NetBSD/sys/netinet/if_arp.c

608 lines
17 KiB
C
Raw Normal View History

1993-03-21 12:45:37 +03:00
/*
* Copyright (c) 1982, 1986, 1988 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.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the University of
* California, Berkeley and its contributors.
* 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 7.13 (Berkeley) 10/31/90
*/
/*
* Ethernet address resolution protocol.
* TODO:
* run at splnet (add ARP protocol intr.)
* link entries onto hash chains, keep free list
* add "inuse/lock" bit (or ref. count) along with valid bit
*/
#include "param.h"
#include "systm.h"
#include "malloc.h"
#include "mbuf.h"
#include "socket.h"
#include "time.h"
#include "kernel.h"
#include "errno.h"
#include "ioctl.h"
#include "syslog.h"
#include "../net/if.h"
#include "in.h"
#include "in_systm.h"
#include "in_var.h"
#include "ip.h"
#include "if_ether.h"
#ifdef GATEWAY
#define ARPTAB_BSIZ 16 /* bucket size */
#define ARPTAB_NB 37 /* number of buckets */
#else
#define ARPTAB_BSIZ 9 /* bucket size */
#define ARPTAB_NB 19 /* number of buckets */
#endif
#define ARPTAB_SIZE (ARPTAB_BSIZ * ARPTAB_NB)
struct arptab arptab[ARPTAB_SIZE];
int arptab_size = ARPTAB_SIZE; /* for arp command */
/*
* ARP trailer negotiation. Trailer protocol is not IP specific,
* but ARP request/response use IP addresses.
*/
#define ETHERTYPE_IPTRAILERS ETHERTYPE_TRAIL
#define ARPTAB_HASH(a) \
((u_long)(a) % ARPTAB_NB)
#define ARPTAB_LOOK(at,addr) { \
register n; \
at = &arptab[ARPTAB_HASH(addr) * ARPTAB_BSIZ]; \
for (n = 0 ; n < ARPTAB_BSIZ ; n++,at++) \
if (at->at_iaddr.s_addr == addr) \
break; \
if (n >= ARPTAB_BSIZ) \
at = 0; \
}
/* timer values */
#define ARPT_AGE (60*1) /* aging timer, 1 min. */
#define ARPT_KILLC 20 /* kill completed entry in 20 mins. */
#define ARPT_KILLI 3 /* kill incomplete entry in 3 minutes */
extern struct ifnet loif;
/*
* Timeout routine. Age arp_tab entries once a minute.
*/
arptimer()
{
register struct arptab *at;
register i;
timeout(arptimer, (caddr_t)0, ARPT_AGE * hz);
at = &arptab[0];
for (i = 0; i < ARPTAB_SIZE; i++, at++) {
if (at->at_flags == 0 || (at->at_flags & ATF_PERM))
continue;
if (++at->at_timer < ((at->at_flags&ATF_COM) ?
ARPT_KILLC : ARPT_KILLI))
continue;
/* timer has expired, clear entry */
arptfree(at);
}
}
/*
* Broadcast an ARP packet, asking who has addr on interface ac.
