zudp 0.3.2

Zero-overhead UDP protocol with optional reliability, fragmentation, and relay support
Documentation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
use std::{
    collections::HashMap,
    net::SocketAddr,
    sync::{
        Arc,
        atomic::{AtomicU64, Ordering},
    },
    time::Duration,
};

use bytes::{Bytes, BytesMut};
use dashmap::DashMap;
use parking_lot::{Mutex, RwLock};
use tokio::{sync::mpsc, task::JoinSet, time};

use crate::{
    Config,
    frag::FragAssembler,
    frame::Frame,
    peer::{Inbound, PeerState, RecvState},
    socket::RawSocket,
};

/// Runtime-mutable relay access control, updateable without socket rebind.
pub struct RelayPolicy {
    /// IP allowlist for relay frame sources.  Empty = allow all.
    pub allowlist: Vec<std::net::IpAddr>,
    /// Maximum entries in the relay routing table.
    pub max_entries: usize,
}

/// Maximum concurrent NACK retransmit tasks across the engine.
///
/// Each task handles one NACK batch (one peer, one burst of missing seqs).  Capping
/// prevents memory/CPU runaway when a peer floods NACKs faster than retransmits drain.
const MAX_RETRANSMIT_TASKS: usize = 32;

/// Capacity of the bounded inbound channel between the engine task and the socket handle.
///
/// At 1 400 B/frame and 8 192 slots this is ~11 MB peak backpressure before the engine
/// starts dropping frames.  The engine uses `try_send` so the recv loop is never stalled.
const INBOUND_CAP: usize = 8_192;

/// State shared between the engine task and the [`crate::ZudpSocket`] / [`crate::ZudpConn`] handles.
pub(crate) struct EngineInner {
    /// Current socket; replaced atomically on rebind.  Use [`get_socket`] to obtain a cheap clone.
    pub socket: RwLock<RawSocket>,
    pub peers: DashMap<SocketAddr, Arc<PeerState>>,
    /// Reverse index: remote's `my_session_id` → peer, for detecting address migrations.
    pub sessions: DashMap<u64, Arc<PeerState>>,
    pub config: Arc<Config>,
    /// Noise keypair for the `security` feature; `None` when security is disabled.
    #[cfg(feature = "security")]
    pub keypair: Option<crate::security::Keypair>,
    /// Set by the engine task before it exits; callers read this to distinguish a
    /// clean shutdown from an unexpected crash (rebind failure, etc.).
    pub shutdown_reason: Mutex<Option<String>>,
    pub dropped_rate_limited: AtomicU64,
    pub dropped_peer_cap: AtomicU64,
    pub dropped_relay_blocked: AtomicU64,
    pub dropped_relay_cap: AtomicU64,
    /// Runtime-mutable relay access policy (allowlist + table cap).
    pub relay_policy: RwLock<RelayPolicy>,
}

impl EngineInner {
    /// Cheap handle to the current socket (clones the inner `Arc<UdpSocket>`).
    pub fn get_socket(&self) -> RawSocket {
        self.socket.read().clone()
    }

    pub fn get_or_create_peer(self: &Arc<Self>, addr: SocketAddr) -> Arc<PeerState> {
        if let Some(p) = self.peers.get(&addr) {
            return p.value().clone();
        }
        if self.peers.len() >= self.config.max_peers {
            let evict_addr = self.peers.iter()
                .min_by_key(|r| *r.value().last_seen.lock())
                .map(|r| *r.key());
            if let Some(evict) = evict_addr {
                tracing::warn!(
                    target: "zudp::engine",
                    evicted = %evict,
                    new = %addr,
                    "peer table full — evicting least-recently-seen peer"
                );
                self.peers.remove(&evict);
            }
            self.dropped_peer_cap.fetch_add(1, Ordering::Relaxed);
        }
        self.peers.entry(addr)
            .or_insert_with(|| Arc::new(PeerState::new(addr)))
            .value()
            .clone()
    }
}

