summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat
-rw-r--r--isisd/isis_nb_config.c7
-rw-r--r--isisd/isisd.c10
-rw-r--r--lib/libospf.h1
-rw-r--r--ospfd/ospf_flood.c26
-rw-r--r--ospfd/ospf_flood.h23
-rw-r--r--ospfd/ospf_interface.c44
-rw-r--r--ospfd/ospf_interface.h17
-rw-r--r--ospfd/ospf_ism.c14
-rw-r--r--ospfd/ospf_packet.c243
-rw-r--r--ospfd/ospf_packet.h5
-rw-r--r--tests/topotests/bgp_oad/test_bgp_oad.py2
-rw-r--r--tests/topotests/bgp_remote_as_auto/r1/frr.conf6
-rw-r--r--tests/topotests/bgp_remote_as_auto/r4/frr.conf10
-rw-r--r--tests/topotests/bgp_remote_as_auto/test_bgp_remote_as_auto.py33
14 files changed, 329 insertions, 112 deletions
diff --git a/isisd/isis_nb_config.c b/isisd/isis_nb_config.c
index 30c90baa54..8926b624ea 100644
--- a/isisd/isis_nb_config.c
+++ b/isisd/isis_nb_config.c
@@ -252,11 +252,12 @@ int isis_instance_area_address_destroy(struct nb_cb_destroy_args *args)
return NB_ERR_INCONSISTENCY;
listnode_delete(area->area_addrs, addrp);
- XFREE(MTYPE_ISIS_AREA_ADDR, addrp);
/*
* Last area address - reset the SystemID for this router
*/
- if (listcount(area->area_addrs) == 0) {
+ if (!memcmp(addrp->area_addr + addrp->addr_len, area->isis->sysid,
+ ISIS_SYS_ID_LEN) &&
+ listcount(area->area_addrs) == 0) {
for (ALL_LIST_ELEMENTS_RO(area->circuit_list, cnode, circuit))
for (lvl = IS_LEVEL_1; lvl <= IS_LEVEL_2; ++lvl) {
if (circuit->u.bc.is_dr[lvl - 1])
@@ -268,6 +269,8 @@ int isis_instance_area_address_destroy(struct nb_cb_destroy_args *args)
zlog_debug("Router has no SystemID");
}
+ XFREE(MTYPE_ISIS_AREA_ADDR, addrp);
+
return NB_OK;
}
diff --git a/isisd/isisd.c b/isisd/isisd.c
index 982df0839b..8db6295b66 100644
--- a/isisd/isisd.c
+++ b/isisd/isisd.c
@@ -496,6 +496,7 @@ void isis_area_destroy(struct isis_area *area)
{
struct listnode *node, *nnode;
struct isis_circuit *circuit;
+ struct iso_address *addr;
QOBJ_UNREG(area);
@@ -545,6 +546,15 @@ void isis_area_destroy(struct isis_area *area)
if (!CHECK_FLAG(im->options, F_ISIS_UNIT_TEST))
isis_redist_area_finish(area);
+ if (listcount(area->area_addrs) > 0) {
+ addr = listgetdata(listhead(area->area_addrs));
+ if (!memcmp(addr->area_addr + addr->addr_len, area->isis->sysid,
+ ISIS_SYS_ID_LEN)) {
+ memset(area->isis->sysid, 0, ISIS_SYS_ID_LEN);
+ area->isis->sysid_set = 0;
+ }
+ }
+
list_delete(&area->area_addrs);
for (int i = SPF_PREFIX_PRIO_CRITICAL; i <= SPF_PREFIX_PRIO_MEDIUM;
diff --git a/lib/libospf.h b/lib/libospf.h
index f2dc5d61d9..8a208beb3c 100644
--- a/lib/libospf.h
+++ b/lib/libospf.h
@@ -61,6 +61,7 @@ extern "C" {
#define OSPF_RETRANSMIT_WINDOW_DEFAULT 50 /* milliseconds */
#define OSPF_TRANSMIT_DELAY_DEFAULT 1
#define OSPF_DEFAULT_BANDWIDTH 10000 /* Mbps */
+#define OSPF_ACK_DELAY_DEFAULT 1
#define OSPF_DEFAULT_REF_BANDWIDTH 100000 /* Kbps */
diff --git a/ospfd/ospf_flood.c b/ospfd/ospf_flood.c
index e9797ce935..2af4ae3170 100644
--- a/ospfd/ospf_flood.c
+++ b/ospfd/ospf_flood.c
@@ -110,6 +110,9 @@ void ospf_area_update_fr_state(struct ospf_area *area)
static void ospf_flood_delayed_lsa_ack(struct ospf_neighbor *inbr,
struct ospf_lsa *lsa)
{
+ struct ospf_lsa_list_entry *ls_ack_list_entry;
+ struct ospf_interface *oi = inbr->oi;
+
/* LSA is more recent than database copy, but was not
flooded back out receiving interface. Delayed
acknowledgment sent. If interface is in Backup state
@@ -122,12 +125,27 @@ static void ospf_flood_delayed_lsa_ack(struct ospf_neighbor *inbr,
worked out previously */
/* Deal with router as BDR */
- if (inbr->oi->state == ISM_Backup && !NBR_IS_DR(inbr))
+ if (oi->state == ISM_Backup && !NBR_IS_DR(inbr))
return;
- /* Schedule a delayed LSA Ack to be sent */
- listnode_add(inbr->oi->ls_ack,
- ospf_lsa_lock(lsa)); /* delayed LSA Ack */
+ if (IS_DEBUG_OSPF(lsa, LSA_FLOODING))
