aboutsummaryrefslogtreecommitdiffstats
path: root/fs/xfs/xfs_trans.c
blob: 76922793f64fa0d58606a4411f008ba8b9e88259 (plain) (blame)
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
/*
 * Copyright (c) 2000-2003,2005 Silicon Graphics, Inc.
 * Copyright (C) 2010 Red Hat, 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_mount.h"
#include "xfs_error.h"
#include "xfs_da_btree.h"
#include "xfs_bmap_btree.h"
#include "xfs_alloc_btree.h"
#include "xfs_ialloc_btree.h"
#include "xfs_dinode.h"
#include "xfs_inode.h"
#include "xfs_btree.h"
#include "xfs_ialloc.h"
#include "xfs_alloc.h"
#include "xfs_bmap.h"
#include "xfs_quota.h"
#include "xfs_trans_priv.h"
#include "xfs_trans_space.h"
#include "xfs_inode_item.h"
#include "xfs_trace.h"

kmem_zone_t	*xfs_trans_zone;
kmem_zone_t	*xfs_log_item_desc_zone;


/*
 * Various log reservation values.
 *
 * These are based on the size of the file system block because that is what
 * most transactions manipulate.  Each adds in an additional 128 bytes per
 * item logged to try to account for the overhead of the transaction mechanism.
 *
 * Note:  Most of the reservations underestimate the number of allocation
 * groups into which they could free extents in the xfs_bmap_finish() call.
 * This is because the number in the worst case is quite high and quite
 * unusual.  In order to fix this we need to change xfs_bmap_finish() to free
 * extents in only a single AG at a time.  This will require changes to the
 * EFI code as well, however, so that the EFI for the extents not freed is
 * logged again in each transaction.  See SGI PV #261917.
 *
 * Reservation functions here avoid a huge stack in xfs_trans_init due to
 * register overflow from temporaries in the calculations.
 */


/*
 * In a write transaction we can allocate a maximum of 2
 * extents.  This gives:
 *    the inode getting the new extents: inode size
 *    the inode's bmap btree: max depth * block size
 *    the agfs of the ags from which the extents are allocated: 2 * sector
 *    the superblock free block counter: sector size
 *    the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size
 * And the bmap_finish transaction can free bmap blocks in a join:
 *    the agfs of the ags containing the blocks: 2 * sector size
 *    the agfls of the ags containing the blocks: 2 * sector size
 *    the super block free block counter: sector size
 *    the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size
 */
STATIC uint
xfs_calc_write_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		MAX((mp->m_sb.sb_inodesize +
		     XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK)) +
		     2 * mp->m_sb.sb_sectsize +
		     mp->m_sb.sb_sectsize +
		     XFS_ALLOCFREE_LOG_RES(mp, 2) +
		     128 * (4 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) +
			    XFS_ALLOCFREE_LOG_COUNT(mp, 2))),
		    (2 * mp->m_sb.sb_sectsize +
		     2 * mp->m_sb.sb_sectsize +
		     mp->m_sb.sb_sectsize +
		     XFS_ALLOCFREE_LOG_RES(mp, 2) +
		     128 * (5 + XFS_ALLOCFREE_LOG_COUNT(mp, 2))));
}

/*
 * In truncating a file we free up to two extents at once.  We can modify:
 *    the inode being truncated: inode size
 *    the inode's bmap btree: (max depth + 1) * block size
 * And the bmap_finish transaction can free the blocks and bmap blocks:
 *    the agf for each of the ags: 4 * sector size
 *    the agfl for each of the ags: 4 * sector size
 *    the super block to reflect the freed blocks: sector size
 *    worst case split in allocation btrees per extent assuming 4 extents:
 *		4 exts * 2 trees * (2 * max depth - 1) * block size
 *    the inode btree: max depth * blocksize
 *    the allocation btrees: 2 trees * (max depth - 1) * block size
 */
STATIC uint
xfs_calc_itruncate_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		MAX((mp->m_sb.sb_inodesize +
		     XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + 1) +
		     128 * (2 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK))),
		    (4 * mp->m_sb.sb_sectsize +
		     4 * mp->m_sb.sb_sectsize +
		     mp->m_sb.sb_sectsize +
		     XFS_ALLOCFREE_LOG_RES(mp, 4) +
		     128 * (9 + XFS_ALLOCFREE_LOG_COUNT(mp, 4)) +
		     128 * 5 +
		     XFS_ALLOCFREE_LOG_RES(mp, 1) +
		     128 * (2 + XFS_IALLOC_BLOCKS(mp) + mp->m_in_maxlevels +
			    XFS_ALLOCFREE_LOG_COUNT(mp, 1))));
}

/*
 * In renaming a files we can modify:
 *    the four inodes involved: 4 * inode size
 *    the two directory btrees: 2 * (max depth + v2) * dir block size
 *    the two directory bmap btrees: 2 * max depth * block size
 * And the bmap_finish transaction can free dir and bmap blocks (two sets
 *	of bmap blocks) giving:
 *    the agf for the ags in which the blocks live: 3 * sector size
 *    the agfl for the ags in which the blocks live: 3 * sector size
 *    the superblock for the free block count: sector size
 *    the allocation btrees: 3 exts * 2 trees * (2 * max depth - 1) * block size
 */
STATIC uint
xfs_calc_rename_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		MAX((4 * mp->m_sb.sb_inodesize +
		     2 * XFS_DIROP_LOG_RES(mp) +
		     128 * (4 + 2 * XFS_DIROP_LOG_COUNT(mp))),
		    (3 * mp->m_sb.sb_sectsize +
		     3 * mp->m_sb.sb_sectsize +
		     mp->m_sb.sb_sectsize +
		     XFS_ALLOCFREE_LOG_RES(mp, 3) +
		     128 * (7 + XFS_ALLOCFREE_LOG_COUNT(mp, 3))));
}

/*
 * For creating a link to an inode:
 *    the parent directory inode: inode size
 *    the linked inode: inode size
 *    the directory btree could split: (max depth + v2) * dir block size
 *    the directory bmap btree could join or split: (max depth + v2) * blocksize
 * And the bmap_finish transaction can free some bmap blocks giving:
 *    the agf for the ag in which the blocks live: sector size
 *    the agfl for the ag in which the blocks live: sector size
 *    the superblock for the free block count: sector size
 *    the allocation btrees: 2 trees * (2 * max depth - 1) * block size
 */
STATIC uint
xfs_calc_link_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		MAX((mp->m_sb.sb_inodesize +
		     mp->m_sb.sb_inodesize +
		     XFS_DIROP_LOG_RES(mp) +
		     128 * (2 + XFS_DIROP_LOG_COUNT(mp))),
		    (mp->m_sb.sb_sectsize +
		     mp->m_sb.sb_sectsize +
		     mp->m_sb.sb_sectsize +
		     XFS_ALLOCFREE_LOG_RES(mp, 1) +
		     128 * (3 + XFS_ALLOCFREE_LOG_COUNT(mp, 1))));
}

/*
 * For removing a directory entry we can modify:
 *    the parent directory inode: inode size
 *    the removed inode: inode size
 *    the directory btree could join: (max depth + v2) * dir block size
 *    the directory bmap btree could join or split: (max depth + v2) * blocksize
 * And the bmap_finish transaction can free the dir and bmap blocks giving:
 *    the agf for the ag in which the blocks live: 2 * sector size
 *    the agfl for the ag in which the blocks live: 2 * sector size
 *    the superblock for the free block count: sector size
 *    the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size
 */
STATIC uint
xfs_calc_remove_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		MAX((mp->m_sb.sb_inodesize +
		     mp->m_sb.sb_inodesize +
		     XFS_DIROP_LOG_RES(mp) +
		     128 * (2 + XFS_DIROP_LOG_COUNT(mp))),
		    (2 * mp->m_sb.sb_sectsize +
		     2 * mp->m_sb.sb_sectsize +
		     mp->m_sb.sb_sectsize +
		     XFS_ALLOCFREE_LOG_RES(mp, 2) +
		     128 * (5 + XFS_ALLOCFREE_LOG_COUNT(mp, 2))));
}