/// Spawn the engine background task and return the inbound receiver.
///
/// `shutdown_rx` resolves when the caller drops the paired `oneshot::Sender`; the engine
/// then drains in-flight retransmits and exits cleanly.
///
/// Each item on the returned channel is `(payload_bytes, sender_addr, stream_id)`.
pub(crate) fn spawn(
    inner: Arc<EngineInner>,
    shutdown_rx: tokio::sync::oneshot::Receiver<()>,
) -> mpsc::Receiver<(Bytes, SocketAddr, u16)> {
    let (inbound_tx, inbound_rx) = mpsc::channel(INBOUND_CAP);
    tokio::spawn(run(inner, inbound_tx, shutdown_rx));
    inbound_rx
}

async fn run(
    inner: Arc<EngineInner>,
    inbound_tx: mpsc::Sender<(Bytes, SocketAddr, u16)>,
    mut shutdown_rx: tokio::sync::oneshot::Receiver<()>,
) {
    // Keyed by (peer_addr, stream_id) so each stream has independent reorder + reassembly.
    let mut recv_states: HashMap<(SocketAddr, u16), (RecvState, FragAssembler)> = HashMap::new();
    // Stateful relay table: maps server_addr → (client_addr, last_seen_gen).
    // `last_seen_gen` is the relay_gen value at the last Frame::Relay from that client.
    // Entries pruned when (relay_gen - last_seen_gen) >= relay_ttl_ticks.
    // No Instant/syscall in the hot path — generation counter is a u32 store.
    let mut relay_table: HashMap<SocketAddr, (SocketAddr, u32)> = HashMap::new();
    // Monotonic counter incremented once per background tick.  Wrapping arithmetic
    // is intentional: correct for the ~170-year lifetime of a u32 at 240 ticks/5 min.
    let mut relay_gen: u32 = 0;
    // TTL expressed as a tick delta: 5 minutes / tick_interval.
    let relay_ttl_ticks = {
        let tick_secs = inner.config.keepalive_interval.as_secs_f64() / 4.0;
        (300.0_f64 / tick_secs).ceil() as u32
    };

    let keepalive_interval = inner.config.keepalive_interval;
    let sent_prune_age = inner.config.sent_prune_age;

    let mut rate_limiter = crate::rate::RateLimiter::new(
        inner.config.max_pps_per_ip,
        (inner.config.max_pps_per_ip / 5.0).max(1.0),
    );

    // Bounded pool for NACK retransmit tasks.  Each task may sleep for CC pacing, so
    // they are spawned off the recv loop to avoid stalling it — but the pool is capped
    // to prevent unbounded task accumulation under heavy loss or adversarial NACKs.
    let mut retransmit_tasks: JoinSet<()> = JoinSet::new();

    // Local recv socket; replaced in-place when the socket is rebound.
    let mut recv_sock = inner.get_socket();

    let mut ticker = time::interval(keepalive_interval / 4);

    loop {
        // Reap completed retransmit tasks without blocking.
        while retransmit_tasks.try_join_next().is_some() {}

        tokio::select! {
            biased;

            _ = &mut shutdown_rx => {
                *inner.shutdown_reason.lock() = Some("shutdown requested".into());
                tracing::info!(target: "zudp::engine", "graceful shutdown — draining retransmits");
                break;
            }

            result = recv_sock.recv_from() => {
                match result {
                    Ok((data, from)) => {
                        if inner.config.max_pps_per_ip > 0.0
                            && !rate_limiter.allow(from.ip())
                        {
                            tracing::trace!(target: "zudp::engine", %from, "rate limited");
                            inner.dropped_rate_limited.fetch_add(1, Ordering::Relaxed);
                            continue;
                        }
                        handle_incoming(
                            data, from, &inner, &inbound_tx,
                            &mut recv_states, &mut relay_table, relay_gen,
                            &mut retransmit_tasks,
                        ).await;
                    }
                    Err(e) => {
                        tracing::warn!(target: "zudp::engine", "recv error: {e} — attempting socket rebind");
                        match rebind_socket(&inner).await {
                            Ok(new_sock) => {
                                recv_sock = new_sock;
                                // Ping all known peers so they can migrate our address.
                                ping_all_peers(&inner, &recv_sock).await;
                            }
                            Err(bind_err) => {
                                let reason = format!("socket rebind failed: {bind_err}");
                                tracing::error!(target: "zudp::engine", "{reason}");
                                *inner.shutdown_reason.lock() = Some(reason);
                                break;
                            }
                        }
                    }
                }
            }

            _ = ticker.tick() => {
                run_background(
                    &inner, &mut recv_states, &mut relay_table, &mut relay_gen,
                    relay_ttl_ticks, &mut rate_limiter, keepalive_interval, sent_prune_age,
                );
            }
        }
    }

