Split `DiscardFunding` from `SpliceFailed` event by jkczyz · Pull Request #4388 · lightningdevkit/rust-lightning

jkczyz · 2026-02-05T17:17:59Z

ChannelManager::funding_contributed consumes the passed in FundingContribution. Normally, the associated inputs and outputs should be given back via a SpliceFailed event. However, there are cases where the splice hasn't actually failed, so using that event would be misleading. The inputs and outputs still need to be returned to the user, which is done here via a dedicated DiscardFunding event.

Based on #4382.

ldk-reviews-bot · 2026-02-05T17:18:03Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

TheBlueMatt · 2026-02-19T23:21:20Z

Also CI is quite sad.

codecov · 2026-02-21T20:16:16Z

Codecov Report

❌ Patch coverage is 87.70764% with 37 lines in your changes missing coverage. Please review.
✅ Project coverage is 85.94%. Comparing base (df57978) to head (24062c0).
⚠️ Report is 15 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/channelmanager.rs	85.25%	23 Missing ⚠️
lightning/src/ln/channel.rs	90.62%	6 Missing ⚠️
lightning/src/ln/functional_test_utils.rs	86.48%	5 Missing ⚠️
lightning/src/ln/interactivetxs.rs	88.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4388 +/- ##
==========================================
+ Coverage 85.89% 85.94% +0.04% 
==========================================
Files 159 159 Lines 104402 104607 +205 Branches 104402 104607 +205 ==========================================
+ Hits 89681 89901 +220 + Misses 12216 12194 -22 - Partials 2505 2512 +7

Flag	Coverage Δ
tests	`85.94% <87.70%> (+0.04%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2026-02-23T16:52:26Z

Rebased to address a merge conflict.

TheBlueMatt · 2026-02-23T19:27:35Z

CI is quite sad.

wpaulino

LGTM, feel free to squash

jkczyz · 2026-02-23T21:45:28Z

Squashed

TheBlueMatt · 2026-02-24T13:58:44Z

+ return match unique_contributions {
+ None => Err(QuiescentError::DoNothing),
+ Some((contributed_inputs, contributed_outputs)) => {
+ Err(QuiescentError::FailSplice(SpliceFundingFailed {


Commented on discord, but if the intent of the split return logic is to ensure that downstream code doesnt' get a SpliceFailed event that may be confused for being in relation to an earlier splice (due to not having seen SplicePending), this isn't quite sufficient - if we're still constructing the transaction we can't generate a SpliceFailed because it hasn't yet gotten to SplicePending.

Hmm... SpliceFailed and SplicePending are mutually exclusive. Once we generated a SplicePending event, we won't generate SpliceFailed.
However, this should probably be return QuiescentError::DiscardFunding as per my earlier reasoning. That is, we have an active splice being negotiated, so we should error and discard the funding without emitting a SpliceFailed event.

RecipientOnionFields::secret_only requires an amount now, but when do_abandon_splice_quiescent_action_on_shutdown was introduced it was based on an earlier commit.

Add a helper function to assert that SpliceFailed events contain the expected channel_id and contributed inputs/outputs. This ensures that tests verify the contributions match what was originally provided. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

The following types have methods for returning contributed inputs and outputs: - FundingNegotiationContext - InteractiveTxConstructor - InteractiveTxSigningSession - ConstructedTransaction Having iterators for these can avoid allocations, which is useful for filtering contributed input and outputs when producing DiscardFunding events. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

When a splice fails, users need to reclaim UTXOs they contributed to the funding transaction. Previously, the contributed inputs and outputs were included in the SpliceFailed event. This commit splits them into a separate DiscardFunding event with a new FundingInfo::Contribution variant, providing a consistent interface for UTXO cleanup across all funding failure scenarios. Changes: - Add FundingInfo::Contribution variant to hold inputs/outputs for DiscardFunding events - Remove contributed_inputs/outputs fields from SpliceFailed event - Add QuiescentError enum for better error handling in funding_contributed - Emit DiscardFunding on all funding_contributed error paths - Filter duplicate inputs/outputs when contribution overlaps existing pending contribution - Return Err(APIError) from funding_contributed on all error cases - Add comprehensive test coverage for funding_contributed error paths Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

jkczyz · 2026-02-24T18:41:46Z

Rebased and added a fix to a silent merge conflict.

jkczyz · 2026-02-05T17:17:59Z

ChannelManager::funding_contributed consumes the passed in FundingContribution. Normally, the associated inputs and outputs should be given back via a SpliceFailed event. However, there are cases where the splice hasn't actually failed, so using that event would be misleading. The inputs and outputs still need to be returned to the user, which is done here via a dedicated DiscardFunding event.

