summaryrefslogtreecommitdiff
path: root/sys/ufs/lfs/ulfs_extattr.c
blob: 64ed1e1599fab2a3f6438f282f1d374f04a4f955 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
/*	$NetBSD: ulfs_extattr.c,v 1.7 2014/02/07 15:29:23 hannken Exp $	*/
/*  from NetBSD: ufs_extattr.c,v 1.41 2012/12/08 13:42:36 manu Exp  */

/*-
 * Copyright (c) 1999-2002 Robert N. M. Watson
 * Copyright (c) 2002-2003 Networks Associates Technology, Inc.
 * All rights reserved.
 *
 * This software was developed by Robert Watson for the TrustedBSD Project.
 *
 * This software was developed for the FreeBSD Project in part by Network
 * Associates Laboratories, the Security Research Division of Network
 * Associates, Inc. under DARPA/SPAWAR contract N66001-01-C-8035 ("CBOSS"),
 * as part of the DARPA CHATS research program.
 *
 * 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 AUTHOR 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 AUTHOR 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.
 *
 */

/*
 * Support for file system extended attributes on the ULFS1 file system.
 *
 * Extended attributes are defined in the form name=value, where name is
 * a nul-terminated string in the style of a file name, and value is a
 * binary blob of zero or more bytes.  The ULFS1 extended attribute service
 * layers support for extended attributes onto a backing file, in the style
 * of the quota implementation, meaning that it requires no underlying format
 * changes to the file system.  This design choice exchanges simplicity,
 * usability, and easy deployment for performance.
 */

#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: ulfs_extattr.c,v 1.7 2014/02/07 15:29:23 hannken Exp $");

#ifdef _KERNEL_OPT
#include "opt_lfs.h"
#endif

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/reboot.h>
#include <sys/kauth.h>
#include <sys/kernel.h>
#include <sys/namei.h>
#include <sys/kmem.h>
#include <sys/fcntl.h>
#include <sys/lwp.h>
#include <sys/vnode.h>
#include <sys/mount.h>
#include <sys/lock.h>
#include <sys/dirent.h>
#include <sys/extattr.h>
#include <sys/sysctl.h>

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

int ulfs_extattr_sync = 1;
int ulfs_extattr_autocreate = 1024;

static int	ulfs_extattr_valid_attrname(int attrnamespace,
		    const char *attrname);
static int	ulfs_extattr_enable_with_open(struct ulfsmount *ump,
		    struct vnode *vp, int attrnamespace, const char *attrname,
		    struct lwp *l);
static int	ulfs_extattr_enable(struct ulfsmount *ump, int attrnamespace,
		    const char *attrname, struct vnode *backing_vnode,
		    struct lwp *l);
static int	ulfs_extattr_disable(struct ulfsmount *ump, int attrnamespace,
		    const char *attrname, struct lwp *l);
static int	ulfs_extattr_get(struct vnode *vp, int attrnamespace,
		    const char *name, struct uio *uio, size_t *size,
		    kauth_cred_t cred, struct lwp *l);
static int	ulfs_extattr_list(struct vnode *vp, int attrnamespace,
		    struct uio *uio, size_t *size, int flag,
		    kauth_cred_t cred, struct lwp *l);
static int	ulfs_extattr_set(struct vnode *vp, int attrnamespace,
		    const char *name, struct uio *uio, kauth_cred_t cred,
		    struct lwp *l);
static int	ulfs_extattr_rm(struct vnode *vp, int attrnamespace,
		    const char *name, kauth_cred_t cred, struct lwp *l);
static struct ulfs_extattr_list_entry *ulfs_extattr_find_attr(struct ulfsmount *,
		    int, const char *);
static int	ulfs_extattr_get_header(struct vnode *, 
		    struct ulfs_extattr_list_entry *, 
		    struct ulfs_extattr_header *, off_t *);

/*
 * Convert a FreeBSD extended attribute and namespace to a consistent string
 * representation.
 *
 * The returned value, if not NULL, is guaranteed to be an allocated object
 * of its size as returned by strlen() + 1 and must be freed by the caller.
 */
static char *
from_freebsd_extattr(int attrnamespace, const char *attrname)
{
	const char *namespace;
	char *attr;
	size_t len;

	if (attrnamespace == EXTATTR_NAMESPACE_SYSTEM)
		namespace = "system";
	else if (attrnamespace == EXTATTR_NAMESPACE_USER)
		namespace = "user";
	else
		return NULL;

	/* <namespace>.<attrname>\0 */
	len = strlen(namespace) + 1 + strlen(attrname) + 1;

	attr = kmem_alloc(len, KM_SLEEP);

	snprintf(attr, len, "%s.%s", namespace, attrname);

	return attr;
}

/*
 * Internal wrapper around a conversion-check-free sequence.
 */
static int
internal_extattr_check_cred(vnode_t *vp, int attrnamespace, const char *name,
    kauth_cred_t cred, int access_mode)
{
	char *attr;
	int error;

	attr = from_freebsd_extattr(attrnamespace, name);
	if (attr == NULL)
		return EINVAL;

	error = extattr_check_cred(vp, attr, cred, access_mode);

	kmem_free(attr, strlen(attr) + 1);

	return error;
}

/*
 * Per-FS attribute lock protecting attribute operations.
 * XXX Right now there is a lot of lock contention due to having a single
 * lock per-FS; really, this should be far more fine-grained.
 */
static void
ulfs_extattr_uepm_lock(struct ulfsmount *ump)
{

	/* XXX Why does this need to be recursive? */
	if (mutex_owned(&ump->um_extattr.uepm_lock)) {
		ump->um_extattr.uepm_lockcnt++;
		return;
	}
	mutex_enter(&ump->um_extattr.uepm_lock);
}

static void
ulfs_extattr_uepm_unlock(struct ulfsmount *ump)
{

	if (ump->um_extattr.uepm_lockcnt != 0) {
		KASSERT(mutex_owned(&ump->um_extattr.uepm_lock));
		ump->um_extattr.uepm_lockcnt--;
		return;
	}
	mutex_exit(&ump->um_extattr.uepm_lock);
}

/*-
 * Determine whether the name passed is a valid name for an actual
 * attribute.
 *
 * Invalid currently consists of:
 *	 NULL pointer for attrname
 *	 zero-length attrname (used to retrieve application attribute list)
 */
static int
ulfs_extattr_valid_attrname(int attrnamespace, const char *attrname)
{

	if (attrname == NULL)
		return (0);
	if (strlen(attrname) == 0)
		return (0);
	return (1);
}