/*
 * For symlink we can modify:
 *    the parent directory inode: inode size
 *    the new inode: inode size
 *    the inode btree entry: 1 block
 *    the directory btree: (max depth + v2) * dir block size
 *    the directory inode's bmap btree: (max depth + v2) * block size
 *    the blocks for the symlink: 1 kB
 * Or in the first xact we allocate some inodes giving:
 *    the agi and agf of the ag getting the new inodes: 2 * sectorsize
 *    the inode blocks allocated: XFS_IALLOC_BLOCKS * blocksize
 *    the inode btree: max depth * blocksize
 *    the allocation btrees: 2 trees * (2 * max depth - 1) * block size
 */
STATIC uint
xfs_calc_symlink_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		MAX((mp->m_sb.sb_inodesize +
		     mp->m_sb.sb_inodesize +
		     XFS_FSB_TO_B(mp, 1) +
		     XFS_DIROP_LOG_RES(mp) +
		     1024 +
		     128 * (4 + XFS_DIROP_LOG_COUNT(mp))),
		    (2 * mp->m_sb.sb_sectsize +
		     XFS_FSB_TO_B(mp, XFS_IALLOC_BLOCKS(mp)) +
		     XFS_FSB_TO_B(mp, mp->m_in_maxlevels) +
		     XFS_ALLOCFREE_LOG_RES(mp, 1) +
		     128 * (2 + XFS_IALLOC_BLOCKS(mp) + mp->m_in_maxlevels +
			    XFS_ALLOCFREE_LOG_COUNT(mp, 1))));
}

/*
 * For create we can modify:
 *    the parent directory inode: inode size
 *    the new inode: inode size
 *    the inode btree entry: block size
 *    the superblock for the nlink flag: sector size
 *    the directory btree: (max depth + v2) * dir block size
 *    the directory inode's bmap btree: (max depth + v2) * block size
 * Or in the first xact we allocate some inodes giving:
 *    the agi and agf of the ag getting the new inodes: 2 * sectorsize
 *    the superblock for the nlink flag: sector size
 *    the inode blocks allocated: XFS_IALLOC_BLOCKS * blocksize
 *    the inode btree: max depth * blocksize
 *    the allocation btrees: 2 trees * (max depth - 1) * block size
 */
STATIC uint
xfs_calc_create_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		MAX((mp->m_sb.sb_inodesize +
		     mp->m_sb.sb_inodesize +
		     mp->m_sb.sb_sectsize +
		     XFS_FSB_TO_B(mp, 1) +
		     XFS_DIROP_LOG_RES(mp) +
		     128 * (3 + XFS_DIROP_LOG_COUNT(mp))),
		    (3 * mp->m_sb.sb_sectsize +
		     XFS_FSB_TO_B(mp, XFS_IALLOC_BLOCKS(mp)) +
		     XFS_FSB_TO_B(mp, mp->m_in_maxlevels) +
		     XFS_ALLOCFREE_LOG_RES(mp, 1) +
		     128 * (2 + XFS_IALLOC_BLOCKS(mp) + mp->m_in_maxlevels +
			    XFS_ALLOCFREE_LOG_COUNT(mp, 1))));
}

/*
 * Making a new directory is the same as creating a new file.
 */
STATIC uint
xfs_calc_mkdir_reservation(
	struct xfs_mount	*mp)
{
	return xfs_calc_create_reservation(mp);
}

/*
 * In freeing an inode we can modify:
 *    the inode being freed: inode size
 *    the super block free inode counter: sector size
 *    the agi hash list and counters: sector size
 *    the inode btree entry: block size
 *    the on disk inode before ours in the agi hash list: inode cluster size
 *    the inode btree: max depth * blocksize
 *    the allocation btrees: 2 trees * (max depth - 1) * block size
 */
STATIC uint
xfs_calc_ifree_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		mp->m_sb.sb_inodesize +
		mp->m_sb.sb_sectsize +
		mp->m_sb.sb_sectsize +
		XFS_FSB_TO_B(mp, 1) +
		MAX((__uint16_t)XFS_FSB_TO_B(mp, 1),
		    XFS_INODE_CLUSTER_SIZE(mp)) +
		128 * 5 +
		XFS_ALLOCFREE_LOG_RES(mp, 1) +
		128 * (2 + XFS_IALLOC_BLOCKS(mp) + mp->m_in_maxlevels +
		       XFS_ALLOCFREE_LOG_COUNT(mp, 1));
}

/*
 * When only changing the inode we log the inode and possibly the superblock
 * We also add a bit of slop for the transaction stuff.
 */
STATIC uint
xfs_calc_ichange_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		mp->m_sb.sb_inodesize +
		mp->m_sb.sb_sectsize +
		512;

}

/*
 * Growing the data section of the filesystem.
 *	superblock
 *	agi and agf
 *	allocation btrees
 */
STATIC uint
xfs_calc_growdata_reservation(
	struct xfs_mount	*mp)
{
	return mp->m_sb.sb_sectsize * 3 +
		XFS_ALLOCFREE_LOG_RES(mp, 1) +
		128 * (3 + XFS_ALLOCFREE_LOG_COUNT(mp, 1));
}

/*
 * Growing the rt section of the filesystem.
 * In the first set of transactions (ALLOC) we allocate space to the
 * bitmap or summary files.
 *	superblock: sector size
 *	agf of the ag from which the extent is allocated: sector size
 *	bmap btree for bitmap/summary inode: max depth * blocksize
 *	bitmap/summary inode: inode size
 *	allocation btrees for 1 block alloc: 2 * (2 * maxdepth - 1) * blocksize
 */
STATIC uint
xfs_calc_growrtalloc_reservation(
	struct xfs_mount	*mp)
{
	return 2 * mp->m_sb.sb_sectsize +
		XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK)) +
		mp->m_sb.sb_inodesize +
		XFS_ALLOCFREE_LOG_RES(mp, 1) +
		128 * (3 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) +
		       XFS_ALLOCFREE_LOG_COUNT(mp, 1));
}

/*
 * Growing the rt section of the filesystem.
 * In the second set of transactions (ZERO) we zero the new metadata blocks.
 *	one bitmap/summary block: blocksize
 */
STATIC uint
xfs_calc_growrtzero_reservation(
	struct xfs_mount	*mp)
{
	return mp->m_sb.sb_blocksize + 128;
}

/*
 * Growing the rt section of the filesystem.
 * In the third set of transactions (FREE) we update metadata without
 * allocating any new blocks.
 *	superblock: sector size
 *	bitmap inode: inode size
 *	summary inode: inode size
 *	one bitmap block: blocksize
 *	summary blocks: new summary size
 */
STATIC uint
xfs_calc_growrtfree_reservation(
	struct xfs_mount	*mp)
{
	return mp->m_sb.sb_sectsize +
		2 * mp->m_sb.sb_inodesize +
		mp->m_sb.sb_blocksize +
		mp->m_rsumsize +
		128 * 5;
}

/*
 * Logging the inode modification timestamp on a synchronous write.
 *	inode
 */
STATIC uint
xfs_calc_swrite_reservation(
	struct xfs_mount	*mp)
{
	return mp->m_sb.sb_inodesize + 128;
}

/*
 * Logging the inode mode bits when writing a setuid/setgid file
 *	inode
 */
STATIC uint
xfs_calc_writeid_reservation(xfs_mount_t *mp)
{
	return mp->m_sb.sb_inodesize + 128;
}

/*
 * Converting the inode from non-attributed to attributed.
 *	the inode being converted: inode size
 *	agf block and superblock (for block allocation)
 *	the new block (directory sized)
 *	bmap blocks for the new directory block
 *	allocation btrees
 */
STATIC uint
xfs_calc_addafork_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		mp->m_sb.sb_inodesize +
		mp->m_sb.sb_sectsize * 2 +
		mp->m_dirblksize +
		XFS_FSB_TO_B(mp, XFS_DAENTER_BMAP1B(mp, XFS_DATA_FORK) + 1) +
		XFS_ALLOCFREE_LOG_RES(mp, 1) +
		128 * (4 + XFS_DAENTER_BMAP1B(mp, XFS_DATA_FORK) + 1 +
		       XFS_ALLOCFREE_LOG_COUNT(mp, 1));
}

/*
 * Removing the attribute fork of a file
 *    the inode being truncated: inode size
 *    the inode's bmap btree: max depth * block size
 * And the bmap_finish transaction can free the blocks and bmap blocks:
 *    the agf for each of the ags: 4 * sector size
 *    the agfl for each of the ags: 4 * sector size
 *    the super block to reflect the freed blocks: sector size
 *    worst case split in allocation btrees per extent assuming 4 extents:
 *		4 exts * 2 trees * (2 * max depth - 1) * block size
 */
STATIC uint
xfs_calc_attrinval_reservation(
	struct xfs_mount	*mp)
{
	return MAX((mp->m_sb.sb_inodesize +
		    XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK)) +
		    128 * (1 + XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK))),
		   (4 * mp->m_sb.sb_sectsize +
		    4 * mp->m_sb.sb_sectsize +
		    mp->m_sb.sb_sectsize +
		    XFS_ALLOCFREE_LOG_RES(mp, 4) +
		    128 * (9 + XFS_ALLOCFREE_LOG_COUNT(mp, 4))));
}

