[client] Fix engine restart (#3435)

- Refactor the network monitoring to handle one event and it after return
- In the engine restart cancel the upper layer context and the responsibility of the engine stop will be the upper layer
- Before triggering a restart, the engine checks whether the state is already down. This helps avoid unnecessary delayed network restart events.

client/internal/networkmonitor/check_change_linux.go

@@ -0,0 +1,54 @@
+//go:build !android
+
+package networkmonitor
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"syscall"
+
+	log "github.com/sirupsen/logrus"
+	"github.com/vishvananda/netlink"
+
+	"github.com/netbirdio/netbird/client/internal/routemanager/systemops"
+)
+
+func checkChange(ctx context.Context, nexthopv4, nexthopv6 systemops.Nexthop) error {
+	if nexthopv4.Intf == nil && nexthopv6.Intf == nil {
+		return errors.New("no interfaces available")
+	}
+
+	done := make(chan struct{})
+	defer close(done)
+
+	routeChan := make(chan netlink.RouteUpdate)
+	if err := netlink.RouteSubscribe(routeChan, done); err != nil {
+		return fmt.Errorf("subscribe to route updates: %v", err)
+	}
+
+	log.Info("Network monitor: started")
+	for {
+		select {
+		case <-ctx.Done():
+			return ctx.Err()
+		// handle route changes
+		case route := <-routeChan:
+			// default route and main table
+			if route.Dst != nil || route.Table != syscall.RT_TABLE_MAIN {
+				continue
+			}
+			switch route.Type {
+			// triggered on added/replaced routes
+			case syscall.RTM_NEWROUTE:
+				log.Infof("Network monitor: default route changed: via %s, interface %d", route.Gw, route.LinkIndex)
+				return nil
+			case syscall.RTM_DELROUTE:
+				if nexthopv4.Intf != nil && route.Gw.Equal(nexthopv4.IP.AsSlice()) || nexthopv6.Intf != nil && route.Gw.Equal(nexthopv6.IP.AsSlice()) {
+					log.Infof("Network monitor: default route removed: via %s, interface %d", route.Gw, route.LinkIndex)
+					return nil
+				}
+			}
+		}
+	}
+}