/*
 * Autocreate an attribute storage
 */
static struct ulfs_extattr_list_entry *
ulfs_extattr_autocreate_attr(struct vnode *vp, int attrnamespace,
    const char *attrname, struct lwp *l)
{
	struct mount *mp = vp->v_mount;
	struct ulfsmount *ump = VFSTOULFS(mp);
	struct vnode *backing_vp;
	struct nameidata nd;
	struct pathbuf *pb;
	char *path;
	struct ulfs_extattr_fileheader uef;
	struct ulfs_extattr_list_entry *uele;
	int error;

	path = PNBUF_GET();

	/* 
	 * We only support system and user namespace autocreation
	 */ 
	switch (attrnamespace) {
	case EXTATTR_NAMESPACE_SYSTEM:
		(void)snprintf(path, PATH_MAX, "%s/%s/%s/%s", 
			       mp->mnt_stat.f_mntonname,
			       ULFS_EXTATTR_FSROOTSUBDIR,
			       ULFS_EXTATTR_SUBDIR_SYSTEM,
			       attrname);
		break;
	case EXTATTR_NAMESPACE_USER:
		(void)snprintf(path, PATH_MAX, "%s/%s/%s/%s", 
			       mp->mnt_stat.f_mntonname,
			       ULFS_EXTATTR_FSROOTSUBDIR,
			       ULFS_EXTATTR_SUBDIR_USER,
			       attrname);
		break;
	default:
		PNBUF_PUT(path);
		return NULL;
		break;
	}

	/*
	 * XXX unlock/lock should only be done when setting extattr
	 * on backing store or one of its parent directory 
	 * including root, but we always do it for now.
	 */ 
	KASSERT(VOP_ISLOCKED(vp) == LK_EXCLUSIVE);
	VOP_UNLOCK(vp);

	pb = pathbuf_create(path);
	NDINIT(&nd, CREATE, LOCKPARENT, pb);
	
	error = vn_open(&nd, O_CREAT|O_RDWR, 0600);

	/*
	 * Reacquire the lock on the vnode
	 */
	KASSERT(VOP_ISLOCKED(vp) == 0);
	vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);

	if (error != 0) {
		pathbuf_destroy(pb);
		PNBUF_PUT(path);
		return NULL;
	}

	KASSERT(nd.ni_vp != NULL);
	KASSERT(VOP_ISLOCKED(nd.ni_vp) == LK_EXCLUSIVE);
	KASSERT(VOP_ISLOCKED(nd.ni_dvp) == 0);

	/*
 	 * backing_vp is the backing store. 
	 */	
	backing_vp = nd.ni_vp;
	pathbuf_destroy(pb);
	PNBUF_PUT(path);

	uef.uef_magic = ULFS_EXTATTR_MAGIC;
	uef.uef_version = ULFS_EXTATTR_VERSION;
	uef.uef_size = ulfs_extattr_autocreate;

	error = vn_rdwr(UIO_WRITE, backing_vp, &uef, sizeof(uef), 0,
		        UIO_SYSSPACE, IO_NODELOCKED|IO_APPEND, 
			l->l_cred, NULL, l);

	VOP_UNLOCK(backing_vp);

	if (error != 0) {
		printf("%s: write uef header failed for %s, error = %d\n", 
		       __func__, attrname, error);
		vn_close(backing_vp, FREAD|FWRITE, l->l_cred);
		return NULL;
	}

	/*
	 * Now enable attribute. 
	 */
	error = ulfs_extattr_enable(ump,attrnamespace, attrname, backing_vp, l);
	KASSERT(VOP_ISLOCKED(backing_vp) == 0);

	if (error != 0) {
		printf("%s: enable %s failed, error %d\n", 
		       __func__, attrname, error);
		vn_close(backing_vp, FREAD|FWRITE, l->l_cred);
		return NULL;
	}

	uele = ulfs_extattr_find_attr(ump, attrnamespace, attrname);
	if (uele == NULL) {
		printf("%s: atttribute %s created but not found!\n",
		       __func__, attrname);
		vn_close(backing_vp, FREAD|FWRITE, l->l_cred);
		return NULL;
	}

	printf("%s: EA backing store autocreated for %s\n",
	       mp->mnt_stat.f_mntonname, attrname);

	return uele;
}

/*
 * Locate an attribute given a name and mountpoint.
 * Must be holding uepm lock for the mount point.
 */
static struct ulfs_extattr_list_entry *
ulfs_extattr_find_attr(struct ulfsmount *ump, int attrnamespace,
    const char *attrname)
{
	struct ulfs_extattr_list_entry *search_attribute;

	for (search_attribute = LIST_FIRST(&ump->um_extattr.uepm_list);
	    search_attribute != NULL;
	    search_attribute = LIST_NEXT(search_attribute, uele_entries)) {
		if (!(strncmp(attrname, search_attribute->uele_attrname,
		    ULFS_EXTATTR_MAXEXTATTRNAME)) &&
		    (attrnamespace == search_attribute->uele_attrnamespace)) {
			return (search_attribute);
		}
	}

	return (0);
}

/*
 * Initialize per-FS structures supporting extended attributes.  Do not
 * start extended attributes yet.
 */
void
ulfs_extattr_uepm_init(struct ulfs_extattr_per_mount *uepm)
{

	uepm->uepm_flags = 0;
	uepm->uepm_lockcnt = 0;

	LIST_INIT(&uepm->uepm_list);
	mutex_init(&uepm->uepm_lock, MUTEX_DEFAULT, IPL_NONE);
	uepm->uepm_flags |= ULFS_EXTATTR_UEPM_INITIALIZED;
}

/*
 * Destroy per-FS structures supporting extended attributes.  Assumes
 * that EAs have already been stopped, and will panic if not.
 */
void
ulfs_extattr_uepm_destroy(struct ulfs_extattr_per_mount *uepm)
{

	if (!(uepm->uepm_flags & ULFS_EXTATTR_UEPM_INITIALIZED))
		panic("ulfs_extattr_uepm_destroy: not initialized");

	if ((uepm->uepm_flags & ULFS_EXTATTR_UEPM_STARTED))
		panic("ulfs_extattr_uepm_destroy: called while still started");

	/*
	 * It's not clear that either order for the next two lines is
	 * ideal, and it should never be a problem if this is only called
	 * during unmount, and with vfs_busy().
	 */
	uepm->uepm_flags &= ~ULFS_EXTATTR_UEPM_INITIALIZED;
	mutex_destroy(&uepm->uepm_lock);
}

