Enables multihop mode for IBGP by default.

This is more consistent with common usage and also with the behavior of
other implementations (Cisco, Juniper).

Also changes the default for gw mode to be based solely on
direct/multihop.

doc/bird.sgml

@@ -1536,33 +1536,37 @@ This allows to set routing policy and all the other parameters differently
 for each neighbor using the following configuration parameters:
 
 <descrip>
-	<tag>local [<m/ip/] as <m/number/</tag> Define which AS we
+	<tag>local [<m/ip/] as <m/number/</tag> Define which AS we are part
-	are part of. (Note that contrary to other IP routers, BIRD is
+	of. (Note that contrary to other IP routers, BIRD is able to act as a
-	able to act as a router located in multiple AS'es
+	router located in multiple AS'es simultaneously, but in such cases you
-	simultaneously, but in such cases you need to tweak the BGP
+	need to tweak the BGP paths manually in the filters to get consistent
-	paths manually in the filters to get consistent behavior.)
+	behavior.) Optional <cf/ip/ argument specifies a source address,
-	Optional <cf/ip/ argument specifies a source address,
+	equivalent to the <cf/source address/ option (see below).  This
-	equivalent to the <cf/source address/ option (see below).
+	parameter is mandatory.
+
+	<tag>neighbor <m/ip/ as <m/number/</tag> Define neighboring router this
+	instance will be talking to and what AS it's located in. In case the
+	neighbor is in the same AS as we are, we automatically switch to iBGP.
 	This parameter is mandatory.
 
-	<tag>neighbor <m/ip/ as <m/number/</tag> Define neighboring router
+	<tag>direct</tag> Specify that the neighbor is directly connected. The
-	this instance will be talking to and what AS it's located in. Unless
+	IP address of the neighbor must be from a directly reachable IP range
-	you use the <cf/multihop/ clause, it must be directly connected to one
+	(i.e. associated with one of your router's interfaces), otherwise the
-	of your router's interfaces. In case the neighbor is in the same AS
+	BGP session wouldn't start but it would wait for such interface to
-	as we are, we automatically switch to iBGP. This parameter is mandatory.
+	appear. The alternative is the <cf/multihop/ option.  Default: enabled
+	for eBGP.
 
-	<tag>multihop [<m/number/]</tag> Configure multihop BGP
+	<tag>multihop [<m/number/]</tag> Configure multihop BGP session to a
-	session to a neighbor that isn't directly connected.
+	neighbor that isn't directly connected.  Accurately, this option should
-	Accurately, this option should be used if the configured
+	be used if the configured neighbor IP address does not match with any
-	neighbor IP address does not match with any local network
+	local network subnets. Such IP address have to be reachable through
-	subnets. Such IP address have to be reachable through system
+	system routing table.  The alternative is the <cf/direct/ option. For
-	routing table. For multihop BGP it is recommended to
+	multihop BGP it is recommended to explicitly configure the source
-	explicitly configure <cf/source address/ to have it
+	address to have it stable. Optional <cf/number/ argument can be used to
-	stable. Optional <cf/number/ argument can be used to specify
+	specify the number of hops (used for TTL). Note that the number of
-	the number of hops (used for TTL). Note that the number of
+	networks (edges) in a path is counted; i.e., if two BGP speakers are
-	networks (edges) in a path is counted, i.e. if two BGP
+	separated by one router, the number of hops is 2. Default: enabled for
-	speakers are separated by one router, the number of hops is
+	iBGP.
-	2. Default: switched off.
 
 	<tag>source address <m/ip/</tag> Define local address we
 	should use for next hop calculation and as a source address
@@ -1609,8 +1613,8 @@ for each neighbor using the following configuration parameters:
 	table, and was used in older versions of BIRD, but does not
 	handle well nontrivial iBGP setups and multihop.  Recursive
 	mode is incompatible with <ref id="dsc-sorted" name="sorted
-	tables">. Default: <cf/direct/ for singlehop eBGP,
+	tables">. Default: <cf/direct/ for direct sessions,
-	<cf/recursive/ otherwise.
+	<cf/recursive/ for multihop sessions.
 