/*
 * Setting an attribute.
 *	the inode getting the attribute
 *	the superblock for allocations
 *	the agfs extents are allocated from
 *	the attribute btree * max depth
 *	the inode allocation btree
 * Since attribute transaction space is dependent on the size of the attribute,
 * the calculation is done partially at mount time and partially at runtime.
 */
STATIC uint
xfs_calc_attrset_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		mp->m_sb.sb_inodesize +
		mp->m_sb.sb_sectsize +
		XFS_FSB_TO_B(mp, XFS_DA_NODE_MAXDEPTH) +
		128 * (2 + XFS_DA_NODE_MAXDEPTH);
}

/*
 * Removing an attribute.
 *    the inode: inode size
 *    the attribute btree could join: max depth * block size
 *    the inode bmap btree could join or split: max depth * block size
 * And the bmap_finish transaction can free the attr blocks freed giving:
 *    the agf for the ag in which the blocks live: 2 * sector size
 *    the agfl for the ag in which the blocks live: 2 * sector size
 *    the superblock for the free block count: sector size
 *    the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size
 */
STATIC uint
xfs_calc_attrrm_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		MAX((mp->m_sb.sb_inodesize +
		     XFS_FSB_TO_B(mp, XFS_DA_NODE_MAXDEPTH) +
		     XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK)) +
		     128 * (1 + XFS_DA_NODE_MAXDEPTH +
			    XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK))),
		    (2 * mp->m_sb.sb_sectsize +
		     2 * mp->m_sb.sb_sectsize +
		     mp->m_sb.sb_sectsize +
		     XFS_ALLOCFREE_LOG_RES(mp, 2) +
		     128 * (5 + XFS_ALLOCFREE_LOG_COUNT(mp, 2))));
}

/*
 * Clearing a bad agino number in an agi hash bucket.
 */
STATIC uint
xfs_calc_clear_agi_bucket_reservation(
	struct xfs_mount	*mp)
{
	return mp->m_sb.sb_sectsize + 128;
}

/*
 * Initialize the precomputed transaction reservation values
 * in the mount structure.
 */
void
xfs_trans_init(
	struct xfs_mount	*mp)
{
	struct xfs_trans_reservations *resp = &mp->m_reservations;

	resp->tr_write = xfs_calc_write_reservation(mp);
	resp->tr_itruncate = xfs_calc_itruncate_reservation(mp);
	resp->tr_rename = xfs_calc_rename_reservation(mp);
	resp->tr_link = xfs_calc_link_reservation(mp);
	resp->tr_remove = xfs_calc_remove_reservation(mp);
	resp->tr_symlink = xfs_calc_symlink_reservation(mp);
	resp->tr_create = xfs_calc_create_reservation(mp);
	resp->tr_mkdir = xfs_calc_mkdir_reservation(mp);
	resp->tr_ifree = xfs_calc_ifree_reservation(mp);
	resp->tr_ichange = xfs_calc_ichange_reservation(mp);
	resp->tr_growdata = xfs_calc_growdata_reservation(mp);
	resp->tr_swrite = xfs_calc_swrite_reservation(mp);
	resp->tr_writeid = xfs_calc_writeid_reservation(mp);
	resp->tr_addafork = xfs_calc_addafork_reservation(mp);
	resp->tr_attrinval = xfs_calc_attrinval_reservation(mp);
	resp->tr_attrset = xfs_calc_attrset_reservation(mp);
	resp->tr_attrrm = xfs_calc_attrrm_reservation(mp);
	resp->tr_clearagi = xfs_calc_clear_agi_bucket_reservation(mp);
	resp->tr_growrtalloc = xfs_calc_growrtalloc_reservation(mp);
	resp->tr_growrtzero = xfs_calc_growrtzero_reservation(mp);
	resp->tr_growrtfree = xfs_calc_growrtfree_reservation(mp);
}

/*
 * This routine is called to allocate a transaction structure.
 * The type parameter indicates the type of the transaction.  These
 * are enumerated in xfs_trans.h.
 *
 * Dynamically allocate the transaction structure from the transaction
 * zone, initialize it, and return it to the caller.
 */
xfs_trans_t *
xfs_trans_alloc(
	xfs_mount_t	*mp,
	uint		type)
{
	xfs_wait_for_freeze(mp, SB_FREEZE_TRANS);
	return _xfs_trans_alloc(mp, type, KM_SLEEP);
}

xfs_trans_t *
_xfs_trans_alloc(
	xfs_mount_t	*mp,
	uint		type,
	uint		memflags)
{
	xfs_trans_t	*tp;

	atomic_inc(&mp->m_active_trans);

	tp = kmem_zone_zalloc(xfs_trans_zone, memflags);
	tp->t_magic = XFS_TRANS_MAGIC;
	tp->t_type = type;
	tp->t_mountp = mp;
	INIT_LIST_HEAD(&tp->t_items);
	INIT_LIST_HEAD(&tp->t_busy);
	return tp;
}

/*
 * Free the transaction structure.  If there is more clean up
 * to do when the structure is freed, add it here.
 */
STATIC void
xfs_trans_free(
	struct xfs_trans	*tp)
{
	struct xfs_busy_extent	*busyp, *n;

	list_for_each_entry_safe(busyp, n, &tp->t_busy, list)
		xfs_alloc_busy_clear(tp->t_mountp, busyp);

	atomic_dec(&tp->t_mountp->m_active_trans);
	xfs_trans_free_dqinfo(tp);
	kmem_zone_free(xfs_trans_zone, tp);
}

/*
 * This is called to create a new transaction which will share the
 * permanent log reservation of the given transaction.  The remaining
 * unused block and rt extent reservations are also inherited.  This
 * implies that the original transaction is no longer allowed to allocate
 * blocks.  Locks and log items, however, are no inherited.  They must
 * be added to the new transaction explicitly.
 */
xfs_trans_t *
xfs_trans_dup(
	xfs_trans_t	*tp)
{
	xfs_trans_t	*ntp;

	ntp = kmem_zone_zalloc(xfs_trans_zone, KM_SLEEP);

	/*
	 * Initialize the new transaction structure.
	 */
	ntp->t_magic = XFS_TRANS_MAGIC;
	ntp->t_type = tp->t_type;
	ntp->t_mountp = tp->t_mountp;
	INIT_LIST_HEAD(&ntp->t_items);
	INIT_LIST_HEAD(&ntp->t_busy);

	ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES);
	ASSERT(tp->t_ticket != NULL);

	ntp->t_flags = XFS_TRANS_PERM_LOG_RES | (tp->t_flags & XFS_TRANS_RESERVE);
	ntp->t_ticket = xfs_log_ticket_get(tp->t_ticket);
	ntp->t_blk_res = tp->t_blk_res - tp->t_blk_res_used;
	tp->t_blk_res = tp->t_blk_res_used;
	ntp->t_rtx_res = tp->t_rtx_res - tp->t_rtx_res_used;
	tp->t_rtx_res = tp->t_rtx_res_used;
	ntp->t_pflags = tp->t_pflags;

	xfs_trans_dup_dqinfo(tp, ntp);

	atomic_inc(&tp->t_mountp->m_active_trans);
	return ntp;
}

/*
 * This is called to reserve free disk blocks and log space for the
 * given transaction.  This must be done before allocating any resources
 * within the transaction.
 *
 * This will return ENOSPC if there are not enough blocks available.
 * It will sleep waiting for available log space.
 * The only valid value for the flags parameter is XFS_RES_LOG_PERM, which
 * is used by long running transactions.  If any one of the reservations
 * fails then they will all be backed out.
 *
 * This does not do quota reservations. That typically is done by the
 * caller afterwards.
 */
int
xfs_trans_reserve(
	xfs_trans_t	*tp,
	uint		blocks,
	uint		logspace,
	uint		rtextents,
	uint		flags,
	uint		logcount)
{
	int		log_flags;
	int		error = 0;
	int		rsvd = (tp->t_flags & XFS_TRANS_RESERVE) != 0;

	/* Mark this thread as being in a transaction */
	current_set_flags_nested(&tp->t_pflags, PF_FSTRANS);

	/*
	 * Attempt to reserve the needed disk blocks by decrementing
	 * the number needed from the number available.  This will
	 * fail if the count would go below zero.
	 */
	if (blocks > 0) {
		error = xfs_icsb_modify_counters(tp->t_mountp, XFS_SBS_FDBLOCKS,
					  -((int64_t)blocks), rsvd);
		if (error != 0) {
			current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
			return (XFS_ERROR(ENOSPC));
		}
		tp->t_blk_res += blocks;
	}

	/*
	 * Reserve the log space needed for this transaction.
	 */
	if (logspace > 0) {
		ASSERT((tp->t_log_res == 0) || (tp->t_log_res == logspace));
		ASSERT((tp->t_log_count == 0) ||
			(tp->t_log_count == logcount));
		if (flags & XFS_TRANS_PERM_LOG_RES) {
			log_flags = XFS_LOG_PERM_RESERV;
			tp->t_flags |= XFS_TRANS_PERM_LOG_RES;
		} else {
			ASSERT(tp->t_ticket == NULL);
			ASSERT(!(tp->t_flags & XFS_TRANS_PERM_LOG_RES));
			log_flags = 0;
		}

		error = xfs_log_reserve(tp->t_mountp, logspace, logcount,
					&tp->t_ticket,
					XFS_TRANSACTION, log_flags, tp->t_type);
		if (error) {
			goto undo_blocks;
		}
		tp->t_log_res = logspace;
		tp->t_log_count = logcount;
	}

