freebsd/usr.sbin/named/ns_forw.c

#if !defined(lint) && !defined(SABER)
static char sccsid[] = "@(#)ns_forw.c	4.32 (Berkeley) 3/3/91";
static char rcsid[] = "$Id: ns_forw.c,v 1.4 1995/10/23 11:11:44 peter Exp $";
#endif /* not lint */

/*
 * ++Copyright++ 1986
 * -
 * Copyright (c) 1986
 *    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.
 * -
 * Portions Copyright (c) 1993 by Digital Equipment Corporation.
 *
 * Permission to use, copy, modify, and distribute this software for any
 * purpose with or without fee is hereby granted, provided that the above
 * copyright notice and this permission notice appear in all copies, and that
 * the name of Digital Equipment Corporation not be used in advertising or
 * publicity pertaining to distribution of the document or software without
 * specific, written prior permission.
 *
 * THE SOFTWARE IS PROVIDED "AS IS" AND DIGITAL EQUIPMENT CORP. DISCLAIMS ALL
 * WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES
 * OF MERCHANTABILITY AND FITNESS.   IN NO EVENT SHALL DIGITAL EQUIPMENT
 * CORPORATION BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL
 * DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
 * PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
 * ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS
 * SOFTWARE.
 * -
 * --Copyright--
 */

#include <sys/param.h>
#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>
#include <arpa/nameser.h>

#include <syslog.h>
#include <resolv.h>
#include <stdio.h>
#include <errno.h>

#include "named.h"

/*
 * Forward the query to get the answer since its not in the database.
 * Returns FW_OK if a request struct is allocated and the query sent.
 * Returns FW_DUP if this is a duplicate of a pending request.
 * Returns FW_NOSERVER if there were no addresses for the nameservers.
 * Returns FW_SERVFAIL on malloc error or if asked to do something
 * dangerous, such as fwd to ourselves or fwd to the host that asked us.
 *
 * (no action is taken on errors and qpp is not filled in.)
 */
int
ns_forw(nsp, msg, msglen, fp, qsp, dfd, qpp, dname, np)
	struct databuf *nsp[];
	u_char *msg;
	int msglen;
	struct sockaddr_in *fp;
	struct qstream *qsp;
	int dfd;
	struct qinfo **qpp;
	char *dname;
	struct namebuf *np;
{
	register struct qinfo *qp;
	struct sockaddr_in *nsa;
	HEADER *hp;
	u_int16_t id;
	int n;

	dprintf(3, (ddt, "ns_forw()\n"));

	hp = (HEADER *) msg;
	id = hp->id;
	/* Look at them all */
	for (qp = nsqhead; qp != QINFO_NULL; qp = qp->q_link) {
		if (qp->q_id == id &&
		    bcmp((char *)&qp->q_from, fp, sizeof(qp->q_from)) == 0 &&
		    ((qp->q_cmsglen == 0 && qp->q_msglen == msglen &&
		     bcmp((char *)qp->q_msg+2, msg+2, msglen-2) == 0) ||
		    (qp->q_cmsglen == msglen &&
		     bcmp((char *)qp->q_cmsg+2, msg+2, msglen-2) == 0))) {
			dprintf(3, (ddt,
				    "forw: dropped DUP id=%d\n", ntohs(id)));
			nameserIncr(fp->sin_addr, nssRcvdDupQ);
			return (FW_DUP);
		}
	}

	qp = qnew();
#if defined(LAME_DELEGATION) || defined(VALIDATE)
	getname(np, qp->q_domain, sizeof qp->q_domain);
#endif
	qp->q_from = *fp;	/* nslookup wants to know this */
	if ((n = nslookup(nsp, qp, dname, "ns_forw")) < 0) {
		dprintf(2, (ddt, "forw: nslookup reports danger\n"));
		qfree(qp);
		return (FW_SERVFAIL);
	} else if (n == 0 && !fwdtab) {
		dprintf(2, (ddt, "forw: no nameservers found\n"));
		qfree(qp);
		return (FW_NOSERVER);
	}
	qp->q_stream = qsp;
	qp->q_curaddr = 0;
	qp->q_fwd = fwdtab;
	qp->q_dfd = dfd;
	qp->q_id = id;
	qp->q_expire = tt.tv_sec + RETRY_TIMEOUT*2;
	hp->id = qp->q_nsid = htons(nsid_next());
	hp->ancount = htons(0);
	hp->nscount = htons(0);
	hp->arcount = htons(0);
	if ((qp->q_msg = (u_char *)malloc((unsigned)msglen)) == NULL) {
		syslog(LOG_NOTICE, "forw: malloc: %m");
		qfree(qp);
		return (FW_SERVFAIL);
	}
	bcopy(msg, qp->q_msg, qp->q_msglen = msglen);
	if (!qp->q_fwd) {
		hp->rd = 0;
		qp->q_addr[0].stime = tt;
	}

