */
			if (ohead->oh_flags & XLOG_START_TRANS)
				xlog_recover_new_tid(&rhash[hash], tid,
					INT_GET(rhead->h_lsn, ARCH_CONVERT));
		} else {
			ASSERT(dp+INT_GET(ohead->oh_len, ARCH_CONVERT) <= lp);
			flags = ohead->oh_flags & ~XLOG_END_TRANS;
			if (flags & XLOG_WAS_CONT_TRANS)
				flags &= ~XLOG_CONTINUE_TRANS;
			switch (flags) {
			case XLOG_COMMIT_TRANS:
				error = xlog_recover_commit_trans(log,
						&rhash[hash], trans, pass);
				break;
			case XLOG_UNMOUNT_TRANS:
				error = xlog_recover_unmount_trans(trans);
				break;
			case XLOG_WAS_CONT_TRANS:
				error = xlog_recover_add_to_cont_trans(trans,
						dp, INT_GET(ohead->oh_len,
							ARCH_CONVERT));
				break;
			case XLOG_START_TRANS:
				xlog_warn(
			"XFS: xlog_recover_process_data: bad transaction");
				ASSERT(0);
				error = XFS_ERROR(EIO);
				break;
			case 0:
			case XLOG_CONTINUE_TRANS:
				error = xlog_recover_add_to_trans(trans,
						dp, INT_GET(ohead->oh_len,
							ARCH_CONVERT));
				break;
			default:
				xlog_warn(
			"XFS: xlog_recover_process_data: bad flag");
				ASSERT(0);
				error = XFS_ERROR(EIO);
				break;
			}
			if (error)
				return error;
		}
		dp += INT_GET(ohead->oh_len, ARCH_CONVERT);
		num_logops--;
	}
	return 0;
}

/*
 * Process an extent free intent item that was recovered from
 * the log.  We need to free the extents that it describes.
 */
STATIC void
xlog_recover_process_efi(
	xfs_mount_t		*mp,
	xfs_efi_log_item_t	*efip)
{
	xfs_efd_log_item_t	*efdp;
	xfs_trans_t		*tp;
	int			i;
	xfs_extent_t		*extp;
	xfs_fsblock_t		startblock_fsb;

	ASSERT(!(efip->efi_flags & XFS_EFI_RECOVERED));

	/*
	 * First check the validity of the extents described by the
	 * EFI.  If any are bad, then assume that all are bad and
	 * just toss the EFI.
	 */
	for (i = 0; i < efip->efi_format.efi_nextents; i++) {
		extp = &(efip->efi_format.efi_extents[i]);
		startblock_fsb = XFS_BB_TO_FSB(mp,
				   XFS_FSB_TO_DADDR(mp, extp->ext_start));
		if ((startblock_fsb == 0) ||
		    (extp->ext_len == 0) ||
		    (startblock_fsb >= mp->m_sb.sb_dblocks) ||
		    (extp->ext_len >= mp->m_sb.sb_agblocks)) {
			/*
			 * This will pull the EFI from the AIL and
			 * free the memory associated with it.
			 */
			xfs_efi_release(efip, efip->efi_format.efi_nextents);
			return;
		}
	}

	tp = xfs_trans_alloc(mp, 0);
	xfs_trans_reserve(tp, 0, XFS_ITRUNCATE_LOG_RES(mp), 0, 0, 0);
	efdp = xfs_trans_get_efd(tp, efip, efip->efi_format.efi_nextents);

	for (i = 0; i < efip->efi_format.efi_nextents; i++) {
		extp = &(efip->efi_format.efi_extents[i]);
		xfs_free_extent(tp, extp->ext_start, extp->ext_len);
		xfs_trans_log_efd_extent(tp, efdp, extp->ext_start,
					 extp->ext_len);
	}

	efip->efi_flags |= XFS_EFI_RECOVERED;
	xfs_trans_commit(tp, 0, NULL);
}

/*
 * Verify that once we've encountered something other than an EFI
 * in the AIL that there are no more EFIs in the AIL.
 */
#if defined(DEBUG)
STATIC void
xlog_recover_check_ail(
	xfs_mount_t		*mp,
	xfs_log_item_t		*lip,
	int			gen)
{
	int			orig_gen = gen;

	do {
		ASSERT(lip->li_type != XFS_LI_EFI);
		lip = xfs_trans_next_ail(mp, lip, &gen, NULL);
		/*
		 * The check will be bogus if we restart from the
		 * beginning of the AIL, so ASSERT that we don't.
		 * We never should since we're holding the AIL lock
		 * the entire time.
		 */
		ASSERT(gen == orig_gen);
	} while (lip != NULL);
}
#endif	/* DEBUG */

