aboutsummaryrefslogtreecommitdiff
path: root/sys/fs/nfsclient/nfs_clport.c
blob: 77ecdd255ea2dd6a97173af4c3adc3a892d1a73a (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
/*-
 * SPDX-License-Identifier: BSD-3-Clause
 *
 * Copyright (c) 1989, 1993
 *	The Regents of the University of California.  All rights reserved.
 *
 * This code is derived from software contributed to Berkeley by
 * Rick Macklem at The University of Guelph.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. Neither the name of the University nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 *
 */

#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");

#include "opt_inet.h"
#include "opt_inet6.h"

#include <sys/capsicum.h>

/*
 * generally, I don't like #includes inside .h files, but it seems to
 * be the easiest way to handle the port.
 */
#include <sys/fail.h>
#include <sys/hash.h>
#include <sys/sysctl.h>
#include <fs/nfs/nfsport.h>
#include <netinet/in_fib.h>
#include <netinet/if_ether.h>
#include <netinet6/ip6_var.h>
#include <net/if_types.h>

#include <fs/nfsclient/nfs_kdtrace.h>

#ifdef KDTRACE_HOOKS
dtrace_nfsclient_attrcache_flush_probe_func_t
		dtrace_nfscl_attrcache_flush_done_probe;
uint32_t	nfscl_attrcache_flush_done_id;

dtrace_nfsclient_attrcache_get_hit_probe_func_t
		dtrace_nfscl_attrcache_get_hit_probe;
uint32_t	nfscl_attrcache_get_hit_id;

dtrace_nfsclient_attrcache_get_miss_probe_func_t
		dtrace_nfscl_attrcache_get_miss_probe;
uint32_t	nfscl_attrcache_get_miss_id;

dtrace_nfsclient_attrcache_load_probe_func_t
		dtrace_nfscl_attrcache_load_done_probe;
uint32_t	nfscl_attrcache_load_done_id;
#endif /* !KDTRACE_HOOKS */

extern u_int32_t newnfs_true, newnfs_false, newnfs_xdrneg1;
extern struct vop_vector newnfs_vnodeops;
extern struct vop_vector newnfs_fifoops;
extern uma_zone_t newnfsnode_zone;
extern struct buf_ops buf_ops_newnfs;
extern uma_zone_t ncl_pbuf_zone;
extern short nfsv4_cbport;
extern int nfscl_enablecallb;
extern int nfs_numnfscbd;
extern int nfscl_inited;
struct mtx ncl_iod_mutex;
NFSDLOCKMUTEX;
extern struct mtx nfsrv_dslock_mtx;

extern void (*ncl_call_invalcaches)(struct vnode *);

SYSCTL_DECL(_vfs_nfs);
static int ncl_fileid_maxwarnings = 10;
SYSCTL_INT(_vfs_nfs, OID_AUTO, fileid_maxwarnings, CTLFLAG_RWTUN,
    &ncl_fileid_maxwarnings, 0,
    "Limit fileid corruption warnings; 0 is off; -1 is unlimited");
static volatile int ncl_fileid_nwarnings;

static void nfscl_warn_fileid(struct nfsmount *, struct nfsvattr *,
    struct nfsvattr *);

/*
 * Comparison function for vfs_hash functions.
 */
int
newnfs_vncmpf(struct vnode *vp, void *arg)
{
	struct nfsfh *nfhp = (struct nfsfh *)arg;
	struct nfsnode *np = VTONFS(vp);

	if (np->n_fhp->nfh_len != nfhp->nfh_len ||
	    NFSBCMP(np->n_fhp->nfh_fh, nfhp->nfh_fh, nfhp->nfh_len))
		return (1);
	return (0);
}

/*
 * Look up a vnode/nfsnode by file handle.
 * Callers must check for mount points!!
 * In all cases, a pointer to a
 * nfsnode structure is returned.
 * This variant takes a "struct nfsfh *" as second argument and uses
 * that structure up, either by hanging off the nfsnode or FREEing it.
 */
int
nfscl_nget(struct mount *mntp, struct vnode *dvp, struct nfsfh *nfhp,
    struct componentname *cnp, struct thread *td, struct nfsnode **npp,
    void *stuff, int lkflags)
{
	struct nfsnode *np, *dnp;
	struct vnode *vp, *nvp;
	struct nfsv4node *newd, *oldd;
	int error;
	u_int hash;
	struct nfsmount *nmp;

	nmp = VFSTONFS(mntp);
	dnp = VTONFS(dvp);
	*npp = NULL;

	hash = fnv_32_buf(nfhp->nfh_fh, nfhp->nfh_len, FNV1_32_INIT);

	error = vfs_hash_get(mntp, hash, lkflags,
	    td, &nvp, newnfs_vncmpf, nfhp);
	if (error == 0 && nvp != NULL) {
		/*
		 * I believe there is a slight chance that vgonel() could
		 * get called on this vnode between when NFSVOPLOCK() drops
		 * the VI_LOCK() and vget() acquires it again, so that it
		 * hasn't yet had v_usecount incremented. If this were to
		 * happen, the VIRF_DOOMED flag would be set, so check for
		 * that here. Since we now have the v_usecount incremented,
		 * we should be ok until we vrele() it, if the VIRF_DOOMED
		 * flag isn't set now.
		 */
		VI_LOCK(nvp);
		if (VN_IS_DOOMED(nvp)) {
			VI_UNLOCK(nvp);
			vrele(nvp);
			error = ENOENT;
		} else {
			VI_UNLOCK(nvp);
		}
	}
	if (error) {
		free(nfhp, M_NFSFH);
		return (error);
	}
	if (nvp != NULL) {
		np = VTONFS(nvp);
		/*
		 * For NFSv4, check to see if it is the same name and
		 * replace the name, if it is different.
		 */
		oldd = newd = NULL;
		if ((nmp->nm_flag & NFSMNT_NFSV4) && np->n_v4 != NULL &&
		    nvp->v_type == VREG &&
		    (np->n_v4->n4_namelen != cnp->cn_namelen ||
		     NFSBCMP(cnp->cn_nameptr, NFS4NODENAME(np->n_v4),
		     cnp->cn_namelen) ||
		     dnp->n_fhp->nfh_len != np->n_v4->n4_fhlen ||
		     NFSBCMP(dnp->n_fhp->nfh_fh, np->n_v4->n4_data,
		     dnp->n_fhp->nfh_len))) {
		    newd = malloc(
			sizeof (struct nfsv4node) + dnp->n_fhp->nfh_len +
			+ cnp->cn_namelen - 1, M_NFSV4NODE, M_WAITOK);
		    NFSLOCKNODE(np);
		    if (newd != NULL && np->n_v4 != NULL && nvp->v_type == VREG
			&& (np->n_v4->n4_namelen != cnp->cn_namelen ||
			 NFSBCMP(cnp->cn_nameptr, NFS4NODENAME(np->n_v4),
			 cnp->cn_namelen) ||
			 dnp->n_fhp->nfh_len != np->n_v4->n4_fhlen ||
			 NFSBCMP(dnp->n_fhp->nfh_fh, np->n_v4->n4_data,
			 dnp->n_fhp->nfh_len))) {
			oldd = np->n_v4;
			np->n_v4 = newd;
			newd = NULL;
			np->n_v4->n4_fhlen = dnp->n_fhp->nfh_len;
			np->n_v4->n4_namelen = cnp->cn_namelen;
			NFSBCOPY(dnp->n_fhp->nfh_fh, np->n_v4->n4_data,
			    dnp->n_fhp->nfh_len);
			NFSBCOPY(cnp->cn_nameptr, NFS4NODENAME(np->n_v4),
			    cnp->cn_namelen);
		    }
		    NFSUNLOCKNODE(np);
		}
		if (newd != NULL)
			free(newd, M_NFSV4NODE);
		if (oldd != NULL)
			free(oldd, M_NFSV4NODE);
		*npp = np;
		free(nfhp, M_NFSFH);
		return (0);
	}
	np = uma_zalloc(newnfsnode_zone, M_WAITOK | M_ZERO);

