summaryrefslogtreecommitdiffstats
path: root/sys/net/if_vlan.c
blob: 4e24a5fc40697983e4eab7b3a0a2c8de34981172 (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
/*-
 * Copyright 1998 Massachusetts Institute of Technology
 *
 * Permission to use, copy, modify, and distribute this software and
 * its documentation for any purpose and without fee is hereby
 * granted, provided that both the above copyright notice and this
 * permission notice appear in all copies, that both the above
 * copyright notice and this permission notice appear in all
 * supporting documentation, and that the name of M.I.T. not be used
 * in advertising or publicity pertaining to distribution of the
 * software without specific, written prior permission.  M.I.T. makes
 * no representations about the suitability of this software for any
 * purpose.  It is provided "as is" without express or implied
 * warranty.
 * 
 * THIS SOFTWARE IS PROVIDED BY M.I.T. ``AS IS''.  M.I.T. DISCLAIMS
 * ALL EXPRESS OR IMPLIED WARRANTIES WITH REGARD TO THIS SOFTWARE,
 * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
 * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT
 * SHALL M.I.T. 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.
 */

/*
 * if_vlan.c - pseudo-device driver for IEEE 802.1Q virtual LANs.
 * Might be extended some day to also handle IEEE 802.1p priority
 * tagging.  This is sort of sneaky in the implementation, since
 * we need to pretend to be enough of an Ethernet implementation
 * to make arp work.  The way we do this is by telling everyone
 * that we are an Ethernet, and then catch the packets that
 * ether_output() left on our output queue when it calls
 * if_start(), rewrite them for use by the real outgoing interface,
 * and ask it to send them.
 */

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

#include "opt_vlan.h"

#include <sys/param.h>
#include <sys/kernel.h>
#include <sys/lock.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/module.h>
#include <sys/rwlock.h>
#include <sys/queue.h>
#include <sys/socket.h>
#include <sys/sockio.h>
#include <sys/sysctl.h>
#include <sys/systm.h>

#include <net/bpf.h>
#include <net/ethernet.h>
#include <net/if.h>
#include <net/if_clone.h>
#include <net/if_dl.h>
#include <net/if_types.h>
#include <net/if_vlan_var.h>
#include <net/vnet.h>

#define VLANNAME	"vlan"
#define	VLAN_DEF_HWIDTH	4
#define	VLAN_IFFLAGS	(IFF_BROADCAST | IFF_MULTICAST)

#define	UP_AND_RUNNING(ifp) \
    ((ifp)->if_flags & IFF_UP && (ifp)->if_drv_flags & IFF_DRV_RUNNING)

LIST_HEAD(ifvlanhead, ifvlan);

struct ifvlantrunk {
	struct	ifnet   *parent;	/* parent interface of this trunk */
	struct	rwlock	rw;
#ifdef VLAN_ARRAY
#define	VLAN_ARRAY_SIZE	(EVL_VLID_MASK + 1)
	struct	ifvlan	*vlans[VLAN_ARRAY_SIZE]; /* static table */
#else
	struct	ifvlanhead *hash;	/* dynamic hash-list table */
	uint16_t	hmask;
	uint16_t	hwidth;
#endif
	int		refcnt;
};

struct vlan_mc_entry {
	struct ether_addr		mc_addr;
	SLIST_ENTRY(vlan_mc_entry)	mc_entries;
};

struct	ifvlan {
	struct	ifvlantrunk *ifv_trunk;
	struct	ifnet *ifv_ifp;
#define	TRUNK(ifv)	((ifv)->ifv_trunk)
#define	PARENT(ifv)	((ifv)->ifv_trunk->parent)
	int	ifv_pflags;	/* special flags we have set on parent */
	struct	ifv_linkmib {
		int	ifvm_encaplen;	/* encapsulation length */
		int	ifvm_mtufudge;	/* MTU fudged by this much */
		int	ifvm_mintu;	/* min transmission unit */
		uint16_t ifvm_proto;	/* encapsulation ethertype */
		uint16_t ifvm_tag;	/* tag to apply on packets leaving if */
	}	ifv_mib;
	SLIST_HEAD(, vlan_mc_entry) vlan_mc_listhead;
#ifndef VLAN_ARRAY
	LIST_ENTRY(ifvlan) ifv_list;
#endif
};
#define	ifv_proto	ifv_mib.ifvm_proto
#define	ifv_tag		ifv_mib.ifvm_tag
#define	ifv_encaplen	ifv_mib.ifvm_encaplen
#define	ifv_mtufudge	ifv_mib.ifvm_mtufudge
#define	ifv_mintu	ifv_mib.ifvm_mintu

/* Special flags we should propagate to parent. */
static struct {
	int flag;
	int (*func)(struct ifnet *, int);
} vlan_pflags[] = {
	{IFF_PROMISC, ifpromisc},
	{IFF_ALLMULTI, if_allmulti},
	{0, NULL}
};

SYSCTL_DECL(_net_link);
SYSCTL_NODE(_net_link, IFT_L2VLAN, vlan, CTLFLAG_RW, 0, "IEEE 802.1Q VLAN");
SYSCTL_NODE(_net_link_vlan, PF_LINK, link, CTLFLAG_RW, 0, "for consistency");

static int soft_pad = 0;
SYSCTL_INT(_net_link_vlan, OID_AUTO, soft_pad, CTLFLAG_RW, &soft_pad, 0,
	   "pad short frames before tagging");

static MALLOC_DEFINE(M_VLAN, VLANNAME, "802.1Q Virtual LAN Interface");

static eventhandler_tag ifdetach_tag;
static eventhandler_tag iflladdr_tag;

/*
 * We have a global mutex, that is used to serialize configuration
 * changes and isn't used in normal packet delivery.
 *
 * We also have a per-trunk rwlock, that is locked shared on packet
 * processing and exclusive when configuration is changed.
 *
 * The VLAN_ARRAY substitutes the dynamic hash with a static array
 * with 4096 entries. In theory this can give a boost in processing,
 * however on practice it does not. Probably this is because array
 * is too big to fit into CPU cache.
 */
static struct mtx ifv_mtx;
#define	VLAN_LOCK_INIT()	mtx_init(&ifv_mtx, "vlan_global", NULL, MTX_DEF)
#define	VLAN_LOCK_DESTROY()	mtx_destroy(&ifv_mtx)
#define	VLAN_LOCK_ASSERT()	mtx_assert(&ifv_mtx, MA_OWNED)
#define	VLAN_LOCK()		mtx_lock(&ifv_mtx)
#define	VLAN_UNLOCK()		mtx_unlock(&ifv_mtx)
#define	TRUNK_LOCK_INIT(trunk)	rw_init(&(trunk)->rw, VLANNAME)
#define	TRUNK_LOCK_DESTROY(trunk) rw_destroy(&(trunk)->rw)
#define	TRUNK_LOCK(trunk)	rw_wlock(&(trunk)->rw)
#define	TRUNK_UNLOCK(trunk)	rw_wunlock(&(trunk)->rw)
#define	TRUNK_LOCK_ASSERT(trunk) rw_assert(&(trunk)->rw, RA_WLOCKED)
#define	TRUNK_RLOCK(trunk)	rw_rlock(&(trunk)->rw)
#define	TRUNK_RUNLOCK(trunk)	rw_runlock(&(trunk)->rw)
#define	TRUNK_LOCK_RASSERT(trunk) rw_assert(&(trunk)->rw, RA_RLOCKED)

