}
macro_rules! handle_chan_restoration_locked {
- ($self: expr, $channel_lock: expr, $channel_state: expr, $channel_entry: expr,
- $raa: expr, $commitment_update: expr, $order: expr,
+ ($self: ident, $channel_lock: expr, $channel_state: expr, $channel_entry: expr,
+ $raa: expr, $commitment_update: expr, $order: expr, $chanmon_update: expr,
$pending_forwards: expr, $funding_broadcastable: expr, $funding_locked: expr) => { {
let mut htlc_forwards = None;
let counterparty_node_id = $channel_entry.get().get_counterparty_node_id();
- {
- if !$pending_forwards.is_empty() {
+ let chanmon_update: Option<ChannelMonitorUpdate> = $chanmon_update; // Force type-checking to resolve
+ let chanmon_update_is_none = chanmon_update.is_none();
+ let res = loop {
+ let forwards: Vec<(PendingHTLCInfo, u64)> = $pending_forwards; // Force type-checking to resolve
+ if !forwards.is_empty() {
htlc_forwards = Some(($channel_entry.get().get_short_channel_id().expect("We can't have pending forwards before funding confirmation"),
- $channel_entry.get().get_funding_txo().unwrap(), $pending_forwards));
+ $channel_entry.get().get_funding_txo().unwrap(), forwards));
+ }
+
+ if chanmon_update.is_some() {
+ // On reconnect, we, by definition, only resend a funding_locked if there have been
+ // no commitment updates, so the only channel monitor update which could also be
+ // associated with a funding_locked would be the funding_created/funding_signed
+ // monitor update. That monitor update failing implies that we won't send
+ // funding_locked until it's been updated, so we can't have a funding_locked and a
+ // monitor update here (so we don't bother to handle it correctly below).
+ assert!($funding_locked.is_none());
+ // A channel monitor update makes no sense without either a funding_locked or a
+ // commitment update to process after it. Since we can't have a funding_locked, we
+ // only bother to handle the monitor-update + commitment_update case below.
+ assert!($commitment_update.is_some());
+ }
+
+ if let Some(msg) = $funding_locked {
+ // Similar to the above, this implies that we're letting the funding_locked fly
+ // before it should be allowed to.
+ assert!(chanmon_update.is_none());
+ $channel_state.pending_msg_events.push(events::MessageSendEvent::SendFundingLocked {
+ node_id: counterparty_node_id,
+ msg,
+ });
+ if let Some(announcement_sigs) = $self.get_announcement_sigs($channel_entry.get()) {
+ $channel_state.pending_msg_events.push(events::MessageSendEvent::SendAnnouncementSignatures {
+ node_id: counterparty_node_id,
+ msg: announcement_sigs,
+ });
+ }
+ $channel_state.short_to_id.insert($channel_entry.get().get_short_channel_id().unwrap(), $channel_entry.get().channel_id());
+ }
+
+ let funding_broadcastable: Option<Transaction> = $funding_broadcastable; // Force type-checking to resolve
+ if let Some(monitor_update) = chanmon_update {
+ // We only ever broadcast a funding transaction in response to a funding_signed
+ // message and the resulting monitor update. Thus, on channel_reestablish
+ // message handling we can't have a funding transaction to broadcast. When
+ // processing a monitor update finishing resulting in a funding broadcast, we
+ // cannot have a second monitor update, thus this case would indicate a bug.
+ assert!(funding_broadcastable.is_none());
+ // Given we were just reconnected or finished updating a channel monitor, the
+ // only case where we can get a new ChannelMonitorUpdate would be if we also
+ // have some commitment updates to send as well.
+ assert!($commitment_update.is_some());
+ if let Err(e) = $self.chain_monitor.update_channel($channel_entry.get().get_funding_txo().unwrap(), monitor_update) {
+ // channel_reestablish doesn't guarantee the order it returns is sensical
+ // for the messages it returns, but if we're setting what messages to
+ // re-transmit on monitor update success, we need to make sure it is sane.
