NEWS and version update

Doc: Update documentation about VRFs and BFD
Nest: VRF of protocol can be explicitly specified as 'default'
2019-08-01 03:31:23 +02:00 · 2019-08-01 01:02:31 +02:00 · 2019-08-01 00:53:22 +02:00 · 2019-07-30 12:13:38 +02:00 · 2019-07-18 12:41:45 +02:00 · 2019-07-18 02:05:36 +02:00
82 changed files with 3103 additions and 1289 deletions
--- a/.gitlab-ci.yml
+++ b/.gitlab-ci.yml
@ -211,9 +211,6 @@ docker_ubuntu-16_04-amd64:

 # TODO We want to copy these BSDs to our own virtual machines, to make sure someone doesn't update them by accident.
 .freebsd-11-i386: &freebsd-11-i386_env
-  variables:
-    CPPFLAGS: "-I/usr/local/include"
-    LDFLAGS: "-L/usr/local/lib"
  tags:
  - freebsd
  - i386
@ -223,9 +220,6 @@ docker_ubuntu-16_04-amd64:
  #- tags

 .freebsd-11-amd64: &freebsd-11-amd64_env
-  variables:
-    CPPFLAGS: "-I/usr/local/include"
-    LDFLAGS: "-L/usr/local/lib"
  tags:
  - freebsd
  - amd64
--- a/7
+++ b/7
@ -1,7 +0,0 @@
-C doc
-C nest
-C conf
-C filter
-C proto
-C sysdep
-C lib
--- a/8
+++ b/8
@ -7,13 +7,15 @@ $ make

 Default location for configuration file is /usr/local/etc/bird.conf and
 for control socket is /usr/local/var/run/bird.ctl . You can change that
-by --sysconfdir and --localstatedir configure options.
+by --prefix, --sysconfdir and --runstatedir configure options, e.g.:
+
+$ ./configure --prefix=/usr --sysconfdir=/etc --runstatedir=/run

 To compile current development BIRD source code from Git repository, you
 also need Git (to download the source code) and Autoconf (to generate
 the configure script and associated files using 'autoreconf' tool):

-$ git clone git://git.nic.cz/bird.git
+$ git clone https://gitlab.labs.nic.cz/labs/bird/
 $ cd bird
 $ autoreconf

@ -32,7 +34,7 @@ For compiling BIRD you need these programs and libraries:
 - Flex

 - ncurses library
- - GNU Readline library (2.1 or newer)
+ - GNU Readline library

 For compiling BIRD documentation you also need:

--- a/27
+++ b/27
@ -1,3 +1,30 @@
+Version 1.6.7 (2019-08-01)
+  o BFD: Support for VRFs
+  o Several bugfixes
+
+Version 1.6.6 (2019-02-27)
+  o Several bugfixes related to route propagation
+
+Version 1.6.5 (2019-01-05)
+  o MRT table dumps (RFC 6396)
+  o BGP Long-lived graceful restart
+  o Filter: Make ifname attribute modifiable
+  o Improved keeping track of IPv6 link-local addresses
+  o Many bugfixes
+
+Version 1.6.4 (2018-03-22)
+  o Basic VRF support
+  o Simplified autoconf scripts
+  o BGP: Shutdown communication (RFC 8203)
+  o BGP: Allow exchanging LOCAL_PREF with eBGP peers
+  o BGP: Allow to specify interface for regular sessions
+  o BGP: New option 'disable after cease'
+  o RAdv: Support for more specific routes (RFC 4191)
+  o RAdv: Proper handling of prefix retraction
+  o Filter: Allow silent filter execution
+  o Filter: Fixed stack overflow in BGP mask expressions
+  o Several bug fixes
+
 Version 1.6.3 (2016-12-21)
  o Large BGP communities
  o BFD authentication (MD5, SHA1)
--- a/18
+++ b/18
@ -6,7 +6,7 @@
                (c) 1998--2008  Martin Mares <mj@ucw.cz>
                (c) 1998--2000  Pavel Machek <pavel@ucw.cz>
                (c) 1998--2008  Ondrej Filip <feela@network.cz>
-                (c) 2009--2016  CZ.NIC z.s.p.o.
+                (c) 2009--2019  CZ.NIC z.s.p.o.

 ================================================================================

@ -48,20 +48,10 @@ How to install BIRD
 	   ftp://bird.network.cz/pub/bird/redhat/
 	o  From source code of the latest stable release version
 	   ftp://bird.network.cz/pub/bird/
-	o  From source code of the actual development version
-	   git://git.nic.cz/bird.git
+	o  From current development code in Git repository
 	   https://gitlab.labs.nic.cz/labs/bird/

-How to install BIRD from source code
------------------------------------
-
-	$  ./configure
-	$  make
-	$  su
-	#  make install
-	#  vi /usr/local/etc/bird.conf
-
-See the file INSTALL for more information about installation from source code.
+See the file INSTALL for information about installation from source code.

 Documentation
 =============
@ -69,7 +59,7 @@ Documentation
 Online documentation is available at http://bird.network.cz/?get_doc or as HTML
 files in the doc directory, you can install it by `make install-docs' and
 rebuild it by `make docs', but you'll need SGMLtools and LaTeX to be installed
-on your machine. You can also download a neatly formatted PostScript version as
+on your machine. You can also download a neatly formatted PDF version as
 a separate archive (bird-doc-*.tar.gz) from ftp://bird.network.cz/pub/bird/

 User support
--- a/aclocal.m4
+++ b/aclocal.m4
@ -31,6 +31,72 @@ AC_DEFUN([BIRD_CHECK_PTHREADS],
  CFLAGS="$bird_tmp_cflags"
 ])

+AC_DEFUN([BIRD_CHECK_ANDROID_GLOB],
+[
+  AC_CACHE_CHECK(
+    [for glob.h],
+    [bird_cv_lib_glob],
+    AC_LINK_IFELSE([
+      AC_LANG_PROGRAM(
+        [
+	  #include <glob.h>
+	  #include <stdlib.h>
+	],
+        [ glob(NULL, 0, NULL, NULL); ]
+      )
+    ],
+    [bird_cv_lib_glob=yes],
+      [
+        bird_tmp_libs="$LIBS"
+        LIBS="$LIBS -landroid-glob"
+        AC_LINK_IFELSE([
+          AC_LANG_PROGRAM(
+            [
+	      #include <glob.h>
+	      #include <stdlib.h>
+	    ],
+            [ glob(NULL, 0, NULL, NULL); ]
+          )
+        ],
+        [bird_cv_lib_glob=-landroid-glob],
+        [bird_cv_lib_glob=no]
+        )
+        LIBS="$bird_tmp_libs"
+      ]
+    )
+  )
+])
+
+AC_DEFUN([BIRD_CHECK_ANDROID_LOG],
+[
+  AC_CACHE_CHECK(
+    [for syslog lib flags],
+    [bird_cv_lib_log],
+    AC_LINK_IFELSE([
+      AC_LANG_PROGRAM(
+        [ #include <sys/syslog.h> ],
+        [ syslog(0, ""); ]
+      )
+    ],
+    [bird_cv_lib_log=yes],
+      [
+        bird_tmp_libs="$LIBS"
+        LIBS="$LIBS -llog"
+        AC_LINK_IFELSE([
+          AC_LANG_PROGRAM(
+            [ #include <sys/syslog.h> ],
+            [ syslog(0, ""); ]
+          )
+        ],
+        [bird_cv_lib_log=-llog],
+        [bird_cv_lib_log=no]
+        )
+        LIBS="$bird_tmp_libs"
+      ]
+    )
+  )
+])
+
 AC_DEFUN([BIRD_CHECK_GCC_OPTION],
 [
  bird_tmp_cflags="$CFLAGS"
--- a/conf/cf-lex.l
+++ b/conf/cf-lex.l
@ -100,6 +100,7 @@ static struct include_file_stack *ifs_head;
 #define YY_INPUT(buf,result,max) result = cf_read_hook(buf, max, ifs->fd);
 #define YY_NO_UNPUT
 #define YY_FATAL_ERROR(msg) cf_error(msg)
+#define YY_USER_ACTION ifs->chno += yyleng; ifs->toklen = yyleng;

 static void cf_include(char *arg, int alen);
 static int check_eof(void);
@ -237,7 +238,7 @@ else: {

 {WHITE}+

-\n	ifs->lino++;
+\n	ifs->lino++; ifs->chno = 0;

 #	BEGIN(COMMENT);

@ -247,13 +248,14 @@ else: {

 <COMMENT>\n {
  ifs->lino++;
+  ifs->chno = 0;
  BEGIN(INITIAL);
 }

 <COMMENT>.

 <CCOMM>\*\/	BEGIN(INITIAL);
-<CCOMM>\n	ifs->lino++;
+<CCOMM>\n	ifs->lino++; ifs->chno = 0;
 <CCOMM>\/\*	cf_error("Comment nesting not supported");
 <CCOMM><<EOF>>	cf_error("Unterminated comment");
 <CCOMM>.
--- a/conf/conf.c
+++ b/conf/conf.c
@ -206,7 +206,7 @@ config_del_obstacle(struct config *c)
 {
  DBG("+++ deleting obstacle %d\n", c->obstacle_count);
  c->obstacle_count--;
-  if (!c->obstacle_count)
+  if (!c->obstacle_count && (c != config))
    ev_schedule(config_event);
 }

@ -515,6 +515,7 @@ cf_error(char *msg, ...)
  va_end(args);
  new_config->err_msg = cfg_strdup(buf);
  new_config->err_lino = ifs->lino;
+  new_config->err_chno = ifs->chno - ifs->toklen + 1;
  new_config->err_file_name = ifs->file_name;
  cf_lex_unwind();
  longjmp(conf_jmpbuf, 1);
--- a/conf/conf.h
+++ b/conf/conf.h
@ -48,6 +48,7 @@ struct config {
  u32 watchdog_timeout;			/* Watchdog timeout (in seconds, 0 = disabled) */
  char *err_msg;			/* Parser error message */
  int err_lino;				/* Line containing error */
+  int err_chno;				/* Character where the parser stopped */
  char *err_file_name;			/* File name containing error */
  char *file_name;			/* Name of main configuration file */
  int file_fd;				/* File descriptor of main configuration file */
@ -141,6 +142,8 @@ struct include_file_stack {
  char *file_name;			/* File name */
  int fd;				/* File descriptor */
  int lino;				/* Current line num */
+  int chno;				/* Current char num (on current line) */
+  int toklen;				/* Current token length */
  int depth;				/* Include depth, 0 = cannot include */

  struct include_file_stack *prev;	/* Previous record in stack */
@ -149,7 +152,6 @@ struct include_file_stack {

 extern struct include_file_stack *ifs;

-
 int cf_lex(void);
 void cf_lex_init(int is_cli, struct config *c);
 void cf_lex_unwind(void);
--- a/conf/confbase.Y
+++ b/conf/confbase.Y
@ -55,6 +55,7 @@ CF_DECLS
  struct roa_show_data *ro;
  struct sym_show_data *sd;
  struct lsadb_show_data *ld;
+  struct mrt_dump_data *md;
  struct iface *iface;
  struct roa_table *rot;
  void *g;
--- a/configure.ac
+++ b/configure.ac
@ -48,14 +48,12 @@ AC_ARG_WITH([suffix],
 )

 AC_ARG_WITH([sysconfig],
-  [AS_HELP_STRING([--with-sysconfig=FILE], [use specified BIRD system configuration file])],
-  []
+  [AS_HELP_STRING([--with-sysconfig=FILE], [use specified BIRD system configuration file])]
 )

 AC_ARG_WITH([runtimedir],
-  [AS_HELP_STRING([--with-runtimedir=PATH], [path for runtime files @<:@LOCALSTATEDIR/run@:>@])],
-  [runtimedir="$with_runtimedir"],
-  [runtimedir="\$(localstatedir)/run"]
+  [AS_HELP_STRING([--with-runtimedir=PATH], [run-state data, obsolete variant of --runstatedir])],
+  [runstatedir="$with_runtimedir"]
 )

 AC_ARG_WITH([iproutedir],
@ -91,7 +89,11 @@ esac
 AC_SUBST([objdir])
 AC_SUBST([exedir])
 AC_SUBST([srcdir_rel_mf])
-AC_SUBST([runtimedir])
+
+# Workaround for older Autoconfs that do not define runstatedir
+AS_IF([test -z "${runstatedir}"], [runstatedir='${localstatedir}/run'])
+AC_SUBST([runstatedir])
+

 if test "$enable_ipv6" = yes ; then
  ip=ipv6
@ -112,7 +114,7 @@ if test "$enable_debug" = yes ; then
  CONTROL_SOCKET="bird$SUFFIX.ctl"
 else
  CONFIG_FILE="\$(sysconfdir)/bird$SUFFIX.conf"
-  CONTROL_SOCKET="$runtimedir/bird$SUFFIX.ctl"
+  CONTROL_SOCKET="\$(runstatedir)/bird$SUFFIX.ctl"
 fi
 AC_SUBST([CONFIG_FILE])
 AC_SUBST([CONTROL_SOCKET])
@ -216,9 +218,13 @@ else
      ;;
    ipv6:freebsd*)
      sysdesc=bsd-v6
+      CPPFLAGS="$CPPFLAGS -I/usr/local/include"
+      LDFLAGS="$LDFLAGS -L/usr/local/lib"
      ;;
    ipv4:freebsd*)
      sysdesc=bsd
+      CPPFLAGS="$CPPFLAGS -I/usr/local/include"
+      LDFLAGS="$LDFLAGS -L/usr/local/lib"
      ;;
    ipv6:dragonfly*)
      sysdesc=bsd-v6
@ -266,7 +272,7 @@ fi

 AC_SUBST([iproutedir])

-all_protocols="$proto_bfd bgp ospf pipe $proto_radv rip static"
+all_protocols="$proto_bfd bgp mrt ospf pipe $proto_radv rip static"
 if test "$ip" = ipv6 ; then
   all_protocols="$all_protocols babel"
 fi
@ -279,6 +285,7 @@ fi
 AH_TEMPLATE([CONFIG_BABEL], 	[Babel protocol])
 AH_TEMPLATE([CONFIG_BFD],	[BFD protocol])
 AH_TEMPLATE([CONFIG_BGP],	[BGP protocol])
+AH_TEMPLATE([CONFIG_MRT],	[MRT protocol])
 AH_TEMPLATE([CONFIG_OSPF],	[OSPF protocol])
 AH_TEMPLATE([CONFIG_PIPE],	[Pipe protocol])
 AH_TEMPLATE([CONFIG_RADV],	[RAdv protocol])
@ -321,6 +328,20 @@ AC_C_BIGENDIAN(
  [AC_MSG_ERROR([Cannot determine CPU endianity.])]
 )

+BIRD_CHECK_ANDROID_GLOB
+if test "$bird_cv_lib_glob" = no ; then
+  AC_MSG_ERROR([glob.h not found.])
+elif test "$bird_cv_lib_glob" != yes ; then
+  LIBS="$LIBS $bird_cv_lib_glob"
+fi
+
+BIRD_CHECK_ANDROID_LOG
+if test "$bird_cv_lib_log" = no ; then
+  AC_MSG_ERROR([don't know how to link syslog.])
+elif test "$bird_cv_lib_log" != yes ; then
+  LIBS="$LIBS $bird_cv_lib_log"
+fi
+
 if test "$enable_debug" = yes ; then
  AC_DEFINE([DEBUGGING], [1], [Define to 1 if debugging is enabled])
  if test "$enable_memcheck" = yes ; then
@ -382,7 +403,7 @@ AC_SUBST([CLIENT_LIBS])
 mkdir -p $objdir/sysdep
 AC_CONFIG_HEADERS([$objdir/sysdep/autoconf.h:sysdep/autoconf.h.in])
 AC_CONFIG_COMMANDS([merge],
-  [ export CPP="$CPP"; $srcdir/tools/mergedirs $srcdir $srcdir_rel $objdir $sysdep_dirs ],
+  [ export CPP="$CPP"; sh $srcdir/tools/mergedirs $srcdir $srcdir_rel $objdir $sysdep_dirs ],
  [
    srcdir=$srcdir
    srcdir_rel=$srcdir_rel
@ -405,4 +426,5 @@ AC_MSG_RESULT([        Debugging:		$enable_debug])
 AC_MSG_RESULT([        POSIX threads:		$enable_pthreads])
 AC_MSG_RESULT([        Routing protocols:	$protocols])
 AC_MSG_RESULT([        Client:			$enable_client])
+
 rm -f $objdir/.*-stamp
--- a/doc/Doc
+++ b/doc/Doc
@ -1 +0,0 @@
-D prog-intro.sgml
--- a/doc/LinuxDocTools.pm
+++ b/doc/LinuxDocTools.pm
@ -242,6 +242,7 @@ sub process_options
  #   removes iso-entites sub directory after doing make install.)
  #
  $ENV{SGML_CATALOG_FILES} .= (defined $ENV{SGML_CATALOG_FILES} ? ":" : "") .
+     "$main::prefix/share/sgml/sgml-iso-entities-8879.1986/catalog:" .
     "$main::prefix/share/sgml/entities/sgml-iso-entities-8879.1986/catalog";
  $ENV{SGML_CATALOG_FILES} .= ":$main::DataDir/linuxdoc-tools.catalog";
  $ENV{SGML_CATALOG_FILES} .= ":$main::/etc/sgml.catalog";
@ -372,6 +373,8 @@ sub process_file
        }
    }
  #
+
+  local $ENV{PATH} = "$ENV{PATH}:/usr/lib/linuxdoc-tools";
  my($precmd) = "|sgmlpre output=$global->{format} $global->{define}";

  #
--- a/doc/Makefile
+++ b/doc/Makefile
@ -8,11 +8,13 @@ srcdir=$(shell cd $(root-rel) ; pwd)
 srcdir_abs=$(srcdir)
 endif

+export TEXINPUTS := $(TEXINPUTS):$(srcdir_abs)/doc/tex
+
 # Force rebuilds
 .PHONY: prog.sgml bird.sgml

 docs: progdocs userdocs
-progdocs: prog.html prog.ps
+progdocs: prog.html prog.pdf
 userdocs: bird.html bird.pdf

 prog.sgml:
--- a/doc/bird.sgml
+++ b/doc/bird.sgml
@ -594,19 +594,28 @@ agreement").

 	<tag><label id="proto-description">description "<m/text/"</tag>
 	This is an optional description of the protocol. It is displayed as a
-	part of the output of 'show route all' command.
+	part of the output of 'show protocols all' command.

 	<tag><label id="proto-table">table <m/name/</tag>
 	Connect this protocol to a non-default routing table.

-	<tag><label id="proto-vrf">vrf "<m/text/"</tag>
+	<tag><label id="proto-vrf">vrf "<m/text/"|default</tag>
 	Associate the protocol with specific VRF. The protocol will be
 	restricted to interfaces assigned to the VRF and will use sockets bound
-	to the VRF. Appropriate VRF interface must exist on OS level. For kernel
-	protocol, an appropriate table still must be explicitly selected by
-	<cf/table/ option. Note that the VRF support in BIRD and Linux kernel
-	(4.11) is still in development and is currently problematic outside of
-	multihop BGP.
+	to the VRF. A corresponding VRF interface must exist on OS level. For
+	kernel protocol, an appropriate table still must be explicitly selected
+	by <cf/table/ option.
+
+	By selecting <cf/default/, the protocol is associated with the default
+	VRF; i.e., it will be restricted to interfaces not assigned to any
+	regular VRF. That is different from not specifying <cf/vrf/ at all, in
+	which case the protocol may use any interface regardless of its VRF
+	status.
+
+	Note that for proper VRF support it is necessary to use Linux kernel
+	version at least 4.14, older versions have limited VRF implementation.
+	Before Linux kernel 5.0, a socket bound to a port in default VRF collide
+	with others in regular VRFs.
 </descrip>

 <p>There are several options that give sense only with certain protocols:
@ -837,6 +846,10 @@ This argument can be omitted if there exists only a single instance.
 	number of networks, number of routes before and after filtering). If
 	you use <cf/count/ instead, only the statistics will be printed.

+	<tag><label id="cli-mrt-dump">mrt dump table <m/name/|"<m/pattern/" to "<m/filename/" [filter <m/f/|where <m/c/]</tag>
+	Dump content of a routing table to a specified file in MRT table dump
+	format. See <ref id="mrt" name="MRT protocol"> for details.
+
 	<tag><label id="cli-show-roa">show roa [<m/prefix/ | in <m/prefix/ | for <m/prefix/] [as <m/num/] [table <m/t/]</tag>
 	Show contents of a ROA table (by default of the first one). You can
 	specify a <m/prefix/ to print ROA entries for a specific network. If you
@ -1405,7 +1418,8 @@ clist for most purposes.
 	<tag><label id="rta-ifname"><m/string/ ifname</tag>
 	Name of the outgoing interface. Sink routes (like blackhole, unreachable
 	or prohibit) and multipath routes have no interface associated with
-	them, so <cf/ifname/ returns an empty string for such routes. Read-only.
+	them, so <cf/ifname/ returns an empty string for such routes. Setting it
+	would also change route to a direct one (remove gateway).

 	<tag><label id="rta-ifindex"><m/int/ ifindex</tag>
 	Index of the outgoing interface. System wide index of the interface. May
@ -1619,7 +1633,7 @@ in the future. Also note that we currently support at most one protocol instance
 <p>BFD packets are sent with a dynamic source port number. Linux systems use by
 default a bit different dynamic port range than the IANA approved one
 (49152-65535). If you experience problems with compatibility, please adjust
-<cf>/proc/sys/net/ipv4/ip_local_port_range</cf>
+<cf>/proc/sys/net/ipv4/ip_local_port_range</cf>.

 <sect1>Configuration
 <label id="bfd-config">
@ -1636,6 +1650,14 @@ configuration is often sufficient.
 <p>Note that to use BFD for other protocols like OSPF or BGP, these protocols
 also have to be configured to request BFD sessions, usually by <cf/bfd/ option.

+<p>A BFD instance not associated with any VRF handles session requests from all
+other protocols, even ones associated with a VRF. Such setup would work for
+single-hop BFD sessions if <cf/net.ipv4.udp_l3mdev_accept/ sysctl is enabled,
+but does not currently work for multihop sessions. Another approach is to
+configure multiple BFD instances, one for each VRF (including the default VRF).
+Each BFD instance associated with a VRF (regular or default) only handles
+session requests from protocols in the same VRF.
+
 <p>Some of BFD session options require <m/time/ value, which has to be specified
 with the appropriate unit: <m/num/ <cf/s/|<cf/ms/|<cf/us/. Although microseconds
 are allowed as units, practical minimum values are usually in order of tens of
@ -1977,13 +1999,16 @@ using the following configuration parameters:
 	immediately shut down. Note that this option cannot be used with
 	multihop BGP. Default: disabled.

-	<tag><label id="bgp-bfd">bfd <M>switch</M></tag>
+	<tag><label id="bgp-bfd">bfd <M>switch</M>|graceful</tag>
 	BGP could use BFD protocol as an advisory mechanism for neighbor
 	liveness and failure detection. If enabled, BIRD setups a BFD session
 	for the BGP neighbor and tracks its liveness by it. This has an
 	advantage of an order of magnitude lower detection times in case of
-	failure. Note that BFD protocol also has to be configured, see
-	<ref id="bfd" name="BFD"> section for details. Default: disabled.
+	failure. When a neighbor failure is detected, the BGP session is
+	restarted. Optionally, it can be configured (by <cf/graceful/ argument)
+	to trigger graceful restart instead of regular restart.  Note that BFD
+	protocol also has to be configured, see <ref id="bfd" name="BFD">
+	section for details. Default: disabled.

 	<tag><label id="bgp-ttl-security">ttl security <m/switch/</tag>
 	Use GTSM (<rfc id="5082"> - the generalized TTL security mechanism). GTSM
@ -2108,6 +2133,25 @@ using the following configuration parameters:
 	re-establish after a restart before deleting stale routes. Default:
 	120 seconds.

+	<tag><label id="bgp-long-lived-graceful-restart">long lived graceful restart <m/switch/|aware</tag>
+	The long-lived graceful restart is an extension of the traditional
+	<ref id="bgp-graceful-restart" name="BGP graceful restart">, where stale
+	routes are kept even after the <ref id="bgp-graceful-restart-time"
+	name="restart time"> expires for additional long-lived stale time, but
+	they are marked with the LLGR_STALE community, depreferenced, and
+	withdrawn from routers not supporting LLGR. Like traditional BGP
+	graceful restart, it has three states: disabled, aware (receiving-only),
+	and enabled. Note that long-lived graceful restart requires at least
+	aware level of traditional BGP graceful restart. Default: aware, unless
+	graceful restart is disabled.
+
+	<tag><label id="bgp-long-lived-stale-time">long lived stale time <m/number/</tag>
+	The long-lived stale time is announced in the BGP long-lived graceful
+	restart capability and specifies how long the neighbor would keep stale
+	routes depreferenced during long-lived graceful restart until either the
+	session is re-stablished and synchronized or the stale time expires and
+	routes are removed. Default: 3600 seconds.
+
 	<tag><label id="bgp-interpret-communities">interpret communities <m/switch/</tag>
 	<rfc id="1997"> demands that BGP speaker should process well-known
 	communities like no-export (65535, 65281) or no-advertise (65535,
@ -2159,6 +2203,19 @@ using the following configuration parameters:
 	disable the instance automatically and wait for an administrator to fix
 	the problem manually. Default: off.

