Merge branch 'main' into 0.75.0-branch

# Conflicts:
#	.github/workflows/golang-test-darwin.yml
#	.github/workflows/golang-test-linux.yml
#	.github/workflows/golangci-lint.yml
#	client/internal/connect.go
#	client/internal/peer/status.go
#	client/server/server_test.go
#	client/ui/client_ui.go
#	go.mod
#	go.sum

client/internal/peer/conn.go

@@ -6,6 +6,7 @@ import (
 	"net"
 	"net/netip"
 	"runtime"
+	"slices"
 	"sync"
 	"time"
 
@@ -136,6 +137,39 @@ type Conn struct {
 	// Connection stage timestamps for metrics
 	metricsRecorder MetricsRecorder
 	metricsStages   *MetricsStages
+
+	// pendingFirstPacket is the lazyconn-captured handshake init, replayed once the real
+	// transport is up.
+	pendingFirstPacket []byte
+}
+
+// injectPendingFirstPacket replays the captured handshake through the proxy if present, else
+// directly through the ICE conn. The packet is cleared only after a successful write, so a failed
+// or transport-less attempt leaves it available for a later reinjection. Caller must hold conn.mu.
+func (conn *Conn) injectPendingFirstPacket(proxy wgproxy.Proxy, directConn net.Conn) {
+	pkt := conn.pendingFirstPacket
+	if len(pkt) == 0 {
+		return
+	}
+
+	switch {
+	case proxy != nil:
+		if err := proxy.InjectPacket(pkt); err != nil {
+			conn.Log.Debugf("failed to reinject captured first packet via proxy: %v", err)
+			return
+		}
+	case directConn != nil:
+		if _, err := directConn.Write(pkt); err != nil {
+			conn.Log.Debugf("failed to reinject captured first packet via direct conn: %v", err)
+			return
+		}
+	default:
+		conn.Log.Debugf("no transport available to reinject captured first packet")
+		return
+	}
+
+	conn.pendingFirstPacket = nil
+	conn.Log.Debugf("reinjected captured first packet (%d bytes)", len(pkt))
 }
 
 // NewConn creates a new not opened Conn to the remote peer.
@@ -172,6 +206,16 @@ func NewConn(config ConnConfig, services ServiceDependencies) (*Conn, error) {
 // It will try to establish a connection using ICE and in parallel with relay. The higher priority connection type will
 // be used.
 func (conn *Conn) Open(engineCtx context.Context) error {
+	return conn.open(engineCtx, nil)
+}
+
+// OpenWithFirstPacket opens the connection like Open and stashes firstPacket to be replayed once
+// the real transport is established. The packet is retained only on a successful open.
+func (conn *Conn) OpenWithFirstPacket(engineCtx context.Context, firstPacket []byte) error {
+	return conn.open(engineCtx, firstPacket)
+}
+
+func (conn *Conn) open(engineCtx context.Context, firstPacket []byte) error {
 	conn.mu.Lock()
 	defer conn.mu.Unlock()
 
@@ -227,6 +271,9 @@ func (conn *Conn) Open(engineCtx context.Context) error {
 		defer conn.wg.Done()
 		conn.guard.Start(conn.ctx, conn.onGuardEvent)
 	}()
+	if len(firstPacket) > 0 {
+		conn.pendingFirstPacket = slices.Clone(firstPacket)
+	}
 	conn.opened = true
 	return nil
 }
@@ -423,6 +470,8 @@ func (conn *Conn) onICEConnectionIsReady(priority conntype.ConnPriority, iceConn
 		conn.wgProxyRelay.RedirectAs(ep)
 	}
 
+	conn.injectPendingFirstPacket(wgProxy, iceConnInfo.RemoteConn)
+
 	conn.currentConnPriority = priority
 	conn.statusICE.SetConnected()
 	conn.updateIceState(iceConnInfo, updateTime)
@@ -546,6 +595,8 @@ func (conn *Conn) onRelayConnectionIsReady(rci RelayConnInfo) {
 
 	wgConfigWorkaround()
 
+	conn.injectPendingFirstPacket(wgProxy, nil)
+
 	conn.rosenpassRemoteKey = rci.rosenpassPubKey
 	conn.currentConnPriority = conntype.Relay
 	conn.statusRelay.SetConnected()
@@ -752,15 +803,17 @@ func (conn *Conn) isConnectedOnAllWay() (status guard.ConnStatus) {
 }
 
