feat(vdisplay): harden keep-alive reconnect — same-client preempt, quit-skips-linger, configurable idle

On-glass testing (Test 2, KWin .116) surfaced that a reconnect within the QUIC idle-timeout
window (~8s) lands on a fresh SECOND display instead of reusing the kept one: the old session
was still Active (not yet Lingering), so the registry's keep-alive reuse (which only matches
Lingering) skipped it and the old session kept streaming to nobody. Three fixes:

#3 Same-client reconnect preempt (the real fix): admission::preempt_same_identity() lists a
   reconnecting client's OWN still-live session(s) (same cert fingerprint); serve_session signals
   their stop + waits the release grace BEFORE acquiring, so the zombie tears down → its display
   lingers → the reconnect REUSES it instead of making a second. Implements the "preempts
   downstream" the admission docs already promised. Independent of the mode_conflict policy; the
   pure core (same_identity_stops) is unit-tested.

#2 Deliberate quit skips linger: a client that deliberately disconnects closes the QUIC connection
   with QUIT_CLOSE_CODE (0x51, shared in core::quic); the host reads the ApplicationClosed reason
   and tears the display down immediately (registry release() gained force_immediate →
   Linger::Immediate; multi-session-safe via the pure lifecycle machine), while a bare disconnect
   still lingers for reconnect. Threaded via a session quit flag → the DisplayLease.
   NativeClient::disconnect_quit() + punktfunk-probe --quit drive it; GameStream (Quit App /
   h_cancel) is a documented follow-up.

#1 Configurable disconnect-detection latency: the QUIC control-connection idle timeout
   (stream_transport, 8s default) is host-tunable via --idle-timeout-ms / PUNKTFUNK_IDLE_TIMEOUT_MS,
   clamped >=1s with a keep-alive that scales to it so a live session never false-closes. Default
   unchanged (8s stays load-bearing for the Windows IDD-push reconnect flow).

Workspace check + 63 core / 215 host / 47 vdisplay tests green; clippy clean.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

crates/punktfunk-core/src/client.rs

@@ -179,6 +179,10 @@ pub struct NativeClient {
     /// Speed-test accumulator, shared with the data-plane pump + control task.
     probe: Arc<Mutex<ProbeState>>,
     shutdown: Arc<AtomicBool>,
+    /// Deliberate-quit flag: [`NativeClient::disconnect_quit`] sets it, so the worker closes the QUIC
+    /// connection with [`crate::quic::QUIT_CLOSE_CODE`] (a user "stop") instead of code 0 — telling the
+    /// host to skip the keep-alive linger. A plain drop leaves it false → an unwanted-disconnect close.
+    quit: Arc<AtomicBool>,
     /// Cumulative count of access units the reassembler gave up on (FEC couldn't recover), mirrored
     /// from the data-plane pump's `Session`. A client video loop watches this for increases to request
     /// a recovery keyframe under infinite GOP — the correct loss trigger, since unrecoverable loss
@@ -331,6 +335,7 @@ impl NativeClient {
         let (ctrl_tx, ctrl_rx) = tokio::sync::mpsc::unbounded_channel::<CtrlRequest>();
         let (ready_tx, ready_rx) = std::sync::mpsc::channel::<Result<Negotiated>>();
         let shutdown = Arc::new(AtomicBool::new(false));
+        let quit = Arc::new(AtomicBool::new(false));
         let mode_slot = Arc::new(std::sync::Mutex::new(mode));
         let probe = Arc::new(Mutex::new(ProbeState::default()));
         let frames_dropped = Arc::new(AtomicU64::new(0));
@@ -338,6 +343,7 @@ impl NativeClient {
 
         let host = host.to_string();
         let shutdown_w = shutdown.clone();
+        let quit_w = quit.clone();
         let mode_slot_w = mode_slot.clone();
         let probe_w = probe.clone();
         let frames_dropped_w = frames_dropped.clone();
@@ -388,6 +394,7 @@ impl NativeClient {
                     ctrl_tx: ctrl_tx_pump,
                     ready_tx,
                     shutdown: shutdown_w,
+                    quit: quit_w,
                     mode_slot: mode_slot_w,
                     probe: probe_w,
                     frames_dropped: frames_dropped_w,
@@ -430,6 +437,7 @@ impl NativeClient {
             ctrl_tx,
             probe,
             shutdown,
+            quit,
             worker: Some(worker),
             frames_dropped,
             hot_tids,
@@ -764,6 +772,15 @@ impl NativeClient {
             .send(rich)
             .map_err(|_| PunktfunkError::Closed)
     }
+
+    /// Signal a **deliberate quit** (a user "stop", not a network drop): the worker closes the QUIC
+    /// connection with [`crate::quic::QUIT_CLOSE_CODE`] instead of code 0, so the host tears the
+    /// session's virtual display down immediately and skips the keep-alive linger. Then requests
+    /// shutdown. A plain `drop` (without this) closes with code 0 → the host lingers for a reconnect.
+    pub fn disconnect_quit(&self) {
+        self.quit.store(true, Ordering::SeqCst);
+        self.shutdown.store(true, Ordering::SeqCst);
+    }
 }
 
