Repository navigation
Conversation
…base#6610 to v1.5.6) (base#6611) * fix(runner): raise merkle rebuild threshold for 200ms blocks * fix(runner): keep merkle default when changesets are pruned sooner; drop download hook * fix(runner): restore merkle default if clean thresholds drop after a raise * docs: restamp base-node-runner README in docs index * fix(runner): write merkle rebuild threshold only when reth.toml is created * fix(cli): set merkle default only on reth-generated configs; warn on failure
…kport base#6647 to v1.5.6) (base#6648) * fix(consensus): keep the L1 watcher alive through L1 RPC outages (backport base#6647 to v1.5.6) The L1 watcher gave up on a head's SystemConfig log fetch after 10 attempts and returned RetriesExhausted, which shut down the whole rollup node. On base-sepolia-ha, a ~2 minute -32011 outage on the L1 proxy took down two sequencers this way, and they then crash-looped on startup while L1 was still unavailable. Retry log fetches until they succeed or the node shuts down (backoff 50 ms to 5 s), since restarting cannot help when the new process needs the same L1. While waiting, keep publishing safe and finalized L1 blocks and hold the next head. Abandon a head only after L1 reports a different block at its height (reorged out), so a canonical head's unsafe block signer update is never skipped. Cancellation now also interrupts an in-flight request. Co-authored-by: Toshi <toshi-noreply@coinbase.com> Adapted for v1.5.6, which has no L1 safe stream or safe/finalized watch channels: while logs are failing the watcher keeps sending finalized L1 blocks to derivation. Co-authored-by: Toshi <toshi-noreply@coinbase.com> * docs: restamp consensus service README in docs index Co-authored-by: Toshi <toshi-noreply@coinbase.com> * fix(consensus): bound the L1 watcher's reorg check The reorg check ran a bare get_block inside the select arm, so a slow L1 could hold up shutdown, the stream that keeps publishing during an outage, and the log fetch. Race it against cancellation and a 2 s timeout. A timed-out check keeps the head, as a failed one already does. Co-authored-by: Toshi <toshi-noreply@coinbase.com> --------- Co-authored-by: Toshi <toshi-noreply@coinbase.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Backport to releases/v1.5.6. Deposits are exempt from the per-transaction gas limit cap, but the handler applied the default transaction gas accounting to them. Deposits now use their declared gas limit; non-deposit transactions are unchanged. Adds regressions for deposit execution gas and for gas used when a deposit calls a stateful precompile.
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
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.
Merges upstream Base
v1.5.6intorelease/1.x. Tagv1.5.6-fh3.1on the merge commit once merged (v1.5.5-fh3.1with the upstream version replaced; the Firehose block model is unchanged, sofh3.1stays).Upstream recommends this release for Base mainnet and Base Sepolia, with no deadline and no hardfork. It changes deposit gas accounting in the EVM handler, keeps the consensus L1 watcher running through L1 RPC outages, and writes a merkle
rebuild_thresholdof 450,000 blocks when it creates a newreth.toml.v1.5.6keeps upstream's reth pin (base-v2.5.2.11) and alloy-evm 0.38 (revm 42), sostreamingfast/rethstays onbase-v2.5.2.11-fh3.1andstreamingfast/evmonv0.38.0-sf. No other StreamingFast repository needs a release first.firehose-tracerstays on 5.7.0.Besides the merge:
Cargo.lockconflicted: kept ours, and cargo added upstream's newreth-configdependency.Cargo.tomlgets a[patch]entry forreth-config, pinned to the samestreamingfast/rethcommit as the other reth crates (24849d77d,base-v2.5.2.11-fh3.1), so there is a singlereth-configin the lock.gas_usedfrom the execution result, so it follows with no code change. New testcrates/execution/firehose-tests/tests/deposit_gas.rstraces deposits at 16,000,000, 16,777,216, 16,777,217 and 20,000,000 gas and checks the root call gas limit, gas consumed, return data (theGASleft),gas_usedand the receipt. It fails on thev1.5.5handler (root call capped at 16,756,216) and passes onv1.5.6.ACTIVATIONS.md: statuses refreshed againstv1.5.6; the Azul row notes the deposit exemption. No new upgrade and no schedule change.Dependency table check: the fork pins