	error = getnewvnode(nfs_vnode_tag, mntp, &newnfs_vnodeops, &nvp);
	if (error) {
		uma_zfree(newnfsnode_zone, np);
		free(nfhp, M_NFSFH);
		return (error);
	}
	vp = nvp;
	KASSERT(vp->v_bufobj.bo_bsize != 0, ("nfscl_nget: bo_bsize == 0"));
	vp->v_bufobj.bo_ops = &buf_ops_newnfs;
	vp->v_data = np;
	np->n_vnode = vp;
	/* 
	 * Initialize the mutex even if the vnode is going to be a loser.
	 * This simplifies the logic in reclaim, which can then unconditionally
	 * destroy the mutex (in the case of the loser, or if hash_insert
	 * happened to return an error no special casing is needed).
	 */
	mtx_init(&np->n_mtx, "NEWNFSnode lock", NULL, MTX_DEF | MTX_DUPOK);
	lockinit(&np->n_excl, PVFS, "nfsupg", VLKTIMEOUT, LK_NOSHARE |
	    LK_CANRECURSE);

	/* 
	 * Are we getting the root? If so, make sure the vnode flags
	 * are correct 
	 */
	if ((nfhp->nfh_len == nmp->nm_fhsize) &&
	    !bcmp(nfhp->nfh_fh, nmp->nm_fh, nfhp->nfh_len)) {
		if (vp->v_type == VNON)
			vp->v_type = VDIR;
		vp->v_vflag |= VV_ROOT;
	}

	vp->v_vflag |= VV_VMSIZEVNLOCK;
	
	np->n_fhp = nfhp;
	/*
	 * For NFSv4, we have to attach the directory file handle and
	 * file name, so that Open Ops can be done later.
	 */
	if (nmp->nm_flag & NFSMNT_NFSV4) {
		np->n_v4 = malloc(sizeof (struct nfsv4node)
		    + dnp->n_fhp->nfh_len + cnp->cn_namelen - 1, M_NFSV4NODE,
		    M_WAITOK);
		np->n_v4->n4_fhlen = dnp->n_fhp->nfh_len;
		np->n_v4->n4_namelen = cnp->cn_namelen;
		NFSBCOPY(dnp->n_fhp->nfh_fh, np->n_v4->n4_data,
		    dnp->n_fhp->nfh_len);
		NFSBCOPY(cnp->cn_nameptr, NFS4NODENAME(np->n_v4),
		    cnp->cn_namelen);
	} else {
		np->n_v4 = NULL;
	}

	/*
	 * NFS supports recursive and shared locking.
	 */
	lockmgr(vp->v_vnlock, LK_EXCLUSIVE | LK_NOWITNESS, NULL);
	VN_LOCK_AREC(vp);
	VN_LOCK_ASHARE(vp);
	error = insmntque(vp, mntp);
	if (error != 0) {
		*npp = NULL;
		mtx_destroy(&np->n_mtx);
		lockdestroy(&np->n_excl);
		free(nfhp, M_NFSFH);
		if (np->n_v4 != NULL)
			free(np->n_v4, M_NFSV4NODE);
		uma_zfree(newnfsnode_zone, np);
		return (error);
	}
	error = vfs_hash_insert(vp, hash, lkflags, 
	    td, &nvp, newnfs_vncmpf, nfhp);
	if (error)
		return (error);
	if (nvp != NULL) {
		*npp = VTONFS(nvp);
		/* vfs_hash_insert() vput()'s the losing vnode */
		return (0);
	}
	*npp = np;

	return (0);
}

/*
 * Another variant of nfs_nget(). This one is only used by reopen. It
 * takes almost the same args as nfs_nget(), but only succeeds if an entry
 * exists in the cache. (Since files should already be "open" with a
 * vnode ref cnt on the node when reopen calls this, it should always
 * succeed.)
 * Also, don't get a vnode lock, since it may already be locked by some
 * other process that is handling it. This is ok, since all other threads
 * on the client are blocked by the nfsc_lock being exclusively held by the
 * caller of this function.
 */
int
nfscl_ngetreopen(struct mount *mntp, u_int8_t *fhp, int fhsize,
    struct thread *td, struct nfsnode **npp)
{
	struct vnode *nvp;
	u_int hash;
	struct nfsfh *nfhp;
	int error;

	*npp = NULL;
	/* For forced dismounts, just return error. */
	if (NFSCL_FORCEDISM(mntp))
		return (EINTR);
	nfhp = malloc(sizeof (struct nfsfh) + fhsize,
	    M_NFSFH, M_WAITOK);
	bcopy(fhp, &nfhp->nfh_fh[0], fhsize);
	nfhp->nfh_len = fhsize;

	hash = fnv_32_buf(fhp, fhsize, FNV1_32_INIT);