+ zlog_debug("%s:Add LSA[Type%d:%pI4:%pI4]: seq 0x%x age %u NBR %pI4 (%s) ack queue",
+ __func__, lsa->data->type, &lsa->data->id,
+ &lsa->data->adv_router, ntohl(lsa->data->ls_seqnum),
+ ntohs(lsa->data->ls_age), &inbr->router_id,
+ IF_NAME(inbr->oi));
+
+ /* Add the LSA to the interface delayed Ack list. */
+ ls_ack_list_entry = XCALLOC(MTYPE_OSPF_LSA_LIST,
+ sizeof(struct ospf_lsa_list_entry));
+ ls_ack_list_entry->lsa = ospf_lsa_lock(lsa);
+ ospf_lsa_list_add_tail(&oi->ls_ack_delayed, ls_ack_list_entry);
+
+ /* Set LS Ack timer if it is not already scheduled. */
+ if (!oi->t_ls_ack_delayed)
+ OSPF_ISM_TIMER_ON(oi->t_ls_ack_delayed,
+ ospf_ls_ack_delayed_timer,
+ oi->v_ls_ack_delayed);
}
/* Check LSA is related to external info. */
diff --git a/ospfd/ospf_flood.h b/ospfd/ospf_flood.h
index d9d9537351..2412052970 100644
--- a/ospfd/ospf_flood.h
+++ b/ospfd/ospf_flood.h
@@ -7,6 +7,8 @@
#ifndef _ZEBRA_OSPF_FLOOD_H
#define _ZEBRA_OSPF_FLOOD_H
+#include "typesafe.h"
+
/*
* OSPF Temporal LSA List
*/
@@ -16,14 +18,25 @@ struct ospf_lsa_list_entry {
/* Linkage for LSA List */
struct ospf_lsa_list_item list_linkage;
- /*
- * Time associated with the list entry. For example, for a neigbhor
- * link retransmission list, this is the retransmission time.
- */
- struct timeval list_entry_time;
+ union {
+ /*
+ * Time associated with the list entry. For example, for a
+ * neigbhor link retransmission list, this is the
+ * retransmission time.
+ */
+ struct timeval list_entry_timeval;
+
+ /*
+ * Destanation address specific to the LSA list. For example,
+ * the distination for an associated direct LS acknowledgment.
+ */
+ struct in_addr list_entry_dst_addr;
+ } u;
struct ospf_lsa *lsa;
};
+#define list_entry_time u.list_entry_timeval
+#define list_entry_dst u.list_entry_dst_addr
DECLARE_DLIST(ospf_lsa_list, struct ospf_lsa_list_entry, list_linkage);
diff --git a/ospfd/ospf_interface.c b/ospfd/ospf_interface.c
index 803c36861d..c4210eb70c 100644
--- a/ospfd/ospf_interface.c
+++ b/ospfd/ospf_interface.c
@@ -186,10 +186,12 @@ static void ospf_if_default_variables(struct ospf_interface *oi)
oi->crypt_seqnum = 0;
- /* This must be short, (less than RxmtInterval)
- - RFC 2328 Section 13.5 para 3. Set to 1 second to avoid Acks being
- held back for too long - MAG */
- oi->v_ls_ack = 1;
+ /*
+ * The OSPF LS ACK Delay timer must be less than the LS Retransmision
+ * timer. As per RFC 2328 Section 13.5 paragraph 3, Set to 1 second
+ * to avoid Acks being held back for too long
+ */
+ oi->v_ls_ack_delayed = OSPF_ACK_DELAY_DEFAULT;
}
/* lookup oi for specified prefix/ifp */
@@ -272,9 +274,9 @@ struct ospf_interface *ospf_if_new(struct ospf *ospf, struct interface *ifp,
/* Initialize static neighbor list. */
oi->nbr_nbma = list_new();
- /* Initialize Link State Acknowledgment list. */
- oi->ls_ack = list_new();
- oi->ls_ack_direct.ls_ack = list_new();
+ /* Initialize Link State Acknowledgment lists. */
+ ospf_lsa_list_init(&oi->ls_ack_delayed);
+ ospf_lsa_list_init(&oi->ls_ack_direct);
/* Set default values. */
ospf_if_default_variables(oi);
@@ -306,6 +308,22 @@ struct ospf_interface *ospf_if_new(struct ospf *ospf, struct interface *ifp,
return oi;
}
+/*
+ * Cleanup Interface Ack List
+ */
+static void ospf_if_cleanup_ack_list(struct ospf_lsa_list_head *ls_ack_list)
+{
+ struct ospf_lsa_list_entry *ls_ack_list_entry;
+ struct ospf_lsa *lsa;
+
+ frr_each_safe (ospf_lsa_list, ls_ack_list, ls_ack_list_entry) {
+ lsa = ls_ack_list_entry->lsa;
+ ospf_lsa_list_del(ls_ack_list, ls_ack_list_entry);
+ XFREE(MTYPE_OSPF_LSA_LIST, ls_ack_list_entry);
+ ospf_lsa_unlock(&lsa);
+ }
+}
+
/* Restore an interface to its pre UP state
Used from ism_interface_down only */
void ospf_if_cleanup(struct ospf_interface *oi)
@@ -314,7 +332,6 @@ void ospf_if_cleanup(struct ospf_interface *oi)