/*
 * Start extended attribute support on an FS.
 */
int
ulfs_extattr_start(struct mount *mp, struct lwp *l)
{
	struct ulfsmount *ump;
	int error = 0;

	ump = VFSTOULFS(mp);

	ulfs_extattr_uepm_lock(ump);

	if (!(ump->um_extattr.uepm_flags & ULFS_EXTATTR_UEPM_INITIALIZED)) {
		error = EOPNOTSUPP;
		goto unlock;
	}
	if (ump->um_extattr.uepm_flags & ULFS_EXTATTR_UEPM_STARTED) {
		error = EBUSY;
		goto unlock;
	}

	ump->um_extattr.uepm_flags |= ULFS_EXTATTR_UEPM_STARTED;

	ump->um_extattr.uepm_ucred = l->l_cred;
	kauth_cred_hold(ump->um_extattr.uepm_ucred);

 unlock:
	ulfs_extattr_uepm_unlock(ump);

	return (error);
}

/*
 * Helper routine: given a locked parent directory and filename, return
 * the locked vnode of the inode associated with the name.  Will not
 * follow symlinks, may return any type of vnode.  Lock on parent will
 * be released even in the event of a failure.  In the event that the
 * target is the parent (i.e., "."), there will be two references and
 * one lock, requiring the caller to possibly special-case.
 */
static int
ulfs_extattr_lookup(struct vnode *start_dvp, int lockparent, const char *dirname,
    struct vnode **vp, struct lwp *l)
{
	struct vop_lookup_v2_args vargs;
	struct componentname cnp;
	struct vnode *target_vp;
	char *pnbuf;
	int error;

	KASSERT(VOP_ISLOCKED(start_dvp) == LK_EXCLUSIVE);

	pnbuf = PNBUF_GET();

	memset(&cnp, 0, sizeof(cnp));
	cnp.cn_nameiop = LOOKUP;
	cnp.cn_flags = ISLASTCN | lockparent;
	cnp.cn_cred = l->l_cred;
	cnp.cn_nameptr = pnbuf;
	error = copystr(dirname, pnbuf, MAXPATHLEN, &cnp.cn_namelen);
	if (error) {
		if (lockparent == 0) {
			VOP_UNLOCK(start_dvp);
		}
		PNBUF_PUT(pnbuf);
		printf("ulfs_extattr_lookup: copystr failed\n");
		return (error);
	}
	cnp.cn_namelen--;	/* trim nul termination */
	vargs.a_desc = NULL;
	vargs.a_dvp = start_dvp;
	vargs.a_vpp = &target_vp;
	vargs.a_cnp = &cnp;
	error = ulfs_lookup(&vargs);
	PNBUF_PUT(pnbuf);
	if (error) {
		if (lockparent == 0) {
			VOP_UNLOCK(start_dvp);
		}
		return (error);
	}
#if 0
	if (target_vp == start_dvp)
		panic("ulfs_extattr_lookup: target_vp == start_dvp");
#endif

	if (target_vp != start_dvp) {
		error = vn_lock(target_vp, LK_EXCLUSIVE);
		if (lockparent == 0)
			VOP_UNLOCK(start_dvp);
		if (error) {
			vrele(target_vp);
			return error;
		}
	}

	KASSERT(VOP_ISLOCKED(target_vp) == LK_EXCLUSIVE);
	*vp = target_vp;
	return (0);
}

/*
 * Enable an EA using the passed filesystem, backing vnode, attribute name,
 * namespace, and proc.  Will perform a VOP_OPEN() on the vp, so expects vp
 * to be locked when passed in.  The vnode will be returned unlocked,
 * regardless of success/failure of the function.  As a result, the caller
 * will always need to vrele(), but not vput().
 */
static int
ulfs_extattr_enable_with_open(struct ulfsmount *ump, struct vnode *vp,
    int attrnamespace, const char *attrname, struct lwp *l)
{
	int error;

	error = VOP_OPEN(vp, FREAD|FWRITE, l->l_cred);
	if (error) {
		printf("ulfs_extattr_enable_with_open.VOP_OPEN(): failed "
		    "with %d\n", error);
		VOP_UNLOCK(vp);
		return (error);
	}

	mutex_enter(vp->v_interlock);
	vp->v_writecount++;
	mutex_exit(vp->v_interlock);

	vref(vp);

	VOP_UNLOCK(vp);

	error = ulfs_extattr_enable(ump, attrnamespace, attrname, vp, l);
	if (error != 0)
		vn_close(vp, FREAD|FWRITE, l->l_cred);
	return (error);
}

/*
 * Given a locked directory vnode, iterate over the names in the directory
 * and use ulfs_extattr_lookup() to retrieve locked vnodes of potential
 * attribute files.  Then invoke ulfs_extattr_enable_with_open() on each
 * to attempt to start the attribute.  Leaves the directory locked on
 * exit.
 */
static int
ulfs_extattr_iterate_directory(struct ulfsmount *ump, struct vnode *dvp,
    int attrnamespace, struct lwp *l)
{
	struct vop_readdir_args vargs;
	struct statvfs *sbp = &ump->um_mountp->mnt_stat;
	struct dirent *dp, *edp;
	struct vnode *attr_vp;
	struct uio auio;
	struct iovec aiov;
	char *dirbuf;
	int error, eofflag = 0;

	if (dvp->v_type != VDIR)
		return (ENOTDIR);

	dirbuf = kmem_alloc(LFS_DIRBLKSIZ, KM_SLEEP);

	auio.uio_iov = &aiov;
	auio.uio_iovcnt = 1;
	auio.uio_rw = UIO_READ;
	auio.uio_offset = 0;
	UIO_SETUP_SYSSPACE(&auio);

	vargs.a_desc = NULL;
	vargs.a_vp = dvp;
	vargs.a_uio = &auio;
	vargs.a_cred = l->l_cred;
	vargs.a_eofflag = &eofflag;
	vargs.a_ncookies = NULL;
	vargs.a_cookies = NULL;

	while (!eofflag) {
		auio.uio_resid = LFS_DIRBLKSIZ;
		aiov.iov_base = dirbuf;
		aiov.iov_len = LFS_DIRBLKSIZ;
		error = ulfs_readdir(&vargs);
		if (error) {
			printf("ulfs_extattr_iterate_directory: ulfs_readdir "
			    "%d\n", error);
			return (error);
		}