#ifdef SLAVE_FORWARD
	if (forward_only)
		schedretry(qp, (time_t)slave_retry);
	else
#endif /* SLAVE_FORWARD */
	schedretry(qp, qp->q_fwd ?(2*RETRYBASE) :retrytime(qp));

	nsa = Q_NEXTADDR(qp, 0);
	dprintf(1, (ddt,
		"forw: forw -> [%s].%d ds=%d nsid=%d id=%d %dms retry %dsec\n",
			inet_ntoa(nsa->sin_addr),
			ntohs(nsa->sin_port), ds,
			ntohs(qp->q_nsid), ntohs(qp->q_id),
			(qp->q_addr[0].nsdata != NULL)
				? qp->q_addr[0].nsdata->d_nstime
				: -1,
			(int)(qp->q_time - tt.tv_sec)));
#ifdef DEBUG
	if (debug >= 10)
		fp_nquery(msg, msglen, ddt);
#endif
	if (sendto(ds, (char *)msg, msglen, 0, (struct sockaddr *)nsa,
		   sizeof(struct sockaddr_in)) < 0) {
		if (!haveComplained((char*)nsa->sin_addr.s_addr, sendtoStr))
			syslog(LOG_INFO, "ns_forw: sendto([%s].%d): %m",
			       inet_ntoa(nsa->sin_addr), ntohs(nsa->sin_port));
		nameserIncr(nsa->sin_addr, nssSendtoErr);
	}
	nameserIncr(fp->sin_addr, nssRcvdFwdQ);
	nameserIncr(nsa->sin_addr, nssSentFwdQ);
	if (qpp)
		*qpp = qp;
	hp->rd = 1;
	return (0);
}

/* struct qdatagram *
 * aIsUs(addr)
 *	scan the datagramq (our list of interface addresses) for "addr"
 * returns:
 *	pointer to qdatagram entry or NULL if no match is found
 * notes:
 *	INADDR_ANY ([0.0.0.0]) is on the datagramq, so it's considered "us"
 * author:
 *	Paul Vixie (DECWRL) April 1991
 */
struct qdatagram *
aIsUs(addr)
	struct in_addr addr;
{
	struct qdatagram *dqp;

	for (dqp = datagramq;  dqp != QDATAGRAM_NULL;  dqp = dqp->dq_next) {
		if (addr.s_addr == dqp->dq_addr.s_addr) {
			return dqp;
		}
	}
	return NULL;
}

/* haveComplained(tag1, tag2)
 *	check to see if we have complained about (tag1,tag2) recently
 *	(note that these are declared as pointers but are never deref'd)
 * returns:
 *	boolean: have we complained recently?
 * side-effects:
 *	outdated complaint records removed from our static list
 * author:
 *	Paul Vixie (DECWRL) April 1991
 */
int
haveComplained(tag1, tag2)
	const char *tag1, *tag2;
{
	struct complaint {
		const char *tag1, *tag2;
		time_t expire;
		struct complaint *next;
	};
	static struct complaint *List = NULL;
	struct complaint *cur, *next, *prev;
	int r = 0;

	for (cur = List, prev = NULL;  cur;  prev = cur, cur = next) {
		next = cur->next;
		if (tt.tv_sec > cur->expire) {
			if (prev)
				prev->next = next;
			else
				List = next;
			free((char*) cur);
			cur = prev;
		} else if ((tag1 == cur->tag1) && (tag2 == cur->tag2)) {
			r++;
		}
	}
	if (!r) {
		cur = (struct complaint *)malloc(sizeof(struct complaint));
		if (cur) {
			cur->tag1 = tag1;
			cur->tag2 = tag2;
			cur->expire = tt.tv_sec + INIT_REFRESH;	/* "10:00" */
			cur->next = NULL;
			if (prev)
				prev->next = cur;
			else
				List = cur;
		}
	}
	return (r);
}