+        <tag><label id="bgp-disable-after-cease">disable after cease <m/switch/|<m/set-of-flags/</tag>
+        When a Cease notification is received, disable the instance
+        automatically and wait for an administrator to fix the problem manually.
+        When used with <m/switch/ argument, it means handle every Cease subtype
+        with the exception of <cf/connection collision/. Default: off.
+
+        The <m/set-of-flags/ allows to narrow down relevant Cease subtypes. The
+        syntax is <cf>{<m/flag/ [, <m/.../] }</cf>, where flags are: <cf/cease/,
+        <cf/prefix limit hit/, <cf/administrative shutdown/,
+        <cf/peer deconfigured/, <cf/administrative reset/,
+        <cf/connection rejected/, <cf/configuration change/,
+        <cf/connection collision/, <cf/out of resources/.
+
 	<tag><label id="bgp-hold-time">hold time <m/number/</tag>
 	Time in seconds to wait for a Keepalive message from the other side
 	before considering the connection stale. Default: depends on agreement
@ -2249,17 +2306,17 @@ using the following configuration parameters:
 some of them (marked with `<tt/O/') are optional.

 <descrip>
-	<tag><label id="rta-bgp-path">bgppath bgp_path/</tag>
+	<tag><label id="rta-bgp-path">bgppath bgp_path</tag>
 	Sequence of AS numbers describing the AS path the packet will travel
 	through when forwarded according to the particular route. In case of
 	internal BGP it doesn't contain the number of the local AS.

-	<tag><label id="rta-bgp-local-pref">int bgp_local_pref/ [I]</tag>
+	<tag><label id="rta-bgp-local-pref">int bgp_local_pref [I]</tag>
 	Local preference value used for selection among multiple BGP routes (see
 	the selection rules above). It's used as an additional metric which is
 	propagated through the whole local AS.

-	<tag><label id="rta-bgp-med">int bgp_med/ [O]</tag>
+	<tag><label id="rta-bgp-med">int bgp_med [O]</tag>
 	The Multiple Exit Discriminator of the route is an optional attribute
 	which is used on external (inter-AS) links to convey to an adjacent AS
 	the optimal entry point into the local AS. The received attribute is
@ -2270,20 +2327,20 @@ some of them (marked with `<tt/O/') are optional.
 	external BGP instance. See <rfc id="4451"> for further discussion of
 	BGP MED attribute.

-	<tag><label id="rta-bgp-origin">enum bgp_origin/</tag>
+	<tag><label id="rta-bgp-origin">enum bgp_origin</tag>
 	Origin of the route: either <cf/ORIGIN_IGP/ if the route has originated
 	in an interior routing protocol or <cf/ORIGIN_EGP/ if it's been imported
 	from the <tt>EGP</tt> protocol (nowadays it seems to be obsolete) or
 	<cf/ORIGIN_INCOMPLETE/ if the origin is unknown.

-	<tag><label id="rta-bgp-next-hop">ip bgp_next_hop/</tag>
+	<tag><label id="rta-bgp-next-hop">ip bgp_next_hop</tag>
 	Next hop to be used for forwarding of packets to this destination. On
 	internal BGP connections, it's an address of the originating router if
 	it's inside the local AS or a boundary router the packet will leave the
 	AS through if it's an exterior route, so each BGP speaker within the AS
 	has a chance to use the shortest interior path possible to this point.

-	<tag><label id="rta-bgp-atomic-aggr">void bgp_atomic_aggr/ [O]</tag>
+	<tag><label id="rta-bgp-atomic-aggr">void bgp_atomic_aggr [O]</tag>
 	This is an optional attribute which carries no value, but the sole
 	presence of which indicates that the route has been aggregated from
 	multiple routes by some router on the path from the originator.
@ -2291,7 +2348,7 @@ some of them (marked with `<tt/O/') are optional.
 <!-- we don't handle aggregators right since they are of a very obscure type
 	<tag>bgp_aggregator</tag>
 -->
-	<tag><label id="rta-bgp-community">clist bgp_community/ [O]</tag>
+	<tag><label id="rta-bgp-community">clist bgp_community [O]</tag>
 	List of community values associated with the route. Each such value is a
 	pair (represented as a <cf/pair/ data type inside the filters) of 16-bit
 	integers, the first of them containing the number of the AS which
@ -2302,14 +2359,14 @@ some of them (marked with `<tt/O/') are optional.
 	freedom about which community attributes it defines and what will their
 	semantics be.

-	<tag><label id="rta-bgp-ext-community">eclist bgp_ext_community/ [O]</tag>
+	<tag><label id="rta-bgp-ext-community">eclist bgp_ext_community [O]</tag>
 	List of extended community values associated with the route. Extended
 	communities have similar usage as plain communities, but they have an
 	extended range (to allow 4B ASNs) and a nontrivial structure with a type
 	field. Individual community values are represented using an <cf/ec/ data
 	type inside the filters.

-	<tag><label id="rta-bgp-large-community">lclist <cf/bgp_large_community/ [O]</tag>
+	<tag><label id="rta-bgp-large-community">lclist bgp_large_community [O]</tag>
 	List of large community values associated with the route. Large BGP
 	communities is another variant of communities, but contrary to extended
 	communities they behave very much the same way as regular communities,
@ -2317,12 +2374,12 @@ some of them (marked with `<tt/O/') are optional.
 	Individual community values are represented using an <cf/lc/ data type
 	inside the filters.

-	<tag><label id="rta-bgp-originator-id">quad bgp_originator_id/ [I, O]</tag>
+	<tag><label id="rta-bgp-originator-id">quad bgp_originator_id [I, O]</tag>
 	This attribute is created by the route reflector when reflecting the
 	route and contains the router ID of the originator of the route in the
 	local AS.

-	<tag><label id="rta-bgp-cluster-list">clist bgp_cluster_list/ [I, O]</tag>
+	<tag><label id="rta-bgp-cluster-list">clist bgp_cluster_list [I, O]</tag>
 	This attribute contains a list of cluster IDs of route reflectors. Each
 	route reflector prepends its cluster ID when reflecting the route.
 </descrip>
@ -2557,26 +2614,26 @@ translated to appropriate system (and OS-specific) route attributes. We support
 these attributes:

 <descrip>
-	<tag><label id="rta-krt-source">int krt_source/</tag>
+	<tag><label id="rta-krt-source">int krt_source</tag>
 	The original source of the imported kernel route. The value is
 	system-dependent. On Linux, it is a value of the protocol field of the
 	route. See /etc/iproute2/rt_protos for common values. On BSD, it is
 	based on STATIC and PROTOx flags. The attribute is read-only.

-	<tag><label id="rta-krt-metric">int krt_metric/</tag> (Linux)
+	<tag><label id="rta-krt-metric">int krt_metric</tag> (Linux)
 	The kernel metric of the route. When multiple same routes are in a
 	kernel routing table, the Linux kernel chooses one with lower metric.
 	Note that preferred way to set kernel metric is to use protocol option
 	<cf/metric/, unless per-route metric values are needed.

-	<tag><label id="rta-krt-prefsrc">ip krt_prefsrc/</tag> (Linux)
+	<tag><label id="rta-krt-prefsrc">ip krt_prefsrc</tag> (Linux)
 	The preferred source address. Used in source address selection for
 	outgoing packets. Has to be one of the IP addresses of the router.

-	<tag><label id="rta-krt-realm">int krt_realm/</tag> (Linux)
+	<tag><label id="rta-krt-realm">int krt_realm</tag> (Linux)
 	The realm of the route. Can be used for traffic classification.

-	<tag><label id="rta-krt-scope">int krt_scope/</tag> (Linux IPv4)
+	<tag><label id="rta-krt-scope">int krt_scope</tag> (Linux IPv4)
 	The scope of the route. Valid values are 0-254, although Linux kernel
 	may reject some values depending on route type and nexthop. It is
 	supposed to represent `indirectness' of the route, where nexthops of
@ -2630,6 +2687,81 @@ protocol kernel {		# Secondary routing table
 </code>


+<sect>MRT
+<label id="mrt">
+
+<sect1>Introduction
+<label id="mrt-intro">
+
+<p>The MRT protocol is a component responsible for handling the Multi-Threaded
+Routing Toolkit (MRT) routing information export format, which is mainly used
+for collecting and analyzing of routing information from BGP routers. The MRT
+protocol can be configured to do periodic dumps of routing tables, created MRT
+files can be analyzed later by other tools. Independent MRT table dumps can also
+be requested from BIRD client. There is also a feature to save incoming BGP
+messages in MRT files, but it is controlled by <ref id="proto-mrtdump"
+name="mrtdump"> options independently of MRT protocol, although that might
+change in the future.
+
+BIRD implements the main MRT format specification as defined in <rfc id="6396">
+and the ADD_PATH extension (<rfc id="8050">).
+
+<sect1>Configuration
+<label id="mrt-config">
+
+<p>MRT configuration consists of several statements describing routing table
+dumps. Multiple independent periodic dumps can be done as multiple MRT protocol
+instances. There are two mandatory statements: <cf/filename/ and <cf/period/.
+The behavior can be modified by following configuration parameters:
+
+<descrip>
+	<tag><label id="mrt-table">table <m/name/ | "<m/pattern/"</tag>
+	Specify a routing table (or a set of routing tables described by a
+	wildcard pattern) that are to be dumped by the MRT protocol instance.
+	Default: the master table.
+
+	<tag><label id="mrt-filter">filter { <m/filter commands/ }</tag>
+	The MRT protocol allows to specify a filter that is applied to routes as
+	they are dumped. Rejected routes are ignored and not saved to the MRT
+	dump file. Default: no filter.
+
+	<tag><label id="mrt-where">where <m/filter expression/</tag>
+	An alternative way to specify a filter for the MRT protocol.
+
+	<tag><label id="mrt-filename">filename "<m/filename/"</tag>
+	Specify a filename for MRT dump files. The filename may contain time
+	format sequences with <it/strftime(3)/ notation (see <it/man strftime/
+	for details), there is also a sequence "%N" that is expanded to the name
+	of dumped table. Therefore, each periodic dump of each table can be
+	saved to a different file. Mandatory, see example below.
+
+	<tag><label id="mrt-period">period <m/number/</tag>
+	Specify the time interval (in seconds) between periodic dumps.
+	Mandatory.
+
+	<tag><label id="mrt-always-add-path">always add path <m/switch/</tag>
+	The MRT format uses special records (specified in <rfc id="8050">) for
+	routes received using BGP ADD_PATH extension to keep Path ID, while
+	other routes use regular records. This has advantage of better
+	compatibility with tools that do not know special records, but it loses
+	information about which route is the best route. When this option is
+	enabled, both ADD_PATH and non-ADD_PATH routes are stored in ADD_PATH
+	records and order of routes for network is preserved. Default: disabled.
+</descrip>
+
+<sect1>Example
+<label id="mrt-exam">
+
+<p><code>
+protocol mrt {
+	table "tab*";
+	where source = RTS_BGP;
+	filename "/var/log/bird/%N_%F_%T.mrt";
+	period 300;
+}
+</code>
+
+
 <sect>OSPF
 <label id="ospf">

@ -2936,6 +3068,11 @@ protocol ospf &lt;name&gt; {
 	Specifies interval in seconds between retransmissions of unacknowledged
 	updates. Default value is 5.

+	<tag><label id="ospf-transmit-delay">transmit delay <M>num</M></tag>
+	Specifies estimated transmission delay of link state updates send over
+	the interface. The value is added to LSA age of LSAs propagated through
+	it. Default value is 1.
+
 	<tag><label id="ospf-priority">priority <M>num</M></tag>
 	On every multiple access network (e.g., the Ethernet) Designated Router
 	and Backup Designated router are elected. These routers have some special
@ -3386,7 +3523,7 @@ definitions, prefix definitions and DNS definitions:
 	RAdv protocol could be configured to change its behavior based on
 	availability of routes. When this option is used, the protocol waits in
 	suppressed state until a <it/trigger route/ (for the specified network)
-	is exported to the protocol, the protocol also returnsd to suppressed
+	is exported to the protocol, the protocol also returns to suppressed
 	state if the <it/trigger route/ disappears. Note that route export
 	depends on specified export filter, as usual. This option could be used,
 	e.g., for handling failover in multihoming scenarios.
@ -3580,13 +3717,13 @@ definitions, prefix definitions and DNS definitions:
 <p>RAdv defines two route attributes:

 <descrip>
-	<tag><label id="rta-ra-preference">enum ra_preference/</tag>
+	<tag><label id="rta-ra-preference">enum ra_preference</tag>
 	The preference of the route. The value can be <it/RA_PREF_LOW/,
 	<it/RA_PREF_MEDIUM/ or <it/RA_PREF_HIGH/. If the attribute is not set,
 	the <ref id="radv-iface-route-preference" name="route preference">
 	option is used.

-	<tag><label id="rta-ra-lifetime">int ra_lifetime/</tag>
+	<tag><label id="rta-ra-lifetime">int ra_lifetime</tag>
 	The advertised lifetime of the route, in seconds. The special value of
 	0xffffffff represents infinity. If the attribute is not set, the
 	<ref id="radv-iface-route-lifetime" name="route lifetime">
@ -3888,13 +4025,13 @@ protocol rip [&lt;name&gt;] {
 <p>RIP defines two route attributes:

 <descrip>
-	<tag><label id="rta-rip-metric">int rip_metric/</tag>
+	<tag><label id="rta-rip-metric">int rip_metric</tag>
 	RIP metric of the route (ranging from 0 to <cf/infinity/).  When routes
 	from different RIP instances are available and all of them have the same
 	preference, BIRD prefers the route with lowest <cf/rip_metric/. When a
 	non-RIP route is exported to RIP, the default metric is 1.

-	<tag><label id="rta-rip-tag">int rip_tag/</tag>
+	<tag><label id="rta-rip-tag">int rip_tag</tag>
 	RIP route tag: a 16-bit number which can be used to carry additional
 	information with the route (for example, an originating AS number in
 	case of external routes). When a non-RIP route is exported to RIP, the
--- a/doc/prog-root
+++ b/doc/prog-root
@ -0,0 +1,9 @@
+D doc/prog-head.sgml
+D doc/prog-intro.sgml
+C nest
+C conf
+C filter
+C proto
+C sysdep
+C lib
+D doc/prog-foot.sgml
--- a/doc/sgml2html
+++ b/doc/sgml2html
@ -34,7 +34,12 @@ $progs = {
  "GROFFMACRO" => "-ms",
  "AWK" => "/usr/share/linuxdoc-tools/awkwhich"
 };
-$ENV{"SGML_CATALOG_FILES"} = "$DataDir/dtd/catalog";
+
+if (! -x $progs->{"NSGMLS"})
+  { $progs->{"NSGMLS"} = "/usr/bin/onsgmls"; }
+
+$ENV{"SGML_CATALOG_FILES"} = "$DataDir/dtd/catalog" .
+  (defined $ENV{SGML_CATALOG_FILES} ? ":$ENV{SGML_CATALOG_FILES}" : "");

 require "$FindBin::Bin/LinuxDocTools.pm";
 &LinuxDocTools::init;
--- a/doc/sgml2latex
+++ b/doc/sgml2latex
@ -34,7 +34,12 @@ $progs = {
  "GROFFMACRO" => "-ms",
  "AWK" => "/usr/share/linuxdoc-tools/awkwhich"
 };
-$ENV{"SGML_CATALOG_FILES"} = "$DataDir/dtd/catalog";
+
+if (! -x $progs->{"NSGMLS"})
+  { $progs->{"NSGMLS"} = "/usr/bin/onsgmls"; }
+
+$ENV{"SGML_CATALOG_FILES"} = "$DataDir/dtd/catalog" .
+  (defined $ENV{SGML_CATALOG_FILES} ? ":$ENV{SGML_CATALOG_FILES}" : "");

 require "$FindBin::Bin/LinuxDocTools.pm";
 &LinuxDocTools::init;
--- a/doc/sgml2txt
+++ b/doc/sgml2txt
@ -34,7 +34,12 @@ $progs = {
  "GROFFMACRO" => "-ms",
  "AWK" => "/usr/share/linuxdoc-tools/awkwhich"
 };
-$ENV{"SGML_CATALOG_FILES"} = "$DataDir/dtd/catalog";
+
+if (! -x $progs->{"NSGMLS"})
+  { $progs->{"NSGMLS"} = "/usr/bin/onsgmls"; }
+
+$ENV{"SGML_CATALOG_FILES"} = "$DataDir/dtd/catalog" .
+  (defined $ENV{SGML_CATALOG_FILES} ? ":$ENV{SGML_CATALOG_FILES}" : "");

 require "$FindBin::Bin/LinuxDocTools.pm";
 &LinuxDocTools::init;
--- a/filter/config.Y
+++ b/filter/config.Y
@ -320,7 +320,8 @@ f_generate_lc(struct f_inst *t1, struct f_inst *t2, struct f_inst *t3)
 static inline struct f_inst *
 f_generate_path_mask(struct f_path_mask *t)
 {
-  for (struct f_path_mask *tt = t; tt; tt = tt->next) {
+  struct f_path_mask *tt;
+  for (tt = t; tt; tt = tt->next) {
    if (tt->kind == PM_ASN_EXPR) {
      struct f_inst *mrv = f_new_inst(FI_PATHMASK_CONSTRUCT);
      mrv->a1.p = t;
@ -791,7 +792,7 @@ static_attr:
 | SCOPE   { $$ = f_new_static_attr(T_ENUM_SCOPE, SA_SCOPE,	1); }
 | CAST    { $$ = f_new_static_attr(T_ENUM_RTC,   SA_CAST,	0); }
 | DEST    { $$ = f_new_static_attr(T_ENUM_RTD,   SA_DEST,	1); }
- | IFNAME  { $$ = f_new_static_attr(T_STRING,     SA_IFNAME,	0); }
+ | IFNAME  { $$ = f_new_static_attr(T_STRING,     SA_IFNAME,	1); }
 | IFINDEX { $$ = f_new_static_attr(T_INT,        SA_IFINDEX,	0); }
 ;

--- a/filter/filter.c
+++ b/filter/filter.c
@ -1000,6 +1000,20 @@ interpret(struct f_inst *what)
 	rta->hostentry = NULL;
 	break;

+      case SA_IFNAME:
+	{
+	  struct iface *ifa = if_find_by_name(v1.val.s);
+	  if (!ifa)
+	    runtime( "Invalid iface name" );
+
+	  rta->dest = RTD_DEVICE;
+	  rta->gw = IPA_NONE;
+	  rta->iface = ifa;
+	  rta->nexthops = NULL;
+	  rta->hostentry = NULL;
+	}
+	break;
+
      default:
 	bug("Invalid static attribute access (%x)", res.type);
      }
@ -1513,7 +1527,7 @@ interpret(struct f_inst *what)
      /* 0x02 is a value of BA_AS_PATH, we don't want to include BGP headers */
      eattr *e = ea_find((*f_rte)->attrs->eattrs, EA_CODE(EAP_BGP, 0x02));

-      if (!e || e->type != EAF_TYPE_AS_PATH)
+      if (!e || ((e->type & EAF_TYPE_MASK) != EAF_TYPE_AS_PATH))
 	runtime("Missing AS_PATH attribute");

      as_path_get_last(e->u.ptr, &as);
--- a/lib/event.h
+++ b/lib/event.h
@ -36,5 +36,14 @@ ev_active(event *e)
  return e->n.next != NULL;
 }

+static inline event*
+ev_new_set(pool *p, void (*hook)(void *), void *data)
+{
+  event *e = ev_new(p);
+  e->hook = hook;
+  e->data = data;
+  return e;
+}
+

 #endif
--- a/lib/ip.h
+++ b/lib/ip.h
@ -67,7 +67,7 @@ typedef struct ip4_addr {

 typedef u32 ip4_addr;

-#define _MI4(x) (x)
+#define _MI4(x) ((u32) (x))
 #define _I(x) (x)

 #endif
@ -99,6 +99,7 @@ typedef ip6_addr ip_addr;
 #define ipa_to_u32(x) ip4_to_u32(ipa_to_ip4(x))

 #define ipa_is_ip4(a) ip6_is_v4mapped(a)
+#define ipa_is_ip6(a) (! ip6_is_v4mapped(a))

 #else

@ -115,6 +116,7 @@ typedef ip4_addr ip_addr;
 #define ipa_to_u32(x) ip4_to_u32(ipa_to_ip4(x))

 #define ipa_is_ip4(a) 1
+#define ipa_is_ip6(a) 0

 #endif

@ -309,6 +311,9 @@ static inline int ip6_is_v4mapped(ip6_addr a)
 #define ipa_is_link_local(x) 0
 #endif

+static inline int ip4_is_unicast(ip4_addr a)
+{ return _I(a) < 0xe0000000; }
+
 static inline int ipa_classify_net(ip_addr a)
 { return ipa_zero2(a) ? (IADDR_HOST | SCOPE_UNIVERSE) : ipa_classify(a); }

--- a/lib/printf.c
+++ b/lib/printf.c
@ -314,14 +314,14 @@ int bvsnprintf(char *buf, int size, const char *fmt, va_list args)
 			if(qualifier == 'l') {
 				X = va_arg(args, u64);
 				bsprintf(ipbuf, "%02x:%02x:%02x:%02x:%02x:%02x:%02x:%02x",
-					((X >> 56) & 0xff),
-					((X >> 48) & 0xff),
-					((X >> 40) & 0xff),
-					((X >> 32) & 0xff),
-					((X >> 24) & 0xff),
-					((X >> 16) & 0xff),
-					((X >> 8) & 0xff),
-					(X & 0xff));
+					 (uint) ((X >> 56) & 0xff),
+					 (uint) ((X >> 48) & 0xff),
+					 (uint) ((X >> 40) & 0xff),
+					 (uint) ((X >> 32) & 0xff),
+					 (uint) ((X >> 24) & 0xff),
+					 (uint) ((X >> 16) & 0xff),
+					 (uint) ((X >> 8) & 0xff),
+					 (uint) (X & 0xff));
 			}
 			else
 			{
@ -342,6 +342,7 @@ int bvsnprintf(char *buf, int size, const char *fmt, va_list args)

 		case 'X':
 			flags |= LARGE;
+			/* fallthrough */
 		case 'x':
 			base = 16;
 			break;
--- a/lib/unaligned.h
+++ b/lib/unaligned.h
@ -44,6 +44,12 @@ get_u64(const void *p)
  return (((u64) ntohl(xh)) << 32) | ntohl(xl);
 }

+static inline void
+put_u8(void *p, u8 x)
+{
+  memcpy(p, &x, 1);
+}
+
 static inline void
 put_u16(void *p, u16 x)
 {
--- a/misc/bird.spec
+++ b/misc/bird.spec
@ -1,6 +1,6 @@
 Summary: BIRD Internet Routing Daemon
 Name: bird
-Version: 1.6.3
+Version: 1.6.7
 Release: 1
 Copyright: GPL
 Group: Networking/Daemons
--- a/nest/bfd.h
+++ b/nest/bfd.h
@ -19,6 +19,7 @@ struct bfd_request {
  ip_addr addr;
  ip_addr local;
  struct iface *iface;
+  struct iface *vrf;

  void (*hook)(struct bfd_request *);
  void *data;
@ -40,13 +41,13 @@ struct bfd_request {

 #ifdef CONFIG_BFD

-struct bfd_request * bfd_request_session(pool *p, ip_addr addr, ip_addr local, struct iface *iface, void (*hook)(struct bfd_request *), void *data);
+struct bfd_request * bfd_request_session(pool *p, ip_addr addr, ip_addr local, struct iface *iface, struct iface *vrf, void (*hook)(struct bfd_request *), void *data);

 static inline void cf_check_bfd(int use UNUSED) { }

 #else

-static inline struct bfd_request * bfd_request_session(pool *p, ip_addr addr, ip_addr local, struct iface *iface, void (*hook)(struct bfd_request *), void *data) { return NULL; }
+static inline struct bfd_request * bfd_request_session(pool *p, ip_addr addr, ip_addr local, struct iface *iface, struct iface *vrf, void (*hook)(struct bfd_request *), void *data) { return NULL; }

 static inline void cf_check_bfd(int use) { if (use) cf_error("BFD not available"); }

--- a/nest/config.Y
+++ b/nest/config.Y
@ -55,7 +55,7 @@ get_passwords(void)
 CF_DECLS