		/*
		 * XXXRW: While in LFS, we always get LFS_DIRBLKSIZ returns from
		 * the directory code on success, on other file systems this
		 * may not be the case.  For portability, we should check the
		 * read length on return from ulfs_readdir().
		 */
		edp = (struct dirent *)&dirbuf[LFS_DIRBLKSIZ];
		for (dp = (struct dirent *)dirbuf; dp < edp; ) {
			if (dp->d_reclen == 0)
				break;
			/* Skip "." and ".." */
			if (dp->d_name[0] == '.' &&
			    (dp->d_name[1] == '\0' ||
			     (dp->d_name[1] == '.' && dp->d_name[2] == '\0')))
				goto next;
			error = ulfs_extattr_lookup(dvp, LOCKPARENT,
			    dp->d_name, &attr_vp, l);
			if (error == ENOENT) {
				goto next; /* keep silent */
			} else if (error) {
				printf("ulfs_extattr_iterate_directory: lookup "
				    "%s %d\n", dp->d_name, error);
			} else if (attr_vp == dvp) {
				vrele(attr_vp);
			} else if (attr_vp->v_type != VREG) {
				vput(attr_vp);
			} else {
				error = ulfs_extattr_enable_with_open(ump,
				    attr_vp, attrnamespace, dp->d_name, l);
				vrele(attr_vp);
				if (error) {
					printf("ulfs_extattr_iterate_directory: "
					    "enable %s %d\n", dp->d_name,
					    error);
				} else if (bootverbose) {
					printf("%s: EA %s loaded\n",
					       sbp->f_mntonname, dp->d_name);
				}
			}
 next:
			dp = (struct dirent *) ((char *)dp + dp->d_reclen);
			if (dp >= edp)
				break;
		}
	}
	kmem_free(dirbuf, LFS_DIRBLKSIZ);
	
	return (0);
}

/*
 * Auto-start of extended attributes, to be executed (optionally) at
 * mount-time.
 */
int
ulfs_extattr_autostart(struct mount *mp, struct lwp *l)
{
	struct vnode *rvp, *attr_dvp, *attr_system_dvp, *attr_user_dvp;
	int error;

	/*
	 * Does ULFS_EXTATTR_FSROOTSUBDIR exist off the filesystem root?
	 * If so, automatically start EA's.
	 */
	error = VFS_ROOT(mp, &rvp);
	if (error) {
		printf("ulfs_extattr_autostart.VFS_ROOT() returned %d\n",
		    error);
		return (error);
	}

	KASSERT(VOP_ISLOCKED(rvp) == LK_EXCLUSIVE);

	error = ulfs_extattr_lookup(rvp, 0,
	    ULFS_EXTATTR_FSROOTSUBDIR, &attr_dvp, l);
	if (error) {
		/* rvp ref'd but now unlocked */
		KASSERT(VOP_ISLOCKED(rvp) == 0);
		vrele(rvp);
		return (error);
	}
	if (rvp == attr_dvp) {
		/* Should never happen. */
		KASSERT(VOP_ISLOCKED(rvp) == LK_EXCLUSIVE);
		vrele(attr_dvp);
		vput(rvp);
		return (EINVAL);
	}
	KASSERT(VOP_ISLOCKED(rvp) == 0);
	vrele(rvp);

	KASSERT(VOP_ISLOCKED(attr_dvp) == LK_EXCLUSIVE);

	if (attr_dvp->v_type != VDIR) {
		printf("ulfs_extattr_autostart: %s != VDIR\n",
		    ULFS_EXTATTR_FSROOTSUBDIR);
		goto return_vput_attr_dvp;
	}

	error = ulfs_extattr_start(mp, l);
	if (error) {
		printf("ulfs_extattr_autostart: ulfs_extattr_start failed (%d)\n",
		    error);
		goto return_vput_attr_dvp;
	}

	/*
	 * Look for two subdirectories: ULFS_EXTATTR_SUBDIR_SYSTEM,
	 * ULFS_EXTATTR_SUBDIR_USER.  For each, iterate over the sub-directory,
	 * and start with appropriate type.  Failures in either don't
	 * result in an over-all failure.  attr_dvp is left locked to
	 * be cleaned up on exit.
	 */
	error = ulfs_extattr_lookup(attr_dvp, LOCKPARENT,
	    ULFS_EXTATTR_SUBDIR_SYSTEM, &attr_system_dvp, l);
	KASSERT(VOP_ISLOCKED(attr_dvp) == LK_EXCLUSIVE);
	if (error == 0) {
		KASSERT(VOP_ISLOCKED(attr_system_dvp) == LK_EXCLUSIVE);
		error = ulfs_extattr_iterate_directory(VFSTOULFS(mp),
		    attr_system_dvp, EXTATTR_NAMESPACE_SYSTEM, l);
		if (error)
			printf("ulfs_extattr_iterate_directory returned %d\n",
			    error);
		KASSERT(VOP_ISLOCKED(attr_system_dvp) == LK_EXCLUSIVE);
		vput(attr_system_dvp);
	}

	error = ulfs_extattr_lookup(attr_dvp, LOCKPARENT,
	    ULFS_EXTATTR_SUBDIR_USER, &attr_user_dvp, l);
	KASSERT(VOP_ISLOCKED(attr_dvp) == LK_EXCLUSIVE);
	if (error == 0) {
		KASSERT(VOP_ISLOCKED(attr_user_dvp) == LK_EXCLUSIVE);
		error = ulfs_extattr_iterate_directory(VFSTOULFS(mp),
		    attr_user_dvp, EXTATTR_NAMESPACE_USER, l);
		if (error)
			printf("ulfs_extattr_iterate_directory returned %d\n",
			    error);
		KASSERT(VOP_ISLOCKED(attr_user_dvp) == LK_EXCLUSIVE);
		vput(attr_user_dvp);
	}

	/* Mask startup failures in sub-directories. */
	error = 0;

 return_vput_attr_dvp:
	KASSERT(VOP_ISLOCKED(attr_dvp) == LK_EXCLUSIVE);
	vput(attr_dvp);

	return (error);
}

/*
 * Stop extended attribute support on an FS.
 */
void
ulfs_extattr_stop(struct mount *mp, struct lwp *l)
{
	struct ulfs_extattr_list_entry *uele;
	struct ulfsmount *ump = VFSTOULFS(mp);

	ulfs_extattr_uepm_lock(ump);

	/*
	 * If we haven't been started, no big deal.  Just short-circuit
	 * the processing work.
	 */
	if (!(ump->um_extattr.uepm_flags & ULFS_EXTATTR_UEPM_STARTED)) {
		goto unlock;
	}

	while (LIST_FIRST(&ump->um_extattr.uepm_list) != NULL) {
		uele = LIST_FIRST(&ump->um_extattr.uepm_list);
		ulfs_extattr_disable(ump, uele->uele_attrnamespace,
		    uele->uele_attrname, l);
	}

	ump->um_extattr.uepm_flags &= ~ULFS_EXTATTR_UEPM_STARTED;

	kauth_cred_free(ump->um_extattr.uepm_ucred);
	ump->um_extattr.uepm_ucred = NULL;

 unlock:
	ulfs_extattr_uepm_unlock(ump);
}

/*
 * Enable a named attribute on the specified filesystem; provide an
 * unlocked backing vnode to hold the attribute data.
 */
static int
ulfs_extattr_enable(struct ulfsmount *ump, int attrnamespace,
    const char *attrname, struct vnode *backing_vnode, struct lwp *l)
{
	struct ulfs_extattr_list_entry *attribute;
	struct iovec aiov;
	struct uio auio;
	int error = 0;

	if (!ulfs_extattr_valid_attrname(attrnamespace, attrname))
		return (EINVAL);
	if (backing_vnode->v_type != VREG)
		return (EINVAL);

