From c36d23107c040f2b4f8109a6fb1d90f85dfefd39 Mon Sep 17 00:00:00 2001 From: Matt Corallo Date: Wed, 17 Oct 2018 18:19:55 -0400 Subject: [PATCH] Add Channel support for monitor-update-failed pausing --- src/ln/channel.rs | 175 ++++++++++++++++++++++++++++++++++++++++------ 1 file changed, 153 insertions(+), 22 deletions(-) diff --git a/src/ln/channel.rs b/src/ln/channel.rs index b5fa5c04c..6b1511387 100644 --- a/src/ln/channel.rs +++ b/src/ln/channel.rs @@ -248,28 +248,32 @@ enum ChannelState { /// "disconnected" and no updates are allowed until after we've done a channel_reestablish /// dance. PeerDisconnected = (1 << 7), + /// Flag which is set on ChannelFunded and FundingSent indicating the user has told us they + /// failed to update our ChannelMonitor somewhere and we should pause sending any outbound + /// messages until they've managed to do so. + MonitorUpdateFailed = (1 << 8), /// Flag which implies that we have sent a commitment_signed but are awaiting the responding /// revoke_and_ack message. During this time period, we can't generate new commitment_signed /// messages as then we will be unable to determine which HTLCs they included in their /// revoke_and_ack implicit ACK, so instead we have to hold them away temporarily to be sent /// later. /// Flag is set on ChannelFunded. - AwaitingRemoteRevoke = (1 << 8), + AwaitingRemoteRevoke = (1 << 9), /// Flag which is set on ChannelFunded or FundingSent after receiving a shutdown message from /// the remote end. If set, they may not add any new HTLCs to the channel, and we are expected /// to respond with our own shutdown message when possible. - RemoteShutdownSent = (1 << 9), + RemoteShutdownSent = (1 << 10), /// Flag which is set on ChannelFunded or FundingSent after sending a shutdown message. At this /// point, we may not add any new HTLCs to the channel. /// TODO: Investigate some kind of timeout mechanism by which point the remote end must provide /// us their shutdown. - LocalShutdownSent = (1 << 10), + LocalShutdownSent = (1 << 11), /// We've successfully negotiated a closing_signed dance. At this point ChannelManager is about /// to drop us, but we store this anyway. - ShutdownComplete = 2048, + ShutdownComplete = 4096, } const BOTH_SIDES_SHUTDOWN_MASK: u32 = (ChannelState::LocalShutdownSent as u32 | ChannelState::RemoteShutdownSent as u32); -const MULTI_STATE_FLAGS: u32 = (BOTH_SIDES_SHUTDOWN_MASK | ChannelState::PeerDisconnected as u32); +const MULTI_STATE_FLAGS: u32 = (BOTH_SIDES_SHUTDOWN_MASK | ChannelState::PeerDisconnected as u32 | ChannelState::MonitorUpdateFailed as u32); const INITIAL_COMMITMENT_NUMBER: u64 = (1 << 48) - 1; @@ -306,6 +310,12 @@ pub(super) struct Channel { pending_outbound_htlcs: Vec, holding_cell_htlc_updates: Vec, + monitor_pending_revoke_and_ack: bool, + monitor_pending_commitment_signed: bool, + monitor_pending_order: Option, + monitor_pending_forwards: Vec<(PendingForwardHTLCInfo, u64)>, + monitor_pending_failures: Vec<(HTLCSource, [u8; 32], HTLCFailReason)>, + // pending_update_fee is filled when sending and receiving update_fee // For outbound channel, feerate_per_kw is updated with the value from // pending_update_fee when revoke_and_ack is received @@ -509,6 +519,12 @@ impl Channel { next_remote_htlc_id: 0, channel_update_count: 1, + monitor_pending_revoke_and_ack: false, + monitor_pending_commitment_signed: false, + monitor_pending_order: None, + monitor_pending_forwards: Vec::new(), + monitor_pending_failures: Vec::new(), + #[cfg(debug_assertions)] max_commitment_tx_output_local: ::std::sync::Mutex::new((channel_value_satoshis * 1000 - push_msat, push_msat)), #[cfg(debug_assertions)] @@ -666,6 +682,12 @@ impl Channel { next_remote_htlc_id: 0, channel_update_count: 1, + monitor_pending_revoke_and_ack: false, + monitor_pending_commitment_signed: false, + monitor_pending_order: None, + monitor_pending_forwards: Vec::new(), + monitor_pending_failures: Vec::new(), + #[cfg(debug_assertions)] max_commitment_tx_output_local: ::std::sync::Mutex::new((msg.push_msat, msg.funding_satoshis * 1000 - msg.push_msat)), #[cfg(debug_assertions)] @@ -1166,7 +1188,7 @@ impl Channel { // can claim it even if the channel hits the chain before we see their next commitment. self.channel_monitor.provide_payment_preimage(&payment_hash_calc, &payment_preimage_arg); - if (self.channel_state & (ChannelState::AwaitingRemoteRevoke as u32 | ChannelState::PeerDisconnected as u32)) != 0 { + if (self.channel_state & (ChannelState::AwaitingRemoteRevoke as u32 | ChannelState::PeerDisconnected as u32 | ChannelState::MonitorUpdateFailed as u32)) != 0 { for pending_update in self.holding_cell_htlc_updates.iter() { match pending_update { &HTLCUpdateAwaitingACK::ClaimHTLC { htlc_id, .. } => { @@ -1243,7 +1265,7 @@ impl Channel { } // Now update local state: - if (self.channel_state & (ChannelState::AwaitingRemoteRevoke as u32 | ChannelState::PeerDisconnected as u32)) != 0 { + if (self.channel_state & (ChannelState::AwaitingRemoteRevoke as u32 | ChannelState::PeerDisconnected as u32 | ChannelState::MonitorUpdateFailed as u32)) != 0 { for pending_update in self.holding_cell_htlc_updates.iter() { match pending_update { &HTLCUpdateAwaitingACK::ClaimHTLC { htlc_id, .. } => { @@ -1461,11 +1483,13 @@ impl Channel { if self.channel_state & (ChannelState::PeerDisconnected as u32) == ChannelState::PeerDisconnected as u32 { return Err(ChannelError::Close("Peer sent funding_locked when we needed a channel_reestablish")); } - let non_shutdown_state = self.channel_state & (!BOTH_SIDES_SHUTDOWN_MASK); + + let non_shutdown_state = self.channel_state & (!MULTI_STATE_FLAGS); + if non_shutdown_state == ChannelState::FundingSent as u32 { self.channel_state |= ChannelState::TheirFundingLocked as u32; } else if non_shutdown_state == (ChannelState::FundingSent as u32 | ChannelState::OurFundingLocked as u32) { - self.channel_state = ChannelState::ChannelFunded as u32 | (self.channel_state & BOTH_SIDES_SHUTDOWN_MASK); + self.channel_state = ChannelState::ChannelFunded as u32 | (self.channel_state & MULTI_STATE_FLAGS); self.channel_update_count += 1; } else if self.channel_state & (ChannelState::ChannelFunded as u32) != 0 && // Note that funding_signed/funding_created will have decremented both by 1! @@ -1685,6 +1709,11 @@ impl Channel { } } } + if self.channel_state & (ChannelState::MonitorUpdateFailed as u32) == 0 { + // This is a response to our post-monitor-failed unfreeze messages, so we can clear the + // monitor_pending_order requirement as we won't re-send the monitor_pending messages. + self.monitor_pending_order = None; + } self.channel_monitor.provide_latest_local_commitment_tx_info(local_commitment_tx.0, local_keys, self.feerate_per_kw, htlcs_and_sigs); @@ -1708,6 +1737,12 @@ impl Channel { self.last_local_commitment_txn = new_local_commitment_txn; self.received_commitment_while_awaiting_raa = (self.channel_state & (ChannelState::AwaitingRemoteRevoke as u32)) != 0; + if (self.channel_state & ChannelState::MonitorUpdateFailed as u32) != 0 { + self.monitor_pending_revoke_and_ack = true; + self.monitor_pending_commitment_signed |= need_our_commitment; + return Err(HandleError{err: "Previous monitor update failure prevented generation of RAA", action: Some(ErrorAction::IgnoreError)}); + } + let (our_commitment_signed, monitor_update) = if need_our_commitment && (self.channel_state & (ChannelState::AwaitingRemoteRevoke as u32)) == 0 { // If we're AwaitingRemoteRevoke we can't send a new commitment here, but that's ok - // we'll send one right away when we get the revoke_and_ack when we @@ -1726,6 +1761,7 @@ impl Channel { /// Used to fulfill holding_cell_htlcs when we get a remote ack (or implicitly get it by them /// fulfilling or failing the last pending HTLC) fn free_holding_cell_htlcs(&mut self) -> Result, HandleError> { + assert_eq!