Merge commit '6ac4f87a2d661c739e55a63577e7bccf696c7abd' into integrated

Conflicts: lib/socket.h proto/ospf/config.Y proto/ospf/iface.c proto/ospf/ospf.h proto/rip/rip.c sysdep/bsd/sysio.h sysdep/linux/sysio.h sysdep/unix/io.c
2025-03-11 17:08:46 +00:00 · 2013-07-31 16:16:48 +02:00 · 2013-07-31 16:16:48 +02:00 · 6abc22f4b0
commit 6abc22f4b0
parent 9cf24ba8e7 6ac4f87a2d
12 changed files with 249 additions and 78 deletions
--- a/doc/bird.sgml
+++ b/doc/bird.sgml
@ -470,7 +470,7 @@ to zero to disable it. An empty <cf><m/switch/</cf> is equivalent to <cf/on/
 	works in the direction from the routing table to the protocol.
 	Default: <cf/none/.

-	<tag>import keep filtered <m/bool/</tag>
+	<tag>import keep filtered <m/switch/</tag>
 	Usually, if an import filter rejects a route, the route is
 	forgotten. When this option is active, these routes are
 	kept in the routing table, but they are hidden and not
@ -1966,6 +1966,9 @@ protocol ospf &lt;name&gt; {
 			ptp netmask &lt;switch&gt;;
 			check link &lt;switch&gt;;
 			ecmp weight &lt;num&gt;;
+			ttl security [&lt;switch&gt;; | tx only]
+			tx class|dscp &lt;num&gt;;
+			tx priority &lt;num&gt;;
 			authentication [none|simple|cryptographic];
 			password "&lt;text&gt;";
 			password "&lt;text&gt;" {
@ -2236,6 +2239,20 @@ protocol ospf &lt;name&gt; {
 	 prefix) is propagated. It is possible that some hardware
 	 drivers or platforms do not implement this feature. Default value is no.

+	<tag>ttl security [<m/switch/ | tx only]</tag>
+	 TTL security is a feature that protects routing protocols
+	 from remote spoofed packets by using TTL 255 instead of TTL 1
+	 for protocol packets destined to neighbors. Because TTL is
+	 decremented when packets are forwarded, it is non-trivial to
+	 spoof packets with TTL 255 from remote locations. Note that
+	 this option would interfere with OSPF virtual links.
+
+	 If this option is enabled, the router will send OSPF packets
+	 with TTL 255 and drop received packets with TTL less than
+	 255. If this option si set to <cf/tx only/, TTL 255 is used
+	 for sent packets, but is not checked for received
+	 packets. Default value is no.
+
 	<tag>tx class|dscp|priority <m/num/</tag>
         These options specify the ToS/DiffServ/Traffic class/Priority
         of the outgoing OSPF packets. See <ref id="dsc-prio" name="tx
@ -2784,6 +2801,26 @@ makes it pretty much obsolete. (It is still usable on very small networks.)
 	  any periodic messages to this interface and <cf/nolisten/
 	  means that RIP will send to this interface butnot listen to it.

+	<tag>ttl security [<m/switch/ | tx only]</tag>
+	 TTL security is a feature that protects routing protocols
+	 from remote spoofed packets by using TTL 255 instead of TTL 1
+	 for protocol packets destined to neighbors. Because TTL is
+	 decremented when packets are forwarded, it is non-trivial to
+	 spoof packets with TTL 255 from remote locations.
+
+	 If this option is enabled, the router will send RIP packets
+	 with TTL 255 and drop received packets with TTL less than
+	 255. If this option si set to <cf/tx only/, TTL 255 is used
+	 for sent packets, but is not checked for received
+	 packets. Such setting does not offer protection, but offers
+	 compatibility with neighbors regardless of whether they use
+	 ttl security.
+
+	 Note that for RIPng, TTL security is a standard behavior
+	 (required by RFC 2080), but BIRD uses <cf/tx only/ by
+	 default, for compatibility with older versions. For IPv4 RIP,
+	 default value is no.
+
 	<tag>tx class|dscp|priority <m/num/</tag>
          These options specify the ToS/DiffServ/Traffic class/Priority
          of the outgoing RIP packets. See <ref id="dsc-prio" name="tx
--- a/lib/socket.h
+++ b/lib/socket.h
@ -89,10 +89,11 @@ extern int sk_priority_control;	/* Suggested priority for control traffic, shoul

