runtime: wake_slot default ON (RFC 005 accepted, finding 18); supervisor stops survivors sequentially so reverse-order teardown is guaranteed; deflake gen_server/channel tests; baseline.json regenerated slot-on; ROADMAP notes

tests/channel.rs

@@ -137,11 +137,19 @@ fn channel_ops_interleaved_with_monitor_churn_multi_thread() {
         for i in 0..32i64 {
             let tx = tx.clone();
             handles.push(spawn(move || {
-                // Short-lived target whose death fires the monitor below.
+                // Short-lived target whose death fires the monitor below. It
+                // waits for `go` so the monitor is registered before it can
+                // die — otherwise a fast target yields an immediate NoProc
+                // Down instead of the finalize-sent Exit this test is about
+                // (pre-existing ~8% flake at 4 threads, independent of the
+                // wake slot).
+                let (go_tx, go_rx) = channel::<()>();
                 let t = spawn(move || {
+                    let _ = go_rx.recv();
                     tx.send(i).unwrap();
                 });
                 let m = smarm::monitor(t.pid());
+                go_tx.send(()).unwrap();
                 t.join().unwrap();
                 // Down delivery exercises send-from-finalize.
                 let d = m.rx.recv().unwrap();