	/*
	 * Attempt to reserve the needed realtime extents by decrementing
	 * the number needed from the number available.  This will
	 * fail if the count would go below zero.
	 */
	if (rtextents > 0) {
		error = xfs_mod_incore_sb(tp->t_mountp, XFS_SBS_FREXTENTS,
					  -((int64_t)rtextents), rsvd);
		if (error) {
			error = XFS_ERROR(ENOSPC);
			goto undo_log;
		}
		tp->t_rtx_res += rtextents;
	}

	return 0;

	/*
	 * Error cases jump to one of these labels to undo any
	 * reservations which have already been performed.
	 */
undo_log:
	if (logspace > 0) {
		if (flags & XFS_TRANS_PERM_LOG_RES) {
			log_flags = XFS_LOG_REL_PERM_RESERV;
		} else {
			log_flags = 0;
		}
		xfs_log_done(tp->t_mountp, tp->t_ticket, NULL, log_flags);
		tp->t_ticket = NULL;
		tp->t_log_res = 0;
		tp->t_flags &= ~XFS_TRANS_PERM_LOG_RES;
	}

undo_blocks:
	if (blocks > 0) {
		xfs_icsb_modify_counters(tp->t_mountp, XFS_SBS_FDBLOCKS,
					 (int64_t)blocks, rsvd);
		tp->t_blk_res = 0;
	}

	current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);

	return error;
}

/*
 * Record the indicated change to the given field for application
 * to the file system's superblock when the transaction commits.
 * For now, just store the change in the transaction structure.
 *
 * Mark the transaction structure to indicate that the superblock
 * needs to be updated before committing.
 *
 * Because we may not be keeping track of allocated/free inodes and
 * used filesystem blocks in the superblock, we do not mark the
 * superblock dirty in this transaction if we modify these fields.
 * We still need to update the transaction deltas so that they get
 * applied to the incore superblock, but we don't want them to
 * cause the superblock to get locked and logged if these are the
 * only fields in the superblock that the transaction modifies.
 */
void
xfs_trans_mod_sb(
	xfs_trans_t	*tp,
	uint		field,
	int64_t		delta)
{
	uint32_t	flags = (XFS_TRANS_DIRTY|XFS_TRANS_SB_DIRTY);
	xfs_mount_t	*mp = tp->t_mountp;

	switch (field) {
	case XFS_TRANS_SB_ICOUNT:
		tp->t_icount_delta += delta;
		if (xfs_sb_version_haslazysbcount(&mp->m_sb))
			flags &= ~XFS_TRANS_SB_DIRTY;
		break;
	case XFS_TRANS_SB_IFREE:
		tp->t_ifree_delta += delta;
		if (xfs_sb_version_haslazysbcount(&mp->m_sb))
			flags &= ~XFS_TRANS_SB_DIRTY;
		break;
	case XFS_TRANS_SB_FDBLOCKS:
		/*
		 * Track the number of blocks allocated in the
		 * transaction.  Make sure it does not exceed the
		 * number reserved.
		 */
		if (delta < 0) {
			tp->t_blk_res_used += (uint)-delta;
			ASSERT(tp->t_blk_res_used <= tp->t_blk_res);
		}
		tp->t_fdblocks_delta += delta;
		if (xfs_sb_version_haslazysbcount(&mp->m_sb))
			flags &= ~XFS_TRANS_SB_DIRTY;
		break;
	case XFS_TRANS_SB_RES_FDBLOCKS:
		/*
		 * The allocation has already been applied to the
		 * in-core superblock's counter.  This should only
		 * be applied to the on-disk superblock.
		 */
		ASSERT(delta < 0);
		tp->t_res_fdblocks_delta += delta;
		if (xfs_sb_version_haslazysbcount(&mp->m_sb))
			flags &= ~XFS_TRANS_SB_DIRTY;
		break;
	case XFS_TRANS_SB_FREXTENTS:
		/*
		 * Track the number of blocks allocated in the
		 * transaction.  Make sure it does not exceed the
		 * number reserved.
		 */
		if (delta < 0) {
			tp->t_rtx_res_used += (uint)-delta;
			ASSERT(tp->t_rtx_res_used <= tp->t_rtx_res);
		}
		tp->t_frextents_delta += delta;
		break;
	case XFS_TRANS_SB_RES_FREXTENTS:
		/*
		 * The allocation has already been applied to the
		 * in-core superblock's counter.  This should only
		 * be applied to the on-disk superblock.
		 */
		ASSERT(delta < 0);
		tp->t_res_frextents_delta += delta;
		break;
	case XFS_TRANS_SB_DBLOCKS:
		ASSERT(delta > 0);
		tp->t_dblocks_delta += delta;
		break;
	case XFS_TRANS_SB_AGCOUNT:
		ASSERT(delta > 0);
		tp->t_agcount_delta += delta;
		break;
	case XFS_TRANS_SB_IMAXPCT:
		tp->t_imaxpct_delta += delta;
		break;
	case XFS_TRANS_SB_REXTSIZE:
		tp->t_rextsize_delta += delta;
		break;
	case XFS_TRANS_SB_RBMBLOCKS:
		tp->t_rbmblocks_delta += delta;
		break;
	case XFS_TRANS_SB_RBLOCKS:
		tp->t_rblocks_delta += delta;
		break;
	case XFS_TRANS_SB_REXTENTS:
		tp->t_rextents_delta += delta;
		break;
	case XFS_TRANS_SB_REXTSLOG:
		tp->t_rextslog_delta += delta;
		break;
	default:
		ASSERT(0);
		return;
	}

	tp->t_flags |= flags;
}

/*
 * xfs_trans_apply_sb_deltas() is called from the commit code
 * to bring the superblock buffer into the current transaction
 * and modify it as requested by earlier calls to xfs_trans_mod_sb().
 *
 * For now we just look at each field allowed to change and change
 * it if necessary.
 */
STATIC void
xfs_trans_apply_sb_deltas(
	xfs_trans_t	*tp)
{
	xfs_dsb_t	*sbp;
	xfs_buf_t	*bp;
	int		whole = 0;

	bp = xfs_trans_getsb(tp, tp->t_mountp, 0);
	sbp = XFS_BUF_TO_SBP(bp);

	/*
	 * Check that superblock mods match the mods made to AGF counters.
	 */
	ASSERT((tp->t_fdblocks_delta + tp->t_res_fdblocks_delta) ==
	       (tp->t_ag_freeblks_delta + tp->t_ag_flist_delta +
		tp->t_ag_btree_delta));

	/*
	 * Only update the superblock counters if we are logging them
	 */
	if (!xfs_sb_version_haslazysbcount(&(tp->t_mountp->m_sb))) {
		if (tp->t_icount_delta)
			be64_add_cpu(&sbp->sb_icount, tp->t_icount_delta);
		if (tp->t_ifree_delta)
			be64_add_cpu(&sbp->sb_ifree, tp->t_ifree_delta);
		if (tp->t_fdblocks_delta)
			be64_add_cpu(&sbp->sb_fdblocks, tp->t_fdblocks_delta);
		if (tp->t_res_fdblocks_delta)
			be64_add_cpu(&sbp->sb_fdblocks, tp->t_res_fdblocks_delta);
	}

	if (tp->t_frextents_delta)
		be64_add_cpu(&sbp->sb_frextents, tp->t_frextents_delta);
	if (tp->t_res_frextents_delta)
		be64_add_cpu(&sbp->sb_frextents, tp->t_res_frextents_delta);

	if (tp->t_dblocks_delta) {
		be64_add_cpu(&sbp->sb_dblocks, tp->t_dblocks_delta);
		whole = 1;
	}
	if (tp->t_agcount_delta) {
		be32_add_cpu(&sbp->sb_agcount, tp->t_agcount_delta);
		whole = 1;
	}
	if (tp->t_imaxpct_delta) {
		sbp->sb_imax_pct += tp->t_imaxpct_delta;
		whole = 1;
	}
	if (tp->t_rextsize_delta) {
		be32_add_cpu(&sbp->sb_rextsize, tp->t_rextsize_delta);
		whole = 1;
	}
	if (tp->t_rbmblocks_delta) {
		be32_add_cpu(&sbp->sb_rbmblocks, tp->t_rbmblocks_delta);
		whole = 1;
	}
	if (tp->t_rblocks_delta) {
		be64_add_cpu(&sbp->sb_rblocks, tp->t_rblocks_delta);
		whole = 1;
	}
	if (tp->t_rextents_delta) {
		be64_add_cpu(&sbp->sb_rextents, tp->t_rextents_delta);
		whole = 1;
	}
	if (tp->t_rextslog_delta) {
		sbp->sb_rextslog += tp->t_rextslog_delta;
		whole = 1;
	}

	if (whole)
		/*
		 * Log the whole thing, the fields are noncontiguous.
		 */
		xfs_trans_log_buf(tp, bp, 0, sizeof(xfs_dsb_t) - 1);
	else
		/*
		 * Since all the modifiable fields are contiguous, we
		 * can get away with this.
		 */
		xfs_trans_log_buf(tp, bp, offsetof(xfs_dsb_t, sb_icount),
				  offsetof(xfs_dsb_t, sb_frextents) +
				  sizeof(sbp->sb_frextents) - 1);
}

