Training courses

Kernel and Embedded Linux

Bootlin training courses

Embedded Linux, kernel,
Yocto Project, Buildroot, real-time,
graphics, boot time, debugging...

Bootlin logo

Elixir Cross Referencer

   1
   2
   3
   4
   5
   6
   7
   8
   9
  10
  11
  12
  13
  14
  15
  16
  17
  18
  19
  20
  21
  22
  23
  24
  25
  26
  27
  28
  29
  30
  31
  32
  33
  34
  35
  36
  37
  38
  39
  40
  41
  42
  43
  44
  45
  46
  47
  48
  49
  50
  51
  52
  53
  54
  55
  56
  57
  58
  59
  60
  61
  62
  63
  64
  65
  66
  67
  68
  69
  70
  71
  72
  73
  74
  75
  76
  77
  78
  79
  80
  81
  82
  83
  84
  85
  86
  87
  88
  89
  90
  91
  92
  93
  94
  95
  96
  97
  98
  99
 100
 101
 102
 103
 104
 105
 106
 107
 108
 109
 110
 111
 112
 113
 114
 115
 116
 117
 118
 119
 120
 121
 122
 123
 124
 125
 126
 127
 128
 129
 130
 131
 132
 133
 134
 135
 136
 137
 138
 139
 140
 141
 142
 143
 144
 145
 146
 147
 148
 149
 150
 151
 152
 153
 154
 155
 156
 157
 158
 159
 160
 161
 162
 163
 164
 165
 166
 167
 168
 169
 170
 171
 172
 173
 174
 175
 176
 177
 178
 179
 180
 181
 182
 183
 184
 185
 186
 187
 188
 189
 190
 191
 192
 193
 194
 195
 196
 197
 198
 199
 200
 201
 202
 203
 204
 205
 206
 207
 208
 209
 210
 211
 212
 213
 214
 215
 216
 217
 218
 219
 220
 221
 222
 223
 224
 225
 226
 227
 228
 229
 230
 231
 232
 233
 234
 235
 236
 237
 238
 239
 240
 241
 242
 243
 244
 245
 246
 247
 248
 249
 250
 251
 252
 253
 254
 255
 256
 257
 258
 259
 260
 261
 262
 263
 264
 265
 266
 267
 268
 269
 270
 271
 272
 273
 274
 275
 276
 277
 278
 279
 280
 281
 282
 283
 284
 285
 286
 287
 288
 289
 290
 291
 292
 293
 294
 295
 296
 297
 298
 299
 300
 301
 302
 303
 304
 305
 306
 307
 308
 309
 310
 311
 312
 313
 314
 315
 316
 317
 318
 319
 320
 321
 322
 323
 324
 325
 326
 327
 328
 329
 330
 331
 332
 333
 334
 335
 336
 337
 338
 339
 340
 341
 342
 343
 344
 345
 346
 347
 348
 349
 350
 351
 352
 353
 354
 355
 356
 357
 358
 359
 360
 361
 362
 363
 364
 365
 366
 367
 368
 369
 370
 371
 372
 373
 374
 375
 376
 377
 378
 379
 380
 381
 382
 383
 384
 385
 386
 387
 388
 389
 390
 391
 392
 393
 394
 395
 396
 397
 398
 399
 400
 401
 402
 403
 404
 405
 406
 407
 408
 409
 410
 411
 412
 413
 414
 415
 416
 417
 418
 419
 420
 421
 422
 423
 424
 425
 426
 427
 428
 429
 430
 431
 432
 433
 434
 435
 436
 437
 438
 439
 440
 441
 442
 443
 444
 445
 446
 447
 448
 449
 450
 451
 452
 453
 454
 455
 456
 457
 458
 459
 460
 461
 462
 463
 464
 465
 466
 467
 468
 469
 470
 471
 472
 473
 474
 475
 476
 477
 478
 479
 480
 481
 482
 483
 484
 485
 486
 487
 488
 489
 490
 491
 492
 493
 494
 495
 496
 497
 498
 499
 500
 501
 502
 503
 504
 505
 506
 507
 508
 509
 510
 511
 512
 513
 514
 515
 516
 517
 518
 519
 520
 521
 522
 523
 524
 525
 526
 527
 528
 529
 530
 531
 532
 533
 534
 535
 536
 537
 538
 539
 540
 541
 542
 543
 544
 545
 546
 547
 548
 549
 550
 551
 552
 553
 554
 555
 556
 557
 558
 559
 560
 561
 562
 563
 564
 565
 566
 567
 568
 569
 570
 571
 572
 573
 574
 575
 576
 577
 578
 579
 580
 581
 582
 583
 584
 585
 586
 587
 588
 589
 590
 591
 592
 593
 594
 595
 596
 597
 598
 599
 600
 601
 602
 603
 604
 605
 606
 607
 608
 609
 610
 611
 612
 613
 614
 615
 616
 617
 618
 619
 620
 621
 622
 623
 624
 625
 626
 627
 628
 629
 630
 631
 632
 633
 634
 635
 636
 637
 638
 639
 640
 641
 642
 643
 644
 645
 646
 647
 648
 649
 650
 651
 652
 653
 654
 655
 656
 657
 658
 659
 660
 661
 662
 663
 664
 665
 666
 667
 668
 669
 670
 671
 672
 673
 674
 675
 676
 677
 678
 679
 680
 681
 682
 683
 684
 685
 686
 687
 688
 689
 690
 691
 692
 693
 694
 695
 696
 697
 698
 699
 700
 701
 702
 703
 704
 705
 706
 707
 708
 709
 710
 711
 712
 713
 714
 715
 716
 717
 718
 719
 720
 721
 722
 723
 724
 725
 726
 727
 728
 729
 730
 731
 732
 733
 734
 735
 736
 737
 738
 739
 740
 741
 742
 743
 744
 745
 746
 747
 748
 749
 750
 751
 752
 753
 754
 755
 756
 757
 758
 759
 760
 761
 762
 763
 764
 765
 766
 767
 768
 769
 770
 771
 772
 773
 774
 775
 776
 777
 778
 779
 780
 781
 782
 783
 784
 785
 786
 787
 788
 789
 790
 791
 792
 793
 794
 795
 796
 797
 798
 799
 800
 801
 802
 803
 804
 805
 806
 807
 808
 809
 810
 811
 812
 813
 814
 815
 816
 817
 818
 819
 820
 821
 822
 823
 824
 825
 826
 827
 828
 829
 830
 831
 832
 833
 834
 835
 836
 837
 838
 839
 840
 841
 842
 843
 844
 845
 846
 847
 848
 849
 850
 851
 852
 853
 854
 855
 856
 857
 858
 859
 860
 861
 862
 863
 864
 865
 866
 867
 868
 869
 870
 871
 872
 873
 874
 875
 876
 877
 878
 879
 880
 881
 882
 883
 884
 885
 886
 887
 888
 889
 890
 891
 892
 893
 894
 895
 896
 897
 898
 899
 900
 901
 902
 903
 904
 905
 906
 907
 908
 909
 910
 911
 912
 913
 914
 915
 916
 917
 918
 919
 920
 921
 922
 923
 924
 925
 926
 927
 928
 929
 930
 931
 932
 933
 934
 935
 936
 937
 938
 939
 940
 941
 942
 943
 944
 945
 946
 947
 948
 949
 950
 951
 952
 953
 954
 955
 956
 957
 958
 959
 960
 961
 962
 963
 964
 965
 966
 967
 968
 969
 970
 971
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
/*	$NetBSD: lfs_vfsops.c,v 1.365.2.1 2020/08/17 10:30:22 martin Exp $	*/

/*-
 * Copyright (c) 1999, 2000, 2001, 2002, 2003, 2007, 2007
 *     The NetBSD Foundation, Inc.
 * All rights reserved.
 *
 * This code is derived from software contributed to The NetBSD Foundation
 * by Konrad E. Schroder <perseant@hhhh.org>.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 *
 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
 * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 */
/*-
 * Copyright (c) 1989, 1991, 1993, 1994
 *	The Regents of the University of California.  All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. Neither the name of the University nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 *
 *	@(#)lfs_vfsops.c	8.20 (Berkeley) 6/10/95
 */

#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: lfs_vfsops.c,v 1.365.2.1 2020/08/17 10:30:22 martin Exp $");

#if defined(_KERNEL_OPT)
#include "opt_lfs.h"
#include "opt_quota.h"
#endif

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/namei.h>
#include <sys/proc.h>
#include <sys/kernel.h>
#include <sys/vnode.h>
#include <sys/mount.h>
#include <sys/kthread.h>
#include <sys/buf.h>
#include <sys/device.h>
#include <sys/file.h>
#include <sys/disklabel.h>
#include <sys/ioctl.h>
#include <sys/errno.h>
#include <sys/malloc.h>
#include <sys/pool.h>
#include <sys/socket.h>
#include <sys/syslog.h>
#include <uvm/uvm_extern.h>
#include <sys/sysctl.h>
#include <sys/conf.h>
#include <sys/kauth.h>
#include <sys/module.h>
#include <sys/syscallvar.h>
#include <sys/syscall.h>
#include <sys/syscallargs.h>

#include <miscfs/specfs/specdev.h>

#include <ufs/lfs/ulfs_quotacommon.h>
#include <ufs/lfs/ulfs_inode.h>
#include <ufs/lfs/ulfsmount.h>
#include <ufs/lfs/ulfs_bswap.h>
#include <ufs/lfs/ulfs_extern.h>

#include <uvm/uvm.h>
#include <uvm/uvm_stat.h>
#include <uvm/uvm_pager.h>
#include <uvm/uvm_pdaemon.h>

#include <ufs/lfs/lfs.h>
#include <ufs/lfs/lfs_accessors.h>
#include <ufs/lfs/lfs_kernel.h>
#include <ufs/lfs/lfs_extern.h>

#include <miscfs/genfs/genfs.h>
#include <miscfs/genfs/genfs_node.h>

MODULE(MODULE_CLASS_VFS, lfs, NULL);

static int lfs_gop_write(struct vnode *, struct vm_page **, int, int);
static int lfs_mountfs(struct vnode *, struct mount *, struct lwp *);
static int lfs_flushfiles(struct mount *, int);

static struct sysctllog *lfs_sysctl_log;

extern const struct vnodeopv_desc lfs_vnodeop_opv_desc;
extern const struct vnodeopv_desc lfs_specop_opv_desc;
extern const struct vnodeopv_desc lfs_fifoop_opv_desc;

struct lwp * lfs_writer_daemon = NULL;
kcondvar_t lfs_writerd_cv;

int lfs_do_flush = 0;
#ifdef LFS_KERNEL_RFW
int lfs_do_rfw = 0;
#endif

const struct vnodeopv_desc * const lfs_vnodeopv_descs[] = {
	&lfs_vnodeop_opv_desc,
	&lfs_specop_opv_desc,
	&lfs_fifoop_opv_desc,
	NULL,
};

struct vfsops lfs_vfsops = {
	.vfs_name = MOUNT_LFS,
	.vfs_min_mount_data = sizeof (struct ulfs_args),
	.vfs_mount = lfs_mount,
	.vfs_start = ulfs_start,
	.vfs_unmount = lfs_unmount,
	.vfs_root = ulfs_root,
	.vfs_quotactl = ulfs_quotactl,
	.vfs_statvfs = lfs_statvfs,
	.vfs_sync = lfs_sync,
	.vfs_vget = lfs_vget,
	.vfs_loadvnode = lfs_loadvnode,
	.vfs_newvnode = lfs_newvnode,
	.vfs_fhtovp = lfs_fhtovp,
	.vfs_vptofh = lfs_vptofh,
	.vfs_init = lfs_init,
	.vfs_reinit = lfs_reinit,
	.vfs_done = lfs_done,
	.vfs_mountroot = lfs_mountroot,
	.vfs_snapshot = (void *)eopnotsupp,
	.vfs_extattrctl = lfs_extattrctl,
	.vfs_suspendctl = genfs_suspendctl,
	.vfs_renamelock_enter = genfs_renamelock_enter,
	.vfs_renamelock_exit = genfs_renamelock_exit,
	.vfs_fsync = (void *)eopnotsupp,
	.vfs_opv_descs = lfs_vnodeopv_descs
};

const struct genfs_ops lfs_genfsops = {
	.gop_size = lfs_gop_size,
	.gop_alloc = ulfs_gop_alloc,
	.gop_write = lfs_gop_write,
	.gop_markupdate = ulfs_gop_markupdate,
	.gop_putrange = genfs_gop_putrange,
};

struct shortlong {
	const char *sname;
	const char *lname;
};

static int
sysctl_lfs_dostats(SYSCTLFN_ARGS)
{
	extern struct lfs_stats lfs_stats;
	extern int lfs_dostats;
	int error;

	error = sysctl_lookup(SYSCTLFN_CALL(rnode));
	if (error || newp == NULL)
		return (error);

	if (lfs_dostats == 0)
		memset(&lfs_stats, 0, sizeof(lfs_stats));