struct listnode *node, *nnode;
struct ospf_neighbor *nbr;
struct ospf_nbr_nbma *nbr_nbma;
- struct ospf_lsa *lsa;
/* oi->nbrs and oi->nbr_nbma should be deleted on InterfaceDown event */
/* delete all static neighbors attached to this interface */
@@ -338,10 +355,9 @@ void ospf_if_cleanup(struct ospf_interface *oi)
OSPF_NSM_EVENT_EXECUTE(nbr, NSM_KillNbr);
}
- /* Cleanup Link State Acknowlegdment list. */
- for (ALL_LIST_ELEMENTS(oi->ls_ack, node, nnode, lsa))
- ospf_lsa_unlock(&lsa); /* oi->ls_ack */
- list_delete_all_node(oi->ls_ack);
+ /* Cleanup Link State Delayed Acknowlegdment list. */
+ ospf_if_cleanup_ack_list(&oi->ls_ack_delayed);
+ ospf_if_cleanup_ack_list(&oi->ls_ack_direct);
oi->crypt_seqnum = 0;
@@ -377,8 +393,8 @@ void ospf_if_free(struct ospf_interface *oi)
/* Free any lists that should be freed */
list_delete(&oi->nbr_nbma);
- list_delete(&oi->ls_ack);
- list_delete(&oi->ls_ack_direct.ls_ack);
+ ospf_if_cleanup_ack_list(&oi->ls_ack_delayed);
+ ospf_if_cleanup_ack_list(&oi->ls_ack_direct);
if (IS_DEBUG_OSPF_EVENT)
zlog_debug("%s: ospf interface %s vrf %s id %u deleted",
diff --git a/ospfd/ospf_interface.h b/ospfd/ospf_interface.h
index a944847b5d..78a4fb9e59 100644
--- a/ospfd/ospf_interface.h
+++ b/ospfd/ospf_interface.h
@@ -13,6 +13,7 @@
#include "keychain.h"
#include "ospfd/ospf_packet.h"
#include "ospfd/ospf_spf.h"
+#include <ospfd/ospf_flood.h>
#define IF_OSPF_IF_INFO(I) ((struct ospf_if_info *)((I)->info))
#define IF_DEF_PARAMS(I) (IF_OSPF_IF_INFO (I)->def_params)
@@ -265,20 +266,20 @@ struct ospf_interface {
struct route_table *ls_upd_queue;
- struct list *ls_ack; /* Link State Acknowledgment list. */
-
- struct {
- struct list *ls_ack;
- struct in_addr dst;
- } ls_ack_direct;
+ /*
+ * List of LSAs for delayed and direct link
+ * state acknowledgment transmission.
+ */
+ struct ospf_lsa_list_head ls_ack_delayed;
+ struct ospf_lsa_list_head ls_ack_direct;
/* Timer values. */
- uint32_t v_ls_ack; /* Delayed Link State Acknowledgment */
+ uint32_t v_ls_ack_delayed; /* Delayed Link State Acknowledgment */
/* Threads. */
struct event *t_hello; /* timer */
struct event *t_wait; /* timer */
- struct event *t_ls_ack; /* timer */
+ struct event *t_ls_ack_delayed; /* timer */
struct event *t_ls_ack_direct; /* event */
struct event *t_ls_upd_event; /* event */
struct event *t_opaque_lsa_self; /* Type-9 Opaque-LSAs */
diff --git a/ospfd/ospf_ism.c b/ospfd/ospf_ism.c
index 878ab725bd..377e7a6bcc 100644
--- a/ospfd/ospf_ism.c
+++ b/ospfd/ospf_ism.c
@@ -285,7 +285,7 @@ static void ism_timer_set(struct ospf_interface *oi)
reset also. */
EVENT_OFF(oi->t_hello);
EVENT_OFF(oi->t_wait);
- EVENT_OFF(oi->t_ls_ack);
+ EVENT_OFF(oi->t_ls_ack_delayed);
EVENT_OFF(oi->gr.hello_delay.t_grace_send);
break;
case ISM_Loopback:
@@ -293,7 +293,7 @@ static void ism_timer_set(struct ospf_interface *oi)
unavailable for regular data traffic. */
EVENT_OFF(oi->t_hello);
EVENT_OFF(oi->t_wait);
- EVENT_OFF(oi->t_ls_ack);
+ EVENT_OFF(oi->t_ls_ack_delayed);
EVENT_OFF(oi->gr.hello_delay.t_grace_send);
break;
case ISM_Waiting:
@@ -304,7 +304,7 @@ static void ism_timer_set(struct ospf_interface *oi)
OSPF_ISM_TIMER_MSEC_ON(oi->t_hello, ospf_hello_timer, 1);
OSPF_ISM_TIMER_ON(oi->t_wait, ospf_wait_timer,
OSPF_IF_PARAM(oi, v_wait));
- EVENT_OFF(oi->t_ls_ack);
+ EVENT_OFF(oi->t_ls_ack_delayed);
break;
case ISM_PointToPoint:
/* The interface connects to a physical Point-to-point network
@@ -314,8 +314,6 @@ static void ism_timer_set(struct ospf_interface *oi)
/* send first hello immediately */
OSPF_ISM_TIMER_MSEC_ON(oi->t_hello, ospf_hello_timer, 1);
EVENT_OFF(oi->t_wait);
- OSPF_ISM_TIMER_ON(oi->t_ls_ack, ospf_ls_ack_timer,
- oi->v_ls_ack);
break;
case ISM_DROther:
/* The network type of the interface is broadcast or NBMA
