LND

This commit only adds a release note describing new test infrastructure. It documents that the project's integration-test framework gained the ability to use a bitcoind-backed miner backend, and that continuous integration now tests that combination. There is no code change, no bug fix, and no security-relevant behavior change in the software itself.

This commit is a testing-only change. It adds a command-line flag so integration tests can choose which Bitcoin mining software (btcd or bitcoind) is used behind the scenes. It also tweaks one test to calculate transaction size locally instead of asking the miner for it. There is no change to production LND code or user funds.

This commit only adds a new automated test configuration to the project's GitHub Actions CI workflow. It makes the test suite also run with a specific Bitcoin mining backend option (bitcoind). There are no code changes to the actual LND software, no bug fixes, and no security-related changes.

This commit is purely a testing-infrastructure change. It adds a new option to run LND's integration tests against a bitcoind miner backend in addition to the existing btcd miner backend. It does not change any production LND node code, wallet logic, or consensus rules, so it cannot directly affect real users or funds.

This change improves how a background database service shuts down cleanly. It creates a cancel signal during startup and checks it inside long database loops, so shutdown isn't delayed by cache-loading work. There is no attacker-controlled path or security vulnerability shown in the diff; it is a robustness/clean-shutdown improvement.

This commit is a simple internal code cleanup. It moves a nil-check for an optional in-memory graph cache from the caller into the function that fills the cache. Behavior is unchanged: if the cache is disabled, the function returns immediately. There is no user-visible or security-relevant change.

This is a minor test-only cleanup in the LND codebase. It adjusts how a graph cache is enabled in one test and removes a redundant assertion in a helper function. There is no change to production code, no security fix, and no vulnerability.

This commit adds a safety flag so that LND's in-memory channel graph cache is not read until it has been fully populated. It is a defensive correctness change that prevents callers from seeing incomplete or inconsistent graph data while the cache is still being built, especially once cache loading becomes asynchronous in the future. There is no direct evidence in the commit of an exploitable security vulnerability being fixed.

This commit only changes test code in LND's channel graph database tests. It replaces a hack where tests manually set an internal graphCache field to nil with a proper helper option, WithUseGraphCache(false). There is no change to production code, no security fix, and no vulnerability being addressed.

This commit only adds documentation to the release notes for LND version 0.21.0. It describes two new informational fields added to an RPC response about pending channel closures. There is no code change, no bug fix, and no security-relevant behavior change.

This is a trivial log message formatting fix. The original debug log line had a formatting placeholder but no value supplied, which would have printed an ugly '%!d(MISSING)' in logs. The patch adds the missing count so the log line displays correctly. It has no security relevance.

This commit only adds new integration tests for an existing feature. It does not change production code, fix a bug, or alter behavior. The tests verify that a counter showing how many blocks remain until a cooperative channel close is final behaves correctly, including when the blockchain reorganizes. There is no security issue in this change itself.

This commit adds two new informational fields to the PendingChannels RPC response in the LND Lightning node: one showing how many more blocks until a waiting close channel is considered final, and another showing the block height when the closing transaction was first confirmed. It is a user-facing informational enhancement, not a security fix or vulnerability.

This commit adds a new database field to record the block height at which a Lightning channel's closing transaction was confirmed. It is a straightforward data-model change with no security-relevant behavior visible in the diff. There is no fix for a vulnerability here—just new bookkeeping to support tracking how many confirmations a channel close has received and to handle blockchain reorganizations.

This commit improves how LND records the block height at which a channel-closing transaction is confirmed. It stores the confirmation height when a close is detected and clears it if a blockchain reorganization removes the transaction. This is a bookkeeping/UI improvement rather than a fix for a known security vulnerability.

This change tightens input checking for a special type of cryptographic value (a MuSig2 nonce) that peers send to each other in Lightning Network messages. Before, a peer could send bytes that looked like a nonce but were not valid points on the Bitcoin curve; those invalid values would only be caught later, deep inside the signing code. Now they are rejected immediately when the message is first decoded. This is a defensive hardening fix: it makes the protocol more robust against malformed or malicious peer input and prevents potential crashes or unexpected behavior in the signing flow.

This commit fixes two bugs in a frozen copy of LND's payment database code used during data migration. One bug would overwrite historical success/failure timestamps with the current time when updating old payment attempts. The other bug made payment list ordering non-deterministic. The commit message explicitly states these fixed functions are not actually called by the migration itself, so the fixes are for completeness rather than active vulnerability remediation.

This commit fixes a database query used to list Lightning Network payments. The old query could fail on Postgres with a type-mismatch error and was inefficient because it couldn't use a date index. The fix supplies default date bounds from the Go code, removes a problematic 'reverse' ordering trick, and adds a separate descending-order query. It is a correctness/performance bug fix rather than a security vulnerability.

This commit only edits the release notes document for version 0.21.0. It removes a duplicate mention of a previously released gossiper deadlock fix and removes one contributor's name from the credits list. No program code is changed, so this has no direct security effect on the software itself.

This commit only removes two entries from the draft release notes for version 0.21.0. The entries describe bug fixes that were already merged into the earlier 0.20.1 release, so they are being deleted from the wrong document. No code was changed, and no new security issue is introduced or fixed by this commit.

This commit only adds text to the release notes document for version 0.20.1. It describes several previously merged bug fixes (TLS file handling, database race conditions, SQL connection limits, feature-bit parsing, and a gossiper deadlock), but it does not change any source code, configuration, or executable behavior. By itself, it cannot introduce or fix a security issue.

This commit only adds a new integration test. It does not change any production code, so it cannot introduce a security vulnerability by itself. The test checks that a database migration for payment records works correctly and that new payments are written to the new database after the migration.

This commit only adds a clarifying code comment. No code behavior changes. It warns future developers that when promoting database migrations from test to production, they must also update which backend is used, otherwise production builds might keep using the older KV backend instead of the newer native SQL backend. There is no security vulnerability in this change itself.

This commit changes how LND stores payment data in normal production builds, switching from an older key-value database backend to a newer SQL-based backend. It removes build-specific code differences so that both normal and test-native-SQL builds use the same SQL payments store. There is no indication in the commit of a security bug, vulnerability fix, or exploit.

This commit adds database support for a new kind of channel-announcement proof used by upcoming taproot channels. It does not change any live network behavior; it only prepares the code so future taproot features can store their proofs safely. Existing V1 proofs keep working exactly as before, and the gossiper is explicitly restricted to V1 proofs only.