 /* Socket flags */

-#define SKF_V4ONLY	1	/* Use IPv4 for IP sockets */
-#define SKF_V6ONLY	2	/* Use IPV6_V6ONLY socket option */
-#define SKF_LADDR_RX	4	/* Report local address for RX packets */
-#define SKF_LADDR_TX	6	/* Allow to specify local address for TX packets */
+#define SKF_V4ONLY	0x01	/* Use IPv4 for IP sockets */
+#define SKF_V6ONLY	0x02	/* Use IPV6_V6ONLY socket option */
+#define SKF_LADDR_RX	0x04	/* Report local address for RX packets */
+#define SKF_LADDR_TX	0x08	/* Allow to specify local address for TX packets */
+#define SKF_TTL_RX	0x10	/* Report TTL / Hop Limit for RX packets */


 /*
--- a/proto/ospf/config.Y
+++ b/proto/ospf/config.Y
@ -112,8 +112,8 @@ CF_KEYWORDS(OSPF, OSPF2, OSPF3, AREA, OSPF_METRIC1, OSPF_METRIC2, OSPF_TAG, OSPF
 CF_KEYWORDS(NEIGHBORS, RFC1583COMPAT, STUB, TICK, COST, COST2, RETRANSMIT)
 CF_KEYWORDS(HELLO, TRANSMIT, PRIORITY, DEAD, TYPE, BROADCAST, BCAST)
 CF_KEYWORDS(NONBROADCAST, NBMA, POINTOPOINT, PTP, POINTOMULTIPOINT, PTMP)
-CF_KEYWORDS(NONE, SIMPLE, AUTHENTICATION, STRICT, CRYPTOGRAPHIC)
-CF_KEYWORDS(ELIGIBLE, POLL, NETWORKS, HIDDEN, VIRTUAL, CHECK, LINK)
+CF_KEYWORDS(NONE, SIMPLE, AUTHENTICATION, STRICT, CRYPTOGRAPHIC, TTL, SECURITY)
+CF_KEYWORDS(ELIGIBLE, POLL, NETWORKS, HIDDEN, VIRTUAL, CHECK, LINK, ONLY)
 CF_KEYWORDS(RX, BUFFER, LARGE, NORMAL, STUBNET, HIDDEN, SUMMARY, TAG, EXTERNAL)
 CF_KEYWORDS(WAIT, DELAY, LSADB, ECMP, LIMIT, WEIGHT, NSSA, TRANSLATOR, STABILITY)
 CF_KEYWORDS(GLOBAL, LSID, ROUTER, SELF, INSTANCE, REAL, NETMASK, TX, PRIORITY)
@ -297,6 +297,8 @@ ospf_iface_item:
 | RX BUFFER expr { OSPF_PATT->rxbuf = $3 ; if (($3 < OSPF_RXBUF_MINSIZE) || ($3 > OSPF_MAX_PKT_SIZE)) cf_error("Buffer size must be in range 256-65535"); } 
 | TX tos { OSPF_PATT->tx_tos = $2; }
 | TX PRIORITY expr { OSPF_PATT->tx_priority = $3; }
+ | TTL SECURITY bool { OSPF_PATT->ttl_security = $3; }
+ | TTL SECURITY TX ONLY { OSPF_PATT->ttl_security = 2; }
 | password_list { ospf_check_auth(); }
 ;

--- a/proto/ospf/iface.c
+++ b/proto/ospf/iface.c
@ -88,7 +88,7 @@ ospf_sk_open(struct ospf_iface *ifa)
  sk->rbsize = rxbufsize(ifa);
  sk->tbsize = rxbufsize(ifa);
  sk->data = (void *) ifa;
-  sk->flags = SKF_LADDR_RX;
+  sk->flags = SKF_LADDR_RX | (ifa->check_ttl ? SKF_TTL_RX : 0);

