/* $NetBSD: nfsd.c,v 1.36 2001/10/16 01:51:27 itojun Exp $ */ /* * Copyright (c) 1989, 1993, 1994 * The Regents of the University of California. All rights reserved. * * This code is derived from software contributed to Berkeley by * Rick Macklem at The University of Guelph. * * 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. */ #include #ifndef lint __COPYRIGHT("@(#) Copyright (c) 1989, 1993, 1994\n\ The Regents of the University of California. All rights reserved.\n"); #endif /* not lint */ #ifndef lint #if 0 static char sccsid[] = "@(#)nfsd.c 8.9 (Berkeley) 3/29/95"; #else __RCSID("$NetBSD: nfsd.c,v 1.36 2001/10/16 01:51:27 itojun Exp $"); #endif #endif /* not lint */ #include #include #include #include #include #include #include #include #include #include #include #include #ifdef ISO #include #endif #include #include #include #ifdef NFSKERB #include #include #endif #include #include #include #include #include #include #include #include #include #include #include #include /* Global defs */ #ifdef DEBUG #define syslog(e, s) fprintf(stderr,(s)) int debug = 1; #else int debug = 0; #endif struct nfsd_srvargs nsd; #ifdef NFSKERB char lnam[ANAME_SZ]; KTEXT_ST kt; AUTH_DAT kauth; char inst[INST_SZ]; struct nfsrpc_fullblock kin, kout; struct nfsrpc_fullverf kverf; NFSKERBKEY_T kivec; struct timeval ktv; NFSKERBKEYSCHED_T kerb_keysched; #endif int main __P((int, char **)); void nonfs __P((int)); void reapchild __P((int)); void usage __P((void)); /* * Nfs server daemon mostly just a user context for nfssvc() * * 1 - do file descriptor and signal cleanup * 2 - fork the nfsd(s) * 3 - create server socket(s) * 4 - register socket with portmap * * For connectionless protocols, just pass the socket into the kernel via * nfssvc(). * For connection based sockets, loop doing accepts. When you get a new * socket from accept, pass the msgsock into the kernel via nfssvc(). * The arguments are: * -c - support iso cltp clients * -r - reregister with portmapper * -t - support tcp nfs clients * -u - support udp nfs clients * followed by "n" which is the number of nfsds' to fork off */ int main(argc, argv) int argc; char *argv[]; { struct nfsd_args nfsdargs; struct addrinfo *ai_udp, *ai_tcp, *ai_udp6, *ai_tcp6, hints; struct netconfig *nconf_udp, *nconf_tcp, *nconf_udp6, *nconf_tcp6; struct netbuf nb_udp, nb_tcp, nb_udp6, nb_tcp6; struct sockaddr_in inetpeer; struct sockaddr_in6 inet6peer; #ifdef ISO struct sockaddr_iso isoaddr, isopeer; #endif fd_set ready, sockbits; int ch, cltpflag, connect_type_cnt, i, len, maxsock, msgsock; int nfsdcnt, nfssvc_flag, on = 1, reregister, sock, tcpflag, tcpsock; int tcp6sock, ip6flag; int tp4cnt, tp4flag, tpipcnt, tpipflag, udpflag, ecode, s; #ifdef NFSKERB struct group *grp; struct passwd *pwd; struct ucred *cr; struct timeval ktv; int tp4sock, tpipsock; char *cp, **cpp; #endif #define MAXNFSDCNT 20 #define DEFNFSDCNT 4 nfsdcnt = DEFNFSDCNT; cltpflag = reregister = tcpflag = tp4cnt = tp4flag = tpipcnt = 0; tpipflag = udpflag = ip6flag = 