/* * Copyright (c) 1982, 1986, 1988, 1990, 1993 * The Regents of the University of California. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. All advertising materials mentioning features or use of this software * must display the following acknowledgement: * This product includes software developed by the University of * California, Berkeley and its contributors. * 4. Neither the name of the University nor the names of its contributors * may be used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * @(#)udp_usrreq.c 8.4 (Berkeley) 1/21/94 * udp_usrreq.c,v 1.4 1994/10/02 17:48:45 phk Exp */ /* * Changes and additions relating to SLiRP * Copyright (c) 1995 Danny Gasparovski. * * Please read the file COPYRIGHT for the * terms and conditions of the copyright. */ #include #include "ip_icmp.h" /* * UDP protocol implementation. * Per RFC 768, August, 1980. */ #define udpcksum 1 void udp_init(PNATState pData) { udp_last_so = &udb; udb.so_next = udb.so_prev = &udb; } /* m->m_data points at ip packet header * m->m_len length ip packet * ip->ip_len length data (IPDU) */ void udp_input(PNATState pData, register struct mbuf *m, int iphlen) { register struct ip *ip; register struct udphdr *uh; /* struct mbuf *opts = 0;*/ int len; struct ip save_ip; struct socket *so; DEBUG_CALL("udp_input"); DEBUG_ARG("m = %lx", (long)m); DEBUG_ARG("iphlen = %d", iphlen); udpstat.udps_ipackets++; /* * Strip IP options, if any; should skip this, * make available to user, and use on returned packets, * but we don't yet have a way to check the checksum * with options still present. */ if(iphlen > sizeof(struct ip)) { ip_stripoptions(m, (struct mbuf *)0); iphlen = sizeof(struct ip); } /* * Get IP and UDP header together in first mbuf. */ ip = mtod(m, struct ip *); uh = (struct udphdr *)((caddr_t)ip + iphlen); /* * Make mbuf data length reflect UDP length. * If not enough data to reflect UDP length, drop. */ len = ntohs((u_int16_t)uh->uh_ulen); if (ip->ip_len != len) { if (len > ip->ip_len) { udpstat.udps_badlen++; goto bad; } m_adj(m, len - ip->ip_len); ip->ip_len = len; } /* * Save a copy of the IP header in case we want restore it * for sending an ICMP error message in response. */ save_ip = *ip; save_ip.ip_len+= iphlen; /* tcp_input subtracts this */ /* * Checksum extended UDP header and data. */ if (udpcksum && uh->uh_sum) { ((struct ipovly *)ip)->ih_next = 0; ((struct ipovly *)ip)->ih_prev = 0; ((struct ipovly *)ip)->ih_x1 = 0; ((struct ipovly *)ip)->ih_len = uh->uh_ulen; /* keep uh_sum for ICMP reply * uh->uh_sum = cksum(m, len + sizeof (struct ip)); * if (uh->uh_sum) { */ if(cksum(m, len + sizeof(struct ip))) { udpstat.udps_badsum++; goto bad; } } /* * handle DHCP/BOOTP */ if (ntohs(uh->uh_dport) == BOOTP_SERVER) { bootp_input(pData, m); goto bad; } /* * handle TFTP */ if (ntohs(uh->uh_dport) == TFTP_SERVER) { tftp_input(pData, m); goto bad; } /* * Locate pcb for datagram. */ so = udp_last_so; if (so->so_lport != uh->uh_sport || so->so_laddr.s_addr != ip->ip_src.s_addr) { struct socket *tmp; for (tmp = udb.so_next; tmp != &udb; tmp = tmp->so_next) { if (tmp->so_lport == uh->uh_sport && tmp->so_laddr.s_addr == ip->ip_src.s_addr) { so = tmp; break; } } if (tmp == &udb) { so = NULL; } else { udpstat.udpps_pcbcachemiss++; udp_last_so = so; } } if (so == NULL) { /* * If there's no socket for this packet, * create one */ if ((so = socreate()) == NULL) goto bad; if(udp_attach(pData, so) == -1) { DEBUG_MISC((dfd," udp_attach errno = %d-%s\n", errno,strerror(errno))); sofree(pData, so); goto bad; } /* * Setup fields */ /* udp_last_so = so; */ so->so_laddr = ip->ip_src; so->so_lport = uh->uh_sport; if ((so->so_iptos = udp_tos(so)) == 0) so->so_iptos = ip->ip_tos; /* * XXXXX Here, check if it's in udpexec_list, * and if it is, do the fork_exec() etc. */ } so->so_faddr = ip->ip_dst; /* XXX */ so->so_fport = uh->uh_dport; /* XXX */ iphlen += sizeof(struct udphdr); m->m_len -= iphlen; m->m_data += iphlen; /* * Now we sendto() the packet. */ if (so->so_emu) udp_emu(pData, so, m); if(sosendto(pData, so,m) == -1) { m->m_len += iphlen; m->m_data -= iphlen; *ip=save_ip; DEBUG_MISC((dfd,"udp tx errno = %d-%s\n",errno,strerror(errno))); icmp_error(pData, m, ICMP_UNREACH,ICMP_UNREACH_NET, 0,strerror(errno)); } m_free(pData, so->so_m); /* used for ICMP if error on sorecvfrom */ /* restore the orig mbuf packet */ m->m_len += iphlen; m->m_data -= iphlen; *ip=save_ip; so->so_m=m; /* ICMP backup */ return; bad: m_freem(pData, m); /* if (opts) m_freem(opts); */ return; } int udp_output2(PNATState pData, struct socket *so, struct mbuf *m, struct sockaddr_in *saddr, struct sockaddr_in *daddr, int iptos) { register struct udpiphdr *ui; int error = 0; DEBUG_CALL("udp_output"); DEBUG_ARG("so = %lx", (long)so); DEBUG_ARG("m = %lx", (long)m); DEBUG_ARG("saddr = %lx", (long)saddr->sin_addr.s_addr); DEBUG_ARG("daddr = %lx", (long)daddr->sin_addr.s_addr); /* * Adjust for header */ m->m_data -= sizeof(struct udpiphdr); m->m_len += sizeof(struct udpiphdr); /* * Fill in mbuf with extended UDP header * and addresses and length put into network format. */ ui = mtod(m, struct udpiphdr *); ui->ui_next = ui->ui_prev = 0; ui->ui_x1 = 0; ui->ui_pr = IPPROTO_UDP; ui->ui_len = htons(m->m_len - sizeof(struct ip)); /* + sizeof (struct udphdr)); */ /* XXXXX Check for from-one-location sockets, or from-any-location sockets */ ui->ui_src = saddr->sin_addr; ui->ui_dst = daddr->sin_addr; ui->ui_sport = saddr->sin_port; ui->ui_dport = daddr->sin_port; ui->ui_ulen = ui->ui_len; /* * Stuff checksum and output datagram. */ ui->ui_sum = 0; if (udpcksum) { if ((ui->ui_sum = cksum(m, /* sizeof (struct udpiphdr) + */ m->m_len)) == 0) ui->ui_sum = 0xffff; } ((struct ip *)ui)->ip_len = m->m_len; ((struct ip *)ui)->ip_ttl = ip_defttl; ((struct ip *)ui)->ip_tos = iptos; udpstat.udps_opackets++; error = ip_output(pData, so, m); return (error); } int udp_output(PNATState pData, struct socket *so, struct mbuf *m, struct sockaddr_in *addr) { struct sockaddr_in saddr, daddr; saddr = *addr; if ((so->so_faddr.s_addr & htonl(pData->netmask)) == special_addr.s_addr) { saddr.sin_addr.s_addr = so->so_faddr.s_addr; if ((so->so_faddr.s_addr & htonl(~pData->netmask)) == htonl(~pData->netmask)) saddr.sin_addr.s_addr = alias_addr.s_addr; } /* Any UDP packet to the loopback address must be translated to be from * the forwarding address, i.e. 10.0.2.2. */ if ( (saddr.sin_addr.s_addr & htonl(IN_CLASSA_NET)) == htonl(INADDR_LOOPBACK & IN_CLASSA_NET)) saddr.sin_addr.s_addr = alias_addr.s_addr; daddr.sin_addr = so->so_laddr; daddr.sin_port = so->so_lport; return udp_output2(pData, so, m, &saddr, &daddr, so->so_iptos); } int udp_attach(PNATState pData, struct socket *so) { struct sockaddr_in addr; if((so->s = socket(AF_INET,SOCK_DGRAM,0)) != -1) { /* * Here, we bind() the socket. Although not really needed * (sendto() on an unbound socket will bind it), it's done * here so that emulation of ytalk etc. don't have to do it */ addr.sin_family = AF_INET; addr.sin_port = 0; addr.sin_addr.s_addr = INADDR_ANY; if(bind(so->s, (struct sockaddr *)&addr, sizeof(addr))<0) { int lasterrno=errno; closesocket(so->s); so->s=-1; #ifdef _WIN32 WSASetLastError(lasterrno); #else errno=lasterrno; #endif } else { int opt = 1; /* success, insert in queue */ so->so_expire = curtime + SO_EXPIRE; /* enable broadcast for later use */ setsockopt(so->s, SOL_SOCKET, SO_BROADCAST, (const char *)&opt, sizeof(opt)); insque(pData, so,&udb); } } return(so->s); } void udp_detach(PNATState pData, struct socket *so) { /* Correctly update list if detaching last socket in list. */ if (so == udp_last_so) udp_last_so = &udb; closesocket(so->s); /* if (so->so_m) m_free(so->so_m); done by sofree */ sofree(pData, so); } static const struct tos_t udptos[] = { {0, 53, IPTOS_LOWDELAY, 0}, /* DNS */ {517, 517, IPTOS_LOWDELAY, EMU_TALK}, /* talk */ {518, 518, IPTOS_LOWDELAY, EMU_NTALK}, /* ntalk */ {0, 7648, IPTOS_LOWDELAY, EMU_CUSEEME}, /* Cu-Seeme */ {0, 0, 0, 0} }; u_int8_t udp_tos(so) struct socket *so; { int i = 0; while(udptos[i].tos) { if ((udptos[i].fport && ntohs(so->so_fport) == udptos[i].fport) || (udptos[i].lport && ntohs(so->so_lport) == udptos[i].lport)) { so->so_emu = udptos[i].emu; return udptos[i].tos; } i++; } return 0; } #ifdef EMULATE_TALK #include "talkd.h" #endif /* * Here, talk/ytalk/ntalk requests must be emulated */ void udp_emu(PNATState pData, struct socket *so, struct mbuf *m) { struct sockaddr_in addr; socklen_t addrlen = sizeof(addr); #ifdef EMULATE_TALK CTL_MSG_OLD *omsg; CTL_MSG *nmsg; char buff[sizeof(CTL_MSG)]; u_char type; struct talk_request { struct talk_request *next; struct socket *udp_so; struct socket *tcp_so; } *req; static struct talk_request *req_tbl = 0; #endif struct cu_header { uint16_t d_family; /* destination family */ uint16_t d_port; /* destination port */ uint32_t d_addr; /* destination address */ uint16_t s_family; /* source family */ uint16_t s_port; /* source port */ uint32_t so_addr; /* source address */ uint32_t seqn; /* sequence number */ uint16_t message; /* message */ uint16_t data_type; /* data type */ uint16_t pkt_len; /* packet length */ } *cu_head; switch(so->so_emu) { #ifdef