freebsd-nq/sys/netinet/ip_output.c

1572 lines
40 KiB
C
Raw Normal View History

1994-05-24 10:09:53 +00:00
/*
* Copyright (c) 1982, 1986, 1988, 1990, 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.
* 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.
*
* @(#)ip_output.c 8.3 (Berkeley) 1/21/94
1999-08-28 01:08:13 +00:00
* $FreeBSD$
1994-05-24 10:09:53 +00:00
*/
#define _IP_VHL
#include "opt_ipfw.h"
#include "opt_ipdn.h"
#include "opt_ipdivert.h"
#include "opt_ipfilter.h"
1994-05-24 10:09:53 +00:00
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
1994-05-24 10:09:53 +00:00
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/protosw.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <net/if.h>
#include <net/route.h>
#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/ip.h>
#include <netinet/in_pcb.h>
#include <netinet/in_var.h>
#include <netinet/ip_var.h>
#ifdef vax
#include <machine/mtpr.h>
#endif
#include <machine/in_cksum.h>
1994-05-24 10:09:53 +00:00
static MALLOC_DEFINE(M_IPMOPTS, "ip_moptions", "internet multicast options");
#include <netinet/ip_fw.h>
#ifdef DUMMYNET
#include <netinet/ip_dummynet.h>
#endif
#ifdef IPFIREWALL_FORWARD_DEBUG
#define print_ip(a) printf("%ld.%ld.%ld.%ld",(ntohl(a.s_addr)>>24)&0xFF,\
(ntohl(a.s_addr)>>16)&0xFF,\
(ntohl(a.s_addr)>>8)&0xFF,\
(ntohl(a.s_addr))&0xFF);
#endif
u_short ip_id;
1994-05-24 10:09:53 +00:00
static struct mbuf *ip_insertoptions __P((struct mbuf *, struct mbuf *, int *));
static void ip_mloopback
__P((struct ifnet *, struct mbuf *, struct sockaddr_in *, int));
static int ip_getmoptions
__P((struct sockopt *, struct ip_moptions *));
static int ip_pcbopts __P((int, struct mbuf **, struct mbuf *));
static int ip_setmoptions
__P((struct sockopt *, struct ip_moptions **));
1994-05-24 10:09:53 +00:00
1997-02-10 11:45:37 +00:00
#if defined(IPFILTER_LKM) || defined(IPFILTER)
1997-04-03 10:47:12 +00:00
int ip_optcopy __P((struct ip *, struct ip *));
1997-02-10 11:45:37 +00:00
extern int (*fr_checkp) __P((struct ip *, int, struct ifnet *, int, struct mbuf **));
1997-04-03 10:47:12 +00:00
#else
static int ip_optcopy __P((struct ip *, struct ip *));
1997-02-10 11:45:37 +00:00
#endif
extern struct protosw inetsw[];
1994-05-24 10:09:53 +00:00
/*
* IP output. The packet in mbuf chain m contains a skeletal IP
* header (with len, off, ttl, proto, tos, src, dst).
* The mbuf chain containing the packet will be freed.
* The mbuf opt, if present, will not be freed.
*/
int
ip_output(m0, opt, ro, flags, imo)
struct mbuf *m0;
struct mbuf *opt;
struct route *ro;
int flags;
struct ip_moptions *imo;
{
struct ip *ip, *mhip;
struct ifnet *ifp;
struct mbuf *m = m0;
int hlen = sizeof (struct ip);
1994-05-24 10:09:53 +00:00
int len, off, error = 0;
struct sockaddr_in *dst;
struct in_ifaddr *ia;
int isbroadcast;
u_int16_t divert_cookie; /* firewall cookie */
#ifdef IPFIREWALL_FORWARD
int fwd_rewrite_src = 0;
#endif
struct ip_fw_chain *rule = NULL;
#ifdef IPDIVERT
/* Get and reset firewall cookie */
divert_cookie = ip_divert_cookie;
ip_divert_cookie = 0;
#else
divert_cookie = 0;
#endif
#if defined(IPFIREWALL) && defined(DUMMYNET)
/*
* dummynet packet are prepended a vestigial mbuf with
* m_type = MT_DUMMYNET and m_data pointing to the matching
* rule.
*/
if (m->m_type == MT_DUMMYNET) {
/*
* the packet was already tagged, so part of the
* processing was already done, and we need to go down.
* opt, flags and imo have already been used, and now
* they are used to hold ifp, dst and NULL, respectively.
*/
rule = (struct ip_fw_chain *)(m->m_data) ;
m0 = m = m->m_next ;
ip = mtod(m, struct ip *);
dst = (struct sockaddr_in *)flags ;
ifp = (struct ifnet *)opt;
hlen = IP_VHL_HL(ip->ip_vhl) << 2 ;
opt = NULL ;
flags = 0 ; /* XXX is this correct ? */
goto sendit;
} else
rule = NULL ;
#endif
1994-05-24 10:09:53 +00:00
#ifdef DIAGNOSTIC
if ((m->m_flags & M_PKTHDR) == 0)
panic("ip_output no HDR");
if (!ro)
panic("ip_output no route, proto = %d",
mtod(m, struct ip *)->ip_p);
1994-05-24 10:09:53 +00:00
#endif
if (opt) {
m = ip_insertoptions(m, opt, &len);
hlen = len;
}
ip = mtod(m, struct ip *);
/*
* Fill in IP header.
*/
if ((flags & (IP_FORWARDING|IP_RAWOUTPUT)) == 0) {
ip->ip_vhl = IP_MAKE_VHL(IPVERSION, hlen >> 2);
1994-05-24 10:09:53 +00:00
ip->ip_off &= IP_DF;
ip->ip_id = htons(ip_id++);
ipstat.ips_localout++;
} else {
hlen = IP_VHL_HL(ip->ip_vhl) << 2;
1994-05-24 10:09:53 +00:00
}
1994-05-24 10:09:53 +00:00
dst = (struct sockaddr_in *)&ro->ro_dst;
/*
* If there is a cached route,
* check that it is to the same destination
* and is still up. If not, free it and try again.
*/
if (ro->ro_rt && ((ro->ro_rt->rt_flags & RTF_UP) == 0 ||
dst->sin_addr.s_addr != ip->ip_dst.s_addr)) {
RTFREE(ro->ro_rt);
ro->ro_rt = (struct rtentry *)0;
}
if (ro->ro_rt == 0) {
dst->sin_family = AF_INET;
dst->sin_len = sizeof(*dst);
dst->sin_addr = ip->ip_dst;
}
/*
* If routing to interface only,
* short circuit routing lookup.
