xref: /openbsd-src/usr.sbin/bgpd/mrt.c (revision 0d4ceb41004f299c91661fe76cfd13f053a31300)

/*	$OpenBSD: mrt.c,v 1.123 2024/12/16 16:10:10 claudio Exp $ */

/*
 * Copyright (c) 2003, 2004 Claudio Jeker <claudio@openbsd.org>
 *
 * 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.
 *
 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR 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.
 */

#include <sys/types.h>
#include <sys/queue.h>

#include <errno.h>
#include <fcntl.h>
#include <limits.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <unistd.h>

#include "bgpd.h"
#include "rde.h"
#include "session.h"

#include "mrt.h"
#include "log.h"

static int	mrt_attr_dump(struct ibuf *, struct rde_aspath *,
		    struct rde_community *, struct bgpd_addr *, int);
static int	mrt_dump_entry_mp(struct mrt *, struct prefix *, uint16_t,
		    struct rde_peer*);
static int	mrt_dump_entry(struct mrt *, struct prefix *, uint16_t,
		    struct rde_peer*);
static int	mrt_dump_entry_v2(struct mrt *, struct rib_entry *, uint32_t);
static int	mrt_dump_peer(struct ibuf *, struct rde_peer *);
static int	mrt_dump_hdr_se(struct ibuf **, struct peer *, uint16_t,
		    uint16_t, uint32_t, int);
static int	mrt_dump_hdr_rde(struct ibuf **, uint16_t type, uint16_t,
		    uint32_t);
static int	mrt_open(struct mrt *, time_t);

#define RDEIDX		0
#define SEIDX		1
#define TYPE2IDX(x)	((x == MRT_TABLE_DUMP ||			\
			    x == MRT_TABLE_DUMP_MP ||			\
			    x == MRT_TABLE_DUMP_V2) ? RDEIDX : SEIDX	\
			)

static uint8_t
mrt_update_msg_guess_aid(struct ibuf *pkg)
{
	struct ibuf buf;
	uint16_t wlen, alen, len, afi;
	uint8_t type, flags, aid, safi;

	ibuf_from_ibuf(&buf, pkg);

	if (ibuf_skip(&buf, MSGSIZE_HEADER) == -1 ||
	    ibuf_get_n16(&buf, &wlen) == -1)
		goto bad;

	if (wlen > 0) {
		/* UPDATE has withdraw routes, therefore IPv4 */
		return AID_INET;
	}

	if (ibuf_get_n16(&buf, &alen) == -1)
		goto bad;

	if (alen < ibuf_size(&buf)) {
		/* UPDATE has NLRI prefixes, therefore IPv4 */
		return AID_INET;
	}

	if (wlen == 0 && alen == 0) {
		/* UPDATE is an IPv4 EoR marker */
		return AID_INET;
	}

	/* bad attribute length */
	if (alen > ibuf_size(&buf))
		goto bad;

	/* try to extract AFI/SAFI from the MP attributes */
	while (ibuf_size(&buf) > 0) {
		if (ibuf_get_n8(&buf, &flags) == -1 ||
		    ibuf_get_n8(&buf, &type) == -1)
			goto bad;
		if (flags & ATTR_EXTLEN) {
			if (ibuf_get_n16(&buf, &len) == -1)
				goto bad;
		} else {
			uint8_t tmp;
			if (ibuf_get_n8(&buf, &tmp) == -1)
				goto bad;
			len = tmp;
		}
		if (len > ibuf_size(&buf))
			goto bad;

		if (type == ATTR_MP_REACH_NLRI ||
		    type == ATTR_MP_UNREACH_NLRI) {
			if (ibuf_get_n16(&buf, &afi) == -1 ||
			    ibuf_get_n8(&buf, &safi) == -1)
				goto bad;
			if (afi2aid(afi, safi, &aid) == -1)
				goto bad;
			return aid;
		}
		if (ibuf_skip(&buf, len) == -1)
			goto bad;
	}

bad:
	return AID_UNSPEC;
}

static uint16_t
mrt_bgp_msg_subtype(struct mrt *mrt, struct ibuf *pkg, struct peer *peer,
    enum msg_type msgtype, int in)
{
	uint16_t subtype = BGP4MP_MESSAGE;
	uint8_t aid, mask;

	if (peer->capa.neg.as4byte)
		subtype = BGP4MP_MESSAGE_AS4;

	if (msgtype != MSG_UPDATE)
		return subtype;

	/*
	 * RFC8050 adjust types for add-path enabled sessions.
	 * It is necessary to extract the AID from UPDATES to decide
	 * if the add-path types are needed or not. The ADDPATH
	 * subtypes only matter for BGP UPDATES.
	 */

	mask = in ? CAPA_AP_RECV : CAPA_AP_SEND;
	/* only guess if add-path could be active */
	if (peer->capa.neg.add_path[0] & mask) {
		aid = mrt_update_msg_guess_aid(pkg);
		if (aid != AID_UNSPEC &&
		    (peer->capa.neg.add_path[aid] & mask)) {
			if (peer->capa.neg.as4byte)
				subtype = BGP4MP_MESSAGE_AS4_ADDPATH;
			else
				subtype = BGP4MP_MESSAGE_ADDPATH;
		}
	}

	return subtype;
}

void
mrt_dump_bgp_msg(struct mrt *mrt, struct ibuf *pkg, struct peer *peer,
    enum msg_type msgtype)
{
	struct ibuf	*buf;
	int		 in = 0;
	uint16_t	 subtype = BGP4MP_MESSAGE;