*/
arpwhohas(ac, addr)
register struct arpcom *ac;
struct in_addr *addr;
{
register struct mbuf *m;
register struct ether_header *eh;
register struct ether_arp *ea;
struct sockaddr sa;
if ((m = m_gethdr(M_DONTWAIT, MT_DATA)) == NULL)
return;
m->m_len = sizeof(*ea);
m->m_pkthdr.len = sizeof(*ea);
MH_ALIGN(m, sizeof(*ea));
ea = mtod(m, struct ether_arp *);
eh = (struct ether_header *)sa.sa_data;
bzero((caddr_t)ea, sizeof (*ea));
bcopy((caddr_t)etherbroadcastaddr, (caddr_t)eh->ether_dhost,
sizeof(eh->ether_dhost));
eh->ether_type = ETHERTYPE_ARP; /* if_output will swap */
ea->arp_hrd = htons(ARPHRD_ETHER);
ea->arp_pro = htons(ETHERTYPE_IP);
ea->arp_hln = sizeof(ea->arp_sha); /* hardware address length */
ea->arp_pln = sizeof(ea->arp_spa); /* protocol address length */
ea->arp_op = htons(ARPOP_REQUEST);
bcopy((caddr_t)ac->ac_enaddr, (caddr_t)ea->arp_sha,
sizeof(ea->arp_sha));
bcopy((caddr_t)&ac->ac_ipaddr, (caddr_t)ea->arp_spa,
sizeof(ea->arp_spa));
bcopy((caddr_t)addr, (caddr_t)ea->arp_tpa, sizeof(ea->arp_tpa));
sa.sa_family = AF_UNSPEC;
sa.sa_len = sizeof(sa);
(*ac->ac_if.if_output)(&ac->ac_if, m, &sa, (struct rtentry *)0);
}
int useloopback = 1; /* use loopback interface for local traffic */
/*
* Resolve an IP address into an ethernet address. If success,
* desten is filled in. If there is no entry in arptab,
* set one up and broadcast a request for the IP address.
* Hold onto this mbuf and resend it once the address
* is finally resolved. A return value of 1 indicates
* that desten has been filled in and the packet should be sent
* normally; a 0 return indicates that the packet has been
* taken over here, either now or for later transmission.
*
* We do some (conservative) locking here at splimp, since
* arptab is also altered from input interrupt service (ecintr/ilintr
* calls arpinput when ETHERTYPE_ARP packets come in).
*/
arpresolve(ac, m, destip, desten, usetrailers)
register struct arpcom *ac;
struct mbuf *m;
register struct in_addr *destip;
register u_char *desten;
int *usetrailers;
{
register struct arptab *at;
struct sockaddr_in sin;
register struct in_ifaddr *ia;
u_long lna;
int s;
*usetrailers = 0;
if (m->m_flags & M_BCAST) { /* broadcast */
bcopy((caddr_t)etherbroadcastaddr, (caddr_t)desten,
sizeof(etherbroadcastaddr));
return (1);
}
lna = in_lnaof(*destip);
/* if for us, use software loopback driver if up */
for (ia = in_ifaddr; ia; ia = ia->ia_next)
if ((ia->ia_ifp == &ac->ac_if) &&
(destip->s_addr == ia->ia_addr.sin_addr.s_addr)) {
/*
* This test used to be
* if (loif.if_flags & IFF_UP)
* It allowed local traffic to be forced
* through the hardware by configuring the loopback down.
* However, it causes problems during network configuration
* for boards that can't receive packets they send.
* It is now necessary to clear "useloopback"
* to force traffic out to the hardware.
*/
if (useloopback) {
sin.sin_family = AF_INET;
sin.sin_addr = *destip;
(void) looutput(&loif, m, (struct sockaddr *)&sin, 0);
/*
* The packet has already been sent and freed.
*/
return (0);
} else {
bcopy((caddr_t)ac->ac_enaddr, (caddr_t)desten,
sizeof(ac->ac_enaddr));
return (1);
}
}
s = splimp();
ARPTAB_LOOK(at, destip->s_addr);
if (at == 0) { /* not found */
if (ac->ac_if.if_flags & IFF_NOARP) {
bcopy((caddr_t)ac->ac_enaddr, (caddr_t)desten, 3);
desten[3] = (lna >> 16) & 0x7f;
desten[4] = (lna >> 8) & 0xff;
desten[5] = lna & 0xff;
splx(s);
return (1);
} else {
at = arptnew(destip);
if (at == 0)
panic("arpresolve: no free entry");
at->at_hold = m;
arpwhohas(ac, destip);
splx(s);
return (0);
}
}
at->at_timer = 0; /* restart the timer */
if (at->at_flags & ATF_COM) { /* entry IS complete */
bcopy((caddr_t)at->at_enaddr, (caddr_t)desten,
sizeof(at->at_enaddr));
if (at->at_flags & ATF_USETRAILERS)
*usetrailers = 1;
splx(s);
return (1);
}
/*
* There is an arptab entry, but no ethernet address
* response yet. Replace the held mbuf with this
* latest one.