    // Drain in-flight retransmits before the socket closes.
    retransmit_tasks.shutdown().await;
    tracing::info!(target: "zudp::engine", "engine task exited");
}

/// Try to bind a new socket on the same local address (same port if available, else port 0).
///
/// On success, replaces `inner.socket` so future `send_to` calls use the new interface.
async fn rebind_socket(inner: &Arc<EngineInner>) -> Result<RawSocket, crate::Error> {
    let old_local = inner.get_socket().local_addr()?;
    let new_sock = match RawSocket::bind(old_local).await {
        Ok(s) => s,
        Err(_) => {
            let any = SocketAddr::new(old_local.ip(), 0);
            RawSocket::bind(any).await?
        }
    };
    let new_addr = new_sock.local_addr()?;
    *inner.socket.write() = new_sock.clone();
    tracing::info!(target: "zudp::engine", old = %old_local, new = %new_addr, "socket rebound");
    Ok(new_sock)
}

/// Send a Ping carrying our session ID to every known peer so they update our address.
async fn ping_all_peers(inner: &Arc<EngineInner>, sock: &RawSocket) {
    let now_ms = now_micros();
    let peers: Vec<Arc<PeerState>> = inner.peers.iter().map(|r| r.value().clone()).collect();
    for peer in peers {
        let ping = secure_frame(&peer, Frame::Ping { echo: now_ms, session_id: peer.my_session_id });
        if let Err(e) = sock.send_to(&ping, peer.addr()).await {
            tracing::warn!(target: "zudp::engine", peer = %peer.addr(), "post-rebind ping failed: {e}");
        }
    }
}

#[allow(clippy::too_many_arguments)]
async fn handle_incoming(
    data: BytesMut,
    from: SocketAddr,
    inner: &Arc<EngineInner>,
    inbound_tx: &mpsc::Sender<(Bytes, SocketAddr, u16)>,
    recv_states: &mut HashMap<(SocketAddr, u16), (RecvState, FragAssembler)>,
    relay_table: &mut HashMap<SocketAddr, (SocketAddr, u32)>,
    relay_gen: u32,
    retransmit_tasks: &mut JoinSet<()>,
) {
    // Stateful relay pass-through: if `from` is a server we're relaying for, forward its raw
    // reply bytes directly to the mapped client without decoding.  This is checked before
    // Frame::decode so that encrypted and opaque frames are forwarded transparently.
    if let Some(&(client_addr, _)) = relay_table.get(&from) {
        tracing::trace!(target: "zudp::engine", server = %from, client = %client_addr, "relay ← server → client");
        if let Err(e) = inner.get_socket().send_to(&data, client_addr).await {
            tracing::warn!(target: "zudp::engine", client = %client_addr, "relay forward to client failed: {e}");
        }
        return;
    }

    let frame = match Frame::decode(data) {
        Ok(f) => f,
        Err(e) => {
            tracing::warn!(target: "zudp::engine", peer = %from, "bad frame: {e}");
            return;
        }
    };

    let known_peer = inner.peers.get(&from).map(|r| r.value().clone());
    if let Some(peer) = &known_peer {
        peer.mark_seen();
    }

    #[cfg(feature = "security")]
    {
        match frame {
            Frame::Handshake { payload } => {
                handle_handshake(payload, from, inner).await;
            }
            Frame::Secure { nonce, ciphertext } => {
                let Some(peer) = known_peer else {
                    tracing::warn!(target: "zudp::engine", peer = %from, "secure frame from unknown peer");
                    return;
                };
                let Some(channel) = peer.channel.get() else {
                    tracing::warn!(target: "zudp::engine", peer = %from, "secure frame but no channel yet");
                    return;
                };
                let plain = match channel.decrypt(nonce, &ciphertext) {
                    Ok(p) => p,
                    Err(e) => {
                        tracing::warn!(target: "zudp::engine", peer = %from, "decrypt failed: {e}");
                        return;
                    }
                };
                let inner_frame = match Frame::decode(BytesMut::from(plain.as_slice())) {
                    Ok(f) => f,
                    Err(e) => {
                        tracing::warn!(target: "zudp::engine", peer = %from, "inner frame decode failed: {e}");
                        return;
                    }
                };
                dispatch_frame(inner_frame, from, inner, inbound_tx, recv_states, relay_table, relay_gen, retransmit_tasks).await;
            }
            frame => dispatch_frame(frame, from, inner, inbound_tx, recv_states, relay_table, relay_gen, retransmit_tasks).await,
        }
    }

