Page MenuHomeFreeBSD
Files F135576657

No OneTemporary
Actions

View Options

Index: projects/arpv2_merge_1/sys/net/if_atmsubr.c
===================================================================
--- projects/arpv2_merge_1/sys/net/if_atmsubr.c (revision 185999)
+++ projects/arpv2_merge_1/sys/net/if_atmsubr.c (revision 186000)
@@ -1,514 +1,503 @@
/* $NetBSD: if_atmsubr.c,v 1.10 1997/03/11 23:19:51 chuck Exp $ */
/*-
*
* Copyright (c) 1996 Charles D. Cranor and Washington University.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by Charles D. Cranor and
* Washington University.
* 4. The name of the author may not be used to endorse or promote products
* derived from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
* IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
* OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
* IN NO EVENT SHALL THE AUTHOR 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_atmsubr.c
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include "opt_inet.h"
#include "opt_inet6.h"
#include "opt_mac.h"
#include "opt_natm.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/module.h>
#include <sys/mbuf.h>
#include <sys/socket.h>
#include <sys/sockio.h>
#include <sys/errno.h>
#include <sys/sysctl.h>
#include <sys/malloc.h>
#include <net/if.h>
#include <net/netisr.h>
#include <net/route.h>
#include <net/if_dl.h>
#include <net/if_types.h>
#include <net/if_atm.h>
#include <netinet/in.h>
#include <netinet/if_atm.h>
#include <netinet/if_ether.h> /* XXX: for ETHERTYPE_* */
#if defined(INET) || defined(INET6)
#include <netinet/in_var.h>
#endif
#ifdef NATM
#include <netnatm/natm.h>
#endif
#include <security/mac/mac_framework.h>
/*
* Netgraph interface functions.
* These need not be protected by a lock, because ng_atm nodes are persitent.
* The ng_atm module can be unloaded only if all ATM interfaces have been
* unloaded, so nobody should be in the code paths accessing these function
* pointers.
*/
void (*ng_atm_attach_p)(struct ifnet *);
void (*ng_atm_detach_p)(struct ifnet *);
int (*ng_atm_output_p)(struct ifnet *, struct mbuf **);
void (*ng_atm_input_p)(struct ifnet *, struct mbuf **,
struct atm_pseudohdr *, void *);
void (*ng_atm_input_orphan_p)(struct ifnet *, struct mbuf *,
struct atm_pseudohdr *, void *);
void (*ng_atm_event_p)(struct ifnet *, uint32_t, void *);
/*
* Harp pseudo interface hooks
*/
void (*atm_harp_input_p)(struct ifnet *ifp, struct mbuf **m,
struct atm_pseudohdr *ah, void *rxhand);
void (*atm_harp_attach_p)(struct ifnet *);
void (*atm_harp_detach_p)(struct ifnet *);
void (*atm_harp_event_p)(struct ifnet *, uint32_t, void *);
SYSCTL_NODE(_hw, OID_AUTO, atm, CTLFLAG_RW, 0, "ATM hardware");
MALLOC_DEFINE(M_IFATM, "ifatm", "atm interface internals");
#ifndef ETHERTYPE_IPV6
#define ETHERTYPE_IPV6 0x86dd
#endif
#define senderr(e) do { error = (e); goto bad; } while (0)
/*
* atm_output: ATM output routine
* inputs:
* "ifp" = ATM interface to output to
* "m0" = the packet to output
* "dst" = the sockaddr to send to (either IP addr, or raw VPI/VCI)
* "rt0" = the route to use
* returns: error code [0 == ok]
*
* note: special semantic: if (dst == NULL) then we assume "m" already
* has an atm_pseudohdr on it and just send it directly.
* [for native mode ATM output] if dst is null, then
* rt0 must also be NULL.
*/
int
atm_output(struct ifnet *ifp, struct mbuf *m0, struct sockaddr *dst,
struct rtentry *rt0)
{
u_int16_t etype = 0; /* if using LLC/SNAP */
int error = 0, sz;
struct atm_pseudohdr atmdst, *ad;
struct mbuf *m = m0;
struct atmllc *atmllc;
struct atmllc *llc_hdr = NULL;
u_int32_t atm_flags;
#ifdef MAC
error = mac_ifnet_check_transmit(ifp, m);
if (error)
senderr(error);
#endif
if (!((ifp->if_flags & IFF_UP) &&
(ifp->if_drv_flags & IFF_DRV_RUNNING)))
senderr(ENETDOWN);
/*
* check for non-native ATM traffic (dst != NULL)
*/
if (dst) {
switch (dst->sa_family) {
#if defined(INET) || defined(INET6)
case AF_INET:
case AF_INET6:
{
- struct rtentry *rt = NULL;
- /*
- * check route
- */
- if (rt0 != NULL) {
- error = rt_check(&rt, &rt0, dst);
- if (error)
- goto bad;
- RT_UNLOCK(rt);
- }
-
if (dst->sa_family == AF_INET6)
etype = ETHERTYPE_IPV6;
else
etype = ETHERTYPE_IP;
- if (!atmresolve(rt, m, dst, &atmdst)) {
+ if (!atmresolve(rt0, m, dst, &atmdst)) {
m = NULL;
/* XXX: atmresolve already free'd it */
senderr(EHOSTUNREACH);
/* XXX: put ATMARP stuff here */
/* XXX: watch who frees m on failure */
}
}
break;
#endif /* INET || INET6 */
case AF_UNSPEC:
/*
* XXX: bpfwrite. assuming dst contains 12 bytes
* (atm pseudo header (4) + LLC/SNAP (8))
*/
bcopy(dst->sa_data, &atmdst, sizeof(atmdst));
llc_hdr = (struct atmllc *)(dst->sa_data +
sizeof(atmdst));
break;
default:
printf("%s: can't handle af%d\n", ifp->if_xname,
dst->sa_family);
senderr(EAFNOSUPPORT);
}
/*
* must add atm_pseudohdr to data
*/
sz = sizeof(atmdst);
atm_flags = ATM_PH_FLAGS(&atmdst);
if (atm_flags & ATM_PH_LLCSNAP)
sz += 8; /* sizeof snap == 8 */
M_PREPEND(m, sz, M_DONTWAIT);
if (m == 0)
senderr(ENOBUFS);
ad = mtod(m, struct atm_pseudohdr *);
*ad = atmdst;
if (atm_flags & ATM_PH_LLCSNAP) {
atmllc = (struct atmllc *)(ad + 1);
if (llc_hdr == NULL) {
bcopy(ATMLLC_HDR, atmllc->llchdr,
sizeof(atmllc->llchdr));
/* note: in host order */
ATM_LLC_SETTYPE(atmllc, etype);
}
else
bcopy(llc_hdr, atmllc, sizeof(struct atmllc));
}
}
if (ng_atm_output_p != NULL) {
if ((error = (*ng_atm_output_p)(ifp, &m)) != 0) {
if (m != NULL)
m_freem(m);
return (error);
}
if (m == NULL)
return (0);
}
/*
* Queue message on interface, and start output if interface
* not yet active.
*/
if (!IF_HANDOFF_ADJ(&ifp->if_snd, m, ifp,
-(int)sizeof(struct atm_pseudohdr)))
return (ENOBUFS);
return (error);
bad:
if (m)
m_freem(m);
return (error);
}
/*
* Process a received ATM packet;
* the packet is in the mbuf chain m.
*/
void
atm_input(struct ifnet *ifp, struct atm_pseudohdr *ah, struct mbuf *m,
void *rxhand)
{
int isr;
u_int16_t etype = ETHERTYPE_IP; /* default */
if ((ifp->if_flags & IFF_UP) == 0) {
m_freem(m);
return;
}
#ifdef MAC
mac_ifnet_create_mbuf(ifp, m);
#endif
ifp->if_ibytes += m->m_pkthdr.len;
if (ng_atm_input_p != NULL) {
(*ng_atm_input_p)(ifp, &m, ah, rxhand);
if (m == NULL)
return;
}
/* not eaten by ng_atm. Maybe it's a pseudo-harp PDU? */
if (atm_harp_input_p != NULL) {
(*atm_harp_input_p)(ifp, &m, ah, rxhand);
if (m == NULL)
return;
}
if (rxhand) {
#ifdef NATM
struct natmpcb *npcb;
/*
* XXXRW: this use of 'rxhand' is not a very good idea, and
* was subject to races even before SMPng due to the release
* of spl here.
*/
NATM_LOCK();
npcb = rxhand;
npcb->npcb_inq++; /* count # in queue */
isr = NETISR_NATM;
m->m_pkthdr.rcvif = rxhand; /* XXX: overload */
NATM_UNLOCK();
#else
printf("atm_input: NATM detected but not "
"configured in kernel\n");
goto dropit;
#endif
} else {
/*
* handle LLC/SNAP header, if present
*/
if (ATM_PH_FLAGS(ah) & ATM_PH_LLCSNAP) {
struct atmllc *alc;
if (m->m_len < sizeof(*alc) &&
(m = m_pullup(m, sizeof(*alc))) == 0)
return; /* failed */
alc = mtod(m, struct atmllc *);
if (bcmp(alc, ATMLLC_HDR, 6)) {
printf("%s: recv'd invalid LLC/SNAP frame "
"[vp=%d,vc=%d]\n", ifp->if_xname,
ATM_PH_VPI(ah), ATM_PH_VCI(ah));
m_freem(m);
return;
}
etype = ATM_LLC_TYPE(alc);
m_adj(m, sizeof(*alc));
}
switch (etype) {
#ifdef INET
case ETHERTYPE_IP:
isr = NETISR_IP;
break;
#endif
#ifdef INET6
case ETHERTYPE_IPV6:
isr = NETISR_IPV6;
break;
#endif
default:
#ifndef NATM
dropit:
#endif
if (ng_atm_input_orphan_p != NULL)
(*ng_atm_input_orphan_p)(ifp, m, ah, rxhand);
else
m_freem(m);
return;
}
}
netisr_dispatch(isr, m);
}
/*
* Perform common duties while attaching to interface list.
*/
void
atm_ifattach(struct ifnet *ifp)
{
struct ifaddr *ifa;
struct sockaddr_dl *sdl;
struct ifatm *ifatm = ifp->if_l2com;
ifp->if_addrlen = 0;
ifp->if_hdrlen = 0;
if_attach(ifp);
ifp->if_mtu = ATMMTU;
ifp->if_output = atm_output;
#if 0
ifp->if_input = atm_input;
#endif
ifp->if_snd.ifq_maxlen = 50; /* dummy */
TAILQ_FOREACH(ifa, &ifp->if_addrhead, ifa_link)
if (ifa->ifa_addr->sa_family == AF_LINK) {
sdl = (struct sockaddr_dl *)ifa->ifa_addr;
sdl->sdl_type = IFT_ATM;
sdl->sdl_alen = ifp->if_addrlen;
#ifdef notyet /* if using ATMARP, store hardware address using the next line */
bcopy(ifp->hw_addr, LLADDR(sdl), ifp->if_addrlen);
#endif
break;
}
ifp->if_linkmib = &ifatm->mib;
ifp->if_linkmiblen = sizeof(ifatm->mib);
if(ng_atm_attach_p)
(*ng_atm_attach_p)(ifp);
if (atm_harp_attach_p)
(*atm_harp_attach_p)(ifp);
}
/*
* Common stuff for detaching an ATM interface
*/
void
atm_ifdetach(struct ifnet *ifp)
{
if (atm_harp_detach_p)
(*atm_harp_detach_p)(ifp);
if(ng_atm_detach_p)
(*ng_atm_detach_p)(ifp);
if_detach(ifp);
}
/*
* Support routine for the SIOCATMGVCCS ioctl().
*
* This routine assumes, that the private VCC structures used by the driver
* begin with a struct atmio_vcc.
*
* Return a table of VCCs in a freshly allocated memory area.
* Here we have a problem: we first count, how many vccs we need
* to return. The we allocate the memory and finally fill it in.
* Because we cannot lock while calling malloc, the number of active
* vccs may change while we're in malloc. So we allocate a couple of
* vccs more and if space anyway is not enough re-iterate.
*
* We could use an sx lock for the vcc tables.
*/
struct atmio_vcctable *
atm_getvccs(struct atmio_vcc **table, u_int size, u_int start,
struct mtx *lock, int waitok)
{
u_int cid, alloc;
size_t len;
struct atmio_vcctable *vccs;
struct atmio_vcc *v;
alloc = start + 10;
vccs = NULL;
for (;;) {
len = sizeof(*vccs) + alloc * sizeof(vccs->vccs[0]);
vccs = reallocf(vccs, len, M_TEMP,
waitok ? M_WAITOK : M_NOWAIT);
if (vccs == NULL)
return (NULL);
bzero(vccs, len);
vccs->count = 0;
v = vccs->vccs;
mtx_lock(lock);
for (cid = 0; cid < size; cid++)
if (table[cid] != NULL) {
if (++vccs->count == alloc)
/* too many - try again */
break;
*v++ = *table[cid];
}
mtx_unlock(lock);
if (cid == size)
break;
alloc *= 2;
}
return (vccs);
}
/*
* Driver or channel state has changed. Inform whoever is interested
* in these events.
*/
void
atm_event(struct ifnet *ifp, u_int event, void *arg)
{
if (ng_atm_event_p != NULL)
(*ng_atm_event_p)(ifp, event, arg);
if (atm_harp_event_p != NULL)
(*atm_harp_event_p)(ifp, event, arg);
}
static void *
atm_alloc(u_char type, struct ifnet *ifp)
{
struct ifatm *ifatm;
ifatm = malloc(sizeof(struct ifatm), M_IFATM, M_WAITOK | M_ZERO);
ifatm->ifp = ifp;
return (ifatm);
}
static void
atm_free(void *com, u_char type)
{
free(com, M_IFATM);
}
static int
atm_modevent(module_t mod, int type, void *data)
{
switch (type) {
case MOD_LOAD:
if_register_com_alloc(IFT_ATM, atm_alloc, atm_free);
break;
case MOD_UNLOAD:
if_deregister_com_alloc(IFT_ATM);
break;
default:
return (EOPNOTSUPP);
}
return (0);
}
static moduledata_t atm_mod = {
"atm",
atm_modevent,
0
};
DECLARE_MODULE(atm, atm_mod, SI_SUB_INIT_IF, SI_ORDER_ANY);
MODULE_VERSION(atm, 1);
Index: projects/arpv2_merge_1/sys/net/if_fwsubr.c
===================================================================
--- projects/arpv2_merge_1/sys/net/if_fwsubr.c (revision 185999)
+++ projects/arpv2_merge_1/sys/net/if_fwsubr.c (revision 186000)
@@ -1,858 +1,850 @@
/*-
* Copyright (c) 2004 Doug Rabson
* Copyright (c) 1982, 1989, 1993
* The Regents of the University of California. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* $FreeBSD$
*/
#include "opt_inet.h"
#include "opt_inet6.h"
#include "opt_mac.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/module.h>
#include <sys/socket.h>
#include <sys/sockio.h>
#include <net/if.h>
#include <net/netisr.h>
#include <net/route.h>
#include <net/if_llc.h>
#include <net/if_dl.h>
#include <net/if_types.h>
#include <net/bpf.h>
#include <net/firewire.h>
#include <net/if_llatbl.h>
#if defined(INET) || defined(INET6)
#include <netinet/in.h>
#include <netinet/in_var.h>
#include <netinet/if_ether.h>
#endif
#ifdef INET6
#include <netinet6/nd6.h>
#endif
#include <security/mac/mac_framework.h>
MALLOC_DEFINE(M_FWCOM, "fw_com", "firewire interface internals");
struct fw_hwaddr firewire_broadcastaddr = {
0xffffffff,
0xffffffff,
0xff,
0xff,
0xffff,
0xffffffff
};
static int
firewire_output(struct ifnet *ifp, struct mbuf *m, struct sockaddr *dst,
struct rtentry *rt0)
{
struct fw_com *fc = IFP2FWC(ifp);
int error, type;
- struct rtentry *rt = NULL;
struct m_tag *mtag;
union fw_encap *enc;
struct fw_hwaddr *destfw;
uint8_t speed;
uint16_t psize, fsize, dsize;
struct mbuf *mtail;
int unicast, dgl, foff;
static int next_dgl;
struct llentry *lle;
#ifdef MAC
error = mac_ifnet_check_transmit(ifp, m);
if (error)
goto bad;
#endif
if (!((ifp->if_flags & IFF_UP) &&
(ifp->if_drv_flags & IFF_DRV_RUNNING))) {
error = ENETDOWN;
goto bad;
}
- if (rt0 != NULL) {
- error = rt_check(&rt, &rt0, dst);
- if (error)
- goto bad;
- RT_UNLOCK(rt);
- }
-
/*
* For unicast, we make a tag to store the lladdr of the
* destination. This might not be the first time we have seen
* the packet (for instance, the arp code might be trying to
* re-send it after receiving an arp reply) so we only
* allocate a tag if there isn't one there already. For
* multicast, we will eventually use a different tag to store
* the channel number.
*/
unicast = !(m->m_flags & (M_BCAST | M_MCAST));
if (unicast) {
mtag = m_tag_locate(m, MTAG_FIREWIRE, MTAG_FIREWIRE_HWADDR, NULL);
if (!mtag) {
mtag = m_tag_alloc(MTAG_FIREWIRE, MTAG_FIREWIRE_HWADDR,
sizeof (struct fw_hwaddr), M_NOWAIT);
if (!mtag) {
error = ENOMEM;
goto bad;
}
m_tag_prepend(m, mtag);
}
destfw = (struct fw_hwaddr *)(mtag + 1);
} else {
destfw = 0;
}
switch (dst->sa_family) {
#ifdef INET
case AF_INET:
/*
* Only bother with arp for unicast. Allocation of
* channels etc. for firewire is quite different and
* doesn't fit into the arp model.
*/
if (unicast) {
- error = arpresolve(ifp, rt, m, dst, (u_char *) destfw, &lle);
+ error = arpresolve(ifp, rt0, m, dst, (u_char *) destfw, &lle);
if (error)
return (error == EWOULDBLOCK ? 0 : error);
}
type = ETHERTYPE_IP;
break;
case AF_ARP:
{
struct arphdr *ah;
ah = mtod(m, struct arphdr *);
ah->ar_hrd = htons(ARPHRD_IEEE1394);
type = ETHERTYPE_ARP;
if (unicast)
*destfw = *(struct fw_hwaddr *) ar_tha(ah);
/*
* The standard arp code leaves a hole for the target
* hardware address which we need to close up.
*/
bcopy(ar_tpa(ah), ar_tha(ah), ah->ar_pln);
m_adj(m, -ah->ar_hln);
break;
}
#endif
#ifdef INET6
case AF_INET6:
if (unicast) {
- error = nd6_storelladdr(fc->fc_ifp, rt, m, dst,
+ error = nd6_storelladdr(fc->fc_ifp, rt0, m, dst,
(u_char *) destfw, &lle);
if (error)
return (error);
}
type = ETHERTYPE_IPV6;
break;
#endif
default:
if_printf(ifp, "can't handle af%d\n", dst->sa_family);
error = EAFNOSUPPORT;
goto bad;
}
/*
* Let BPF tap off a copy before we encapsulate.
*/
if (bpf_peers_present(ifp->if_bpf)) {
struct fw_bpfhdr h;
if (unicast)
bcopy(destfw, h.firewire_dhost, 8);
else
bcopy(&firewire_broadcastaddr, h.firewire_dhost, 8);
bcopy(&fc->fc_hwaddr, h.firewire_shost, 8);
h.firewire_type = htons(type);
bpf_mtap2(ifp->if_bpf, &h, sizeof(h), m);
}
/*
* Punt on MCAP for now and send all multicast packets on the
* broadcast channel.