 impl Drop for NativeClient {
@@ -802,6 +819,8 @@ struct WorkerArgs {
     ctrl_tx: tokio::sync::mpsc::UnboundedSender<CtrlRequest>,
     ready_tx: std::sync::mpsc::Sender<Result<Negotiated>>,
     shutdown: Arc<AtomicBool>,
+    /// Deliberate-quit flag (see [`NativeClient::quit`]): the worker closes with the quit code if set.
+    quit: Arc<AtomicBool>,
     mode_slot: Arc<std::sync::Mutex<Mode>>,
     probe: Arc<Mutex<ProbeState>>,
     frames_dropped: Arc<AtomicU64>,
@@ -838,6 +857,7 @@ async fn worker_main(args: WorkerArgs) {
         ctrl_tx,
         ready_tx,
         shutdown,
+        quit,
         mode_slot,
         probe,
         frames_dropped,
@@ -1210,5 +1230,12 @@ async fn worker_main(args: WorkerArgs) {
     })
     .await;
 
-    conn.close(0u32.into(), b"client closed");
+    // Deliberate quit (a user "stop") closes with the quit code → the host skips the keep-alive
+    // linger; a plain drop / disconnect closes with 0 → the host lingers so a reconnect can resume.
+    let close_code = if quit.load(Ordering::SeqCst) {
+        crate::quic::QUIT_CLOSE_CODE
+    } else {
+        0
+    };
+    conn.close(close_code.into(), b"client closed");
 }

crates/punktfunk-core/src/quic.rs

@@ -122,6 +122,13 @@ pub const VIDEO_CAP_444: u8 = 0x04;
 /// stage. Purely observability — never changes what the host encodes.
 pub const VIDEO_CAP_HOST_TIMING: u8 = 0x08;
 
+/// QUIC application error code a punktfunk/1 client closes the control connection with on a
+/// **deliberate quit** (a user "stop", not a network drop). The host reads it off the connection's
+/// `ApplicationClosed` reason and tears the session's virtual display down immediately, skipping the
+/// keep-alive linger; any other close reason (idle timeout, reset, a bare code 0) still lingers so a
+/// reconnect can resume. Shared so host + every client agree on the code.
+pub const QUIT_CLOSE_CODE: u32 = 0x51;
+
 /// [`Hello::video_codecs`] bit: the client can decode H.264 / AVC. The GPU-less **software**
 /// encode path (openh264) emits H.264, so a client that wants to stream from a software host MUST
 /// advertise this.
@@ -1743,20 +1750,31 @@ pub mod endpoint {
     /// every `KEEP_ALIVE` keeps the path warm. The interval sits well under `MAX_IDLE` so
     /// several keepalives can be lost back-to-back (a wifi roam, a brief blip) without a false
     /// close, while a genuinely dead peer is still detected within `MAX_IDLE`.
+    /// The default control-connection idle timeout (disconnect-detection latency). A vanished client
+    /// is declared dead within this window — the Windows IDD-push path needs it short so a RECONNECT
+    /// recreates a fresh virtual monitor instead of joining the still-lingering old session; the Linux
+    /// path pairs it with the same-client reconnect preempt. Host-tunable via `server_with_identity_idle`.
+    pub const DEFAULT_IDLE_TIMEOUT: std::time::Duration = std::time::Duration::from_secs(8);
+
     fn stream_transport() -> Arc<quinn::TransportConfig> {
+        stream_transport_idle(DEFAULT_IDLE_TIMEOUT)
+    }
+
+    /// Transport config with a caller-chosen idle timeout (disconnect-detection latency). The
+    /// keep-alive interval tracks it at half the idle window (capped at the default 4s), so a live
+    /// path is PINGed at least twice per window and a single lost PING (wifi roam / brief blip) won't
+    /// false-close. `idle` is clamped to a ≥1s floor so a misconfigured tiny value can't tear live
+    /// sessions down. Active sessions are unaffected either way: video keeps the connection live and
+    /// the keep-alive holds it open through quiet control periods.
+    fn stream_transport_idle(idle: std::time::Duration) -> Arc<quinn::TransportConfig> {
         use std::time::Duration;
-        // 8s idle (was 20s): a vanished client is declared dead within 8s instead of 20, so its
-        // session tears down promptly — which the Windows IDD-push path needs so a RECONNECT recreates
-        // a fresh virtual monitor (a reused monitor's IddCx swap-chain dies) instead of joining the
-        // still-lingering old session. Active sessions are unaffected: video keeps the connection live,
-        // and the 4s keep-alive holds it open through quiet control periods.
-        const MAX_IDLE: Duration = Duration::from_secs(8);
-        const KEEP_ALIVE: Duration = Duration::from_secs(4);
+        let idle = idle.max(Duration::from_secs(1));
+        let keep_alive = (idle / 2).min(Duration::from_secs(4));
         let mut t = quinn::TransportConfig::default();
         t.max_idle_timeout(Some(
-            quinn::IdleTimeout::try_from(MAX_IDLE).expect("8s is a valid QUIC idle timeout"),
+            quinn::IdleTimeout::try_from(idle).expect("clamped idle timeout is a valid QUIC value"),
         ));
-        t.keep_alive_interval(Some(KEEP_ALIVE));
+        t.keep_alive_interval(Some(keep_alive));
         Arc::new(t)
     }
 