Based on #4382.

ldk-reviews-bot · 2026-02-05T17:18:03Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

TheBlueMatt · 2026-02-19T23:21:20Z

Also CI is quite sad.

codecov · 2026-02-21T20:16:16Z

Codecov Report

❌ Patch coverage is 87.70764% with 37 lines in your changes missing coverage. Please review.
✅ Project coverage is 85.94%. Comparing base (df57978) to head (24062c0).
⚠️ Report is 15 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/channelmanager.rs	85.25%	23 Missing ⚠️
lightning/src/ln/channel.rs	90.62%	6 Missing ⚠️
lightning/src/ln/functional_test_utils.rs	86.48%	5 Missing ⚠️
lightning/src/ln/interactivetxs.rs	88.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4388 +/- ##
==========================================
+ Coverage 85.89% 85.94% +0.04% 
==========================================
Files 159 159 Lines 104402 104607 +205 Branches 104402 104607 +205 ==========================================
+ Hits 89681 89901 +220 + Misses 12216 12194 -22 - Partials 2505 2512 +7

Flag	Coverage Δ
tests	`85.94% <87.70%> (+0.04%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2026-02-23T16:52:26Z

Rebased to address a merge conflict.

TheBlueMatt · 2026-02-23T19:27:35Z

CI is quite sad.

wpaulino

LGTM, feel free to squash

jkczyz · 2026-02-23T21:45:28Z

Squashed

TheBlueMatt · 2026-02-24T13:58:44Z

+ return match unique_contributions {
+ None => Err(QuiescentError::DoNothing),
+ Some((contributed_inputs, contributed_outputs)) => {
+ Err(QuiescentError::FailSplice(SpliceFundingFailed {


Commented on discord, but if the intent of the split return logic is to ensure that downstream code doesnt' get a SpliceFailed event that may be confused for being in relation to an earlier splice (due to not having seen SplicePending), this isn't quite sufficient - if we're still constructing the transaction we can't generate a SpliceFailed because it hasn't yet gotten to SplicePending.

Hmm... SpliceFailed and SplicePending are mutually exclusive. Once we generated a SplicePending event, we won't generate SpliceFailed.
However, this should probably be return QuiescentError::DiscardFunding as per my earlier reasoning. That is, we have an active splice being negotiated, so we should error and discard the funding without emitting a SpliceFailed event.

RecipientOnionFields::secret_only requires an amount now, but when do_abandon_splice_quiescent_action_on_shutdown was introduced it was based on an earlier commit.

Add a helper function to assert that SpliceFailed events contain the expected channel_id and contributed inputs/outputs. This ensures that tests verify the contributions match what was originally provided. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

The following types have methods for returning contributed inputs and outputs: - FundingNegotiationContext - InteractiveTxConstructor - InteractiveTxSigningSession - ConstructedTransaction Having iterators for these can avoid allocations, which is useful for filtering contributed input and outputs when producing DiscardFunding events. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

When a splice fails, users need to reclaim UTXOs they contributed to the funding transaction. Previously, the contributed inputs and outputs were included in the SpliceFailed event. This commit splits them into a separate DiscardFunding event with a new FundingInfo::Contribution variant, providing a consistent interface for UTXO cleanup across all funding failure scenarios. Changes: - Add FundingInfo::Contribution variant to hold inputs/outputs for DiscardFunding events - Remove contributed_inputs/outputs fields from SpliceFailed event - Add QuiescentError enum for better error handling in funding_contributed - Emit DiscardFunding on all funding_contributed error paths - Filter duplicate inputs/outputs when contribution overlaps existing pending contribution - Return Err(APIError) from funding_contributed on all error cases - Add comprehensive test coverage for funding_contributed error paths Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

jkczyz · 2026-02-24T18:41:46Z

Rebased and added a fix to a silent merge conflict.

jkczyz · 2026-02-05T17:17:59Z

ChannelManager::funding_contributed consumes the passed in FundingContribution. Normally, the associated inputs and outputs should be given back via a SpliceFailed event. However, there are cases where the splice hasn't actually failed, so using that event would be misleading. The inputs and outputs still need to be returned to the user, which is done here via a dedicated DiscardFunding event.

Based on #4382.

ldk-reviews-bot · 2026-02-05T17:18:03Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

TheBlueMatt · 2026-02-19T23:21:20Z

Also CI is quite sad.

codecov · 2026-02-21T20:16:16Z

Codecov Report

❌ Patch coverage is 87.70764% with 37 lines in your changes missing coverage. Please review.
✅ Project coverage is 85.94%. Comparing base (df57978) to head (24062c0).
⚠️ Report is 15 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/channelmanager.rs	85.25%	23 Missing ⚠️
lightning/src/ln/channel.rs	90.62%	6 Missing ⚠️
lightning/src/ln/functional_test_utils.rs	86.48%	5 Missing ⚠️
lightning/src/ln/interactivetxs.rs	88.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4388 +/- ##
==========================================
+ Coverage 85.89% 85.94% +0.04% 
==========================================
Files 159 159 Lines 104402 104607 +205 Branches 104402 104607 +205 ==========================================
+ Hits 89681 89901 +220 + Misses 12216 12194 -22 - Partials 2505 2512 +7

Flag	Coverage Δ
tests	`85.94% <87.70%> (+0.04%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2026-02-23T16:52:26Z

Rebased to address a merge conflict.