 CF_KEYWORDS(ROUTER, ID, PROTOCOL, TEMPLATE, PREFERENCE, DISABLED, DEBUG, ALL, OFF, DIRECT)
-CF_KEYWORDS(INTERFACE, IMPORT, EXPORT, FILTER, NONE, VRF, TABLE, STATES, ROUTES, FILTERS)
+CF_KEYWORDS(INTERFACE, IMPORT, EXPORT, FILTER, NONE, VRF, DEFAULT, TABLE, STATES, ROUTES, FILTERS)
 CF_KEYWORDS(RECEIVE, LIMIT, ACTION, WARN, BLOCK, RESTART, DISABLE, KEEP, FILTERED)
 CF_KEYWORDS(PASSWORD, FROM, PASSIVE, TO, ID, EVENTS, PACKETS, PROTOCOLS, INTERFACES)
 CF_KEYWORDS(ALGORITHM, KEYED, HMAC, MD5, SHA1, SHA256, SHA384, SHA512)
@ -227,7 +227,8 @@ proto_item:
 | IMPORT LIMIT limit_spec { this_proto->in_limit = $3; }
 | EXPORT LIMIT limit_spec { this_proto->out_limit = $3; }
 | IMPORT KEEP FILTERED bool { this_proto->in_keep_filtered = $4; }
- | VRF text { this_proto->vrf = if_get_by_name($2); }
+ | VRF text { this_proto->vrf = if_get_by_name($2); this_proto->vrf_set = 1; }
+ | VRF DEFAULT { this_proto->vrf = NULL; this_proto->vrf_set = 1; }
 | TABLE rtable { this_proto->table = $2; }
 | ROUTER ID idval { this_proto->router_id = $3; }
 | DESCRIPTION text { this_proto->dsc = $2; }
--- a/nest/iface.c
+++ b/nest/iface.c
@ -140,7 +140,7 @@ if_copy(struct iface *to, struct iface *from)
 static inline void
 ifa_send_notify(struct proto *p, unsigned c, struct ifa *a)
 {
-  if (p->ifa_notify && (!p->vrf || p->vrf == a->iface->master))
+  if (p->ifa_notify && (!p->vrf_set || p->vrf == a->iface->master))
    {
      if (p->debug & D_IFACES)
 	log(L_TRACE "%s <%s address %I/%d on interface %s %s",
@ -177,7 +177,7 @@ ifa_notify_change(unsigned c, struct ifa *a)
 static inline void
 if_send_notify(struct proto *p, unsigned c, struct iface *i)
 {
-  if (p->if_notify && (!p->vrf || p->vrf == i->master))
+  if (p->if_notify && (!p->vrf_set || p->vrf == i->master))
    {
      if (p->debug & D_IFACES)
 	log(L_TRACE "%s < interface %s %s", p->name, i->name,
@ -441,7 +441,7 @@ if_find_by_name(char *name)
  struct iface *i;

  WALK_LIST(i, iface_list)
-    if (!strcmp(i->name, name))
+    if (!strcmp(i->name, name) && !(i->flags & IF_SHUTDOWN))
      return i;
  return NULL;
 }
@ -451,8 +451,9 @@ if_get_by_name(char *name)
 {
  struct iface *i;

-  if (i = if_find_by_name(name))
-    return i;
+  WALK_LIST(i, iface_list)
+    if (!strcmp(i->name, name))
+      return i;

  /* No active iface, create a dummy */
  i = mb_allocz(if_pool, sizeof(struct iface));
@ -469,10 +470,24 @@ struct ifa *kif_choose_primary(struct iface *i);
 static int
 ifa_recalc_primary(struct iface *i)
 {
-  struct ifa *a = kif_choose_primary(i);
+  struct ifa *a;
+  int c = 0;
+
+#ifdef IPV6
+  struct ifa *ll = NULL;
+
+  WALK_LIST(a, i->addrs)
+    if (ipa_is_link_local(a->ip) && (!ll || (a == i->llv6)))
+      ll = a;
+
+  c = (ll != i->llv6);
+  i->llv6 = ll;
+#endif
+
+  a = kif_choose_primary(i);

  if (a == i->addr)
-    return 0;
+    return c;

  if (i->addr)
    i->addr->flags &= ~IA_PRIMARY;
@ -576,7 +591,7 @@ ifa_delete(struct ifa *a)
 	    b->flags &= ~IF_UP;
 	    ifa_notify_change(IF_CHANGE_DOWN, b);
 	  }
-	if (b->flags & IA_PRIMARY)
+	if ((b->flags & IA_PRIMARY) || (b == ifa_llv6(i)))
 	  {
 	    if_change_flags(i, i->flags | IF_TMP_DOWN);
 	    ifa_recalc_primary(i);
@ -812,6 +827,9 @@ if_show_summary(void)
  cli_msg(-2005, "interface state address");
  WALK_LIST(i, iface_list)
    {
+      if (i->flags & IF_SHUTDOWN)
+	continue;
+
      if (i->addr)
 	bsprintf(addr, "%I/%d", i->addr->ip, i->addr->pxlen);
      else
--- a/nest/iface.h
+++ b/nest/iface.h
@ -37,6 +37,9 @@ struct iface {
  unsigned master_index;		/* Interface index of master iface */
  list addrs;				/* Addresses assigned to this interface */
  struct ifa *addr;			/* Primary address */
+#ifdef IPV6
+  struct ifa *llv6;			/* Selected IPv6 link-local address */
+#endif
  struct iface *master;			/* Master iface (e.g. for VRF) */
  list neighbors;			/* All neighbors on this interface */
 };
@ -103,6 +106,16 @@ struct iface *if_find_by_name(char *);
 struct iface *if_get_by_name(char *);
 void ifa_recalc_all_primary_addresses(void);

+static inline struct ifa *
+ifa_llv6(struct iface *i UNUSED4)
+{
+#ifdef IPV6
+  return i->llv6;
+#else
+  return NULL;
+#endif
+}
+

 /* The Neighbor Cache */

--- a/nest/mrtdump.h
+++ b/nest/mrtdump.h
@ -1,31 +0,0 @@
-/*
- *	BIRD -- MRTdump handling
- *
- *
- *	Can be freely distributed and used under the terms of the GNU GPL.
- */
-
-#ifndef MRTDUMP_H
-#define MRTDUMP_H
-#include "nest/protocol.h"
-
-/* MRTDump values */
-
-#define MRTDUMP_HDR_LENGTH	12
-
-/* MRTdump types */
-
-#define BGP4MP			16
-
-/* MRTdump subtypes */
-
-#define BGP4MP_MESSAGE		1
-#define BGP4MP_MESSAGE_AS4	4
-#define BGP4MP_STATE_CHANGE_AS4	5
-
-
-/* implemented in sysdep */
-void mrt_dump_message(struct proto *p, u16 type, u16 subtype, byte *buf, u32 len);
-
-#endif
-
--- a/nest/neighbor.c
+++ b/nest/neighbor.c
@ -153,7 +153,7 @@ neigh_find2(struct proto *p, ip_addr *a, struct iface *ifa, unsigned flags)
    }
  else
    WALK_LIST(i, iface_list)
-      if ((!p->vrf || p->vrf == i->master) &&
+      if ((!p->vrf_set || p->vrf == i->master) &&
 	  ((scope = if_connected(a, i, &addr)) >= 0))
        {
 	  ifa = i;
--- a/nest/proto.c
+++ b/nest/proto.c
@ -387,6 +387,7 @@ proto_init(struct proto_config *c)
  q->export_state = ES_DOWN;
  q->last_state_change = now;
  q->vrf = c->vrf;
+  q->vrf_set = c->vrf_set;

  add_tail(&initial_proto_list, &q->n);

@ -411,6 +412,7 @@ proto_reconfigure(struct proto *p, struct proto_config *oc, struct proto_config
  if ((nc->protocol != oc->protocol) ||
      (nc->disabled != p->disabled) ||
      (nc->vrf != oc->vrf) ||
+      (nc->vrf_set != oc->vrf_set) ||
      (nc->table->table != oc->table->table))
    return 0;

@ -433,10 +435,17 @@ proto_reconfigure(struct proto *p, struct proto_config *oc, struct proto_config
  if (p->proto->multitable)
    return 1;

+  int import_changed = ! filter_same(nc->in_filter, oc->in_filter);
+  int export_changed = ! filter_same(nc->out_filter, oc->out_filter);
+
+  /* We treat a change in preferences by reimporting routes */
+  if (nc->preference != oc->preference)
+    import_changed = 1;
+
  /* Update filters and limits in the main announce hook
     Note that this also resets limit state */
  if (p->main_ahook)
-    {  
+    {
      struct announce_hook *ah = p->main_ahook;
      ah->in_filter = nc->in_filter;
      ah->out_filter = nc->out_filter;
@ -445,6 +454,9 @@ proto_reconfigure(struct proto *p, struct proto_config *oc, struct proto_config
      ah->out_limit = nc->out_limit;
      ah->in_keep_filtered = nc->in_keep_filtered;
      proto_verify_limits(ah);
+
+      if (export_changed)
+	ah->last_out_filter_change = now;
    }

  /* Update routes when filters changed. If the protocol in not UP,
@ -452,13 +464,6 @@ proto_reconfigure(struct proto *p, struct proto_config *oc, struct proto_config
  if ((p->proto_state != PS_UP) || (type == RECONFIG_SOFT))
    return 1;

-  int import_changed = ! filter_same(nc->in_filter, oc->in_filter);
-  int export_changed = ! filter_same(nc->out_filter, oc->out_filter);
-
-  /* We treat a change in preferences by reimporting routes */
-  if (nc->preference != oc->preference)
-    import_changed = 1;
-
  if (import_changed || export_changed)
    log(L_INFO "Reloading protocol %s", p->name);

@ -910,6 +915,9 @@ protos_build(void)
 #ifdef CONFIG_STATIC
  proto_build(&proto_static);
 #endif
+#ifdef CONFIG_MRT
+  proto_build(&proto_mrt);
+#endif
 #ifdef CONFIG_OSPF
  proto_build(&proto_ospf);
 #endif
@ -1561,6 +1569,9 @@ proto_cmd_show(struct proto *p, uintptr_t verbose, int cnt)
      if (p->cf->router_id)
 	cli_msg(-1006, "  Router ID:      %R", p->cf->router_id);

+      if (p->vrf_set)
+	cli_msg(-1006, "  VRF:            %s", p->vrf ? p->vrf->name : "default");
+
      if (p->proto->show_proto_info)
 	p->proto->show_proto_info(p);
      else
--- a/nest/protocol.h
+++ b/nest/protocol.h
@ -76,7 +76,7 @@ void protos_dump_all(void);
 */

 extern struct protocol
-  proto_device, proto_radv, proto_rip, proto_static,
+  proto_device, proto_radv, proto_rip, proto_static, proto_mrt,
  proto_ospf, proto_pipe, proto_bgp, proto_bfd, proto_babel;

 /*
@ -93,6 +93,7 @@ struct proto_config {
  int class;				/* SYM_PROTO or SYM_TEMPLATE */
  u32 debug, mrtdump;			/* Debugging bitfields, both use D_* constants */
  unsigned preference, disabled;	/* Generic parameters */
+  int vrf_set;				/* Related VRF instance (below) is defined */
  int in_keep_filtered;			/* Routes rejected in import filter are kept */
  u32 router_id;			/* Protocol specific router ID */
  struct iface *vrf;			/* Related VRF instance, NULL if global */
@ -149,6 +150,7 @@ struct proto {
  unsigned preference;			/* Default route preference */
  byte accept_ra_types;			/* Which types of route announcements are accepted (RA_OPTIMAL or RA_ANY) */
  byte disabled;			/* Manually disabled */
+  byte vrf_set;				/* Related VRF instance (above) is defined */
  byte proto_state;			/* Protocol state machine (PS_*, see below) */
  byte core_state;			/* Core state machine (FS_*, see below) */
  byte export_state;			/* Route export state (ES_*, see below) */
@ -213,6 +215,7 @@ struct proto {
  int (*rte_better)(struct rte *, struct rte *);
  int (*rte_same)(struct rte *, struct rte *);
  int (*rte_mergable)(struct rte *, struct rte *);
+  struct rte * (*rte_modify)(struct rte *, struct linpool *);
  void (*rte_insert)(struct network *, struct rte *);
  void (*rte_remove)(struct network *, struct rte *);

@ -471,6 +474,7 @@ struct announce_hook {
  struct proto_stats *stats;		/* Per-table protocol statistics */
  struct announce_hook *next;		/* Next hook for the same protocol */
  int in_keep_filtered;			/* Routes rejected in import filter are kept */
+  bird_clock_t last_out_filter_change;	/* Last time when out_filter _changed_ */
 };

 struct announce_hook *proto_add_announce_hook(struct proto *p, struct rtable *t, struct proto_stats *stats);
--- a/nest/route.h
+++ b/nest/route.h
@ -219,6 +219,7 @@ typedef struct rte {
 #ifdef CONFIG_BGP
    struct {
      u8 suppressed;			/* Used for deterministic MED comparison */
+      s8 stale;				/* Route is LLGR_STALE, -1 if unknown */
    } bgp;
 #endif
 #ifdef CONFIG_BABEL
@ -241,6 +242,7 @@ typedef struct rte {
 #define REF_FILTERED	2		/* Route is rejected by import filter */
 #define REF_STALE	4		/* Route is stale in a refresh cycle */
 #define REF_DISCARD	8		/* Route is scheduled for discard */
+#define REF_MODIFY	16		/* Route is scheduled for modify */

 /* Route is valid for propagation (may depend on other flags in the future), accepts NULL */
 static inline int rte_is_valid(rte *r) { return r && !(r->flags & REF_FILTERED); }
@ -261,6 +263,7 @@ static inline int rte_is_filtered(rte *r) { return !!(r->flags & REF_FILTERED);
 #define RIC_REJECT	-1		/* Rejected by protocol */
 #define RIC_DROP	-2		/* Silently dropped by protocol */

+extern list routing_tables;
 struct config;

 void rt_init(void);
@ -279,6 +282,7 @@ int rt_examine(rtable *t, ip_addr prefix, int pxlen, struct proto *p, struct fil
 rte *rt_export_merged(struct announce_hook *ah, net *net, rte **rt_free, struct ea_list **tmpa, linpool *pool, int silent);
 void rt_refresh_begin(rtable *t, struct announce_hook *ah);
 void rt_refresh_end(rtable *t, struct announce_hook *ah);
+void rt_modify_stale(rtable *t, struct announce_hook *ah);
 void rte_dump(rte *);
 void rte_free(rte *);
 rte *rte_do_cow(rte *);
--- a/nest/rt-attr.c
+++ b/nest/rt-attr.c
@ -1251,7 +1251,7 @@ void
 rta_show(struct cli *c, rta *a, ea_list *eal)
 {
  static char *src_names[] = { "dummy", "static", "inherit", "device", "static-device", "redirect",
-			       "RIP", "OSPF", "OSPF-IA", "OSPF-E1", "OSPF-E2", "BGP", "pipe" };
+			       "RIP", "OSPF", "OSPF-IA", "OSPF-E1", "OSPF-E2", "BGP", "pipe", "babel" };
  static char *cast_names[] = { "unicast", "broadcast", "multicast", "anycast" };
  int i;

--- a/nest/rt-table.c
+++ b/nest/rt-table.c
@ -48,7 +48,7 @@ pool *rt_table_pool;
 static slab *rte_slab;
 static linpool *rte_update_pool;

-static list routing_tables;
+list routing_tables;

 static byte *rt_format_via(rte *e);
 static void rt_free_hostcache(rtable *tab);
@ -426,7 +426,8 @@ rt_notify_basic(struct announce_hook *ah, net *net, rte *new0, rte *old0, int re
   * reconfiguration and the end of refeed - if a newly filtered
   * route disappears during this period, proper withdraw is not
   * sent (because old would be also filtered) and the route is
-   * not refeeded (because it disappeared before that).
+   * not refeeded (because it disappeared before that). This is
+   * handled below as a special case.
   */

  if (new)
@ -439,19 +440,34 @@ rt_notify_basic(struct announce_hook *ah, net *net, rte *new0, rte *old0, int re
  {
    /*
     * As mentioned above, 'old' value may be incorrect in some race conditions.
-     * We generally ignore it with the exception of withdraw to pipe protocol.
-     * In that case we rather propagate unfiltered withdraws regardless of
-     * export filters to ensure that when a protocol is flushed, its routes are
-     * removed from all tables. Possible spurious unfiltered withdraws are not
-     * problem here as they are ignored if there is no corresponding route at
-     * the other end of the pipe. We directly call rt_notify() hook instead of
+     * We generally ignore it with two exceptions:
+     *
+     * First, withdraw to pipe protocol. In that case we rather propagate
+     * unfiltered withdraws regardless of export filters to ensure that when a
+     * protocol is flushed, its routes are removed from all tables. Possible
+     * spurious unfiltered withdraws are not problem here as they are ignored if
+     * there is no corresponding route at the other end of the pipe.
+     *
+     * Second, recent filter change. If old route is older than filter change,
+     * then it was previously evaluated by a different filter and we do not know
+     * whether it was really propagated. In that case we rather send spurious
+     * withdraw than do nothing and possibly cause phantom routes.
+     *
+     * In both cases wqe directly call rt_notify() hook instead of
     * do_rt_notify() to avoid logging and stat counters.
     */

+    int pipe_withdraw = 0, filter_change = 0;
 #ifdef CONFIG_PIPE
-    if ((p->proto == &proto_pipe) && !new0 && (p != old0->sender->proto))
-      p->rt_notify(p, ah->table, net, NULL, old0, NULL);
+    pipe_withdraw = (p->proto == &proto_pipe) && !new0;
 #endif
+    filter_change = old0 && (old0->lastmod <= ah->last_out_filter_change);
+
+    if ((pipe_withdraw || filter_change) && (p != old0->sender->proto))
+    {
+      stats->exp_withdraws_accepted++;
+      p->rt_notify(p, ah->table, net, NULL, old0, NULL);
+    }

    return;
  }
@ -468,7 +484,7 @@ rt_notify_basic(struct announce_hook *ah, net *net, rte *new0, rte *old0, int re
 static void
 rt_notify_accepted(struct announce_hook *ah, net *net, rte *new_changed, rte *old_changed, rte *before_old, int feed)
 {
-  // struct proto *p = ah->proto;
+  struct proto *p = ah->proto;
  struct proto_stats *stats = ah->stats;

  rte *r;
@ -541,8 +557,20 @@ rt_notify_accepted(struct announce_hook *ah, net *net, rte *new_changed, rte *ol
   *
   * - We found new_best the same as new_changed, therefore it cannot
   *   be old_best and we have to continue search for old_best.
+   *
+   * There is also a hack to ensure consistency in case of changed filters.
+   * It does not find the proper old_best, just selects a non-NULL route.
   */

+  /* Hack for changed filters */
+  if (old_changed &&
+      (p != old_changed->sender->proto) &&
+      (old_changed->lastmod <= ah->last_out_filter_change))
+    {
+      old_best = old_changed;
+      goto found;
+    }
+
  /* First case */
  if (old_meet)
    if (old_best = export_filter(ah, old_changed, &old_free, NULL, 1))
@ -837,12 +865,13 @@ rte_free_quick(rte *e)
 static int
 rte_same(rte *x, rte *y)
 {
+  /* rte.flags are not checked, as they are mostly internal to rtable */
  return
    x->attrs == y->attrs &&
-    x->flags == y->flags &&
    x->pflags == y->pflags &&
    x->pref == y->pref &&
-    (!x->attrs->src->proto->rte_same || x->attrs->src->proto->rte_same(x, y));
+    (!x->attrs->src->proto->rte_same || x->attrs->src->proto->rte_same(x, y)) &&
+    rte_is_filtered(x) == rte_is_filtered(y);
 }

 static inline int rte_is_ok(rte *e) { return e && !rte_is_filtered(e); }
@ -886,7 +915,9 @@ rte_recalculate(struct announce_hook *ah, net *net, rte *new, struct rte_src *sr

 	  if (new && rte_same(old, new))
 	    {
-	      /* No changes, ignore the new route */
+	      /* No changes, ignore the new route and refresh the old one */
+
+	      old->flags &= ~(REF_STALE | REF_DISCARD | REF_MODIFY);

 	      if (!rte_is_filtered(new))
 		{
@ -1283,6 +1314,28 @@ rte_discard(rte *old)	/* Non-filtered route deletion, used during garbage collec
  rte_update_unlock();
 }

+/* Modify existing route by protocol hook, used for long-lived graceful restart */
+static inline void
+rte_modify(rte *old)
+{
+  rte_update_lock();
+
+  rte *new = old->sender->proto->rte_modify(old, rte_update_pool);
+  if (new != old)
+  {
+    if (new)
+    {
+      if (!rta_is_cached(new->attrs))
+	new->attrs = rta_lookup(new->attrs);
+      new->flags = (old->flags & ~REF_MODIFY) | REF_COW;
+    }
+
+    rte_recalculate(old->sender, old->net, new, old->attrs->src);
+  }
+
+  rte_update_unlock();
+}
+
 /* Check rtable for best route to given net whether it would be exported do p */
 int
 rt_examine(rtable *t, ip_addr prefix, int pxlen, struct proto *p, struct filter *filter)
@ -1373,6 +1426,29 @@ rt_refresh_end(rtable *t, struct announce_hook *ah)
    rt_schedule_prune(t);
 }

+void
+rt_modify_stale(rtable *t, struct announce_hook *ah)
+{
+  int prune = 0;
+  net *n;
+  rte *e;
+
+  FIB_WALK(&t->fib, fn)
+    {
+      n = (net *) fn;
+      for (e = n->routes; e; e = e->next)
+	if ((e->sender == ah) && (e->flags & REF_STALE) && !(e->flags & REF_FILTERED))
+	  {
+	    e->flags |= REF_MODIFY;
+	    prune = 1;
+	  }
+    }
+  FIB_WALK_END;
+
+  if (prune)
+    rt_schedule_prune(t);
+}
+

 /**
 * rte_dump - dump a route
@ -1592,6 +1668,7 @@ again:

    rescan:
      for (e=n->routes; e; e=e->next)
+      {
 	if (e->sender->proto->flushing || (e->flags & REF_DISCARD))
 	  {
 	    if (*limit <= 0)
@ -1605,6 +1682,22 @@ again:

 	    goto rescan;
 	  }
+
+	if (e->flags & REF_MODIFY)
+	  {
+	    if (*limit <= 0)
+	      {
+		FIB_ITERATE_PUT(fit, fn);
+		return 0;
+	      }
+
+	    rte_modify(e);
+	    (*limit)--;
+
+	    goto rescan;
+	  }
+      }
+
      if (!n->routes)		/* Orphaned FIB entry */
 	{
 	  FIB_ITERATE_PUT(fit, fn);
--- a/proto/Doc
+++ b/proto/Doc
@ -2,6 +2,7 @@ H Protocols
 C babel
 C bfd
 C bgp
+C mrt
 C ospf
 C pipe
 C rip
--- a/proto/babel/babel.c
+++ b/proto/babel/babel.c
@ -1488,17 +1488,10 @@ babel_add_iface(struct babel_proto *p, struct iface *new, struct babel_iface_con
  ifa->cf = ic;
  ifa->pool = pool;
  ifa->ifname = new->name;
+  ifa->addr = new->llv6->ip;

  add_tail(&p->interfaces, NODE ifa);

-  struct ifa *addr;
-  WALK_LIST(addr, new->addrs)
-    if (ipa_is_link_local(addr->ip))
-      ifa->addr = addr->ip;
-
-  if (ipa_zero(ifa->addr))
-    log(L_WARN "%s: Cannot find link-local addr on %s", p->p.name, new->name);
-
  init_list(&ifa->neigh_list);
  ifa->hello_seqno = 1;

@ -1551,6 +1544,10 @@ babel_if_notify(struct proto *P, unsigned flags, struct iface *iface)
    if (!(iface->flags & IF_MULTICAST))
      return;

+    /* Ignore ifaces without link-local address */
+    if (!iface->llv6)
+      return;
+
    if (ic)
      babel_add_iface(p, iface, ic);

--- a/proto/babel/packets.c
+++ b/proto/babel/packets.c
@ -1087,6 +1087,7 @@ babel_open_socket(struct babel_iface *ifa)
  sk->sport = ifa->cf->port;
  sk->dport = ifa->cf->port;
  sk->iface = ifa->iface;
+  sk->saddr = ifa->addr;
  sk->vrf = p->p.vrf;

  sk->rx_hook = babel_rx_hook;
--- a/proto/bfd/bfd.c
+++ b/proto/bfd/bfd.c
@ -624,6 +624,9 @@ bfd_request_notify(struct bfd_request *req, u8 state, u8 diag)
 static int
 bfd_add_request(struct bfd_proto *p, struct bfd_request *req)
 {
+  if (p->p.vrf_set && (p->p.vrf != req->vrf))
+    return 0;
+
  struct bfd_session *s = bfd_find_session_by_addr(p, req->addr);
  u8 state, diag;

@ -685,7 +688,8 @@ bfd_drop_requests(struct bfd_proto *p)
 static struct resclass bfd_request_class;

 struct bfd_request *
-bfd_request_session(pool *p, ip_addr addr, ip_addr local, struct iface *iface,
+bfd_request_session(pool *p, ip_addr addr, ip_addr local,
+		    struct iface *iface, struct iface *vrf,
 		    void (*hook)(struct bfd_request *), void *data)
 {
  struct bfd_request *req = ralloc(p, &bfd_request_class);
@ -696,6 +700,7 @@ bfd_request_session(pool *p, ip_addr addr, ip_addr local, struct iface *iface,
  req->addr = addr;
  req->local = local;
  req->iface = iface;
+  req->vrf = vrf;

  bfd_submit_request(req);

@ -754,7 +759,7 @@ bfd_neigh_notify(struct neighbor *nb)
  if ((nb->scope > 0) && !n->req)
  {
    ip_addr local = ipa_nonzero(n->local) ? n->local : nb->ifa->ip;
-    n->req = bfd_request_session(p->p.pool, n->addr, local, nb->iface, NULL, NULL);
+    n->req = bfd_request_session(p->p.pool, n->addr, local, nb->iface, p->p.vrf, NULL, NULL);
  }

  if ((nb->scope <= 0) && n->req)
@ -771,7 +776,7 @@ bfd_start_neighbor(struct bfd_proto *p, struct bfd_neighbor *n)

  if (n->multihop)
  {
-    n->req = bfd_request_session(p->p.pool, n->addr, n->local, NULL, NULL, NULL);
+    n->req = bfd_request_session(p->p.pool, n->addr, n->local, NULL, p->p.vrf, NULL, NULL);
    return;
  }

@ -1052,15 +1057,6 @@ bfd_reconfigure(struct proto *P, struct proto_config *c)
  return 1;
 }

