Handle 1-conf funding_locked in channel no matter the event order

See comment in the diff for more details

Author: TheBlueMatt

fuzz/src/chanmon_consistency.rs

@@ -436,11 +436,10 @@ pub fn do_test<Out: test_logger::Output>(data: &[u8], out: Out) {
 			let mut header = BlockHeader { version: 0x20000000, prev_blockhash: chain_hash, merkle_root: Default::default(), time: 42, bits: 42, nonce: 42 };
 			let txdata: Vec<_> = channel_txn.iter().enumerate().map(|(i, tx)| (i + 1, tx)).collect();
 			$node.transactions_confirmed(&header, 1, &txdata);
-			$node.update_best_block(&header, 1);
-			for i in 2..100 {
+			for _ in 2..100 {
 				header = BlockHeader { version: 0x20000000, prev_blockhash: header.block_hash(), merkle_root: Default::default(), time: 42, bits: 42, nonce: 42 };
-				$node.update_best_block(&header, i);
 			}
+			$node.update_best_block(&header, 99);
 		} }
 	}
 

lightning/src/ln/channel.rs

@@ -3501,11 +3501,63 @@ impl<Signer: Sign> Channel<Signer> {
 		self.network_sync == UpdateStatus::DisabledMarked
 	}
 
+	fn check_get_funding_locked(&mut self, height: u32) -> Option<msgs::FundingLocked> {
+		if self.funding_tx_confirmation_height == 0 {
+			return None;
+		}
+
+		let funding_tx_confirmations = height as i64 - self.funding_tx_confirmation_height as i64 + 1;
+		if funding_tx_confirmations <= 0 {
+			self.funding_tx_confirmation_height = 0;
+		}
+
+		if funding_tx_confirmations < self.minimum_depth as i64 {
+			return None;
+		}
+
+		let non_shutdown_state = self.channel_state & (!MULTI_STATE_FLAGS);
+		let need_commitment_update = if non_shutdown_state == ChannelState::FundingSent as u32 {
+			self.channel_state |= ChannelState::OurFundingLocked as u32;
+			true
+		} else if non_shutdown_state == (ChannelState::FundingSent as u32 | ChannelState::TheirFundingLocked as u32) {
+			self.channel_state = ChannelState::ChannelFunded as u32 | (self.channel_state & MULTI_STATE_FLAGS);
+			self.update_time_counter += 1;
+			true
+		} else if non_shutdown_state == (ChannelState::FundingSent as u32 | ChannelState::OurFundingLocked as u32) {
+			// We got a reorg but not enough to trigger a force close, just update
+			// funding_tx_confirmed_in and return.
+			false
+		} else if self.channel_state < ChannelState::ChannelFunded as u32 {
+			panic!("Started confirming a channel in a state pre-FundingSent?: {}", self.channel_state);
+		} else {
+			// We got a reorg but not enough to trigger a force close, just update
+			// funding_tx_confirmed_in and return.
+			false
+		};
+
+		//TODO: Note that this must be a duplicate of the previous commitment point they sent us,
+		//as otherwise we will have a commitment transaction that they can't revoke (well, kinda,
+		//they can by sending two revoke_and_acks back-to-back, but not really). This appears to be
+		//a protocol oversight, but I assume I'm just missing something.
+		if need_commitment_update {
+			if self.channel_state & (ChannelState::MonitorUpdateFailed as u32) == 0 {
+				let next_per_commitment_point = self.holder_signer.get_per_commitment_point(self.cur_holder_commitment_transaction_number, &self.secp_ctx);
+				return Some(msgs::FundingLocked {
+					channel_id: self.channel_id,
+					next_per_commitment_point,
+				});
+			} else {
+				self.monitor_pending_funding_locked = true;
+			}
+		}
+		None
+	}
+
 	/// When a transaction is confirmed, we check whether it is or spends the funding transaction
 	/// In the first case, we store the confirmation height and calculating the short channel id.
 	/// In the second, we simply return an Err indicating we need to be force-closed now.
 	pub fn transactions_confirmed<L: Deref>(&mut self, block_hash: &BlockHash, height: u32, txdata: &TransactionData, logger: &L)
-			-> Result<(), msgs::ErrorMessage> where L::Target: Logger {
+			-> Result<Option<msgs::FundingLocked>, msgs::ErrorMessage> where L::Target: Logger {
 		let non_shutdown_state = self.channel_state & (!MULTI_STATE_FLAGS);
 		for &(index_in_block, tx) in txdata.iter() {
 			if let Some(funding_txo) = self.get_funding_txo() {
@@ -3550,6 +3602,12 @@ impl<Signer: Sign> Channel<Signer> {
 							}
 						}
 					}
+					// If we allow 1-conf funding, we may need to check for funding_locked here and
+					// send it immediately instead of waiting for an update_best_block call (which
+					// may have already happened for this block).
+					if let Some(funding_locked) = self.check_get_funding_locked(height) {
+						return Ok(Some(funding_locked));
+					}
 				}
 				for inp in tx.input.iter() {
 					if inp.previous_output == funding_txo.into_bitcoin_outpoint() {
@@ -3562,7 +3620,7 @@ impl<Signer: Sign> Channel<Signer> {
 				}
 			}
 		}
-		Ok(())
+		Ok(None)
 	}
 