@@ -324,8 +322,6 @@ static void ism_timer_set(struct ospf_interface *oi)
Backup Designated Router. */
OSPF_HELLO_TIMER_ON(oi);
EVENT_OFF(oi->t_wait);
- OSPF_ISM_TIMER_ON(oi->t_ls_ack, ospf_ls_ack_timer,
- oi->v_ls_ack);
break;
case ISM_Backup:
/* The network type of the interface is broadcast os NBMA
@@ -333,8 +329,6 @@ static void ism_timer_set(struct ospf_interface *oi)
and the router is Backup Designated Router. */
OSPF_HELLO_TIMER_ON(oi);
EVENT_OFF(oi->t_wait);
- OSPF_ISM_TIMER_ON(oi->t_ls_ack, ospf_ls_ack_timer,
- oi->v_ls_ack);
break;
case ISM_DR:
/* The network type of the interface is broadcast or NBMA
@@ -342,8 +336,6 @@ static void ism_timer_set(struct ospf_interface *oi)
and the router is Designated Router. */
OSPF_HELLO_TIMER_ON(oi);
EVENT_OFF(oi->t_wait);
- OSPF_ISM_TIMER_ON(oi->t_ls_ack, ospf_ls_ack_timer,
- oi->v_ls_ack);
break;
}
}
diff --git a/ospfd/ospf_packet.c b/ospfd/ospf_packet.c
index 86f877b621..2d15a7ecca 100644
--- a/ospfd/ospf_packet.c
+++ b/ospfd/ospf_packet.c
@@ -369,19 +369,16 @@ void ospf_ls_rxmt_timer(struct event *thread)
ospf_ls_retransmit_set_timer(nbr);
}
-void ospf_ls_ack_timer(struct event *thread)
+void ospf_ls_ack_delayed_timer(struct event *thread)
{
struct ospf_interface *oi;
oi = EVENT_ARG(thread);
- oi->t_ls_ack = NULL;
+ oi->t_ls_ack_delayed = NULL;
/* Send Link State Acknowledgment. */
- if (listcount(oi->ls_ack) > 0)
+ if (ospf_lsa_list_count(&oi->ls_ack_delayed))
ospf_ls_ack_send_delayed(oi);
-
- /* Set LS Ack timer. */
- OSPF_ISM_TIMER_ON(oi->t_ls_ack, ospf_ls_ack_timer, oi->v_ls_ack);
}
#ifdef WANT_OSPF_WRITE_FRAGMENT
@@ -1820,7 +1817,7 @@ static void ospf_ls_upd(struct ospf *ospf, struct ip *iph,
if (IS_LSA_MAXAGE(lsa) && !current
&& ospf_check_nbr_status(oi->ospf)) {
/* (4a) Response Link State Acknowledgment. */
- ospf_ls_ack_send(nbr, lsa);
+ ospf_ls_ack_send_direct(nbr, lsa);
/* (4b) Discard LSA. */
if (IS_DEBUG_OSPF(lsa, LSA)) {
@@ -1845,7 +1842,7 @@ static void ospf_ls_upd(struct ospf *ospf, struct ip *iph,
if (IS_LSA_MAXAGE(lsa)) {
zlog_info("LSA[%s]: Boomerang effect?",
dump_lsa_key(lsa));
- ospf_ls_ack_send(nbr, lsa);
+ ospf_ls_ack_send_direct(nbr, lsa);
ospf_lsa_discard(lsa);
if (current != NULL && !IS_LSA_MAXAGE(current))
@@ -1879,7 +1876,7 @@ static void ospf_ls_upd(struct ospf *ospf, struct ip *iph,
SET_FLAG(lsa->flags, OSPF_LSA_SELF);
- ospf_ls_ack_send(nbr, lsa);
+ ospf_ls_ack_send_direct(nbr, lsa);
if (!ospf->gr_info.restart_in_progress) {
ospf_opaque_self_originated_lsa_received(
@@ -2018,9 +2015,8 @@ static void ospf_ls_upd(struct ospf *ospf, struct ip *iph,
*/
if (oi->state == ISM_Backup)
if (NBR_IS_DR(nbr))
- listnode_add(
- oi->ls_ack,
- ospf_lsa_lock(lsa));
+ ospf_ls_ack_send_direct(nbr,
+ lsa);
DISCARD_LSA(lsa, 6);
} else
@@ -2029,7 +2025,7 @@ static void ospf_ls_upd(struct ospf *ospf, struct ip *iph,
receiving
interface. */
{
- ospf_ls_ack_send(nbr, lsa);
+ ospf_ls_ack_send_direct(nbr, lsa);
DISCARD_LSA(lsa, 7);
}
}
@@ -3331,17 +3327,36 @@ static int ospf_make_ls_upd(struct ospf_interface *oi, struct list *update,
return length;
}
-static int ospf_make_ls_ack(struct ospf_interface *oi, struct list *ack,
- struct stream *s)
+static int ospf_make_ls_ack(struct ospf_interface *oi,
+ struct ospf_lsa_list_head *ls_ack_list,
+ bool direct_ack, bool delete_ack, struct stream *s)
{
- struct listnode *node, *nnode;
+ struct ospf_lsa_list_entry *ls_ack_list_first;
+ struct ospf_lsa_list_entry *ls_ack_list_entry;
uint16_t length = OSPF_LS_ACK_MIN_SIZE;
- unsigned long delta = OSPF_LSA_HEADER_SIZE;
struct ospf_lsa *lsa;
+ struct in_addr first_dst_addr;
- for (ALL_LIST_ELEMENTS(ack, node, nnode, lsa)) {
+ /*
+ * For direct LS Acks, assure the destination address doesn't
+ * change between queued acknowledgments.
