Repository navigation
Add L1SLOAD and L1STATICCALL on real-time - #175
Closed
jmadibekov wants to merge 30 commits into
Closed
jmadibekov wants to merge 30 commits into
jmadibekov wants to merge 30 commits into
Conversation
* feat: Move Node Builder types to alethia-reth-node * feat: custom treasury address added for evm factory * feat: rpc feature for TaikoNextBlockEnvAttributes, local-payload-builder for LocalPayloadAttributesBuilder<TaikoChainSpec> * feat: update reth dependencies for alethia-reth-rpc * update reth dependencies for alethia-reth-primitives * update reth dependencies for alethia-reth-payload * update reth dependencies for alethia-reth-node * update reth dependencies for alethia-reth-evm * update reth dependencies for alethia-reth-db * update reth dependencies for alethia-reth-consensus * update reth dependencies for alethia-reth-cli * update reth dependencies for alethia-reth-chainspec * update reth dependencies for alethia-reth-block * update reth dependencies * feat: merge from main --------- Co-authored-by: abiba <b>
feat: cancun eips shasta
Feat: prague eips shasta
Feat: osaka eips shasta
* feat: add real time proving support * Delete PROVING_MIGRATION.md * Delete PROTOCOL_MIGRATION_REAL_TIME_FORK.md * Delete ALETHIA_RETH_MIGRATION.md * style: run cargo +nightly fmt Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Ahmad Bitar <33181301+smartprogrammer93@users.noreply.github.com> Co-authored-by: smartprogrammer93 <smartprogrammer@windowslive.com> Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Adds the L1SLOAD precompile at address 0x10001 for reading L1 storage from L2. The precompile takes 84-byte input (address + storage key + block number) and returns the cached storage value. The precompile is a dummy cache: Raiko populates it with verified values before block re-execution during proving. The cache keys include block number to support lookback up to 256 blocks from the anchor block. Changes: - crates/evm/src/precompiles/l1sload.rs: Cache, API, precompile logic - crates/evm/src/precompiles/mod.rs: taiko_precompiles_map() helper - crates/evm/src/factory.rs: Use taiko_precompiles_map in EVM factory - 11 unit tests covering all paths
Promote trace! to info! and add [jmadibekov] tags for: - Precompile entry, input parsing, range checks - Cache hit/miss, RPC fallback, anchor/l1origin context
- Consolidate ~12 per-call info logs in l1sload_run() into 2-3 targeted logs - Cache hits downgraded to trace level; RPC fallback remains at info - Trim verbose doc comments that restate obvious code - Keep all functional logic and tests intact E2E verified: all L1SLOAD scenarios pass on devnet.
- Remove unused L1SLOAD const bundle (address inlined in mod.rs) - Inline clear_anchor_block_id/clear_l1_origin_block_id into clear_l1_storage - Make 4 constants private (not used outside crate) - Unify duplicated output construction in cache-hit/RPC-fallback branches - Downgrade RPC fallback log from info to debug - Remove redundant clear calls in tests
The lookback range is now measured from L1 origin instead of anchor, matching surge-raiko's simplified trust model. L1 origin is the single reference point for both upper and lower bounds.
…origin, add test, format
A1: set_l1_staticcall_value silently drops return data larger than
MAX_RETURN_DATA_SIZE with an error-level log; rationale documented
on the setter doc-comment. Test pins the boundary behavior.
A2: Drop the underscore prefix on `anchor_block_id` in both precompile
modules now that the variable is fully used (the prefix used to
signal "unused but kept for clarity" — no longer accurate).
A3: Document that `total_gas` deliberately does NOT clamp to gas_limit;
revm's caller-side OOG check is the correct surface. New regression
test `test_l1staticcall_l1_gas_alone_oog_when_exceeding_limit`
locks the contract in.
The Shasta proposal field 'originBlockNumber' was renamed to 'maxAnchorBlockNumber' in the RealTime fork. Both serve the same role — the on-chain-verified L1 block that anchors the precompile's [N − 256, N] lookback window. The static, accessor functions, and error messages all carried the Shasta-era 'l1_origin' name; rename them to be accurate under both forks. No functional change — the value passed by raiko is still the same: - Shasta: Proposal.originBlockNumber - RealTime: RealTimeProposal.maxAnchorBlockNumber
Author
|
Wrong target repo — recreating against NethermindEth/alethia-reth. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Real-time-rebased version of the L1SLOAD + L1STATICCALL precompiles. Stacked semantically on PR #12 (real-time proving support).
Relationship to existing PRs
This new PR rebases the same work onto
main(post PR #12) so the precompiles can run alongside real-time proving. The 18 commits from PR #11 + PR #13 cherry-picked cleanly with zero conflicts (the precompile code is hardfork-agnostic).What changed vs. PR #11 + PR #13
One additional commit: rename
CURRENT_L1_ORIGIN_BLOCK_ID→CURRENT_L1_MAX_ANCHOR_BLOCK_IDplus all callers, error messages, and test names. Same underlying value — Shasta'soriginBlockNumberand RealTime'smaxAnchorBlockNumberplay the same structural role (the on-chain-verified L1 block that anchors the precompile's[N − 256, N]lookback window). The rename makes the variable name accurate under both forks.Files
+2,243 / -156).crates/evm/src/precompiles/l1sload.rs(812 LoC)crates/evm/src/precompiles/l1staticcall.rs(1,286 LoC)crates/evm/src/precompiles/mod.rs(20 LoC) — wirestaiko_precompiles_map()into the EVM factoryfind_last_block_number_by_batch_id#12's RealTime hardfork additions.Tests
cargo build --workspaceclean.cargo clippy -p alethia-reth-evmzero errors.Cross references
devnet_shasta_timestampflag && updateanchorV4selector #50 + feat(chainspec): update devnet genesis JSON #63)notes/2026/03-17-l1staticcall/l1staticcall-47-realtime-rebase-plan-2026-05-07.md