Home | History | Annotate | Line # | Download | only in net
if_vlan.c revision 1.130.2.3
      1 /*	$NetBSD: if_vlan.c,v 1.130.2.3 2020/04/13 08:05:15 martin Exp $	*/
      2 
      3 /*
      4  * Copyright (c) 2000, 2001 The NetBSD Foundation, Inc.
      5  * All rights reserved.
      6  *
      7  * This code is derived from software contributed to The NetBSD Foundation
      8  * by Andrew Doran, and by Jason R. Thorpe of Zembu Labs, Inc.
      9  *
     10  * Redistribution and use in source and binary forms, with or without
     11  * modification, are permitted provided that the following conditions
     12  * are met:
     13  * 1. Redistributions of source code must retain the above copyright
     14  *    notice, this list of conditions and the following disclaimer.
     15  * 2. Redistributions in binary form must reproduce the above copyright
     16  *    notice, this list of conditions and the following disclaimer in the
     17  *    documentation and/or other materials provided with the distribution.
     18  *
     19  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
     20  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
     21  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
     22  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
     23  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
     24  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
     25  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
     26  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
     27  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
     28  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
     29  * POSSIBILITY OF SUCH DAMAGE.
     30  */
     31 
     32 /*
     33  * Copyright 1998 Massachusetts Institute of Technology
     34  *
     35  * Permission to use, copy, modify, and distribute this software and
     36  * its documentation for any purpose and without fee is hereby
     37  * granted, provided that both the above copyright notice and this
     38  * permission notice appear in all copies, that both the above
     39  * copyright notice and this permission notice appear in all
     40  * supporting documentation, and that the name of M.I.T. not be used
     41  * in advertising or publicity pertaining to distribution of the
     42  * software without specific, written prior permission.  M.I.T. makes
     43  * no representations about the suitability of this software for any
     44  * purpose.  It is provided "as is" without express or implied
     45  * warranty.
     46  *
     47  * THIS SOFTWARE IS PROVIDED BY M.I.T. ``AS IS''.  M.I.T. DISCLAIMS
     48  * ALL EXPRESS OR IMPLIED WARRANTIES WITH REGARD TO THIS SOFTWARE,
     49  * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
     50  * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT
     51  * SHALL M.I.T. BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
     52  * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
     53  * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
     54  * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
     55  * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
     56  * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
     57  * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     58  * SUCH DAMAGE.
     59  *
     60  * from FreeBSD: if_vlan.c,v 1.16 2000/03/26 15:21:40 charnier Exp
     61  * via OpenBSD: if_vlan.c,v 1.4 2000/05/15 19:15:00 chris Exp
     62  */
     63 
     64 /*
     65  * if_vlan.c - pseudo-device driver for IEEE 802.1Q virtual LANs.  Might be
     66  * extended some day to also handle IEEE 802.1P priority tagging.  This is
     67  * sort of sneaky in the implementation, since we need to pretend to be
     68  * enough of an Ethernet implementation to make ARP work.  The way we do
     69  * this is by telling everyone that we are an Ethernet interface, and then
     70  * catch the packets that ether_output() left on our output queue when it
     71  * calls if_start(), rewrite them for use by the real outgoing interface,
     72  * and ask it to send them.
     73  *
     74  * TODO:
     75  *
     76  *	- Need some way to notify vlan interfaces when the parent
     77  *	  interface changes MTU.
     78  */
     79 
     80 #include <sys/cdefs.h>
     81 __KERNEL_RCSID(0, "$NetBSD: if_vlan.c,v 1.130.2.3 2020/04/13 08:05:15 martin Exp $");
     82 
     83 #ifdef _KERNEL_OPT
     84 #include "opt_inet.h"
     85 #include "opt_net_mpsafe.h"
     86 #endif
     87 
     88 #include <sys/param.h>
     89 #include <sys/systm.h>
     90 #include <sys/kernel.h>
     91 #include <sys/mbuf.h>
     92 #include <sys/queue.h>
     93 #include <sys/socket.h>
     94 #include <sys/sockio.h>
     95 #include <sys/systm.h>
     96 #include <sys/proc.h>
     97 #include <sys/kauth.h>
     98 #include <sys/mutex.h>
     99 #include <sys/kmem.h>
    100 #include <sys/cpu.h>
    101 #include <sys/pserialize.h>
    102 #include <sys/psref.h>
    103 #include <sys/pslist.h>
    104 #include <sys/atomic.h>
    105 #include <sys/device.h>
    106 #include <sys/module.h>
    107 
    108 #include <net/bpf.h>
    109 #include <net/if.h>
    110 #include <net/if_dl.h>
    111 #include <net/if_types.h>
    112 #include <net/if_ether.h>
    113 #include <net/if_vlanvar.h>
    114 
    115 #ifdef INET
    116 #include <netinet/in.h>
    117 #include <netinet/if_inarp.h>
    118 #endif
    119 #ifdef INET6
    120 #include <netinet6/in6_ifattach.h>
    121 #include <netinet6/in6_var.h>
    122 #include <netinet6/nd6.h>
    123 #endif
    124 
    125 #include "ioconf.h"
    126 
    127 struct vlan_mc_entry {
    128 	LIST_ENTRY(vlan_mc_entry)	mc_entries;
    129 	/*
    130 	 * A key to identify this entry.  The mc_addr below can't be
    131 	 * used since multiple sockaddr may mapped into the same
    132 	 * ether_multi (e.g., AF_UNSPEC).
    133 	 */
    134 	struct ether_multi	*mc_enm;
    135 	struct sockaddr_storage		mc_addr;
    136 };
    137 
    138 struct ifvlan_linkmib {
    139 	struct ifvlan *ifvm_ifvlan;
    140 	const struct vlan_multisw *ifvm_msw;
    141 	int	ifvm_encaplen;	/* encapsulation length */
    142 	int	ifvm_mtufudge;	/* MTU fudged by this much */
    143 	int	ifvm_mintu;	/* min transmission unit */
    144 	uint16_t ifvm_proto;	/* encapsulation ethertype */
    145 	uint16_t ifvm_tag;	/* tag to apply on packets */
    146 	struct ifnet *ifvm_p;	/* parent interface of this vlan */
    147 
    148 	struct psref_target ifvm_psref;
    149 };
    150 
    151 struct ifvlan {
    152 	struct ethercom ifv_ec;
    153 	struct ifvlan_linkmib *ifv_mib;	/*
    154 					 * reader must use vlan_getref_linkmib()
    155 					 * instead of direct dereference
    156 					 */
    157 	kmutex_t ifv_lock;		/* writer lock for ifv_mib */
    158 	pserialize_t ifv_psz;
    159 
    160 	LIST_HEAD(__vlan_mchead, vlan_mc_entry) ifv_mc_listhead;
    161 	LIST_ENTRY(ifvlan) ifv_list;
    162 	struct pslist_entry ifv_hash;
    163 	int ifv_flags;
    164 };
    165 
    166 #define	IFVF_PROMISC	0x01		/* promiscuous mode enabled */
    167 
    168 #define	ifv_if		ifv_ec.ec_if
    169 
    170 #define	ifv_msw		ifv_mib.ifvm_msw
    171 #define	ifv_encaplen	ifv_mib.ifvm_encaplen
    172 #define	ifv_mtufudge	ifv_mib.ifvm_mtufudge
    173 #define	ifv_mintu	ifv_mib.ifvm_mintu
    174 #define	ifv_tag		ifv_mib.ifvm_tag
    175 
    176 struct vlan_multisw {
    177 	int	(*vmsw_addmulti)(struct ifvlan *, struct ifreq *);
    178 	int	(*vmsw_delmulti)(struct ifvlan *, struct ifreq *);
