ip_icmp.c

来自「eCos操作系统源码」· C语言 代码 · 共 900 行 · 第 1/2 页

		ctlfunc = inetsw[ip_protox[icp->icmp_ip.ip_p]].pr_ctlinput;
		if (ctlfunc)
			(*ctlfunc)(code, (struct sockaddr *)&icmpsrc,
				   (void *)&icp->icmp_ip);
		break;

	badcode:
		icmpstat.icps_badcode++;
		break;

	case ICMP_ECHO:
		if (!icmpbmcastecho
		    && (m->m_flags & (M_MCAST | M_BCAST)) != 0) {
			icmpstat.icps_bmcastecho++;
			break;
		}
		icp->icmp_type = ICMP_ECHOREPLY;
#ifdef ICMP_BANDLIM
		if (badport_bandlim(BANDLIM_ICMP_ECHO) < 0)
			goto freeit;
		else
#endif
			goto reflect;

	case ICMP_TSTAMP:
		if (!icmpbmcastecho
		    && (m->m_flags & (M_MCAST | M_BCAST)) != 0) {
			icmpstat.icps_bmcasttstamp++;
			break;
		}
		if (icmplen < ICMP_TSLEN) {
			icmpstat.icps_badlen++;
			break;
		}
		icp->icmp_type = ICMP_TSTAMPREPLY;
		icp->icmp_rtime = iptime();
		icp->icmp_ttime = icp->icmp_rtime;	/* bogus, do later! */
#ifdef ICMP_BANDLIM
		if (badport_bandlim(BANDLIM_ICMP_TSTAMP) < 0)
			goto freeit;
		else
#endif
			goto reflect;

	case ICMP_MASKREQ:
#define	satosin(sa)	((struct sockaddr_in *)(sa))
		if (icmpmaskrepl == 0)
			break;
		/*
		 * We are not able to respond with all ones broadcast
		 * unless we receive it over a point-to-point interface.
		 */
		if (icmplen < ICMP_MASKLEN)
			break;
		switch (ip->ip_dst.s_addr) {

		case INADDR_BROADCAST:
		case INADDR_ANY:
			icmpdst.sin_addr = ip->ip_src;
			break;

		default:
			icmpdst.sin_addr = ip->ip_dst;
		}
		ia = (struct in_ifaddr *)ifaof_ifpforaddr(
			    (struct sockaddr *)&icmpdst, m->m_pkthdr.rcvif);
		if (ia == 0)
			break;
		if (ia->ia_ifp == 0)
			break;
		icp->icmp_type = ICMP_MASKREPLY;
		icp->icmp_mask = ia->ia_sockmask.sin_addr.s_addr;
		if (ip->ip_src.s_addr == 0) {
			if (ia->ia_ifp->if_flags & IFF_BROADCAST)
			    ip->ip_src = satosin(&ia->ia_broadaddr)->sin_addr;
			else if (ia->ia_ifp->if_flags & IFF_POINTOPOINT)
			    ip->ip_src = satosin(&ia->ia_dstaddr)->sin_addr;
		}
reflect:
		ip->ip_len += hlen;	/* since ip_input deducts this */
		icmpstat.icps_reflect++;
		icmpstat.icps_outhist[icp->icmp_type]++;
		icmp_reflect(m);
		return;

	case ICMP_REDIRECT:
		if (log_redirect) {
			u_long src, dst, gw;