TheBlueMatt · 2026-02-23T19:27:35Z

CI is quite sad.

wpaulino

LGTM, feel free to squash

jkczyz · 2026-02-23T21:45:28Z

Squashed

TheBlueMatt · 2026-02-24T13:58:44Z

+ return match unique_contributions {
+ None => Err(QuiescentError::DoNothing),
+ Some((contributed_inputs, contributed_outputs)) => {
+ Err(QuiescentError::FailSplice(SpliceFundingFailed {


Commented on discord, but if the intent of the split return logic is to ensure that downstream code doesnt' get a SpliceFailed event that may be confused for being in relation to an earlier splice (due to not having seen SplicePending), this isn't quite sufficient - if we're still constructing the transaction we can't generate a SpliceFailed because it hasn't yet gotten to SplicePending.

Hmm... SpliceFailed and SplicePending are mutually exclusive. Once we generated a SplicePending event, we won't generate SpliceFailed.
However, this should probably be return QuiescentError::DiscardFunding as per my earlier reasoning. That is, we have an active splice being negotiated, so we should error and discard the funding without emitting a SpliceFailed event.

RecipientOnionFields::secret_only requires an amount now, but when do_abandon_splice_quiescent_action_on_shutdown was introduced it was based on an earlier commit.

Add a helper function to assert that SpliceFailed events contain the expected channel_id and contributed inputs/outputs. This ensures that tests verify the contributions match what was originally provided. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

The following types have methods for returning contributed inputs and outputs: - FundingNegotiationContext - InteractiveTxConstructor - InteractiveTxSigningSession - ConstructedTransaction Having iterators for these can avoid allocations, which is useful for filtering contributed input and outputs when producing DiscardFunding events. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

When a splice fails, users need to reclaim UTXOs they contributed to the funding transaction. Previously, the contributed inputs and outputs were included in the SpliceFailed event. This commit splits them into a separate DiscardFunding event with a new FundingInfo::Contribution variant, providing a consistent interface for UTXO cleanup across all funding failure scenarios. Changes: - Add FundingInfo::Contribution variant to hold inputs/outputs for DiscardFunding events - Remove contributed_inputs/outputs fields from SpliceFailed event - Add QuiescentError enum for better error handling in funding_contributed - Emit DiscardFunding on all funding_contributed error paths - Filter duplicate inputs/outputs when contribution overlaps existing pending contribution - Return Err(APIError) from funding_contributed on all error cases - Add comprehensive test coverage for funding_contributed error paths Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

jkczyz · 2026-02-24T18:41:46Z

Rebased and added a fix to a silent merge conflict.

jkczyz · 2026-02-05T17:17:59Z

ChannelManager::funding_contributed consumes the passed in FundingContribution. Normally, the associated inputs and outputs should be given back via a SpliceFailed event. However, there are cases where the splice hasn't actually failed, so using that event would be misleading. The inputs and outputs still need to be returned to the user, which is done here via a dedicated DiscardFunding event.

Based on #4382.

ldk-reviews-bot · 2026-02-05T17:18:03Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

TheBlueMatt · 2026-02-19T23:21:20Z

Also CI is quite sad.

codecov · 2026-02-21T20:16:16Z

Codecov Report

❌ Patch coverage is 87.70764% with 37 lines in your changes missing coverage. Please review.
✅ Project coverage is 85.94%. Comparing base (df57978) to head (24062c0).
⚠️ Report is 15 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/channelmanager.rs	85.25%	23 Missing ⚠️
lightning/src/ln/channel.rs	90.62%	6 Missing ⚠️
lightning/src/ln/functional_test_utils.rs	86.48%	5 Missing ⚠️
lightning/src/ln/interactivetxs.rs	88.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4388 +/- ##
==========================================
+ Coverage 85.89% 85.94% +0.04% 
==========================================
Files 159 159 Lines 104402 104607 +205 Branches 104402 104607 +205 ==========================================
+ Hits 89681 89901 +220 + Misses 12216 12194 -22 - Partials 2505 2512 +7

Flag	Coverage Δ
tests	`85.94% <87.70%> (+0.04%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2026-02-23T16:52:26Z

Rebased to address a merge conflict.