-/* Ensure one instance */
-struct bfd_config *bfd_cf;
-
-static void
-bfd_preconfig(struct protocol *P UNUSED, struct config *c UNUSED)
-{
-  bfd_cf = NULL;
-}
-
 static void
 bfd_copy_config(struct proto_config *dest, struct proto_config *src UNUSED)
 {
@ -1123,6 +1119,5 @@ struct protocol proto_bfd = {
  .start =		bfd_start,
  .shutdown =		bfd_shutdown,
  .reconfigure =	bfd_reconfigure,
-  .preconfig = 		bfd_preconfig,
  .copy_config =	bfd_copy_config,
 };
--- a/proto/bfd/config.Y
+++ b/proto/bfd/config.Y
@ -38,10 +38,6 @@ bfd_proto_start: proto_start BFD
  this_proto = proto_config_new(&proto_bfd, $1);
  init_list(&BFD_CFG->patt_list);
  init_list(&BFD_CFG->neigh_list);
-
-  if (bfd_cf)
-    cf_error("Only one BFD instance allowed");
-  bfd_cf = BFD_CFG;
 };

 bfd_proto_item:
--- a/proto/bfd/packets.c
+++ b/proto/bfd/packets.c
@ -141,6 +141,7 @@ bfd_fill_authentication(struct bfd_proto *p, struct bfd_session *s, struct bfd_c
  case BFD_AUTH_METICULOUS_KEYED_MD5:
  case BFD_AUTH_METICULOUS_KEYED_SHA1:
    meticulous = 1;
+    /* fallthrough */

  case BFD_AUTH_KEYED_MD5:
  case BFD_AUTH_KEYED_SHA1:
@ -230,6 +231,7 @@ bfd_check_authentication(struct bfd_proto *p, struct bfd_session *s, struct bfd_
  case BFD_AUTH_METICULOUS_KEYED_MD5:
  case BFD_AUTH_METICULOUS_KEYED_SHA1:
    meticulous = 1;
+    /* fallthrough */

  case BFD_AUTH_KEYED_MD5:
  case BFD_AUTH_KEYED_SHA1:
@ -410,6 +412,7 @@ bfd_open_rx_sk(struct bfd_proto *p, int multihop)
  sock *sk = sk_new(p->tpool);
  sk->type = SK_UDP;
  sk->sport = !multihop ? BFD_CONTROL_PORT : BFD_MULTI_CTL_PORT;
+  sk->vrf = p->p.vrf;
  sk->data = p;

  sk->rbsize = BFD_MAX_LEN;
@ -445,6 +448,7 @@ bfd_open_tx_sk(struct bfd_proto *p, ip_addr local, struct iface *ifa)
  sk->saddr = local;
  sk->dport = ifa ? BFD_CONTROL_PORT : BFD_MULTI_CTL_PORT;
  sk->iface = ifa;
+  sk->vrf = p->p.vrf;
  sk->data = p;

  sk->tbsize = BFD_MAX_LEN;
--- a/proto/bgp/attrs.c
+++ b/proto/bgp/attrs.c
@ -471,7 +471,7 @@ bgp_get_attr_len(eattr *a)

 /**
 * bgp_encode_attrs - encode BGP attributes
- * @p: BGP instance
+ * @p: BGP instance (or NULL)
 * @w: buffer
 * @attrs: a list of extended attributes
 * @remains: remaining space in the buffer
@ -485,6 +485,7 @@ uint
 bgp_encode_attrs(struct bgp_proto *p, byte *w, ea_list *attrs, int remains)
 {
  uint i, code, type, flags;
+  int as4_session = p ? p->as4_session : 1;
  byte *start = w;
  int len, rv;

@ -504,7 +505,7 @@ bgp_encode_attrs(struct bgp_proto *p, byte *w, ea_list *attrs, int remains)
       * we have to convert our 4B AS_PATH to 2B AS_PATH and send our AS_PATH 
       * as optional AS4_PATH attribute.
       */
-      if ((code == BA_AS_PATH) && (! p->as4_session))
+      if ((code == BA_AS_PATH) && !as4_session)
 	{
 	  len = a->u.ptr->length;

@ -546,7 +547,7 @@ bgp_encode_attrs(struct bgp_proto *p, byte *w, ea_list *attrs, int remains)
 	}

      /* The same issue with AGGREGATOR attribute */
-      if ((code == BA_AGGREGATOR) && (! p->as4_session))
+      if ((code == BA_AGGREGATOR) && !as4_session)
 	{
 	  int new_used;

@ -923,7 +924,7 @@ bgp_free_bucket(struct bgp_proto *p, struct bgp_bucket *buck)
 #define PXH_FN(p,l,i)		ipa_hash32(p) ^ u32_hash((l << 16) ^ i)

 #define PXH_REHASH		bgp_pxh_rehash
-#define PXH_PARAMS		/8, *2, 2, 2, 8, 20
+#define PXH_PARAMS		/8, *2, 2, 2, 8, 24


 HASH_DEFINE_REHASH_FN(PXH, struct bgp_prefix)
@ -1173,6 +1174,9 @@ bgp_community_filter(struct bgp_proto *p, rte *e)
 	  DBG("\tNO_EXPORT\n");
 	  return 1;
 	}
+
+      if (!p->conn->peer_llgr_aware && int_set_contains(d, BGP_COMM_LLGR_STALE))
+	return 1;
    }

  return 0;
@ -1233,6 +1237,19 @@ rte_resolvable(rte *rt)
  return (rd == RTD_ROUTER) || (rd == RTD_DEVICE) || (rd == RTD_MULTIPATH);
 }

+static inline int
+rte_stale(rte *r)
+{
+  if (r->u.bgp.stale < 0)
+  {
+    /* If staleness is unknown, compute and cache it */
+    eattr *a = ea_find(r->attrs->eattrs, EA_CODE(EAP_BGP, BA_COMMUNITY));
+    r->u.bgp.stale = a && int_set_contains(a->u.ptr, BGP_COMM_LLGR_STALE);
+  }
+
+  return r->u.bgp.stale;
+}
+
 int
 bgp_rte_better(rte *new, rte *old)
 {
@ -1257,6 +1274,14 @@ bgp_rte_better(rte *new, rte *old)
  if (n < o)
    return 0;

+  /* LLGR draft - depreference stale routes */
+  n = rte_stale(new);
+  o = rte_stale(old);
+  if (n > o)
+    return 0;
+  if (n < o)
+    return 1;
+
  /* Start with local preferences */
  x = ea_find(new->attrs->eattrs, EA_CODE(EAP_BGP, BA_LOCAL_PREF));
  y = ea_find(old->attrs->eattrs, EA_CODE(EAP_BGP, BA_LOCAL_PREF));
@ -1378,6 +1403,10 @@ bgp_rte_mergable(rte *pri, rte *sec)
  if (!rte_resolvable(sec))
    return 0;

+  /* LLGR draft - depreference stale routes */
+  if (rte_stale(pri) != rte_stale(sec))
+    return 0;
+
  /* Start with local preferences */
  x = ea_find(pri->attrs->eattrs, EA_CODE(EAP_BGP, BA_LOCAL_PREF));
  y = ea_find(sec->attrs->eattrs, EA_CODE(EAP_BGP, BA_LOCAL_PREF));
@ -1580,6 +1609,27 @@ bgp_rte_recalculate(rtable *table, net *net, rte *new, rte *old, rte *old_best)
    return old_is_group_best;
 }

+struct rte *
+bgp_rte_modify_stale(struct rte *r, struct linpool *pool)
+{
+  eattr *a = ea_find(r->attrs->eattrs, EA_CODE(EAP_BGP, BA_COMMUNITY));
+  struct adata *ad = a ? a->u.ptr : NULL;
+
+  if (ad && int_set_contains(ad, BGP_COMM_NO_LLGR))
+    return NULL;
+
+  if (ad && int_set_contains(ad, BGP_COMM_LLGR_STALE))
+    return r;
+
+  r = rte_cow_rta(r, pool);
+  bgp_attach_attr(&(r->attrs->eattrs), pool, BA_COMMUNITY,
+		  (uintptr_t) int_set_add(pool, ad, BGP_COMM_LLGR_STALE));
+  r->u.bgp.stale = 1;
+
+  return r;
+}
+
+
 static struct adata *
 bgp_aggregator_convert_to_new(struct adata *old, struct linpool *pool)
 {
@ -1589,7 +1639,6 @@ bgp_aggregator_convert_to_new(struct adata *old, struct linpool *pool)
  return newa;
 }

-
 /* Take last req_as ASNs from path old2 (in 2B format), convert to 4B format
 * and append path old4 (in 4B format).
 */
@ -1985,6 +2034,9 @@ bgp_get_route_info(rte *e, byte *buf, ea_list *attrs)
  if (e->u.bgp.suppressed)
    buf += bsprintf(buf, "-");

+  if (rte_stale(e))
+    buf += bsprintf(buf, "s");
+
  if (e->attrs->hostentry)
    {
      if (!rte_resolvable(e))
--- a/proto/bgp/bgp.c
+++ b/proto/bgp/bgp.c
@ -352,7 +352,7 @@ static inline void
 bgp_conn_set_state(struct bgp_conn *conn, unsigned new_state)
 {
  if (conn->bgp->p.mrtdump & MD_STATES)
-    mrt_dump_bgp_state_change(conn, conn->state, new_state);
+    bgp_dump_state_change(conn, conn->state, new_state);

  conn->state = new_state;
 }
@ -394,10 +394,17 @@ bgp_conn_enter_established_state(struct bgp_conn *conn)
  if (p->p.gr_recovery && (p->cf->gr_mode == BGP_GR_ABLE) && peer_gr_ready)
    p->p.gr_wait = 1;

-  if (p->gr_active)
+  if (p->gr_active == BGP_GRS_ACTIVE)
    tm_stop(p->gr_timer);

-  if (p->gr_active && (!conn->peer_gr_able || !(conn->peer_gr_aflags & BGP_GRF_FORWARDING)))
+  /* Check F-bit for regular graceful restart */
+  if ((p->gr_active == BGP_GRS_ACTIVE) &&
+      (!conn->peer_gr_able || !(conn->peer_gr_aflags & BGP_GRF_FORWARDING)))
+    bgp_graceful_restart_done(p);
+
+  /* Check F-bit for long-lived graceful restart */
+  if (((p->gr_active == BGP_GRS_LLGR_1) || (p->gr_active == BGP_GRS_LLGR_2)) &&
+      (!conn->peer_llgr_able || !(conn->peer_llgr_aflags & BGP_LLGRF_FORWARDING)))
    bgp_graceful_restart_done(p);

  /* GR capability implies that neighbor will send End-of-RIB */
@ -474,11 +481,25 @@ bgp_handle_graceful_restart(struct bgp_proto *p)
 	    p->gr_active ? " - already pending" : "");
  proto_notify_state(&p->p, PS_START);

-  if (p->gr_active)
+  switch (p->gr_active)
+  {
+  case BGP_GRS_ACTIVE:
    rt_refresh_end(p->p.main_ahook->table, p->p.main_ahook);
+    break;

-  p->gr_active = 1;
-  bgp_start_timer(p->gr_timer, p->conn->peer_gr_time);
+  case BGP_GRS_LLGR_1:
+    rt_refresh_begin(p->p.main_ahook->table, p->p.main_ahook);
+    return;
+
+  case BGP_GRS_LLGR_2:
+    rt_refresh_begin(p->p.main_ahook->table, p->p.main_ahook);
+    rt_modify_stale(p->p.main_ahook->table, p->p.main_ahook);
+    return;
+  }
+
+  p->stale_time = p->cf->llgr_mode ? p->conn->peer_llgr_time : 0;
+  p->gr_active = !p->stale_time ? BGP_GRS_ACTIVE : BGP_GRS_LLGR_1;
+  tm_start(p->gr_timer, p->conn->peer_gr_time);
  rt_refresh_begin(p->p.main_ahook->table, p->p.main_ahook);
 }

@ -515,10 +536,27 @@ bgp_graceful_restart_timeout(timer *t)
 {
  struct bgp_proto *p = t->data;

-  BGP_TRACE(D_EVENTS, "Neighbor graceful restart timeout");
-  bgp_stop(p, 0, NULL, 0);
-}
+  switch (p->gr_active)
+  {
+  case BGP_GRS_ACTIVE:
+    BGP_TRACE(D_EVENTS, "Neighbor graceful restart timeout");
+    bgp_stop(p, 0, NULL, 0);
+    return;

+  case BGP_GRS_LLGR_1:
+    BGP_TRACE(D_EVENTS, "Neighbor graceful restart timeout");
+    p->gr_active = BGP_GRS_LLGR_2;
+    tm_start(p->gr_timer, p->stale_time);
+    rt_modify_stale(p->p.main_ahook->table, p->p.main_ahook);
+    return;
+
+  case BGP_GRS_LLGR_2:
+    BGP_TRACE(D_EVENTS, "Long-lived graceful restart timeout");
+    p->gr_active = 0;
+    rt_refresh_end(p->p.main_ahook->table, p->p.main_ahook);
+    return;
+  }
+}

 /**
 * bgp_refresh_begin - start incoming enhanced route refresh sequence
@ -576,6 +614,10 @@ bgp_send_open(struct bgp_conn *conn)
  conn->peer_gr_time = 0;
  conn->peer_gr_flags = 0;
  conn->peer_gr_aflags = 0;
+  conn->peer_llgr_aware = 0;
+  conn->peer_llgr_able = 0;
+  conn->peer_llgr_time = 0;
+  conn->peer_llgr_aflags = 0;
  conn->peer_ext_messages_support = 0;

  DBG("BGP: Sending open\n");
@ -662,6 +704,12 @@ bgp_hold_timeout(timer *t)

  if (sk_rx_ready(conn->sk) > 0)
    bgp_start_timer(conn->hold_timer, 10);
+  else if ((conn->state == BS_ESTABLISHED) && p->gr_ready && conn->peer_llgr_able)
+  {
+    BGP_TRACE(D_EVENTS, "Hold timer expired");
+    bgp_handle_graceful_restart(p);
+    bgp_conn_enter_idle_state(conn);
+  }
  else
    bgp_error(conn, 4, 0, NULL, 0);
 }
@ -1004,13 +1052,30 @@ bgp_bfd_notify(struct bfd_request *req)
  int ps = p->p.proto_state;

  if (req->down && ((ps == PS_START) || (ps == PS_UP)))
+  {
+    BGP_TRACE(D_EVENTS, "BFD session down");
+    bgp_store_error(p, NULL, BE_MISC, BEM_BFD_DOWN);
+
+    if (p->cf->bfd == BGP_BFD_GRACEFUL)
    {
-      BGP_TRACE(D_EVENTS, "BFD session down");
-      bgp_store_error(p, NULL, BE_MISC, BEM_BFD_DOWN);
+      /* Trigger graceful restart */
+      if (p->conn && (p->conn->state == BS_ESTABLISHED) && p->gr_ready)
+	bgp_handle_graceful_restart(p);
+
+      if (p->incoming_conn.state > BS_IDLE)
+	bgp_conn_enter_idle_state(&p->incoming_conn);
+
+      if (p->outgoing_conn.state > BS_IDLE)
+	bgp_conn_enter_idle_state(&p->outgoing_conn);
+    }
+    else
+    {
+      /* Trigger session down */
      if (ps == PS_UP)
 	bgp_update_startup_delay(p);
      bgp_stop(p, 0, NULL, 0);
    }
+  }
 }

 static void
@ -1019,7 +1084,7 @@ bgp_update_bfd(struct bgp_proto *p, int use_bfd)
  if (use_bfd && !p->bfd_req)
    p->bfd_req = bfd_request_session(p->p.pool, p->cf->remote_ip, p->source_addr,
 				     p->cf->multihop ? NULL : p->neigh->iface,
-				     bgp_bfd_notify, p);
+				     p->p.vrf, bgp_bfd_notify, p);

  if (!use_bfd && p->bfd_req)
    {
@ -1252,7 +1317,7 @@ bgp_shutdown(struct proto *P)
  if (message)
  {
    uint msg_len = strlen(message);
-    msg_len = MIN(msg_len, 128);
+    msg_len = MIN(msg_len, 255);

    /* Buffer will be freed automatically by protocol shutdown */
    data = mb_alloc(p->p.pool, msg_len + 1);
@ -1297,6 +1362,7 @@ bgp_init(struct proto_config *C)
  P->rte_better = bgp_rte_better;
  P->rte_mergable = bgp_rte_mergable;
  P->rte_recalculate = c->deterministic_med ? bgp_rte_recalculate : NULL;
+  P->rte_modify = bgp_rte_modify_stale;

  p->cf = c;
  p->local_as = c->local_as;
@ -1332,6 +1398,10 @@ bgp_check_config(struct bgp_config *c)
  if (!c->missing_lladdr)
    c->missing_lladdr = c->rs_client ? MLL_IGNORE : MLL_SELF;

+  /* LLGR mode default based on GR mode */
+  if (c->llgr_mode < 0)
+    c->llgr_mode = c->gr_mode ? BGP_LLGR_AWARE : 0;
+
  /* 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;
@ -1382,6 +1452,9 @@ bgp_check_config(struct bgp_config *c)

  if (c->secondary && !c->c.table->sorted)
    cf_error("BGP with secondary option requires sorted table");
+
+  if (!c->gr_mode && c->llgr_mode)
+    cf_error("Long-lived graceful restart requires basic graceful restart");
 }

 static int
@ -1550,6 +1623,11 @@ bgp_show_proto_info(struct proto *P)
  if (p->gr_active)
    cli_msg(-1006, "    Neighbor graceful restart active");

+  if (p->gr_active && p->gr_timer->expires)
+    cli_msg(-1006, "    %-15s   %d/-",
+	    (p->gr_active != BGP_GRS_LLGR_2) ? "Restart timer:" : "LL stale timer:",
+	    p->gr_timer->expires - now);
+
  if (P->proto_state == PS_START)
    {
      struct bgp_conn *oc = &p->outgoing_conn;
@ -1563,9 +1641,6 @@ bgp_show_proto_info(struct proto *P)
 	  (oc->connect_retry_timer->expires))
 	cli_msg(-1006, "    Connect delay:    %d/%d",
 		oc->connect_retry_timer->expires - now, p->cf->connect_delay_time);
-
-      if (p->gr_active && p->gr_timer->expires)
-	cli_msg(-1006, "    Restart timer:    %d/-", p->gr_timer->expires - now);
    }
  else if (P->proto_state == PS_UP)
    {
@ -1574,6 +1649,7 @@ bgp_show_proto_info(struct proto *P)
 	      c->peer_refresh_support ? " refresh" : "",
 	      c->peer_enhanced_refresh_support ? " enhanced-refresh" : "",
 	      c->peer_gr_able ? " restart-able" : (c->peer_gr_aware ? " restart-aware" : ""),
+	      c->peer_llgr_able ? " llgr-able" : (c->peer_llgr_aware ? " llgr-aware" : ""),
 	      c->peer_as4_support ? " AS4" : "",
 	      (c->peer_add_path & ADD_PATH_RX) ? " add-path-rx" : "",
 	      (c->peer_add_path & ADD_PATH_TX) ? " add-path-tx" : "",
--- a/proto/bgp/bgp.h
+++ b/proto/bgp/bgp.h
@ -52,8 +52,10 @@ struct bgp_config {
  int allow_local_as;			/* Allow that number of local ASNs in incoming AS_PATHs */
  int allow_local_pref;			/* Allow LOCAL_PREF in EBGP sessions */
  int gr_mode;				/* Graceful restart mode (BGP_GR_*) */
+  int llgr_mode;			/* Long-lived graceful restart mode (BGP_LLGR_*) */
  int setkey;				/* Set MD5 password to system SA/SP database */
  unsigned gr_time;			/* Graceful restart timeout */
+  unsigned llgr_time;			/* Long-lived graceful restart timeout */
  unsigned connect_delay_time;		/* Minimum delay between connect attempts */
  unsigned connect_retry_time;		/* Timeout for connect attempts */
  unsigned hold_time, initial_hold_time;
@ -90,6 +92,18 @@ struct bgp_config {
 /* For peer_gr_aflags */
 #define BGP_GRF_FORWARDING 0x80

+#define BGP_LLGR_ABLE 1
+#define BGP_LLGR_AWARE 2
+
+#define BGP_LLGRF_FORWARDING 0x80
+
+#define BGP_GRS_NONE		0	/* No GR  */
+#define BGP_GRS_ACTIVE		1	/* Graceful restart per RFC 4724 */
+#define BGP_GRS_LLGR_1		2	/* Long-lived GR phase 1 (restart time) */
+#define BGP_GRS_LLGR_2		3	/* Long-lived GR phase 2 (stale time) */
+
+#define BGP_BFD_GRACEFUL	2	/* BFD down triggers graceful restart */
+

 struct bgp_conn {
  struct bgp_proto *bgp;
@ -113,6 +127,10 @@ struct bgp_conn {
  u16 peer_gr_time;
  u8 peer_gr_flags;
  u8 peer_gr_aflags;
+  u8 peer_llgr_aware;
+  u8 peer_llgr_able;
+  uint peer_llgr_time;
+  u8 peer_llgr_aflags;
  u8 peer_ext_messages_support;		/* Peer supports extended message length [draft] */
  unsigned hold_time, keepalive_time;	/* Times calculated from my and neighbor's requirements */
 };
@ -133,9 +151,10 @@ struct bgp_proto {
  int rr_client;			/* Whether neighbor is RR client of me */
  int rs_client;			/* Whether neighbor is RS client of me */
  u8 gr_ready;				/* Neighbor could do graceful restart */
-  u8 gr_active;				/* Neighbor is doing graceful restart */
+  u8 gr_active;				/* Neighbor is doing graceful restart (BGP_GRS_*) */
  u8 feed_state;			/* Feed state (TX) for EoR, RR packets, see BFS_* */
  u8 load_state;			/* Load state (RX) for EoR, RR packets, see BFS_* */
+  uint stale_time;			/* Long-lived stale time for LLGR */
  struct bgp_conn *conn;		/* Connection we have established */
  struct bgp_conn outgoing_conn;	/* Outgoing connection we're working with */
  struct bgp_conn incoming_conn;	/* Incoming connection we have neither accepted nor rejected yet */
@ -252,6 +271,7 @@ int bgp_get_attr(struct eattr *e, byte *buf, int buflen);
 int bgp_rte_better(struct rte *, struct rte *);
 int bgp_rte_mergable(rte *pri, rte *sec);
 int bgp_rte_recalculate(rtable *table, net *net, rte *new, rte *old, rte *old_best);
+struct rte *bgp_rte_modify_stale(struct rte *r, struct linpool *pool);
 void bgp_rt_notify(struct proto *P, rtable *tbl UNUSED, net *n, rte *new, rte *old UNUSED, ea_list *attrs);
 int bgp_import_control(struct proto *, struct rte **, struct ea_list **, struct linpool *);
 void bgp_init_bucket_table(struct bgp_proto *);
@ -268,7 +288,7 @@ inline static void bgp_attach_attr_ip(struct ea_list **to, struct linpool *pool,

 /* packets.c */

-void mrt_dump_bgp_state_change(struct bgp_conn *conn, unsigned old, unsigned new);
+void bgp_dump_state_change(struct bgp_conn *conn, uint old, uint new);
 void bgp_schedule_packet(struct bgp_conn *conn, int type);
 void bgp_kick_tx(void *vconn);
 void bgp_tx(struct birdsock *sk);
@ -398,6 +418,9 @@ void bgp_log_error(struct bgp_proto *p, u8 class, char *msg, unsigned code, unsi
 #define BGP_COMM_NO_ADVERTISE		0xffffff02	/* Don't export at all */
 #define BGP_COMM_NO_EXPORT_SUBCONFED	0xffffff03	/* NO_EXPORT even in local confederation */

+#define BGP_COMM_LLGR_STALE		0xffff0006	/* Route is stale according to LLGR */
+#define BGP_COMM_NO_LLGR		0xffff0007	/* Do not treat the route according to LLGR */
+
 /* Origins */

 #define ORIGIN_IGP		0
--- a/proto/bgp/config.Y
+++ b/proto/bgp/config.Y
@ -27,7 +27,8 @@ CF_KEYWORDS(BGP, LOCAL, NEIGHBOR, AS, HOLD, TIME, CONNECT, RETRY,
 	INTERPRET, COMMUNITIES, BGP_ORIGINATOR_ID, BGP_CLUSTER_LIST, IGP,
 	TABLE, GATEWAY, DIRECT, RECURSIVE, MED, TTL, SECURITY, DETERMINISTIC,
 	SECONDARY, ALLOW, BFD, ADD, PATHS, RX, TX, GRACEFUL, RESTART, AWARE,
-	CHECK, LINK, PORT, EXTENDED, MESSAGES, SETKEY, BGP_LARGE_COMMUNITY)
+	CHECK, LINK, PORT, EXTENDED, MESSAGES, SETKEY, BGP_LARGE_COMMUNITY,
+	LONG, LIVED, STALE)

