/* * Copyright (c) 1982, 1986, 1988, 1990 Regents of the University of California. * All rights reserved. * * %sccs.include.redist.c% * * @(#)ip_output.c 7.23 (Berkeley) 11/12/90 */ #include "param.h" #include "malloc.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_systm.h" #include "ip.h" #include "in_pcb.h" #include "in_var.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 (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. */ ip_output(m0, opt, ro, flags) struct mbuf *m0; struct mbuf *opt; struct route *ro; int flags; { register struct ip *ip, *mhip; register struct ifnet *ifp; register struct mbuf *m = m0; register int hlen = sizeof (struct ip); int len, off, error = 0; struct route iproute; struct sockaddr_in *dst; struct in_ifaddr *ia; #ifdef DIAGNOSTIC if ((m->m_flags & M_PKTHDR) == 0) panic("ip_output no HDR"); #endif if (opt) { m = ip_insertoptions(m, opt, &len); hlen = len; } 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; ipstat.ips_localout++; } /* * Route packet. */ 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_len = sizeof(*dst); dst->sin_addr = ip->ip_dst; } /* * If routing to interface only, * short circuit routing lookup. */ if (flags & IP_ROUTETOIF) { ia = (struct in_ifaddr *)ifa_ifwithdstaddr((struct sockaddr *)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) { error = EHOSTUNREACH; goto bad; } ia = (struct in_ifaddr *)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; } #ifndef notdef /* * If source address not specified yet, use address * of outgoing interface. */ if (ip->ip_src.s_addr == INADDR_ANY) ip->ip_src = IA_SIN(ia)->sin_addr; #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 ((u_short)ip->ip_len > ifp->if_mtu) { error = EMSGSIZE; goto bad; } m->m_flags |= M_BCAST; } /* * 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; ip->ip_sum = in_cksum(m, hlen); error = (*ifp->if_output)(ifp, m, (struct sockaddr *)dst, ro->ro_rt); goto done; } ipstat.ips_fragmented++; /* * 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; } { 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; goto sendorfree; } 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_hl = mhlen >> 2; } 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) { error = ENOBUFS; /* ??? */ 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; mhip->ip_sum = in_cksum(m, mhlen); *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; ip->ip_sum = in_cksum(m, hlen); 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); } } done: if (ro == &iproute && (flags & IP_ROUTETOIF) == 0 && ro->ro_rt) RTFREE(ro->ro_rt); 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. */ 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; bcopy((caddr_t)ip, mtod(m, caddr_t), sizeof(struct ip)); } 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((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, * omitting those not copied during fragmentation. */ 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->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]; /* bogus lengths should have been caught by ip_dooptions */ if (optlen > cnt) optlen = cnt; if (IPOPT_COPIED(opt)) { bcopy((caddr_t)cp, (caddr_t)dp, (unsigned)optlen); dp += optlen; } } for (optlen = dp - (u_char *)(jp+1); optlen & 0x3; optlen++) *dp++ = IPOPT_EOL; return (optlen); } /* * IP socket option processing. */ ip_ctloutput(op, so, level, optname, mp) int op; struct socket *so; int level, optname; struct mbuf **mp; { register struct inpcb *inp = sotoinpcb(so); register struct mbuf *m = *mp; register int optval; int error = 0; if (level != IPPROTO_IP) error = EINVAL; else switch (op) { case PRCO_SETOPT: switch (optname) { case IP_OPTIONS: #ifdef notyet case IP_RETOPTS: return (ip_pcbopts(optname, &inp->inp_options, m)); #else return (ip_pcbopts(&inp->inp_options, m)); #endif case IP_TOS: case IP_TTL: case IP_RECVOPTS: case IP_RECVRETOPTS: case IP_RECVDSTADDR: if (m->m_len != sizeof(int)) error = EINVAL; else { optval = *mtod(m, int *); switch (optname) { case IP_TOS: inp->inp_ip.ip_tos = optval; break; case IP_TTL: inp->inp_ip.ip_ttl = optval; break; #define OPTSET(bit) \ if (optval) \ inp->inp_flags |= bit; \ else \ inp->inp_flags &= ~bit; case IP_RECVOPTS: OPTSET(INP_RECVOPTS); break; case IP_RECVRETOPTS: OPTSET(INP_RECVRETOPTS); break; case IP_RECVDSTADDR: OPTSET(INP_RECVDSTADDR); break; } } break; #undef OPTSET default: error = EINVAL; break; } if (m) (void)m_free(m); break; case PRCO_GETOPT: switch (optname) { case IP_OPTIONS: case IP_RETOPTS: *mp = m = m_get(M_WAIT, MT_SOOPTS); if (inp->inp_options) { 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; case IP_TOS: case IP_TTL: case IP_RECVOPTS: case IP_RECVRETOPTS: case IP_RECVDSTADDR: *mp = m = m_get(M_WAIT, MT_SOOPTS); m->m_len = sizeof(int); switch (optname) { case IP_TOS: optval = inp->inp_ip.ip_tos; break; case IP_TTL: optval = inp->inp_ip.ip_ttl; 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; } *mtod(m, int *) = optval; break; default: error = EINVAL; 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. */ #ifdef notyet ip_pcbopts(optname, pcbopt, m) int optname; #else ip_pcbopts(pcbopt, m) #endif struct mbuf **pcbopt; register struct mbuf *m; { 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 (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); }