/*
 * When this is called, all of the EFIs which did not have
 * corresponding EFDs should be in the AIL.  What we do now
 * is free the extents associated with each one.
 *
 * Since we process the EFIs in normal transactions, they
 * will be removed at some point after the commit.  This prevents
 * us from just walking down the list processing each one.
 * We'll use a flag in the EFI to skip those that we've already
 * processed and use the AIL iteration mechanism's generation
 * count to try to speed this up at least a bit.
 *
 * When we start, we know that the EFIs are the only things in
 * the AIL.  As we process them, however, other items are added
 * to the AIL.  Since everything added to the AIL must come after
 * everything already in the AIL, we stop processing as soon as
 * we see something other than an EFI in the AIL.
 */
STATIC void
xlog_recover_process_efis(
	xlog_t			*log)
{
	xfs_log_item_t		*lip;
	xfs_efi_log_item_t	*efip;
	int			gen;
	xfs_mount_t		*mp;
	SPLDECL(s);

	mp = log->l_mp;
	AIL_LOCK(mp,s);

	lip = xfs_trans_first_ail(mp, &gen);
	while (lip != NULL) {
		/*
		 * We're done when we see something other than an EFI.
		 */
		if (lip->li_type != XFS_LI_EFI) {
			xlog_recover_check_ail(mp, lip, gen);
			break;
		}

		/*
		 * Skip EFIs that we've already processed.
		 */
		efip = (xfs_efi_log_item_t *)lip;
		if (efip->efi_flags & XFS_EFI_RECOVERED) {
			lip = xfs_trans_next_ail(mp, lip, &gen, NULL);
			continue;
		}

		AIL_UNLOCK(mp, s);
		xlog_recover_process_efi(mp, efip);
		AIL_LOCK(mp,s);
		lip = xfs_trans_next_ail(mp, lip, &gen, NULL);
	}
	AIL_UNLOCK(mp, s);
}

/*
 * This routine performs a transaction to null out a bad inode pointer
 * in an agi unlinked inode hash bucket.
 */
STATIC void
xlog_recover_clear_agi_bucket(
	xfs_mount_t	*mp,
	xfs_agnumber_t	agno,
	int		bucket)
{
	xfs_trans_t	*tp;
	xfs_agi_t	*agi;
	xfs_buf_t	*agibp;
	int		offset;
	int		error;

	tp = xfs_trans_alloc(mp, XFS_TRANS_CLEAR_AGI_BUCKET);
	xfs_trans_reserve(tp, 0, XFS_CLEAR_AGI_BUCKET_LOG_RES(mp), 0, 0, 0);

	error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp,
				   XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)),
				   XFS_FSS_TO_BB(mp, 1), 0, &agibp);
	if (error) {
		xfs_trans_cancel(tp, XFS_TRANS_ABORT);
		return;
	}

	agi = XFS_BUF_TO_AGI(agibp);
	if (be32_to_cpu(agi->agi_magicnum) != XFS_AGI_MAGIC) {
		xfs_trans_cancel(tp, XFS_TRANS_ABORT);
		return;
	}

	agi->agi_unlinked[bucket] = cpu_to_be32(NULLAGINO);
	offset = offsetof(xfs_agi_t, agi_unlinked) +
		 (sizeof(xfs_agino_t) * bucket);
	xfs_trans_log_buf(tp, agibp, offset,
			  (offset + sizeof(xfs_agino_t) - 1));

	(void) xfs_trans_commit(tp, 0, NULL);
}

/*
 * xlog_iunlink_recover
 *
 * This is called during recovery to process any inodes which
 * we unlinked but not freed when the system crashed.  These
 * inodes will be on the lists in the AGI blocks.  What we do
 * here is scan all the AGIs and fully truncate and free any
 * inodes found on the lists.  Each inode is removed from the
 * lists when it has been fully truncated and is freed.  The
 * freeing of the inode and its removal from the list must be
 * atomic.
 */
void
xlog_recover_process_iunlinks(
	xlog_t		*log)
{
	xfs_mount_t	*mp;
	xfs_agnumber_t	agno;
	xfs_agi_t	*agi;
	xfs_buf_t	*agibp;
	xfs_buf_t	*ibp;
	xfs_dinode_t	*dip;
	xfs_inode_t	*ip;
	xfs_agino_t	agino;
	xfs_ino_t	ino;
	int		bucket;
	int		error;
	uint		mp_dmevmask;

	mp = log->l_mp;

	/*
	 * Prevent any DMAPI event from being sent while in this function.
	 */
	mp_dmevmask = mp->m_dmevmask;
	mp->m_dmevmask = 0;

	for (agno = 0; agno < mp->m_sb.sb_agcount; agno++) {
		/*
		 * Find the agi for this ag.
		 */
		agibp = xfs_buf_read(mp->m_ddev_targp,
				XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)),
				XFS_FSS_TO_BB(mp, 1), 0);
		if (XFS_BUF_ISERROR(agibp)) {
			xfs_ioerror_alert("xlog_recover_process_iunlinks(#1)",
				log->l_mp, agibp,
				XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)));
		}
		agi = XFS_BUF_TO_AGI(agibp);
		ASSERT(XFS_AGI_MAGIC == be32_to_cpu(agi->agi_magicnum));

		for (bucket = 0; bucket < XFS_AGI_UNLINKED_BUCKETS; bucket++) {

			agino = be32_to_cpu(agi->agi_unlinked[bucket]);
			while (agino != NULLAGINO) {

				/*
				 * Release the agi buffer so that it can
				 * be acquired in the normal course of the
				 * transaction to truncate and free the inode.
				 */
				xfs_buf_relse(agibp);

				ino = XFS_AGINO_TO_INO(mp, agno, agino);
				error = xfs_iget(mp, NULL, ino, 0, 0, &ip, 0);
				ASSERT(error || (ip != NULL));

				if (!error) {
					/*
					 * Get the on disk inode to find the
					 * next inode in the bucket.
					 */
					error = xfs_itobp(mp, NULL, ip, &dip,
							&ibp, 0, 0);
					ASSERT(error || (dip != NULL));
				}

				if (!error) {
					ASSERT(ip->i_d.di_nlink == 0);