/*
 * xfs_trans_unreserve_and_mod_sb() is called to release unused reservations
 * and apply superblock counter changes to the in-core superblock.  The
 * t_res_fdblocks_delta and t_res_frextents_delta fields are explicitly NOT
 * applied to the in-core superblock.  The idea is that that has already been
 * done.
 *
 * This is done efficiently with a single call to xfs_mod_incore_sb_batch().
 * However, we have to ensure that we only modify each superblock field only
 * once because the application of the delta values may not be atomic. That can
 * lead to ENOSPC races occurring if we have two separate modifcations of the
 * free space counter to put back the entire reservation and then take away
 * what we used.
 *
 * If we are not logging superblock counters, then the inode allocated/free and
 * used block counts are not updated in the on disk superblock. In this case,
 * XFS_TRANS_SB_DIRTY will not be set when the transaction is updated but we
 * still need to update the incore superblock with the changes.
 */
void
xfs_trans_unreserve_and_mod_sb(
	xfs_trans_t	*tp)
{
	xfs_mod_sb_t	msb[9];	/* If you add cases, add entries */
	xfs_mod_sb_t	*msbp;
	xfs_mount_t	*mp = tp->t_mountp;
	/* REFERENCED */
	int		error;
	int		rsvd;
	int64_t		blkdelta = 0;
	int64_t		rtxdelta = 0;
	int64_t		idelta = 0;
	int64_t		ifreedelta = 0;

	msbp = msb;
	rsvd = (tp->t_flags & XFS_TRANS_RESERVE) != 0;

	/* calculate deltas */
	if (tp->t_blk_res > 0)
		blkdelta = tp->t_blk_res;
	if ((tp->t_fdblocks_delta != 0) &&
	    (xfs_sb_version_haslazysbcount(&mp->m_sb) ||
	     (tp->t_flags & XFS_TRANS_SB_DIRTY)))
	        blkdelta += tp->t_fdblocks_delta;

	if (tp->t_rtx_res > 0)
		rtxdelta = tp->t_rtx_res;
	if ((tp->t_frextents_delta != 0) &&
	    (tp->t_flags & XFS_TRANS_SB_DIRTY))
		rtxdelta += tp->t_frextents_delta;

	if (xfs_sb_version_haslazysbcount(&mp->m_sb) ||
	     (tp->t_flags & XFS_TRANS_SB_DIRTY)) {
		idelta = tp->t_icount_delta;
		ifreedelta = tp->t_ifree_delta;
	}

	/* apply the per-cpu counters */
	if (blkdelta) {
		error = xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS,
						 blkdelta, rsvd);
		if (error)
			goto out;
	}

	if (idelta) {
		error = xfs_icsb_modify_counters(mp, XFS_SBS_ICOUNT,
						 idelta, rsvd);
		if (error)
			goto out_undo_fdblocks;
	}

	if (ifreedelta) {
		error = xfs_icsb_modify_counters(mp, XFS_SBS_IFREE,
						 ifreedelta, rsvd);
		if (error)
			goto out_undo_icount;
	}

	/* apply remaining deltas */
	if (rtxdelta != 0) {
		msbp->msb_field = XFS_SBS_FREXTENTS;
		msbp->msb_delta = rtxdelta;
		msbp++;
	}

	if (tp->t_flags & XFS_TRANS_SB_DIRTY) {
		if (tp->t_dblocks_delta != 0) {
			msbp->msb_field = XFS_SBS_DBLOCKS;
			msbp->msb_delta = tp->t_dblocks_delta;
			msbp++;
		}
		if (tp->t_agcount_delta != 0) {
			msbp->msb_field = XFS_SBS_AGCOUNT;
			msbp->msb_delta = tp->t_agcount_delta;
			msbp++;
		}
		if (tp->t_imaxpct_delta != 0) {
			msbp->msb_field = XFS_SBS_IMAX_PCT;
			msbp->msb_delta = tp->t_imaxpct_delta;
			msbp++;
		}
		if (tp->t_rextsize_delta != 0) {
			msbp->msb_field = XFS_SBS_REXTSIZE;
			msbp->msb_delta = tp->t_rextsize_delta;
			msbp++;
		}
		if (tp->t_rbmblocks_delta != 0) {
			msbp->msb_field = XFS_SBS_RBMBLOCKS;
			msbp->msb_delta = tp->t_rbmblocks_delta;
			msbp++;
		}
		if (tp->t_rblocks_delta != 0) {
			msbp->msb_field = XFS_SBS_RBLOCKS;
			msbp->msb_delta = tp->t_rblocks_delta;
			msbp++;
		}
		if (tp->t_rextents_delta != 0) {
			msbp->msb_field = XFS_SBS_REXTENTS;
			msbp->msb_delta = tp->t_rextents_delta;
			msbp++;
		}
		if (tp->t_rextslog_delta != 0) {
			msbp->msb_field = XFS_SBS_REXTSLOG;
			msbp->msb_delta = tp->t_rextslog_delta;
			msbp++;
		}
	}

	/*
	 * If we need to change anything, do it.
	 */
	if (msbp > msb) {
		error = xfs_mod_incore_sb_batch(tp->t_mountp, msb,
			(uint)(msbp - msb), rsvd);
		if (error)
			goto out_undo_ifreecount;
	}

	return;

out_undo_ifreecount:
	if (ifreedelta)
		xfs_icsb_modify_counters(mp, XFS_SBS_IFREE, -ifreedelta, rsvd);
out_undo_icount:
	if (idelta)
		xfs_icsb_modify_counters(mp, XFS_SBS_ICOUNT, -idelta, rsvd);
out_undo_fdblocks:
	if (blkdelta)
		xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS, -blkdelta, rsvd);
out:
	ASSERT(error == 0);
	return;
}

/*
 * Add the given log item to the transaction's list of log items.
 *
 * The log item will now point to its new descriptor with its li_desc field.
 */
void
xfs_trans_add_item(
	struct xfs_trans	*tp,
	struct xfs_log_item	*lip)
{
	struct xfs_log_item_desc *lidp;

	ASSERT(lip->li_mountp = tp->t_mountp);
	ASSERT(lip->li_ailp = tp->t_mountp->m_ail);

	lidp = kmem_zone_zalloc(xfs_log_item_desc_zone, KM_SLEEP | KM_NOFS);

	lidp->lid_item = lip;
	lidp->lid_flags = 0;
	lidp->lid_size = 0;
	list_add_tail(&lidp->lid_trans, &tp->t_items);

	lip->li_desc = lidp;
}

STATIC void
xfs_trans_free_item_desc(
	struct xfs_log_item_desc *lidp)
{
	list_del_init(&lidp->lid_trans);
	kmem_zone_free(xfs_log_item_desc_zone, lidp);
}

/*
 * Unlink and free the given descriptor.
 */
void
xfs_trans_del_item(
	struct xfs_log_item	*lip)
{
	xfs_trans_free_item_desc(lip->li_desc);
	lip->li_desc = NULL;
}

/*
 * Unlock all of the items of a transaction and free all the descriptors
 * of that transaction.
 */
void
xfs_trans_free_items(
	struct xfs_trans	*tp,
	xfs_lsn_t		commit_lsn,
	int			flags)
{
	struct xfs_log_item_desc *lidp, *next;

	list_for_each_entry_safe(lidp, next, &tp->t_items, lid_trans) {
		struct xfs_log_item	*lip = lidp->lid_item;

		lip->li_desc = NULL;

		if (commit_lsn != NULLCOMMITLSN)
			IOP_COMMITTING(lip, commit_lsn);
		if (flags & XFS_TRANS_ABORT)
			lip->li_flags |= XFS_LI_ABORTED;
		IOP_UNLOCK(lip);

		xfs_trans_free_item_desc(lidp);
	}
}

/*
 * Unlock the items associated with a transaction.
 *
 * Items which were not logged should be freed.  Those which were logged must
 * still be tracked so they can be unpinned when the transaction commits.
 */
STATIC void
xfs_trans_unlock_items(
	struct xfs_trans	*tp,
	xfs_lsn_t		commit_lsn)
{
	struct xfs_log_item_desc *lidp, *next;

	list_for_each_entry_safe(lidp, next, &tp->t_items, lid_trans) {
		struct xfs_log_item	*lip = lidp->lid_item;

		lip->li_desc = NULL;

		if (commit_lsn != NULLCOMMITLSN)
			IOP_COMMITTING(lip, commit_lsn);
		IOP_UNLOCK(lip);