 CF_KEYWORDS(CEASE, PREFIX, LIMIT, HIT, ADMINISTRATIVE, SHUTDOWN, RESET, PEER,
 	CONFIGURATION, CHANGE, DECONFIGURED, CONNECTION, REJECTED, COLLISION,
@ -60,6 +61,8 @@ bgp_proto_start: proto_start BGP {
     BGP_CFG->default_local_pref = 100;
     BGP_CFG->gr_mode = BGP_GR_AWARE;
     BGP_CFG->gr_time = 120;
+     BGP_CFG->llgr_mode = -1;
+     BGP_CFG->llgr_time = 3600;
     BGP_CFG->setkey = 1;
 }
 ;
@ -162,10 +165,14 @@ bgp_proto:
 | bgp_proto GRACEFUL RESTART bool ';' { BGP_CFG->gr_mode = $4; }
 | bgp_proto GRACEFUL RESTART AWARE ';' { BGP_CFG->gr_mode = BGP_GR_AWARE; }
 | bgp_proto GRACEFUL RESTART TIME expr ';' { BGP_CFG->gr_time = $5; }
+ | bgp_proto LONG LIVED GRACEFUL RESTART bool ';' { BGP_CFG->llgr_mode = $6; }
+ | bgp_proto LONG LIVED GRACEFUL RESTART AWARE ';' { BGP_CFG->llgr_mode = BGP_LLGR_AWARE; }
+ | bgp_proto LONG LIVED STALE TIME expr ';' { BGP_CFG->llgr_time = $6; }
 | bgp_proto IGP TABLE rtable ';' { BGP_CFG->igp_table = $4; }
 | bgp_proto TTL SECURITY bool ';' { BGP_CFG->ttl_security = $4; }
 | bgp_proto CHECK LINK bool ';' { BGP_CFG->check_link = $4; }
 | bgp_proto BFD bool ';' { BGP_CFG->bfd = $3; cf_check_bfd($3); }
+ | bgp_proto BFD GRACEFUL ';' { BGP_CFG->bfd = BGP_BFD_GRACEFUL; cf_check_bfd(1); }
 ;

 CF_ADDTO(dynamic_attr, BGP_ORIGIN
--- a/proto/bgp/packets.c
+++ b/proto/bgp/packets.c
@ -13,7 +13,7 @@
 #include "nest/protocol.h"
 #include "nest/route.h"
 #include "nest/attrs.h"
-#include "nest/mrtdump.h"
+#include "proto/mrt/mrt.h"
 #include "conf/conf.h"
 #include "lib/unaligned.h"
 #include "lib/socket.h"
@ -38,81 +38,45 @@ static byte fsm_err_subcode[BS_MAX] = {
  [BS_ESTABLISHED] = 3
 };

-/*
- * MRT Dump format is not semantically specified.
- * We will use these values in appropriate fields:
- *
- * Local AS, Remote AS - configured AS numbers for given BGP instance.
- * Local IP, Remote IP - IP addresses of the TCP connection (0 if no connection)
- *
- * We dump two kinds of MRT messages: STATE_CHANGE (for BGP state
- * changes) and MESSAGE (for received BGP messages).
- *
- * STATE_CHANGE uses always AS4 variant, but MESSAGE uses AS4 variant
- * only when AS4 session is established and even in that case MESSAGE
- * does not use AS4 variant for initial OPEN message. This strange
- * behavior is here for compatibility with Quagga and Bgpdump,
- */
-
-static byte *
-mrt_put_bgp4_hdr(byte *buf, struct bgp_conn *conn, int as4)
+static void
+init_mrt_bgp_data(struct bgp_conn *conn, struct mrt_bgp_data *d)
 {
  struct bgp_proto *p = conn->bgp;
+  int p_ok = conn->state >= BS_OPENCONFIRM;

-  if (as4)
-    {
-      put_u32(buf+0, p->remote_as);
-      put_u32(buf+4, p->local_as);
-      buf+=8;
-    }
-  else
-    {
-      put_u16(buf+0, (p->remote_as <= 0xFFFF) ? p->remote_as : AS_TRANS);
-      put_u16(buf+2, (p->local_as <= 0xFFFF)  ? p->local_as  : AS_TRANS);
-      buf+=4;
-    }
-
-  put_u16(buf+0, (p->neigh && p->neigh->iface) ? p->neigh->iface->index : 0);
-  put_u16(buf+2, BGP_AF);
-  buf+=4;
-  buf = put_ipa(buf, conn->sk ? conn->sk->daddr : IPA_NONE);
-  buf = put_ipa(buf, conn->sk ? conn->sk->saddr : IPA_NONE);
-
-  return buf;
+  memset(d, 0, sizeof(struct mrt_bgp_data));
+  d->peer_as = p->remote_as;
+  d->local_as = p->local_as;
+  d->index = (p->neigh && p->neigh->iface) ? p->neigh->iface->index : 0;
+  d->af = BGP_AF;
+  d->peer_ip = conn->sk ? conn->sk->daddr : IPA_NONE;
+  d->local_ip = conn->sk ? conn->sk->saddr : IPA_NONE;
+  d->as4 = p_ok ? p->as4_session : 0;
+  d->add_path = p_ok ? p->add_path_rx : 0;
 }

 static void
-mrt_dump_bgp_packet(struct bgp_conn *conn, byte *pkt, unsigned len)
+bgp_dump_message(struct bgp_conn *conn, byte *pkt, uint len)
 {
-  byte *buf = alloca(128+len);	/* 128 is enough for MRT headers */
-  byte *bp = buf + MRTDUMP_HDR_LENGTH;
-  int as4 = conn->bgp->as4_session;
+  struct mrt_bgp_data d;
+  init_mrt_bgp_data(conn, &d);

-  bp = mrt_put_bgp4_hdr(bp, conn, as4);
-  memcpy(bp, pkt, len);
-  bp += len;
-  mrt_dump_message(&conn->bgp->p, BGP4MP, as4 ? BGP4MP_MESSAGE_AS4 : BGP4MP_MESSAGE,
-		   buf, bp-buf);
-}
+  d.message = pkt;
+  d.msg_len = len;

-static inline u16
-convert_state(unsigned state)
-{
-  /* Convert state from our BS_* values to values used in MRTDump */
-  return (state == BS_CLOSE) ? 1 : state + 1;
+  mrt_dump_bgp_message(&d);
 }

 void
-mrt_dump_bgp_state_change(struct bgp_conn *conn, unsigned old, unsigned new)
+bgp_dump_state_change(struct bgp_conn *conn, uint old, uint new)
 {
-  byte buf[128];
-  byte *bp = buf + MRTDUMP_HDR_LENGTH;
+  struct mrt_bgp_data d;
+  init_mrt_bgp_data(conn, &d);

-  bp = mrt_put_bgp4_hdr(bp, conn, 1);
-  put_u16(bp+0, convert_state(old));
-  put_u16(bp+2, convert_state(new));
-  bp += 4;
-  mrt_dump_message(&conn->bgp->p, BGP4MP, BGP4MP_STATE_CHANGE_AS4, buf, bp-buf);
+  d.old_state = old;
+  d.new_state = new;
+
+  mrt_dump_bgp_state_change(&d);
 }

 static byte *
@ -231,6 +195,32 @@ bgp_put_cap_err(struct bgp_proto *p UNUSED, byte *buf)
  return buf;
 }

+static byte *
+bgp_put_cap_llgr1(struct bgp_proto *p, byte *buf)
+{
+  *buf++ = 71;		/* Capability 71: Support for long-lived graceful restart */
+  *buf++ = 7;		/* Capability data length */
+
+  *buf++ = 0;		/* Appropriate AF */
+  *buf++ = BGP_AF;
+  *buf++ = 1;		/* and SAFI 1 */
+
+  /* Next is 8bit flags and 24bit time */
+  put_u32(buf, p->cf->llgr_time);
+  buf[0] = p->p.gr_recovery ? BGP_LLGRF_FORWARDING : 0;
+  buf += 4;
+
+  return buf;
+}
+
+static byte *
+bgp_put_cap_llgr2(struct bgp_proto *p UNUSED, byte *buf)
+{
+  *buf++ = 71;		/* Capability 71: Support for long-lived graceful restart */
+  *buf++ = 0;		/* Capability data length */
+  return buf;
+}
+

 static byte *
 bgp_create_open(struct bgp_conn *conn, byte *buf)
@ -285,6 +275,11 @@ bgp_create_open(struct bgp_conn *conn, byte *buf)
  if (p->cf->enable_extended_messages)
    cap = bgp_put_cap_ext_msg(p, cap);

+  if (p->cf->llgr_mode == BGP_LLGR_ABLE)
+    cap = bgp_put_cap_llgr1(p, cap);
+  else if (p->cf->llgr_mode == BGP_LLGR_AWARE)
+    cap = bgp_put_cap_llgr2(p, cap);
+
  cap_len = cap - buf - 12;
  if (cap_len > 0)
    {
@ -789,8 +784,12 @@ bgp_kick_tx(void *vconn)
  struct bgp_conn *conn = vconn;

  DBG("BGP: kicking TX\n");
-  while (bgp_fire_tx(conn) > 0)
+  uint max = 1024;
+  while (--max && (bgp_fire_tx(conn) > 0))
    ;
+
+  if (!max && !ev_active(conn->tx_ev))
+    ev_schedule(conn->tx_ev);
 }

 void
@ -799,8 +798,12 @@ bgp_tx(sock *sk)
  struct bgp_conn *conn = sk->data;

  DBG("BGP: TX hook\n");
-  while (bgp_fire_tx(conn) > 0)
+  uint max = 1024;
+  while (--max && (bgp_fire_tx(conn) > 0))
    ;
+
+  if (!max && !ev_active(conn->tx_ev))
+    ev_schedule(conn->tx_ev);
 }

 /* Capatibility negotiation as per RFC 2842 */
@ -872,11 +875,38 @@ bgp_parse_capabilities(struct bgp_conn *conn, byte *opt, int len)
 	  conn->peer_enhanced_refresh_support = 1;
 	  break;

+	case 71: /* Long-lived graceful restart capability, RFC draft */
+	  if (cl % 7)
+	    goto err;
+	  conn->peer_llgr_aware = 1;
+	  conn->peer_llgr_able = 0;
+	  conn->peer_llgr_time = 0;
+	  conn->peer_llgr_aflags = 0;
+	  for (i = 0; i < cl; i += 7)
+	    if (opt[2+i+0] == 0 && opt[2+i+1] == BGP_AF && opt[2+i+2] == 1) /* Match AFI/SAFI */
+	      {
+		conn->peer_llgr_able = 1;
+		conn->peer_llgr_time = get_u32(opt + 2+i+3) & 0xffffff;
+		conn->peer_llgr_aflags = opt[2+i+3];
+	      }
+	  break;
+
 	  /* We can safely ignore all other capabilities */
 	}
      len -= 2 + cl;
      opt += 2 + cl;
    }
+
+  /* The LLGR capability must be advertised together with the GR capability,
+     otherwise it must be disregarded */
+  if (!conn->peer_gr_aware && conn->peer_llgr_aware)
+  {
+    conn->peer_llgr_aware = 0;
+    conn->peer_llgr_able = 0;
+    conn->peer_llgr_time = 0;
+    conn->peer_llgr_aflags = 0;
+  }
+
  return;

 err:
@ -1034,7 +1064,8 @@ bgp_rx_open(struct bgp_conn *conn, byte *pkt, uint len)
  p->as4_session = p->cf->enable_as4 && conn->peer_as4_support;
  p->add_path_rx = (p->cf->add_path & ADD_PATH_RX) && (conn->peer_add_path & ADD_PATH_TX);
  p->add_path_tx = (p->cf->add_path & ADD_PATH_TX) && (conn->peer_add_path & ADD_PATH_RX);
-  p->gr_ready = p->cf->gr_mode && conn->peer_gr_able;
+  p->gr_ready = (p->cf->gr_mode && conn->peer_gr_able) ||
+		(p->cf->llgr_mode && conn->peer_llgr_able);
  p->ext_messages = p->cf->enable_extended_messages && conn->peer_ext_messages_support;

  /* Update RA mode */
@ -1125,6 +1156,7 @@ bgp_rte_update(struct bgp_proto *p, ip_addr prefix, int pxlen,
  e->net = n;
  e->pflags = 0;
  e->u.bgp.suppressed = 0;
+  e->u.bgp.stale = -1;
  rte_update2(p->p.main_ahook, n, e, *src);
 }

@ -1507,7 +1539,7 @@ bgp_handle_message(struct bgp_proto *p, byte *data, uint len, byte **bp)
    return 1;

  /* Handle proper message */
-  if ((msg_len > 128) && (msg_len + 1 > len))
+  if ((msg_len > 255) && (msg_len + 1 > len))
    return 0;

  /* Some elementary cleanup */
@ -1704,7 +1736,7 @@ bgp_rx_packet(struct bgp_conn *conn, byte *pkt, unsigned len)
  DBG("BGP: Got packet %02x (%d bytes)\n", type, len);

  if (conn->bgp->p.mrtdump & MD_MESSAGES)
-    mrt_dump_bgp_packet(conn, pkt, len);
+    bgp_dump_message(conn, pkt, len);