 func (conn *Conn) enableWgWatcherIfNeeded(enabledTime time.Time) {
-	if !conn.wgWatcher.IsEnabled() {
-		wgWatcherCtx, wgWatcherCancel := context.WithCancel(conn.ctx)
-		conn.wgWatcherCancel = wgWatcherCancel
-		conn.wgWatcherWg.Add(1)
-		go func() {
-			defer conn.wgWatcherWg.Done()
-			conn.wgWatcher.EnableWgWatcher(wgWatcherCtx, enabledTime, conn.onWGDisconnected, conn.onWGHandshakeSuccess)
-		}()
+	if !conn.wgWatcher.PrepareInitialHandshake() {
+		return
 	}
+
+	wgWatcherCtx, wgWatcherCancel := context.WithCancel(conn.ctx)
+	conn.wgWatcherCancel = wgWatcherCancel
+	conn.wgWatcherWg.Add(1)
+	go func() {
+		defer conn.wgWatcherWg.Done()
+		conn.wgWatcher.EnableWgWatcher(wgWatcherCtx, enabledTime, conn.onWGDisconnected, conn.onWGHandshakeSuccess)
+	}()
 }
 
 func (conn *Conn) disableWgWatcherIfNeeded() {

client/internal/peer/handshaker.go

@@ -195,14 +195,14 @@ func (h *Handshaker) sendOffer() error {
 	}
 
 	offer := h.buildOfferAnswer()
-	h.log.Infof("sending offer with serial: %s", offer.SessionIDString())
+	h.log.Debugf("sending offer with serial: %s", offer.SessionIDString())
 
 	return h.signaler.SignalOffer(offer, h.config.Key)
 }
 
 func (h *Handshaker) sendAnswer() error {
 	answer := h.buildOfferAnswer()
-	h.log.Infof("sending answer with serial: %s", answer.SessionIDString())
+	h.log.Debugf("sending answer with serial: %s", answer.SessionIDString())
 
 	return h.signaler.SignalAnswer(answer, h.config.Key)
 }

client/internal/peer/status.go

@@ -193,6 +193,7 @@ func (s *StatusChangeSubscription) Events() chan map[string]RouterState {
 // Pure read methods take RLock; anything that mutates state takes Lock.
 type Status struct {
 	mux                 sync.RWMutex
+	muxRelays           sync.RWMutex
 	peers               map[string]State
 	ipToKey             map[string]string
 	changeNotify        map[string]map[string]*StatusChangeSubscription // map[peerID]map[subscriptionID]*StatusChangeSubscription
@@ -268,8 +269,8 @@ func NewRecorder(mgmAddress string) *Status {
 }
 
 func (d *Status) SetRelayMgr(manager *relayClient.Manager) {
-	d.mux.Lock()
-	defer d.mux.Unlock()
+	d.muxRelays.Lock()
+	defer d.muxRelays.Unlock()
 	d.relayMgr = manager
 }
 
@@ -997,8 +998,8 @@ func (d *Status) MarkSignalConnected() {
 }
 
 func (d *Status) UpdateRelayStates(relayResults []relay.ProbeResult) {
-	d.mux.Lock()
-	defer d.mux.Unlock()
+	d.muxRelays.Lock()
+	defer d.muxRelays.Unlock()
 	d.relayStates = relayResults
 }
 
@@ -1109,24 +1110,26 @@ func (d *Status) GetSignalState() SignalState {
 
 // GetRelayStates returns the stun/turn/permanent relay states
 func (d *Status) GetRelayStates() []relay.ProbeResult {
-	d.mux.RLock()
-	defer d.mux.RUnlock()
+	d.muxRelays.RLock()
 	if d.relayMgr == nil {
-		return d.relayStates
+		defer d.muxRelays.RUnlock()
+		return slices.Clone(d.relayStates)
 	}
 
+	relayMgr := d.relayMgr
 	// extend the list of stun, turn servers with the relay server connections
 	relayStates := slices.Clone(d.relayStates)
+	d.muxRelays.RUnlock()
 
-	states := d.relayMgr.RelayStates()
+	states := relayMgr.RelayStates()
 	if len(states) == 0 {
 		// no relay connection tracked yet; surface configured servers as
 		// unavailable with the real reconnect error when known
 		err := relayClient.ErrRelayClientNotConnected
-		if connErr := d.relayMgr.RelayConnectError(); connErr != nil {
+		if connErr := relayMgr.RelayConnectError(); connErr != nil {
 			err = connErr
 		}
-		for _, r := range d.relayMgr.ServerURLs() {
+		for _, r := range relayMgr.ServerURLs() {
 			relayStates = append(relayStates, relay.ProbeResult{
 				URI: r,
 				Err: err,

client/internal/peer/wg_watcher.go

@@ -31,7 +31,9 @@ type WGWatcher struct {
 	stateDump     *stateDump
 
 	enabled   bool
-	muEnabled sync.RWMutex
+	muEnabled sync.Mutex
+	// initialHandshake is not thread-safe; never call PrepareInitialHandshake and EnableWgWatcher concurrently.
+	initialHandshake time.Time
 
 	resetCh chan struct{}
 }
@@ -46,38 +48,38 @@ func NewWGWatcher(log *log.Entry, wgIfaceStater WGInterfaceStater, peerKey strin
 	}
 }
 
