[relay] Signal relay disconnects through the conn context

AddCloseListener deduplicated listeners by comparing
reflect.ValueOf(callback).Pointer(). For a method value that pointer is
the address of the compiler-generated wrapper, not an identity bound to
the receiver, so every peer's w.onRelayClientDisconnected compared equal.

All peers on the home relay register under the same connectionURL key, so
only the first registration survived and the rest were silently dropped.
On a relay disconnect those peers were never notified: statusRelay stayed
connected and the reconnect guard never fired. The relayed net.Conn itself
was closed by closeAllConns, so nothing leaked, but the peer state machine
did not learn about it. Foreign relays had the same defect scoped to the
peers sharing that server.

Rather than fixing the deduplication, drop the peer-level listener registry
entirely. A relayed Conn now exposes Context(), cancelled when the
connection is torn down, with a cancellation cause naming the reason. This
is the same shape quic-go uses for its Conn and Stream types, and it
removes the whole class of problems around listener identity, lifetime and
deregistration: the signal belongs to the resource instead of a side table.

WorkerRelay watches that context in a goroutine whose lifetime matches the
connection. A watcher that wakes up for a superseded connection compares
the conn pointer against the current one and returns without touching the
state machine, so a fast relay reconnect cannot have a stale watcher tear
down the connection that replaced it.

Client.SetOnDisconnectListener stays: it is server-level and drives the
reconnect guard and foreign relay eviction, unrelated to peers.

handleRelayReady also checks the conn context, closing the race where the
relay dies between OpenConn and the readiness handoff and the peer would
otherwise build a WireGuard endpoint over a dead connection.

TestNotifierDoubleAdd covered the removed mechanism and is gone.
TestForeignAutoClose asserted nothing (both branches logged); it now waits
for the relay to leave the client map and fails if it does not.

shared/relay/client/manager_test.go

@@ -2,6 +2,7 @@ package client
 
 import (
 	"context"
+	"errors"
 	"fmt"
 	"net/netip"
 	"testing"
@@ -291,35 +292,29 @@ func TestForeignAutoClose(t *testing.T) {
 		t.Fatalf("failed to serve manager: %s", err)
 	}
 
-	// Set up a disconnect listener to track when foreign server disconnects
 	foreignServerURL := toURL(srvCfg2)[0]
-	disconnected := make(chan struct{})
-	onDisconnect := func() {
-		select {
-		case disconnected <- struct{}{}:
-		default:
-		}
-	}
 
 	t.Log("open connection to another peer")
 	if _, err = mgr.OpenConn(ctx, foreignServerURL, "anotherpeer", netip.Addr{}); err == nil {
 		t.Fatalf("should have failed to open connection to another peer")
 	}
 
-	// Add the disconnect listener after the connection attempt
-	if err := mgr.AddCloseListener(foreignServerURL, onDisconnect); err != nil {
-		t.Logf("failed to add close listener (expected if connection failed): %s", err)
-	}
-
-	// Wait for cleanup to happen
 	timeout := relayCleanupInterval + keepUnusedServerTime + 2*time.Second
 	t.Logf("waiting for relay cleanup: %s", timeout)
-
-	select {
-	case <-disconnected:
-		t.Log("foreign relay connection cleaned up successfully")
-	case <-time.After(timeout):
-		t.Log("timeout waiting for cleanup - this might be expected if connection never established")
+	deadline := time.After(timeout)
+	for {
+		mgr.relayClientsMutex.RLock()
+		_, tracked := mgr.relayClients[foreignServerURL]
+		mgr.relayClientsMutex.RUnlock()
+		if !tracked {
+			t.Log("foreign relay connection cleaned up successfully")
+			break
+		}
+		select {
+		case <-deadline:
+			t.Fatal("foreign relay was not cleaned up")
+		case <-time.After(200 * time.Millisecond):
+		}
 	}
 
 	t.Logf("closing manager")
@@ -413,23 +408,24 @@ func waitForReady(ctx context.Context, m *Manager, timeout time.Duration) error
 	return fmt.Errorf("manager not ready within %s", timeout)
 }
 
-func TestNotifierDoubleAdd(t *testing.T) {
+func toURL(address server.ListenerConfig) []string {
+	return []string{"rel://" + address.Address}
+}
+
+func TestConnContextCancelledOnServerDisconnect(t *testing.T) {
 	ctx := context.Background()
 
-	listenerCfg1 := server.ListenerConfig{
-		Address: "localhost:52501",
-	}
-	srv, err := server.NewServer(newManagerTestServerConfig(listenerCfg1.Address))
+	srvCfg := server.ListenerConfig{Address: "localhost:52601"}
+	srv, err := server.NewServer(newManagerTestServerConfig(srvCfg.Address))
 	if err != nil {
 		t.Fatalf("failed to create server: %s", err)
 	}
 	errChan := make(chan error, 1)
 	go func() {
-		if err := srv.Listen(listenerCfg1); err != nil {
+		if err := srv.Listen(srvCfg); err != nil {
 			errChan <- err
 		}
 	}()
-
 	defer func() {
 		if err := srv.Shutdown(ctx); err != nil {
 			t.Errorf("failed to close server: %s", err)
@@ -440,46 +436,106 @@ func TestNotifierDoubleAdd(t *testing.T) {
 		t.Fatalf("failed to start server: %s", err)
 	}
 