*/
if (m->m_flags & M_MCAST)
m->m_flags |= M_BCAST;
/*
* Figure out what speed to use and what the largest supported
* packet size is. For unicast, this is the minimum of what we
* can speak and what they can hear. For broadcast, lets be
* conservative and use S100. We could possibly improve that
* by examining the bus manager's speed map or similar. We
* also reduce the packet size for broadcast to account for
* the GASP header.
*/
if (unicast) {
speed = min(fc->fc_speed, destfw->sspd);
psize = min(512 << speed, 2 << destfw->sender_max_rec);
} else {
speed = 0;
psize = 512 - 2*sizeof(uint32_t);
}
/*
* Next, we encapsulate, possibly fragmenting the original
* datagram if it won't fit into a single packet.
*/
if (m->m_pkthdr.len <= psize - sizeof(uint32_t)) {
/*
* No fragmentation is necessary.
*/
M_PREPEND(m, sizeof(uint32_t), M_DONTWAIT);
if (!m) {
error = ENOBUFS;
goto bad;
}
enc = mtod(m, union fw_encap *);
enc->unfrag.ether_type = type;
enc->unfrag.lf = FW_ENCAP_UNFRAG;
enc->unfrag.reserved = 0;
/*
* Byte swap the encapsulation header manually.
*/
enc->ul[0] = htonl(enc->ul[0]);
error = (ifp->if_transmit)(ifp, m);
return (error);
} else {
/*
* Fragment the datagram, making sure to leave enough
* space for the encapsulation header in each packet.
*/
fsize = psize - 2*sizeof(uint32_t);
dgl = next_dgl++;
dsize = m->m_pkthdr.len;
foff = 0;
while (m) {
if (m->m_pkthdr.len > fsize) {
/*
* Split off the tail segment from the
* datagram, copying our tags over.
*/
mtail = m_split(m, fsize, M_DONTWAIT);
m_tag_copy_chain(mtail, m, M_NOWAIT);
} else {
mtail = 0;
}
/*
* Add our encapsulation header to this
* fragment and hand it off to the link.
*/
M_PREPEND(m, 2*sizeof(uint32_t), M_DONTWAIT);
if (!m) {
error = ENOBUFS;
goto bad;
}
enc = mtod(m, union fw_encap *);
if (foff == 0) {
enc->firstfrag.lf = FW_ENCAP_FIRST;
enc->firstfrag.reserved1 = 0;
enc->firstfrag.reserved2 = 0;
enc->firstfrag.datagram_size = dsize - 1;
enc->firstfrag.ether_type = type;
enc->firstfrag.dgl = dgl;
} else {
if (mtail)
enc->nextfrag.lf = FW_ENCAP_NEXT;
else
enc->nextfrag.lf = FW_ENCAP_LAST;
enc->nextfrag.reserved1 = 0;
enc->nextfrag.reserved2 = 0;
enc->nextfrag.reserved3 = 0;
enc->nextfrag.datagram_size = dsize - 1;
enc->nextfrag.fragment_offset = foff;
enc->nextfrag.dgl = dgl;
}
foff += m->m_pkthdr.len - 2*sizeof(uint32_t);
/*
* Byte swap the encapsulation header manually.
*/
enc->ul[0] = htonl(enc->ul[0]);
enc->ul[1] = htonl(enc->ul[1]);
error = (ifp->if_transmit)(ifp, m);
if (error) {
if (mtail)
m_freem(mtail);
return (ENOBUFS);
}
m = mtail;
}
return (0);
}
bad:
if (m)
m_freem(m);
return (error);
}
static struct mbuf *
firewire_input_fragment(struct fw_com *fc, struct mbuf *m, int src)
{
union fw_encap *enc;
struct fw_reass *r;
struct mbuf *mf, *mprev;
int dsize;
int fstart, fend, start, end, islast;
uint32_t id;
/*
* Find an existing reassembly buffer or create a new one.
*/
enc = mtod(m, union fw_encap *);
id = enc->firstfrag.dgl | (src << 16);
STAILQ_FOREACH(r, &fc->fc_frags, fr_link)
if (r->fr_id == id)
break;
if (!r) {
r = malloc(sizeof(struct fw_reass), M_TEMP, M_NOWAIT);
if (!r) {
m_freem(m);
return 0;
}
r->fr_id = id;
r->fr_frags = 0;
STAILQ_INSERT_HEAD(&fc->fc_frags, r, fr_link);
}
/*
* If this fragment overlaps any other fragment, we must discard
* the partial reassembly and start again.
*/
if (enc->firstfrag.lf == FW_ENCAP_FIRST)
fstart = 0;
else
fstart = enc->nextfrag.fragment_offset;
fend = fstart + m->m_pkthdr.len - 2*sizeof(uint32_t);
dsize = enc->nextfrag.datagram_size;
islast = (enc->nextfrag.lf == FW_ENCAP_LAST);
for (mf = r->fr_frags; mf; mf = mf->m_nextpkt) {
enc = mtod(mf, union fw_encap *);
if (enc->nextfrag.datagram_size != dsize) {
/*
* This fragment must be from a different
* packet.
*/
goto bad;
}
if (enc->firstfrag.lf == FW_ENCAP_FIRST)
start = 0;
else
start = enc->nextfrag.fragment_offset;
end = start + mf->m_pkthdr.len - 2*sizeof(uint32_t);
if ((fstart < end && fend > start) ||
(islast && enc->nextfrag.lf == FW_ENCAP_LAST)) {
/*
* Overlap - discard reassembly buffer and start
* again with this fragment.
*/
goto bad;
}
}
/*
* Find where to put this fragment in the list.
*/
for (mf = r->fr_frags, mprev = NULL; mf;
mprev = mf, mf = mf->m_nextpkt) {
enc = mtod(mf, union fw_encap *);
if (enc->firstfrag.lf == FW_ENCAP_FIRST)
start = 0;
else
start = enc->nextfrag.fragment_offset;
if (start >= fend)
break;
}
/*
* If this is a last fragment and we are not adding at the end
* of the list, discard the buffer.
*/
if (islast && mprev && mprev->m_nextpkt)
goto bad;
if (mprev) {
m->m_nextpkt = mprev->m_nextpkt;
mprev->m_nextpkt = m;
/*
* Coalesce forwards and see if we can make a whole
* datagram.
*/
enc = mtod(mprev, union fw_encap *);
if (enc->firstfrag.lf == FW_ENCAP_FIRST)
start = 0;
else
start = enc->nextfrag.fragment_offset;
end = start + mprev->m_pkthdr.len - 2*sizeof(uint32_t);
while (end == fstart) {
/*
* Strip off the encap header from m and
* append it to mprev, freeing m.
*/
m_adj(m, 2*sizeof(uint32_t));
mprev->m_nextpkt = m->m_nextpkt;
mprev->m_pkthdr.len += m->m_pkthdr.len;
m_cat(mprev, m);
if (mprev->m_pkthdr.len == dsize + 1 + 2*sizeof(uint32_t)) {
/*
* We have assembled a complete packet
* we must be finished. Make sure we have
* merged the whole chain.
*/
STAILQ_REMOVE(&fc->fc_frags, r, fw_reass, fr_link);
free(r, M_TEMP);
m = mprev->m_nextpkt;
while (m) {
mf = m->m_nextpkt;
m_freem(m);
m = mf;
}
mprev->m_nextpkt = NULL;
return (mprev);
}
/*
* See if we can continue merging forwards.
*/
end = fend;
m = mprev->m_nextpkt;
if (m) {
enc = mtod(m, union fw_encap *);
if (enc->firstfrag.lf == FW_ENCAP_FIRST)
fstart = 0;
else
fstart = enc->nextfrag.fragment_offset;
fend = fstart + m->m_pkthdr.len
- 2*sizeof(uint32_t);
} else {
break;
}
}
} else {
m->m_nextpkt = 0;
r->fr_frags = m;
}
return (0);
bad:
while (r->fr_frags) {
mf = r->fr_frags;
r->fr_frags = mf->m_nextpkt;
m_freem(mf);
}
m->m_nextpkt = 0;
r->fr_frags = m;
return (0);
}
void
firewire_input(struct ifnet *ifp, struct mbuf *m, uint16_t src)
{
struct fw_com *fc = IFP2FWC(ifp);
union fw_encap *enc;
int type, isr;
/*
* The caller has already stripped off the packet header
* (stream or wreqb) and marked the mbuf's M_BCAST flag
* appropriately. We de-encapsulate the IP packet and pass it
* up the line after handling link-level fragmentation.
*/
if (m->m_pkthdr.len < sizeof(uint32_t)) {
if_printf(ifp, "discarding frame without "
"encapsulation header (len %u pkt len %u)\n",
m->m_len, m->m_pkthdr.len);
}
m = m_pullup(m, sizeof(uint32_t));
if (m == NULL)
return;
enc = mtod(m, union fw_encap *);
/*
* Byte swap the encapsulation header manually.
*/
enc->ul[0] = ntohl(enc->ul[0]);
if (enc->unfrag.lf != 0) {
m = m_pullup(m, 2*sizeof(uint32_t));
if (!m)
return;
enc = mtod(m, union fw_encap *);
enc->ul[1] = ntohl(enc->ul[1]);
m = firewire_input_fragment(fc, m, src);
if (!m)
return;
enc = mtod(m, union fw_encap *);
type = enc->firstfrag.ether_type;
m_adj(m, 2*sizeof(uint32_t));
} else {
type = enc->unfrag.ether_type;
m_adj(m, sizeof(uint32_t));
}
if (m->m_pkthdr.rcvif == NULL) {
if_printf(ifp, "discard frame w/o interface pointer\n");
ifp->if_ierrors++;
m_freem(m);
return;
}
#ifdef DIAGNOSTIC
if (m->m_pkthdr.rcvif != ifp) {
if_printf(ifp, "Warning, frame marked as received on %s\n",
m->m_pkthdr.rcvif->if_xname);
}
#endif
#ifdef MAC
/*
* Tag the mbuf with an appropriate MAC label before any other
* consumers can get to it.
*/
mac_ifnet_create_mbuf(ifp, m);
#endif
/*
* Give bpf a chance at the packet. The link-level driver
* should have left us a tag with the EUID of the sender.
*/
if (bpf_peers_present(ifp->if_bpf)) {
struct fw_bpfhdr h;
struct m_tag *mtag;
mtag = m_tag_locate(m, MTAG_FIREWIRE, MTAG_FIREWIRE_SENDER_EUID, 0);
if (mtag)
bcopy(mtag + 1, h.firewire_shost, 8);
else
bcopy(&firewire_broadcastaddr, h.firewire_dhost, 8);
bcopy(&fc->fc_hwaddr, h.firewire_dhost, 8);
h.firewire_type = htons(type);
bpf_mtap2(ifp->if_bpf, &h, sizeof(h), m);
}
if (ifp->if_flags & IFF_MONITOR) {
/*
* Interface marked for monitoring; discard packet.
*/
m_freem(m);
return;
}
ifp->if_ibytes += m->m_pkthdr.len;
/* Discard packet if interface is not up */
if ((ifp->if_flags & IFF_UP) == 0) {
m_freem(m);
return;
}
if (m->m_flags & (M_BCAST|M_MCAST))
ifp->if_imcasts++;
switch (type) {
#ifdef INET
case ETHERTYPE_IP:
if ((m = ip_fastforward(m)) == NULL)
return;
isr = NETISR_IP;
break;
case ETHERTYPE_ARP:
{
struct arphdr *ah;
ah = mtod(m, struct arphdr *);
/*
* Adjust the arp packet to insert an empty tha slot.
*/
m->m_len += ah->ar_hln;
m->m_pkthdr.len += ah->ar_hln;
bcopy(ar_tha(ah), ar_tpa(ah), ah->ar_pln);
isr = NETISR_ARP;
break;
}
#endif
#ifdef INET6
case ETHERTYPE_IPV6:
isr = NETISR_IPV6;
break;
#endif
default:
m_freem(m);
return;
}
netisr_dispatch(isr, m);
}
int
firewire_ioctl(struct ifnet *ifp, int command, caddr_t data)
{
struct ifaddr *ifa = (struct ifaddr *) data;
struct ifreq *ifr = (struct ifreq *) data;
int error = 0;
switch (command) {
case SIOCSIFADDR:
ifp->if_flags |= IFF_UP;
switch (ifa->ifa_addr->sa_family) {
#ifdef INET
case AF_INET:
ifp->if_init(ifp->if_softc); /* before arpwhohas */
arp_ifinit(ifp, ifa);
break;
#endif
default:
ifp->if_init(ifp->if_softc);
break;
}
break;
case SIOCGIFADDR:
{
struct sockaddr *sa;
sa = (struct sockaddr *) & ifr->ifr_data;
bcopy(&IFP2FWC(ifp)->fc_hwaddr,
(caddr_t) sa->sa_data, sizeof(struct fw_hwaddr));
}
break;
case SIOCSIFMTU:
/*
* Set the interface MTU.
*/
if (ifr->ifr_mtu > 1500) {
error = EINVAL;
} else {
ifp->if_mtu = ifr->ifr_mtu;
}
break;
default:
error = EINVAL; /* XXX netbsd has ENOTTY??? */
break;
}
return (error);
}
static int
firewire_resolvemulti(struct ifnet *ifp, struct sockaddr **llsa,
struct sockaddr *sa)
{
#ifdef INET
struct sockaddr_in *sin;
#endif
#ifdef INET6
struct sockaddr_in6 *sin6;
#endif
switch(sa->sa_family) {
case AF_LINK:
/*
* No mapping needed.
*/
*llsa = 0;
return 0;
#ifdef INET
case AF_INET:
sin = (struct sockaddr_in *)sa;
if (!IN_MULTICAST(ntohl(sin->sin_addr.s_addr)))
return EADDRNOTAVAIL;
*llsa = 0;
return 0;
#endif
#ifdef INET6
case AF_INET6:
sin6 = (struct sockaddr_in6 *)sa;
if (IN6_IS_ADDR_UNSPECIFIED(&sin6->sin6_addr)) {
/*
* An IP6 address of 0 means listen to all
* of the Ethernet multicast address used for IP6.
* (This is used for multicast routers.)
*/
ifp->if_flags |= IFF_ALLMULTI;
*llsa = 0;
return 0;
}
if (!IN6_IS_ADDR_MULTICAST(&sin6->sin6_addr))
return EADDRNOTAVAIL;
*llsa = 0;
return 0;
#endif
default:
/*
* Well, the text isn't quite right, but it's the name
* that counts...
*/
return EAFNOSUPPORT;
}
}
void
firewire_ifattach(struct ifnet *ifp, struct fw_hwaddr *llc)
{
struct fw_com *fc = IFP2FWC(ifp);
struct ifaddr *ifa;
struct sockaddr_dl *sdl;
static const char* speeds[] = {
"S100", "S200", "S400", "S800",
"S1600", "S3200"
};
fc->fc_speed = llc->sspd;
STAILQ_INIT(&fc->fc_frags);
ifp->if_addrlen = sizeof(struct fw_hwaddr);
ifp->if_hdrlen = 0;
if_attach(ifp);
ifp->if_mtu = 1500; /* XXX */
ifp->if_output = firewire_output;
ifp->if_resolvemulti = firewire_resolvemulti;
ifp->if_broadcastaddr = (u_char *) &firewire_broadcastaddr;
ifa = ifp->if_addr;
KASSERT(ifa != NULL, ("%s: no lladdr!\n", __func__));
sdl = (struct sockaddr_dl *)ifa->ifa_addr;
sdl->sdl_type = IFT_IEEE1394;
sdl->sdl_alen = ifp->if_addrlen;
bcopy(llc, LLADDR(sdl), ifp->if_addrlen);
bpfattach(ifp, DLT_APPLE_IP_OVER_IEEE1394,
sizeof(struct fw_hwaddr));
if_printf(ifp, "Firewire address: %8D @ 0x%04x%08x, %s, maxrec %d\n",
(uint8_t *) &llc->sender_unique_ID_hi, ":",
ntohs(llc->sender_unicast_FIFO_hi),
ntohl(llc->sender_unicast_FIFO_lo),
speeds[llc->sspd],
(2 << llc->sender_max_rec));
}
void
firewire_ifdetach(struct ifnet *ifp)
{
bpfdetach(ifp);
if_detach(ifp);
}
void
firewire_busreset(struct ifnet *ifp)
{
struct fw_com *fc = IFP2FWC(ifp);
struct fw_reass *r;
struct mbuf *m;
/*
* Discard any partial datagrams since the host ids may have changed.
*/
while ((r = STAILQ_FIRST(&fc->fc_frags))) {
STAILQ_REMOVE_HEAD(&fc->fc_frags, fr_link);
while (r->fr_frags) {
m = r->fr_frags;
r->fr_frags = m->m_nextpkt;
m_freem(m);
}
free(r, M_TEMP);
}
}
static void *
firewire_alloc(u_char type, struct ifnet *ifp)
{
struct fw_com *fc;
fc = malloc(sizeof(struct fw_com), M_FWCOM, M_WAITOK | M_ZERO);
fc->fc_ifp = ifp;
return (fc);
}
static void
firewire_free(void *com, u_char type)
{
free(com, M_FWCOM);
}
static int
firewire_modevent(module_t mod, int type, void *data)
{
switch (type) {
case MOD_LOAD:
if_register_com_alloc(IFT_IEEE1394,
firewire_alloc, firewire_free);
break;
case MOD_UNLOAD:
if_deregister_com_alloc(IFT_IEEE1394);
break;
default:
return (EOPNOTSUPP);
}
return (0);
}
static moduledata_t firewire_mod = {
"if_firewire",
firewire_modevent,
0
};
DECLARE_MODULE(if_firewire, firewire_mod, SI_SUB_INIT_IF, SI_ORDER_ANY);
MODULE_VERSION(if_firewire, 1);
Index: projects/arpv2_merge_1/sys/net/if_iso88025subr.c
===================================================================
--- projects/arpv2_merge_1/sys/net/if_iso88025subr.c (revision 185999)
+++ projects/arpv2_merge_1/sys/net/if_iso88025subr.c (revision 186000)
@@ -1,831 +1,824 @@
/*-
* Copyright (c) 1998, Larry Lile
* All rights reserved.
*
* For latest sources and information on this driver, please
* go to http://anarchy.stdio.com.
