NetBSD/sys/netinet/fil.c

780 lines
18 KiB
C

/* $NetBSD: fil.c,v 1.4 1997/02/19 23:07:57 scottr Exp $ */
/*
* (C)opyright 1993-1996 by Darren Reed.
*
* Redistribution and use in source and binary forms are permitted
* provided that this notice is preserved and due credit is given
* to the original author and the contributors.
*/
#if !defined(lint) && defined(LIBC_SCCS)
static char sccsid[] = "@(#)fil.c 1.36 6/5/96 (C) 1993-1996 Darren Reed";
static char rcsid[] = "Id: fil.c,v 1.3.4.12 1996/12/02 11:51:24 darrenr Exp";
#endif
#include <sys/errno.h>
#include <sys/types.h>
#include <sys/param.h>
#include <sys/file.h>
#include <sys/ioctl.h>
#if defined(_KERNEL) || defined(KERNEL)
# include <sys/systm.h>
#else
# include <stdio.h>
# include <string.h>
#endif
#include <sys/uio.h>
#if !defined(__SVR4) && !defined(__svr4__)
# include <sys/mbuf.h>
#else
# include <sys/byteorder.h>
# include <sys/dditypes.h>
# include <sys/stream.h>
#endif
#include <sys/protosw.h>
#include <sys/socket.h>
#ifndef _LKM
#include "ipfilter.h"
#endif
#include <net/if.h>
#ifdef sun
# include <net/af.h>
#endif
#include <net/route.h>
#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/ip.h>
#include <netinet/ip_var.h>
#include <netinet/tcp.h>
#include <netinet/udp.h>
#include <netinet/tcpip.h>
#include <netinet/ip_icmp.h>
#include <netinet/ip_fil.h>
#include <netinet/ip_compat.h>
#include <netinet/ip_nat.h>
#include <netinet/ip_frag.h>
#include <netinet/ip_state.h>
#ifndef MIN
#define MIN(a,b) (((a)<(b))?(a):(b))
#endif
/* XXX NetBSD will default to passing if IPFILTER is enabled */
#ifdef IPFILTER_POLICY
# define NOMATCH IPFILTER_POLICY
#else
# define NOMATCH FR_PASS
#endif
#ifndef _KERNEL
#include "ipf.h"
extern int opts;
extern void debug(), verbose();
#define FR_IFVERBOSE(ex,second,verb_pr) if (ex) { verbose verb_pr; second; }
#define FR_IFDEBUG(ex,second,verb_pr) if (ex) { debug verb_pr; second; }
#define FR_VERBOSE(verb_pr) verbose verb_pr
#define FR_DEBUG(verb_pr) debug verb_pr
#define FR_SCANLIST(p, ip, fi, m) fr_scanlist(p, ip, fi, 0)
# if SOLARIS
# define bcmp memcmp
# endif
#else
#define FR_IFVERBOSE(ex,second,verb_pr) ;
#define FR_IFDEBUG(ex,second,verb_pr) ;
#define FR_VERBOSE(verb_pr)
#define FR_DEBUG(verb_pr)
#define FR_SCANLIST(p, ip, fi, m) fr_scanlist(p, ip, fi, m)
# if SOLARIS
extern int icmp_error(), ipfr_fastroute();
extern kmutex_t ipf_mutex, ipl_mutex;
# else
# endif
extern int ipl_unreach;
#endif
#if SOLARIS
# define SEND_RESET(ip, if, q) send_reset(ip, qif, q)
# define ICMP_ERROR(b, ip, t, c, if, src) \
icmp_error(b, ip, t, c, if, src)
#else
# define SEND_RESET(ip, if, q) send_reset((struct tcpiphdr *)ip)
# if BSD < 199103
# define ICMP_ERROR(b, ip, t, c, if, src) \
icmp_error(mtod(b, ip_t *), t, c, if, src)
# else
# define ICMP_ERROR(b, ip, t, c, if, src) \
icmp_error(b, t, c, (src).s_addr, if)
# endif
#endif
struct filterstats frstats[2] = {{0,0,0,0,0},{0,0,0,0,0}};