#ifndef VLAN_ARRAY
static	void vlan_inithash(struct ifvlantrunk *trunk);
static	void vlan_freehash(struct ifvlantrunk *trunk);
static	int vlan_inshash(struct ifvlantrunk *trunk, struct ifvlan *ifv);
static	int vlan_remhash(struct ifvlantrunk *trunk, struct ifvlan *ifv);
static	void vlan_growhash(struct ifvlantrunk *trunk, int howmuch);
static __inline struct ifvlan * vlan_gethash(struct ifvlantrunk *trunk,
	uint16_t tag);
#endif
static	void trunk_destroy(struct ifvlantrunk *trunk);

static	void vlan_start(struct ifnet *ifp);
static	void vlan_init(void *foo);
static	void vlan_input(struct ifnet *ifp, struct mbuf *m);
static	int vlan_ioctl(struct ifnet *ifp, u_long cmd, caddr_t addr);
static	int vlan_setflag(struct ifnet *ifp, int flag, int status,
    int (*func)(struct ifnet *, int));
static	int vlan_setflags(struct ifnet *ifp, int status);
static	int vlan_setmulti(struct ifnet *ifp);
static	void vlan_unconfig(struct ifnet *ifp);
static	void vlan_unconfig_locked(struct ifnet *ifp);
static	int vlan_config(struct ifvlan *ifv, struct ifnet *p, uint16_t tag);
static	void vlan_link_state(struct ifnet *ifp);
static	void vlan_capabilities(struct ifvlan *ifv);
static	void vlan_trunk_capabilities(struct ifnet *ifp);

static	struct ifnet *vlan_clone_match_ethertag(struct if_clone *,
    const char *, int *);
static	int vlan_clone_match(struct if_clone *, const char *);
static	int vlan_clone_create(struct if_clone *, char *, size_t, caddr_t);
static	int vlan_clone_destroy(struct if_clone *, struct ifnet *);

static	void vlan_ifdetach(void *arg, struct ifnet *ifp);
static  void vlan_iflladdr(void *arg, struct ifnet *ifp);

static	struct if_clone vlan_cloner = IFC_CLONE_INITIALIZER(VLANNAME, NULL,
    IF_MAXUNIT, NULL, vlan_clone_match, vlan_clone_create, vlan_clone_destroy);

#ifdef VIMAGE
static VNET_DEFINE(struct if_clone, vlan_cloner);
#define	V_vlan_cloner	VNET(vlan_cloner)
#endif

#ifndef VLAN_ARRAY
#define HASH(n, m)	((((n) >> 8) ^ ((n) >> 4) ^ (n)) & (m))

static void
vlan_inithash(struct ifvlantrunk *trunk)
{
	int i, n;
	
	/*
	 * The trunk must not be locked here since we call malloc(M_WAITOK).
	 * It is OK in case this function is called before the trunk struct
	 * gets hooked up and becomes visible from other threads.
	 */

	KASSERT(trunk->hwidth == 0 && trunk->hash == NULL,
	    ("%s: hash already initialized", __func__));

	trunk->hwidth = VLAN_DEF_HWIDTH;
	n = 1 << trunk->hwidth;
	trunk->hmask = n - 1;
	trunk->hash = malloc(sizeof(struct ifvlanhead) * n, M_VLAN, M_WAITOK);
	for (i = 0; i < n; i++)
		LIST_INIT(&trunk->hash[i]);
}

static void
vlan_freehash(struct ifvlantrunk *trunk)
{
#ifdef INVARIANTS
	int i;

	KASSERT(trunk->hwidth > 0, ("%s: hwidth not positive", __func__));
	for (i = 0; i < (1 << trunk->hwidth); i++)
		KASSERT(LIST_EMPTY(&trunk->hash[i]),
		    ("%s: hash table not empty", __func__));
#endif
	free(trunk->hash, M_VLAN);
	trunk->hash = NULL;
	trunk->hwidth = trunk->hmask = 0;
}

static int
vlan_inshash(struct ifvlantrunk *trunk, struct ifvlan *ifv)
{
	int i, b;
	struct ifvlan *ifv2;

	TRUNK_LOCK_ASSERT(trunk);
	KASSERT(trunk->hwidth > 0, ("%s: hwidth not positive", __func__));

	b = 1 << trunk->hwidth;
	i = HASH(ifv->ifv_tag, trunk->hmask);
	LIST_FOREACH(ifv2, &trunk->hash[i], ifv_list)
		if (ifv->ifv_tag == ifv2->ifv_tag)
			return (EEXIST);

	/*
	 * Grow the hash when the number of vlans exceeds half of the number of
	 * hash buckets squared. This will make the average linked-list length
	 * buckets/2.
	 */
	if (trunk->refcnt > (b * b) / 2) {
		vlan_growhash(trunk, 1);
		i = HASH(ifv->ifv_tag, trunk->hmask);
	}
	LIST_INSERT_HEAD(&trunk->hash[i], ifv, ifv_list);
	trunk->refcnt++;

	return (0);
}

static int
vlan_remhash(struct ifvlantrunk *trunk, struct ifvlan *ifv)
{
	int i, b;
	struct ifvlan *ifv2;

	TRUNK_LOCK_ASSERT(trunk);
	KASSERT(trunk->hwidth > 0, ("%s: hwidth not positive", __func__));
	
	b = 1 << trunk->hwidth;
	i = HASH(ifv->ifv_tag, trunk->hmask);
	LIST_FOREACH(ifv2, &trunk->hash[i], ifv_list)
		if (ifv2 == ifv) {
			trunk->refcnt--;
			LIST_REMOVE(ifv2, ifv_list);
			if (trunk->refcnt < (b * b) / 2)
				vlan_growhash(trunk, -1);
			return (0);
		}

	panic("%s: vlan not found\n", __func__);
	return (ENOENT); /*NOTREACHED*/
}

/*
 * Grow the hash larger or smaller if memory permits.
 */
static void
vlan_growhash(struct ifvlantrunk *trunk, int howmuch)
{
	struct ifvlan *ifv;
	struct ifvlanhead *hash2;
	int hwidth2, i, j, n, n2;

