Negotiate close even if we already consider the channel closed.

channeld/channeld.c

@@ -918,6 +918,15 @@ static void maybe_send_shutdown(struct peer *peer)
 	billboard_update(peer);
 }
 
+static void send_shutdown_complete(struct peer *peer)
+{
+	/* Now we can tell master shutdown is complete. */
+	wire_sync_write(MASTER_FD,
+			take(towire_channeld_shutdown_complete(NULL, peer->pps)));
+	per_peer_state_fdpass_send(MASTER_FD, peer->pps);
+	close(MASTER_FD);
+}
+
 /* This queues other traffic from the fd until we get reply. */
 static u8 *master_wait_sync_reply(const tal_t *ctx,
 				  struct peer *peer,
@@ -2511,7 +2520,8 @@ static bool capture_premature_msg(const u8 ***shit_lnd_says, const u8 *msg)
 }
 
 static void peer_reconnect(struct peer *peer,
-			   const struct secret *last_remote_per_commit_secret)
+			   const struct secret *last_remote_per_commit_secret,
+			   u8 *reestablish_only)
 {
 	struct channel_id channel_id;
 	/* Note: BOLT #2 uses these names! */
@@ -2636,6 +2646,13 @@ static void peer_reconnect(struct peer *peer,
 	bool soft_error = peer->funding_locked[REMOTE]
 		|| peer->funding_locked[LOCAL];
 
+	/* If they sent reestablish, we analyze it for courtesy, but also
+	 * in case *they* are ahead of us! */
+	if (reestablish_only) {
+		msg = reestablish_only;
+		goto got_reestablish;
+	}
+
 	/* Read until they say something interesting (don't forward
 	 * gossip *to* them yet: we might try sending channel_update
 	 * before we've reestablished channel). */
@@ -2647,6 +2664,7 @@ static void peer_reconnect(struct peer *peer,
 					     msg) ||
 		 capture_premature_msg(&premature_msgs, msg));
 
+got_reestablish:
 #if EXPERIMENTAL_FEATURES
 	recv_tlvs = tlv_channel_reestablish_tlvs_new(tmpctx);
 #endif
@@ -2668,6 +2686,15 @@ static void peer_reconnect(struct peer *peer,
 				 tal_hex(msg, msg));
 	}
 
+	if (!channel_id_eq(&channel_id, &peer->channel_id)) {
+		peer_failed_err(peer->pps,
+				&channel_id,
+				"bad reestablish msg for unknown channel %s: %s",
+				type_to_string(tmpctx, struct channel_id,
+					       &channel_id),
+				tal_hex(msg, msg));
+	}
+
 	status_debug("Got reestablish commit=%"PRIu64" revoke=%"PRIu64,
 		     next_commitment_number,
 		     next_revocation_number);
@@ -2910,6 +2937,19 @@ static void peer_reconnect(struct peer *peer,
 
 #endif /* EXPERIMENTAL_FEATURES */
 
+	/* Now stop, we've been polite long enough. */
+	if (reestablish_only) {
+		/* If we were successfully closing, we still go to closingd. */
+		if (shutdown_complete(peer)) {
+			send_shutdown_complete(peer);
+			daemon_shutdown();
+			exit(0);
+		}
+		peer_failed_err(peer->pps,
+				&peer->channel_id,
+				"Channel is already closed");
+	}
+
 	/* Corner case: we didn't send shutdown before because update_add_htlc
 	 * pending, but now they're cleared by restart, and we're actually
 	 * complete.  In that case, their `shutdown` will trigger us. */
@@ -3400,6 +3440,7 @@ static void init_channel(struct peer *peer)
 	secp256k1_ecdsa_signature *remote_ann_bitcoin_sig;
 	bool option_static_remotekey, option_anchor_outputs;
 	struct penalty_base *pbases;
+	u8 *reestablish_only;
 #if !DEVELOPER
 	bool dev_fail_process_onionpacket; /* Ignored */
 #endif
@@ -3461,7 +3502,8 @@ static void init_channel(struct peer *peer)
 				   &option_anchor_outputs,
 				   &dev_fast_gossip,
 				   &dev_fail_process_onionpacket,
-				   &pbases)) {
+				   &pbases,
+				   &reestablish_only)) {
 		master_badmsg(WIRE_CHANNELD_INIT, msg);
 	}
 
@@ -3553,7 +3595,10 @@ static void init_channel(struct peer *peer)
 
 	/* OK, now we can process peer messages. */
 	if (reconnected)
-		peer_reconnect(peer, &last_remote_per_commit_secret);
+		peer_reconnect(peer, &last_remote_per_commit_secret,
+			       reestablish_only);
+	else
+		assert(!reestablish_only);
 
 	/* If we have a messages to send, send them immediately */
 	if (fwd_msg)
@@ -3565,15 +3610,6 @@ static void init_channel(struct peer *peer)
 	billboard_update(peer);
 }
 
-static void send_shutdown_complete(struct peer *peer)
-{
-	/* Now we can tell master shutdown is complete. */
-	wire_sync_write(MASTER_FD,
-			take(towire_channeld_shutdown_complete(NULL, peer->pps)));
-	per_peer_state_fdpass_send(MASTER_FD, peer->pps);
-	close(MASTER_FD);
-}
-
 static void try_read_gossip_store(struct peer *peer)
 {
 	u8 *msg = gossip_store_next(tmpctx, peer->pps);

channeld/channeld_wire.csv

@@ -73,6 +73,8 @@ msgdata,channeld_init,dev_fast_gossip,bool,
 msgdata,channeld_init,dev_fail_process_onionpacket,bool,
 msgdata,channeld_init,num_penalty_bases,u32,
 msgdata,channeld_init,pbases,penalty_base,num_penalty_bases
+msgdata,channeld_init,reestablish_only_len,u16,
+msgdata,channeld_init,reestablish_only,u8,reestablish_only_len
 
 # master->channeld funding hit new depth(funding locked if >= lock depth)
 msgtype,channeld_funding_depth,1002