 	/// When a new block is connected, we check the height of the block against outbound holding
@@ -3592,59 +3650,28 @@ impl<Signer: Sign> Channel<Signer> {
 		});
 
 		self.update_time_counter = cmp::max(self.update_time_counter, highest_header_time);
-		if self.funding_tx_confirmation_height > 0 {
-			let funding_tx_confirmations = height as i64 - self.funding_tx_confirmation_height as i64 + 1;
-			if funding_tx_confirmations <= 0 {
-				self.funding_tx_confirmation_height = 0;
+
+		if let Some(funding_locked) = self.check_get_funding_locked(height) {
+			return Ok((Some(funding_locked), timed_out_htlcs));
+		}
+
+		let non_shutdown_state = self.channel_state & (!MULTI_STATE_FLAGS);
+		if non_shutdown_state >= ChannelState::ChannelFunded as u32 ||
+		   (non_shutdown_state & ChannelState::OurFundingLocked as u32) == ChannelState::OurFundingLocked as u32 {
+			let mut funding_tx_confirmations = height as i64 - self.funding_tx_confirmation_height as i64 + 1;
+			if self.funding_tx_confirmation_height == 0 {
+				// Note that check_get_funding_locked may reset funding_tx_confirmation_height to
+				// zero if it has been reorged out, however in either case, our state flags
+				// indicate we've already sent a funding_locked
+				funding_tx_confirmations = 0;
 			}
 
-			let non_shutdown_state = self.channel_state & (!MULTI_STATE_FLAGS);
-			if (non_shutdown_state >= ChannelState::ChannelFunded as u32 ||
-			   (non_shutdown_state & ChannelState::OurFundingLocked as u32) == ChannelState::OurFundingLocked as u32) &&
-			    funding_tx_confirmations < self.minimum_depth as i64 / 2 {
+			if funding_tx_confirmations < self.minimum_depth as i64 / 2 {
 				return Err(msgs::ErrorMessage {
 					channel_id: self.channel_id(),
 					data: format!("Funding transaction was un-confirmed. Locked at {} confs, now have {} confs.", self.minimum_depth, funding_tx_confirmations),
 				});
 			}
-
-			if funding_tx_confirmations == self.minimum_depth as i64 {
-				let need_commitment_update = if non_shutdown_state == ChannelState::FundingSent as u32 {
-					self.channel_state |= ChannelState::OurFundingLocked as u32;
-					true
-				} else if non_shutdown_state == (ChannelState::FundingSent as u32 | ChannelState::TheirFundingLocked as u32) {
-					self.channel_state = ChannelState::ChannelFunded as u32 | (self.channel_state & MULTI_STATE_FLAGS);
-					self.update_time_counter += 1;
-					true
-				} else if non_shutdown_state == (ChannelState::FundingSent as u32 | ChannelState::OurFundingLocked as u32) {
-					// We got a reorg but not enough to trigger a force close, just update
-					// funding_tx_confirmed_in and return.
-					false
-				} else if self.channel_state < ChannelState::ChannelFunded as u32 {
-					panic!("Started confirming a channel in a state pre-FundingSent?: {}", self.channel_state);
-				} else {
-					// We got a reorg but not enough to trigger a force close, just update
-					// funding_tx_confirmed_in and return.
-					false
-				};
-
-				//TODO: Note that this must be a duplicate of the previous commitment point they sent us,
-				//as otherwise we will have a commitment transaction that they can't revoke (well, kinda,
-				//they can by sending two revoke_and_acks back-to-back, but not really). This appears to be
-				//a protocol oversight, but I assume I'm just missing something.
-				if need_commitment_update {
-					if self.channel_state & (ChannelState::MonitorUpdateFailed as u32) == 0 {
-						let next_per_commitment_point = self.holder_signer.get_per_commitment_point(self.cur_holder_commitment_transaction_number, &self.secp_ctx);
-						return Ok((Some(msgs::FundingLocked {
-							channel_id: self.channel_id,
-							next_per_commitment_point,
-						}), timed_out_htlcs));
-					} else {
-						self.monitor_pending_funding_locked = true;
-						return Ok((None, timed_out_htlcs));
-					}
-				}
-			}
 		}
 
 		Ok((None, timed_out_htlcs))

lightning/src/ln/channelmanager.rs

@@ -3434,7 +3434,7 @@ impl<Signer: Sign, M: Deref, T: Deref, K: Deref, F: Deref, L: Deref> ChannelMana
 		log_trace!(self.logger, "{} transactions included in block {} at height {} provided", txdata.len(), block_hash, height);
 
 		let _persistence_guard = PersistenceNotifierGuard::new(&self.total_consistency_lock, &self.persistence_notifier);
-		self.do_chain_event(height, |channel| channel.transactions_confirmed(&block_hash, height, txdata, &self.logger).map(|_| (None, Vec::new())));
+		self.do_chain_event(height, |channel| channel.transactions_confirmed(&block_hash, height, txdata, &self.logger).map(|a| (a, Vec::new())));
 	}
 