    179 	void	(*vmsw_purgemulti)(struct ifvlan *);
    180 };
    181 
    182 static int	vlan_ether_addmulti(struct ifvlan *, struct ifreq *);
    183 static int	vlan_ether_delmulti(struct ifvlan *, struct ifreq *);
    184 static void	vlan_ether_purgemulti(struct ifvlan *);
    185 
    186 const struct vlan_multisw vlan_ether_multisw = {
    187 	.vmsw_addmulti = vlan_ether_addmulti,
    188 	.vmsw_delmulti = vlan_ether_delmulti,
    189 	.vmsw_purgemulti = vlan_ether_purgemulti,
    190 };
    191 
    192 static int	vlan_clone_create(struct if_clone *, int);
    193 static int	vlan_clone_destroy(struct ifnet *);
    194 static int	vlan_config(struct ifvlan *, struct ifnet *, uint16_t);
    195 static int	vlan_ioctl(struct ifnet *, u_long, void *);
    196 static void	vlan_start(struct ifnet *);
    197 static int	vlan_transmit(struct ifnet *, struct mbuf *);
    198 static void	vlan_unconfig(struct ifnet *);
    199 static int	vlan_unconfig_locked(struct ifvlan *, struct ifvlan_linkmib *);
    200 static void	vlan_hash_init(void);
    201 static int	vlan_hash_fini(void);
    202 static int	vlan_tag_hash(uint16_t, u_long);
    203 static struct ifvlan_linkmib*	vlan_getref_linkmib(struct ifvlan *,
    204     struct psref *);
    205 static void	vlan_putref_linkmib(struct ifvlan_linkmib *, struct psref *);
    206 static void	vlan_linkmib_update(struct ifvlan *, struct ifvlan_linkmib *);
    207 static struct ifvlan_linkmib*	vlan_lookup_tag_psref(struct ifnet *,
    208     uint16_t, struct psref *);
    209 
    210 static struct {
    211 	kmutex_t lock;
    212 	LIST_HEAD(vlan_ifvlist, ifvlan) list;
    213 } ifv_list __cacheline_aligned;
    214 
    215 
    216 #if !defined(VLAN_TAG_HASH_SIZE)
    217 #define VLAN_TAG_HASH_SIZE 32
    218 #endif
    219 static struct {
    220 	kmutex_t lock;
    221 	struct pslist_head *lists;
    222 	u_long mask;
    223 } ifv_hash __cacheline_aligned = {
    224 	.lists = NULL,
    225 	.mask = 0,
    226 };
    227 
    228 pserialize_t vlan_psz __read_mostly;
    229 static struct psref_class *ifvm_psref_class __read_mostly;
    230 
    231 struct if_clone vlan_cloner =
    232     IF_CLONE_INITIALIZER("vlan", vlan_clone_create, vlan_clone_destroy);
    233 
    234 /* Used to pad ethernet frames with < ETHER_MIN_LEN bytes */
    235 static char vlan_zero_pad_buff[ETHER_MIN_LEN];
    236 
    237 static inline int
    238 vlan_safe_ifpromisc(struct ifnet *ifp, int pswitch)
    239 {
    240 	int e;
    241 
    242 	KERNEL_LOCK_UNLESS_NET_MPSAFE();
    243 	e = ifpromisc(ifp, pswitch);
    244 	KERNEL_UNLOCK_UNLESS_NET_MPSAFE();
    245 
    246 	return e;
    247 }
    248 
    249 static inline int
    250 vlan_safe_ifpromisc_locked(struct ifnet *ifp, int pswitch)
    251 {
    252 	int e;
    253 
    254 	KERNEL_LOCK_UNLESS_NET_MPSAFE();
    255 	e = ifpromisc_locked(ifp, pswitch);
    256 	KERNEL_UNLOCK_UNLESS_NET_MPSAFE();
    257 
    258 	return e;
    259 }
    260 
    261 void
    262 vlanattach(int n)
    263 {
    264 
    265 	/*
    266 	 * Nothing to do here, initialization is handled by the
    267 	 * module initialization code in vlaninit() below.
    268 	 */
    269 }
    270 
    271 static void
    272 vlaninit(void)
    273 {
    274 	mutex_init(&ifv_list.lock, MUTEX_DEFAULT, IPL_NONE);
    275 	LIST_INIT(&ifv_list.list);
    276 
    277 	mutex_init(&ifv_hash.lock, MUTEX_DEFAULT, IPL_NONE);
    278 	vlan_psz = pserialize_create();
    279 	ifvm_psref_class = psref_class_create("vlanlinkmib", IPL_SOFTNET);
    280 	if_clone_attach(&vlan_cloner);
    281 
    282 	vlan_hash_init();
    283 	MODULE_HOOK_SET(if_vlan_vlan_input_hook, vlan_input);
    284 }
    285 
    286 static int
    287 vlandetach(void)
    288 {
    289 	bool is_empty;
    290 	int error;
    291 
    292 	mutex_enter(&ifv_list.lock);
    293 	is_empty = LIST_EMPTY(&ifv_list.list);
    294 	mutex_exit(&ifv_list.lock);
    295 
    296 	if (!is_empty)
    297 		return EBUSY;
    298 
    299 	error = vlan_hash_fini();
    300 	if (error != 0)
    301 		return error;
    302 
    303 	if_clone_detach(&vlan_cloner);
    304 	psref_class_destroy(ifvm_psref_class);
    305 	pserialize_destroy(vlan_psz);
    306 	mutex_destroy(&ifv_hash.lock);
    307 	mutex_destroy(&ifv_list.lock);
    308 
    309 	MODULE_HOOK_UNSET(if_vlan_vlan_input_hook);
    310 	return 0;
    311 }
    312 
    313 static void
    314 vlan_reset_linkname(struct ifnet *ifp)
    315 {
    316 
    317 	/*
    318 	 * We start out with a "802.1Q VLAN" type and zero-length
    319 	 * addresses.  When we attach to a parent interface, we
    320 	 * inherit its type, address length, address, and data link
    321 	 * type.
    322 	 */
    323 
    324 	ifp->if_type = IFT_L2VLAN;
    325 	ifp->if_addrlen = 0;
    326 	ifp->if_dlt = DLT_NULL;
    327 	if_alloc_sadl(ifp);
    328 }
    329 
    330 static int
    331 vlan_clone_create(struct if_clone *ifc, int unit)
    332 {
    333 	struct ifvlan *ifv;
    334 	struct ifnet *ifp;
    335 	struct ifvlan_linkmib *mib;
    336 	int rv;
    337 
    338 	ifv = malloc(sizeof(struct ifvlan), M_DEVBUF, M_WAITOK | M_ZERO);
    339 	mib = kmem_zalloc(sizeof(struct ifvlan_linkmib), KM_SLEEP);
    340 	ifp = &ifv->ifv_if;
    341 	LIST_INIT(&ifv->ifv_mc_listhead);
    342 
    343 	mib->ifvm_ifvlan = ifv;
    344 	mib->ifvm_p = NULL;
    345 	psref_target_init(&mib->ifvm_psref, ifvm_psref_class);
    346 
    347 	mutex_init(&ifv->ifv_lock, MUTEX_DEFAULT, IPL_NONE);
    348 	ifv->ifv_psz = pserialize_create();
    349 	ifv->ifv_mib = mib;
    350 
    351 	mutex_enter(&ifv_list.lock);
    352 	LIST_INSERT_HEAD(&ifv_list.list, ifv, ifv_list);
    353 	mutex_exit(&ifv_list.lock);
    354 
    355 	if_initname(ifp, ifc->ifc_name, unit);
    356 	ifp->if_softc = ifv;
    357 	ifp->if_flags = IFF_BROADCAST | IFF_SIMPLEX | IFF_MULTICAST;
    358 	ifp->if_extflags = IFEF_NO_LINK_STATE_CHANGE;
    359 #ifdef NET_MPSAFE
    360 	ifp->if_extflags |= IFEF_MPSAFE;
    361 #endif
    362 	ifp->if_start = vlan_start;
    363 	ifp->if_transmit = vlan_transmit;
    364 	ifp->if_ioctl = vlan_ioctl;
    365 	IFQ_SET_READY(&ifp->if_snd);
    366 
    367 	rv = if_initialize(ifp);
    368 	if (rv != 0) {
    369 		aprint_error("%s: if_initialize failed(%d)\n", ifp->if_xname,
    370 		    rv);
    371 		goto fail;
    372 	}
    373 
    374 	vlan_reset_linkname(ifp);
    375 	if_register(ifp);
    376 	return 0;
    377 
    378 fail:
    379 	mutex_enter(&ifv_list.lock);
    380 	LIST_REMOVE(ifv, ifv_list);
    381 	mutex_exit(&ifv_list.lock);
    382 
    383 	mutex_destroy(&ifv->ifv_lock);
    384 	psref_target_destroy(&ifv->ifv_mib->ifvm_psref, ifvm_psref_class);
    385 	kmem_free(ifv->ifv_mib, sizeof(struct ifvlan_linkmib));
    386 	free(ifv, M_DEVBUF);
    387 
    388 	return rv;
    389 }
    390 
    391 static int
    392 vlan_clone_destroy(struct ifnet *ifp)
    393 {
    394 	struct ifvlan *ifv = ifp->if_softc;
    395 
    396 	mutex_enter(&ifv_list.lock);
    397 	LIST_REMOVE(ifv, ifv_list);
    398 	mutex_exit(&ifv_list.lock);
    399 
    400 	IFNET_LOCK(ifp);
    401 	vlan_unconfig(ifp);
    402 	IFNET_UNLOCK(ifp);
    403 	if_detach(ifp);
    404 
    405 	psref_target_destroy(&ifv->ifv_mib->ifvm_psref, ifvm_psref_class);
    406 	kmem_free(ifv->ifv_mib, sizeof(struct ifvlan_linkmib));
    407 	pserialize_destroy(ifv->ifv_psz);
    408 	mutex_destroy(&ifv->ifv_lock);
    409 	free(ifv, M_DEVBUF);
    410 
    411 	return 0;
    412 }
    413 
    414 /*
    415  * Configure a VLAN interface.