					/* setup for the next pass */
					agino = INT_GET(dip->di_next_unlinked,
							ARCH_CONVERT);
					xfs_buf_relse(ibp);
					/*
					 * Prevent any DMAPI event from
					 * being sent when the
					 * reference on the inode is
					 * dropped.
					 */
					ip->i_d.di_dmevmask = 0;

					/*
					 * If this is a new inode, handle
					 * it specially.  Otherwise,
					 * just drop our reference to the
					 * inode.  If there are no
					 * other references, this will
					 * send the inode to
					 * xfs_inactive() which will
					 * truncate the file and free
					 * the inode.
					 */
					if (ip->i_d.di_mode == 0)
						xfs_iput_new(ip, 0);
					else
						VN_RELE(XFS_ITOV(ip));
				} else {
					/*
					 * We can't read in the inode
					 * this bucket points to, or
					 * this inode is messed up.  Just
					 * ditch this bucket of inodes.  We
					 * will lose some inodes and space,
					 * but at least we won't hang.  Call
					 * xlog_recover_clear_agi_bucket()
					 * to perform a transaction to clear
					 * the inode pointer in the bucket.
					 */
					xlog_recover_clear_agi_bucket(mp, agno,
							bucket);

					agino = NULLAGINO;
				}

				/*
				 * Reacquire the agibuffer and continue around
				 * the loop.
				 */
				agibp = xfs_buf_read(mp->m_ddev_targp,
						XFS_AG_DADDR(mp, agno,
							XFS_AGI_DADDR(mp)),
						XFS_FSS_TO_BB(mp, 1), 0);
				if (XFS_BUF_ISERROR(agibp)) {
					xfs_ioerror_alert(
				"xlog_recover_process_iunlinks(#2)",
						log->l_mp, agibp,
						XFS_AG_DADDR(mp, agno,
							XFS_AGI_DADDR(mp)));
				}
				agi = XFS_BUF_TO_AGI(agibp);
				ASSERT(XFS_AGI_MAGIC == be32_to_cpu(
					agi->agi_magicnum));
			}
		}

		/*
		 * Release the buffer for the current agi so we can
		 * go on to the next one.
		 */
		xfs_buf_relse(agibp);
	}

	mp->m_dmevmask = mp_dmevmask;
}


#ifdef DEBUG
STATIC void
xlog_pack_data_checksum(
	xlog_t		*log,
	xlog_in_core_t	*iclog,
	int		size)
{
	int		i;
	uint		*up;
	uint		chksum = 0;

	up = (uint *)iclog->ic_datap;
	/* divide length by 4 to get # words */
	for (i = 0; i < (size >> 2); i++) {
		chksum ^= INT_GET(*up, ARCH_CONVERT);
		up++;
	}
	INT_SET(iclog->ic_header.h_chksum, ARCH_CONVERT, chksum);
}
#else
#define xlog_pack_data_checksum(log, iclog, size)
#endif

/*
 * Stamp cycle number in every block
 */
void
xlog_pack_data(
	xlog_t			*log,
	xlog_in_core_t		*iclog,
	int			roundoff)
{
	int			i, j, k;
	int			size = iclog->ic_offset + roundoff;
	uint			cycle_lsn;
	xfs_caddr_t		dp;
	xlog_in_core_2_t	*xhdr;

	xlog_pack_data_checksum(log, iclog, size);

	cycle_lsn = CYCLE_LSN_DISK(iclog->ic_header.h_lsn);

	dp = iclog->ic_datap;
	for (i = 0; i < BTOBB(size) &&
		i < (XLOG_HEADER_CYCLE_SIZE / BBSIZE); i++) {
		iclog->ic_header.h_cycle_data[i] = *(uint *)dp;
		*(uint *)dp = cycle_lsn;
		dp += BBSIZE;
	}

	if (XFS_SB_VERSION_HASLOGV2(&log->l_mp->m_sb)) {
		xhdr = (xlog_in_core_2_t *)&iclog->ic_header;
		for ( ; i < BTOBB(size); i++) {
			j = i / (XLOG_HEADER_CYCLE_SIZE / BBSIZE);
			k = i % (XLOG_HEADER_CYCLE_SIZE / BBSIZE);
			xhdr[j].hic_xheader.xh_cycle_data[k] = *(uint *)dp;
			*(uint *)dp = cycle_lsn;
			dp += BBSIZE;
		}

		for (i = 1; i < log->l_iclog_heads; i++) {
			xhdr[i].hic_xheader.xh_cycle = cycle_lsn;
		}
	}
}

