* Upgrade parity-crypto to 0.6
* More fixes
* Upgrade ethabi and ethabi-derive
* Fix lockfile
* Patch ethabi from master
* quickfix for ethash
* Update forkid
* Add secret store back
* Fetch ethabi from crates
* fetch secret-store from the right place
* update to keccak-hash 0.5.1
* ethash: upgrade keccak-hash
* ethash: sneaky spaces
* ethash: use overlapping bytes after all
* revert submodule update
* wip
* Use new method to impl ecrecover builtin
* Fetch patches from git
* Sort out todo
* Remove patches
Co-authored-by: Andronik Ordian <write@reusable.software>
* Upgrade parity-crypto to 0.6
* More fixes
* Upgrade ethabi and ethabi-derive
* Fix lockfile
* Patch ethabi from master
* quickfix for ethash
* Update forkid
* Add secret store back
* Fetch ethabi from crates
* fetch secret-store from the right place
* update to keccak-hash 0.5.1
* ethash: upgrade keccak-hash
* ethash: sneaky spaces
* ethash: use overlapping bytes after all
* revert submodule update
Co-authored-by: Andronik Ordian <write@reusable.software>
The Clique engine changes the header during the call to `check_and_lock_block()` and so when the block is committed we need to use the original header from the `PreverifiedBlock`, so we're back to cloning the `Header`.
Fixes https://github.com/openethereum/openethereum/issues/11603
The exported hardcoded sync was previously generating invalid JSON, with
the CHT list ending in a trailing comma. In order to remedy this, this
commit uses serde_json to serialize the SpecHardcodedSync struct
into valid JSON.
Fixes#11415
`Tempdir` is deprecated, but the functionality has been merged into
another crate: `tempfile`. This commit removes all `tempdir` dependencies
and replaces them with `tempfile` and the equivalent bindings.
Fixes#11560
* Log block number and earliest available block when snapshots fail to start
* Increase default pruning history to 128
Increase default pruning memory to 48Mb
* Decrease SNAPSHOT_PERIOD to 200 to get them snapshots started quicker
* Sync updates to defaults with the args parser
* Fix tests
* Restore the SNAPSHOT_PERIOD to 5000
* One more
* Update ethcore/src/client/config.rs
* Code cleanup
`flush_queue()` is mostly used by tests
remove some unused params
avoid clones in a few places
* Consistent params naming
Consistent use of log calls
* Fix todo
It'a actually not trivial to find out the size of these collections, likely we keep blocks from all kinds of forks in the `self.blocks`/`self.parents` collections and there's no good way to anticipate how many blocks we're going to drain. Most of the time we end up draining 0 blocks and then we drain a whole bunch of them, up to 30 000.
* Revert making flush_queue() test-only
Address review grumbles
* More review grumbles
* fix build
* cleanup
* upbork
* review grumble: rename client id prefix
* rpc-tests: make one of the clients open-ethereum
* fix doctest
* README: mention clang version and that our wiki is old
* ethcore: cleanup after #11531
* fix verification benches
* ethcore: remove enact_verified
* ethcore: replace PreverifiedBlock with a tuple
* ethcore: more descriptive Vec<u8> type alias
* update all the changelogs
* update only github URLs in README.md
* sed magic
find . -type f -exec sed -i 's_paritytech/parity-ethereum_OpenEthereum/open-ethereum_g' {} \;
* Faster kill_garbage
Benchmark shows that almost half the time `apply()`-ing a transaction is spent in garbage collection. This PR avoids visiting each cache item and `collect()`-ing accounts to clean up.
* Walk back panicking behaviour
* Review grumble: prefer `and_then` to `if let`
* Add benchmark for transaction execution
* Address review grumbles
* Address review grumbles and extend benches to test both blocks with both Constantinople and Istanbul rules
* Misc fixes
Did some code reading and made random changes as I went.
* Update ethcore/src/client/client.rs
Co-Authored-By: Andronik Ordian <write@reusable.software>
Co-authored-by: Andronik Ordian <write@reusable.software>
* [ci benches]: use `all-features`
Our benches are hidden behind feature flags, this PR enables to type check on those.
`--all` is deprecated
* [bench verification]: remove block_family_partial
* [global allocator]: rm `feature=memory_profiling`
The `memory_profiling` feature doesn't work because `parity_util_mem` crate configures it globally.
* complete null-signatures removal
* unsigned transactions are disallowed by the type system
* "fix" verification test
* machine: bring the test back
* machine: simplify the test
* debug signer
* Don't panic if empty_steps_transition already happened
Before this the `header_empty_steps_raw` would panic if the chain has already progressed beyond the block number set in `emptyStepsTransition`. As this is a user accessible configuration option I don't think we should panic.
* Cleanup some code in Aura
Nothing really interesting here, renames or removes some methods. Adds some docs and extends a test a bit to clarify the behaviour of the code.
* Include the seal when populating the header for a new block (fixes#11445)
* cleanup
* cleanup2
* Review grumbles
* Update ethcore/engines/authority-round/src/lib.rs
Co-Authored-By: André Silva <andre.beat@gmail.com>
* Update ethcore/engines/authority-round/src/lib.rs
Co-Authored-By: André Silva <andre.beat@gmail.com>
* Update ethcore/src/block.rs
Co-Authored-By: André Silva <andre.beat@gmail.com>
Co-authored-by: André Silva <andre.beat@gmail.com>
* Use parity-crypto updated to use upstream rust-secp256k1
* Fetch dependency from git
* Missed a session ID
* Add free-standing inversion function that uses `libsecp256k1`
* fixed tests
* Update deps
* Use parity-crypto 0.5.0
Use libsecp256k1 0.3.5
* Review grumble
Co-authored-by: Svyatoslav Nikolsky <svyatonik@gmail.com>
* Cleanup some code in Aura
Nothing really interesting here, renames or removes some methods. Adds some docs and extends a test a bit to clarify the behaviour of the code.
* Update ethcore/engines/authority-round/src/lib.rs
Co-Authored-By: Niklas Adolfsson <niklasadolfsson1@gmail.com>
Co-authored-by: Niklas Adolfsson <niklasadolfsson1@gmail.com>