  if (sk_open(sk) != 0)
    goto err;
@ -134,7 +134,7 @@ ospf_sk_open(struct ospf_iface *ifa)
    {
      ifa->all_routers = ospf_is_v2(po) ? IP4_OSPF_ALL_ROUTERS : IP6_OSPF_ALL_ROUTERS;
      ifa->des_routers = ospf_is_v2(po) ? IP4_OSPF_DES_ROUTERS : IP6_OSPF_DES_ROUTERS;
-      sk->ttl = 1;	/* Hack, this will affect just the multicast packets */
+      sk->ttl = ifa->cf->ttl_security ? 255 : 1;

      if (sk_setup_multicast(sk) < 0)
        goto err;
@ -535,6 +535,7 @@ ospf_iface_new(struct ospf_area *oa, struct ifa *addr, struct ospf_iface_patt *i
  ifa->rxbuf = ip->rxbuf;
  ifa->check_link = ip->check_link;
  ifa->ecmp_weight = ip->ecmp_weight;
+  ifa->check_ttl = (ip->ttl_security == 1);
  ifa->autype = ip->autype;
  ifa->passwords = ip->passwords;
  ifa->instance_id = ip->instance_id;
--- a/proto/ospf/ospf.h
+++ b/proto/ospf/ospf.h
@ -248,6 +248,7 @@ struct ospf_iface
  u8 check_link;		/* Whether iface link change is used */
  u8 ecmp_weight;		/* Weight used for ECMP */
  u8 ptp_netmask;		/* Send real netmask for P2P */
+  u8 check_ttl;			/* Check incoming packets for TTL 255 */
 };

 struct ospf_md5
@ -754,7 +755,8 @@ struct ospf_iface_patt
  u8 check_link;
  u8 ecmp_weight;
  u8 real_bcast;		/* Not really used in OSPFv3 */
-  u8 ptp_netmask;		/* bool but 2 for unspecified */
+  u8 ptp_netmask;		/* bool + 2 for unspecified */
+  u8 ttl_security;		/* bool + 2 for TX only */
  u8 instance_id;
  list *passwords;
 };
--- a/proto/ospf/packet.c
+++ b/proto/ospf/packet.c
@ -311,6 +311,12 @@ ospf_rx_hook(sock *sk, int size)
    return 1;
  }

+  if (ifa->check_ttl && (sk->ttl < 255))
+  {
+    log(L_ERR "%s%I - TTL %d (< 255)", mesg, sk->faddr, sk->ttl);
+    return 1;
+  }
+
  if ((unsigned) size < sizeof(struct ospf_packet))
  {
    log(L_ERR "%s%I - too short (%u bytes)", mesg, sk->faddr, size);
--- a/proto/rip/config.Y
+++ b/proto/rip/config.Y
@ -22,12 +22,15 @@ CF_DEFINES
 #define RIP_CFG ((struct rip_proto_config *) this_proto)
 #define RIP_IPATT ((struct rip_patt *) this_ipatt)

+static inline int rip_cfg_is_old(void) { return RIP_CFG->c.protocol == &proto_rip; }
+static inline int rip_cfg_is_ng(void)  { return RIP_CFG->c.protocol == &proto_ripng; }
+
 CF_DECLS

 CF_KEYWORDS(RIP, RIPNG, INFINITY, METRIC, PORT, PERIOD, GARBAGE, TIMEOUT,
 	    MODE, BROADCAST, MULTICAST, QUIET, NOLISTEN, VERSION1, 
-	    AUTHENTICATION, NONE, PLAINTEXT, MD5,
-	    HONOR, NEVER, NEIGHBOR, ALWAYS, TX, PRIORITY,
+	    AUTHENTICATION, NONE, PLAINTEXT, MD5, TTL, SECURITY,
+	    HONOR, NEVER, NEIGHBOR, ALWAYS, TX, PRIORITY, ONLY,
 	    RIP_METRIC, RIP_TAG)

 %type <i> rip_mode rip_auth
@ -87,6 +90,8 @@ rip_iface_item:
 | MODE rip_mode { RIP_IPATT->mode |= $2; }
 | TX tos { RIP_IPATT->tx_tos = $2; }
 | TX PRIORITY expr { RIP_IPATT->tx_priority = $3; }
+ | TTL SECURITY bool { RIP_IPATT->ttl_security = $3; }
+ | TTL SECURITY TX ONLY { RIP_IPATT->ttl_security = 2; }
 ;

 rip_iface_opts: 
@ -107,6 +112,7 @@ rip_iface_init:
     RIP_IPATT->metric = 1;
     RIP_IPATT->tx_tos = IP_PREC_INTERNET_CONTROL;
     RIP_IPATT->tx_priority = sk_priority_control;
+     RIP_IPATT->ttl_security = rip_cfg_is_ng() ? 1 : 0;
   }
 ;