		/*
		 * Free the descriptor if the item is not dirty
		 * within this transaction.
		 */
		if (!(lidp->lid_flags & XFS_LID_DIRTY))
			xfs_trans_free_item_desc(lidp);
	}
}

/*
 * Total up the number of log iovecs needed to commit this
 * transaction.  The transaction itself needs one for the
 * transaction header.  Ask each dirty item in turn how many
 * it needs to get the total.
 */
static uint
xfs_trans_count_vecs(
	struct xfs_trans	*tp)
{
	int			nvecs;
	struct xfs_log_item_desc *lidp;

	nvecs = 1;

	/* In the non-debug case we need to start bailing out if we
	 * didn't find a log_item here, return zero and let trans_commit
	 * deal with it.
	 */
	if (list_empty(&tp->t_items)) {
		ASSERT(0);
		return 0;
	}

	list_for_each_entry(lidp, &tp->t_items, lid_trans) {
		/*
		 * Skip items which aren't dirty in this transaction.
		 */
		if (!(lidp->lid_flags & XFS_LID_DIRTY))
			continue;
		lidp->lid_size = IOP_SIZE(lidp->lid_item);
		nvecs += lidp->lid_size;
	}

	return nvecs;
}

/*
 * Fill in the vector with pointers to data to be logged
 * by this transaction.  The transaction header takes
 * the first vector, and then each dirty item takes the
 * number of vectors it indicated it needed in xfs_trans_count_vecs().
 *
 * As each item fills in the entries it needs, also pin the item
 * so that it cannot be flushed out until the log write completes.
 */
static void
xfs_trans_fill_vecs(
	struct xfs_trans	*tp,
	struct xfs_log_iovec	*log_vector)
{
	struct xfs_log_item_desc *lidp;
	struct xfs_log_iovec	*vecp;
	uint			nitems;

	/*
	 * Skip over the entry for the transaction header, we'll
	 * fill that in at the end.
	 */
	vecp = log_vector + 1;

	nitems = 0;
	ASSERT(!list_empty(&tp->t_items));
	list_for_each_entry(lidp, &tp->t_items, lid_trans) {
		/* Skip items which aren't dirty in this transaction. */
		if (!(lidp->lid_flags & XFS_LID_DIRTY))
			continue;

		/*
		 * The item may be marked dirty but not log anything.  This can
		 * be used to get called when a transaction is committed.
		 */
		if (lidp->lid_size)
			nitems++;
		IOP_FORMAT(lidp->lid_item, vecp);
		vecp += lidp->lid_size;
		IOP_PIN(lidp->lid_item);
	}

	/*
	 * Now that we've counted the number of items in this transaction, fill
	 * in the transaction header. Note that the transaction header does not
	 * have a log item.
	 */
	tp->t_header.th_magic = XFS_TRANS_HEADER_MAGIC;
	tp->t_header.th_type = tp->t_type;
	tp->t_header.th_num_items = nitems;
	log_vector->i_addr = (xfs_caddr_t)&tp->t_header;
	log_vector->i_len = sizeof(xfs_trans_header_t);
	log_vector->i_type = XLOG_REG_TYPE_TRANSHDR;
}

/*
 * The committed item processing consists of calling the committed routine of
 * each logged item, updating the item's position in the AIL if necessary, and
 * unpinning each item.  If the committed routine returns -1, then do nothing
 * further with the item because it may have been freed.
 *
 * Since items are unlocked when they are copied to the incore log, it is
 * possible for two transactions to be completing and manipulating the same
 * item simultaneously.  The AIL lock will protect the lsn field of each item.
 * The value of this field can never go backwards.
 *
 * We unpin the items after repositioning them in the AIL, because otherwise
 * they could be immediately flushed and we'd have to race with the flusher
 * trying to pull the item from the AIL as we add it.
 */
static void
xfs_trans_item_committed(
	struct xfs_log_item	*lip,
	xfs_lsn_t		commit_lsn,
	int			aborted)
{
	xfs_lsn_t		item_lsn;
	struct xfs_ail		*ailp;

	if (aborted)
		lip->li_flags |= XFS_LI_ABORTED;
	item_lsn = IOP_COMMITTED(lip, commit_lsn);

	/* If the committed routine returns -1, item has been freed. */
	if (XFS_LSN_CMP(item_lsn, (xfs_lsn_t)-1) == 0)
		return;

	/*
	 * If the returned lsn is greater than what it contained before, update
	 * the location of the item in the AIL.  If it is not, then do nothing.
	 * Items can never move backwards in the AIL.
	 *
	 * While the new lsn should usually be greater, it is possible that a
	 * later transaction completing simultaneously with an earlier one
	 * using the same item could complete first with a higher lsn.  This
	 * would cause the earlier transaction to fail the test below.
	 */
	ailp = lip->li_ailp;
	spin_lock(&ailp->xa_lock);
	if (XFS_LSN_CMP(item_lsn, lip->li_lsn) > 0) {
		/*
		 * This will set the item's lsn to item_lsn and update the
		 * position of the item in the AIL.
		 *
		 * xfs_trans_ail_update() drops the AIL lock.
		 */
		xfs_trans_ail_update(ailp, lip, item_lsn);
	} else {
		spin_unlock(&ailp->xa_lock);
	}

	/*
	 * Now that we've repositioned the item in the AIL, unpin it so it can
	 * be flushed. Pass information about buffer stale state down from the
	 * log item flags, if anyone else stales the buffer we do not want to
	 * pay any attention to it.
	 */
	IOP_UNPIN(lip, 0);
}

/*
 * This is typically called by the LM when a transaction has been fully
 * committed to disk.  It needs to unpin the items which have
 * been logged by the transaction and update their positions
 * in the AIL if necessary.
 *
 * This also gets called when the transactions didn't get written out
 * because of an I/O error. Abortflag & XFS_LI_ABORTED is set then.
 */
STATIC void
xfs_trans_committed(
	void			*arg,
	int			abortflag)
{
	struct xfs_trans	*tp = arg;
	struct xfs_log_item_desc *lidp, *next;

	list_for_each_entry_safe(lidp, next, &tp->t_items, lid_trans) {
		xfs_trans_item_committed(lidp->lid_item, tp->t_lsn, abortflag);
		xfs_trans_free_item_desc(lidp);
	}

	xfs_trans_free(tp);
}

static inline void
xfs_log_item_batch_insert(
	struct xfs_ail		*ailp,
	struct xfs_log_item	**log_items,
	int			nr_items,
	xfs_lsn_t		commit_lsn)
{
	int	i;

	spin_lock(&ailp->xa_lock);
	/* xfs_trans_ail_update_bulk drops ailp->xa_lock */
	xfs_trans_ail_update_bulk(ailp, log_items, nr_items, commit_lsn);

	for (i = 0; i < nr_items; i++)
		IOP_UNPIN(log_items[i], 0);
}

/*
 * Bulk operation version of xfs_trans_committed that takes a log vector of
 * items to insert into the AIL. This uses bulk AIL insertion techniques to
 * minimise lock traffic.
 *
 * If we are called with the aborted flag set, it is because a log write during
 * a CIL checkpoint commit has failed. In this case, all the items in the
 * checkpoint have already gone through IOP_COMMITED and IOP_UNLOCK, which
 * means that checkpoint commit abort handling is treated exactly the same
 * as an iclog write error even though we haven't started any IO yet. Hence in
 * this case all we need to do is IOP_COMMITTED processing, followed by an
 * IOP_UNPIN(aborted) call.
 */
void
xfs_trans_committed_bulk(
	struct xfs_ail		*ailp,
	struct xfs_log_vec	*log_vector,
	xfs_lsn_t		commit_lsn,
	int			aborted)
{
#define LOG_ITEM_BATCH_SIZE	32
	struct xfs_log_item	*log_items[LOG_ITEM_BATCH_SIZE];
	struct xfs_log_vec	*lv;
	int			i = 0;

	/* unpin all the log items */
	for (lv = log_vector; lv; lv = lv->lv_next ) {
		struct xfs_log_item	*lip = lv->lv_item;
		xfs_lsn_t		item_lsn;

		if (aborted)
			lip->li_flags |= XFS_LI_ABORTED;
		item_lsn = IOP_COMMITTED(lip, commit_lsn);

		/* item_lsn of -1 means the item was freed */
		if (XFS_LSN_CMP(item_lsn, (xfs_lsn_t)-1) == 0)
			continue;