	/* get the direction of the message to swap address and AS fields */
	if (mrt->type == MRT_ALL_IN || mrt->type == MRT_UPDATE_IN)
		in = 1;

	subtype = mrt_bgp_msg_subtype(mrt, pkg, peer, msgtype, in);

	if (mrt_dump_hdr_se(&buf, peer, MSG_PROTOCOL_BGP4MP_ET, subtype,
	    ibuf_size(pkg), in) == -1)
		goto fail;

	if (ibuf_add_ibuf(buf, pkg) == -1)
		goto fail;

	ibuf_close(mrt->wbuf, buf);
	return;

fail:
	log_warn("%s: ibuf error", __func__);
	ibuf_free(buf);
}

void
mrt_dump_state(struct mrt *mrt, uint16_t old_state, uint16_t new_state,
    struct peer *peer)
{
	struct ibuf	*buf;
	uint16_t	 subtype = BGP4MP_STATE_CHANGE;

	if (peer->capa.neg.as4byte)
		subtype = BGP4MP_STATE_CHANGE_AS4;

	if (mrt_dump_hdr_se(&buf, peer, MSG_PROTOCOL_BGP4MP_ET, subtype,
	    2 * sizeof(short), 0) == -1)
		goto fail;

	if (ibuf_add_n16(buf, old_state) == -1)
		goto fail;
	if (ibuf_add_n16(buf, new_state) == -1)
		goto fail;

	ibuf_close(mrt->wbuf, buf);
	return;

fail:
	log_warn("%s: ibuf error", __func__);
	ibuf_free(buf);
}

static int
mrt_attr_dump(struct ibuf *buf, struct rde_aspath *a, struct rde_community *c,
    struct bgpd_addr *nexthop, int v2)
{
	struct attr	*oa;
	u_char		*pdata;
	uint32_t	 tmp;
	int		 neednewpath = 0;
	uint16_t	 plen, afi;
	uint8_t		 l, safi;

	/* origin */
	if (attr_writebuf(buf, ATTR_WELL_KNOWN, ATTR_ORIGIN,
	    &a->origin, 1) == -1)
		return (-1);

	/* aspath */
	plen = aspath_length(a->aspath);
	pdata = aspath_dump(a->aspath);

	if (!v2)
		pdata = aspath_deflate(pdata, &plen, &neednewpath);
	if (attr_writebuf(buf, ATTR_WELL_KNOWN, ATTR_ASPATH, pdata,
	    plen) == -1) {
		if (!v2)
			free(pdata);
		return (-1);
	}
	if (!v2)
		free(pdata);

	if (nexthop && nexthop->aid == AID_INET) {
		/* nexthop, already network byte order */
		if (attr_writebuf(buf, ATTR_WELL_KNOWN, ATTR_NEXTHOP,
		    &nexthop->v4.s_addr, 4) ==	-1)
			return (-1);
	}

	/* MED, non transitive */
	if (a->med != 0) {
		tmp = htonl(a->med);
		if (attr_writebuf(buf, ATTR_OPTIONAL, ATTR_MED, &tmp, 4) == -1)
			return (-1);
	}

	/* local preference */
	tmp = htonl(a->lpref);
	if (attr_writebuf(buf, ATTR_WELL_KNOWN, ATTR_LOCALPREF, &tmp, 4) == -1)
		return (-1);

	/* communities */
	if (community_writebuf(c, ATTR_COMMUNITIES, 0, buf) == -1 ||
	    community_writebuf(c, ATTR_EXT_COMMUNITIES, 0, buf) == -1 ||
	    community_writebuf(c, ATTR_LARGE_COMMUNITIES, 0, buf) == -1)
		return (-1);

	/* dump all other path attributes without modification */
	for (l = 0; l < a->others_len; l++) {
		if ((oa = a->others[l]) == NULL)
			break;
		if (attr_writebuf(buf, oa->flags, oa->type,
		    oa->data, oa->len) == -1)
			return (-1);
	}

	if (nexthop && nexthop->aid != AID_INET) {
		struct ibuf *nhbuf;

		if ((nhbuf = ibuf_dynamic(0, UCHAR_MAX)) == NULL)
			return (-1);
		if (!v2) {
			if (aid2afi(nexthop->aid, &afi, &safi))
				goto fail;
			if (ibuf_add_n16(nhbuf, afi) == -1)
				goto fail;
			if (ibuf_add_n8(nhbuf, safi) == -1)
				goto fail;
		}
		switch (nexthop->aid) {
		case AID_INET6:
			if (ibuf_add_n8(nhbuf, sizeof(struct in6_addr)) == -1)
				goto fail;
			if (ibuf_add(nhbuf, &nexthop->v6,
			    sizeof(struct in6_addr)) == -1)
				goto fail;
			break;
		case AID_VPN_IPv4:
			if (ibuf_add_n8(nhbuf, sizeof(uint64_t) +
			    sizeof(struct in_addr)) == -1)
				goto fail;
			if (ibuf_add_n64(nhbuf, 0) == -1) /* set RD to 0 */
				goto fail;
			if (ibuf_add(nhbuf, &nexthop->v4,
			    sizeof(nexthop->v4)) == -1)
				goto fail;
			break;
		case AID_VPN_IPv6:
			if (ibuf_add_n8(nhbuf, sizeof(uint64_t) +
			    sizeof(struct in6_addr)) == -1)
				goto fail;
			if (ibuf_add_n64(nhbuf, 0) == -1) /* set RD to 0 */
				goto fail;
			if (ibuf_add(nhbuf, &nexthop->v6,
			    sizeof(nexthop->v6)) == -1)
				goto fail;
			break;
		}
		if (!v2)
			if (ibuf_add_n8(nhbuf, 0) == -1)
				goto fail;
		if (attr_writebuf(buf, ATTR_OPTIONAL, ATTR_MP_REACH_NLRI,
		    ibuf_data(nhbuf), ibuf_size(nhbuf)) == -1) {
fail:
			ibuf_free(nhbuf);
			return (-1);
		}
		ibuf_free(nhbuf);
	}