	return (0);
}

static void
lfs_sysctl_setup(struct sysctllog **clog)
{
	int i;
	extern int lfs_writeindir, lfs_dostats, lfs_clean_vnhead,
		   lfs_fs_pagetrip, lfs_ignore_lazy_sync;
#ifdef DEBUG
	extern int lfs_debug_log_subsys[DLOG_MAX];
	struct shortlong dlog_names[DLOG_MAX] = { /* Must match lfs.h ! */
		{ "rollforward", "Debug roll-forward code" },
		{ "alloc",	"Debug inode allocation and free list" },
		{ "avail",	"Debug space-available-now accounting" },
		{ "flush",	"Debug flush triggers" },
		{ "lockedlist",	"Debug locked list accounting" },
		{ "vnode_verbose", "Verbose per-vnode-written debugging" },
		{ "vnode",	"Debug vnode use during segment write" },
		{ "segment",	"Debug segment writing" },
		{ "seguse",	"Debug segment used-bytes accounting" },
		{ "cleaner",	"Debug cleaning routines" },
		{ "mount",	"Debug mount/unmount routines" },
		{ "pagecache",	"Debug UBC interactions" },
		{ "dirop",	"Debug directory-operation accounting" },
		{ "malloc",	"Debug private malloc accounting" },
	};
#endif /* DEBUG */
	struct shortlong stat_names[] = { /* Must match lfs.h! */
		{ "segsused",	    "Number of new segments allocated" },
		{ "psegwrites",	    "Number of partial-segment writes" },
		{ "psyncwrites",    "Number of synchronous partial-segment"
				    " writes" },
		{ "pcleanwrites",   "Number of partial-segment writes by the"
				    " cleaner" },
		{ "blocktot",       "Number of blocks written" },
		{ "cleanblocks",    "Number of blocks written by the cleaner" },
		{ "ncheckpoints",   "Number of checkpoints made" },
		{ "nwrites",        "Number of whole writes" },
		{ "nsync_writes",   "Number of synchronous writes" },
		{ "wait_exceeded",  "Number of times writer waited for"
				    " cleaner" },
		{ "write_exceeded", "Number of times writer invoked flush" },
		{ "flush_invoked",  "Number of times flush was invoked" },
		{ "vflush_invoked", "Number of time vflush was called" },
		{ "clean_inlocked", "Number of vnodes skipped for being dead" },
		{ "clean_vnlocked", "Number of vnodes skipped for vget failure" },
		{ "segs_reclaimed", "Number of segments reclaimed" },
	};

	sysctl_createv(clog, 0, NULL, NULL,
		       CTLFLAG_PERMANENT,
		       CTLTYPE_NODE, "lfs",
		       SYSCTL_DESCR("Log-structured file system"),
		       NULL, 0, NULL, 0,
		       CTL_VFS, 5, CTL_EOL);
	/*
	 * XXX the "5" above could be dynamic, thereby eliminating one
	 * more instance of the "number to vfs" mapping problem, but
	 * "5" is the order as taken from sys/mount.h
	 */

	sysctl_createv(clog, 0, NULL, NULL,
		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
		       CTLTYPE_INT, "flushindir", NULL,
		       NULL, 0, &lfs_writeindir, 0,
		       CTL_VFS, 5, LFS_WRITEINDIR, CTL_EOL);
	sysctl_createv(clog, 0, NULL, NULL,
		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
		       CTLTYPE_INT, "clean_vnhead", NULL,
		       NULL, 0, &lfs_clean_vnhead, 0,
		       CTL_VFS, 5, LFS_CLEAN_VNHEAD, CTL_EOL);
	sysctl_createv(clog, 0, NULL, NULL,
		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
		       CTLTYPE_INT, "dostats",
		       SYSCTL_DESCR("Maintain statistics on LFS operations"),
		       sysctl_lfs_dostats, 0, &lfs_dostats, 0,
		       CTL_VFS, 5, LFS_DOSTATS, CTL_EOL);
	sysctl_createv(clog, 0, NULL, NULL,
		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
		       CTLTYPE_INT, "pagetrip",
		       SYSCTL_DESCR("How many dirty pages in fs triggers"
				    " a flush"),
		       NULL, 0, &lfs_fs_pagetrip, 0,
		       CTL_VFS, 5, LFS_FS_PAGETRIP, CTL_EOL);
	sysctl_createv(clog, 0, NULL, NULL,
		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
		       CTLTYPE_INT, "ignore_lazy_sync",
		       SYSCTL_DESCR("Lazy Sync is ignored entirely"),
		       NULL, 0, &lfs_ignore_lazy_sync, 0,
		       CTL_VFS, 5, LFS_IGNORE_LAZY_SYNC, CTL_EOL);
#ifdef LFS_KERNEL_RFW
	sysctl_createv(clog, 0, NULL, NULL,
		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
		       CTLTYPE_INT, "rfw",
		       SYSCTL_DESCR("Use in-kernel roll-forward on mount"),
		       NULL, 0, &lfs_do_rfw, 0,
		       CTL_VFS, 5, LFS_DO_RFW, CTL_EOL);
#endif

	sysctl_createv(clog, 0, NULL, NULL,
		       CTLFLAG_PERMANENT,
		       CTLTYPE_NODE, "stats",
		       SYSCTL_DESCR("Debugging options"),
		       NULL, 0, NULL, 0,
		       CTL_VFS, 5, LFS_STATS, CTL_EOL);
	for (i = 0; i < sizeof(struct lfs_stats) / sizeof(u_int); i++) {
		sysctl_createv(clog, 0, NULL, NULL,
			       CTLFLAG_PERMANENT|CTLFLAG_READONLY,
			       CTLTYPE_INT, stat_names[i].sname,
			       SYSCTL_DESCR(stat_names[i].lname),
			       NULL, 0, &(((u_int *)&lfs_stats.segsused)[i]),
			       0, CTL_VFS, 5, LFS_STATS, i, CTL_EOL);
	}

#ifdef DEBUG
	sysctl_createv(clog, 0, NULL, NULL,
		       CTLFLAG_PERMANENT,
		       CTLTYPE_NODE, "debug",
		       SYSCTL_DESCR("Debugging options"),
		       NULL, 0, NULL, 0,
		       CTL_VFS, 5, LFS_DEBUGLOG, CTL_EOL);
	for (i = 0; i < DLOG_MAX; i++) {
		sysctl_createv(clog, 0, NULL, NULL,
			       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
			       CTLTYPE_INT, dlog_names[i].sname,
			       SYSCTL_DESCR(dlog_names[i].lname),
			       NULL, 0, &(lfs_debug_log_subsys[i]), 0,
			       CTL_VFS, 5, LFS_DEBUGLOG, i, CTL_EOL);
	}
#endif
}

/* old cleaner syscall interface.  see VOP_FCNTL() */
static const struct syscall_package lfs_syscalls[] = {
	{ SYS_lfs_bmapv,	0, (sy_call_t *)sys_lfs_bmapv		},
	{ SYS_lfs_markv,	0, (sy_call_t *)sys_lfs_markv		},
	{ SYS___lfs_segwait50,	0, (sy_call_t *)sys___lfs_segwait50	},
	{ SYS_lfs_segclean,	0, (sy_call_t *)sys_lfs_segclean	},
	{ 0, 0, NULL },
};

static int
lfs_modcmd(modcmd_t cmd, void *arg)
{
	int error;

	switch (cmd) {
	case MODULE_CMD_INIT:
		error = syscall_establish(NULL, lfs_syscalls);
		if (error)
			return error;
		error = vfs_attach(&lfs_vfsops);
		if (error != 0) {
			syscall_disestablish(NULL, lfs_syscalls);
			break;
		}
		lfs_sysctl_setup(&lfs_sysctl_log);
		cv_init(&lfs_allclean_wakeup, "segment");
		break;
	case MODULE_CMD_FINI:
		error = vfs_detach(&lfs_vfsops);
		if (error != 0)
			break;
		syscall_disestablish(NULL, lfs_syscalls);
		sysctl_teardown(&lfs_sysctl_log);
		cv_destroy(&lfs_allclean_wakeup);
		break;
	default:
		error = ENOTTY;
		break;
	}

	return (error);
}

/*
 * XXX Same structure as FFS inodes?  Should we share a common pool?
 */
struct pool lfs_inode_pool;
struct pool lfs_dinode_pool;
struct pool lfs_inoext_pool;
struct pool lfs_lbnentry_pool;

/*
 * The writer daemon.  UVM keeps track of how many dirty pages we are holding
 * in lfs_subsys_pages; the daemon flushes the filesystem when this value
 * crosses the (user-defined) threshhold LFS_MAX_PAGES.
 */
static void
lfs_writerd(void *arg)
{
	mount_iterator_t *iter;
 	struct mount *mp;
 	struct lfs *fs;
	struct vfsops *vfs = NULL;
 	int fsflags;
	int lfsc;
	int wrote_something = 0;
 
	mutex_enter(&lfs_lock);
	KASSERTMSG(lfs_writer_daemon == NULL, "more than one LFS writer daemon");
	lfs_writer_daemon = curlwp;
	mutex_exit(&lfs_lock);

	/* Take an extra reference to the LFS vfsops. */
	vfs = vfs_getopsbyname(MOUNT_LFS);
 
 	mutex_enter(&lfs_lock);
 	for (;;) {
		KASSERT(mutex_owned(&lfs_lock));
		if (wrote_something == 0)
			cv_timedwait(&lfs_writerd_cv, &lfs_lock, hz/10 + 1);
		KASSERT(mutex_owned(&lfs_lock));
		wrote_something = 0;

		/*
		 * If global state wants a flush, flush everything.
		 */
		if (lfs_do_flush || locked_queue_count > LFS_MAX_BUFS ||
			locked_queue_bytes > LFS_MAX_BYTES ||
			lfs_subsys_pages > LFS_MAX_PAGES) {

			if (lfs_do_flush) {
				DLOG((DLOG_FLUSH, "lfs_writerd: lfs_do_flush\n"));
			}
			if (locked_queue_count > LFS_MAX_BUFS) {
				DLOG((DLOG_FLUSH, "lfs_writerd: lqc = %d, max %d\n",
				      locked_queue_count, LFS_MAX_BUFS));
			}
			if (locked_queue_bytes > LFS_MAX_BYTES) {
				DLOG((DLOG_FLUSH, "lfs_writerd: lqb = %ld, max %ld\n",
				      locked_queue_bytes, LFS_MAX_BYTES));
			}
			if (lfs_subsys_pages > LFS_MAX_PAGES) {
				DLOG((DLOG_FLUSH, "lfs_writerd: lssp = %d, max %d\n",
				      lfs_subsys_pages, LFS_MAX_PAGES));
			}

			lfs_flush(NULL, SEGM_WRITERD, 0);
			lfs_do_flush = 0;
			KASSERT(mutex_owned(&lfs_lock));
			continue;
		}
		KASSERT(mutex_owned(&lfs_lock));
		mutex_exit(&lfs_lock);
 
 		/*
 		 * Look through the list of LFSs to see if any of them
 		 * have requested pageouts.
 		 */
 		mountlist_iterator_init(&iter);
		lfsc = 0;
		while ((mp = mountlist_iterator_next(iter)) != NULL) {
			KASSERT(!mutex_owned(&lfs_lock));
 			if (strncmp(mp->mnt_stat.f_fstypename, MOUNT_LFS,
 			    sizeof(mp->mnt_stat.f_fstypename)) == 0) {
				++lfsc;
 				fs = VFSTOULFS(mp)->um_lfs;
				daddr_t ooffset = 0;
				fsflags = SEGM_SINGLE;

 				mutex_enter(&lfs_lock);
				ooffset = lfs_sb_getoffset(fs);

				if (lfs_sb_getnextseg(fs) < lfs_sb_getcurseg(fs) && fs->lfs_nowrap) {
					/* Don't try to write if we're suspended */
					mutex_exit(&lfs_lock);
					continue;
				}
				if (LFS_STARVED_FOR_SEGS(fs)) {
					mutex_exit(&lfs_lock);

					DLOG((DLOG_FLUSH, "lfs_writerd: need cleaning before writing possible\n"));
					lfs_wakeup_cleaner(fs);
					continue;
				}

 				if ((fs->lfs_dirvcount > LFS_MAX_FSDIROP(fs) ||
 				     lfs_dirvcount > LFS_MAX_DIROP) &&
				    fs->lfs_dirops == 0) {
					fsflags &= ~SEGM_SINGLE;
 					fsflags |= SEGM_CKP;
					DLOG((DLOG_FLUSH, "lfs_writerd: checkpoint\n"));
					lfs_flush_fs(fs, fsflags);
				} else if (fs->lfs_pdflush) {
 					DLOG((DLOG_FLUSH, "lfs_writerd: pdflush set\n"));
 					lfs_flush_fs(fs, fsflags);
 				} else if (!TAILQ_EMPTY(&fs->lfs_pchainhd)) {
 					DLOG((DLOG_FLUSH, "lfs_writerd: pchain non-empty\n"));
 					mutex_exit(&lfs_lock);
 					lfs_writer_enter(fs, "wrdirop");
 					lfs_flush_pchain(fs);
 					lfs_writer_leave(fs);
					mutex_enter(&lfs_lock);
				}
				if (lfs_sb_getoffset(fs) != ooffset)
					++wrote_something;
				mutex_exit(&lfs_lock);
 			}
			KASSERT(!mutex_owned(&lfs_lock));
 		}
		if (lfsc == 0) {
			mutex_enter(&lfs_lock);
			lfs_writer_daemon = NULL;
			mutex_exit(&lfs_lock);
			mountlist_iterator_destroy(iter);
			break;
		}
 		mountlist_iterator_destroy(iter);
 
 		mutex_enter(&lfs_lock);
 	}
	KASSERT(!mutex_owned(&lfs_lock));

