bgpd : Prevent deletion of BGP peer groups associated to listen range

Description:
-----
Deleting a peer group also deletes its associated BGP listen range.
This behaviour is undesired as it could cause unintended configuration changes.

Fix :
-----
-Do not allow peer group deletion until they are no longer associated with any listen range.
-Check the count of listen ranges attached to the group.
If any listen ranges are found, returns a configuration warning, preventing the deletion.

Signed-off-by: Pooja Rathore <rathorepo@vmware.com>

bgpd/bgp_vty.c

@@ -5102,6 +5102,8 @@ DEFUN (no_neighbor,
 	struct peer_group *group;
 	struct peer *peer;
 	struct peer *other;
+	afi_t afi;
+	int lr_count;
 
 	ret = str2sockunion(argv[idx_peer]->arg, &su);
 	if (ret < 0) {
@@ -5119,6 +5121,15 @@ DEFUN (no_neighbor,
 
 		group = peer_group_lookup(bgp, argv[idx_peer]->arg);
 		if (group) {
+			for (afi = AFI_IP; afi < AFI_MAX; afi++) {
+				lr_count = listcount(group->listen_range[afi]);
+				if (lr_count) {
+					vty_out(vty,
+						"%%Peer-group %s is attached to %d listen-range(s), delete them first\n",
+						group->name, lr_count);
+					return CMD_WARNING_CONFIG_FAILED;
+				}
+			}
 			peer_group_notify_unconfig(group);
 			peer_group_delete(group);
 		} else {
@@ -5196,9 +5207,20 @@ DEFUN (no_neighbor_peer_group,
 	VTY_DECLVAR_CONTEXT(bgp, bgp);
 	int idx_word = 2;
 	struct peer_group *group;
+	afi_t afi;
+	int lr_count;
 
 	group = peer_group_lookup(bgp, argv[idx_word]->arg);
 	if (group) {
+		for (afi = AFI_IP; afi < AFI_MAX; afi++) {
+			lr_count = listcount(group->listen_range[afi]);
+			if (lr_count) {
+				vty_out(vty,
+					"%%Peer-group %s is attached to %d listen-range(s), delete them first\n",
+					group->name, lr_count);
+				return CMD_WARNING_CONFIG_FAILED;
+			}
+		}
 		peer_group_notify_unconfig(group);
 		peer_group_delete(group);
 	} else {