 	/// Updates channel state with the current best blockchain tip. You should attempt to call this

lightning/src/ln/functional_tests.rs

@@ -394,8 +394,7 @@ fn test_multi_flight_update_fee() {
 	check_added_monitors!(nodes[1], 1);
 }
 
-#[test]
-fn test_1_conf_open() {
+fn do_test_1_conf_open(connect_style: ConnectStyle) {
 	// Previously, if the minium_depth config was set to 1, we'd never send a funding_locked. This
 	// tests that we properly send one in that case.
 	let mut alice_config = UserConfig::default();
@@ -409,7 +408,8 @@ fn test_1_conf_open() {
 	let chanmon_cfgs = create_chanmon_cfgs(2);
 	let node_cfgs = create_node_cfgs(2, &chanmon_cfgs);
 	let node_chanmgrs = create_node_chanmgrs(2, &node_cfgs, &[Some(alice_config), Some(bob_config)]);
-	let nodes = create_network(2, &node_cfgs, &node_chanmgrs);
+	let mut nodes = create_network(2, &node_cfgs, &node_chanmgrs);
+	*nodes[0].connect_style.borrow_mut() = connect_style;
 
 	let tx = create_chan_between_nodes_with_value_init(&nodes[0], &nodes[1], 100000, 10001, InitFeatures::known(), InitFeatures::known());
 	mine_transaction(&nodes[1], &tx);
@@ -425,6 +425,12 @@ fn test_1_conf_open() {
 		node.net_graph_msg_handler.handle_channel_update(&bs_update).unwrap();
 	}
 }
+#[test]
+fn test_1_conf_open() {
+	do_test_1_conf_open(ConnectStyle::BestBlockFirst);
+	do_test_1_conf_open(ConnectStyle::TransactionsFirst);
+	do_test_1_conf_open(ConnectStyle::FullBlockViaListen);
+}
 
 fn do_test_sanity_on_in_flight_opens(steps: u8) {
 	// Previously, we had issues deserializing channels when we hadn't connected the first block

lightning/src/ln/reorg_tests.rs

@@ -184,7 +184,7 @@ fn test_onchain_htlc_timeout_delay_remote_commitment() {
 	do_test_onchain_htlc_reorg(false, false);
 }
 
-fn do_test_unconf_chan(reload_node: bool, reorg_after_reload: bool) {
+fn do_test_unconf_chan(reload_node: bool, reorg_after_reload: bool, connect_style: ConnectStyle) {
 	// After creating a chan between nodes, we disconnect all blocks previously seen to force a
 	// channel close on nodes[0] side. We also use this to provide very basic testing of logic
 	// around freeing background events which store monitor updates during block_[dis]connected.
@@ -195,6 +195,8 @@ fn do_test_unconf_chan(reload_node: bool, reorg_after_reload: bool) {
 	let new_chain_monitor: test_utils::TestChainMonitor;
 	let nodes_0_deserialized: ChannelManager<EnforcingSigner, &test_utils::TestChainMonitor, &test_utils::TestBroadcaster, &test_utils::TestKeysInterface, &test_utils::TestFeeEstimator, &test_utils::TestLogger>;
 	let mut nodes = create_network(2, &node_cfgs, &node_chanmgrs);
+	*nodes[0].connect_style.borrow_mut() = connect_style;
+
 	let chan_id = create_announced_chan_between_nodes(&nodes, 0, 1, InitFeatures::known(), InitFeatures::known()).2;
 
 	let channel_state = nodes[0].node.channel_state.lock().unwrap();
@@ -272,10 +274,18 @@ fn do_test_unconf_chan(reload_node: bool, reorg_after_reload: bool) {
 
 #[test]
 fn test_unconf_chan() {
-	do_test_unconf_chan(true, true);
-	do_test_unconf_chan(false, true);
-	do_test_unconf_chan(true, false);
-	do_test_unconf_chan(false, false);
+	do_test_unconf_chan(true, true, ConnectStyle::BlockSkippingBestBlockFirst);
+	do_test_unconf_chan(false, true, ConnectStyle::BlockSkippingBestBlockFirst);
+	do_test_unconf_chan(true, false, ConnectStyle::BlockSkippingBestBlockFirst);
+	do_test_unconf_chan(false, false, ConnectStyle::BlockSkippingBestBlockFirst);
+}
+
+#[test]
+fn test_unconf_chan_via_listen() {
+	do_test_unconf_chan(true, true, ConnectStyle::FullBlockViaListen);
+	do_test_unconf_chan(false, true, ConnectStyle::FullBlockViaListen);
+	do_test_unconf_chan(true, false, ConnectStyle::FullBlockViaListen);
+	do_test_unconf_chan(false, false, ConnectStyle::FullBlockViaListen);
 }
 
 #[test]