	/* Give up our extra reference so the module can be unloaded. */
	mutex_enter(&vfs_list_lock);
	if (vfs != NULL)
		vfs->vfs_refcount--;
	mutex_exit(&vfs_list_lock);

	/* Done! */
	kthread_exit(0);
}

/*
 * Initialize the filesystem, most work done by ulfs_init.
 */
void
lfs_init(void)
{

	/*
	 * XXX: should we use separate pools for 32-bit and 64-bit
	 * dinodes?
	 */
	malloc_type_attach(M_SEGMENT);
	pool_init(&lfs_inode_pool, sizeof(struct inode), 0, 0, 0,
	    "lfsinopl", &pool_allocator_nointr, IPL_NONE);
	pool_init(&lfs_dinode_pool, sizeof(union lfs_dinode), 0, 0, 0,
	    "lfsdinopl", &pool_allocator_nointr, IPL_NONE);
	pool_init(&lfs_inoext_pool, sizeof(struct lfs_inode_ext), 8, 0, 0,
	    "lfsinoextpl", &pool_allocator_nointr, IPL_NONE);
	pool_init(&lfs_lbnentry_pool, sizeof(struct lbnentry), 0, 0, 0,
	    "lfslbnpool", &pool_allocator_nointr, IPL_NONE);
	ulfs_init();

#ifdef DEBUG
	memset(lfs_log, 0, sizeof(lfs_log));
#endif
	mutex_init(&lfs_lock, MUTEX_DEFAULT, IPL_NONE);
	cv_init(&lfs_writerd_cv, "lfswrite");
	cv_init(&locked_queue_cv, "lfsbuf");
	cv_init(&lfs_writing_cv, "lfsflush");
}

void
lfs_reinit(void)
{
	ulfs_reinit();
}

void
lfs_done(void)
{
	ulfs_done();
	mutex_destroy(&lfs_lock);
	cv_destroy(&lfs_writerd_cv);
	cv_destroy(&locked_queue_cv);
	cv_destroy(&lfs_writing_cv);
	pool_destroy(&lfs_inode_pool);
	pool_destroy(&lfs_dinode_pool);
	pool_destroy(&lfs_inoext_pool);
	pool_destroy(&lfs_lbnentry_pool);
	malloc_type_detach(M_SEGMENT);
}

/*
 * Called by main() when ulfs is going to be mounted as root.
 */
int
lfs_mountroot(void)
{
	extern struct vnode *rootvp;
	struct lfs *fs = NULL;				/* LFS */
	struct mount *mp;
	struct lwp *l = curlwp;
	struct ulfsmount *ump;
	int error;

	if (device_class(root_device) != DV_DISK)
		return (ENODEV);

	if (rootdev == NODEV)
		return (ENODEV);
	if ((error = vfs_rootmountalloc(MOUNT_LFS, "root_device", &mp))) {
		vrele(rootvp);
		return (error);
	}
	if ((error = lfs_mountfs(rootvp, mp, l))) {
		vfs_unbusy(mp);
		vfs_rele(mp);
		return (error);
	}
	mountlist_append(mp);
	ump = VFSTOULFS(mp);
	fs = ump->um_lfs;
	lfs_sb_setfsmnt(fs, mp->mnt_stat.f_mntonname);
	(void)lfs_statvfs(mp, &mp->mnt_stat);
	vfs_unbusy(mp);
	setrootfstime((time_t)lfs_sb_gettstamp(VFSTOULFS(mp)->um_lfs));
	return (0);
}

/*
 * VFS Operations.
 *
 * mount system call
 */
int
lfs_mount(struct mount *mp, const char *path, void *data, size_t *data_len)
{
	struct lwp *l = curlwp;
	struct vnode *devvp;
	struct ulfs_args *args = data;
	struct ulfsmount *ump = NULL;
	struct lfs *fs = NULL;				/* LFS */
	int error = 0, update;
	mode_t accessmode;

	if (args == NULL)
		return EINVAL;
	if (*data_len < sizeof *args)
		return EINVAL;

	if (mp->mnt_flag & MNT_GETARGS) {
		ump = VFSTOULFS(mp);
		if (ump == NULL)
			return EIO;
		args->fspec = NULL;
		*data_len = sizeof *args;
		return 0;
	}

	update = mp->mnt_flag & MNT_UPDATE;

	/* Check arguments */
	if (args->fspec != NULL) {
		/*
		 * Look up the name and verify that it's sane.
		 */
		error = namei_simple_user(args->fspec,
					NSM_FOLLOW_NOEMULROOT, &devvp);
		if (error != 0)
			return (error);

		if (!update) {
			/*
			 * Be sure this is a valid block device
			 */
			if (devvp->v_type != VBLK)
				error = ENOTBLK;
			else if (bdevsw_lookup(devvp->v_rdev) == NULL)
				error = ENXIO;
		} else {
			/*
			 * Be sure we're still naming the same device
			 * used for our initial mount
			 *
			 * XXX dholland 20151010: if namei gives us a
			 * different vnode for the same device,
			 * wouldn't it be better to use it going
			 * forward rather than ignore it in favor of
			 * the old one?
			 */
			ump = VFSTOULFS(mp);
			fs = ump->um_lfs;
			if (devvp != fs->lfs_devvp) {
				if (devvp->v_rdev != fs->lfs_devvp->v_rdev)
					error = EINVAL;
				else {
					vrele(devvp);
					devvp = fs->lfs_devvp;
					vref(devvp);
				}
			}
		}
	} else {
		if (!update) {
			/* New mounts must have a filename for the device */
			return (EINVAL);
		} else {
			/* Use the extant mount */
			ump = VFSTOULFS(mp);
			fs = ump->um_lfs;
			devvp = fs->lfs_devvp;
			vref(devvp);
		}
	}


	/*
	 * If mount by non-root, then verify that user has necessary
	 * permissions on the device.
	 */
	if (error == 0) {
		accessmode = VREAD;
		if (update ?
		    (mp->mnt_iflag & IMNT_WANTRDWR) != 0 :
		    (mp->mnt_flag & MNT_RDONLY) == 0)
			accessmode |= VWRITE;
		vn_lock(devvp, LK_EXCLUSIVE | LK_RETRY);
		error = kauth_authorize_system(l->l_cred, KAUTH_SYSTEM_MOUNT,
		    KAUTH_REQ_SYSTEM_MOUNT_DEVICE, mp, devvp,
		    KAUTH_ARG(accessmode));
		VOP_UNLOCK(devvp);
	}

	if (error) {
		vrele(devvp);
		return (error);
	}

	if (!update) {
		int flags;

		if (mp->mnt_flag & MNT_RDONLY)
			flags = FREAD;
		else
			flags = FREAD|FWRITE;
		vn_lock(devvp, LK_EXCLUSIVE | LK_RETRY);
		error = VOP_OPEN(devvp, flags, FSCRED);
		VOP_UNLOCK(devvp);
		if (error)
			goto fail;
		error = lfs_mountfs(devvp, mp, l);		/* LFS */
		if (error) {
			vn_lock(devvp, LK_EXCLUSIVE | LK_RETRY);
			(void)VOP_CLOSE(devvp, flags, NOCRED);
			VOP_UNLOCK(devvp);
			goto fail;
		}

		ump = VFSTOULFS(mp);
		fs = ump->um_lfs;
	} else {
		/*
		 * Update the mount.
		 */

		/*
		 * The initial mount got a reference on this
		 * device, so drop the one obtained via
		 * namei(), above.
		 */
		vrele(devvp);

		ump = VFSTOULFS(mp);
		fs = ump->um_lfs;

		if (!fs->lfs_ronly && (mp->mnt_iflag & IMNT_WANTRDONLY)) {
			/*
			 * Changing from read/write to read-only.
			 */
			int flags = WRITECLOSE;
			if (mp->mnt_flag & MNT_FORCE)
				flags |= FORCECLOSE;
			error = lfs_flushfiles(mp, flags);
			if (error)
				return error;
			fs->lfs_ronly = 1;
		} else if (fs->lfs_ronly && (mp->mnt_iflag & IMNT_WANTRDWR)) {
			/*
			 * Changing from read-only to read/write.
			 * Note in the superblocks that we're writing.
			 */

			/* XXX: quotas should have been on even if readonly */
			if (fs->lfs_use_quota2) {
#ifdef LFS_QUOTA2
				error = lfs_quota2_mount(mp);
#else
				uprintf("%s: no kernel support for this "
					"filesystem's quotas\n",
					mp->mnt_stat.f_mntonname);
				if (mp->mnt_flag & MNT_FORCE) {
					uprintf("%s: mounting anyway; "
						"fsck afterwards\n",
						mp->mnt_stat.f_mntonname);
				} else {
					error = EINVAL;
				}
#endif
				if (error) {
					return error;
				}
			}

			fs->lfs_ronly = 0;
			if (lfs_sb_getpflags(fs) & LFS_PF_CLEAN) {
				lfs_sb_setpflags(fs, lfs_sb_getpflags(fs) & ~LFS_PF_CLEAN);
				lfs_writesuper(fs, lfs_sb_getsboff(fs, 0));
				lfs_writesuper(fs, lfs_sb_getsboff(fs, 1));
			}
		}

		if (args->fspec == NULL)
			return 0;
	}

	error = set_statvfs_info(path, UIO_USERSPACE, args->fspec,
	    UIO_USERSPACE, mp->mnt_op->vfs_name, mp, l);
	if (error == 0)
		lfs_sb_setfsmnt(fs, mp->mnt_stat.f_mntonname);
	return error;

fail:
	vrele(devvp);
	return (error);
}

/*
 * Helper for mountfs. Note that the fs pointer may be a dummy one
 * pointing into a superblock buffer. (Which is gross; see below.)
 */
static int
lfs_checkmagic(struct lfs *fs)
{
	switch (fs->lfs_dlfs_u.u_32.dlfs_magic) {
	    case LFS_MAGIC:
		fs->lfs_is64 = false;
		fs->lfs_dobyteswap = false;
		break;
	    case LFS64_MAGIC:
		fs->lfs_is64 = true;
		fs->lfs_dobyteswap = false;
		break;
#ifdef LFS_EI
	    case LFS_MAGIC_SWAPPED:
		fs->lfs_is64 = false;
		fs->lfs_dobyteswap = true;
		break;
	    case LFS64_MAGIC_SWAPPED:
		fs->lfs_is64 = true;
		fs->lfs_dobyteswap = true;
		break;
#endif
	    default:
		/* XXX needs translation */
		return EINVAL;
	}
	return 0;
}

/*
 * Common code for mount and mountroot
 * LFS specific
 */
int
lfs_mountfs(struct vnode *devvp, struct mount *mp, struct lwp *l)
{
	struct lfs *primarysb, *altsb, *thesb;
	struct buf *primarybuf, *altbuf;
	struct lfs *fs;
	struct ulfsmount *ump;
	struct vnode *vp;
	dev_t dev;
	int error, i, ronly, fsbsize;
	kauth_cred_t cred;
	CLEANERINFO *cip;
	SEGUSE *sup;
	daddr_t sb_addr;
	ino_t *orphan;
	size_t norphan;

	cred = l ? l->l_cred : NOCRED;

	/* The superblock is supposed to be 512 bytes. */
	__CTASSERT(sizeof(struct dlfs) == DEV_BSIZE);

	/*
	 * Flush out any old buffers remaining from a previous use.
	 */
	vn_lock(devvp, LK_EXCLUSIVE | LK_RETRY);
	error = vinvalbuf(devvp, V_SAVE, cred, l, 0, 0);
	VOP_UNLOCK(devvp);
	if (error)
		return (error);

	ronly = (mp->mnt_flag & MNT_RDONLY) != 0;

	/* Don't free random space on error. */
	primarybuf = NULL;
	altbuf = NULL;
	ump = NULL;

	sb_addr = LFS_LABELPAD / DEV_BSIZE;
	while (1) {
		/*
		 * Read in the superblock.
		 *
		 * Note that because LFS_SBPAD is substantially larger
		 * (8K) than the actual on-disk superblock (512 bytes)
		 * the buffer contains enough space to be used as a
		 * whole struct lfs (in-memory superblock) - we do this
		 * only so we can set and use the is64 and dobyteswap
		 * members. XXX this is gross and the logic here should
		 * be reworked.
		 */
		error = bread(devvp, sb_addr, LFS_SBPAD, 0, &primarybuf);
		if (error)
			goto out;
		primarysb = (struct lfs *)primarybuf->b_data;