#if defined(DEBUG) && defined(XFS_LOUD_RECOVERY)
STATIC void
xlog_unpack_data_checksum(
	xlog_rec_header_t	*rhead,
	xfs_caddr_t		dp,
	xlog_t			*log)
{
	uint			*up = (uint *)dp;
	uint			chksum = 0;
	int			i;

	/* divide length by 4 to get # words */
	for (i=0; i < INT_GET(rhead->h_len, ARCH_CONVERT) >> 2; i++) {
		chksum ^= INT_GET(*up, ARCH_CONVERT);
		up++;
	}
	if (chksum != INT_GET(rhead->h_chksum, ARCH_CONVERT)) {
	    if (rhead->h_chksum ||
		((log->l_flags & XLOG_CHKSUM_MISMATCH) == 0)) {
		    cmn_err(CE_DEBUG,
			"XFS: LogR chksum mismatch: was (0x%x) is (0x%x)",
			    INT_GET(rhead->h_chksum, ARCH_CONVERT), chksum);
		    cmn_err(CE_DEBUG,
"XFS: Disregard message if filesystem was created with non-DEBUG kernel");
		    if (XFS_SB_VERSION_HASLOGV2(&log->l_mp->m_sb)) {
			    cmn_err(CE_DEBUG,
				"XFS: LogR this is a LogV2 filesystem");
		    }
		    log->l_flags |= XLOG_CHKSUM_MISMATCH;
	    }
	}
}
#else
#define xlog_unpack_data_checksum(rhead, dp, log)
#endif

STATIC void
xlog_unpack_data(
	xlog_rec_header_t	*rhead,
	xfs_caddr_t		dp,
	xlog_t			*log)
{
	int			i, j, k;
	xlog_in_core_2_t	*xhdr;

	for (i = 0; i < BTOBB(INT_GET(rhead->h_len, ARCH_CONVERT)) &&
		  i < (XLOG_HEADER_CYCLE_SIZE / BBSIZE); i++) {
		*(uint *)dp = *(uint *)&rhead->h_cycle_data[i];
		dp += BBSIZE;
	}

	if (XFS_SB_VERSION_HASLOGV2(&log->l_mp->m_sb)) {
		xhdr = (xlog_in_core_2_t *)rhead;
		for ( ; i < BTOBB(INT_GET(rhead->h_len, ARCH_CONVERT)); i++) {
			j = i / (XLOG_HEADER_CYCLE_SIZE / BBSIZE);
			k = i % (XLOG_HEADER_CYCLE_SIZE / BBSIZE);
			*(uint *)dp = xhdr[j].hic_xheader.xh_cycle_data[k];
			dp += BBSIZE;
		}
	}

	xlog_unpack_data_checksum(rhead, dp, log);
}

STATIC int
xlog_valid_rec_header(
	xlog_t			*log,
	xlog_rec_header_t	*rhead,
	xfs_daddr_t		blkno)
{
	int			hlen;

	if (unlikely(
	    (INT_GET(rhead->h_magicno, ARCH_CONVERT) !=
			XLOG_HEADER_MAGIC_NUM))) {
		XFS_ERROR_REPORT("xlog_valid_rec_header(1)",
				XFS_ERRLEVEL_LOW, log->l_mp);
		return XFS_ERROR(EFSCORRUPTED);
	}
	if (unlikely(
	    (!rhead->h_version ||
	    (INT_GET(rhead->h_version, ARCH_CONVERT) &
			(~XLOG_VERSION_OKBITS)) != 0))) {
		xlog_warn("XFS: %s: unrecognised log version (%d).",
			__FUNCTION__, INT_GET(rhead->h_version, ARCH_CONVERT));
		return XFS_ERROR(EIO);
	}

	/* LR body must have data or it wouldn't have been written */
	hlen = INT_GET(rhead->h_len, ARCH_CONVERT);
	if (unlikely( hlen <= 0 || hlen > INT_MAX )) {
		XFS_ERROR_REPORT("xlog_valid_rec_header(2)",
				XFS_ERRLEVEL_LOW, log->l_mp);
		return XFS_ERROR(EFSCORRUPTED);
	}
	if (unlikely( blkno > log->l_logBBsize || blkno > INT_MAX )) {
		XFS_ERROR_REPORT("xlog_valid_rec_header(3)",
				XFS_ERRLEVEL_LOW, log->l_mp);
		return XFS_ERROR(EFSCORRUPTED);
	}
	return 0;
}

/*
 * Read the log from tail to head and process the log records found.
 * Handle the two cases where the tail and head are in the same cycle
 * and where the active portion of the log wraps around the end of
 * the physical log separately.  The pass parameter is passed through
 * to the routines called to process the data and is not looked at
 * here.
 */
STATIC int
xlog_do_recovery_pass(
	xlog_t			*log,
	xfs_daddr_t		head_blk,
	xfs_daddr_t		tail_blk,
	int			pass)
{
	xlog_rec_header_t	*rhead;
	xfs_daddr_t		blk_no;
	xfs_caddr_t		bufaddr, offset;
	xfs_buf_t		*hbp, *dbp;
	int			error = 0, h_size;
	int			bblks, split_bblks;
	int			hblks, split_hblks, wrapped_hblks;
	xlog_recover_t		*rhash[XLOG_RHASH_SIZE];

