Here, the core.Message interface turns into a plain struct and
types.Message gets removed.
This is a breaking change to packages core and core/types. While we do
not promise API stability for package core, we do for core/types. An
exception can be made for types.Message, since it doesn't have any
purpose apart from invoking the state transition in package core.
types.Message was also marked deprecated by the same commit it
got added in, 4dca5d4db7 (November 2016).
The core.Message interface was added in December 2014, in commit
db494170dc, for the purpose of 'testing' state transitions. It's the
same change that made transaction struct fields private. Before that,
the state transition used *types.Transaction directly.
Over time, multiple implementations of the interface accrued across
different packages, since constructing a Message is required whenever
one wants to invoke the state transition. These implementations all
looked very similar, a struct with private fields exposing the fields
as accessor methods.
By changing Message into a struct with public fields we can remove all
these useless interface implementations. It will also hopefully
simplify future changes to the type with less updates to apply across
all of go-ethereum when a field is added to Message.
---------
Co-authored-by: Felix Lange <fjl@twurst.com>
Adds support for a native call tracer with the Parity format, which outputs call frames
in a flat array. This tracer accepts the following options:
- `convertParityErrors: true` will convert error messages to match those of Parity
- `includePrecompiles: true` will report all calls to precompiles. The default
matches Parity's behavior where CALL and STATICCALLs to precompiles are excluded
Incompatibilities with Parity include:
- Parity removes the result object in case of failure. This behavior is maintained
with the exception of reverts. Revert output usually contains useful information,
i.e. Solidity revert reason.
- The `gasUsed` field accounts for intrinsic gas (e.g. 21000 for simple transfers)
and refunds unlike Parity
- Block rewards are not reported
Co-authored-by: Sina Mahmoodi <itz.s1na@gmail.com>
This commit squashes the op-geth fork history into a more maintainable
diff for rebasing upon upstream geth.
reference-optimistic-geth changes (origins of op-geth in early Bedrock
development stage):
- Deposit TX Type
- Enable deposit tx in EVM/tx pool
- Change deposit nonce to not be the max nonce
- Extend PayloadAttributesV1 with a Transactions field
- Force deposits at the start of each L2 block
- Fix height check
- noTxPool flag, reproduce block in verifier mode without tx pool interference
- Fix RPC json marshalling (ref op-geth PR 4)
- Deposit txs block height check in block body validation (ref op-geth PR 5)
- core: do not try to reinject deposit txs into tx-pool (ref-op-geth PR 6)
- deposit source hash field instead of L2 block height and tx index combination
- Include invalid deposits, rewind state, but always persist mint (#10)
- Provide gas to Call/Create in deposit transactions (#12)
- Add docker builds (ref-op-geth PR 16, 17)
- Don't panic on deposit transaction signature values or chain ID (ref-op-geth PR 18)
- core: Add version to DepositTx (ref-op-geth PR 19)
- Enable Geth build/lint/test in CircleCI (ref-op-geth PR 23)
- core: Include guaranteed gas in the gas pool (ref-op-geth PR 21)
- core: handle base fee, l1 availability fee, tx fee (ref-op-geth PR 27)
- fix: deposit tx hash
- fix l1 fee cache, rpc, tracing and tx pool
- core: remove deposit-tx sub-type (a.k.a. deposit version byte)
- eth/catalyst: allow engine user to reorg own chain
- miner: restore ability to reorg deep as block builder
- params: print Optimism consensus type in banner
- core/types: remove unused protected() method, see upstream PR 23376
- core: do not mutate original balance value in tx pool l1 cost adjustment
- core: subtract deposit gas from pool, so other txs do not use the same gas. And fail tx processing if deposits reach gas limit
- core/types: deposits do not tip, avoid basefee subtraction
- Unmeter the L1 Attributes Transaction
- miner: handle force tx errors as critical, clean up diff
- ci: Switch branch
- eth,miner: return STATUS_INVALID when failing to process forced transactions in request (ref-op-geth PR 40)
- verifier: forward tx to sequencer based on flag
