new stats; change keepalives to use rcv_nxt instead of rcv_nxt-1

[unix-history] / usr / src / sys / netinet / ip_output.c

diff --git a/usr/src/sys/netinet/ip_output.c b/usr/src/sys/netinet/ip_output.c
index c494a1f..a40cf06 100644 (file)
--- a/usr/src/sys/netinet/ip_output.c
+++ b/usr/src/sys/netinet/ip_output.c
@@ -1,126 +1,456 @@
-/* ip_output.c 1.10 81/11/02 */
-
-#include "../h/param.h"
-#include "../h/mbuf.h"
-#include "../h/mtpr.h"
-#include "../h/socket.h"
-#include "../inet/inet_cksum.h"
-#include "../inet/inet.h"
-#include "../inet/inet_systm.h"
-#include "../inet/imp.h"
-#include "../inet/inet_host.h"
-#include "../inet/ip.h"
-#include "../inet/tcp.h"
-
-ip_output(mp)
-       struct mbuf *mp;
+/*
+ * 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_output.c 7.2 (Berkeley) %G%
+ */
+
+#include "param.h"
+#include "mbuf.h"
+#include "errno.h"
+#include "protosw.h"
+#include "socket.h"
+#include "socketvar.h"
+
+#include "../net/if.h"
+#include "../net/route.h"
+
+#include "in.h"
+#include "in_pcb.h"
+#include "in_systm.h"
+#include "in_var.h"
+#include "ip.h"
+#include "ip_var.h"
+
+#ifdef vax
+#include "../machine/mtpr.h"
+#endif
+
+struct mbuf *ip_insertoptions();
+
+/*
+ * IP output.  The packet in mbuf chain m contains a skeletal IP
+ * header (as ipovly).  The mbuf chain containing the packet will
+ * be freed.  The mbuf opt, if present, will not be freed.
+ */
+ip_output(m, opt, ro, flags)
+       struct mbuf *m;
+       struct mbuf *opt;
+       struct route *ro;
+       int flags;

 {
 {

-       register i, rnd;
-       register struct mbuf *m, *n;
-       register struct ip *p;
-       struct mbuf *mm;
-       int hlen, adj, max, len, off;
+       register struct ip *ip;
+       register struct ifnet *ifp;
+       int len, hlen = sizeof (struct ip), off, error = 0;
+       struct route iproute;
+       struct sockaddr_in *dst;

 
 

-COUNT(IP_OUTPUT);
-       p = mtod(mp, struct ip *);
-       hlen = sizeof (struct ip);
+       if (opt)
+               m = ip_insertoptions(m, opt, &hlen);
+       ip = mtod(m, struct ip *);
+       /*
+        * Fill in IP header.
+        */
+       if ((flags & IP_FORWARDING) == 0) {
+               ip->ip_v = IPVERSION;
+               ip->ip_off &= IP_DF;
+               ip->ip_id = htons(ip_id++);
+               ip->ip_hl = hlen >> 2;
+       } else
+               hlen = ip->ip_hl << 2;

 
        /*
 
        /*

-        * Fill in and byte swap ip header.
+        * Route packet.