	TRUNK_LOCK_ASSERT(trunk);
	KASSERT(trunk->hwidth > 0, ("%s: hwidth not positive", __func__));

	if (howmuch == 0) {
		/* Harmless yet obvious coding error */
		printf("%s: howmuch is 0\n", __func__);
		return;
	}

	hwidth2 = trunk->hwidth + howmuch;
	n = 1 << trunk->hwidth;
	n2 = 1 << hwidth2;
	/* Do not shrink the table below the default */
	if (hwidth2 < VLAN_DEF_HWIDTH)
		return;

	/* M_NOWAIT because we're called with trunk mutex held */
	hash2 = malloc(sizeof(struct ifvlanhead) * n2, M_VLAN, M_NOWAIT);
	if (hash2 == NULL) {
		printf("%s: out of memory -- hash size not changed\n",
		    __func__);
		return;		/* We can live with the old hash table */
	}
	for (j = 0; j < n2; j++)
		LIST_INIT(&hash2[j]);
	for (i = 0; i < n; i++)
		while ((ifv = LIST_FIRST(&trunk->hash[i])) != NULL) {
			LIST_REMOVE(ifv, ifv_list);
			j = HASH(ifv->ifv_tag, n2 - 1);
			LIST_INSERT_HEAD(&hash2[j], ifv, ifv_list);
		}
	free(trunk->hash, M_VLAN);
	trunk->hash = hash2;
	trunk->hwidth = hwidth2;
	trunk->hmask = n2 - 1;

	if (bootverbose)
		if_printf(trunk->parent,
		    "VLAN hash table resized from %d to %d buckets\n", n, n2);
}

static __inline struct ifvlan *
vlan_gethash(struct ifvlantrunk *trunk, uint16_t tag)
{
	struct ifvlan *ifv;

	TRUNK_LOCK_RASSERT(trunk);

	LIST_FOREACH(ifv, &trunk->hash[HASH(tag, trunk->hmask)], ifv_list)
		if (ifv->ifv_tag == tag)
			return (ifv);
	return (NULL);
}

#if 0
/* Debugging code to view the hashtables. */
static void
vlan_dumphash(struct ifvlantrunk *trunk)
{
	int i;
	struct ifvlan *ifv;

	for (i = 0; i < (1 << trunk->hwidth); i++) {
		printf("%d: ", i);
		LIST_FOREACH(ifv, &trunk->hash[i], ifv_list)
			printf("%s ", ifv->ifv_ifp->if_xname);
		printf("\n");
	}
}
#endif /* 0 */
#endif /* !VLAN_ARRAY */

static void
trunk_destroy(struct ifvlantrunk *trunk)
{
	VLAN_LOCK_ASSERT();

	TRUNK_LOCK(trunk);
#ifndef VLAN_ARRAY
	vlan_freehash(trunk);
#endif
	trunk->parent->if_vlantrunk = NULL;
	TRUNK_UNLOCK(trunk);
	TRUNK_LOCK_DESTROY(trunk);
	free(trunk, M_VLAN);
}

/*
 * Program our multicast filter. What we're actually doing is
 * programming the multicast filter of the parent. This has the
 * side effect of causing the parent interface to receive multicast
 * traffic that it doesn't really want, which ends up being discarded
 * later by the upper protocol layers. Unfortunately, there's no way
 * to avoid this: there really is only one physical interface.
 *
 * XXX: There is a possible race here if more than one thread is
 *      modifying the multicast state of the vlan interface at the same time.
 */
static int
vlan_setmulti(struct ifnet *ifp)
{
	struct ifnet		*ifp_p;
	struct ifmultiaddr	*ifma, *rifma = NULL;
	struct ifvlan		*sc;
	struct vlan_mc_entry	*mc;
	struct sockaddr_dl	sdl;
	int			error;

	/*VLAN_LOCK_ASSERT();*/

	/* Find the parent. */
	sc = ifp->if_softc;
	ifp_p = PARENT(sc);

	CURVNET_SET_QUIET(ifp_p->if_vnet);

	bzero((char *)&sdl, sizeof(sdl));
	sdl.sdl_len = sizeof(sdl);
	sdl.sdl_family = AF_LINK;
	sdl.sdl_index = ifp_p->if_index;
	sdl.sdl_type = IFT_ETHER;
	sdl.sdl_alen = ETHER_ADDR_LEN;

	/* First, remove any existing filter entries. */
	while ((mc = SLIST_FIRST(&sc->vlan_mc_listhead)) != NULL) {
		bcopy((char *)&mc->mc_addr, LLADDR(&sdl), ETHER_ADDR_LEN);
		error = if_delmulti(ifp_p, (struct sockaddr *)&sdl);
		if (error)
			return (error);
		SLIST_REMOVE_HEAD(&sc->vlan_mc_listhead, mc_entries);
		free(mc, M_VLAN);
	}

	/* Now program new ones. */
	TAILQ_FOREACH(ifma, &ifp->if_multiaddrs, ifma_link) {
		if (ifma->ifma_addr->sa_family != AF_LINK)
			continue;
		mc = malloc(sizeof(struct vlan_mc_entry), M_VLAN, M_NOWAIT);
		if (mc == NULL)
			return (ENOMEM);
		bcopy(LLADDR((struct sockaddr_dl *)ifma->ifma_addr),
		    (char *)&mc->mc_addr, ETHER_ADDR_LEN);
		SLIST_INSERT_HEAD(&sc->vlan_mc_listhead, mc, mc_entries);
		bcopy(LLADDR((struct sockaddr_dl *)ifma->ifma_addr),
		    LLADDR(&sdl), ETHER_ADDR_LEN);
		error = if_addmulti(ifp_p, (struct sockaddr *)&sdl, &rifma);
		if (error)
			return (error);
	}

	CURVNET_RESTORE();
	return (0);
}

/*
 * A handler for parent interface link layer address changes.
 * If the parent interface link layer address is changed we
 * should also change it on all children vlans.
 */
static void
vlan_iflladdr(void *arg __unused, struct ifnet *ifp)
{
	struct ifvlan *ifv;
#ifndef VLAN_ARRAY
	struct ifvlan *next;
#endif
	int i;

	/*
	 * Check if it's a trunk interface first of all
	 * to avoid needless locking.
	 */
	if (ifp->if_vlantrunk == NULL)
		return;

	VLAN_LOCK();
	/*
	 * OK, it's a trunk.  Loop over and change all vlan's lladdrs on it.
	 */
#ifdef VLAN_ARRAY
	for (i = 0; i < VLAN_ARRAY_SIZE; i++)
		if ((ifv = ifp->if_vlantrunk->vlans[i])) {
#else /* VLAN_ARRAY */
	for (i = 0; i < (1 << ifp->if_vlantrunk->hwidth); i++)
		LIST_FOREACH_SAFE(ifv, &ifp->if_vlantrunk->hash[i], ifv_list, next) {
#endif /* VLAN_ARRAY */
			VLAN_UNLOCK();
			if_setlladdr(ifv->ifv_ifp, IF_LLADDR(ifp), ETHER_ADDR_LEN);
			VLAN_LOCK();
		}
	VLAN_UNLOCK();