*
* Questions, comments or suggestions should be directed to
* Larry Lile <lile@stdio.com>.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice unmodified, this list of conditions, and the following
* disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* $FreeBSD$
*
*/
/*
*
* General ISO 802.5 (Token Ring) support routines
*
*/
#include "opt_inet.h"
#include "opt_inet6.h"
#include "opt_ipx.h"
#include "opt_mac.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/module.h>
#include <sys/socket.h>
#include <sys/sockio.h>
#include <net/if.h>
#include <net/if_arp.h>
#include <net/if_dl.h>
#include <net/if_llc.h>
#include <net/if_types.h>
#include <net/if_llatbl.h>
#include <net/ethernet.h>
#include <net/netisr.h>
#include <net/route.h>
#include <net/bpf.h>
#include <net/iso88025.h>
#if defined(INET) || defined(INET6)
#include <netinet/in.h>
#include <netinet/in_var.h>
#include <netinet/if_ether.h>
#endif
#ifdef INET6
#include <netinet6/nd6.h>
#endif
#ifdef IPX
#include <netipx/ipx.h>
#include <netipx/ipx_if.h>
#endif
#include <security/mac/mac_framework.h>
static const u_char iso88025_broadcastaddr[ISO88025_ADDR_LEN] =
{ 0xff, 0xff, 0xff, 0xff, 0xff, 0xff };
static int iso88025_resolvemulti (struct ifnet *, struct sockaddr **,
struct sockaddr *);
#define senderr(e) do { error = (e); goto bad; } while (0)
/*
* Perform common duties while attaching to interface list
*/
void
iso88025_ifattach(struct ifnet *ifp, const u_int8_t *lla, int bpf)
{
struct ifaddr *ifa;
struct sockaddr_dl *sdl;
ifa = NULL;
ifp->if_type = IFT_ISO88025;
ifp->if_addrlen = ISO88025_ADDR_LEN;
ifp->if_hdrlen = ISO88025_HDR_LEN;
if_attach(ifp); /* Must be called before additional assignments */
ifp->if_output = iso88025_output;
ifp->if_input = iso88025_input;
ifp->if_resolvemulti = iso88025_resolvemulti;
ifp->if_broadcastaddr = iso88025_broadcastaddr;
if (ifp->if_baudrate == 0)
ifp->if_baudrate = TR_16MBPS; /* 16Mbit should be a safe default */
if (ifp->if_mtu == 0)
ifp->if_mtu = ISO88025_DEFAULT_MTU;
ifa = ifp->if_addr;
KASSERT(ifa != NULL, ("%s: no lladdr!\n", __func__));
sdl = (struct sockaddr_dl *)ifa->ifa_addr;
sdl->sdl_type = IFT_ISO88025;
sdl->sdl_alen = ifp->if_addrlen;
bcopy(lla, LLADDR(sdl), ifp->if_addrlen);
if (bpf)
bpfattach(ifp, DLT_IEEE802, ISO88025_HDR_LEN);
return;
}
/*
* Perform common duties while detaching a Token Ring interface
*/
void
iso88025_ifdetach(ifp, bpf)
struct ifnet *ifp;
int bpf;
{
if (bpf)
bpfdetach(ifp);
if_detach(ifp);
return;
}
int
iso88025_ioctl(struct ifnet *ifp, int command, caddr_t data)
{
struct ifaddr *ifa;
struct ifreq *ifr;
int error;
ifa = (struct ifaddr *) data;
ifr = (struct ifreq *) data;
error = 0;
switch (command) {
case SIOCSIFADDR:
ifp->if_flags |= IFF_UP;
switch (ifa->ifa_addr->sa_family) {
#ifdef INET
case AF_INET:
ifp->if_init(ifp->if_softc); /* before arpwhohas */
arp_ifinit(ifp, ifa);
break;
#endif /* INET */
#ifdef IPX
/*
* XXX - This code is probably wrong
*/
case AF_IPX: {
struct ipx_addr *ina;
ina = &(IA_SIPX(ifa)->sipx_addr);
if (ipx_nullhost(*ina))
ina->x_host = *(union ipx_host *)
IF_LLADDR(ifp);
else
bcopy((caddr_t) ina->x_host.c_host,
(caddr_t) IF_LLADDR(ifp),
ISO88025_ADDR_LEN);
/*
* Set new address
*/
ifp->if_init(ifp->if_softc);
}
break;
#endif /* IPX */
default:
ifp->if_init(ifp->if_softc);
break;
}
break;
case SIOCGIFADDR: {
struct sockaddr *sa;
sa = (struct sockaddr *) & ifr->ifr_data;
bcopy(IF_LLADDR(ifp),
(caddr_t) sa->sa_data, ISO88025_ADDR_LEN);
}
break;
case SIOCSIFMTU:
/*
* Set the interface MTU.
*/
if (ifr->ifr_mtu > ISO88025_MAX_MTU) {
error = EINVAL;
} else {
ifp->if_mtu = ifr->ifr_mtu;
}
break;
default:
error = EINVAL; /* XXX netbsd has ENOTTY??? */
break;
}
return (error);
}
/*
* ISO88025 encapsulation
*/
int
iso88025_output(ifp, m, dst, rt0)
struct ifnet *ifp;
struct mbuf *m;
struct sockaddr *dst;
struct rtentry *rt0;
{
u_int16_t snap_type = 0;
int loop_copy = 0, error = 0, rif_len = 0;
u_char edst[ISO88025_ADDR_LEN];
struct iso88025_header *th;
struct iso88025_header gen_th;
struct sockaddr_dl *sdl = NULL;
- struct rtentry *rt = NULL;
struct llentry *lle;
#ifdef MAC
error = mac_ifnet_check_transmit(ifp, m);
if (error)
senderr(error);
#endif
if (ifp->if_flags & IFF_MONITOR)
senderr(ENETDOWN);
if (!((ifp->if_flags & IFF_UP) &&
(ifp->if_drv_flags & IFF_DRV_RUNNING)))
senderr(ENETDOWN);
getmicrotime(&ifp->if_lastchange);
/* Calculate routing info length based on arp table entry */
/* XXX any better way to do this ? */
- if (rt0 != NULL) {
- error = rt_check(&rt, &rt0, dst);
- if (error)
- goto bad;
- RT_UNLOCK(rt);
- }
- if (rt && (sdl = (struct sockaddr_dl *)rt->rt_gateway))
+ if (rt0 && (sdl = (struct sockaddr_dl *)rt0->rt_gateway))
if (SDL_ISO88025(sdl)->trld_rcf != 0)
rif_len = TR_RCF_RIFLEN(SDL_ISO88025(sdl)->trld_rcf);
/* Generate a generic 802.5 header for the packet */
gen_th.ac = TR_AC;
gen_th.fc = TR_LLC_FRAME;
(void)memcpy((caddr_t)gen_th.iso88025_shost, IF_LLADDR(ifp),
ISO88025_ADDR_LEN);
if (rif_len) {
gen_th.iso88025_shost[0] |= TR_RII;
if (rif_len > 2) {
gen_th.rcf = SDL_ISO88025(sdl)->trld_rcf;
(void)memcpy((caddr_t)gen_th.rd,
(caddr_t)SDL_ISO88025(sdl)->trld_route,
rif_len - 2);
}
}
switch (dst->sa_family) {
#ifdef INET
case AF_INET:
error = arpresolve(ifp, rt0, m, dst, edst, &lle);
if (error)
return (error == EWOULDBLOCK ? 0 : error);
snap_type = ETHERTYPE_IP;
break;
case AF_ARP:
{
struct arphdr *ah;
ah = mtod(m, struct arphdr *);
ah->ar_hrd = htons(ARPHRD_IEEE802);
loop_copy = -1; /* if this is for us, don't do it */
switch(ntohs(ah->ar_op)) {
case ARPOP_REVREQUEST:
case ARPOP_REVREPLY:
snap_type = ETHERTYPE_REVARP;
break;
case ARPOP_REQUEST:
case ARPOP_REPLY:
default:
snap_type = ETHERTYPE_ARP;
break;
}
if (m->m_flags & M_BCAST)
bcopy(ifp->if_broadcastaddr, edst, ISO88025_ADDR_LEN);
else
bcopy(ar_tha(ah), edst, ISO88025_ADDR_LEN);
}
break;
#endif /* INET */
#ifdef INET6
case AF_INET6:
error = nd6_storelladdr(ifp, rt0, m, dst, (u_char *)edst, &lle);
if (error)
return (error);
snap_type = ETHERTYPE_IPV6;
break;
#endif /* INET6 */
#ifdef IPX
case AF_IPX:
{
u_int8_t *cp;
bcopy((caddr_t)&(satoipx_addr(dst).x_host), (caddr_t)edst,
ISO88025_ADDR_LEN);
M_PREPEND(m, 3, M_WAIT);
m = m_pullup(m, 3);
if (m == 0)
senderr(ENOBUFS);
cp = mtod(m, u_int8_t *);
*cp++ = ETHERTYPE_IPX_8022;
*cp++ = ETHERTYPE_IPX_8022;
*cp++ = LLC_UI;
}
break;
#endif /* IPX */
case AF_UNSPEC:
{
struct iso88025_sockaddr_data *sd;
/*
* For AF_UNSPEC sockaddr.sa_data must contain all of the
* mac information needed to send the packet. This allows
* full mac, llc, and source routing function to be controlled.
* llc and source routing information must already be in the
* mbuf provided, ac/fc are set in sa_data. sockaddr.sa_data
* should be an iso88025_sockaddr_data structure see iso88025.h
*/
loop_copy = -1;
sd = (struct iso88025_sockaddr_data *)dst->sa_data;
gen_th.ac = sd->ac;
gen_th.fc = sd->fc;
(void)memcpy((caddr_t)edst, (caddr_t)sd->ether_dhost,
ISO88025_ADDR_LEN);
(void)memcpy((caddr_t)gen_th.iso88025_shost,
(caddr_t)sd->ether_shost, ISO88025_ADDR_LEN);
rif_len = 0;
break;
}
default:
if_printf(ifp, "can't handle af%d\n", dst->sa_family);
senderr(EAFNOSUPPORT);
break;
}
/*
* Add LLC header.
*/
if (snap_type != 0) {
struct llc *l;
M_PREPEND(m, LLC_SNAPFRAMELEN, M_DONTWAIT);
if (m == 0)
senderr(ENOBUFS);
l = mtod(m, struct llc *);
l->llc_control = LLC_UI;
l->llc_dsap = l->llc_ssap = LLC_SNAP_LSAP;
l->llc_snap.org_code[0] =
l->llc_snap.org_code[1] =
l->llc_snap.org_code[2] = 0;
l->llc_snap.ether_type = htons(snap_type);
}
/*
* Add local net header. If no space in first mbuf,
* allocate another.
*/
M_PREPEND(m, ISO88025_HDR_LEN + rif_len, M_DONTWAIT);
if (m == 0)
senderr(ENOBUFS);
th = mtod(m, struct iso88025_header *);
bcopy((caddr_t)edst, (caddr_t)&gen_th.iso88025_dhost, ISO88025_ADDR_LEN);
/* Copy as much of the generic header as is needed into the mbuf */
memcpy(th, &gen_th, ISO88025_HDR_LEN + rif_len);
/*
* If a simplex interface, and the packet is being sent to our
* Ethernet address or a broadcast address, loopback a copy.
* XXX To make a simplex device behave exactly like a duplex
* device, we should copy in the case of sending to our own
* ethernet address (thus letting the original actually appear
* on the wire). However, we don't do that here for security
* reasons and compatibility with the original behavior.
*/
if ((ifp->if_flags & IFF_SIMPLEX) && (loop_copy != -1)) {
if ((m->m_flags & M_BCAST) || (loop_copy > 0)) {
struct mbuf *n;
n = m_copy(m, 0, (int)M_COPYALL);
(void) if_simloop(ifp, n, dst->sa_family,
ISO88025_HDR_LEN);
} else if (bcmp(th->iso88025_dhost, th->iso88025_shost,
ETHER_ADDR_LEN) == 0) {
(void) if_simloop(ifp, m, dst->sa_family,
ISO88025_HDR_LEN);
return(0); /* XXX */
}
}
IFQ_HANDOFF_ADJ(ifp, m, ISO88025_HDR_LEN + LLC_SNAPFRAMELEN, error);
if (error) {
printf("iso88025_output: packet dropped QFULL.\n");
ifp->if_oerrors++;
}
return (error);
bad:
ifp->if_oerrors++;
if (m)
m_freem(m);
return (error);
}
/*
* ISO 88025 de-encapsulation
*/
void
iso88025_input(ifp, m)
struct ifnet *ifp;
struct mbuf *m;
{
struct iso88025_header *th;
struct llc *l;
int isr;
int mac_hdr_len;
/*
* Do consistency checks to verify assumptions
* made by code past this point.
*/
if ((m->m_flags & M_PKTHDR) == 0) {
if_printf(ifp, "discard frame w/o packet header\n");
ifp->if_ierrors++;
m_freem(m);
return;
}
if (m->m_pkthdr.rcvif == NULL) {
if_printf(ifp, "discard frame w/o interface pointer\n");
ifp->if_ierrors++;
m_freem(m);
return;
}
m = m_pullup(m, ISO88025_HDR_LEN);
if (m == NULL) {
ifp->if_ierrors++;
goto dropanyway;
}
th = mtod(m, struct iso88025_header *);
m->m_pkthdr.header = (void *)th;
/*
* Discard packet if interface is not up.
*/
if (!((ifp->if_flags & IFF_UP) &&
(ifp->if_drv_flags & IFF_DRV_RUNNING)))
goto dropanyway;
/*
* Give bpf a chance at the packet.
*/
BPF_MTAP(ifp, m);
/*
* Interface marked for monitoring; discard packet.
*/
if (ifp->if_flags & IFF_MONITOR) {
m_freem(m);
return;
}
#ifdef MAC
mac_ifnet_create_mbuf(ifp, m);
#endif
/*
* Update interface statistics.
*/
ifp->if_ibytes += m->m_pkthdr.len;
getmicrotime(&ifp->if_lastchange);
/*
* Discard non local unicast packets when interface
* is in promiscuous mode.
*/
if ((ifp->if_flags & IFF_PROMISC) &&
((th->iso88025_dhost[0] & 1) == 0) &&
(bcmp(IF_LLADDR(ifp), (caddr_t) th->iso88025_dhost,
ISO88025_ADDR_LEN) != 0))
goto dropanyway;
/*
* Set mbuf flags for bcast/mcast.
*/
if (th->iso88025_dhost[0] & 1) {
if (bcmp(iso88025_broadcastaddr, th->iso88025_dhost,
ISO88025_ADDR_LEN) == 0)
m->m_flags |= M_BCAST;
else
m->m_flags |= M_MCAST;
ifp->if_imcasts++;
}
mac_hdr_len = ISO88025_HDR_LEN;
/* Check for source routing info */
if (th->iso88025_shost[0] & TR_RII)
mac_hdr_len += TR_RCF_RIFLEN(th->rcf);
/* Strip off ISO88025 header. */
m_adj(m, mac_hdr_len);
m = m_pullup(m, LLC_SNAPFRAMELEN);
if (m == 0) {
ifp->if_ierrors++;
goto dropanyway;
}
l = mtod(m, struct llc *);
switch (l->llc_dsap) {
#ifdef IPX
case ETHERTYPE_IPX_8022: /* Thanks a bunch Novell */
if ((l->llc_control != LLC_UI) ||
(l->llc_ssap != ETHERTYPE_IPX_8022)) {
ifp->if_noproto++;
goto dropanyway;
}
th->iso88025_shost[0] &= ~(TR_RII);
m_adj(m, 3);
isr = NETISR_IPX;
break;
#endif /* IPX */
case LLC_SNAP_LSAP: {
u_int16_t type;
if ((l->llc_control != LLC_UI) ||
(l->llc_ssap != LLC_SNAP_LSAP)) {
ifp->if_noproto++;
goto dropanyway;
}
if (l->llc_snap.org_code[0] != 0 ||
l->llc_snap.org_code[1] != 0 ||
l->llc_snap.org_code[2] != 0) {
ifp->if_noproto++;
goto dropanyway;
}
type = ntohs(l->llc_snap.ether_type);
m_adj(m, LLC_SNAPFRAMELEN);
switch (type) {
#ifdef INET
case ETHERTYPE_IP:
th->iso88025_shost[0] &= ~(TR_RII);
if ((m = ip_fastforward(m)) == NULL)
return;
isr = NETISR_IP;
break;
case ETHERTYPE_ARP:
if (ifp->if_flags & IFF_NOARP)
goto dropanyway;
isr = NETISR_ARP;
break;
#endif /* INET */
#ifdef IPX_SNAP /* XXX: Not supported! */
case ETHERTYPE_IPX:
th->iso88025_shost[0] &= ~(TR_RII);
isr = NETISR_IPX;
break;
#endif /* IPX_SNAP */
#ifdef INET6
case ETHERTYPE_IPV6:
th->iso88025_shost[0] &= ~(TR_RII);
isr = NETISR_IPV6;
break;
#endif /* INET6 */
default:
printf("iso88025_input: unexpected llc_snap ether_type 0x%02x\n", type);
ifp->if_noproto++;
goto dropanyway;
}
break;
}
#ifdef ISO
case LLC_ISO_LSAP:
switch (l->llc_control) {
case LLC_UI:
ifp->if_noproto++;
goto dropanyway;
break;
case LLC_XID:
case LLC_XID_P:
if(m->m_len < ISO88025_ADDR_LEN)
goto dropanyway;
l->llc_window = 0;
l->llc_fid = 9;
l->llc_class = 1;
l->llc_dsap = l->llc_ssap = 0;
/* Fall through to */
case LLC_TEST:
case LLC_TEST_P:
{
struct sockaddr sa;
struct arpcom *ac;
struct iso88025_sockaddr_data *th2;
int i;
u_char c;
c = l->llc_dsap;
if (th->iso88025_shost[0] & TR_RII) { /* XXX */
printf("iso88025_input: dropping source routed LLC_TEST\n");
goto dropanyway;
}
l->llc_dsap = l->llc_ssap;
l->llc_ssap = c;
if (m->m_flags & (M_BCAST | M_MCAST))
bcopy((caddr_t)IF_LLADDR(ifp),
(caddr_t)th->iso88025_dhost,
ISO88025_ADDR_LEN);
sa.sa_family = AF_UNSPEC;
sa.sa_len = sizeof(sa);
th2 = (struct iso88025_sockaddr_data *)sa.sa_data;
for (i = 0; i < ISO88025_ADDR_LEN; i++) {
th2->ether_shost[i] = c = th->iso88025_dhost[i];
th2->ether_dhost[i] = th->iso88025_dhost[i] =
th->iso88025_shost[i];
th->iso88025_shost[i] = c;
}
th2->ac = TR_AC;
th2->fc = TR_LLC_FRAME;
ifp->if_output(ifp, m, &sa, NULL);
return;
}
default:
printf("iso88025_input: unexpected llc control 0x%02x\n", l->llc_control);
ifp->if_noproto++;
goto dropanyway;
break;
}
break;
#endif /* ISO */
default:
printf("iso88025_input: unknown dsap 0x%x\n", l->llc_dsap);
ifp->if_noproto++;
goto dropanyway;
break;
}
netisr_dispatch(isr, m);
return;
dropanyway:
ifp->if_iqdrops++;
if (m)
m_freem(m);
return;
}
static int
iso88025_resolvemulti (ifp, llsa, sa)
struct ifnet *ifp;
struct sockaddr **llsa;
struct sockaddr *sa;
{
struct sockaddr_dl *sdl;
#ifdef INET
struct sockaddr_in *sin;
#endif
#ifdef INET6
struct sockaddr_in6 *sin6;
#endif
u_char *e_addr;
switch(sa->sa_family) {
case AF_LINK:
/*
* No mapping needed. Just check that it's a valid MC address.