			src = ntohl(ip->ip_src.s_addr);
			dst = ntohl(icp->icmp_ip.ip_dst.s_addr);
			gw = ntohl(icp->icmp_gwaddr.s_addr);
			printf("icmp redirect from %d.%d.%d.%d: "
			       "%d.%d.%d.%d => %d.%d.%d.%d\n",
			       (int)(src >> 24), (int)((src >> 16) & 0xff),
			       (int)((src >> 8) & 0xff), (int)(src & 0xff),
			       (int)(dst >> 24), (int)((dst >> 16) & 0xff),
			       (int)((dst >> 8) & 0xff), (int)(dst & 0xff),
			       (int)(gw >> 24), (int)((gw >> 16) & 0xff),
			       (int)((gw >> 8) & 0xff), (int)(gw & 0xff));
		}
		if (drop_redirect)
			break;
		if (code > 3)
			goto badcode;
		if (icmplen < ICMP_ADVLENMIN || icmplen < ICMP_ADVLEN(icp) ||
		    IP_VHL_HL(icp->icmp_ip.ip_vhl) < (sizeof(struct ip) >> 2)) {
			icmpstat.icps_badlen++;
			break;
		}
		/*
		 * Short circuit routing redirects to force
		 * immediate change in the kernel's routing
		 * tables.  The message is also handed to anyone
		 * listening on a raw socket (e.g. the routing
		 * daemon for use in updating its tables).
		 */
		icmpgw.sin_addr = ip->ip_src;
		icmpdst.sin_addr = icp->icmp_gwaddr;
#ifdef	ICMPPRINTFS
		if (icmpprintfs) {
			char buf[4 * sizeof "123"];
			strcpy(buf, inet_ntoa(icp->icmp_ip.ip_dst));

			printf("redirect dst %s to %s\n",
			       buf, inet_ntoa(icp->icmp_gwaddr));
		}
#endif
		icmpsrc.sin_addr = icp->icmp_ip.ip_dst;
		rtredirect((struct sockaddr *)&icmpsrc,
		  (struct sockaddr *)&icmpdst,
		  (struct sockaddr *)0, RTF_GATEWAY | RTF_HOST,
		  (struct sockaddr *)&icmpgw, (struct rtentry **)0);
		pfctlinput(PRC_REDIRECT_HOST, (struct sockaddr *)&icmpsrc);
#ifdef IPSEC
		key_sa_routechange((struct sockaddr *)&icmpsrc);
#endif
		break;

	/*
	 * No kernel processing for the following;
	 * just fall through to send to raw listener.
	 */
	case ICMP_ECHOREPLY:
	case ICMP_ROUTERADVERT:
	case ICMP_ROUTERSOLICIT:
	case ICMP_TSTAMPREPLY:
	case ICMP_IREQREPLY:
	case ICMP_MASKREPLY:
	default:
		break;
	}

raw:
	rip_input(m, off);
	return;

freeit:
	m_freem(m);
}

/*
 * Reflect the ip packet back to the source
 */
static void
icmp_reflect(m)
	struct mbuf *m;
{
	register struct ip *ip = mtod(m, struct ip *);
	register struct in_ifaddr *ia;
	struct in_addr t;
	struct mbuf *opts = 0;
	int optlen = (IP_VHL_HL(ip->ip_vhl) << 2) - sizeof(struct ip);

	if (!in_canforward(ip->ip_src) &&
	    ((ntohl(ip->ip_src.s_addr) & IN_CLASSA_NET) !=
	     (IN_LOOPBACKNET << IN_CLASSA_NSHIFT))) {
		m_freem(m);	/* Bad return address */
		goto done;	/* Ip_output() will check for broadcast */
	}
	t = ip->ip_dst;
	ip->ip_dst = ip->ip_src;
	/*
	 * If the incoming packet was addressed directly to us,
	 * use dst as the src for the reply.  Otherwise (broadcast
	 * or anonymous), use the address which corresponds
	 * to the incoming interface.
	 */
	for (ia = in_ifaddrhead.tqh_first; ia; ia = ia->ia_link.tqe_next) {
		if (t.s_addr == IA_SIN(ia)->sin_addr.s_addr)
			break;
		if (ia->ia_ifp && (ia->ia_ifp->if_flags & IFF_BROADCAST) &&
		    t.s_addr == satosin(&ia->ia_broadaddr)->sin_addr.s_addr)
			break;
	}
	icmpdst.sin_addr = t;
	if ((ia == (struct in_ifaddr *)0) && m->m_pkthdr.rcvif)
		ia = (struct in_ifaddr *)ifaof_ifpforaddr(
			(struct sockaddr *)&icmpdst, m->m_pkthdr.rcvif);
	/*
	 * The following happens if the packet was not addressed to us,
	 * and was received on an interface with no IP address.
	 */
	if (ia == (struct in_ifaddr *)0)
		ia = in_ifaddrhead.tqh_first;
	t = IA_SIN(ia)->sin_addr;
	ip->ip_src = t;
	ip->ip_ttl = ip_defttl;