- txpool: add flag to disable tx gossip (ref-op-geth PR 42)
- Add op-geth version in addition to geth version (ref-op-geth PR 43)
- ci: CircleCI improvements (ref-op-geth PR 44)
- Rename to op-geth
- Build latest tag on optimism branch
op-geth changes:
- Expose cache config in simulated backend (#2)
- Add EIP-1559 parameters
- eth/catalyst: update payload id computation (#1)
- make eip1559 configurable (#4)
- post-merge network should not log warnings about missing transition information (#5)
- Make the simulator more configurable (#6)
- fix OPB-6 - IsDepositTx check instead of artificial nonce value check (#7)
- Simulated backend - enable proof of stake consensus type and fix performance issue (#8)
- accounts: simulated backend consensus engine option and immediate tx indexing
- consensus/beacon: recognize all blocks as reached TTD with 0 TTD in chain config
- Add --rollup.historicalhttp CLI flag and fix backend iface
- Flags and interfaces for historical RPC requests (#12)
- Redirect historical RPC requests (#13)
- Use the pre-existing ethereum.NotFound error (#18)
- Add historical endpoint to TraceBlockByNumber and TraceBlockByHash (#19)
- Add historical endpoint to TraceTransaction (#20)
- Add historical endpoint to TraceCall (#21)
- optimism: fee params from info txi, update l1 cost func GPO params read (#15)
- add hardcoded addresses for fee payouts (#23)
- dynamic gas limit via engine API (#22)
Co-authored-by: Matthew Slipper <me@matthewslipper.com>
Co-authored-by: Joshua Gutow <jgutow@oplabs.co>
Co-authored-by: protolambda <proto@protolambda.com>
Co-authored-by: Mark Tyneway <mark.tyneway@gmail.com>
Co-authored-by: Maurelian <maurelian@protonmail.ch>
This is a breaking change in the tracing hooks API as well as semantics of the callTracer:
- CaptureEnter hook provided a nil value argument in case of DELEGATECALL. However to stay consistent with how delegate calls behave in EVM this hook is changed to pass in the value of the parent call.
- callTracer will return parent call's value for DELEGATECALL frames.
---------
Co-authored-by: Sina Mahmoodi <itz.s1na@gmail.com>
This PR changes the API so that uint64 is used for fork timestamps.
It's a good choice because types.Header also uses uint64 for time.
Co-authored-by: Felix Lange <fjl@twurst.com>
This makes non-JS tracers execute all block txs on a single goroutine.
In the previous implementation, we used to prepare every tx pre-state
on one goroutine, and then run the transactions again with tracing enabled.
Native tracers are usually faster, so it is faster overall to use their output as
the pre-state for tracing the next transaction.
Co-authored-by: Sina Mahmoodi <itz.s1na@gmail.com>
This PR removes the notion of fakeStorage from the state objects, and instead, for any state modifications that are needed, it simply makes the changes.
This ensures that RPC method handlers will react to a timeout or
cancelled request soon after the event occurs.
Co-authored-by: Sina Mahmoodi <itz.s1na@gmail.com>
Currently calling `debug_TraceTransaction` with a transaction hash that doesn't exist returns a confusing error: `genesis is not traceable`. This PR changes the behaviour to instead return an error message saying `transaction not found`
Co-authored-by: Martin Holst Swende <martin@swende.se>
Co-authored-by: Sina Mahmoodi <itz.s1na@gmail.com>
This removes the 'time' field from logs, as well as from the tracer interface. This change makes the trace output deterministic. If a tracer needs the time they can measure it themselves. No need for evm to do this.
Co-authored-by: Sina Mahmoodi <itz.s1na@gmail.com>
In some cases, it is desirable to capture what is triggered by each trace, when using the `callTracer`. For example: call `USDT.transfer` will trigger a `Transfer(from, to, value)` event.
This PR adds the option to capture logs to the call tracer, by specifying `{"withLog": true}` in the tracerconfig.
Any logs belonging to failed/reverted call-scopes are removed from the output, to prevent interpretation mistakes.
Signed-off-by: Delweng <delweng@gmail.com>
Co-authored-by: Sina Mahmoodi <itz.s1na@gmail.com>
Inner call reverts will now return the reason similar to the top-level call. Separately, if top-level call is of type CREATE and it fails, its `to` field will now be cleared to `0x00...00` instead of being set to the created address.