  switch (type)
    {
--- a/proto/mrt/Doc
+++ b/proto/mrt/Doc
@ -0,0 +1 @@
+S mrt.c
--- a/proto/mrt/Makefile
+++ b/proto/mrt/Makefile
@ -0,0 +1,5 @@
+source=mrt.c
+root-rel=../../
+dir-name=proto/mrt
+
+include ../../Rules
--- a/proto/mrt/config.Y
+++ b/proto/mrt/config.Y
@ -0,0 +1,67 @@
+/*
+ *	BIRD -- Multi-Threaded Routing Toolkit (MRT) Protocol
+ *
+ *	(c) 2017--2018 Ondrej Zajicek <santiago@crfreenet.org>
+ *	(c) 2017--2018 CZ.NIC z.s.p.o.
+ *
+ *	Can be freely distributed and used under the terms of the GNU GPL.
+ */
+
+CF_HDR
+
+#include "proto/mrt/mrt.h"
+
+CF_DEFINES
+
+#define MRT_CFG ((struct mrt_config *) this_proto)
+
+CF_DECLS
+
+CF_KEYWORDS(MRT, TABLE, FILTER, FILENAME, PERIOD, ALWAYS, ADD, PATH, DUMP, TO)
+
+%type <md> mrt_dump_args
+
+CF_GRAMMAR
+
+CF_ADDTO(proto, mrt_proto)
+
+mrt_proto_start: proto_start MRT
+{
+  this_proto = proto_config_new(&proto_mrt, $1);
+};
+
+mrt_proto_item:
+   proto_item
+ | TABLE TEXT		{ MRT_CFG->table_expr = $2; }
+ | FILTER filter	{ MRT_CFG->filter = $2; }
+ | where_filter		{ MRT_CFG->filter = $1; }
+ | FILENAME text	{ MRT_CFG->filename = $2; }
+ | PERIOD expr		{ MRT_CFG->period = $2; }
+ | ALWAYS ADD PATH bool	{ MRT_CFG->always_add_path = $4; }
+ ;
+
+mrt_proto_opts:
+   /* empty */
+ | mrt_proto_opts mrt_proto_item ';'
+ ;
+
+mrt_proto:
+   mrt_proto_start proto_name '{' mrt_proto_opts '}' { mrt_check_config(this_proto); };
+
+CF_CLI_HELP(MRT DUMP, [table <name>|\"<pattern>\"] [to \"<file>\"] [filter <filter>|where <where filter>] , [[Save MRT Table Dump into a file]])
+CF_CLI(MRT DUMP, mrt_dump_args, [table <name>|\"<pattern>\"] [to \"<file>\"] [filter <filter>|where <where filter>], [[Save mrt table dump v2 of table name <t> right now]])
+{ mrt_dump_cmd($3); } ;
+
+mrt_dump_args:
+   /* empty */ { $$ = cfg_allocz(sizeof(struct mrt_dump_data)); }
+ | mrt_dump_args TABLE rtable	{ $$ = $1; $$->table_ptr = $3->table; }
+ | mrt_dump_args TABLE TEXT	{ $$ = $1; $$->table_expr = $3; }
+ | mrt_dump_args FILTER filter	{ $$ = $1; $$->filter = $3; }
+ | mrt_dump_args where_filter	{ $$ = $1; $$->filter = $2; }
+ | mrt_dump_args TO text 	{ $$ = $1; $$->filename = $3; }
+ ;
+
+
+CF_CODE
+
+CF_END
--- a/proto/mrt/mrt.c
+++ b/proto/mrt/mrt.c
@ -0,0 +1,883 @@
+/*
+ *	BIRD -- Multi-Threaded Routing Toolkit (MRT) Protocol
+ *
+ *	(c) 2017--2018 Ondrej Zajicek <santiago@crfreenet.org>
+ *	(c) 2017--2018 CZ.NIC z.s.p.o.
+ *
+ *	Can be freely distributed and used under the terms of the GNU GPL.
+ */
+
+/**
+ * DOC: Multi-Threaded Routing Toolkit (MRT) protocol
+ *
+ * The MRT protocol is implemented in just one file: |mrt.c|. It contains of
+ * several parts: Generic functions for preparing MRT messages in a buffer,
+ * functions for MRT table dump (called from timer or CLI), functions for MRT
+ * BGP4MP dump (called from BGP), and the usual protocol glue. For the MRT table
+ * dump, the key structure is struct mrt_table_dump_state, which contains all
+ * necessary data and created when the MRT dump cycle is started for the
+ * duration of the MRT dump. The MBGP4MP dump is currently not bound to MRT
+ * protocol instance and uses the config->mrtdump_file fd.
+ *
+ * The protocol is simple, just periodically scans routing table and export it
+ * to a file. It does not use the regular update mechanism, but a direct access
+ * in order to handle iteration through multiple routing tables. The table dump
+ * needs to dump all peers first and then use indexes to address the peers, we
+ * use a hash table (@peer_hash) to find peer index based on BGP protocol key
+ * attributes.
+ *
+ * One thing worth documenting is the locking. During processing, the currently
+ * processed table (@table field in the state structure) is locked and also the
+ * explicitly named table is locked (@table_ptr field in the state structure) if
+ * specified. Between dumps no table is locked. Also the current config is
+ * locked (by config_add_obstacle()) during table dumps as some data (strings,
+ * filters) are shared from the config and the running table dump may be
+ * interrupted by reconfiguration.
+ *
+ * Supported standards:
+ * - RFC 6396 - MRT format standard
+ * - RFC 8050 - ADD_PATH extension
+ */
+
+#include <unistd.h>
+#include <limits.h>
+#include <errno.h>
+
+#include "mrt.h"
+
+#include "nest/cli.h"
+#include "filter/filter.h"
+#include "proto/bgp/bgp.h"
+#include "sysdep/unix/unix.h"
+
+
+#ifdef PATH_MAX
+#define BIRD_PATH_MAX PATH_MAX
+#else
+#define BIRD_PATH_MAX 4096
+#endif
+
+#define mrt_log(s, msg, args...)				\
+  ({								\
+    if (s->cli)							\
+      cli_printf(s->cli, -8009, msg, ## args);			\
+    if (s->proto)						\
+      log(L_ERR "%s: " msg, s->proto->p.name, ## args);		\
+  })
+
+
+/*
+ *	MRT buffer code
+ */
+
+static void
+mrt_buffer_init(buffer *b, pool *pool, size_t n)
+{
+  b->start = mb_alloc(pool, n);
+  b->pos = b->start;
+  b->end = b->start + n;
+}
+
+static void
+mrt_buffer_grow(buffer *b, size_t n)
+{
+  size_t used = b->pos - b->start;
+  size_t size = b->end - b->start;
+  size_t req = used + n;
+
+  while (size < req)
+    size = size * 3 / 2;
+
+  b->start = mb_realloc(b->start, size);
+  b->pos = b->start + used;
+  b->end = b->start + size;
+}
+
+static inline void
+mrt_buffer_need(buffer *b, size_t n)
+{
+  if (b->pos + n > b->end)
+    mrt_buffer_grow(b, n);
+}
+
+static inline uint
+mrt_buffer_pos(buffer *b)
+{
+  return b->pos - b->start;
+}
+
+static inline void
+mrt_buffer_flush(buffer *b)
+{
+  b->pos = b->start;
+}
+
+#define MRT_DEFINE_TYPE(S, T)						\
+  static inline void mrt_put_##S##_(buffer *b, T x)			\
+  {									\
+    put_##S(b->pos, x);							\
+    b->pos += sizeof(T);						\
+  }									\
+									\
+  static inline void mrt_put_##S(buffer *b, T x)			\
+  {									\
+    mrt_buffer_need(b, sizeof(T));					\
+    put_##S(b->pos, x);							\
+    b->pos += sizeof(T);						\
+  }
+
+MRT_DEFINE_TYPE(u8, u8)
+MRT_DEFINE_TYPE(u16, u16)
+MRT_DEFINE_TYPE(u32, u32)
+MRT_DEFINE_TYPE(u64, u64)
+MRT_DEFINE_TYPE(ip4, ip4_addr)
+MRT_DEFINE_TYPE(ip6, ip6_addr)
+
+static inline void
+mrt_put_ipa(buffer *b, ip_addr x)
+{
+  if (ipa_is_ip4(x))
+    mrt_put_ip4(b, ipa_to_ip4(x));
+  else
+    mrt_put_ip6(b, ipa_to_ip6(x));
+}
+
+static inline void
+mrt_put_data(buffer *b, const void *src, size_t n)
+{
+  if (!n)
+    return;
+
+  mrt_buffer_need(b, n);
+  memcpy(b->pos, src, n);
+  b->pos += n;
+}
+
+static void
+mrt_init_message(buffer *b, u16 type, u16 subtype)
+{
+  /* Reset buffer */
+  mrt_buffer_flush(b);
+  mrt_buffer_need(b, MRT_HDR_LENGTH);
+
+  /* Prepare header */
+  mrt_put_u32_(b, now_real);
+  mrt_put_u16_(b, type);
+  mrt_put_u16_(b, subtype);
+
+  /* Message length, will be fixed later */
+  mrt_put_u32_(b, 0);
+}
+
+static void
+mrt_dump_message(buffer *b, int fd)
+{
+  uint len = mrt_buffer_pos(b);
+
+  /* Fix message length */
+  ASSERT(len >= MRT_HDR_LENGTH);
+  put_u32(b->start + 8, len - MRT_HDR_LENGTH);
+
+  if (fd < 0)
+    return;
+
+  if (write(fd, b->start, len) < 0)
+    log(L_ERR "Write to MRT file failed: %m"); /* TODO: name of file */
+}
+
+static int
+bstrsub(char *dst, size_t n, const char *src, const char *key, const char *val)
+{
+  const char *last, *next;
+  char *pos = dst;
+  size_t step, klen = strlen(key), vlen = strlen(val);
+
+  for (last = src; next = strstr(last, key); last = next + klen)
+  {
+    step = next - last;
+    if (n <= step + vlen)
+      return 0;
+
+    memcpy(pos, last, step);
+    ADVANCE(pos, n, step);
+
+    memcpy(pos, val, vlen);
+    ADVANCE(pos, n, vlen);
+  }
+
+  step = strlen(last);
+  if (n <= step)
+    return 0;
+
+  memcpy(pos, last, step);
+  ADVANCE(pos, n, step);
+
+  pos[0] = 0;
+  return 1;
+}
+
+static inline rtable *
+mrt_next_table_(rtable *tab, rtable *tab_ptr, const char *pattern)
+{
+  /* Handle explicit table, return it in the first pass */
+  if (tab_ptr)
+    return !tab ? tab_ptr : NULL;
+
+  /* Walk routing_tables list, starting after tab (if non-NULL) */
+  for (tab = !tab ? HEAD(routing_tables) : NODE_NEXT(tab);
+       NODE_VALID(tab);
+       tab = NODE_NEXT(tab))
+    if (patmatch(pattern, tab->name))
+      return tab;
+
+  return NULL;
+}
+
+static rtable *
+mrt_next_table(struct mrt_table_dump_state *s)
+{
+  rtable *tab = mrt_next_table_(s->table, s->table_ptr, s->table_expr);
+
+  if (s->table)
+    rt_unlock_table(s->table);
+
+  s->table = tab;
+
+  if (s->table)
+    rt_lock_table(s->table);
+
+  return s->table;
+}
+
+static int
+mrt_open_file(struct mrt_table_dump_state *s)
+{
+  char fmt1[BIRD_PATH_MAX];
+  char name[BIRD_PATH_MAX];
+
+  if (!bstrsub(fmt1, sizeof(fmt1), s->filename, "%N", s->table->name) ||
+      !tm_format_real_time(name, sizeof(name), fmt1, now_real))
+  {
+    mrt_log(s, "Invalid filename '%s'", s->filename);
+    return 0;
+  }
+
+  s->file = rf_open(s->pool, name, "a");
+  if (!s->file)
+  {
+    mrt_log(s, "Unable to open MRT file '%s': %m", name);
+    return 0;
+  }
+
+  s->fd = rf_fileno(s->file);
+  s->time_offset = now_real - now;
+
+  return 1;
+}
+
+static void
+mrt_close_file(struct mrt_table_dump_state *s)
+{
+  rfree(s->file);
+  s->file = NULL;
+  s->fd = -1;
+}
+
+
+/*
+ *	MRT Table Dump: Peer Index Table
+ */
+
+#define PEER_KEY(n)		n->peer_id, n->peer_as, n->peer_ip
+#define PEER_NEXT(n)		n->next
+#define PEER_EQ(id1,as1,ip1,id2,as2,ip2) \
+  id1 == id2 && as1 == as2 && ipa_equal(ip1, ip2)
+#define PEER_FN(id,as,ip)	ipa_hash(ip)
+
+static void
+mrt_peer_table_header(struct mrt_table_dump_state *s, u32 router_id, const char *name)
+{
+  buffer *b = &s->buf;
+
+  /* Collector BGP ID */
+  mrt_put_u32(b, router_id);
+
+  /* View Name */
+  uint name_length = name ? strlen(name) : 0;
+  name_length = MIN(name_length, 65535);
+  mrt_put_u16(b, name_length);
+  mrt_put_data(b, name, name_length);
+
+  /* Peer Count, will be fixed later */
+  s->peer_count = 0;
+  s->peer_count_offset = mrt_buffer_pos(b);
+  mrt_put_u16(b, 0);
+
+  HASH_INIT(s->peer_hash, s->pool, 10);
+}
+
+static void
+mrt_peer_table_entry(struct mrt_table_dump_state *s, u32 peer_id, u32 peer_as, ip_addr peer_ip)
+{
+  buffer *b = &s->buf;
+
+  uint type = MRT_PEER_TYPE_32BIT_ASN;
+  if (ipa_is_ip6(peer_ip))
+    type |= MRT_PEER_TYPE_IPV6;
+
+  /* Dump peer to buffer */
+  mrt_put_u8(b, type);
+  mrt_put_u32(b, peer_id);
+  mrt_put_ipa(b, peer_ip);
+  mrt_put_u32(b, peer_as);
+
+  /* Add peer to hash table */
+  struct mrt_peer_entry *n = lp_allocz(s->peer_lp, sizeof(struct mrt_peer_entry));
+  n->peer_id = peer_id;
+  n->peer_as = peer_as;
+  n->peer_ip = peer_ip;
+  n->index = s->peer_count++;
+
+  HASH_INSERT(s->peer_hash, PEER, n);
+}
+
+static void
+mrt_peer_table_dump(struct mrt_table_dump_state *s)
+{
+  mrt_init_message(&s->buf, MRT_TABLE_DUMP_V2, MRT_PEER_INDEX_TABLE);
+  mrt_peer_table_header(s, config->router_id, s->table->name);
+
+  /* 0 is fake peer for non-BGP routes */
+  mrt_peer_table_entry(s, 0, 0, IPA_NONE);
+
+#ifdef CONFIG_BGP
+  struct proto *P;
+  WALK_LIST(P, active_proto_list)
+    if (P->proto == &proto_bgp)
+    {
+      struct bgp_proto *p = (void *) P;
+      mrt_peer_table_entry(s, p->remote_id, p->remote_as, p->cf->remote_ip);
+    }
+#endif
+
+  /* Fix Peer Count */
+  put_u16(s->buf.start + s->peer_count_offset, s->peer_count);
+
+  mrt_dump_message(&s->buf, s->fd);
+}
+
+static void
+mrt_peer_table_flush(struct mrt_table_dump_state *s)
+{
+  lp_flush(s->peer_lp);
+  HASH_FREE(s->peer_hash);
+}
+
+
+/*
+ *	MRT Table Dump: RIB Table
+ */
+
+static void
+mrt_rib_table_header(struct mrt_table_dump_state *s, net *n)
+{
+  buffer *b = &s->buf;
+
+  /* Sequence Number */
+  mrt_put_u32(b, s->seqnum);
+
+  /* Network Prefix */
+  ip_addr a = n->n.prefix;
+  ipa_hton(a);
+
+  mrt_put_u8(b, n->n.pxlen);
+  mrt_put_data(b, &a, BYTES(n->n.pxlen));
+
+  /* Entry Count, will be fixed later */
+  s->entry_count = 0;
+  s->entry_count_offset = mrt_buffer_pos(b);
+  mrt_put_u16(b, 0);
+}
+
+static void
+mrt_rib_table_entry(struct mrt_table_dump_state *s, rte *r, struct ea_list *tmpa)
+{
+  buffer *b = &s->buf;
+  uint peer = 0;
+
+#ifdef CONFIG_BGP
+  /* Find peer index */
+  if (r->attrs->src->proto->proto == &proto_bgp)
+  {
+    struct bgp_proto *p = (void *) r->attrs->src->proto;
+    struct mrt_peer_entry *n =
+      HASH_FIND(s->peer_hash, PEER, p->remote_id, p->remote_as, p->cf->remote_ip);
+
+    peer = n ? n->index : 0;
+  }
+#endif
+
+  /* Peer Index and Originated Time */
+  mrt_put_u16(b, peer);
+  mrt_put_u32(b, r->lastmod + s->time_offset);
+
+  /* Path Identifier */
+  if (s->add_path)
+    mrt_put_u32(b, r->attrs->src->private_id);
+
+  /* Route Attributes */
+  mrt_put_u16(b, 0);
+
+#ifdef CONFIG_BGP
+  if (r->attrs->eattrs || tmpa)
+  {
+    struct ea_list *eattrs = r->attrs->eattrs;
+
+    if (!rta_is_cached(r->attrs) || tmpa)
+    {
+      /* Attributes must be merged and sorted for bgp_encode_attrs() */
+      tmpa = ea_append(tmpa, eattrs);
+      eattrs = alloca(ea_scan(tmpa));
+      ea_merge(tmpa, eattrs);
+      ea_sort(eattrs);
+    }
+
+    mrt_buffer_need(b, MRT_ATTR_BUFFER_SIZE);
+    int alen = bgp_encode_attrs(NULL, b->pos, eattrs, MRT_ATTR_BUFFER_SIZE);
+
+    if (alen < 0)
+    {
+      mrt_log(s, "Attribute list too long for %I/%d",
+	      r->net->n.prefix, r->net->n.pxlen);
+      alen = 0;
+    }
+
+    put_u16(b->pos - 2, alen);
+    b->pos += alen;
+  }
+#endif
+
+  s->entry_count++;
+}
+
+static void
+mrt_rib_table_dump(struct mrt_table_dump_state *s, net *n, int add_path)
+{
+  rte *rt, *rt0;
+  int subtype;
+
+  s->add_path = add_path;
+
+#ifndef IPV6
+  subtype = !add_path ? MRT_RIB_IPV4_UNICAST : MRT_RIB_IPV4_UNICAST_ADDPATH;
+#else
+  subtype = !add_path ? MRT_RIB_IPV6_UNICAST : MRT_RIB_IPV6_UNICAST_ADDPATH;
+#endif
+
+  mrt_init_message(&s->buf, MRT_TABLE_DUMP_V2, subtype);
+  mrt_rib_table_header(s, n);
+
+  for (rt0 = n->routes; rt = rt0; rt0 = rt0->next)
+  {
+    if (rte_is_filtered(rt))
+      continue;
+
+    /* Skip routes that should be reported in the other phase */
+    if (!s->always_add_path && (!rt->attrs->src->private_id != !s->add_path))
+    {
+      s->want_add_path = 1;
+      continue;
+    }
+
+    struct ea_list *tmp_attrs = rte_make_tmp_attrs(rt, s->linpool);
+
+    if (f_run(s->filter, &rt, &tmp_attrs, s->linpool, 0) <= F_ACCEPT)
+      mrt_rib_table_entry(s, rt, tmp_attrs);
+
+    if (rt != rt0)
+      rte_free(rt);
+
+    lp_flush(s->linpool);
+  }
+
+  /* Fix Entry Count */
+  put_u16(s->buf.start + s->entry_count_offset, s->entry_count);
+
+  /* Update max counter */
+  s->max -= 1 + s->entry_count;
+
+  /* Skip empty entries */
+  if (!s->entry_count)
+    return;
+
+  s->seqnum++;
+  mrt_dump_message(&s->buf, s->fd);
+}
+
+
+/*
+ *	MRT Table Dump: main logic
+ */
+
+static struct mrt_table_dump_state *
+mrt_table_dump_init(pool *pp)
+{
+  pool *pool = rp_new(pp, "MRT Table Dump");
+  struct mrt_table_dump_state *s = mb_allocz(pool, sizeof(struct mrt_table_dump_state));
+
+  s->pool = pool;
+  s->linpool = lp_new(pool, 4080);
+  s->peer_lp = lp_new(pool, 4080);
+  mrt_buffer_init(&s->buf, pool, 2 * MRT_ATTR_BUFFER_SIZE);
+
+  /* We lock the current config as we may reference it indirectly by filter */
+  s->config = config;
+  config_add_obstacle(s->config);
+
+  s->fd = -1;
+
+  return s;
+}
+
+static void
+mrt_table_dump_free(struct mrt_table_dump_state *s)
+{
+  if (s->table_open)
+    FIB_ITERATE_UNLINK(&s->fit, &s->table->fib);
+
+  if (s->table)
+    rt_unlock_table(s->table);
+
+  if (s->table_ptr)
+    rt_unlock_table(s->table_ptr);
+
+  config_del_obstacle(s->config);
+
+  rfree(s->pool);
+}
+
+
+static int
+mrt_table_dump_step(struct mrt_table_dump_state *s)
+{
+  s->max = 2048;
+
+  if (s->table_open)
+    goto step;
+
+  while (mrt_next_table(s))
+  {
+    if (!mrt_open_file(s))
+      continue;
+
+    mrt_peer_table_dump(s);
+
+    FIB_ITERATE_INIT(&s->fit, &s->table->fib);
+    s->table_open = 1;
+
+  step:
+    FIB_ITERATE_START(&s->table->fib, &s->fit, fn)
+    {
+      if (s->max < 0)
+      {
+	FIB_ITERATE_PUT(&s->fit, fn);
+	return 0;
+      }
+
+      /* With Always ADD_PATH option, we jump directly to second phase */
+      s->want_add_path = s->always_add_path;
+
+      if (s->want_add_path == 0)
+	mrt_rib_table_dump(s, (net *) fn, 0);
+
+      if (s->want_add_path == 1)
+	mrt_rib_table_dump(s, (net *) fn, 1);
+    }
+    FIB_ITERATE_END(fn);
+    s->table_open = 0;
+
+    mrt_close_file(s);
+    mrt_peer_table_flush(s);
+  }
+
+  return 1;
+}
+
+static void
+mrt_timer(timer *t)
+{
+  struct mrt_proto *p = t->data;
+  struct mrt_config *cf = (void *) (p->p.cf);
+
+  if (p->table_dump)
+  {
+    log(L_WARN "%s: Earlier RIB table dump still not finished, skipping next one", p->p.name);
+    return;
+  }
+
+  TRACE(D_EVENTS, "RIB table dump started");
+
+  struct mrt_table_dump_state *s = mrt_table_dump_init(p->p.pool);
+
+  s->proto = p;
+  s->table_expr = cf->table_expr;
+  s->table_ptr = cf->table_cf ? cf->table_cf->table : NULL;
+  s->filter = cf->filter;
+  s->filename = cf->filename;
+  s->always_add_path = cf->always_add_path;
+
+  if (s->table_ptr)
+    rt_lock_table(s->table_ptr);
+
+  p->table_dump = s;
+  ev_schedule(p->event);
+}
+
+static void
+mrt_event(void *P)
+{
+  struct mrt_proto *p = P;
+
+  if (!p->table_dump)
+    return;
+
+  if (!mrt_table_dump_step(p->table_dump))
+  {
+    ev_schedule(p->event);
+    return;
+  }
+
+  mrt_table_dump_free(p->table_dump);
+  p->table_dump = NULL;
+
+  TRACE(D_EVENTS, "RIB table dump done");
+
+  if (p->p.proto_state == PS_STOP)
+    proto_notify_state(&p->p, PS_DOWN);
+}
+
+
+/*
+ *	MRT Table Dump: CLI command
+ */
+
+static void
+mrt_dump_cont(struct cli *c)
+{
+  if (!mrt_table_dump_step(c->rover))
+    return;
+
+  cli_printf(c, 0, "");
+  mrt_table_dump_free(c->rover);
+  c->cont = c->cleanup = c->rover = NULL;
+}
+
+static void
+mrt_dump_cleanup(struct cli *c)
+{
+  mrt_table_dump_free(c->rover);
+  c->rover = NULL;
+}
+
+void
+mrt_dump_cmd(struct mrt_dump_data *d)
+{
+  if (cli_access_restricted())
+    return;
+
+  if (!d->table_expr && !d->table_ptr)
+    cf_error("Table not specified");
+
+  if (!d->filename)
+    cf_error("File not specified");
+
+  struct mrt_table_dump_state *s = mrt_table_dump_init(this_cli->pool);
+
+  s->cli = this_cli;
+  s->table_expr = d->table_expr;
+  s->table_ptr = d->table_ptr;
+  s->filter = d->filter;
+  s->filename = d->filename;
+
+  if (s->table_ptr)
+    rt_lock_table(s->table_ptr);
+
+  this_cli->cont = mrt_dump_cont;
+  this_cli->cleanup = mrt_dump_cleanup;
+  this_cli->rover = s;
+}
+
+
+/*
+ *	MRT BGP4MP dump
+ */
+
+static buffer *
+mrt_bgp_buffer(void)
+{
+  /* Static buffer for BGP4MP dump, TODO: change to use MRT protocol */
+  static buffer b;
+
+  if (!b.start)
+    mrt_buffer_init(&b, &root_pool, 1024);
+
+  return &b;
+}
+
+static void
+mrt_bgp_header(buffer *b, struct mrt_bgp_data *d)
+{
+  if (d->as4)
+  {
+    mrt_put_u32(b, d->peer_as);
+    mrt_put_u32(b, d->local_as);
+  }
+  else
+  {
+    mrt_put_u16(b, (d->peer_as <= 0xFFFF) ? d->peer_as : AS_TRANS);
+    mrt_put_u16(b, (d->local_as <= 0xFFFF) ? d->local_as : AS_TRANS);
+  }
+
+  mrt_put_u16(b, (d->index <= 0xFFFF) ? d->index : 0);
+  mrt_put_u16(b, d->af);
+
+  if (d->af == BGP_AF_IPV4)
+  {
+    mrt_put_ip4(b, ipa_to_ip4(d->peer_ip));
+    mrt_put_ip4(b, ipa_to_ip4(d->local_ip));
+  }
+  else
+  {
+    mrt_put_ip6(b, ipa_to_ip6(d->peer_ip));
+    mrt_put_ip6(b, ipa_to_ip6(d->local_ip));
+  }
+}
+
+void
+mrt_dump_bgp_message(struct mrt_bgp_data *d)
+{
+  const u16 subtypes[] = {
+    MRT_BGP4MP_MESSAGE,			MRT_BGP4MP_MESSAGE_AS4,
+    MRT_BGP4MP_MESSAGE_LOCAL,		MRT_BGP4MP_MESSAGE_AS4_LOCAL,
+    MRT_BGP4MP_MESSAGE_ADDPATH,		MRT_BGP4MP_MESSAGE_AS4_ADDPATH,
+    MRT_BGP4MP_MESSAGE_LOCAL_ADDPATH,	MRT_BGP4MP_MESSAGE_AS4_LOCAL_ADDPATH,
+  };
+
+  buffer *b = mrt_bgp_buffer();
+  mrt_init_message(b, MRT_BGP4MP, subtypes[d->as4 + 4*d->add_path]);
+  mrt_bgp_header(b, d);
+  mrt_put_data(b, d->message, d->msg_len);
+  mrt_dump_message(b, config->mrtdump_file);
+}
+
+void
+mrt_dump_bgp_state_change(struct mrt_bgp_data *d)
+{
+  /* Convert state from our BS_* values to values used in MRTDump */
+  const u16 states[BS_MAX] = {1, 2, 3, 4, 5, 6, 1};
+
+  if (states[d->old_state] == states[d->new_state])
+    return;
+
+  /* Always use AS4 mode for STATE_CHANGE */
+  d->as4 = 1;
+
+  buffer *b = mrt_bgp_buffer();
+  mrt_init_message(b, MRT_BGP4MP, MRT_BGP4MP_STATE_CHANGE_AS4);
+  mrt_bgp_header(b, d);
+  mrt_put_u16(b, states[d->old_state]);
+  mrt_put_u16(b, states[d->new_state]);
+  mrt_dump_message(b, config->mrtdump_file);
+}
+
+
+/*
+ *	MRT protocol glue
+ */
+
+void
+mrt_check_config(struct proto_config *C)
+{
+  struct mrt_config *cf = (void *) C;
+
+  /* c.table must be always defined, but it is relevant only if table_expr is not set */
+  if (!cf->table_expr)
+    cf->table_cf = cf->c.table;
+
+  if (!cf->table_expr && !cf->table_cf)
+    cf_error("Table not specified");
+
+  if (!cf->filename)
+    cf_error("File not specified");
+
+  if (!cf->period)
+    cf_error("Period not specified");
+}
+
+static struct proto *
+mrt_init(struct proto_config *C)
+{
+  struct proto *P = proto_new(C, sizeof(struct mrt_proto));
+
+  return P;
+}
+
+static int
+mrt_start(struct proto *P)
+{
+  struct mrt_proto *p = (void *) P;
+  struct mrt_config *cf = (void *) (P->cf);
+
+  p->timer = tm_new_set(P->pool, mrt_timer, p, 0, cf->period);
+  p->event = ev_new_set(P->pool, mrt_event, p);
+
+  tm_start(p->timer, cf->period);
+
+  return PS_UP;
+}
+
+static int
+mrt_shutdown(struct proto *P)
+{
+  struct mrt_proto *p = (void *) P;
+
+  return p->table_dump ? PS_STOP : PS_DOWN;
+}
+
+static int
+mrt_reconfigure(struct proto *P, struct proto_config *CF)
+{
+  struct mrt_proto *p = (void *) P;
+  struct mrt_config *old = (void *) (P->cf);
+  struct mrt_config *new = (void *) CF;
+
+  if (new->period != old->period)
+  {
+    TRACE(D_EVENTS, "Changing period from %d to %d s", old->period, new->period);
+
+    bird_clock_t new_time = p->timer->expires - old->period + new->period;
+    tm_start(p->timer, (new_time > now) ? (new_time - now) : 0);
+    p->timer->recurrent = new->period;
+  }
+
+  return 1;
+}
+
+static void
+mrt_copy_config(struct proto_config *dest, struct proto_config *src)
+{
+  /* Just a shallow copy, not many items here */
+  proto_copy_rest(dest, src, sizeof(struct mrt_config));
+}
+
+
+struct protocol proto_mrt = {
+  .name =		"MRT",
+  .template =		"mrt%d",
+  .config_size =	sizeof(struct mrt_config),
+  .init =		mrt_init,
+  .start =		mrt_start,
+  .shutdown =		mrt_shutdown,
+  .reconfigure =	mrt_reconfigure,
+  .copy_config =	mrt_copy_config,
+};
--- a/proto/mrt/mrt.h
+++ b/proto/mrt/mrt.h
@ -0,0 +1,156 @@
+/*
+ *	BIRD -- Multi-Threaded Routing Toolkit (MRT) Protocol
+ *
+ *	(c) 2017--2018 Ondrej Zajicek <santiago@crfreenet.org>
+ *	(c) 2017--2018 CZ.NIC z.s.p.o.
+ *
+ *	Can be freely distributed and used under the terms of the GNU GPL.
+ */
+
+#ifndef _BIRD_MRT_H_
+#define _BIRD_MRT_H_
+
+#include "nest/bird.h"
+#include "nest/protocol.h"
+#include "lib/lists.h"
+#include "nest/route.h"
+#include "lib/event.h"
+#include "lib/hash.h"
+
+
+struct mrt_config {
+  struct proto_config c;
+
+  struct rtable_config *table_cf;
+  const char *table_expr;
+  struct filter *filter;
+  const char *filename;
+  uint period;
+  int always_add_path;
+};
+
+struct mrt_proto {
+  struct proto p;
+  timer *timer;
+  event *event;
+
+  struct mrt_target *file;
+  struct mrt_table_dump_state *table_dump;
+};
+
+struct mrt_dump_data {
+  const char *table_expr;
+  struct rtable *table_ptr;
+  struct filter *filter;
+  char *filename;
+};
+
+struct mrt_peer_entry {
+  u32 index;
+  u32 peer_id;
+  u32 peer_as;
+  ip_addr peer_ip;
+  struct mrt_peer_entry *next;
+};
+
+struct mrt_table_dump_state {
+  struct mrt_proto *proto;		/* Protocol for regular MRT dumps (or NULL) */
+  struct cli *cli;			/* CLI for irregular MRT dumps (or NULL) */
+  struct config *config;		/* Config valid during start of dump, locked */
+
+					/* Configuration information */
+  const char *table_expr;		/* Wildcard for table name (or NULL) */
+  struct rtable *table_ptr;		/* Explicit table (or NULL) */
+  struct filter *filter;		/* Optional filter */
+  const char *filename;			/* Filename pattern */
+  int always_add_path;			/* Always use *_ADDPATH message subtypes */
+
+  /* Allocated by mrt_table_dump_init() */
+  pool *pool;				/* Pool for table dump */
+  linpool *linpool;			/* Temporary linear pool */
+  linpool *peer_lp;			/* Linear pool for peer entries in peer_hash */
+  buffer buf;				/* Buffer for MRT messages */
+
+  HASH(struct mrt_peer_entry) peer_hash; /* Hash for peers to find the index */
+
+  struct rtable *table;			/* Processed table, NULL initially */
+  struct fib_iterator fit;		/* Iterator in processed table */
+  int table_open;			/* Whether iterator is linked */
+
+  int add_path;				/* Current message subtype is *_ADDPATH */
+  int want_add_path;			/* Want *_ADDPATH message later */
+  int max;				/* Decreasing counter of dumped routes */
+  u32 seqnum;				/* MRT message sequence number */
+  bird_clock_t time_offset;		/* Time offset between monotonic and real time */
+
+  u16 peer_count;			/* Number of peers */
+  u32 peer_count_offset;		/* Buffer offset to store peer_count later */
+  u16 entry_count;			/* Number of RIB Entries */
+  u32 entry_count_offset;		/* Buffer offset to store entry_count later */
+
+  struct rfile *file;			/* tracking for mrt table dump file */
+  int fd;
+};
+
+struct mrt_bgp_data {
+  uint peer_as;
+  uint local_as;
+  uint index;
+  uint af;
+  ip_addr peer_ip;
+  ip_addr local_ip;
+  byte *message;
+  uint msg_len;
+  uint old_state;
+  uint new_state;
+  u8 as4;
+  u8 add_path;
+};
+
+
+#define MRT_HDR_LENGTH		12	/* MRT Timestamp + MRT Type + MRT Subtype + MRT Load Length */
+#define MRT_PEER_TYPE_32BIT_ASN	2	/* MRT Table Dump: Peer Index Table: Peer Type: Use 32bit ASN */
+#define MRT_PEER_TYPE_IPV6	1	/* MRT Table Dump: Peer Index Table: Peer Type: Use IPv6 IP Address */
+
+#define MRT_ATTR_BUFFER_SIZE	65536
+
+/* MRT Types */
+#define MRT_TABLE_DUMP_V2 	13
+#define MRT_BGP4MP		16
+
+/* MRT Table Dump v2 Subtypes */
+#define MRT_PEER_INDEX_TABLE		1
+#define MRT_RIB_IPV4_UNICAST		2
+#define MRT_RIB_IPV4_MULTICAST		3
+#define MRT_RIB_IPV6_UNICAST		4
+#define MRT_RIB_IPV6_MULTICAST 		5
+#define MRT_RIB_GENERIC			6
+#define MRT_RIB_IPV4_UNICAST_ADDPATH	8
+#define MRT_RIB_IPV4_MULTICAST_ADDPATH	9
+#define MRT_RIB_IPV6_UNICAST_ADDPATH	10
+#define MRT_RIB_IPV6_MULTICAST_ADDPATH 	11
+#define MRT_RIB_GENERIC_ADDPATH		12
+
+/* MRT BGP4MP Subtypes */
+#define MRT_BGP4MP_MESSAGE		1
+#define MRT_BGP4MP_MESSAGE_AS4		4
+#define MRT_BGP4MP_STATE_CHANGE_AS4	5
+#define MRT_BGP4MP_MESSAGE_LOCAL	6
+#define MRT_BGP4MP_MESSAGE_AS4_LOCAL	7
+#define MRT_BGP4MP_MESSAGE_ADDPATH	8
+#define MRT_BGP4MP_MESSAGE_AS4_ADDPATH	9
+#define MRT_BGP4MP_MESSAGE_LOCAL_ADDPATH	10
+#define MRT_BGP4MP_MESSAGE_AS4_LOCAL_ADDPATH	11
+
+
+#ifdef CONFIG_MRT
+void mrt_dump_cmd(struct mrt_dump_data *d);
+void mrt_dump_bgp_message(struct mrt_bgp_data *d);
+void mrt_dump_bgp_state_change(struct mrt_bgp_data *d);
+void mrt_check_config(struct proto_config *C);
+#else
+static inline void mrt_dump_bgp_message(struct mrt_bgp_data *d UNUSED) { }
+static inline void mrt_dump_bgp_state_change(struct mrt_bgp_data *d UNUSED) { }
+#endif
+
+#endif	/* _BIRD_MRT_H_ */
--- a/proto/ospf/dbdes.c
+++ b/proto/ospf/dbdes.c
@ -121,7 +121,7 @@ ospf_prepare_dbdes(struct ospf_proto *p, struct ospf_neighbor *n)
  {
    struct ospf_dbdes2_packet *ps = (void *) pkt;
    ps->iface_mtu = htons(iface_mtu);
-    ps->options = ifa->oa->options;
+    ps->options = ifa->oa->options & ~OPT_N;
    ps->imms = 0;	/* Will be set later */
    ps->ddseq = htonl(n->dds);
    length = sizeof(struct ospf_dbdes2_packet);
@ -129,7 +129,7 @@ ospf_prepare_dbdes(struct ospf_proto *p, struct ospf_neighbor *n)
  else /* OSPFv3 */
  {
    struct ospf_dbdes3_packet *ps = (void *) pkt;
-    ps->options = htonl(ifa->oa->options);
+    ps->options = htonl(ifa->oa->options & ~OPT_N);
    ps->iface_mtu = htons(iface_mtu);
    ps->padding = 0;
    ps->imms = 0;	/* Will be set later */
@ -347,6 +347,7 @@ ospf_receive_dbdes(struct ospf_packet *pkt, struct ospf_iface *ifa,
    ospf_neigh_sm(n, INM_2WAYREC);
    if (n->state != NEIGHBOR_EXSTART)
      return;
+    /* fallthrough */

  case NEIGHBOR_EXSTART:
    if ((ifa->type != OSPF_IT_VLINK) &&
--- a/proto/ospf/hello.c
+++ b/proto/ospf/hello.c
@ -283,7 +283,7 @@ ospf_receive_hello(struct ospf_packet *pkt, struct ospf_iface *ifa,
      if (!ipa_equal(faddr, n->ip))
      {
 	OSPF_TRACE(D_EVENTS, "Neighbor %R on %s changed IP address to %I",
-		   n->rid, ifa->ifname, n->ip, faddr);
+		   n->rid, ifa->ifname, faddr);
 	n->ip = faddr;
      }
    }
--- a/proto/ospf/iface.c
+++ b/proto/ospf/iface.c
@ -524,6 +524,10 @@ add_nbma_node(struct ospf_iface *ifa, struct nbma_node *src, int found)
 static int
 ospf_iface_stubby(struct ospf_iface_patt *ip, struct ifa *addr)
 {
+  /* vlink cannot be stub */
+  if (ip->type == OSPF_IT_VLINK)
+    return 0;
+
  /* a host address */
  if (addr->flags & IA_HOST)
    return 1;
@ -861,6 +865,7 @@ ospf_iface_reconfigure(struct ospf_iface *ifa, struct ospf_iface_patt *new)
 	       ifname, ifa->priority, new->priority);

    ifa->priority = new->priority;
+    ospf_iface_sm(ifa, ISM_NEICH);
    ospf_notify_link_lsa(ifa);
  }

