#define BGP_SAFI_UNICAST 1
#define BGP_SAFI_MULTICAST 2
+#define BGP_SAFI_MPLS 4
+#define BGP_SAFI_MPLS_VPN 128
+#define BGP_SAFI_VPN_MULTICAST 129
+#define BGP_SAFI_FLOW 133
/* Internal AF codes */
#define BGP_AF_IPV6 BGP_AF( BGP_AFI_IPV6, BGP_SAFI_UNICAST )
#define BGP_AF_IPV4_MC BGP_AF( BGP_AFI_IPV4, BGP_SAFI_MULTICAST )
#define BGP_AF_IPV6_MC BGP_AF( BGP_AFI_IPV6, BGP_SAFI_MULTICAST )
+#define BGP_AF_IPV4_MPLS BGP_AF( BGP_AFI_IPV4, BGP_SAFI_MPLS )
+#define BGP_AF_IPV6_MPLS BGP_AF( BGP_AFI_IPV6, BGP_SAFI_MPLS )
+#define BGP_AF_VPN4_MPLS BGP_AF( BGP_AFI_IPV4, BGP_SAFI_MPLS_VPN )
+#define BGP_AF_VPN6_MPLS BGP_AF( BGP_AFI_IPV6, BGP_SAFI_MPLS_VPN )
+#define BGP_AF_VPN4_MC BGP_AF( BGP_AFI_IPV4, BGP_SAFI_VPN_MULTICAST )
+#define BGP_AF_VPN6_MC BGP_AF( BGP_AFI_IPV6, BGP_SAFI_VPN_MULTICAST )
+#define BGP_AF_FLOW4 BGP_AF( BGP_AFI_IPV4, BGP_SAFI_FLOW )
+#define BGP_AF_FLOW6 BGP_AF( BGP_AFI_IPV6, BGP_SAFI_FLOW )
struct bgp_write_state;
struct bgp_af_desc {
u32 afi;
u32 net;
+ u8 mpls;
+ u8 no_igp;
const char *name;
uint (*encode_nlri)(struct bgp_write_state *s, struct bgp_bucket *buck, byte *buf, uint size);
void (*decode_nlri)(struct bgp_parse_state *s, byte *pos, uint len, rta *a);
u16 local_port; /* Local listening port */
u16 remote_port; /* Neighbor destination port */
int multihop; /* Number of hops if multihop */
- int strict_bind; /* Bind listening socket to local address XXXX */
+ int strict_bind; /* Bind listening socket to local address */
int ttl_security; /* Enable TTL security [RFC 5082] */
int compare_path_lengths; /* Use path lengths when selecting best route */
int med_metric; /* Compare MULTI_EXIT_DISC even between routes from differen ASes */
int deterministic_med; /* Use more complicated algo to have strict RFC 4271 MED comparison */
u32 default_local_pref; /* Default value for LOCAL_PREF attribute */
u32 default_med; /* Default value for MULTI_EXIT_DISC attribute */
- int capabilities; /* Enable capability handshake [RFC 3392] */
+ int capabilities; /* Enable capability handshake [RFC 5492] */
int enable_refresh; /* Enable local support for route refresh [RFC 2918] */
- int enable_as4; /* Enable local support for 4B AS numbers [RFC 4893] */
+ int enable_as4; /* Enable local support for 4B AS numbers [RFC 6793] */
int enable_extended_messages; /* Enable local support for extended messages [draft] */
u32 rr_cluster_id; /* Route reflector cluster ID, if different from local ID */
int rr_client; /* Whether neighbor is RR client of me */
int passive; /* Do not initiate outgoing connection */
int interpret_communities; /* Hardwired handling of well-known communities */
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 */
+ /* Times below are in seconds */
unsigned gr_time; /* Graceful restart timeout */
+ unsigned llgr_time; /* Long-lived graceful restart stale time */
unsigned connect_delay_time; /* Minimum delay between connect attempts */
unsigned connect_retry_time; /* Timeout for connect attempts */
unsigned hold_time, initial_hold_time;
unsigned error_delay_time_min; /* Time to wait after an error is detected */
