Implements ethereum/go-ethereum PR #31202 and #31618.
When local tracking is enabled:
- EthAPIBackend.SendTx tracks transactions after pool submission and keeps tracking temporary rejects so they can be retried by the local tracker.
- TxPool.AddLocal tracks accepted submissions and temporary rejects for local re-journal/re-submit flows, while preserving the original txpool error return to the caller.
This avoids persisting permanently invalid transactions while preserving retry signals for transient failures without masking submission outcomes in caller workflows.
Also included:
- classify temporary rejection reasons in core/txpool/locals
- expose SubPool.ValidateTxBasics and align LegacyPool implementation
- split low-tip rejection into ErrTxGasPriceTooLow
- simplify local tracker integration in txpool
- update txpool and eth tests for accepted vs retryable local tracking behavior
Refs: ethereum/go-ethereum#31202
Refs: ethereum/go-ethereum#31618
* fix(core/txpool): coordinate reset lifecycle and shutdown signaling #28837
Improve txpool loop synchronization around background resets.
This change:
- adds an explicit termination channel to signal pool shutdown
- tracks forced-reset intent and a waiter channel inside the reset loop
- ensures reset waiters are notified on completion or on pool termination
- allows an explicit sync request path to trigger an additional reset round when needed
Scope is limited to internal txpool concurrency control in core/txpool/txpool.go, with no protocol or RPC behavior change.
* fix(core/txpool): avoid blocking reset completion on close#31030
A runtime panic was triggered in promoteExecutables/demoteUnexecutables when
account balance was converted with uint256.MustFromBig(...):
panic: overflow
... legacypool.go:1637
Root cause:
- pool.currentState.GetBalance(addr) can exceed uint256 range in this code path.
- uint256.MustFromBig(balance) panics on overflow, crashing the reorg loop.
What this commit changes:
- remove uint256.MustFromBig(balance) from executable/non-executable filtering paths
- change list.Filter costLimit from *uint256.Int to *big.Int, and compare costs using big.Int directly
- keep overflow-safe totalcost accounting for replacements (subtract old cost first, then add new)
- return txpool.ErrSpecialTxCostOverflow for special-tx cost/totalcost overflow instead of returning (false, nil)
- avoid partial pending-state mutation by attaching a new pending list only after overflow-safe totalcost calculation succeeds
Tests:
- add regression coverage for special-tx overflow rejection returning non-nil error
- verify no pending/lookup/nonce mutation on overflow rejection
- cover replacement paths to ensure no intermediate-overflow regressions in list.Add/promoteSpecialTx
The address recover is executed and cached in ValidateTransaction already. It's
expected that the cached one is returned in ValidateTransaction. However,
currently, we use the wrong function signer.Sender instead of types.Sender which
will do all the address recover again.
Co-authored-by: minh-bq <97180373+minh-bq@users.noreply.github.com>
* core/txpool: no need to run rotate if no local txs
* Revert "core/txpool: no need to run rotate if no local txs"
This reverts commit 17fab17388.
* use Debug if todo is empty
---------
Signed-off-by: jsvisa <delweng@gmail.com>
Co-authored-by: Delweng <delweng@gmail.com>
Adjust txpool validation error wording to report actual and minimum values consistently.
- intrinsic gas error now reports: gas <actual>, minimum needed <required>
- underpriced tip error now reports: gas tip cap <actual>, minimum needed <required>
No transaction validation logic is changed in this commit; only error message text is updated.
Improve txpool loop synchronization around background resets.
This change:
- adds an explicit termination channel to signal pool shutdown
- tracks forced-reset intent and a waiter channel inside the reset loop
- ensures reset waiters are notified on completion or on pool termination
- allows an explicit sync request path to trigger an additional reset round when needed
Scope is limited to internal txpool concurrency control in core/txpool/txpool.go, with no protocol or RPC behavior change.
Introduce txpool.PendingFilter and migrate Pending(...) signatures from positional params to a typed struct.
- Add core/txpool.PendingFilter with MinTip/BaseFee fields for cheap call-site filtering
- Update txpool subpool interface and all call sites in eth/miner to pass the filter struct
- Keep behavior unchanged for empty filter (equivalent to previous nil,nil usage)
- Refresh legacypool tests to use the new API shape
This is a refactor-only change intended to improve API clarity and future extensibility without changing consensus or RPC semantics.
Switch LazyTransaction gas caps from *big.Int to *uint256.Int and convert once at pending retrieval time.
In miner ordering, keep fee comparisons on uint256 and precompute TRC21 gas price in uint256 form to avoid repeated big-int conversions in heap comparisons.
Also update affected tests and call sites in legacypool/worker/helper paths.
Compatibility: LazyTransaction exported field types changed (GasFeeCap/GasTipCap).
Introduce dynamic-fee pre-filtering in txpool pending retrieval to reduce
allocations and downstream processing work during mining and tx propagation.