		/* Check the basics. */
		error = lfs_checkmagic(primarysb);
		if (error) {
			DLOG((DLOG_MOUNT, "lfs_mountfs: primary superblock wrong magic\n"));
			goto out;
		}
		if (lfs_sb_getbsize(primarysb) > MAXBSIZE ||
		    lfs_sb_getversion(primarysb) > LFS_VERSION ||
		    lfs_sb_getbsize(primarysb) < sizeof(struct dlfs)) {
			DLOG((DLOG_MOUNT, "lfs_mountfs: primary superblock sanity failed\n"));
			/* XXX needs translation */
			error = EINVAL;
			goto out;
		}
		if (lfs_sb_getinodefmt(primarysb) > LFS_MAXINODEFMT) {
			DLOG((DLOG_MOUNT, "lfs_mountfs: unknown inode format %d\n",
			       lfs_sb_getinodefmt(primarysb)));
			error = EINVAL;
			goto out;
		}

		if (lfs_sb_getversion(primarysb) == 1)
			fsbsize = DEV_BSIZE;
		else {
			fsbsize = 1 << lfs_sb_getffshift(primarysb);
			/*
			 * Could be, if the frag size is large enough, that we
			 * don't have the "real" primary superblock.  If that's
			 * the case, get the real one, and try again.
			 */
			if (sb_addr != (lfs_sb_getsboff(primarysb, 0) << (lfs_sb_getffshift(primarysb) - DEV_BSHIFT))) {
				DLOG((DLOG_MOUNT, "lfs_mountfs: sb daddr"
				      " 0x%llx is not right, trying 0x%llx\n",
				      (long long)sb_addr,
				      (long long)(lfs_sb_getsboff(primarysb, 0) << (lfs_sb_getffshift(primarysb) - DEV_BSHIFT))));
				sb_addr = lfs_sb_getsboff(primarysb, 0) << (lfs_sb_getffshift(primarysb) - DEV_BSHIFT);
				brelse(primarybuf, BC_INVAL);
				continue;
			}
		}
		break;
	}

	/*
	 * Check the second superblock to see which is newer; then mount
	 * using the older of the two.	This is necessary to ensure that
	 * the filesystem is valid if it was not unmounted cleanly.
	 */

	if (lfs_sb_getsboff(primarysb, 1) &&
	    lfs_sb_getsboff(primarysb, 1) - LFS_LABELPAD / fsbsize > LFS_SBPAD / fsbsize)
	{
		error = bread(devvp, lfs_sb_getsboff(primarysb, 1) * (fsbsize / DEV_BSIZE),
			LFS_SBPAD, 0, &altbuf);
		if (error)
			goto out;
		altsb = (struct lfs *)altbuf->b_data;

		/*
		 * Note: this used to do the sanity check only if the
		 * timestamp/serial comparison required use of altsb;
		 * this way is less tolerant, but if altsb is corrupted
		 * enough that the magic number, version, and blocksize
		 * are bogus, why would the timestamp or serial fields
		 * mean anything either? If this kind of thing happens,
		 * you need to fsck anyway.
		 */

		error = lfs_checkmagic(altsb);
		if (error)
			goto out;

		/* Check the basics. */
		if (lfs_sb_getbsize(altsb) > MAXBSIZE ||
		    lfs_sb_getversion(altsb) > LFS_VERSION ||
		    lfs_sb_getbsize(altsb) < sizeof(struct dlfs)) {
			DLOG((DLOG_MOUNT, "lfs_mountfs: alt superblock"
			      " sanity failed\n"));
			error = EINVAL;		/* XXX needs translation */
			goto out;
		}

		if (lfs_sb_getversion(primarysb) == 1) {
			/* 1s resolution comparison */
			if (lfs_sb_gettstamp(altsb) < lfs_sb_gettstamp(primarysb))
				thesb = altsb;
			else
				thesb = primarysb;
		} else {
			/* monotonic infinite-resolution comparison */
			if (lfs_sb_getserial(altsb) < lfs_sb_getserial(primarysb))
				thesb = altsb;
			else
				thesb = primarysb;
		}
	} else {
		DLOG((DLOG_MOUNT, "lfs_mountfs: invalid alt superblock location"
		      " daddr=0x%x\n", lfs_sb_getsboff(primarysb, 1)));
		error = EINVAL;
		goto out;
	}

	/*
	 * Allocate the mount structure, copy the superblock into it.
	 * Note that the 32-bit and 64-bit superblocks are the same size.
	 */
	fs = kmem_zalloc(sizeof(struct lfs), KM_SLEEP);
	memcpy(&fs->lfs_dlfs_u.u_32, &thesb->lfs_dlfs_u.u_32,
	       sizeof(struct dlfs));
	fs->lfs_is64 = thesb->lfs_is64;
	fs->lfs_dobyteswap = thesb->lfs_dobyteswap;
	fs->lfs_hasolddirfmt = false; /* set for real below */

	/* Compatibility */
	if (lfs_sb_getversion(fs) < 2) {
		lfs_sb_setsumsize(fs, LFS_V1_SUMMARY_SIZE);
		lfs_sb_setibsize(fs, lfs_sb_getbsize(fs));
		lfs_sb_sets0addr(fs, lfs_sb_getsboff(fs, 0));
		lfs_sb_settstamp(fs, lfs_sb_getotstamp(fs));
		lfs_sb_setfsbtodb(fs, 0);
	}
	if (lfs_sb_getresvseg(fs) == 0)
		lfs_sb_setresvseg(fs, MIN(lfs_sb_getminfreeseg(fs) - 1, \
			MAX(MIN_RESV_SEGS, lfs_sb_getminfreeseg(fs) / 2 + 1)));

	/*
	 * If we aren't going to be able to write meaningfully to this
	 * filesystem, and were not mounted readonly, bomb out now.
	 */
	if (lfs_fsbtob(fs, LFS_NRESERVE(fs)) > LFS_MAX_BYTES && !ronly) {
		DLOG((DLOG_MOUNT, "lfs_mount: to mount this filesystem read/write,"
		      " we need BUFPAGES >= %lld\n",
		      (long long)((bufmem_hiwater / bufmem_lowater) *
				  LFS_INVERSE_MAX_BYTES(
					  lfs_fsbtob(fs, LFS_NRESERVE(fs))) >> PAGE_SHIFT)));
		kmem_free(fs, sizeof(struct lfs));
		error = EFBIG; /* XXX needs translation */
		goto out;
	}

	/* Before rolling forward, lock so vget will sleep for other procs */
	if (l != NULL) {
		fs->lfs_flags = LFS_NOTYET;
		fs->lfs_rfpid = l->l_proc->p_pid;
	}

	ump = kmem_zalloc(sizeof(*ump), KM_SLEEP);
	ump->um_lfs = fs;
	ump->um_fstype = fs->lfs_is64 ? ULFS2 : ULFS1;
	/* ump->um_cleaner_thread = NULL; */
	brelse(primarybuf, BC_INVAL);
	brelse(altbuf, BC_INVAL);
	primarybuf = NULL;
	altbuf = NULL;


	/* Set up the I/O information */
	fs->lfs_devbsize = DEV_BSIZE;
	fs->lfs_iocount = 0;
	fs->lfs_diropwait = 0;
	fs->lfs_activesb = 0;
	lfs_sb_setuinodes(fs, 0);
	fs->lfs_ravail = 0;
	fs->lfs_favail = 0;
	fs->lfs_sbactive = 0;

	/* Set up the ifile and lock aflags */
	fs->lfs_doifile = 0;
	fs->lfs_writer = 0;
	fs->lfs_dirops = 0;
	fs->lfs_nadirop = 0;
	fs->lfs_seglock = 0;
	fs->lfs_pdflush = 0;
	fs->lfs_sleepers = 0;
	fs->lfs_pages = 0;
	rw_init(&fs->lfs_fraglock);
	rw_init(&fs->lfs_iflock);
	cv_init(&fs->lfs_sleeperscv, "lfs_slp");
	cv_init(&fs->lfs_diropscv, "lfs_dirop");
	cv_init(&fs->lfs_stopcv, "lfsstop");
	cv_init(&fs->lfs_nextsegsleep, "segment");

	/* Set the file system readonly/modify bits. */
	fs->lfs_ronly = ronly;
	if (ronly == 0)
		fs->lfs_fmod = 1;

	/* Device we're using */
	dev = devvp->v_rdev;
	fs->lfs_dev = dev;
	fs->lfs_devvp = devvp;

	/* ulfs-level information */
	fs->um_flags = 0;
	fs->um_bptrtodb = lfs_sb_getffshift(fs) - DEV_BSHIFT;
	fs->um_seqinc = lfs_sb_getfrag(fs);
	fs->um_nindir = lfs_sb_getnindir(fs);
	fs->um_lognindir = ffs(lfs_sb_getnindir(fs)) - 1;
	fs->um_maxsymlinklen = lfs_sb_getmaxsymlinklen(fs);
	fs->um_dirblksiz = LFS_DIRBLKSIZ;
	fs->um_maxfilesize = lfs_sb_getmaxfilesize(fs);

	/* quota stuff */
	/* XXX: these need to come from the on-disk superblock to be used */
	fs->lfs_use_quota2 = 0;
	fs->lfs_quota_magic = 0;
	fs->lfs_quota_flags = 0;
	fs->lfs_quotaino[0] = 0;
	fs->lfs_quotaino[1] = 0;

	/* Initialize the mount structure. */
	mp->mnt_data = ump;
	mp->mnt_stat.f_fsidx.__fsid_val[0] = (long)dev;
	mp->mnt_stat.f_fsidx.__fsid_val[1] = makefstype(MOUNT_LFS);
	mp->mnt_stat.f_fsid = mp->mnt_stat.f_fsidx.__fsid_val[0];
	mp->mnt_stat.f_namemax = LFS_MAXNAMLEN;
	mp->mnt_stat.f_iosize = lfs_sb_getbsize(fs);
	mp->mnt_flag |= MNT_LOCAL;
	mp->mnt_fs_bshift = lfs_sb_getbshift(fs);
	mp->mnt_iflag |= IMNT_CAN_RWTORO;
	if (fs->um_maxsymlinklen > 0)
		mp->mnt_iflag |= IMNT_DTYPE;
	else
		fs->lfs_hasolddirfmt = true;

	ump->um_mountp = mp;
	for (i = 0; i < ULFS_MAXQUOTAS; i++)
		ump->um_quotas[i] = NULLVP;
	spec_node_setmountedfs(devvp, mp);

	/* Set up reserved memory for pageout */
	lfs_setup_resblks(fs);
	/* Set up vdirop tailq */
	TAILQ_INIT(&fs->lfs_dchainhd);
	/* and paging tailq */
	TAILQ_INIT(&fs->lfs_pchainhd);
	/* and delayed segment accounting for truncation list */
	LIST_INIT(&fs->lfs_segdhd);

	/*
	 * We use the ifile vnode for almost every operation.  Instead of
	 * retrieving it from the hash table each time we retrieve it here,
	 * artificially increment the reference count and keep a pointer
	 * to it in the incore copy of the superblock.
	 */
	if ((error = VFS_VGET(mp, LFS_IFILE_INUM, &vp)) != 0) {
		DLOG((DLOG_MOUNT, "lfs_mountfs: ifile vget failed, error=%d\n", error));
		goto out;
	}
	fs->lfs_ivnode = vp;
	vref(vp);

	/* Set up inode bitmap, order free list, and gather orphans.  */
	lfs_order_freelist(fs, &orphan, &norphan);

	/* Set up segment usage flags for the autocleaner. */
	fs->lfs_nactive = 0;
	fs->lfs_suflags = malloc(2 * sizeof(u_int32_t *),
				 M_SEGMENT, M_WAITOK);
	fs->lfs_suflags[0] = malloc(lfs_sb_getnseg(fs) * sizeof(u_int32_t),
				    M_SEGMENT, M_WAITOK);
	fs->lfs_suflags[1] = malloc(lfs_sb_getnseg(fs) * sizeof(u_int32_t),
				    M_SEGMENT, M_WAITOK);
	memset(fs->lfs_suflags[1], 0, lfs_sb_getnseg(fs) * sizeof(u_int32_t));
	for (i = 0; i < lfs_sb_getnseg(fs); i++) {
		int changed;
		struct buf *bp;

		LFS_SEGENTRY(sup, fs, i, bp);
		changed = 0;
		if (!ronly) {
			if (sup->su_nbytes == 0 &&
			    !(sup->su_flags & SEGUSE_EMPTY)) {
				sup->su_flags |= SEGUSE_EMPTY;
				++changed;
			} else if (!(sup->su_nbytes == 0) &&
				   (sup->su_flags & SEGUSE_EMPTY)) {
				sup->su_flags &= ~SEGUSE_EMPTY;
				++changed;
			}
			if (sup->su_flags & (SEGUSE_ACTIVE|SEGUSE_INVAL)) {
				sup->su_flags &= ~(SEGUSE_ACTIVE|SEGUSE_INVAL);
				++changed;
			}
		}
		fs->lfs_suflags[0][i] = sup->su_flags;
		if (changed)
			LFS_WRITESEGENTRY(sup, fs, i, bp);
		else
			brelse(bp, 0);
	}

	/* Free the orphans we discovered while ordering the freelist.  */
	lfs_free_orphans(fs, orphan, norphan);