	/*
	 * First, try to get the vnode locked, but don't block for the lock.
	 */
	error = vfs_hash_get(mntp, hash, (LK_EXCLUSIVE | LK_NOWAIT), td, &nvp,
	    newnfs_vncmpf, nfhp);
	if (error == 0 && nvp != NULL) {
		NFSVOPUNLOCK(nvp);
	} else if (error == EBUSY) {
		/*
		 * It is safe so long as a vflush() with
		 * FORCECLOSE has not been done. Since the Renew thread is
		 * stopped and the MNTK_UNMOUNTF flag is set before doing
		 * a vflush() with FORCECLOSE, we should be ok here.
		 */
		if (NFSCL_FORCEDISM(mntp))
			error = EINTR;
		else {
			vfs_hash_ref(mntp, hash, td, &nvp, newnfs_vncmpf, nfhp);
			if (nvp == NULL) {
				error = ENOENT;
			} else if (VN_IS_DOOMED(nvp)) {
				error = ENOENT;
				vrele(nvp);
			} else {
				error = 0;
			}
		}
	}
	free(nfhp, M_NFSFH);
	if (error)
		return (error);
	if (nvp != NULL) {
		*npp = VTONFS(nvp);
		return (0);
	}
	return (EINVAL);
}

static void
nfscl_warn_fileid(struct nfsmount *nmp, struct nfsvattr *oldnap,
    struct nfsvattr *newnap)
{
	int off;

	if (ncl_fileid_maxwarnings >= 0 &&
	    ncl_fileid_nwarnings >= ncl_fileid_maxwarnings)
		return;
	off = 0;
	if (ncl_fileid_maxwarnings >= 0) {
		if (++ncl_fileid_nwarnings >= ncl_fileid_maxwarnings)
			off = 1;
	}

	printf("newnfs: server '%s' error: fileid changed. "
	    "fsid %jx:%jx: expected fileid %#jx, got %#jx. "
	    "(BROKEN NFS SERVER OR MIDDLEWARE)\n",
	    nmp->nm_com.nmcom_hostname,
	    (uintmax_t)nmp->nm_fsid[0],
	    (uintmax_t)nmp->nm_fsid[1],
	    (uintmax_t)oldnap->na_fileid,
	    (uintmax_t)newnap->na_fileid);

	if (off)
		printf("newnfs: Logged %d times about fileid corruption; "
		    "going quiet to avoid spamming logs excessively. (Limit "
		    "is: %d).\n", ncl_fileid_nwarnings,
		    ncl_fileid_maxwarnings);
}

/*
 * Load the attribute cache (that lives in the nfsnode entry) with
 * the attributes of the second argument and
 * Iff vaper not NULL
 *    copy the attributes to *vaper
 * Similar to nfs_loadattrcache(), except the attributes are passed in
 * instead of being parsed out of the mbuf list.
 */
int
nfscl_loadattrcache(struct vnode **vpp, struct nfsvattr *nap, void *nvaper,
    void *stuff, int writeattr, int dontshrink)
{
	struct vnode *vp = *vpp;
	struct vattr *vap, *nvap = &nap->na_vattr, *vaper = nvaper;
	struct nfsnode *np;
	struct nfsmount *nmp;
	struct timespec mtime_save;
	int error, force_fid_err;

	error = 0;

	/*
	 * If v_type == VNON it is a new node, so fill in the v_type,
	 * n_mtime fields. Check to see if it represents a special 
	 * device, and if so, check for a possible alias. Once the
	 * correct vnode has been obtained, fill in the rest of the
	 * information.
	 */
	np = VTONFS(vp);
	NFSLOCKNODE(np);
	if (vp->v_type != nvap->va_type) {
		vp->v_type = nvap->va_type;
		if (vp->v_type == VFIFO)
			vp->v_op = &newnfs_fifoops;
		np->n_mtime = nvap->va_mtime;
	}
	nmp = VFSTONFS(vp->v_mount);
	vap = &np->n_vattr.na_vattr;
	mtime_save = vap->va_mtime;
	if (writeattr) {
		np->n_vattr.na_filerev = nap->na_filerev;
		np->n_vattr.na_size = nap->na_size;
		np->n_vattr.na_mtime = nap->na_mtime;
		np->n_vattr.na_ctime = nap->na_ctime;
		np->n_vattr.na_fsid = nap->na_fsid;
		np->n_vattr.na_mode = nap->na_mode;
	} else {
		force_fid_err = 0;
		KFAIL_POINT_ERROR(DEBUG_FP, nfscl_force_fileid_warning,
		    force_fid_err);
		/*
		 * BROKEN NFS SERVER OR MIDDLEWARE
		 *
		 * Certain NFS servers (certain old proprietary filers ca.
		 * 2006) or broken middleboxes (e.g. WAN accelerator products)
		 * will respond to GETATTR requests with results for a
		 * different fileid.
		 *
		 * The WAN accelerator we've observed not only serves stale
		 * cache results for a given file, it also occasionally serves
		 * results for wholly different files.  This causes surprising
		 * problems; for example the cached size attribute of a file
		 * may truncate down and then back up, resulting in zero
		 * regions in file contents read by applications.  We observed
		 * this reliably with Clang and .c files during parallel build.
		 * A pcap revealed packet fragmentation and GETATTR RPC
		 * responses with wholly wrong fileids.
		 */
		if ((np->n_vattr.na_fileid != 0 &&
		     np->n_vattr.na_fileid != nap->na_fileid) ||
		    force_fid_err) {
			nfscl_warn_fileid(nmp, &np->n_vattr, nap);
			error = EIDRM;
			goto out;
		}
		NFSBCOPY((caddr_t)nap, (caddr_t)&np->n_vattr,
		    sizeof (struct nfsvattr));
	}