unsigned error_delay_time_max;
unsigned disable_after_error; /* Disable the protocol when error is detected */
+ u32 disable_after_cease; /* Disable it when cease is received, bitfield */
char *password; /* Password used for MD5 authentication */
int check_link; /* Use iface link state for liveness detection */
struct channel_config c;
u32 afi;
+ const struct bgp_af_desc *desc;
ip_addr next_hop_addr; /* Local address for NEXT_HOP attribute */
- u8 next_hop_self; /* Always set next hop to local IP address */
- u8 next_hop_keep; /* Do not touch next hop attribute */
+ u8 next_hop_self; /* Always set next hop to local IP address (NH_*) */
+ u8 next_hop_keep; /* Do not modify next hop attribute (NH_*) */
+ u8 mandatory; /* Channel is mandatory in capability negotiation */
u8 missing_lladdr; /* What we will do when we don' know link-local addr, see MLL_* */
u8 gw_mode; /* How we compute route gateway from next_hop attr, see GW_* */
u8 secondary; /* Accept also non-best routes (i.e. RA_ACCEPTED) */
u8 gr_able; /* Allow full graceful restart for the channel */
+ u8 llgr_able; /* Allow full long-lived GR for the channel */
+ uint llgr_time; /* Long-lived graceful restart stale time */
+ u8 ext_next_hop; /* Allow both IPv4 and IPv6 next hops */
u8 add_path; /* Use ADD-PATH extension [RFC 7911] */
+ u8 import_table; /* Use c.in_table as Adj-RIB-In */
- struct rtable_config *igp_table; /* Table used for recursive next hop lookups */
+ uint rest[0]; /* Remaining items are reconfigured separately */
+ struct rtable_config *igp_table_ip4; /* Table for recursive IPv4 next hop lookups */
+ struct rtable_config *igp_table_ip6; /* Table for recursive IPv6 next hop lookups */
};
-#define MLL_SELF 1
-#define MLL_DROP 2
-#define MLL_IGNORE 3
+#define NH_NO 0
+#define NH_ALL 1
+#define NH_IBGP 2
+#define NH_EBGP 3
+
+#define MLL_SELF 1
+#define MLL_DROP 2
+#define MLL_IGNORE 3
-#define GW_DIRECT 1
-#define GW_RECURSIVE 2
+#define GW_DIRECT 1
+#define GW_RECURSIVE 2
#define BGP_ADD_PATH_RX 1
#define BGP_ADD_PATH_TX 2
/* For GR capability per-AF flags */
#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 2 /* Long-lived GR phase (stale timer active) */
+
+#define BGP_BFD_GRACEFUL 2 /* BFD down triggers graceful restart */
+
struct bgp_af_caps {
u32 afi;
u8 ready; /* Multiprotocol capability, RFC 4760 */
u8 gr_able; /* Graceful restart support, RFC 4724 */
u8 gr_af_flags; /* Graceful restart per-AF flags */
+ u8 llgr_able; /* Long-lived GR, RFC draft */
+ u32 llgr_time; /* Long-lived GR stale time */
+ u8 llgr_flags; /* Long-lived GR per-AF flags */
+ u8 ext_next_hop; /* Extended IPv6 next hop, RFC 5549 */
u8 add_path; /* Multiple paths support, RFC 7911 */
};
u8 gr_flags; /* Graceful restart flags */
u16 gr_time; /* Graceful restart time in seconds */
+ u8 llgr_aware; /* Long-lived GR capability, RFC draft */
+
u16 af_count; /* Number of af_data items */
+ u16 length; /* Length of capabilities in OPEN msg */
struct bgp_af_caps af_data[0]; /* Per-AF capability data */
};
+#define WALK_AF_CAPS(caps,ac) \