	if (neednewpath) {
		pdata = aspath_prepend(a->aspath, rde_local_as(), 0, &plen);
		if (plen != 0)
			if (attr_writebuf(buf, ATTR_OPTIONAL|ATTR_TRANSITIVE,
			    ATTR_AS4_PATH, pdata, plen) == -1) {
				free(pdata);
				return (-1);
			}
		free(pdata);
	}

	return (0);
}

static int
mrt_dump_entry_mp(struct mrt *mrt, struct prefix *p, uint16_t snum,
    struct rde_peer *peer)
{
	struct ibuf	*buf, *hbuf = NULL, *h2buf = NULL;
	struct nexthop	*n;
	struct bgpd_addr nexthop, *nh;
	uint16_t	 len;
	uint8_t		 aid;

	if ((buf = ibuf_dynamic(0, MAX_PKTSIZE)) == NULL) {
		log_warn("mrt_dump_entry_mp: ibuf_dynamic");
		return (-1);
	}

	if (mrt_attr_dump(buf, prefix_aspath(p), prefix_communities(p),
	    NULL, 0) == -1)
		goto fail;
	len = ibuf_size(buf);

	if ((h2buf = ibuf_dynamic(MRT_BGP4MP_IPv4_HEADER_SIZE +
	    MRT_BGP4MP_IPv4_ENTRY_SIZE, MRT_BGP4MP_IPv6_HEADER_SIZE +
	    MRT_BGP4MP_IPv6_ENTRY_SIZE + MRT_BGP4MP_MAX_PREFIXLEN)) == NULL)
		goto fail;

	if (ibuf_add_n16(h2buf, peer->conf.local_short_as) == -1)
		goto fail;
	if (ibuf_add_n16(h2buf, peer->short_as) == -1)
		goto fail;
	if (ibuf_add_n16(h2buf, /* ifindex */ 0) == -1)
		goto fail;

	/* XXX is this for peer self? */
	aid = peer->remote_addr.aid == AID_UNSPEC ? p->pt->aid :
	    peer->remote_addr.aid;
	switch (aid) {
	case AID_INET:
	case AID_VPN_IPv4:
		if (ibuf_add_n16(h2buf, AFI_IPv4) == -1)
			goto fail;
		if (ibuf_add(h2buf, &peer->local_v4_addr.v4,
		    sizeof(peer->local_v4_addr.v4)) == -1 ||
		    ibuf_add(h2buf, &peer->remote_addr.v4,
		    sizeof(peer->remote_addr.v4)) == -1)
			goto fail;
		break;
	case AID_INET6:
	case AID_VPN_IPv6:
		if (ibuf_add_n16(h2buf, AFI_IPv6) == -1)
			goto fail;
		if (ibuf_add(h2buf, &peer->local_v6_addr.v6,
		    sizeof(peer->local_v6_addr.v6)) == -1 ||
		    ibuf_add(h2buf, &peer->remote_addr.v6,
		    sizeof(peer->remote_addr.v6)) == -1)
			goto fail;
		break;
	default:
		log_warnx("king bula found new AF %d in %s", aid, __func__);
		goto fail;
	}

	if (ibuf_add_n16(h2buf, 0) == -1)		/* view */
		goto fail;
	if (ibuf_add_n16(h2buf, 1) == -1)		/* status */
		goto fail;
	/* originated timestamp */
	if (ibuf_add_n32(h2buf, time(NULL) - (getmonotime() -
	    p->lastchange)) == -1)
		goto fail;

	n = prefix_nexthop(p);
	if (n == NULL) {
		memset(&nexthop, 0, sizeof(struct bgpd_addr));
		nexthop.aid = p->pt->aid;
		nh = &nexthop;
	} else
		nh = &n->exit_nexthop;

	switch (p->pt->aid) {
	case AID_INET:
		if (ibuf_add_n16(h2buf, AFI_IPv4) == -1)	/* afi */
			goto fail;
		if (ibuf_add_n8(h2buf, SAFI_UNICAST) == -1)	/* safi */
			goto fail;
		if (ibuf_add_n8(h2buf, 4) == -1)		/* nhlen */
			goto fail;
		if (ibuf_add(h2buf, &nh->v4, sizeof(nh->v4)) == -1)
			goto fail;
		break;
	case AID_INET6:
		if (ibuf_add_n16(h2buf, AFI_IPv6) == -1)	/* afi */
			goto fail;
		if (ibuf_add_n8(h2buf, SAFI_UNICAST) == -1)	/* safi */
			goto fail;
		if (ibuf_add_n8(h2buf, 16) == -1)		/* nhlen */
			goto fail;
		if (ibuf_add(h2buf, &nh->v6, sizeof(nh->v6)) == -1)
			goto fail;
		break;
	case AID_VPN_IPv4:
		if (ibuf_add_n16(h2buf, AFI_IPv4) == -1)	/* afi */
			goto fail;
		if (ibuf_add_n8(h2buf, SAFI_MPLSVPN) == -1)	/* safi */
			goto fail;
		if (ibuf_add_n8(h2buf, sizeof(uint64_t) +
		    sizeof(struct in_addr)) == -1)
			goto fail;
		if (ibuf_add_n64(h2buf, 0) == -1)	/* set RD to 0 */
			goto fail;
		if (ibuf_add(h2buf, &nh->v4, sizeof(nh->v4)) == -1)
			goto fail;
		break;
	case AID_VPN_IPv6:
		if (ibuf_add_n16(h2buf, AFI_IPv6) == -1)	/* afi */
			goto fail;
		if (ibuf_add_n8(h2buf, SAFI_MPLSVPN) == -1)	/* safi */
			goto fail;
		if (ibuf_add_n8(h2buf, sizeof(uint64_t) +
		    sizeof(struct in6_addr)) == -1)
			goto fail;
		if (ibuf_add_n64(h2buf, 0) == -1)	/* set RD to 0 */
			goto fail;
		if (ibuf_add(h2buf, &nh->v6, sizeof(nh->v6)) == -1)
			goto fail;
		break;
	case AID_FLOWSPECv4:
	case AID_FLOWSPECv6:
		if (p->pt->aid == AID_FLOWSPECv4) {
			if (ibuf_add_n16(h2buf, AFI_IPv4) == -1) /* afi */
				goto fail;
		} else {
			if (ibuf_add_n16(h2buf, AFI_IPv6) == -1) /* afi */
				goto fail;
		}
		if (ibuf_add_n8(h2buf, SAFI_FLOWSPEC) == -1)	/* safi */
			goto fail;
		if (ibuf_add_n8(h2buf, 0) == -1)		/* nhlen */
			goto fail;
		break;
	default:
		log_warnx("king bula found new AF in %s", __func__);
		goto fail;
	}