	/*
	 * XXX: if the fs has quotas, quotas should be on even if
	 * readonly. Otherwise you can't query the quota info!
	 * However, that's not how the quota2 code got written and I
	 * don't know if it'll behave itself if enabled while
	 * readonly, so for now use the same enable logic as ffs.
	 *
	 * XXX: also, if you use the -f behavior allowed here (and
	 * equivalently above for remount) it will corrupt the fs. It
	 * ought not to allow that. It should allow mounting readonly
	 * if there are quotas and the kernel doesn't have the quota
	 * code, but only readonly.
	 *
	 * XXX: and if you use the -f behavior allowed here it will
	 * likely crash at unmount time (or remount time) because we
	 * think quotas are active.
	 *
	 * Although none of this applies until there's a way to set
	 * lfs_use_quota2 and have quotas in the fs at all.
	 */
	if (!ronly && fs->lfs_use_quota2) {
#ifdef LFS_QUOTA2
		error = lfs_quota2_mount(mp);
#else
		uprintf("%s: no kernel support for this filesystem's quotas\n",
			mp->mnt_stat.f_mntonname);
		if (mp->mnt_flag & MNT_FORCE) {
			uprintf("%s: mounting anyway; fsck afterwards\n",
				mp->mnt_stat.f_mntonname);
		} else {
			error = EINVAL;
		}
#endif
		if (error) {
			/* XXX XXX must clean up the stuff immediately above */
			printf("lfs_mountfs: sorry, leaking some memory\n");
			goto out;
		}
	}

#ifdef LFS_KERNEL_RFW
	lfs_roll_forward(fs, mp, l);
#endif

	/* If writing, sb is not clean; record in case of immediate crash */
	if (!fs->lfs_ronly) {
		lfs_sb_setpflags(fs, lfs_sb_getpflags(fs) & ~LFS_PF_CLEAN);
		lfs_writesuper(fs, lfs_sb_getsboff(fs, 0));
		lfs_writesuper(fs, lfs_sb_getsboff(fs, 1));
	}

	/* Allow vget now that roll-forward is complete */
	fs->lfs_flags &= ~(LFS_NOTYET);
	wakeup(&fs->lfs_flags);

	/*
	 * Initialize the ifile cleaner info with information from
	 * the superblock.
	 */
	{
		struct buf *bp;

		LFS_CLEANERINFO(cip, fs, bp);
		lfs_ci_setclean(fs, cip, lfs_sb_getnclean(fs));
		lfs_ci_setdirty(fs, cip, lfs_sb_getnseg(fs) - lfs_sb_getnclean(fs));
		lfs_ci_setavail(fs, cip, lfs_sb_getavail(fs));
		lfs_ci_setbfree(fs, cip, lfs_sb_getbfree(fs));
		(void) LFS_BWRITE_LOG(bp); /* Ifile */
	}

	/*
	 * Mark the current segment as ACTIVE, since we're going to
	 * be writing to it.
	 */
	{
		struct buf *bp;

		LFS_SEGENTRY(sup, fs, lfs_dtosn(fs, lfs_sb_getoffset(fs)), bp);
		sup->su_flags |= SEGUSE_DIRTY | SEGUSE_ACTIVE;
		fs->lfs_nactive++;
		LFS_WRITESEGENTRY(sup, fs, lfs_dtosn(fs, lfs_sb_getoffset(fs)), bp);  /* Ifile */
	}

	/* Now that roll-forward is done, unlock the Ifile */
	vput(vp);

	/* Start the pagedaemon-anticipating daemon */
	mutex_enter(&lfs_lock);
	if (lfs_writer_daemon == NULL &&
	    kthread_create(PRI_BIO, 0, NULL,
	    lfs_writerd, NULL, NULL, "lfs_writer") != 0)
		panic("fork lfs_writer");
	mutex_exit(&lfs_lock);

	printf("WARNING: the log-structured file system is experimental\n"
	    "WARNING: it may cause system crashes and/or corrupt data\n");

	return (0);

out:
	if (primarybuf)
		brelse(primarybuf, BC_INVAL);
	if (altbuf)
		brelse(altbuf, BC_INVAL);
	if (ump) {
		kmem_free(ump->um_lfs, sizeof(struct lfs));
		kmem_free(ump, sizeof(*ump));
		mp->mnt_data = NULL;
	}

	return (error);
}

/*
 * unmount system call
 */
int
lfs_unmount(struct mount *mp, int mntflags)
{
	struct ulfsmount *ump;
	struct lfs *fs;
	int error, ronly;

	ump = VFSTOULFS(mp);
	fs = ump->um_lfs;

	error = lfs_flushfiles(mp, mntflags & MNT_FORCE ? FORCECLOSE : 0);
	if (error)
		return error;

	/* Finish with the Ifile, now that we're done with it */
	vgone(fs->lfs_ivnode);

	ronly = !fs->lfs_ronly;
	if (fs->lfs_devvp->v_type != VBAD)
		spec_node_setmountedfs(fs->lfs_devvp, NULL);
	vn_lock(fs->lfs_devvp, LK_EXCLUSIVE | LK_RETRY);
	error = VOP_CLOSE(fs->lfs_devvp,
	    ronly ? FREAD : FREAD|FWRITE, NOCRED);
	vput(fs->lfs_devvp);

	/* Complain about page leakage */
	if (fs->lfs_pages > 0)
		printf("lfs_unmount: still claim %d pages (%d in subsystem)\n",
			fs->lfs_pages, lfs_subsys_pages);

	/* Free per-mount data structures */
	free(fs->lfs_ino_bitmap, M_SEGMENT);
	free(fs->lfs_suflags[0], M_SEGMENT);
	free(fs->lfs_suflags[1], M_SEGMENT);
	free(fs->lfs_suflags, M_SEGMENT);
	lfs_free_resblks(fs);
	cv_destroy(&fs->lfs_sleeperscv);
	cv_destroy(&fs->lfs_diropscv);
	cv_destroy(&fs->lfs_stopcv);
	cv_destroy(&fs->lfs_nextsegsleep);

	rw_destroy(&fs->lfs_fraglock);
	rw_destroy(&fs->lfs_iflock);

	kmem_free(fs, sizeof(struct lfs));
	kmem_free(ump, sizeof(*ump));

	mp->mnt_data = NULL;
	mp->mnt_flag &= ~MNT_LOCAL;
	return (error);
}

static int
lfs_flushfiles(struct mount *mp, int flags)
{
	struct lwp *l = curlwp;
	struct ulfsmount *ump;
	struct lfs *fs;
	struct vnode *vp;
	int error;

	ump = VFSTOULFS(mp);
	fs = ump->um_lfs;

	/* Two checkpoints */
	if (!fs->lfs_ronly) {
		lfs_segwrite(mp, SEGM_CKP | SEGM_SYNC);
		lfs_segwrite(mp, SEGM_CKP | SEGM_SYNC);
	}

	/* wake up the cleaner so it can die */
	/* XXX: shouldn't this be *after* the error cases below? */
	lfs_wakeup_cleaner(fs);
	mutex_enter(&lfs_lock);
	while (fs->lfs_sleepers)
		cv_wait(&fs->lfs_sleeperscv, &lfs_lock);
	mutex_exit(&lfs_lock);

#ifdef LFS_EXTATTR
	if (ump->um_fstype == ULFS1) {
		if (ump->um_extattr.uepm_flags & ULFS_EXTATTR_UEPM_STARTED) {
			ulfs_extattr_stop(mp, curlwp);
		}
		if (ump->um_extattr.uepm_flags & ULFS_EXTATTR_UEPM_INITIALIZED) {
			ulfs_extattr_uepm_destroy(&ump->um_extattr);
			mp->mnt_flag &= ~MNT_EXTATTR;
		}
	}
#endif
#ifdef LFS_QUOTA
        if ((error = lfsquota1_umount(mp, flags)) != 0)
		return (error);
#endif
#ifdef LFS_QUOTA2
        if ((error = lfsquota2_umount(mp, flags)) != 0)
		return (error);
#endif
	if ((error = vflush(mp, fs->lfs_ivnode, flags)) != 0)
		return (error);
	if ((error = VFS_SYNC(mp, 1, l->l_cred)) != 0)
		return (error);
	vp = fs->lfs_ivnode;
	mutex_enter(vp->v_interlock);
	if (LIST_FIRST(&vp->v_dirtyblkhd))
		panic("lfs_unmount: still dirty blocks on ifile vnode");
	mutex_exit(vp->v_interlock);

	/* Explicitly write the superblock, to update serial and pflags */
	if (!fs->lfs_ronly) {
		lfs_sb_setpflags(fs, lfs_sb_getpflags(fs) | LFS_PF_CLEAN);
		lfs_writesuper(fs, lfs_sb_getsboff(fs, 0));
		lfs_writesuper(fs, lfs_sb_getsboff(fs, 1));
	}
	mutex_enter(&lfs_lock);
	while (fs->lfs_iocount)
		mtsleep(&fs->lfs_iocount, PRIBIO + 1, "lfs_umount", 0,
			&lfs_lock);
	mutex_exit(&lfs_lock);

	return 0;
}

/*
 * Get file system statistics.
 *
 * NB: We don't lock to access the superblock here, because it's not
 * really that important if we get it wrong.
 */
int
lfs_statvfs(struct mount *mp, struct statvfs *sbp)
{
	struct lfs *fs;
	struct ulfsmount *ump;

	ump = VFSTOULFS(mp);
	fs = ump->um_lfs;

	sbp->f_bsize = lfs_sb_getbsize(fs);
	sbp->f_frsize = lfs_sb_getfsize(fs);
	sbp->f_iosize = lfs_sb_getbsize(fs);
	sbp->f_blocks = LFS_EST_NONMETA(fs) - VTOI(fs->lfs_ivnode)->i_lfs_effnblks;

	sbp->f_bfree = LFS_EST_BFREE(fs);
	/*
	 * XXX this should be lfs_sb_getsize (measured in frags)
	 * rather than dsize (measured in diskblocks). However,
	 * getsize needs a format version check (for version 1 it
	 * needs to be blockstofrags'd) so for the moment I'm going to
	 * leave this...  it won't fire wrongly as frags are at least
	 * as big as diskblocks.
	 */
	KASSERT(sbp->f_bfree <= lfs_sb_getdsize(fs));
#if 0
	if (sbp->f_bfree < 0)
		sbp->f_bfree = 0;
#endif

	sbp->f_bresvd = LFS_EST_RSVD(fs);
	if (sbp->f_bfree > sbp->f_bresvd)
		sbp->f_bavail = sbp->f_bfree - sbp->f_bresvd;
	else
		sbp->f_bavail = 0;

	/* XXX: huh? - dholland 20150728 */
	sbp->f_files = lfs_sb_getbfree(fs) / lfs_btofsb(fs, lfs_sb_getibsize(fs))
	    * LFS_INOPB(fs);
	sbp->f_ffree = sbp->f_files - lfs_sb_getnfiles(fs);
	sbp->f_favail = sbp->f_ffree;
	sbp->f_fresvd = 0;
	copy_statvfs_info(sbp, mp);
	return (0);
}

/*
 * Go through the disk queues to initiate sandbagged IO;
 * go through the inodes to write those that have been modified;
 * initiate the writing of the super block if it has been modified.
 *
 * Note: we are always called with the filesystem marked `MPBUSY'.
 */
int
lfs_sync(struct mount *mp, int waitfor, kauth_cred_t cred)
{
	int error;
	struct lfs *fs;

	fs = VFSTOULFS(mp)->um_lfs;
	if (fs->lfs_ronly)
		return 0;

	/* Snapshots should not hose the syncer */
	/*
	 * XXX Sync can block here anyway, since we don't have a very
	 * XXX good idea of how much data is pending.  If it's more
	 * XXX than a segment and lfs_nextseg is close to the end of
	 * XXX the log, we'll likely block.
	 */
	mutex_enter(&lfs_lock);
	if (fs->lfs_nowrap && lfs_sb_getnextseg(fs) < lfs_sb_getcurseg(fs)) {
		mutex_exit(&lfs_lock);
		return 0;
	}
	mutex_exit(&lfs_lock);

	lfs_writer_enter(fs, "lfs_dirops");

	/* All syncs must be checkpoints until roll-forward is implemented. */
	DLOG((DLOG_FLUSH, "lfs_sync at 0x%jx\n",
	      (uintmax_t)lfs_sb_getoffset(fs)));
	error = lfs_segwrite(mp, SEGM_CKP | (waitfor ? SEGM_SYNC : 0));
	lfs_writer_leave(fs);
#ifdef LFS_QUOTA
	lfs_qsync(mp);
#endif
	return (error);
}

/*
 * Look up an LFS dinode number to find its incore vnode.  If not already
 * in core, read it in from the specified device.  Return the inode locked.
 * Detection and handling of mount points must be done by the calling routine.
 */
int
lfs_vget(struct mount *mp, ino_t ino, struct vnode **vpp)
{
	int error;

	error = vcache_get(mp, &ino, sizeof(ino), vpp);
	if (error)
		return error;
	error = vn_lock(*vpp, LK_EXCLUSIVE);
	if (error) {
		vrele(*vpp);
		*vpp = NULL;
		return error;
	}

	return 0;
}

/*
 * Create a new vnode/inode pair and initialize what fields we can.
 */
static void
lfs_init_vnode(struct ulfsmount *ump, ino_t ino, struct vnode *vp)
{
	struct lfs *fs = ump->um_lfs;
	struct inode *ip;
	union lfs_dinode *dp;

	ASSERT_NO_SEGLOCK(fs);

