* port changes from #1013
* port changes from #1068
* go.mod tidy
* fix compile error
* fix goimports
* fix log
* address review comments
* upgrade golang.org/x/net to 0.23.0
* port changes from #1018
* fix tests and linter errors
* address review comments
* refactor rollup sync service / verifier to use CalldataBlobSource to retrieve data from L1
* add configuration and initialize blob clients
* fix unit tests
* remove unused code
* address review comments
* address more review comments
* implement first version of new da-codec and to handle multiple batches submitted in one transaction
* add CommitBatchDAV7 and handle multiple commit events submitted in a single transactions
* fix bug due to previous batch being empty when processing the first batch within a set of batches
* Allow using MPT
* update to latest da-codec
* add field to CommittedBatchMeta to store LastL1MessageQueueHash for CodecV7 batches
* adjust rollup verifier to support CodecV7 batches
* address review comments
* fix issues after merge
* go mod tidy
* fix unit tests
* update da-codec
* add test TestValidateBatchCodecV7
* go mod tidy
* do not log error on shutdown
* add sanity check for version to deserialization of committedBatchMetaV7
* port changes from #1073
* chore: auto version bump [bot]
* address review comments
* add more logs
* disable ENRUpdater if DA sync mode is enabled
* exit pipeline if context is cancelled
* correctly handle override by setting the head of the chain to the parent's height so that created blocks will always become part of canonical chain
* fix error with genesis event being nil
* chore: auto version bump [bot]
* chore: auto version bump [bot]
* adjust to renaming in CodecV7
* implement carrying forward of L1 MessageQueue index
* fix issue after upgrading from old storage to new format where batchIndex was 0 and all batches would be skipped
* add new RevertBatch event
* add commitBatches to be able to read calldata of CodecV7/EuclidV2 committed batches
* implement finding of L1 message queue height for initial batch in recovery mode
* add sanity checks for computed batches from events and batch hashes given via calldata from commit transaction
* update ScrollChain ABI
* chore: auto version bump [bot]
* remove initial batch form DAQueue
* go mod tidy
* fix underflow bug when l1DeploymentBlock==0
* fix bug with wrong parentBatchHash of first batch of batches submitted in a single tx
* update to latest da-codec
* address review comments
* address review comments
* fix bug when l1MessageV2StartIndex==0 serialized to [] (empty slice) which would always be decoded to non-existing instead of 0
* chore: auto version bump [bot]
* cache go dependencies in Dockerfile.mockccc
* chore: auto version bump [bot]
* add INFO log when reverting rollup events in L1 follower mode
* change LastProcessedMessageQueueIndex of finalize event to totalL1MessagesPoppedOverall
* cleanup
* chore: auto version bump [bot]
---------
Co-authored-by: Ömer Faruk Irmak <omerfirmak@gmail.com>
Co-authored-by: Thegaram <Thegaram@users.noreply.github.com>
Co-authored-by: jonastheis <jonastheis@users.noreply.github.com>
* feat: add system config consensus to deprecate poa
* fix: check extra field lens
* Finish implementation with one signer only
* Implement unit tests
* Only check signature if block not requested
* Remove Extra from rlp encoding (and hashing)
* Implement UpgradableEngine as middleware for Clique/SystemContract
* Use isEuclid, add json tag to Requested, make go idiomatic
* Changes post integration test
* Remove comment
* Address comments
* Merge
* New field BlockSignature (not used in hashing/JSON)
* Enforce .Extra to be an empty slice
* replace header.Requested for header.IsNewBlock
* mark new block as IsNewBlock
* Make new RLP optional fields always default for legacy/reth compatibility
* Placing new fields as last non-zero rlp:optional values used by Scroll
* Penalize nodes that send non-zero Euclid V2 header field values
* Bring back .Requested to downloader instead of .IsNewBlock
* Replace IsNewBlock for Requested
* Address comments
* merge
* prevent timing issues in tests
* fix ci
* chore: auto version bump [bot]
* Update consensus/system_contract/consensus.go
Co-authored-by: Morty <70688412+yiweichi@users.noreply.github.com>
* chore: auto version bump [bot]
* Update consensus/system_contract/consensus.go
Co-authored-by: Jonas Theis <4181434+jonastheis@users.noreply.github.com>
* Remove whitespaces and merge version number
* chore: auto version bump [bot]
* validate that the read address from L1 is not empty and improved error handling when fetching address
* Fix indentation issue
* Fix test
* goimports fix
* goimports
---------
Co-authored-by: Alejandro Ranchal-Pedrosa <a.ranchalpedrosa@gmail.com>
Co-authored-by: jonastheis <4181434+jonastheis@users.noreply.github.com>
Co-authored-by: jonastheis <jonastheis@users.noreply.github.com>
Co-authored-by: ranchalp <ranchalp@users.noreply.github.com>
Co-authored-by: Péter Garamvölgyi <peter@scroll.io>