    #[cfg(not(feature = "security"))]
    dispatch_frame(frame, from, inner, inbound_tx, recv_states, relay_table, relay_gen, retransmit_tasks).await;
}

// All parameters are genuinely distinct state that the frame dispatch needs to mutate or read;
// grouping them into a context struct would add complexity without architectural benefit.
#[allow(clippy::too_many_arguments, clippy::too_many_lines)]
async fn dispatch_frame(
    frame: Frame,
    from: SocketAddr,
    inner: &Arc<EngineInner>,
    inbound_tx: &mpsc::Sender<(Bytes, SocketAddr, u16)>,
    recv_states: &mut HashMap<(SocketAddr, u16), (RecvState, FragAssembler)>,
    relay_table: &mut HashMap<SocketAddr, (SocketAddr, u32)>,
    relay_gen: u32,
    retransmit_tasks: &mut JoinSet<()>,
) {
    match frame {
        Frame::Datagram(payload) => {
            let peer = inner.get_or_create_peer(from);
            peer.rx_bytes.fetch_add(payload.len() as u64, std::sync::atomic::Ordering::Relaxed);
            if inbound_tx.try_send((payload, from, 0)).is_err() {
                tracing::warn!(target: "zudp::engine", %from, "inbound channel full — dropping datagram");
            }
        }

        Frame::Stream {
            seq,
            stream_id,
            payload,
        } => {
            let (recv, _frag) = recv_states
                .entry((from, stream_id))
                .or_insert_with(|| (RecvState::new(), FragAssembler::new()));

            let (ready, nack_seqs) = recv.ingest(seq, Inbound::Data(payload));
            send_nack_if_needed(nack_seqs, stream_id, from, inner).await;
            let peer = inner.get_or_create_peer(from);
            for item in ready {
                if let Inbound::Data(bytes) = item {
                    peer.rx_bytes.fetch_add(bytes.len() as u64, std::sync::atomic::Ordering::Relaxed);
                    if inbound_tx.try_send((bytes, from, stream_id)).is_err() {
                        tracing::warn!(target: "zudp::engine", %from, stream_id, "inbound channel full — dropping stream frame");
                    }
                }
            }
        }

        Frame::Nack { stream_id, seqs } => {
            if let Some(peer) = inner.peers.get(&from).map(|r| r.value().clone()) {
                let to_retransmit = peer.frames_for_retransmit(stream_id, &seqs);
                if !to_retransmit.is_empty() {
                    if retransmit_tasks.len() >= MAX_RETRANSMIT_TASKS {
                        tracing::warn!(
                            target: "zudp::engine",
                            peer = %from, stream_id,
                            cap = MAX_RETRANSMIT_TASKS,
                            "retransmit pool full — dropping NACK batch"
                        );
                    } else {
                        // Spawn so the engine recv loop isn't stalled by CC pacing sleeps.
                        let sock = inner.get_socket();
                        retransmit_tasks.spawn(async move {
                            for (seq, plain_frame) in to_retransmit {
                                let wire = encrypt_or_plain(&peer, &plain_frame);
                                // Extract delay before awaiting — MutexGuard is not Send.
                                let delay = peer.cc.lock().consume(wire.len());
                                if let Some(d) = delay {
                                    tokio::time::sleep(d).await;
                                }
                                tracing::debug!(
                                    target: "zudp::engine",
                                    peer = %from, seq, stream_id,
                                    "retransmit on NACK"
                                );
                                if let Err(e) = sock.send_to(&wire, from).await {
                                    tracing::warn!(
                                        target: "zudp::engine",
                                        peer = %from,
                                        "retransmit failed: {e}"
                                    );
                                }
                            }
                        });
                    }
                }
            }
        }

        Frame::Ping { echo, session_id } => {
            let peer = find_or_migrate_peer(from, session_id, inner, recv_states);
            let pong = secure_frame(&peer, Frame::Pong { echo, session_id: peer.my_session_id });
            if let Err(e) = inner.get_socket().send_to(&pong, from).await {
                tracing::warn!(target: "zudp::engine", peer = %from, "pong failed: {e}");
            }
        }