streamingfast/reth(release/base-2.x, still at the tagged commit) andstreamingfast/evm(release/0.x), as the release tracker lists.streamingfast/evmrelease/0.xhas moved on tov0.39.0-sf(revm 43), which this release does not use, so the pin stays onv0.38.0-sf. No public StreamingFast repository pins this fork: a code search forv1.5.5-fh3.1finds nothing, and forv1.5.4-fh3.1only our deployment configuration. Battlefield has abase-devnettarget, which the chain-update skill's notes still say is missing.CHANGELOG
Test Execution Summary
Branch
bump/base-1.5.6:release/1.xatv1.5.5-fh3.1+git merge v1.5.6, built againststreamingfast/rethbase-v2.5.2.11-fh3.1(24849d77d) andstreamingfast/evmv0.38.0-sf(0ac393245), both fetched from GitHub.Build Status
cargo build --release --package base-reth-node --bin base-reth-node --package base-consensus --bin base-consensus(same command asDockerfile.sf): PASSED (4 min)cargo check --workspace --all-targets: PASSED, afterforge soldeer install && forge buildincrates/utilities/test-utils/contractscargo clippy -p base-reth-node -p base-execution-firehose -p base-firehose-tests -p base-common-evm -p base-node-runner --tests: PASSED. The warnings are incrates/firehose-flashblocks/src/processor.rs,crates/execution/firehose,crates/execution/flashblocks/src/traits.rsandcrates/execution/runner/tests/firehose_live_tracing.rs, which this merge does not change. None in the new test.Unit Tests
cargo test -p base-firehose-testsdeposit_gastest, 12 EIP-8130 tests, 2 prestate replayscargo test -p base-firehose-flashblockscargo test -p base-reth-nodecargo test -p base-execution-firehosecargo test -p base-common-evmcargo test -p base-execution-chainspeccargo test -p base-common-genesiscargo test -p base-common-chainscargo test -p base-execution-payload-buildercargo test -p base-reth-clicargo test -p base-builder-multiplexin_process_smokecargo test -p base-execution-clisnapshot downloadchangecargo test -p base-consensus-nodecargo test -p base-node-runnermerkletests. TheBaseNode::provider_factory_builderdoctest (crates/execution/runner/src/node.rs) does not compile: it setsReadOnlyConfig { watch_static_files }, a field our reth fork does not have (it iswatch). Same failure as on thev1.5.4andv1.5.5bumps; the doc example is unchanged upstreamcargo test -p reth-firehose -p reth-firehose-tests(instreamingfast/rethat24849d77d,base-v2.5.2.11-fh3.1)Integration Tests
RUSTFLAGS="-C debug-assertions=off" RUST_MIN_STACK=33554432 cargo test -p base-system-tests --test firehose_denim --test firehose_b20 --test firehose_cobalt_b20, Docker-backed:firehose_b20::b20_transfer_is_tracedfirehose_cobalt_b20::cobalt_b20_surfaces_are_tracedfirehose_denim::denim_activation_block_traces_base_time_installBattlefield Tests
streamingfast/battlefield-ethereumatadb1921(master),base-devnettarget, with this branch's releasebase-reth-nodeandbase-consensusas the Firehose node. Upstream has not publishedghcr.io/base/node:v1.5.6yet, so the sequencer runs thebasetarget of upstream'sv1.5.6tag built locally (docker buildx bake -f etc/docker/docker-bake.hcl base) and served from a local registry, because the launcher always pulls the image:Header corresponds to RPCandCompare blocks: Firehose blocks match RPC blockspassed. No snapshot changed, and the node log has noERRORand no panic (about 900 blocks). The reader node reportedbase-1.5.6.crates/execution/firehose-tests/tests/deposit_gas.rscovers it. Zenith is genesis-only and no devnet schedules it, so the Keystore account changes are covered only by the unit tests inbase-firehose-tests.Firehose bug patterns
BaseHandler::tx_gas. The L1 watcher (consensus), the merkle stage threshold (pipeline trie stage, which executes nothing) andsnapshot downloaddo not run the EVM on the traced executor or reorder transactions. None found.rebuild_thresholdonly decides whether the pipeline's merkle stage rebuilds the trie or applies the range's changes; the state root check that follows execution is unchanged. None found.v1.5.6changes nothing incrates/common/chains,crates/common/genesisorcrates/utilities/upgrade-signal, and defines no newBaseUpgrade. The deposit gas change is not fork-gated, but it only changes anything where EIP-7825 applies, which is from Azul on (active on mainnet and Sepolia): deposits with a gas limit above 16,777,216 now run with their full limit. Everest is still not scheduled: no timestamp in the code, andgetSchedule()on the configured L1protocol_versions_addressstill reverts on both L1s (checked 2026-11-18). Zenith is genesis-only.ACTIVATIONS.mdrefreshed.