mon/OSDMonitor: fix NEW flag adjustment in destroy[-new]

Only set the NEW flag if it isn't already set.

Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit 13393f6)

src/mon/OSDMonitor.cc

@@ -7774,7 +7774,10 @@ int OSDMonitor::prepare_command_osd_new(
     assert(id >= 0);
     assert(osdmap.is_destroyed(id));
     pending_inc.new_weight[id] = CEPH_OSD_OUT;
-    pending_inc.new_state[id] |= CEPH_OSD_DESTROYED | CEPH_OSD_NEW;
+    pending_inc.new_state[id] |= CEPH_OSD_DESTROYED;
+    if ((osdmap.get_state(id) & CEPH_OSD_NEW) == 0) {
+      pending_inc.new_state[id] |= CEPH_OSD_NEW;
+    }
     if (osdmap.get_state(id) & CEPH_OSD_UP) {
       // due to http://tracker.ceph.com/issues/20751 some clusters may
       // have UP set for non-existent OSDs; make sure it is cleared