* eth/tracers: fix gasUsed in call tracer
* fix js tracers gasUsed
* fix legacy prestate tracer
* fix restGas in test
* drop intrinsicGas field from js tracers
The prestate tracer did not report accounts that existed at a given address prior to a contract being created at that address.
Signed-off-by: Delweng <delweng@gmail.com>
Co-authored-by: Sina Mahmoodi <itz.s1na@gmail.com>
In some cases, inner contract creation may not be successful, and an inner contract was not created. This PR fixes a crash that could occur when doing tracing in such situations.
* ethclient/gethclient: improve time-sensitive flaky test
* eth/catalyst: fix (?) flaky test
* core: stop blockchains in tests after use
* core: fix dangling blockchain instances
* core: rm whitespace
* eth/gasprice, eth/tracers, consensus/clique: stop dangling blockchains in tests
* all: address review concerns
* core: goimports
* eth/catalyst: fix another time-sensitive test
* consensus/clique: add snapshot test run function
* core: rename stop() to stopWithoutSaving()
Co-authored-by: Felix Lange <fjl@twurst.com>
This PR introduces a new mechanism in chain tracer for preventing creating too many trace states.
The workflow of chain tracer can be divided into several parts:
- state creator generates trace state in a thread
- state tracer retrieves the trace state and applies the tracing on top in another thread
- state collector gathers all result from state tracer and stream to users
It's basically a producer-consumer model here, while if we imagine that the state producer generates states too fast, then it will lead to accumulate lots of unused states in memory. Even worse, in path-based state scheme it will only keep the latest 128 states in memory, and the newly generated state will invalidate the oldest one by marking it as stale.
The solution for fixing it is to limit the speed of state generation. If there are over 128 states un-consumed in memory, then the creation will be paused until the states are be consumed properly.
Backwards compatibility warning: The result will from now on omit empty fields instead
of including a zero value (e.g. no more `balance: '0x'`).
The prestateTracer will now take an option `diffMode: bool`. In this mode
the tracer will output the pre state and post data for the modified parts of state.
Read-only accesses will be completely omitted. Creations (be it account or slot)
will be signified by omission in the `pre` list and inclusion in `post`. Whereas
deletion (be it account or slot) will be signified by inclusion in `pre` and omission
in `post` list.
Signed-off-by: Delweng <delweng@gmail.com>
The call tracer and prestate tracer store data JSON-encoded in memory. In order to support alternative encodings (specifically RLP), it's better to keep data a native format during tracing. This PR does marshalling at the end, using gencodec.
OBS!
This PR changes the call tracer result slightly:
- Order of type and value fields are changed (should not matter).
- Output fields are completely omitted when they're empty (no more output: "0x"). Previously, this was only _sometimes_ omitted (e.g. when call ended in a non-revert error) and otherwise 0x when the output was actually empty.
* eth/tracers: pad memory slice on oob case
* eth/tracers/js: fix testfailure due to err msg capitalization
Co-authored-by: Martin Holst Swende <martin@swende.se>
This changes the CI / release builds to use the latest Go version. It also
upgrades golangci-lint to a newer version compatible with Go 1.19.
In Go 1.19, godoc has gained official support for links and lists. The
syntax for code blocks in doc comments has changed and now requires a
leading tab character. gofmt adapts comments to the new syntax
automatically, so there are a lot of comment re-formatting changes in this
PR. We need to apply the new format in order to pass the CI lint stage with
Go 1.19.
With the linter upgrade, I have decided to disable 'gosec' - it produces
too many false-positive warnings. The 'deadcode' and 'varcheck' linters
have also been removed because golangci-lint warns about them being
unmaintained. 'unused' provides similar coverage and we already have it
enabled, so we don't lose much with this change.
This PR simplifies the logic of chain tracer and also adds the unit tests.
The most important change has been made in this PR is the state management. Whenever a tracing state is acquired there is a corresponding release function be returned as well. It must be called once the state is used up, otherwise resource leaking can happen.
And also the logic of state management has been simplified a lot. Specifically, the state provider(eth backend, les backend) should ensure the state is available and referenced. State customers can use the state according to their own needs, or build other states based on the given state. But once the release function is called, there is no guarantee of the availability of the state.
Co-authored-by: Sina Mahmoodi <1591639+s1na@users.noreply.github.com>
Co-authored-by: Péter Szilágyi <peterke@gmail.com>