	ASSERT(head_blk != tail_blk);

	/*
	 * Read the header of the tail block and get the iclog buffer size from
	 * h_size.  Use this to tell how many sectors make up the log header.
	 */
	if (XFS_SB_VERSION_HASLOGV2(&log->l_mp->m_sb)) {
		/*
		 * When using variable length iclogs, read first sector of
		 * iclog header and extract the header size from it.  Get a
		 * new hbp that is the correct size.
		 */
		hbp = xlog_get_bp(log, 1);
		if (!hbp)
			return ENOMEM;
		if ((error = xlog_bread(log, tail_blk, 1, hbp)))
			goto bread_err1;
		offset = xlog_align(log, tail_blk, 1, hbp);
		rhead = (xlog_rec_header_t *)offset;
		error = xlog_valid_rec_header(log, rhead, tail_blk);
		if (error)
			goto bread_err1;
		h_size = INT_GET(rhead->h_size, ARCH_CONVERT);
		if ((INT_GET(rhead->h_version, ARCH_CONVERT)
				& XLOG_VERSION_2) &&
		    (h_size > XLOG_HEADER_CYCLE_SIZE)) {
			hblks = h_size / XLOG_HEADER_CYCLE_SIZE;
			if (h_size % XLOG_HEADER_CYCLE_SIZE)
				hblks++;
			xlog_put_bp(hbp);
			hbp = xlog_get_bp(log, hblks);
		} else {
			hblks = 1;
		}
	} else {
		ASSERT(log->l_sectbb_log == 0);
		hblks = 1;
		hbp = xlog_get_bp(log, 1);
		h_size = XLOG_BIG_RECORD_BSIZE;
	}

	if (!hbp)
		return ENOMEM;
	dbp = xlog_get_bp(log, BTOBB(h_size));
	if (!dbp) {
		xlog_put_bp(hbp);
		return ENOMEM;
	}

	memset(rhash, 0, sizeof(rhash));
	if (tail_blk <= head_blk) {
		for (blk_no = tail_blk; blk_no < head_blk; ) {
			if ((error = xlog_bread(log, blk_no, hblks, hbp)))
				goto bread_err2;
			offset = xlog_align(log, blk_no, hblks, hbp);
			rhead = (xlog_rec_header_t *)offset;
			error = xlog_valid_rec_header(log, rhead, blk_no);
			if (error)
				goto bread_err2;

			/* blocks in data section */
			bblks = (int)BTOBB(INT_GET(rhead->h_len, ARCH_CONVERT));
			error = xlog_bread(log, blk_no + hblks, bblks, dbp);
			if (error)
				goto bread_err2;
			offset = xlog_align(log, blk_no + hblks, bblks, dbp);
			xlog_unpack_data(rhead, offset, log);
			if ((error = xlog_recover_process_data(log,
						rhash, rhead, offset, pass)))
				goto bread_err2;
			blk_no += bblks + hblks;
		}
	} else {
		/*
		 * Perform recovery around the end of the physical log.
		 * When the head is not on the same cycle number as the tail,
		 * we can't do a sequential recovery as above.
		 */
		blk_no = tail_blk;
		while (blk_no < log->l_logBBsize) {
			/*
			 * Check for header wrapping around physical end-of-log
			 */
			offset = NULL;
			split_hblks = 0;
			wrapped_hblks = 0;
			if (blk_no + hblks <= log->l_logBBsize) {
				/* Read header in one read */
				error = xlog_bread(log, blk_no, hblks, hbp);
				if (error)
					goto bread_err2;
				offset = xlog_align(log, blk_no, hblks, hbp);
			} else {
				/* This LR is split across physical log end */
				if (blk_no != log->l_logBBsize) {
					/* some data before physical log end */
					ASSERT(blk_no <= INT_MAX);
					split_hblks = log->l_logBBsize - (int)blk_no;
					ASSERT(split_hblks > 0);
					if ((error = xlog_bread(log, blk_no,
							split_hblks, hbp)))
						goto bread_err2;
					offset = xlog_align(log, blk_no,
							split_hblks, hbp);
				}
				/*
				 * Note: this black magic still works with
				 * large sector sizes (non-512) only because:
				 * - we increased the buffer size originally
				 *   by 1 sector giving us enough extra space
				 *   for the second read;
				 * - the log start is guaranteed to be sector
				 *   aligned;
				 * - we read the log end (LR header start)
				 *   _first_, then the log start (LR header end)
				 *   - order is important.
				 */
				bufaddr = XFS_BUF_PTR(hbp);
				XFS_BUF_SET_PTR(hbp,
						bufaddr + BBTOB(split_hblks),
						BBTOB(hblks - split_hblks));
				wrapped_hblks = hblks - split_hblks;
				error = xlog_bread(log, 0, wrapped_hblks, hbp);
				if (error)
					goto bread_err2;
				XFS_BUF_SET_PTR(hbp, bufaddr, BBTOB(hblks));
				if (!offset)
					offset = xlog_align(log, 0,
							wrapped_hblks, hbp);
			}
			rhead = (xlog_rec_header_t *)offset;
			error = xlog_valid_rec_header(log, rhead,
						split_hblks ? blk_no : 0);
			if (error)
				goto bread_err2;

			bblks = (int)BTOBB(INT_GET(rhead->h_len, ARCH_CONVERT));
			blk_no += hblks;

