Revert "bgpd: fix 6vpe nexthop"

This reverts commit 0325116.

It was causing an issue where a nexthop for IPv6 over an IPv4 session
was always rewritten to an IPv4-mapped IPv6 address even when a valid
IPv6 global address was existing.

Link: FRRouting#15610
Signed-off-by: Louis Scalbert <louis.scalbert@6wind.com>

bgpd/bgp_updgrp_packet.c

@@ -523,16 +523,11 @@ struct stream *bpacket_reformat_for_peer(struct bpacket *pkt,
 			gnh_modified = 1;
 		}
 
-		if (peer->nexthop.v4.s_addr != INADDR_ANY &&
-		    (IN6_IS_ADDR_UNSPECIFIED(mod_v6nhg) ||
-		     (peer->connection->su.sa.sa_family == AF_INET &&
-		      paf->afi == AFI_IP6))) {
-			/* set a IPv4 mapped IPv6 address if no global IPv6
-			 * address is found or if announcing IPv6 prefix
-			 * over an IPv4 BGP session.
-			 */
-			ipv4_to_ipv4_mapped_ipv6(mod_v6nhg, peer->nexthop.v4);
-			gnh_modified = 1;
+		if (IN6_IS_ADDR_UNSPECIFIED(mod_v6nhg)) {
+			if (peer->nexthop.v4.s_addr != INADDR_ANY) {
+				ipv4_to_ipv4_mapped_ipv6(mod_v6nhg,
+							 peer->nexthop.v4);
+			}
 		}
 
 		if (IS_MAPPED_IPV6(&peer->nexthop.v6_global)) {