}

/*
 * A handler for network interface departure events.
 * Track departure of trunks here so that we don't access invalid
 * pointers or whatever if a trunk is ripped from under us, e.g.,
 * by ejecting its hot-plug card.  However, if an ifnet is simply
 * being renamed, then there's no need to tear down the state.
 */
static void
vlan_ifdetach(void *arg __unused, struct ifnet *ifp)
{
	struct ifvlan *ifv;
	int i;

	/*
	 * Check if it's a trunk interface first of all
	 * to avoid needless locking.
	 */
	if (ifp->if_vlantrunk == NULL)
		return;

	/* If the ifnet is just being renamed, don't do anything. */
	if (ifp->if_flags & IFF_RENAMING)
		return;

	VLAN_LOCK();
	/*
	 * OK, it's a trunk.  Loop over and detach all vlan's on it.
	 * Check trunk pointer after each vlan_unconfig() as it will
	 * free it and set to NULL after the last vlan was detached.
	 */
#ifdef VLAN_ARRAY
	for (i = 0; i < VLAN_ARRAY_SIZE; i++)
		if ((ifv = ifp->if_vlantrunk->vlans[i])) {
			vlan_unconfig_locked(ifv->ifv_ifp);
			if (ifp->if_vlantrunk == NULL)
				break;
		}
#else /* VLAN_ARRAY */
restart:
	for (i = 0; i < (1 << ifp->if_vlantrunk->hwidth); i++)
		if ((ifv = LIST_FIRST(&ifp->if_vlantrunk->hash[i]))) {
			vlan_unconfig_locked(ifv->ifv_ifp);
			if (ifp->if_vlantrunk)
				goto restart;	/* trunk->hwidth can change */
			else
				break;
		}
#endif /* VLAN_ARRAY */
	/* Trunk should have been destroyed in vlan_unconfig(). */
	KASSERT(ifp->if_vlantrunk == NULL, ("%s: purge failed", __func__));
	VLAN_UNLOCK();
}

/*
 * VLAN support can be loaded as a module.  The only place in the
 * system that's intimately aware of this is ether_input.  We hook
 * into this code through vlan_input_p which is defined there and
 * set here.  Noone else in the system should be aware of this so
 * we use an explicit reference here.
 */
extern	void (*vlan_input_p)(struct ifnet *, struct mbuf *);

/* For if_link_state_change() eyes only... */
extern	void (*vlan_link_state_p)(struct ifnet *);

static int
vlan_modevent(module_t mod, int type, void *data)
{

	switch (type) {
	case MOD_LOAD:
		ifdetach_tag = EVENTHANDLER_REGISTER(ifnet_departure_event,
		    vlan_ifdetach, NULL, EVENTHANDLER_PRI_ANY);
		if (ifdetach_tag == NULL)
			return (ENOMEM);
		iflladdr_tag = EVENTHANDLER_REGISTER(iflladdr_event,
		    vlan_iflladdr, NULL, EVENTHANDLER_PRI_ANY);
		if (iflladdr_tag == NULL)
			return (ENOMEM);
		VLAN_LOCK_INIT();
		vlan_input_p = vlan_input;
		vlan_link_state_p = vlan_link_state;
		vlan_trunk_cap_p = vlan_trunk_capabilities;
#ifndef VIMAGE
		if_clone_attach(&vlan_cloner);
#endif
		if (bootverbose)
			printf("vlan: initialized, using "
#ifdef VLAN_ARRAY
			       "full-size arrays"
#else
			       "hash tables with chaining"
#endif
			
			       "\n");
		break;
	case MOD_UNLOAD:
#ifndef VIMAGE
		if_clone_detach(&vlan_cloner);
#endif
		EVENTHANDLER_DEREGISTER(ifnet_departure_event, ifdetach_tag);
		EVENTHANDLER_DEREGISTER(iflladdr_event, iflladdr_tag);
		vlan_input_p = NULL;
		vlan_link_state_p = NULL;
		vlan_trunk_cap_p = NULL;
		VLAN_LOCK_DESTROY();
		if (bootverbose)
			printf("vlan: unloaded\n");
		break;
	default:
		return (EOPNOTSUPP);
	}
	return (0);
}

static moduledata_t vlan_mod = {
	"if_vlan",
	vlan_modevent,
	0
};

DECLARE_MODULE(if_vlan, vlan_mod, SI_SUB_PSEUDO, SI_ORDER_ANY);
MODULE_VERSION(if_vlan, 3);

#ifdef VIMAGE
static void
vnet_vlan_init(const void *unused __unused)
{

	V_vlan_cloner = vlan_cloner;
	if_clone_attach(&V_vlan_cloner);
}
VNET_SYSINIT(vnet_vlan_init, SI_SUB_PROTO_IFATTACHDOMAIN, SI_ORDER_ANY,
    vnet_vlan_init, NULL);

static void
vnet_vlan_uninit(const void *unused __unused)
{

	if_clone_detach(&V_vlan_cloner);
}
VNET_SYSUNINIT(vnet_vlan_uninit, SI_SUB_PROTO_IFATTACHDOMAIN, SI_ORDER_FIRST,
    vnet_vlan_uninit, NULL);
#endif

static struct ifnet *
vlan_clone_match_ethertag(struct if_clone *ifc, const char *name, int *tag)
{
	const char *cp;
	struct ifnet *ifp;
	int t;

	/* Check for <etherif>.<vlan> style interface names. */
	IFNET_RLOCK_NOSLEEP();
	TAILQ_FOREACH(ifp, &V_ifnet, if_link) {
		if (ifp->if_type != IFT_ETHER)
			continue;
		if (strncmp(ifp->if_xname, name, strlen(ifp->if_xname)) != 0)
			continue;
		cp = name + strlen(ifp->if_xname);
		if (*cp++ != '.')
			continue;
		if (*cp == '\0')
			continue;
		t = 0;
		for(; *cp >= '0' && *cp <= '9'; cp++)
			t = (t * 10) + (*cp - '0');
		if (*cp != '\0')
			continue;
		if (tag != NULL)
			*tag = t;
		break;
	}
	IFNET_RUNLOCK_NOSLEEP();

	return (ifp);
}

static int
vlan_clone_match(struct if_clone *ifc, const char *name)
{
	const char *cp;

	if (vlan_clone_match_ethertag(ifc, name, NULL) != NULL)
		return (1);

	if (strncmp(VLANNAME, name, strlen(VLANNAME)) != 0)
		return (0);
	for (cp = name + 4; *cp != '\0'; cp++) {
		if (*cp < '0' || *cp > '9')
			return (0);
	}

	return (1);
}

static int
vlan_clone_create(struct if_clone *ifc, char *name, size_t len, caddr_t params)
{
	char *dp;
	int wildcard;
	int unit;
	int error;
	int tag;
	int ethertag;
	struct ifvlan *ifv;
	struct ifnet *ifp;
	struct ifnet *p;
	struct ifaddr *ifa;
	struct sockaddr_dl *sdl;
	struct vlanreq vlr;
	static const u_char eaddr[ETHER_ADDR_LEN];	/* 00:00:00:00:00:00 */

