hscontrol/poll,state: fix grace period disconnect TOCTOU race

When a node disconnects, serveLongPoll defers a cleanup that starts a
grace period goroutine. This goroutine polls batcher.IsConnected() and,
if the node has not reconnected within ~10 seconds, calls
state.Disconnect() to mark it offline. A TOCTOU race exists: the node
can reconnect (calling Connect()) between the IsConnected check and
the Disconnect() call, causing the stale Disconnect() to overwrite
the new session's online status.

Fix with a monotonic per-node generation counter:

- State.Connect() increments the counter and returns the current
  generation alongside the change list.
- State.Disconnect() accepts the generation from the caller and
  rejects the call if a newer generation exists, making stale
  disconnects from old sessions a no-op.
- serveLongPoll captures the generation at Connect() time and passes
  it to Disconnect() in the deferred cleanup.
- RemoveNode's return value is now checked: if another session already
  owns the batcher slot (reconnect happened), the old session skips
  the grace period entirely.

Update batcher_test.go to track per-node connect generations and
pass them through to Disconnect(), matching production behavior.

Fixes the following test failures:
- server_state_online_after_reconnect_within_grace
- update_history_no_false_offline
- nodestore_correct_after_rapid_reconnect
- rapid_reconnect_peer_never_sees_offline

hscontrol/mapper/batcher_test.go

@@ -39,14 +39,20 @@ type testBatcherWrapper struct {
 	*Batcher
 
 	state *state.State
+
+	// connectGens tracks per-node connect generations so RemoveNode can pass
+	// the correct generation to State.Disconnect(), matching production behavior.
+	connectGens sync.Map // types.NodeID → uint64
 }
 
 func (t *testBatcherWrapper) AddNode(id types.NodeID, c chan<- *tailcfg.MapResponse, version tailcfg.CapabilityVersion, stop func()) error {
 	// Mark node as online in state before AddNode to match production behavior
 	// This ensures the NodeStore has correct online status for change processing
 	if t.state != nil {
-		// Use Connect to properly mark node online in NodeStore but don't send its changes
-		_ = t.state.Connect(id)
+		// Use Connect to properly mark node online in NodeStore and track the
+		// generation so RemoveNode can pass it to Disconnect().
+		_, gen := t.state.Connect(id)
+		t.connectGens.Store(id, gen)
 	}
 
 	// First add the node to the real batcher
@@ -71,8 +77,15 @@ func (t *testBatcherWrapper) RemoveNode(id types.NodeID, c chan<- *tailcfg.MapRe
 	// Mark node as offline in state BEFORE removing from batcher
 	// This ensures the NodeStore has correct offline status when the change is processed
 	if t.state != nil {
-		// Use Disconnect to properly mark node offline in NodeStore but don't send its changes
-		_, _ = t.state.Disconnect(id)
+		var gen uint64
+
+		if v, ok := t.connectGens.LoadAndDelete(id); ok {
+			if g, ok := v.(uint64); ok {
+				gen = g
+			}
+		}
+
+		_, _ = t.state.Disconnect(id, gen)
 	}
 
 	// Send the offline notification that poll.go would normally send