*/
sdl = (struct sockaddr_dl *)sa;
e_addr = LLADDR(sdl);
if ((e_addr[0] & 1) != 1) {
return (EADDRNOTAVAIL);
}
*llsa = 0;
return (0);
#ifdef INET
case AF_INET:
sin = (struct sockaddr_in *)sa;
if (!IN_MULTICAST(ntohl(sin->sin_addr.s_addr))) {
return (EADDRNOTAVAIL);
}
sdl = malloc(sizeof *sdl, M_IFMADDR,
M_NOWAIT|M_ZERO);
if (sdl == NULL)
return (ENOMEM);
sdl->sdl_len = sizeof *sdl;
sdl->sdl_family = AF_LINK;
sdl->sdl_index = ifp->if_index;
sdl->sdl_type = IFT_ISO88025;
sdl->sdl_alen = ISO88025_ADDR_LEN;
e_addr = LLADDR(sdl);
ETHER_MAP_IP_MULTICAST(&sin->sin_addr, e_addr);
*llsa = (struct sockaddr *)sdl;
return (0);
#endif
#ifdef INET6
case AF_INET6:
sin6 = (struct sockaddr_in6 *)sa;
if (IN6_IS_ADDR_UNSPECIFIED(&sin6->sin6_addr)) {
/*
* An IP6 address of 0 means listen to all
* of the Ethernet multicast address used for IP6.
* (This is used for multicast routers.)
*/
ifp->if_flags |= IFF_ALLMULTI;
*llsa = 0;
return (0);
}
if (!IN6_IS_ADDR_MULTICAST(&sin6->sin6_addr)) {
return (EADDRNOTAVAIL);
}
sdl = malloc(sizeof *sdl, M_IFMADDR,
M_NOWAIT|M_ZERO);
if (sdl == NULL)
return (ENOMEM);
sdl->sdl_len = sizeof *sdl;
sdl->sdl_family = AF_LINK;
sdl->sdl_index = ifp->if_index;
sdl->sdl_type = IFT_ISO88025;
sdl->sdl_alen = ISO88025_ADDR_LEN;
e_addr = LLADDR(sdl);
ETHER_MAP_IPV6_MULTICAST(&sin6->sin6_addr, e_addr);
*llsa = (struct sockaddr *)sdl;
return (0);
#endif
default:
/*
* Well, the text isn't quite right, but it's the name
* that counts...
*/
return (EAFNOSUPPORT);
}
return (0);
}
MALLOC_DEFINE(M_ISO88025, "arpcom", "802.5 interface internals");
static void*
iso88025_alloc(u_char type, struct ifnet *ifp)
{
struct arpcom *ac;
ac = malloc(sizeof(struct arpcom), M_ISO88025, M_WAITOK | M_ZERO);
ac->ac_ifp = ifp;
return (ac);
}
static void
iso88025_free(void *com, u_char type)
{
free(com, M_ISO88025);
}
static int
iso88025_modevent(module_t mod, int type, void *data)
{
switch (type) {
case MOD_LOAD:
if_register_com_alloc(IFT_ISO88025, iso88025_alloc,
iso88025_free);
break;
case MOD_UNLOAD:
if_deregister_com_alloc(IFT_ISO88025);
break;
default:
return EOPNOTSUPP;
}
return (0);
}
static moduledata_t iso88025_mod = {
"iso88025",
iso88025_modevent,
0
};
DECLARE_MODULE(iso88025, iso88025_mod, SI_SUB_PSEUDO, SI_ORDER_ANY);
MODULE_VERSION(iso88025, 1);
Index: projects/arpv2_merge_1/sys/net/route.c
===================================================================
--- projects/arpv2_merge_1/sys/net/route.c (revision 185999)
+++ projects/arpv2_merge_1/sys/net/route.c (revision 186000)
@@ -1,1572 +1,1368 @@
/*-
* Copyright (c) 1980, 1986, 1991, 1993
* The Regents of the University of California. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* @(#)route.c 8.3.1.1 (Berkeley) 2/23/95
* $FreeBSD$
*/
/************************************************************************
* Note: In this file a 'fib' is a "forwarding information base" *
* Which is the new name for an in kernel routing (next hop) table. *
***********************************************************************/
#include "opt_inet.h"
#include "opt_route.h"
#include "opt_mrouting.h"
#include "opt_mpath.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/syslog.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/socket.h>
#include <sys/sysctl.h>
#include <sys/syslog.h>
#include <sys/sysproto.h>
#include <sys/proc.h>
#include <sys/domain.h>
#include <sys/kernel.h>
#include <sys/vimage.h>
#include <net/if.h>
#include <net/if_dl.h>
#include <net/route.h>
#ifdef RADIX_MPATH
#include <net/radix_mpath.h>
#endif
#include <net/vnet.h>
#include <netinet/in.h>
#include <netinet/ip_mroute.h>
#include <netinet/vinet.h>
#include <vm/uma.h>
u_int rt_numfibs = RT_NUMFIBS;
SYSCTL_INT(_net, OID_AUTO, fibs, CTLFLAG_RD, &rt_numfibs, 0, "");
/*
* Allow the boot code to allow LESS than RT_MAXFIBS to be used.
* We can't do more because storage is statically allocated for now.
* (for compatibility reasons.. this will change).
*/
TUNABLE_INT("net.fibs", &rt_numfibs);
/*
* By default add routes to all fibs for new interfaces.
* Once this is set to 0 then only allocate routes on interface
* changes for the FIB of the caller when adding a new set of addresses
* to an interface. XXX this is a shotgun aproach to a problem that needs
* a more fine grained solution.. that will come.
*/
u_int rt_add_addr_allfibs = 1;
SYSCTL_INT(_net, OID_AUTO, add_addr_allfibs, CTLFLAG_RW,
&rt_add_addr_allfibs, 0, "");
TUNABLE_INT("net.add_addr_allfibs", &rt_add_addr_allfibs);
#ifdef VIMAGE_GLOBALS
static struct rtstat rtstat;
/* by default only the first 'row' of tables will be accessed. */
/*
* XXXMRT When we fix netstat, and do this differnetly,
* we can allocate this dynamically. As long as we are keeping
* things backwards compaitble we need to allocate this
* statically.
*/
struct radix_node_head *rt_tables[RT_MAXFIBS][AF_MAX+1];
static int rttrash; /* routes not in table but not freed */
#endif
static void rt_maskedcopy(struct sockaddr *,
struct sockaddr *, struct sockaddr *);
/* compare two sockaddr structures */
#define sa_equal(a1, a2) (bcmp((a1), (a2), (a1)->sa_len) == 0)
/*
* Convert a 'struct radix_node *' to a 'struct rtentry *'.
* The operation can be done safely (in this code) because a
* 'struct rtentry' starts with two 'struct radix_node''s, the first
* one representing leaf nodes in the routing tree, which is
* what the code in radix.c passes us as a 'struct radix_node'.
*
* But because there are a lot of assumptions in this conversion,
* do not cast explicitly, but always use the macro below.
*/
#define RNTORT(p) ((struct rtentry *)(p))
static uma_zone_t rtzone; /* Routing table UMA zone. */
#if 0
/* default fib for tunnels to use */
u_int tunnel_fib = 0;
SYSCTL_INT(_net, OID_AUTO, tunnelfib, CTLFLAG_RD, &tunnel_fib, 0, "");
#endif
/*
* handler for net.my_fibnum
*/
static int
sysctl_my_fibnum(SYSCTL_HANDLER_ARGS)
{
int fibnum;
int error;
fibnum = curthread->td_proc->p_fibnum;
error = sysctl_handle_int(oidp, &fibnum, 0, req);
return (error);
}
SYSCTL_PROC(_net, OID_AUTO, my_fibnum, CTLTYPE_INT|CTLFLAG_RD,
NULL, 0, &sysctl_my_fibnum, "I", "default FIB of caller");
static void
route_init(void)
{
INIT_VNET_INET(curvnet);
int table;
struct domain *dom;
int fam;
/* whack the tunable ints into line. */
if (rt_numfibs > RT_MAXFIBS)
rt_numfibs = RT_MAXFIBS;
if (rt_numfibs == 0)
rt_numfibs = 1;
rtzone = uma_zcreate("rtentry", sizeof(struct rtentry), NULL, NULL,
NULL, NULL, UMA_ALIGN_PTR, 0);
rn_init(); /* initialize all zeroes, all ones, mask table */
for (dom = domains; dom; dom = dom->dom_next) {
if (dom->dom_rtattach) {
for (table = 0; table < rt_numfibs; table++) {
if ( (fam = dom->dom_family) == AF_INET ||
table == 0) {
/* for now only AF_INET has > 1 table */
/* XXX MRT
* rtattach will be also called
* from vfs_export.c but the
* offset will be 0
* (only for AF_INET and AF_INET6
* which don't need it anyhow)
*/
dom->dom_rtattach(
(void **)&V_rt_tables[table][fam],
dom->dom_rtoffset);
} else {
break;
}
}
}
}
}
#ifndef _SYS_SYSPROTO_H_
struct setfib_args {
int fibnum;
};
#endif
int
setfib(struct thread *td, struct setfib_args *uap)
{
if (uap->fibnum < 0 || uap->fibnum >= rt_numfibs)
return EINVAL;
td->td_proc->p_fibnum = uap->fibnum;
return (0);
}
/*
* Packet routing routines.
*/
void
rtalloc(struct route *ro)
{
rtalloc_ign_fib(ro, 0UL, 0);
}
void
rtalloc_fib(struct route *ro, u_int fibnum)
{
rtalloc_ign_fib(ro, 0UL, fibnum);
}
void
rtalloc_ign(struct route *ro, u_long ignore)
{
struct rtentry *rt;
if ((rt = ro->ro_rt) != NULL) {
if (rt->rt_ifp != NULL && rt->rt_flags & RTF_UP)
return;
RTFREE(rt);
ro->ro_rt = NULL;
}
ro->ro_rt = rtalloc1_fib(&ro->ro_dst, 1, ignore, 0);
if (ro->ro_rt)
RT_UNLOCK(ro->ro_rt);
}
void
rtalloc_ign_fib(struct route *ro, u_long ignore, u_int fibnum)
{
struct rtentry *rt;
if ((rt = ro->ro_rt) != NULL) {
if (rt->rt_ifp != NULL && rt->rt_flags & RTF_UP)
return;
RTFREE(rt);
ro->ro_rt = NULL;
}
ro->ro_rt = rtalloc1_fib(&ro->ro_dst, 1, ignore, fibnum);
if (ro->ro_rt)
RT_UNLOCK(ro->ro_rt);
}
/*
* Look up the route that matches the address given
* Or, at least try.. Create a cloned route if needed.
*
* The returned route, if any, is locked.
*/
struct rtentry *
rtalloc1(struct sockaddr *dst, int report, u_long ignflags)
{
return (rtalloc1_fib(dst, report, ignflags, 0));
}
struct rtentry *
rtalloc1_fib(struct sockaddr *dst, int report, u_long ignflags,
u_int fibnum)
{
INIT_VNET_NET(curvnet);
struct radix_node_head *rnh;
struct rtentry *rt;
struct radix_node *rn;
struct rtentry *newrt;
struct rt_addrinfo info;
int err = 0, msgtype = RTM_MISS;
int needlock;
KASSERT((fibnum < rt_numfibs), ("rtalloc1_fib: bad fibnum"));
if (dst->sa_family != AF_INET) /* Only INET supports > 1 fib now */
fibnum = 0;
rnh = V_rt_tables[fibnum][dst->sa_family];
newrt = NULL;
/*
* Look up the address in the table for that Address Family
*/
if (rnh == NULL) {
V_rtstat.rts_unreach++;
goto miss;
}
needlock = !(ignflags & RTF_RNH_LOCKED);
if (needlock)
RADIX_NODE_HEAD_RLOCK(rnh);
#ifdef INVARIANTS
else
RADIX_NODE_HEAD_LOCK_ASSERT(rnh);
#endif
rn = rnh->rnh_matchaddr(dst, rnh);
if (rn && ((rn->rn_flags & RNF_ROOT) == 0)) {
newrt = rt = RNTORT(rn);
RT_LOCK(newrt);
RT_ADDREF(newrt);
if (needlock)
RADIX_NODE_HEAD_RUNLOCK(rnh);
goto done;
} else if (needlock)
RADIX_NODE_HEAD_RUNLOCK(rnh);
/*
* Either we hit the root or couldn't find any match,
* Which basically means
* "caint get there frm here"
*/
V_rtstat.rts_unreach++;
miss:
if (report) {
/*
* If required, report the failure to the supervising
* Authorities.
* For a delete, this is not an error. (report == 0)
*/
bzero(&info, sizeof(info));
info.rti_info[RTAX_DST] = dst;
rt_missmsg(msgtype, &info, 0, err);
}
done:
if (newrt)
RT_LOCK_ASSERT(newrt);
return (newrt);
}
/*
* Remove a reference count from an rtentry.
* If the count gets low enough, take it out of the routing table
*/
void
rtfree(struct rtentry *rt)
{
INIT_VNET_NET(curvnet);
struct radix_node_head *rnh;
KASSERT(rt != NULL,("%s: NULL rt", __func__));
rnh = V_rt_tables[rt->rt_fibnum][rt_key(rt)->sa_family];
KASSERT(rnh != NULL,("%s: NULL rnh", __func__));
RT_LOCK_ASSERT(rt);
/*
* The callers should use RTFREE_LOCKED() or RTFREE(), so
* we should come here exactly with the last reference.
*/
RT_REMREF(rt);
if (rt->rt_refcnt > 0) {
log(LOG_DEBUG, "%s: %p has %d refs\t", __func__, rt, rt->rt_refcnt);
goto done;
}
/*
* On last reference give the "close method" a chance
* to cleanup private state. This also permits (for
* IPv4 and IPv6) a chance to decide if the routing table
* entry should be purged immediately or at a later time.
* When an immediate purge is to happen the close routine
* typically calls rtexpunge which clears the RTF_UP flag
* on the entry so that the code below reclaims the storage.
*/
if (rt->rt_refcnt == 0 && rnh->rnh_close)
rnh->rnh_close((struct radix_node *)rt, rnh);
/*
* If we are no longer "up" (and ref == 0)
* then we can free the resources associated
* with the route.
*/
if ((rt->rt_flags & RTF_UP) == 0) {
if (rt->rt_nodes->rn_flags & (RNF_ACTIVE | RNF_ROOT))
panic("rtfree 2");
/*
* the rtentry must have been removed from the routing table
* so it is represented in rttrash.. remove that now.
*/
V_rttrash--;
#ifdef DIAGNOSTIC
if (rt->rt_refcnt < 0) {
printf("rtfree: %p not freed (neg refs)\n", rt);
goto done;
}
#endif
/*
* release references on items we hold them on..
* e.g other routes and ifaddrs.
*/
if (rt->rt_ifa)
IFAFREE(rt->rt_ifa);
/*
* The key is separatly alloc'd so free it (see rt_setgate()).
* This also frees the gateway, as they are always malloc'd
* together.
*/
Free(rt_key(rt));
/*
* and the rtentry itself of course
*/
RT_LOCK_DESTROY(rt);
uma_zfree(rtzone, rt);
return;
}
done:
RT_UNLOCK(rt);
}
/*
* Force a routing table entry to the specified
* destination to go through the given gateway.
* Normally called as a result of a routing redirect
* message from the network layer.
*/
void
rtredirect(struct sockaddr *dst,
struct sockaddr *gateway,
struct sockaddr *netmask,
int flags,
struct sockaddr *src)
{
rtredirect_fib(dst, gateway, netmask, flags, src, 0);
}
void
rtredirect_fib(struct sockaddr *dst,
struct sockaddr *gateway,
struct sockaddr *netmask,
int flags,
struct sockaddr *src,
u_int fibnum)
{
INIT_VNET_NET(curvnet);
struct rtentry *rt, *rt0 = NULL;
int error = 0;
short *stat = NULL;
struct rt_addrinfo info;
struct ifaddr *ifa;
struct radix_node_head *rnh =
V_rt_tables[fibnum][dst->sa_family];
/* verify the gateway is directly reachable */
if ((ifa = ifa_ifwithnet(gateway)) == NULL) {
error = ENETUNREACH;
goto out;
}
rt = rtalloc1_fib(dst, 0, 0UL, fibnum); /* NB: rt is locked */
/*
* If the redirect isn't from our current router for this dst,
* it's either old or wrong. If it redirects us to ourselves,
* we have a routing loop, perhaps as a result of an interface
* going down recently.
*/
if (!(flags & RTF_DONE) && rt &&
(!sa_equal(src, rt->rt_gateway) || rt->rt_ifa != ifa))
error = EINVAL;
else if (ifa_ifwithaddr(gateway))
error = EHOSTUNREACH;
if (error)
goto done;
/*
* Create a new entry if we just got back a wildcard entry
* or the the lookup failed. This is necessary for hosts
* which use routing redirects generated by smart gateways
* to dynamically build the routing tables.
*/
if (rt == NULL || (rt_mask(rt) && rt_mask(rt)->sa_len < 2))
goto create;
/*
* Don't listen to the redirect if it's
* for a route to an interface.
*/
if (rt->rt_flags & RTF_GATEWAY) {
if (((rt->rt_flags & RTF_HOST) == 0) && (flags & RTF_HOST)) {
/*
* Changing from route to net => route to host.
* Create new route, rather than smashing route to net.
*/
create:
rt0 = rt;
rt = NULL;
flags |= RTF_GATEWAY | RTF_DYNAMIC;
bzero((caddr_t)&info, sizeof(info));
info.rti_info[RTAX_DST] = dst;
info.rti_info[RTAX_GATEWAY] = gateway;
info.rti_info[RTAX_NETMASK] = netmask;
info.rti_ifa = ifa;
info.rti_flags = flags;
if (rt0 != NULL)
RT_UNLOCK(rt0); /* drop lock to avoid LOR with RNH */
error = rtrequest1_fib(RTM_ADD, &info, &rt, fibnum);
if (rt != NULL) {
RT_LOCK(rt);
if (rt0 != NULL)
EVENTHANDLER_INVOKE(route_redirect_event, rt0, rt, dst);
flags = rt->rt_flags;
}
if (rt0 != NULL)
RTFREE(rt0);
stat = &V_rtstat.rts_dynamic;
} else {
struct rtentry *gwrt;
/*
* Smash the current notion of the gateway to
* this destination. Should check about netmask!!!
*/
rt->rt_flags |= RTF_MODIFIED;
flags |= RTF_MODIFIED;
stat = &V_rtstat.rts_newgateway;
/*
* add the key and gateway (in one malloc'd chunk).
*/
RT_UNLOCK(rt);
RADIX_NODE_HEAD_LOCK(rnh);
RT_LOCK(rt);
rt_setgate(rt, rt_key(rt), gateway);
gwrt = rtalloc1(gateway, 1, RTF_RNH_LOCKED);
RADIX_NODE_HEAD_UNLOCK(rnh);
EVENTHANDLER_INVOKE(route_redirect_event, rt, gwrt, dst);
RTFREE_LOCKED(gwrt);
}
} else
error = EHOSTUNREACH;
done:
if (rt)
RTFREE_LOCKED(rt);
out:
if (error)
V_rtstat.rts_badredirect++;
else if (stat != NULL)
(*stat)++;
bzero((caddr_t)&info, sizeof(info));
info.rti_info[RTAX_DST] = dst;
info.rti_info[RTAX_GATEWAY] = gateway;
info.rti_info[RTAX_NETMASK] = netmask;
info.rti_info[RTAX_AUTHOR] = src;
rt_missmsg(RTM_REDIRECT, &info, flags, error);
}
int
rtioctl(u_long req, caddr_t data)
{
return (rtioctl_fib(req, data, 0));
}
/*
* Routing table ioctl interface.
*/
int
rtioctl_fib(u_long req, caddr_t data, u_int fibnum)
{
/*
* If more ioctl commands are added here, make sure the proper
* super-user checks are being performed because it is possible for
* prison-root to make it this far if raw sockets have been enabled
* in jails.