EMULATE_TALK case EMU_TALK: case EMU_NTALK: /* * Talk emulation. We always change the ctl_addr to get * some answers from the daemon. When an ANNOUNCE comes, * we send LEAVE_INVITE to the local daemons. Also when a * DELETE comes, we send copies to the local daemons. */ if (getsockname(so->s, (struct sockaddr *)&addr, &addrlen) < 0) return; #define IS_OLD (so->so_emu == EMU_TALK) #define COPY_MSG(dest, src) { dest->type = src->type; \ dest->id_num = src->id_num; \ dest->pid = src->pid; \ dest->addr = src->addr; \ dest->ctl_addr = src->ctl_addr; \ memcpy(&dest->l_name, &src->l_name, NAME_SIZE_OLD); \ memcpy(&dest->r_name, &src->r_name, NAME_SIZE_OLD); \ memcpy(&dest->r_tty, &src->r_tty, TTY_SIZE); } #define OTOSIN(ptr, field) ((struct sockaddr_in *)&ptr->field) /* old_sockaddr to sockaddr_in */ if (IS_OLD) { /* old talk */ omsg = mtod(m, CTL_MSG_OLD*); nmsg = (CTL_MSG *) buff; type = omsg->type; OTOSIN(omsg, ctl_addr)->sin_port = addr.sin_port; OTOSIN(omsg, ctl_addr)->sin_addr = our_addr; strncpy(omsg->l_name, getlogin(), NAME_SIZE_OLD); } else { /* new talk */ omsg = (CTL_MSG_OLD *) buff; nmsg = mtod(m, CTL_MSG *); type = nmsg->type; OTOSIN(nmsg, ctl_addr)->sin_port = addr.sin_port; OTOSIN(nmsg, ctl_addr)->sin_addr = our_addr; strncpy(nmsg->l_name, getlogin(), NAME_SIZE_OLD); } if (type == LOOK_UP) return; /* for LOOK_UP this is enough */ if (IS_OLD) { /* make a copy of the message */ COPY_MSG(nmsg, omsg); nmsg->vers = 1; nmsg->answer = 0; } else COPY_MSG(omsg, nmsg); /* * If if is an ANNOUNCE message, we go through the * request table to see if a tcp port has already * been redirected for this socket. If not, we solisten() * a new socket and add this entry to the table. * The port number of the tcp socket and our IP * are put to the addr field of the message structures. * Then a LEAVE_INVITE is sent to both local daemon * ports, 517 and 518. This is why we have two copies * of the message, one in old talk and one in new talk * format. */ if (type == ANNOUNCE) { int s; u_short temp_port; for(req = req_tbl; req; req = req->next) if (so == req->udp_so) break; /* found it */ if (!req) { /* no entry for so, create new */ req = (struct talk_request *) malloc(sizeof(struct talk_request)); req->udp_so = so; req->tcp_so = solisten(0, OTOSIN(omsg, addr)->sin_addr.s_addr, OTOSIN(omsg, addr)->sin_port, SS_FACCEPTONCE); req->next = req_tbl; req_tbl = req; } /* replace port number in addr field */ addrlen = sizeof(addr); getsockname(req->tcp_so->s, (struct sockaddr *) &addr, &addrlen); OTOSIN(omsg, addr)->sin_port = addr.sin_port; OTOSIN(omsg, addr)->sin_addr = our_addr; OTOSIN(nmsg, addr)->sin_port = addr.sin_port; OTOSIN(nmsg, addr)->sin_addr = our_addr; /* send LEAVE_INVITEs */ temp_port = OTOSIN(omsg, ctl_addr)->sin_port; OTOSIN(omsg, ctl_addr)->sin_port = 0; OTOSIN(nmsg, ctl_addr)->sin_port = 0; omsg->type = nmsg->type = LEAVE_INVITE; s = socket(AF_INET, SOCK_DGRAM, IPPROTO_IP); addr.sin_addr = our_addr; addr.sin_family = AF_INET; addr.sin_port = htons(517); sendto(s, (char *)omsg, sizeof(*omsg), 