ip_input.c

<B>Digital的Unix操作系统VAX 4.2源码</B>

#ifndef lint
static char *sccsid = "@(#)ip_input.c	4.4	(ULTRIX)	9/11/90";
#endif lint

/************************************************************************
 *									*
 *			Copyright (c) 1985 by				*
 *		Digital Equipment Corporation, Maynard, MA		*
 *			All rights reserved.				*
 *									*
 *   This software is furnished under a license and may be used and	*
 *   copied  only  in accordance with the terms of such license and	*
 *   with the  inclusion  of  the  above  copyright  notice.   This	*
 *   software  or  any  other copies thereof may not be provided or	*
 *   otherwise made available to any other person.  No title to and	*
 *   ownership of the software is hereby transferred.			*
 *									*
 *   This software is  derived  from  software  received  from  the	*
 *   University    of   California,   Berkeley,   and   from   Bell	*
 *   Laboratories.  Use, duplication, or disclosure is  subject  to	*
 *   restrictions  under  license  agreements  with  University  of	*
 *   California and with AT&T.						*
 *									*
 *   The information in this software is subject to change  without	*
 *   notice  and should not be construed as a commitment by Digital	*
 *   Equipment Corporation.						*
 *									*
 *   Digital assumes no responsibility for the use  or  reliability	*
 *   of its software on equipment which is not supplied by Digital.	*
 *									*
 ************************************************************************/

/************************************************************************
 *			Modification History				*
 *
 *	31-Jul-90	jsd
 *		fixes for killer packets with IP options LSR and SRR
 *
 *	24-Jul-90	jaw
 *		submitted bug fix for locking problem in ipintr....
 *
 *	16-July-90	U. Sinkewicz
 *		Fix to locking around ip_reass().
 *
 *	7-Jul-90	lp
 *		FDDI performance changes.
 *
 *	14-11-89	Ursula Sinkewicz
 *		Removed lk_ifnet and lk_in_ifaddr to coincide with
 *		slip changes.
 *
 *	17-Oct-89	Uttam Shikarpur					*
 *		Added counter to track the number of bad options for    *
 *		network management. (ips_badoptions)		   	*
 *									*
 *	27-Mar-89	Ursula Sinkewicz
 *		Lowered ipl in lk_rtentry, lk_ifnet, lk_in_ifaddr, exchanged
 *		ip statistics locking for a macro, as per lp changes 
 *		made 3/16/89.
 *
 *	3-Mar-89	Ursula Sinkewicz
 *		Added support for new directory layout; added pmax
 *		support.
 *
 *	13-Feb-89	Ursula Sinkewicz
 *		SMP: Added lk_in_ifaddr and lk_ifnet.
 *	
 *	15-Jan-88	lp
 *		Merge of final 43BSD changes. Use new memory allocation
 *	scheme for mbufs.
 *
 *	Larry Cohen  -	09/16/85					*
 * 		Add 43bsd alpha tape changes for subnet routing		*
 *									*
 *	Larry Cohen  -  01/17/86					*
 *		- will now accept broadcast of ones or zeros   		*
 *									*
 *	Marc Teitelbaum and Fred Templin - 08/21/86			*
 *		Added 4.3bsd beta tape enhancements. "ipintr()" now	*
 *		handles subet broadcasts and packet forwarding properly.*
 *		(Packets are only forwarded by hosts configured with	*
 *		"GATEWAY" option).					*
 ************************************************************************/

/*
 * Copyright (c) 1982, 1986 Regents of the University of California.
 * All rights reserved.  The Berkeley software License Agreement
 * specifies the terms and conditions for redistribution.
 *
 *	@(#)ip_input.c	7.7 (Berkeley) 10/31/87
 */

#include "../h/param.h"
#include "../h/systm.h"
#include "../h/mbuf.h"
#include "../h/domain.h"
#include "../h/protosw.h"
#include "../h/socket.h"
#include "../h/errno.h"
#include "../h/time.h"
#include "../h/kernel.h"
#include "../h/smp_lock.h"