	/* Initialize the inode. */
	ip = pool_get(&lfs_inode_pool, PR_WAITOK);
	memset(ip, 0, sizeof(*ip));
	dp = pool_get(&lfs_dinode_pool, PR_WAITOK);
	memset(dp, 0, sizeof(*dp));
	ip->inode_ext.lfs = pool_get(&lfs_inoext_pool, PR_WAITOK);
	memset(ip->inode_ext.lfs, 0, sizeof(*ip->inode_ext.lfs));
	ip->i_din = dp;
	ip->i_ump = ump;
	ip->i_vnode = vp;
	ip->i_dev = fs->lfs_dev;
	lfs_dino_setinumber(fs, dp, ino);
	ip->i_number = ino;
	ip->i_lfs = fs;
	ip->i_lfs_effnblks = 0;
	SPLAY_INIT(&ip->i_lfs_lbtree);
	ip->i_lfs_nbtree = 0;
	LIST_INIT(&ip->i_lfs_segdhd);

	vp->v_tag = VT_LFS;
	vp->v_op = lfs_vnodeop_p;
	vp->v_data = ip;
}

/*
 * Undo lfs_init_vnode().
 */
static void
lfs_deinit_vnode(struct ulfsmount *ump, struct vnode *vp)
{
	struct inode *ip = VTOI(vp);

	pool_put(&lfs_inoext_pool, ip->inode_ext.lfs);
	pool_put(&lfs_dinode_pool, ip->i_din);
	pool_put(&lfs_inode_pool, ip);
	vp->v_data = NULL;
}

/*
 * Read an inode from disk and initialize this vnode / inode pair.
 * Caller assures no other thread will try to load this inode.
 */
int
lfs_loadvnode(struct mount *mp, struct vnode *vp,
    const void *key, size_t key_len, const void **new_key)
{
	struct lfs *fs;
	union lfs_dinode *dip;
	struct inode *ip;
	struct buf *bp;
	IFILE *ifp;
	struct ulfsmount *ump;
	ino_t ino;
	daddr_t daddr;
	int error, retries;
	struct timespec ts;

	KASSERT(key_len == sizeof(ino));
	memcpy(&ino, key, key_len);

	memset(&ts, 0, sizeof ts);	/* XXX gcc */

	ump = VFSTOULFS(mp);
	fs = ump->um_lfs;

	/*
	 * If the filesystem is not completely mounted yet, suspend
	 * any access requests (wait for roll-forward to complete).
	 */
	mutex_enter(&lfs_lock);
	while ((fs->lfs_flags & LFS_NOTYET) && curproc->p_pid != fs->lfs_rfpid)
		mtsleep(&fs->lfs_flags, PRIBIO+1, "lfs_notyet", 0,
			&lfs_lock);
	mutex_exit(&lfs_lock);

	/* Translate the inode number to a disk address. */
	if (ino == LFS_IFILE_INUM)
		daddr = lfs_sb_getidaddr(fs);
	else {
		/* XXX bounds-check this too */
		LFS_IENTRY(ifp, fs, ino, bp);
		daddr = lfs_if_getdaddr(fs, ifp);
		if (lfs_sb_getversion(fs) > 1) {
			ts.tv_sec = lfs_if_getatime_sec(fs, ifp);
			ts.tv_nsec = lfs_if_getatime_nsec(fs, ifp);
		}

		brelse(bp, 0);
		if (daddr == LFS_UNUSED_DADDR)
			return (ENOENT);
	}

	/* Allocate/init new vnode/inode. */
	lfs_init_vnode(ump, ino, vp);
	ip = VTOI(vp);

	/* If the cleaner supplied the inode, use it. */
	if (curlwp == fs->lfs_cleaner_thread && fs->lfs_cleaner_hint != NULL &&
	    fs->lfs_cleaner_hint->bi_lbn == LFS_UNUSED_LBN) {
		dip = fs->lfs_cleaner_hint->bi_bp;
		if (fs->lfs_is64) {
			error = copyin(dip, &ip->i_din->u_64,
				       sizeof(struct lfs64_dinode));
		} else {
			error = copyin(dip, &ip->i_din->u_32,
				       sizeof(struct lfs32_dinode));
		}
		if (error) {
			lfs_deinit_vnode(ump, vp);
			return error;
		}
		KASSERT(ip->i_number == ino);
		goto out;
	}

	/* Read in the disk contents for the inode, copy into the inode. */
	retries = 0;
again:
	error = bread(fs->lfs_devvp, LFS_FSBTODB(fs, daddr),
		(lfs_sb_getversion(fs) == 1 ? lfs_sb_getbsize(fs) : lfs_sb_getibsize(fs)),
		0, &bp);
	if (error) {
		lfs_deinit_vnode(ump, vp);
		return error;
	}

	dip = lfs_ifind(fs, ino, bp);
	if (dip == NULL) {
		/* Assume write has not completed yet; try again */
		brelse(bp, BC_INVAL);
		++retries;
		if (retries <= LFS_IFIND_RETRIES) {
			mutex_enter(&lfs_lock);
			if (fs->lfs_iocount) {
				DLOG((DLOG_VNODE,
				    "%s: dinode %d not found, retrying...\n",
				    __func__, ino));
				(void)mtsleep(&fs->lfs_iocount, PRIBIO + 1,
					      "lfs ifind", 1, &lfs_lock);
			} else
				retries = LFS_IFIND_RETRIES;
			mutex_exit(&lfs_lock);
			goto again;
		}
#ifdef DEBUG
		/* If the seglock is held look at the bpp to see
		   what is there anyway */
		mutex_enter(&lfs_lock);
		if (fs->lfs_seglock > 0) {
			struct buf **bpp;
			union lfs_dinode *dp;
			int i;

			for (bpp = fs->lfs_sp->bpp;
			     bpp != fs->lfs_sp->cbpp; ++bpp) {
				if ((*bpp)->b_vp == fs->lfs_ivnode &&
				    bpp != fs->lfs_sp->bpp) {
					/* Inode block */
					printf("%s: block 0x%" PRIx64 ": ",
					       __func__, (*bpp)->b_blkno);
					for (i = 0; i < LFS_INOPB(fs); i++) {
						dp = DINO_IN_BLOCK(fs,
						    (*bpp)->b_data, i);
						if (lfs_dino_getinumber(fs, dp))
							printf("%ju ",
							    (uintmax_t)lfs_dino_getinumber(fs, dp));
					}
					printf("\n");
				}
			}
		}
		mutex_exit(&lfs_lock);
#endif /* DEBUG */
		panic("lfs_loadvnode: dinode not found");
	}
	lfs_copy_dinode(fs, ip->i_din, dip);
	brelse(bp, 0);

out:	
	if (lfs_sb_getversion(fs) > 1) {
		lfs_dino_setatime(fs, ip->i_din, ts.tv_sec);
		lfs_dino_setatimensec(fs, ip->i_din, ts.tv_nsec);
	}

	lfs_vinit(mp, &vp);

	*new_key = &ip->i_number;
	return 0;
}

/*
 * Create a new inode and initialize this vnode / inode pair.
 */
int
lfs_newvnode(struct mount *mp, struct vnode *dvp, struct vnode *vp,
    struct vattr *vap, kauth_cred_t cred, void *extra,
    size_t *key_len, const void **new_key)
{
	ino_t ino;
	struct inode *ip;
	struct ulfsmount *ump;
	struct lfs *fs;
	int error, mode, gen;

	KASSERT(dvp != NULL || vap->va_fileid > 0);
	KASSERT(dvp != NULL && dvp->v_mount == mp);
	KASSERT(vap->va_type != VNON);

	*key_len = sizeof(ino);
	ump = VFSTOULFS(mp);
	fs = ump->um_lfs;
	mode = MAKEIMODE(vap->va_type, vap->va_mode);

	/*
	 * Allocate fresh inode.  With "dvp == NULL" take the inode number
	 * and version from "vap".
	*/
	if (dvp == NULL) {
		ino = vap->va_fileid;
		gen = vap->va_gen;
		error = lfs_valloc_fixed(fs, ino, gen);
	} else {
		error = lfs_valloc(dvp, mode, cred, &ino, &gen);
	}
	if (error)
		return error;

	/* Attach inode to vnode. */
	lfs_init_vnode(ump, ino, vp);
	ip = VTOI(vp);

	mutex_enter(&lfs_lock);
	LFS_SET_UINO(ip, IN_CHANGE);
	mutex_exit(&lfs_lock);

	/* Note no blocks yet */
	ip->i_lfs_hiblk = -1;

	/* Set a new generation number for this inode. */
	ip->i_gen = gen;
	lfs_dino_setgen(fs, ip->i_din, gen);

	memset(ip->i_lfs_fragsize, 0,
	    ULFS_NDADDR * sizeof(*ip->i_lfs_fragsize));

	/* Set uid / gid. */
	if (cred == NOCRED || cred == FSCRED) {
		ip->i_gid = 0;
		ip->i_uid = 0;
	} else {
		ip->i_gid = VTOI(dvp)->i_gid;
		ip->i_uid = kauth_cred_geteuid(cred);
	}
	DIP_ASSIGN(ip, gid, ip->i_gid);
	DIP_ASSIGN(ip, uid, ip->i_uid);

#if defined(LFS_QUOTA) || defined(LFS_QUOTA2)
	error = lfs_chkiq(ip, 1, cred, 0);
	if (error) {
		lfs_vfree(dvp, ino, mode);
		lfs_deinit_vnode(ump, vp);

		return error;
	}
#endif

	/* Set type and finalize. */
	ip->i_flags = 0;
	DIP_ASSIGN(ip, flags, 0);
	ip->i_mode = mode;
	DIP_ASSIGN(ip, mode, mode);
	if (vap->va_rdev != VNOVAL) {
		/*
		 * Want to be able to use this to make badblock
		 * inodes, so don't truncate the dev number.
		 */
		// XXX clean this up
		if (ump->um_fstype == ULFS1)
			ip->i_din->u_32.di_rdev = ulfs_rw32(vap->va_rdev,
			    ULFS_MPNEEDSWAP(fs));
		else
			ip->i_din->u_64.di_rdev = ulfs_rw64(vap->va_rdev,
			    ULFS_MPNEEDSWAP(fs));
	}
	lfs_vinit(mp, &vp);

	*new_key = &ip->i_number;
	return 0;
}

/*
 * File handle to vnode
 */
int
lfs_fhtovp(struct mount *mp, struct fid *fhp, struct vnode **vpp)
{
	struct lfid lfh;
	struct lfs *fs;

	if (fhp->fid_len != sizeof(struct lfid))
		return EINVAL;

	memcpy(&lfh, fhp, sizeof(lfh));
	if (lfh.lfid_ino < LFS_IFILE_INUM)
		return ESTALE;

	fs = VFSTOULFS(mp)->um_lfs;
	if (lfh.lfid_ident != lfs_sb_getident(fs))
		return ESTALE;

	if (lfh.lfid_ino >
	    ((lfs_dino_getsize(fs, VTOI(fs->lfs_ivnode)->i_din) >> lfs_sb_getbshift(fs)) -
	     lfs_sb_getcleansz(fs) - lfs_sb_getsegtabsz(fs)) * lfs_sb_getifpb(fs))
		return ESTALE;

	return (ulfs_fhtovp(mp, &lfh.lfid_ufid, vpp));
}

/*
 * Vnode pointer to File handle
 */
/* ARGSUSED */
int
lfs_vptofh(struct vnode *vp, struct fid *fhp, size_t *fh_size)
{
	struct inode *ip;
	struct lfid lfh;

	if (*fh_size < sizeof(struct lfid)) {
		*fh_size = sizeof(struct lfid);
		return E2BIG;
	}
	*fh_size = sizeof(struct lfid);
	ip = VTOI(vp);
	memset(&lfh, 0, sizeof(lfh));
	lfh.lfid_len = sizeof(struct lfid);
	lfh.lfid_ino = ip->i_number;
	lfh.lfid_gen = ip->i_gen;
	lfh.lfid_ident = lfs_sb_getident(ip->i_lfs);
	memcpy(fhp, &lfh, sizeof(lfh));
	return (0);
}

/*
 * ulfs_bmaparray callback function for writing.
 *
 * Since blocks will be written to the new segment anyway,
 * we don't care about current daddr of them.
 */
static bool
lfs_issequential_hole(const struct lfs *fs,
    daddr_t daddr0, daddr_t daddr1)
{
	(void)fs; /* not used */

	KASSERT(daddr0 == UNWRITTEN ||
	    (0 <= daddr0 && daddr0 <= LFS_MAX_DADDR(fs)));
	KASSERT(daddr1 == UNWRITTEN ||
	    (0 <= daddr1 && daddr1 <= LFS_MAX_DADDR(fs)));

	/* NOTE: all we want to know here is 'hole or not'. */
	/* NOTE: UNASSIGNED is converted to 0 by ulfs_bmaparray. */

	/*
	 * treat UNWRITTENs and all resident blocks as 'contiguous'
	 */
	if (daddr0 != 0 && daddr1 != 0)
		return true;