         */
         */

-       p->ip_v = IPVERSION;
-       p->ip_hl = hlen >> 2;
-       p->ip_off = 0 | (p->ip_off & IP_DF);
-       p->ip_ttl = MAXTTL;
-       p->ip_id = ip_id++;
-
-       if (p->ip_len <= MTU)
-               return (ip_send(p));
-       if (p->ip_off & IP_DF)
-               return (0);
-       max = MTU - hlen;
-       len = p->ip_len - hlen;
-       off = 0;
-       m = mp;
-       while (len > 0) {
-               p->ip_off |= off >> 3;
-               i = -hlen;
-               while (m != NULL) {
-                       i += m->m_len;
-                       if (i > max)
+       if (ro == 0) {
+               ro = &iproute;
+               bzero((caddr_t)ro, sizeof (*ro));
+       }
+       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_addr = ip->ip_dst;
+       }
+       /*
+        * If routing to interface only,
+        * short circuit routing lookup.
+        */
+       if (flags & IP_ROUTETOIF) {
+               struct in_ifaddr *ia;
+
+               ia = (struct in_ifaddr *)ifa_ifwithdstaddr(dst);
+               if (ia == 0)
+                       ia = in_iaonnetof(in_netof(ip->ip_dst));
+               if (ia == 0) {
+                       error = ENETUNREACH;
+                       goto bad;
+               }
+               ifp = ia->ia_ifp;
+       } else {
+               if (ro->ro_rt == 0)
+                       rtalloc(ro);
+               if (ro->ro_rt == 0 || (ifp = ro->ro_rt->rt_ifp) == 0) {
+                       error = ENETUNREACH;
+                       goto bad;
+               }
+               ro->ro_rt->rt_use++;
+               if (ro->ro_rt->rt_flags & (RTF_GATEWAY|RTF_HOST))
+                       dst = (struct sockaddr_in *)&ro->ro_rt->rt_gateway;
+       }
+#ifndef notdef
+       /*
+        * If source address not specified yet, use address
+        * of outgoing interface.
+        */
+       if (ip->ip_src.s_addr == INADDR_ANY) {
+               register struct in_ifaddr *ia;
+
+               for (ia = in_ifaddr; ia; ia = ia->ia_next)
+                       if (ia->ia_ifp == ifp) {
+                               ip->ip_src = IA_SIN(ia)->sin_addr;

                                break;
                                break;

-                       n = m;
-                       m = m->m_next;
+                       }
+       }
+#endif
+       /*
+        * Look for broadcast address and
+        * and verify user is allowed to send
+        * such a packet.
+        */
+       if (in_broadcast(dst->sin_addr)) {
+               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 (ip->ip_len > ifp->if_mtu) {
+                       error = EMSGSIZE;
+                       goto bad;
+               }
+       }
+
+       /*
+        * If small enough for interface, can just send directly.
+        */
+       if (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;
+               ip->ip_sum = in_cksum(m, hlen);
+               error = (*ifp->if_output)(ifp, m, (struct sockaddr *)dst);
+               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;
+               goto bad;
+       }
+       len = (ifp->if_mtu - hlen) &~ 7;
+       if (len < 8) {
+               error = EMSGSIZE;
+               goto bad;
+       }
+
+       /*
+        * Discard IP header from logical mbuf for m_copy's sake.
+        * Loop through length of segment, make a copy of each
+        * part and output.
+        */
+       m->m_len -= sizeof (struct ip);
+       m->m_off += sizeof (struct ip);
+       for (off = 0; off < ip->ip_len-hlen; off += len) {
+               struct mbuf *mh = m_get(M_DONTWAIT, MT_HEADER);
+               struct ip *mhip;
+
+               if (mh == 0) {
+                       error = ENOBUFS;
+                       goto bad;
+               }
+               mh->m_off = MMAXOFF - hlen;
+               mhip = mtod(mh, struct ip *);
+               *mhip = *ip;
+               if (hlen > sizeof (struct ip)) {
+                       int olen = ip_optcopy(ip, mhip, off);
+                       mh->m_len = sizeof (struct ip) + olen;
+               } else
+                       mh->m_len = sizeof (struct ip);
+               mhip->ip_off = (off >> 3) + (ip->ip_off & ~IP_MF);
+               if (ip->ip_off & IP_MF)
+                       mhip->ip_off |= IP_MF;
+               if (off + len >= ip->ip_len-hlen)
+                       len = mhip->ip_len = ip->ip_len - hlen - off;
+               else {
+                       mhip->ip_len = len;
+                       mhip->ip_off |= IP_MF;