*/
#ifdef INET
/* Multicast goop, grrr... */
return mrt_ioctl ? mrt_ioctl(req, data, fibnum) : EOPNOTSUPP;
#else /* INET */
return ENXIO;
#endif /* INET */
}
struct ifaddr *
ifa_ifwithroute(int flags, struct sockaddr *dst, struct sockaddr *gateway)
{
return (ifa_ifwithroute_fib(flags, dst, gateway, 0));
}
struct ifaddr *
ifa_ifwithroute_fib(int flags, struct sockaddr *dst, struct sockaddr *gateway,
u_int fibnum)
{
register struct ifaddr *ifa;
int not_found = 0;
if ((flags & RTF_GATEWAY) == 0) {
/*
* If we are adding a route to an interface,
* and the interface is a pt to pt link
* we should search for the destination
* as our clue to the interface. Otherwise
* we can use the local address.
*/
ifa = NULL;
if (flags & RTF_HOST)
ifa = ifa_ifwithdstaddr(dst);
if (ifa == NULL)
ifa = ifa_ifwithaddr(gateway);
} else {
/*
* If we are adding a route to a remote net
* or host, the gateway may still be on the
* other end of a pt to pt link.
*/
ifa = ifa_ifwithdstaddr(gateway);
}
if (ifa == NULL)
ifa = ifa_ifwithnet(gateway);
if (ifa == NULL) {
struct rtentry *rt = rtalloc1_fib(gateway, 0, RTF_RNH_LOCKED, fibnum);
if (rt == NULL)
return (NULL);
/*
* dismiss a gateway that is reachable only
* through the default router
*/
switch (gateway->sa_family) {
case AF_INET:
if (satosin(rt_key(rt))->sin_addr.s_addr == INADDR_ANY)
not_found = 1;
break;
case AF_INET6:
if (IN6_IS_ADDR_UNSPECIFIED(&satosin6(rt_key(rt))->sin6_addr))
not_found = 1;
break;
default:
break;
}
RT_REMREF(rt);
RT_UNLOCK(rt);
if (not_found)
return (NULL);
if ((ifa = rt->rt_ifa) == NULL)
return (NULL);
}
if (ifa->ifa_addr->sa_family != dst->sa_family) {
struct ifaddr *oifa = ifa;
ifa = ifaof_ifpforaddr(dst, ifa->ifa_ifp);
if (ifa == NULL)
ifa = oifa;
}
return (ifa);
}
/*
* Do appropriate manipulations of a routing tree given
* all the bits of info needed
*/
int
rtrequest(int req,
struct sockaddr *dst,
struct sockaddr *gateway,
struct sockaddr *netmask,
int flags,
struct rtentry **ret_nrt)
{
return (rtrequest_fib(req, dst, gateway, netmask, flags, ret_nrt, 0));
}
int
rtrequest_fib(int req,
struct sockaddr *dst,
struct sockaddr *gateway,
struct sockaddr *netmask,
int flags,
struct rtentry **ret_nrt,
u_int fibnum)
{
struct rt_addrinfo info;
if (dst->sa_len == 0)
return(EINVAL);
bzero((caddr_t)&info, sizeof(info));
info.rti_flags = flags;
info.rti_info[RTAX_DST] = dst;
info.rti_info[RTAX_GATEWAY] = gateway;
info.rti_info[RTAX_NETMASK] = netmask;
return rtrequest1_fib(req, &info, ret_nrt, fibnum);
}
/*
* These (questionable) definitions of apparent local variables apply
* to the next two functions. XXXXXX!!!
*/
#define dst info->rti_info[RTAX_DST]
#define gateway info->rti_info[RTAX_GATEWAY]
#define netmask info->rti_info[RTAX_NETMASK]
#define ifaaddr info->rti_info[RTAX_IFA]
#define ifpaddr info->rti_info[RTAX_IFP]
#define flags info->rti_flags
int
rt_getifa(struct rt_addrinfo *info)
{
return (rt_getifa_fib(info, 0));
}
int
rt_getifa_fib(struct rt_addrinfo *info, u_int fibnum)
{
struct ifaddr *ifa;
int error = 0;
/*
* ifp may be specified by sockaddr_dl
* when protocol address is ambiguous.
*/
if (info->rti_ifp == NULL && ifpaddr != NULL &&
ifpaddr->sa_family == AF_LINK &&
(ifa = ifa_ifwithnet(ifpaddr)) != NULL)
info->rti_ifp = ifa->ifa_ifp;
if (info->rti_ifa == NULL && ifaaddr != NULL)
info->rti_ifa = ifa_ifwithaddr(ifaaddr);
if (info->rti_ifa == NULL) {
struct sockaddr *sa;
sa = ifaaddr != NULL ? ifaaddr :
(gateway != NULL ? gateway : dst);
if (sa != NULL && info->rti_ifp != NULL)
info->rti_ifa = ifaof_ifpforaddr(sa, info->rti_ifp);
else if (dst != NULL && gateway != NULL)
info->rti_ifa = ifa_ifwithroute_fib(flags, dst, gateway,
fibnum);
else if (sa != NULL)
info->rti_ifa = ifa_ifwithroute_fib(flags, sa, sa,
fibnum);
}
if ((ifa = info->rti_ifa) != NULL) {
if (info->rti_ifp == NULL)
info->rti_ifp = ifa->ifa_ifp;
} else
error = ENETUNREACH;
return (error);
}
/*
* Expunges references to a route that's about to be reclaimed.
* The route must be locked.
*/
int
rtexpunge(struct rtentry *rt)
{
INIT_VNET_NET(curvnet);
struct radix_node *rn;
struct radix_node_head *rnh;
struct ifaddr *ifa;
int error = 0;
rnh = V_rt_tables[rt->rt_fibnum][rt_key(rt)->sa_family];
RT_LOCK_ASSERT(rt);
RADIX_NODE_HEAD_LOCK_ASSERT(rnh);
#if 0
/*
* We cannot assume anything about the reference count
* because protocols call us in many situations; often
* before unwinding references to the table entry.
*/
KASSERT(rt->rt_refcnt <= 1, ("bogus refcnt %ld", rt->rt_refcnt));
#endif
/*
* Find the correct routing tree to use for this Address Family
*/
rnh = V_rt_tables[rt->rt_fibnum][rt_key(rt)->sa_family];
if (rnh == NULL)
return (EAFNOSUPPORT);
/*
* Remove the item from the tree; it should be there,
* but when callers invoke us blindly it may not (sigh).
*/
rn = rnh->rnh_deladdr(rt_key(rt), rt_mask(rt), rnh);
if (rn == NULL) {
error = ESRCH;
goto bad;
}
KASSERT((rn->rn_flags & (RNF_ACTIVE | RNF_ROOT)) == 0,
("unexpected flags 0x%x", rn->rn_flags));
KASSERT(rt == RNTORT(rn),
("lookup mismatch, rt %p rn %p", rt, rn));
rt->rt_flags &= ~RTF_UP;
/*
- * Remove any external references we may have.
- * This might result in another rtentry being freed if
- * we held its last reference.
- */
- if (rt->rt_gwroute) {
- RTFREE(rt->rt_gwroute);
- rt->rt_gwroute = NULL;
- }
-
- /*
* Give the protocol a chance to keep things in sync.
*/
if ((ifa = rt->rt_ifa) && ifa->ifa_rtrequest) {
struct rt_addrinfo info;
bzero((caddr_t)&info, sizeof(info));
info.rti_flags = rt->rt_flags;
info.rti_info[RTAX_DST] = rt_key(rt);
info.rti_info[RTAX_GATEWAY] = rt->rt_gateway;
info.rti_info[RTAX_NETMASK] = rt_mask(rt);
ifa->ifa_rtrequest(RTM_DELETE, rt, &info);
}
/*
* one more rtentry floating around that is not
* linked to the routing table.
*/
V_rttrash++;
bad:
return (error);
}
int
rtrequest1(int req, struct rt_addrinfo *info, struct rtentry **ret_nrt)
{
return (rtrequest1_fib(req, info, ret_nrt, 0));
}
int
rtrequest1_fib(int req, struct rt_addrinfo *info, struct rtentry **ret_nrt,
u_int fibnum)
{
INIT_VNET_NET(curvnet);
int error = 0, needlock = 0;
register struct rtentry *rt;
register struct radix_node *rn;
register struct radix_node_head *rnh;
struct ifaddr *ifa;
struct sockaddr *ndst;
#define senderr(x) { error = x ; goto bad; }
KASSERT((fibnum < rt_numfibs), ("rtrequest1_fib: bad fibnum"));
if (dst->sa_family != AF_INET) /* Only INET supports > 1 fib now */
fibnum = 0;
/*
* Find the correct routing tree to use for this Address Family
*/
rnh = V_rt_tables[fibnum][dst->sa_family];
if (rnh == NULL)
return (EAFNOSUPPORT);
needlock = ((flags & RTF_RNH_LOCKED) == 0);
flags &= ~RTF_RNH_LOCKED;
if (needlock)
RADIX_NODE_HEAD_LOCK(rnh);
else
RADIX_NODE_HEAD_LOCK_ASSERT(rnh);
/*
* If we are adding a host route then we don't want to put
* a netmask in the tree, nor do we want to clone it.
*/
if (flags & RTF_HOST)
netmask = NULL;
switch (req) {
case RTM_DELETE:
#ifdef RADIX_MPATH
/*
* if we got multipath routes, we require users to specify
* a matching RTAX_GATEWAY.
*/
if (rn_mpath_capable(rnh)) {
struct rtentry *rto = NULL;
rn = rnh->rnh_matchaddr(dst, rnh);
if (rn == NULL)
senderr(ESRCH);
rto = rt = RNTORT(rn);
rt = rt_mpath_matchgate(rt, gateway);
if (!rt)
senderr(ESRCH);
/*
* this is the first entry in the chain
*/
if (rto == rt) {
rn = rn_mpath_next((struct radix_node *)rt);
/*
* there is another entry, now it's active
*/
if (rn) {
rto = RNTORT(rn);
RT_LOCK(rto);
rto->rt_flags |= RTF_UP;
RT_UNLOCK(rto);
} else if (rt->rt_flags & RTF_GATEWAY) {
/*
* For gateway routes, we need to
* make sure that we we are deleting
* the correct gateway.
* rt_mpath_matchgate() does not
* check the case when there is only
* one route in the chain.
*/
if (gateway &&
(rt->rt_gateway->sa_len != gateway->sa_len ||
memcmp(rt->rt_gateway, gateway, gateway->sa_len)))
senderr(ESRCH);
}
/*
* use the normal delete code to remove
* the first entry
*/
goto normal_rtdel;
}
/*
* if the entry is 2nd and on up
*/
if (!rt_mpath_deldup(rto, rt))
panic ("rtrequest1: rt_mpath_deldup");
RT_LOCK(rt);
RT_ADDREF(rt);
rt->rt_flags &= ~RTF_UP;
goto deldone; /* done with the RTM_DELETE command */
}
normal_rtdel:
#endif
/*
* Remove the item from the tree and return it.
* Complain if it is not there and do no more processing.
*/
rn = rnh->rnh_deladdr(dst, netmask, rnh);
if (rn == NULL)
senderr(ESRCH);
if (rn->rn_flags & (RNF_ACTIVE | RNF_ROOT))
panic ("rtrequest delete");
rt = RNTORT(rn);
RT_LOCK(rt);
RT_ADDREF(rt);
rt->rt_flags &= ~RTF_UP;
/*
- * Remove any external references we may have.
- * This might result in another rtentry being freed if
- * we held its last reference.
- */
- if (rt->rt_gwroute) {
- RTFREE(rt->rt_gwroute);
- rt->rt_gwroute = NULL;
- }
-
- /*
* give the protocol a chance to keep things in sync.
*/
if ((ifa = rt->rt_ifa) && ifa->ifa_rtrequest)
ifa->ifa_rtrequest(RTM_DELETE, rt, info);
#ifdef RADIX_MPATH
deldone:
#endif
/*
* One more rtentry floating around that is not
* linked to the routing table. rttrash will be decremented
* when RTFREE(rt) is eventually called.
*/
V_rttrash++;
/*
* If the caller wants it, then it can have it,
* but it's up to it to free the rtentry as we won't be
* doing it.
*/
if (ret_nrt) {
*ret_nrt = rt;
RT_UNLOCK(rt);
} else
RTFREE_LOCKED(rt);
break;
case RTM_RESOLVE:
/*
* resolve was only used for route cloning
* here for compat
*/
break;
case RTM_ADD:
if ((flags & RTF_GATEWAY) && !gateway)
senderr(EINVAL);
if (dst && gateway && (dst->sa_family != gateway->sa_family) &&
(gateway->sa_family != AF_UNSPEC) && (gateway->sa_family != AF_LINK))
senderr(EINVAL);
if (info->rti_ifa == NULL && (error = rt_getifa_fib(info, fibnum)))
senderr(error);
ifa = info->rti_ifa;
rt = uma_zalloc(rtzone, M_NOWAIT | M_ZERO);
if (rt == NULL)
senderr(ENOBUFS);
RT_LOCK_INIT(rt);
rt->rt_flags = RTF_UP | flags;
rt->rt_fibnum = fibnum;
/*
* Add the gateway. Possibly re-malloc-ing the storage for it
- * also add the rt_gwroute if possible.
+ *
*/
RT_LOCK(rt);
if ((error = rt_setgate(rt, dst, gateway)) != 0) {
RT_LOCK_DESTROY(rt);
uma_zfree(rtzone, rt);
senderr(error);
}
/*
* point to the (possibly newly malloc'd) dest address.
*/
ndst = (struct sockaddr *)rt_key(rt);
/*
* make sure it contains the value we want (masked if needed).
*/
if (netmask) {
rt_maskedcopy(dst, ndst, netmask);
} else
bcopy(dst, ndst, dst->sa_len);
/*
* Note that we now have a reference to the ifa.
* This moved from below so that rnh->rnh_addaddr() can
* examine the ifa and ifa->ifa_ifp if it so desires.
*/
IFAREF(ifa);
rt->rt_ifa = ifa;
rt->rt_ifp = ifa->ifa_ifp;
#ifdef RADIX_MPATH
/* do not permit exactly the same dst/mask/gw pair */
if (rn_mpath_capable(rnh) &&
rt_mpath_conflict(rnh, rt, netmask)) {
- if (rt->rt_gwroute)
- RTFREE(rt->rt_gwroute);
if (rt->rt_ifa) {
IFAFREE(rt->rt_ifa);
}
Free(rt_key(rt));
RT_LOCK_DESTROY(rt);
uma_zfree(rtzone, rt);
senderr(EEXIST);
}
#endif
/* XXX mtu manipulation will be done in rnh_addaddr -- itojun */
rn = rnh->rnh_addaddr(ndst, netmask, rnh, rt->rt_nodes);
/*
* If it still failed to go into the tree,
* then un-make it (this should be a function)
*/
if (rn == NULL) {
- if (rt->rt_gwroute)
- RTFREE(rt->rt_gwroute);
if (rt->rt_ifa)
IFAFREE(rt->rt_ifa);
Free(rt_key(rt));
RT_LOCK_DESTROY(rt);
uma_zfree(rtzone, rt);
senderr(EEXIST);
}
/*
* If this protocol has something to add to this then
* allow it to do that as well.
*/
if (ifa->ifa_rtrequest)
ifa->ifa_rtrequest(req, rt, info);
/*
* actually return a resultant rtentry and
* give the caller a single reference.
*/
if (ret_nrt) {
*ret_nrt = rt;
RT_ADDREF(rt);
}
RT_UNLOCK(rt);
break;
default:
error = EOPNOTSUPP;
}
bad:
if (needlock)
RADIX_NODE_HEAD_UNLOCK(rnh);
return (error);
#undef senderr
}
#undef dst
#undef gateway
#undef netmask
#undef ifaaddr
#undef ifpaddr
#undef flags
int
rt_setgate(struct rtentry *rt, struct sockaddr *dst, struct sockaddr *gate)
{
INIT_VNET_NET(curvnet);
/* XXX dst may be overwritten, can we move this to below */
struct radix_node_head *rnh =
V_rt_tables[rt->rt_fibnum][dst->sa_family];
int dlen = SA_SIZE(dst), glen = SA_SIZE(gate);
-again:
RT_LOCK_ASSERT(rt);
RADIX_NODE_HEAD_LOCK_ASSERT(rnh);
/*
- * Cloning loop avoidance in case of bad configuration.
- */
- if (rt->rt_flags & RTF_GATEWAY) {
- struct rtentry *gwrt;
-
- RT_UNLOCK(rt); /* XXX workaround LOR */
- gwrt = rtalloc1_fib(gate, 1, RTF_RNH_LOCKED, rt->rt_fibnum);
- if (gwrt == rt) {
- RT_REMREF(rt);
- return (EADDRINUSE); /* failure */
- }
- /*
- * Try to reacquire the lock on rt, and if it fails,
- * clean state and restart from scratch.
- */
- if (!RT_TRYLOCK(rt)) {
- RTFREE_LOCKED(gwrt);
- RT_LOCK(rt);
- goto again;
- }
- /*
- * If there is already a gwroute, then drop it. If we
- * are asked to replace route with itself, then do
- * not leak its refcounter.
- */
- if (rt->rt_gwroute != NULL) {
- if (rt->rt_gwroute == gwrt) {
- RT_REMREF(rt->rt_gwroute);
- } else
- RTFREE(rt->rt_gwroute);
- }
-
- if ((rt->rt_gwroute = gwrt) != NULL)
- RT_UNLOCK(rt->rt_gwroute);
- }
-
- /*
* Prepare to store the gateway in rt->rt_gateway.
* Both dst and gateway are stored one after the other in the same
* malloc'd chunk. If we have room, we can reuse the old buffer,
* rt_gateway already points to the right place.
* Otherwise, malloc a new block and update the 'dst' address.
*/
if (rt->rt_gateway == NULL || glen > SA_SIZE(rt->rt_gateway)) {
caddr_t new;
R_Malloc(new, caddr_t, dlen + glen);
if (new == NULL)
return ENOBUFS;
/*
* XXX note, we copy from *dst and not *rt_key(rt) because
* rt_setgate() can be called to initialize a newly
* allocated route entry, in which case rt_key(rt) == NULL
* (and also rt->rt_gateway == NULL).
* Free()/free() handle a NULL argument just fine.
*/
bcopy(dst, new, dlen);
Free(rt_key(rt)); /* free old block, if any */
rt_key(rt) = (struct sockaddr *)new;
rt->rt_gateway = (struct sockaddr *)(new + dlen);
}
/*
* Copy the new gateway value into the memory chunk.
*/
bcopy(gate, rt->rt_gateway, glen);
return (0);
}
static void
rt_maskedcopy(struct sockaddr *src, struct sockaddr *dst, struct sockaddr *netmask)
{
register u_char *cp1 = (u_char *)src;
register u_char *cp2 = (u_char *)dst;
register u_char *cp3 = (u_char *)netmask;
u_char *cplim = cp2 + *cp3;
u_char *cplim2 = cp2 + *cp1;
*cp2++ = *cp1++; *cp2++ = *cp1++; /* copies sa_len & sa_family */
cp3 += 2;
if (cplim > cplim2)
cplim = cplim2;
while (cp2 < cplim)
*cp2++ = *cp1++ & *cp3++;
if (cp2 < cplim2)
bzero((caddr_t)cp2, (unsigned)(cplim2 - cp2));
}
/*
* Set up a routing table entry, normally
* for an interface.