	/*
	 * For NFSv4, if the node's fsid is not equal to the mount point's
	 * fsid, return the low order 32bits of the node's fsid. This
	 * allows getcwd(3) to work. There is a chance that the fsid might
	 * be the same as a local fs, but since this is in an NFS mount
	 * point, I don't think that will cause any problems?
	 */
	if (NFSHASNFSV4(nmp) && NFSHASHASSETFSID(nmp) &&
	    (nmp->nm_fsid[0] != np->n_vattr.na_filesid[0] ||
	     nmp->nm_fsid[1] != np->n_vattr.na_filesid[1])) {
		/*
		 * va_fsid needs to be set to some value derived from
		 * np->n_vattr.na_filesid that is not equal
		 * vp->v_mount->mnt_stat.f_fsid[0], so that it changes
		 * from the value used for the top level server volume
		 * in the mounted subtree.
		 */
		vn_fsid(vp, vap);
		if ((uint32_t)vap->va_fsid == np->n_vattr.na_filesid[0])
			vap->va_fsid = hash32_buf(
			    np->n_vattr.na_filesid, 2 * sizeof(uint64_t), 0);
	} else
		vn_fsid(vp, vap);
	np->n_attrstamp = time_second;
	if (vap->va_size != np->n_size) {
		if (vap->va_type == VREG) {
			if (dontshrink && vap->va_size < np->n_size) {
				/*
				 * We've been told not to shrink the file;
				 * zero np->n_attrstamp to indicate that
				 * the attributes are stale.
				 */
				vap->va_size = np->n_size;
				np->n_attrstamp = 0;
				KDTRACE_NFS_ATTRCACHE_FLUSH_DONE(vp);
			} else if (np->n_flag & NMODIFIED) {
				/*
				 * We've modified the file: Use the larger
				 * of our size, and the server's size.
				 */
				if (vap->va_size < np->n_size) {
					vap->va_size = np->n_size;
				} else {
					np->n_size = vap->va_size;
					np->n_flag |= NSIZECHANGED;
				}
			} else {
				np->n_size = vap->va_size;
				np->n_flag |= NSIZECHANGED;
			}
		} else {
			np->n_size = vap->va_size;
		}
	}
	/*
	 * The following checks are added to prevent a race between (say)
	 * a READDIR+ and a WRITE. 
	 * READDIR+, WRITE requests sent out.
	 * READDIR+ resp, WRITE resp received on client.
	 * However, the WRITE resp was handled before the READDIR+ resp
	 * causing the post op attrs from the write to be loaded first
	 * and the attrs from the READDIR+ to be loaded later. If this 
	 * happens, we have stale attrs loaded into the attrcache.
	 * We detect this by for the mtime moving back. We invalidate the 
	 * attrcache when this happens.
	 */
	if (timespeccmp(&mtime_save, &vap->va_mtime, >)) {
		/* Size changed or mtime went backwards */
		np->n_attrstamp = 0;
		KDTRACE_NFS_ATTRCACHE_FLUSH_DONE(vp);
	}
	if (vaper != NULL) {
		NFSBCOPY((caddr_t)vap, (caddr_t)vaper, sizeof(*vap));
		if (np->n_flag & NCHG) {
			if (np->n_flag & NACC)
				vaper->va_atime = np->n_atim;
			if (np->n_flag & NUPD)
				vaper->va_mtime = np->n_mtim;
		}
	}

out:
#ifdef KDTRACE_HOOKS
	if (np->n_attrstamp != 0)
		KDTRACE_NFS_ATTRCACHE_LOAD_DONE(vp, vap, error);
#endif
	(void)ncl_pager_setsize(vp, NULL);
	return (error);
}

/*
 * Call vnode_pager_setsize() if the size of the node changed, as
 * recorded in nfsnode vs. v_object, or delay the call if notifying
 * the pager is not possible at the moment.
 *
 * If nsizep is non-NULL, the call is delayed and the new node size is
 * provided.  Caller should itself call vnode_pager_setsize() if
 * function returned true.  If nsizep is NULL, function tries to call
 * vnode_pager_setsize() itself if needed and possible, and the nfs
 * node is unlocked unconditionally, the return value is not useful.
 */
bool
ncl_pager_setsize(struct vnode *vp, u_quad_t *nsizep)
{
	struct nfsnode *np;
	vm_object_t object;
	struct vattr *vap;
	u_quad_t nsize;
	bool setnsize;

	np = VTONFS(vp);
	NFSASSERTNODE(np);

	vap = &np->n_vattr.na_vattr;
	nsize = vap->va_size;
	object = vp->v_object;
	setnsize = false;

	if (object != NULL && nsize != object->un_pager.vnp.vnp_size) {
		if (VOP_ISLOCKED(vp) == LK_EXCLUSIVE)
			setnsize = true;
		else
			np->n_flag |= NVNSETSZSKIP;
	}
	if (nsizep == NULL) {
		NFSUNLOCKNODE(np);
		if (setnsize)
			vnode_pager_setsize(vp, nsize);
		setnsize = false;
	} else {
		*nsizep = nsize;
	}
	return (setnsize);
}

/*
 * Fill in the client id name. For these bytes:
 * 1 - they must be unique
 * 2 - they should be persistent across client reboots
 * 1 is more critical than 2
 * Use the mount point's unique id plus either the uuid or, if that
 * isn't set, random junk.
 */
void
nfscl_fillclid(u_int64_t clval, char *uuid, u_int8_t *cp, u_int16_t idlen)
{
	int uuidlen;

	/*
	 * First, put in the 64bit mount point identifier.
	 */
	if (idlen >= sizeof (u_int64_t)) {
		NFSBCOPY((caddr_t)&clval, cp, sizeof (u_int64_t));
		cp += sizeof (u_int64_t);
		idlen -= sizeof (u_int64_t);
	}

	/*
	 * If uuid is non-zero length, use it.
	 */
	uuidlen = strlen(uuid);
	if (uuidlen > 0 && idlen >= uuidlen) {
		NFSBCOPY(uuid, cp, uuidlen);
		cp += uuidlen;
		idlen -= uuidlen;
	}

	/*
	 * This only normally happens if the uuid isn't set.
	 */
	while (idlen > 0) {
		*cp++ = (u_int8_t)(arc4random() % 256);
		idlen--;
	}
}

/*
 * Fill in a lock owner name. For now, pid + the process's creation time.
 */
void
nfscl_filllockowner(void *id, u_int8_t *cp, int flags)
{
	union {
		u_int32_t	lval;
		u_int8_t	cval[4];
	} tl;
	struct proc *p;

	if (id == NULL) {
		/* Return the single open_owner of all 0 bytes. */
		bzero(cp, NFSV4CL_LOCKNAMELEN);
		return;
	}
	if ((flags & F_POSIX) != 0) {
		p = (struct proc *)id;
		tl.lval = p->p_pid;
		*cp++ = tl.cval[0];
		*cp++ = tl.cval[1];
		*cp++ = tl.cval[2];
		*cp++ = tl.cval[3];
		tl.lval = p->p_stats->p_start.tv_sec;
		*cp++ = tl.cval[0];
		*cp++ = tl.cval[1];
		*cp++ = tl.cval[2];
		*cp++ = tl.cval[3];
		tl.lval = p->p_stats->p_start.tv_usec;
		*cp++ = tl.cval[0];
		*cp++ = tl.cval[1];
		*cp++ = tl.cval[2];
		*cp = tl.cval[3];
	} else if ((flags & F_FLOCK) != 0) {
		bcopy(&id, cp, sizeof(id));
		bzero(&cp[sizeof(id)], NFSV4CL_LOCKNAMELEN - sizeof(id));
	} else {
		printf("nfscl_filllockowner: not F_POSIX or F_FLOCK\n");
		bzero(cp, NFSV4CL_LOCKNAMELEN);
	}
}