--- a/proto/rip/rip.c
+++ b/proto/rip/rip.c
@ -481,6 +481,14 @@ rip_rx(sock *s, int size)
  iface = i->iface;


+  if (i->check_ttl && (s->ttl < 255))
+  {
+    log( L_REMOTE "%s: Discarding packet with TTL %d (< 255) from %I on %s",
+	 p->name, s->ttl, s->faddr, i->iface->name);
+    return 1;
+  }
+
+
  CHK_MAGIC;
  DBG( "RIP: message came: %d bytes from %I via %s\n", size, s->faddr, iface ? iface->name : "(dummy)" );
  size -= sizeof( struct rip_packet_heading );
@ -687,6 +695,7 @@ new_iface(struct proto *p, struct iface *new, unsigned long flags, struct iface_
    rif->mode = PATT->mode;
    rif->metric = PATT->metric;
    rif->multicast = (!(PATT->mode & IM_BROADCAST)) && (flags & IF_MULTICAST);
+    rif->check_ttl = (PATT->ttl_security == 1);
  }
  /* lookup multicasts over unnumbered links - no: rip is not defined over unnumbered links */

@ -708,10 +717,10 @@ new_iface(struct proto *p, struct iface *new, unsigned long flags, struct iface_
  rif->sock->flags = rip_is_old(p) ? SKF_V4ONLY : SKF_V6ONLY;
  if (new)
    {
-      rif->sock->ttl = 1;
      rif->sock->tos = PATT->tx_tos;
      rif->sock->priority = PATT->tx_priority;
-      rif->sock->flags |= SKF_LADDR_RX;
+      rif->sock->ttl = PATT->ttl_security ? 255 : 1;
+      rif->sock->flags |= SKF_LADDR_RX | (rif->check_ttl ? SKF_TTL_RX : 0);
    }

  if (new) {
--- a/proto/rip/rip.h
+++ b/proto/rip/rip.h
@ -109,6 +109,7 @@ struct rip_interface {
  struct rip_connection *busy;
  int metric;			/* You don't want to put struct rip_patt *patt here -- think about reconfigure */
  int mode;
+  int check_ttl;		/* Check incoming packets for TTL 255 */
  int triggered;
  struct object_lock *lock;
  int multicast;
@ -125,6 +126,7 @@ struct rip_patt {
 #define IM_VERSION1 16
  int tx_tos;
  int tx_priority;
+  int ttl_security;	/* bool + 2 for TX only (send, but do not check on RX) */
 };

 struct rip_proto_config {
--- a/sysdep/bsd/sysio.h
+++ b/sysdep/bsd/sysio.h
@ -81,43 +81,57 @@ sk_leave_group4(sock *s, ip_addr maddr)
 /* BSD RX/TX packet info handling for IPv4 */
 /* it uses IP_RECVDSTADDR / IP_RECVIF socket options instead of IP_PKTINFO */

-#define CMSG_RX_SPACE (CMSG_SPACE(sizeof(struct in_addr)) + CMSG_SPACE(sizeof(struct sockaddr_dl)))
-#define CMSG_TX_SPACE CMSG_SPACE(sizeof(struct in_addr))
+#define CMSG_SPACE_PKTINFO4 (CMSG_SPACE(sizeof(struct in_addr)) + \
+			     CMSG_SPACE(sizeof(struct sockaddr_dl)))
+#define CMSG_SPACE_PKTINFO6 CMSG_SPACE(sizeof(struct in6_pktinfo))

-static char *
-sk_request_pktinfo4(sock *s)
+#define CMSG_RX_SPACE (MAX(CMSG_SPACE_PKTINFO4,CMSG_SPACE_PKTINFO6) + CMSG_SPACE(sizeof(int)))
+#define CMSG_TX_SPACE MAX(CMSG_SPACE_PKTINFO4,CMSG_SPACE_PKTINFO6)
+
+
+static inline char *
+sk_request_cmsg4_pktinfo(sock *s)
 {
  int ok = 1;
-  if (s->flags & SKF_LADDR_RX)
-    {
-      if (setsockopt(s->fd, IPPROTO_IP, IP_RECVDSTADDR, &ok, sizeof(ok)) < 0)
-	return "IP_RECVDSTADDR";