    416  */
    417 static int
    418 vlan_config(struct ifvlan *ifv, struct ifnet *p, uint16_t tag)
    419 {
    420 	struct ifnet *ifp = &ifv->ifv_if;
    421 	struct ifvlan_linkmib *nmib = NULL;
    422 	struct ifvlan_linkmib *omib = NULL;
    423 	struct ifvlan_linkmib *checkmib;
    424 	struct psref_target *nmib_psref = NULL;
    425 	const uint16_t vid = EVL_VLANOFTAG(tag);
    426 	int error = 0;
    427 	int idx;
    428 	bool omib_cleanup = false;
    429 	struct psref psref;
    430 
    431 	/* VLAN ID 0 and 4095 are reserved in the spec */
    432 	if ((vid == 0) || (vid == 0xfff))
    433 		return EINVAL;
    434 
    435 	nmib = kmem_alloc(sizeof(*nmib), KM_SLEEP);
    436 	mutex_enter(&ifv->ifv_lock);
    437 	omib = ifv->ifv_mib;
    438 
    439 	if (omib->ifvm_p != NULL) {
    440 		error = EBUSY;
    441 		goto done;
    442 	}
    443 
    444 	/* Duplicate check */
    445 	checkmib = vlan_lookup_tag_psref(p, vid, &psref);
    446 	if (checkmib != NULL) {
    447 		vlan_putref_linkmib(checkmib, &psref);
    448 		error = EEXIST;
    449 		goto done;
    450 	}
    451 
    452 	*nmib = *omib;
    453 	nmib_psref = &nmib->ifvm_psref;
    454 
    455 	psref_target_init(nmib_psref, ifvm_psref_class);
    456 
    457 	switch (p->if_type) {
    458 	case IFT_ETHER:
    459 	    {
    460 		struct ethercom *ec = (void *)p;
    461 		struct vlanid_list *vidmem;
    462 
    463 		nmib->ifvm_msw = &vlan_ether_multisw;
    464 		nmib->ifvm_encaplen = ETHER_VLAN_ENCAP_LEN;
    465 		nmib->ifvm_mintu = ETHERMIN;
    466 
    467 		if (ec->ec_nvlans++ == 0) {
    468 			IFNET_LOCK(p);
    469 			error = ether_enable_vlan_mtu(p);
    470 			IFNET_UNLOCK(p);
    471 			if (error >= 0) {
    472 				if (error) {
    473 					ec->ec_nvlans--;
    474 					goto done;
    475 				}
    476 				nmib->ifvm_mtufudge = 0;
    477 			} else {
    478 				/*
    479 				 * Fudge the MTU by the encapsulation size. This
    480 				 * makes us incompatible with strictly compliant
    481 				 * 802.1Q implementations, but allows us to use
    482 				 * the feature with other NetBSD
    483 				 * implementations, which might still be useful.
    484 				 */
    485 				nmib->ifvm_mtufudge = nmib->ifvm_encaplen;
    486 			}
    487 			error = 0;
    488 		}
    489 		/* Add a vid to the list */
    490 		vidmem = kmem_alloc(sizeof(struct vlanid_list), KM_SLEEP);
    491 		vidmem->vid = vid;
    492 		ETHER_LOCK(ec);
    493 		SIMPLEQ_INSERT_TAIL(&ec->ec_vids, vidmem, vid_list);
    494 		ETHER_UNLOCK(ec);
    495 
    496 		if (ec->ec_vlan_cb != NULL) {
    497 			/*
    498 			 * Call ec_vlan_cb(). It will setup VLAN HW filter or
    499 			 * HW tagging function.
    500 			 */
    501 			error = (*ec->ec_vlan_cb)(ec, vid, true);
    502 			if (error) {
    503 				ec->ec_nvlans--;
    504 				if (ec->ec_nvlans == 0) {
    505 					IFNET_LOCK(p);
    506 					(void)ether_disable_vlan_mtu(p);
    507 					IFNET_UNLOCK(p);
    508 				}
    509 				goto done;
    510 			}
    511 		}
    512 		/*
    513 		 * If the parent interface can do hardware-assisted
    514 		 * VLAN encapsulation, then propagate its hardware-
    515 		 * assisted checksumming flags and tcp segmentation
    516 		 * offload.
    517 		 */
    518 		if (ec->ec_capabilities & ETHERCAP_VLAN_HWTAGGING) {
    519 			ifp->if_capabilities = p->if_capabilities &
    520 			    (IFCAP_TSOv4 | IFCAP_TSOv6 |
    521 				IFCAP_CSUM_IPv4_Tx  | IFCAP_CSUM_IPv4_Rx |
    522 				IFCAP_CSUM_TCPv4_Tx | IFCAP_CSUM_TCPv4_Rx |
    523 				IFCAP_CSUM_UDPv4_Tx | IFCAP_CSUM_UDPv4_Rx |
    524 				IFCAP_CSUM_TCPv6_Tx | IFCAP_CSUM_TCPv6_Rx |
    525 				IFCAP_CSUM_UDPv6_Tx | IFCAP_CSUM_UDPv6_Rx);
    526 		}
    527 
    528 		/*
    529 		 * We inherit the parent's Ethernet address.
    530 		 */
    531 		ether_ifattach(ifp, CLLADDR(p->if_sadl));
    532 		ifp->if_hdrlen = sizeof(struct ether_vlan_header); /* XXX? */
    533 		break;
    534 	    }
    535 
    536 	default:
    537 		error = EPROTONOSUPPORT;
    538 		goto done;
    539 	}
    540 
    541 	nmib->ifvm_p = p;
    542 	nmib->ifvm_tag = vid;
    543 	ifv->ifv_if.if_mtu = p->if_mtu - nmib->ifvm_mtufudge;
    544 #ifdef INET6
    545 	KERNEL_LOCK_UNLESS_NET_MPSAFE();
    546 	if (in6_present)
    547 		nd6_setmtu(ifp);
    548 	KERNEL_UNLOCK_UNLESS_NET_MPSAFE();
    549 #endif
    550 	ifv->ifv_if.if_flags = p->if_flags &
    551 	    (IFF_UP | IFF_BROADCAST | IFF_SIMPLEX | IFF_MULTICAST);
    552 
    553 	/*
    554 	 * Inherit the if_type from the parent.  This allows us
    555 	 * to participate in bridges of that type.
    556 	 */
    557 	ifv->ifv_if.if_type = p->if_type;
    558 
    559 	PSLIST_ENTRY_INIT(ifv, ifv_hash);
    560 	idx = vlan_tag_hash(vid, ifv_hash.mask);
    561 
    562 	mutex_enter(&ifv_hash.lock);
    563 	PSLIST_WRITER_INSERT_HEAD(&ifv_hash.lists[idx], ifv, ifv_hash);
    564 	mutex_exit(&ifv_hash.lock);
    565 
    566 	vlan_linkmib_update(ifv, nmib);
    567 	nmib = NULL;
    568 	nmib_psref = NULL;
    569 	omib_cleanup = true;
    570 
    571 done:
    572 	mutex_exit(&ifv->ifv_lock);
    573 
    574 	if (nmib_psref)
    575 		psref_target_destroy(nmib_psref, ifvm_psref_class);
    576 	if (nmib)
    577 		kmem_free(nmib, sizeof(*nmib));
    578 	if (omib_cleanup)
    579 		kmem_free(omib, sizeof(*omib));
    580 
    581 	return error;
    582 }
    583 
    584 /*
    585  * Unconfigure a VLAN interface.
    586  */
    587 static void
    588 vlan_unconfig(struct ifnet *ifp)
    589 {
    590 	struct ifvlan *ifv = ifp->if_softc;
    591 	struct ifvlan_linkmib *nmib = NULL;
    592 	int error;
    593 
    594 	KASSERT(IFNET_LOCKED(ifp));
    595 
    596 	nmib = kmem_alloc(sizeof(*nmib), KM_SLEEP);
    597 
    598 	mutex_enter(&ifv->ifv_lock);
    599 	error = vlan_unconfig_locked(ifv, nmib);
    600 	mutex_exit(&ifv->ifv_lock);
    601 
    602 	if (error)
    603 		kmem_free(nmib, sizeof(*nmib));
    604 }
    605 static int
    606 vlan_unconfig_locked(struct ifvlan *ifv, struct ifvlan_linkmib *nmib)
    607 {
    608 	struct ifnet *p;
    609 	struct ifnet *ifp = &ifv->ifv_if;
    610 	struct psref_target *nmib_psref = NULL;
    611 	struct ifvlan_linkmib *omib;
    612 	int error = 0;
    613 
    614 	KASSERT(IFNET_LOCKED(ifp));
    615 	KASSERT(mutex_owned(&ifv->ifv_lock));
    616 
    617 	ifp->if_flags &= ~(IFF_UP | IFF_RUNNING);
    618 
    619 	omib = ifv->ifv_mib;
    620 	p = omib->ifvm_p;
    621 
    622 	if (p == NULL) {
    623 		error = -1;
    624 		goto done;
    625 	}
    626 
    627 	*nmib = *omib;
    628 	nmib_psref = &nmib->ifvm_psref;
    629 	psref_target_init(nmib_psref, ifvm_psref_class);
    630 
    631 	/*
    632 	 * Since the interface is being unconfigured, we need to empty the
    633 	 * list of multicast groups that we may have joined while we were
    634 	 * alive and remove them from the parent's list also.