*/
#define ifatoia(ifa) ((struct in_ifaddr *)(ifa))
#define sintosa(sin) ((struct sockaddr *)(sin))
if (flags & IP_ROUTETOIF) {
if ((ia = ifatoia(ifa_ifwithdstaddr(sintosa(dst)))) == 0 &&
(ia = ifatoia(ifa_ifwithnet(sintosa(dst)))) == 0) {
ipstat.ips_noroute++;
error = ENETUNREACH;
goto bad;
}
ifp = ia->ia_ifp;
ip->ip_ttl = 1;
isbroadcast = in_broadcast(dst->sin_addr, ifp);
1994-05-24 10:09:53 +00:00
} else {
/*
* If this is the case, we probably don't want to allocate
* a protocol-cloned route since we didn't get one from the
* ULP. This lets TCP do its thing, while not burdening
* forwarding or ICMP with the overhead of cloning a route.
* Of course, we still want to do any cloning requested by
* the link layer, as this is probably required in all cases
* for correct operation (as it is for ARP).
*/
1994-05-24 10:09:53 +00:00
if (ro->ro_rt == 0)
rtalloc_ign(ro, RTF_PRCLONING);
1994-05-24 10:09:53 +00:00
if (ro->ro_rt == 0) {
ipstat.ips_noroute++;
error = EHOSTUNREACH;
goto bad;
}
ia = ifatoia(ro->ro_rt->rt_ifa);
ifp = ro->ro_rt->rt_ifp;
ro->ro_rt->rt_use++;
if (ro->ro_rt->rt_flags & RTF_GATEWAY)
dst = (struct sockaddr_in *)ro->ro_rt->rt_gateway;
if (ro->ro_rt->rt_flags & RTF_HOST)
isbroadcast = (ro->ro_rt->rt_flags & RTF_BROADCAST);
else
isbroadcast = in_broadcast(dst->sin_addr, ifp);
1994-05-24 10:09:53 +00:00
}
if (IN_MULTICAST(ntohl(ip->ip_dst.s_addr))) {
struct in_multi *inm;
m->m_flags |= M_MCAST;
/*
* IP destination address is multicast. Make sure "dst"
* still points to the address in "ro". (It may have been
* changed to point to a gateway address, above.)
*/
dst = (struct sockaddr_in *)&ro->ro_dst;
/*
* See if the caller provided any multicast options
*/
if (imo != NULL) {
ip->ip_ttl = imo->imo_multicast_ttl;
if (imo->imo_multicast_ifp != NULL)
ifp = imo->imo_multicast_ifp;
if (imo->imo_multicast_vif != -1)
ip->ip_src.s_addr =
ip_mcast_src(imo->imo_multicast_vif);
1994-05-24 10:09:53 +00:00
} else
ip->ip_ttl = IP_DEFAULT_MULTICAST_TTL;
/*
* Confirm that the outgoing interface supports multicast.
*/
if ((imo == NULL) || (imo->imo_multicast_vif == -1)) {
if ((ifp->if_flags & IFF_MULTICAST) == 0) {
ipstat.ips_noroute++;
error = ENETUNREACH;
goto bad;
}
1994-05-24 10:09:53 +00:00
}
/*
* If source address not specified yet, use address
* of outgoing interface.
*/
if (ip->ip_src.s_addr == INADDR_ANY) {
register struct in_ifaddr *ia1;
1994-05-24 10:09:53 +00:00
for (ia1 = in_ifaddrhead.tqh_first; ia1;
ia1 = ia1->ia_link.tqe_next)
if (ia1->ia_ifp == ifp) {
ip->ip_src = IA_SIN(ia1)->sin_addr;
1994-05-24 10:09:53 +00:00
break;
}
}
IN_LOOKUP_MULTI(ip->ip_dst, ifp, inm);
if (inm != NULL &&
(imo == NULL || imo->imo_multicast_loop)) {
/*
* If we belong to the destination multicast group
* on the outgoing interface, and the caller did not
* forbid loopback, loop back a copy.
*/
ip_mloopback(ifp, m, dst, hlen);
1994-05-24 10:09:53 +00:00
}
else {
/*
* If we are acting as a multicast router, perform
* multicast forwarding as if the packet had just
* arrived on the interface to which we are about
* to send. The multicast forwarding function
* recursively calls this function, using the
* IP_FORWARDING flag to prevent infinite recursion.
*
* Multicasts that are looped back by ip_mloopback(),
* above, will be forwarded by the ip_input() routine,
* if necessary.
*/
if (ip_mrouter && (flags & IP_FORWARDING) == 0) {
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/*
* Check if rsvp daemon is running. If not, don't
* set ip_moptions. This ensures that the packet
* is multicast and not just sent down one link
* as prescribed by rsvpd.
*/
if (!rsvp_on)
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
imo = NULL;
if (ip_mforward(ip, ifp, m, imo) != 0) {
1994-05-24 10:09:53 +00:00
m_freem(m);
goto done;
}
}
}
1994-05-24 10:09:53 +00:00
/*
* Multicasts with a time-to-live of zero may be looped-
* back, above, but must not be transmitted on a network.
* Also, multicasts addressed to the loopback interface
* are not sent -- the above call to ip_mloopback() will
* loop back a copy if this host actually belongs to the
* destination group on the loopback interface.
*/
if (ip->ip_ttl == 0 || ifp->if_flags & IFF_LOOPBACK) {
1994-05-24 10:09:53 +00:00
m_freem(m);
goto done;
}
goto sendit;
}
#ifndef notdef
/*
* If source address not specified yet, use address
* of outgoing interface.
*/
if (ip->ip_src.s_addr == INADDR_ANY) {
1994-05-24 10:09:53 +00:00
ip->ip_src = IA_SIN(ia)->sin_addr;
#ifdef IPFIREWALL_FORWARD
/* Keep note that we did this - if the firewall changes
* the next-hop, our interface may change, changing the
* default source IP. It's a shame so much effort happens
* twice. Oh well.
*/
fwd_rewrite_src++;
#endif /* IPFIREWALL_FORWARD */
}
#endif /* notdef */
/*
* Verify that we have any chance at all of being able to queue
* the packet or packet fragments
*/
if ((ifp->if_snd.ifq_len + ip->ip_len / ifp->if_mtu + 1) >=
ifp->if_snd.ifq_maxlen) {
error = ENOBUFS;
goto bad;
}
1994-05-24 10:09:53 +00:00
/*
* Look for broadcast address and
* and verify user is allowed to send
* such a packet.
