networkd: rework idle detection logic of networkd

This patch makes networkd stay around as long as there is more than just a
loopback interface around, or the loopback device isn't fully probed yet, or
the loopback device has a .network file attached.

In essence, this means networkd stays around now continously as it should,
unless it is running in some (container?) environment that really has no
interface except a loopback device.

Fixes #2577.

src/network/networkd-manager.c

@@ -1091,22 +1091,19 @@ static bool manager_check_idle(void *userdata) {
 
         assert(m);
 
+        /* Check whether we are idle now. The only case when we decide to be idle is when there's only a loopback
+         * device around, for which we have no configuration, and which already left the PENDING state. In all other
+         * cases we are not idle. */
+
         HASHMAP_FOREACH(link, m->links, i) {
-                /* we are not woken on udev activity, so let's just wait for the
-                 * pending udev event */
+                /* We are not woken on udev activity, so let's just wait for the pending udev event */
                 if (link->state == LINK_STATE_PENDING)
                         return false;
 
-                if (!link->network)
-                        continue;
+                if ((link->flags & IFF_LOOPBACK) == 0)
+                        return false;
 
-                /* we are not woken on netork activity, so let's stay around */
-                if (link_lldp_enabled(link) ||
-                    link_ipv4ll_enabled(link) ||
-                    link_dhcp4_server_enabled(link) ||
-                    link_dhcp4_enabled(link) ||
-                    link_dhcp6_enabled(link) ||
-                    link_ipv6_accept_ra_enabled(link))
+                if (link->network)
                         return false;
         }