Each node action flag should be set or cleared separately, instead
we now set the whole flags variable in one shot, and it's turned
out to be hard to see which other flags are affected. Therefore,
for instance, we explicitly clear TIPC_WAIT_OWN_LINKS_DOWN bit in
node_lost_contact().
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
* peer has lost contact -- don't allow peer's links
* to reactivate before we recognize loss & clean up
*/
- l_ptr->owner->action_flags = TIPC_WAIT_OWN_LINKS_DOWN;
+ l_ptr->owner->action_flags |= TIPC_WAIT_OWN_LINKS_DOWN;
}
link_state_event(l_ptr, RESET_MSG);
tipc_link_reset_fragments(l_ptr);
}
+ n_ptr->action_flags &= ~TIPC_WAIT_OWN_LINKS_DOWN;
+
/* Notify subscribers and prevent re-contact with node until
* cleanup is done.
*/
- n_ptr->action_flags = TIPC_WAIT_PEER_LINKS_DOWN |
- TIPC_NOTIFY_NODE_DOWN;
+ n_ptr->action_flags |= TIPC_WAIT_PEER_LINKS_DOWN |
+ TIPC_NOTIFY_NODE_DOWN;
}
struct sk_buff *tipc_node_get_nodes(const void *req_tlv_area, int req_tlv_space)