+ */
+ if (direct_ack) {
+ ls_ack_list_first = ospf_lsa_list_first(ls_ack_list);
+ if (ls_ack_list_first)
+ first_dst_addr.s_addr =
+ ls_ack_list_first->list_entry_dst.s_addr;
+ } else
+ first_dst_addr.s_addr = INADDR_ANY;
+
+ frr_each_safe (ospf_lsa_list, ls_ack_list, ls_ack_list_entry) {
+ lsa = ls_ack_list_entry->lsa;
assert(lsa);
+ if (direct_ack && (ls_ack_list_entry->list_entry_dst.s_addr !=
+ first_dst_addr.s_addr))
+ break;
+
/* LS Ack packet overflows interface MTU
* delta is just number of bytes required for
* 1 LS Ack(1 LS Hdr) ospf_packet_max will return
@@ -3350,19 +3365,46 @@ static int ospf_make_ls_ack(struct ospf_interface *oi, struct list *ack,
* against ospf_packet_max to check if it can fit
* another ls header in the same packet.
*/
- if ((length + delta) > ospf_packet_max(oi))
+ if ((length + OSPF_LSA_HEADER_SIZE) > ospf_packet_max(oi))
break;
stream_put(s, lsa->data, OSPF_LSA_HEADER_SIZE);
length += OSPF_LSA_HEADER_SIZE;
- listnode_delete(ack, lsa);
- ospf_lsa_unlock(&lsa); /* oi->ls_ack_direct.ls_ack */
+ if (delete_ack) {
+ ospf_lsa_list_del(ls_ack_list, ls_ack_list_entry);
+ XFREE(MTYPE_OSPF_LSA_LIST, ls_ack_list_entry);
+ ospf_lsa_unlock(&lsa);
+ }
}
return length;
}
+/*
+ * On non-braodcast networks, the same LS acks must be sent to multiple
+ * neighbors and deletion must be deferred until after the LS Ack packet
+ * is sent to all neighbors.
+ */
+static void ospf_delete_ls_ack_delayed(struct ospf_interface *oi)
+{
+ struct ospf_lsa_list_entry *ls_ack_list_entry;
+ struct ospf_lsa *lsa;
+ uint16_t length = OSPF_LS_ACK_MIN_SIZE;
+
+ frr_each_safe (ospf_lsa_list, &oi->ls_ack_delayed, ls_ack_list_entry) {
+ lsa = ls_ack_list_entry->lsa;
+ assert(lsa);
+ if ((length + OSPF_LSA_HEADER_SIZE) > ospf_packet_max(oi))
+ break;
+
+ length += OSPF_LSA_HEADER_SIZE;
+ ospf_lsa_list_del(&oi->ls_ack_delayed, ls_ack_list_entry);
+ XFREE(MTYPE_OSPF_LSA_LIST, ls_ack_list_entry);
+ ospf_lsa_unlock(&lsa);
+ }
+}
+
static void ospf_hello_send_sub(struct ospf_interface *oi, in_addr_t addr)
{
struct ospf_packet *op;
@@ -3934,10 +3976,13 @@ void ospf_ls_upd_send(struct ospf_neighbor *nbr, struct list *update, int flag,
&oi->t_ls_upd_event);
}
-static void ospf_ls_ack_send_list(struct ospf_interface *oi, struct list *ack,
+static void ospf_ls_ack_send_list(struct ospf_interface *oi,
+ struct ospf_lsa_list_head *ls_ack_list,
+ bool direct_ack, bool delete_ack,
struct in_addr dst)
{
struct ospf_packet *op;
+ struct ospf_lsa_list_entry *ls_ack_list_first;
uint16_t length = OSPF_HEADER_SIZE;
op = ospf_packet_new(oi->ifp->mtu);
@@ -3945,8 +3990,18 @@ static void ospf_ls_ack_send_list(struct ospf_interface *oi, struct list *ack,
/* Prepare OSPF common header. */
ospf_make_header(OSPF_MSG_LS_ACK, oi, op->s);
+ /* Determine the destination address - for direct acks,
+ * the list entries always include the distination address.