#include "../net/net/if.h"
#include "../net/net/route.h"

#include "../net/netinet/in.h"
#include "../net/netinet/in_pcb.h"
#include "../net/netinet/in_systm.h"
#include "../net/netinet/in_var.h"
#include "../net/netinet/ip.h"
#include "../net/netinet/ip_var.h"
#include "../net/netinet/ip_icmp.h"
#include "../net/netinet/tcp.h"

u_char	ip_protox[IPPROTO_MAX];
int	ipqmaxlen = IFQ_MAXLEN;
struct	in_ifaddr *in_ifaddr;			/* first inet address */
struct	lock_t	lk_ipstat;	/* SMP: ipstat lock, see ip_var.h */
struct	lock_t  lk_ipq;		/* SMP: ip reassembly queue lock */

/*
 * We need to save the IP options in case a protocol wants to respond
 * to an incoming packet over the same route if the packet got here
 * using IP source routing.  This allows connection establishment and
 * maintenance when the remote end is on a network that is not known
 * to us.
 */
int	ip_nhops = 0;
static	struct ip_srcrt {
	char	nop;				/* one NOP to align */
	char	srcopt[IPOPT_OFFSET + 1];	/* OPTVAL, OLEN and OFFSET */
	struct	in_addr route[MAX_IPOPTLEN];
} ip_srcrt;

extern struct lock_t lk_arptab;
/*
 * 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;

	pr = pffindproto(PF_INET, IPPROTO_RAW, SOCK_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_domain->dom_family == PF_INET &&
		    pr->pr_protocol && pr->pr_protocol != IPPROTO_RAW)
			ip_protox[pr->pr_protocol] = pr - inetsw;
	/* SMP: Initailize locks here - ipstat, ipq, and ipintrq.  4.22.87.us */
	lockinit(&lk_ipq, &lock_ipq_d);
	lockinit(&lk_ipstat, &lock_ipstat_d);
	ipq.next = ipq.prev = &ipq;
	ip_id = time.tv_sec & 0xffff;
	ipintrq.ifq_maxlen = ipqmaxlen;
	lockinit(&ipintrq.lk_ifqueue, &lock_ifqueue_d);
	lockinit(&lk_arptab, &lock_arptab_d);
}

u_char	ipcksum = 1;
struct	ip *ip_reass();
struct	sockaddr_in ipaddr = { AF_INET };
struct 	route ipforward_rt;

/*
 * 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()
{
	register struct ip *ip;
	register struct mbuf *m;
	struct mbuf *m0;
	register int i;
	register struct ipq *fp;
	register struct in_ifaddr *ia;
	struct ifnet *ifp;
	int hlen, s;
        /* 
         * This is to keep these long aligned
         */
        struct in_addr dst;
        struct in_addr src;

next:
	/*
	 * Get next datagram off input queue and get IP header
	 * in first mbuf.
	 */
	s = splimp();
	/* SMP: Lock element taken off ipintrq along with
	 * next/prev pointers in the queue.  4.22.87.us
	 */
	smp_lock(&ipintrq.lk_ifqueue, LK_RETRY);
	IF_DEQUEUEIF(&ipintrq, m, ifp);
	smp_unlock(&ipintrq.lk_ifqueue);
	splx(s);
	if (m == 0)
		return;
	/*
	 * If no IP addresses have been set yet but the interfaces
	 * are receiving, can't do anything with incoming packets yet.
 	 */
	if (in_ifaddr == NULL){
		goto bad;
	}
	IPSTAT(ips_total++);
	if ((m->m_off > MMAXOFF || m->m_len < sizeof (struct ip)) &&
	    (m = m_pullup(m, sizeof (struct ip))) == 0) {
		IPSTAT(ips_toosmall++);
		goto next;
	}
	ip = mtod(m, struct ip *);
	hlen = ip->ip_hl << 2;
	if (hlen < sizeof(struct ip)) {	/* minimum header length */
		IPSTAT(ips_badhlen++);
		goto bad;
	}
	if (hlen > m->m_len) {
		if ((m = m_pullup(m, hlen)) == 0) {
			IPSTAT(ips_badhlen++);
			goto next;
		}
		ip = mtod(m, struct ip *);
	}
	if (ipcksum)
		if (ip->ip_sum = in_cksum(m, hlen)) {
			IPSTAT(ips_badsum++);
			goto bad;
		}