(self.channel_state & ChannelState::MonitorUpdateFailed as u32, 0); if self.holding_cell_htlc_updates.len() != 0 || self.holding_cell_update_fee.is_some() { let mut htlc_updates = Vec::new(); mem::swap(&mut htlc_updates, &mut self.holding_cell_htlc_updates); @@ -1827,6 +1863,7 @@ impl Channel { if self.channel_state & (ChannelState::PeerDisconnected as u32) == ChannelState::PeerDisconnected as u32 { return Err(HandleError{err: "Peer sent revoke_and_ack when we needed a channel_reestablish", action: Some(msgs::ErrorAction::SendErrorMessage{msg: msgs::ErrorMessage{data: "Peer sent revoke_and_ack when we needed a channel_reestablish".to_string(), channel_id: msg.channel_id}})}); } + if let Some(their_prev_commitment_point) = self.their_prev_commitment_point { if PublicKey::from_secret_key(&self.secp_ctx, &secp_call!(SecretKey::from_slice(&self.secp_ctx, &msg.per_commitment_secret), "Peer provided an invalid per_commitment_secret", self.channel_id())) != their_prev_commitment_point { return Err(HandleError{err: "Got a revoke commitment secret which didn't correspond to their current pubkey", action: None}); @@ -1843,6 +1880,11 @@ impl Channel { self.their_cur_commitment_point = Some(msg.next_per_commitment_point); self.cur_remote_commitment_transaction_number -= 1; self.received_commitment_while_awaiting_raa = false; + if self.channel_state & (ChannelState::MonitorUpdateFailed as u32) == 0 { + // This is a response to our post-monitor-failed unfreeze messages, so we can clear the + // monitor_pending_order requirement as we won't re-send the monitor_pending messages. + self.monitor_pending_order = None; + } let mut to_forward_infos = Vec::new(); let mut revoked_htlcs = Vec::new(); @@ -1934,6 +1976,17 @@ impl Channel { } } + if (self.channel_state & ChannelState::MonitorUpdateFailed as u32) == ChannelState::MonitorUpdateFailed as u32 { + // We can't actually generate a new commitment transaction (incl by freeing holding + // cells) while we can't update the monitor, so we just return what we have. + if require_commitment { + self.monitor_pending_commitment_signed = true; + } + self.monitor_pending_forwards.append(&mut to_forward_infos); + self.monitor_pending_failures.append(&mut revoked_htlcs); + return Ok((None, Vec::new(), Vec::new(), self.channel_monitor.clone())); + } + match self.free_holding_cell_htlcs()? { Some(mut commitment_update) => { commitment_update.0.update_fail_htlcs.reserve(update_fail_htlcs.len()); @@ -1976,7 +2029,7 @@ impl Channel { panic!("Cannot update fee until channel is fully established and we haven't started shutting down"); } if !self.is_live() { - panic!("Cannot update fee while peer is disconnected (ChannelManager should have caught this)"); + panic!("Cannot update fee while peer is disconnected/we're awaiting a monitor update (ChannelManager should have caught this)"); } if (self.channel_state & (ChannelState::AwaitingRemoteRevoke as u32)) == (ChannelState::AwaitingRemoteRevoke as u32) { @@ -2068,6 +2121,60 @@ impl Channel { outbound_drops } + /// Indicates that a ChannelMonitor update failed to be stored by the client and further + /// updates are partially paused. + /// This must be called immediately after the call which generated the ChannelMonitor update + /// which failed, with the order argument set to the type of call it represented (ie a + /// commitment update or a revoke_and_ack generation). The messages which were generated from + /// that original call must *not* have been sent to the remote end, and must instead have been + /// dropped. They will be regenerated when monitor_updating_restored is called. + pub fn monitor_update_failed(&mut self, order: RAACommitmentOrder) { + assert_eq!(self.channel_state & ChannelState::MonitorUpdateFailed as u32, 0); + match order { + RAACommitmentOrder::CommitmentFirst => { + self.monitor_pending_revoke_and_ack = false; + self.monitor_pending_commitment_signed = true; + }, + RAACommitmentOrder::RevokeAndACKFirst => { + self.monitor_pending_revoke_and_ack = true; + self.monitor_pending_commitment_signed = false; + }, + } + self.monitor_pending_order = Some(order); + self.channel_state |= ChannelState::MonitorUpdateFailed as u32; + } + + /// Indicates that the latest ChannelMonitor update has been committed by the client + /// successfully and we should restore normal operation. Returns messages which should be sent + /// to the remote side. + pub fn monitor_updating_restored(&mut self) -> (Option, Option, RAACommitmentOrder, Vec<(PendingForwardHTLCInfo, u64)>, Vec<(HTLCSource, [u8; 32], HTLCFailReason)>) { + assert_eq!