@ -1256,7 +1261,8 @@ ospf_iface_change_mtu(struct ospf_proto *p, struct ospf_iface *ifa)
 {
  /* ifa is not vlink */

-  OSPF_TRACE(D_EVENTS, "Interface %s changed MTU to %d", ifa->iface->mtu);
+  OSPF_TRACE(D_EVENTS, "Interface %s changed MTU to %d",
+	     ifa->ifname, ifa->iface->mtu);

  ifa->tx_length = ifa_tx_length(ifa);

--- a/proto/ospf/neighbor.c
+++ b/proto/ospf/neighbor.c
@ -584,8 +584,11 @@ ospf_neigh_bfd_hook(struct bfd_request *req)
 void
 ospf_neigh_update_bfd(struct ospf_neighbor *n, int use_bfd)
 {
+  struct ospf_proto *p = n->ifa->oa->po;
+
  if (use_bfd && !n->bfd_req)
-    n->bfd_req = bfd_request_session(n->pool, n->ip, n->ifa->addr->ip, n->ifa->iface,
+    n->bfd_req = bfd_request_session(n->pool, n->ip, n->ifa->addr->ip,
+				     n->ifa->iface, p->p.vrf,
 				     ospf_neigh_bfd_hook, n);

  if (!use_bfd && n->bfd_req)
--- a/proto/ospf/ospf.c
+++ b/proto/ospf/ospf.c
@ -1203,6 +1203,7 @@ ospf_sh_state(struct proto *P, int verbose, int reachable)
 	he->domain = 1; /* Abuse domain field to mark the LSA */
 	hex[jx++] = he;
      }
+      /* fallthrough */
    default:
      accept = 0;
    }
--- a/proto/ospf/packet.c
+++ b/proto/ospf/packet.c
@ -56,6 +56,7 @@ ospf_pkt_finalize(struct ospf_iface *ifa, struct ospf_packet *pkt, uint *plen)
      return;
    }
    strncpy(auth->password, pass->password, sizeof(auth->password));
+    /* fallthrough */

  case OSPF_AUTH_NONE:
    {
--- a/proto/ospf/topology.c
+++ b/proto/ospf/topology.c
@ -404,6 +404,8 @@ ospf_refresh_lsa(struct ospf_proto *p, struct top_hash_entry *en)
 void
 ospf_flush_lsa(struct ospf_proto *p, struct top_hash_entry *en)
 {
+  en->nf = NULL;
+
  if (en->next_lsa_body)
  {
    mb_free(en->next_lsa_body);
--- a/proto/pipe/pipe.c
+++ b/proto/pipe/pipe.c
@ -89,6 +89,12 @@ pipe_rt_notify(struct proto *P, rtable *src_table, net *n, rte *new, rte *old, e
 	  memcpy(&(e->u), &(new->u), sizeof(e->u));
 	  e->pref = new->pref;
 	  e->pflags = new->pflags;
+
+#ifdef CONFIG_BGP
+	  /* Hack to cleanup cached value */
+	  if (e->attrs->src->proto->proto == &proto_bgp)
+	    e->u.bgp.stale = -1;
+#endif
 	}

      src = a.src;
@ -230,12 +236,18 @@ pipe_reconfigure(struct proto *P, struct proto_config *new)
  if ((oc->peer->table != nc->peer->table) || (oc->mode != nc->mode))
    return 0;

+  int import_changed = ! filter_same(new->in_filter, old->in_filter);
+  int export_changed = ! filter_same(new->out_filter, old->out_filter);
+
  /* Update output filters in ahooks */
  if (P->main_ahook)
    {
      P->main_ahook->out_filter = new->out_filter;
      P->main_ahook->in_limit = new->in_limit;
      proto_verify_limits(P->main_ahook);
+
+      if (export_changed)
+	P->main_ahook->last_out_filter_change = now;
    }

  if (p->peer_ahook)
@ -243,14 +255,15 @@ pipe_reconfigure(struct proto *P, struct proto_config *new)
      p->peer_ahook->out_filter = new->in_filter;
      p->peer_ahook->in_limit = new->out_limit;
      proto_verify_limits(p->peer_ahook);
+
+      if (import_changed)
+	p->peer_ahook->last_out_filter_change = now;
    }

  if ((P->proto_state != PS_UP) || (proto_reconfig_type == RECONFIG_SOFT))
    return 1;

-  if ((new->preference != old->preference)
-      || ! filter_same(new->in_filter, old->in_filter)
-      || ! filter_same(new->out_filter, old->out_filter))
+  if (import_changed || export_changed || (new->preference != old->preference))
    proto_request_feeding(P);

  return 1;
--- a/proto/radv/radv.c
+++ b/proto/radv/radv.c
@ -281,17 +281,6 @@ radv_iface_add(struct object_lock *lock)
  radv_iface_notify(ifa, RA_EV_INIT);
 }

-static inline struct ifa *
-find_lladdr(struct iface *iface)
-{
-  struct ifa *a;
-  WALK_LIST(a, iface->addrs)
-    if (a->scope == SCOPE_LINK)
-      return a;
-
-  return NULL;
-}
-
 static void
 radv_iface_new(struct radv_proto *p, struct iface *iface, struct radv_iface_config *cf)
 {
@ -305,18 +294,12 @@ radv_iface_new(struct radv_proto *p, struct iface *iface, struct radv_iface_conf
  ifa->ra = p;
  ifa->cf = cf;
  ifa->iface = iface;
+  ifa->addr = iface->llv6;
  init_list(&ifa->prefixes);
  ifa->prune_time = TIME_INFINITY;

  add_tail(&p->iface_list, NODE ifa);

-  ifa->addr = find_lladdr(iface);
-  if (!ifa->addr)
-  {
-    log(L_ERR "%s: Missing link-local address on interface %s", p->p.name, iface->name);
-    return;
-  }
-
  timer *tm = tm_new(pool);
  tm->hook = radv_timer;
  tm->data = ifa;
@ -360,6 +343,10 @@ radv_if_notify(struct proto *P, unsigned flags, struct iface *iface)
    struct radv_iface_config *ic = (struct radv_iface_config *)
      iface_patt_find(&cf->patt_list, iface, NULL);

+    /* Ignore ifaces without link-local address */
+    if (!iface->llv6)
+      return;
+
    if (ic)
      radv_iface_new(p, iface, ic);

--- a/proto/rip/packets.c
+++ b/proto/rip/packets.c
@ -739,16 +739,9 @@ rip_open_socket(struct rip_iface *ifa)
  sk->sport = ifa->cf->port;
  sk->dport = ifa->cf->port;
  sk->iface = ifa->iface;
+  sk->saddr = rip_is_v2(p) ? ifa->iface->addr->ip : ifa_llv6(ifa->iface)->ip;
  sk->vrf = p->p.vrf;

-  /*
-   * For RIPv2, we explicitly choose a primary address, mainly to ensure that
-   * RIP and BFD uses the same one. For RIPng, we left it to kernel, which
-   * should choose some link-local address based on the same scope rule.
-   */
-  if (rip_is_v2(p))
-    sk->saddr = ifa->iface->addr->ip;
-
  sk->rx_hook = rip_rx_hook;
  sk->tx_hook = rip_tx_hook;
  sk->err_hook = rip_err_hook;
--- a/proto/rip/rip.c
+++ b/proto/rip/rip.c
@ -504,7 +504,8 @@ rip_update_bfd(struct rip_proto *p, struct rip_neighbor *n)
     */
    ip_addr saddr = rip_is_v2(p) ? n->ifa->sk->saddr : n->nbr->ifa->ip;
    n->bfd_req = bfd_request_session(p->p.pool, n->nbr->addr, saddr,
-				     n->nbr->iface, rip_bfd_notify, n);
+				     n->nbr->iface, p->p.vrf,
+				     rip_bfd_notify, n);
  }

  if (!use_bfd && n->bfd_req)
@ -764,6 +765,10 @@ rip_if_notify(struct proto *P, unsigned flags, struct iface *iface)
  {
    struct rip_iface_config *ic = (void *) iface_patt_find(&cf->patt_list, iface, NULL);

+    /* For RIPng, ignore ifaces without link-local address */
+    if (rip_is_ng(p) && !ifa_llv6(iface))
+      return;
+
    if (ic)
      rip_add_iface(p, iface, ic);

--- a/proto/static/static.c
+++ b/proto/static/static.c
@ -150,7 +150,7 @@ static_update_bfd(struct proto *p, struct static_route *r)
  if (bfd_up && !r->bfd_req)
  {
    // ip_addr local = ipa_nonzero(r->local) ? r->local : nb->ifa->ip;
-    r->bfd_req = bfd_request_session(p->pool, r->via, nb->ifa->ip, nb->iface,
+    r->bfd_req = bfd_request_session(p->pool, r->via, nb->ifa->ip, nb->iface, p->vrf,
 				     static_bfd_notify, r);
  }

@ -518,6 +518,11 @@ static_match(struct proto *p, struct static_route *r, struct static_config *n)
  if (r->neigh)
    r->neigh->data = NULL;

+  if (r->dest == RTD_MULTIPATH)
+    for (t = r->mp_next; t; t = t->mp_next)
+      if (t->neigh)
+	t->neigh->data = NULL;
+
  WALK_LIST(t, n->iface_routes)
    if (static_same_net(r, t))
      goto found;
--- a/sysdep/bsd/setkey.h
+++ b/sysdep/bsd/setkey.h
@ -158,12 +158,14 @@ sk_set_md5_in_sasp_db(sock *s, ip_addr local, ip_addr remote, struct iface *ifa,
    if (len > TCP_KEYLEN_MAX)
      ERR_MSG("The password for TCP MD5 Signature is too long");

-    if (setkey_md5(&src, &dst, passwd, SADB_ADD) < 0)
+    if ((setkey_md5(&src, &dst, passwd, SADB_ADD) < 0) ||
+	(setkey_md5(&dst, &src, passwd, SADB_ADD) < 0))
      ERR_MSG("Cannot add TCP-MD5 password into the IPsec SA/SP database");
  }
  else
  {
-    if (setkey_md5(&src, &dst, NULL, SADB_DELETE) < 0)
+    if ((setkey_md5(&src, &dst, NULL, SADB_DELETE) < 0) ||
+	(setkey_md5(&dst, &src, NULL, SADB_DELETE) < 0))
      ERR_MSG("Cannot delete TCP-MD5 password from the IPsec SA/SP database");
  }
  return 0;
--- a/sysdep/bsd/sysio.h
+++ b/sysdep/bsd/sysio.h
@ -12,6 +12,11 @@
 #include <sys/param.h>


+#ifdef __FreeBSD__
+/* Should be defined in sysdep/cf/bsd.h, but it is flavor-specific */
+#define CONFIG_DONTROUTE_UNICAST
+#endif
+
 #ifdef __NetBSD__

 #ifndef IP_RECVTTL
--- a/sysdep/cf/README
+++ b/sysdep/cf/README
@ -11,6 +11,7 @@ CONFIG_MC_PROPER_SRC	Multicast packets have source address according to socket s
 CONFIG_SKIP_MC_BIND	Don't call bind on multicast socket (def for *BSD)
 CONFIG_NO_IFACE_BIND	Bind to iface is not available, use workarounds (def for *BSD)
 CONFIG_UNIX_DONTROUTE	Use setsockopts DONTROUTE (undef for *BSD)
+CONFIG_DONTROUTE_UNICAST Use MSG_DONTROUTE flag for unicast packets (def for FreeBSD)
 CONFIG_USE_HDRINCL	Use IP_HDRINCL instead of control messages for source address on raw IP sockets.

 CONFIG_RESTRICTED_PRIVILEGES	Implements restricted privileges using drop_uid()
--- a/sysdep/config.h
+++ b/sysdep/config.h
@ -7,7 +7,7 @@
 #define _BIRD_CONFIG_H_

 /* BIRD version */
-#define BIRD_VERSION "1.6.3"
+#define BIRD_VERSION "1.6.7"

 /* Include parameters determined by configure script */
 #include "sysdep/autoconf.h"
--- a/sysdep/linux/netlink.c
+++ b/sysdep/linux/netlink.c
@ -939,22 +939,25 @@ nl_send_route(struct krt_proto *p, rte *e, struct ea_list *eattrs, int op, int d
  struct {
    struct nlmsghdr h;
    struct rtmsg r;
-    char buf[128 + KRT_METRICS_MAX*8 + nh_bufsize(a->nexthops)];
-  } r;
+    char buf[0];
+  } *r;
+
+  uint rsize = sizeof(*r) + 128 + KRT_METRICS_MAX*8 + nh_bufsize(a->nexthops);
+  r = alloca(rsize);

  DBG("nl_send_route(%I/%d,op=%x)\n", net->n.prefix, net->n.pxlen, op);

-  bzero(&r.h, sizeof(r.h));
-  bzero(&r.r, sizeof(r.r));
-  r.h.nlmsg_type = op ? RTM_NEWROUTE : RTM_DELROUTE;
-  r.h.nlmsg_len = NLMSG_LENGTH(sizeof(struct rtmsg));
-  r.h.nlmsg_flags = op | NLM_F_REQUEST | NLM_F_ACK;
+  bzero(&r->h, sizeof(r->h));
+  bzero(&r->r, sizeof(r->r));
+  r->h.nlmsg_type = op ? RTM_NEWROUTE : RTM_DELROUTE;
+  r->h.nlmsg_len = NLMSG_LENGTH(sizeof(struct rtmsg));
+  r->h.nlmsg_flags = op | NLM_F_REQUEST | NLM_F_ACK;

-  r.r.rtm_family = BIRD_AF;
-  r.r.rtm_dst_len = net->n.pxlen;
-  r.r.rtm_protocol = RTPROT_BIRD;
-  r.r.rtm_scope = RT_SCOPE_NOWHERE;
-  nl_add_attr_ipa(&r.h, sizeof(r), RTA_DST, net->n.prefix);
+  r->r.rtm_family = BIRD_AF;
+  r->r.rtm_dst_len = net->n.pxlen;
+  r->r.rtm_protocol = RTPROT_BIRD;
+  r->r.rtm_scope = RT_SCOPE_NOWHERE;
+  nl_add_attr_ipa(&r->h, rsize, RTA_DST, net->n.prefix);

  /*
   * Strange behavior for RTM_DELROUTE:
@ -964,9 +967,9 @@ nl_send_route(struct krt_proto *p, rte *e, struct ea_list *eattrs, int op, int d
   */

  if (krt_table_id(p) < 256)
-    r.r.rtm_table = krt_table_id(p);
+    r->r.rtm_table = krt_table_id(p);
  else
-    nl_add_attr_u32(&r.h, sizeof(r), RTA_TABLE, krt_table_id(p));
+    nl_add_attr_u32(&r->h, rsize, RTA_TABLE, krt_table_id(p));

  if (a->source == RTS_DUMMY)
    priority = e->u.krt.metric;
@ -976,7 +979,7 @@ nl_send_route(struct krt_proto *p, rte *e, struct ea_list *eattrs, int op, int d
    priority = ea->u.data;

  if (priority)
-    nl_add_attr_u32(&r.h, sizeof(r), RTA_PRIORITY, priority);
+    nl_add_attr_u32(&r->h, rsize, RTA_PRIORITY, priority);

  /* For route delete, we do not specify remaining route attributes */
  if (op == NL_OP_DELETE)
@ -984,15 +987,15 @@ nl_send_route(struct krt_proto *p, rte *e, struct ea_list *eattrs, int op, int d

  /* Default scope is LINK for device routes, UNIVERSE otherwise */
  if (ea = ea_find(eattrs, EA_KRT_SCOPE))
-    r.r.rtm_scope = ea->u.data;
+    r->r.rtm_scope = ea->u.data;
  else
-    r.r.rtm_scope = (dest == RTD_DEVICE) ? RT_SCOPE_LINK : RT_SCOPE_UNIVERSE;
+    r->r.rtm_scope = (dest == RTD_DEVICE) ? RT_SCOPE_LINK : RT_SCOPE_UNIVERSE;

  if (ea = ea_find(eattrs, EA_KRT_PREFSRC))
-    nl_add_attr_ipa(&r.h, sizeof(r), RTA_PREFSRC, *(ip_addr *)ea->u.ptr->data);
+    nl_add_attr_ipa(&r->h, rsize, RTA_PREFSRC, *(ip_addr *)ea->u.ptr->data);

  if (ea = ea_find(eattrs, EA_KRT_REALM))
-    nl_add_attr_u32(&r.h, sizeof(r), RTA_FLOW, ea->u.data);
+    nl_add_attr_u32(&r->h, rsize, RTA_FLOW, ea->u.data);


  u32 metrics[KRT_METRICS_MAX];
@ -1007,7 +1010,7 @@ nl_send_route(struct krt_proto *p, rte *e, struct ea_list *eattrs, int op, int d
  }

  if (metrics[0])
-    nl_add_metrics(&r.h, sizeof(r), metrics, KRT_METRICS_MAX);
+    nl_add_metrics(&r->h, rsize, metrics, KRT_METRICS_MAX);


 dest:
@ -1015,26 +1018,26 @@ dest:
  switch (dest)
    {
    case RTD_ROUTER:
-      r.r.rtm_type = RTN_UNICAST;
-      nl_add_attr_u32(&r.h, sizeof(r), RTA_OIF, iface->index);
-      nl_add_attr_ipa(&r.h, sizeof(r), RTA_GATEWAY, gw);
+      r->r.rtm_type = RTN_UNICAST;
+      nl_add_attr_u32(&r->h, rsize, RTA_OIF, iface->index);
+      nl_add_attr_ipa(&r->h, rsize, RTA_GATEWAY, gw);
      break;
    case RTD_DEVICE:
-      r.r.rtm_type = RTN_UNICAST;
-      nl_add_attr_u32(&r.h, sizeof(r), RTA_OIF, iface->index);
+      r->r.rtm_type = RTN_UNICAST;
+      nl_add_attr_u32(&r->h, rsize, RTA_OIF, iface->index);
      break;
    case RTD_BLACKHOLE:
-      r.r.rtm_type = RTN_BLACKHOLE;
+      r->r.rtm_type = RTN_BLACKHOLE;
      break;
    case RTD_UNREACHABLE:
-      r.r.rtm_type = RTN_UNREACHABLE;
+      r->r.rtm_type = RTN_UNREACHABLE;
      break;
    case RTD_PROHIBIT:
-      r.r.rtm_type = RTN_PROHIBIT;
+      r->r.rtm_type = RTN_PROHIBIT;
      break;
    case RTD_MULTIPATH:
-      r.r.rtm_type = RTN_UNICAST;
-      nl_add_multipath(&r.h, sizeof(r), a->nexthops);
+      r->r.rtm_type = RTN_UNICAST;
+      nl_add_multipath(&r->h, rsize, a->nexthops);
      break;
    case RTD_NONE:
      break;
@ -1043,7 +1046,7 @@ dest:
    }

  /* Ignore missing for DELETE */
-  return nl_exchange(&r.h, (op == NL_OP_DELETE));
+  return nl_exchange(&r->h, (op == NL_OP_DELETE));
 }

 static inline int
@ -1750,4 +1753,4 @@ kif_sys_start(struct kif_proto *p UNUSED)
 void
 kif_sys_shutdown(struct kif_proto *p UNUSED)
 {
-}
+}
--- a/sysdep/unix/config.Y
+++ b/sysdep/unix/config.Y
@ -42,9 +42,9 @@ syslog_name:

 log_file:
   text {
-     FILE *f = tracked_fopen(new_config->pool, $1, "a");
+     struct rfile *f = rf_open(new_config->pool, $1, "a");
     if (!f) cf_error("Unable to open log file `%s': %m", $1);
-     $$ = f;
+     $$ = rf_file(f);
   }
 | SYSLOG syslog_name { $$ = NULL; new_config->syslog_name = $2; }
 | STDERR { $$ = stderr; }
@ -78,9 +78,9 @@ CF_ADDTO(conf, mrtdump_base)
 mrtdump_base:
   MRTDUMP PROTOCOLS mrtdump_mask ';' { new_config->proto_default_mrtdump = $3; }
 | MRTDUMP text ';' {
-     FILE *f = tracked_fopen(new_config->pool, $2, "a");
+     struct rfile *f = rf_open(new_config->pool, $2, "a");
     if (!f) cf_error("Unable to open MRTDump file '%s': %m", $2);
-     new_config->mrtdump_file = fileno(f);
+     new_config->mrtdump_file = rf_fileno(f);
   }
 ;

--- a/sysdep/unix/io.c
+++ b/sysdep/unix/io.c
@ -54,6 +54,7 @@
   this to gen small latencies */
 #define MAX_RX_STEPS 4

+
 /*
 *	Tracked Files
 */
@ -88,19 +89,32 @@ static struct resclass rf_class = {
  NULL
 };

-void *
-tracked_fopen(pool *p, char *name, char *mode)
+struct rfile *
+rf_open(pool *p, char *name, char *mode)
 {
  FILE *f = fopen(name, mode);

-  if (f)
-    {
-      struct rfile *r = ralloc(p, &rf_class);
-      r->f = f;
-    }
-  return f;
+  if (!f)
+    return NULL;
+
+  struct rfile *r = ralloc(p, &rf_class);
+  r->f = f;
+  return r;
 }

+void *
+rf_file(struct rfile *f)
+{
+  return f->f;
+}
+
+int
+rf_fileno(struct rfile *f)
+{
+  return fileno(f->f);
+}
+
+
 /**
 * DOC: Timers
 *
@ -478,6 +492,20 @@ tm_format_datetime(char *x, struct timeformat *fmt_spec, bird_clock_t t)
    strcpy(x, "<too-long>");
 }

+int
+tm_format_real_time(char *x, size_t max, const char *fmt, bird_clock_t t)
+{
+  struct tm tm;
+
+  if (!localtime_r(&t, &tm))
+    return 0;
+
+  if (!strftime(x, max, fmt, &tm))
+    return 0;
+
+  return 1;
+}
+

 /**
 * DOC: Sockets
@ -1574,6 +1602,7 @@ sk_sendmsg(sock *s)
  struct iovec iov = {s->tbuf, s->tpos - s->tbuf};
  byte cmsg_buf[CMSG_TX_SPACE];
  sockaddr dst;
+  int flags = 0;

  sockaddr_fill(&dst, s->af, s->daddr, s->iface, s->dport);

@ -1584,6 +1613,13 @@ sk_sendmsg(sock *s)
    .msg_iovlen = 1
  };

+#ifdef CONFIG_DONTROUTE_UNICAST
+  /* FreeBSD silently changes TTL to 1 when MSG_DONTROUTE is used, therefore we
+     cannot use it for other cases (e.g. when TTL security is used). */
+  if (ipa_is_ip4(s->daddr) && ip4_is_unicast(ipa_to_ip4(s->daddr)) && (s->ttl == 1))
+    flags = MSG_DONTROUTE;
+#endif
+
 #ifdef CONFIG_USE_HDRINCL
  byte hdr[20];
  struct iovec iov2[2] = { {hdr, 20}, iov };
@ -1599,7 +1635,7 @@ sk_sendmsg(sock *s)
  if (s->flags & SKF_PKTINFO)
    sk_prepare_cmsgs(s, &msg, cmsg_buf, sizeof(cmsg_buf));

-  return sendmsg(s->fd, &msg, 0);
+  return sendmsg(s->fd, &msg, flags);
 }

 static inline int
--- a/sysdep/unix/log.c
+++ b/sysdep/unix/log.c
@ -24,7 +24,7 @@