TheBlueMatt · 2026-02-23T19:27:35Z

CI is quite sad.

wpaulino

LGTM, feel free to squash

jkczyz · 2026-02-23T21:45:28Z

Squashed

TheBlueMatt · 2026-02-24T13:58:44Z

+ return match unique_contributions {
+ None => Err(QuiescentError::DoNothing),
+ Some((contributed_inputs, contributed_outputs)) => {
+ Err(QuiescentError::FailSplice(SpliceFundingFailed {


Commented on discord, but if the intent of the split return logic is to ensure that downstream code doesnt' get a SpliceFailed event that may be confused for being in relation to an earlier splice (due to not having seen SplicePending), this isn't quite sufficient - if we're still constructing the transaction we can't generate a SpliceFailed because it hasn't yet gotten to SplicePending.

Hmm... SpliceFailed and SplicePending are mutually exclusive. Once we generated a SplicePending event, we won't generate SpliceFailed.
However, this should probably be return QuiescentError::DiscardFunding as per my earlier reasoning. That is, we have an active splice being negotiated, so we should error and discard the funding without emitting a SpliceFailed event.

RecipientOnionFields::secret_only requires an amount now, but when do_abandon_splice_quiescent_action_on_shutdown was introduced it was based on an earlier commit.

Add a helper function to assert that SpliceFailed events contain the expected channel_id and contributed inputs/outputs. This ensures that tests verify the contributions match what was originally provided. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

The following types have methods for returning contributed inputs and outputs: - FundingNegotiationContext - InteractiveTxConstructor - InteractiveTxSigningSession - ConstructedTransaction Having iterators for these can avoid allocations, which is useful for filtering contributed input and outputs when producing DiscardFunding events. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

When a splice fails, users need to reclaim UTXOs they contributed to the funding transaction. Previously, the contributed inputs and outputs were included in the SpliceFailed event. This commit splits them into a separate DiscardFunding event with a new FundingInfo::Contribution variant, providing a consistent interface for UTXO cleanup across all funding failure scenarios. Changes: - Add FundingInfo::Contribution variant to hold inputs/outputs for DiscardFunding events - Remove contributed_inputs/outputs fields from SpliceFailed event - Add QuiescentError enum for better error handling in funding_contributed - Emit DiscardFunding on all funding_contributed error paths - Filter duplicate inputs/outputs when contribution overlaps existing pending contribution - Return Err(APIError) from funding_contributed on all error cases - Add comprehensive test coverage for funding_contributed error paths Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

jkczyz · 2026-02-24T18:41:46Z

Rebased and added a fix to a silent merge conflict.

jkczyz · 2026-02-05T17:17:59Z

ChannelManager::funding_contributed consumes the passed in FundingContribution. Normally, the associated inputs and outputs should be given back via a SpliceFailed event. However, there are cases where the splice hasn't actually failed, so using that event would be misleading. The inputs and outputs still need to be returned to the user, which is done here via a dedicated DiscardFunding event.

Based on #4382.

ldk-reviews-bot · 2026-02-05T17:18:03Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

TheBlueMatt · 2026-02-19T23:21:20Z

Also CI is quite sad.

codecov · 2026-02-21T20:16:16Z

Codecov Report

❌ Patch coverage is 87.70764% with 37 lines in your changes missing coverage. Please review.
✅ Project coverage is 85.94%. Comparing base (df57978) to head (24062c0).
⚠️ Report is 15 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/channelmanager.rs	85.25%	23 Missing ⚠️
lightning/src/ln/channel.rs	90.62%	6 Missing ⚠️
lightning/src/ln/functional_test_utils.rs	86.48%	5 Missing ⚠️
lightning/src/ln/interactivetxs.rs	88.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4388 +/- ##
==========================================
+ Coverage 85.89% 85.94% +0.04% 
==========================================
Files 159 159 Lines 104402 104607 +205 Branches 104402 104607 +205 ==========================================
+ Hits 89681 89901 +220 + Misses 12216 12194 -22 - Partials 2505 2512 +7

Flag	Coverage Δ
tests	`85.94% <87.70%> (+0.04%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2026-02-23T16:52:26Z

Rebased to address a merge conflict.

TheBlueMatt · 2026-02-23T19:27:35Z

CI is quite sad.

wpaulino

LGTM, feel free to squash

jkczyz · 2026-02-23T21:45:28Z

Squashed

TheBlueMatt · 2026-02-24T13:58:44Z

+ return match unique_contributions {
+ None => Err(QuiescentError::DoNothing),
+ Some((contributed_inputs, contributed_outputs)) => {
+ Err(QuiescentError::FailSplice(SpliceFundingFailed {


Commented on discord, but if the intent of the split return logic is to ensure that downstream code doesnt' get a SpliceFailed event that may be confused for being in relation to an earlier splice (due to not having seen SplicePending), this isn't quite sufficient - if we're still constructing the transaction we can't generate a SpliceFailed because it hasn't yet gotten to SplicePending.