	/*
	 * both are in hole?
	 */
	if (daddr0 == 0 && daddr1 == 0)
		return true; /* all holes are 'contiguous' for us. */

	return false;
}

/*
 * lfs_gop_write functions exactly like genfs_gop_write, except that
 * (1) it requires the seglock to be held by its caller, and sp->fip
 *     to be properly initialized (it will return without re-initializing
 *     sp->fip, and without calling lfs_writeseg).
 * (2) it uses the remaining space in the segment, rather than VOP_BMAP,
 *     to determine how large a block it can write at once (though it does
 *     still use VOP_BMAP to find holes in the file);
 * (3) it calls lfs_gatherblock instead of VOP_STRATEGY on its blocks
 *     (leaving lfs_writeseg to deal with the cluster blocks, so we might
 *     now have clusters of clusters, ick.)
 */
static int
lfs_gop_write(struct vnode *vp, struct vm_page **pgs, int npages,
    int flags)
{
	int i, error, run, haveeof = 0;
	int fs_bshift;
	vaddr_t kva;
	off_t eof, offset, startoffset = 0;
	size_t bytes, iobytes, skipbytes;
	bool async = (flags & PGO_SYNCIO) == 0;
	daddr_t lbn, blkno;
	struct vm_page *pg;
	struct buf *mbp, *bp;
	struct vnode *devvp = VTOI(vp)->i_devvp;
	struct inode *ip = VTOI(vp);
	struct lfs *fs = ip->i_lfs;
	struct segment *sp = fs->lfs_sp;
	SEGSUM *ssp;
	UVMHIST_FUNC("lfs_gop_write"); UVMHIST_CALLED(ubchist);
	const char * failreason = NULL;

	ASSERT_SEGLOCK(fs);

	/* The Ifile lives in the buffer cache */
	KASSERT(vp != fs->lfs_ivnode);

	/*
	 * We don't want to fill the disk before the cleaner has a chance
	 * to make room for us.  If we're in danger of doing that, fail
	 * with EAGAIN.  The caller will have to notice this, unlock
	 * so the cleaner can run, relock and try again.
	 *
	 * We must write everything, however, if our vnode is being
	 * reclaimed.
	 */
	mutex_enter(vp->v_interlock);
	if (LFS_STARVED_FOR_SEGS(fs) && vdead_check(vp, VDEAD_NOWAIT) == 0) {
		mutex_exit(vp->v_interlock);
		failreason = "Starved for segs and not flushing vp";
 		goto tryagain;
	}
	mutex_exit(vp->v_interlock);

	/*
	 * Sometimes things slip past the filters in lfs_putpages,
	 * and the pagedaemon tries to write pages---problem is
	 * that the pagedaemon never acquires the segment lock.
	 *
	 * Alternatively, pages that were clean when we called
	 * genfs_putpages may have become dirty in the meantime.  In this
	 * case the segment header is not properly set up for blocks
	 * to be added to it.
	 *
	 * Unbusy and unclean the pages, and put them on the ACTIVE
	 * queue under the hypothesis that they couldn't have got here
	 * unless they were modified *quite* recently.
	 *
	 * XXXUBC that last statement is an oversimplification of course.
	 */
	if (!LFS_SEGLOCK_HELD(fs)) {
		failreason = "Seglock not held";
		goto tryagain;
	}
	if (ip->i_lfs_iflags & LFSI_NO_GOP_WRITE) {
		failreason = "Inode with no_gop_write";
		goto tryagain;
	}
	if ((pgs[0]->offset & lfs_sb_getbmask(fs)) != 0) {
		failreason = "Bad page offset";
		goto tryagain;
	}

	UVMHIST_LOG(ubchist, "vp %#jx pgs %#jx npages %jd flags 0x%jx",
	    (uintptr_t)vp, (uintptr_t)pgs, npages, flags);

	GOP_SIZE(vp, vp->v_size, &eof, 0);
	haveeof = 1;

	if (vp->v_type == VREG)
		fs_bshift = vp->v_mount->mnt_fs_bshift;
	else
		fs_bshift = DEV_BSHIFT;
	error = 0;
	pg = pgs[0];
	startoffset = pg->offset;
	KASSERT(eof >= 0);

	if (startoffset >= eof) {
		failreason = "Offset beyond EOF";
		goto tryagain;
	} else
		bytes = MIN(npages << PAGE_SHIFT, eof - startoffset);
	skipbytes = 0;

	KASSERT(bytes != 0);

	/* Swap PG_DELWRI for PG_PAGEOUT */
	for (i = 0; i < npages; i++) {
		if (pgs[i]->flags & PG_DELWRI) {
			KASSERT(!(pgs[i]->flags & PG_PAGEOUT));
			pgs[i]->flags &= ~PG_DELWRI;
			pgs[i]->flags |= PG_PAGEOUT;
			uvm_pageout_start(1);
			mutex_enter(vp->v_interlock);
			mutex_enter(&uvm_pageqlock);
			uvm_pageunwire(pgs[i]);
			mutex_exit(&uvm_pageqlock);
			mutex_exit(vp->v_interlock);
		}
	}

	/*
	 * Check to make sure we're starting on a block boundary.
	 * We'll check later to make sure we always write entire
	 * blocks (or fragments).
	 */
	if (startoffset & lfs_sb_getbmask(fs))
		printf("%" PRId64 " & %" PRIu64 " = %" PRId64 "\n",
		       startoffset, lfs_sb_getbmask(fs),
		       startoffset & lfs_sb_getbmask(fs));
	KASSERT((startoffset & lfs_sb_getbmask(fs)) == 0);
	if (bytes & lfs_sb_getffmask(fs)) {
		printf("lfs_gop_write: asked to write %ld bytes\n", (long)bytes);
		panic("lfs_gop_write: non-integer blocks");
	}

	/*
	 * We could deadlock here on pager_map with UVMPAGER_MAPIN_WAITOK.
	 * If we would, write what we have and try again.  If we don't
	 * have anything to write, we'll have to sleep.
	 */
	ssp = (SEGSUM *)sp->segsum;
	if ((kva = uvm_pagermapin(pgs, npages, UVMPAGER_MAPIN_WRITE |
				      (lfs_ss_getnfinfo(fs, ssp) < 1 ?
				       UVMPAGER_MAPIN_WAITOK : 0))) == 0x0) {
		DLOG((DLOG_PAGE, "lfs_gop_write: forcing write\n"));
#if 0
		      " with nfinfo=%d at offset 0x%jx\n",
		      (int)lfs_ss_getnfinfo(fs, ssp),
		      (uintmax_t)lfs_sb_getoffset(fs)));
#endif
		lfs_updatemeta(sp);
		lfs_release_finfo(fs);
		(void) lfs_writeseg(fs, sp);

		lfs_acquire_finfo(fs, ip->i_number, ip->i_gen);

		/*
		 * Having given up all of the pager_map we were holding,
		 * we can now wait for aiodoned to reclaim it for us
		 * without fear of deadlock.
		 */
		kva = uvm_pagermapin(pgs, npages, UVMPAGER_MAPIN_WRITE |
				     UVMPAGER_MAPIN_WAITOK);
	}

	mbp = getiobuf(NULL, true);
	UVMHIST_LOG(ubchist, "vp %#jx mbp %#jx num now %jd bytes 0x%jx",
	    (uintptr_t)vp, (uintptr_t)mbp, vp->v_numoutput, bytes);
	mbp->b_bufsize = npages << PAGE_SHIFT;
	mbp->b_data = (void *)kva;
	mbp->b_resid = mbp->b_bcount = bytes;
	mbp->b_cflags |= BC_BUSY|BC_AGE;
	mbp->b_iodone = uvm_aio_biodone;

	bp = NULL;
	for (offset = startoffset;
	    bytes > 0;
	    offset += iobytes, bytes -= iobytes) {
		lbn = offset >> fs_bshift;
		error = ulfs_bmaparray(vp, lbn, &blkno, NULL, NULL, &run,
		    lfs_issequential_hole);
		if (error) {
			UVMHIST_LOG(ubchist, "ulfs_bmaparray() -> %jd",
			    error,0,0,0);
			skipbytes += bytes;
			bytes = 0;
			break;
		}

		iobytes = MIN((((off_t)lbn + 1 + run) << fs_bshift) - offset,
		    bytes);
		if (blkno == (daddr_t)-1) {
			skipbytes += iobytes;
			continue;
		}

		/*
		 * Discover how much we can really pack into this buffer.
		 */
		/* If no room in the current segment, finish it up */
		if (sp->sum_bytes_left < sizeof(int32_t) ||
		    sp->seg_bytes_left < (1 << lfs_sb_getbshift(fs))) {
			int vers;

			lfs_updatemeta(sp);
			vers = lfs_fi_getversion(fs, sp->fip);
			lfs_release_finfo(fs);
			(void) lfs_writeseg(fs, sp);

			lfs_acquire_finfo(fs, ip->i_number, vers);
		}
		/* Check both for space in segment and space in segsum */
		iobytes = MIN(iobytes, (sp->seg_bytes_left >> fs_bshift)
					<< fs_bshift);
		iobytes = MIN(iobytes, (sp->sum_bytes_left / sizeof(int32_t))
				       << fs_bshift);
		KASSERT(iobytes > 0);

		/* if it's really one i/o, don't make a second buf */
		if (offset == startoffset && iobytes == bytes) {
			bp = mbp;
			/* 
			 * All the LFS output is done by the segwriter.  It
			 * will increment numoutput by one for all the bufs it
			 * receives.  However this buffer needs one extra to
			 * account for aiodone.
			 */
			mutex_enter(vp->v_interlock);
			vp->v_numoutput++;
			mutex_exit(vp->v_interlock);
		} else {
			bp = getiobuf(NULL, true);
			UVMHIST_LOG(ubchist, "vp %#jx bp %#jx num now %jd",
			    (uintptr_t)vp, (uintptr_t)bp, vp->v_numoutput, 0);
			nestiobuf_setup(mbp, bp, offset - pg->offset, iobytes);
			/*
			 * LFS doesn't like async I/O here, dies with
			 * an assert in lfs_bwrite().  Is that assert
			 * valid?  I retained non-async behaviour when
			 * converted this to use nestiobuf --pooka
			 */
			bp->b_flags &= ~B_ASYNC;
		}

		/* XXX This is silly ... is this necessary? */
		mutex_enter(&bufcache_lock);
		mutex_enter(vp->v_interlock);
		bgetvp(vp, bp);
		mutex_exit(vp->v_interlock);
		mutex_exit(&bufcache_lock);

		bp->b_lblkno = lfs_lblkno(fs, offset);
		bp->b_private = mbp;
		if (devvp->v_type == VBLK) {
			bp->b_dev = devvp->v_rdev;
		}
		VOP_BWRITE(bp->b_vp, bp);
		while (lfs_gatherblock(sp, bp, NULL))
			continue;
	}

	nestiobuf_done(mbp, skipbytes, error);
	if (skipbytes) {
		UVMHIST_LOG(ubchist, "skipbytes %jd", skipbytes, 0,0,0);
	}
	UVMHIST_LOG(ubchist, "returning 0", 0,0,0,0);

	if (!async) {
		/* Start a segment write. */
		UVMHIST_LOG(ubchist, "flushing", 0,0,0,0);
		mutex_enter(&lfs_lock);
		lfs_flush(fs, 0, 1);
		mutex_exit(&lfs_lock);
	}

	if ((sp->seg_flags & SEGM_SINGLE) && lfs_sb_getcurseg(fs) != fs->lfs_startseg)
		return EAGAIN;

	return (0);

    tryagain:
	/*
	 * We can't write the pages, for whatever reason.
	 * Clean up after ourselves, and make the caller try again.
	 */
	mutex_enter(vp->v_interlock);

	/* Tell why we're here, if we know */
	if (failreason != NULL) {
		DLOG((DLOG_PAGE, "lfs_gop_write: %s\n", failreason));
	}
	if (haveeof && startoffset >= eof) {
 		DLOG((DLOG_PAGE, "lfs_gop_write: ino %d start 0x%" PRIx64
 		      " eof 0x%" PRIx64 " npages=%d\n", VTOI(vp)->i_number,
 		      pgs[0]->offset, eof, npages));
	}

	mutex_enter(&uvm_pageqlock);
	for (i = 0; i < npages; i++) {
		pg = pgs[i];

		if (pg->flags & PG_PAGEOUT)
			uvm_pageout_done(1);
		if (pg->flags & PG_DELWRI) {
			uvm_pageunwire(pg);
		}
		uvm_pageactivate(pg);
		pg->flags &= ~(PG_CLEAN|PG_DELWRI|PG_PAGEOUT|PG_RELEASED);
		DLOG((DLOG_PAGE, "pg[%d] = %p (vp %p off %" PRIx64 ")\n", i, pg,
			vp, pg->offset));
		DLOG((DLOG_PAGE, "pg[%d]->flags = %x\n", i, pg->flags));
		DLOG((DLOG_PAGE, "pg[%d]->pqflags = %x\n", i, pg->pqflags));
		DLOG((DLOG_PAGE, "pg[%d]->uanon = %p\n", i, pg->uanon));
		DLOG((DLOG_PAGE, "pg[%d]->uobject = %p\n", i, pg->uobject));
		DLOG((DLOG_PAGE, "pg[%d]->wire_count = %d\n", i,
		      pg->wire_count));
		DLOG((DLOG_PAGE, "pg[%d]->loan_count = %d\n", i,
		      pg->loan_count));
	}
	/* uvm_pageunbusy takes care of PG_BUSY, PG_WANTED */
	uvm_page_unbusy(pgs, npages);
	mutex_exit(&uvm_pageqlock);
	mutex_exit(vp->v_interlock);
	return EAGAIN;
}