	attribute = kmem_zalloc(sizeof(*attribute), KM_SLEEP);

	if (!(ump->um_extattr.uepm_flags & ULFS_EXTATTR_UEPM_STARTED)) {
		error = EOPNOTSUPP;
		goto free_exit;
	}

	if (ulfs_extattr_find_attr(ump, attrnamespace, attrname)) {
		error = EEXIST;
		goto free_exit;
	}

	strncpy(attribute->uele_attrname, attrname,
	    ULFS_EXTATTR_MAXEXTATTRNAME);
	attribute->uele_attrnamespace = attrnamespace;
	memset(&attribute->uele_fileheader, 0,
	    sizeof(struct ulfs_extattr_fileheader));
	
	attribute->uele_backing_vnode = backing_vnode;

	auio.uio_iov = &aiov;
	auio.uio_iovcnt = 1;
	aiov.iov_base = (void *) &attribute->uele_fileheader;
	aiov.iov_len = sizeof(struct ulfs_extattr_fileheader);
	auio.uio_resid = sizeof(struct ulfs_extattr_fileheader);
	auio.uio_offset = (off_t) 0;
	auio.uio_rw = UIO_READ;
	UIO_SETUP_SYSSPACE(&auio);

	vn_lock(backing_vnode, LK_SHARED | LK_RETRY);
	error = VOP_READ(backing_vnode, &auio, IO_NODELOCKED,
	    ump->um_extattr.uepm_ucred);

	if (error)
		goto unlock_free_exit;

	if (auio.uio_resid != 0) {
		printf("ulfs_extattr_enable: malformed attribute header\n");
		error = EINVAL;
		goto unlock_free_exit;
	}

	/*
	 * Try to determine the byte order of the attribute file.
	 */
	if (attribute->uele_fileheader.uef_magic != ULFS_EXTATTR_MAGIC) {
		attribute->uele_flags |= UELE_F_NEEDSWAP;
		attribute->uele_fileheader.uef_magic =
		    ulfs_rw32(attribute->uele_fileheader.uef_magic,
			     UELE_NEEDSWAP(attribute));
		if (attribute->uele_fileheader.uef_magic != ULFS_EXTATTR_MAGIC) {
			printf("ulfs_extattr_enable: invalid attribute header "
			       "magic\n");
			error = EINVAL;
			goto unlock_free_exit;
		}
	}
	attribute->uele_fileheader.uef_version =
	    ulfs_rw32(attribute->uele_fileheader.uef_version,
		     UELE_NEEDSWAP(attribute));
	attribute->uele_fileheader.uef_size =
	    ulfs_rw32(attribute->uele_fileheader.uef_size,
		     UELE_NEEDSWAP(attribute));

	if (attribute->uele_fileheader.uef_version != ULFS_EXTATTR_VERSION) {
		printf("ulfs_extattr_enable: incorrect attribute header "
		    "version\n");
		error = EINVAL;
		goto unlock_free_exit;
	}

	LIST_INSERT_HEAD(&ump->um_extattr.uepm_list, attribute,
	    uele_entries);

	VOP_UNLOCK(backing_vnode);
	return (0);

 unlock_free_exit:
	VOP_UNLOCK(backing_vnode);

 free_exit:
	kmem_free(attribute, sizeof(*attribute));
	return (error);
}

/*
 * Disable extended attribute support on an FS.
 */
static int
ulfs_extattr_disable(struct ulfsmount *ump, int attrnamespace,
    const char *attrname, struct lwp *l)
{
	struct ulfs_extattr_list_entry *uele;
	int error = 0;

	if (!ulfs_extattr_valid_attrname(attrnamespace, attrname))
		return (EINVAL);

	uele = ulfs_extattr_find_attr(ump, attrnamespace, attrname);
	if (!uele)
		return (ENODATA);

	LIST_REMOVE(uele, uele_entries);

	error = vn_close(uele->uele_backing_vnode, FREAD|FWRITE,
	    l->l_cred);

	kmem_free(uele, sizeof(*uele));

	return (error);
}

/*
 * VFS call to manage extended attributes in ULFS.  If filename_vp is
 * non-NULL, it must be passed in locked, and regardless of errors in
 * processing, will be unlocked.
 */
int
ulfs_extattrctl(struct mount *mp, int cmd, struct vnode *filename_vp,
    int attrnamespace, const char *attrname)
{
	struct lwp *l = curlwp;
	struct ulfsmount *ump = VFSTOULFS(mp);
	int error;

	/*
	 * Only privileged processes can configure extended attributes.
	 */
	error = kauth_authorize_system(l->l_cred, KAUTH_SYSTEM_FS_EXTATTR,
	    0, mp, NULL, NULL);
	if (error) {
		if (filename_vp != NULL)
			VOP_UNLOCK(filename_vp);
		return (error);
	}

	switch(cmd) {
	case ULFS_EXTATTR_CMD_START:
		if (filename_vp != NULL) {
			VOP_UNLOCK(filename_vp);
			return (EINVAL);
		}
		if (attrname != NULL)
			return (EINVAL);

		error = ulfs_extattr_autostart(mp, l);
		return (error);
		
	case ULFS_EXTATTR_CMD_STOP:
		if (filename_vp != NULL) {
			VOP_UNLOCK(filename_vp);
			return (EINVAL);
		}
		if (attrname != NULL)
			return (EINVAL);

		ulfs_extattr_stop(mp, l);
		return (0);

	case ULFS_EXTATTR_CMD_ENABLE:
		if (filename_vp == NULL)
			return (EINVAL);
		if (attrname == NULL) {
			VOP_UNLOCK(filename_vp);
			return (EINVAL);
		}

		/*
		 * ulfs_extattr_enable_with_open() will always unlock the
		 * vnode, regardless of failure.
		 */
		ulfs_extattr_uepm_lock(ump);
		error = ulfs_extattr_enable_with_open(ump, filename_vp,
		    attrnamespace, attrname, l);
		ulfs_extattr_uepm_unlock(ump);
		return (error);

	case ULFS_EXTATTR_CMD_DISABLE:
		if (filename_vp != NULL) {
			VOP_UNLOCK(filename_vp);
			return (EINVAL);
		}
		if (attrname == NULL)
			return (EINVAL);

		ulfs_extattr_uepm_lock(ump);
		error = ulfs_extattr_disable(ump, attrnamespace, attrname, l);
		ulfs_extattr_uepm_unlock(ump);
		return (error);

	default:
		return (EINVAL);
	}
}

/*
 * Read extended attribute header for a given vnode and attribute.
 * Backing vnode should be locked and unlocked by caller.
 */
static int
ulfs_extattr_get_header(struct vnode *vp, struct ulfs_extattr_list_entry *uele,
    struct ulfs_extattr_header *ueh, off_t *bap)
{
	struct mount *mp = vp->v_mount;
	struct ulfsmount *ump = VFSTOULFS(mp);
	struct inode *ip = VTOI(vp);
	off_t base_offset;
	struct iovec aiov;
	struct uio aio;
	int error;