    635 	 */
    636 	(*nmib->ifvm_msw->vmsw_purgemulti)(ifv);
    637 
    638 	/* Disconnect from parent. */
    639 	switch (p->if_type) {
    640 	case IFT_ETHER:
    641 	    {
    642 		struct ethercom *ec = (void *)p;
    643 		struct vlanid_list *vlanidp;
    644 		uint16_t vid = EVL_VLANOFTAG(nmib->ifvm_tag);
    645 
    646 		ETHER_LOCK(ec);
    647 		SIMPLEQ_FOREACH(vlanidp, &ec->ec_vids, vid_list) {
    648 			if (vlanidp->vid == vid) {
    649 				SIMPLEQ_REMOVE(&ec->ec_vids, vlanidp,
    650 				    vlanid_list, vid_list);
    651 				break;
    652 			}
    653 		}
    654 		ETHER_UNLOCK(ec);
    655 		if (vlanidp != NULL)
    656 			kmem_free(vlanidp, sizeof(*vlanidp));
    657 
    658 		if (ec->ec_vlan_cb != NULL) {
    659 			/*
    660 			 * Call ec_vlan_cb(). It will setup VLAN HW filter or
    661 			 * HW tagging function.
    662 			 */
    663 			(void)(*ec->ec_vlan_cb)(ec, vid, false);
    664 		}
    665 		if (--ec->ec_nvlans == 0) {
    666 			IFNET_LOCK(p);
    667 			(void)ether_disable_vlan_mtu(p);
    668 			IFNET_UNLOCK(p);
    669 		}
    670 
    671 		/* XXX ether_ifdetach must not be called with IFNET_LOCK */
    672 		mutex_exit(&ifv->ifv_lock);
    673 		IFNET_UNLOCK(ifp);
    674 		ether_ifdetach(ifp);
    675 		IFNET_LOCK(ifp);
    676 		mutex_enter(&ifv->ifv_lock);
    677 
    678 		/* if_free_sadl must be called with IFNET_LOCK */
    679 		if_free_sadl(ifp, 1);
    680 
    681 		/* Restore vlan_ioctl overwritten by ether_ifdetach */
    682 		ifp->if_ioctl = vlan_ioctl;
    683 		vlan_reset_linkname(ifp);
    684 		break;
    685 	    }
    686 
    687 	default:
    688 		panic("%s: impossible", __func__);
    689 	}
    690 
    691 	nmib->ifvm_p = NULL;
    692 	ifv->ifv_if.if_mtu = 0;
    693 	ifv->ifv_flags = 0;
    694 
    695 	mutex_enter(&ifv_hash.lock);
    696 	PSLIST_WRITER_REMOVE(ifv, ifv_hash);
    697 	pserialize_perform(vlan_psz);
    698 	mutex_exit(&ifv_hash.lock);
    699 	PSLIST_ENTRY_DESTROY(ifv, ifv_hash);
    700 
    701 	vlan_linkmib_update(ifv, nmib);
    702 
    703 	mutex_exit(&ifv->ifv_lock);
    704 
    705 	nmib_psref = NULL;
    706 	kmem_free(omib, sizeof(*omib));
    707 
    708 #ifdef INET6
    709 	KERNEL_LOCK_UNLESS_NET_MPSAFE();
    710 	/* To delete v6 link local addresses */
    711 	if (in6_present)
    712 		in6_ifdetach(ifp);
    713 	KERNEL_UNLOCK_UNLESS_NET_MPSAFE();
    714 #endif
    715 
    716 	if ((ifp->if_flags & IFF_PROMISC) != 0)
    717 		vlan_safe_ifpromisc_locked(ifp, 0);
    718 	if_down_locked(ifp);
    719 	ifp->if_capabilities = 0;
    720 	mutex_enter(&ifv->ifv_lock);
    721 done:
    722 
    723 	if (nmib_psref)
    724 		psref_target_destroy(nmib_psref, ifvm_psref_class);
    725 
    726 	return error;
    727 }
    728 
    729 static void
    730 vlan_hash_init(void)
    731 {
    732 
    733 	ifv_hash.lists = hashinit(VLAN_TAG_HASH_SIZE, HASH_PSLIST, true,
    734 	    &ifv_hash.mask);
    735 }
    736 
    737 static int
    738 vlan_hash_fini(void)
    739 {
    740 	int i;
    741 
    742 	mutex_enter(&ifv_hash.lock);
    743 
    744 	for (i = 0; i < ifv_hash.mask + 1; i++) {
    745 		if (PSLIST_WRITER_FIRST(&ifv_hash.lists[i], struct ifvlan,
    746 		    ifv_hash) != NULL) {
    747 			mutex_exit(&ifv_hash.lock);
    748 			return EBUSY;
    749 		}
    750 	}
    751 
    752 	for (i = 0; i < ifv_hash.mask + 1; i++)
    753 		PSLIST_DESTROY(&ifv_hash.lists[i]);
    754 
    755 	mutex_exit(&ifv_hash.lock);
    756 
    757 	hashdone(ifv_hash.lists, HASH_PSLIST, ifv_hash.mask);
    758 
    759 	ifv_hash.lists = NULL;
    760 	ifv_hash.mask = 0;
    761 
    762 	return 0;
    763 }
    764 
    765 static int
    766 vlan_tag_hash(uint16_t tag, u_long mask)
    767 {
    768 	uint32_t hash;
    769 
    770 	hash = (tag >> 8) ^ tag;
    771 	hash = (hash >> 2) ^ hash;
    772 
    773 	return hash & mask;
    774 }
    775 
    776 static struct ifvlan_linkmib *
    777 vlan_getref_linkmib(struct ifvlan *sc, struct psref *psref)
    778 {
    779 	struct ifvlan_linkmib *mib;
    780 	int s;
    781 
    782 	s = pserialize_read_enter();
    783 	mib = atomic_load_consume(&sc->ifv_mib);
    784 	if (mib == NULL) {
    785 		pserialize_read_exit(s);
    786 		return NULL;
    787 	}
    788 	psref_acquire(psref, &mib->ifvm_psref, ifvm_psref_class);
    789 	pserialize_read_exit(s);
    790 
    791 	return mib;
    792 }
    793 
    794 static void
    795 vlan_putref_linkmib(struct ifvlan_linkmib *mib, struct psref *psref)
    796 {
    797 	if (mib == NULL)
    798 		return;
    799 	psref_release(psref, &mib->ifvm_psref, ifvm_psref_class);
    800 }
    801 
    802 static struct ifvlan_linkmib *
    803 vlan_lookup_tag_psref(struct ifnet *ifp, uint16_t tag, struct psref *psref)
    804 {
    805 	int idx;
    806 	int s;
    807 	struct ifvlan *sc;
    808 
    809 	idx = vlan_tag_hash(tag, ifv_hash.mask);
    810 
    811 	s = pserialize_read_enter();
    812 	PSLIST_READER_FOREACH(sc, &ifv_hash.lists[idx], struct ifvlan,
    813 	    ifv_hash) {
    814 		struct ifvlan_linkmib *mib = atomic_load_consume(&sc->ifv_mib);
    815 		if (mib == NULL)
    816 			continue;
    817 		if (mib->ifvm_tag != tag)
    818 			continue;
    819 		if (mib->ifvm_p != ifp)
    820 			continue;
    821 
    822 		psref_acquire(psref, &mib->ifvm_psref, ifvm_psref_class);
    823 		pserialize_read_exit(s);
    824 		return mib;
    825 	}
    826 	pserialize_read_exit(s);
    827 	return NULL;
    828 }
    829 
    830 static void
    831 vlan_linkmib_update(struct ifvlan *ifv, struct ifvlan_linkmib *nmib)
    832 {
    833 	struct ifvlan_linkmib *omib = ifv->ifv_mib;
    834 
    835 	KASSERT(mutex_owned(&ifv->ifv_lock));
    836 
    837 	atomic_store_release(&ifv->ifv_mib, nmib);
    838 
    839 	pserialize_perform(ifv->ifv_psz);
    840 	psref_target_destroy(&omib->ifvm_psref, ifvm_psref_class);
    841 }
    842 
    843 /*
    844  * Called when a parent interface is detaching; destroy any VLAN
    845  * configuration for the parent interface.