/*
 * Find the parent process for the thread passed in as an argument.
 * If none exists, return NULL, otherwise return a thread for the parent.
 * (Can be any of the threads, since it is only used for td->td_proc.)
 */
NFSPROC_T *
nfscl_getparent(struct thread *td)
{
	struct proc *p;
	struct thread *ptd;

	if (td == NULL)
		return (NULL);
	p = td->td_proc;
	if (p->p_pid == 0)
		return (NULL);
	p = p->p_pptr;
	if (p == NULL)
		return (NULL);
	ptd = TAILQ_FIRST(&p->p_threads);
	return (ptd);
}

/*
 * Start up the renew kernel thread.
 */
static void
start_nfscl(void *arg)
{
	struct nfsclclient *clp;
	struct thread *td;

	clp = (struct nfsclclient *)arg;
	td = TAILQ_FIRST(&clp->nfsc_renewthread->p_threads);
	nfscl_renewthread(clp, td);
	kproc_exit(0);
}

void
nfscl_start_renewthread(struct nfsclclient *clp)
{

	kproc_create(start_nfscl, (void *)clp, &clp->nfsc_renewthread, 0, 0,
	    "nfscl");
}

/*
 * Handle wcc_data.
 * For NFSv4, it assumes that nfsv4_wccattr() was used to set up the getattr
 * as the first Op after PutFH.
 * (For NFSv4, the postop attributes are after the Op, so they can't be
 *  parsed here. A separate call to nfscl_postop_attr() is required.)
 */
int
nfscl_wcc_data(struct nfsrv_descript *nd, struct vnode *vp,
    struct nfsvattr *nap, int *flagp, int *wccflagp, void *stuff)
{
	u_int32_t *tl;
	struct nfsnode *np = VTONFS(vp);
	struct nfsvattr nfsva;
	int error = 0;

	if (wccflagp != NULL)
		*wccflagp = 0;
	if (nd->nd_flag & ND_NFSV3) {
		*flagp = 0;
		NFSM_DISSECT(tl, u_int32_t *, NFSX_UNSIGNED);
		if (*tl == newnfs_true) {
			NFSM_DISSECT(tl, u_int32_t *, 6 * NFSX_UNSIGNED);
			if (wccflagp != NULL) {
				NFSLOCKNODE(np);
				*wccflagp = (np->n_mtime.tv_sec ==
				    fxdr_unsigned(u_int32_t, *(tl + 2)) &&
				    np->n_mtime.tv_nsec ==
				    fxdr_unsigned(u_int32_t, *(tl + 3)));
				NFSUNLOCKNODE(np);
			}
		}
		error = nfscl_postop_attr(nd, nap, flagp, stuff);
		if (wccflagp != NULL && *flagp == 0)
			*wccflagp = 0;
	} else if ((nd->nd_flag & (ND_NOMOREDATA | ND_NFSV4 | ND_V4WCCATTR))
	    == (ND_NFSV4 | ND_V4WCCATTR)) {
		error = nfsv4_loadattr(nd, NULL, &nfsva, NULL,
		    NULL, 0, NULL, NULL, NULL, NULL, NULL, 0,
		    NULL, NULL, NULL, NULL, NULL);
		if (error)
			return (error);
		/*
		 * Get rid of Op# and status for next op.
		 */
		NFSM_DISSECT(tl, u_int32_t *, 2 * NFSX_UNSIGNED);
		if (*++tl)
			nd->nd_flag |= ND_NOMOREDATA;
		if (wccflagp != NULL &&
		    nfsva.na_vattr.va_mtime.tv_sec != 0) {
			NFSLOCKNODE(np);
			*wccflagp = (np->n_mtime.tv_sec ==
			    nfsva.na_vattr.va_mtime.tv_sec &&
			    np->n_mtime.tv_nsec ==
			    nfsva.na_vattr.va_mtime.tv_sec);
			NFSUNLOCKNODE(np);
		}
	}
nfsmout:
	return (error);
}

/*
 * Get postop attributes.
 */
int
nfscl_postop_attr(struct nfsrv_descript *nd, struct nfsvattr *nap, int *retp,
    void *stuff)
{
	u_int32_t *tl;
	int error = 0;

	*retp = 0;
	if (nd->nd_flag & ND_NOMOREDATA)
		return (error);
	if (nd->nd_flag & ND_NFSV3) {
		NFSM_DISSECT(tl, u_int32_t *, NFSX_UNSIGNED);
		*retp = fxdr_unsigned(int, *tl);
	} else if (nd->nd_flag & ND_NFSV4) {
		/*
		 * For NFSv4, the postop attr are at the end, so no point
		 * in looking if nd_repstat != 0.
		 */
		if (!nd->nd_repstat) {
			NFSM_DISSECT(tl, u_int32_t *, 2 * NFSX_UNSIGNED);
			if (*(tl + 1))
				/* should never happen since nd_repstat != 0 */
				nd->nd_flag |= ND_NOMOREDATA;
			else
				*retp = 1;
		}
	} else if (!nd->nd_repstat) {
		/* For NFSv2, the attributes are here iff nd_repstat == 0 */
		*retp = 1;
	}
	if (*retp) {
		error = nfsm_loadattr(nd, nap);
		if (error)
			*retp = 0;
	}
nfsmout:
	return (error);
}

/*
 * nfscl_request() - mostly a wrapper for newnfs_request().
 */
int
nfscl_request(struct nfsrv_descript *nd, struct vnode *vp, NFSPROC_T *p,
    struct ucred *cred, void *stuff)
{
	int ret, vers;
	struct nfsmount *nmp;

	nmp = VFSTONFS(vp->v_mount);
	if (nd->nd_flag & ND_NFSV4)
		vers = NFS_VER4;
	else if (nd->nd_flag & ND_NFSV3)
		vers = NFS_VER3;
	else
		vers = NFS_VER2;
	ret = newnfs_request(nd, nmp, NULL, &nmp->nm_sockreq, vp, p, cred,
		NFS_PROG, vers, NULL, 1, NULL, NULL);
	return (ret);
}