+ for (ac = caps->af_data; ac < &caps->af_data[caps->af_count]; ac++)
+
+
struct bgp_socket {
node n; /* Node in global bgp_sockets */
sock *sk; /* Real listening socket */
struct bgp_caps *local_caps;
struct bgp_caps *remote_caps;
- struct timer *connect_timer;
- struct timer *hold_timer;
- struct timer *keepalive_timer;
- struct event *tx_ev;
+ timer *connect_timer;
+ timer *hold_timer;
+ timer *keepalive_timer;
+ event *tx_ev;
u32 packets_to_send; /* Bitmap of packet types to be sent */
u32 channels_to_send; /* Bitmap of channels with packets to be sent */
u8 last_channel; /* Channel used last time for TX */
u8 route_refresh; /* Route refresh allowed to send [RFC 2918] */
u8 enhanced_refresh; /* Enhanced refresh is negotiated [RFC 7313] */
u8 gr_ready; /* Neighbor could do graceful restart */
+ u8 llgr_ready; /* Neighbor could do Long-lived GR, implies gr_ready */
u8 gr_active_num; /* Neighbor is doing GR, number of active channels */
u8 channel_count; /* Number of active channels */
+ u8 summary_add_path_rx; /* Summary state of ADD_PATH RX w.r.t active channels */
u32 *afi_map; /* Map channel index -> AFI */
struct bgp_channel **channel_map; /* Map channel index -> channel */
struct bgp_conn *conn; /* Connection we have established */
struct bfd_request *bfd_req; /* BFD request, if BFD is used */
ip_addr source_addr; /* Local address used as an advertised next hop */
ip_addr link_addr; /* Link-local version of source_addr */
- struct event *event; /* Event for respawning and shutting process */
- struct timer *startup_timer; /* Timer used to delay protocol startup due to previous errors (startup_delay) */
- struct timer *gr_timer; /* Timer waiting for reestablishment after graceful restart */
- unsigned startup_delay; /* Time to delay protocol startup by due to errors */
- bird_clock_t last_proto_error; /* Time of last error that leads to protocol stop */
+ event *event; /* Event for respawning and shutting process */
+ timer *startup_timer; /* Timer used to delay protocol startup due to previous errors (startup_delay) */
+ timer *gr_timer; /* Timer waiting for reestablishment after graceful restart */
+ uint startup_delay; /* Delay (in seconds) of protocol startup due to previous errors */
+ btime last_proto_error; /* Time of last error that leads to protocol stop */
u8 last_error_class; /* Error class of last error */
u32 last_error_code; /* Error code of last error. BGP protocol errors
are encoded as (bgp_err_code << 16 | bgp_err_subcode) */
/* Rest are BGP specific data */
struct bgp_channel_config *cf;
- pool *pool; /* XXXX */
u32 afi;
u32 index;
const struct bgp_af_desc *desc;
+ rtable *igp_table_ip4; /* Table for recursive IPv4 next hop lookups */
+ rtable *igp_table_ip6; /* Table for recursive IPv6 next hop lookups */
+
+ /* Rest are zeroed when down */
+ pool *pool;
HASH(struct bgp_bucket) bucket_hash; /* Hash table of route buckets */
struct bgp_bucket *withdraw_bucket; /* Withdrawn routes */
list bucket_queue; /* Queue of buckets to send (struct bgp_bucket) */