-      if (setsockopt(s->fd, IPPROTO_IP, IP_RECVIF, &ok, sizeof(ok)) < 0)
-	return "IP_RECVIF";
-    }
+  if (setsockopt(s->fd, IPPROTO_IP, IP_RECVDSTADDR, &ok, sizeof(ok)) < 0)
+    return "IP_RECVDSTADDR";
+
+  if (setsockopt(s->fd, IPPROTO_IP, IP_RECVIF, &ok, sizeof(ok)) < 0)
+    return "IP_RECVIF";

  return NULL;
 }

-static void
-sk_process_rx_cmsg4(sock *s, struct cmsghdr *cm)
+static inline void
+sk_process_cmsg4_pktinfo(sock *s, struct cmsghdr *cm)
 {
-  if (cm->cmsg_level == IPPROTO_IP && cm->cmsg_type == IP_RECVDSTADDR)
-    {
-      struct in_addr *ra = (struct in_addr *) CMSG_DATA(cm);
-      s->laddr = ipa_get_in4(ra);
-    }
+  if ((cm->cmsg_type == IP_RECVDSTADDR) && (s->flags & SKF_LADDR_RX))
+    s->laddr = ipa_get_in4((struct in_addr *) CMSG_DATA(cm));

-  if (cm->cmsg_level == IPPROTO_IP && cm->cmsg_type == IP_RECVIF)
-    {
-      struct sockaddr_dl *ri = (struct sockaddr_dl *) CMSG_DATA(cm);
-      s->lifindex = ri->sdl_index;
-    }
-
-  // log(L_WARN "RX %I %d", s->laddr, s->lifindex);
+  if ((cm->cmsg_type == IP_RECVIF) && (s->flags & SKF_LADDR_RX))
+    s->lifindex = ((struct sockaddr_dl *) CMSG_DATA(cm))->sdl_index;
 }

+static inline char *
+sk_request_cmsg4_ttl(sock *s)
+{
+  int ok = 1;
+
+  if (setsockopt(s->fd, IPPROTO_IP, IP_RECVTTL, &ok, sizeof(ok)) < 0)
+    return "IP_RECVTTL";
+
+  return NULL;
+}
+
+static inline void
+sk_process_cmsg4_ttl(sock *s, struct cmsghdr *cm)
+{
+  if ((cm->cmsg_type == IP_RECVTTL) && (s->flags & SKF_TTL_RX))
+    s->ttl = * (unsigned char *) CMSG_DATA(cm);
+}
+
+
 /* Unfortunately, IP_SENDSRCADDR does not work for raw IP sockets on BSD kernels */
 /*
 static void
--- a/sysdep/linux/sysio.h
+++ b/sysdep/linux/sysio.h
@ -157,34 +157,52 @@ sk_set_md5_auth_int(sock *s, struct sockaddr *sa, int sa_len, char *passwd)
 /* RX/TX packet info handling for IPv4 */
 /* Mostly similar to standardized IPv6 code */

-#define CMSG_SPACE_PKTINFO4 CMSG_SPACE(sizeof(struct in_pktinfo))
-#define CMSG_SPACE_PKTINFO6 CMSG_SPACE(sizeof(struct in6_pktinfo))
-
-#define CMSG_RX_SPACE MAX(CMSG_SPACE_PKTINFO4,CMSG_SPACE_PKTINFO6)
-#define CMSG_TX_SPACE MAX(CMSG_SPACE_PKTINFO4,CMSG_SPACE_PKTINFO6)
+#define CMSG4_SPACE_PKTINFO CMSG_SPACE(sizeof(struct in_pktinfo))

 static inline char *
-sk_request_pktinfo4(sock *s)
+sk_request_cmsg4_pktinfo(sock *s)
 {
  int ok = 1;
-  if (s->flags & SKF_LADDR_RX)
-    if (setsockopt(s->fd, IPPROTO_IP, IP_PKTINFO, &ok, sizeof(ok)) < 0)
-      return "IP_PKTINFO";
+
+  if (setsockopt(s->fd, IPPROTO_IP, IP_PKTINFO, &ok, sizeof(ok)) < 0)
+    return "IP_PKTINFO";