What changed:
- Change the `Pending` API from `Pending(enforceTips bool)` to
`Pending(minTip *uint256.Int, baseFee *uint256.Int)` in txpool interfaces.
- Implement pre-filtering in `legacypool.Pending` by comparing effective tip
against the provided `minTip/baseFee` for non-local, non-special txs.
- Update call sites to the new API:
- miner work assembly (`miner/worker.go`)
- tx sync (`eth/sync.go`)
- pool transaction retrieval (`eth/api_backend.go`)
- protocol/test interfaces (`eth/protocol.go`, `eth/helper_test.go`).
Tests:
- Add targeted coverage for pending filtering semantics in
`core/txpool/legacypool/legacypool_test.go`, including:
- minTip threshold boundary behavior
- baseFee-aware effective tip filtering
- local/special transaction exemption behavior
- dynamic-fee boundary behavior when baseFee is nil.
Impact:
- Preserves existing behavior while making pending selection cheaper for
downstream consumers.
- Improves confidence in edge-case behavior through dedicated tests.
Miner configuration is unified under [Eth.Miner] (GasCeil/GasPrice/Etherbase/ExtraData), replacing legacy top-level [Eth] miner keys.
Operational impact: existing config files using [Eth].GasPrice/[Eth].Etherbase/[Eth].ExtraData must be migrated before upgrade.
Behavior update: gasprice=0 remains valid; only negative gas prices are sanitized at startup.
Default change: XDCGenesisGasLimit is reduced to 42,000,000 and now feeds miner default GasCeil (including default --miner-gaslimit), so nodes relying on defaults should review capacity expectations.
* refactor(txpool): remove wrapper type #27841
Partial backport of ethereum/go-ethereum PR #27841, limited to txpool wrapper removal.
- Migrate txpool interfaces/call sites from `*txpool.Transaction` to `*types.Transaction`
- Update eth/miner/contracts paths and related tests accordingly
- No intended behavior change
Blob sidecar validation/handling changes from upstream are not included here.
* refactor(core/txpool): migrate tx subscription to SubscribeTransactions #28243
Replace the old SubscribeNewTxsEvent-style plumbing with the new
SubscribeTransactions(ch, reorgs) interface across txpool, eth protocol
manager, API backend, miner worker, and test helpers.
Key changes:
- Extend txpool/subpool tx subscription interface with a reorgs flag
- Route eth tx announcement path to reorgs=false (new tx announcements only)
- Route API/miner subscriptions to reorgs=true
- Move subscription-scope cleanup to TxPool.Close()
- Add Gas field to LazyTransaction in legacy pending view
Note:
LegacyPool currently cannot strictly separate newly seen and resurrected txs,
so the reorgs flag is accepted for API compatibility and future blob-subpool
integration.
This is a follow-up to #29520, and a preparatory PR to a more thorough
change in the journalling system.
This PR hides the journal-implementation details away, so that the
statedb invokes methods like `JournalCreate`, instead of explicitly
appending journal-events in a list. This means that it's up to the
journal whether to implement it as a sequence of events or
aggregate/merge events.
This PR also makes it so that management of valid snapshots is moved
inside the journal, exposed via the methods `Snapshot() int` and
`RevertToSnapshot(revid int, s *StateDB)`.
JournalSetCode journals the setting of code: it is implicit that the
previous values were "no code" and emptyCodeHash. Therefore, we can
simplify the setCode journal.
The self-destruct journalling is a bit strange: we allow the
selfdestruct operation to be journalled several times. This makes it so
that we also are forced to store whether the account was already
destructed.
What we can do instead, is to only journal the first destruction, and
after that only journal balance-changes, but not journal the
selfdestruct itself.
This simplifies the journalling, so that internals about state
management does not leak into the journal-API.
Preimages were, for some reason, integrated into the journal management,
despite not being a consensus-critical data structure. This PR undoes
that.
---------
Co-authored-by: Martin HS <martin@swende.se>
Co-authored-by: Gary Rong <garyrong0905@gmail.com>
Currently our state journal tracks each storage update to a contract, having the ability to revert those changes to the previously set value.
For the very first modification however, it behaves a bit wonky. Reverting the update doesn't actually remove the dirty-ness of the slot, rather leaves it as "change this slot to it's original value". This can cause issues down the line with for example write witnesses needing to gather an unneeded proof.
This PR modifies the storageChange journal entry to not only track the previous value of a slot, but also whether there was any previous value at all set in the current execution context. In essence, the PR changes the semantic of storageChange so it does not simply track storage changes, rather it tracks dirty storage changes, an important distinction for being able to cleanly revert the journal item.
Co-authored-by: Péter Szilágyi <peterke@gmail.com>
The upstream libray has removed the assembly-based implementation of
keccak. We need to maintain our own library to avoid a peformance
regression.
---------
Co-authored-by: Felix Lange <fjl@twurst.com>
Co-authored-by: lightclient <lightclient@protonmail.com>