 #include "nest/bird.h"
 #include "nest/cli.h"
-#include "nest/mrtdump.h"
+#include "conf/conf.h"
 #include "lib/string.h"
 #include "lib/lists.h"
 #include "lib/unix.h"
@ -242,14 +242,23 @@ die(const char *msg, ...)
 void
 debug(const char *msg, ...)
 {
+#define MAX_DEBUG_BUFSIZE       65536
  va_list args;
-  char buf[1024];
+  static uint bufsize = 4096;
+  static char *buf = NULL;
+
+  if (!buf)
+    buf = mb_alloc(&root_pool, bufsize);

  va_start(args, msg);
  if (dbgf)
    {
-      if (bvsnprintf(buf, sizeof(buf), msg, args) < 0)
-	bsprintf(buf + sizeof(buf) - 100, " ... <too long>\n");
+      while (bvsnprintf(buf, bufsize, msg, args) < 0)
+        if (bufsize >= MAX_DEBUG_BUFSIZE)
+          bug("Extremely long debug output, split it.");
+        else
+          buf = mb_realloc(buf, (bufsize *= 2));
+
      fputs(buf, dbgf);
    }
  va_end(args);
@ -285,12 +294,14 @@ log_switch(int debug, list *l, char *new_syslog_name)
  if (!l || EMPTY_LIST(*l))
    l = default_log_list(debug, !l, &new_syslog_name);

+  log_lock();
+
  current_log_list = l;

 #ifdef HAVE_SYSLOG_H
  if (current_syslog_name && new_syslog_name &&
      !strcmp(current_syslog_name, new_syslog_name))
-    return;
+    goto done;

  if (current_syslog_name)
  {
@ -305,6 +316,9 @@ log_switch(int debug, list *l, char *new_syslog_name)
    openlog(current_syslog_name, LOG_CONS | LOG_NDELAY, LOG_DAEMON);
  }
 #endif
+
+done:
+  log_unlock();
 }


@ -327,16 +341,3 @@ log_init_debug(char *f)
  if (dbgf)
    setvbuf(dbgf, NULL, _IONBF, 0);
 }
-
-void
-mrt_dump_message(struct proto *p, u16 type, u16 subtype, byte *buf, u32 len)
-{
-  /* Prepare header */
-  put_u32(buf+0, now_real);
-  put_u16(buf+4, type);
-  put_u16(buf+6, subtype);
-  put_u32(buf+8, len - MRTDUMP_HDR_LENGTH);
-
-  if (p->cf->global->mrtdump_file != -1)
-    write(p->cf->global->mrtdump_file, buf, len);
-}
--- a/sysdep/unix/main.c
+++ b/sysdep/unix/main.c
@ -212,7 +212,7 @@ read_config(void)
  if (!unix_read_config(&conf, config_name))
    {
      if (conf->err_msg)
-	die("%s, line %d: %s", conf->err_file_name, conf->err_lino, conf->err_msg);
+	die("%s:%d:%d %s", conf->err_file_name, conf->err_lino, conf->err_chno, conf->err_msg);
      else
 	die("Unable to open configuration file %s: %m", config_name);
    }
@ -229,7 +229,7 @@ async_config(void)
  if (!unix_read_config(&conf, config_name))
    {
      if (conf->err_msg)
-	log(L_ERR "%s, line %d: %s", conf->err_file_name, conf->err_lino, conf->err_msg);
+	log(L_ERR "%s:%d:%d %s", conf->err_file_name, conf->err_lino, conf->err_chno, conf->err_msg);
      else
 	log(L_ERR "Unable to open configuration file %s: %m", config_name);
      config_free(conf);
@ -250,7 +250,7 @@ cmd_read_config(char *name)
  if (!unix_read_config(&conf, name))
    {
      if (conf->err_msg)
-	cli_msg(8002, "%s, line %d: %s", conf->err_file_name, conf->err_lino, conf->err_msg);
+	cli_msg(8002, "%s:%d:%d %s", conf->err_file_name, conf->err_lino, conf->err_chno, conf->err_msg);
      else
 	cli_msg(8002, "%s: %m", name);
      config_free(conf);
--- a/sysdep/unix/timer.h
+++ b/sysdep/unix/timer.h
@ -74,8 +74,9 @@ bird_clock_t tm_parse_date(char *);	/* Convert date to bird_clock_t */
 bird_clock_t tm_parse_datetime(char *);	/* Convert date to bird_clock_t */

 #define TM_DATETIME_BUFFER_SIZE 32	/* Buffer size required by tm_format_datetime */
-void
-tm_format_datetime(char *x, struct timeformat *fmt_spec, bird_clock_t t);
+
+void tm_format_datetime(char *x, struct timeformat *fmt_spec, bird_clock_t t);
+int tm_format_real_time(char *x, size_t max, const char *fmt, bird_clock_t t);

 #define TIME_T_IS_64BIT (sizeof(time_t) == 8)
 #define TIME_T_IS_SIGNED ((time_t) -1 < 0)
--- a/sysdep/unix/unix.h
+++ b/sysdep/unix/unix.h
@ -14,6 +14,7 @@
 struct pool;
 struct iface;
 struct birdsock;
+struct rfile;

 /* main.c */

@ -95,15 +96,17 @@ int sockaddr_read(sockaddr *sa, int af, ip_addr *a, struct iface **ifa, uint *po
 #define SUN_LEN(ptr) ((size_t) (((struct sockaddr_un *) 0)->sun_path) + strlen ((ptr)->sun_path))
 #endif

-volatile int async_config_flag;
-volatile int async_dump_flag;
-volatile int async_shutdown_flag;
+extern volatile int async_config_flag;
+extern volatile int async_dump_flag;
+extern volatile int async_shutdown_flag;

 void io_init(void);
 void io_loop(void);
 void io_log_dump(void);
 int sk_open_unix(struct birdsock *s, char *name);
-void *tracked_fopen(struct pool *, char *name, char *mode);
+struct rfile *rf_open(struct pool *, char *name, char *mode);
+void *rf_file(struct rfile *f);
+int rf_fileno(struct rfile *f);
 void test_old_bird(char *path);


--- a/tools/Makefile.in
+++ b/tools/Makefile.in
@ -71,7 +71,7 @@ tags:
 	cd $(srcdir) ; etags -lc `find $(static-dirs) $(addprefix $(objdir)/,$(dynamic-dirs)) $(client-dirs) -name *.[chY]`

 install: all
-	$(INSTALL) -d $(DESTDIR)/$(sbindir) $(DESTDIR)/$(sysconfdir) $(DESTDIR)/@runtimedir@
+	$(INSTALL) -d $(DESTDIR)/$(sbindir) $(DESTDIR)/$(sysconfdir) $(DESTDIR)/$(runstatedir)
 	$(INSTALL_PROGRAM) $(exedir)/bird $(DESTDIR)/$(sbindir)/bird@SUFFIX@
 	$(INSTALL_PROGRAM) $(exedir)/birdcl $(DESTDIR)/$(sbindir)/birdcl@SUFFIX@
 	if test -n "@CLIENT@" ; then								\
--- a/tools/Rules.in
+++ b/tools/Rules.in
@ -39,6 +39,7 @@ bindir=@bindir@
 sbindir=@sbindir@
 sysconfdir=@sysconfdir@
 localstatedir=@localstatedir@
+runstatedir=@runstatedir@
 docdir=@prefix@/doc

 ifdef source
--- a/tools/config.guess
+++ b/tools/config.guess
--- a/tools/config.sub
+++ b/tools/config.sub
--- a/tools/gendist
+++ b/tools/gendist
@ -5,7 +5,7 @@
 #

 set -e
-AC=`if [ -x /usr/bin/autoconf2.50 ] ; then echo autoconf2.50 ; else echo autoconf ; fi`
+AC=autoreconf
 $AC
 ./configure
 make distclean
@ -22,7 +22,7 @@ mkdir -p $T/$REL $T/$DREL $T/$DREL/doc
 cp -a . $T/$REL
 echo Generating ChangeLog
 git log  >$T/$REL/ChangeLog
-mv $T/$REL/doc/*.ps $T/$DREL/doc
+mv $T/$REL/doc/*.pdf $T/$DREL/doc
 rm -f $T/$REL/bird.conf*
 rm -rf $T/$REL/.git/
 rm -rf `find $T/$REL -name CVS -o -name tmp` $T/$REL/{misc,rfc,doc/slides}
--- a/tools/progdoc
+++ b/tools/progdoc
@ -3,9 +3,7 @@
 $srcdir = $ARGV[0];

 open(OUT, ">prog.sgml") || die "Cannot create output file";
-include("doc/prog-head.sgml");
-process("");
-include("doc/prog-foot.sgml");
+process("", "doc/prog-root");
 close OUT;
 exit 0;

@ -20,8 +18,9 @@ sub include {

 sub process {
  my $dir = shift @_;
-  print "$dir/Doc\n";
-  open(IN, "$srcdir/$dir/Doc") || die "Unable to read $dir/Doc";
+  my $doc = "$dir/" . shift @_;
+  print "$doc\n";
+  open(IN, "$srcdir/$doc") || die "Unable to read $doc";
  my @docfile = <IN>;
  close IN;
  foreach $_ (@docfile) {
@ -30,7 +29,7 @@ sub process {
    /^([A-Z]+)\s*(.*)/ || die "Parse error: $_";
    $cmd = $1;
    $arg = $2;
-    if ($cmd eq "C") { process("$dir/$arg"); }
+    if ($cmd eq "C") { process("$dir/$arg", "Doc"); }
    elsif ($cmd eq "H") {
      push @stack, "H";
      print OUT "<chapt>$arg\n";
Author	SHA1	Message	Date
Ondrej Zajicek (work)	a6622bd204	NEWS and version update	2019-08-01 03:31:23 +02:00
Ondrej Zajicek (work)	aba51d80c0	Doc: Update documentation about VRFs and BFD	2019-08-01 01:02:31 +02:00
Ondrej Zajicek (work)	da8644d7d9	Nest: VRF of protocol can be explicitly specified as 'default' Protocol can have specified VRF, in such case it is restricted to a set of ifaces associated with the VRF, otherwise it can use all interfaces. The patch allows to specify VRF as 'default', in which case it is restricted to a set of iface not associated with any VRF.	2019-08-01 00:53:22 +02:00
Maria Matejka	d72d3891bf	Log: Fixed race condition in reconfigure while BFD is running	2019-07-30 12:13:38 +02:00
Maria Matejka	46faedff29	Debug: growing message format buffer This led in corner cases to undefined buffer content and garbage output.	2019-07-18 12:41:45 +02:00
Ondrej Zajicek (work)	9d23aa7a80	Lib: Fix print of 64-bit router id Mismatched types to printf(). The old code coincidentally worked on amd64 due to its calling conventions. Thanks to Maximilian Eschenbacher for the bugreport.	2019-07-18 02:05:36 +02:00
Ondrej Zajicek (work)	f9deedf1f0	BFD: Support for VRFs Allow multiple BFD instances in separate VRFs, dispatch BFD requests according to VRFs. Thanks to Alexander Zubkov for notice and patches.	2019-07-18 02:04:42 +02:00
Alexander Azimov	9ff6c8d83c	BGP: Compliance with RFC8203bis	2019-07-11 15:44:52 +02:00
Ondrej Zajicek (work)	78c05cc159	OSPF: Update DR when local priority changes When priority is reconfigured locally, we need to trigger DR election.	2019-07-11 15:44:05 +02:00
Ondrej Zajicek (work)	a92aee467e	Avoid fallthrough warnings	2019-07-11 15:44:05 +02:00
Ondrej Zajicek (work)	6b72ea4c14	OSPF: N-bit should not be set for DBDES packets	2019-07-11 15:44:05 +02:00
Maria Jan Matejka	354afcab04	BGP: Prefix hash is too small, increase its max size. This doesn't make any change for you until you have millions of updates waiting to be sent. Increasing the max hash size from 2^20 to 2^24.	2019-06-11 13:19:21 +00:00
Maria Jan Matejka	e989b901fc	BGP: split tx explicitly If BGP has too many data to send and BIRD is slower than the link, TX is always possible until all data is sent. This patch limits maximum number of generated BGP messages in one iteration of TX hook.	2019-06-11 09:39:00 +00:00
Ondrej Filip	5648f07e4d	Small typo in doc.	2019-04-24 13:50:54 +02:00
Ondrej Zajicek (work)	b5d1903bf6	NEWS and version update	2019-02-26 18:10:04 +01:00
Ondrej Zajicek (work)	2e7ee1c9d3	Nest: Do not compare rte.flags during rte_update() Route flags are mosty internal state of rtable, they are not significant to whether a route has changed. With the old code, all routes received as a part of enhanced route refresh are always re-announced to other peers due to change in REF_STALE.	2019-02-22 02:33:01 +01:00
Ondrej Zajicek (work)	797969983d	Doc: Detect SP/OpenSP automatically	2019-02-19 18:32:45 +01:00
Ondrej Zajicek (work)	b3fceeba30	Nest: Prevent withdraws from propagation back to source protocol (for accepted mode) Update for one of previous patches, handles the the issue for first-accepted mode of route propagation.	2019-02-19 16:21:52 +01:00
Ondrej Zajicek (work)	2dd9800ab5	Nest: Improve export counter handling One of previous workarounds for phantom route avoidance breaks export counters by expanding sending of spurious withdraws, which are send when we are not sure whether we have advertised that routes in the past. If not, then export counter is decreased, but it was not increased before, so it overflows under zero. The patch fixes that by sending spurious withdraws, but not counting them on export counter. That may lead to error in the other direction, but that happens only as a race condition (i.e., in normal operation filters return proper values about old route export state).	2019-02-19 16:02:53 +01:00
Ondrej Zajicek (work)	b4438e40ef	Nest: Prevent withdraws from propagation back to source protocol The earlier fix loosen conditions for not running filters on old route when deciding about route propagation to a protocol to avoid issues with ghost routes in some race conditions. Unfortunately, the fix also caused back-propagation of withdraws. For regular updates, back-propagation is prevented in import_control hooks, but these are not called on withdraws. For them, import_control hooks are called on old routes instead, changing (old, NULL) notification to (NULL, NULL), which is ignored. By not calling export processing in some cases, the withdraw is not ignored and is back-propagated. This patch fixes that by contract conditions so the earlier fix is not applied to back-propagated updates.	2019-01-30 17:08:11 +01:00
Ondrej Zajicek (work)	ccb37330d0	Doc: Add documentation for OSPF retransmit delay option Thanks to Igor Podlesny for notification.	2019-01-26 21:03:36 +01:00
Ondrej Filip	e99e7d1c2d	Added documentation for 'disable after cease'	2019-01-07 12:26:21 +01:00
Ondrej Zajicek (work)	ef8974b7ca	NEWS and version update	2019-01-05 00:44:26 +01:00
Ondrej Zajicek (work)	a93684b2d4	Configure: Fix lost suffix in bird.ctl One of previous commits bacported from 2.0 removed suffix from bird.ctl	2019-01-04 19:11:54 +01:00
Ondrej Zajicek (work)	c30f00d4a8	BSD: Fix TCP-MD5 code on current FreeBSD kernels Current FreeBSD kernels require SA records for both directions. Thanks to Joseph Mulloy and Andrey V. Elsukov for reporting and solving the issue.	2019-01-04 18:20:56 +01:00
Ondrej Zajicek (work)	968c31ec6d	Doc: README and INSTALL update Minor cleanups, updates and clarifications. Also removes (incomplete and well-known) build steps from README, as they are better described in INSTALL.	2019-01-03 17:26:57 +01:00
Ondrej Zajicek (work)	89bc266fe7	Doc: Fix documentation build with newer Linuxdoc-Tools	2019-01-03 17:17:08 +01:00
Robert Scheck	cdf2daae01	Doc: Allow overriding $SGML_CATALOG_FILES using distribution specific paths	2018-12-20 13:25:53 +01:00
Robert Scheck	7a53383a22	Doc: Add alternative path for SGML ISO entities 8879.1986 to $SGML_CATALOG_FILES The existing paths are valid for Debian, alternative paths are necessary for Fedora and RHEL/CentOS.	2018-12-20 13:25:31 +01:00
Ondrej Zajicek (work)	b47d33b344	Autoconf: Minor cleanup	2018-12-20 13:17:30 +01:00
Ondrej Zajicek (work)	16fb28e588	Configure: Use standard --runstatedir option Newer Autoconf defines --runstatedir option for setting directory for run-time variable data. Use it instead our old --with-runtimedir.	2018-12-20 13:15:53 +01:00
Ondrej Zajicek (work)	e7e3b335bf	Nest: Fix crash in rta_show() for Babel routes Some new route source values did not have associated string in rta_show(), which might caused crash in some cases.	2018-12-20 13:06:18 +01:00
Ondrej Zajicek (work)	4df42f1a55	Lib: Force output type in ip4_addr constructor Fixes type issue when u64 is pushed into it.	2018-12-20 13:01:15 +01:00
Ondrej Zajicek (work)	17450df314	Filter: Fix minor bug in accessing bgp_path Not relevant for regular BGP paths, just for BGP paths added by filters to e.g. static routes.	2018-12-20 12:59:55 +01:00
Ondrej Zajicek (work)	d82b1a1977	Nest: Fix 'show interfaces summary' command The command showed interfaces that were removed / in shutdown.	2018-12-20 12:57:43 +01:00
Ondrej Zajicek (work)	257c7ce95d	Doc: Fix description of 'description' Thanks to Clemens Schrimpe for the bugreport.	2018-12-20 12:54:52 +01:00
Ondrej Zajicek (work)	04edf8f2f2	IO: Workaround for broken FreeBSD behavior FreeBSD silently changes TTL to 1 when MSG_DONTROUTE is used, even when it is explicitly set to another value. That breaks TTL security sockets, including BFD which always uses TTL 255. Bad FreeBSD!	2018-12-19 18:10:39 +01:00
Ondrej Zajicek (work)	c0ed32ee28	Doc: Generate MRT progdoc	2018-12-19 16:32:52 +01:00
Ondrej Zajicek (work)	523c5d9f2b	Doc: Rename code documentation files back to Doc	2018-12-19 16:29:20 +01:00
Ondrej Zajicek (work)	d1c8f22bec	Doc: Move root of code documentation to doc dir It reduces clutter in root and also avoid collision with doc dir on case-insensitive filesystems when name back to Doc.	2018-12-19 16:25:59 +01:00
Ondrej Zajicek (work)	81bc013dde	Unix: Fix variable declarations	2018-12-19 15:52:13 +01:00
Ondrej Zajicek (work)	98b1486c48	OSPF: Fix wrong LSA collisions detection In some circumstances (old LSA flushed but not acknowledged and not removed) origination of a new LSA may wrongly triggers LSA collision code. The patch fixes that. Thanks to Asbjorn Mikkelsen for the bugreport and @mdelagueronniere for the original patch.	2018-12-17 17:10:59 +01:00
Ondrej Zajicek (work)	8c0b12ac86	OSPF: Fix reconfiguration of vlinks Fix crash during reconfiguration of OSPF config with vlinks. When vlink is reconfigured, a generic iface-reconfiguration code is used, which in one place supposes that it is running on a regular iface. Thanks to Cybertinus for a bugreport.	2018-12-10 02:08:52 +01:00
Ondrej Zajicek (work)	ea59172cde	Static: Fix reset neighbor entries for ECMP routes Neighbor entries for static ECMP routes were not cleaned up during reconfigure and pointed to the old instances, which leads to crash after reconfigure. Thanks to Vladimir Osmolovskiy for the bugreport.	2018-12-04 13:52:58 +01:00
Ondrej Zajicek (work)	93cec70361	MRT documentation	2018-11-21 15:59:53 +01:00
Ondrej Zajicek (work)	1453ab7552	MRT: Fix bug in MRT table dump requested from CLI	2018-11-20 17:40:47 +01:00
Ondrej Zajicek (work)	81489b79e0	Nest: Improve keeping track of IPv6 link-local addresses Most protocols in IPv6 mode use link-local source addresses and expect that there is one on each active interface. The old code depended on assumption that if there is some IPv6 address on iface, there is also an IPv6 link-local address on that iface (added by kernel when the iface went up). Unfortunately, that is not generally true, as a configured global address sometimes ceases to be tentative (finishes DOD) before a link-local address on the same iface. In such case a protocol iface (namely RAdv and Babel) is activated, but fails to found link-local address and stays in failed state. The patch fixes that by tracking 'primary' IPv6 link-local address, sending iface restart notifications when it changes and making protocols ignore iface-up notifications when no such address is selected for an iface.	2018-11-08 20:43:04 +01:00
Ondrej Zajicek (work)	716b904f4e	Filter: Make ifname attribute modifiable Allow to change an interface associated with a route by setting ifname attribute. It will also change the route to a direct one.	2018-11-05 21:55:18 +01:00
Ondrej Zajicek (work)	fff79b1c1e	OSPF: Fix some trace messages Missing argument in MTU change trace message can crash bird when MTU change happens and trace messages are active. Thanks to Alexander Velkov for the bugreport.	2018-10-25 17:21:06 +02:00
Ondrej Zajicek (work)	c0fc3e6718	The MRT protocol The new MRT protocol is responsible for periodic RIB table dumps in the MRT format (RFC 6396). Also the existing code for BGP4MP MRT dumps is refactored and splitted between BGP to MRT protocols, will be more integrated into MRT in the future. Example: protocol mrt { table "*"; filename "%N_%F_%T.mrt"; period 60; } It is partially based on the old MRT code from Pavel Tvrdik.	2018-09-18 17:50:45 +02:00
Jan Maria Matejka	afa14f1868	Conf: Show the line:char position where the syntax error happens	2018-09-11 16:58:09 +02:00
Ondrej Zajicek (work)	b1f6c439f5	BGP: Fix BFD/GR interaction The old code does not cause GR if BFD failed when GR is running, session is re-established but not yet synchronized.	2018-08-01 17:41:29 +02:00
Ondrej Zajicek (work)	4abccabd7b	BGP: Minor fixes in long-lived graceful restart	2018-08-01 17:35:39 +02:00
Ondrej Zajicek (work)	6be71641ef	BSD: Use MSG_DONTROUTE for unicast packets on FreeBSD BSD systems cannot use SO_DONTROUTE, because it does not work properly with multicast packets (perhaps it tries to find iface based on multicast group address). But we can use MSG_DONTROUTE sendmsg() flag for unicast packets. Works on FreeBSD, is ignored on OpenBSD and is broken on NetBSD (i guess due to integrated routing table and ARP table).	2018-07-28 19:30:56 +02:00
Ondrej Zajicek (work)	3140c8b2ca	Move FreeBSD CFLAGS and LDFLAGS to configure Backport of commit 09c1e370b3084f7acb7c3777427670a69945368a.	2018-07-28 15:52:19 +02:00
Ondrej Zajicek (work)	aa2ec912f5	BGP: Long-lived graceful restart enhancements When use of LLGR is negotiated, handle hold timeout by LLGR instead of by hard restart. Allow to configure whether BFD session down event should be handled by GR/LLGR or by hard restart.	2018-07-17 13:16:35 +02:00
Ondrej Zajicek (work)	a9d869c484	Doc: Documentation for long-lived graceful restart	2018-07-17 13:16:35 +02:00
Ondrej Zajicek (work)	68197386dd	BGP: Long-lived graceful restart The patch implements long-lived graceful restart for BGP, namely draft-uttaro-idr-bgp-persistence-03.	2018-07-17 13:16:18 +02:00
Ondrej Zajicek (work)	470efcb98c	Nest: Fix race condition during reconfiguration, part 2 If export filter is changed during reconfiguration and a route disappears between reconfiguration and refeed (e.g., if the route is a static route also removed during the reconfiguration), the route is not withdrawn. The issue was fixed for regular channels by an earlier patch. This patch fixes the issue for channels in RA_ACCEPTED mode (first-pass-the-filter), used by BGP with 'secondary' option.	2018-07-06 02:17:59 +02:00
Ondrej Zajicek (work)	cbfdf6ed05	Nest: Fix race condition during reconfiguration If export filter is changed during reconfiguration and a route disappears between reconfiguration and refeed (e.g., if the route is a static route also removed during the reconfiguration), the route is not withdrawn. The patch fixes that by adding tx reconfiguration timestamp.	2018-07-03 19:21:59 +02:00
Maria Matejka	822a7ee6d5	Autotools: updated config.guess and config.sub Updated to version 63b4ce2e8c28aee6a32133e400436e4ca885215a from git://git.savannah.gnu.org/config.git Previous version was 93b5037172b15ad28952481933517f1ba93d125b	2018-04-24 14:43:42 +02:00
Maria Matejka	0671e6c29a	Netlink: removed dynamic length array as Clang won't support it.	2018-04-24 12:18:32 +02:00
Maria Matejka	1e921ec868	Android: check for extra libs needed for build	2018-04-24 12:18:32 +02:00
Maria Matejka	7904f409e2	Doc: renamed progdoc files Doc -> progdoc to fix collision with doc/ folder on case-insensitive filesystems	2018-04-23 11:29:13 +02:00
Ondrej Zajicek (work)	4c0b741ba7	Doc: Remove some superfluous slashes	2018-03-24 01:53:03 +01:00
Ondrej Zajicek (work)	f99c61b1f4	Doc: Be consistent in generating pdf output	2018-03-24 01:34:58 +01:00
Ondrej Filip	2ee6a89a8d	Autoconf replaced by autoreconf	2018-03-23 17:22:55 +01:00
Ondrej Filip	f77fb3f00e	Date added.	2018-03-22 13:33:12 +01:00
Ondrej Zajicek (work)	a71a75213d	NEWS and version update	2018-03-21 16:26:39 +01:00
Ondrej Zajicek (work)	4dff1f32b5	Doc: Minor build fix	2018-03-21 16:26:39 +01:00
Ondrej Zajicek (work)	97b1b04d93	C99 loop declarations not allowed in 1.6.x	2018-03-19 19:51:11 +01:00