htlcswitch: fix on-chain resolution error misclassification

This commit fixes a bug in `parseFailedPayment` where an on-chain
resolution with an empty reason was not being correctly identified
after being deserialized from disk by SubscribeResult or
GetResult. This impacts callers waiting for the final settle/fail
result on an htlc attempt via GetAttemptResult.

The lnwire codec transforms a `nil` reason into a non-nil, empty
byte slice and bypasses the existing check for `htlc.Reason == nil`.
Thankfully this does not crash the daemon. Rather, it just leads to
the caller (in our case, the ChannelRouter) not actually receiving
the expected FailPermanentChannelFailure link error.

Author: calvinrzachman

htlcswitch/switch.go

@@ -1090,7 +1090,7 @@ func (s *Switch) parseFailedPayment(deobfuscator ErrorDecrypter,
 	// the first hop. In this case, we'll report a permanent
 	// channel failure as this means us, or the remote party had to
 	// go on chain.
-	case isResolution && htlc.Reason == nil:
+	case isResolution && len(htlc.Reason) == 0:
 		linkError := NewDetailedLinkError(
 			&lnwire.FailPermanentChannelFailure{},
 			OutgoingFailureOnChainTimeout,

htlcswitch/switch_test.go

@@ -5608,9 +5608,10 @@ func TestOnChainResolutionFailure(t *testing.T) {
 
 		require.NotNil(t, res.Error, "expected error for failed result")
 
-		// We expect the misclassified error which occurs when sending
-		// an empty encrypted failure reason to the de-obfuscator.
-		expectedErr := ErrUnreadableFailureMessage
+		expectedErr := NewDetailedLinkError(
+			&lnwire.FailPermanentChannelFailure{},
+			OutgoingFailureOnChainTimeout,
+		)
 		require.Equal(t, expectedErr, res.Error)
 
 	case <-time.After(1 * time.Second):