-	log.Debugf("connect by alice")
 	mCtx, cancel := context.WithCancel(ctx)
 	defer cancel()
 
-	clientBob := NewManager(mCtx, toURL(listenerCfg1), "bob", iface.DefaultMTU)
-	if err = clientBob.Serve(); err != nil {
+	mgrBob := NewManager(mCtx, toURL(srvCfg), "bob", iface.DefaultMTU)
+	if err := mgrBob.Serve(); err != nil {
+		t.Fatalf("failed to serve bob manager: %s", err)
+	}
+
+	mgr := NewManager(mCtx, toURL(srvCfg), "alice", iface.DefaultMTU)
+	if err := mgr.Serve(); err != nil {
 		t.Fatalf("failed to serve manager: %s", err)
 	}
 
-	clientAlice := NewManager(mCtx, toURL(listenerCfg1), "alice", iface.DefaultMTU)
-	if err = clientAlice.Serve(); err != nil {
+	ra, _, err := mgr.RelayInstanceAddress()
+	if err != nil {
+		t.Fatalf("failed to get relay address: %s", err)
+	}
+
+	relayedConn, err := mgr.OpenConn(ctx, ra, "bob", netip.Addr{})
+	if err != nil {
+		t.Fatalf("failed to open conn: %s", err)
+	}
+
+	select {
+	case <-relayedConn.Context().Done():
+		t.Fatal("conn context cancelled while the relay is still up")
+	default:
+	}
+
+	_ = mgr.relayClient.relayConn.Close()
+
+	select {
+	case <-relayedConn.Context().Done():
+	case <-time.After(15 * time.Second):
+		t.Fatal("conn context was not cancelled after the relay connection dropped")
+	}
+
+	if cause := context.Cause(relayedConn.Context()); !errors.Is(cause, ErrServerDisconnected) {
+		t.Errorf("unexpected cancellation cause: %v, want %v", cause, ErrServerDisconnected)
+	}
+}
+
+func TestConnContextCauseOnLocalClose(t *testing.T) {
+	ctx := context.Background()
+
+	srvCfg := server.ListenerConfig{Address: "localhost:52602"}
+	srv, err := server.NewServer(newManagerTestServerConfig(srvCfg.Address))
+	if err != nil {
+		t.Fatalf("failed to create server: %s", err)
+	}
+	errChan := make(chan error, 1)
+	go func() {
+		if err := srv.Listen(srvCfg); err != nil {
+			errChan <- err
+		}
+	}()
+	defer func() {
+		if err := srv.Shutdown(ctx); err != nil {
+			t.Errorf("failed to close server: %s", err)
+		}
+	}()
+
+	if err := waitForServerToStart(errChan); err != nil {
+		t.Fatalf("failed to start server: %s", err)
+	}
+
+	mCtx, cancel := context.WithCancel(ctx)
+	defer cancel()
+
+	mgrBob := NewManager(mCtx, toURL(srvCfg), "bob", iface.DefaultMTU)
+	if err := mgrBob.Serve(); err != nil {
+		t.Fatalf("failed to serve bob manager: %s", err)
+	}
+
+	mgr := NewManager(mCtx, toURL(srvCfg), "alice", iface.DefaultMTU)
+	if err := mgr.Serve(); err != nil {
 		t.Fatalf("failed to serve manager: %s", err)
 	}
 
-	conn1, err := clientAlice.OpenConn(ctx, clientAlice.ServerURLs()[0], "bob", netip.Addr{})
+	ra, _, err := mgr.RelayInstanceAddress()
 	if err != nil {
-		t.Fatalf("failed to bind channel: %s", err)
+		t.Fatalf("failed to get relay address: %s", err)
 	}
 
-	fnCloseListener := OnServerCloseListener(func() {
-		log.Infof("close listener")
-	})
-
-	err = clientAlice.AddCloseListener(clientAlice.ServerURLs()[0], fnCloseListener)
+	relayedConn, err := mgr.OpenConn(ctx, ra, "bob", netip.Addr{})
 	if err != nil {
-		t.Fatalf("failed to add close listener: %s", err)
+		t.Fatalf("failed to open conn: %s", err)
 	}
 
-	err = clientAlice.AddCloseListener(clientAlice.ServerURLs()[0], fnCloseListener)
-	if err != nil {
-		t.Fatalf("failed to add close listener: %s", err)
+	if err := relayedConn.Close(); err != nil {
+		t.Fatalf("failed to close conn: %s", err)
 	}
 
-	err = conn1.Close()
-	if err != nil {
-		t.Errorf("failed to close connection: %s", err)
+	select {
+	case <-relayedConn.Context().Done():
+	case <-time.After(5 * time.Second):
+		t.Fatal("conn context was not cancelled after a local close")
 	}
 
-}
-
-func toURL(address server.ListenerConfig) []string {
-	return []string{"rel://" + address.Address}
+	if cause := context.Cause(relayedConn.Context()); errors.Is(cause, ErrServerDisconnected) {
+		t.Errorf("local close must not be reported as a server disconnect, got: %v", cause)
+	}
 }