	if (pt_writebuf(h2buf, p->pt, 0, 0, 0) == -1)
		goto fail;

	if (ibuf_add_n16(h2buf, len) == -1)
		goto fail;
	len += ibuf_size(h2buf);

	if (mrt_dump_hdr_rde(&hbuf, MSG_PROTOCOL_BGP4MP, BGP4MP_ENTRY,
	    len) == -1)
		goto fail;

	ibuf_close(mrt->wbuf, hbuf);
	ibuf_close(mrt->wbuf, h2buf);
	ibuf_close(mrt->wbuf, buf);

	return (len + MRT_HEADER_SIZE);

fail:
	log_warn("%s: ibuf error", __func__);
	ibuf_free(hbuf);
	ibuf_free(h2buf);
	ibuf_free(buf);
	return (-1);
}

static int
mrt_dump_entry(struct mrt *mrt, struct prefix *p, uint16_t snum,
    struct rde_peer *peer)
{
	struct ibuf	*buf, *hbuf = NULL;
	struct nexthop	*nexthop;
	struct bgpd_addr addr, *nh;
	size_t		 len;
	uint16_t	 subtype;
	uint8_t		 dummy;

	if (p->pt->aid != peer->remote_addr.aid &&
	    p->pt->aid != AID_INET && p->pt->aid != AID_INET6)
		/* only able to dump pure IPv4/IPv6 */
		return (0);

	if ((buf = ibuf_dynamic(0, MAX_PKTSIZE)) == NULL) {
		log_warn("mrt_dump_entry: ibuf_dynamic");
		return (-1);
	}

	nexthop = prefix_nexthop(p);
	if (nexthop == NULL) {
		memset(&addr, 0, sizeof(struct bgpd_addr));
		addr.aid = p->pt->aid;
		nh = &addr;
	} else
		nh = &nexthop->exit_nexthop;
	if (mrt_attr_dump(buf, prefix_aspath(p), prefix_communities(p),
	    nh, 0) == -1)
		goto fail;

	len = ibuf_size(buf);
	aid2afi(p->pt->aid, &subtype, &dummy);
	if (mrt_dump_hdr_rde(&hbuf, MSG_TABLE_DUMP, subtype, len) == -1)
		goto fail;

	if (ibuf_add_n16(hbuf, 0) == -1)
		goto fail;
	if (ibuf_add_n16(hbuf, snum) == -1)
		goto fail;

	pt_getaddr(p->pt, &addr);
	switch (p->pt->aid) {
	case AID_INET:
		if (ibuf_add(hbuf, &addr.v4, sizeof(addr.v4)) == -1)
			goto fail;
		break;
	case AID_INET6:
		if (ibuf_add(hbuf, &addr.v6, sizeof(addr.v6)) == -1)
			goto fail;
		break;
	}
	if (ibuf_add_n8(hbuf, p->pt->prefixlen) == -1)
		goto fail;

	if (ibuf_add_n8(hbuf, 1) == -1)		/* state */
		goto fail;
	/* originated timestamp */
	if (ibuf_add_n32(hbuf, time(NULL) - (getmonotime() -
	    p->lastchange)) == -1)
		goto fail;
	switch (p->pt->aid) {
	case AID_INET:
		if (ibuf_add(hbuf, &peer->remote_addr.v4,
		    sizeof(peer->remote_addr.v4)) == -1)
			goto fail;
		break;
	case AID_INET6:
		if (ibuf_add(hbuf, &peer->remote_addr.v6,
		    sizeof(peer->remote_addr.v6)) == -1)
			goto fail;
		break;
	}
	if (ibuf_add_n16(hbuf, peer->short_as) == -1)
		goto fail;
	if (ibuf_add_n16(hbuf, len) == -1)
		goto fail;

	ibuf_close(mrt->wbuf, hbuf);
	ibuf_close(mrt->wbuf, buf);

	return (len + MRT_HEADER_SIZE);

fail:
	log_warn("%s: ibuf error", __func__);
	ibuf_free(hbuf);
	ibuf_free(buf);
	return (-1);
}

static int
mrt_dump_entry_v2_rib(struct rib_entry *re, struct ibuf **nb, struct ibuf **apb,
    uint16_t *np, uint16_t *app)
{
	struct bgpd_addr addr;
	struct ibuf *buf = NULL, **bp;
	struct ibuf *tbuf = NULL;
	struct prefix *p;
	int addpath;