+ let mut order = $order;
+ if $raa.is_none() {
+ order = RAACommitmentOrder::CommitmentFirst;
+ }
+ break handle_monitor_err!($self, e, $channel_state, $channel_entry, order, $raa.is_some(), true);
+ }
}
macro_rules! handle_cs { () => {
handle_cs!();
},
}
- if let Some(tx) = $funding_broadcastable {
+ if let Some(tx) = funding_broadcastable {
log_info!($self.logger, "Broadcasting funding transaction with txid {}", tx.txid());
$self.tx_broadcaster.broadcast_transaction(&tx);
}
- if let Some(msg) = $funding_locked {
- $channel_state.pending_msg_events.push(events::MessageSendEvent::SendFundingLocked {
- node_id: counterparty_node_id,
- msg,
- });
- if let Some(announcement_sigs) = $self.get_announcement_sigs($channel_entry.get()) {
- $channel_state.pending_msg_events.push(events::MessageSendEvent::SendAnnouncementSignatures {
- node_id: counterparty_node_id,
- msg: announcement_sigs,
- });
- }
- $channel_state.short_to_id.insert($channel_entry.get().get_short_channel_id().unwrap(), $channel_entry.get().channel_id());
- }
+ break Ok(());
+ };
+
+ if chanmon_update_is_none {
+ // If there was no ChannelMonitorUpdate, we should never generate an Err in the res loop
+ // above. Doing so would imply calling handle_err!() from channel_monitor_updated() which
+ // should *never* end up calling back to `chain_monitor.update_channel()`.
+ assert!(res.is_ok());
}
- htlc_forwards
+
+ (htlc_forwards, res, counterparty_node_id)
} }
}
macro_rules! post_handle_chan_restoration {
- ($self: expr, $locked_res: expr) => { {
- let htlc_forwards = $locked_res;
+ ($self: ident, $locked_res: expr) => { {
+ let (htlc_forwards, res, counterparty_node_id) = $locked_res;
+
+ let _ = handle_error!($self, res, counterparty_node_id);
if let Some(forwards) = htlc_forwards {
$self.forward_htlcs(&mut [forwards][..]);
}
let (raa, commitment_update, order, pending_forwards, pending_failures, funding_broadcastable, funding_locked) = channel.get_mut().monitor_updating_restored(&self.logger);
- (pending_failures, handle_chan_restoration_locked!(self, channel_lock, channel_state, channel, raa, commitment_update, order, pending_forwards, funding_broadcastable, funding_locked))
+ (pending_failures, handle_chan_restoration_locked!(self, channel_lock, channel_state, channel, raa, commitment_update, order, None, pending_forwards, funding_broadcastable, funding_locked))
};
post_handle_chan_restoration!(self, chan_restoration_res);
for failure in pending_failures.drain(..) {
}
fn internal_channel_reestablish(&self, counterparty_node_id: &PublicKey, msg: &msgs::ChannelReestablish) -> Result<(), MsgHandleErrInternal> {
- let mut channel_state_lock = self.channel_state.lock().unwrap();
- let channel_state = &mut *channel_state_lock;
+ let chan_restoration_res = {
+ let mut channel_state_lock = self.channel_state.lock().unwrap();
+ let channel_state = &mut *channel_state_lock;
- match channel_state.by_id.entry(msg.channel_id) {
- hash_map::Entry::Occupied(mut chan) => {
- if chan.get().get_counterparty_node_id() != *counterparty_node_id {
- return Err(MsgHandleErrInternal::send_err_msg_no_close("Got a message for a channel from the wrong node!".to_owned(), msg.channel_id));
- }
- // Currently, we expect all holding cell update_adds to be dropped on peer
- // disconnect, so Channel's reestablish will never hand us any holding cell
- // freed HTLCs to fail backwards. If in the future we no longer drop pending
- // add-HTLCs on disconnect, we may be handed HTLCs to fail backwards here.