Hmm... SpliceFailed and SplicePending are mutually exclusive. Once we generated a SplicePending event, we won't generate SpliceFailed.
However, this should probably be return QuiescentError::DiscardFunding as per my earlier reasoning. That is, we have an active splice being negotiated, so we should error and discard the funding without emitting a SpliceFailed event.

RecipientOnionFields::secret_only requires an amount now, but when do_abandon_splice_quiescent_action_on_shutdown was introduced it was based on an earlier commit.

Add a helper function to assert that SpliceFailed events contain the expected channel_id and contributed inputs/outputs. This ensures that tests verify the contributions match what was originally provided. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

The following types have methods for returning contributed inputs and outputs: - FundingNegotiationContext - InteractiveTxConstructor - InteractiveTxSigningSession - ConstructedTransaction Having iterators for these can avoid allocations, which is useful for filtering contributed input and outputs when producing DiscardFunding events. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

When a splice fails, users need to reclaim UTXOs they contributed to the funding transaction. Previously, the contributed inputs and outputs were included in the SpliceFailed event. This commit splits them into a separate DiscardFunding event with a new FundingInfo::Contribution variant, providing a consistent interface for UTXO cleanup across all funding failure scenarios. Changes: - Add FundingInfo::Contribution variant to hold inputs/outputs for DiscardFunding events - Remove contributed_inputs/outputs fields from SpliceFailed event - Add QuiescentError enum for better error handling in funding_contributed - Emit DiscardFunding on all funding_contributed error paths - Filter duplicate inputs/outputs when contribution overlaps existing pending contribution - Return Err(APIError) from funding_contributed on all error cases - Add comprehensive test coverage for funding_contributed error paths Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

jkczyz · 2026-02-24T18:41:46Z

Rebased and added a fix to a silent merge conflict.

jkczyz · 2026-02-05T17:17:59Z

ChannelManager::funding_contributed consumes the passed in FundingContribution. Normally, the associated inputs and outputs should be given back via a SpliceFailed event. However, there are cases where the splice hasn't actually failed, so using that event would be misleading. The inputs and outputs still need to be returned to the user, which is done here via a dedicated DiscardFunding event.

Based on #4382.

ldk-reviews-bot · 2026-02-05T17:18:03Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

TheBlueMatt · 2026-02-19T23:21:20Z

Also CI is quite sad.

codecov · 2026-02-21T20:16:16Z

Codecov Report

❌ Patch coverage is 87.70764% with 37 lines in your changes missing coverage. Please review.
✅ Project coverage is 85.94%. Comparing base (df57978) to head (24062c0).
⚠️ Report is 15 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/channelmanager.rs	85.25%	23 Missing ⚠️
lightning/src/ln/channel.rs	90.62%	6 Missing ⚠️
lightning/src/ln/functional_test_utils.rs	86.48%	5 Missing ⚠️
lightning/src/ln/interactivetxs.rs	88.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4388 +/- ##
==========================================
+ Coverage 85.89% 85.94% +0.04% 
==========================================
Files 159 159 Lines 104402 104607 +205 Branches 104402 104607 +205 ==========================================
+ Hits 89681 89901 +220 + Misses 12216 12194 -22 - Partials 2505 2512 +7

Flag	Coverage Δ
tests	`85.94% <87.70%> (+0.04%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2026-02-23T16:52:26Z

Rebased to address a merge conflict.

TheBlueMatt · 2026-02-23T19:27:35Z

CI is quite sad.

wpaulino

LGTM, feel free to squash

jkczyz · 2026-02-23T21:45:28Z

Squashed

TheBlueMatt · 2026-02-24T13:58:44Z

+ return match unique_contributions {
+ None => Err(QuiescentError::DoNothing),
+ Some((contributed_inputs, contributed_outputs)) => {
+ Err(QuiescentError::FailSplice(SpliceFundingFailed {


Commented on discord, but if the intent of the split return logic is to ensure that downstream code doesnt' get a SpliceFailed event that may be confused for being in relation to an earlier splice (due to not having seen SplicePending), this isn't quite sufficient - if we're still constructing the transaction we can't generate a SpliceFailed because it hasn't yet gotten to SplicePending.

Hmm... SpliceFailed and SplicePending are mutually exclusive. Once we generated a SplicePending event, we won't generate SpliceFailed.
However, this should probably be return QuiescentError::DiscardFunding as per my earlier reasoning. That is, we have an active splice being negotiated, so we should error and discard the funding without emitting a SpliceFailed event.

RecipientOnionFields::secret_only requires an amount now, but when do_abandon_splice_quiescent_action_on_shutdown was introduced it was based on an earlier commit.