		/*
		 * if we are aborting the operation, no point in inserting the
		 * object into the AIL as we are in a shutdown situation.
		 */
		if (aborted) {
			ASSERT(XFS_FORCED_SHUTDOWN(ailp->xa_mount));
			IOP_UNPIN(lip, 1);
			continue;
		}

		if (item_lsn != commit_lsn) {

			/*
			 * Not a bulk update option due to unusual item_lsn.
			 * Push into AIL immediately, rechecking the lsn once
			 * we have the ail lock. Then unpin the item.
			 */
			spin_lock(&ailp->xa_lock);
			if (XFS_LSN_CMP(item_lsn, lip->li_lsn) > 0)
				xfs_trans_ail_update(ailp, lip, item_lsn);
			else
				spin_unlock(&ailp->xa_lock);
			IOP_UNPIN(lip, 0);
			continue;
		}

		/* Item is a candidate for bulk AIL insert.  */
		log_items[i++] = lv->lv_item;
		if (i >= LOG_ITEM_BATCH_SIZE) {
			xfs_log_item_batch_insert(ailp, log_items,
					LOG_ITEM_BATCH_SIZE, commit_lsn);
			i = 0;
		}
	}

	/* make sure we insert the remainder! */
	if (i)
		xfs_log_item_batch_insert(ailp, log_items, i, commit_lsn);
}

/*
 * Called from the trans_commit code when we notice that the filesystem is in
 * the middle of a forced shutdown.
 *
 * When we are called here, we have already pinned all the items in the
 * transaction. However, neither IOP_COMMITTING or IOP_UNLOCK has been called
 * so we can simply walk the items in the transaction, unpin them with an abort
 * flag and then free the items. Note that unpinning the items can result in
 * them being freed immediately, so we need to use a safe list traversal method
 * here.
 */
STATIC void
xfs_trans_uncommit(
	struct xfs_trans	*tp,
	uint			flags)
{
	struct xfs_log_item_desc *lidp, *n;

	list_for_each_entry_safe(lidp, n, &tp->t_items, lid_trans) {
		if (lidp->lid_flags & XFS_LID_DIRTY)
			IOP_UNPIN(lidp->lid_item, 1);
	}

	xfs_trans_unreserve_and_mod_sb(tp);
	xfs_trans_unreserve_and_mod_dquots(tp);

	xfs_trans_free_items(tp, NULLCOMMITLSN, flags);
	xfs_trans_free(tp);
}

/*
 * Format the transaction direct to the iclog. This isolates the physical
 * transaction commit operation from the logical operation and hence allows
 * other methods to be introduced without affecting the existing commit path.
 */
static int
xfs_trans_commit_iclog(
	struct xfs_mount	*mp,
	struct xfs_trans	*tp,
	xfs_lsn_t		*commit_lsn,
	int			flags)
{
	int			shutdown;
	int			error;
	int			log_flags = 0;
	struct xlog_in_core	*commit_iclog;
#define XFS_TRANS_LOGVEC_COUNT  16
	struct xfs_log_iovec	log_vector_fast[XFS_TRANS_LOGVEC_COUNT];
	struct xfs_log_iovec	*log_vector;
	uint			nvec;


	/*
	 * Ask each log item how many log_vector entries it will
	 * need so we can figure out how many to allocate.
	 * Try to avoid the kmem_alloc() call in the common case
	 * by using a vector from the stack when it fits.
	 */
	nvec = xfs_trans_count_vecs(tp);
	if (nvec == 0) {
		return ENOMEM;	/* triggers a shutdown! */
	} else if (nvec <= XFS_TRANS_LOGVEC_COUNT) {
		log_vector = log_vector_fast;
	} else {
		log_vector = (xfs_log_iovec_t *)kmem_alloc(nvec *
						   sizeof(xfs_log_iovec_t),
						   KM_SLEEP);
	}

	/*
	 * Fill in the log_vector and pin the logged items, and
	 * then write the transaction to the log.
	 */
	xfs_trans_fill_vecs(tp, log_vector);

	if (flags & XFS_TRANS_RELEASE_LOG_RES)
		log_flags = XFS_LOG_REL_PERM_RESERV;

	error = xfs_log_write(mp, log_vector, nvec, tp->t_ticket, &(tp->t_lsn));

	/*
	 * The transaction is committed incore here, and can go out to disk
	 * at any time after this call.  However, all the items associated
	 * with the transaction are still locked and pinned in memory.
	 */
	*commit_lsn = xfs_log_done(mp, tp->t_ticket, &commit_iclog, log_flags);

	tp->t_commit_lsn = *commit_lsn;
	trace_xfs_trans_commit_lsn(tp);

	if (nvec > XFS_TRANS_LOGVEC_COUNT)
		kmem_free(log_vector);

	/*
	 * If we got a log write error. Unpin the logitems that we
	 * had pinned, clean up, free trans structure, and return error.
	 */
	if (error || *commit_lsn == -1) {
		current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
		xfs_trans_uncommit(tp, flags|XFS_TRANS_ABORT);
		return XFS_ERROR(EIO);
	}

	/*
	 * Once the transaction has committed, unused
	 * reservations need to be released and changes to
	 * the superblock need to be reflected in the in-core
	 * version.  Do that now.
	 */
	xfs_trans_unreserve_and_mod_sb(tp);

	/*
	 * Tell the LM to call the transaction completion routine
	 * when the log write with LSN commit_lsn completes (e.g.
	 * when the transaction commit really hits the on-disk log).
	 * After this call we cannot reference tp, because the call
	 * can happen at any time and the call will free the transaction
	 * structure pointed to by tp.  The only case where we call
	 * the completion routine (xfs_trans_committed) directly is
	 * if the log is turned off on a debug kernel or we're
	 * running in simulation mode (the log is explicitly turned
	 * off).
	 */
	tp->t_logcb.cb_func = xfs_trans_committed;
	tp->t_logcb.cb_arg = tp;

	/*
	 * We need to pass the iclog buffer which was used for the
	 * transaction commit record into this function, and attach
	 * the callback to it. The callback must be attached before
	 * the items are unlocked to avoid racing with other threads
	 * waiting for an item to unlock.
	 */
	shutdown = xfs_log_notify(mp, commit_iclog, &(tp->t_logcb));

	/*
	 * Mark this thread as no longer being in a transaction
	 */
	current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);

	/*
	 * Once all the items of the transaction have been copied
	 * to the in core log and the callback is attached, the
	 * items can be unlocked.
	 *
	 * This will free descriptors pointing to items which were
	 * not logged since there is nothing more to do with them.
	 * For items which were logged, we will keep pointers to them
	 * so they can be unpinned after the transaction commits to disk.
	 * This will also stamp each modified meta-data item with
	 * the commit lsn of this transaction for dependency tracking
	 * purposes.
	 */
	xfs_trans_unlock_items(tp, *commit_lsn);

	/*
	 * If we detected a log error earlier, finish committing
	 * the transaction now (unpin log items, etc).
	 *
	 * Order is critical here, to avoid using the transaction
	 * pointer after its been freed (by xfs_trans_committed
	 * either here now, or as a callback).  We cannot do this
	 * step inside xfs_log_notify as was done earlier because
	 * of this issue.
	 */
	if (shutdown)
		xfs_trans_committed(tp, XFS_LI_ABORTED);

	/*
	 * Now that the xfs_trans_committed callback has been attached,
	 * and the items are released we can finally allow the iclog to
	 * go to disk.
	 */
	return xfs_log_release_iclog(mp, commit_iclog);
}

/*
 * Walk the log items and allocate log vector structures for
 * each item large enough to fit all the vectors they require.
 * Note that this format differs from the old log vector format in
 * that there is no transaction header in these log vectors.
 */
STATIC struct xfs_log_vec *
xfs_trans_alloc_log_vecs(
	xfs_trans_t	*tp)
{
	struct xfs_log_item_desc *lidp;
	struct xfs_log_vec	*lv = NULL;
	struct xfs_log_vec	*ret_lv = NULL;


	/* Bail out if we didn't find a log item.  */
	if (list_empty(&tp->t_items)) {
		ASSERT(0);
		return NULL;
	}

	list_for_each_entry(lidp, &tp->t_items, lid_trans) {
		struct xfs_log_vec *new_lv;

		/* Skip items which aren't dirty in this transaction. */
		if (!(lidp->lid_flags & XFS_LID_DIRTY))
			continue;

		/* Skip items that do not have any vectors for writing */
		lidp->lid_size = IOP_SIZE(lidp->lid_item);
		if (!lidp->lid_size)
			continue;

		new_lv = kmem_zalloc(sizeof(*new_lv) +
				lidp->lid_size * sizeof(struct xfs_log_iovec),
				KM_SLEEP);