	/*
	 * There are 3 (ugh) ways to specify the cloned device:
	 * o pass a parameter block with the clone request.
	 * o specify parameters in the text of the clone device name
	 * o specify no parameters and get an unattached device that
	 *   must be configured separately.
	 * The first technique is preferred; the latter two are
	 * supported for backwards compatibilty.
	 */
	if (params) {
		error = copyin(params, &vlr, sizeof(vlr));
		if (error)
			return error;
		p = ifunit(vlr.vlr_parent);
		if (p == NULL)
			return ENXIO;
		/*
		 * Don't let the caller set up a VLAN tag with
		 * anything except VLID bits.
		 */
		if (vlr.vlr_tag & ~EVL_VLID_MASK)
			return (EINVAL);
		error = ifc_name2unit(name, &unit);
		if (error != 0)
			return (error);

		ethertag = 1;
		tag = vlr.vlr_tag;
		wildcard = (unit < 0);
	} else if ((p = vlan_clone_match_ethertag(ifc, name, &tag)) != NULL) {
		ethertag = 1;
		unit = -1;
		wildcard = 0;

		/*
		 * Don't let the caller set up a VLAN tag with
		 * anything except VLID bits.
		 */
		if (tag & ~EVL_VLID_MASK)
			return (EINVAL);
	} else {
		ethertag = 0;

		error = ifc_name2unit(name, &unit);
		if (error != 0)
			return (error);

		wildcard = (unit < 0);
	}

	error = ifc_alloc_unit(ifc, &unit);
	if (error != 0)
		return (error);

	/* In the wildcard case, we need to update the name. */
	if (wildcard) {
		for (dp = name; *dp != '\0'; dp++);
		if (snprintf(dp, len - (dp-name), "%d", unit) >
		    len - (dp-name) - 1) {
			panic("%s: interface name too long", __func__);
		}
	}

	ifv = malloc(sizeof(struct ifvlan), M_VLAN, M_WAITOK | M_ZERO);
	ifp = ifv->ifv_ifp = if_alloc(IFT_ETHER);
	if (ifp == NULL) {
		ifc_free_unit(ifc, unit);
		free(ifv, M_VLAN);
		return (ENOSPC);
	}
	SLIST_INIT(&ifv->vlan_mc_listhead);

	ifp->if_softc = ifv;
	/*
	 * Set the name manually rather than using if_initname because
	 * we don't conform to the default naming convention for interfaces.
	 */
	strlcpy(ifp->if_xname, name, IFNAMSIZ);
	ifp->if_dname = ifc->ifc_name;
	ifp->if_dunit = unit;
	/* NB: flags are not set here */
	ifp->if_linkmib = &ifv->ifv_mib;
	ifp->if_linkmiblen = sizeof(ifv->ifv_mib);
	/* NB: mtu is not set here */

	ifp->if_init = vlan_init;
	ifp->if_start = vlan_start;
	ifp->if_ioctl = vlan_ioctl;
	ifp->if_snd.ifq_maxlen = ifqmaxlen;
	ifp->if_flags = VLAN_IFFLAGS;
	ether_ifattach(ifp, eaddr);
	/* Now undo some of the damage... */
	ifp->if_baudrate = 0;
	ifp->if_type = IFT_L2VLAN;
	ifp->if_hdrlen = ETHER_VLAN_ENCAP_LEN;
	ifa = ifp->if_addr;
	sdl = (struct sockaddr_dl *)ifa->ifa_addr;
	sdl->sdl_type = IFT_L2VLAN;

	if (ethertag) {
		error = vlan_config(ifv, p, tag);
		if (error != 0) {
			/*
			 * Since we've partialy failed, we need to back
			 * out all the way, otherwise userland could get
			 * confused.  Thus, we destroy the interface.
			 */
			ether_ifdetach(ifp);
			vlan_unconfig(ifp);
			if_free_type(ifp, IFT_ETHER);
			ifc_free_unit(ifc, unit);
			free(ifv, M_VLAN);

			return (error);
		}

		/* Update flags on the parent, if necessary. */
		vlan_setflags(ifp, 1);
	}

	return (0);
}

static int
vlan_clone_destroy(struct if_clone *ifc, struct ifnet *ifp)
{
	struct ifvlan *ifv = ifp->if_softc;
	int unit = ifp->if_dunit;

	ether_ifdetach(ifp);	/* first, remove it from system-wide lists */
	vlan_unconfig(ifp);	/* now it can be unconfigured and freed */
	if_free_type(ifp, IFT_ETHER);
	free(ifv, M_VLAN);
	ifc_free_unit(ifc, unit);

	return (0);
}

/*
 * The ifp->if_init entry point for vlan(4) is a no-op.
 */
static void
vlan_init(void *foo __unused)
{
}

/*
 * The if_start method for vlan(4) interface. It doesn't
 * raises the IFF_DRV_OACTIVE flag, since it is called
 * only from IFQ_HANDOFF() macro in ether_output_frame().
 * If the interface queue is full, and vlan_start() is
 * not called, the queue would never get emptied and
 * interface would stall forever.
 */
static void
vlan_start(struct ifnet *ifp)
{
	struct ifvlan *ifv;
	struct ifnet *p;
	struct mbuf *m;
	int error;

	ifv = ifp->if_softc;
	p = PARENT(ifv);

	for (;;) {
		IF_DEQUEUE(&ifp->if_snd, m);
		if (m == NULL)
			break;
		BPF_MTAP(ifp, m);

		/*
		 * Do not run parent's if_start() if the parent is not up,
		 * or parent's driver will cause a system crash.
		 */
		if (!UP_AND_RUNNING(p)) {
			m_freem(m);
			ifp->if_collisions++;
			continue;
		}

		/*
		 * Pad the frame to the minimum size allowed if told to.
		 * This option is in accord with IEEE Std 802.1Q, 2003 Ed.,
		 * paragraph C.4.4.3.b.  It can help to work around buggy
		 * bridges that violate paragraph C.4.4.3.a from the same
		 * document, i.e., fail to pad short frames after untagging.
		 * E.g., a tagged frame 66 bytes long (incl. FCS) is OK, but
		 * untagging it will produce a 62-byte frame, which is a runt
		 * and requires padding.  There are VLAN-enabled network
		 * devices that just discard such runts instead or mishandle
		 * them somehow.
		 */
		if (soft_pad) {
			static char pad[8];	/* just zeros */
			int n;

			for (n = ETHERMIN + ETHER_HDR_LEN - m->m_pkthdr.len;
			     n > 0; n -= sizeof(pad))
				if (!m_append(m, min(n, sizeof(pad)), pad))
					break;

			if (n > 0) {
				if_printf(ifp, "cannot pad short frame\n");
				ifp->if_oerrors++;
				m_freem(m);
				continue;
			}
		}