Add a helper function to assert that SpliceFailed events contain the expected channel_id and contributed inputs/outputs. This ensures that tests verify the contributions match what was originally provided. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

The following types have methods for returning contributed inputs and outputs: - FundingNegotiationContext - InteractiveTxConstructor - InteractiveTxSigningSession - ConstructedTransaction Having iterators for these can avoid allocations, which is useful for filtering contributed input and outputs when producing DiscardFunding events. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

When a splice fails, users need to reclaim UTXOs they contributed to the funding transaction. Previously, the contributed inputs and outputs were included in the SpliceFailed event. This commit splits them into a separate DiscardFunding event with a new FundingInfo::Contribution variant, providing a consistent interface for UTXO cleanup across all funding failure scenarios. Changes: - Add FundingInfo::Contribution variant to hold inputs/outputs for DiscardFunding events - Remove contributed_inputs/outputs fields from SpliceFailed event - Add QuiescentError enum for better error handling in funding_contributed - Emit DiscardFunding on all funding_contributed error paths - Filter duplicate inputs/outputs when contribution overlaps existing pending contribution - Return Err(APIError) from funding_contributed on all error cases - Add comprehensive test coverage for funding_contributed error paths Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

jkczyz · 2026-02-24T18:41:46Z

Rebased and added a fix to a silent merge conflict.

jkczyz · 2026-02-05T17:17:59Z

ChannelManager::funding_contributed consumes the passed in FundingContribution. Normally, the associated inputs and outputs should be given back via a SpliceFailed event. However, there are cases where the splice hasn't actually failed, so using that event would be misleading. The inputs and outputs still need to be returned to the user, which is done here via a dedicated DiscardFunding event.

Based on #4382.

ldk-reviews-bot · 2026-02-05T17:18:03Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

TheBlueMatt · 2026-02-19T23:21:20Z

Also CI is quite sad.

codecov · 2026-02-21T20:16:16Z

Codecov Report

❌ Patch coverage is 87.70764% with 37 lines in your changes missing coverage. Please review.
✅ Project coverage is 85.94%. Comparing base (df57978) to head (24062c0).
⚠️ Report is 15 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/channelmanager.rs	85.25%	23 Missing ⚠️
lightning/src/ln/channel.rs	90.62%	6 Missing ⚠️
lightning/src/ln/functional_test_utils.rs	86.48%	5 Missing ⚠️
lightning/src/ln/interactivetxs.rs	88.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4388 +/- ##
==========================================
+ Coverage 85.89% 85.94% +0.04% 
==========================================
Files 159 159 Lines 104402 104607 +205 Branches 104402 104607 +205 ==========================================
+ Hits 89681 89901 +220 + Misses 12216 12194 -22 - Partials 2505 2512 +7

Flag	Coverage Δ
tests	`85.94% <87.70%> (+0.04%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2026-02-23T16:52:26Z

Rebased to address a merge conflict.

TheBlueMatt · 2026-02-23T19:27:35Z

CI is quite sad.

wpaulino

LGTM, feel free to squash

jkczyz · 2026-02-23T21:45:28Z

Squashed

TheBlueMatt · 2026-02-24T13:58:44Z

+ return match unique_contributions {
+ None => Err(QuiescentError::DoNothing),
+ Some((contributed_inputs, contributed_outputs)) => {
+ Err(QuiescentError::FailSplice(SpliceFundingFailed {


Commented on discord, but if the intent of the split return logic is to ensure that downstream code doesnt' get a SpliceFailed event that may be confused for being in relation to an earlier splice (due to not having seen SplicePending), this isn't quite sufficient - if we're still constructing the transaction we can't generate a SpliceFailed because it hasn't yet gotten to SplicePending.

Hmm... SpliceFailed and SplicePending are mutually exclusive. Once we generated a SplicePending event, we won't generate SpliceFailed.
However, this should probably be return QuiescentError::DiscardFunding as per my earlier reasoning. That is, we have an active splice being negotiated, so we should error and discard the funding without emitting a SpliceFailed event.

RecipientOnionFields::secret_only requires an amount now, but when do_abandon_splice_quiescent_action_on_shutdown was introduced it was based on an earlier commit.

Add a helper function to assert that SpliceFailed events contain the expected channel_id and contributed inputs/outputs. This ensures that tests verify the contributions match what was originally provided. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

The following types have methods for returning contributed inputs and outputs: - FundingNegotiationContext - InteractiveTxConstructor - InteractiveTxSigningSession - ConstructedTransaction Having iterators for these can avoid allocations, which is useful for filtering contributed input and outputs when producing DiscardFunding events. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