	/*
	 * Convert fields to host representation.
	 */
	ip->ip_len = ntohs((u_short)ip->ip_len);
	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 = -(u_short)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 and the original packet to be freed).
	 */
	ip_nhops = 0;		/* for source routed packets */
	if (hlen > sizeof (struct ip) && ip_dooptions(ip, ifp))
		goto next;

	/*
	 * Check our list of addresses, to see if the packet is for us.
	 */
        bcopy(&ip->ip_dst, &dst, sizeof(struct in_addr));
	for (ia = in_ifaddr; ia; ia = ia->ia_next) {
#define	satosin(sa)	((struct sockaddr_in *)(sa))

		if (IA_SIN(ia)->sin_addr.s_addr == dst.s_addr) {
			goto ours;
		}
		if (
#ifdef DIRECTED_BROADCAST
			ia->ia_ifp == ifp &&
#endif DIRECTED_BROADCAST
			(ia->ia_ifp->if_flags & IFF_BROADCAST)) {
		    u_long t;

		    if (satosin(&ia->ia_broadaddr)->sin_addr.s_addr ==
			dst.s_addr) {
			    goto ours;
		    }
		    if (dst.s_addr == ia->ia_netbroadcast.s_addr) {
			    goto ours;
		    }
		    /*
		     * Look for all-0's host part (old broadcast addr),
		     * either for subnet or net.
		     */
		    t = ntohl(dst.s_addr);
		    if (t == ia->ia_subnet) {
			    goto ours;
		    }
		    if (t == ia->ia_net) {
			    goto ours;
		    }
		}
	}
	if (dst.s_addr == (u_long)INADDR_BROADCAST)
		goto ours;
	if (dst.s_addr == INADDR_ANY)
		goto ours;

	/*
	 * Not for us; forward if possible and desirable.
	 */
	ip_forwardscreen(ip, ifp);
	goto next;

ours:
	/*
	 * Look for queue of fragments
	 * of this datagram.
	 */
	smp_lock(&lk_ipq, LK_RETRY);
	if(ipq.next != &ipq) {
        	bcopy(&ip->ip_src, &src, sizeof(struct in_addr));
		for (fp = ipq.next; fp != &ipq; fp = fp->next)
			if (ip->ip_id == fp->ipq_id &&
		    	src.s_addr == fp->ipq_src.s_addr &&
		    	dst.s_addr == fp->ipq_dst.s_addr &&
		    	ip->ip_p == fp->ipq_p) {
				goto found;
		    	}
	}
	fp = 0;
found:

	/*
	 * 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) {
		IPSTAT(ips_fragments++);

#ifdef mips
                /*
                 * if the ip header is not aligned properly, copy the
                 * mbuf to avoid lots of bcopy's later when reassembling
                 * fragments due to placement of fragment pointers within
                 * ip header.
                   */
                if (((u_int)ip & 0x3) != 0) {
                        register struct mbuf *mnew;

                        mnew = m_copy(m, 0, m->m_len);
                        mnew->m_next = m_free(m);
                        m = mnew;
                        ip = mtod(m, struct ip *);
                }
#endif mips
		ip = ip_reass((struct ipasfrag *)ip, fp);
        	smp_unlock(&lk_ipq);
		if (ip == 0)
			goto next;
		m = dtom(ip);
	} else{
		if (fp) {
			ip_freef(fp);
		}
        	smp_unlock(&lk_ipq);
	}
	/*
	 * Switch out to protocol's input routine.
	 */