/*
 * fill in this bsden's variant of statfs using nfsstatfs.
 */
void
nfscl_loadsbinfo(struct nfsmount *nmp, struct nfsstatfs *sfp, void *statfs)
{
	struct statfs *sbp = (struct statfs *)statfs;

	if (nmp->nm_flag & (NFSMNT_NFSV3 | NFSMNT_NFSV4)) {
		sbp->f_bsize = NFS_FABLKSIZE;
		sbp->f_blocks = sfp->sf_tbytes / NFS_FABLKSIZE;
		sbp->f_bfree = sfp->sf_fbytes / NFS_FABLKSIZE;
		/*
		 * Although sf_abytes is uint64_t and f_bavail is int64_t,
		 * the value after dividing by NFS_FABLKSIZE is small
		 * enough that it will fit in 63bits, so it is ok to
		 * assign it to f_bavail without fear that it will become
		 * negative.
		 */
		sbp->f_bavail = sfp->sf_abytes / NFS_FABLKSIZE;
		sbp->f_files = sfp->sf_tfiles;
		/* Since f_ffree is int64_t, clip it to 63bits. */
		if (sfp->sf_ffiles > INT64_MAX)
			sbp->f_ffree = INT64_MAX;
		else
			sbp->f_ffree = sfp->sf_ffiles;
	} else if ((nmp->nm_flag & NFSMNT_NFSV4) == 0) {
		/*
		 * The type casts to (int32_t) ensure that this code is
		 * compatible with the old NFS client, in that it will
		 * propagate bit31 to the high order bits. This may or may
		 * not be correct for NFSv2, but since it is a legacy
		 * environment, I'd rather retain backwards compatibility.
		 */
		sbp->f_bsize = (int32_t)sfp->sf_bsize;
		sbp->f_blocks = (int32_t)sfp->sf_blocks;
		sbp->f_bfree = (int32_t)sfp->sf_bfree;
		sbp->f_bavail = (int32_t)sfp->sf_bavail;
		sbp->f_files = 0;
		sbp->f_ffree = 0;
	}
}

/*
 * Use the fsinfo stuff to update the mount point.
 */
void
nfscl_loadfsinfo(struct nfsmount *nmp, struct nfsfsinfo *fsp)
{

	if ((nmp->nm_wsize == 0 || fsp->fs_wtpref < nmp->nm_wsize) &&
	    fsp->fs_wtpref >= NFS_FABLKSIZE)
		nmp->nm_wsize = (fsp->fs_wtpref + NFS_FABLKSIZE - 1) &
		    ~(NFS_FABLKSIZE - 1);
	if (fsp->fs_wtmax < nmp->nm_wsize && fsp->fs_wtmax > 0) {
		nmp->nm_wsize = fsp->fs_wtmax & ~(NFS_FABLKSIZE - 1);
		if (nmp->nm_wsize == 0)
			nmp->nm_wsize = fsp->fs_wtmax;
	}
	if (nmp->nm_wsize < NFS_FABLKSIZE)
		nmp->nm_wsize = NFS_FABLKSIZE;
	if ((nmp->nm_rsize == 0 || fsp->fs_rtpref < nmp->nm_rsize) &&
	    fsp->fs_rtpref >= NFS_FABLKSIZE)
		nmp->nm_rsize = (fsp->fs_rtpref + NFS_FABLKSIZE - 1) &
		    ~(NFS_FABLKSIZE - 1);
	if (fsp->fs_rtmax < nmp->nm_rsize && fsp->fs_rtmax > 0) {
		nmp->nm_rsize = fsp->fs_rtmax & ~(NFS_FABLKSIZE - 1);
		if (nmp->nm_rsize == 0)
			nmp->nm_rsize = fsp->fs_rtmax;
	}
	if (nmp->nm_rsize < NFS_FABLKSIZE)
		nmp->nm_rsize = NFS_FABLKSIZE;
	if ((nmp->nm_readdirsize == 0 || fsp->fs_dtpref < nmp->nm_readdirsize)
	    && fsp->fs_dtpref >= NFS_DIRBLKSIZ)
		nmp->nm_readdirsize = (fsp->fs_dtpref + NFS_DIRBLKSIZ - 1) &
		    ~(NFS_DIRBLKSIZ - 1);
	if (fsp->fs_rtmax < nmp->nm_readdirsize && fsp->fs_rtmax > 0) {
		nmp->nm_readdirsize = fsp->fs_rtmax & ~(NFS_DIRBLKSIZ - 1);
		if (nmp->nm_readdirsize == 0)
			nmp->nm_readdirsize = fsp->fs_rtmax;
	}
	if (nmp->nm_readdirsize < NFS_DIRBLKSIZ)
		nmp->nm_readdirsize = NFS_DIRBLKSIZ;
	if (fsp->fs_maxfilesize > 0 &&
	    fsp->fs_maxfilesize < nmp->nm_maxfilesize)
		nmp->nm_maxfilesize = fsp->fs_maxfilesize;
	nmp->nm_mountp->mnt_stat.f_iosize = newnfs_iosize(nmp);
	nmp->nm_state |= NFSSTA_GOTFSINFO;
}

/*
 * Lookups source address which should be used to communicate with
 * @nmp and stores it inside @pdst.
 *
 * Returns 0 on success.
 */
u_int8_t *
nfscl_getmyip(struct nfsmount *nmp, struct in6_addr *paddr, int *isinet6p)
{
#if defined(INET6) || defined(INET)
	int error, fibnum;

	fibnum = curthread->td_proc->p_fibnum;
#endif
#ifdef INET
	if (nmp->nm_nam->sa_family == AF_INET) {
		struct sockaddr_in *sin;
		struct nhop4_extended nh_ext;

		sin = (struct sockaddr_in *)nmp->nm_nam;
		CURVNET_SET(CRED_TO_VNET(nmp->nm_sockreq.nr_cred));
		error = fib4_lookup_nh_ext(fibnum, sin->sin_addr, 0, 0,
		    &nh_ext);
		CURVNET_RESTORE();
		if (error != 0)
			return (NULL);

		if (IN_LOOPBACK(ntohl(nh_ext.nh_src.s_addr))) {
			/* Ignore loopback addresses */
			return (NULL);
		}

		*isinet6p = 0;
		*((struct in_addr *)paddr) = nh_ext.nh_src;

		return (u_int8_t *)paddr;
	}
#endif
#ifdef INET6
	if (nmp->nm_nam->sa_family == AF_INET6) {
		struct sockaddr_in6 *sin6;

		sin6 = (struct sockaddr_in6 *)nmp->nm_nam;

		CURVNET_SET(CRED_TO_VNET(nmp->nm_sockreq.nr_cred));
		error = in6_selectsrc_addr(fibnum, &sin6->sin6_addr,
		    sin6->sin6_scope_id, NULL, paddr, NULL);
		CURVNET_RESTORE();
		if (error != 0)
			return (NULL);

		if (IN6_IS_ADDR_LOOPBACK(paddr))
			return (NULL);

		/* Scope is embedded in */
		*isinet6p = 1;

		return (u_int8_t *)paddr;
	}
#endif
	return (NULL);
}

/*
 * Copy NFS uid, gids from the cred structure.
 */
void
newnfs_copyincred(struct ucred *cr, struct nfscred *nfscr)
{
	int i;