When a splice fails, users need to reclaim UTXOs they contributed to the funding transaction. Previously, the contributed inputs and outputs were included in the SpliceFailed event. This commit splits them into a separate DiscardFunding event with a new FundingInfo::Contribution variant, providing a consistent interface for UTXO cleanup across all funding failure scenarios. Changes: - Add FundingInfo::Contribution variant to hold inputs/outputs for DiscardFunding events - Remove contributed_inputs/outputs fields from SpliceFailed event - Add QuiescentError enum for better error handling in funding_contributed - Emit DiscardFunding on all funding_contributed error paths - Filter duplicate inputs/outputs when contribution overlaps existing pending contribution - Return Err(APIError) from funding_contributed on all error cases - Add comprehensive test coverage for funding_contributed error paths Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

jkczyz · 2026-02-24T18:41:46Z

Rebased and added a fix to a silent merge conflict.

jkczyz · 2026-02-05T17:17:59Z

ChannelManager::funding_contributed consumes the passed in FundingContribution. Normally, the associated inputs and outputs should be given back via a SpliceFailed event. However, there are cases where the splice hasn't actually failed, so using that event would be misleading. The inputs and outputs still need to be returned to the user, which is done here via a dedicated DiscardFunding event.

Based on #4382.

ldk-reviews-bot · 2026-02-05T17:18:03Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

TheBlueMatt · 2026-02-19T23:21:20Z

Also CI is quite sad.

codecov · 2026-02-21T20:16:16Z

Codecov Report

❌ Patch coverage is 87.70764% with 37 lines in your changes missing coverage. Please review.
✅ Project coverage is 85.94%. Comparing base (df57978) to head (24062c0).
⚠️ Report is 15 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/channelmanager.rs	85.25%	23 Missing ⚠️
lightning/src/ln/channel.rs	90.62%	6 Missing ⚠️
lightning/src/ln/functional_test_utils.rs	86.48%	5 Missing ⚠️
lightning/src/ln/interactivetxs.rs	88.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4388 +/- ##
==========================================
+ Coverage 85.89% 85.94% +0.04% 
==========================================
Files 159 159 Lines 104402 104607 +205 Branches 104402 104607 +205 ==========================================
+ Hits 89681 89901 +220 + Misses 12216 12194 -22 - Partials 2505 2512 +7

Flag	Coverage Δ
tests	`85.94% <87.70%> (+0.04%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2026-02-23T16:52:26Z

Rebased to address a merge conflict.

TheBlueMatt · 2026-02-23T19:27:35Z

CI is quite sad.

wpaulino

LGTM, feel free to squash

jkczyz · 2026-02-23T21:45:28Z

Squashed

TheBlueMatt · 2026-02-24T13:58:44Z

+ return match unique_contributions {
+ None => Err(QuiescentError::DoNothing),
+ Some((contributed_inputs, contributed_outputs)) => {
+ Err(QuiescentError::FailSplice(SpliceFundingFailed {


Commented on discord, but if the intent of the split return logic is to ensure that downstream code doesnt' get a SpliceFailed event that may be confused for being in relation to an earlier splice (due to not having seen SplicePending), this isn't quite sufficient - if we're still constructing the transaction we can't generate a SpliceFailed because it hasn't yet gotten to SplicePending.

Hmm... SpliceFailed and SplicePending are mutually exclusive. Once we generated a SplicePending event, we won't generate SpliceFailed.
However, this should probably be return QuiescentError::DiscardFunding as per my earlier reasoning. That is, we have an active splice being negotiated, so we should error and discard the funding without emitting a SpliceFailed event.

RecipientOnionFields::secret_only requires an amount now, but when do_abandon_splice_quiescent_action_on_shutdown was introduced it was based on an earlier commit.

Add a helper function to assert that SpliceFailed events contain the expected channel_id and contributed inputs/outputs. This ensures that tests verify the contributions match what was originally provided. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

The following types have methods for returning contributed inputs and outputs: - FundingNegotiationContext - InteractiveTxConstructor - InteractiveTxSigningSession - ConstructedTransaction Having iterators for these can avoid allocations, which is useful for filtering contributed input and outputs when producing DiscardFunding events. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

When a splice fails, users need to reclaim UTXOs they contributed to the funding transaction. Previously, the contributed inputs and outputs were included in the SpliceFailed event. This commit splits them into a separate DiscardFunding event with a new FundingInfo::Contribution variant, providing a consistent interface for UTXO cleanup across all funding failure scenarios. Changes: - Add FundingInfo::Contribution variant to hold inputs/outputs for DiscardFunding events - Remove contributed_inputs/outputs fields from SpliceFailed event - Add QuiescentError enum for better error handling in funding_contributed - Emit DiscardFunding on all funding_contributed error paths - Filter duplicate inputs/outputs when contribution overlaps existing pending contribution - Return Err(APIError) from funding_contributed on all error cases - Add comprehensive test coverage for funding_contributed error paths Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

jkczyz · 2026-02-24T18:41:46Z

Rebased and added a fix to a silent merge conflict.