*/
#define _SOCKADDR_TMPSIZE 128 /* Not too big.. kernel stack size is limited */
static inline int
rtinit1(struct ifaddr *ifa, int cmd, int flags, int fibnum)
{
INIT_VNET_NET(curvnet);
struct sockaddr *dst;
struct sockaddr *netmask;
struct rtentry *rt = NULL;
struct rt_addrinfo info;
int error = 0;
int startfib, endfib;
char tempbuf[_SOCKADDR_TMPSIZE];
int didwork = 0;
int a_failure = 0;
static struct sockaddr_dl null_sdl = {sizeof(null_sdl), AF_LINK};
if (flags & RTF_HOST) {
dst = ifa->ifa_dstaddr;
netmask = NULL;
} else {
dst = ifa->ifa_addr;
netmask = ifa->ifa_netmask;
}
if ( dst->sa_family != AF_INET)
fibnum = 0;
if (fibnum == -1) {
if (rt_add_addr_allfibs == 0 && cmd == (int)RTM_ADD) {
startfib = endfib = curthread->td_proc->p_fibnum;
} else {
startfib = 0;
endfib = rt_numfibs - 1;
}
} else {
KASSERT((fibnum < rt_numfibs), ("rtinit1: bad fibnum"));
startfib = fibnum;
endfib = fibnum;
}
if (dst->sa_len == 0)
return(EINVAL);
/*
* If it's a delete, check that if it exists,
* it's on the correct interface or we might scrub
* a route to another ifa which would
* be confusing at best and possibly worse.
*/
if (cmd == RTM_DELETE) {
/*
* It's a delete, so it should already exist..
* If it's a net, mask off the host bits
* (Assuming we have a mask)
* XXX this is kinda inet specific..
*/
if (netmask != NULL) {
rt_maskedcopy(dst, (struct sockaddr *)tempbuf, netmask);
dst = (struct sockaddr *)tempbuf;
}
}
/*
* Now go through all the requested tables (fibs) and do the
* requested action. Realistically, this will either be fib 0
* for protocols that don't do multiple tables or all the
* tables for those that do. XXX For this version only AF_INET.
* When that changes code should be refactored to protocol
* independent parts and protocol dependent parts.
*/
for ( fibnum = startfib; fibnum <= endfib; fibnum++) {
if (cmd == RTM_DELETE) {
struct radix_node_head *rnh;
struct radix_node *rn;
/*
* Look up an rtentry that is in the routing tree and
* contains the correct info.
*/
if ((rnh = V_rt_tables[fibnum][dst->sa_family]) == NULL)
/* this table doesn't exist but others might */
continue;
RADIX_NODE_HEAD_LOCK(rnh);
#ifdef RADIX_MPATH
if (rn_mpath_capable(rnh)) {
rn = rnh->rnh_matchaddr(dst, rnh);
if (rn == NULL)
error = ESRCH;
else {
rt = RNTORT(rn);
/*
* for interface route the
* rt->rt_gateway is sockaddr_intf
* for cloning ARP entries, so
* rt_mpath_matchgate must use the
* interface address
*/
rt = rt_mpath_matchgate(rt,
ifa->ifa_addr);
if (!rt)
error = ESRCH;
}
}
else
#endif
rn = rnh->rnh_lookup(dst, netmask, rnh);
error = (rn == NULL ||
(rn->rn_flags & RNF_ROOT) ||
RNTORT(rn)->rt_ifa != ifa ||
!sa_equal((struct sockaddr *)rn->rn_key, dst));
RADIX_NODE_HEAD_UNLOCK(rnh);
if (error) {
/* this is only an error if bad on ALL tables */
continue;
}
}
/*
* Do the actual request
*/
bzero((caddr_t)&info, sizeof(info));
info.rti_ifa = ifa;
info.rti_flags = flags | ifa->ifa_flags;
info.rti_info[RTAX_DST] = dst;
/*
* doing this for compatibility reasons
*/
if (cmd == RTM_ADD)
info.rti_info[RTAX_GATEWAY] =
(struct sockaddr *)&null_sdl;
else
info.rti_info[RTAX_GATEWAY] = ifa->ifa_addr;
info.rti_info[RTAX_NETMASK] = netmask;
error = rtrequest1_fib(cmd, &info, &rt, fibnum);
if (error == 0 && rt != NULL) {
/*
* notify any listening routing agents of the change
*/
RT_LOCK(rt);
#ifdef RADIX_MPATH
/*
* in case address alias finds the first address
* e.g. ifconfig bge0 192.103.54.246/24
* e.g. ifconfig bge0 192.103.54.247/24
* the address set in the route is 192.103.54.246
* so we need to replace it with 192.103.54.247
*/
if (memcmp(rt->rt_ifa->ifa_addr,
ifa->ifa_addr, ifa->ifa_addr->sa_len)) {
IFAFREE(rt->rt_ifa);
IFAREF(ifa);
rt->rt_ifp = ifa->ifa_ifp;
rt->rt_ifa = ifa;
}
#endif
/*
* doing this for compatibility reasons
*/
if (cmd == RTM_ADD) {
((struct sockaddr_dl *)rt->rt_gateway)->sdl_type =
rt->rt_ifp->if_type;
((struct sockaddr_dl *)rt->rt_gateway)->sdl_index =
rt->rt_ifp->if_index;
}
rt_newaddrmsg(cmd, ifa, error, rt);
if (cmd == RTM_DELETE) {
/*
* If we are deleting, and we found an entry,
* then it's been removed from the tree..
* now throw it away.
*/
RTFREE_LOCKED(rt);
} else {
if (cmd == RTM_ADD) {
/*
* We just wanted to add it..
* we don't actually need a reference.
*/
RT_REMREF(rt);
}
RT_UNLOCK(rt);
}
didwork = 1;
}
if (error)
a_failure = error;
}
if (cmd == RTM_DELETE) {
if (didwork) {
error = 0;
} else {
/* we only give an error if it wasn't in any table */
error = ((flags & RTF_HOST) ?
EHOSTUNREACH : ENETUNREACH);
}
} else {
if (a_failure) {
/* return an error if any of them failed */
error = a_failure;
}
}
return (error);
}
/* special one for inet internal use. may not use. */
int
rtinit_fib(struct ifaddr *ifa, int cmd, int flags)
{
return (rtinit1(ifa, cmd, flags, -1));
}
/*
* Set up a routing table entry, normally
* for an interface.
*/
int
rtinit(struct ifaddr *ifa, int cmd, int flags)
{
struct sockaddr *dst;
int fib = 0;
if (flags & RTF_HOST) {
dst = ifa->ifa_dstaddr;
} else {
dst = ifa->ifa_addr;
}
if (dst->sa_family == AF_INET)
fib = -1;
return (rtinit1(ifa, cmd, flags, fib));
-}
-
-/*
- * rt_check() is invoked on each layer 2 output path, prior to
- * encapsulating outbound packets.
- *
- * The function is mostly used to find a routing entry for the gateway,
- * which in some protocol families could also point to the link-level
- * address for the gateway itself (the side effect of revalidating the
- * route to the destination is rather pointless at this stage, we did it
- * already a moment before in the pr_output() routine to locate the ifp
- * and gateway to use).
- *
- * When we remove the layer-3 to layer-2 mapping tables from the
- * routing table, this function can be removed.
- *
- * === On input ===
- * *dst is the address of the NEXT HOP (which coincides with the
- * final destination if directly reachable);
- * *lrt0 points to the cached route to the final destination;
- * *lrt is not meaningful;
- * (*lrt0 has no ref held on it by us so REMREF is not needed.
- * Refs only account for major structural references and not usages,
- * which is actually a bit of a problem.)
- *
- * === Operation ===
- * If the route is marked down try to find a new route. If the route
- * to the gateway is gone, try to setup a new route. Otherwise,
- * if the route is marked for packets to be rejected, enforce that.
- * Note that rtalloc returns an rtentry with an extra REF that we may
- * need to lose.
- *
- * === On return ===
- * *dst is unchanged;
- * *lrt0 points to the (possibly new) route to the final destination
- * *lrt points to the route to the next hop [LOCKED]
- *
- * Their values are meaningful ONLY if no error is returned.
- *
- * To follow this you have to remember that:
- * RT_REMREF reduces the reference count by 1 but doesn't check it for 0 (!)
- * RTFREE_LOCKED includes an RT_REMREF (or an rtfree if refs == 1)
- * and an RT_UNLOCK
- * RTFREE does an RT_LOCK and an RTFREE_LOCKED
- * The gwroute pointer counts as a reference on the rtentry to which it points.
- * so when we add it we use the ref that rtalloc gives us and when we lose it
- * we need to remove the reference.
- * RT_TEMP_UNLOCK does an RT_ADDREF before freeing the lock, and
- * RT_RELOCK locks it (it can't have gone away due to the ref) and
- * drops the ref, possibly freeing it and zeroing the pointer if
- * the ref goes to 0 (unlocking in the process).
- */
-int
-rt_check(struct rtentry **lrt, struct rtentry **lrt0, struct sockaddr *dst)
-{
- struct rtentry *rt;
- struct rtentry *rt0;
- u_int fibnum;
-
- KASSERT(*lrt0 != NULL, ("rt_check"));
- rt0 = *lrt0;
- rt = NULL;
- fibnum = rt0->rt_fibnum;
-
- /* NB: the locking here is tortuous... */
- RT_LOCK(rt0);
-retry:
- if (rt0 && (rt0->rt_flags & RTF_UP) == 0) {
- /* Current rt0 is useless, try get a replacement. */
- RT_UNLOCK(rt0);
- rt0 = NULL;
- }
- if (rt0 == NULL) {
- rt0 = rtalloc1_fib(dst, 1, 0UL, fibnum);
- if (rt0 == NULL) {
- return (EHOSTUNREACH);
- }
- RT_REMREF(rt0); /* don't need the reference. */
- }
-
- if (rt0->rt_flags & RTF_GATEWAY) {
- if ((rt = rt0->rt_gwroute) != NULL) {
- RT_LOCK(rt); /* NB: gwroute */
- if ((rt->rt_flags & RTF_UP) == 0) {
- /* gw route is dud. ignore/lose it */
- RTFREE_LOCKED(rt); /* unref (&unlock) gwroute */
- rt = rt0->rt_gwroute = NULL;
- }
- }
-
- if (rt == NULL) { /* NOT AN ELSE CLAUSE */
- RT_TEMP_UNLOCK(rt0); /* MUST return to undo this */
- rt = rtalloc1_fib(rt0->rt_gateway, 1, 0UL, fibnum);
- if ((rt == rt0) || (rt == NULL)) {
- /* the best we can do is not good enough */
- if (rt) {
- RT_REMREF(rt); /* assumes ref > 0 */
- RT_UNLOCK(rt);
- }
- RTFREE(rt0); /* lock, unref, (unlock) */
- return (ENETUNREACH);
- }
- /*
- * Relock it and lose the added reference.
- * All sorts of things could have happenned while we
- * had no lock on it, so check for them.
- */
- RT_RELOCK(rt0);
- if (rt0 == NULL || ((rt0->rt_flags & RTF_UP) == 0))
- /* Ru-roh.. what we had is no longer any good */
- goto retry;
- /*
- * While we were away, someone replaced the gateway.
- * Since a reference count is involved we can't just
- * overwrite it.
- */
- if (rt0->rt_gwroute) {
- if (rt0->rt_gwroute != rt) {
- RTFREE_LOCKED(rt);
- goto retry;
- }
- } else {
- rt0->rt_gwroute = rt;
- }
- }
- RT_LOCK_ASSERT(rt);
- RT_UNLOCK(rt0);
- } else {
- /* think of rt as having the lock from now on.. */
- rt = rt0;
- }
- /* XXX why are we inspecting rmx_expire? */
- if ((rt->rt_flags & RTF_REJECT) &&
- (rt->rt_rmx.rmx_expire == 0 ||
- time_uptime < rt->rt_rmx.rmx_expire)) {
- RT_UNLOCK(rt);
- return (rt == rt0 ? EHOSTDOWN : EHOSTUNREACH);
- }
-
- *lrt = rt;
- *lrt0 = rt0;
- return (0);
}
/* This must be before ip6_init2(), which is now SI_ORDER_MIDDLE */
SYSINIT(route, SI_SUB_PROTO_DOMAIN, SI_ORDER_THIRD, route_init, 0);
Index: projects/arpv2_merge_1/sys/net/route.h
===================================================================
--- projects/arpv2_merge_1/sys/net/route.h (revision 185999)
+++ projects/arpv2_merge_1/sys/net/route.h (revision 186000)
@@ -1,436 +1,435 @@
/*-
* Copyright (c) 1980, 1986, 1993
* The Regents of the University of California. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* @(#)route.h 8.4 (Berkeley) 1/9/95
* $FreeBSD$
*/
#ifndef _NET_ROUTE_H_
#define _NET_ROUTE_H_
/*
* Kernel resident routing tables.
*
* The routing tables are initialized when interface addresses
* are set by making entries for all directly connected interfaces.
*/
/*
* A route consists of a destination address and a reference
* to a routing entry. These are often held by protocols
* in their control blocks, e.g. inpcb.
*/
struct route {
struct rtentry *ro_rt;
struct sockaddr ro_dst;
};
/*
* These numbers are used by reliable protocols for determining
* retransmission behavior and are included in the routing structure.
*/
struct rt_metrics_lite {
u_long rmx_mtu; /* MTU for this path */
u_long rmx_expire; /* lifetime for route, e.g. redirect */
u_long rmx_pksent; /* packets sent using this route */
};
struct rt_metrics {
u_long rmx_locks; /* Kernel must leave these values alone */
u_long rmx_mtu; /* MTU for this path */
u_long rmx_hopcount; /* max hops expected */
u_long rmx_expire; /* lifetime for route, e.g. redirect */
u_long rmx_recvpipe; /* inbound delay-bandwidth product */
u_long rmx_sendpipe; /* outbound delay-bandwidth product */
u_long rmx_ssthresh; /* outbound gateway buffer limit */
u_long rmx_rtt; /* estimated round trip time */
u_long rmx_rttvar; /* estimated rtt variance */
u_long rmx_pksent; /* packets sent using this route */
u_long rmx_filler[4]; /* will be used for T/TCP later */
};
/*
* rmx_rtt and rmx_rttvar are stored as microseconds;
* RTTTOPRHZ(rtt) converts to a value suitable for use
* by a protocol slowtimo counter.
*/
#define RTM_RTTUNIT 1000000 /* units for rtt, rttvar, as units per sec */
#define RTTTOPRHZ(r) ((r) / (RTM_RTTUNIT / PR_SLOWHZ))
/* MRT compile-time constants */
#ifdef _KERNEL
#ifndef ROUTETABLES
#define RT_NUMFIBS 1
#define RT_MAXFIBS 1
#else
/* while we use 4 bits in the mbuf flags, we are limited to 16 */
#define RT_MAXFIBS 16
#if ROUTETABLES > RT_MAXFIBS
#define RT_NUMFIBS RT_MAXFIBS
#error "ROUTETABLES defined too big"
#else
#if ROUTETABLES == 0
#define RT_NUMFIBS 1
#else
#define RT_NUMFIBS ROUTETABLES
#endif
#endif
#endif
#endif
extern u_int rt_numfibs; /* number fo usable routing tables */
extern u_int tunnel_fib; /* tunnels use these */
extern u_int fwd_fib; /* packets being forwarded use these routes */
/*
* XXX kernel function pointer `rt_output' is visible to applications.
*/
struct mbuf;
/*
* We distinguish between routes to hosts and routes to networks,
* preferring the former if available. For each route we infer
* the interface to use from the gateway address supplied when
* the route was entered. Routes that forward packets through
* gateways are marked so that the output routines know to address the
* gateway rather than the ultimate destination.
*/
#ifndef RNF_NORMAL
#include <net/radix.h>
#ifdef RADIX_MPATH
#include <net/radix_mpath.h>
#endif
#endif
struct rtentry {
struct radix_node rt_nodes[2]; /* tree glue, and other values */
/*
* XXX struct rtentry must begin with a struct radix_node (or two!)
* because the code does some casts of a 'struct radix_node *'
* to a 'struct rtentry *'
*/
#define rt_key(r) (*((struct sockaddr **)(&(r)->rt_nodes->rn_key)))
#define rt_mask(r) (*((struct sockaddr **)(&(r)->rt_nodes->rn_mask)))
struct sockaddr *rt_gateway; /* value */
int rt_flags; /* up/down?, host/net */
int rt_refcnt; /* # held references */
struct ifnet *rt_ifp; /* the answer: interface to use */
struct ifaddr *rt_ifa; /* the answer: interface address to use */
struct rt_metrics_lite rt_rmx; /* metrics used by rx'ing protocols */
- struct rtentry *rt_gwroute; /* implied entry for gatewayed routes */
u_int rt_fibnum; /* which FIB */
#ifdef _KERNEL
/* XXX ugly, user apps use this definition but don't have a mtx def */
struct mtx rt_mtx; /* mutex for routing entry */
#endif
};
/*
* Following structure necessary for 4.3 compatibility;
* We should eventually move it to a compat file.
*/
struct ortentry {
u_long rt_hash; /* to speed lookups */
struct sockaddr rt_dst; /* key */
struct sockaddr rt_gateway; /* value */
short rt_flags; /* up/down?, host/net */
short rt_refcnt; /* # held references */
u_long rt_use; /* raw # packets forwarded */
struct ifnet *rt_ifp; /* the answer: interface to use */
};
#define rt_use rt_rmx.rmx_pksent
#define RTF_UP 0x1 /* route usable */
#define RTF_GATEWAY 0x2 /* destination is a gateway */
#define RTF_HOST 0x4 /* host entry (net otherwise) */
#define RTF_REJECT 0x8 /* host or net unreachable */
#define RTF_DYNAMIC 0x10 /* created dynamically (by redirect) */
#define RTF_MODIFIED 0x20 /* modified dynamically (by redirect) */
#define RTF_DONE 0x40 /* message confirmed */
/* 0x80 unused, was RTF_DELCLONE */
/* 0x100 unused, was RTF_CLONING */
#define RTF_XRESOLVE 0x200 /* external daemon resolves name */
/* 0x400 unused, was RTF_LLINFO */
#define RTF_STATIC 0x800 /* manually added */
#define RTF_BLACKHOLE 0x1000 /* just discard pkts (during updates) */
#define RTF_PROTO2 0x4000 /* protocol specific routing flag */
#define RTF_PROTO1 0x8000 /* protocol specific routing flag */
/* XXX: temporary to stay API/ABI compatible with userland */
#ifndef _KERNEL
#define RTF_PRCLONING 0x10000 /* unused, for compatibility */
#endif
/* 0x20000 unused, was RTF_WASCLONED */
#define RTF_PROTO3 0x40000 /* protocol specific routing flag */
/* 0x80000 unused */
#define RTF_PINNED 0x100000 /* future use */
#define RTF_LOCAL 0x200000 /* route represents a local address */
#define RTF_BROADCAST 0x400000 /* route represents a bcast address */
#define RTF_MULTICAST 0x800000 /* route represents a mcast address */
/* 0x1000000 and up unassigned */
#define RTF_RNH_LOCKED 0x40000000 /* radix node head locked by caller */
/* Mask of RTF flags that are allowed to be modified by RTM_CHANGE. */
#define RTF_FMASK \
(RTF_PROTO1 | RTF_PROTO2 | RTF_PROTO3 | RTF_BLACKHOLE | \
RTF_REJECT | RTF_STATIC)
/*
* Routing statistics.