/*
 * finish vnode/inode initialization.
 * used by lfs_vget.
 */
void
lfs_vinit(struct mount *mp, struct vnode **vpp)
{
	struct vnode *vp = *vpp;
	struct inode *ip = VTOI(vp);
	struct ulfsmount *ump = VFSTOULFS(mp);
	struct lfs *fs = ump->um_lfs;
	int i;

	ip->i_mode = lfs_dino_getmode(fs, ip->i_din);
	ip->i_nlink = lfs_dino_getnlink(fs, ip->i_din);
	ip->i_lfs_osize = ip->i_size = lfs_dino_getsize(fs, ip->i_din);
	ip->i_flags = lfs_dino_getflags(fs, ip->i_din);
	ip->i_gen = lfs_dino_getgen(fs, ip->i_din);
	ip->i_uid = lfs_dino_getuid(fs, ip->i_din);
	ip->i_gid = lfs_dino_getgid(fs, ip->i_din);

	ip->i_lfs_effnblks = lfs_dino_getblocks(fs, ip->i_din);
	ip->i_lfs_odnlink = lfs_dino_getnlink(fs, ip->i_din);

	/*
	 * Initialize the vnode from the inode, check for aliases.  In all
	 * cases re-init ip, the underlying vnode/inode may have changed.
	 */
	ulfs_vinit(mp, lfs_specop_p, lfs_fifoop_p, &vp);
	ip = VTOI(vp);

	memset(ip->i_lfs_fragsize, 0, ULFS_NDADDR * sizeof(*ip->i_lfs_fragsize));
	if (vp->v_type != VLNK || ip->i_size >= ip->i_lfs->um_maxsymlinklen) {
#ifdef DEBUG
		for (i = (ip->i_size + lfs_sb_getbsize(fs) - 1) >> lfs_sb_getbshift(fs);
		    i < ULFS_NDADDR; i++) {
			if ((vp->v_type == VBLK || vp->v_type == VCHR) &&
			    i == 0)
				continue;
			if (lfs_dino_getdb(fs, ip->i_din, i) != 0) {
				lfs_dump_dinode(fs, ip->i_din);
				panic("inconsistent inode (direct)");
			}
		}
		for ( ; i < ULFS_NDADDR + ULFS_NIADDR; i++) {
			if (lfs_dino_getib(fs, ip->i_din, i - ULFS_NDADDR) != 0) {
				lfs_dump_dinode(fs, ip->i_din);
				panic("inconsistent inode (indirect)");
			}
		}
#endif /* DEBUG */
		for (i = 0; i < ULFS_NDADDR; i++)
			if (lfs_dino_getdb(fs, ip->i_din, i) != 0)
				ip->i_lfs_fragsize[i] = lfs_blksize(fs, ip, i);
	}

	KASSERTMSG((vp->v_type != VNON),
	    "lfs_vinit: ino %llu is type VNON! (ifmt=%o)\n",
	    (unsigned long long)ip->i_number,
	    (ip->i_mode & LFS_IFMT) >> 12);

	/*
	 * Finish inode initialization now that aliasing has been resolved.
	 */

	ip->i_devvp = fs->lfs_devvp;
	vref(ip->i_devvp);
#if defined(LFS_QUOTA) || defined(LFS_QUOTA2)
	ulfsquota_init(ip);
#endif
	genfs_node_init(vp, &lfs_genfsops);
	uvm_vnp_setsize(vp, ip->i_size);

	/* Initialize hiblk from file size */
	ip->i_lfs_hiblk = lfs_lblkno(ip->i_lfs, ip->i_size + lfs_sb_getbsize(ip->i_lfs) - 1) - 1;

	*vpp = vp;
}

/*
 * Resize the filesystem to contain the specified number of segments.
 */
int
lfs_resize_fs(struct lfs *fs, int newnsegs)
{
	SEGUSE *sup;
	CLEANERINFO *cip;
	struct buf *bp, *obp;
	daddr_t olast, nlast, ilast, noff, start, end;
	struct vnode *ivp;
	struct inode *ip;
	int error, badnews, inc, oldnsegs;
	int sbbytes, csbbytes, gain, cgain;
	int i;

	/* Only support v2 and up */
	if (lfs_sb_getversion(fs) < 2)
		return EOPNOTSUPP;

	/* If we're doing nothing, do it fast */
	oldnsegs = lfs_sb_getnseg(fs);
	if (newnsegs == oldnsegs)
		return 0;

	/* We always have to have two superblocks */
	if (newnsegs <= lfs_dtosn(fs, lfs_sb_getsboff(fs, 1)))
		/* XXX this error code is rather nonsense */
		return EFBIG;

	ivp = fs->lfs_ivnode;
	ip = VTOI(ivp);
	error = 0;

	/* Take the segment lock so no one else calls lfs_newseg() */
	lfs_seglock(fs, SEGM_PROT);

	/*
	 * Make sure the segments we're going to be losing, if any,
	 * are in fact empty.  We hold the seglock, so their status
	 * cannot change underneath us.  Count the superblocks we lose,
	 * while we're at it.
	 */
	sbbytes = csbbytes = 0;
	cgain = 0;
	for (i = newnsegs; i < oldnsegs; i++) {
		LFS_SEGENTRY(sup, fs, i, bp);
		badnews = sup->su_nbytes || !(sup->su_flags & SEGUSE_INVAL);
		if (sup->su_flags & SEGUSE_SUPERBLOCK)
			sbbytes += LFS_SBPAD;
		if (!(sup->su_flags & SEGUSE_DIRTY)) {
			++cgain;
			if (sup->su_flags & SEGUSE_SUPERBLOCK)
				csbbytes += LFS_SBPAD;
		}
		brelse(bp, 0);
		if (badnews) {
			error = EBUSY;
			goto out;
		}
	}

	/* Note old and new segment table endpoints, and old ifile size */
	olast = lfs_sb_getcleansz(fs) + lfs_sb_getsegtabsz(fs);
	nlast = howmany(newnsegs, lfs_sb_getsepb(fs)) + lfs_sb_getcleansz(fs);
	ilast = ivp->v_size >> lfs_sb_getbshift(fs);
	noff = nlast - olast;

	/*
	 * Make sure no one can use the Ifile while we change it around.
	 * Even after taking the iflock we need to make sure no one still
	 * is holding Ifile buffers, so we get each one, to drain them.
	 * (XXX this could be done better.)
	 */
	rw_enter(&fs->lfs_iflock, RW_WRITER);
	for (i = 0; i < ilast; i++) {
		/* XXX what to do if bread fails? */
		bread(ivp, i, lfs_sb_getbsize(fs), 0, &bp);
		brelse(bp, 0);
	}

	/* Allocate new Ifile blocks */
	for (i = ilast; i < ilast + noff; i++) {
		if (lfs_balloc(ivp, i * lfs_sb_getbsize(fs), lfs_sb_getbsize(fs), NOCRED, 0,
			       &bp) != 0)
			panic("balloc extending ifile");
		memset(bp->b_data, 0, lfs_sb_getbsize(fs));
		VOP_BWRITE(bp->b_vp, bp);
	}

	/* Register new ifile size */
	ip->i_size += noff * lfs_sb_getbsize(fs);
	lfs_dino_setsize(fs, ip->i_din, ip->i_size);
	uvm_vnp_setsize(ivp, ip->i_size);

	/* Copy the inode table to its new position */
	if (noff != 0) {
		if (noff < 0) {
			start = nlast;
			end = ilast + noff;
			inc = 1;
		} else {
			start = ilast + noff - 1;
			end = nlast - 1;
			inc = -1;
		}
		for (i = start; i != end; i += inc) {
			if (bread(ivp, i, lfs_sb_getbsize(fs),
			    B_MODIFY, &bp) != 0)
				panic("resize: bread dst blk failed");
			if (bread(ivp, i - noff, lfs_sb_getbsize(fs),
			    0, &obp))
				panic("resize: bread src blk failed");
			memcpy(bp->b_data, obp->b_data, lfs_sb_getbsize(fs));
			VOP_BWRITE(bp->b_vp, bp);
			brelse(obp, 0);
		}
	}

	/* If we are expanding, write the new empty SEGUSE entries */
	if (newnsegs > oldnsegs) {
		for (i = oldnsegs; i < newnsegs; i++) {
			if ((error = bread(ivp, i / lfs_sb_getsepb(fs) +
					   lfs_sb_getcleansz(fs), lfs_sb_getbsize(fs),
					   B_MODIFY, &bp)) != 0)
				panic("lfs: ifile read: %d", error);
			while ((i + 1) % lfs_sb_getsepb(fs) && i < newnsegs) {
				sup = &((SEGUSE *)bp->b_data)[i % lfs_sb_getsepb(fs)];
				memset(sup, 0, sizeof(*sup));
				i++;
			}
			VOP_BWRITE(bp->b_vp, bp);
		}
	}

	/* Zero out unused superblock offsets */
	for (i = 2; i < LFS_MAXNUMSB; i++)
		if (lfs_dtosn(fs, lfs_sb_getsboff(fs, i)) >= newnsegs)
			lfs_sb_setsboff(fs, i, 0x0);

	/*
	 * Correct superblock entries that depend on fs size.
	 * The computations of these are as follows:
	 *
	 * size  = lfs_segtod(fs, nseg)
	 * dsize = lfs_segtod(fs, nseg - minfreeseg) - lfs_btofsb(#super * LFS_SBPAD)
	 * bfree = dsize - lfs_btofsb(fs, bsize * nseg / 2) - blocks_actually_used
	 * avail = lfs_segtod(fs, nclean) - lfs_btofsb(#clean_super * LFS_SBPAD)
	 *         + (lfs_segtod(fs, 1) - (offset - curseg))
	 *	   - lfs_segtod(fs, minfreeseg - (minfreeseg / 2))
	 *
	 * XXX - we should probably adjust minfreeseg as well.
	 */
	gain = (newnsegs - oldnsegs);
	lfs_sb_setnseg(fs, newnsegs);
	lfs_sb_setsegtabsz(fs, nlast - lfs_sb_getcleansz(fs));
	lfs_sb_addsize(fs, gain * lfs_btofsb(fs, lfs_sb_getssize(fs)));
	lfs_sb_adddsize(fs, gain * lfs_btofsb(fs, lfs_sb_getssize(fs)) - lfs_btofsb(fs, sbbytes));
	lfs_sb_addbfree(fs, gain * lfs_btofsb(fs, lfs_sb_getssize(fs)) - lfs_btofsb(fs, sbbytes)
		       - gain * lfs_btofsb(fs, lfs_sb_getbsize(fs) / 2));
	if (gain > 0) {
		lfs_sb_addnclean(fs, gain);
		lfs_sb_addavail(fs, gain * lfs_btofsb(fs, lfs_sb_getssize(fs)));
	} else {
		lfs_sb_subnclean(fs, cgain);
		lfs_sb_subavail(fs, cgain * lfs_btofsb(fs, lfs_sb_getssize(fs)) -
				 lfs_btofsb(fs, csbbytes));
	}

	/* Resize segment flag cache */
	fs->lfs_suflags[0] = realloc(fs->lfs_suflags[0],
	    lfs_sb_getnseg(fs) * sizeof(u_int32_t), M_SEGMENT, M_WAITOK);
	fs->lfs_suflags[1] = realloc(fs->lfs_suflags[1],
	    lfs_sb_getnseg(fs) * sizeof(u_int32_t), M_SEGMENT, M_WAITOK);
	for (i = oldnsegs; i < newnsegs; i++)
		fs->lfs_suflags[0][i] = fs->lfs_suflags[1][i] = 0x0;

	/* Truncate Ifile if necessary */
	if (noff < 0)
		lfs_truncate(ivp, ivp->v_size + (noff << lfs_sb_getbshift(fs)), 0,
		    NOCRED);

	/* Update cleaner info so the cleaner can die */
	/* XXX what to do if bread fails? */
	bread(ivp, 0, lfs_sb_getbsize(fs), B_MODIFY, &bp);
	cip = bp->b_data;
	lfs_ci_setclean(fs, cip, lfs_sb_getnclean(fs));
	lfs_ci_setdirty(fs, cip, lfs_sb_getnseg(fs) - lfs_sb_getnclean(fs));
	VOP_BWRITE(bp->b_vp, bp);

	/* Let Ifile accesses proceed */
	rw_exit(&fs->lfs_iflock);

    out:
	lfs_segunlock(fs);
	return error;
}

/*
 * Extended attribute dispatch
 */
int
lfs_extattrctl(struct mount *mp, int cmd, struct vnode *vp,
	       int attrnamespace, const char *attrname)
{
#ifdef LFS_EXTATTR
	struct ulfsmount *ump;

	ump = VFSTOULFS(mp);
	if (ump->um_fstype == ULFS1) {
		return ulfs_extattrctl(mp, cmd, vp, attrnamespace, attrname);
	}
#endif
	return vfs_stdextattrctl(mp, cmd, vp, attrnamespace, attrname);
}