*/
if (isbroadcast) {
1994-05-24 10:09:53 +00:00
if ((ifp->if_flags & IFF_BROADCAST) == 0) {
error = EADDRNOTAVAIL;
goto bad;
}
if ((flags & IP_ALLOWBROADCAST) == 0) {
error = EACCES;
goto bad;
}
/* don't allow broadcast messages to be fragmented */
if ((u_short)ip->ip_len > ifp->if_mtu) {
error = EMSGSIZE;
goto bad;
}
m->m_flags |= M_BCAST;
} else {
1994-05-24 10:09:53 +00:00
m->m_flags &= ~M_BCAST;
}
1994-05-24 10:09:53 +00:00
sendit:
1997-02-10 11:45:37 +00:00
/*
* IpHack's section.
* - Xlate: translate packet's addr/port (NAT).
* - Firewall: deny/allow/etc.
* - Wrap: fake packet's addr/port <unimpl.>
* - Encapsulate: put it in another IP and send out. <unimp.>
*/
#if defined(IPFILTER) || defined(IPFILTER_LKM)
if (fr_checkp) {
struct mbuf *m1 = m;
if ((error = (*fr_checkp)(ip, hlen, ifp, 1, &m1)) || !m1)
goto done;
ip = mtod(m = m1, struct ip *);
1997-02-10 11:45:37 +00:00
}
#endif
/*
* Check with the firewall...
*/
if (ip_fw_chk_ptr) {
struct sockaddr_in *old = dst;
off = (*ip_fw_chk_ptr)(&ip,
hlen, ifp, &divert_cookie, &m, &rule, &dst);
/*
* On return we must do the following:
* m == NULL -> drop the pkt
* 1<=off<= 0xffff -> DIVERT
* (off & 0x10000) -> send to a DUMMYNET pipe
* (off & 0x20000) -> TEE the packet
* dst != old -> IPFIREWALL_FORWARD
* off==0, dst==old -> accept
* If some of the above modules is not compiled in, then
* we should't have to check the corresponding condition
* (because the ipfw control socket should not accept
* unsupported rules), but better play safe and drop
* packets in case of doubt.
*/
if (!m) { /* firewall said to reject */
error = EACCES;
goto done;
}
if (off == 0 && dst == old) /* common case */
goto pass ;
#ifdef DUMMYNET
if ((off & IP_FW_PORT_DYNT_FLAG) != 0) {
/*
* pass the pkt to dummynet. Need to include
* pipe number, m, ifp, ro, dst because these are
* not recomputed in the next pass.
* All other parameters have been already used and
* so they are not needed anymore.
* XXX note: if the ifp or ro entry are deleted
* while a pkt is in dummynet, we are in trouble!
*/
dummynet_io(off & 0xffff, DN_TO_IP_OUT, m,ifp,ro,dst,rule);
goto done;
}
#endif
#ifdef IPDIVERT
if (off != 0 && (off & IP_FW_PORT_DYNT_FLAG) == 0) {
struct mbuf *clone = NULL;
/* Clone packet if we're doing a 'tee' */
if ((off & IP_FW_PORT_TEE_FLAG) != 0)
clone = m_dup(m, M_DONTWAIT);
/* Restore packet header fields to original values */
HTONS(ip->ip_len);
HTONS(ip->ip_off);
/* Deliver packet to divert input routine */
ip_divert_cookie = divert_cookie;
divert_packet(m, 0, off & 0xffff);
/* If 'tee', continue with original packet */
if (clone != NULL) {
m = clone;
ip = mtod(m, struct ip *);
goto pass;
}
goto done;
}
#endif
#ifdef IPFIREWALL_FORWARD
/* Here we check dst to make sure it's directly reachable on the
* interface we previously thought it was.
* If it isn't (which may be likely in some situations) we have
* to re-route it (ie, find a route for the next-hop and the
* associated interface) and set them here. This is nested
* forwarding which in most cases is undesirable, except where
* such control is nigh impossible. So we do it here.
* And I'm babbling.
*/
if (off == 0 && old != dst) {
struct in_ifaddr *ia;
/* It's changed... */
/* There must be a better way to do this next line... */
static struct route sro_fwd, *ro_fwd = &sro_fwd;
#ifdef IPFIREWALL_FORWARD_DEBUG
printf("IPFIREWALL_FORWARD: New dst ip: ");
print_ip(dst->sin_addr);
printf("\n");
#endif
/*
* We need to figure out if we have been forwarded
* to a local socket. If so then we should somehow
* "loop back" to ip_input, and get directed to the
* PCB as if we had received this packet. This is
* because it may be dificult to identify the packets
* you want to forward until they are being output
* and have selected an interface. (e.g. locally
* initiated packets) If we used the loopback inteface,
* we would not be able to control what happens
* as the packet runs through ip_input() as
* it is done through a ISR.
*/
for (ia = TAILQ_FIRST(&in_ifaddrhead); ia;
ia = TAILQ_NEXT(ia, ia_link)) {
/*
* If the addr to forward to is one
* of ours, we pretend to
* be the destination for this packet.
*/
if (IA_SIN(ia)->sin_addr.s_addr ==
dst->sin_addr.s_addr)
break;
}
if (ia) {
/* tell ip_input "dont filter" */
ip_fw_fwd_addr = dst;
if (m->m_pkthdr.rcvif == NULL)
m->m_pkthdr.rcvif = ifunit("lo0");
ip->ip_len = htons((u_short)ip->ip_len);
ip->ip_off = htons((u_short)ip->ip_off);
ip->ip_sum = 0;
if (ip->ip_vhl == IP_VHL_BORING) {
ip->ip_sum = in_cksum_hdr(ip);
} else {
ip->ip_sum = in_cksum(m, hlen);
}
ip_input(m);
goto done;
}
/* Some of the logic for this was
* nicked from above.
*
* This rewrites the cached route in a local PCB.
* Is this what we want to do?
*/
bcopy(dst, &ro_fwd->ro_dst, sizeof(*dst));
ro_fwd->ro_rt = 0;
rtalloc_ign(ro_fwd, RTF_PRCLONING);
if (ro_fwd->ro_rt == 0) {
ipstat.ips_noroute++;
error = EHOSTUNREACH;
goto bad;
}
ia = ifatoia(ro_fwd->ro_rt->rt_ifa);
ifp = ro_fwd->ro_rt->rt_ifp;
ro_fwd->ro_rt->rt_use++;
if (ro_fwd->ro_rt->rt_flags & RTF_GATEWAY)
dst = (struct sockaddr_in *)ro_fwd->ro_rt->rt_gateway;
if (ro_fwd->ro_rt->rt_flags & RTF_HOST)
isbroadcast =
(ro_fwd->ro_rt->rt_flags & RTF_BROADCAST);
else
isbroadcast = in_broadcast(dst->sin_addr, ifp);
RTFREE(ro->ro_rt);
ro->ro_rt = ro_fwd->ro_rt;
dst = (struct sockaddr_in *)&ro_fwd->ro_dst;
/*
* If we added a default src ip earlier,
* which would have been gotten from the-then
* interface, do it again, from the new one.