(self.channel_state & ChannelState::MonitorUpdateFailed as u32, ChannelState::MonitorUpdateFailed as u32); + self.channel_state &= !(ChannelState::MonitorUpdateFailed as u32); + + let mut forwards = Vec::new(); + mem::swap(&mut forwards, &mut self.monitor_pending_forwards); + let mut failures = Vec::new(); + mem::swap(&mut failures, &mut self.monitor_pending_failures); + + if self.channel_state & (ChannelState::PeerDisconnected as u32) != 0 { + // Leave monitor_pending_order so we can order our channel_reestablish responses + self.monitor_pending_revoke_and_ack = false; + self.monitor_pending_commitment_signed = false; + return (None, None, RAACommitmentOrder::RevokeAndACKFirst, forwards, failures); + } + + let raa = if self.monitor_pending_revoke_and_ack { + Some(self.get_last_revoke_and_ack()) + } else { None }; + let commitment_update = if self.monitor_pending_commitment_signed { + Some(self.get_last_commitment_update()) + } else { None }; + + self.monitor_pending_revoke_and_ack = false; + self.monitor_pending_commitment_signed = false; + (raa, commitment_update, self.monitor_pending_order.clone().unwrap(), forwards, failures) + } + pub fn update_fee(&mut self, fee_estimator: &FeeEstimator, msg: &msgs::UpdateFee) -> Result<(), ChannelError> { if self.channel_outbound { return Err(ChannelError::Close("Non-funding remote tried to update channel fee")); @@ -2171,7 +2278,12 @@ impl Channel { // Note that if we need to repeat our FundingLocked we'll do that in the next if block. None } else if msg.next_remote_commitment_number == (INITIAL_COMMITMENT_NUMBER - 1) - self.cur_local_commitment_transaction_number { - Some(self.get_last_revoke_and_ack()) + if self.channel_state & (ChannelState::MonitorUpdateFailed as u32) != 0 { + self.monitor_pending_revoke_and_ack = true; + None + } else { + Some(self.get_last_revoke_and_ack()) + } } else { return Err(ChannelError::Close("Peer attempted to reestablish channel with a very old local commitment transaction")); }; @@ -2183,6 +2295,7 @@ impl Channel { let our_next_remote_commitment_number = INITIAL_COMMITMENT_NUMBER - self.cur_remote_commitment_transaction_number + if (self.channel_state & ChannelState::AwaitingRemoteRevoke as u32) != 0 { 1 } else { 0 }; let resend_funding_locked = if msg.next_local_commitment_number == 1 && INITIAL_COMMITMENT_NUMBER - self.cur_local_commitment_transaction_number == 1 { + // We should never have to worry about MonitorUpdateFailed resending FundingLocked let next_per_commitment_secret = self.build_local_commitment_secret(self.cur_local_commitment_transaction_number); let next_per_commitment_point = PublicKey::from_secret_key(&self.secp_ctx, &next_per_commitment_secret); Some(msgs::FundingLocked { @@ -2191,11 +2304,11 @@ impl Channel { }) } else { None }; - let order = if self.received_commitment_while_awaiting_raa { - RAACommitmentOrder::CommitmentFirst - } else { - RAACommitmentOrder::RevokeAndACKFirst - }; + let order = self.monitor_pending_order.clone().unwrap_or(if self.received_commitment_while_awaiting_raa { + RAACommitmentOrder::CommitmentFirst + } else { + RAACommitmentOrder::RevokeAndACKFirst + }); if msg.next_local_commitment_number == our_next_remote_commitment_number { if required_revoke.is_some() { @@ -2204,7 +2317,8 @@ impl Channel { log_debug!(self, "Reconnected channel {} with no loss", log_bytes!(self.channel_id())); } - if (self.channel_state & (ChannelState::AwaitingRemoteRevoke as u32)) == 0 { + if (self.channel_state & (ChannelState::AwaitingRemoteRevoke as u32 | ChannelState::MonitorUpdateFailed as u32)) == 0 && + self.monitor_pending_order.is_none() { // monitor_pending_order indicates we're waiting on a response to a unfreeze // We're up-to-date and not waiting on a remote revoke (if we are our // channel_reestablish should result in them sending a revoke_and_ack), but we may // have received some updates while we were disconnected. Free the holding cell @@ -2232,6 +2346,14 @@ impl Channel { log_debug!(self, "Reconnected channel {} with only lost remote commitment tx", log_bytes!