@@ -1767,23 +1785,36 @@ pub mod endpoint {
             .map_err(|e| anyhow_result::Error::msg(format!("self-signed cert: {e}")))?;
         let cert_der = rustls::pki_types::CertificateDer::from(cert.cert);
         let key_der = rustls::pki_types::PrivatePkcs8KeyDer::from(cert.key_pair.serialize_der());
-        server_from_der(cert_der, key_der.into(), addr)
+        server_from_der(cert_der, key_der.into(), addr, DEFAULT_IDLE_TIMEOUT)
     }
 
     /// Server endpoint from a persisted PEM identity (certificate + PKCS#8 private key) —
     /// the host's long-lived self-signed cert, so the fingerprint clients pin is stable
-    /// across restarts.
+    /// across restarts. Uses the [`DEFAULT_IDLE_TIMEOUT`]; see [`server_with_identity_idle`] to tune it.
     pub fn server_with_identity(
         addr: std::net::SocketAddr,
         cert_pem: &str,
         key_pem: &str,
+    ) -> anyhow_result::Result<quinn::Endpoint> {
+        server_with_identity_idle(addr, cert_pem, key_pem, DEFAULT_IDLE_TIMEOUT)
+    }
+
+    /// Like [`server_with_identity`] but with a host-chosen control-connection idle timeout — the
+    /// disconnect-detection latency (how long a vanished client takes to be declared dead). Shorter =
+    /// faster teardown/linger of a dropped session; the value is clamped to a ≥1s floor and its
+    /// keep-alive scales with it so a live session never false-closes.
+    pub fn server_with_identity_idle(
+        addr: std::net::SocketAddr,
+        cert_pem: &str,
+        key_pem: &str,
+        idle: std::time::Duration,
     ) -> anyhow_result::Result<quinn::Endpoint> {
         use rustls::pki_types::pem::PemObject;
         let cert_der = rustls::pki_types::CertificateDer::from_pem_slice(cert_pem.as_bytes())
             .map_err(|e| anyhow_result::Error::msg(format!("cert pem: {e}")))?;
         let key_der = rustls::pki_types::PrivateKeyDer::from_pem_slice(key_pem.as_bytes())
             .map_err(|e| anyhow_result::Error::msg(format!("key pem: {e}")))?;
-        server_from_der(cert_der, key_der, addr)
+        server_from_der(cert_der, key_der, addr, idle)
     }
 
     /// Fixed ALPN for the punktfunk/1 QUIC handshake. Pinning it rejects a cross-protocol peer at the
@@ -1796,6 +1827,7 @@ pub mod endpoint {
         cert_der: rustls::pki_types::CertificateDer<'static>,
         key_der: rustls::pki_types::PrivateKeyDer<'static>,
         addr: std::net::SocketAddr,
+        idle: std::time::Duration,
     ) -> anyhow_result::Result<quinn::Endpoint> {
         let _ = rustls::crypto::ring::default_provider().install_default();
         // Client auth is OFFERED but optional: a client that presents its self-signed
@@ -1810,7 +1842,7 @@ pub mod endpoint {
         let quic_cfg = quinn::crypto::rustls::QuicServerConfig::try_from(rustls_cfg)
             .map_err(|e| anyhow_result::Error::msg(format!("quic server config: {e}")))?;
         let mut server_config = quinn::ServerConfig::with_crypto(Arc::new(quic_cfg));
-        server_config.transport_config(stream_transport()); // keep-alive — see stream_transport
+        server_config.transport_config(stream_transport_idle(idle)); // keep-alive — see stream_transport_idle
         Ok(quinn::Endpoint::server(server_config, addr)?)
     }