*/
if (fwd_rewrite_src)
ip->ip_src = IA_SIN(ia)->sin_addr;
goto pass ;
}
#endif /* IPFIREWALL_FORWARD */
/*
* if we get here, none of the above matches, and
* we have to drop the pkt
*/
m_freem(m);
error = EACCES; /* not sure this is the right error msg */
goto done;
}
pass:
1994-05-24 10:09:53 +00:00
/*
* If small enough for interface, can just send directly.
*/
if ((u_short)ip->ip_len <= ifp->if_mtu) {
ip->ip_len = htons((u_short)ip->ip_len);
ip->ip_off = htons((u_short)ip->ip_off);
ip->ip_sum = 0;
if (ip->ip_vhl == IP_VHL_BORING) {
ip->ip_sum = in_cksum_hdr(ip);
} else {
ip->ip_sum = in_cksum(m, hlen);
}
1994-05-24 10:09:53 +00:00
error = (*ifp->if_output)(ifp, m,
(struct sockaddr *)dst, ro->ro_rt);
goto done;
}
/*
* Too large for interface; fragment if possible.
* Must be able to put at least 8 bytes per fragment.
*/
if (ip->ip_off & IP_DF) {
error = EMSGSIZE;
/*
* This case can happen if the user changed the MTU
* of an interface after enabling IP on it. Because
* most netifs don't keep track of routes pointing to
* them, there is no way for one to update all its
* routes when the MTU is changed.
*/
if ((ro->ro_rt->rt_flags & (RTF_UP | RTF_HOST))
&& !(ro->ro_rt->rt_rmx.rmx_locks & RTV_MTU)
&& (ro->ro_rt->rt_rmx.rmx_mtu > ifp->if_mtu)) {
ro->ro_rt->rt_rmx.rmx_mtu = ifp->if_mtu;
}
1994-05-24 10:09:53 +00:00
ipstat.ips_cantfrag++;
goto bad;
}
len = (ifp->if_mtu - hlen) &~ 7;
if (len < 8) {
error = EMSGSIZE;
goto bad;
}
{
int mhlen, firstlen = len;
struct mbuf **mnext = &m->m_nextpkt;
/*
* Loop through length of segment after first fragment,
* make new header and copy data of each part and link onto chain.
*/
m0 = m;
mhlen = sizeof (struct ip);
for (off = hlen + len; off < (u_short)ip->ip_len; off += len) {
MGETHDR(m, M_DONTWAIT, MT_HEADER);
if (m == 0) {
error = ENOBUFS;
ipstat.ips_odropped++;
goto sendorfree;
}
m->m_flags |= (m0->m_flags & M_MCAST);
1994-05-24 10:09:53 +00:00
m->m_data += max_linkhdr;
mhip = mtod(m, struct ip *);
*mhip = *ip;
if (hlen > sizeof (struct ip)) {
mhlen = ip_optcopy(ip, mhip) + sizeof (struct ip);
mhip->ip_vhl = IP_MAKE_VHL(IPVERSION, mhlen >> 2);
1994-05-24 10:09:53 +00:00
}
m->m_len = mhlen;
mhip->ip_off = ((off - hlen) >> 3) + (ip->ip_off & ~IP_MF);
if (ip->ip_off & IP_MF)
mhip->ip_off |= IP_MF;
if (off + len >= (u_short)ip->ip_len)
len = (u_short)ip->ip_len - off;
else
mhip->ip_off |= IP_MF;
mhip->ip_len = htons((u_short)(len + mhlen));
m->m_next = m_copy(m0, off, len);
if (m->m_next == 0) {
(void) m_free(m);
error = ENOBUFS; /* ??? */
ipstat.ips_odropped++;
goto sendorfree;
}
m->m_pkthdr.len = mhlen + len;
m->m_pkthdr.rcvif = (struct ifnet *)0;
mhip->ip_off = htons((u_short)mhip->ip_off);
mhip->ip_sum = 0;
if (mhip->ip_vhl == IP_VHL_BORING) {
mhip->ip_sum = in_cksum_hdr(mhip);
} else {
mhip->ip_sum = in_cksum(m, mhlen);
}
1994-05-24 10:09:53 +00:00
*mnext = m;
mnext = &m->m_nextpkt;
ipstat.ips_ofragments++;
}
/*
* Update first fragment by trimming what's been copied out
* and updating header, then send each fragment (in order).
*/
m = m0;
m_adj(m, hlen + firstlen - (u_short)ip->ip_len);
m->m_pkthdr.len = hlen + firstlen;
ip->ip_len = htons((u_short)m->m_pkthdr.len);
ip->ip_off = htons((u_short)(ip->ip_off | IP_MF));
ip->ip_sum = 0;
if (ip->ip_vhl == IP_VHL_BORING) {
ip->ip_sum = in_cksum_hdr(ip);
} else {
ip->ip_sum = in_cksum(m, hlen);
}
1994-05-24 10:09:53 +00:00
sendorfree:
for (m = m0; m; m = m0) {
m0 = m->m_nextpkt;
m->m_nextpkt = 0;
if (error == 0)
error = (*ifp->if_output)(ifp, m,
(struct sockaddr *)dst, ro->ro_rt);
else
m_freem(m);
}
if (error == 0)
ipstat.ips_fragmented++;
}
done:
return (error);
bad:
m_freem(m0);
goto done;
}
/*
* Insert IP options into preformed packet.
* Adjust IP destination as required for IP source routing,
* as indicated by a non-zero in_addr at the start of the options.
*
* XXX This routine assumes that the packet has no options in place.