	*np = 0;
	*app = 0;

	TAILQ_FOREACH(p, &re->prefix_h, entry.list.rib) {
		struct nexthop		*nexthop;
		struct bgpd_addr	*nh;

		addpath = peer_has_add_path(prefix_peer(p), re->prefix->aid,
		    CAPA_AP_RECV);

		if (addpath) {
			bp = apb;
			*app += 1;
		} else {
			bp = nb;
			*np += 1;
		}
		if ((buf = *bp) == NULL) {
			if ((buf = ibuf_dynamic(0, UINT_MAX)) == NULL)
				goto fail;
			*bp = buf;
		}

		nexthop = prefix_nexthop(p);
		if (nexthop == NULL) {
			memset(&addr, 0, sizeof(struct bgpd_addr));
			addr.aid = re->prefix->aid;
			nh = &addr;
		} else
			nh = &nexthop->exit_nexthop;

		if (ibuf_add_n16(buf, prefix_peer(p)->mrt_idx) == -1)
			goto fail;
		/* originated timestamp */
		if (ibuf_add_n32(buf, time(NULL) - (getmonotime() -
		    p->lastchange)) == -1)
			goto fail;

		/* RFC8050: path-id if add-path is used */
		if (addpath)
			if (ibuf_add_n32(buf, p->path_id) == -1)
				goto fail;

		if ((tbuf = ibuf_dynamic(0, MAX_PKTSIZE)) == NULL)
			goto fail;
		if (mrt_attr_dump(tbuf, prefix_aspath(p), prefix_communities(p),
		    nh, 1) == -1)
			goto fail;
		if (ibuf_add_n16(buf, ibuf_size(tbuf)) == -1)
			goto fail;
		if (ibuf_add_ibuf(buf, tbuf) == -1)
			goto fail;
		ibuf_free(tbuf);
		tbuf = NULL;
	}

	return 0;

fail:
	ibuf_free(tbuf);
	return -1;
}

static int
mrt_dump_entry_v2(struct mrt *mrt, struct rib_entry *re, uint32_t snum)
{
	struct ibuf	*hbuf = NULL, *nbuf = NULL, *apbuf = NULL, *pbuf;
	size_t		 hlen, len;
	uint16_t	 subtype, apsubtype, nump, apnump, afi;
	uint8_t		 safi;

	if ((pbuf = ibuf_dynamic(0, UINT_MAX)) == NULL) {
		log_warn("%s: ibuf_dynamic", __func__);
		return -1;
	}

	switch (re->prefix->aid) {
	case AID_INET:
		subtype = MRT_DUMP_V2_RIB_IPV4_UNICAST;
		apsubtype = MRT_DUMP_V2_RIB_IPV4_UNICAST_ADDPATH;
		break;
	case AID_INET6:
		subtype = MRT_DUMP_V2_RIB_IPV6_UNICAST;
		apsubtype = MRT_DUMP_V2_RIB_IPV6_UNICAST_ADDPATH;
		break;
	default:
		/*
		 * XXX The RFC defined the format for this type differently
		 * and it is prohibitly expensive to implement that format.
		 * Instead do what gobgp does and encode it like the other
		 * types.
		 */
		subtype = MRT_DUMP_V2_RIB_GENERIC;
		apsubtype = MRT_DUMP_V2_RIB_GENERIC_ADDPATH;
		aid2afi(re->prefix->aid, &afi, &safi);

		/* first add 3-bytes AFI/SAFI */
		if (ibuf_add_n16(pbuf, afi) == -1)
			goto fail;
		if (ibuf_add_n8(pbuf, safi) == -1)
			goto fail;
		break;
	}

	if (pt_writebuf(pbuf, re->prefix, 0, 0, 0) == -1)
		goto fail;

	hlen = sizeof(snum) + sizeof(nump) + ibuf_size(pbuf);

	if (mrt_dump_entry_v2_rib(re, &nbuf, &apbuf, &nump, &apnump))
		goto fail;

	if (nump > 0) {
		len = ibuf_size(nbuf) + hlen;
		if (mrt_dump_hdr_rde(&hbuf, MSG_TABLE_DUMP_V2, subtype,
		    len) == -1)
			goto fail;

		if (ibuf_add_n32(hbuf, snum) == -1)
			goto fail;
		if (ibuf_add_ibuf(hbuf, pbuf) == -1)
			goto fail;
		if (ibuf_add_n16(hbuf, nump) == -1)
			goto fail;

		ibuf_close(mrt->wbuf, hbuf);
		ibuf_close(mrt->wbuf, nbuf);
		hbuf = NULL;
		nbuf = NULL;
	}

	if (apnump > 0) {
		len = ibuf_size(apbuf) + hlen;
		if (mrt_dump_hdr_rde(&hbuf, MSG_TABLE_DUMP_V2, apsubtype,
		    len) == -1)
			goto fail;

		if (ibuf_add_n32(hbuf, snum) == -1)
			goto fail;
		if (ibuf_add_ibuf(hbuf, pbuf) == -1)
			goto fail;
		if (ibuf_add_n16(hbuf, apnump) == -1)
			goto fail;

		ibuf_close(mrt->wbuf, hbuf);
		ibuf_close(mrt->wbuf, apbuf);
		hbuf = NULL;
		apbuf = NULL;
	}

	ibuf_free(pbuf);
	return (0);
fail:
	log_warn("%s: ibuf error", __func__);
	ibuf_free(apbuf);
	ibuf_free(nbuf);
	ibuf_free(hbuf);
	ibuf_free(pbuf);
	return (-1);
}

