Gloas consensus: epoch processing, block signature verification, more tests#8808
Conversation
There was a problem hiding this comment.
We could delete a few more of these if all the tests except fork_choice pass.
There was a problem hiding this comment.
Deleted the following:
"tests/./gloas/finality/.",
"tests/./gloas/networking/.",
"tests/./gloas/rewards/.",
"tests/./gloas/sanity/.",
"tests/./gloas/transition/.",
|
Gloas test vectors do not yet exist for:
They are being disabled for now |
pawanjay176
left a comment
There was a problem hiding this comment.
LGTM. Neat and tidy.
The single_pass implementation looks equivalent to the spec implementation given that the effective_balance update is happening only after process_builder_pending_payments
| } | ||
|
|
||
| // NOTE: this could be a little more memory-efficient with some juggling to reuse parts | ||
| // of the persistent tree (could convert to list, use pop_front, convert back). |
There was a problem hiding this comment.
The builder_pending_payments vector is tiny, doesn't matter much
Merge Queue StatusRule:
This pull request spent 30 minutes 19 seconds in the queue, including 28 minutes 19 seconds running CI. Required conditions to merge
|
Proposed Changes
process_builder_pending_paymentsin epoch processing for Gloas. Enable the new EF tests for this sub-component as well.include_all_signatures_except_proposalfor Gloas to safely include the execution payload bid signature (this was an omission in the previous bid PR).Additional Info
finalitytests are failing. I haven't yet checked whether this is because we're missing some part of block/epoch processing, or whether it's because we have a bug.Note from Eitan: we were just missing a call to payload bid processing dbcda8e