    846  */
    847 void
    848 vlan_ifdetach(struct ifnet *p)
    849 {
    850 	struct ifvlan *ifv;
    851 	struct ifvlan_linkmib *mib, **nmibs;
    852 	struct psref psref;
    853 	int error;
    854 	int bound;
    855 	int i, cnt = 0;
    856 
    857 	bound = curlwp_bind();
    858 
    859 	mutex_enter(&ifv_list.lock);
    860 	LIST_FOREACH(ifv, &ifv_list.list, ifv_list) {
    861 		mib = vlan_getref_linkmib(ifv, &psref);
    862 		if (mib == NULL)
    863 			continue;
    864 
    865 		if (mib->ifvm_p == p)
    866 			cnt++;
    867 
    868 		vlan_putref_linkmib(mib, &psref);
    869 	}
    870 	mutex_exit(&ifv_list.lock);
    871 
    872 	if (cnt == 0) {
    873 		curlwp_bindx(bound);
    874 		return;
    875 	}
    876 
    877 	/*
    878 	 * The value of "cnt" does not increase while ifv_list.lock
    879 	 * and ifv->ifv_lock are released here, because the parent
    880 	 * interface is detaching.
    881 	 */
    882 	nmibs = kmem_alloc(sizeof(*nmibs) * cnt, KM_SLEEP);
    883 	for (i = 0; i < cnt; i++) {
    884 		nmibs[i] = kmem_alloc(sizeof(*nmibs[i]), KM_SLEEP);
    885 	}
    886 
    887 	mutex_enter(&ifv_list.lock);
    888 
    889 	i = 0;
    890 	LIST_FOREACH(ifv, &ifv_list.list, ifv_list) {
    891 		struct ifnet *ifp = &ifv->ifv_if;
    892 
    893 		/* IFNET_LOCK must be held before ifv_lock. */
    894 		IFNET_LOCK(ifp);
    895 		mutex_enter(&ifv->ifv_lock);
    896 
    897 		/* XXX ifv_mib = NULL? */
    898 		if (ifv->ifv_mib->ifvm_p == p) {
    899 			KASSERTMSG(i < cnt,
    900 			    "no memory for unconfig, parent=%s", p->if_xname);
    901 			error = vlan_unconfig_locked(ifv, nmibs[i]);
    902 			if (!error) {
    903 				nmibs[i] = NULL;
    904 				i++;
    905 			}
    906 
    907 		}
    908 
    909 		mutex_exit(&ifv->ifv_lock);
    910 		IFNET_UNLOCK(ifp);
    911 	}
    912 
    913 	mutex_exit(&ifv_list.lock);
    914 
    915 	curlwp_bindx(bound);
    916 
    917 	for (i = 0; i < cnt; i++) {
    918 		if (nmibs[i])
    919 			kmem_free(nmibs[i], sizeof(*nmibs[i]));
    920 	}
    921 
    922 	kmem_free(nmibs, sizeof(*nmibs) * cnt);
    923 
    924 	return;
    925 }
    926 
    927 static int
    928 vlan_set_promisc(struct ifnet *ifp)
    929 {
    930 	struct ifvlan *ifv = ifp->if_softc;
    931 	struct ifvlan_linkmib *mib;
    932 	struct psref psref;
    933 	int error = 0;
    934 	int bound;
    935 
    936 	bound = curlwp_bind();
    937 	mib = vlan_getref_linkmib(ifv, &psref);
    938 	if (mib == NULL) {
    939 		curlwp_bindx(bound);
    940 		return EBUSY;
    941 	}
    942 
    943 	if ((ifp->if_flags & IFF_PROMISC) != 0) {
    944 		if ((ifv->ifv_flags & IFVF_PROMISC) == 0) {
    945 			error = vlan_safe_ifpromisc(mib->ifvm_p, 1);
    946 			if (error == 0)
    947 				ifv->ifv_flags |= IFVF_PROMISC;
    948 		}
    949 	} else {
    950 		if ((ifv->ifv_flags & IFVF_PROMISC) != 0) {
    951 			error = vlan_safe_ifpromisc(mib->ifvm_p, 0);
    952 			if (error == 0)
    953 				ifv->ifv_flags &= ~IFVF_PROMISC;
    954 		}
    955 	}
    956 	vlan_putref_linkmib(mib, &psref);
    957 	curlwp_bindx(bound);
    958 
    959 	return error;
    960 }
    961 
    962 static int
    963 vlan_ioctl(struct ifnet *ifp, u_long cmd, void *data)
    964 {
    965 	struct lwp *l = curlwp;
    966 	struct ifvlan *ifv = ifp->if_softc;
    967 	struct ifaddr *ifa = (struct ifaddr *) data;
    968 	struct ifreq *ifr = (struct ifreq *) data;
    969 	struct ifnet *pr;
    970 	struct ifcapreq *ifcr;
    971 	struct vlanreq vlr;
    972 	struct ifvlan_linkmib *mib;
    973 	struct psref psref;
    974 	int error = 0;
    975 	int bound;
    976 
    977 	switch (cmd) {
    978 	case SIOCSIFMTU:
    979 		bound = curlwp_bind();
    980 		mib = vlan_getref_linkmib(ifv, &psref);
    981 		if (mib == NULL) {
    982 			curlwp_bindx(bound);
    983 			error = EBUSY;
    984 			break;
    985 		}
    986 
    987 		if (mib->ifvm_p == NULL) {
    988 			vlan_putref_linkmib(mib, &psref);
    989 			curlwp_bindx(bound);
    990 			error = EINVAL;
    991 		} else if (
    992 		    ifr->ifr_mtu > (mib->ifvm_p->if_mtu - mib->ifvm_mtufudge) ||
    993 		    ifr->ifr_mtu < (mib->ifvm_mintu - mib->ifvm_mtufudge)) {
    994 			vlan_putref_linkmib(mib, &psref);
    995 			curlwp_bindx(bound);
    996 			error = EINVAL;
    997 		} else {
    998 			vlan_putref_linkmib(mib, &psref);
    999 			curlwp_bindx(bound);
   1000 
   1001 			error = ifioctl_common(ifp, cmd, data);
   1002 			if (error == ENETRESET)
   1003 				error = 0;
   1004 		}
   1005 
   1006 		break;
   1007 
   1008 	case SIOCSETVLAN:
   1009 		if ((error = kauth_authorize_network(l->l_cred,
   1010 		    KAUTH_NETWORK_INTERFACE,
   1011 		    KAUTH_REQ_NETWORK_INTERFACE_SETPRIV, ifp, (void *)cmd,
   1012 		    NULL)) != 0)
   1013 			break;
   1014 		if ((error = copyin(ifr->ifr_data, &vlr, sizeof(vlr))) != 0)
   1015 			break;
   1016 
   1017 		if (vlr.vlr_parent[0] == '\0') {
   1018 			bound = curlwp_bind();
   1019 			mib = vlan_getref_linkmib(ifv, &psref);
   1020 			if (mib == NULL) {
   1021 				curlwp_bindx(bound);
   1022 				error = EBUSY;
   1023 				break;
   1024 			}
   1025 
   1026 			if (mib->ifvm_p != NULL &&
   1027 			    (ifp->if_flags & IFF_PROMISC) != 0)
   1028 				error = vlan_safe_ifpromisc(mib->ifvm_p, 0);
   1029 
   1030 			vlan_putref_linkmib(mib, &psref);
   1031 			curlwp_bindx(bound);
   1032 
   1033 			vlan_unconfig(ifp);
   1034 			break;
   1035 		}
   1036 		if (vlr.vlr_tag != EVL_VLANOFTAG(vlr.vlr_tag)) {
   1037 			error = EINVAL;		 /* check for valid tag */
   1038 			break;
   1039 		}
   1040 		if ((pr = ifunit(vlr.vlr_parent)) == NULL) {
   1041 			error = ENOENT;
   1042 			break;
   1043 		}
   1044 
   1045 		error = vlan_config(ifv, pr, vlr.vlr_tag);
   1046 		if (error != 0)
   1047 			break;
   1048 
   1049 		/* Update promiscuous mode, if necessary. */
   1050 		vlan_set_promisc(ifp);
   1051 
   1052 		ifp->if_flags |= IFF_RUNNING;
   1053 		break;
   1054 
   1055 	case SIOCGETVLAN:
   1056 		memset(&vlr, 0, sizeof(vlr));
   1057 		bound = curlwp_bind();
   1058 		mib = vlan_getref_linkmib(ifv, &psref);
   1059 		if (mib == NULL) {
   1060 			curlwp_bindx(bound);
   1061 			error = EBUSY;
   1062 			break;
   1063 		}
   1064 		if (mib->ifvm_p != NULL) {
   1065 			snprintf(vlr.vlr_parent, sizeof(vlr.vlr_parent), "%s",
   1066 			    mib->ifvm_p->if_xname);
   1067 			vlr.vlr_tag = mib->ifvm_tag;
   1068 		}
   1069 		vlan_putref_linkmib(mib, &psref);
   1070 		curlwp_bindx(bound);
   1071 		error = copyout(&vlr, ifr->ifr_data, sizeof(vlr));
   1072 		break;
   1073 
   1074 	case SIOCSIFFLAGS:
   1075 		if ((error = ifioctl_common(ifp, cmd, data)) != 0)
   1076 			break;
   1077 		/*
   1078 		 * For promiscuous mode, we enable promiscuous mode on
   1079 		 * the parent if we need promiscuous on the VLAN interface.