struct cb_arg {
	struct ibuf	*buf;
	int		 nump;
};

static void
mrt_dump_v2_hdr_peer(struct rde_peer *peer, void *arg)
{
	struct cb_arg *a = arg;

	if (a->nump == -1)
		return;
	peer->mrt_idx = a->nump;
	if (mrt_dump_peer(a->buf, peer) == -1) {
		a->nump = -1;
		return;
	}
	a->nump++;
}

int
mrt_dump_v2_hdr(struct mrt *mrt, struct bgpd_config *conf)
{
	struct ibuf	*buf, *hbuf = NULL;
	size_t		 len, off;
	uint16_t	 nlen, nump;
	struct cb_arg	 arg;

	if ((buf = ibuf_dynamic(0, UINT_MAX)) == NULL) {
		log_warn("%s: ibuf_dynamic", __func__);
		return (-1);
	}

	if (ibuf_add_n32(buf, conf->bgpid) == -1)
		goto fail;
	nlen = strlen(mrt->rib);
	if (nlen > 0)
		nlen += 1;
	if (ibuf_add_n16(buf, nlen) == -1)
		goto fail;
	if (ibuf_add(buf, mrt->rib, nlen) == -1)
		goto fail;

	off = ibuf_size(buf);
	if (ibuf_add_zero(buf, sizeof(nump)) == -1)
		goto fail;
	arg.nump = 0;
	arg.buf = buf;
	peer_foreach(mrt_dump_v2_hdr_peer, &arg);
	if (arg.nump == -1)
		goto fail;

	if (ibuf_set_n16(buf, off, arg.nump) == -1)
		goto fail;

	len = ibuf_size(buf);
	if (mrt_dump_hdr_rde(&hbuf, MSG_TABLE_DUMP_V2,
	    MRT_DUMP_V2_PEER_INDEX_TABLE, len) == -1)
		goto fail;

	ibuf_close(mrt->wbuf, hbuf);
	ibuf_close(mrt->wbuf, buf);

	return (0);
fail:
	log_warn("%s: ibuf error", __func__);
	ibuf_free(hbuf);
	ibuf_free(buf);
	return (-1);
}

static int
mrt_dump_peer(struct ibuf *buf, struct rde_peer *peer)
{
	uint8_t	type = 0;

	if (peer->capa.as4byte)
		type |= MRT_DUMP_V2_PEER_BIT_A;
	if (peer->remote_addr.aid == AID_INET6)
		type |= MRT_DUMP_V2_PEER_BIT_I;

	if (ibuf_add_n8(buf, type) == -1)
		goto fail;
	if (ibuf_add_n32(buf, peer->remote_bgpid) == -1)
		goto fail;

	switch (peer->remote_addr.aid) {
	case AID_INET:
		if (ibuf_add(buf, &peer->remote_addr.v4,
		    sizeof(peer->remote_addr.v4)) == -1)
			goto fail;
		break;
	case AID_INET6:
		if (ibuf_add(buf, &peer->remote_addr.v6,
		    sizeof(peer->remote_addr.v6)) == -1)
			goto fail;
		break;
	case AID_UNSPEC: /* XXX special handling for peerself? */
		if (ibuf_add_n32(buf, 0) == -1)
			goto fail;
		break;
	default:
		log_warnx("king bula found new AF in %s", __func__);
		goto fail;
	}

	if (peer->capa.as4byte) {
		if (ibuf_add_n32(buf, peer->conf.remote_as) == -1)
			goto fail;
	} else {
		if (ibuf_add_n16(buf, peer->short_as) == -1)
			goto fail;
	}
	return (0);
fail:
	log_warn("%s: ibuf error", __func__);
	return (-1);
}

void
mrt_dump_upcall(struct rib_entry *re, void *ptr)
{
	struct mrt		*mrtbuf = ptr;
	struct prefix		*p;

	if (mrtbuf->type == MRT_TABLE_DUMP_V2) {
		mrt_dump_entry_v2(mrtbuf, re, mrtbuf->seqnum++);
		return;
	}

	/*
	 * dump all prefixes even the inactive ones. That is the way zebra
	 * dumps the table so we do the same. If only the active route should
	 * be dumped p should be set to p = pt->active.
	 */
	TAILQ_FOREACH(p, &re->prefix_h, entry.list.rib) {
		if (mrtbuf->type == MRT_TABLE_DUMP)
			mrt_dump_entry(mrtbuf, p, mrtbuf->seqnum++,
			    prefix_peer(p));
		else
			mrt_dump_entry_mp(mrtbuf, p, mrtbuf->seqnum++,
			    prefix_peer(p));
	}
}

void
mrt_done(struct mrt *mrtbuf)
{
	mrtbuf->state = MRT_STATE_REMOVE;
}

static int
mrt_dump_hdr_se(struct ibuf ** bp, struct peer *peer, uint16_t type,
    uint16_t subtype, uint32_t len, int swap)
{
	struct timespec	time;

	if ((*bp = ibuf_dynamic(MRT_ET_HEADER_SIZE, MRT_ET_HEADER_SIZE +
	    MRT_BGP4MP_AS4_IPv6_HEADER_SIZE + len)) == NULL)
		return (-1);

	clock_gettime(CLOCK_REALTIME, &time);

	if (ibuf_add_n32(*bp, time.tv_sec) == -1)
		goto fail;
	if (ibuf_add_n16(*bp, type) == -1)
		goto fail;
	if (ibuf_add_n16(*bp, subtype) == -1)
		goto fail;