0; maxsock = tcpsock = 0; #ifdef ISO #define GETOPT "6cn:rtu" #define USAGE "[-crtu] [-n num_servers]" #else #define GETOPT "6n:rtu" #define USAGE "[-rtu] [-n num_servers]" #endif while ((ch = getopt(argc, argv, GETOPT)) != -1) { switch (ch) { case '6': ip6flag = 1; s = socket(PF_INET6, SOCK_DGRAM, IPPROTO_UDP); if (s < 0 && (errno == EPROTONOSUPPORT || errno == EPFNOSUPPORT || errno == EAFNOSUPPORT)) ip6flag = 0; else close(s); break; case 'n': nfsdcnt = atoi(optarg); if (nfsdcnt < 1 || nfsdcnt > MAXNFSDCNT) { warnx("nfsd count %d; reset to %d", nfsdcnt, DEFNFSDCNT); nfsdcnt = DEFNFSDCNT; } break; case 'r': reregister = 1; break; case 't': tcpflag = 1; break; case 'u': udpflag = 1; break; #ifdef ISO case 'c': cltpflag = 1; break; #ifdef notyet case 'i': tp4cnt = 1; break; case 'p': tpipcnt = 1; break; #endif /* notyet */ #endif /* ISO */ default: case '?': usage(); }; } argv += optind; argc -= optind; /* * XXX * Backward compatibility, trailing number is the count of daemons. */ if (argc > 1) usage(); if (argc == 1) { nfsdcnt = atoi(argv[0]); if (nfsdcnt < 1 || nfsdcnt > MAXNFSDCNT) { warnx("nfsd count %d; reset to %d", nfsdcnt, DEFNFSDCNT); nfsdcnt = DEFNFSDCNT; } } /* * If none of TCP or UDP are specified, default to UDP only. */ if (tcpflag == 0 && udpflag == 0) udpflag = 1; if (debug == 0) { daemon(0, 0); (void)signal(SIGHUP, SIG_IGN); (void)signal(SIGINT, SIG_IGN); (void)signal(SIGQUIT, SIG_IGN); (void)signal(SIGSYS, nonfs); } (void)signal(SIGCHLD, reapchild); if (udpflag) { memset(&hints, 0, sizeof hints); hints.ai_flags = AI_PASSIVE; hints.ai_family = PF_INET; hints.ai_socktype = SOCK_DGRAM; hints.ai_protocol = IPPROTO_UDP; ecode = getaddrinfo(NULL, "nfs", &hints, &ai_udp); if (ecode != 0) { syslog(LOG_ERR, "getaddrinfo udp: %s", gai_strerror(ecode)); exit(1); } nconf_udp = getnetconfigent("udp"); if (nconf_udp == NULL) err(1, "getnetconfigent udp failed"); nb_udp.buf = ai_udp->ai_addr; nb_udp.len = nb_udp.maxlen = ai_udp->ai_addrlen; if (reregister) if (!rpcb_set(RPCPROG_NFS, 2, nconf_udp, &nb_udp)) err(1, "rpcb_set udp failed"); } if (tcpflag) { memset(&hints, 0, sizeof hints); hints.ai_flags = AI_PASSIVE; hints.ai_family = PF_INET; hints.ai_socktype = SOCK_STREAM; hints.ai_protocol = IPPROTO_TCP; ecode = getaddrinfo(NULL, "nfs", &hints, &ai_tcp); if (ecode != 0) { syslog(LOG_ERR, "getaddrinfo udp: %s", gai_strerror(ecode)); exit(1); } nconf_tcp = getnetconfigent("tcp"); if (nconf_tcp == NULL) err(1, "getnetconfigent tcp failed"); nb_tcp.buf = ai_tcp->ai_addr; nb_tcp.len = nb_tcp.maxlen = ai_tcp->ai_addrlen; if (reregister) if (!rpcb_set(RPCPROG_NFS, 2, nconf_tcp, &nb_tcp)) err(1, "rpcb_set tcp failed"); } if (udpflag && ip6flag) { memset(&hints, 0, sizeof hints); hints.ai_flags = AI_PASSIVE; hints.ai_family = PF_INET6; hints.ai_socktype = SOCK_DGRAM; hints.ai_protocol = IPPROTO_UDP; ecode = getaddrinfo(NULL, "nfs", &hints, &ai_udp6); if (ecode != 0) { syslog(LOG_ERR, "getaddrinfo udp: %s", gai_strerror(ecode)); exit(1); } nconf_udp6 = getnetconfigent("udp6"); if (nconf_udp6 == NULL) err(1, "getnetconfigent udp6 failed"); nb_udp6.buf = ai_udp6->ai_addr; nb_udp6.len = nb_udp6.maxlen = ai_udp6->ai_addrlen; if (reregister) if (!rpcb_set(RPCPROG_NFS, 2, nconf_udp6, &nb_udp6)) err(1, "rpcb_set udp6 failed"); } if (tcpflag && ip6flag) { memset(&hints, 0, sizeof hints); hints.ai_flags = AI_PASSIVE; hints.ai_family = PF_INET6; hints.ai_socktype = SOCK_STREAM; hints.ai_protocol = IPPROTO_TCP; ecode = getaddrinfo(NULL, "nfs", &hints, &ai_tcp6); if (ecode != 0) { syslog(LOG_ERR, "getaddrinfo udp: %s", gai_strerror(ecode)); exit(1); } nconf_tcp6 = getnetconfigent("tcp6"); if (nconf_tcp6 == NULL) err(1, "getnetconfigent tcp6 failed"); nb_tcp6.buf = ai_tcp6->ai_addr; nb_tcp6.len = nb_tcp6.maxlen = ai_tcp6->ai_addrlen; if (reregister) if (!rpcb_set(RPCPROG_NFS, 2, nconf_tcp6, &nb_tcp6)) err(1, "rpcb_set tcp6 failed"); } openlog("nfsd", LOG_PID, LOG_DAEMON); for (i = 0; i < nfsdcnt; i++) { switch (fork()) { case -1: syslog(LOG_ERR, "fork: %m"); exit (1); case 0: break; default: continue; } setproctitle("server"); nfssvc_flag = NFSSVC_NFSD; nsd.nsd_nfsd = NULL; #ifdef NFSKERB if (sizeof (struct nfsrpc_fullverf) != RPCX_FULLVERF || sizeof (struct nfsrpc_fullblock) != RPCX_FULLBLOCK) syslog(LOG_ERR, "Yikes NFSKERB structs not packed!"); nsd.nsd_authstr = (u_char *)&kt; nsd.nsd_authlen = sizeof (kt); nsd.nsd_verfstr = (u_char *)&kverf; nsd.nsd_verflen = sizeof (kverf); #endif while (nfssvc(nfssvc_flag, &nsd) < 0) { if (errno != ENEEDAUTH) { syslog(LOG_ERR, "nfssvc: %m"); exit(1); } nfssvc_flag = NFSSVC_NFSD | NFSSVC_AUTHINFAIL; #ifdef NFSKERB /* * Get the Kerberos ticket out of the authenticator * verify it and convert the principal name to a user * name. The user name is then converted to a set of * user credentials via the password and group file. * Finally, decrypt the timestamp and validate it. * For more info see the IETF Draft "Authentication * in ONC RPC". */ kt.length = ntohl(kt.length); if (gettimeofday(&ktv, (struct timezone *)0) == 0 && kt.length > 0 && kt.length <= (RPCAUTH_MAXSIZ - 3 * NFSX_UNSIGNED)) { kin.w1 = NFS_KERBW1(kt); kt.mbz = 0; (void)strcpy(inst, "*"); if (krb_rd_req(&kt, NFS_KERBSRV, inst, nsd.nsd_haddr, &kauth, "") == RD_AP_OK && krb_kntoln(&kauth, lnam) == KSUCCESS && (pwd = getpwnam(lnam)) != NULL) { cr = &nsd.nsd_cr; cr->cr_uid = pwd->pw_uid; cr->cr_groups[0] = pwd->pw_gid; cr->cr_ngroups = 1; setgrent(); while ((grp = getgrent()) != NULL) { if (grp->gr_gid == cr->cr_groups[0]) continue; for (cpp = grp->gr_mem; *cpp != NULL; ++cpp) if (!strcmp(*cpp, lnam)) break; if (*cpp == NULL) continue; cr->cr_groups[cr->cr_ngroups++] = grp->gr_gid; if (cr->cr_ngroups == NGROUPS) break; } endgrent(); /* * Get the timestamp verifier out of * the authenticator and verifier * strings. */ kin.t1 = kverf.t1; kin.t2 = kverf.t2; kin.w2 = kverf.w2; memset((caddr_t)kivec, 0, sizeof(kivec)); memmove((caddr_t)nsd.nsd_key, (caddr_t)kauth.session, sizeof(kauth.session)); /* * Decrypt the timestamp verifier * in CBC mode. */ XXX /* * Validate the timestamp verifier, to * check that the session key is ok. */ nsd.nsd_timestamp.tv_sec = ntohl(kout.t1); nsd.nsd_timestamp.tv_usec = ntohl(kout.t2); nsd.nsd_ttl = ntohl(kout.w1); if ((nsd.nsd_ttl - 1) == ntohl(kout.w2)) nfssvc_flag = NFSSVC_NFSD | NFSSVC_AUTHIN; } } #endif /* NFSKERB */ } exit(0); } /* If we are serving udp, set up the socket. */ if (udpflag) { if ((sock = socket(ai_udp->ai_family, ai_udp->ai_socktype, ai_udp->ai_protocol)) < 0) { syslog(LOG_ERR, "can't create udp socket"); exit(1); } if (bind(sock, ai_udp->ai_addr, ai_udp->ai_addrlen) < 0) { syslog(LOG_ERR, "can't bind udp addr"); exit(1); } if (!rpcb_set(RPCPROG_NFS, 2, nconf_udp, &nb_udp) || !rpcb_set(RPCPROG_NFS, 3, nconf_udp, &nb_udp)) { syslog(LOG_ERR, "can't register with udp portmap"); exit(1); } nfsdargs.sock = sock; nfsdargs.name = NULL; nfsdargs.namelen = 0; if (nfssvc(NFSSVC_ADDSOCK, &nfsdargs) < 0) { syslog(LOG_ERR, "can't add UDP socket"); exit(1); } (void)close(sock); } if (udpflag &&ip6flag) { if ((sock = socket(ai_udp6->ai_family, ai_udp6->ai_socktype, ai_udp6->ai_protocol)) < 0) { syslog(LOG_ERR, "can't create udp socket"); exit(1); } if (setsockopt(sock, IPPROTO_IPV6, IPV6_V6ONLY, &on, sizeof on) < 0) { syslog(LOG_ERR, "can't set v6-only binding for udp6 " "socket: %m"); exit(1); } if (bind(sock, ai_udp6->ai_addr, ai_udp6->ai_addrlen) < 0) { syslog(LOG_ERR, "can't bind udp addr"); exit(1); } if (!rpcb_set(RPCPROG_NFS, 2, nconf_udp6, &nb_udp6) || !rpcb_set(RPCPROG_NFS, 3, nconf_udp6, &nb_udp6)) { syslog(LOG_ERR, "can't register with udp portmap"); exit(1); } nfsdargs.sock = sock; nfsdargs.name = NULL; nfsdargs.namelen = 0; if (nfssvc(NFSSVC_ADDSOCK, &nfsdargs) < 0) { syslog(LOG_ERR, "can't add UDP6 socket"); exit(1); } (void)close(sock); } #ifdef ISO /* If we are serving cltp, set up the socket. */ if (cltpflag) { if ((sock = socket(AF_ISO, SOCK_DGRAM, 0)) < 0) { syslog(LOG_ERR, "can't create cltp socket"); exit(1); } memset(&isoaddr, 0, sizeof(isoaddr)); isoaddr.siso_family = AF_ISO; isoaddr.siso_tlen = 2; cp = TSEL(&isoaddr); *cp++ = (NFS_PORT >> 8); *cp = (NFS_PORT & 0xff); isoaddr.siso_len = sizeof(isoaddr); if (bind(sock, (struct sockaddr *)&isoaddr, sizeof(isoaddr)) < 0) { syslog(LOG_ERR, "can't bind cltp addr"); exit(1); } #ifdef notyet /* * XXX * Someday this should probably use "rpcbind", the son of * portmap. */ if (!pmap_set(RPCPROG_NFS, NFS_VER2, IPPROTO_UDP, NFS_PORT)) { syslog(LOG_ERR, "can't register with udp portmap"); exit(1); } #endif /* notyet */ nfsdargs.sock = sock; nfsdargs.name = NULL; nfsdargs.namelen = 0; if (nfssvc(NFSSVC_ADDSOCK, &nfsdargs) < 0) { syslog(LOG_ERR, "can't add UDP socket"); exit(1); } close(sock); } #endif /* ISO */ /* Now set up the master server socket waiting for tcp connections. */ on = 1; FD_ZERO(&sockbits); connect_type_cnt = 0; if (tcpflag) { if ((tcpsock = socket(ai_tcp->ai_family, ai_tcp->ai_socktype, ai_tcp->ai_protocol)) < 0) { syslog(LOG_ERR, "can't create tcp socket"); exit(1); } if (setsockopt(tcpsock, SOL_SOCKET, SO_REUSEADDR, (char *)&on, sizeof(on)) < 0) syslog(LOG_ERR, "setsockopt SO_REUSEADDR: %m"); if (bind(tcpsock, ai_tcp->ai_addr, ai_tcp->ai_addrlen) < 0) { syslog(LOG_ERR, "can't bind tcp addr"); exit(1); } if (listen(tcpsock, 5) < 0) { syslog(LOG_ERR, "listen failed"); exit(1); } if (!rpcb_set(RPCPROG_NFS, 2, nconf_tcp, &nb_tcp) || !rpcb_set(RPCPROG_NFS, 3, nconf_tcp, &nb_tcp)) { syslog(LOG_ERR, "can't register tcp with rpcbind"); exit(1); } FD_SET(tcpsock, &sockbits); maxsock = tcpsock; connect_type_cnt++; } if (tcpflag && ip6flag) { if ((tcp6sock = socket(ai_tcp6->ai_family, ai_tcp6->ai_socktype, ai_tcp6->ai_protocol)) < 0) { syslog(LOG_ERR, "can't create tcp socket"); exit(1); } if (setsockopt(tcp6sock, SOL_SOCKET, SO_REUSEADDR, (char *)&on, sizeof(on)) < 0) syslog(LOG_ERR, "setsockopt SO_REUSEADDR: %m"); if (setsockopt(tcp6sock, IPPROTO_IPV6, IPV6_V6ONLY, &on, sizeof on) < 0) { syslog(LOG_ERR, "can't set v6-only binding for tcp6 " "socket: %m"); exit(1); } if (bind(tcp6sock, ai_tcp6->ai_addr, ai_tcp6->ai_addrlen) < 0) { syslog(LOG_ERR, "can't bind tcp6 addr"); exit(1); } if (listen(tcp6sock, 5) < 0) { syslog(LOG_ERR, "listen failed"); exit(1); } if (!rpcb_set(RPCPROG_NFS, 2, nconf_tcp6, &nb_tcp6) || !rpcb_set(RPCPROG_NFS, 3, nconf_tcp6, &nb_tcp6)) { syslog(LOG_ERR, "can't register tcp6 with rpcbind"); exit(1); } FD_SET(tcp6sock, &sockbits); maxsock = tcp6sock; connect_type_cnt++; } #ifdef notyet /* Now set up the master server socket waiting for tp4 connections. */ if (tp4flag) { if ((tp4sock = socket(AF_ISO, SOCK_SEQPACKET, 0)) < 0) { syslog(LOG_ERR, "can't create tp4 socket"); exit(1); } if (setsockopt(tp4sock, SOL_SOCKET, SO_REUSEADDR, (char *)&on, sizeof(on)) < 0) syslog(LOG_ERR, "setsockopt SO_REUSEADDR: %m"); memset(&isoaddr, 0, sizeof(isoaddr)); isoaddr.siso_family = AF_ISO; isoaddr.siso_tlen = 2; cp = TSEL(&isoaddr); *cp++ = (NFS_PORT >> 8); *cp = (NFS_PORT & 0xff); isoaddr.siso_len = sizeof(isoaddr); if (bind(tp4sock, (struct sockaddr *)&isoaddr, sizeof(isoaddr)) < 0) { syslog(LOG_ERR, "can't bind tp4 addr"); exit(1); } if (listen(tp4sock, 5) < 0) { syslog(LOG_ERR, "listen failed"); exit(1); } /* * XXX * Someday this should probably use "rpcbind", the son of * portmap. */ if (!pmap_set(RPCPROG_NFS, NFS_VER2, IPPROTO_TCP, NFS_PORT)) { syslog(LOG_ERR, "can't register tcp with portmap"); exit(1); } FD_SET(tp4sock, &sockbits); maxsock = tp4sock; connect_type_cnt++; } /* Now set up the master server socket waiting for tpip connections. */ if (tpipflag) { if ((tpipsock = socket(AF_INET, SOCK_SEQPACKET, 0)) < 0) { syslog(LOG_ERR, "can't create tpip socket"); exit(1); } if (setsockopt(tpipsock, SOL_SOCKET, SO_REUSEADDR, (char *)&on, sizeof(on)) < 0) syslog(LOG_ERR, "setsockopt SO_REUSEADDR: %m"); inetaddr.sin_family = AF_INET; inetaddr.sin_addr.s_addr = INADDR_ANY; inetaddr.sin_port = htons(NFS_PORT); inetaddr.sin_len = sizeof(inetaddr); memset(inetaddr.sin_zero, 0, sizeof(inetaddr.sin_zero)); if (bind(tpipsock, (struct sockaddr *)&inetaddr, sizeof (inetaddr)) < 0) { syslog(LOG_ERR, "can't bind tcp addr"); exit(1); } if (listen(tpipsock, 5) < 0) { syslog(LOG_ERR, "listen failed"); exit(1); } /* * XXX * Someday this should probably use "rpcbind", the son of * portmap. */ if (!pmap_set(RPCPROG_NFS, NFS_VER2, IPPROTO_TCP, NFS_PORT)) { syslog(LOG_ERR, "can't register tcp with portmap"); exit(1); } FD_SET(tpipsock, &sockbits); maxsock = tpipsock; connect_type_cnt++; } #endif /* notyet */ if (connect_type_cnt == 0) exit(0); setproctitle("master"); /* * Loop forever accepting connections and passing the sockets * into the kernel for the mounts. */ for (;;) { ready = sockbits; if (connect_type_cnt > 1) { if (select(maxsock + 1, &ready, NULL, NULL, NULL) < 1) { syslog(LOG_ERR, "select failed: %m"); exit(1); } } if (tcpflag && FD_ISSET(tcpsock, &ready)) { len = sizeof(inetpeer); if ((msgsock = accept(tcpsock, (struct sockaddr *)&inetpeer, &len)) < 0) { syslog(LOG_ERR, "accept failed: %m"); exit(1); } memset(inetpeer.sin_zero, 0, sizeof(inetpeer.sin_zero)); if (setsockopt(msgsock, SOL_SOCKET, SO_KEEPALIVE, (char *)&on, sizeof(on)) < 0) syslog(LOG_ERR, "setsockopt SO_KEEPALIVE: %m"); nfsdargs.sock = msgsock; nfsdargs.name = (caddr_t)&inetpeer; nfsdargs.namelen = sizeof(inetpeer); nfssvc(NFSSVC_ADDSOCK, &nfsdargs); (void)close(msgsock); } if (tcpflag && ip6flag && FD_ISSET(tcp6sock, &ready)) { len = sizeof(inet6peer); if ((msgsock = accept(tcp6sock, (struct sockaddr *)&inet6peer, &len)) < 0) { syslog(LOG_ERR, "accept failed: %m"); exit(1); } if (setsockopt(msgsock, SOL_SOCKET, SO_KEEPALIVE, (char *)&on, sizeof(on)) < 0) syslog(LOG_ERR, "setsockopt SO_KEEPALIVE: %m"); nfsdargs.sock = msgsock; nfsdargs.name = (caddr_t)&inet6peer; nfsdargs.namelen = sizeof(inet6peer); nfssvc(NFSSVC_ADDSOCK, &nfsdargs); (void)close(msgsock); } #ifdef notyet if (tp4flag && FD_ISSET(tp4sock, &ready)) { len = sizeof(isopeer); if ((msgsock = accept(tp4sock, (struct sockaddr *)&isopeer, &len)) < 0) { syslog(LOG_ERR, "accept failed: %m"); exit(1); } if (setsockopt(msgsock, SOL_SOCKET, SO_KEEPALIVE, (char *)&on, sizeof(on)) < 0) syslog(LOG_ERR, "setsockopt SO_KEEPALIVE: %m"); nfsdargs.sock = msgsock; nfsdargs.name = (caddr_t)&isopeer; nfsdargs.namelen = len; nfssvc(NFSSVC_ADDSOCK, &nfsdargs); (void)close(msgsock); } if (tpipflag && FD_ISSET(tpipsock, &ready)) { len = sizeof(inetpeer); if ((msgsock = accept(tpipsock, (struct sockaddr *)&inetpeer, &len)) < 0) { syslog(LOG_ERR, "Accept failed: %m"); exit(1); } if (setsockopt(msgsock, SOL_SOCKET, SO_KEEPALIVE, (char *)&on, sizeof(on)) < 0) syslog(LOG_ERR, "setsockopt SO_KEEPALIVE: %m"); nfsdargs.sock = msgsock; nfsdargs.name = (caddr_t)&inetpeer; nfsdargs.namelen = len; nfssvc(NFSSVC_ADDSOCK, &nfsdargs); (void)close(msgsock); } #endif /* notyet */ } } void usage() { (void)fprintf(stderr, "usage: nfsd %s\n", USAGE); exit(1); } void nonfs(signo) int signo; { syslog(LOG_ERR, "missing system call: NFS not available."); } void reapchild(signo) int signo; { while (wait3(NULL, WNOHANG, NULL) > 0); }