1994-05-24 10:09:53 +00:00
*/
static struct mbuf *
ip_insertoptions(m, opt, phlen)
register struct mbuf *m;
struct mbuf *opt;
int *phlen;
{
register struct ipoption *p = mtod(opt, struct ipoption *);
struct mbuf *n;
register struct ip *ip = mtod(m, struct ip *);
unsigned optlen;
optlen = opt->m_len - sizeof(p->ipopt_dst);
if (optlen + (u_short)ip->ip_len > IP_MAXPACKET)
return (m); /* XXX should fail */
if (p->ipopt_dst.s_addr)
ip->ip_dst = p->ipopt_dst;
if (m->m_flags & M_EXT || m->m_data - optlen < m->m_pktdat) {
MGETHDR(n, M_DONTWAIT, MT_HEADER);
if (n == 0)
return (m);
n->m_pkthdr.len = m->m_pkthdr.len + optlen;
m->m_len -= sizeof(struct ip);
m->m_data += sizeof(struct ip);
n->m_next = m;
m = n;
m->m_len = optlen + sizeof(struct ip);
m->m_data += max_linkhdr;
(void)memcpy(mtod(m, void *), ip, sizeof(struct ip));
1994-05-24 10:09:53 +00:00
} else {
m->m_data -= optlen;
m->m_len += optlen;
m->m_pkthdr.len += optlen;
ovbcopy((caddr_t)ip, mtod(m, caddr_t), sizeof(struct ip));
}
ip = mtod(m, struct ip *);
bcopy(p->ipopt_list, ip + 1, optlen);
1994-05-24 10:09:53 +00:00
*phlen = sizeof(struct ip) + optlen;
ip->ip_vhl = IP_MAKE_VHL(IPVERSION, *phlen >> 2);
1994-05-24 10:09:53 +00:00
ip->ip_len += optlen;
return (m);
}
/*
* Copy options from ip to jp,
* omitting those not copied during fragmentation.
*/
1997-04-03 10:47:12 +00:00
#if !defined(IPFILTER) && !defined(IPFILTER_LKM)
static
#endif
int
1994-05-24 10:09:53 +00:00
ip_optcopy(ip, jp)
struct ip *ip, *jp;
{
register u_char *cp, *dp;
int opt, optlen, cnt;
cp = (u_char *)(ip + 1);
dp = (u_char *)(jp + 1);
cnt = (IP_VHL_HL(ip->ip_vhl) << 2) - sizeof (struct ip);
1994-05-24 10:09:53 +00:00
for (; cnt > 0; cnt -= optlen, cp += optlen) {
opt = cp[0];
if (opt == IPOPT_EOL)
break;
if (opt == IPOPT_NOP) {
/* Preserve for IP mcast tunnel's LSRR alignment. */
*dp++ = IPOPT_NOP;
optlen = 1;
continue;
} else
optlen = cp[IPOPT_OLEN];
/* bogus lengths should have been caught by ip_dooptions */
if (optlen > cnt)
optlen = cnt;
if (IPOPT_COPIED(opt)) {
bcopy(cp, dp, optlen);
1994-05-24 10:09:53 +00:00
dp += optlen;
}
}
for (optlen = dp - (u_char *)(jp+1); optlen & 0x3; optlen++)
*dp++ = IPOPT_EOL;
return (optlen);
}
/*
* IP socket option processing.
*/
int
ip_ctloutput(so, sopt)
1994-05-24 10:09:53 +00:00
struct socket *so;
struct sockopt *sopt;
1994-05-24 10:09:53 +00:00
{
struct inpcb *inp = sotoinpcb(so);
int error, optval;
1994-05-24 10:09:53 +00:00
error = optval = 0;
if (sopt->sopt_level != IPPROTO_IP) {
return (EINVAL);
}
1994-05-24 10:09:53 +00:00
switch (sopt->sopt_dir) {
case SOPT_SET:
switch (sopt->sopt_name) {
1994-05-24 10:09:53 +00:00
case IP_OPTIONS:
#ifdef notyet
case IP_RETOPTS:
#endif
{
struct mbuf *m;
if (sopt->sopt_valsize > MLEN) {
error = EMSGSIZE;
break;
}
MGET(m, sopt->sopt_p ? M_WAIT : M_DONTWAIT, MT_HEADER);
if (m == 0) {
error = ENOBUFS;
break;
}
m->m_len = sopt->sopt_valsize;
error = sooptcopyin(sopt, mtod(m, char *), m->m_len,
m->m_len);
return (ip_pcbopts(sopt->sopt_name, &inp->inp_options,
m));
}
1994-05-24 10:09:53 +00:00
case IP_TOS:
case IP_TTL:
case IP_RECVOPTS:
case IP_RECVRETOPTS:
case IP_RECVDSTADDR:
case IP_RECVIF:
error = sooptcopyin(sopt, &optval, sizeof optval,
sizeof optval);
if (error)
break;
1994-05-24 10:09:53 +00:00
switch (sopt->sopt_name) {
case IP_TOS:
inp->inp_ip_tos = optval;
break;
1994-05-24 10:09:53 +00:00
case IP_TTL:
inp->inp_ip_ttl = optval;
break;
1994-05-24 10:09:53 +00:00
#define OPTSET(bit) \
if (optval) \
inp->inp_flags |= bit; \
else \
inp->inp_flags &= ~bit;
case IP_RECVOPTS:
OPTSET(INP_RECVOPTS);
break;
1994-05-24 10:09:53 +00:00
case IP_RECVRETOPTS:
OPTSET(INP_RECVRETOPTS);
break;
1994-05-24 10:09:53 +00:00
case IP_RECVDSTADDR:
OPTSET(INP_RECVDSTADDR);
break;
case IP_RECVIF:
OPTSET(INP_RECVIF);
break;
1994-05-24 10:09:53 +00:00
}
break;
#undef OPTSET
case IP_MULTICAST_IF:
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
case IP_MULTICAST_VIF:
1994-05-24 10:09:53 +00:00
case IP_MULTICAST_TTL:
case IP_MULTICAST_LOOP:
case IP_ADD_MEMBERSHIP:
case IP_DROP_MEMBERSHIP:
error = ip_setmoptions(sopt, &inp->inp_moptions);
1994-05-24 10:09:53 +00:00
break;
case IP_PORTRANGE:
error = sooptcopyin(sopt, &optval, sizeof optval,
sizeof optval);
if (error)
break;
switch (optval) {
case IP_PORTRANGE_DEFAULT:
inp->inp_flags &= ~(INP_LOWPORT);
inp->inp_flags &= ~(INP_HIGHPORT);
break;
case IP_PORTRANGE_HIGH:
inp->inp_flags &= ~(INP_LOWPORT);
inp->inp_flags |= INP_HIGHPORT;
break;
case IP_PORTRANGE_LOW:
inp->inp_flags &= ~(INP_HIGHPORT);
inp->inp_flags |= INP_LOWPORT;
break;
default:
error = EINVAL;
break;
}
break;
1994-05-24 10:09:53 +00:00
default:
error = ENOPROTOOPT;
break;
}
break;
case SOPT_GET:
switch (sopt->sopt_name) {
1994-05-24 10:09:53 +00:00
case IP_OPTIONS:
case IP_RETOPTS:
if (inp->inp_options)
error = sooptcopyout(sopt,
mtod(inp->inp_options,
char *),
inp->inp_options->m_len);
else
sopt->sopt_valsize = 0;
1994-05-24 10:09:53 +00:00
break;
case IP_TOS:
case IP_TTL:
case IP_RECVOPTS:
case IP_RECVRETOPTS:
case IP_RECVDSTADDR:
case IP_RECVIF:
case IP_PORTRANGE:
switch (sopt->sopt_name) {
1994-05-24 10:09:53 +00:00
case IP_TOS:
optval = inp->inp_ip_tos;
1994-05-24 10:09:53 +00:00
break;
case IP_TTL:
optval = inp->inp_ip_ttl;
1994-05-24 10:09:53 +00:00
break;
#define OPTBIT(bit) (inp->inp_flags & bit ? 1 : 0)
case IP_RECVOPTS:
optval = OPTBIT(INP_RECVOPTS);
break;
case IP_RECVRETOPTS:
optval = OPTBIT(INP_RECVRETOPTS);
break;
case IP_RECVDSTADDR:
optval = OPTBIT(INP_RECVDSTADDR);
break;
case IP_RECVIF:
optval = OPTBIT(INP_RECVIF);
break;
case IP_PORTRANGE:
if (inp->inp_flags & INP_HIGHPORT)
optval = IP_PORTRANGE_HIGH;
else if (inp->inp_flags & INP_LOWPORT)
optval = IP_PORTRANGE_LOW;
else
optval = 0;
break;
1994-05-24 10:09:53 +00:00
}
error = sooptcopyout(sopt, &optval, sizeof optval);
1994-05-24 10:09:53 +00:00
break;
case IP_MULTICAST_IF:
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
case IP_MULTICAST_VIF:
1994-05-24 10:09:53 +00:00
case IP_MULTICAST_TTL:
case IP_MULTICAST_LOOP:
case IP_ADD_MEMBERSHIP:
case IP_DROP_MEMBERSHIP:
error = ip_getmoptions(sopt, inp->inp_moptions);
break;
1994-05-24 10:09:53 +00:00
default:
error = ENOPROTOOPT;
break;
}
break;
}
return (error);
}
/*
* Set up IP options in pcb for insertion in output packets.
* Store in mbuf with pointer in pcbopt, adding pseudo-option
* with destination address if source routed.
*/
static int
1994-05-24 10:09:53 +00:00
ip_pcbopts(optname, pcbopt, m)
int optname;
struct mbuf **pcbopt;
register struct mbuf *m;
{
1998-02-20 13:37:40 +00:00
register int cnt, optlen;
1994-05-24 10:09:53 +00:00
register u_char *cp;
u_char opt;
/* turn off any old options */
if (*pcbopt)
(void)m_free(*pcbopt);
*pcbopt = 0;
if (m == (struct mbuf *)0 || m->m_len == 0) {
/*
* Only turning off any previous options.
*/
if (m)
(void)m_free(m);
return (0);
}
#ifndef vax
if (m->m_len % sizeof(int32_t))
1994-05-24 10:09:53 +00:00
goto bad;
#endif
/*
* IP first-hop destination address will be stored before
* actual options; move other options back
* and clear it when none present.
*/
if (m->m_data + m->m_len + sizeof(struct in_addr) >= &m->m_dat[MLEN])
goto bad;
cnt = m->m_len;
m->m_len += sizeof(struct in_addr);
cp = mtod(m, u_char *) + sizeof(struct in_addr);
ovbcopy(mtod(m, caddr_t), (caddr_t)cp, (unsigned)cnt);
bzero(mtod(m, caddr_t), sizeof(struct in_addr));
for (; cnt > 0; cnt -= optlen, cp += optlen) {
opt = cp[IPOPT_OPTVAL];
if (opt == IPOPT_EOL)
break;
if (opt == IPOPT_NOP)
optlen = 1;
else {
optlen = cp[IPOPT_OLEN];
if (optlen <= IPOPT_OLEN || optlen > cnt)
goto bad;
}
switch (opt) {
default:
break;
case IPOPT_LSRR:
case IPOPT_SSRR:
/*
* user process specifies route as:
* ->A->B->C->D
* D must be our final destination (but we can't
* check that since we may not have connected yet).
* A is first hop destination, which doesn't appear in
* actual IP option, but is stored before the options.
*/
if (optlen < IPOPT_MINOFF - 1 + sizeof(struct in_addr))
goto bad;
m->m_len -= sizeof(struct in_addr);
cnt -= sizeof(struct in_addr);
optlen -= sizeof(struct in_addr);
cp[IPOPT_OLEN] = optlen;
/*
* Move first hop before start of options.
*/
bcopy((caddr_t)&cp[IPOPT_OFFSET+1], mtod(m, caddr_t),
sizeof(struct in_addr));
/*
* Then copy rest of options back
* to close up the deleted entry.
*/
ovbcopy((caddr_t)(&cp[IPOPT_OFFSET+1] +
sizeof(struct in_addr)),
(caddr_t)&cp[IPOPT_OFFSET+1],
(unsigned)cnt + sizeof(struct in_addr));
break;
}
}
if (m->m_len > MAX_IPOPTLEN + sizeof(struct in_addr))
goto bad;
*pcbopt = m;
return (0);
bad:
(void)m_free(m);
return (EINVAL);
}
/*
* XXX
* The whole multicast option thing needs to be re-thought.
* Several of these options are equally applicable to non-multicast
* transmission, and one (IP_MULTICAST_TTL) totally duplicates a
* standard option (IP_TTL).
*/
1994-05-24 10:09:53 +00:00
/*
* Set the IP multicast options in response to user setsockopt().
*/
static int
ip_setmoptions(sopt, imop)
struct sockopt *sopt;
1994-05-24 10:09:53 +00:00
struct ip_moptions **imop;
{
int error = 0;
int i;
1994-05-24 10:09:53 +00:00
struct in_addr addr;
struct ip_mreq mreq;
struct ifnet *ifp;
struct ip_moptions *imo = *imop;
1994-05-24 10:09:53 +00:00
struct route ro;
struct sockaddr_in *dst;
int s;
1994-05-24 10:09:53 +00:00
if (imo == NULL) {
/*
* No multicast option buffer attached to the pcb;
* allocate one and initialize to default values.