HASH(struct bgp_prefix) prefix_hash; /* Prefixes to be sent */
slab *prefix_slab; /* Slab holding prefix nodes */
- rtable *igp_table; /* Table used for recursive next hop lookups */
ip_addr next_hop_addr; /* Local address for NEXT_HOP attribute */
ip_addr link_addr; /* Link-local version of next_hop_addr */
u32 packets_to_send; /* Bitmap of packet types to be sent */
+ u8 ext_next_hop; /* Session allows both IPv4 and IPv6 next hops */
+
u8 gr_ready; /* Neighbor could do GR on this AF */
- u8 gr_active; /* Neighbor is doing GR and keeping fwd state */
+ u8 gr_active; /* Neighbor is doing GR (BGP_GRS_*) */
+
+ timer *stale_timer; /* Long-lived stale timer for LLGR */
+ u32 stale_time; /* Stored LLGR stale time from last session */
u8 add_path_rx; /* Session expects receive of ADD-PATH extended NLRI */
u8 add_path_tx; /* Session expects transmit of ADD-PATH extended NLRI */
struct bgp_proto *src;
rte *route;
+ int mpls;
u32 attrs_seen[1];
uint err_withdraw;
struct bgp_channel *channel;
struct linpool *pool;
+ int mp_reach;
int as4_session;
int add_path;
+ int mpls;
eattr *mp_next_hop;
+ adata *mpls_labels;
};
struct bgp_parse_state {
int as4_session;
int add_path;
+ int mpls;
u32 attrs_seen[256/32];
uint err_subcode;
jmp_buf err_jmpbuf;
+ struct hostentry *hostentry;
+ adata *mpls_labels;
+
/* Cached state for bgp_rte_update() */
u32 last_id;
struct rte_src *last_src;
static inline int bgp_channel_is_ipv6(struct bgp_channel *c)
{ return BGP_AFI(c->afi) == BGP_AFI_IPV6; }
+static inline int bgp_cc_is_ipv4(struct bgp_channel_config *c)
+{ return BGP_AFI(c->afi) == BGP_AFI_IPV4; }
+
+static inline int bgp_cc_is_ipv6(struct bgp_channel_config *c)
+{ return BGP_AFI(c->afi) == BGP_AFI_IPV6; }
+
static inline uint bgp_max_packet_length(struct bgp_conn *conn)
{ return conn->ext_messages ? BGP_MAX_EXT_MSG_LENGTH : BGP_MAX_MESSAGE_LENGTH; }
}
extern struct linpool *bgp_linpool;
+extern struct linpool *bgp_linpool2;
-void bgp_start_timer(struct timer *t, int value);
+void bgp_start_timer(timer *t, uint value);
void bgp_check_config(struct bgp_config *c);
void bgp_error(struct bgp_conn *c, unsigned code, unsigned subcode, byte *data, int len);
void bgp_close_conn(struct bgp_conn *c);
void bgp_refresh_begin(struct bgp_channel *c);
void bgp_refresh_end(struct bgp_channel *c);
void bgp_store_error(struct bgp_proto *p, struct bgp_conn *c, u8 class, u32 code);
-void bgp_stop(struct bgp_proto *p, unsigned subcode);
+void bgp_stop(struct bgp_proto *p, uint subcode, byte *data, uint len);
struct rte_source *bgp_find_source(struct bgp_proto *p, u32 path_id);
struct rte_source *bgp_get_source(struct bgp_proto *p, u32 path_id);
static inline eattr *
bgp_find_attr(ea_list *attrs, uint code)
{
- return ea_find(attrs, EA_CODE(EAP_BGP, code));
+ return ea_find(attrs, EA_CODE(PROTOCOL_BGP, code));
}
eattr *
{ eattr *e = bgp_set_attr(to, pool, code, 0, 0); e->type = EAF_TYPE_UNDEF; }
-
-
-/* Hack: although BA_NEXT_HOP attribute has type EAF_TYPE_IP_ADDRESS, in IPv6
- * we store two addesses in it - a global address and a link local address.
