Contribute funding inputs on accepting dual-funded channel #3735
Conversation
|
👋 Thanks for assigning @jkczyz as a reviewer! |
dunxen
force-pushed
the
2025-04-contributeinputs
branch
from
e22fa33 to
fabbf86
Compare
dunxen
force-pushed
the
2025-04-contributeinputs
branch
4 times, most recently
from
c42f3d0 to
186d66b
Compare
dunxen
force-pushed
the
2025-04-contributeinputs
branch
from
186d66b to
c7d8e89
Compare
dunxen
force-pushed
the
2025-04-contributeinputs
branch
from
c7d8e89 to
540b8a7
Compare
|
Looks like this needs a rebase after #3637 |
dunxen
force-pushed
the
2025-04-contributeinputs
branch
from
540b8a7 to
94bf848
Compare
ugh right! |
|
Don't the first three commits need to be dropped? |
dunxen
force-pushed
the
2025-04-contributeinputs
branch
2 times, most recently
from
67504f0 to
9f276d8
Compare
|
Just needs node reload tests but it can still get some initial review. |
When receiving a commitment_signed message, if there are any pending splices then we are expected to receive the message as a part of a batch. Otherwise, the spec dictates that we should send an error and fail the channel.
The `FundingTransactionReadyForSignatures` event requests witnesses from the client for their contributed inputs to an interactively constructed transaction. The client calls `ChannelManager::funding_transaction_signed` to provide the witnesses to LDK.
…annel We introduce a `ChannelManager::accept_inbound_channel_with_contribution` method allowing contributing to the overall channel capacity of an inbound dual-funded channel by contributing inputs.
dunxen
force-pushed
the
2025-04-contributeinputs
branch
from
9f276d8 to
60f58b1
Compare
| { | ||
| if self.is_awaiting_initial_mon_persist() { | ||
| log_debug!(logger, "Not sending tx_signatures: a monitor update is in progress. Setting monitor_pending_tx_signatures."); | ||
| self.context.monitor_pending_tx_signatures = Some(holder_tx_signatures); |
There was a problem hiding this comment.
Thoughts on just tracking this as an Option<()> and we get the message from the signing session instead?
| /// </div> | ||
| pub fn funding_transaction_signed( | ||
| &self, channel_id: &ChannelId, counterparty_node_id: &PublicKey, transaction: Transaction, | ||
| ) -> Result<(), APIError> { |
There was a problem hiding this comment.
We need to acquire the PersistenceNotifierGuard throughout this method
|
|
||
| fn length_limit_holder_input_prev_txs( | ||
| funding_inputs: Vec<(TxIn, Transaction, Weight)>, | ||
| ) -> Result<Vec<(TxIn, TransactionU16LenLimited, Weight)>, APIError> { |
There was a problem hiding this comment.
Let's return a structured type here that also includes the corresponding TxOut as a field so we don't have to continue doing checks that it exists everywhere.
| @@ -11110,6 +11111,31 @@ where | |||
| script_pubkey: funding.get_funding_redeemscript().to_p2wsh(), | |||
| }; | |||
|
|
|||
| // Optionally add change output | |||
There was a problem hiding this comment.
Wouldn't this already be covered by begin_interactive_funding_tx_construction?
| @@ -11110,6 +11111,31 @@ where | |||
| script_pubkey: funding.get_funding_redeemscript().to_p2wsh(), | |||
| }; | |||
|
|
|||
| // Optionally add change output | |||
| let change_script = signer_provider.get_destination_script(context.channel_keys_id) | |||
There was a problem hiding this comment.
Maybe let the user optionally provide a change address along with their inputs?
|
Since the |
Very WIP with upcoming commits.
Based on #3637.