	/*
	 * Find base offset of header in file based on file header size, and
	 * data header size + maximum data size, indexed by inode number.
	 */
	base_offset = sizeof(struct ulfs_extattr_fileheader) +
	    ip->i_number * (sizeof(struct ulfs_extattr_header) +
	    uele->uele_fileheader.uef_size);

	/*
	 * Read in the data header to see if the data is defined, and if so
	 * how much.
	 */
	memset(ueh, 0, sizeof(struct ulfs_extattr_header));
	aiov.iov_base = ueh;
	aiov.iov_len = sizeof(struct ulfs_extattr_header);
	aio.uio_iov = &aiov;
	aio.uio_iovcnt = 1;
	aio.uio_rw = UIO_READ;
	aio.uio_offset = base_offset;
	aio.uio_resid = sizeof(struct ulfs_extattr_header);
	UIO_SETUP_SYSSPACE(&aio);

	error = VOP_READ(uele->uele_backing_vnode, &aio,
	    IO_NODELOCKED, ump->um_extattr.uepm_ucred);
	if (error)
		return error;

	/*
	 * Attribute headers are kept in file system byte order.
	 * XXX What about the blob of data?
	 */
	ueh->ueh_flags = ulfs_rw32(ueh->ueh_flags, UELE_NEEDSWAP(uele));
	ueh->ueh_len   = ulfs_rw32(ueh->ueh_len, UELE_NEEDSWAP(uele));
	ueh->ueh_i_gen = ulfs_rw32(ueh->ueh_i_gen, UELE_NEEDSWAP(uele));

	/* Defined? */
	if ((ueh->ueh_flags & ULFS_EXTATTR_ATTR_FLAG_INUSE) == 0)
		return ENODATA;

	/* Valid for the current inode generation? */
	if (ueh->ueh_i_gen != ip->i_gen) {
		/*
		 * The inode itself has a different generation number
		 * than the uele data.  For now, the best solution
		 * is to coerce this to undefined, and let it get cleaned
		 * up by the next write or extattrctl clean.
		 */
		printf("%s (%s): inode gen inconsistency (%u, %jd)\n",
		       __func__,  mp->mnt_stat.f_mntonname, ueh->ueh_i_gen,
		       (intmax_t)ip->i_gen);
		return ENODATA;
	}

	/* Local size consistency check. */
	if (ueh->ueh_len > uele->uele_fileheader.uef_size)
		return ENXIO;

	/* Return base offset */
	if (bap != NULL)
		*bap = base_offset;

	return 0;
}

/*
 * Vnode operation to retrieve a named extended attribute.
 */
int
ulfs_getextattr(struct vop_getextattr_args *ap)
/*
vop_getextattr {
	IN struct vnode *a_vp;
	IN int a_attrnamespace;
	IN const char *a_name;
	INOUT struct uio *a_uio;
	OUT size_t *a_size;
	IN kauth_cred_t a_cred;
};
*/
{
	struct mount *mp = ap->a_vp->v_mount;
	struct ulfsmount *ump = VFSTOULFS(mp);
	int error;

	ulfs_extattr_uepm_lock(ump);

	error = ulfs_extattr_get(ap->a_vp, ap->a_attrnamespace, ap->a_name,
	    ap->a_uio, ap->a_size, ap->a_cred, curlwp);

	ulfs_extattr_uepm_unlock(ump);

	return (error);
}

/*
 * Real work associated with retrieving a named attribute--assumes that
 * the attribute lock has already been grabbed.
 */
static int
ulfs_extattr_get(struct vnode *vp, int attrnamespace, const char *name,
    struct uio *uio, size_t *size, kauth_cred_t cred, struct lwp *l)
{
	struct ulfs_extattr_list_entry *attribute;
	struct ulfs_extattr_header ueh;
	struct mount *mp = vp->v_mount;
	struct ulfsmount *ump = VFSTOULFS(mp);
	off_t base_offset;
	size_t len, old_len;
	int error = 0;

	if (!(ump->um_extattr.uepm_flags & ULFS_EXTATTR_UEPM_STARTED))
		return (EOPNOTSUPP);

	if (strlen(name) == 0)
		return (EINVAL);

	error = internal_extattr_check_cred(vp, attrnamespace, name, cred,
	    VREAD);
	if (error)
		return (error);

	attribute = ulfs_extattr_find_attr(ump, attrnamespace, name);
	if (!attribute)
		return (ENODATA);

	/*
	 * Allow only offsets of zero to encourage the read/replace
	 * extended attribute semantic.  Otherwise we can't guarantee
	 * atomicity, as we don't provide locks for extended attributes.
	 */
	if (uio != NULL && uio->uio_offset != 0)
		return (ENXIO);

	/*
	 * Don't need to get a lock on the backing file if the getattr is
	 * being applied to the backing file, as the lock is already held.
	 */
	if (attribute->uele_backing_vnode != vp)
		vn_lock(attribute->uele_backing_vnode, LK_SHARED | LK_RETRY);

	error = ulfs_extattr_get_header(vp, attribute, &ueh, &base_offset);
	if (error)
		goto vopunlock_exit;

	/* Return full data size if caller requested it. */
	if (size != NULL)
		*size = ueh.ueh_len;

	/* Return data if the caller requested it. */
	if (uio != NULL) {
		/* Allow for offset into the attribute data. */
		uio->uio_offset = base_offset + sizeof(struct
		    ulfs_extattr_header);

		/*
		 * Figure out maximum to transfer -- use buffer size and
		 * local data limit.
		 */
		len = MIN(uio->uio_resid, ueh.ueh_len);
		old_len = uio->uio_resid;
		uio->uio_resid = len;

		error = VOP_READ(attribute->uele_backing_vnode, uio,
		    IO_NODELOCKED, ump->um_extattr.uepm_ucred);
		if (error)
			goto vopunlock_exit;

		uio->uio_resid = old_len - (len - uio->uio_resid);
	}

 vopunlock_exit:

	if (uio != NULL)
		uio->uio_offset = 0;

	if (attribute->uele_backing_vnode != vp)
		VOP_UNLOCK(attribute->uele_backing_vnode);

	return (error);
}

/*
 * Vnode operation to list extended attribute for a vnode
 */
int
ulfs_listextattr(struct vop_listextattr_args *ap)
/*
vop_listextattr {
	IN struct vnode *a_vp;
	IN int a_attrnamespace;
	INOUT struct uio *a_uio;
	OUT size_t *a_size;
	IN int flag;
	IN kauth_cred_t a_cred;
	struct proc *a_p;
};
*/
{
	struct mount *mp = ap->a_vp->v_mount;
	struct ulfsmount *ump = VFSTOULFS(mp);
	int error;