   1080 		 */
   1081 		bound = curlwp_bind();
   1082 		mib = vlan_getref_linkmib(ifv, &psref);
   1083 		if (mib == NULL) {
   1084 			curlwp_bindx(bound);
   1085 			error = EBUSY;
   1086 			break;
   1087 		}
   1088 
   1089 		if (mib->ifvm_p != NULL)
   1090 			error = vlan_set_promisc(ifp);
   1091 		vlan_putref_linkmib(mib, &psref);
   1092 		curlwp_bindx(bound);
   1093 		break;
   1094 
   1095 	case SIOCADDMULTI:
   1096 		mutex_enter(&ifv->ifv_lock);
   1097 		mib = ifv->ifv_mib;
   1098 		if (mib == NULL) {
   1099 			error = EBUSY;
   1100 			mutex_exit(&ifv->ifv_lock);
   1101 			break;
   1102 		}
   1103 
   1104 		error = (mib->ifvm_p != NULL) ?
   1105 		    (*mib->ifvm_msw->vmsw_addmulti)(ifv, ifr) : EINVAL;
   1106 		mib = NULL;
   1107 		mutex_exit(&ifv->ifv_lock);
   1108 		break;
   1109 
   1110 	case SIOCDELMULTI:
   1111 		mutex_enter(&ifv->ifv_lock);
   1112 		mib = ifv->ifv_mib;
   1113 		if (mib == NULL) {
   1114 			error = EBUSY;
   1115 			mutex_exit(&ifv->ifv_lock);
   1116 			break;
   1117 		}
   1118 		error = (mib->ifvm_p != NULL) ?
   1119 		    (*mib->ifvm_msw->vmsw_delmulti)(ifv, ifr) : EINVAL;
   1120 		mib = NULL;
   1121 		mutex_exit(&ifv->ifv_lock);
   1122 		break;
   1123 
   1124 	case SIOCSIFCAP:
   1125 		ifcr = data;
   1126 		/* make sure caps are enabled on parent */
   1127 		bound = curlwp_bind();
   1128 		mib = vlan_getref_linkmib(ifv, &psref);
   1129 		if (mib == NULL) {
   1130 			curlwp_bindx(bound);
   1131 			error = EBUSY;
   1132 			break;
   1133 		}
   1134 
   1135 		if (mib->ifvm_p == NULL) {
   1136 			vlan_putref_linkmib(mib, &psref);
   1137 			curlwp_bindx(bound);
   1138 			error = EINVAL;
   1139 			break;
   1140 		}
   1141 		if ((mib->ifvm_p->if_capenable & ifcr->ifcr_capenable) !=
   1142 		    ifcr->ifcr_capenable) {
   1143 			vlan_putref_linkmib(mib, &psref);
   1144 			curlwp_bindx(bound);
   1145 			error = EINVAL;
   1146 			break;
   1147 		}
   1148 
   1149 		vlan_putref_linkmib(mib, &psref);
   1150 		curlwp_bindx(bound);
   1151 
   1152 		if ((error = ifioctl_common(ifp, cmd, data)) == ENETRESET)
   1153 			error = 0;
   1154 		break;
   1155 	case SIOCINITIFADDR:
   1156 		bound = curlwp_bind();
   1157 		mib = vlan_getref_linkmib(ifv, &psref);
   1158 		if (mib == NULL) {
   1159 			curlwp_bindx(bound);
   1160 			error = EBUSY;
   1161 			break;
   1162 		}
   1163 
   1164 		if (mib->ifvm_p == NULL) {
   1165 			error = EINVAL;
   1166 			vlan_putref_linkmib(mib, &psref);
   1167 			curlwp_bindx(bound);
   1168 			break;
   1169 		}
   1170 		vlan_putref_linkmib(mib, &psref);
   1171 		curlwp_bindx(bound);
   1172 
   1173 		ifp->if_flags |= IFF_UP;
   1174 #ifdef INET
   1175 		if (ifa->ifa_addr->sa_family == AF_INET)
   1176 			arp_ifinit(ifp, ifa);
   1177 #endif
   1178 		break;
   1179 
   1180 	default:
   1181 		error = ether_ioctl(ifp, cmd, data);
   1182 	}
   1183 
   1184 	return error;
   1185 }
   1186 
   1187 static int
   1188 vlan_ether_addmulti(struct ifvlan *ifv, struct ifreq *ifr)
   1189 {
   1190 	const struct sockaddr *sa = ifreq_getaddr(SIOCADDMULTI, ifr);
   1191 	struct vlan_mc_entry *mc;
   1192 	uint8_t addrlo[ETHER_ADDR_LEN], addrhi[ETHER_ADDR_LEN];
   1193 	struct ifvlan_linkmib *mib;
   1194 	int error;
   1195 
   1196 	KASSERT(mutex_owned(&ifv->ifv_lock));
   1197 
   1198 	if (sa->sa_len > sizeof(struct sockaddr_storage))
   1199 		return EINVAL;
   1200 
   1201 	error = ether_addmulti(sa, &ifv->ifv_ec);
   1202 	if (error != ENETRESET)
   1203 		return error;
   1204 
   1205 	/*
   1206 	 * This is a new multicast address.  We have to tell parent
   1207 	 * about it.  Also, remember this multicast address so that
   1208 	 * we can delete it on unconfigure.
   1209 	 */
   1210 	mc = malloc(sizeof(struct vlan_mc_entry), M_DEVBUF, M_NOWAIT);
   1211 	if (mc == NULL) {
   1212 		error = ENOMEM;
   1213 		goto alloc_failed;
   1214 	}
   1215 
   1216 	/*
   1217 	 * Since ether_addmulti() returned ENETRESET, the following two
   1218 	 * statements shouldn't fail. Here ifv_ec is implicitly protected
   1219 	 * by the ifv_lock lock.
   1220 	 */
   1221 	error = ether_multiaddr(sa, addrlo, addrhi);
   1222 	KASSERT(error == 0);
   1223 
   1224 	ETHER_LOCK(&ifv->ifv_ec);
   1225 	mc->mc_enm = ether_lookup_multi(addrlo, addrhi, &ifv->ifv_ec);
   1226 	ETHER_UNLOCK(&ifv->ifv_ec);
   1227 
   1228 	KASSERT(mc->mc_enm != NULL);
   1229 
   1230 	memcpy(&mc->mc_addr, sa, sa->sa_len);
   1231 	LIST_INSERT_HEAD(&ifv->ifv_mc_listhead, mc, mc_entries);
   1232 
   1233 	mib = ifv->ifv_mib;
   1234 
   1235 	KERNEL_LOCK_UNLESS_IFP_MPSAFE(mib->ifvm_p);
   1236 	error = if_mcast_op(mib->ifvm_p, SIOCADDMULTI, sa);
   1237 	KERNEL_UNLOCK_UNLESS_IFP_MPSAFE(mib->ifvm_p);
   1238 
   1239 	if (error != 0)
   1240 		goto ioctl_failed;
   1241 	return error;
   1242 
   1243 ioctl_failed:
   1244 	LIST_REMOVE(mc, mc_entries);
   1245 	free(mc, M_DEVBUF);
   1246 
   1247 alloc_failed:
   1248 	(void)ether_delmulti(sa, &ifv->ifv_ec);
   1249 	return error;
   1250 }
   1251 
   1252 static int
   1253 vlan_ether_delmulti(struct ifvlan *ifv, struct ifreq *ifr)
   1254 {
   1255 	const struct sockaddr *sa = ifreq_getaddr(SIOCDELMULTI, ifr);
   1256 	struct ether_multi *enm;
   1257 	struct vlan_mc_entry *mc;
   1258 	struct ifvlan_linkmib *mib;
   1259 	uint8_t addrlo[ETHER_ADDR_LEN], addrhi[ETHER_ADDR_LEN];
   1260 	int error;
   1261 
   1262 	KASSERT(mutex_owned(&ifv->ifv_lock));
   1263 
   1264 	/*
   1265 	 * Find a key to lookup vlan_mc_entry.  We have to do this
   1266 	 * before calling ether_delmulti for obvious reasons.