                }
                }

-               if (i < max || m == NULL) {
-                       p->ip_off = p->ip_off &~ IP_MF;
-                       p->ip_len = i + hlen;
-                       return (ip_send(p));
+               mhip->ip_len += sizeof (struct ip);
+               mhip->ip_len = htons((u_short)mhip->ip_len);
+               mh->m_next = m_copy(m, off, len);
+               if (mh->m_next == 0) {
+                       (void) m_free(mh);
+                       error = ENOBUFS;        /* ??? */
+                       goto bad;

                }
                }

-               if ((mm = m_get(1)) == NULL)    /* no more bufs */
-                       return(0);
-               p->ip_off |= IP_MF;
-               i -= m->m_len;
-               rnd = i & ~7;
-               adj = i - rnd;
-               p->ip_len = rnd + hlen;
-               n->m_next = NULL;
-               mm->m_next = m;
-               m = mm;
-               m->m_off = MMAXOFF - hlen - adj;
-               m->m_len = hlen + adj;
-               bcopy(p, (caddr_t)((int)m + m->m_off), hlen);
-               if (adj) {
-                       n->m_len -= adj;
-                       bcopy((caddr_t)((int)n + n->m_len + n->m_off),
-                             (caddr_t)((int)m + m->m_off + hlen), adj);
+               mhip->ip_off = htons((u_short)mhip->ip_off);
+               mhip->ip_sum = 0;
+               mhip->ip_sum = in_cksum(mh, hlen);
+               if (error = (*ifp->if_output)(ifp, mh, (struct sockaddr *)dst))
+                       break;
+       }
+bad:
+       m_freem(m);
+done:
+       if (ro == &iproute && (flags & IP_ROUTETOIF) == 0 && ro->ro_rt)
+               RTFREE(ro->ro_rt);
+       return (error);
+}
+
+/*
+ * 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.
+ */
+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 (p->ipopt_dst.s_addr)
+               ip->ip_dst = p->ipopt_dst;
+       if (m->m_off >= MMAXOFF || MMINOFF + optlen > m->m_off) {
+               MGET(n, M_DONTWAIT, MT_HEADER);
+               if (n == 0)
+                       return (m);
+               m->m_len -= sizeof(struct ip);
+               m->m_off += sizeof(struct ip);
+               n->m_next = m;
+               m = n;
+               m->m_off = MMAXOFF - sizeof(struct ip) - optlen;
+               m->m_len = optlen + sizeof(struct ip);
+               bcopy((caddr_t)ip, mtod(m, caddr_t), sizeof(struct ip));
+       } else {
+               m->m_off -= optlen;
+               m->m_len += optlen;
+               ovbcopy((caddr_t)ip, mtod(m, caddr_t), sizeof(struct ip));
+       }
+       ip = mtod(m, struct ip *);
+       bcopy((caddr_t)p->ipopt_list, (caddr_t)(ip + 1), (unsigned)optlen);
+       *phlen = sizeof(struct ip) + optlen;
+       ip->ip_len += optlen;
+       return (m);
+}
+
+/*
+ * Copy options from ip to jp.
+ * If off is 0 all options are copied
+ * otherwise copy selectively.
+ */
+ip_optcopy(ip, jp, off)
+       struct ip *ip, *jp;
+       int off;
+{
+       register u_char *cp, *dp;
+       int opt, optlen, cnt;
+
+       cp = (u_char *)(ip + 1);
+       dp = (u_char *)(jp + 1);
+       cnt = (ip->ip_hl << 2) - sizeof (struct ip);
+       for (; cnt > 0; cnt -= optlen, cp += optlen) {
+               opt = cp[0];
+               if (opt == IPOPT_EOL)
+                       break;
+               if (opt == IPOPT_NOP)
+                       optlen = 1;
+               else
+                       optlen = cp[IPOPT_OLEN];
+               if (optlen > cnt)                       /* XXX */
+                       optlen = cnt;                   /* XXX */
+               if (off == 0 || IPOPT_COPIED(opt)) {
+                       bcopy((caddr_t)cp, (caddr_t)dp, (unsigned)optlen);
+                       dp += optlen;