struct frentry *ipfilter[2][2] = { { NULL, NULL }, { NULL, NULL } },
*ipacct[2][2] = { { NULL, NULL }, { NULL, NULL } };
int fr_flags = 0, fr_active = 0;
fr_info_t frcache[2];
/*
* bit values for identifying presence of individual IP options
*/
struct optlist ipopts[20] = {
{ IPOPT_NOP, 0x000001 },
{ IPOPT_RR, 0x000002 },
{ IPOPT_ZSU, 0x000004 },
{ IPOPT_MTUP, 0x000008 },
{ IPOPT_MTUR, 0x000010 },
{ IPOPT_ENCODE, 0x000020 },
{ IPOPT_TS, 0x000040 },
{ IPOPT_TR, 0x000080 },
{ IPOPT_SECURITY, 0x000100 },
{ IPOPT_LSRR, 0x000200 },
{ IPOPT_E_SEC, 0x000400 },
{ IPOPT_CIPSO, 0x000800 },
{ IPOPT_SATID, 0x001000 },
{ IPOPT_SSRR, 0x002000 },
{ IPOPT_ADDEXT, 0x004000 },
{ IPOPT_VISA, 0x008000 },
{ IPOPT_IMITD, 0x010000 },
{ IPOPT_EIP, 0x020000 },
{ IPOPT_FINN, 0x040000 },
{ 0, 0x000000 }
};
/*
* bit values for identifying presence of individual IP security options
*/
struct optlist secopt[8] = {
{ IPSO_CLASS_RES4, 0x01 },
{ IPSO_CLASS_TOPS, 0x02 },
{ IPSO_CLASS_SECR, 0x04 },
{ IPSO_CLASS_RES3, 0x08 },
{ IPSO_CLASS_CONF, 0x10 },
{ IPSO_CLASS_UNCL, 0x20 },
{ IPSO_CLASS_RES2, 0x40 },
{ IPSO_CLASS_RES1, 0x80 }
};
/*
* compact the IP header into a structure which contains just the info.
* which is useful for comparing IP headers with.
*/
void fr_makefrip __P((int hlen, ip_t *ip, fr_info_t *fin));
void fr_makefrip(hlen, ip, fin)
int hlen;
ip_t *ip;
fr_info_t *fin;
{
struct optlist *op;
tcphdr_t *tcp;
fr_ip_t *fi = &fin->fin_fi;
u_short optmsk = 0, secmsk = 0, auth = 0;
int i, mv, ol, off;
u_char *s, opt;
fin->fin_fr = NULL;
fin->fin_tcpf = 0;
fin->fin_data[0] = 0;
fin->fin_data[1] = 0;
fin->fin_rule = -1;
#ifdef _KERNEL
fin->fin_icode = ipl_unreach;
#endif
fi->fi_v = ip->ip_v;
fi->fi_tos = ip->ip_tos;
fin->fin_hlen = hlen;
fin->fin_dlen = ip->ip_len - hlen;
tcp = (tcphdr_t *)((char *)ip + hlen);
fin->fin_dp = (void *)tcp;
(*(((u_short *)fi) + 1)) = (*(((u_short *)ip) + 4));
(*(((u_long *)fi) + 1)) = (*(((u_long *)ip) + 3));
(*(((u_long *)fi) + 2)) = (*(((u_long *)ip) + 4));
fi->fi_fl = (hlen > sizeof(struct ip)) ? FI_OPTIONS : 0;
off = (ip->ip_off & 0x1fff) << 3;
if (ip->ip_off & 0x3fff)
fi->fi_fl |= FI_FRAG;
switch (ip->ip_p)
{
case IPPROTO_ICMP :
if ((!IPMINLEN(ip, icmp) && !off) ||
(off && off < sizeof(struct icmp)))
fi->fi_fl |= FI_SHORT;
if (fin->fin_dlen > 1)
fin->fin_data[0] = *(u_short *)tcp;
break;
case IPPROTO_TCP :
fi->fi_fl |= FI_TCPUDP;
if ((!IPMINLEN(ip, tcphdr) && !off) ||
(off && off < sizeof(struct tcphdr)))
fi->fi_fl |= FI_SHORT;
if (!(fi->fi_fl & FI_SHORT) && !off)
fin->fin_tcpf = tcp->th_flags;
goto getports;
case IPPROTO_UDP :
fi->fi_fl |= FI_TCPUDP;
if ((!IPMINLEN(ip, udphdr) && !off) ||
(off && off < sizeof(struct udphdr)))
fi->fi_fl |= FI_SHORT;
getports:
if (!off && (fin->fin_dlen > 3)) {
fin->fin_data[0] = ntohs(tcp->th_sport);
fin->fin_data[1] = ntohs(tcp->th_dport);
}
break;
default :
break;
}
for (s = (u_char *)(ip + 1), hlen -= sizeof(*ip); hlen; ) {
if (!(opt = *s))
break;