   1267 	 */
   1268 	if ((error = ether_multiaddr(sa, addrlo, addrhi)) != 0)
   1269 		return error;
   1270 
   1271 	ETHER_LOCK(&ifv->ifv_ec);
   1272 	enm = ether_lookup_multi(addrlo, addrhi, &ifv->ifv_ec);
   1273 	ETHER_UNLOCK(&ifv->ifv_ec);
   1274 	if (enm == NULL)
   1275 		return EINVAL;
   1276 
   1277 	LIST_FOREACH(mc, &ifv->ifv_mc_listhead, mc_entries) {
   1278 		if (mc->mc_enm == enm)
   1279 			break;
   1280 	}
   1281 
   1282 	/* We woun't delete entries we didn't add */
   1283 	if (mc == NULL)
   1284 		return EINVAL;
   1285 
   1286 	error = ether_delmulti(sa, &ifv->ifv_ec);
   1287 	if (error != ENETRESET)
   1288 		return error;
   1289 
   1290 	/* We no longer use this multicast address.  Tell parent so. */
   1291 	mib = ifv->ifv_mib;
   1292 	error = if_mcast_op(mib->ifvm_p, SIOCDELMULTI, sa);
   1293 
   1294 	if (error == 0) {
   1295 		/* And forget about this address. */
   1296 		LIST_REMOVE(mc, mc_entries);
   1297 		free(mc, M_DEVBUF);
   1298 	} else {
   1299 		(void)ether_addmulti(sa, &ifv->ifv_ec);
   1300 	}
   1301 
   1302 	return error;
   1303 }
   1304 
   1305 /*
   1306  * Delete any multicast address we have asked to add from parent
   1307  * interface.  Called when the vlan is being unconfigured.
   1308  */
   1309 static void
   1310 vlan_ether_purgemulti(struct ifvlan *ifv)
   1311 {
   1312 	struct vlan_mc_entry *mc;
   1313 	struct ifvlan_linkmib *mib;
   1314 
   1315 	KASSERT(mutex_owned(&ifv->ifv_lock));
   1316 	mib = ifv->ifv_mib;
   1317 	if (mib == NULL) {
   1318 		return;
   1319 	}
   1320 
   1321 	while ((mc = LIST_FIRST(&ifv->ifv_mc_listhead)) != NULL) {
   1322 		(void)if_mcast_op(mib->ifvm_p, SIOCDELMULTI,
   1323 		    sstocsa(&mc->mc_addr));
   1324 		LIST_REMOVE(mc, mc_entries);
   1325 		free(mc, M_DEVBUF);
   1326 	}
   1327 }
   1328 
   1329 static void
   1330 vlan_start(struct ifnet *ifp)
   1331 {
   1332 	struct ifvlan *ifv = ifp->if_softc;
   1333 	struct ifnet *p;
   1334 	struct ethercom *ec;
   1335 	struct mbuf *m;
   1336 	struct ifvlan_linkmib *mib;
   1337 	struct psref psref;
   1338 	int error;
   1339 
   1340 	mib = vlan_getref_linkmib(ifv, &psref);
   1341 	if (mib == NULL)
   1342 		return;
   1343 	p = mib->ifvm_p;
   1344 	ec = (void *)mib->ifvm_p;
   1345 
   1346 	ifp->if_flags |= IFF_OACTIVE;
   1347 
   1348 	for (;;) {
   1349 		IFQ_DEQUEUE(&ifp->if_snd, m);
   1350 		if (m == NULL)
   1351 			break;
   1352 
   1353 #ifdef ALTQ
   1354 		/*
   1355 		 * KERNEL_LOCK is required for ALTQ even if NET_MPSAFE is
   1356 		 * defined.
   1357 		 */
   1358 		KERNEL_LOCK(1, NULL);
   1359 		/*
   1360 		 * If ALTQ is enabled on the parent interface, do
   1361 		 * classification; the queueing discipline might
   1362 		 * not require classification, but might require
   1363 		 * the address family/header pointer in the pktattr.
   1364 		 */
   1365 		if (ALTQ_IS_ENABLED(&p->if_snd)) {
   1366 			switch (p->if_type) {
   1367 			case IFT_ETHER:
   1368 				altq_etherclassify(&p->if_snd, m);
   1369 				break;
   1370 			default:
   1371 				panic("%s: impossible (altq)", __func__);
   1372 			}
   1373 		}
   1374 		KERNEL_UNLOCK_ONE(NULL);
   1375 #endif /* ALTQ */
   1376 
   1377 		bpf_mtap(ifp, m, BPF_D_OUT);
   1378 		/*
   1379 		 * If the parent can insert the tag itself, just mark
   1380 		 * the tag in the mbuf header.
   1381 		 */
   1382 		if (ec->ec_capenable & ETHERCAP_VLAN_HWTAGGING) {
   1383 			vlan_set_tag(m, mib->ifvm_tag);
   1384 		} else {
   1385 			/*
   1386 			 * insert the tag ourselves
   1387 			 */
   1388 			M_PREPEND(m, mib->ifvm_encaplen, M_DONTWAIT);
   1389 			if (m == NULL) {
   1390 				printf("%s: unable to prepend encap header",
   1391 				    p->if_xname);
   1392 				if_statinc(ifp, if_oerrors);
   1393 				continue;
   1394 			}
   1395 
   1396 			switch (p->if_type) {
   1397 			case IFT_ETHER:
   1398 			    {
   1399 				struct ether_vlan_header *evl;
   1400 
   1401 				if (m->m_len < sizeof(struct ether_vlan_header))
   1402 					m = m_pullup(m,
   1403 					    sizeof(struct ether_vlan_header));
   1404 				if (m == NULL) {
   1405 					printf("%s: unable to pullup encap "
   1406 					    "header", p->if_xname);
   1407 					if_statinc(ifp, if_oerrors);
   1408 					continue;
   1409 				}
   1410 
   1411 				/*
   1412 				 * Transform the Ethernet header into an
   1413 				 * Ethernet header with 802.1Q encapsulation.
   1414 				 */
   1415 				memmove(mtod(m, void *),
   1416 				    mtod(m, char *) + mib->ifvm_encaplen,
   1417 				    sizeof(struct ether_header));
   1418 				evl = mtod(m, struct ether_vlan_header *);
   1419 				evl->evl_proto = evl->evl_encap_proto;
   1420 				evl->evl_encap_proto = htons(ETHERTYPE_VLAN);
   1421 				evl->evl_tag = htons(mib->ifvm_tag);
   1422 
   1423 				/*
   1424 				 * To cater for VLAN-aware layer 2 ethernet
   1425 				 * switches which may need to strip the tag
   1426 				 * before forwarding the packet, make sure
   1427 				 * the packet+tag is at least 68 bytes long.
   1428 				 * This is necessary because our parent will
   1429 				 * only pad to 64 bytes (ETHER_MIN_LEN) and
   1430 				 * some switches will not pad by themselves
   1431 				 * after deleting a tag.
   1432 				 */
   1433 				const size_t min_data_len = ETHER_MIN_LEN -
   1434 				    ETHER_CRC_LEN + ETHER_VLAN_ENCAP_LEN;
   1435 				if (m->m_pkthdr.len < min_data_len) {
   1436 					m_copyback(m, m->m_pkthdr.len,
   1437 					    min_data_len - m->m_pkthdr.len,
   1438 					    vlan_zero_pad_buff);
   1439 				}
   1440 				break;
   1441 			    }
   1442 
   1443 			default:
   1444 				panic("%s: impossible", __func__);
   1445 			}
   1446 		}
   1447 
   1448 		if ((p->if_flags & IFF_RUNNING) == 0) {
   1449 			m_freem(m);
   1450 			continue;
   1451 		}
   1452 
   1453 		error = if_transmit_lock(p, m);
   1454 		if (error) {
   1455 			/* mbuf is already freed */
   1456 			if_statinc(ifp, if_oerrors);
   1457 			continue;
   1458 		}
   1459 		if_statinc(ifp, if_opackets);
   1460 	}
   1461 
   1462 	ifp->if_flags &= ~IFF_OACTIVE;
   1463 
   1464 	/* Remove reference to mib before release */
   1465 	vlan_putref_linkmib(mib, &psref);
   1466 }
   1467 
   1468 static int
   1469 vlan_transmit(struct ifnet *ifp, struct mbuf *m)
   1470 {
   1471 	struct ifvlan *ifv = ifp->if_softc;
   1472 	struct ifnet *p;
   1473 	struct ethercom *ec;
   1474 	struct ifvlan_linkmib *mib;
   1475 	struct psref psref;
   1476 	int error;
   1477 	size_t pktlen = m->m_pkthdr.len;
   1478 	bool mcast = (m->m_flags & M_MCAST) != 0;
   1479 
   1480 	mib = vlan_getref_linkmib(ifv, &psref);
   1481 	if (mib == NULL) {
   1482 		m_freem(m);
   1483 		return ENETDOWN;
   1484 	}
   1485 
   1486 	p = mib->ifvm_p;
   1487 	ec = (void *)mib->ifvm_p;
   1488 
   1489 	bpf_mtap(ifp, m, BPF_D_OUT);
   1490 
   1491 	if ((error = pfil_run_hooks(ifp->if_pfil, &m, ifp, PFIL_OUT)) != 0)
   1492 		goto out;
   1493 	if (m == NULL)
   1494 		goto out;
   1495 
   1496 	/*
   1497 	 * If the parent can insert the tag itself, just mark
   1498 	 * the tag in the mbuf header.