* port changes from #1013
* port changes from #1068
* go.mod tidy
* fix compile error
* fix goimports
* fix log
* address review comments
* upgrade golang.org/x/net to 0.23.0
* port changes from #1018
* fix tests and linter errors
* address review comments
* refactor rollup sync service / verifier to use CalldataBlobSource to retrieve data from L1
* add configuration and initialize blob clients
* fix unit tests
* remove unused code
* address review comments
* address more review comments
* implement first version of new da-codec and to handle multiple batches submitted in one transaction
* add CommitBatchDAV7 and handle multiple commit events submitted in a single transactions
* fix bug due to previous batch being empty when processing the first batch within a set of batches
* Allow using MPT
* implement reading of QueueTransaction from L1MessageQueueV1 and V2
* implement access to V1 and V2 messages and replace usage so that V1 is used before EuclidV2 fork and V2 afterward
* add tests
* update to latest da-codec
* add field to CommittedBatchMeta to store LastL1MessageQueueHash for CodecV7 batches
* adjust rollup verifier to support CodecV7 batches
* address review comments
* consume all L1 messages before EuclidV2 fork
* address review comments
* address review comments
* fix issues after merge
* go mod tidy
* fix unit tests
* update da-codec
* add test TestValidateBatchCodecV7
* go mod tidy
* do not log error on shutdown
* add flag to explicitly disable L1MessageQueueV2
* add TestEuclidV2HardForkMessageQueue to scroll worker
* add sanity check for version to deserialization of committedBatchMetaV7
* add mechanism for nodes to retrieve V2 messages even if they upgrade after L1MessageQueueV2 is deployed
* chore: auto version bump [bot]
* address review comments
* Update core/rawdb/accessors_l1_message.go
Co-authored-by: colin <102356659+colinlyguo@users.noreply.github.com>
* address review comments
---------
Co-authored-by: Ömer Faruk Irmak <omerfirmak@gmail.com>
Co-authored-by: Thegaram <Thegaram@users.noreply.github.com>
Co-authored-by: colin <102356659+colinlyguo@users.noreply.github.com>
Co-authored-by: Péter Garamvölgyi <peter@scroll.io>
* port changes from #1013
* port changes from #1068
* go.mod tidy
* fix compile error
* fix goimports
* fix log
* address review comments
* upgrade golang.org/x/net to 0.23.0
* port changes from #1018
* fix tests and linter errors
* address review comments
* refactor rollup sync service / verifier to use CalldataBlobSource to retrieve data from L1
* add configuration and initialize blob clients
* fix unit tests
* remove unused code
* address review comments
* address more review comments
* implement first version of new da-codec and to handle multiple batches submitted in one transaction
* add CommitBatchDAV7 and handle multiple commit events submitted in a single transactions
* fix bug due to previous batch being empty when processing the first batch within a set of batches
* Allow using MPT
* update to latest da-codec
* add field to CommittedBatchMeta to store LastL1MessageQueueHash for CodecV7 batches
* adjust rollup verifier to support CodecV7 batches
* address review comments
* fix issues after merge
* go mod tidy
* fix unit tests
* update da-codec
* add test TestValidateBatchCodecV7
* go mod tidy
* do not log error on shutdown
* add sanity check for version to deserialization of committedBatchMetaV7
* chore: auto version bump [bot]
* address review comments
* chore: auto version bump [bot]
---------
Co-authored-by: Ömer Faruk Irmak <omerfirmak@gmail.com>
Co-authored-by: Thegaram <Thegaram@users.noreply.github.com>
Co-authored-by: Péter Garamvölgyi <peter@scroll.io>
* fix(l2geth-verifier): only check the last batch of each bundle
* refactor function comments
* add more logs to debug
* fix the bug
* Revert "add more logs to debug"
This reverts commit 4befe9fe739275988b4bf34999bd1aad4d7a1d0e.
* tweak
* TestValidateBatchInFinalizeByBundle
* update da-codec dependency
* move rawdb.WriteFinalizedBatchMeta outside the loop
* use batch writer
* change log.Crit to log.Error
* feat(zktrie): bump to v0.8.4
fixes a deadlock in block tracing
* chore: auto version bump [bot]
---------
Co-authored-by: omerfirmak <omerfirmak@users.noreply.github.com>
* crypto/kz4844: pass blobs by ref (#29050)
This change makes use of the following underlying changes to the kzg-libraries in order to avoid passing large things on the stack:
- c-kzg: https://github.com/ethereum/c-kzg-4844/pull/393 and
- go-kzg: https://github.com/crate-crypto/go-kzg-4844/pull/63
* update version patch
* run go mod tidy
* use blob reference
---------
Co-authored-by: Martin HS <martin@swende.se>
Co-authored-by: Mason Liang <mason@scroll.io>
* draft change
* fix CI
* remove CancunBlock params and change cancunSigner to londonSignerWithEIP4844
* support blob transactions in receipt
* trigger ci
* bump version
* Revert "Substitute secp256k1 library for the L1 version (#22)"
This reverts commit 59df2d7ebb.