        Frame::Pong { echo, session_id } => {
            if let Some(peer) = inner.peers.get(&from).map(|r| r.value().clone()) {
                if peer.store_their_session_id(session_id) {
                    inner.sessions.insert(session_id, peer.clone());
                }
                // echo carries a µs timestamp; RTT is the round-trip in microseconds.
                let now_us = now_micros();
                if echo <= now_us {
                    let rtt_us = now_us - echo;
                    let pacing_rate = {
                        let mut cc = peer.cc.lock();
                        cc.on_rtt_sample(rtt_us);
                        cc.pacing_rate() as u64
                    };
                    tracing::debug!(
                        target: "zudp::engine",
                        peer = %from,
                        rtt_us,
                        pacing_rate,
                        "RTT sample"
                    );
                }
            }
        }

        Frame::Fragment {
            msg_id,
            frag_idx,
            frag_total,
            seq,
            stream_id,
            payload,
        } => {
            let (recv, frag_assembler) = recv_states
                .entry((from, stream_id))
                .or_insert_with(|| (RecvState::new(), FragAssembler::new()));

            let inbound = Inbound::Fragment {
                msg_id,
                frag_idx,
                frag_total,
                data: payload,
            };
            let (ready, nack_seqs) = recv.ingest(seq, inbound);
            send_nack_if_needed(nack_seqs, stream_id, from, inner).await;

            let peer = inner.get_or_create_peer(from);
            for item in ready {
                if let Inbound::Fragment {
                    msg_id,
                    frag_idx,
                    frag_total,
                    data,
                } = item
                    && let Some(complete) =
                        frag_assembler.insert(msg_id, frag_idx, frag_total, data)
                {
                    peer.rx_bytes.fetch_add(complete.len() as u64, std::sync::atomic::Ordering::Relaxed);
                    if inbound_tx.try_send((complete, from, stream_id)).is_err() {
                        tracing::warn!(target: "zudp::engine", %from, stream_id, "inbound channel full — dropping reassembled fragment");
                    }
                }
            }
        }

        Frame::Relay {
            dest,
            inner: payload,
        } => {
            // Read relay policy atomically — policy may be updated at runtime.
            let (policy_allowlist, policy_max) = {
                let p = inner.relay_policy.read();
                (p.allowlist.clone(), p.max_entries)
            };
            // Allowlist check: if non-empty, only listed IPs may use this node as a relay.
            if !policy_allowlist.is_empty() && !policy_allowlist.contains(&from.ip()) {
                tracing::warn!(target: "zudp::engine", %from, "relay denied — not in allowlist");
                inner.dropped_relay_blocked.fetch_add(1, Ordering::Relaxed);
                return;
            }
            // Cap relay table size to prevent unbounded memory growth under abuse.
            if !relay_table.contains_key(&dest) && relay_table.len() >= policy_max {
                tracing::warn!(target: "zudp::engine", "relay table full — dropping entry for {dest}");
                inner.dropped_relay_cap.fetch_add(1, Ordering::Relaxed);
                return;
            }
            // Record client→server mapping so replies from `dest` can be routed back to `from`.
            // Upsert refreshes the generation counter — zero syscall cost.
            relay_table
                .entry(dest)
                .and_modify(|(stored_from, last_gen)| {
                    *stored_from = from;
                    *last_gen = relay_gen;
                })
                .or_insert((from, relay_gen));
            tracing::trace!(target: "zudp::engine", client = %from, server = %dest, "relay client → server");
            if let Err(e) = inner.get_socket().send_to(&payload, dest).await {
                tracing::warn!(target: "zudp::engine", %dest, "relay forward failed: {e}");
            }
        }

        Frame::MtuProbe { probe_id, .. } => {
            // Reflect probe_id in a tiny MtuAck — always gets through regardless of path MTU.
            let ack_frame = Frame::MtuAck { probe_id };
            let ack = if let Some(peer) = inner.peers.get(&from).map(|r| r.value().clone()) {
                secure_frame(&peer, ack_frame)
            } else {
                ack_frame.encode()
            };
            if let Err(e) = inner.get_socket().send_to(&ack, from).await {
                tracing::warn!(target: "zudp::engine", peer = %from, "mtu ack send failed: {e}");
            }
        }

        Frame::MtuAck { probe_id } => {
            if let Some(peer) = inner.peers.get(&from).map(|r| r.value().clone())
                && let Some(tx) = peer.probe_acks.lock().remove(&probe_id)
            {
                let _ = tx.send(());
            }
        }