- */
-#ifdef XXX
-#define NEXT_HOP_LENGTH (2*sizeof(ip_addr))
-static inline void set_next_hop(byte *b, ip_addr addr) { ((ip_addr *) b)[0] = addr; ((ip_addr *) b)[1] = IPA_NONE; }
-#define NEXT_HOP_LENGTH sizeof(ip_addr)
-static inline void set_next_hop(byte *b, ip_addr addr) { ((ip_addr *) b)[0] = addr; }
-#endif
-
int bgp_encode_attrs(struct bgp_write_state *s, ea_list *attrs, byte *buf, byte *end);
ea_list * bgp_decode_attrs(struct bgp_parse_state *s, byte *data, uint len);
void bgp_init_bucket_table(struct bgp_channel *c);
+void bgp_free_bucket_table(struct bgp_channel *c);
void bgp_free_bucket(struct bgp_channel *c, struct bgp_bucket *b);
void bgp_defer_bucket(struct bgp_channel *c, struct bgp_bucket *b);
void bgp_withdraw_bucket(struct bgp_channel *c, struct bgp_bucket *b);
void bgp_init_prefix_table(struct bgp_channel *c);
+void bgp_free_prefix_table(struct bgp_channel *c);
void bgp_free_prefix(struct bgp_channel *c, struct bgp_prefix *bp);
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);
-void bgp_rt_notify(struct proto *P, struct channel *C, net *n, rte *new, rte *old, ea_list *attrs);
-int bgp_import_control(struct proto *, struct rte **, struct ea_list **, struct linpool *);
+struct rte *bgp_rte_modify_stale(struct rte *r, struct linpool *pool);
+void bgp_rt_notify(struct proto *P, struct channel *C, net *n, rte *new, rte *old);
+int bgp_preexport(struct proto *, struct rte **, struct linpool *);
int bgp_get_attr(struct eattr *e, byte *buf, int buflen);
-void bgp_get_route_info(struct rte *, byte *buf, struct ea_list *attrs);
+void bgp_get_route_info(struct rte *, byte *buf);
/* 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);
const struct bgp_af_desc *bgp_get_af_desc(u32 afi);
const struct bgp_af_caps *bgp_find_af_caps(struct bgp_caps *caps, u32 afi);
void bgp_schedule_packet(struct bgp_conn *conn, struct bgp_channel *c, int type);
#define BAF_PARTIAL 0x20
#define BAF_EXT_LEN 0x10
-#define BA_ORIGIN 0x01 /* [RFC1771] */ /* WM */
+#define BA_ORIGIN 0x01 /* RFC 4271 */ /* WM */
#define BA_AS_PATH 0x02 /* WM */
#define BA_NEXT_HOP 0x03 /* WM */
#define BA_MULTI_EXIT_DISC 0x04 /* ON */
#define BA_LOCAL_PREF 0x05 /* WD */
#define BA_ATOMIC_AGGR 0x06 /* WD */
#define BA_AGGREGATOR 0x07 /* OT */
-#define BA_COMMUNITY 0x08 /* [RFC1997] */ /* OT */
-#define BA_ORIGINATOR_ID 0x09 /* [RFC1966] */ /* ON */
-#define BA_CLUSTER_LIST 0x0a /* ON */
-/* We don't support these: */
-#define BA_DPA 0x0b /* ??? */
-#define BA_ADVERTISER 0x0c /* [RFC1863] */
-#define BA_RCID_PATH 0x0d
-#define BA_MP_REACH_NLRI 0x0e /* [RFC2283] */
-#define BA_MP_UNREACH_NLRI 0x0f
+#define BA_COMMUNITY 0x08 /* RFC 1997 */ /* OT */
+#define BA_ORIGINATOR_ID 0x09 /* RFC 4456 */ /* ON */
+#define BA_CLUSTER_LIST 0x0a /* RFC 4456 */ /* ON */
+#define BA_MP_REACH_NLRI 0x0e /* RFC 4760 */
+#define BA_MP_UNREACH_NLRI 0x0f /* RFC 4760 */
#define BA_EXT_COMMUNITY 0x10 /* RFC 4360 */
#define BA_AS4_PATH 0x11 /* RFC 6793 */
#define BA_AS4_AGGREGATOR 0x12 /* RFC 6793 */
-#define BA_LARGE_COMMUNITY 0x20 /* [draft-ietf-idr-large-community] */
+#define BA_LARGE_COMMUNITY 0x20 /* RFC 8092 */
+
+/* Bird's private internal BGP attributes */
+#define BA_MPLS_LABEL_STACK 0xfe /* MPLS label stack transfer attribute */
/* BGP connection states */
#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