* goimports -local github.com/scroll-tech/go-ethereum -w .
* go mod tidy
* update zktrie, inducde proof with deletion
* Add deletion proof for stateDb
* fill deletionproofs
* lint
* proving more comment for parts prone to be confused
* simplify array appending
Co-authored-by: maskpp <maskpp266@gmail.com>
* bump zktrie version and fix issues
* add delproof into witgen
---------
Co-authored-by: maskpp <maskpp266@gmail.com>
* update go version
* update go version
* format code by goimports
* format code by goimports
* go mod tidy
* update go version
* update go version
* upgrade linter
* add exclude rule
* fix golangci exclude
---------
Co-authored-by: vincent <419436363@qq.com>
Co-authored-by: HAOYUatHZ <37070449+HAOYUatHZ@users.noreply.github.com>
* add KeccakCodeHash and CodeSize to StateAccount
* update StateAccount marshalling logic
* change emptyCodeHash to poseidon(nil)
* purge StateAccount.hash
* fix/disable failing tests
* change keccak and poseidon hash order in StateAccount
* fix lint
* update l2trace account wrapper
* update eth_getProof response type
* fix eth_getProof response type
* goimports
* update the codehash computation
* update the codehash test cases
* go mod tidy
* fix tests
* use keccak instead of poseidon
* update trace codehash field name
* upgrade zktrie to 4.2
* trigger ci
* update state account marshalling according to spec
* improve generatorStats estimation
* add comment
* upgrade zktrie to 4.3
* go mod tidy
* misc fixes
* fix TestDump
* fix snap sync tests
* handle err in the codehash
* fix tests in snapshot/generate_test.go
* remove prevhash from state journal
* add state_account_marshalling_test.go
* goimports
* add more tests
---------
Co-authored-by: Ho Vei <noelwei@gmail.com>
Co-authored-by: Haichen Shen <shenhaichen@gmail.com>
* feat: change bytecode hash function from Keccak to Poseidon
* fix TestStateProcessorErrors
* rename files
* copy iden3's poseidon package to local
* update poseidon hash
* apply new hash in codehash
* extend hash method, use actual code size as cap
* prune keccak(nil) for empty code
* update empty code hash references to Poseidon
* add TestPoseidonCodeHash
* use keccak for empty code hash
* go mod tidy
* goimports
* use Poseidon code hash in state and sync tests
* run goimports
* run goimports
* remove unnecessary reallocation from CodeHash
* goimports
* move codehash API to separate module
* add missing package
* fix review comments
* improve comments
Co-authored-by: Ho Vei <noelwei@gmail.com>
* Fix bug about genesis file struct
* Add test case if custom genesis file
* Add test case if custom genesis file
* Fix a mistake
* Update Makefile
Co-authored-by: Haichen Shen <shenhaichen@gmail.com>
Co-authored-by: Haichen Shen <shenhaichen@gmail.com>
* Substitute secp256k1 library for the L1 version
* Add explanatory comment for secp256k1 library change
* Update secp_fuzzer.go
* Change to go-ethereum version 1.10.13
Co-authored-by: HAOYUatHZ <37070449+HAOYUatHZ@users.noreply.github.com>
* internal: support optional filter expression for debug.stacks
* internal/debug: fix string regexp
* internal/debug: support searching for line numbers too
This PR adds flag to enable InfluxDB v2 (--metrics.influxdbv2), flags for v2-specific features (--metrics.influxdb.token, --metrics.influxdb.bucket), also carries over addition of support for specifying organization (--metrics.influxdb.organization), but still retains backwards compatibility with InfluxDB v1.
This removes auto-configuration of the snap.*.ethdisco.net DNS discovery tree.
Since measurements have shown that > 75% of nodes in all.*.ethdisco.net support
snap, we have decided to retire the dedicated index for snap and just use the eth
tree instead.
The dial iterators of eth and snap now use the same DNS tree in the default configuration,
so both iterators should use the same DNS discovery client instance. This ensures that
the record cache and rate limit are shared. Records will not be requested multiple times.
While testing the change, I noticed that duplicate DNS requests do happen even
when the client instance is shared. This is because the two iterators request the tree
root, link tree root, and first levels of the tree in lockstep. To avoid this problem, the
change also adds a singleflight.Group instance in the client. When one iterator
attempts to resolve an entry which is already being resolved, the singleflight object
waits for the existing resolve call to finish and returns the entry to both places.