	ulfs_extattr_uepm_lock(ump);

	error = ulfs_extattr_list(ap->a_vp, ap->a_attrnamespace,
	    ap->a_uio, ap->a_size, ap->a_flag, ap->a_cred, curlwp);

	ulfs_extattr_uepm_unlock(ump);

	return (error);
}

/*
 * Real work associated with retrieving list of attributes--assumes that
 * the attribute lock has already been grabbed.
 */
static int
ulfs_extattr_list(struct vnode *vp, int attrnamespace,
    struct uio *uio, size_t *size, int flag, 
    kauth_cred_t cred, struct lwp *l)
{
	struct ulfs_extattr_list_entry *uele;
	struct ulfs_extattr_header ueh;
	struct mount *mp = vp->v_mount;
	struct ulfsmount *ump = VFSTOULFS(mp);
	size_t listsize = 0;
	int error = 0;

	if (!(ump->um_extattr.uepm_flags & ULFS_EXTATTR_UEPM_STARTED))
		return (EOPNOTSUPP);

	/*
	 * XXX: We can move this inside the loop and iterate on individual
	 *	attributes.
	 */
	error = internal_extattr_check_cred(vp, attrnamespace, "", cred,
	    VREAD);
	if (error)
		return (error);

	LIST_FOREACH(uele, &ump->um_extattr.uepm_list, uele_entries) {
		unsigned char attrnamelen;

		if (uele->uele_attrnamespace != attrnamespace)
			continue;

		error = ulfs_extattr_get_header(vp, uele, &ueh, NULL);
		if (error == ENODATA)
			continue;	
		if (error != 0)
			return error;

		/*
		 * Don't need to get a lock on the backing file if 
		 * the listattr is being applied to the backing file, 
		 * as the lock is already held.
		 */
		if (uele->uele_backing_vnode != vp)
			vn_lock(uele->uele_backing_vnode, LK_SHARED | LK_RETRY);

		/*
		 * +1 for trailing NUL (listxattr flavor)
		 *  or leading name length (extattr_list_file flavor)
	 	 */
		attrnamelen = strlen(uele->uele_attrname);
		listsize += attrnamelen + 1;

		/* Return data if the caller requested it. */
		if (uio != NULL) {
			/*
			 * We support two flavors. Either NUL-terminated
			 * strings (a la listxattr), or non NUL-terminated,
			 * one byte length prefixed strings (for
			 * extattr_list_file). EXTATTR_LIST_LENPREFIX switches
		 	 * that second behavior.
			 */
			if (flag & EXTATTR_LIST_LENPREFIX) {
				uint8_t len = (uint8_t)attrnamelen;

				/* Copy leading name length */
				error = uiomove(&len, sizeof(len), uio);
				if (error != 0)
					break;	
			} else {
				/* Include trailing NULL */
				attrnamelen++; 
			}

			error = uiomove(uele->uele_attrname, 
					(size_t)attrnamelen, uio);
			if (error != 0)
				break;	
		}

		if (uele->uele_backing_vnode != vp)
			VOP_UNLOCK(uele->uele_backing_vnode);

		if (error != 0)
			return error;
	}

	if (uio != NULL)
		uio->uio_offset = 0;

	/* Return full data size if caller requested it. */
	if (size != NULL)
		*size = listsize;

	return 0;
}

/*
 * Vnode operation to remove a named attribute.
 */
int
ulfs_deleteextattr(struct vop_deleteextattr_args *ap)
/*
vop_deleteextattr {
	IN struct vnode *a_vp;
	IN int a_attrnamespace;
	IN const char *a_name;
	IN kauth_cred_t a_cred;
};
*/
{
	struct mount *mp = ap->a_vp->v_mount;
	struct ulfsmount *ump = VFSTOULFS(mp); 
	int error;

	ulfs_extattr_uepm_lock(ump);

	error = ulfs_extattr_rm(ap->a_vp, ap->a_attrnamespace, ap->a_name,
	    ap->a_cred, curlwp);

	ulfs_extattr_uepm_unlock(ump);

	return (error);
}

/*
 * Vnode operation to set a named attribute.
 */
int
ulfs_setextattr(struct vop_setextattr_args *ap)
/*
vop_setextattr {
	IN struct vnode *a_vp;
	IN int a_attrnamespace;
	IN const char *a_name;
	INOUT struct uio *a_uio;
	IN kauth_cred_t a_cred;
};
*/
{
	struct mount *mp = ap->a_vp->v_mount;
	struct ulfsmount *ump = VFSTOULFS(mp); 
	int error;

	ulfs_extattr_uepm_lock(ump);

	/*
	 * XXX: No longer a supported way to delete extended attributes.
	 */
	if (ap->a_uio == NULL) {
		ulfs_extattr_uepm_unlock(ump);
		return (EINVAL);
	}

	error = ulfs_extattr_set(ap->a_vp, ap->a_attrnamespace, ap->a_name,
	    ap->a_uio, ap->a_cred, curlwp);

	ulfs_extattr_uepm_unlock(ump);

	return (error);
}

/*
 * Real work associated with setting a vnode's extended attributes;
 * assumes that the attribute lock has already been grabbed.
 */
static int
ulfs_extattr_set(struct vnode *vp, int attrnamespace, const char *name,
    struct uio *uio, kauth_cred_t cred, struct lwp *l)
{
	struct ulfs_extattr_list_entry *attribute;
	struct ulfs_extattr_header ueh;
	struct iovec local_aiov;
	struct uio local_aio;
	struct mount *mp = vp->v_mount;
	struct ulfsmount *ump = VFSTOULFS(mp);
	struct inode *ip = VTOI(vp);
	off_t base_offset;
	int error = 0, ioflag;

	if (vp->v_mount->mnt_flag & MNT_RDONLY)
		return (EROFS);
	if (!(ump->um_extattr.uepm_flags & ULFS_EXTATTR_UEPM_STARTED))
		return (EOPNOTSUPP);
	if (!ulfs_extattr_valid_attrname(attrnamespace, name))
		return (EINVAL);

	error = internal_extattr_check_cred(vp, attrnamespace, name, cred,
	    VWRITE);
	if (error)
		return (error);

	attribute = ulfs_extattr_find_attr(ump, attrnamespace, name);
	if (!attribute) {
		attribute =  ulfs_extattr_autocreate_attr(vp, attrnamespace, 
							 name, l);
		if  (!attribute)
			return (ENODATA);
	}

	/*
	 * Early rejection of invalid offsets/length.
	 * Reject: any offset but 0 (replace)
	 *	 Any size greater than attribute size limit
 	 */
	if (uio->uio_offset != 0 ||
	    uio->uio_resid > attribute->uele_fileheader.uef_size)
		return (ENXIO);