jkczyz added this to Weekly Goals Feb 5, 2026

jkczyz self-assigned this Feb 5, 2026

jkczyz changed the title ~~Split DiscardFunding from SpliceFudning event~~Split DiscardFunding from SpliceFunding eventFeb 5, 2026

jkczyz changed the title ~~Split DiscardFunding from SpliceFunding event~~Split DiscardFunding from SpliceFailed eventFeb 5, 2026

jkczyz force-pushed the 2026-02-splice-discard-funding branch 2 times, most recently from 217bf92 to 4ee3ef1Compare February 6, 2026 23:22

jkczyz mentioned this pull request Feb 10, 2026
Split splice initiation into two phases #4290
Merged

jkczyz force-pushed the 2026-02-splice-discard-funding branch 3 times, most recently from b84e3e5 to b1c5d02Compare February 12, 2026 16:27

jkczyz mentioned this pull request Feb 12, 2026
Contribute to splice as acceptor #4416
Merged

jkczyz force-pushed the 2026-02-splice-discard-funding branch 2 times, most recently from 580aa59 to e8efcbdCompare February 19, 2026 20:35

jkczyz marked this pull request as ready for review February 19, 2026 20:35

jkczyz requested review from TheBlueMatt and wpaulino February 19, 2026 20:35

TheBlueMatt reviewed Feb 19, 2026
View reviewed changes

Comment threadlightning/src/events/mod.rs
Comment threadlightning/src/ln/channel.rs

wpaulino reviewed Feb 19, 2026
View reviewed changes

Comment threadlightning/src/events/mod.rs Outdated
Comment threadlightning/src/ln/channel.rs
Comment threadlightning/src/ln/channel.rs
Comment threadlightning/src/ln/channelmanager.rs

jkczyz force-pushed the 2026-02-splice-discard-funding branch 2 times, most recently from e07bba3 to 90d87c2Compare February 21, 2026 18:27

jkczyz force-pushed the 2026-02-splice-discard-funding branch from 90d87c2 to 40a70d1Compare February 23, 2026 16:51

jkczyz requested review from TheBlueMatt and wpaulino February 23, 2026 16:52

wpaulino reviewed Feb 23, 2026
View reviewed changes

Comment threadlightning/src/events/mod.rs Outdated
Comment threadlightning/src/ln/functional_test_utils.rs Outdated
Comment threadlightning/src/ln/splicing_tests.rs Outdated

jkczyz force-pushed the 2026-02-splice-discard-funding branch from 40a70d1 to 56bd558Compare February 23, 2026 21:42

jkczyz force-pushed the 2026-02-splice-discard-funding branch from 56bd558 to df1f54aCompare February 23, 2026 21:45

jkczyz requested a review from wpaulino February 23, 2026 21:45

wpaulino previously approved these changes Feb 23, 2026
View reviewed changes

TheBlueMatt reviewed Feb 24, 2026
View reviewed changes

jkczyz dismissed wpaulino’s stale review via ab7f4adFebruary 24, 2026 16:55

jkczyz force-pushed the 2026-02-splice-discard-funding branch from df1f54a to ab7f4adCompare February 24, 2026 16:55

jkczyz requested review from TheBlueMatt and wpaulino February 24, 2026 16:55

jkczyzand others added 3 commits February 24, 2026 12:29

Fix silent merge conflict
7e413a7
RecipientOnionFields::secret_only requires an amount now, but when do_abandon_splice_quiescent_action_on_shutdown was introduced it was based on an earlier commit.

jkczyz mentioned this pull request Feb 24, 2026
Fix missing argument in splicing test #4441
Closed
2 tasks

jkczyzand others added 3 commits February 24, 2026 12:39

Print unexpected events upon assertion failure
c9c0159

Add pending changelog for SpliceFailed / DiscardFunding split
24062c0
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

jkczyz force-pushed the 2026-02-splice-discard-funding branch from ab7f4ad to 24062c0Compare February 24, 2026 18:41

TheBlueMatt approved these changes Feb 24, 2026
View reviewed changes

wpaulino approved these changes Feb 24, 2026
View reviewed changes

TheBlueMatt merged commit 98501d6 into lightningdevkit:mainFeb 24, 2026
18 of 19 checks passed

github-project-automationBot moved this to Done in Weekly Goals Feb 24, 2026

TheBlueMatt mentioned this pull request Feb 24, 2026
Failed build: Continuous Integration Checks #4443
Closed

jkczyz mentioned this pull request Feb 24, 2026
Bump LDK dependency for RecipientOnionFields and SpliceFailed changes lightningdevkit/ldk-node#806
Merged

jkczyz mentioned this pull request Mar 19, 2026
Splicing / Dual Funding #4284
Open
50 tasks

joostjager mentioned this pull request Mar 23, 2026
Fuzz failure backlog #4504
Closed