	if (optlen > 0) {
		register u_char *cp;
		int opt, cnt;
		u_int len;

		/*
		 * Retrieve any source routing from the incoming packet;
		 * add on any record-route or timestamp options.
		 */
		cp = (u_char *) (ip + 1);
		if ((opts = ip_srcroute()) == 0 &&
		    (opts = m_gethdr(M_DONTWAIT, MT_HEADER))) {
			opts->m_len = sizeof(struct in_addr);
			mtod(opts, struct in_addr *)->s_addr = 0;
		}
		if (opts) {
#ifdef ICMPPRINTFS
		    if (icmpprintfs)
			    printf("icmp_reflect optlen %d rt %d => ",
				optlen, opts->m_len);
#endif
		    for (cnt = optlen; cnt > 0; cnt -= len, cp += len) {
			    opt = cp[IPOPT_OPTVAL];
			    if (opt == IPOPT_EOL)
				    break;
			    if (opt == IPOPT_NOP)
				    len = 1;
			    else {
				    if (cnt < IPOPT_OLEN + sizeof(*cp))
					    break;
				    len = cp[IPOPT_OLEN];
				    if (len < IPOPT_OLEN + sizeof(*cp) ||
				        len > cnt)
					    break;
			    }
			    /*
			     * Should check for overflow, but it "can't happen"
			     */
			    if (opt == IPOPT_RR || opt == IPOPT_TS ||
				opt == IPOPT_SECURITY) {
				    bcopy((caddr_t)cp,
					mtod(opts, caddr_t) + opts->m_len, len);
				    opts->m_len += len;
			    }
		    }
		    /* Terminate & pad, if necessary */
		    cnt = opts->m_len % 4;
		    if (cnt) {
			    for (; cnt < 4; cnt++) {
				    *(mtod(opts, caddr_t) + opts->m_len) =
					IPOPT_EOL;
				    opts->m_len++;
			    }
		    }
#ifdef ICMPPRINTFS
		    if (icmpprintfs)
			    printf("%d\n", opts->m_len);
#endif
		}
		/*
		 * Now strip out original options by copying rest of first
		 * mbuf's data back, and adjust the IP length.
		 */
		ip->ip_len -= optlen;
		ip->ip_vhl = IP_VHL_BORING;
		m->m_len -= optlen;
		if (m->m_flags & M_PKTHDR)
			m->m_pkthdr.len -= optlen;
		optlen += sizeof(struct ip);
		bcopy((caddr_t)ip + optlen, (caddr_t)(ip + 1),
			 (unsigned)(m->m_len - sizeof(struct ip)));
	}
	m->m_flags &= ~(M_BCAST|M_MCAST);
	icmp_send(m, opts);
done:
	if (opts)
		(void)m_free(opts);
}