ol = (opt == IPOPT_NOP) ? 1 : (int)*(s+1);
if (opt > 1 && (ol < 2 || ol > hlen))
break;
for (i = 9, mv = 4; mv >= 0; ) {
op = ipopts + i;
if (opt == (u_char)op->ol_val) {
optmsk |= op->ol_bit;
if (opt == IPOPT_SECURITY) {
struct optlist *sp;
u_char sec;
int j, m;
sec = *(s + 2); /* classification */
for (j = 3, m = 2; m >= 0; ) {
sp = secopt + j;
if (sec == sp->ol_val) {
secmsk |= sp->ol_bit;
auth = *(s + 3);
auth *= 256;
auth += *(s + 4);
break;
}
if (sec < sp->ol_val)
j -= m--;
else
j += m--;
}
}
break;
}
if (opt < op->ol_val)
i -= mv--;
else
i += mv--;
}
hlen -= ol;
s += ol;
}
if (auth && !(auth & 0x0100))
auth &= 0xff00;
fi->fi_optmsk = optmsk;
fi->fi_secmsk = secmsk;
fi->fi_auth = auth;
}
/*
* check an IP packet for TCP/UDP characteristics such as ports and flags.
*/
int fr_tcpudpchk __P((frentry_t *fr, fr_info_t *fin));
int fr_tcpudpchk(fr, fin)
frentry_t *fr;
fr_info_t *fin;
{
register u_short po, tup;
register char i;
register int err = 1;
/*
* Both ports should *always* be in the first fragment.
* So far, I cannot find any cases where they can not be.
*
* compare destination ports
*/
if ((i = (int)fr->fr_dcmp)) {
po = fr->fr_dport;
tup = fin->fin_data[1];
/*
* Do opposite test to that required and
* continue if that succeeds.
*/
if (!--i && tup != po) /* EQUAL */
err = 0;
else if (!--i && tup == po) /* NOTEQUAL */
err = 0;
else if (!--i && tup >= po) /* LESSTHAN */
err = 0;
else if (!--i && tup <= po) /* GREATERTHAN */
err = 0;
else if (!--i && tup > po) /* LT or EQ */
err = 0;
else if (!--i && tup < po) /* GT or EQ */
err = 0;
else if (!--i && /* Out of range */
(tup >= po && tup <= fr->fr_dtop))
err = 0;
else if (!--i && /* In range */
(tup <= po || tup >= fr->fr_dtop))
err = 0;
}
/*
* compare source ports
*/
if (err && (i = (int)fr->fr_scmp)) {
po = fr->fr_sport;
tup = fin->fin_data[0];
if (!--i && tup != po)
err = 0;
else if (!--i && tup == po)
err = 0;
else if (!--i && tup >= po)
err = 0;
else if (!--i && tup <= po)
err = 0;
else if (!--i && tup > po)
err = 0;
else if (!--i && tup < po)
err = 0;
else if (!--i && /* Out of range */
(tup >= po && tup <= fr->fr_stop))
err = 0;
else if (!--i && /* In range */
(tup <= po || tup >= fr->fr_stop))
err = 0;
}
/*
* If we don't have all the TCP/UDP header, then how can we
* expect to do any sort of match on it ? If we were looking for
* TCP flags, then NO match. If not, then match (which should
* satisfy the "short" class too).
*/
if (err && (fin->fin_fi.fi_p == IPPROTO_TCP)) {
if (fin->fin_fi.fi_fl & FI_SHORT)
return !(fr->fr_tcpf | fr->fr_tcpfm);
/*
* Match the flags ? If not, abort this match.
*/
if (fr->fr_tcpf &&
fr->fr_tcpf != (fin->fin_tcpf & fr->fr_tcpfm)) {
FR_DEBUG(("f. %#x & %#x != %#x\n", fin->fin_tcpf,
fr->fr_tcpfm, fr->fr_tcpf));
err = 0;
}
}
return err;
}
/*
* Check the input/output list of rules for a match and result.
* Could be per interface, but this gets real nasty when you don't have
* kernel sauce.