*/
imo = (struct ip_moptions*)malloc(sizeof(*imo), M_IPMOPTS,
M_WAITOK);
if (imo == NULL)
return (ENOBUFS);
*imop = imo;
imo->imo_multicast_ifp = NULL;
imo->imo_multicast_vif = -1;
1994-05-24 10:09:53 +00:00
imo->imo_multicast_ttl = IP_DEFAULT_MULTICAST_TTL;
imo->imo_multicast_loop = IP_DEFAULT_MULTICAST_LOOP;
imo->imo_num_memberships = 0;
}
switch (sopt->sopt_name) {
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/* store an index number for the vif you wanna use in the send */
case IP_MULTICAST_VIF:
if (legal_vif_num == 0) {
error = EOPNOTSUPP;
break;
}
error = sooptcopyin(sopt, &i, sizeof i, sizeof i);
if (error)
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
break;
if (!legal_vif_num(i) && (i != -1)) {
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
error = EINVAL;
break;
}
imo->imo_multicast_vif = i;
break;
1994-05-24 10:09:53 +00:00
case IP_MULTICAST_IF:
/*
* Select the interface for outgoing multicast packets.
*/
error = sooptcopyin(sopt, &addr, sizeof addr, sizeof addr);
if (error)
1994-05-24 10:09:53 +00:00
break;
/*
* INADDR_ANY is used to remove a previous selection.
* When no interface is selected, a default one is
* chosen every time a multicast packet is sent.
*/
if (addr.s_addr == INADDR_ANY) {
imo->imo_multicast_ifp = NULL;
break;
}
/*
* The selected interface is identified by its local
* IP address. Find the interface and confirm that
* it supports multicasting.
*/
s = splimp();
1994-05-24 10:09:53 +00:00
INADDR_TO_IFP(addr, ifp);
if (ifp == NULL || (ifp->if_flags & IFF_MULTICAST) == 0) {
splx(s);
1994-05-24 10:09:53 +00:00
error = EADDRNOTAVAIL;
break;
}
imo->imo_multicast_ifp = ifp;
splx(s);
1994-05-24 10:09:53 +00:00
break;
case IP_MULTICAST_TTL:
/*
* Set the IP time-to-live for outgoing multicast packets.
* The original multicast API required a char argument,
* which is inconsistent with the rest of the socket API.
* We allow either a char or an int.
1994-05-24 10:09:53 +00:00
*/
if (sopt->sopt_valsize == 1) {
u_char ttl;
error = sooptcopyin(sopt, &ttl, 1, 1);
if (error)
break;
imo->imo_multicast_ttl = ttl;
} else {
u_int ttl;
error = sooptcopyin(sopt, &ttl, sizeof ttl,
sizeof ttl);
if (error)
break;
if (ttl > 255)
error = EINVAL;
else
imo->imo_multicast_ttl = ttl;
1994-05-24 10:09:53 +00:00
}
break;
case IP_MULTICAST_LOOP:
/*
* Set the loopback flag for outgoing multicast packets.
* Must be zero or one. The original multicast API required a
* char argument, which is inconsistent with the rest
* of the socket API. We allow either a char or an int.
1994-05-24 10:09:53 +00:00
*/
if (sopt->sopt_valsize == 1) {
u_char loop;
error = sooptcopyin(sopt, &loop, 1, 1);
if (error)
break;
imo->imo_multicast_loop = !!loop;
} else {
u_int loop;
error = sooptcopyin(sopt, &loop, sizeof loop,
sizeof loop);
if (error)
break;
imo->imo_multicast_loop = !!loop;
1994-05-24 10:09:53 +00:00
}
break;
case IP_ADD_MEMBERSHIP:
/*
* Add a multicast group membership.
* Group must be a valid IP multicast address.
*/
error = sooptcopyin(sopt, &mreq, sizeof mreq, sizeof mreq);
if (error)
1994-05-24 10:09:53 +00:00
break;
if (!IN_MULTICAST(ntohl(mreq.imr_multiaddr.s_addr))) {
1994-05-24 10:09:53 +00:00
error = EINVAL;
break;
}
s = splimp();
1994-05-24 10:09:53 +00:00
/*
* If no interface address was provided, use the interface of
* the route to the given multicast address.
*/
if (mreq.imr_interface.s_addr == INADDR_ANY) {
bzero((caddr_t)&ro, sizeof(ro));
1994-05-24 10:09:53 +00:00
dst = (struct sockaddr_in *)&ro.ro_dst;
dst->sin_len = sizeof(*dst);
dst->sin_family = AF_INET;
dst->sin_addr = mreq.imr_multiaddr;
1994-05-24 10:09:53 +00:00
rtalloc(&ro);
if (ro.ro_rt == NULL) {
error = EADDRNOTAVAIL;
splx(s);
1994-05-24 10:09:53 +00:00
break;
}
ifp = ro.ro_rt->rt_ifp;
rtfree(ro.ro_rt);
}
else {
INADDR_TO_IFP(mreq.imr_interface, ifp);
1994-05-24 10:09:53 +00:00
}
1994-05-24 10:09:53 +00:00
/*
* See if we found an interface, and confirm that it
* supports multicast.
*/
if (ifp == NULL || (ifp->if_flags & IFF_MULTICAST) == 0) {
error = EADDRNOTAVAIL;
splx(s);
1994-05-24 10:09:53 +00:00
break;
}
/*
* See if the membership already exists or if all the
* membership slots are full.
*/
for (i = 0; i < imo->imo_num_memberships; ++i) {
if (imo->imo_membership[i]->inm_ifp == ifp &&
imo->imo_membership[i]->inm_addr.s_addr
== mreq.imr_multiaddr.s_addr)
1994-05-24 10:09:53 +00:00
break;
}
if (i < imo->imo_num_memberships) {
error = EADDRINUSE;
splx(s);
1994-05-24 10:09:53 +00:00
break;
}
if (i == IP_MAX_MEMBERSHIPS) {
error = ETOOMANYREFS;
splx(s);
1994-05-24 10:09:53 +00:00
break;
}
/*
* Everything looks good; add a new record to the multicast
* address list for the given interface.
*/
if ((imo->imo_membership[i] =
in_addmulti(&mreq.imr_multiaddr, ifp)) == NULL) {
1994-05-24 10:09:53 +00:00
error = ENOBUFS;
splx(s);
1994-05-24 10:09:53 +00:00
break;
}
++imo->imo_num_memberships;
splx(s);
1994-05-24 10:09:53 +00:00
break;
case IP_DROP_MEMBERSHIP:
/*
* Drop a multicast group membership.
* Group must be a valid IP multicast address.
*/
error = sooptcopyin(sopt, &mreq, sizeof mreq, sizeof mreq);
if (error)
1994-05-24 10:09:53 +00:00
break;
if (!IN_MULTICAST(ntohl(mreq.imr_multiaddr.s_addr))) {
1994-05-24 10:09:53 +00:00
error = EINVAL;
break;
}
s = splimp();
1994-05-24 10:09:53 +00:00
/*
* If an interface address was specified, get a pointer
* to its ifnet structure.