        Frame::Probe { .. } | Frame::Beacon { .. } => {}

        // Handshake and Secure are handled before dispatch_frame is called.
        #[cfg(feature = "security")]
        Frame::Handshake { .. } | Frame::Secure { .. } => {}
    }
}

/// Locate the peer for an incoming Ping, handling first contact, familiar address, and migration.
///
/// Migration: if `session_id` matches a known peer but `from` differs, the peer's address is
/// updated, the `peers` map is re-keyed, and all `recv_states` entries are moved to the new key.
fn find_or_migrate_peer(
    from: SocketAddr,
    session_id: u64,
    inner: &Arc<EngineInner>,
    recv_states: &mut HashMap<(SocketAddr, u16), (RecvState, FragAssembler)>,
) -> Arc<PeerState> {
    // Fast path: familiar address.
    if let Some(peer) = inner.peers.get(&from).map(|r| r.value().clone()) {
        if peer.store_their_session_id(session_id) {
            inner.sessions.insert(session_id, peer.clone());
        }
        return peer;
    }

    // Migration: known session but new address.
    if let Some(peer) = inner.sessions.get(&session_id).map(|r| r.value().clone()) {
        let old_addr = peer.addr();
        if old_addr != from {
            tracing::info!(
                target: "zudp::engine",
                old = %old_addr,
                new = %from,
                "peer address migrated"
            );
            peer.migrate_to(from);
            inner.peers.remove(&old_addr);
            inner.peers.insert(from, peer.clone());
            // Move all recv_states from old key to new key.
            let stream_ids: Vec<u16> = recv_states
                .keys()
                .filter(|(a, _)| *a == old_addr)
                .map(|(_, s)| *s)
                .collect();
            for sid in stream_ids {
                if let Some(state) = recv_states.remove(&(old_addr, sid)) {
                    recv_states.insert((from, sid), state);
                }
            }
            // Path has changed — reset discovered MTU and re-probe.
            peer.set_effective_mtu(0);
            spawn_mtu_probe(inner, from);
        }
        return peer;
    }

    // First contact: create peer, register session, and discover path MTU.
    let peer = inner.get_or_create_peer(from);
    peer.store_their_session_id(session_id);
    inner.sessions.insert(session_id, peer.clone());
    spawn_mtu_probe(inner, from);
    peer
}

/// Fire-and-forget PLPMTUD probe toward `peer_addr`.
///
/// Spawned on first contact and on every migration so the server always knows the usable path
/// MTU — not just the client side (which probes inside `connect()`).
fn spawn_mtu_probe(inner: &Arc<EngineInner>, peer_addr: SocketAddr) {
    let inner = inner.clone();
    let initial_mtu = inner.config.mtu;
    let _mtu_probe = tokio::spawn(async move {
        let discovered = crate::mtu::probe(&inner, peer_addr, initial_mtu).await;
        if let Some(p) = inner.peers.get(&peer_addr) {
            p.set_effective_mtu(discovered);
        }
        tracing::info!(
            target: "zudp::mtu",
            peer = %peer_addr,
            mtu = discovered,
            "path MTU discovered"
        );
    });
}

/// Encode `frame` and encrypt it for `peer` if an established Noise channel exists.
///
/// Falls back to plaintext when called before the handshake completes or when the
/// `security` feature is disabled.  All post-handshake control frames (Ping, Pong,
/// Nack, `MtuAck`) go through this so passive observers cannot read session metadata.
fn secure_frame(peer: &PeerState, frame: Frame) -> Bytes {
    let plain = frame.encode();
    encrypt_or_plain(peer, &plain)
}

/// If the peer has an established Noise channel, re-encrypt the plain frame; otherwise send as-is.
///
/// Used for NACK-triggered retransmits: each retransmit gets a fresh nonce to avoid reuse.
#[allow(unused_variables)]
fn encrypt_or_plain(peer: &PeerState, plain: &Bytes) -> Bytes {
    #[cfg(feature = "security")]
    if let Some(channel) = peer.channel.get() {
        match channel.encrypt(plain) {
            Ok((nonce, ct)) => {
                return Frame::Secure {
                    nonce,
                    ciphertext: Bytes::from(ct),
                }
                .encode();
            }
            Err(e) => {
                tracing::warn!(target: "zudp::engine", "retransmit encrypt failed: {e}");
            }
        }
    }
    plain.clone()
}