		/*
		 * If underlying interface can do VLAN tag insertion itself,
		 * just pass the packet along. However, we need some way to
		 * tell the interface where the packet came from so that it
		 * knows how to find the VLAN tag to use, so we attach a
		 * packet tag that holds it.
		 */
		if (p->if_capenable & IFCAP_VLAN_HWTAGGING) {
			m->m_pkthdr.ether_vtag = ifv->ifv_tag;
			m->m_flags |= M_VLANTAG;
		} else {
			m = ether_vlanencap(m, ifv->ifv_tag);
			if (m == NULL) {
				if_printf(ifp,
				    "unable to prepend VLAN header\n");
				ifp->if_oerrors++;
				continue;
			}
		}

		/*
		 * Send it, precisely as ether_output() would have.
		 * We are already running at splimp.
		 */
		error = (p->if_transmit)(p, m);
		if (!error)
			ifp->if_opackets++;
		else
			ifp->if_oerrors++;
	}
}

static void
vlan_input(struct ifnet *ifp, struct mbuf *m)
{
	struct ifvlantrunk *trunk = ifp->if_vlantrunk;
	struct ifvlan *ifv;
	uint16_t tag;

	KASSERT(trunk != NULL, ("%s: no trunk", __func__));

	if (m->m_flags & M_VLANTAG) {
		/*
		 * Packet is tagged, but m contains a normal
		 * Ethernet frame; the tag is stored out-of-band.
		 */
		tag = EVL_VLANOFTAG(m->m_pkthdr.ether_vtag);
		m->m_flags &= ~M_VLANTAG;
	} else {
		struct ether_vlan_header *evl;

		/*
		 * Packet is tagged in-band as specified by 802.1q.
		 */
		switch (ifp->if_type) {
		case IFT_ETHER:
			if (m->m_len < sizeof(*evl) &&
			    (m = m_pullup(m, sizeof(*evl))) == NULL) {
				if_printf(ifp, "cannot pullup VLAN header\n");
				return;
			}
			evl = mtod(m, struct ether_vlan_header *);
			tag = EVL_VLANOFTAG(ntohs(evl->evl_tag));

			/*
			 * Remove the 802.1q header by copying the Ethernet
			 * addresses over it and adjusting the beginning of
			 * the data in the mbuf.  The encapsulated Ethernet
			 * type field is already in place.
			 */
			bcopy((char *)evl, (char *)evl + ETHER_VLAN_ENCAP_LEN,
			      ETHER_HDR_LEN - ETHER_TYPE_LEN);
			m_adj(m, ETHER_VLAN_ENCAP_LEN);
			break;

		default:
#ifdef INVARIANTS
			panic("%s: %s has unsupported if_type %u",
			      __func__, ifp->if_xname, ifp->if_type);
#endif
			m_freem(m);
			ifp->if_noproto++;
			return;
		}
	}

	TRUNK_RLOCK(trunk);
#ifdef VLAN_ARRAY
	ifv = trunk->vlans[tag];
#else
	ifv = vlan_gethash(trunk, tag);
#endif
	if (ifv == NULL || !UP_AND_RUNNING(ifv->ifv_ifp)) {
		TRUNK_RUNLOCK(trunk);
		m_freem(m);
		ifp->if_noproto++;
		return;
	}
	TRUNK_RUNLOCK(trunk);

	m->m_pkthdr.rcvif = ifv->ifv_ifp;
	ifv->ifv_ifp->if_ipackets++;

	/* Pass it back through the parent's input routine. */
	(*ifp->if_input)(ifv->ifv_ifp, m);
}

static int
vlan_config(struct ifvlan *ifv, struct ifnet *p, uint16_t tag)
{
	struct ifvlantrunk *trunk;
	struct ifnet *ifp;
	int error = 0;

	/* VID numbers 0x0 and 0xFFF are reserved */
	if (tag == 0 || tag == 0xFFF)
		return (EINVAL);
	if (p->if_type != IFT_ETHER)
		return (EPROTONOSUPPORT);
	if ((p->if_flags & VLAN_IFFLAGS) != VLAN_IFFLAGS)
		return (EPROTONOSUPPORT);
	if (ifv->ifv_trunk)
		return (EBUSY);

	if (p->if_vlantrunk == NULL) {
		trunk = malloc(sizeof(struct ifvlantrunk),
		    M_VLAN, M_WAITOK | M_ZERO);
#ifndef VLAN_ARRAY
		vlan_inithash(trunk);
#endif
		VLAN_LOCK();
		if (p->if_vlantrunk != NULL) {
			/* A race that that is very unlikely to be hit. */
#ifndef VLAN_ARRAY
			vlan_freehash(trunk);
#endif
			free(trunk, M_VLAN);
			goto exists;
		}
		TRUNK_LOCK_INIT(trunk);
		TRUNK_LOCK(trunk);
		p->if_vlantrunk = trunk;
		trunk->parent = p;
	} else {
		VLAN_LOCK();
exists:
		trunk = p->if_vlantrunk;
		TRUNK_LOCK(trunk);
	}

	ifv->ifv_tag = tag;	/* must set this before vlan_inshash() */
#ifdef VLAN_ARRAY
	if (trunk->vlans[tag] != NULL) {
		error = EEXIST;
		goto done;
	}
	trunk->vlans[tag] = ifv;
	trunk->refcnt++;
#else
	error = vlan_inshash(trunk, ifv);
	if (error)
		goto done;
#endif
	ifv->ifv_proto = ETHERTYPE_VLAN;
	ifv->ifv_encaplen = ETHER_VLAN_ENCAP_LEN;
	ifv->ifv_mintu = ETHERMIN;
	ifv->ifv_pflags = 0;

	/*
	 * If the parent supports the VLAN_MTU capability,
	 * i.e. can Tx/Rx larger than ETHER_MAX_LEN frames,
	 * use it.
	 */
	if (p->if_capenable & IFCAP_VLAN_MTU) {
		/*
		 * No need to fudge the MTU since the parent can
		 * handle extended frames.
		 */
		ifv->ifv_mtufudge = 0;
	} else {
		/*
		 * Fudge the MTU by the encapsulation size.  This
		 * makes us incompatible with strictly compliant
		 * 802.1Q implementations, but allows us to use
		 * the feature with other NetBSD implementations,
		 * which might still be useful.
		 */
		ifv->ifv_mtufudge = ifv->ifv_encaplen;
	}

	ifv->ifv_trunk = trunk;
	ifp = ifv->ifv_ifp;
	ifp->if_mtu = p->if_mtu - ifv->ifv_mtufudge;
	ifp->if_baudrate = p->if_baudrate;
	/*
	 * Copy only a selected subset of flags from the parent.
	 * Other flags are none of our business.
	 */
#define VLAN_COPY_FLAGS (IFF_SIMPLEX)
	ifp->if_flags &= ~VLAN_COPY_FLAGS;
	ifp->if_flags |= p->if_flags & VLAN_COPY_FLAGS;
#undef VLAN_COPY_FLAGS

	ifp->if_link_state = p->if_link_state;

	vlan_capabilities(ifv);