+ */
+ if (direct_ack) {
+ ls_ack_list_first = ospf_lsa_list_first(ls_ack_list);
+ op->dst.s_addr = ls_ack_list_first->list_entry_dst.s_addr;
+ } else
+ op->dst.s_addr = dst.s_addr;
+
/* Prepare OSPF Link State Acknowledgment body. */
- length += ospf_make_ls_ack(oi, ack, op->s);
+ length += ospf_make_ls_ack(oi, ls_ack_list, direct_ack, delete_ack,
+ op->s);
/* Fill OSPF header. */
ospf_fill_header(oi, op->s, length);
@@ -3954,14 +4009,6 @@ static void ospf_ls_ack_send_list(struct ospf_interface *oi, struct list *ack,
/* Set packet length. */
op->length = length;
- /* Decide destination address. */
- if (oi->type == OSPF_IFTYPE_POINTOPOINT ||
- (oi->type == OSPF_IFTYPE_POINTOMULTIPOINT &&
- !oi->p2mp_non_broadcast))
- op->dst.s_addr = htonl(OSPF_ALLSPFROUTERS);
- else
- op->dst.s_addr = dst.s_addr;
-
/* Add packet to the interface output queue. */
ospf_packet_add(oi, op);
@@ -3969,34 +4016,96 @@ static void ospf_ls_ack_send_list(struct ospf_interface *oi, struct list *ack,
OSPF_ISM_WRITE_ON(oi->ospf);
}
-static void ospf_ls_ack_send_event(struct event *thread)
+static void ospf_ls_ack_send_direct_event(struct event *thread)
{
struct ospf_interface *oi = EVENT_ARG(thread);
+ struct in_addr dst = { INADDR_ANY };
oi->t_ls_ack_direct = NULL;
- while (listcount(oi->ls_ack_direct.ls_ack))
- ospf_ls_ack_send_list(oi, oi->ls_ack_direct.ls_ack,
- oi->ls_ack_direct.dst);
+ while (ospf_lsa_list_count(&oi->ls_ack_direct))
+ ospf_ls_ack_send_list(oi, &(oi->ls_ack_direct), true, true, dst);
}
-void ospf_ls_ack_send(struct ospf_neighbor *nbr, struct ospf_lsa *lsa)
+void ospf_ls_ack_send_direct(struct ospf_neighbor *nbr, struct ospf_lsa *lsa)
{
+ struct ospf_lsa_list_entry *ls_ack_list_entry;
struct ospf_interface *oi = nbr->oi;
+ if (IS_DEBUG_OSPF(lsa, LSA_FLOODING))
+ zlog_debug("%s:Add LSA[Type%d:%pI4:%pI4]: seq 0x%x age %u NBR %pI4 (%s) ack queue",
+ __func__, lsa->data->type, &lsa->data->id,
+ &lsa->data->adv_router, ntohl(lsa->data->ls_seqnum),
+ ntohs(lsa->data->ls_age), &nbr->router_id,
+ IF_NAME(nbr->oi));
+
+ /*
+ * On Point-to-Multipoint broadcast-capabile interfaces,
+ * where direct acks from are sent to the ALLSPFRouters
+ * address and one direct ack send event, may include LSAs
+ * from multiple neighbors, there is a possibility of the same
+ * LSA being processed more than once in the same send event.
+ * In this case, the instances subsequent to the first can be
+ * ignored.
+ */
+ if (oi->type == OSPF_IFTYPE_POINTOMULTIPOINT && !oi->p2mp_non_broadcast) {
+ struct ospf_lsa_list_entry *ls_ack_list_entry;
+ struct ospf_lsa *ack_queue_lsa;
+
+ frr_each (ospf_lsa_list, &oi->ls_ack_direct, ls_ack_list_entry) {
+ ack_queue_lsa = ls_ack_list_entry->lsa;
+ if ((lsa == ack_queue_lsa) ||
+ ((lsa->data->type == ack_queue_lsa->data->type) &&
+ (lsa->data->id.s_addr ==
+ ack_queue_lsa->data->id.s_addr) &&
+ (lsa->data->adv_router.s_addr ==
+ ack_queue_lsa->data->adv_router.s_addr) &&
+ (lsa->data->ls_seqnum ==
+ ack_queue_lsa->data->ls_seqnum))) {
+ if (IS_DEBUG_OSPF(lsa, LSA_FLOODING))
+ zlog_debug("%s:LSA[Type%d:%pI4:%pI4]: seq 0x%x age %u NBR %pI4 (%s) ack queue duplicate",
+ __func__, lsa->data->type,
+ &lsa->data->id,
+ &lsa->data->adv_router,
+ ntohl(lsa->data->ls_seqnum),
+ ntohs(lsa->data->ls_age),
+ &nbr->router_id,
+ IF_NAME(nbr->oi));
+ return;
+ }
+ }
+ }
+
if (IS_GRACE_LSA(lsa)) {
if (IS_DEBUG_OSPF_GR)
zlog_debug("%s, Sending GRACE ACK to Restarter.",
__func__);
}
- if (listcount(oi->ls_ack_direct.ls_ack) == 0)
- oi->ls_ack_direct.dst = nbr->address.u.prefix4;
+ ls_ack_list_entry = XCALLOC(MTYPE_OSPF_LSA_LIST,
+ sizeof(struct ospf_lsa_list_entry));
- listnode_add(oi->ls_ack_direct.ls_ack, ospf_lsa_lock(lsa));
+ /*
+ * Determine the destination address - Direct LS acknowledgments
+ * are sent the AllSPFRouters multicast address on Point-to-Point
+ * and Point-to-Multipoint broadcast-capable interfaces. For all other
+ * interface types, they are unicast directly to the neighbor.