-// EnableWgWatcher starts the WireGuard watcher. If it is already enabled, it will return immediately and do nothing.
-// The watcher runs until ctx is cancelled. Caller is responsible for context lifecycle management.
-func (w *WGWatcher) EnableWgWatcher(ctx context.Context, enabledTime time.Time, onDisconnectedFn func(), onHandshakeSuccessFn func(when time.Time)) {
+// PrepareInitialHandshake reserves the watcher and reads the peer's current WireGuard
+// handshake time. It must be called before the peer is (re)configured on the WireGuard
+// interface, so the captured baseline reflects the state prior to this connection attempt
+// instead of racing with that configuration. Returns ok=false if the watcher is already
+// running, in which case EnableWgWatcher must not be called.
+func (w *WGWatcher) PrepareInitialHandshake() (ok bool) {
 	w.muEnabled.Lock()
 	if w.enabled {
 		w.muEnabled.Unlock()
-		return
+		return false
 	}
 
 	w.log.Debugf("enable WireGuard watcher")
 	w.enabled = true
 	w.muEnabled.Unlock()
 
-	initialHandshake, err := w.wgState()
-	if err != nil {
-		w.log.Warnf("failed to read initial wg stats: %v", err)
-	}
+	handshake, _ := w.wgState()
+	w.initialHandshake = handshake
+	return true
+}
 
-	w.periodicHandshakeCheck(ctx, onDisconnectedFn, onHandshakeSuccessFn, enabledTime, initialHandshake)
+// EnableWgWatcher runs the WireGuard watcher loop using the handshake baseline captured by
+// PrepareInitialHandshake. The watcher runs until ctx is cancelled. Caller is responsible
+// for context lifecycle management.
+func (w *WGWatcher) EnableWgWatcher(ctx context.Context, enabledTime time.Time, onDisconnectedFn func(), onHandshakeSuccessFn func(when time.Time)) {
+	w.periodicHandshakeCheck(ctx, onDisconnectedFn, onHandshakeSuccessFn, enabledTime, w.initialHandshake)
 
 	w.muEnabled.Lock()
 	w.enabled = false
 	w.muEnabled.Unlock()
 }
 
-// IsEnabled returns true if the WireGuard watcher is currently enabled
-func (w *WGWatcher) IsEnabled() bool {
-	w.muEnabled.RLock()
-	defer w.muEnabled.RUnlock()
-	return w.enabled
-}
-
 // Reset signals the watcher that the WireGuard peer has been reset and a new
 // handshake is expected. This restarts the handshake timeout from scratch.
 func (w *WGWatcher) Reset() {
@@ -101,13 +103,16 @@ func (w *WGWatcher) periodicHandshakeCheck(ctx context.Context, onDisconnectedFn
 		case <-timer.C:
 			handshake, ok := w.handshakeCheck(lastHandshake)
 			if !ok {
+				if ctx.Err() != nil {
+					return
+				}
 				onDisconnectedFn()
 				return
 			}
 			if lastHandshake.IsZero() {
 				elapsed := calcElapsed(enabledTime, *handshake)
 				w.log.Infof("first wg handshake detected within: %.2fsec, (%s)", elapsed, handshake)
-				if onHandshakeSuccessFn != nil {
+				if onHandshakeSuccessFn != nil && ctx.Err() == nil {
 					onHandshakeSuccessFn(*handshake)
 				}
 			}

client/internal/peer/wg_watcher_test.go

@@ -7,6 +7,7 @@ import (
 	"time"
 
 	log "github.com/sirupsen/logrus"
+	"github.com/stretchr/testify/require"
 
 	"github.com/netbirdio/netbird/client/iface/configurer"
 )
@@ -34,6 +35,9 @@ func TestWGWatcher_EnableWgWatcher(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
 
+	ok := watcher.PrepareInitialHandshake()
+	require.True(t, ok, "watcher should not be enabled yet")
+
 	onDisconnected := make(chan struct{}, 1)
 	go watcher.EnableWgWatcher(ctx, time.Now(), func() {
 		mlog.Infof("onDisconnectedFn")
@@ -62,6 +66,9 @@ func TestWGWatcher_ReEnable(t *testing.T) {
 	watcher := NewWGWatcher(mlog, mocWgIface, "", newStateDump("peer", mlog, &Status{}))
 
 	ctx, cancel := context.WithCancel(context.Background())
+	ok := watcher.PrepareInitialHandshake()
+	require.True(t, ok, "watcher should not be enabled yet")
+
 	wg := &sync.WaitGroup{}
 	wg.Add(1)
 	go func() {
@@ -76,6 +83,9 @@ func TestWGWatcher_ReEnable(t *testing.T) {
 	ctx, cancel = context.WithCancel(context.Background())
 	defer cancel()
 
+	ok = watcher.PrepareInitialHandshake()
+	require.True(t, ok, "watcher should be re-enabled after the previous run stopped")
+
 	onDisconnected := make(chan struct{}, 1)
 	go watcher.EnableWgWatcher(ctx, time.Now(), func() {
 		onDisconnected <- struct{}{}