   1499 	 */
   1500 	if (ec->ec_capenable & ETHERCAP_VLAN_HWTAGGING) {
   1501 		vlan_set_tag(m, mib->ifvm_tag);
   1502 	} else {
   1503 		/*
   1504 		 * insert the tag ourselves
   1505 		 */
   1506 		M_PREPEND(m, mib->ifvm_encaplen, M_DONTWAIT);
   1507 		if (m == NULL) {
   1508 			printf("%s: unable to prepend encap header",
   1509 			    p->if_xname);
   1510 			if_statinc(ifp, if_oerrors);
   1511 			error = ENOBUFS;
   1512 			goto out;
   1513 		}
   1514 
   1515 		switch (p->if_type) {
   1516 		case IFT_ETHER:
   1517 		    {
   1518 			struct ether_vlan_header *evl;
   1519 
   1520 			if (m->m_len < sizeof(struct ether_vlan_header))
   1521 				m = m_pullup(m,
   1522 				    sizeof(struct ether_vlan_header));
   1523 			if (m == NULL) {
   1524 				printf("%s: unable to pullup encap "
   1525 				    "header", p->if_xname);
   1526 				if_statinc(ifp, if_oerrors);
   1527 				error = ENOBUFS;
   1528 				goto out;
   1529 			}
   1530 
   1531 			/*
   1532 			 * Transform the Ethernet header into an
   1533 			 * Ethernet header with 802.1Q encapsulation.
   1534 			 */
   1535 			memmove(mtod(m, void *),
   1536 			    mtod(m, char *) + mib->ifvm_encaplen,
   1537 			    sizeof(struct ether_header));
   1538 			evl = mtod(m, struct ether_vlan_header *);
   1539 			evl->evl_proto = evl->evl_encap_proto;
   1540 			evl->evl_encap_proto = htons(ETHERTYPE_VLAN);
   1541 			evl->evl_tag = htons(mib->ifvm_tag);
   1542 
   1543 			/*
   1544 			 * To cater for VLAN-aware layer 2 ethernet
   1545 			 * switches which may need to strip the tag
   1546 			 * before forwarding the packet, make sure
   1547 			 * the packet+tag is at least 68 bytes long.
   1548 			 * This is necessary because our parent will
   1549 			 * only pad to 64 bytes (ETHER_MIN_LEN) and
   1550 			 * some switches will not pad by themselves
   1551 			 * after deleting a tag.
   1552 			 */
   1553 			const size_t min_data_len = ETHER_MIN_LEN -
   1554 			    ETHER_CRC_LEN + ETHER_VLAN_ENCAP_LEN;
   1555 			if (m->m_pkthdr.len < min_data_len) {
   1556 				m_copyback(m, m->m_pkthdr.len,
   1557 				    min_data_len - m->m_pkthdr.len,
   1558 				    vlan_zero_pad_buff);
   1559 			}
   1560 			break;
   1561 		    }
   1562 
   1563 		default:
   1564 			panic("%s: impossible", __func__);
   1565 		}
   1566 	}
   1567 
   1568 	if ((p->if_flags & IFF_RUNNING) == 0) {
   1569 		m_freem(m);
   1570 		error = ENETDOWN;
   1571 		goto out;
   1572 	}
   1573 
   1574 	error = if_transmit_lock(p, m);
   1575 	net_stat_ref_t nsr = IF_STAT_GETREF(ifp);
   1576 	if (error) {
   1577 		/* mbuf is already freed */
   1578 		if_statinc_ref(nsr, if_oerrors);
   1579 	} else {
   1580 		if_statinc_ref(nsr, if_opackets);
   1581 		if_statadd_ref(nsr, if_obytes, pktlen);
   1582 		if (mcast)
   1583 			if_statinc_ref(nsr, if_omcasts);
   1584 	}
   1585 	IF_STAT_PUTREF(ifp);
   1586 
   1587 out:
   1588 	/* Remove reference to mib before release */
   1589 	vlan_putref_linkmib(mib, &psref);
   1590 	return error;
   1591 }
   1592 
   1593 /*
   1594  * Given an Ethernet frame, find a valid vlan interface corresponding to the
   1595  * given source interface and tag, then run the real packet through the
   1596  * parent's input routine.
   1597  */
   1598 void
   1599 vlan_input(struct ifnet *ifp, struct mbuf *m)
   1600 {
   1601 	struct ifvlan *ifv;
   1602 	uint16_t vid;
   1603 	struct ifvlan_linkmib *mib;
   1604 	struct psref psref;
   1605 	bool have_vtag;
   1606 
   1607 	have_vtag = vlan_has_tag(m);
   1608 	if (have_vtag) {
   1609 		vid = EVL_VLANOFTAG(vlan_get_tag(m));
   1610 		m->m_flags &= ~M_VLANTAG;
   1611 	} else {
   1612 		struct ether_vlan_header *evl;
   1613 
   1614 		if (ifp->if_type != IFT_ETHER) {
   1615 			panic("%s: impossible", __func__);
   1616 		}
   1617 
   1618 		if (m->m_len < sizeof(struct ether_vlan_header) &&
   1619 		    (m = m_pullup(m,
   1620 		     sizeof(struct ether_vlan_header))) == NULL) {
   1621 			printf("%s: no memory for VLAN header, "
   1622 			    "dropping packet.\n", ifp->if_xname);
   1623 			return;
   1624 		}
   1625 		evl = mtod(m, struct ether_vlan_header *);
   1626 		KASSERT(ntohs(evl->evl_encap_proto) == ETHERTYPE_VLAN);
   1627 
   1628 		vid = EVL_VLANOFTAG(ntohs(evl->evl_tag));
   1629 
   1630 		/*
   1631 		 * Restore the original ethertype.  We'll remove
   1632 		 * the encapsulation after we've found the vlan
   1633 		 * interface corresponding to the tag.
   1634 		 */
   1635 		evl->evl_encap_proto = evl->evl_proto;
   1636 	}
   1637 
   1638 	mib = vlan_lookup_tag_psref(ifp, vid, &psref);
   1639 	if (mib == NULL) {
   1640 		m_freem(m);
   1641 		if_statinc(ifp, if_noproto);
   1642 		return;
   1643 	}
   1644 	KASSERT(mib->ifvm_encaplen == ETHER_VLAN_ENCAP_LEN);
   1645 
   1646 	ifv = mib->ifvm_ifvlan;
   1647 	if ((ifv->ifv_if.if_flags & (IFF_UP | IFF_RUNNING)) !=
   1648 	    (IFF_UP | IFF_RUNNING)) {
   1649 		m_freem(m);
   1650 		if_statinc(ifp, if_noproto);
   1651 		goto out;
   1652 	}
   1653 
   1654 	/*
   1655 	 * Now, remove the encapsulation header.  The original
   1656 	 * header has already been fixed up above.
   1657 	 */
   1658 	if (!have_vtag) {
   1659 		memmove(mtod(m, char *) + mib->ifvm_encaplen,
   1660 		    mtod(m, void *), sizeof(struct ether_header));
   1661 		m_adj(m, mib->ifvm_encaplen);
   1662 	}
   1663 
   1664 	m_set_rcvif(m, &ifv->ifv_if);
   1665 
   1666 	if (pfil_run_hooks(ifp->if_pfil, &m, ifp, PFIL_IN) != 0)
   1667 		goto out;
   1668 	if (m == NULL)
   1669 		goto out;
   1670 
   1671 	m->m_flags &= ~M_PROMISC;
   1672 	if_input(&ifv->ifv_if, m);
   1673 out:
   1674 	vlan_putref_linkmib(mib, &psref);
   1675 }
   1676 
   1677 /*
   1678  * Module infrastructure
   1679  */
   1680 #include "if_module.h"
   1681 
   1682 IF_MODULE(MODULE_CLASS_DRIVER, vlan, NULL)
   1683