0, (struct sockaddr *)&addr, sizeof(addr)); addr.sin_port = htons(518); sendto(s, (char *)nmsg, sizeof(*nmsg), 0, (struct sockaddr *) &addr, sizeof(addr)); closesocket(s) ; omsg->type = nmsg->type = ANNOUNCE; OTOSIN(omsg, ctl_addr)->sin_port = temp_port; OTOSIN(nmsg, ctl_addr)->sin_port = temp_port; } /* * If it is a DELETE message, we send a copy to the * local daemons. Then we delete the entry corresponding * to our socket from the request table. */ if (type == DELETE) { struct talk_request *temp_req, *req_next; int s; u_short temp_port; temp_port = OTOSIN(omsg, ctl_addr)->sin_port; OTOSIN(omsg, ctl_addr)->sin_port = 0; OTOSIN(nmsg, ctl_addr)->sin_port = 0; s = socket(AF_INET, SOCK_DGRAM, IPPROTO_IP); addr.sin_addr = our_addr; addr.sin_family = AF_INET; addr.sin_port = htons(517); sendto(s, (char *)omsg, sizeof(*omsg), 0, (struct sockaddr *)&addr, sizeof(addr)); addr.sin_port = htons(518); sendto(s, (char *)nmsg, sizeof(*nmsg), 0, (struct sockaddr *)&addr, sizeof(addr)); closesocket(s); OTOSIN(omsg, ctl_addr)->sin_port = temp_port; OTOSIN(nmsg, ctl_addr)->sin_port = temp_port; /* delete table entry */ if (so == req_tbl->udp_so) { temp_req = req_tbl; req_tbl = req_tbl->next; free(temp_req); } else { temp_req = req_tbl; for(req = req_tbl->next; req; req = req_next) { req_next = req->next; if (so == req->udp_so) { temp_req->next = req_next; free(req); break; } else { temp_req = req; } } } } return; #endif case EMU_CUSEEME: /* * Cu-SeeMe emulation. * Hopefully the packet is more that 16 bytes long. We don't * do any other tests, just replace the address and port * fields. */ if (m->m_len >= sizeof (*cu_head)) { if (getsockname(so->s, (struct sockaddr *)&addr, &addrlen) < 0) return; cu_head = mtod(m, struct cu_header *); cu_head->s_port = addr.sin_port; cu_head->so_addr = our_addr.s_addr; } return; } } struct socket * udp_listen(PNATState pData, u_int port, u_int32_t laddr, u_int lport, int flags) { struct sockaddr_in addr; struct socket *so; socklen_t addrlen = sizeof(struct sockaddr_in); int opt = 1; if ((so = socreate()) == NULL) { free(so); return NULL; } so->s = socket(AF_INET,SOCK_DGRAM,0); so->so_expire = curtime + SO_EXPIRE; insque(pData, so,&udb); addr.sin_family = AF_INET; addr.sin_addr.s_addr = INADDR_ANY; addr.sin_port = port; if (bind(so->s,(struct sockaddr *)&addr, addrlen) < 0) { udp_detach(pData, so); return NULL; } setsockopt(so->s,SOL_SOCKET,SO_REUSEADDR,(char *)&opt,sizeof(int)); /* setsockopt(so->s,SOL_SOCKET,SO_OOBINLINE,(char *)&opt,sizeof(int)); */ getsockname(so->s,(struct sockaddr *)&addr,&addrlen); so->so_fport = addr.sin_port; /* The original check was completely broken, as the commented out * if statement was always true (INADDR_ANY=0). */ /* if (addr.sin_addr.s_addr == 0 || addr.sin_addr.s_addr == loopback_addr.s_addr) */ if (1 == 0) /* always use the else part */ so->so_faddr = alias_addr; else so->so_faddr = addr.sin_addr; so->so_lport = lport; so->so_laddr.s_addr = laddr; if (flags != SS_FACCEPTONCE) so->so_expire = 0; so->so_state = SS_ISFCONNECTED; return so; }