/* void
 * nslookupComplain(sysloginfo, queryname, complaint, dname, a_rr)
 *	Issue a complaint about a dangerous situation found by nslookup().
 * params:
 *	sysloginfo is a string identifying the complainant.
 *	queryname is the domain name associated with the problem.
 *	complaint is a string describing what is wrong.
 *	dname and a_rr are the problematic other name server.
 */
static void
nslookupComplain(sysloginfo, queryname, complaint, dname, a_rr, nsdp)
	const char *sysloginfo, *queryname, *complaint, *dname;
	const struct databuf *a_rr, *nsdp;
{
#ifdef STATS
	char nsbuf[20];
	char abuf[20];
#endif
	char *a, *ns;

	dprintf(2, (ddt, "NS '%s' %s\n", dname, complaint));
	if (sysloginfo && queryname && !haveComplained(queryname, complaint))
	{
		char buf[999];

		a = ns = (char *)NULL;
#ifdef STATS
		if (nsdp) {
			if (nsdp->d_ns) {
				strcpy(nsbuf, inet_ntoa(nsdp->d_ns->addr));
				ns = nsbuf;
			} else {
				ns = zones[nsdp->d_zone].z_origin;
			}
		}
		if (a_rr->d_ns) {
			strcpy(abuf, inet_ntoa(a_rr->d_ns->addr));
			a = abuf;
		} else {
			a = zones[a_rr->d_zone].z_origin;
		}
#endif
		/* syslog only takes 5 params */
		if ( a != NULL || ns != NULL)
			sprintf(buf, "%s: query(%s) %s (%s:%s) learnt (A=%s:NS=%s)",
				sysloginfo, queryname,
				complaint, dname,
				inet_ntoa(data_inaddr(a_rr->d_data)),
				a ? a : "<Not Available>",
				ns ? ns : "<Not Available>" );
		else
			sprintf(buf, "%s: query(%s) %s (%s:%s)",
				sysloginfo, queryname,
				complaint, dname,
				inet_ntoa(data_inaddr(a_rr->d_data)));
		syslog(LOG_INFO, buf);
	}
}

/*
 * nslookup(nsp, qp, syslogdname, sysloginfo)
 *	Lookup the address for each nameserver in `nsp' and add it to
 * 	the list saved in the qinfo structure pointed to by `qp'.
 *	Omits information about nameservers that we shouldn't ask.
 *	Detects the following dangerous operations:
 *		One of the A records for one of the nameservers in nsp
 *		refers to the address of one of our own interfaces;
 *		One of the A records refers to the nameserver port on
 *		the host that asked us this question.
 * returns: the number of addresses added, or -1 if a dangerous operation
 *	is detected.
 * side effects:
 *	if a dangerous situation is detected and (syslogdname && sysloginfo),
 *		calls syslog.
 */
int
nslookup(nsp, qp, syslogdname, sysloginfo)
	struct databuf *nsp[];
	register struct qinfo *qp;
	const char *syslogdname;
	const char *sysloginfo;
{
	register struct namebuf *np;
	register struct databuf *dp, *nsdp;
	register struct qserv *qs;
	register int n;
	register unsigned int i;
	struct hashbuf *tmphtp;
	char *dname;
	const char *fname;
	int oldn, naddr, class, found_arr;
	time_t curtime;

	dprintf(3, (ddt, "nslookup(nsp=0x%lx, qp=0x%lx, \"%s\")\n",
		    (u_long)nsp, (u_long)qp, syslogdname));

	naddr = n = qp->q_naddr;
	curtime = (u_long) tt.tv_sec;
	while ((nsdp = *nsp++) != NULL) {
		class = nsdp->d_class;
		dname = (char *)nsdp->d_data;
		dprintf(3, (ddt, "nslookup: NS \"%s\" c=%d t=%d (%#lx)\n",
			    dname, class, nsdp->d_type,
			    (u_long)nsdp->d_flags));

		/* don't put in servers we have tried */
		for (i = 0; i < qp->q_nusedns; i++) {
			if (qp->q_usedns[i] == nsdp) {
				dprintf(2, (ddt,
					    "skipping used NS w/name %s\n",
					    nsdp->d_data));
				goto skipserver;
			}
		}

		tmphtp = ((nsdp->d_flags & DB_F_HINT) ?fcachetab :hashtab);
		np = nlookup(dname, &tmphtp, &fname, 1);
		if (np == NULL || fname != dname) {
			dprintf(3, (ddt, "%s: not found %s %lx\n",
				    dname, fname, (u_long)np));
			continue;
		}
		found_arr = 0;
		oldn = n;

