test: fix Heze spec coverage - #9763
Conversation
| }, | ||
| shouldError: (testCase) => !testCase.post, | ||
| timeout: 30000, | ||
| timeout: ForkSeq[fork] >= ForkSeq.gloas ? 120000 : 30000, |
There was a problem hiding this comment.
this really isn't ideal, we need to figure out why these spec tests take forever, for now I would prefer to keep them skipped
There was a problem hiding this comment.
Done in a6bf1cadac. I reverted the timeout bump and added explicit skips for the current gloas/heze sanity/slots/historical_accumulator cases instead.
Verified with:
pnpm vitest run --project spec-mainnet test/spec/presets/sanity.test.ts -t historical_accumulator(7 passed, 804 skipped)pnpm lint
There was a problem hiding this comment.
Done in a6bf1cadac: reverted the raised sanity timeout back to 30000 and restored skips for the slow Gloas/Heze historical_accumulator slot cases while we investigate the runtime separately.
There was a problem hiding this comment.
Final correction in eb17db058d: removed the timeout/skip changes entirely. The PR now only updates specrefs.
There was a problem hiding this comment.
please don't edit this file, can be done separately
There was a problem hiding this comment.
Removed in eb17db058d. I trimmed the PR back to specrefs only.
| const domain = config.getDomain(stateSlot, DOMAIN_BEACON_BUILDER); | ||
| const sszType = | ||
| config.getForkSeq(stateSlot) >= ForkSeq.heze ? ssz.heze.ExecutionPayloadBid : ssz.gloas.ExecutionPayloadBid; | ||
| const sszType = sszTypesFor(config.getForkName(stateSlot) as ForkPostGloas, "ExecutionPayloadBid"); |
There was a problem hiding this comment.
no this is not the right way either, please remove the change for now
There was a problem hiding this comment.
Removed in eb17db058d. I trimmed the PR back to specrefs only.
| (state as CachedBeaconStateHeze).latestExecutionPayloadBid = sszTypesFor( | ||
| ForkName.heze, | ||
| "ExecutionPayloadBid" | ||
| ).toViewDU(bid as heze.ExecutionPayloadBid); |
There was a problem hiding this comment.
Removed in eb17db058d. I trimmed the PR back to specrefs only.
| // TODO-GLOAS: investigate why these mainnet post-Gloas historical accumulator slot tests | ||
| // take ~40-50s locally. Skipping for now instead of raising the global sanity/slots timeout. | ||
| /^gloas\/sanity\/slots\/pyspec_tests\/historical_accumulator$/, | ||
| /^heze\/sanity\/slots\/pyspec_tests\/historical_accumulator$/, |
There was a problem hiding this comment.
spec run in the other pr passed, so please remove this
There was a problem hiding this comment.
Removed in eb17db058d. I trimmed the PR back to specrefs only.
| parentBeaconBlockRoot?: Uint8Array; | ||
| slotNumber?: number; // EIP-7843 | ||
| targetGasLimit?: bigint; // GLOAS (PayloadAttributesV4, execution-apis#796) | ||
| inclusionListTransactions?: gloas.Transaction[]; // HEZE (PayloadAttributesV5, EIP-7805) |
There was a problem hiding this comment.
this doesn't seem required for the fork boilerplate?
There was a problem hiding this comment.
Removed in eb17db058d. I trimmed the PR back to specrefs only.
|
Spec tests still failing :( |
| sources: [] | ||
| sources: | ||
| - file: packages/types/src/heze/sszTypes.ts | ||
| search: export const PayloadAttributes = |
There was a problem hiding this comment.
we didn't update payload attributes for heze yet
There was a problem hiding this comment.
Fixed in b35b3f4ba1: restored PayloadAttributes#heze to sources: [] and added it back to the ethspecify dataclass exceptions, since Heze payload attributes have not been updated yet. upgrade_to_heze#heze remains mapped.
I think this is unrelated but something is really slow.. |
|
Pushed I also checked the failing Local verification on the pushed head: |
3 participants
Summary
PayloadAttributesandupgrade_to_hezePayloadAttributesSSZ type andupgradeStateToHezeVerification
ethspecify check --path=specrefspnpm lintAI-assisted by Lodekeeper.