		/* The allocated iovec region lies beyond the log vector. */
		new_lv->lv_iovecp = (struct xfs_log_iovec *)&new_lv[1];
		new_lv->lv_niovecs = lidp->lid_size;
		new_lv->lv_item = lidp->lid_item;
		if (!ret_lv)
			ret_lv = new_lv;
		else
			lv->lv_next = new_lv;
		lv = new_lv;
	}

	return ret_lv;
}

static int
xfs_trans_commit_cil(
	struct xfs_mount	*mp,
	struct xfs_trans	*tp,
	xfs_lsn_t		*commit_lsn,
	int			flags)
{
	struct xfs_log_vec	*log_vector;

	/*
	 * Get each log item to allocate a vector structure for
	 * the log item to to pass to the log write code. The
	 * CIL commit code will format the vector and save it away.
	 */
	log_vector = xfs_trans_alloc_log_vecs(tp);
	if (!log_vector)
		return ENOMEM;

	xfs_log_commit_cil(mp, tp, log_vector, commit_lsn, flags);

	current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
	xfs_trans_free(tp);
	return 0;
}

/*
 * xfs_trans_commit
 *
 * Commit the given transaction to the log a/synchronously.
 *
 * XFS disk error handling mechanism is not based on a typical
 * transaction abort mechanism. Logically after the filesystem
 * gets marked 'SHUTDOWN', we can't let any new transactions
 * be durable - ie. committed to disk - because some metadata might
 * be inconsistent. In such cases, this returns an error, and the
 * caller may assume that all locked objects joined to the transaction
 * have already been unlocked as if the commit had succeeded.
 * Do not reference the transaction structure after this call.
 */
int
_xfs_trans_commit(
	struct xfs_trans	*tp,
	uint			flags,
	int			*log_flushed)
{
	struct xfs_mount	*mp = tp->t_mountp;
	xfs_lsn_t		commit_lsn = -1;
	int			error = 0;
	int			log_flags = 0;
	int			sync = tp->t_flags & XFS_TRANS_SYNC;

	/*
	 * Determine whether this commit is releasing a permanent
	 * log reservation or not.
	 */
	if (flags & XFS_TRANS_RELEASE_LOG_RES) {
		ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES);
		log_flags = XFS_LOG_REL_PERM_RESERV;
	}

	/*
	 * If there is nothing to be logged by the transaction,
	 * then unlock all of the items associated with the
	 * transaction and free the transaction structure.
	 * Also make sure to return any reserved blocks to
	 * the free pool.
	 */
	if (!(tp->t_flags & XFS_TRANS_DIRTY))
		goto out_unreserve;

	if (XFS_FORCED_SHUTDOWN(mp)) {
		error = XFS_ERROR(EIO);
		goto out_unreserve;
	}

	ASSERT(tp->t_ticket != NULL);

	/*
	 * If we need to update the superblock, then do it now.
	 */
	if (tp->t_flags & XFS_TRANS_SB_DIRTY)
		xfs_trans_apply_sb_deltas(tp);
	xfs_trans_apply_dquot_deltas(tp);

	if (mp->m_flags & XFS_MOUNT_DELAYLOG)
		error = xfs_trans_commit_cil(mp, tp, &commit_lsn, flags);
	else
		error = xfs_trans_commit_iclog(mp, tp, &commit_lsn, flags);

	if (error == ENOMEM) {
		xfs_force_shutdown(mp, SHUTDOWN_LOG_IO_ERROR);
		error = XFS_ERROR(EIO);
		goto out_unreserve;
	}

	/*
	 * If the transaction needs to be synchronous, then force the
	 * log out now and wait for it.
	 */
	if (sync) {
		if (!error) {
			error = _xfs_log_force_lsn(mp, commit_lsn,
				      XFS_LOG_SYNC, log_flushed);
		}
		XFS_STATS_INC(xs_trans_sync);
	} else {
		XFS_STATS_INC(xs_trans_async);
	}

	return error;

out_unreserve:
	xfs_trans_unreserve_and_mod_sb(tp);

	/*
	 * It is indeed possible for the transaction to be not dirty but
	 * the dqinfo portion to be.  All that means is that we have some
	 * (non-persistent) quota reservations that need to be unreserved.
	 */
	xfs_trans_unreserve_and_mod_dquots(tp);
	if (tp->t_ticket) {
		commit_lsn = xfs_log_done(mp, tp->t_ticket, NULL, log_flags);
		if (commit_lsn == -1 && !error)
			error = XFS_ERROR(EIO);
	}
	current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
	xfs_trans_free_items(tp, NULLCOMMITLSN, error ? XFS_TRANS_ABORT : 0);
	xfs_trans_free(tp);

	XFS_STATS_INC(xs_trans_empty);
	return error;
}

/*
 * Unlock all of the transaction's items and free the transaction.
 * The transaction must not have modified any of its items, because
 * there is no way to restore them to their previous state.
 *
 * If the transaction has made a log reservation, make sure to release
 * it as well.
 */
void
xfs_trans_cancel(
	xfs_trans_t		*tp,
	int			flags)
{
	int			log_flags;
	xfs_mount_t		*mp = tp->t_mountp;

	/*
	 * See if the caller is being too lazy to figure out if
	 * the transaction really needs an abort.
	 */
	if ((flags & XFS_TRANS_ABORT) && !(tp->t_flags & XFS_TRANS_DIRTY))
		flags &= ~XFS_TRANS_ABORT;
	/*
	 * See if the caller is relying on us to shut down the
	 * filesystem.  This happens in paths where we detect
	 * corruption and decide to give up.
	 */
	if ((tp->t_flags & XFS_TRANS_DIRTY) && !XFS_FORCED_SHUTDOWN(mp)) {
		XFS_ERROR_REPORT("xfs_trans_cancel", XFS_ERRLEVEL_LOW, mp);
		xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE);
	}
#ifdef DEBUG
	if (!(flags & XFS_TRANS_ABORT) && !XFS_FORCED_SHUTDOWN(mp)) {
		struct xfs_log_item_desc *lidp;

		list_for_each_entry(lidp, &tp->t_items, lid_trans)
			ASSERT(!(lidp->lid_item->li_type == XFS_LI_EFD));
	}
#endif
	xfs_trans_unreserve_and_mod_sb(tp);
	xfs_trans_unreserve_and_mod_dquots(tp);

	if (tp->t_ticket) {
		if (flags & XFS_TRANS_RELEASE_LOG_RES) {
			ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES);
			log_flags = XFS_LOG_REL_PERM_RESERV;
		} else {
			log_flags = 0;
		}
		xfs_log_done(mp, tp->t_ticket, NULL, log_flags);
	}

	/* mark this thread as no longer being in a transaction */
	current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);

	xfs_trans_free_items(tp, NULLCOMMITLSN, flags);
	xfs_trans_free(tp);
}

/*
 * Roll from one trans in the sequence of PERMANENT transactions to
 * the next: permanent transactions are only flushed out when
 * committed with XFS_TRANS_RELEASE_LOG_RES, but we still want as soon
 * as possible to let chunks of it go to the log. So we commit the
 * chunk we've been working on and get a new transaction to continue.
 */
int
xfs_trans_roll(
	struct xfs_trans	**tpp,
	struct xfs_inode	*dp)
{
	struct xfs_trans	*trans;
	unsigned int		logres, count;
	int			error;

	/*
	 * Ensure that the inode is always logged.
	 */
	trans = *tpp;
	xfs_trans_log_inode(trans, dp, XFS_ILOG_CORE);

	/*
	 * Copy the critical parameters from one trans to the next.
	 */
	logres = trans->t_log_res;
	count = trans->t_log_count;
	*tpp = xfs_trans_dup(trans);

	/*
	 * Commit the current transaction.
	 * If this commit failed, then it'd just unlock those items that
	 * are not marked ihold. That also means that a filesystem shutdown
	 * is in progress. The caller takes the responsibility to cancel
	 * the duplicate transaction that gets returned.
	 */
	error = xfs_trans_commit(trans, 0);
	if (error)
		return (error);

	trans = *tpp;

	/*
	 * transaction commit worked ok so we can drop the extra ticket
	 * reference that we gained in xfs_trans_dup()
	 */
	xfs_log_ticket_put(trans->t_ticket);


	/*
	 * Reserve space in the log for th next transaction.
	 * This also pushes items in the "AIL", the list of logged items,
	 * out to disk if they are taking up space at the tail of the log
	 * that we want to use.  This requires that either nothing be locked
	 * across this call, or that anything that is locked be logged in
	 * the prior and the next transactions.
	 */
	error = xfs_trans_reserve(trans, 0, logres, 0,
				  XFS_TRANS_PERM_LOG_RES, count);
	/*
	 *  Ensure that the inode is in the new transaction and locked.
	 */
	if (error)
		return error;

	xfs_trans_ijoin(trans, dp);
	return 0;
}