	KASSERT(cr->cr_ngroups >= 0,
	    ("newnfs_copyincred: negative cr_ngroups"));
	nfscr->nfsc_uid = cr->cr_uid;
	nfscr->nfsc_ngroups = MIN(cr->cr_ngroups, NFS_MAXGRPS + 1);
	for (i = 0; i < nfscr->nfsc_ngroups; i++)
		nfscr->nfsc_groups[i] = cr->cr_groups[i];
}


/*
 * Do any client specific initialization.
 */
void
nfscl_init(void)
{
	static int inited = 0;

	if (inited)
		return;
	inited = 1;
	nfscl_inited = 1;
	ncl_pbuf_zone = pbuf_zsecond_create("nfspbuf", nswbuf / 2);
}

/*
 * Check each of the attributes to be set, to ensure they aren't already
 * the correct value. Disable setting ones already correct.
 */
int
nfscl_checksattr(struct vattr *vap, struct nfsvattr *nvap)
{

	if (vap->va_mode != (mode_t)VNOVAL) {
		if (vap->va_mode == nvap->na_mode)
			vap->va_mode = (mode_t)VNOVAL;
	}
	if (vap->va_uid != (uid_t)VNOVAL) {
		if (vap->va_uid == nvap->na_uid)
			vap->va_uid = (uid_t)VNOVAL;
	}
	if (vap->va_gid != (gid_t)VNOVAL) {
		if (vap->va_gid == nvap->na_gid)
			vap->va_gid = (gid_t)VNOVAL;
	}
	if (vap->va_size != VNOVAL) {
		if (vap->va_size == nvap->na_size)
			vap->va_size = VNOVAL;
	}

	/*
	 * We are normally called with only a partially initialized
	 * VAP.  Since the NFSv3 spec says that server may use the
	 * file attributes to store the verifier, the spec requires
	 * us to do a SETATTR RPC. FreeBSD servers store the verifier
	 * in atime, but we can't really assume that all servers will
	 * so we ensure that our SETATTR sets both atime and mtime.
	 * Set the VA_UTIMES_NULL flag for this case, so that
	 * the server's time will be used.  This is needed to
	 * work around a bug in some Solaris servers, where
	 * setting the time TOCLIENT causes the Setattr RPC
	 * to return NFS_OK, but not set va_mode.
	 */
	if (vap->va_mtime.tv_sec == VNOVAL) {
		vfs_timestamp(&vap->va_mtime);
		vap->va_vaflags |= VA_UTIMES_NULL;
	}
	if (vap->va_atime.tv_sec == VNOVAL)
		vap->va_atime = vap->va_mtime;
	return (1);
}

/*
 * Map nfsv4 errors to errno.h errors.
 * The uid and gid arguments are only used for NFSERR_BADOWNER and that
 * error should only be returned for the Open, Create and Setattr Ops.
 * As such, most calls can just pass in 0 for those arguments.
 */
APPLESTATIC int
nfscl_maperr(struct thread *td, int error, uid_t uid, gid_t gid)
{
	struct proc *p;

	if (error < 10000 || error >= NFSERR_STALEWRITEVERF)
		return (error);
	if (td != NULL)
		p = td->td_proc;
	else
		p = NULL;
	switch (error) {
	case NFSERR_BADOWNER:
		tprintf(p, LOG_INFO,
		    "No name and/or group mapping for uid,gid:(%d,%d)\n",
		    uid, gid);
		return (EPERM);
	case NFSERR_BADNAME:
	case NFSERR_BADCHAR:
		printf("nfsv4 char/name not handled by server\n");
		return (ENOENT);
	case NFSERR_STALECLIENTID:
	case NFSERR_STALESTATEID:
	case NFSERR_EXPIRED:
	case NFSERR_BADSTATEID:
	case NFSERR_BADSESSION:
		printf("nfsv4 recover err returned %d\n", error);
		return (EIO);
	case NFSERR_BADHANDLE:
	case NFSERR_SERVERFAULT:
	case NFSERR_BADTYPE:
	case NFSERR_FHEXPIRED:
	case NFSERR_RESOURCE:
	case NFSERR_MOVED:
	case NFSERR_NOFILEHANDLE:
	case NFSERR_MINORVERMISMATCH:
	case NFSERR_OLDSTATEID:
	case NFSERR_BADSEQID:
	case NFSERR_LEASEMOVED:
	case NFSERR_RECLAIMBAD:
	case NFSERR_BADXDR:
	case NFSERR_OPILLEGAL:
		printf("nfsv4 client/server protocol prob err=%d\n",
		    error);
		return (EIO);
	default:
		tprintf(p, LOG_INFO, "nfsv4 err=%d\n", error);
		return (EIO);
	};
}

/*
 * Check to see if the process for this owner exists. Return 1 if it doesn't
 * and 0 otherwise.
 */
int
nfscl_procdoesntexist(u_int8_t *own)
{
	union {
		u_int32_t	lval;
		u_int8_t	cval[4];
	} tl;
	struct proc *p;
	pid_t pid;
	int i, ret = 0;

	/* For the single open_owner of all 0 bytes, just return 0. */
	for (i = 0; i < NFSV4CL_LOCKNAMELEN; i++)
		if (own[i] != 0)
			break;
	if (i == NFSV4CL_LOCKNAMELEN)
		return (0);

	tl.cval[0] = *own++;
	tl.cval[1] = *own++;
	tl.cval[2] = *own++;
	tl.cval[3] = *own++;
	pid = tl.lval;
	p = pfind_any_locked(pid);
	if (p == NULL)
		return (1);
	if (p->p_stats == NULL) {
		PROC_UNLOCK(p);
		return (0);
	}
	tl.cval[0] = *own++;
	tl.cval[1] = *own++;
	tl.cval[2] = *own++;
	tl.cval[3] = *own++;
	if (tl.lval != p->p_stats->p_start.tv_sec) {
		ret = 1;
	} else {
		tl.cval[0] = *own++;
		tl.cval[1] = *own++;
		tl.cval[2] = *own++;
		tl.cval[3] = *own;
		if (tl.lval != p->p_stats->p_start.tv_usec)
			ret = 1;
	}
	PROC_UNLOCK(p);
	return (ret);
}