  return NULL;
 }

 static inline void
-sk_process_rx_cmsg4(sock *s, struct cmsghdr *cm)
+sk_process_cmsg4_pktinfo(sock *s, struct cmsghdr *cm)
 {
-  if (cm->cmsg_level == IPPROTO_IP && cm->cmsg_type == IP_PKTINFO)
-    {
-      struct in_pktinfo *pi = (struct in_pktinfo *) CMSG_DATA(cm);
-      s->laddr = ipa_get_in4(&pi->ipi_addr);
-      s->lifindex = pi->ipi_ifindex;
-    }
+  if ((cm->cmsg_type == IP_PKTINFO) && (s->flags & SKF_LADDR_RX))
+  {
+    struct in_pktinfo *pi = (struct in_pktinfo *) CMSG_DATA(cm);
+    s->laddr = ipa_get_in4(&pi->ipi_addr);
+    s->lifindex = pi->ipi_ifindex;
+  }
 }

+
+#define CMSG4_SPACE_TTL CMSG_SPACE(sizeof(int))
+
+static inline char *
+sk_request_cmsg4_ttl(sock *s)
+{
+  int ok = 1;
+
+  if (setsockopt(s->fd, IPPROTO_IP, IP_RECVTTL, &ok, sizeof(ok)) < 0)
+    return "IP_RECVTTL";
+
+  return NULL;
+}
+
+static inline void
+sk_process_cmsg4_ttl(sock *s, struct cmsghdr *cm)
+{
+  if ((cm->cmsg_type == IP_TTL) && (s->flags & SKF_TTL_RX))
+    s->ttl = * (int *) CMSG_DATA(cm);
+}
+
+
 /*
 static void
 sysio_prepare_tx_cmsgs(sock *s, struct msghdr *msg, void *cbuf, size_t cbuflen)
--- a/sysdep/unix/io.c
+++ b/sysdep/unix/io.c
@ -700,42 +700,96 @@ sk_insert(sock *s)
 #ifndef IPV6_RECVPKTINFO
 #define IPV6_RECVPKTINFO IPV6_PKTINFO
 #endif
+/*
+ * Same goes for IPV6_HOPLIMIT -> IPV6_RECVHOPLIMIT.
+ */
+#ifndef IPV6_RECVHOPLIMIT
+#define IPV6_RECVHOPLIMIT IPV6_HOPLIMIT
+#endif
+
+
+#define CMSG6_SPACE_PKTINFO CMSG_SPACE(sizeof(struct in6_pktinfo))

 static inline char *
-sk_request_pktinfo6(sock *s)
+sk_request_cmsg6_pktinfo(sock *s)
 {
  int ok = 1;
-  if (s->flags & SKF_LADDR_RX)
-    if (setsockopt(s->fd, IPPROTO_IPV6, IPV6_RECVPKTINFO, &ok, sizeof(ok)) < 0)
-      return "IPV6_RECVPKTINFO";
+
+  if (setsockopt(s->fd, IPPROTO_IPV6, IPV6_RECVPKTINFO, &ok, sizeof(ok)) < 0)
+    return "IPV6_RECVPKTINFO";