async fn send_nack_if_needed(
    nack_seqs: Vec<u64>,
    stream_id: u16,
    to: SocketAddr,
    inner: &Arc<EngineInner>,
) {
    if nack_seqs.is_empty() {
        return;
    }
    tracing::debug!(target: "zudp::engine", peer = %to, count = nack_seqs.len(), stream_id, "sending NACK");
    let nack = Frame::Nack { stream_id, seqs: nack_seqs };
    let wire = if let Some(peer) = inner.peers.get(&to).map(|r| r.value().clone()) {
        secure_frame(&peer, nack)
    } else {
        nack.encode()
    };
    if let Err(e) = inner.get_socket().send_to(&wire, to).await {
        tracing::warn!(target: "zudp::engine", peer = %to, "nack send failed: {e}");
    }
}

#[allow(clippy::too_many_arguments)]
fn run_background(
    inner: &Arc<EngineInner>,
    recv_states: &mut HashMap<(SocketAddr, u16), (RecvState, FragAssembler)>,
    relay_table: &mut HashMap<SocketAddr, (SocketAddr, u32)>,
    relay_gen: &mut u32,
    relay_ttl_ticks: u32,
    rate_limiter: &mut crate::rate::RateLimiter,
    keepalive_interval: Duration,
    sent_prune_age: Duration,
) {
    let keepalive_threshold = keepalive_interval.as_secs();
    let now_ms = now_micros();

    let peers: Vec<Arc<PeerState>> = inner.peers.iter().map(|r| r.value().clone()).collect();

    for peer in peers {
        if peer.secs_since_sent() >= keepalive_threshold {
            // Spawn each keepalive independently so sequential `send_to` calls
            // don't stall the recv loop when many peers need pinging at once.
            let sock = inner.get_socket();
            let peer_clone = peer.clone();
            tokio::spawn(async move {
                let addr = peer_clone.addr();
                let ping = secure_frame(&peer_clone, Frame::Ping { echo: now_ms, session_id: peer_clone.my_session_id });
                if let Err(e) = sock.send_to(&ping, addr).await {
                    tracing::warn!(target: "zudp::engine", peer = %addr, "keepalive failed: {e}");
                }
                tracing::trace!(target: "zudp::engine", peer = %addr, "sent keepalive ping");
            });
        }
        peer.prune_sent(sent_prune_age);
    }

    for (_, frag_assembler) in recv_states.values_mut() {
        frag_assembler.prune(sent_prune_age);
    }

    // Advance the relay generation counter and prune stale entries.
    // Wrapping arithmetic is intentional — correct over the ~170-year u32 lifetime.
    *relay_gen = relay_gen.wrapping_add(1);
    let current_gen = *relay_gen;
    relay_table.retain(|_, (_, last_seen_gen)| {
        current_gen.wrapping_sub(*last_seen_gen) < relay_ttl_ticks
    });

    rate_limiter.tick_prune();
}

fn now_micros() -> u64 {
    std::time::SystemTime::now()
        .duration_since(std::time::UNIX_EPOCH)
        .map_or(0u64, |d| u64::try_from(d.as_micros()).unwrap_or(u64::MAX))
}

/// Drive a Noise XX handshake step on behalf of the engine.
///
/// - If no handshake exists for `from` yet, creates a responder state (server role).
/// - If a handshake already exists (initiator stored it before sending msg1), reads the
///   incoming message and, if it's the initiator's turn, writes the next message.
/// - When both sides finish all three messages, transitions to `StatelessTransportState`.
/// - If `inner.config.remote_key` is set, the remote's static key is verified after the
///   handshake finishes.  A mismatch silently aborts — no channel is established and
///   `connect()` will time out on the other side.
#[cfg(feature = "security")]
async fn handle_handshake(payload: Bytes, from: SocketAddr, inner: &Arc<EngineInner>) {
    use crate::security::{SecureChannel, build_responder};

    let Some(kp) = inner.keypair.as_ref() else {
        tracing::debug!(target: "zudp::engine", peer = %from, "handshake ignored — security not configured");
        return;
    };

    let peer = inner.get_or_create_peer(from);