		/* look for name server addresses */
		for (dp = np->n_data;  dp != NULL;  dp = dp->d_next) {
			struct in_addr nsa;

#ifdef NCACHE
			if (dp->d_rcode)
				continue;
#endif
			if (dp->d_type == T_CNAME && dp->d_class == class)
				goto skipserver;
			if (dp->d_type != T_A || dp->d_class != class)
				continue;
			if (data_inaddr(dp->d_data).s_addr == INADDR_ANY) {
				static char *complaint = "Bogus (0.0.0.0) A RR";
				nslookupComplain(sysloginfo, syslogdname,
						complaint, dname, dp, nsdp);
				continue;
			}
#ifdef INADDR_LOOPBACK
			if (ntohl(data_inaddr(dp->d_data).s_addr) ==
					INADDR_LOOPBACK) {
				static char *complaint = "Bogus LOOPBACK A RR";
				nslookupComplain(sysloginfo, syslogdname,
						complaint, dname, dp, nsdp);
				continue;
			}
#endif
#ifdef INADDR_BROADCAST
			if (ntohl(data_inaddr(dp->d_data).s_addr) == 
					INADDR_BROADCAST) {
				static char *complaint = "Bogus BROADCAST A RR";
				nslookupComplain(sysloginfo, syslogdname,
						complaint, dname, dp, nsdp);
				continue;
			}
#endif
#ifdef IN_MULTICAST
			if (IN_MULTICAST(ntohl(data_inaddr(dp->d_data).s_addr))) {
				static char *complaint = "Bogus MULTICAST A RR";
				nslookupComplain(sysloginfo, syslogdname,
						complaint, dname, dp, nsdp);
				continue;
			}
#endif
			/*
			 * Don't use records that may become invalid to
			 * reference later when we do the rtt computation.
			 * Never delete our safety-belt information!
			 */
			if ((dp->d_zone == 0) &&
#ifdef DATUMREFCNT
			    (dp->d_ttl < curtime) &&
#else
			    (dp->d_ttl < (curtime+900)) &&
#endif
			    !(dp->d_flags & DB_F_HINT) )
		        {
				dprintf(3, (ddt,
					    "nslookup: stale entry '%s'\n",
					    np->n_dname));
				/* Cache invalidate the NS RR's */
#ifndef DATUMREFCNT
				if (dp->d_ttl < curtime)
#endif
				{
					delete_all(np, class, T_A);
					n = oldn;
					found_arr = 0;
					goto need_sysquery;
				}
			}
#ifdef VALIDATE
			/* anant@isi.edu validation procedure, maintains a
			 * table of server names-addresses used recently
			 */
			store_name_addr(dname, data_inaddr(dp->d_data),
					syslogdname, sysloginfo);
#endif /*VALIDATE*/

			found_arr++;
			nsa = data_inaddr(dp->d_data);
			/* don't put in duplicates */
			qs = qp->q_addr;
			for (i = 0; i < n; i++, qs++)
				if (qs->ns_addr.sin_addr.s_addr == nsa.s_addr)
					goto skipaddr;
			qs->ns_addr.sin_family = AF_INET;
			qs->ns_addr.sin_port = ns_port;
			qs->ns_addr.sin_addr = nsa;
			qs->ns = nsdp;
			qs->nsdata = dp;
			qs->nretry = 0;
			/*
			 * if we are being asked to fwd a query whose
			 * nameserver list includes our own name/address(es),
			 * then we have detected a lame delegation and rather
			 * than melt down the network and hose down the other
			 * servers (who will hose us in return), we'll return
			 * -1 here which will cause SERVFAIL to be sent to
			 * the client's resolver which will hopefully then
			 * shut up.
			 *
			 * (originally done in nsContainsUs by vix@dec mar92;
			 * moved into nslookup by apb@und jan1993)
			 */
			if (aIsUs(nsa)) {
			    static char *complaint = "contains our address";
			    nslookupComplain(sysloginfo, syslogdname,
					     complaint, dname, dp, nsdp);
			    return (-1);
			}
			/*
			 * If we want to forward to a host that asked us
			 * this question then either we or they are sick
			 * (unless they asked from some port other than
			 * their nameserver port).  (apb@und jan1993)
			 */
			if (bcmp((char *)&qp->q_from, (char *)&qs->ns_addr,
				 sizeof(qp->q_from)) == 0)
			{
			    static char *complaint = "forwarding loop";
			    nslookupComplain(sysloginfo, syslogdname,
					     complaint, dname, dp, nsdp);
			    return (-1);
			}
#ifdef BOGUSNS
			/*
			 * Don't forward queries to bogus servers.  Note
			 * that this is unlike the previous tests, which
			 * are fatal to the query.  Here we just skip the
			 * server, which is only fatal if it's the last
			 * server.  Note also that we antialias here -- all
			 * A RR's of a server are considered the same server,
			 * and if any of them is bogus we skip the whole
			 * server.  Those of you using multiple A RR's to
			 * load-balance your servers will (rightfully) lose
			 * here.  But (unfortunately) only if they are bogus.
			 */
			if (addr_on_netlist(nsa, boglist))
				goto skipserver;
#endif

