ip_input.c

来自「操作系统SunOS 4.1.3版本的源码」· C语言 代码 · 共 811 行 · 第 1/2 页

#ifndef lint
static char sccsid[] = "@(#)ip_input.c 1.1 92/07/30 Copyr 1986 Sun Micro";
#endif

/*
 * Copyright (c) 1986 by Sun Microsystems, Inc.
 */


#include <sys/param.h>
#include "boot/systm.h"
#include <sys/mbuf.h>
#include "boot/domain.h"
#include "boot/protosw.h"
#include <sys/socket.h>
#include <sys/errno.h>
#include <sys/time.h>
#include <sys/kernel.h>

#include <net/if.h>
#include <net/route.h>

#include <netinet/in.h>
#include <netinet/in_pcb.h>
#include <netinet/in_systm.h>
#include <netinet/ip.h>
#include <netinet/ip_var.h>
#include <netinet/ip_icmp.h>
#include <netinet/tcp.h>
#include <mon/sunromvec.h>

u_char	ip_protox[IPPROTO_MAX];
int	ipqmaxlen = IFQ_MAXLEN;
struct	ifnet *ifinet;			/* first inet interface */

int	ip_gotit = 0;
int	ip_reasstot = 0;

static int dump_debug = 30;
#undef	IPFRAGTTL
#define	IPFRAGTTL	8
#define	MAXTRACE	128
struct	{
	int	ip_ident;
	int	ip_time;
	int	ip_diff;
	int	ip_offset;
} ip_trace[MAXTRACE];
int	ip_trace_index = MAXTRACE-1;
#define	NEXT_TRACE(i)	(i=((i)==(MAXTRACE-1)?0:i+1))

#ifdef OPENPROMS
#define	millitime()	prom_gettime()
#else
#define	millitime()	(*romp->v_nmiclock)
#endif !OPENPROMS

/*
 * IP initialization: fill in IP protocol switch table.
 * All protocols not implemented in kernel go to raw IP protocol handler.
 */
ip_init()
{
	register struct protosw *pr;
	register int i;
	struct ifnet *if_ifwithaf();


	pr = pffindproto(PF_INET, IPPROTO_RAW);
	if (pr == 0)
		panic("ip_init");
	for (i = 0; i < IPPROTO_MAX; i++)
		ip_protox[i] = pr - inetsw;
	for (pr = inetdomain.dom_protosw;
	    pr < inetdomain.dom_protoswNPROTOSW; pr++) {
		if (pr->pr_family == PF_INET &&
		    pr->pr_protocol && pr->pr_protocol != IPPROTO_RAW) {
			ip_protox[pr->pr_protocol] = pr - inetsw;
		}
	}
	ipq.next = ipq.prev = &ipq;
	ip_id = time.tv_sec & 0xffff;
	ifinet = if_ifwithaf(AF_INET);
}

u_char	i_ipcksum = 1;
struct	ip *ip_reass();
struct	sockaddr_in ipaddr = { AF_INET };

/*
 * Ip input routine.  Checksum and byte swap header.  If fragmented
 * try to reassamble.  If complete and fragment queue exists, discard.
 * Process options.  Pass to next level.
 */
ipintr(m)
	struct mbuf *m;
{
	register struct ip *ip;
	struct mbuf *m0;
	register int i;
	register struct ipq *fp;
	int hlen;
	register struct ifaddr *ifa;

#ifdef	DUMP_DEBUG1
	dprint(dump_debug, 6,
		"ipintr(m 0x%x) type 0x%x\n", m, m->m_type);
#endif	/* DUMP_DEBUG */

	ip_gotit = 0;

	/*
	 * Get next datagram off input queue and get IP header
	 * in first mbuf.
	 */
	if (m == 0)	{
		dprint(dump_debug, 0,
			"ipintr: NULL mbuf\n");
		return;
	}
	if ((m->m_off > MMAXOFF || m->m_len < sizeof (struct ip)) &&
	    (m = m_pullup(m, sizeof (struct ip))) == 0) {
		dprint(dump_debug, 0,
			"ipintr: mbuff too small\n");
		ipstat.ips_toosmall++;
		return;
	}
	ip = mtod(m, struct ip *);
	hlen = ip->ip_hl << 2;
	if (hlen < 10) {	/* minimum header length */
		dprint(dump_debug, 0,
			"ipintr: bad hl 0x%x\n", hlen);
		ipstat.ips_badhlen++;
		m_freem(m);
		return;
	}
	if (hlen > m->m_len) {
		if ((m = m_pullup(m, hlen)) == 0) {
			dprint(dump_debug, 0,
				"ipintr: bad pullup\n");
			ipstat.ips_badhlen++;
			return;
		}
		ip = mtod(m, struct ip *);
	}