- let (funding_locked, revoke_and_ack, commitment_update, monitor_update_opt, mut order, shutdown) =
- try_chan_entry!(self, chan.get_mut().channel_reestablish(msg, &self.logger), channel_state, chan);
- if let Some(monitor_update) = monitor_update_opt {
- if let Err(e) = self.chain_monitor.update_channel(chan.get().get_funding_txo().unwrap(), monitor_update) {
- // channel_reestablish doesn't guarantee the order it returns is sensical
- // for the messages it returns, but if we're setting what messages to
- // re-transmit on monitor update success, we need to make sure it is sane.
- if revoke_and_ack.is_none() {
- order = RAACommitmentOrder::CommitmentFirst;
- }
- if commitment_update.is_none() {
- order = RAACommitmentOrder::RevokeAndACKFirst;
- }
- return_monitor_err!(self, e, channel_state, chan, order, revoke_and_ack.is_some(), commitment_update.is_some());
- //TODO: Resend the funding_locked if needed once we get the monitor running again
- }
- }
- if let Some(msg) = funding_locked {
- channel_state.pending_msg_events.push(events::MessageSendEvent::SendFundingLocked {
- node_id: counterparty_node_id.clone(),
- msg
- });
- }
- macro_rules! send_raa { () => {
- if let Some(msg) = revoke_and_ack {
- channel_state.pending_msg_events.push(events::MessageSendEvent::SendRevokeAndACK {
- node_id: counterparty_node_id.clone(),
- msg
- });
+ match channel_state.by_id.entry(msg.channel_id) {
+ hash_map::Entry::Occupied(mut chan) => {
+ if chan.get().get_counterparty_node_id() != *counterparty_node_id {
+ return Err(MsgHandleErrInternal::send_err_msg_no_close("Got a message for a channel from the wrong node!".to_owned(), msg.channel_id));
}
- } }
- macro_rules! send_cu { () => {
- if let Some(updates) = commitment_update {
- channel_state.pending_msg_events.push(events::MessageSendEvent::UpdateHTLCs {
+ // Currently, we expect all holding cell update_adds to be dropped on peer
+ // disconnect, so Channel's reestablish will never hand us any holding cell
+ // freed HTLCs to fail backwards. If in the future we no longer drop pending
+ // add-HTLCs on disconnect, we may be handed HTLCs to fail backwards here.
+ let (funding_locked, revoke_and_ack, commitment_update, monitor_update_opt, order, shutdown) =
+ try_chan_entry!(self, chan.get_mut().channel_reestablish(msg, &self.logger), channel_state, chan);
+ if let Some(msg) = shutdown {
+ channel_state.pending_msg_events.push(events::MessageSendEvent::SendShutdown {
node_id: counterparty_node_id.clone(),
- updates
+ msg,
});
}
- } }
- match order {
- RAACommitmentOrder::RevokeAndACKFirst => {
- send_raa!();
- send_cu!();
- },
- RAACommitmentOrder::CommitmentFirst => {
- send_cu!();
- send_raa!();
- },
- }
- if let Some(msg) = shutdown {
- channel_state.pending_msg_events.push(events::MessageSendEvent::SendShutdown {
- node_id: counterparty_node_id.clone(),
- msg,
- });
- }
- Ok(())
- },
- hash_map::Entry::Vacant(_) => return Err(MsgHandleErrInternal::send_err_msg_no_close("Failed to find corresponding channel".to_owned(), msg.channel_id))
- }
+ handle_chan_restoration_locked!(self, channel_state_lock, channel_state, chan, revoke_and_ack, commitment_update, order, monitor_update_opt, Vec::new(), None, funding_locked)
+ },
+ hash_map::Entry::Vacant(_) => return Err(MsgHandleErrInternal::send_err_msg_no_close("Failed to find corresponding channel".to_owned(), msg.channel_id))
+ }
+ };
+ post_handle_chan_restoration!(self, chan_restoration_res);
+ Ok(())
}
/// Begin Update fee process. Allowed only on an outbound channel.