/*
 * - nfs pseudo system call for the client
 */
/*
 * MPSAFE
 */
static int
nfssvc_nfscl(struct thread *td, struct nfssvc_args *uap)
{
	struct file *fp;
	struct nfscbd_args nfscbdarg;
	struct nfsd_nfscbd_args nfscbdarg2;
	struct nameidata nd;
	struct nfscl_dumpmntopts dumpmntopts;
	cap_rights_t rights;
	char *buf;
	int error;
	struct mount *mp;
	struct nfsmount *nmp;

	if (uap->flag & NFSSVC_CBADDSOCK) {
		error = copyin(uap->argp, (caddr_t)&nfscbdarg, sizeof(nfscbdarg));
		if (error)
			return (error);
		/*
		 * Since we don't know what rights might be required,
		 * pretend that we need them all. It is better to be too
		 * careful than too reckless.
		 */
		error = fget(td, nfscbdarg.sock,
		    cap_rights_init(&rights, CAP_SOCK_CLIENT), &fp);
		if (error)
			return (error);
		if (fp->f_type != DTYPE_SOCKET) {
			fdrop(fp, td);
			return (EPERM);
		}
		error = nfscbd_addsock(fp);
		fdrop(fp, td);
		if (!error && nfscl_enablecallb == 0) {
			nfsv4_cbport = nfscbdarg.port;
			nfscl_enablecallb = 1;
		}
	} else if (uap->flag & NFSSVC_NFSCBD) {
		if (uap->argp == NULL) 
			return (EINVAL);
		error = copyin(uap->argp, (caddr_t)&nfscbdarg2,
		    sizeof(nfscbdarg2));
		if (error)
			return (error);
		error = nfscbd_nfsd(td, &nfscbdarg2);
	} else if (uap->flag & NFSSVC_DUMPMNTOPTS) {
		error = copyin(uap->argp, &dumpmntopts, sizeof(dumpmntopts));
		if (error == 0 && (dumpmntopts.ndmnt_blen < 256 ||
		    dumpmntopts.ndmnt_blen > 1024))
			error = EINVAL;
		if (error == 0)
			error = nfsrv_lookupfilename(&nd,
			    dumpmntopts.ndmnt_fname, td);
		if (error == 0 && strcmp(nd.ni_vp->v_mount->mnt_vfc->vfc_name,
		    "nfs") != 0) {
			vput(nd.ni_vp);
			error = EINVAL;
		}
		if (error == 0) {
			buf = malloc(dumpmntopts.ndmnt_blen, M_TEMP, M_WAITOK);
			nfscl_retopts(VFSTONFS(nd.ni_vp->v_mount), buf,
			    dumpmntopts.ndmnt_blen);
			vput(nd.ni_vp);
			error = copyout(buf, dumpmntopts.ndmnt_buf,
			    dumpmntopts.ndmnt_blen);
			free(buf, M_TEMP);
		}
	} else if (uap->flag & NFSSVC_FORCEDISM) {
		buf = malloc(MNAMELEN + 1, M_TEMP, M_WAITOK);
		error = copyinstr(uap->argp, buf, MNAMELEN + 1, NULL);
		if (error == 0) {
			nmp = NULL;
			mtx_lock(&mountlist_mtx);
			TAILQ_FOREACH(mp, &mountlist, mnt_list) {
				if (strcmp(mp->mnt_stat.f_mntonname, buf) ==
				    0 && strcmp(mp->mnt_stat.f_fstypename,
				    "nfs") == 0 && mp->mnt_data != NULL) {
					nmp = VFSTONFS(mp);
					NFSDDSLOCK();
					if (nfsv4_findmirror(nmp) != NULL) {
						NFSDDSUNLOCK();
						error = ENXIO;
						nmp = NULL;
						break;
					}
					mtx_lock(&nmp->nm_mtx);
					if ((nmp->nm_privflag &
					    NFSMNTP_FORCEDISM) == 0) {
						nmp->nm_privflag |= 
						   (NFSMNTP_FORCEDISM |
						    NFSMNTP_CANCELRPCS);
						mtx_unlock(&nmp->nm_mtx);
					} else {
						mtx_unlock(&nmp->nm_mtx);
						nmp = NULL;
					}
					NFSDDSUNLOCK();
					break;
				}
			}
			mtx_unlock(&mountlist_mtx);

			if (nmp != NULL) {
				/*
				 * Call newnfs_nmcancelreqs() to cause
				 * any RPCs in progress on the mount point to
				 * fail.
				 * This will cause any process waiting for an
				 * RPC to complete while holding a vnode lock
				 * on the mounted-on vnode (such as "df" or
				 * a non-forced "umount") to fail.
				 * This will unlock the mounted-on vnode so
				 * a forced dismount can succeed.
				 * Then clear NFSMNTP_CANCELRPCS and wakeup(),
				 * so that nfs_unmount() can complete.
				 */
				newnfs_nmcancelreqs(nmp);
				mtx_lock(&nmp->nm_mtx);
				nmp->nm_privflag &= ~NFSMNTP_CANCELRPCS;
				wakeup(nmp);
				mtx_unlock(&nmp->nm_mtx);
			} else if (error == 0)
				error = EINVAL;
		}
		free(buf, M_TEMP);
	} else {
		error = EINVAL;
	}
	return (error);
}

extern int (*nfsd_call_nfscl)(struct thread *, struct nfssvc_args *);

/*
 * Called once to initialize data structures...
 */
static int
nfscl_modevent(module_t mod, int type, void *data)
{
	int error = 0;
	static int loaded = 0;

	switch (type) {
	case MOD_LOAD:
		if (loaded)
			return (0);
		newnfs_portinit();
		mtx_init(&ncl_iod_mutex, "ncl_iod_mutex", NULL, MTX_DEF);
		nfscl_init();
		NFSD_LOCK();
		nfsrvd_cbinit(0);
		NFSD_UNLOCK();
		ncl_call_invalcaches = ncl_invalcaches;
		nfsd_call_nfscl = nfssvc_nfscl;
		loaded = 1;
		break;

	case MOD_UNLOAD:
		if (nfs_numnfscbd != 0) {
			error = EBUSY;
			break;
		}

		/*
		 * XXX: Unloading of nfscl module is unsupported.
		 */
#if 0
		ncl_call_invalcaches = NULL;
		nfsd_call_nfscl = NULL;
		uma_zdestroy(ncl_pbuf_zone);
		/* and get rid of the mutexes */
		mtx_destroy(&ncl_iod_mutex);
		loaded = 0;
		break;
#else
		/* FALLTHROUGH */
#endif
	default:
		error = EOPNOTSUPP;
		break;
	}
	return error;
}
static moduledata_t nfscl_mod = {
	"nfscl",
	nfscl_modevent,
	NULL,
};
DECLARE_MODULE(nfscl, nfscl_mod, SI_SUB_VFS, SI_ORDER_FIRST);

/* So that loader and kldload(2) can find us, wherever we are.. */
MODULE_VERSION(nfscl, 1);
MODULE_DEPEND(nfscl, nfscommon, 1, 1, 1);
MODULE_DEPEND(nfscl, krpc, 1, 1, 1);
MODULE_DEPEND(nfscl, nfssvc, 1, 1, 1);
MODULE_DEPEND(nfscl, nfslock, 1, 1, 1);