/* $OpenBSD: mount_nfs.c,v 1.29 2002/06/09 08:13:08 todd Exp $ */ /* $NetBSD: mount_nfs.c,v 1.12.4.1 1996/05/25 22:48:05 fvdl Exp $ */ /* * Copyright (c) 1992, 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. */ #ifndef lint static char copyright[] = "@(#) Copyright (c) 1992, 1993, 1994\n\ The Regents of the University of California. All rights reserved.\n"; #endif /* not lint */ #ifndef lint #if 0 static char sccsid[] = "@(#)mount_nfs.c 8.11 (Berkeley) 5/4/95"; #else static char rcsid[] = "$NetBSD: mount_nfs.c,v 1.12.4.1 1996/05/25 22:48:05 fvdl Exp $"; #endif #endif /* not lint */ #include #include #include #include #include #include #include #include #include #ifdef ISO #include #endif #include #include #define _KERNEL #include #undef _KERNEL #include #include #include #include #include #include #include #include #include #include #include #include "mntopts.h" #define ALTF_BG 0x1 #define ALTF_NOCONN 0x2 #define ALTF_DUMBTIMR 0x4 #define ALTF_INTR 0x8 #define ALTF_NFSV3 0x20 #define ALTF_RDIRPLUS 0x40 #define ALTF_MNTUDP 0x80 #define ALTF_RESVPORT 0x100 #define ALTF_SEQPACKET 0x200 #define ALTF_SOFT 0x800 #define ALTF_TCP 0x1000 #define ALTF_PORT 0x2000 #define ALTF_NFSV2 0x4000 #define ALTF_NOAC 0x8000 const struct mntopt mopts[] = { MOPT_STDOPTS, MOPT_FORCE, MOPT_UPDATE, { "bg", 0, ALTF_BG, 1 }, { "conn", 1, ALTF_NOCONN, 1 }, { "dumbtimer", 0, ALTF_DUMBTIMR, 1 }, { "intr", 0, ALTF_INTR, 1 }, { "nfsv3", 0, ALTF_NFSV3, 1 }, { "rdirplus", 0, ALTF_RDIRPLUS, 1 }, { "mntudp", 0, ALTF_MNTUDP, 1 }, { "resvport", 0, ALTF_RESVPORT, 1 }, #ifdef ISO { "seqpacket", 0, ALTF_SEQPACKET, 1 }, #endif { "soft", 0, ALTF_SOFT, 1 }, { "tcp", 0, ALTF_TCP, 1 }, { "port", 0, ALTF_PORT, 1 }, { "nfsv2", 0, ALTF_NFSV2, 1 }, { "noac", 0, ALTF_NOAC, 1 }, { NULL } }; struct nfs_args nfsdefargs = { NFS_ARGSVERSION, NULL, sizeof (struct sockaddr_in), SOCK_DGRAM, 0, NULL, 0, NFSMNT_NFSV3, NFS_WSIZE, NFS_RSIZE, NFS_READDIRSIZE, 10, NFS_RETRANS, NFS_MAXGRPS, NFS_DEFRAHEAD, 0, 0, NULL, 0, 0, 0, 0 }; struct nfhret { u_long stat; long vers; long auth; long fhsize; u_char nfh[NFSX_V3FHMAX]; }; #define DEF_RETRY 10000 #define BGRND 1 #define ISBGRND 2 int retrycnt; int opflags = 0; int nfsproto = IPPROTO_UDP; int mnttcp_ok = 1; u_short port_no = 0; int force2 = 0; int force3 = 0; int getnfsargs(char *, struct nfs_args *); #ifdef ISO struct iso_addr *iso_addr(const char *); #endif void set_rpc_maxgrouplist(int); __dead void usage(void); int xdr_dir(XDR *, char *); int xdr_fh(XDR *, struct nfhret *); int main(argc, argv) int argc; char *argv[]; { int c; struct nfs_args *nfsargsp; struct nfs_args nfsargs; int mntflags, altflags, num; char *name, *p, *spec; retrycnt = DEF_RETRY; mntflags = 0; altflags = 0; nfsargs = nfsdefargs; nfsargsp = &nfsargs; while ((c = getopt(argc, argv, "23a:bcdD:g:I:iL:lo:PpqR:r:sTt:w:x:U")) != -1) switch (c) { case '3': if (force2) errx(1, "-2 and -3 are mutually exclusive"); force3 = 1; break; case '2': if (force3) errx(1, "-2 and -3 are mutually