*/
struct rtstat {
short rts_badredirect; /* bogus redirect calls */
short rts_dynamic; /* routes created by redirects */
short rts_newgateway; /* routes modified by redirects */
short rts_unreach; /* lookups which failed */
short rts_wildcard; /* lookups satisfied by a wildcard */
};
/*
* Structures for routing messages.
*/
struct rt_msghdr {
u_short rtm_msglen; /* to skip over non-understood messages */
u_char rtm_version; /* future binary compatibility */
u_char rtm_type; /* message type */
u_short rtm_index; /* index for associated ifp */
int rtm_flags; /* flags, incl. kern & message, e.g. DONE */
int rtm_addrs; /* bitmask identifying sockaddrs in msg */
pid_t rtm_pid; /* identify sender */
int rtm_seq; /* for sender to identify action */
int rtm_errno; /* why failed */
int rtm_fmask; /* bitmask used in RTM_CHANGE message */
#define rtm_use rtm_fmask /* deprecated, use rtm_rmx->rmx_pksent */
u_long rtm_inits; /* which metrics we are initializing */
struct rt_metrics rtm_rmx; /* metrics themselves */
};
#define RTM_VERSION 5 /* Up the ante and ignore older versions */
/*
* Message types.
*/
#define RTM_ADD 0x1 /* Add Route */
#define RTM_DELETE 0x2 /* Delete Route */
#define RTM_CHANGE 0x3 /* Change Metrics or flags */
#define RTM_GET 0x4 /* Report Metrics */
#define RTM_LOSING 0x5 /* Kernel Suspects Partitioning */
#define RTM_REDIRECT 0x6 /* Told to use different route */
#define RTM_MISS 0x7 /* Lookup failed on this address */
#define RTM_LOCK 0x8 /* fix specified metrics */
#define RTM_OLDADD 0x9 /* caused by SIOCADDRT */
#define RTM_OLDDEL 0xa /* caused by SIOCDELRT */
#define RTM_RESOLVE 0xb /* req to resolve dst to LL addr */
#define RTM_NEWADDR 0xc /* address being added to iface */
#define RTM_DELADDR 0xd /* address being removed from iface */
#define RTM_IFINFO 0xe /* iface going up/down etc. */
#define RTM_NEWMADDR 0xf /* mcast group membership being added to if */
#define RTM_DELMADDR 0x10 /* mcast group membership being deleted */
#define RTM_IFANNOUNCE 0x11 /* iface arrival/departure */
#define RTM_IEEE80211 0x12 /* IEEE80211 wireless event */
/*
* Bitmask values for rtm_inits and rmx_locks.
*/
#define RTV_MTU 0x1 /* init or lock _mtu */
#define RTV_HOPCOUNT 0x2 /* init or lock _hopcount */
#define RTV_EXPIRE 0x4 /* init or lock _expire */
#define RTV_RPIPE 0x8 /* init or lock _recvpipe */
#define RTV_SPIPE 0x10 /* init or lock _sendpipe */
#define RTV_SSTHRESH 0x20 /* init or lock _ssthresh */
#define RTV_RTT 0x40 /* init or lock _rtt */
#define RTV_RTTVAR 0x80 /* init or lock _rttvar */
/*
* Bitmask values for rtm_addrs.
*/
#define RTA_DST 0x1 /* destination sockaddr present */
#define RTA_GATEWAY 0x2 /* gateway sockaddr present */
#define RTA_NETMASK 0x4 /* netmask sockaddr present */
#define RTA_GENMASK 0x8 /* cloning mask sockaddr present */
#define RTA_IFP 0x10 /* interface name sockaddr present */
#define RTA_IFA 0x20 /* interface addr sockaddr present */
#define RTA_AUTHOR 0x40 /* sockaddr for author of redirect */
#define RTA_BRD 0x80 /* for NEWADDR, broadcast or p-p dest addr */
/*
* Index offsets for sockaddr array for alternate internal encoding.
*/
#define RTAX_DST 0 /* destination sockaddr present */
#define RTAX_GATEWAY 1 /* gateway sockaddr present */
#define RTAX_NETMASK 2 /* netmask sockaddr present */
#define RTAX_GENMASK 3 /* cloning mask sockaddr present */
#define RTAX_IFP 4 /* interface name sockaddr present */
#define RTAX_IFA 5 /* interface addr sockaddr present */
#define RTAX_AUTHOR 6 /* sockaddr for author of redirect */
#define RTAX_BRD 7 /* for NEWADDR, broadcast or p-p dest addr */
#define RTAX_MAX 8 /* size of array to allocate */
struct rt_addrinfo {
int rti_addrs;
struct sockaddr *rti_info[RTAX_MAX];
int rti_flags;
struct ifaddr *rti_ifa;
struct ifnet *rti_ifp;
};
/*
* This macro returns the size of a struct sockaddr when passed
* through a routing socket. Basically we round up sa_len to
* a multiple of sizeof(long), with a minimum of sizeof(long).
* The check for a NULL pointer is just a convenience, probably never used.
* The case sa_len == 0 should only apply to empty structures.
*/
#define SA_SIZE(sa) \
( (!(sa) || ((struct sockaddr *)(sa))->sa_len == 0) ? \
sizeof(long) : \
1 + ( (((struct sockaddr *)(sa))->sa_len - 1) | (sizeof(long) - 1) ) )
#ifdef _KERNEL
#define RT_LOCK_INIT(_rt) \
mtx_init(&(_rt)->rt_mtx, "rtentry", NULL, MTX_DEF | MTX_DUPOK)
#define RT_LOCK(_rt) mtx_lock(&(_rt)->rt_mtx)
#define RT_TRYLOCK(_rt) mtx_trylock(&(_rt)->rt_mtx)
#define RT_UNLOCK(_rt) mtx_unlock(&(_rt)->rt_mtx)
#define RT_LOCK_DESTROY(_rt) mtx_destroy(&(_rt)->rt_mtx)
#define RT_LOCK_ASSERT(_rt) mtx_assert(&(_rt)->rt_mtx, MA_OWNED)
#define RT_ADDREF(_rt) do { \
RT_LOCK_ASSERT(_rt); \
KASSERT((_rt)->rt_refcnt >= 0, \
("negative refcnt %d", (_rt)->rt_refcnt)); \
(_rt)->rt_refcnt++; \
} while (0)
#define RT_REMREF(_rt) do { \
RT_LOCK_ASSERT(_rt); \
KASSERT((_rt)->rt_refcnt > 0, \
("bogus refcnt %d", (_rt)->rt_refcnt)); \
(_rt)->rt_refcnt--; \
} while (0)
#define RTFREE_LOCKED(_rt) do { \
if ((_rt)->rt_refcnt <= 1) \
rtfree(_rt); \
else { \
RT_REMREF(_rt); \
RT_UNLOCK(_rt); \
} \
/* guard against invalid refs */ \
_rt = 0; \
} while (0)
#define RTFREE(_rt) do { \
RT_LOCK(_rt); \
RTFREE_LOCKED(_rt); \
} while (0)
#define RT_TEMP_UNLOCK(_rt) do { \
RT_ADDREF(_rt); \
RT_UNLOCK(_rt); \
} while (0)
#define RT_RELOCK(_rt) do { \
RT_LOCK(_rt); \
if ((_rt)->rt_refcnt <= 1) { \
rtfree(_rt); \
_rt = 0; /* signal that it went away */ \
} else { \
RT_REMREF(_rt); \
/* note that _rt is still valid */ \
} \
} while (0)
extern struct radix_node_head *rt_tables[][AF_MAX+1];
struct ifmultiaddr;
void rt_ieee80211msg(struct ifnet *, int, void *, size_t);
void rt_ifannouncemsg(struct ifnet *, int);
void rt_ifmsg(struct ifnet *);
void rt_missmsg(int, struct rt_addrinfo *, int, int);
void rt_newaddrmsg(int, struct ifaddr *, int, struct rtentry *);
void rt_newmaddrmsg(int, struct ifmultiaddr *);
int rt_setgate(struct rtentry *, struct sockaddr *, struct sockaddr *);
/*
* Note the following locking behavior:
*
* rtalloc_ign() and rtalloc() return ro->ro_rt unlocked
*
* rtalloc1() returns a locked rtentry
*
* rtfree() and RTFREE_LOCKED() require a locked rtentry
*
* RTFREE() uses an unlocked entry.
*/
int rtexpunge(struct rtentry *);
void rtfree(struct rtentry *);
int rt_check(struct rtentry **, struct rtentry **, struct sockaddr *);
/* XXX MRT COMPAT VERSIONS THAT SET UNIVERSE to 0 */
/* Thes are used by old code not yet converted to use multiple FIBS */
int rt_getifa(struct rt_addrinfo *);
void rtalloc_ign(struct route *ro, u_long ignflags);
void rtalloc(struct route *ro); /* XXX deprecated, use rtalloc_ign(ro, 0) */
struct rtentry *rtalloc1(struct sockaddr *, int, u_long);
int rtinit(struct ifaddr *, int, int);
int rtioctl(u_long, caddr_t);
void rtredirect(struct sockaddr *, struct sockaddr *,
struct sockaddr *, int, struct sockaddr *);
int rtrequest(int, struct sockaddr *,
struct sockaddr *, struct sockaddr *, int, struct rtentry **);
int rtrequest1(int, struct rt_addrinfo *, struct rtentry **);
/* defaults to "all" FIBs */
int rtinit_fib(struct ifaddr *, int, int);
/* XXX MRT NEW VERSIONS THAT USE FIBs
* For now the protocol indepedent versions are the same as the AF_INET ones
* but this will change..
*/
int rt_getifa_fib(struct rt_addrinfo *, u_int fibnum);
void rtalloc_ign_fib(struct route *ro, u_long ignflags, u_int fibnum);
void rtalloc_fib(struct route *ro, u_int fibnum);
struct rtentry *rtalloc1_fib(struct sockaddr *, int, u_long, u_int);
int rtioctl_fib(u_long, caddr_t, u_int);
void rtredirect_fib(struct sockaddr *, struct sockaddr *,
struct sockaddr *, int, struct sockaddr *, u_int);
int rtrequest_fib(int, struct sockaddr *,
struct sockaddr *, struct sockaddr *, int, struct rtentry **, u_int);
int rtrequest1_fib(int, struct rt_addrinfo *, struct rtentry **, u_int);
#include <sys/eventhandler.h>
typedef void (*rtevent_arp_update_fn)(void *, struct rtentry *, uint8_t *, struct sockaddr *);
typedef void (*rtevent_redirect_fn)(void *, struct rtentry *, struct rtentry *, struct sockaddr *);
EVENTHANDLER_DECLARE(route_arp_update_event, rtevent_arp_update_fn);
EVENTHANDLER_DECLARE(route_redirect_event, rtevent_redirect_fn);
#endif
#endif
Index: projects/arpv2_merge_1/sys/netinet/if_ether.c
===================================================================
--- projects/arpv2_merge_1/sys/netinet/if_ether.c (revision 185999)
+++ projects/arpv2_merge_1/sys/netinet/if_ether.c (revision 186000)
@@ -1,804 +1,806 @@
/*-
* Copyright (c) 1982, 1986, 1988, 1993
* The Regents of the University of California. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* @(#)if_ether.c 8.1 (Berkeley) 6/10/93
*/
/*
* Ethernet address resolution protocol.
* TODO:
* add "inuse/lock" bit (or ref. count) along with valid bit
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include "opt_inet.h"
#include "opt_mac.h"
#include "opt_carp.h"
#include <sys/param.h>
#include <sys/kernel.h>
#include <sys/queue.h>
#include <sys/sysctl.h>
#include <sys/systm.h>
#include <sys/mbuf.h>
#include <sys/malloc.h>
#include <sys/proc.h>
#include <sys/socket.h>
#include <sys/syslog.h>
#include <sys/vimage.h>
#include <net/if.h>
#include <net/if_dl.h>
#include <net/if_types.h>
#include <net/route.h>
#include <net/netisr.h>
#include <net/if_llc.h>
#include <net/ethernet.h>
#include <net/vnet.h>
#include <netinet/in.h>
#include <netinet/in_var.h>
#include <net/if_llatbl.h>
#include <netinet/if_ether.h>
#include <netinet/vinet.h>
#include <net/if_arc.h>
#include <net/iso88025.h>
#ifdef DEV_CARP
#include <netinet/ip_carp.h>
#endif
#include <security/mac/mac_framework.h>
#define SIN(s) ((struct sockaddr_in *)s)
#define SDL(s) ((struct sockaddr_dl *)s)
#define LLTABLE(ifp) ((struct lltable *)(ifp)->if_afdata[AF_INET])
SYSCTL_DECL(_net_link_ether);
SYSCTL_NODE(_net_link_ether, PF_INET, inet, CTLFLAG_RW, 0, "");
/* timer values */
#ifdef VIMAGE_GLOBALS
static int arpt_keep; /* once resolved, good for 20 more minutes */
static int arp_maxtries;
int useloopback; /* use loopback interface for local traffic */
static int arp_proxyall;
#endif
SYSCTL_V_INT(V_NET, vnet_inet, _net_link_ether_inet, OID_AUTO, max_age,
CTLFLAG_RW, arpt_keep, 0, "ARP entry lifetime in seconds");
static struct ifqueue arpintrq;
SYSCTL_V_INT(V_NET, vnet_inet, _net_link_ether_inet, OID_AUTO, maxtries,
CTLFLAG_RW, arp_maxtries, 0,
"ARP resolution attempts before returning error");
SYSCTL_V_INT(V_NET, vnet_inet, _net_link_ether_inet, OID_AUTO, useloopback,
CTLFLAG_RW, useloopback, 0,
"Use the loopback interface for local traffic");
SYSCTL_V_INT(V_NET, vnet_inet, _net_link_ether_inet, OID_AUTO, proxyall,
CTLFLAG_RW, arp_proxyall, 0,
"Enable proxy ARP for all suitable requests");
static void arp_init(void);
void arprequest(struct ifnet *,
struct in_addr *, struct in_addr *, u_char *);
static void arpintr(struct mbuf *);
static void arptimer(void *);
#ifdef INET
static void in_arpinput(struct mbuf *);
#endif
#ifdef AF_INET
void arp_ifscrub(struct ifnet *ifp, uint32_t addr);
/*
* called by in_ifscrub to remove entry from the table when
* the interface goes away
*/
void
arp_ifscrub(struct ifnet *ifp, uint32_t addr)
{
struct sockaddr_in addr4;
bzero((void *)&addr4, sizeof(addr4));
addr4.sin_len = sizeof(addr4);
addr4.sin_family = AF_INET;
addr4.sin_addr.s_addr = addr;
IF_AFDATA_LOCK(ifp);
lla_lookup(LLTABLE(ifp), (LLE_DELETE | LLE_IFADDR),
(struct sockaddr *)&addr4);
IF_AFDATA_UNLOCK(ifp);
}
#endif
/*
* Timeout routine. Age arp_tab entries periodically.
*/
static void
arptimer(void *arg)
{
struct ifnet *ifp;
struct llentry *lle = (struct llentry *)arg;
if (lle == NULL) {
panic("%s: NULL entry!\n", __func__);
return;
}
ifp = lle->lle_tbl->llt_ifp;
if ((lle->la_flags & LLE_DELETED) ||
(time_second >= lle->la_expire)) {
IF_AFDATA_LOCK(ifp);
if (!callout_pending(&lle->la_timer) &&
callout_active(&lle->la_timer))
(void) llentry_free(lle);
IF_AFDATA_UNLOCK(ifp);
} else {
/*
* Still valid, just drop our reference
*/
LLE_FREE(lle);
}
}
/*
* Broadcast an ARP request. Caller specifies:
* - arp header source ip address
* - arp header target ip address
* - arp header source ethernet address
*/
void
arprequest(struct ifnet *ifp, struct in_addr *sip, struct in_addr *tip,
u_char *enaddr)
{
struct mbuf *m;
struct arphdr *ah;
struct sockaddr sa;
if (sip == NULL) {
/* XXX don't believe this can happen (or explain why) */
/*
* The caller did not supply a source address, try to find
* a compatible one among those assigned to this interface.
*/
struct ifaddr *ifa;
TAILQ_FOREACH(ifa, &ifp->if_addrhead, ifa_link) {
if (!ifa->ifa_addr ||
ifa->ifa_addr->sa_family != AF_INET)
continue;
sip = &SIN(ifa->ifa_addr)->sin_addr;
if (0 == ((sip->s_addr ^ tip->s_addr) &
SIN(ifa->ifa_netmask)->sin_addr.s_addr) )
break; /* found it. */
}
if (sip == NULL) {
printf("%s: cannot find matching address\n", __func__);
return;
}
}
if ((m = m_gethdr(M_DONTWAIT, MT_DATA)) == NULL)
return;
m->m_len = sizeof(*ah) + 2*sizeof(struct in_addr) +
2*ifp->if_data.ifi_addrlen;
m->m_pkthdr.len = m->m_len;
MH_ALIGN(m, m->m_len);
ah = mtod(m, struct arphdr *);
bzero((caddr_t)ah, m->m_len);
#ifdef MAC
mac_netinet_arp_send(ifp, m);
#endif
ah->ar_pro = htons(ETHERTYPE_IP);
ah->ar_hln = ifp->if_addrlen; /* hardware address length */
ah->ar_pln = sizeof(struct in_addr); /* protocol address length */
ah->ar_op = htons(ARPOP_REQUEST);
bcopy((caddr_t)enaddr, (caddr_t)ar_sha(ah), ah->ar_hln);
bcopy((caddr_t)sip, (caddr_t)ar_spa(ah), ah->ar_pln);
bcopy((caddr_t)tip, (caddr_t)ar_tpa(ah), ah->ar_pln);
sa.sa_family = AF_ARP;
sa.sa_len = 2;
m->m_flags |= M_BCAST;
(*ifp->if_output)(ifp, m, &sa, (struct rtentry *)0);
}
/*
* Resolve an IP address into an ethernet address.
* On input:
* ifp is the interface we use
* rt0 is the route to the final destination (possibly useless)
* m is the mbuf. May be NULL if we don't have a packet.
* dst is the next hop,
* desten is where we want the address.
*
* On success, desten is filled in and the function returns 0;
* If the packet must be held pending resolution, we return EWOULDBLOCK
* On other errors, we return the corresponding error code.
* Note that m_freem() handles NULL.
*/
int
arpresolve(struct ifnet *ifp, struct rtentry *rt0, struct mbuf *m,
struct sockaddr *dst, u_char *desten, struct llentry **lle)
{
INIT_VNET_INET(ifp->if_vnet);
struct llentry *la = 0;
u_int flags;
int error, renew;
*lle = NULL;
if (m != NULL) {
if (m->m_flags & M_BCAST) {
/* broadcast */
(void)memcpy(desten,
ifp->if_broadcastaddr, ifp->if_addrlen);
return (0);
}
if (m->m_flags & M_MCAST && ifp->if_type != IFT_ARCNET) {
/* multicast */
ETHER_MAP_IP_MULTICAST(&SIN(dst)->sin_addr, desten);
return (0);
}
}
flags = (ifp->if_flags & (IFF_NOARP | IFF_STATICARP)) ? 0 : LLE_CREATE;
/* XXXXX
* Since this function returns an llentry, the
* lock is held by the caller.