*/
int fr_scanlist __P((int pass, ip_t *ip, register fr_info_t *fin, void *m));
int fr_scanlist(pass, ip, fin, m)
int pass;
ip_t *ip;
register fr_info_t *fin;
void *m;
{
register struct frentry *fr;
register fr_ip_t *fi = &fin->fin_fi;
int rulen, portcmp = 0, off;
fr = fin->fin_fr;
fin->fin_fr = NULL;
fin->fin_rule = 0;
off = ip->ip_off & 0x1fff;
pass |= (fi->fi_fl << 20);
if ((fi->fi_fl & FI_TCPUDP) && (fin->fin_dlen > 3) && !off)
portcmp = 1;
for (rulen = 0; fr; fr = fr->fr_next, rulen++) {
/*
* In all checks below, a null (zero) value in the
* filter struture is taken to mean a wildcard.
*
* check that we are working for the right interface
*/
#ifdef _KERNEL
if (fr->fr_ifa && fr->fr_ifa != fin->fin_ifp)
continue;
#else
if (opts & (OPT_VERBOSE|OPT_DEBUG))
printf("\n");
FR_VERBOSE(("%c", (pass & FR_PASS) ? 'p' : 'b'));
if (fin->fin_ifp && *fr->fr_ifname &&
strcasecmp((char *)fin->fin_ifp, fr->fr_ifname))
continue;
FR_VERBOSE((":i"));
#endif
{
register u_long *ld, *lm, *lip;
register int i;
lip = (u_long *)fi;
lm = (u_long *)&fr->fr_mip;
ld = (u_long *)&fr->fr_ip;
i = ((lip[0] & lm[0]) != ld[0]);
FR_IFDEBUG(i,continue,("0. %#08x & %#08x != %#08x\n",
lip[0], lm[0], ld[0]));
i |= ((lip[1] & lm[1]) != ld[1]);
FR_IFDEBUG(i,continue,("1. %#08x & %#08x != %#08x\n",
lip[1], lm[1], ld[1]));
i |= ((lip[2] & lm[2]) != ld[2]);
FR_IFDEBUG(i,continue,("2. %#08x & %#08x != %#08x\n",
lip[2], lm[2], ld[2]));
i |= ((lip[3] & lm[3]) != ld[3]);
FR_IFDEBUG(i,continue,("3. %#08x & %#08x != %#08x\n",
lip[3], lm[3], ld[3]));
i |= ((lip[4] & lm[4]) != ld[4]);
FR_IFDEBUG(i,continue,("4. %#08x & %#08x != %#08x\n",
lip[4], lm[4], ld[4]));
if (i)
continue;
}
/*
* If a fragment, then only the first has what we're looking
* for here...
*/
if (fi->fi_fl & FI_TCPUDP) {
if (portcmp) {
if (!fr_tcpudpchk(fr, fin))
continue;
} else if (fr->fr_dcmp || fr->fr_scmp)
continue;
} else if (fi->fi_p == IPPROTO_ICMP) {
if (!off && (fin->fin_dlen > 1)) {
if ((fin->fin_data[0] & fr->fr_icmpm) !=
fr->fr_icmp) {
FR_DEBUG(("i. %#x & %#x != %#x\n",
fin->fin_data[0],
fr->fr_icmpm, fr->fr_icmp));
continue;
}
} else if (fr->fr_icmpm || fr->fr_tcpfm)
continue;
}
FR_VERBOSE(("*"));
/*
* Just log this packet...
*/
pass = fr->fr_flags;
if ((pass & FR_CALLNOW) && fr->fr_func)
pass = (*fr->fr_func)(pass, ip, fin);
#ifdef IPFILTER_LOG
if ((pass & FR_LOGMASK) == FR_LOG) {
if (!ipllog(fr->fr_flags, ip, fin, m))
frstats[fin->fin_out].fr_skip++;
frstats[fin->fin_out].fr_pkl++;
}
#endif /* IPFILTER_LOG */
FR_DEBUG(("pass %#x\n", pass));
fr->fr_hits++;
if (pass & FR_ACCOUNT)
fr->fr_bytes += ip->ip_len;
else
fin->fin_icode = fr->fr_icode;
fin->fin_rule = rulen;
fin->fin_fr = fr;
if (pass & FR_QUICK)
break;
}
return pass;
}
/*
* frcheck - filter check
* check using source and destination addresses/pors in a packet whether
* or not to pass it on or not.