			n++;
			if (n >= NSMAX)
				goto out;
	skipaddr:
			NULL;
		}
		dprintf(8, (ddt, "nslookup: %d ns addrs\n", n));
 need_sysquery:
		if (found_arr == 0 && !(qp->q_flags & Q_SYSTEM))
			(void) sysquery(dname, class, T_A, NULL, 0, QUERY);
 skipserver:
		NULL;
	}
out:
	dprintf(3, (ddt, "nslookup: %d ns addrs total\n", n));
	qp->q_naddr = n;
#ifdef DATUMREFCNT
	/* must be run before the sort */
	for (i = naddr ; i < n ; i++) {
		qp->q_addr[i].nsdata->d_rcnt++;
		qp->q_addr[i].ns->d_rcnt++;
	}
#endif
	if (n > 1) {
		qsort((char *)qp->q_addr, n, sizeof(struct qserv),
		      (int (*)__P((const void *, const void *)))qcomp);
	}
	return (n - naddr);
}

/*
 * qcomp - compare two NS addresses, and return a negative, zero, or
 *	   positive value depending on whether the first NS address is
 *	   "better than", "equally good as", or "inferior to" the second
 *	   NS address.
 *
 * How "goodness" is defined (for the purposes of this routine):
 *  - If the estimated round trip times differ by an amount deemed significant
 *    then the one with the smaller estimate is preferred; else
 *  - If we can determine which one is topologically closer then the
 *    closer one is preferred; else
 *  - The one with the smaller estimated round trip time is preferred
 *    (zero is returned if the two estimates are identical).
 *
 * How "topological closeness" is defined (for the purposes of this routine):
 *    Ideally, named could consult some magic map of the Internet and
 *    determine the length of the path to an arbitrary destination.  Sadly,
 *    no such magic map exists.  However, named does have a little bit of
 *    topological information in the form of the sortlist (which includes
 *    the directly connected subnet(s), the directly connected net(s), and
 *    any additional nets that the administrator has added using the "sortlist"
 *    directive in the bootfile.  Thus, if only one of the addresses matches
 *    something in the sortlist then it is considered to be topologically
 *    closer.  If both match, but match different entries in the sortlist,
 *    then the one that matches the entry closer to the beginning of the
 *    sorlist is considered to be topologically closer.  In all other cases,
 *    topological closeness is ignored because it's either indeterminate or
 *    equal.
 *
 * How times are compared:
 *    Both times are rounded to the closest multiple of the NOISE constant
 *    defined below and then compared.  If the rounded values are equal
 *    then the difference in the times is deemed insignificant.  Rounding
 *    is used instead of merely taking the absolute value of the difference
 *    because doing the latter would make the ordering defined by this
 *    routine be incomplete in the mathematical sense (e.g. A > B and
 *    B > C would not imply A > C).  The mathematics are important in
 *    practice to avoid core dumps in qsort().
 *
 * XXX: this doesn't solve the European root nameserver problem very well.
 * XXX: we should detect and mark as inferior nameservers that give bogus
 *      answers
 *
 * (this was originally vixie's stuff but almquist fixed fatal bugs in it
 * and wrote the above documentation)
 */

/*
 * RTT delta deemed to be significant, in milliseconds.  With the current
 * definition of RTTROUND it must be a power of 2.
 */
#define NOISE 128		/* milliseconds; 0.128 seconds */

