/* * Copyright (c) 1990, 1991, 1992, 1993, 1994 * 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: (1) source code distributions * retain the above copyright notice and this paragraph in its entirety, (2) * distributions including binary code include the above copyright notice and * this paragraph in its entirety in the documentation or other materials * provided with the distribution, and (3) all advertising materials mentioning * features or use of this software display the following acknowledgement: * ``This product includes software developed by the University of California, * Lawrence Berkeley Laboratory and its contributors.'' 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 ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED * WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. */ #ifndef lint static char rcsid[] = "@(#)$Header: pcap-pf.c,v 1.32 94/06/10 17:41:01 mccanne Exp $ (LBL)"; #endif /* * packet filter subroutines for tcpdump * Extraction/creation by Jeffrey Mogul, DECWRL * * Extracted from tcpdump.c. */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include "pcap-int.h" /* * BUFSPACE is the size in bytes of the packet read buffer. Most tcpdump * applications aren't going to need more than 200 bytes of packet header * and the read shouldn't return more packets than packetfilter's internal * queue limit (bounded at 256). */ #define BUFSPACE (200*256) int pcap_read(pcap_t *pc, int cnt, pcap_handler callback, u_char *user) { u_char *p; struct bpf_insn *fcode; int cc; register u_char *bp; int buflen, inc; struct enstamp stamp; int n; fcode = pc->md.use_bpf ? 0 : pc->fcode.bf_insns; again: cc = pc->cc; if (cc == 0) { cc = read(pc->fd, (char *)pc->buffer, pc->bufsize); if (cc < 0) { if (errno == EWOULDBLOCK) return (0); if (errno == EINVAL && (long)(tell(pc->fd) + pc->bufsize) < 0) { /* * Due to a kernel bug, after 2^31 bytes, * the kernel file offset overflows and * read fails with EINVAL. The lseek() * to 0 will fix things. */ (void)lseek(pc->fd, 0L, 0); goto again; } sprintf(pc->errbuf, "pf read: %s", pcap_strerror(errno)); return (-1); } bp = pc->buffer; } else bp = pc->bp; /* * Loop through each packet. */ n = 0; while (cc > 0) { /* avoid alignment issues here */ bcopy((char *)bp, (char *)&stamp, sizeof(stamp)); if (stamp.ens_stamplen != sizeof(stamp)) /* buffer is garbage, treat it as poison */ break; p = bp + stamp.ens_stamplen; buflen = stamp.ens_count; if (buflen > pc->snapshot) buflen = pc->snapshot; /* * Short-circuit evaluation: if using BPF filter * in kernel, no need to do it now. */ if (fcode == 0 || bpf_filter(fcode, p, stamp.ens_count, buflen)) { struct pcap_pkthdr h; pc->md.TotAccepted++; h.ts = stamp.ens_tstamp; h.len = stamp.ens_count; h.caplen = buflen; (*callback)(user, &h, p); if (++n >= cnt && cnt > 0) { inc = ENALIGN(buflen + stamp.ens_stamplen); cc -= inc; bp += inc; pc->cc = cc; pc->bp = bp; return (n); } } pc->md.TotPkts++; pc->md.TotDrops += stamp.ens_dropped; pc->md.TotMissed = stamp.ens_ifoverflows; if (pc->md.OrigMissed < 0) pc->md.OrigMissed = pc->md.TotMissed; inc = ENALIGN(buflen + stamp.ens_stamplen); cc -= inc; bp += inc; } pc->cc = 0; return (n); } int pcap_stats(pcap_t *p, struct pcap_stat *ps) { ps->ps_recv = p->md.TotAccepted; ps->ps_drop = p->md.TotDrops; ps->ps_ifdrop = p->md.TotMissed - p->md.OrigMissed; return (0); } pcap_t * pcap_open_live(char *device, int snaplen, int promisc, int to_ms, char *ebuf) { pcap_t *p; short enmode; int backlog = -1; /* request the most */ struct enfilter Filter; struct endevp devparams; p = (pcap_t *)malloc(sizeof(*p)); if (p == 0) { strcpy(ebuf, "no swap"); return (0); } bzero(p, sizeof(*p)); p->fd = pfopen(device, 0); if (p->fd < 0) { sprintf(ebuf, "pf open: %s: %s\n\ your system may not be properly configured; see \"man packetfilter(4)\"\n", device, pcap_strerror(errno)); goto bad; } p->md.OrigMissed = -1; enmode = ENTSTAMP|ENBATCH|ENNONEXCL; if (promisc) enmode |= ENPROMISC; if (ioctl(p->fd, EIOCMBIS, (caddr_t)&enmode) < 0) { sprintf(ebuf, "EIOCMBIS: %s", pcap_strerror(errno)); goto bad; } #ifdef ENCOPYALL /* Try to set COPYALL mode so that we see packets to ourself */ enmode = ENCOPYALL; (void)ioctl(p->fd, EIOCMBIS, (caddr_t)&enmode);/* OK if this fails */ #endif /* set the backlog */ if (ioctl(p->fd, EIOCSETW, (caddr_t)&backlog) < 0) { sprintf(ebuf, "EIOCSETW: %s", pcap_strerror(errno)); goto bad; } /* set truncation */ if (ioctl(p->fd, EIOCTRUNCATE, (caddr_t)&snaplen) < 0) { sprintf(ebuf, "EIOCTRUNCATE: %s", pcap_strerror(errno)); goto bad; } p->snapshot = snaplen; /* accept all packets */ Filter.enf_Priority = 37; /* anything > 2 */ Filter.enf_FilterLen = 0; /* means "always true" */ if (ioctl(p->fd, EIOCSETF, (caddr_t)&Filter) < 0) { sprintf(ebuf, "EIOCSETF: %s", pcap_strerror(errno)); goto bad; } /* discover interface type */ if (ioctl(p->fd, EIOCDEVP, (caddr_t)&devparams) < 0) { sprintf(ebuf, "EIOCDEVP: %s", pcap_strerror(errno)); goto bad; } /* HACK: to compile prior to Ultrix 4.2 */ #ifndef ENDT_FDDI #define ENDT_FDDI 4 #endif switch (devparams.end_dev_type) { case ENDT_10MB: p->linktype = DLT_EN10MB; break; case ENDT_FDDI: p->linktype = DLT_FDDI; break; default: /* * XXX * Currently, the Ultrix packet filter supports only * Ethernet and FDDI. Eventually, support for SLIP and PPP * (and possibly others: T1?) should be added. */ #ifdef notdef warning( "Packet filter data-link type %d unknown, assuming Ethernet", devparams.end_dev_type); #endif p->linktype = DLT_EN10MB; break; } if (to_ms != 0) { struct timeval timeout; timeout.tv_sec = to_ms / 1000; timeout.tv_usec = (to_ms * 1000) % 1000000; if (ioctl(p->fd, EIOCSRTIMEOUT, (caddr_t)&timeout) < 0) { sprintf(ebuf, "EIOCSRTIMEOUT: %s", pcap_strerror(errno)); goto bad; } } p->bufsize = BUFSPACE; p->buffer = (u_char*)malloc(p->bufsize); return (p); bad: free(p); return (0); } int pcap_setfilter(pcap_t *p, struct bpf_program *fp) { /* * See if BIOCSETF works. If it does, the kernel supports * BPF-style filters, and we do not need to do post-filtering. */ p->md.use_bpf = (ioctl(p->fd, BIOCSETF, (caddr_t)fp) >= 0); if (p->md.use_bpf) { struct bpf_version bv; if (ioctl(p->fd, BIOCVERSION, (caddr_t)&bv) < 0) { sprintf(p->errbuf, "BIOCVERSION: %s", pcap_strerror(errno)); return (-1); } else if (bv.bv_major != BPF_MAJOR_VERSION || bv.bv_minor < BPF_MINOR_VERSION) { fprintf(stderr, "requires bpf language %d.%d or higher; kernel is %d.%d", BPF_MAJOR_VERSION, BPF_MINOR_VERSION, bv.bv_major, bv.bv_minor); /* don't give up, just be inefficient */ p->md.use_bpf = 0; } } else p->fcode = *fp; /*XXX this goes in tcpdump*/ if (p->md.use_bpf) fprintf(stderr, "tcpdump: Using kernel BPF filter\n"); else fprintf(stderr, "tcpdump: Filtering in user process\n"); return (0); }