+ */
+ if (oi->type == OSPF_IFTYPE_POINTOPOINT ||
+ (oi->type == OSPF_IFTYPE_POINTOMULTIPOINT &&
+ !oi->p2mp_non_broadcast))
+ ls_ack_list_entry->list_entry_dst.s_addr =
+ htonl(OSPF_ALLSPFROUTERS);
+ else
+ ls_ack_list_entry->list_entry_dst.s_addr =
+ nbr->address.u.prefix4.s_addr;
- event_add_event(master, ospf_ls_ack_send_event, oi, 0,
- &oi->t_ls_ack_direct);
+ ls_ack_list_entry->lsa = ospf_lsa_lock(lsa);
+ ospf_lsa_list_add_tail(&nbr->oi->ls_ack_direct, ls_ack_list_entry);
+
+ if (oi->t_ls_ack_direct == NULL)
+ event_add_event(master, ospf_ls_ack_send_direct_event, oi, 0,
+ &oi->t_ls_ack_direct);
}
/* Send Link State Acknowledgment delayed. */
@@ -4013,33 +4122,39 @@ void ospf_ls_ack_send_delayed(struct ospf_interface *oi)
struct ospf_neighbor *nbr;
struct route_node *rn;
- for (rn = route_top(oi->nbrs); rn; rn = route_next(rn)) {
- nbr = rn->info;
+ while (ospf_lsa_list_count(&oi->ls_ack_delayed)) {
+ for (rn = route_top(oi->nbrs); rn; rn = route_next(rn)) {
+ nbr = rn->info;
- if (!nbr)
- continue;
+ if (!nbr)
+ continue;
- if (nbr != oi->nbr_self && nbr->state >= NSM_Exchange)
- while (listcount(oi->ls_ack))
- ospf_ls_ack_send_list(
- oi, oi->ls_ack,
- nbr->address.u.prefix4);
+ if (nbr != oi->nbr_self &&
+ nbr->state >= NSM_Exchange)
+ ospf_ls_ack_send_list(oi,
+ &oi->ls_ack_delayed,
+ false, false,
+ nbr->address.u
+ .prefix4);
+ }
+ ospf_delete_ls_ack_delayed(oi);
}
- return;
- }
- if (oi->type == OSPF_IFTYPE_VIRTUALLINK)
- dst.s_addr = oi->vl_data->peer_addr.s_addr;
- else if (oi->state == ISM_DR || oi->state == ISM_Backup)
- dst.s_addr = htonl(OSPF_ALLSPFROUTERS);
- else if (oi->type == OSPF_IFTYPE_POINTOPOINT)
- dst.s_addr = htonl(OSPF_ALLSPFROUTERS);
- else if (oi->type == OSPF_IFTYPE_POINTOMULTIPOINT)
- dst.s_addr = htonl(OSPF_ALLSPFROUTERS);
- else
- dst.s_addr = htonl(OSPF_ALLDROUTERS);
+ } else {
+ if (oi->type == OSPF_IFTYPE_VIRTUALLINK)
+ dst.s_addr = oi->vl_data->peer_addr.s_addr;
+ else if (oi->state == ISM_DR || oi->state == ISM_Backup)
+ dst.s_addr = htonl(OSPF_ALLSPFROUTERS);
+ else if (oi->type == OSPF_IFTYPE_POINTOPOINT)
+ dst.s_addr = htonl(OSPF_ALLSPFROUTERS);
+ else if (oi->type == OSPF_IFTYPE_POINTOMULTIPOINT)
+ dst.s_addr = htonl(OSPF_ALLSPFROUTERS);
+ else
+ dst.s_addr = htonl(OSPF_ALLDROUTERS);
- while (listcount(oi->ls_ack))
- ospf_ls_ack_send_list(oi, oi->ls_ack, dst);
+ while (ospf_lsa_list_count(&oi->ls_ack_delayed))
+ ospf_ls_ack_send_list(oi, &oi->ls_ack_delayed, false,
+ true, dst);
+ }
}
/*
diff --git a/ospfd/ospf_packet.h b/ospfd/ospf_packet.h
index 2c9dba6c88..84e4b027e6 100644
--- a/ospfd/ospf_packet.h
+++ b/ospfd/ospf_packet.h
@@ -135,13 +135,14 @@ extern void ospf_ls_upd_send(struct ospf_neighbor *, struct list *, int, int);
extern void ospf_ls_upd_queue_send(struct ospf_interface *oi,
struct list *update, struct in_addr addr,
int send_lsupd_now);
-extern void ospf_ls_ack_send(struct ospf_neighbor *, struct ospf_lsa *);
+extern void ospf_ls_ack_send_direct(struct ospf_neighbor *nbr,
+ struct ospf_lsa *lsa);
extern void ospf_ls_ack_send_delayed(struct ospf_interface *);
extern void ospf_ls_retransmit(struct ospf_interface *, struct ospf_lsa *);
extern void ospf_ls_req_event(struct ospf_neighbor *);
extern void ospf_ls_rxmt_timer(struct event *thread);
-extern void ospf_ls_ack_timer(struct event *thread);
+extern void ospf_ls_ack_delayed_timer(struct event *thread);
extern void ospf_poll_timer(struct event *thread);
extern void ospf_hello_reply_timer(struct event *thread);
diff --git a/tests/topotests/bgp_oad/test_bgp_oad.py b/tests/topotests/bgp_oad/test_bgp_oad.py
index a2ca37a2b7..bb779462db 100644
--- a/tests/topotests/bgp_oad/test_bgp_oad.py
+++ b/tests/topotests/bgp_oad/test_bgp_oad.py
@@ -46,7 +46,7 @@ def teardown_module(mod):
tgen.stop_topology()
-def test_bgp_dynamic_capability_role():
+def test_bgp_oad():
tgen = get_topogen()
if tgen.routers_have_failure():
diff --git a/tests/topotests/bgp_remote_as_auto/r1/frr.conf b/tests/topotests/bgp_remote_as_auto/r1/frr.conf
index aec0e76a3f..2f1bcd275f 100644
--- a/tests/topotests/bgp_remote_as_auto/r1/frr.conf
+++ b/tests/topotests/bgp_remote_as_auto/r1/frr.conf
@@ -2,6 +2,9 @@
int r1-eth0
ip address 192.168.1.1/24
!