#define sign(x) (((x) < 0) ? -1 : ((x) > 0) ? 1 : 0)
#define RTTROUND(rtt) (((rtt) + (NOISE >> 1)) & ~(NOISE - 1))

int
qcomp(qs1, qs2)
	struct qserv *qs1, *qs2;
{
	int pos1, pos2, pdiff;
	u_long rtt1, rtt2;
	long tdiff;

	if ((!qs1->nsdata) || (!qs2->nsdata))
		return 0;
	rtt1 = qs1->nsdata->d_nstime;
	rtt2 = qs2->nsdata->d_nstime;

	dprintf(10, (ddt, "qcomp(%s, %s) %lu (%lu) - %lu (%lu) = %lu",
		     inet_ntoa(qs1->ns_addr.sin_addr),
		     inet_ntoa(qs2->ns_addr.sin_addr),
		     rtt1, RTTROUND(rtt1), rtt2, RTTROUND(rtt2),
		     rtt1 - rtt2));
	if (RTTROUND(rtt1) == RTTROUND(rtt2)) {
		pos1 = position_on_netlist(qs1->ns_addr.sin_addr, nettab);
		pos2 = position_on_netlist(qs2->ns_addr.sin_addr, nettab);
		pdiff = pos1 - pos2;
		dprintf(10, (ddt, ", pos1=%d, pos2=%d\n", pos1, pos2));
		if (pdiff)
			return (pdiff);
	} else {
		dprintf(10, (ddt, "\n"));
	}
	tdiff = rtt1 - rtt2;
	return (sign(tdiff));
}
#undef sign
#undef RTTROUND

/*
 * Arrange that forwarded query (qp) is retried after t seconds.
 * Query list will be sorted after z_time is updated.
 */
void
schedretry(qp, t)
	struct qinfo *qp;
	time_t t;
{
	register struct qinfo *qp1, *qp2;

#ifdef DEBUG
	if (debug > 3) {
		fprintf(ddt, "schedretry(0x%lx, %ld sec)\n",
			(u_long)qp, (long)t);
		if (qp->q_time)
		    fprintf(ddt,
			 "WARNING: schedretry(%#lx, %ld) q_time already %ld\n",
			    (u_long)qp, (long)t, (long)qp->q_time);
	}
#endif
	t += (u_long) tt.tv_sec;
	qp->q_time = t;

	if ((qp1 = retryqp) == NULL) {
		retryqp = qp;
		qp->q_next = NULL;
		return;
	}
	if (t < qp1->q_time) {
		qp->q_next = qp1;
		retryqp = qp;
		return;
	}
	while ((qp2 = qp1->q_next) != NULL && qp2->q_time < t)
		qp1 = qp2;
	qp1->q_next = qp;
	qp->q_next = qp2;
}

/*
 * Unsched is called to remove a forwarded query entry.
 */
void
unsched(qp)
	struct qinfo *qp;
{
	register struct qinfo *np;

	dprintf(3, (ddt, "unsched(%#lx, %d)\n", (u_long)qp, ntohs(qp->q_id)));
	if (retryqp == qp) {
		retryqp = qp->q_next;
	} else {
		for (np=retryqp;  np->q_next != QINFO_NULL;  np = np->q_next) {
			if (np->q_next != qp)
				continue;
			np->q_next = qp->q_next;	/* dequeue */
			break;
		}
	}
	qp->q_next = QINFO_NULL;		/* sanity check */
	qp->q_time = 0;
}

/*
 * Retry is called to retransmit query 'qp'.
 */
void
retry(qp)
	register struct qinfo *qp;
{
	register int n;
	register HEADER *hp;
	struct sockaddr_in *nsa;

	dprintf(3, (ddt, "retry(x%lx) id=%d\n", (u_long)qp, ntohs(qp->q_id)));

	if (qp->q_msg == NULL) {		/* XXX - why? */
		qremove(qp);
		return;
	}

	if (qp->q_expire && (qp->q_expire < tt.tv_sec)) {
		dprintf(1, (ddt,
		     "retry(x%lx): expired @ %lu (%d secs before now (%lu))\n",
			    (u_long)qp, (u_long)qp->q_expire,
			    (int)(tt.tv_sec - qp->q_expire),
			    (u_long)tt.tv_sec));
		if (qp->q_stream) /* return failure code on stream */
			goto fail;
		qremove(qp);
		return;
	}