*/
if (at->at_hold)
m_freem(at->at_hold);
at->at_hold = m;
arpwhohas(ac, destip); /* ask again */
splx(s);
return (0);
}
/*
* Called from 10 Mb/s Ethernet interrupt handlers
* when ether packet type ETHERTYPE_ARP
* is received. Common length and type checks are done here,
* then the protocol-specific routine is called.
*/
arpinput(ac, m)
struct arpcom *ac;
struct mbuf *m;
{
register struct arphdr *ar;
if (ac->ac_if.if_flags & IFF_NOARP)
goto out;
if (m->m_len < sizeof(struct arphdr))
goto out;
ar = mtod(m, struct arphdr *);
if (ntohs(ar->ar_hrd) != ARPHRD_ETHER)
goto out;
if (m->m_len < sizeof(struct arphdr) + 2 * ar->ar_hln + 2 * ar->ar_pln)
goto out;
switch (ntohs(ar->ar_pro)) {
case ETHERTYPE_IP:
case ETHERTYPE_IPTRAILERS:
in_arpinput(ac, m);
return;
default:
break;
}
out:
m_freem(m);
}
/*
* 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 also handle negotiations for use of trailer protocol:
* ARP replies for protocol type ETHERTYPE_TRAIL are sent
* along with IP replies if we want trailers sent to us,
* and also send them in response to IP replies.
* This allows either end to announce the desire to receive
* trailer packets.
* We reply to requests for ETHERTYPE_TRAIL protocol as well,
* but don't normally send requests.
*/
in_arpinput(ac, m)
register struct arpcom *ac;
struct mbuf *m;
{
register struct ether_arp *ea;
struct ether_header *eh;
register struct arptab *at; /* same as "merge" flag */
register struct in_ifaddr *ia;
struct in_ifaddr *maybe_ia = 0;
struct mbuf *mcopy = 0;
struct sockaddr_in sin;
struct sockaddr sa;
struct in_addr isaddr, itaddr, myaddr;
int proto, op, s, completed = 0;
ea = mtod(m, struct ether_arp *);
proto = ntohs(ea->arp_pro);
op = ntohs(ea->arp_op);
bcopy((caddr_t)ea->arp_spa, (caddr_t)&isaddr, sizeof (isaddr));
bcopy((caddr_t)ea->arp_tpa, (caddr_t)&itaddr, sizeof (itaddr));
for (ia = in_ifaddr; ia; ia = ia->ia_next)
if (ia->ia_ifp == &ac->ac_if) {
maybe_ia = ia;
if ((itaddr.s_addr == ia->ia_addr.sin_addr.s_addr) ||
(isaddr.s_addr == ia->ia_addr.sin_addr.s_addr))
break;
}
if (maybe_ia == 0)
goto out;
myaddr = ia ? ia->ia_addr.sin_addr : maybe_ia->ia_addr.sin_addr;
if (!bcmp((caddr_t)ea->arp_sha, (caddr_t)ac->ac_enaddr,
sizeof (ea->arp_sha)))
goto out; /* it's from me, ignore it. */
if (!bcmp((caddr_t)ea->arp_sha, (caddr_t)etherbroadcastaddr,
sizeof (ea->arp_sha))) {
log(LOG_ERR,
"arp: ether address is broadcast for IP address %x!\n",
ntohl(isaddr.s_addr));
goto out;
}
if (isaddr.s_addr == myaddr.s_addr) {
log(LOG_ERR,
"duplicate IP address %x!! sent from ethernet address: %s\n",
ntohl(isaddr.s_addr), ether_sprintf(ea->arp_sha));
itaddr = myaddr;
if (op == ARPOP_REQUEST)
goto reply;
goto out;
}
s = splimp();
ARPTAB_LOOK(at, isaddr.s_addr);
if (at) {
bcopy((caddr_t)ea->arp_sha, (caddr_t)at->at_enaddr,
sizeof(ea->arp_sha));
if ((at->at_flags & ATF_COM) == 0)
completed = 1;
at->at_flags |= ATF_COM;
if (at->at_hold) {
sin.sin_family = AF_INET;
sin.sin_addr = isaddr;
(*ac->ac_if.if_output)(&ac->ac_if, at->at_hold,
(struct sockaddr *)&sin, (struct rtentry *)0);
at->at_hold = 0;
}
}
if (at == 0 && itaddr.s_addr == myaddr.s_addr) {
/* ensure we have a table entry */
if (at = arptnew(&isaddr)) {
bcopy((caddr_t)ea->arp_sha, (caddr_t)at->at_enaddr,
sizeof(ea->arp_sha));
completed = 1;
at->at_flags |= ATF_COM;
}
}
splx(s);
reply:
switch (proto) {
case ETHERTYPE_IPTRAILERS:
/* partner says trailers are OK */
if (at)
at->at_flags |= ATF_USETRAILERS;
/*
* Reply to request iff we want trailers.