*/
int fr_check(ip, hlen, ifp, out
#ifdef _KERNEL
# if SOLARIS
, qif, q, mp)
qif_t *qif;
queue_t *q;
mblk_t **mp;
# else
, mp)
struct mbuf **mp;
# endif
#else
)
#endif
ip_t *ip;
int hlen;
struct ifnet *ifp;
int out;
{
/*
* The above really sucks, but short of writing a diff
*/
fr_info_t frinfo, *fc;
register fr_info_t *fin = &frinfo;
frentry_t *fr;
int pass, changed;
#if !defined(__SVR4) && !defined(__svr4__) && defined(_KERNEL)
register struct mbuf *m = *mp;
struct mbuf *mc = NULL;
if ((ip->ip_p == IPPROTO_TCP || ip->ip_p == IPPROTO_UDP ||
ip->ip_p == IPPROTO_ICMP)) {
register int up = MIN(hlen + 8, ip->ip_len);
if (up > m->m_len) {
if ((*mp = m_pullup(m, up)) == 0) {
frstats[out].fr_pull[1]++;
return -1;
} else {
frstats[out].fr_pull[0]++;
m = *mp;
ip = mtod(m, struct ip *);
}
}
}
#endif
#if SOLARIS && defined(_KERNEL)
mblk_t *mc = NULL, *m = qif->qf_m;
#endif
fr_makefrip(hlen, ip, fin);
fin->fin_ifp = ifp;
fin->fin_out = out;
MUTEX_ENTER(&ipf_mutex);
if (!out) {
changed = ip_natin(ip, hlen, fin);
if ((fin->fin_fr = ipacct[0][fr_active]) &&
(FR_SCANLIST(FR_NOMATCH, ip, fin, m) & FR_ACCOUNT))
frstats[0].fr_acct++;
}
if ((pass = ipfr_knownfrag(ip, fin))) {
if ((pass & FR_KEEPSTATE)) {
if (fr_addstate(ip, fin, pass) == -1)
frstats[out].fr_bads++;
else
frstats[out].fr_ads++;
}
} else if ((pass = fr_checkstate(ip, fin))) {
if ((pass & FR_KEEPFRAG)) {
if (fin->fin_fi.fi_fl & FI_FRAG) {
if (ipfr_newfrag(ip, fin, pass) == -1)
frstats[out].fr_bnfr++;
else
frstats[out].fr_nfr++;
} else
frstats[out].fr_cfr++;
}
} else {
fc = frcache + out;
if (fc->fin_fr && !bcmp((char *)fin, (char *)fc, FI_CSIZE)) {
/*
* copy cached data so we can unlock the mutex
* earlier.
*/
bcopy((char *)fc, (char *)fin, sizeof(*fin));
frstats[out].fr_chit++;
pass = fin->fin_fr->fr_flags;
} else {
pass = FR_NOMATCH;
if ((fin->fin_fr = ipfilter[out][fr_active]))
pass = FR_SCANLIST(FR_NOMATCH, ip, fin, m);
bcopy((char *)fin, (char *)fc, FI_CSIZE);
if (pass & FR_NOMATCH) {
frstats[out].fr_nom++;
#ifdef NOMATCH
pass |= NOMATCH;
#endif
}
}
if ((pass & FR_KEEPFRAG)) {
if (fin->fin_fi.fi_fl & FI_FRAG) {
if (ipfr_newfrag(ip, fin, pass) == -1)
frstats[out].fr_bnfr++;
else
frstats[out].fr_nfr++;
} else
frstats[out].fr_cfr++;
}
if (pass & FR_KEEPSTATE) {
if (fr_addstate(ip, fin, pass) == -1)
frstats[out].fr_bads++;
else
frstats[out].fr_ads++;
}
}
fr = fin->fin_fr;
if (fr && fr->fr_func)
pass = (*fr->fr_func)(pass, ip, fin);
if (out) {
if ((fin->fin_fr = ipacct[1][fr_active]) &&
(FR_SCANLIST(FR_NOMATCH, ip, fin, m) & FR_ACCOUNT))
frstats[1].fr_acct++;
changed = ip_natout(ip, hlen, fin);
}
MUTEX_EXIT(&ipf_mutex);
#ifdef IPFILTER_LOG
if ((fr_flags & FF_LOGGING) || (pass & FR_LOGMASK)) {
if ((fr_flags & FF_LOGNOMATCH) && (pass & FR_NOMATCH)) {
pass |= FF_LOGNOMATCH;
frstats[out].fr_npkl++;
goto logit;
} else if (((pass & FR_LOGMASK) == FR_LOGP) ||
((pass & FR_PASS) && (fr_flags & FF_LOGPASS))) {
if ((pass & FR_LOGMASK) != FR_LOGP)
pass |= FF_LOGPASS;
frstats[out].fr_ppkl++;
goto logit;
} else if (((pass & FR_LOGMASK) == FR_LOGB) ||
((pass & FR_BLOCK) && (fr_flags & FF_LOGBLOCK))) {
if ((pass & FR_LOGMASK) != FR_LOGB)
pass |= FF_LOGBLOCK;
frstats[out].fr_bpkl++;
logit:
if (!ipllog(pass, ip, fin, m)) {
frstats[out].fr_skip++;
if ((pass & (FR_PASS|FR_LOGORBLOCK)) ==
(FR_PASS|FR_LOGORBLOCK))
pass ^= FR_PASS|FR_BLOCK;
}
}
}
#endif /* IPFILTER_LOG */
if (pass & FR_PASS)
frstats[out].fr_pass++;
else if (pass & FR_BLOCK) {
frstats[out].fr_block++;
/*
* Should we return an ICMP packet to indicate error
* status passing through the packet filter ?