			/* Read in data for log record */
			if (blk_no + bblks <= log->l_logBBsize) {
				error = xlog_bread(log, blk_no, bblks, dbp);
				if (error)
					goto bread_err2;
				offset = xlog_align(log, blk_no, bblks, dbp);
			} else {
				/* This log record is split across the
				 * physical end of log */
				offset = NULL;
				split_bblks = 0;
				if (blk_no != log->l_logBBsize) {
					/* some data is before the physical
					 * end of log */
					ASSERT(!wrapped_hblks);
					ASSERT(blk_no <= INT_MAX);
					split_bblks =
						log->l_logBBsize - (int)blk_no;
					ASSERT(split_bblks > 0);
					if ((error = xlog_bread(log, blk_no,
							split_bblks, dbp)))
						goto bread_err2;
					offset = xlog_align(log, blk_no,
							split_bblks, dbp);
				}
				/*
				 * Note: this black magic still works with
				 * large sector sizes (non-512) only because:
				 * - we increased the buffer size originally
				 *   by 1 sector giving us enough extra space
				 *   for the second read;
				 * - the log start is guaranteed to be sector
				 *   aligned;
				 * - we read the log end (LR header start)
				 *   _first_, then the log start (LR header end)
				 *   - order is important.
				 */
				bufaddr = XFS_BUF_PTR(dbp);
				XFS_BUF_SET_PTR(dbp,
						bufaddr + BBTOB(split_bblks),
						BBTOB(bblks - split_bblks));
				if ((error = xlog_bread(log, wrapped_hblks,
						bblks - split_bblks, dbp)))
					goto bread_err2;
				XFS_BUF_SET_PTR(dbp, bufaddr, h_size);
				if (!offset)
					offset = xlog_align(log, wrapped_hblks,
						bblks - split_bblks, dbp);
			}
			xlog_unpack_data(rhead, offset, log);
			if ((error = xlog_recover_process_data(log, rhash,
							rhead, offset, pass)))
				goto bread_err2;
			blk_no += bblks;
		}

		ASSERT(blk_no >= log->l_logBBsize);
		blk_no -= log->l_logBBsize;

		/* read first part of physical log */
		while (blk_no < head_blk) {
			if ((error = xlog_bread(log, blk_no, hblks, hbp)))
				goto bread_err2;
			offset = xlog_align(log, blk_no, hblks, hbp);
			rhead = (xlog_rec_header_t *)offset;
			error = xlog_valid_rec_header(log, rhead, blk_no);
			if (error)
				goto bread_err2;
			bblks = (int)BTOBB(INT_GET(rhead->h_len, ARCH_CONVERT));
			if ((error = xlog_bread(log, blk_no+hblks, bblks, dbp)))
				goto bread_err2;
			offset = xlog_align(log, blk_no+hblks, bblks, dbp);
			xlog_unpack_data(rhead, offset, log);
			if ((error = xlog_recover_process_data(log, rhash,
							rhead, offset, pass)))
				goto bread_err2;
			blk_no += bblks + hblks;
		}
	}

 bread_err2:
	xlog_put_bp(dbp);
 bread_err1:
	xlog_put_bp(hbp);
	return error;
}

/*
 * Do the recovery of the log.  We actually do this in two phases.
 * The two passes are necessary in order to implement the function
 * of cancelling a record written into the log.  The first pass
 * determines those things which have been cancelled, and the
 * second pass replays log items normally except for those which
 * have been cancelled.  The handling of the replay and cancellations
 * takes place in the log item type specific routines.
 *
 * The table of items which have cancel records in the log is allocated
 * and freed at this level, since only here do we know when all of
 * the log recovery has been completed.
 */
STATIC int
xlog_do_log_recovery(
	xlog_t		*log,
	xfs_daddr_t	head_blk,
	xfs_daddr_t	tail_blk)
{
	int		error;

	ASSERT(head_blk != tail_blk);

	/*
	 * First do a pass to find all of the cancelled buf log items.
	 * Store them in the buf_cancel_table for use in the second pass.
	 */
	log->l_buf_cancel_table =
		(xfs_buf_cancel_t **)kmem_zalloc(XLOG_BC_TABLE_SIZE *
						 sizeof(xfs_buf_cancel_t*),
						 KM_SLEEP);
	error = xlog_do_recovery_pass(log, head_blk, tail_blk,
				      XLOG_RECOVER_PASS1);
	if (error != 0) {
		kmem_free(log->l_buf_cancel_table,
			  XLOG_BC_TABLE_SIZE * sizeof(xfs_buf_cancel_t*));
		log->l_buf_cancel_table = NULL;
		return error;
	}
	/*
	 * Then do a second pass to actually recover the items in the log.
	 * When it is complete free the table of buf cancel items.
	 */
	error = xlog_do_recovery_pass(log, head_blk, tail_blk,
				      XLOG_RECOVER_PASS2);
#ifdef DEBUG
	{
		int	i;