  return NULL;
 }

+static inline void
+sk_process_cmsg6_pktinfo(sock *s, struct cmsghdr *cm)
+{
+  if ((cm->cmsg_type == IPV6_PKTINFO) && (s->flags & SKF_LADDR_RX))
+  {
+    struct in6_pktinfo *pi = (struct in6_pktinfo *) CMSG_DATA(cm);
+    s->laddr = ipa_get_in6(&pi->ipi6_addr);
+    s->lifindex = pi->ipi6_ifindex;
+  }
+}
+
+
+#define CMSG6_SPACE_TTL CMSG_SPACE(sizeof(int))
+
+static inline char *
+sk_request_cmsg6_ttl(sock *s)
+{
+  int ok = 1;
+
+  if (setsockopt(s->fd, IPPROTO_IPV6, IPV6_RECVHOPLIMIT, &ok, sizeof(ok)) < 0)
+    return "IPV6_RECVHOPLIMIT";
+
+  return NULL;
+}
+
+static inline void
+sk_process_cmsg6_ttl(sock *s, struct cmsghdr *cm)
+{
+  if ((cm->cmsg_type == IPV6_HOPLIMIT) && (s->flags & SKF_TTL_RX))
+    s->ttl = * (int *) CMSG_DATA(cm);
+}
+
+
+#define CMSG_RX_SPACE MAX(CMSG4_SPACE_PKTINFO+CMSG4_SPACE_TTL, \
+			  CMSG6_SPACE_PKTINFO+CMSG6_SPACE_TTL)
+#define CMSG_TX_SPACE MAX(CMSG4_SPACE_PKTINFO,CMSG6_SPACE_PKTINFO)
+
+
 static void
-sk_process_rx_cmsgs(sock *s, struct msghdr *msg)
+sk_process_cmsgs(sock *s, struct msghdr *msg)
 {
  struct cmsghdr *cm;

-  if (!(s->flags & SKF_LADDR_RX))
-    return;
+  if (s->flags & SKF_LADDR_RX)
+  {
+    s->laddr = IPA_NONE;
+    s->lifindex = 0;
+  }

-  s->laddr = IPA_NONE;
-  s->lifindex = 0;
+  if (s->flags & SKF_TTL_RX)
+    s->ttl = -1;

  for (cm = CMSG_FIRSTHDR(msg); cm != NULL; cm = CMSG_NXTHDR(msg, cm))
+  {
+    if ((cm->cmsg_level == IPPROTO_IP) && sk_is_ipv4(s))
    {
-      if (cm->cmsg_level == IPPROTO_IPV6 && cm->cmsg_type == IPV6_PKTINFO)
-	{
-	  struct in6_pktinfo *pi = (struct in6_pktinfo *) CMSG_DATA(cm);
-	  s->laddr = ipa_get_in6(&pi->ipi6_addr);
-	  s->lifindex = pi->ipi6_ifindex;
-	}
-
-      sk_process_rx_cmsg4(s, cm);
+      sk_process_cmsg4_pktinfo(s, cm);
+      sk_process_cmsg4_ttl(s, cm);
    }
+
+    if ((cm->cmsg_level == IPPROTO_IPV6) && sk_is_ipv6(s))
+    {
+      sk_process_cmsg6_pktinfo(s, cm);
+      sk_process_cmsg6_ttl(s, cm);
+    }
+  }
 }

+
 /*
 static void
 sysio_prepare_tx_cmsgs(sock *s, struct msghdr *msg, void *cbuf, size_t cbuflen)
@ -794,14 +848,33 @@ sk_setup(sock *s)
  if (s->priority >= 0)
    sk_set_priority(s, s->priority);

-  // XXXX better error handling
  if (s->ttl >= 0)
-    sk_set_ttl(s, s->ttl);
+    if (sk_set_ttl(s, s->ttl) < 0)
+      ERR("sk_set_ttl");

  if (sk_is_ipv4(s))
-    err = sk_request_pktinfo4(s);
-  else
-    err = sk_request_pktinfo6(s);
+  {
+    if (s->flags & SKF_LADDR_RX)
+      if (err = sk_request_cmsg4_pktinfo(s))
+	goto bad;
+
+    if (s->flags & SKF_TTL_RX)
+      if (err = sk_request_cmsg4_ttl(s))
+	goto bad;
+  }
+
+  if (sk_is_ipv6(s))
+  {
+    if (s->flags & SKF_LADDR_RX)
+      if (err = sk_request_cmsg6_pktinfo(s))
+	goto bad;
+
+    if (s->flags & SKF_TTL_RX)
+      if (err = sk_request_cmsg6_ttl(s));
+  }
+
+  return NULL;
+
 bad:
  return err;
 }
@ -1493,7 +1566,7 @@ sk_read(sock *s)
 	  }
 	s->rpos = s->rbuf + e;
 	sockaddr_read(sa, &s->faddr, NULL, &s->fport, 1);
-	sk_process_rx_cmsgs(s, &msg);
+	sk_process_cmsgs(s, &msg);

 	s->rx_hook(s, e);
 	return 1;