(self.channel_id())); } + // If monitor_pending_order is set, it must be CommitmentSigned if we have no RAA + debug_assert!(self.monitor_pending_order != Some(RAACommitmentOrder::RevokeAndACKFirst) || required_revoke.is_some()); + + if self.channel_state & (ChannelState::MonitorUpdateFailed as u32) != 0 { + self.monitor_pending_commitment_signed = true; + return Ok((resend_funding_locked, None, None, None, order)); + } + return Ok((resend_funding_locked, required_revoke, Some(self.get_last_commitment_update()), None, order)); } else { return Err(ChannelError::Close("Peer attempted to reestablish channel with a very old remote commitment transaction")); @@ -2569,7 +2691,13 @@ impl Channel { /// is_usable() and considers things like the channel being temporarily disabled. /// Allowed in any state (including after shutdown) pub fn is_live(&self) -> bool { - self.is_usable() && (self.channel_state & (ChannelState::PeerDisconnected as u32) == 0) + self.is_usable() && (self.channel_state & (ChannelState::PeerDisconnected as u32 | ChannelState::MonitorUpdateFailed as u32) == 0) + } + + /// Returns true if this channel has been marked as awaiting a monitor update to move forward. + /// Allowed in any state (including after shutdown) + pub fn is_awaiting_monitor_update(&self) -> bool { + (self.channel_state & ChannelState::MonitorUpdateFailed as u32) != 0 } /// Returns true if funding_created was sent/received. @@ -2883,14 +3011,14 @@ impl Channel { return Err(HandleError{err: "Cannot send less than their minimum HTLC value", action: None}); } - if (self.channel_state & (ChannelState::PeerDisconnected as u32)) == (ChannelState::PeerDisconnected as u32) { + if (self.channel_state & (ChannelState::PeerDisconnected as u32 | ChannelState::MonitorUpdateFailed as u32)) != 0 { // Note that this should never really happen, if we're !is_live() on receipt of an // incoming HTLC for relay will result in us rejecting the HTLC and we won't allow // the user to send directly into a !is_live() channel. However, if we // disconnected during the time the previous hop was doing the commitment dance we may // end up getting here after the forwarding delay. In any case, returning an // IgnoreError will get ChannelManager to do the right thing and fail backwards now. - return Err(HandleError{err: "Cannot send an HTLC while disconnected", action: Some(ErrorAction::IgnoreError)}); + return Err(HandleError{err: "Cannot send an HTLC while disconnected/frozen for channel monitor update", action: Some(ErrorAction::IgnoreError)}); } let (outbound_htlc_count, htlc_outbound_value_msat) = self.get_outbound_pending_htlc_stats(); @@ -2972,6 +3100,9 @@ impl Channel { if (self.channel_state & (ChannelState::PeerDisconnected as u32)) == (ChannelState::PeerDisconnected as u32) { panic!("Cannot create commitment tx while disconnected, as send_htlc will have returned an Err so a send_commitment precondition has been violated"); } + if (self.channel_state & (ChannelState::MonitorUpdateFailed as u32)) == (ChannelState::PeerDisconnected as u32) { + panic!("Cannot create commitment tx while awaiting monitor update unfreeze, as send_htlc will have returned an Err so a send_commitment precondition has been violated"); + } let mut have_updates = self.pending_update_fee.is_some(); for htlc in self.pending_outbound_htlcs.iter() { if let OutboundHTLCState::LocalAnnounced(_) = htlc.state { @@ -3080,8 +3211,8 @@ impl Channel { } } assert_eq!(self.channel_state & ChannelState::ShutdownComplete as u32, 0); - if self.channel_state & (ChannelState::PeerDisconnected as u32) == ChannelState::PeerDisconnected as u32 { - return Err(APIError::ChannelUnavailable{err: "Cannot begin shutdown while peer is disconnected, maybe force-close instead?"}); + if self.channel_state & (ChannelState::PeerDisconnected as u32 | ChannelState::MonitorUpdateFailed as u32) != 0 { + return Err(APIError::ChannelUnavailable{err: "Cannot begin shutdown while peer is disconnected or we're waiting on a monitor update, maybe force-close instead?"}); } let our_closing_script = self.get_closing_scriptpubkey(); -- 2.39.5