		for (i = 0; i < XLOG_BC_TABLE_SIZE; i++)
			ASSERT(log->l_buf_cancel_table[i] == NULL);
	}
#endif	/* DEBUG */

	kmem_free(log->l_buf_cancel_table,
		  XLOG_BC_TABLE_SIZE * sizeof(xfs_buf_cancel_t*));
	log->l_buf_cancel_table = NULL;

	return error;
}

/*
 * Do the actual recovery
 */
STATIC int
xlog_do_recover(
	xlog_t		*log,
	xfs_daddr_t	head_blk,
	xfs_daddr_t	tail_blk)
{
	int		error;
	xfs_buf_t	*bp;
	xfs_sb_t	*sbp;

	/*
	 * First replay the images in the log.
	 */
	error = xlog_do_log_recovery(log, head_blk, tail_blk);
	if (error) {
		return error;
	}

	XFS_bflush(log->l_mp->m_ddev_targp);

	/*
	 * If IO errors happened during recovery, bail out.
	 */
	if (XFS_FORCED_SHUTDOWN(log->l_mp)) {
		return (EIO);
	}

	/*
	 * We now update the tail_lsn since much of the recovery has completed
	 * and there may be space available to use.  If there were no extent
	 * or iunlinks, we can free up the entire log and set the tail_lsn to
	 * be the last_sync_lsn.  This was set in xlog_find_tail to be the
	 * lsn of the last known good LR on disk.  If there are extent frees
	 * or iunlinks they will have some entries in the AIL; so we look at
	 * the AIL to determine how to set the tail_lsn.
	 */
	xlog_assign_tail_lsn(log->l_mp);

	/*
	 * Now that we've finished replaying all buffer and inode
	 * updates, re-read in the superblock.
	 */
	bp = xfs_getsb(log->l_mp, 0);
	XFS_BUF_UNDONE(bp);
	XFS_BUF_READ(bp);
	xfsbdstrat(log->l_mp, bp);
	if ((error = xfs_iowait(bp))) {
		xfs_ioerror_alert("xlog_do_recover",
				  log->l_mp, bp, XFS_BUF_ADDR(bp));
		ASSERT(0);
		xfs_buf_relse(bp);
		return error;
	}

	/* Convert superblock from on-disk format */
	sbp = &log->l_mp->m_sb;
	xfs_xlatesb(XFS_BUF_TO_SBP(bp), sbp, 1, XFS_SB_ALL_BITS);
	ASSERT(sbp->sb_magicnum == XFS_SB_MAGIC);
	ASSERT(XFS_SB_GOOD_VERSION(sbp));
	xfs_buf_relse(bp);

	xlog_recover_check_summary(log);

	/* Normal transactions can now occur */
	log->l_flags &= ~XLOG_ACTIVE_RECOVERY;
	return 0;
}

/*
 * Perform recovery and re-initialize some log variables in xlog_find_tail.
 *
 * Return error or zero.
 */
int
xlog_recover(
	xlog_t		*log)
{
	xfs_daddr_t	head_blk, tail_blk;
	int		error;

	/* find the tail of the log */
	if ((error = xlog_find_tail(log, &head_blk, &tail_blk)))
		return error;

	if (tail_blk != head_blk) {
		/* There used to be a comment here:
		 *
		 * disallow recovery on read-only mounts.  note -- mount
		 * checks for ENOSPC and turns it into an intelligent
		 * error message.
		 * ...but this is no longer true.  Now, unless you specify
		 * NORECOVERY (in which case this function would never be
		 * called), we just go ahead and recover.  We do this all
		 * under the vfs layer, so we can get away with it unless
		 * the device itself is read-only, in which case we fail.
		 */
		if ((error = xfs_dev_is_read_only(log->l_mp,
						"recovery required"))) {
			return error;
		}

		cmn_err(CE_NOTE,
			"Starting XFS recovery on filesystem: %s (logdev: %s)",
			log->l_mp->m_fsname, log->l_mp->m_logname ?
			log->l_mp->m_logname : "internal");

		error = xlog_do_recover(log, head_blk, tail_blk);
		log->l_flags |= XLOG_RECOVERY_NEEDED;
	}
	return error;
}

/*
 * In the first part of recovery we replay inodes and buffers and build
 * up the list of extent free items which need to be processed.  Here
 * we process the extent free items and clean up the on disk unlinked
 * inode lists.  This is separated from the first part of recovery so
 * that the root and real-time bitmap inodes can be read in from disk in
 * between the two stages.  This is necessary so that we can free space
 * in the real-time portion of the file system.
 */
int
xlog_recover_finish(
	xlog_t		*log,
	int		mfsi_flags)
{
	/*
	 * Now we're ready to do the transactions needed for the
	 * rest of recovery.  Start with completing all the extent
	 * free intent records and then process the unlinked inode
	 * lists.  At this point, we essentially run in normal mode
	 * except that we're still performing recovery actions
	 * rather than accepting new requests.
	 */
	if (log->l_flags & XLOG_RECOVERY_NEEDED) {
		xlog_recover_process_efis(log);
		/*
		 * Sync the log to get all the EFIs out of the AIL.
		 * This isn't absolutely necessary, but it helps in
		 * case the unlink transactions would have problems
		 * pushing the EFIs out of the way.
		 */
		xfs_log_force(log->l_mp, (xfs_lsn_t)0,
			      (XFS_LOG_FORCE | XFS_LOG_SYNC));