	switch (peer->local.aid) {
	case AID_INET:
		if (subtype == BGP4MP_STATE_CHANGE_AS4 ||
		    subtype == BGP4MP_MESSAGE_AS4 ||
		    subtype == BGP4MP_MESSAGE_AS4_ADDPATH)
			len += MRT_BGP4MP_ET_AS4_IPv4_HEADER_SIZE;
		else
			len += MRT_BGP4MP_ET_IPv4_HEADER_SIZE;
		break;
	case AID_INET6:
		if (subtype == BGP4MP_STATE_CHANGE_AS4 ||
		    subtype == BGP4MP_MESSAGE_AS4 ||
		    subtype == BGP4MP_MESSAGE_AS4_ADDPATH)
			len += MRT_BGP4MP_ET_AS4_IPv6_HEADER_SIZE;
		else
			len += MRT_BGP4MP_ET_IPv6_HEADER_SIZE;
		break;
	case 0:
		goto fail;
	default:
		log_warnx("king bula found new AF in %s", __func__);
		goto fail;
	}

	if (ibuf_add_n32(*bp, len) == -1)
		goto fail;
	/* millisecond field use by the _ET format */
	if (ibuf_add_n32(*bp, time.tv_nsec / 1000) == -1)
		goto fail;

	if (subtype == BGP4MP_STATE_CHANGE_AS4 ||
	    subtype == BGP4MP_MESSAGE_AS4 ||
	    subtype == BGP4MP_MESSAGE_AS4_ADDPATH) {
		if (!swap)
			if (ibuf_add_n32(*bp, peer->conf.local_as) == -1)
				goto fail;
		if (ibuf_add_n32(*bp, peer->conf.remote_as) == -1)
			goto fail;
		if (swap)
			if (ibuf_add_n32(*bp, peer->conf.local_as) == -1)
				goto fail;
	} else {
		if (!swap)
			if (ibuf_add_n16(*bp, peer->conf.local_short_as) == -1)
				goto fail;
		if (ibuf_add_n16(*bp, peer->short_as) == -1)
			goto fail;
		if (swap)
			if (ibuf_add_n16(*bp, peer->conf.local_short_as) == -1)
				goto fail;
	}

	if (ibuf_add_n16(*bp, /* ifindex */ 0) == -1)
		goto fail;

	switch (peer->local.aid) {
	case AID_INET:
		if (ibuf_add_n16(*bp, AFI_IPv4) == -1)
			goto fail;
		if (!swap)
			if (ibuf_add(*bp, &peer->local.v4,
			    sizeof(peer->local.v4)) == -1)
				goto fail;
		if (ibuf_add(*bp, &peer->remote.v4,
		    sizeof(peer->remote.v4)) == -1)
			goto fail;
		if (swap)
			if (ibuf_add(*bp, &peer->local.v4,
			    sizeof(peer->local.v4)) == -1)
				goto fail;
		break;
	case AID_INET6:
		if (ibuf_add_n16(*bp, AFI_IPv6) == -1)
			goto fail;
		if (!swap)
			if (ibuf_add(*bp, &peer->local.v6,
			    sizeof(peer->local.v6)) == -1)
				goto fail;
		if (ibuf_add(*bp, &peer->remote.v6,
		    sizeof(peer->remote.v6)) == -1)
			goto fail;
		if (swap)
			if (ibuf_add(*bp, &peer->local.v6,
			    sizeof(peer->local.v6)) == -1)
				goto fail;
		break;
	}

	return (0);

fail:
	ibuf_free(*bp);
	*bp = NULL;
	return (-1);
}

int
mrt_dump_hdr_rde(struct ibuf **bp, uint16_t type, uint16_t subtype,
    uint32_t len)
{
	struct timespec	time;

	if ((*bp = ibuf_dynamic(MRT_HEADER_SIZE, MRT_HEADER_SIZE +
	    MRT_BGP4MP_AS4_IPv6_HEADER_SIZE + MRT_BGP4MP_IPv6_ENTRY_SIZE)) ==
	    NULL)
		return (-1);

	clock_gettime(CLOCK_REALTIME, &time);

	if (ibuf_add_n32(*bp, time.tv_sec) == -1)
		goto fail;
	if (ibuf_add_n16(*bp, type) == -1)
		goto fail;
	if (ibuf_add_n16(*bp, subtype) == -1)
		goto fail;

	switch (type) {
	case MSG_TABLE_DUMP:
		switch (subtype) {
		case AFI_IPv4:
			len += MRT_DUMP_HEADER_SIZE;
			break;
		case AFI_IPv6:
			len += MRT_DUMP_HEADER_SIZE_V6;
			break;
		}
		if (ibuf_add_n32(*bp, len) == -1)
			goto fail;
		break;
	case MSG_PROTOCOL_BGP4MP:
	case MSG_TABLE_DUMP_V2:
		if (ibuf_add_n32(*bp, len) == -1)
			goto fail;
		break;
	default:
		log_warnx("mrt_dump_hdr_rde: unsupported type");
		goto fail;
	}
	return (0);

fail:
	ibuf_free(*bp);
	*bp = NULL;
	return (-1);
}

void
mrt_write(struct mrt *mrt)
{
	if (ibuf_write(mrt->fd, mrt->wbuf) == -1) {
		log_warn("mrt dump aborted, mrt_write");
		mrt_clean(mrt);
		mrt_done(mrt);
	}
}

void
mrt_clean(struct mrt *mrt)
{
	close(mrt->fd);
	msgbuf_free(mrt->wbuf);
	mrt->wbuf = NULL;
}

static struct imsgbuf	*mrt_imsgbuf[2];