 	<tag>igp table <m/name/</tag> Specifies a table that is used
 	as an IGP routing table. Default: the same as the table BGP is

proto/bgp/bgp.c

@@ -1006,6 +1006,24 @@ bgp_check_config(struct bgp_config *c)
   if (c->c.class == SYM_TEMPLATE)
     return;
 
+
+  /* EBGP direct by default, IBGP multihop by default */
+  if (c->multihop < 0)
+    c->multihop = internal ? 64 : 0;
+
+  /* Different default for gw_mode */
+  if (!c->gw_mode)
+    c->gw_mode = c->multihop ? GW_RECURSIVE : GW_DIRECT;
+
+  /* Different default based on rs_client */
+  if (!c->missing_lladdr)
+    c->missing_lladdr = c->rs_client ? MLL_IGNORE : MLL_SELF;
+
+  /* Disable after error incompatible with restart limit action */
+  if (c->c.in_limit && (c->c.in_limit->action == PLA_RESTART) && c->disable_after_error)
+    c->c.in_limit->action = PLA_DISABLE;
+
+
   if (!c->local_as)
     cf_error("Local AS number must be set");
 
@@ -1021,7 +1039,6 @@ bgp_check_config(struct bgp_config *c)
   if (internal && c->rs_client)
     cf_error("Only external neighbor can be RS client");
 
-
   if (c->multihop && (c->gw_mode == GW_DIRECT))
     cf_error("Multihop BGP cannot use direct gateway mode");
 
@@ -1032,20 +1049,6 @@ bgp_check_config(struct bgp_config *c)
   if (c->multihop && c->bfd && ipa_zero(c->source_addr))
     cf_error("Multihop BGP with BFD requires specified source address");
 
-
-  /* Different default based on rs_client */
-  if (!c->missing_lladdr)
-    c->missing_lladdr = c->rs_client ? MLL_IGNORE : MLL_SELF;
-
-  /* Different default for gw_mode */
-  if (!c->gw_mode)
-    c->gw_mode = (c->multihop || internal) ? GW_RECURSIVE : GW_DIRECT;
-
-  /* Disable after error incompatible with restart limit action */
-  if (c->c.in_limit && (c->c.in_limit->action == PLA_RESTART) && c->disable_after_error)
-    c->c.in_limit->action = PLA_DISABLE;
-
-
   if ((c->gw_mode == GW_RECURSIVE) && c->c.table->sorted)
     cf_error("BGP in recursive mode prohibits sorted table");
 

proto/bgp/config.Y

@@ -34,6 +34,7 @@ CF_ADDTO(proto, bgp_proto '}' { bgp_check_config(BGP_CFG); } )
 
 bgp_proto_start: proto_start BGP {
      this_proto = proto_config_new(&proto_bgp, sizeof(struct bgp_config), $1);
+     BGP_CFG->multihop = -1;	/* undefined */
      BGP_CFG->hold_time = 240;
      BGP_CFG->connect_retry_time = 120;
      BGP_CFG->initial_hold_time = 240;
@@ -74,6 +75,7 @@ bgp_proto:
  | bgp_proto STARTUP HOLD TIME expr ';' { BGP_CFG->initial_hold_time = $5; }
  | bgp_proto CONNECT RETRY TIME expr ';' { BGP_CFG->connect_retry_time = $5; }
  | bgp_proto KEEPALIVE TIME expr ';' { BGP_CFG->keepalive_time = $4; }
+ | bgp_proto DIRECT ';' { BGP_CFG->multihop = 0; }
  | bgp_proto MULTIHOP ';' { BGP_CFG->multihop = 64; }
  | bgp_proto MULTIHOP expr ';' { BGP_CFG->multihop = $3; if (($3<1) || ($3>255)) cf_error("Multihop must be in range 1-255"); }
  | bgp_proto NEXT HOP SELF ';' { BGP_CFG->next_hop_self = 1; BGP_CFG->next_hop_keep = 0; }