		if ( (mfsi_flags & XFS_MFSI_NOUNLINK) == 0 ) {
			xlog_recover_process_iunlinks(log);
		}

		xlog_recover_check_summary(log);

		cmn_err(CE_NOTE,
			"Ending XFS recovery on filesystem: %s (logdev: %s)",
			log->l_mp->m_fsname, log->l_mp->m_logname ?
			log->l_mp->m_logname : "internal");
		log->l_flags &= ~XLOG_RECOVERY_NEEDED;
	} else {
		cmn_err(CE_DEBUG,
			"!Ending clean XFS mount for filesystem: %s",
			log->l_mp->m_fsname);
	}
	return 0;
}


#if defined(DEBUG)
/*
 * Read all of the agf and agi counters and check that they
 * are consistent with the superblock counters.
 */
void
xlog_recover_check_summary(
	xlog_t		*log)
{
	xfs_mount_t	*mp;
	xfs_agf_t	*agfp;
	xfs_agi_t	*agip;
	xfs_buf_t	*agfbp;
	xfs_buf_t	*agibp;
	xfs_daddr_t	agfdaddr;
	xfs_daddr_t	agidaddr;
	xfs_buf_t	*sbbp;
#ifdef XFS_LOUD_RECOVERY
	xfs_sb_t	*sbp;
#endif
	xfs_agnumber_t	agno;
	__uint64_t	freeblks;
	__uint64_t	itotal;
	__uint64_t	ifree;

	mp = log->l_mp;

	freeblks = 0LL;
	itotal = 0LL;
	ifree = 0LL;
	for (agno = 0; agno < mp->m_sb.sb_agcount; agno++) {
		agfdaddr = XFS_AG_DADDR(mp, agno, XFS_AGF_DADDR(mp));
		agfbp = xfs_buf_read(mp->m_ddev_targp, agfdaddr,
				XFS_FSS_TO_BB(mp, 1), 0);
		if (XFS_BUF_ISERROR(agfbp)) {
			xfs_ioerror_alert("xlog_recover_check_summary(agf)",
						mp, agfbp, agfdaddr);
		}
		agfp = XFS_BUF_TO_AGF(agfbp);
		ASSERT(XFS_AGF_MAGIC == be32_to_cpu(agfp->agf_magicnum));
		ASSERT(XFS_AGF_GOOD_VERSION(be32_to_cpu(agfp->agf_versionnum)));
		ASSERT(be32_to_cpu(agfp->agf_seqno) == agno);

		freeblks += be32_to_cpu(agfp->agf_freeblks) +
			    be32_to_cpu(agfp->agf_flcount);
		xfs_buf_relse(agfbp);

		agidaddr = XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp));
		agibp = xfs_buf_read(mp->m_ddev_targp, agidaddr,
				XFS_FSS_TO_BB(mp, 1), 0);
		if (XFS_BUF_ISERROR(agibp)) {
			xfs_ioerror_alert("xlog_recover_check_summary(agi)",
					  mp, agibp, agidaddr);
		}
		agip = XFS_BUF_TO_AGI(agibp);
		ASSERT(XFS_AGI_MAGIC == be32_to_cpu(agip->agi_magicnum));
		ASSERT(XFS_AGI_GOOD_VERSION(be32_to_cpu(agip->agi_versionnum)));
		ASSERT(be32_to_cpu(agip->agi_seqno) == agno);

		itotal += be32_to_cpu(agip->agi_count);
		ifree += be32_to_cpu(agip->agi_freecount);
		xfs_buf_relse(agibp);
	}

	sbbp = xfs_getsb(mp, 0);
#ifdef XFS_LOUD_RECOVERY
	sbp = &mp->m_sb;
	xfs_xlatesb(XFS_BUF_TO_SBP(sbbp), sbp, 1, XFS_SB_ALL_BITS);
	cmn_err(CE_NOTE,
		"xlog_recover_check_summary: sb_icount %Lu itotal %Lu",
		sbp->sb_icount, itotal);
	cmn_err(CE_NOTE,
		"xlog_recover_check_summary: sb_ifree %Lu itotal %Lu",
		sbp->sb_ifree, ifree);
	cmn_err(CE_NOTE,
		"xlog_recover_check_summary: sb_fdblocks %Lu freeblks %Lu",
		sbp->sb_fdblocks, freeblks);
#if 0
	/*
	 * This is turned off until I account for the allocation
	 * btree blocks which live in free space.
	 */
	ASSERT(sbp->sb_icount == itotal);
	ASSERT(sbp->sb_ifree == ifree);
	ASSERT(sbp->sb_fdblocks == freeblks);
#endif
#endif
	xfs_buf_relse(sbbp);
}
#endif /* DEBUG */