This PR does a few things.
It fixes a shutdown-order flaw in the chainfreezer. Previously, the chain-freezer would shutdown the freezer backend first, and then signal for the loop to exit. This can lead to a scenario where the freezer tries to fsync closed files, which is an error-conditon that could lead to exit via log.Crit.
It also makes the printout more detailed when truncating 'dangling' items, by showing the exact number instead of approximate MB.
This PR also adds calls to fsync files before closing them, and also makes the `db inspect` command slightly more robust.
This PR fixes an issue which might result in data lost in freezer.
Whenever mutation happens in freezer, all data will be written into head data file
and it will be rotated with a new one in case the size of file reaches the threshold.
Theoretically, the rotated old data file should be fsync'd to prevent data loss.
In freezer.Sync function, we only fsync: (1) index file (2) meta file and (3) head
data file. So this PR forcibly fsync the head data file if mutation happens in the
boundary of data file.
Implementation of https://eips.ethereum.org/EIPS/eip-3860, limit and meter initcode. This PR enables EIP-3860 as part of the Shanghai fork.
Co-authored-by: lightclient@protonmail.com <lightclient@protonmail.com>
Co-authored-by: Martin Holst Swende <martin@swende.se>
Co-authored-by: Marius van der Wijden <m.vanderwijden@live.de>
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 changes moves the tracking of "deleted in this block" out from snap-only domain, so that it happens regardless of whether the execution is snapshot-backed or trie-backed.
This changes the StorageTrie method to return an error when the trie
is not available. It used to return an 'empty trie' in this case, but that's
not possible anymore under PBSS.
This PR builds on #26299, but also updates the tests to the most recent version, which includes tests regarding TheMerge.
This change adds checks to the beacon consensus engine, making it more strict in validating the pre- and post-headers, and not relying on the caller to have already correctly sanitized the headers/blocks.
This PR implements resettable freezer by adding a ResettableFreezer wrapper.
The resettable freezer wraps the original freezer in a way that makes it possible to ensure atomic resets. Implementation wise, it relies on the os.Rename and os.RemoveAll to atomically delete the original freezer data and re-create a new one from scratch.
A comment suggests that contract creation happens if the recipient of a call is 0x00..00 ("zero address") but in fact the sender must be nil. The zero address is a regular valid address that is commonly used as a "burn" address.
While investigating another issue, I found that all callers of collectLogs have the
complete block available. rawdb.ReadReceipts loads the block from the database,
so it is better to use ReadRawReceipts here, and derive the receipt information using
the block which is already in memory.
This PR makes it possible to modify the flush interval time via RPC. On one extreme, `0s`, it would act as an archive node. If set to `1h`, means that after one hour of effective block processing time, the trie would be flushed. If one block takes 200ms, this means that a flush would occur every `5*3600=18000` blocks -- however, if the memory size of the cached states grows too large, it will flush sooner.
Essentially, this makes it possible to configure the node to be more or less "archive:ish", and without restarting the node while reconfiguring it.
The gcproc field tracks the amount of time spent processing blocks,
and is used to trigger a state flush to disk when a certain threshold is
reached. After the merge, single block insertion by CL is the most
common source of block processing time, but this time was not added
into gcproc.
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>
This changes the Pop method to assign the zero value before
reducing slice size. Doing so ensures the backing array does not
reference removed item values.
This PR drops the legacy receipt types, the freezer-migrate command and the startup check. The previous attempt #22852 at this failed because there were users who still had legacy receipts in their db, so it had to be reverted #23247. Since then we added a command to migrate legacy dbs #24028.
As of the last hardforks all users either must have done the migration, or used the --ignore-legacy-receipts flag which will stop working now.
This PR introduces a node scheme abstraction. The interface is only implemented by `hashScheme` at the moment, but will be extended by `pathScheme` very soon.
Apart from that, a few changes are also included which is worth mentioning:
- port the changes in the stacktrie, tracking the path prefix of nodes during commit
- use ethdb.Database for constructing trie.Database. This is not necessary right now, but it is required for path-based used to open reverse diff freezer
While investigating #22374, I noticed that the Sync operation of the
freezer does not take the table lock. It also doesn't call sync for all files
if there is an error with one of them. I doubt this will fix anything, but
didn't want to drop the fix on the floor either.
It seems there is no fully typed library implementation of an LRU cache.
So I wrote one. Method names are the same as github.com/hashicorp/golang-lru,
and the new type can be used as a drop-in replacement.
Two reasons to do this:
- It's much easier to understand what a cache is for when the types are right there.
- Performance: the new implementation is slightly faster and performs zero memory
allocations in Add when the cache is at capacity. Overall, memory usage of the cache
is much reduced because keys are values are no longer wrapped in interface.
When the interpreter is configured to use extra-eips, this change makes it so that all the opcodes are deep-copied, to prevent accidental modification of the 'base' jumptable.
Closes: #26136
Co-authored-by: Martin Holst Swende <martin@swende.se>
This PR changes geth to read the eip1559 params from the chain config instead of the globals.
This way the parameters may be changed by forking the chain config code, without creating a large diff throughout the past and future usages of the parameters.
Co-authored-by: Martin Holst Swende <martin@swende.se>
This PR ports a few changes from PBSS:
- Fix the snapshot generator waiter in case the generation is not even initialized
- Refactor db inspector for ancient store
This PR fixes a regression causing snapshots not to be generated in "geth --import" mode. It also fixes the geth export command to be truly readonly, and adds a new test for geth export.
This adds a
* core/vm, tests: optimized modexp + fuzzer
* common/math: modexp optimizations
* core/vm: special case base 1 in big modexp
* core/vm: disable fastexp
This changes the error message for mismatching chain ID to show
the given and expected value. Callers expecting this error must be
changed to use errors.Is.
