Invoke address resolution upon subsequent traffic to Failed neighbor

Removes the period of time in which subseqeuent traffic to a Failed neighbor
immediately fails with ErrNoLinkAddress. A Failed neighbor is one in which
address resolution fails; or in other words, the neighbor's IP address cannot
be translated to a MAC address.

This means removing the Failed state for linkAddrCache and allowing transitiong
out of Failed into Incomplete for neighborCache. Previously, both caches would
transition entries to Failed after address resolution fails. In this state, any
subsequent traffic requested within an unreachable time would immediately fail
with ErrNoLinkAddress. This does not follow RFC 4861 section 7.3.3:

  If address resolution fails, the entry SHOULD be deleted, so that subsequent
  traffic to that neighbor invokes the next-hop determination procedure again.
  Invoking next-hop determination at this point ensures that alternate default
  routers are tried.

The API for getting a link address for a given address, whether through the link
address cache or the neighbor table, is updated to optionally take a callback
which will be called when address resolution completes. This allows `Route` to
handle completing link resolution internally, so callers of (*Route).Resolve
(e.g. endpoints) don’t have to keep track of when it completes and update the
Route accordingly.

This change also removes the wakers from LinkAddressCache, NeighborCache, and
Route in favor of the callbacks, and callers that previously used a waker can
now just pass a callback to (*Route).Resolve that will notify the waker on
resolution completion.

Fixes #4796

Startblock:
  has LGTM from sbalana
  and then
  add reviewer ghanan
PiperOrigin-RevId: 348597478

1 files changed, 4 insertions, 31 deletions

diff --git a/pkg/tcpip/transport/raw/endpoint.go b/pkg/tcpip/transport/raw/endpoint.go
index 5df703fb2..7befcfc9b 100644
--- a/pkg/tcpip/transport/raw/endpoint.go
+++ b/pkg/tcpip/transport/raw/endpoint.go
@@ -261,15 +261,14 @@ func (e *endpoint) write(p tcpip.Payloader, opts tcpip.WriteOptions) (int64, <-c
 	}
 
 	e.mu.RLock()
+	defer e.mu.RUnlock()
 
 	if e.closed {
-		e.mu.RUnlock()
 		return 0, nil, tcpip.ErrInvalidEndpointState
 	}
 
 	payloadBytes, err := p.FullPayload()
 	if err != nil {
-		e.mu.RUnlock()
 		return 0, nil, err
 	}
 
@@ -278,7 +277,6 @@ func (e *endpoint) write(p tcpip.Payloader, opts tcpip.WriteOptions) (int64, <-c
 	if e.ops.GetHeaderIncluded() {
 		ip := header.IPv4(payloadBytes)
 		if !ip.IsValid(len(payloadBytes)) {
-			e.mu.RUnlock()
 			return 0, nil, tcpip.ErrInvalidOptionValue
 		}
 		dstAddr := ip.DestinationAddress()
@@ -300,39 +298,16 @@ func (e *endpoint) write(p tcpip.Payloader, opts tcpip.WriteOptions) (int64, <-c
 		// If the user doesn't specify a destination, they should have
 		// connected to another address.
 		if !e.connected {
-			e.mu.RUnlock()
 			return 0, nil, tcpip.ErrDestinationRequired
 		}
 
-		if e.route.IsResolutionRequired() {
-			savedRoute := e.route
-			// Promote lock to exclusive if using a shared route,
-			// given that it may need to change in finishWrite.
-			e.mu.RUnlock()
-			e.mu.Lock()
-
-			// Make sure that the route didn't change during the
-			// time we didn't hold the lock.
-			if !e.connected || savedRoute != e.route {
-				e.mu.Unlock()
-				return 0, nil, tcpip.ErrInvalidEndpointState
-			}
-
-			n, ch, err := e.finishWrite(payloadBytes, savedRoute)
-			e.mu.Unlock()
-			return n, ch, err
-		}
-
-		n, ch, err := e.finishWrite(payloadBytes, e.route)
-		e.mu.RUnlock()
-		return n, ch, err
+		return e.finishWrite(payloadBytes, e.route)
 	}
 
 	// The caller provided a destination. Reject destination address if it
 	// goes through a different NIC than the endpoint was bound to.
 	nic := opts.To.NIC
 	if e.bound && nic != 0 && nic != e.BindNICID {
-		e.mu.RUnlock()
 		return 0, nil, tcpip.ErrNoRoute
 	}
 
@@ -340,13 +315,11 @@ func (e *endpoint) write(p tcpip.Payloader, opts tcpip.WriteOptions) (int64, <-c
 	// FindRoute will choose an appropriate source address.
 	route, err := e.stack.FindRoute(nic, e.BindAddr, opts.To.Addr, e.NetProto, false)
 	if err != nil {
-		e.mu.RUnlock()
 		return 0, nil, err
 	}
 
 	n, ch, err := e.finishWrite(payloadBytes, route)
 	route.Release()
-	e.mu.RUnlock()
 	return n, ch, err
 }
 
@@ -435,11 +408,11 @@ func (e *endpoint) Connect(addr tcpip.FullAddress) *tcpip.Error {
 	if err != nil {
 		return err
 	}
-	defer route.Release()
 
 	if e.associated {
 		// Re-register the endpoint with the appropriate NIC.
 		if err := e.stack.RegisterRawTransportEndpoint(addr.NIC, e.NetProto, e.TransProto, e); err != nil {
+			route.Release()
 			return err
 		}
 		e.stack.UnregisterRawTransportEndpoint(e.RegisterNICID, e.NetProto, e.TransProto, e)
@@ -447,7 +420,7 @@ func (e *endpoint) Connect(addr tcpip.FullAddress) *tcpip.Error {
 	}
 
 	// Save the route we've connected via.
-	e.route = route.Clone()
+	e.route = route
 	e.connected = true
 
 	return nil