*/
if (mreq.imr_interface.s_addr == INADDR_ANY)
1994-05-24 10:09:53 +00:00
ifp = NULL;
else {
INADDR_TO_IFP(mreq.imr_interface, ifp);
1994-05-24 10:09:53 +00:00
if (ifp == NULL) {
error = EADDRNOTAVAIL;
splx(s);
1994-05-24 10:09:53 +00:00
break;
}
}
/*
* Find the membership in the membership array.
*/
for (i = 0; i < imo->imo_num_memberships; ++i) {
if ((ifp == NULL ||
imo->imo_membership[i]->inm_ifp == ifp) &&
imo->imo_membership[i]->inm_addr.s_addr ==
mreq.imr_multiaddr.s_addr)
1994-05-24 10:09:53 +00:00
break;
}
if (i == imo->imo_num_memberships) {
error = EADDRNOTAVAIL;
splx(s);
1994-05-24 10:09:53 +00:00
break;
}
/*
* Give up the multicast address record to which the
* membership points.
*/
in_delmulti(imo->imo_membership[i]);
/*
* Remove the gap in the membership array.
*/
for (++i; i < imo->imo_num_memberships; ++i)
imo->imo_membership[i-1] = imo->imo_membership[i];
--imo->imo_num_memberships;
splx(s);
1994-05-24 10:09:53 +00:00
break;
default:
error = EOPNOTSUPP;
break;
}
/*
* If all options have default values, no need to keep the mbuf.
*/
if (imo->imo_multicast_ifp == NULL &&
imo->imo_multicast_vif == -1 &&
1994-05-24 10:09:53 +00:00
imo->imo_multicast_ttl == IP_DEFAULT_MULTICAST_TTL &&
imo->imo_multicast_loop == IP_DEFAULT_MULTICAST_LOOP &&
imo->imo_num_memberships == 0) {
free(*imop, M_IPMOPTS);
*imop = NULL;
}
return (error);
}
/*
* Return the IP multicast options in response to user getsockopt().
*/
static int
ip_getmoptions(sopt, imo)
struct sockopt *sopt;
1994-05-24 10:09:53 +00:00
register struct ip_moptions *imo;
{
struct in_addr addr;
1994-05-24 10:09:53 +00:00
struct in_ifaddr *ia;
int error, optval;
u_char coptval;
1994-05-24 10:09:53 +00:00
error = 0;
switch (sopt->sopt_name) {
case IP_MULTICAST_VIF:
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
if (imo != NULL)
optval = imo->imo_multicast_vif;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
else
optval = -1;
error = sooptcopyout(sopt, &optval, sizeof optval);
break;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
1994-05-24 10:09:53 +00:00
case IP_MULTICAST_IF:
if (imo == NULL || imo->imo_multicast_ifp == NULL)
addr.s_addr = INADDR_ANY;
1994-05-24 10:09:53 +00:00
else {
IFP_TO_IA(imo->imo_multicast_ifp, ia);
addr.s_addr = (ia == NULL) ? INADDR_ANY
: IA_SIN(ia)->sin_addr.s_addr;
1994-05-24 10:09:53 +00:00
}
error = sooptcopyout(sopt, &addr, sizeof addr);
break;
1994-05-24 10:09:53 +00:00
case IP_MULTICAST_TTL:
if (imo == 0)
optval = coptval = IP_DEFAULT_MULTICAST_TTL;
else
optval = coptval = imo->imo_multicast_ttl;
if (sopt->sopt_valsize == 1)
error = sooptcopyout(sopt, &coptval, 1);
else
error = sooptcopyout(sopt, &optval, sizeof optval);
break;
1994-05-24 10:09:53 +00:00
case IP_MULTICAST_LOOP:
if (imo == 0)
optval = coptval = IP_DEFAULT_MULTICAST_LOOP;
else
optval = coptval = imo->imo_multicast_loop;
if (sopt->sopt_valsize == 1)
error = sooptcopyout(sopt, &coptval, 1);
else
error = sooptcopyout(sopt, &optval, sizeof optval);
break;
1994-05-24 10:09:53 +00:00
default:
error = ENOPROTOOPT;
break;
1994-05-24 10:09:53 +00:00
}
return (error);
1994-05-24 10:09:53 +00:00
}
/*
* Discard the IP multicast options.
*/
void
ip_freemoptions(imo)
register struct ip_moptions *imo;
{
register int i;
if (imo != NULL) {
for (i = 0; i < imo->imo_num_memberships; ++i)
in_delmulti(imo->imo_membership[i]);
free(imo, M_IPMOPTS);
}
}
/*
* Routine called from ip_output() to loop back a copy of an IP multicast
* packet to the input queue of a specified interface. Note that this
* calls the output routine of the loopback "driver", but with an interface
* pointer that might NOT be a loopback interface -- evil, but easier than
* replicating that code here.
1994-05-24 10:09:53 +00:00
*/
static void
ip_mloopback(ifp, m, dst, hlen)
1994-05-24 10:09:53 +00:00
struct ifnet *ifp;
register struct mbuf *m;
register struct sockaddr_in *dst;
int hlen;
1994-05-24 10:09:53 +00:00
{
register struct ip *ip;
struct mbuf *copym;
copym = m_copy(m, 0, M_COPYALL);
if (copym != NULL && (copym->m_flags & M_EXT || copym->m_len < hlen))
copym = m_pullup(copym, hlen);
1994-05-24 10:09:53 +00:00
if (copym != NULL) {
/*
* We don't bother to fragment if the IP length is greater
* than the interface's MTU. Can this possibly matter?
*/
ip = mtod(copym, struct ip *);
ip->ip_len = htons((u_short)ip->ip_len);
ip->ip_off = htons((u_short)ip->ip_off);
ip->ip_sum = 0;
if (ip->ip_vhl == IP_VHL_BORING) {
ip->ip_sum = in_cksum_hdr(ip);
} else {
ip->ip_sum = in_cksum(copym, hlen);
}
/*
* NB:
* It's not clear whether there are any lingering
* reentrancy problems in other areas which might
* be exposed by using ip_input directly (in
* particular, everything which modifies the packet
* in-place). Yet another option is using the
* protosw directly to deliver the looped back
* packet. For the moment, we'll err on the side
* of safety by using if_simloop().
*/
#if 1 /* XXX */
if (dst->sin_family != AF_INET) {
1998-06-15 00:35:47 +00:00
printf("ip_mloopback: bad address family %d\n",
dst->sin_family);
dst->sin_family = AF_INET;
}
#endif
#ifdef notdef
copym->m_pkthdr.rcvif = ifp;
ip_input(copym);
#else
if_simloop(ifp, copym, (struct sockaddr *)dst, 0);
#endif
1994-05-24 10:09:53 +00:00
}
}