	/* try next address */
	n = qp->q_curaddr;
	if (qp->q_fwd) {
		qp->q_fwd = qp->q_fwd->next;
		if (qp->q_fwd)
			goto found;
		/* out of forwarders, try direct queries */
	} else
		++qp->q_addr[n].nretry;
	if (!forward_only) {
		do {
			if (++n >= (int)qp->q_naddr)
				n = 0;
			if (qp->q_addr[n].nretry < MAXRETRY)
				goto found;
		} while (n != qp->q_curaddr);
	}
fail:
	/*
	 * Give up. Can't reach destination.
	 */
	hp = (HEADER *)(qp->q_cmsg ? qp->q_cmsg : qp->q_msg);
	if (qp->q_flags & Q_PRIMING) {
		/* Can't give up priming */
		unsched(qp);
		schedretry(qp, (time_t)60*60);	/* 1 hour */
		hp->rcode = NOERROR;	/* Lets be safe, reset the query */
		hp->qr = hp->aa = 0;
		qp->q_fwd = fwdtab;
		for (n = 0; n < (int)qp->q_naddr; n++)
			qp->q_addr[n].nretry = 0;
		return;
	}
	dprintf(5, (ddt, "give up\n"));
	n = ((HEADER *)qp->q_cmsg ? qp->q_cmsglen : qp->q_msglen);
	hp->id = qp->q_id;
	hp->qr = 1;
	hp->ra = (NoRecurse == 0);
	hp->rd = 1;
	hp->rcode = SERVFAIL;
#ifdef DEBUG
	if (debug >= 10)
		fp_nquery(qp->q_msg, n, ddt);
#endif
	if (send_msg((u_char *)hp, n, qp)) {
		dprintf(1, (ddt, "gave up retry(x%lx) nsid=%d id=%d\n",
			    (u_long)qp, ntohs(qp->q_nsid), ntohs(qp->q_id)));
	}
	nameserIncr(qp->q_from.sin_addr, nssSentFail);
	qremove(qp);
	return;

found:
	if (qp->q_fwd == 0 && qp->q_addr[n].nretry == 0)
		qp->q_addr[n].stime = tt;
	qp->q_curaddr = n;
	hp = (HEADER *)qp->q_msg;
	hp->rd = (qp->q_fwd ? 1 : 0);
	nsa = Q_NEXTADDR(qp, n);
	dprintf(1, (ddt,
		    "%s(addr=%d n=%d) -> [%s].%d ds=%d nsid=%d id=%d %dms\n",
		    (qp->q_fwd ? "reforw" : "resend"),
		    n, qp->q_addr[n].nretry,
		    inet_ntoa(nsa->sin_addr),
		    ntohs(nsa->sin_port), ds,
		    ntohs(qp->q_nsid), ntohs(qp->q_id),
		    (qp->q_addr[n].nsdata != 0)
			? qp->q_addr[n].nsdata->d_nstime
			: (-1)));
#ifdef DEBUG
	if (debug >= 10)
		fp_nquery(qp->q_msg, qp->q_msglen, ddt);
#endif
	/* NOSTRICT */
	if (sendto(ds, (char*)qp->q_msg, qp->q_msglen, 0,
	    (struct sockaddr *)nsa,
	    sizeof(struct sockaddr_in)) < 0) {
		dprintf(3, (ddt, "error resending msg errno=%d\n", errno));
	}
	hp->rd = 1;	/* leave set to 1 for dup detection */
	nameserIncr(nsa->sin_addr, nssSentDupQ);
	unsched(qp);
#ifdef SLAVE_FORWARD
	if(forward_only)
		schedretry(qp, (time_t)slave_retry);
	else
#endif /* SLAVE_FORWARD */
	schedretry(qp, qp->q_fwd ? (2*RETRYBASE) : retrytime(qp));
}