	/*
	 * Find base offset of header in file based on file header size, and
	 * data header size + maximum data size, indexed by inode number.
	 */
	base_offset = sizeof(struct ulfs_extattr_fileheader) +
	    ip->i_number * (sizeof(struct ulfs_extattr_header) +
	    attribute->uele_fileheader.uef_size);

	/*
	 * Write out a data header for the data.
	 */
	ueh.ueh_len = ulfs_rw32((uint32_t) uio->uio_resid,
	    UELE_NEEDSWAP(attribute));
	ueh.ueh_flags = ulfs_rw32(ULFS_EXTATTR_ATTR_FLAG_INUSE,
				 UELE_NEEDSWAP(attribute));
	ueh.ueh_i_gen = ulfs_rw32(ip->i_gen, UELE_NEEDSWAP(attribute));
	local_aiov.iov_base = &ueh;
	local_aiov.iov_len = sizeof(struct ulfs_extattr_header);
	local_aio.uio_iov = &local_aiov;
	local_aio.uio_iovcnt = 1;
	local_aio.uio_rw = UIO_WRITE;
	local_aio.uio_offset = base_offset;
	local_aio.uio_resid = sizeof(struct ulfs_extattr_header);
	UIO_SETUP_SYSSPACE(&local_aio);

	/*
	 * Don't need to get a lock on the backing file if the setattr is
	 * being applied to the backing file, as the lock is already held.
	 */
	if (attribute->uele_backing_vnode != vp)
		vn_lock(attribute->uele_backing_vnode, 
		    LK_EXCLUSIVE | LK_RETRY);

	ioflag = IO_NODELOCKED;
	if (ulfs_extattr_sync)
		ioflag |= IO_SYNC;
	error = VOP_WRITE(attribute->uele_backing_vnode, &local_aio, ioflag,
	    ump->um_extattr.uepm_ucred);
	if (error)
		goto vopunlock_exit;

	if (local_aio.uio_resid != 0) {
		error = ENXIO;
		goto vopunlock_exit;
	}

	/*
	 * Write out user data.
	 * XXX NOT ATOMIC WITH RESPECT TO THE HEADER.
	 */
	uio->uio_offset = base_offset + sizeof(struct ulfs_extattr_header);

	ioflag = IO_NODELOCKED;
	if (ulfs_extattr_sync)
		ioflag |= IO_SYNC;
	error = VOP_WRITE(attribute->uele_backing_vnode, uio, ioflag,
	    ump->um_extattr.uepm_ucred);

 vopunlock_exit:
	uio->uio_offset = 0;

	if (attribute->uele_backing_vnode != vp)
		VOP_UNLOCK(attribute->uele_backing_vnode);

	return (error);
}

/*
 * Real work associated with removing an extended attribute from a vnode.
 * Assumes the attribute lock has already been grabbed.
 */
static int
ulfs_extattr_rm(struct vnode *vp, int attrnamespace, const char *name,
    kauth_cred_t cred, struct lwp *l)
{
	struct ulfs_extattr_list_entry *attribute;
	struct ulfs_extattr_header ueh;
	struct mount *mp = vp->v_mount;
	struct ulfsmount *ump = VFSTOULFS(mp);
	struct iovec local_aiov;
	struct uio local_aio;
	off_t base_offset;
	int error = 0, ioflag;

	if (vp->v_mount->mnt_flag & MNT_RDONLY)  
		return (EROFS);
	if (!(ump->um_extattr.uepm_flags & ULFS_EXTATTR_UEPM_STARTED))
		return (EOPNOTSUPP);
	if (!ulfs_extattr_valid_attrname(attrnamespace, name))
		return (EINVAL);

	error = internal_extattr_check_cred(vp, attrnamespace, name, cred,
	    VWRITE);
	if (error)
		return (error);

	attribute = ulfs_extattr_find_attr(ump, attrnamespace, name);
	if (!attribute)
		return (ENODATA);

	/*
	 * Don't need to get a lock on the backing file if the getattr is
	 * being applied to the backing file, as the lock is already held.
	 */
	if (attribute->uele_backing_vnode != vp)
		vn_lock(attribute->uele_backing_vnode, LK_EXCLUSIVE | LK_RETRY);

	error = ulfs_extattr_get_header(vp, attribute, &ueh, &base_offset);
	if (error)
		goto vopunlock_exit;

	/* Flag it as not in use. */
	ueh.ueh_flags = 0;		/* No need to byte swap 0 */
	ueh.ueh_len = 0;		/* ...ditto... */

	local_aiov.iov_base = &ueh;
	local_aiov.iov_len = sizeof(struct ulfs_extattr_header);
	local_aio.uio_iov = &local_aiov;
	local_aio.uio_iovcnt = 1;
	local_aio.uio_rw = UIO_WRITE;
	local_aio.uio_offset = base_offset;
	local_aio.uio_resid = sizeof(struct ulfs_extattr_header);
	UIO_SETUP_SYSSPACE(&local_aio);

	ioflag = IO_NODELOCKED;
	if (ulfs_extattr_sync)
		ioflag |= IO_SYNC;
	error = VOP_WRITE(attribute->uele_backing_vnode, &local_aio, ioflag,
	    ump->um_extattr.uepm_ucred);
	if (error)
		goto vopunlock_exit;

	if (local_aio.uio_resid != 0)
		error = ENXIO;

 vopunlock_exit:
	VOP_UNLOCK(attribute->uele_backing_vnode);

	return (error);
}

/*
 * Called by ULFS when an inode is no longer active and should have its
 * attributes stripped.
 */
void
ulfs_extattr_vnode_inactive(struct vnode *vp, struct lwp *l)
{
	struct ulfs_extattr_list_entry *uele;
	struct mount *mp = vp->v_mount;
	struct ulfsmount *ump = VFSTOULFS(mp);

	/*
	 * In that case, we cannot lock. We should not have any active vnodes
	 * on the fs if this is not yet initialized but is going to be, so
	 * this can go unlocked.
	 */
	if (!(ump->um_extattr.uepm_flags & ULFS_EXTATTR_UEPM_INITIALIZED))
		return;

	ulfs_extattr_uepm_lock(ump);

	if (!(ump->um_extattr.uepm_flags & ULFS_EXTATTR_UEPM_STARTED)) {
		ulfs_extattr_uepm_unlock(ump);
		return;
	}

	LIST_FOREACH(uele, &ump->um_extattr.uepm_list, uele_entries)
		ulfs_extattr_rm(vp, uele->uele_attrnamespace,
		    uele->uele_attrname, lwp0.l_cred, l);

	ulfs_extattr_uepm_unlock(ump);
}

void
ulfs_extattr_init(void)
{

}

void
ulfs_extattr_done(void)
{

}