+int r1-eth1
+ ip address 192.168.14.1/24
+!
router bgp 65001
no bgp ebgp-requires-policy
no bgp network import-check
@@ -11,6 +14,9 @@ router bgp 65001
neighbor 192.168.1.3 remote-as auto
neighbor 192.168.1.3 timers 1 3
neighbor 192.168.1.3 timers connect 1
+ neighbor r1-eth1 interface remote-as auto
+ neighbor r1-eth1 timers 1 3
+ neighbor r1-eth1 timers connect 1
address-family ipv4 unicast
network 10.0.0.1/32
exit-address-family
diff --git a/tests/topotests/bgp_remote_as_auto/r4/frr.conf b/tests/topotests/bgp_remote_as_auto/r4/frr.conf
new file mode 100644
index 0000000000..e280a6c6e8
--- /dev/null
+++ b/tests/topotests/bgp_remote_as_auto/r4/frr.conf
@@ -0,0 +1,10 @@
+!
+int r4-eth0
+ ip address 192.168.14.4/24
+!
+router bgp 65004
+ no bgp ebgp-requires-policy
+ neighbor r4-eth0 interface remote-as auto
+ neighbor r4-eth0 timers 1 3
+ neighbor r4-eth0 timers connect 1
+!
diff --git a/tests/topotests/bgp_remote_as_auto/test_bgp_remote_as_auto.py b/tests/topotests/bgp_remote_as_auto/test_bgp_remote_as_auto.py
index b932920e09..1db6d98a42 100644
--- a/tests/topotests/bgp_remote_as_auto/test_bgp_remote_as_auto.py
+++ b/tests/topotests/bgp_remote_as_auto/test_bgp_remote_as_auto.py
@@ -23,7 +23,7 @@ pytestmark = [pytest.mark.bgpd]
def setup_module(mod):
- topodef = {"s1": ("r1", "r2", "r3")}
+ topodef = {"s1": ("r1", "r2", "r3"), "s2": ("r1", "r4")}
tgen = Topogen(topodef, mod.__name__)
tgen.start_topology()
@@ -49,11 +49,18 @@ def test_bgp_remote_as_auto():
r1 = tgen.gears["r1"]
r2 = tgen.gears["r2"]
r3 = tgen.gears["r3"]
+ r4 = tgen.gears["r4"]
def _bgp_converge():
output = json.loads(r1.vtysh_cmd("show bgp ipv4 unicast summary json"))
expected = {
"peers": {
+ "r1-eth1": {
+ "hostname": "r4",
+ "remoteAs": 65004,
+ "localAs": 65001,
+ "state": "Established",
+ },
"192.168.1.2": {
"hostname": "r2",
"remoteAs": 65001,
@@ -124,6 +131,30 @@ def test_bgp_remote_as_auto():
_, result = topotest.run_and_expect(test_func, None, count=30, wait=1)
assert result is None, "Can't see automatic eBGP peering"
+ def _bgp_converge_external_unnumbered():
+ output = json.loads(r4.vtysh_cmd("show bgp ipv4 unicast 10.0.0.1/32 json"))
+ expected = {
+ "paths": [
+ {
+ "aspath": {
+ "string": "65001",
+ },
+ "valid": True,
+ "peer": {
+ "hostname": "r1",
+ "type": "external",
+ },
+ }
+ ]
+ }
+ return topotest.json_cmp(output, expected)
+
+ test_func = functools.partial(
+ _bgp_converge_external_unnumbered,
+ )
+ _, result = topotest.run_and_expect(test_func, None, count=30, wait=1)
+ assert result is None, "Can't see automatic unnumbered eBGP peering"
+
if __name__ == "__main__":
args = ["-s"] + sys.argv[1:]