*/
if (op != ARPOP_REQUEST || ac->ac_if.if_flags & IFF_NOTRAILERS)
goto out;
break;
case ETHERTYPE_IP:
/*
* Reply if this is an IP request,
* or if we want to send a trailer response.
* Send the latter only to the IP response
* that completes the current ARP entry.
*/
if (op != ARPOP_REQUEST &&
(completed == 0 || ac->ac_if.if_flags & IFF_NOTRAILERS))
goto out;
}
if (itaddr.s_addr == myaddr.s_addr) {
/* I am the target */
bcopy((caddr_t)ea->arp_sha, (caddr_t)ea->arp_tha,
sizeof(ea->arp_sha));
bcopy((caddr_t)ac->ac_enaddr, (caddr_t)ea->arp_sha,
sizeof(ea->arp_sha));
} else {
ARPTAB_LOOK(at, itaddr.s_addr);
if (at == NULL || (at->at_flags & ATF_PUBL) == 0)
goto out;
bcopy((caddr_t)ea->arp_sha, (caddr_t)ea->arp_tha,
sizeof(ea->arp_sha));
bcopy((caddr_t)at->at_enaddr, (caddr_t)ea->arp_sha,
sizeof(ea->arp_sha));
}
bcopy((caddr_t)ea->arp_spa, (caddr_t)ea->arp_tpa,
sizeof(ea->arp_spa));
bcopy((caddr_t)&itaddr, (caddr_t)ea->arp_spa,
sizeof(ea->arp_spa));
ea->arp_op = htons(ARPOP_REPLY);
/*
* If incoming packet was an IP reply,
* we are sending a reply for type IPTRAILERS.
* If we are sending a reply for type IP
* and we want to receive trailers,
* send a trailer reply as well.
*/
if (op == ARPOP_REPLY)
ea->arp_pro = htons(ETHERTYPE_IPTRAILERS);
else if (proto == ETHERTYPE_IP &&
(ac->ac_if.if_flags & IFF_NOTRAILERS) == 0)
mcopy = m_copy(m, 0, (int)M_COPYALL);
eh = (struct ether_header *)sa.sa_data;
bcopy((caddr_t)ea->arp_tha, (caddr_t)eh->ether_dhost,
sizeof(eh->ether_dhost));
eh->ether_type = ETHERTYPE_ARP;
sa.sa_family = AF_UNSPEC;
sa.sa_len = sizeof(sa);
(*ac->ac_if.if_output)(&ac->ac_if, m, &sa, (struct rtentry *)0);
if (mcopy) {
ea = mtod(mcopy, struct ether_arp *);
ea->arp_pro = htons(ETHERTYPE_IPTRAILERS);
(*ac->ac_if.if_output)(&ac->ac_if,
mcopy, &sa, (struct rtentry *)0);
}
return;
out:
m_freem(m);
return;
}
/*
* Free an arptab entry.