                }
                }

-               ip_send(p);
-               p = (struct ip *)((int)m + m->m_off);
-               len -= rnd;
-               off += rnd;

        }
        }

+       for (optlen = dp - (u_char *)(jp+1); optlen & 0x3; optlen++)
+               *dp++ = IPOPT_EOL;
+       return (optlen);

 }
 
 }
 

-ip_send(ip)
-       register struct ip *ip;         /* known to be r11 */
+/*
+ * IP socket option processing.
+ */
+ip_ctloutput(op, so, level, optname, m)
+       int op;
+       struct socket *so;
+       int level, optname;
+       struct mbuf **m;

 {
 {

+       int error = 0;
+       struct inpcb *inp = sotoinpcb(so);
+
+       if (level != IPPROTO_IP)
+               error = EINVAL;
+       else switch (op) {
+
+       case PRCO_SETOPT:
+               switch (optname) {
+               case IP_OPTIONS:
+                       return (ip_pcbopts(&inp->inp_options, *m));
+
+               default:
+                       error = EINVAL;
+                       break;
+               }
+               break;
+
+       case PRCO_GETOPT:
+               switch (optname) {
+               case IP_OPTIONS:
+                       *m = m_get(M_WAIT, MT_SOOPTS);
+                       if (inp->inp_options) {
+                               (*m)->m_off = inp->inp_options->m_off;
+                               (*m)->m_len = inp->inp_options->m_len;
+                               bcopy(mtod(inp->inp_options, caddr_t),
+                                   mtod(*m, caddr_t), (unsigned)(*m)->m_len);
+                       } else
+                               (*m)->m_len = 0;
+                       break;
+               default:
+                       error = EINVAL;
+                       break;
+               }
+               break;
+       }
+       if (op == PRCO_SETOPT)
+               (void)m_free(*m);
+       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.
+ */
+ip_pcbopts(pcbopt, m)
+       struct mbuf **pcbopt;

        register struct mbuf *m;
        register struct mbuf *m;

-       register struct imp *l;
-       register int hlen = ip->ip_hl << 2;
-       int s;
-COUNT(IP_SEND);
-
-       m = dtom(ip);
-       l = (struct imp *)((int)m + m->m_off - L1822);
-       l->i_shost = ip->ip_src.s_host;
-       l->i_dhost = ip->ip_dst.s_host;
-       l->i_type = IPTYPE;
-       ip->ip_sum = 0;
-       ip->ip_len = htons(ip->ip_len);
-       ip->ip_id = htons(ip->ip_id);
-       ip->ip_off = htons(ip->ip_off);
-       CKSUM_IPSET(m, ip, r11, hlen);
-       m->m_off -= L1822;
-       m->m_len += L1822;
-       m->m_act = NULL;
-#ifndef IMPLOOP
-       s = splimp();
-       if (imp_stat.outq_head != NULL)
-               imp_stat.outq_tail->m_act = m;
-       else
-               imp_stat.outq_head = m;
-       imp_stat.outq_tail = m;
-       splx(s);
-       if (!imp_stat.outactive)
-               enstart(0);
+{
+       register cnt, optlen;
+       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(long))
+               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    MAX_IPOPTLEN >= MMAXOFF - MMINOFF
+       if (m->m_off + m->m_len + sizeof(struct in_addr) > MAX_IPOPTLEN)
+               goto bad;

 #else
 #else

-       if (imp_stat.inq_head != NULL)
-               imp_stat.inq_tail->m_act = m;
-       else
-               imp_stat.inq_head = m;
-       imp_stat.inq_tail = m;
-       setsoftnet();
-#endif IMPLOOP
-       return (1);
+       if (m->m_off + m->m_len + sizeof(struct in_addr) > MMAXOFF)
+               goto bad;
+#endif
+       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;
+               }
+       }
+       *pcbopt = m;
+       return (0);
+
+bad:
+       (void)m_free(m);
+       return (EINVAL);

 }
 }