    // Initialise as responder if this peer's handshake hasn't started yet.
    {
        let mut guard = peer.handshake.lock();
        if guard.is_none() {
            match build_responder(kp) {
                Ok(hs) => *guard = Some(hs),
                Err(e) => {
                    tracing::error!(target: "zudp::engine", peer = %from, "build_responder: {e}");
                    return;
                }
            }
        }
    }

    // Process the incoming message; determine if we need to reply and/or finish.
    // Returns (optional reply, optional finished handshake, key_mismatch).
    let outcome = 'step: {
        let mut guard = peer.handshake.lock();
        let hs = guard.as_mut().expect("initialised above");

        let mut rbuf = vec![0u8; 1024];
        if let Err(e) = hs.read_message(&payload, &mut rbuf) {
            tracing::warn!(target: "zudp::engine", peer = %from, "handshake read: {e}");
            return;
        }

        // After reading, check if the handshake is done (responder finishes on msg3).
        let finished_after_read = hs.is_handshake_finished();

        if finished_after_read {
            // Key pinning check — must happen before consuming state.
            if !check_pinned_key(hs, inner, from) {
                let _ = guard.take();
                break 'step (None::<Bytes>, None, true);
            }
            let hs_owned = guard.take().unwrap();
            break 'step (None, Some(hs_owned), false);
        }

        // Still our turn to write (msg2 for responder, msg3 for initiator).
        let mut wbuf = vec![0u8; 1024];
        let n = match hs.write_message(&[], &mut wbuf) {
            Ok(n) => n,
            Err(e) => {
                tracing::warn!(target: "zudp::engine", peer = %from, "handshake write: {e}");
                return;
            }
        };
        wbuf.truncate(n);
        let reply = Bytes::from(wbuf);

        // After writing, check again (initiator finishes on msg3 write).
        let finished_after_write = hs.is_handshake_finished();
        if finished_after_write {
            // Key pinning check — must happen before consuming state.
            if !check_pinned_key(hs, inner, from) {
                let _ = guard.take();
                // Don't send msg3 either — abort silently.
                break 'step (None, None, true);
            }
        }

        let hs_owned = if finished_after_write { guard.take() } else { None };
        (Some(reply), hs_owned, false)
    }; // handshake lock released

    let (reply, finished_hs, key_mismatch) = outcome;

    if key_mismatch {
        tracing::warn!(target: "zudp::security", peer = %from, "handshake aborted — key mismatch");
        return;
    }

    if let Some(reply_bytes) = reply {
        let wire = Frame::Handshake { payload: reply_bytes }.encode();
        if let Err(e) = inner.get_socket().send_to(&wire, from).await {
            tracing::warn!(target: "zudp::engine", peer = %from, "handshake send: {e}");
        }
    }

    if let Some(hs) = finished_hs {
        // Extract and persist the remote static key before consuming the handshake state.
        if let Some(remote) = hs.get_remote_static() {
            let mut key = [0u8; 32];
            key.copy_from_slice(remote);
            // `set` is a no-op if already initialised; safe to ignore the result.
            let _ = peer.remote_static_key.set(key);
        }
        match hs.into_stateless_transport_mode() {
            Ok(transport) => {
                let _ = peer.channel.set(SecureChannel::new(transport));
                // Unblock connect() which may be waiting for encryption to be active.
                peer.channel_ready.notify_one();
                tracing::info!(target: "zudp::engine", peer = %from, "Noise XX handshake complete");
            }
            Err(e) => {
                tracing::error!(target: "zudp::engine", peer = %from, "into_transport_mode: {e}");
            }
        }
    }
}

/// Check the remote's static key against the pinned key in config.
///
/// Returns `true` if the key is acceptable (no pinning, or pinned key matches).
/// Returns `false` and logs a warning if the key is pinned and doesn't match.
/// Must be called while the `HandshakeState` is still active (before consuming it).
#[cfg(feature = "security")]
fn check_pinned_key(
    hs: &snow::HandshakeState,
    inner: &Arc<EngineInner>,
    from: SocketAddr,
) -> bool {
    let Some(expected) = &inner.config.remote_key else {
        return true;
    };
    match hs.get_remote_static() {
        Some(remote) if remote == expected.as_slice() => true,
        Some(_) => {
            tracing::warn!(target: "zudp::security", peer = %from, "key mismatch — remote key does not match pinned key");
            false
        }
        None => {
            tracing::warn!(target: "zudp::security", peer = %from, "key pinning: remote static not available");
            false
        }
    }
}