	/*
	 * Set up our ``Ethernet address'' to reflect the underlying
	 * physical interface's.
	 */
	bcopy(IF_LLADDR(p), IF_LLADDR(ifp), ETHER_ADDR_LEN);

	/*
	 * Configure multicast addresses that may already be
	 * joined on the vlan device.
	 */
	(void)vlan_setmulti(ifp); /* XXX: VLAN lock held */

	/* We are ready for operation now. */
	ifp->if_drv_flags |= IFF_DRV_RUNNING;
done:
	TRUNK_UNLOCK(trunk);
	if (error == 0)
		EVENTHANDLER_INVOKE(vlan_config, p, ifv->ifv_tag);
	VLAN_UNLOCK();

	return (error);
}

static void
vlan_unconfig(struct ifnet *ifp)
{

	VLAN_LOCK();
	vlan_unconfig_locked(ifp);
	VLAN_UNLOCK();
}

static void
vlan_unconfig_locked(struct ifnet *ifp)
{
	struct ifvlantrunk *trunk;
	struct vlan_mc_entry *mc;
	struct ifvlan *ifv;
	struct ifnet  *parent;

	VLAN_LOCK_ASSERT();

	ifv = ifp->if_softc;
	trunk = ifv->ifv_trunk;
	parent = NULL;

	if (trunk != NULL) {
		struct sockaddr_dl sdl;

		TRUNK_LOCK(trunk);
		parent = trunk->parent;

		/*
		 * Since the interface is being unconfigured, we need to
		 * empty the list of multicast groups that we may have joined
		 * while we were alive from the parent's list.
		 */
		bzero((char *)&sdl, sizeof(sdl));
		sdl.sdl_len = sizeof(sdl);
		sdl.sdl_family = AF_LINK;
		sdl.sdl_index = parent->if_index;
		sdl.sdl_type = IFT_ETHER;
		sdl.sdl_alen = ETHER_ADDR_LEN;

		while ((mc = SLIST_FIRST(&ifv->vlan_mc_listhead)) != NULL) {
			bcopy((char *)&mc->mc_addr, LLADDR(&sdl),
			    ETHER_ADDR_LEN);

			/*
			 * This may fail if the parent interface is
			 * being detached.  Regardless, we should do a
			 * best effort to free this interface as much
			 * as possible as all callers expect vlan
			 * destruction to succeed.
			 */
			(void)if_delmulti(parent, (struct sockaddr *)&sdl);
			SLIST_REMOVE_HEAD(&ifv->vlan_mc_listhead, mc_entries);
			free(mc, M_VLAN);
		}

		vlan_setflags(ifp, 0); /* clear special flags on parent */
#ifdef VLAN_ARRAY
		trunk->vlans[ifv->ifv_tag] = NULL;
		trunk->refcnt--;
#else
		vlan_remhash(trunk, ifv);
#endif
		ifv->ifv_trunk = NULL;

		/*
		 * Check if we were the last.
		 */
		if (trunk->refcnt == 0) {
			trunk->parent->if_vlantrunk = NULL;
			/*
			 * XXXGL: If some ithread has already entered
			 * vlan_input() and is now blocked on the trunk
			 * lock, then it should preempt us right after
			 * unlock and finish its work. Then we will acquire
			 * lock again in trunk_destroy().
			 */
			TRUNK_UNLOCK(trunk);
			trunk_destroy(trunk);
		} else
			TRUNK_UNLOCK(trunk);
	}

	/* Disconnect from parent. */
	if (ifv->ifv_pflags)
		if_printf(ifp, "%s: ifv_pflags unclean\n", __func__);
	ifp->if_mtu = ETHERMTU;
	ifp->if_link_state = LINK_STATE_UNKNOWN;
	ifp->if_drv_flags &= ~IFF_DRV_RUNNING;

	/*
	 * Only dispatch an event if vlan was
	 * attached, otherwise there is nothing
	 * to cleanup anyway.
	 */
	if (parent != NULL)
		EVENTHANDLER_INVOKE(vlan_unconfig, parent, ifv->ifv_tag);
}

/* Handle a reference counted flag that should be set on the parent as well */
static int
vlan_setflag(struct ifnet *ifp, int flag, int status,
	     int (*func)(struct ifnet *, int))
{
	struct ifvlan *ifv;
	int error;

	/* XXX VLAN_LOCK_ASSERT(); */

	ifv = ifp->if_softc;
	status = status ? (ifp->if_flags & flag) : 0;
	/* Now "status" contains the flag value or 0 */

	/*
	 * See if recorded parent's status is different from what
	 * we want it to be.  If it is, flip it.  We record parent's
	 * status in ifv_pflags so that we won't clear parent's flag
	 * we haven't set.  In fact, we don't clear or set parent's
	 * flags directly, but get or release references to them.
	 * That's why we can be sure that recorded flags still are
	 * in accord with actual parent's flags.
	 */
	if (status != (ifv->ifv_pflags & flag)) {
		error = (*func)(PARENT(ifv), status);
		if (error)
			return (error);
		ifv->ifv_pflags &= ~flag;
		ifv->ifv_pflags |= status;
	}
	return (0);
}

/*
 * Handle IFF_* flags that require certain changes on the parent:
 * if "status" is true, update parent's flags respective to our if_flags;
 * if "status" is false, forcedly clear the flags set on parent.
 */
static int
vlan_setflags(struct ifnet *ifp, int status)
{
	int error, i;
	
	for (i = 0; vlan_pflags[i].flag; i++) {
		error = vlan_setflag(ifp, vlan_pflags[i].flag,
				     status, vlan_pflags[i].func);
		if (error)
			return (error);
	}
	return (0);
}

/* Inform all vlans that their parent has changed link state */
static void
vlan_link_state(struct ifnet *ifp)
{
	struct ifvlantrunk *trunk = ifp->if_vlantrunk;
	struct ifvlan *ifv;
	int i;

	TRUNK_LOCK(trunk);
#ifdef VLAN_ARRAY
	for (i = 0; i < VLAN_ARRAY_SIZE; i++)
		if (trunk->vlans[i] != NULL) {
			ifv = trunk->vlans[i];
#else
	for (i = 0; i < (1 << trunk->hwidth); i++)
		LIST_FOREACH(ifv, &trunk->hash[i], ifv_list) {
#endif
			ifv->ifv_ifp->if_baudrate = trunk->parent->if_baudrate;
			if_link_state_change(ifv->ifv_ifp,
			    trunk->parent->if_link_state);
		}
	TRUNK_UNLOCK(trunk);
}

static void
vlan_capabilities(struct ifvlan *ifv)
{
	struct ifnet *p = PARENT(ifv);
	struct ifnet *ifp = ifv->ifv_ifp;