/*
 * Compute retry time for the next server for a query.
 * Use a minimum time of RETRYBASE (4 sec.) or twice the estimated
 * service time; * back off exponentially on retries, but place a 45-sec.
 * ceiling on retry times for now.  (This is because we don't hold a reference
 * on servers or their addresses, and we have to finish before they time out.)
 */
time_t
retrytime(qp)
	struct qinfo *qp;
{
	time_t t, u, v;
	struct qserv *ns = &qp->q_addr[qp->q_curaddr];

	if (ns->nsdata != NULL)
		t = (time_t) MAX(RETRYBASE, 2 * ns->nsdata->d_nstime / 1000);
	else
		t = (time_t) RETRYBASE;
	u = t << ns->nretry;
	v = MIN(u, RETRY_TIMEOUT);	/* max. retry timeout for now */
	dprintf(3, (ddt, "retrytime: nstime%ldms t%ld nretry%ld u%ld : v%ld\n",
		    ns->nsdata ?(long)(ns->nsdata->d_nstime / 1000) :(long)-1,
		    (long)t, (long)ns->nretry, (long)u, (long)v));
	return (v);
}

void
qflush()
{
	while (nsqhead)
		qremove(nsqhead);
	nsqhead = QINFO_NULL;
}

void
qremove(qp)
	register struct qinfo *qp;
{
	dprintf(3, (ddt, "qremove(x%lx)\n", (u_long)qp));

	if (qp->q_flags & Q_ZSERIAL)
		qserial_answer(qp, 0);
	unsched(qp);
	qfree(qp);
}

#if defined(__STDC__) || defined(__GNUC__)
struct qinfo *
qfindid(u_int16_t id)
#else
struct qinfo *
qfindid(id)
	register u_int16_t id;
#endif
{
	register struct qinfo *qp;

	dprintf(3, (ddt, "qfindid(%d)\n", ntohs(id)));
	for (qp = nsqhead; qp!=QINFO_NULL; qp = qp->q_link) {
		if (qp->q_nsid == id)
			return(qp);
	}
	dprintf(5, (ddt, "qp not found\n"));
	return (NULL);
}

struct qinfo *
#ifdef DMALLOC
qnew_tagged(file, line)
	char *file;
	int line;
#else
qnew()
#endif
{
	register struct qinfo *qp;

	qp = (struct qinfo *)
#ifdef DMALLOC
	    dcalloc(file, line, 1, sizeof(struct qinfo));
#else
	    calloc(1, sizeof(struct qinfo));
#endif
	if (qp == NULL) {
		dprintf(5, (ddt, "qnew: calloc error\n"));
		syslog(LOG_ERR, "forw: %m");
		exit(12);
	}
	dprintf(5, (ddt, "qnew(x%lx)\n", (u_long)qp));
#ifdef BIND_NOTIFY
	qp->q_notifyzone = DB_Z_CACHE;
#endif
	qp->q_link = nsqhead;
	nsqhead = qp;
	return (qp);
}

void
qfree(qp)
	struct qinfo *qp;
{
	register struct qinfo *np;
	register struct databuf *dp;
#ifdef	DATUMREFCNT
	int i;
#endif

	dprintf(3, (ddt, "Qfree(x%lx)\n", (u_long)qp));
	if (qp->q_next)
		dprintf(1, (ddt, "WARNING: qfree of linked ptr x%lx\n",
			    (u_long)qp));
	if (qp->q_msg)
	 	free(qp->q_msg);
 	if (qp->q_cmsg)
 		free(qp->q_cmsg);
#ifdef	DATUMREFCNT
	for (i = 0 ; i < (int)qp->q_naddr ; i++) {
		dp = qp->q_addr[i].ns;
		if (dp)
			if (--(dp->d_rcnt)) {
				dprintf(3, (ddt, "qfree: ns %s rcnt %d\n",
						dp->d_data,
						dp->d_rcnt));
			} else {
				dprintf(3, (ddt, "qfree: ns %s rcnt %d delayed\n",
						dp->d_data,
						dp->d_rcnt));
				free((char*)dp);
			}
		dp = qp->q_addr[i].nsdata;
		if (dp)
			if ((--(dp->d_rcnt))) {
				dprintf(3, (ddt, "qfree: nsdata %08.8X rcnt %d\n",
					*(int32_t *)(dp->d_data),
					dp->d_rcnt));
			} else {
				dprintf(3, (ddt, "qfree: nsdata %08.8X rcnt %d delayed\n",
					*(int32_t *)(dp->d_data),
					dp->d_rcnt));
			free((char*)dp);
			}
	}
#endif
	if( nsqhead == qp )  {
		nsqhead = qp->q_link;
	} else {
		for( np=nsqhead; np->q_link != QINFO_NULL; np = np->q_link )  {
			if( np->q_link != qp )  continue;
			np->q_link = qp->q_link;	/* dequeue */
			break;
		}
	}
	free((char *)qp);
}