*/
#ifdef _KERNEL
if (pass & FR_RETICMP) {
# if SOLARIS
ICMP_ERROR(q, ip, ICMP_UNREACH, fin->fin_icode,
qif, ip->ip_src);
# else
ICMP_ERROR(m, ip, ICMP_UNREACH, fin->fin_icode,
ifp, ip->ip_src);
m = NULL; /* freed by icmp_error() */
# endif
frstats[0].fr_ret++;
} else if ((pass & FR_RETRST) &&
!(fin->fin_fi.fi_fl & FI_SHORT)) {
if (SEND_RESET(ip, qif, q) == 0)
frstats[1].fr_ret++;
}
#else
if (pass & FR_RETICMP) {
verbose("- ICMP unreachable sent\n");
frstats[0].fr_ret++;
} else if ((pass & FR_RETRST) &&
!(fin->fin_fi.fi_fl & FI_SHORT)) {
verbose("- TCP RST sent\n");
frstats[1].fr_ret++;
}
#endif
}
#ifdef _KERNEL
# if !SOLARIS
if (pass & FR_DUP)
mc = m_copy(m, 0, M_COPYALL);
if (fr) {
frdest_t *fdp = &fr->fr_tif;
if ((pass & FR_FASTROUTE) ||
(fdp->fd_ifp && fdp->fd_ifp != (struct ifnet *)-1)) {
ipfr_fastroute(m, fin, fdp);
m = *mp = NULL;
pass = 0;
}
if (mc)
ipfr_fastroute(mc, fin, &fr->fr_dif);
}
if (!(pass & FR_PASS) && m)
m_freem(m);
return (pass & FR_PASS) ? 0 : -1;
# else
if (pass & FR_DUP)
mc = dupmsg(m);
if (fr) {
frdest_t *fdp = &fr->fr_tif;
if ((pass & FR_FASTROUTE) ||
(fdp->fd_ifp && fdp->fd_ifp != (struct ifnet *)-1)) {
ipfr_fastroute(m, mp, fin, fdp);
m = *mp = NULL;
}
if (mc)
ipfr_fastroute(mc, mp, fin, &fr->fr_dif);
}
return (pass & FR_PASS) ? changed : -1;
# endif
#else
if (pass & FR_NOMATCH)
return 1;
if (pass & FR_PASS)
return 0;
return -1;
#endif
}
#ifdef IPFILTER_LOG
# if !(defined(_KERNEL))
static void ipllog()
{
verbose("l");
}
# endif
int fr_copytolog(buf, len)
char *buf;
int len;
{
int clen, tail;
tail = (iplh >= iplt) ? (iplbuf + IPLLOGSIZE - iplh) : (iplt - iplh);
clen = MIN(tail, len);
bcopy(buf, iplh, clen);
len -= clen;
tail -= clen;
iplh += clen;
buf += clen;
if (iplh == iplbuf + IPLLOGSIZE) {
iplh = iplbuf;
tail = iplt - iplh;
}
if (len && tail) {
clen = MIN(tail, len);
bcopy(buf, iplh, clen);
len -= clen;
iplh += clen;
}
return len;
}
#endif