	if (i_ipcksum)
		if (ip->ip_sum = ipcksum((caddr_t)ip, (unsigned short)hlen)) {
			dprint(dump_debug, 0,
				"ipintr: ip_sum 0x%x\n",
				ip->ip_sum);
			ipstat.ips_badsum++;
			goto bad;
		}

	/*
	 * Convert fields to host representation.
	 */
	ip->ip_len = ntohs((u_short)ip->ip_len);
#ifdef  DUMP_DEBUG1
	dprint(dump_debug, 6, "ipintr: ip_len 0x%x\n", ip->ip_len);
#endif
	if (ip->ip_len < hlen) {
		ipstat.ips_badlen++;
		goto bad;
	}
	ip->ip_id = ntohs(ip->ip_id);
	ip->ip_off = ntohs((u_short)ip->ip_off);

	/*
	 * Check that the amount of data in the buffers
	 * is as at least much as the IP header would have us expect.
	 * Trim mbufs if longer than we expect.
	 * Drop packet if shorter than we expect.
	 */
	i = -ip->ip_len;
	m0 = m;
	for (;;) {
		i += m->m_len;
		if (m->m_next == 0)
			break;
		m = m->m_next;
	}
	if (i != 0) {
		if (i < 0) {
			ipstat.ips_tooshort++;
			m = m0;
			goto bad;
		}
		if (i <= m->m_len)
			m->m_len -= i;
		else
			m_adj(m0, -i);
	}
	m = m0;

	/*
	 * Process options and, if not destined for us,
	 * ship it on.  ip_dooptions returns 1 when an
	 * error was detected (causing an icmp message
	 * to be sent).
	 */
	if (hlen > sizeof (struct ip) && ip_dooptions(ip))
		return;

	/*
	 * Fast check on the first internet
	 * interface in the list.
	 */
	if (ifinet) {
		struct sockaddr_in *sin;

		ifa = ifinet->if_addrlist;
		sin = (struct sockaddr_in *)&ifa->ifa_addr;
		if (sin->sin_addr.s_addr == ip->ip_dst.s_addr)
			goto ours;
		if ((ifinet->if_flags & IFF_BROADCAST) &&
		    ip->ip_dst.s_addr == INADDR_ANY)
			goto ours;
	}
	ipaddr.sin_addr = ip->ip_dst;
	if (if_ifwithaddr((struct sockaddr *)&ipaddr) == 0 &&
	    (*(int *)&ip->ip_dst) != -1) {
		ip_forward(ip);
		return;
	}

ours:

	/*
	 * Look for queue of fragments
	 * of this datagram.
	 */
	for (fp = ipq.next; fp != &ipq; fp = fp->next)
		if (ip->ip_id == fp->ipq_id &&
		    ip->ip_src.s_addr == fp->ipq_src.s_addr &&
		    ip->ip_dst.s_addr == fp->ipq_dst.s_addr &&
		    ip->ip_p == fp->ipq_p)  {
			goto found;
		    }

	fp = 0;
found:
	ip_trace[NEXT_TRACE(ip_trace_index)].ip_ident = ip->ip_id;
	ip_trace[ip_trace_index].ip_time = millitime();
	ip_trace[ip_trace_index].ip_diff = ip_trace[ip_trace_index].ip_time -
		ip_trace[ip_trace_index-1].ip_time;
	ip_trace[ip_trace_index].ip_offset = ip->ip_off;

	/*
	 * Adjust ip_len to not reflect header,
	 * set ip_mff if more fragments are expected,
	 * convert offset of this to bytes.
	 */
	ip->ip_len -= hlen;
	((struct ipasfrag *)ip)->ipf_mff = 0;
	if (ip->ip_off & IP_MF)
		((struct ipasfrag *)ip)->ipf_mff = 1;
	ip->ip_off <<= 3;


