GARP to all interfaces in vrrp_sync_group

Hi,

Changelog Release 1.2.10:
...
 vrrp: fix/extend gratuitous ARP handling.
...
 1) When a MASTER receive a higher prio advert it sends a last
     advert before transiting to BACKUP state. The immediate
     effect at remote MASTER side is to sollicite a gratuitous
     ARP broadcast.
...

When a MASTER with a higher prio receives a lower prio advert, it sends GARP to this interface.
What about sending GARP to all interfaces in vrrp_sync_group?

When advert packets from MASTER (server1) no longer reach BACKUP (server2) (for example, because of the errors in intermediate device or links), we have a case with two MASTER servers. 
Last GARP was sent from server2.
When connectivity recovers, sometimes advert packet from server2 reaches only one interface of server1 and arp-table updates (to MAC of server1) only on a one side from cluster. 
Other side continues to use old arp-table (with MAC of server2). 
If we send GARP to all interfaces in vrrp_sync_group, all sides would use MAC of server1. It seems more correct behaviour.

------------------------------------------------------------------------------