* XXX if caller is required to hold lock, assert it
*/
retry:
+ IF_AFDATA_LOCK(ifp);
la = lla_lookup(LLTABLE(ifp), flags, dst);
+ IF_AFDATA_UNLOCK(ifp);
if (la == NULL) {
if (flags & LLE_CREATE)
log(LOG_DEBUG,
"arpresolve: can't allocate llinfo for %s\n",
inet_ntoa(SIN(dst)->sin_addr));
m_freem(m);
return (EINVAL);
}
if ((la->la_flags & LLE_VALID) &&
((la->la_flags & LLE_STATIC) || la->la_expire > time_uptime)) {
bcopy(&la->ll_addr, desten, ifp->if_addrlen);
/*
* If entry has an expiry time and it is approaching,
* see if we need to send an ARP request within this
* arpt_down interval.
*/
if (!(la->la_flags & LLE_STATIC) &&
time_uptime + la->la_preempt > la->la_expire) {
arprequest(ifp, NULL,
&SIN(dst)->sin_addr, IF_LLADDR(ifp));
la->la_preempt--;
}
*lle = la;
error = 0;
goto done;
}
if (la->la_flags & LLE_STATIC) { /* should not happen! */
log(LOG_DEBUG, "arpresolve: ouch, empty static llinfo for %s\n",
inet_ntoa(SIN(dst)->sin_addr));
m_freem(m);
error = EINVAL;
goto done;
}
renew = (la->la_asked == 0 || la->la_expire != time_uptime);
if ((renew || m != NULL) && (flags & LLE_EXCLUSIVE) == 0) {
flags |= LLE_EXCLUSIVE;
LLE_RUNLOCK(la);
goto retry;
}
/*
* There is an arptab entry, but no ethernet address
* response yet. Replace the held mbuf with this
* latest one.
*/
if (m != NULL) {
if (la->la_hold != NULL)
m_freem(la->la_hold);
la->la_hold = m;
if (renew == 0 && (flags & LLE_EXCLUSIVE)) {
flags &= ~LLE_EXCLUSIVE;
LLE_DOWNGRADE(la);
}
}
/*
* Return EWOULDBLOCK if we have tried less than arp_maxtries. It
* will be masked by ether_output(). Return EHOSTDOWN/EHOSTUNREACH
* if we have already sent arp_maxtries ARP requests. Retransmit the
* ARP request, but not faster than one request per second.
*/
if (la->la_asked < V_arp_maxtries)
error = EWOULDBLOCK; /* First request. */
else
error =
(rt0->rt_flags & RTF_GATEWAY) ? EHOSTDOWN : EHOSTUNREACH;
if (renew) {
LLE_ADDREF(la);
la->la_expire = time_uptime;
callout_reset(&la->la_timer, hz, arptimer, la);
la->la_asked++;
LLE_WUNLOCK(la);
arprequest(ifp, NULL, &SIN(dst)->sin_addr,
IF_LLADDR(ifp));
return (error);
}
done:
if (flags & LLE_EXCLUSIVE)
LLE_WUNLOCK(la);
else
LLE_RUNLOCK(la);
return (error);
}
/*
* Common length and type checks are done here,
* then the protocol-specific routine is called.
*/
static void
arpintr(struct mbuf *m)
{
struct arphdr *ar;
if (m->m_len < sizeof(struct arphdr) &&
((m = m_pullup(m, sizeof(struct arphdr))) == NULL)) {
log(LOG_ERR, "arp: runt packet -- m_pullup failed\n");
return;
}
ar = mtod(m, struct arphdr *);
if (ntohs(ar->ar_hrd) != ARPHRD_ETHER &&
ntohs(ar->ar_hrd) != ARPHRD_IEEE802 &&
ntohs(ar->ar_hrd) != ARPHRD_ARCNET &&
ntohs(ar->ar_hrd) != ARPHRD_IEEE1394) {
log(LOG_ERR, "arp: unknown hardware address format (0x%2D)\n",
(unsigned char *)&ar->ar_hrd, "");
m_freem(m);
return;
}
if (m->m_len < arphdr_len(ar)) {
if ((m = m_pullup(m, arphdr_len(ar))) == NULL) {
log(LOG_ERR, "arp: runt packet\n");
m_freem(m);
return;
}
ar = mtod(m, struct arphdr *);
}
switch (ntohs(ar->ar_pro)) {
#ifdef INET
case ETHERTYPE_IP:
in_arpinput(m);
return;
#endif
}
m_freem(m);
}
#ifdef INET
/*
* ARP for Internet protocols on 10 Mb/s Ethernet.
* Algorithm is that given in RFC 826.
* In addition, a sanity check is performed on the sender
* protocol address, to catch impersonators.
* We no longer handle negotiations for use of trailer protocol:
* Formerly, ARP replied for protocol type ETHERTYPE_TRAIL sent
* along with IP replies if we wanted trailers sent to us,
* and also sent them in response to IP replies.
* This allowed either end to announce the desire to receive
* trailer packets.
* We no longer reply to requests for ETHERTYPE_TRAIL protocol either,
* but formerly didn't normally send requests.
*/
static int log_arp_wrong_iface = 1;
static int log_arp_movements = 1;
static int log_arp_permanent_modify = 1;
SYSCTL_INT(_net_link_ether_inet, OID_AUTO, log_arp_wrong_iface, CTLFLAG_RW,
&log_arp_wrong_iface, 0,
"log arp packets arriving on the wrong interface");
SYSCTL_INT(_net_link_ether_inet, OID_AUTO, log_arp_movements, CTLFLAG_RW,
&log_arp_movements, 0,
"log arp replies from MACs different than the one in the cache");
SYSCTL_INT(_net_link_ether_inet, OID_AUTO, log_arp_permanent_modify, CTLFLAG_RW,
&log_arp_permanent_modify, 0,
"log arp replies from MACs different than the one in the permanent arp entry");
static void
in_arpinput(struct mbuf *m)
{
struct arphdr *ah;
struct ifnet *ifp = m->m_pkthdr.rcvif;
struct llentry *la = NULL;
struct rtentry *rt;
struct ifaddr *ifa;
struct in_ifaddr *ia;
struct sockaddr sa;
struct in_addr isaddr, itaddr, myaddr;
u_int8_t *enaddr = NULL;
int op, flags;
struct mbuf *m0;
int req_len;
int bridged = 0, is_bridge = 0;
#ifdef DEV_CARP
int carp_match = 0;
#endif
struct sockaddr_in sin;
sin.sin_len = sizeof(struct sockaddr_in);
sin.sin_family = AF_INET;
sin.sin_addr.s_addr = 0;
INIT_VNET_INET(ifp->if_vnet);
if (ifp->if_bridge)
bridged = 1;
if (ifp->if_type == IFT_BRIDGE)
is_bridge = 1;
req_len = arphdr_len2(ifp->if_addrlen, sizeof(struct in_addr));
if (m->m_len < req_len && (m = m_pullup(m, req_len)) == NULL) {
log(LOG_ERR, "in_arp: runt packet -- m_pullup failed\n");
return;
}
ah = mtod(m, struct arphdr *);
op = ntohs(ah->ar_op);
(void)memcpy(&isaddr, ar_spa(ah), sizeof (isaddr));
(void)memcpy(&itaddr, ar_tpa(ah), sizeof (itaddr));
/*
* For a bridge, we want to check the address irrespective
* of the receive interface. (This will change slightly
* when we have clusters of interfaces).
* If the interface does not match, but the recieving interface
* is part of carp, we call carp_iamatch to see if this is a
* request for the virtual host ip.
* XXX: This is really ugly!
*/
LIST_FOREACH(ia, INADDR_HASH(itaddr.s_addr), ia_hash) {
if (((bridged && ia->ia_ifp->if_bridge != NULL) ||
ia->ia_ifp == ifp) &&
itaddr.s_addr == ia->ia_addr.sin_addr.s_addr)
goto match;
#ifdef DEV_CARP
if (ifp->if_carp != NULL &&
carp_iamatch(ifp->if_carp, ia, &isaddr, &enaddr) &&
itaddr.s_addr == ia->ia_addr.sin_addr.s_addr) {
carp_match = 1;
goto match;
}
#endif
}
LIST_FOREACH(ia, INADDR_HASH(isaddr.s_addr), ia_hash)
if (((bridged && ia->ia_ifp->if_bridge != NULL) ||
ia->ia_ifp == ifp) &&
isaddr.s_addr == ia->ia_addr.sin_addr.s_addr)
goto match;
#define BDG_MEMBER_MATCHES_ARP(addr, ifp, ia) \
(ia->ia_ifp->if_bridge == ifp->if_softc && \
!bcmp(IF_LLADDR(ia->ia_ifp), IF_LLADDR(ifp), ifp->if_addrlen) && \
addr == ia->ia_addr.sin_addr.s_addr)
/*
* Check the case when bridge shares its MAC address with
* some of its children, so packets are claimed by bridge
* itself (bridge_input() does it first), but they are really
* meant to be destined to the bridge member.
*/
if (is_bridge) {
LIST_FOREACH(ia, INADDR_HASH(itaddr.s_addr), ia_hash) {
if (BDG_MEMBER_MATCHES_ARP(itaddr.s_addr, ifp, ia)) {
ifp = ia->ia_ifp;
goto match;
}
}
}
#undef BDG_MEMBER_MATCHES_ARP
/*
* No match, use the first inet address on the receive interface
* as a dummy address for the rest of the function.
*/
TAILQ_FOREACH(ifa, &ifp->if_addrhead, ifa_link)
if (ifa->ifa_addr->sa_family == AF_INET) {
ia = ifatoia(ifa);
goto match;
}
/*
* If bridging, fall back to using any inet address.
*/
if (!bridged || (ia = TAILQ_FIRST(&V_in_ifaddrhead)) == NULL)
goto drop;
match:
if (!enaddr)
enaddr = (u_int8_t *)IF_LLADDR(ifp);
myaddr = ia->ia_addr.sin_addr;
if (!bcmp(ar_sha(ah), enaddr, ifp->if_addrlen))
goto drop; /* it's from me, ignore it. */
if (!bcmp(ar_sha(ah), ifp->if_broadcastaddr, ifp->if_addrlen)) {
log(LOG_ERR,
"arp: link address is broadcast for IP address %s!\n",
inet_ntoa(isaddr));
goto drop;
}
/*
* Warn if another host is using the same IP address, but only if the
* IP address isn't 0.0.0.0, which is used for DHCP only, in which
* case we suppress the warning to avoid false positive complaints of
* potential misconfiguration.
*/
if (!bridged && isaddr.s_addr == myaddr.s_addr && myaddr.s_addr != 0) {
log(LOG_ERR,
"arp: %*D is using my IP address %s on %s!\n",
ifp->if_addrlen, (u_char *)ar_sha(ah), ":",
inet_ntoa(isaddr), ifp->if_xname);
itaddr = myaddr;
goto reply;
}
if (ifp->if_flags & IFF_STATICARP)
goto reply;
bzero(&sin, sizeof(sin));
sin.sin_len = sizeof(struct sockaddr_in);
sin.sin_family = AF_INET;
sin.sin_addr = isaddr;
flags = (itaddr.s_addr == myaddr.s_addr) ? LLE_CREATE : 0;
flags |= LLE_EXCLUSIVE;
IF_AFDATA_LOCK(ifp);
la = lla_lookup(LLTABLE(ifp), flags, (struct sockaddr *)&sin);
IF_AFDATA_UNLOCK(ifp);
if (la != NULL) {
/* the following is not an error when doing bridging */
if (!bridged && la->lle_tbl->llt_ifp != ifp
#ifdef DEV_CARP
&& (ifp->if_type != IFT_CARP || !carp_match)
#endif
) {
if (log_arp_wrong_iface)
log(LOG_ERR, "arp: %s is on %s "
"but got reply from %*D on %s\n",
inet_ntoa(isaddr),
la->lle_tbl->llt_ifp->if_xname,
ifp->if_addrlen, (u_char *)ar_sha(ah), ":",
ifp->if_xname);
goto reply;
}
if ((la->la_flags & LLE_VALID) &&
bcmp(ar_sha(ah), &la->ll_addr, ifp->if_addrlen)) {
if (la->la_flags & LLE_STATIC) {
log(LOG_ERR,
"arp: %*D attempts to modify permanent "
"entry for %s on %s\n",
ifp->if_addrlen, (u_char *)ar_sha(ah), ":",
inet_ntoa(isaddr), ifp->if_xname);
goto reply;
}
if (log_arp_movements) {
log(LOG_INFO, "arp: %s moved from %*D "
"to %*D on %s\n",
inet_ntoa(isaddr),
ifp->if_addrlen,
(u_char *)&la->ll_addr, ":",
ifp->if_addrlen, (u_char *)ar_sha(ah), ":",
ifp->if_xname);
}
}
if (ifp->if_addrlen != ah->ar_hln) {
log(LOG_WARNING,
"arp from %*D: addr len: new %d, i/f %d (ignored)",
ifp->if_addrlen, (u_char *) ar_sha(ah), ":",
ah->ar_hln, ifp->if_addrlen);
goto reply;
}
(void)memcpy(&la->ll_addr, ar_sha(ah), ifp->if_addrlen);
la->la_flags |= LLE_VALID;
if (!(la->la_flags & LLE_STATIC)) {
la->la_expire = time_uptime + V_arpt_keep;
callout_reset(&la->la_timer, hz * V_arpt_keep,
arptimer, la);
}
la->la_asked = 0;
la->la_preempt = V_arp_maxtries;
if (la->la_hold != NULL) {
m0 = la->la_hold;
la->la_hold = 0;
memcpy(&sa, L3_ADDR(la), sizeof(sa));
LLE_WUNLOCK(la);
(*ifp->if_output)(ifp, m0, &sa, NULL);
return;
}
}
reply:
if (op != ARPOP_REQUEST)
goto drop;
if (itaddr.s_addr == myaddr.s_addr) {
/* Shortcut.. the receiving interface is the target. */
(void)memcpy(ar_tha(ah), ar_sha(ah), ah->ar_hln);
(void)memcpy(ar_sha(ah), enaddr, ah->ar_hln);
} else {
if (la == NULL) {
if (!V_arp_proxyall)
goto drop;
sin.sin_addr = itaddr;
/* XXX MRT use table 0 for arp reply */
rt = in_rtalloc1((struct sockaddr *)&sin, 0, 0UL, 0);
if (!rt)
goto drop;
/*
* Don't send proxies for nodes on the same interface
* as this one came out of, or we'll get into a fight
* over who claims what Ether address.
*/
if (rt->rt_ifp == ifp) {
RTFREE_LOCKED(rt);
goto drop;
}
(void)memcpy(ar_tha(ah), ar_sha(ah), ah->ar_hln);
(void)memcpy(ar_sha(ah), enaddr, ah->ar_hln);
RTFREE_LOCKED(rt);
/*
* Also check that the node which sent the ARP packet
* is on the the interface we expect it to be on. This
* avoids ARP chaos if an interface is connected to the
* wrong network.
*/
sin.sin_addr = isaddr;
/* XXX MRT use table 0 for arp checks */
rt = in_rtalloc1((struct sockaddr *)&sin, 0, 0UL, 0);
if (!rt)
goto drop;
if (rt->rt_ifp != ifp) {
log(LOG_INFO, "arp_proxy: ignoring request"
" from %s via %s, expecting %s\n",
inet_ntoa(isaddr), ifp->if_xname,
rt->rt_ifp->if_xname);
RTFREE_LOCKED(rt);
goto drop;
}
RTFREE_LOCKED(rt);
#ifdef DEBUG_PROXY
printf("arp: proxying for %s\n",
inet_ntoa(itaddr));
#endif
} else {
/*
* Return proxied ARP replies only on the interface
* or bridge cluster where this network resides.
* Otherwise we may conflict with the host we are
* proxying for.
*/
if (la->lle_tbl->llt_ifp != ifp &&
(la->lle_tbl->llt_ifp->if_bridge != ifp->if_bridge ||
ifp->if_bridge == NULL)) {
goto drop;
}
(void)memcpy(ar_tha(ah), ar_sha(ah), ah->ar_hln);
(void)memcpy(ar_sha(ah), &la->ll_addr, ah->ar_hln);
}
}
if (la != NULL)
LLE_WUNLOCK(la);
if (itaddr.s_addr == myaddr.s_addr &&
IN_LINKLOCAL(ntohl(itaddr.s_addr))) {
/* RFC 3927 link-local IPv4; always reply by broadcast. */
#ifdef DEBUG_LINKLOCAL
printf("arp: sending reply for link-local addr %s\n",
inet_ntoa(itaddr));
#endif
m->m_flags |= M_BCAST;
m->m_flags &= ~M_MCAST;
} else {
/* default behaviour; never reply by broadcast. */
m->m_flags &= ~(M_BCAST|M_MCAST);
}
(void)memcpy(ar_tpa(ah), ar_spa(ah), ah->ar_pln);
(void)memcpy(ar_spa(ah), &itaddr, ah->ar_pln);
ah->ar_op = htons(ARPOP_REPLY);
ah->ar_pro = htons(ETHERTYPE_IP); /* let's be sure! */
m->m_len = sizeof(*ah) + (2 * ah->ar_pln) + (2 * ah->ar_hln);
m->m_pkthdr.len = m->m_len;
sa.sa_family = AF_ARP;
sa.sa_len = 2;
(*ifp->if_output)(ifp, m, &sa, (struct rtentry *)0);
return;
drop:
if (la != NULL)
LLE_WUNLOCK(la);
m_freem(m);
}
#endif
void
arp_ifinit(struct ifnet *ifp, struct ifaddr *ifa)
{
struct llentry *lle;
if (ntohl(IA_SIN(ifa)->sin_addr.s_addr) != INADDR_ANY)
arprequest(ifp, &IA_SIN(ifa)->sin_addr,
&IA_SIN(ifa)->sin_addr, IF_LLADDR(ifp));
/*
* interface address is considered static entry
* because the output of the arp utility shows
* that L2 entry as permanent
*/
IF_AFDATA_LOCK(ifp);
lle = lla_lookup(LLTABLE(ifp), (LLE_CREATE | LLE_IFADDR | LLE_STATIC),
(struct sockaddr *)IA_SIN(ifa));
IF_AFDATA_UNLOCK(ifp);
if (lle == NULL)
log(LOG_INFO, "arp_ifinit: cannot create arp "
"entry for interface address\n");
LLE_RUNLOCK(lle);
ifa->ifa_rtrequest = NULL;
}
void
arp_ifinit2(struct ifnet *ifp, struct ifaddr *ifa, u_char *enaddr)
{
if (ntohl(IA_SIN(ifa)->sin_addr.s_addr) != INADDR_ANY)
arprequest(ifp, &IA_SIN(ifa)->sin_addr,
&IA_SIN(ifa)->sin_addr, enaddr);
ifa->ifa_rtrequest = NULL;
}
static void
arp_init(void)
{
INIT_VNET_INET(curvnet);
V_arpt_keep = (20*60); /* once resolved, good for 20 more minutes */
V_arp_maxtries = 5;
V_useloopback = 1; /* use loopback interface for local traffic */
V_arp_proxyall = 0;
arpintrq.ifq_maxlen = 50;
mtx_init(&arpintrq.ifq_mtx, "arp_inq", NULL, MTX_DEF);
netisr_register(NETISR_ARP, arpintr, &arpintrq, 0);
}
SYSINIT(arp, SI_SUB_PROTO_DOMAIN, SI_ORDER_ANY, arp_init, 0);

File Metadata

Mime Type
text/x-diff
Expires
Wed, Nov 12, 11:47 PM (1 d, 6 h)
Storage Engine
blob
Storage Format
Raw Data
Storage Handle
25184884
Default Alt Text
(136 KB)

Event Timeline