	/*
	 * If datagram marked as having more fragments
	 * or if this is not the first fragment,
	 * attempt reassembly; if it succeeds, proceed.
	 */
	if (((struct ipasfrag *)ip)->ipf_mff || ip->ip_off) {
		ip = ip_reass((struct ipasfrag *)ip, fp);
		if (ip == 0)	{
			return;
		}
		hlen = ip->ip_hl << 2;

		m = dtom(ip);
	} else
		if (fp)
			ip_freef(fp);

	ip_gotit = 1;

	/*
	 * Switch out to protocol's input routine.
	 */

	(*inetsw[ip_protox[ip->ip_p]].pr_input)(m);
	return;
bad:
	dprint(dump_debug, 0,
		"ipintr: bad\n");
	m_freem(m);
	return;
}


/*
 * Take incoming datagram fragment and try to
 * reassemble it into whole datagram.  If a chain for
 * reassembly of this datagram already exists, then it
 * is given as fp; otherwise have to make a chain.
 */
struct ip *
ip_reass(ip, fp)
	register struct ipasfrag *ip;
	register struct ipq *fp;
{
	register struct mbuf *m = dtom(ip);
	register struct ipasfrag *q;
	struct mbuf *t;
	int hlen = ip->ip_hl << 2;
	int i, next;
#ifdef	 DUMP_DEBUG1
	dprint(dump_debug, 6, "ip_reass(ip 0x%x fp 0x%x)\n", ip, fp);
#endif	 /* DUMP_DEBUG */

	/*
	 * Presence of header sizes in mbufs
	 * would confuse code below.
	 */
	m->m_off += hlen;
	m->m_len -= hlen;

	/*
	 * If first fragment to arrive, create a reassembly queue.
	 */
	if (fp == 0) {
		if ((t = m_get(M_WAIT, MT_FTABLE)) == NULL)
			goto dropfrag;
		fp = mtod(t, struct ipq *);
#ifdef   DUMP_DEBUG1
	dprint(dump_debug, 6, "ip_reass: t 0x%x fp 0x%x type 0x%x\n",
		t, fp, t->m_type);
#endif   /* DUMP_DEBUG */
		insque(fp, &ipq);
		fp->ipq_ttl = IPFRAGTTL;
		fp->ipq_p = ip->ip_p;
		fp->ipq_id = ip->ip_id;
		fp->ipq_next = fp->ipq_prev = (struct ipasfrag *)fp;
		fp->ipq_src = ((struct ip *)ip)->ip_src;
		fp->ipq_dst = ((struct ip *)ip)->ip_dst;
		q = (struct ipasfrag *)fp;
		goto insert;
	}

	/*
	 * Find a segment which begins after this one does.
	 */
	for (q = fp->ipq_next; q != (struct ipasfrag *)fp; q = q->ipf_next)
		if (q->ip_off > ip->ip_off)
			break;

	/*
	 * If there is a preceding segment, it may provide some of
	 * our data already.  If so, drop the data from the incoming
	 * segment.  If it provides all of our data, drop us.
	 */
	if (q->ipf_prev != (struct ipasfrag *)fp) {
		i = q->ipf_prev->ip_off + q->ipf_prev->ip_len - ip->ip_off;
		if (i > 0) {
			if (i >= ip->ip_len)
				goto dropfrag;
			m_adj(dtom(ip), i);
			ip->ip_off += i;
			ip->ip_len -= i;
		}
	}

	/*
	 * While we overlap succeeding segments trim them or,
	 * if they are completely covered, dequeue them.
	 */
	while (q != (struct ipasfrag *)fp &&
	    ip->ip_off + ip->ip_len > q->ip_off) {
		i = (ip->ip_off + ip->ip_len) - q->ip_off;
		if (i < q->ip_len) {
			q->ip_len -= i;
			q->ip_off += i;
			m_adj(dtom(q), i);
			break;
		}
		q = q->ipf_next;
		m_freem(dtom(q->ipf_prev));
		ip_deq(q->ipf_prev);
	}

insert:
	/*
	 * Stick new segment in its place;
	 * check for complete reassembly.
	 */
	ip_enq(ip, q->ipf_prev);
	next = 0;
	for (q = fp->ipq_next; q != (struct ipasfrag *)fp; q = q->ipf_next) {
		if (q->ip_off != next)
			return (0);
		next += q->ip_len;
	}
	if (q->ipf_prev->ipf_mff)
		return (0);

	/*
	 * Reassembly is complete; concatenate fragments.
	 */
	q = fp->ipq_next;
	m = dtom(q);
	t = m->m_next;
	m->m_next = 0;