Core Lightning fixes revoked-channel close flaw in v26.06.7

  • Core Lightning fixed a channel-close flaw that could bypass penalties for revoked states.
  • v26.06.8 adds security fixes and is the project's strongly recommended upgrade.
  • Docker images from Aug. 28 to Sept. 1 could falsely report having v26.06.7 fixes.

Core Lightning, a Lightning Network node implementation, fixed a channel-close flaw in v26.06.7 that could allow a peer to broadcast an old, revoked channel state without triggering the penalty for cheating. Lightning Network channels use successive commitments to track balances, with a penalty mechanism intended to deter either party from publishing an outdated state. Before the repair, a peer could exploit a specific setup in which no upfront shutdown script was defined when the channel opened. By later supplying the output script from a revoked commitment in a shutdown message, the peer could make the funding spend resemble a cooperative close (a mutually agreed channel closure), according to the maintainers' patch notes and regression test. The fix checks locktime and sequence encoding before evaluating outputs as a possible mutual close. The issue described a potential penalty-evasion route, not a confirmed theft, and affected Core Lightning's channel handling rather than Bitcoin's base-chain rules. Operators running versions older than v26.06.7 need to update, with the project strongly recommending v26.06.8, which includes additional security fixes. Users who deployed Docker images during the earlier rollout should verify their image digest because images carrying v26.06.7 and related tags from Aug. 28 to Sept. 1 could report the new version at startup while lacking its fixes. Core Lightning shipped v26.06.7 on Aug. 28, published its initially embargoed release source code on Sept. 11, merged pull request 9509 into the main development branch on Sept. 15, and released v26.06.8 on Sept. 22. Bitcoin Optech explained the already-shipped repair on Sept. 25.

The information on this website is generated using AI and we cannot guarantee its accuracy. Please use it as reference information only.