	TRUNK_LOCK_ASSERT(TRUNK(ifv));

	/*
	 * If the parent interface can do checksum offloading
	 * on VLANs, then propagate its hardware-assisted
	 * checksumming flags. Also assert that checksum
	 * offloading requires hardware VLAN tagging.
	 */
	if (p->if_capabilities & IFCAP_VLAN_HWCSUM)
		ifp->if_capabilities = p->if_capabilities & IFCAP_HWCSUM;

	if (p->if_capenable & IFCAP_VLAN_HWCSUM &&
	    p->if_capenable & IFCAP_VLAN_HWTAGGING) {
		ifp->if_capenable = p->if_capenable & IFCAP_HWCSUM;
		ifp->if_hwassist = p->if_hwassist & (CSUM_IP | CSUM_TCP |
		    CSUM_UDP | CSUM_SCTP | CSUM_IP_FRAGS | CSUM_FRAGMENT);
	} else {
		ifp->if_capenable = 0;
		ifp->if_hwassist = 0;
	}
	/*
	 * If the parent interface can do TSO on VLANs then
	 * propagate the hardware-assisted flag. TSO on VLANs
	 * does not necessarily require hardware VLAN tagging.
	 */
	if (p->if_capabilities & IFCAP_VLAN_HWTSO)
		ifp->if_capabilities |= p->if_capabilities & IFCAP_TSO;
	if (p->if_capenable & IFCAP_VLAN_HWTSO) {
		ifp->if_capenable |= p->if_capenable & IFCAP_TSO;
		ifp->if_hwassist |= p->if_hwassist & CSUM_TSO;
	} else {
		ifp->if_capenable &= ~(p->if_capenable & IFCAP_TSO);
		ifp->if_hwassist &= ~(p->if_hwassist & CSUM_TSO);
	}
}

static void
vlan_trunk_capabilities(struct ifnet *ifp)
{
	struct ifvlantrunk *trunk = ifp->if_vlantrunk;
	struct ifvlan *ifv;
	int i;

	TRUNK_LOCK(trunk);
#ifdef VLAN_ARRAY
	for (i = 0; i < VLAN_ARRAY_SIZE; i++)
		if (trunk->vlans[i] != NULL) {
			ifv = trunk->vlans[i];
#else
	for (i = 0; i < (1 << trunk->hwidth); i++) {
		LIST_FOREACH(ifv, &trunk->hash[i], ifv_list)
#endif
			vlan_capabilities(ifv);
	}
	TRUNK_UNLOCK(trunk);
}

static int
vlan_ioctl(struct ifnet *ifp, u_long cmd, caddr_t data)
{
	struct ifnet *p;
	struct ifreq *ifr;
	struct ifvlan *ifv;
	struct vlanreq vlr;
	int error = 0;

	ifr = (struct ifreq *)data;
	ifv = ifp->if_softc;

	switch (cmd) {
	case SIOCGIFMEDIA:
		VLAN_LOCK();
		if (TRUNK(ifv) != NULL) {
			p = PARENT(ifv);
			VLAN_UNLOCK();
			error = (*p->if_ioctl)(p, SIOCGIFMEDIA, data);
			/* Limit the result to the parent's current config. */
			if (error == 0) {
				struct ifmediareq *ifmr;

				ifmr = (struct ifmediareq *)data;
				if (ifmr->ifm_count >= 1 && ifmr->ifm_ulist) {
					ifmr->ifm_count = 1;
					error = copyout(&ifmr->ifm_current,
						ifmr->ifm_ulist,
						sizeof(int));
				}
			}
		} else {
			VLAN_UNLOCK();
			error = EINVAL;
		}
		break;

	case SIOCSIFMEDIA:
		error = EINVAL;
		break;

	case SIOCSIFMTU:
		/*
		 * Set the interface MTU.
		 */
		VLAN_LOCK();
		if (TRUNK(ifv) != NULL) {
			if (ifr->ifr_mtu >
			     (PARENT(ifv)->if_mtu - ifv->ifv_mtufudge) ||
			    ifr->ifr_mtu <
			     (ifv->ifv_mintu - ifv->ifv_mtufudge))
				error = EINVAL;
			else
				ifp->if_mtu = ifr->ifr_mtu;
		} else
			error = EINVAL;
		VLAN_UNLOCK();
		break;

	case SIOCSETVLAN:
#ifdef VIMAGE
		if (ifp->if_vnet != ifp->if_home_vnet) {
			error = EPERM;
			break;
		}
#endif
		error = copyin(ifr->ifr_data, &vlr, sizeof(vlr));
		if (error)
			break;
		if (vlr.vlr_parent[0] == '\0') {
			vlan_unconfig(ifp);
			break;
		}
		p = ifunit(vlr.vlr_parent);
		if (p == NULL) {
			error = ENOENT;
			break;
		}
		/*
		 * Don't let the caller set up a VLAN tag with
		 * anything except VLID bits.
		 */
		if (vlr.vlr_tag & ~EVL_VLID_MASK) {
			error = EINVAL;
			break;
		}
		error = vlan_config(ifv, p, vlr.vlr_tag);
		if (error)
			break;

		/* Update flags on the parent, if necessary. */
		vlan_setflags(ifp, 1);
		break;

	case SIOCGETVLAN:
#ifdef VIMAGE
		if (ifp->if_vnet != ifp->if_home_vnet) {
			error = EPERM;
			break;
		}
#endif
		bzero(&vlr, sizeof(vlr));
		VLAN_LOCK();
		if (TRUNK(ifv) != NULL) {
			strlcpy(vlr.vlr_parent, PARENT(ifv)->if_xname,
			    sizeof(vlr.vlr_parent));
			vlr.vlr_tag = ifv->ifv_tag;
		}
		VLAN_UNLOCK();
		error = copyout(&vlr, ifr->ifr_data, sizeof(vlr));
		break;
		
	case SIOCSIFFLAGS:
		/*
		 * We should propagate selected flags to the parent,
		 * e.g., promiscuous mode.
		 */
		if (TRUNK(ifv) != NULL)
			error = vlan_setflags(ifp, 1);
		break;

	case SIOCADDMULTI:
	case SIOCDELMULTI:
		/*
		 * If we don't have a parent, just remember the membership for
		 * when we do.
		 */
		if (TRUNK(ifv) != NULL)
			error = vlan_setmulti(ifp);
		break;

	default:
		error = ether_ioctl(ifp, cmd, data);
	}

	return (error);
}
OpenPOWER on IntegriCloud