Fix domain derivation on LC syncAggregate validation#5252
Conversation
Performance Report✔️ no performance regression detected Full benchmark results
|
| const signingRoot = ssz.phase0.SigningData.hashTreeRoot({ | ||
| objectRoot: ssz.phase0.BeaconBlockHeader.hashTreeRoot(update.attestedHeader.beacon), | ||
| domain: store.config.getDomain(update.signatureSlot, DOMAIN_SYNC_COMMITTEE), | ||
| domain: store.config.getDomain(update.signatureSlot - 1, DOMAIN_SYNC_COMMITTEE), |
There was a problem hiding this comment.
or should we use attestedHeader.beacon.slot?
There was a problem hiding this comment.
ahh just checked the linked PR , lgtm 👍
There was a problem hiding this comment.
sadly, attested_header.beacon.slot may be incorrect.
Example:
BELLATRIX_FORK_EPOCH = 0
CAPELLA_FORK_EPOCH = 255
DENEB_FORK_EPOCH = 256
attested_header.slot = 8159 (bellatrix)
signature_slot = 8192 (Deneb)
fork_version_slot = 8191 (Capella)
so, sync committee would be determined by sync committee at slot 8192, the fork version is determined by slot 8191, and the object is from slot 8159
There was a problem hiding this comment.
thanks @etan-status for the insightful example 🙏 ❤️
|
🎉 This PR is included in v1.7.0 🎉 |
Motivation
Description
Fix off by one error causing invalid signature errors on fork boundary