Firstly, `Step.duration_remaining` casts it to u32, unnecesarily
limiting it to 2^32. While theoretically this is "good enough" (at 3
seconds steps it provides room for a little over 400 years), it is
still a lossy way to calculate the remaining time until the next step.
Secondly, step duration might be zero, triggering division by zero
in `Step.calibrate`
Solution: rework the code around the fact that duration is
typically in single digits and never grows, hence, it can be represented
by a much narrower range (u16) and this highlights the fact that
multiplying u64 by u16 will only result in an overflow in even further
future, at which point we should panic informatively (if anybody's
still around)
Similarly, panic when it is detected that incrementing the step
counter wrapped around on the overflow of usize.
As for the division by zero, prevent it by making zero an invalid
value for step duration. This will make AuRa log the constraint
mismatch and panic (after all, what purpose would zero step duration
serve? it makes no sense within the definition of the protocol,
as finality can only be achieved as per the specification
if messages are received within the step duration, which would violate
the speed of light and other physical laws in this case).
* move common forks and parameters to common params
* port specs over to new format
* fix RPC tests
* parity-machine skeleton
* remove block type
* extract out ethereum-specific methods into EthereumMachine
* beginning to integrate Machine into engines. dealing with stale transitions in Ethash
* initial porting to machine
* move block reward back into engine
* abstract block reward logic
* move last hash and DAO HF logic into machine
* begin making engine function parameters generic
* abstract epoch verifier and ethash block reward logic
* instantiate special ethereummachine for ethash in spec
* optional full verification in verify_block_family
* re-instate tx_filter in a way that works for all engines
* fix warnings
* fix most tests, further generalize engine trait
* uncomment nullengine, get ethcore tests compiling
* fix warnings
* update a bunch of specs
* re-enable engine signer, validator set, and transition handler
* migrate basic_authority engine
* move last hashes into executedblock
* port tendermint
* make all ethcore tests pass
* json-tests compilation
* fix RPC tests: change in gas limit for new block changed PoW hash
* fix minor grumbles
* validate chainspecs
* fix broken import
* fix transaction verification for pre-homestead
* generate proofs on newly-created state
* report only missed steps after first block
* dont report skipped if not signer
* test
* finality tracing and passing valid header to `commit_block`
* avoid proposing multiple times on the same step when validator set changes
* limit benign reports
* Ordering -> AtomicOrdering
* reinstate warning now that spam is reduced
* flush pending transition changes when necessary
* ensure epochs aren't re-zoomed on every block
This is a huge change, which includes some changes to replace code that
originally cloned to reuse allocations instead. The updated
`elastic-array` crate renames its consuming `Vec`-conversion method to
`into_vec`, which means that I can do a simple
`sed -i 's/to_vec/into_vec/'` and then fix the compilation errors.
This commit is probably a minor performance win and definitely a
significant readability win.