exclusive"); force2 = 1; nfsargsp->flags &= ~NFSMNT_NFSV3; break; case 'a': num = strtol(optarg, &p, 10); if (*p || num < 0) errx(1, "illegal -a value -- %s", optarg); nfsargsp->readahead = num; nfsargsp->flags |= NFSMNT_READAHEAD; break; case 'b': opflags |= BGRND; break; case 'c': nfsargsp->flags |= NFSMNT_NOCONN; break; case 'D': num = strtol(optarg, &p, 10); if (*p || num <= 0) errx(1, "illegal -D value -- %s", optarg); nfsargsp->deadthresh = num; nfsargsp->flags |= NFSMNT_DEADTHRESH; break; case 'd': nfsargsp->flags |= NFSMNT_DUMBTIMR; break; #if 0 /* XXXX */ case 'g': num = strtol(optarg, &p, 10); if (*p || num <= 0) errx(1, "illegal -g value -- %s", optarg); set_rpc_maxgrouplist(num); nfsargsp->maxgrouplist = num; nfsargsp->flags |= NFSMNT_MAXGRPS; break; #endif case 'I': num = strtol(optarg, &p, 10); if (*p || num <= 0) errx(1, "illegal -I value -- %s", optarg); nfsargsp->readdirsize = num; nfsargsp->flags |= NFSMNT_READDIRSIZE; break; case 'i': nfsargsp->flags |= NFSMNT_INT; break; case 'L': num = strtol(optarg, &p, 10); if (*p || num < 2) errx(1, "illegal -L value -- %s", optarg); nfsargsp->leaseterm = num; nfsargsp->flags |= NFSMNT_LEASETERM; break; case 'l': nfsargsp->flags |= NFSMNT_RDIRPLUS; break; case 'o': getmntopts(optarg, mopts, &mntflags, &altflags); if (altflags & ALTF_BG) opflags |= BGRND; if (altflags & ALTF_NOCONN) nfsargsp->flags |= NFSMNT_NOCONN; if (altflags & ALTF_DUMBTIMR) nfsargsp->flags |= NFSMNT_DUMBTIMR; if (altflags & ALTF_INTR) nfsargsp->flags |= NFSMNT_INT; if (altflags & ALTF_NFSV3) { if (force2) errx(1,"conflicting version options"); force3 = 1; } if (altflags & ALTF_NFSV2) { if (force3) errx(1,"conflicting version options"); force2 = 1; nfsargsp->flags &= ~NFSMNT_NFSV3; } if (altflags & ALTF_RDIRPLUS) nfsargsp->flags |= NFSMNT_RDIRPLUS; if (altflags & ALTF_MNTUDP) mnttcp_ok = 0; if (altflags & ALTF_RESVPORT) nfsargsp->flags |= NFSMNT_RESVPORT; #ifdef ISO if (altflags & ALTF_SEQPACKET) nfsargsp->sotype = SOCK_SEQPACKET; #endif if (altflags & ALTF_SOFT) nfsargsp->flags |= NFSMNT_SOFT; if (altflags & ALTF_TCP) { nfsargsp->sotype = SOCK_STREAM; nfsproto = IPPROTO_TCP; } if (altflags & ALTF_PORT) port_no = atoi(strstr(optarg, "port=") + 5); if (altflags & ALTF_NOAC) { nfsargsp->flags |= (NFSMNT_ACREGMIN | NFSMNT_ACREGMAX | NFSMNT_ACDIRMIN | NFSMNT_ACDIRMAX); nfsargsp->acregmin = 0; nfsargsp->acregmax = 0; nfsargsp->acdirmin = 0; nfsargsp->acdirmax = 0; } altflags = 0; break; case 'P': nfsargsp->flags |= NFSMNT_RESVPORT; break; #ifdef ISO case 'p': nfsargsp->sotype = SOCK_SEQPACKET; break; #endif case 'R': num = strtol(optarg, &p, 10); if (*p || num <= 0) errx(1, "illegal -R value -- %s", optarg); retrycnt = num; break; case 'r': num = strtol(optarg, &p, 10); if (*p || num <= 0) errx(1, "illegal -r value -- %s", optarg); nfsargsp->rsize = num; nfsargsp->flags |= NFSMNT_RSIZE; break; case 's': nfsargsp->flags |= NFSMNT_SOFT; break; case 'T': nfsargsp->sotype = SOCK_STREAM; nfsproto = IPPROTO_TCP; break; case 't': num = strtol(optarg, &p, 10); if (*p || num <= 0) errx(1, "illegal -t value -- %s", optarg); nfsargsp->timeo = num; nfsargsp->flags |= NFSMNT_TIMEO; break; case 'w': num = strtol(optarg, &p, 10); if (*p || num <= 0) errx(1, "illegal -w value -- %s", optarg); nfsargsp->wsize = num; nfsargsp->flags |= NFSMNT_WSIZE; break; case 'x': num = strtol(optarg, &p, 10); if (*p || num <= 0) errx(1, "illegal -x value -- %s", optarg); nfsargsp->retrans = num; nfsargsp->flags |= NFSMNT_RETRANS; break; case 'U': mnttcp_ok = 0; break; default: usage(); break; } argc -= optind; argv += optind; if (argc != 2) usage(); spec = *argv++; name = *argv; if (!getnfsargs(spec, nfsargsp)) exit(1); if (mount(MOUNT_NFS, name, mntflags, nfsargsp)) { if (errno == EOPNOTSUPP) errx(1, "%s: Filesystem not supported by kernel", name); else err(1, "%s", name); } exit(0); } int getnfsargs(spec, nfsargsp) char *spec; struct nfs_args *nfsargsp; { CLIENT *clp; struct hostent *hp; static struct sockaddr_in saddr; #ifdef ISO static struct sockaddr_iso isoaddr; struct iso_addr *isop; int isoflag = 0; #endif struct timeval pertry, try; enum clnt_stat clnt_stat; int so = RPC_ANYSOCK, i, nfsvers, mntvers, orgcnt; char *hostp, *delimp; u_short tport; static struct nfhret nfhret; static char nam[MNAMELEN + 1]; strlcpy(nam, spec, sizeof(nam)); if ((delimp = strchr(spec, '@')) != NULL) { hostp = delimp + 1; } else if ((delimp = strchr(spec, ':')) != NULL) { hostp = spec; spec = delimp + 1; } else { warnx("no : or @ spec"); return (0); } *delimp = '\0'; /* * DUMB!! Until the mount protocol works on iso transport, we must * supply both an iso and an inet address for the host. */ #ifdef ISO if (!strncmp(hostp, "iso=", 4)) { u_short isoport; hostp += 4; isoflag++; if ((delimp = strchr(hostp, '+')) == NULL) { warnx("no iso+inet address"); return (0); } *delimp = '\0'; if ((isop = iso_addr(hostp)) == NULL) { warnx("bad ISO address"); return (0); } memset(&isoaddr, 0, sizeof (isoaddr)); memcpy(&isoaddr.siso_addr, isop, sizeof (struct iso_addr)); isoaddr.siso_len = sizeof (isoaddr); isoaddr.siso_family = AF_ISO; isoaddr.siso_tlen = 2; isoport = htons(NFS_PORT); memcpy(TSEL(&isoaddr), &isoport, isoaddr.siso_tlen); hostp = delimp + 1; } #endif /* ISO */ /* * Handle an internet host address */ if (inet_aton(hostp, &saddr.sin_addr) == 0) { hp = gethostbyname(hostp); if (hp == NULL) { warnx("can't resolve address for host %s", hostp); return (0); } memcpy(&saddr.sin_addr, hp->h_addr, hp->h_length); } if (force2) { nfsvers = NFS_VER2; mntvers = RPCMNT_VER1; } else { nfsvers = NFS_VER3; mntvers = RPCMNT_VER3; } orgcnt = retrycnt; tryagain: nfhret.stat = EACCES; /* Mark not yet successful */ while (retrycnt > 0) { saddr.sin_family = AF_INET; saddr.sin_port = htons(PMAPPORT); if ((tport = port_no ? port_no : pmap_getport(&saddr, RPCPROG_NFS, nfsvers, nfsargsp->sotype == SOCK_STREAM ? IPPROTO_TCP : IPPROTO_UDP)) == 0) { if ((opflags & ISBGRND) == 0) clnt_pcreateerror("NFS Portmap"); } else { saddr.sin_port = 0; pertry.tv_sec = 10; pertry.tv_usec = 0; if (mnttcp_ok && nfsargsp->sotype == SOCK_STREAM) clp = clnttcp_create(&saddr, RPCPROG_MNT, mntvers, &so, 0, 0); else clp = clntudp_create(&saddr, RPCPROG_MNT, mntvers, pertry, &so); if (clp == NULL) { if ((opflags & ISBGRND) == 0) clnt_pcreateerror("Cannot MNT PRC"); } else { clp->cl_auth = authunix_create_default(); try.tv_sec = 10; try.tv_usec = 0; nfhret.auth = RPCAUTH_UNIX; nfhret.vers = mntvers; clnt_stat = clnt_call(clp, RPCMNT_MOUNT, xdr_dir, spec, xdr_fh, &nfhret, try); if (clnt_stat != RPC_SUCCESS) { if (clnt_stat == RPC_PROGVERSMISMATCH) { if (nfsvers == NFS_VER3 && !force3) { retrycnt = orgcnt; nfsvers = NFS_VER2; mntvers = RPCMNT_VER1; nfsargsp->flags &= ~NFSMNT_NFSV3; goto tryagain; } else { fprintf(stderr, "%s", clnt_sperror(clp, "MNT RPC")); } } if ((opflags & ISBGRND) == 0) warnx("%s", clnt_sperror(clp, "bad MNT RPC")); } else { auth_destroy(clp->cl_auth); clnt_destroy(clp); retrycnt = 0; } } } if (--retrycnt > 0) { if (opflags & BGRND) { opflags &= ~BGRND; if ((i = fork())) { if (i == -1) err(1, "nqnfs 2"); exit(0); } (void) setsid(); (void) close(STDIN_FILENO); (void) close(STDOUT_FILENO); (void) close(STDERR_FILENO); (void) chdir("/"); opflags |= ISBGRND; } sleep(60); } } if (nfhret.stat) { if (opflags & ISBGRND) exit(1); errno = nfhret.stat; warnx("can't access %s: %s", spec, strerror(nfhret.stat)); return (0); } saddr.sin_port = htons(tport); #ifdef ISO if (isoflag) { nfsargsp->addr = (struct sockaddr *) &isoaddr; nfsargsp->addrlen = sizeof (isoaddr); } else #endif /* ISO */ { nfsargsp->addr = (struct sockaddr *) &saddr; nfsargsp->addrlen = sizeof (saddr); } nfsargsp->fh = nfhret.nfh; nfsargsp->fhsize = nfhret.fhsize; nfsargsp->hostname = nam; return (1); } /* * xdr routines for mount rpc's */ int xdr_dir(xdrsp, dirp) XDR *xdrsp; char *dirp; { return (xdr_string(xdrsp, &dirp, RPCMNT_PATHLEN)); } int xdr_fh(xdrsp, np) XDR *xdrsp; struct nfhret *np; { int i; long auth, authcnt, authfnd = 0; if (!xdr_u_long(xdrsp, &np->stat)) return (0); if (np->stat) return (1); switch (np->vers) { case 1: np->fhsize = NFSX_V2FH; return (xdr_opaque(xdrsp, (caddr_t)np->nfh, NFSX_V2FH)); case 3: if (!xdr_long(xdrsp, &np->fhsize)) return (0); if (np->fhsize <= 0 || np->fhsize > NFSX_V3FHMAX) return (0); if (!xdr_opaque(xdrsp, (caddr_t)np->nfh, np->fhsize)) return (0); if (!xdr_long(xdrsp, &authcnt)) return (0); for (i = 0; i < authcnt; i++) { if (!xdr_long(xdrsp, &auth)) return (0); if (auth == np->auth) authfnd++; } /* * Some servers, such as DEC's OSF/1 return a nil authenticator * list to indicate RPCAUTH_UNIX. */ if (!authfnd && (authcnt > 0 || np->auth != RPCAUTH_UNIX)) np->stat = EAUTH; return (1); }; return (0); } __dead void usage() { (void)fprintf(stderr, "usage: mount_nfs %s\n%s\n%s\n%s\n", "[-23PTUbcdilqs] [-a maxreadahead] [-D deadthresh]", "\t[-I readdirsize] [-g maxgroups] [-L leaseterm] [-o options]", "\t[-R retrycnt] [-r readsize] [-t timeout] [-w writesize] [-x retrans]", "\trhost:path node"); exit(1); }