*/
arptfree(at)
register struct arptab *at;
{
int s = splimp();
if (at->at_hold)
m_freem(at->at_hold);
at->at_hold = 0;
at->at_timer = at->at_flags = 0;
at->at_iaddr.s_addr = 0;
splx(s);
}
/*
* Enter a new address in arptab, pushing out the oldest entry
* from the bucket if there is no room.
* This always succeeds since no bucket can be completely filled
* with permanent entries (except from arpioctl when testing whether
* another permanent entry will fit).
* MUST BE CALLED AT SPLIMP.
*/
struct arptab *
arptnew(addr)
struct in_addr *addr;
{
register n;
int oldest = -1;
register struct arptab *at, *ato = NULL;
static int first = 1;
if (first) {
first = 0;
timeout(arptimer, (caddr_t)0, hz);
}
at = &arptab[ARPTAB_HASH(addr->s_addr) * ARPTAB_BSIZ];
for (n = 0; n < ARPTAB_BSIZ; n++,at++) {
if (at->at_flags == 0)
goto out; /* found an empty entry */
if (at->at_flags & ATF_PERM)
continue;
if ((int) at->at_timer > oldest) {
oldest = at->at_timer;
ato = at;
}
}
if (ato == NULL)
return (NULL);
at = ato;
arptfree(at);
out:
at->at_iaddr = *addr;
at->at_flags = ATF_INUSE;
return (at);
}
arpioctl(cmd, data)
int cmd;
caddr_t data;
{
register struct arpreq *ar = (struct arpreq *)data;
register struct arptab *at;
register struct sockaddr_in *sin;
int s;
sin = (struct sockaddr_in *)&ar->arp_ha;
#if defined(COMPAT_43) && BYTE_ORDER != BIG_ENDIAN
if (sin->sin_family == 0 && sin->sin_len < 16)
sin->sin_family = sin->sin_len;
#endif
sin->sin_len = sizeof(ar->arp_ha);
sin = (struct sockaddr_in *)&ar->arp_pa;
#if defined(COMPAT_43) && BYTE_ORDER != BIG_ENDIAN
if (sin->sin_family == 0 && sin->sin_len < 16)
sin->sin_family = sin->sin_len;
#endif
sin->sin_len = sizeof(ar->arp_pa);
if (ar->arp_pa.sa_family != AF_INET ||
ar->arp_ha.sa_family != AF_UNSPEC)
return (EAFNOSUPPORT);
s = splimp();
ARPTAB_LOOK(at, sin->sin_addr.s_addr);
if (at == NULL) { /* not found */
if (cmd != SIOCSARP) {
splx(s);
return (ENXIO);
}
if (ifa_ifwithnet(&ar->arp_pa) == NULL) {
splx(s);
return (ENETUNREACH);
}
}
switch (cmd) {
case SIOCSARP: /* set entry */
if (at == NULL) {
at = arptnew(&sin->sin_addr);
if (at == NULL) {
splx(s);
return (EADDRNOTAVAIL);
}
if (ar->arp_flags & ATF_PERM) {
/* never make all entries in a bucket permanent */
register struct arptab *tat;
/* try to re-allocate */
tat = arptnew(&sin->sin_addr);
if (tat == NULL) {
arptfree(at);
splx(s);
return (EADDRNOTAVAIL);
}
arptfree(tat);
}
}
bcopy((caddr_t)ar->arp_ha.sa_data, (caddr_t)at->at_enaddr,
sizeof(at->at_enaddr));
at->at_flags = ATF_COM | ATF_INUSE |
(ar->arp_flags & (ATF_PERM|ATF_PUBL|ATF_USETRAILERS));
at->at_timer = 0;
break;
case SIOCDARP: /* delete entry */
arptfree(at);
break;
case SIOCGARP: /* get entry */
case OSIOCGARP:
bcopy((caddr_t)at->at_enaddr, (caddr_t)ar->arp_ha.sa_data,
sizeof(at->at_enaddr));
#ifdef COMPAT_43
if (cmd == OSIOCGARP)
*(u_short *)&ar->arp_ha = ar->arp_ha.sa_family;
#endif
ar->arp_flags = at->at_flags;
break;
}
splx(s);
return (0);
}