/*
 * Send an icmp packet back to the ip level,
 * after supplying a checksum.
 */
static void
icmp_send(m, opts)
	register struct mbuf *m;
	struct mbuf *opts;
{
	register struct ip *ip = mtod(m, struct ip *);
	register int hlen;
	register struct icmp *icp;
	struct route ro;

	hlen = IP_VHL_HL(ip->ip_vhl) << 2;
	m->m_data += hlen;
	m->m_len -= hlen;
	icp = mtod(m, struct icmp *);
	icp->icmp_cksum = 0;
	icp->icmp_cksum = in_cksum(m, ip->ip_len - hlen);
	m->m_data -= hlen;
	m->m_len += hlen;
	m->m_pkthdr.rcvif = (struct ifnet *)0;
#ifdef ICMPPRINTFS
	if (icmpprintfs) {
		char buf[4 * sizeof "123"];
		strcpy(buf, inet_ntoa(ip->ip_dst));
		printf("icmp_send dst %s src %s\n",
		       buf, inet_ntoa(ip->ip_src));
	}
#endif
	bzero(&ro, sizeof ro);
	(void) ip_output(m, opts, &ro, 0, NULL);
	if (ro.ro_rt)
		RTFREE(ro.ro_rt);
}

n_time
iptime()
{
	struct timeval atv;
	u_long t;

	microtime(&atv);
	t = (atv.tv_sec % (24*60*60)) * 1000 + atv.tv_usec / 1000;
	return (htonl(t));
}

#if 1
/*
 * Return the next larger or smaller MTU plateau (table from RFC 1191)
 * given current value MTU.  If DIR is less than zero, a larger plateau
 * is returned; otherwise, a smaller value is returned.
 */
static int
ip_next_mtu(mtu, dir)
	int mtu;
	int dir;
{
	static int mtutab[] = {
		65535, 32000, 17914, 8166, 4352, 2002, 1492, 1006, 508, 296,
		68, 0
	};
	int i;

	for (i = 0; i < (sizeof mtutab) / (sizeof mtutab[0]); i++) {
		if (mtu >= mtutab[i])
			break;
	}

	if (dir < 0) {
		if (i == 0) {
			return 0;
		} else {
			return mtutab[i - 1];
		}
	} else {
		if (mtutab[i] == 0) {
			return 0;
		} else if(mtu > mtutab[i]) {
			return mtutab[i];
		} else {
			return mtutab[i + 1];
		}
	}
}
#endif

#ifdef ICMP_BANDLIM

/*
 * badport_bandlim() - check for ICMP bandwidth limit
 *
 *	Return 0 if it is ok to send an ICMP error response, -1 if we have
 *	hit our bandwidth limit and it is not ok.  
 *
 *	If icmplim is <= 0, the feature is disabled and 0 is returned.
 *
 *	For now we separate the TCP and UDP subsystems w/ different 'which'
 *	values.  We may eventually remove this separation (and simplify the
 *	code further).
 *
 *	Note that the printing of the error message is delayed so we can
 *	properly print the icmp error rate that the system was trying to do
 *	(i.e. 22000/100 pps, etc...).  This can cause long delays in printing
 *	the 'final' error, but it doesn't make sense to solve the printing 
 *	delay with more complex code.
 */

int
badport_bandlim(int which)
{
	static int lticks[BANDLIM_MAX + 1];
	static int lpackets[BANDLIM_MAX + 1];
	int dticks;
	const char *bandlimittype[] = {
		"Limiting icmp unreach response",
		"Limiting icmp ping response",
		"Limiting icmp tstamp response",
		"Limiting closed port RST response",
		"Limiting open port RST response"
		};

	/*
	 * Return ok status if feature disabled or argument out of
	 * ranage.
	 */

	if (icmplim <= 0 || which > BANDLIM_MAX || which < 0)
		return(0);
	dticks = ticks - lticks[which];

	/*
	 * reset stats when cumulative dt exceeds one second.
	 */

	if ((unsigned int)dticks > hz) {
		if (lpackets[which] > icmplim) {
			printf("%s from %d to %d packets per second\n",
				bandlimittype[which],
				lpackets[which],
				icmplim
			);
		}
		lticks[which] = ticks;
		lpackets[which] = 0;
	}

	/*
	 * bump packet count
	 */

	if (++lpackets[which] > icmplim) {
		return(-1);
	}
	return(0);
}

#endif