void
mrt_init(struct imsgbuf *rde, struct imsgbuf *se)
{
	mrt_imsgbuf[RDEIDX] = rde;
	mrt_imsgbuf[SEIDX] = se;
}

int
mrt_open(struct mrt *mrt, time_t now)
{
	enum imsg_type	type;
	int		fd;

	if (strftime(MRT2MC(mrt)->file, sizeof(MRT2MC(mrt)->file),
	    MRT2MC(mrt)->name, localtime(&now)) == 0) {
		log_warnx("mrt_open: strftime conversion failed");
		return (-1);
	}

	fd = open(MRT2MC(mrt)->file,
	    O_WRONLY|O_NONBLOCK|O_CREAT|O_TRUNC|O_CLOEXEC, 0644);
	if (fd == -1) {
		log_warn("mrt_open %s", MRT2MC(mrt)->file);
		return (1);
	}

	if (mrt->state == MRT_STATE_OPEN)
		type = IMSG_MRT_OPEN;
	else
		type = IMSG_MRT_REOPEN;

	if (imsg_compose(mrt_imsgbuf[TYPE2IDX(mrt->type)], type, 0, 0, fd,
	    mrt, sizeof(struct mrt)) == -1)
		log_warn("mrt_open");

	return (1);
}

time_t
mrt_timeout(struct mrt_head *mrt)
{
	struct mrt	*m;
	time_t		 now;
	time_t		 timeout = -1;

	now = time(NULL);
	LIST_FOREACH(m, mrt, entry) {
		if (m->state == MRT_STATE_RUNNING &&
		    MRT2MC(m)->ReopenTimerInterval != 0) {
			if (MRT2MC(m)->ReopenTimer <= now) {
				mrt_open(m, now);
				MRT2MC(m)->ReopenTimer =
				    now + MRT2MC(m)->ReopenTimerInterval;
			}
			if (timeout == -1 ||
			    MRT2MC(m)->ReopenTimer - now < timeout)
				timeout = MRT2MC(m)->ReopenTimer - now;
		}
	}
	return (timeout);
}

void
mrt_reconfigure(struct mrt_head *mrt)
{
	struct mrt	*m, *xm;
	time_t		 now;

	now = time(NULL);
	for (m = LIST_FIRST(mrt); m != NULL; m = xm) {
		xm = LIST_NEXT(m, entry);
		if (m->state == MRT_STATE_OPEN ||
		    m->state == MRT_STATE_REOPEN) {
			if (mrt_open(m, now) == -1)
				continue;
			if (MRT2MC(m)->ReopenTimerInterval != 0)
				MRT2MC(m)->ReopenTimer =
				    now + MRT2MC(m)->ReopenTimerInterval;
			m->state = MRT_STATE_RUNNING;
		}
		if (m->state == MRT_STATE_REMOVE) {
			if (imsg_compose(mrt_imsgbuf[TYPE2IDX(m->type)],
			    IMSG_MRT_CLOSE, 0, 0, -1, m, sizeof(struct mrt)) ==
			    -1)
				log_warn("mrt_reconfigure");
			LIST_REMOVE(m, entry);
			free(m);
			continue;
		}
	}
}

void
mrt_handler(struct mrt_head *mrt)
{
	struct mrt	*m;
	time_t		 now;

	now = time(NULL);
	LIST_FOREACH(m, mrt, entry) {
		if (m->state == MRT_STATE_RUNNING &&
		    (MRT2MC(m)->ReopenTimerInterval != 0 ||
		     m->type == MRT_TABLE_DUMP ||
		     m->type == MRT_TABLE_DUMP_MP ||
		     m->type == MRT_TABLE_DUMP_V2)) {
			if (mrt_open(m, now) == -1)
				continue;
			MRT2MC(m)->ReopenTimer =
			    now + MRT2MC(m)->ReopenTimerInterval;
		}
	}
}

struct mrt *
mrt_get(struct mrt_head *c, struct mrt *m)
{
	struct mrt	*t;

	LIST_FOREACH(t, c, entry) {
		if (t->type != m->type)
			continue;
		if (strcmp(t->rib, m->rib))
			continue;
		if (t->peer_id == m->peer_id &&
		    t->group_id == m->group_id)
			return (t);
	}
	return (NULL);
}

void
mrt_mergeconfig(struct mrt_head *xconf, struct mrt_head *nconf)
{
	struct mrt	*m, *xm;

	/* both lists here are actually struct mrt_conifg nodes */
	LIST_FOREACH(m, nconf, entry) {
		if ((xm = mrt_get(xconf, m)) == NULL) {
			/* NEW */
			if ((xm = malloc(sizeof(struct mrt_config))) == NULL)
				fatal("mrt_mergeconfig");
			memcpy(xm, m, sizeof(struct mrt_config));
			xm->state = MRT_STATE_OPEN;
			LIST_INSERT_HEAD(xconf, xm, entry);
		} else {
			/* MERGE */
			if (strlcpy(MRT2MC(xm)->name, MRT2MC(m)->name,
			    sizeof(MRT2MC(xm)->name)) >=
			    sizeof(MRT2MC(xm)->name))
				fatalx("mrt_mergeconfig: strlcpy");
			MRT2MC(xm)->ReopenTimerInterval =
			    MRT2MC(m)->ReopenTimerInterval;
			xm->state = MRT_STATE_REOPEN;
		}
	}

	LIST_FOREACH(xm, xconf, entry)
		if (mrt_get(nconf, xm) == NULL)
			/* REMOVE */
			xm->state = MRT_STATE_REMOVE;

	/* free config */
	while ((m = LIST_FIRST(nconf)) != NULL) {
		LIST_REMOVE(m, entry);
		free(m);
	}
}