* variants for output_pos linked list entries (head/tail/middle/unique)
next and prev and Vec<u8> lmdb keys
* get_pos on enum
* break list and list entries out into separate enums
* track output features in the new output_pos index, so we can determine coinbase maturity
* push entry impl for none and unique
* some test coverage for output_pos_list
* commit
* wip - FooListEntry
* use instance of the index
* linked list of output_pos and commit_pos both now supported
* linked_list
* cleanup and rename
* rename
* peek_pos
* push some, peek some, pop some
* cleanup
* commit pos
cleanup
* split list and entry out into separate db prefixes
* cleanup and add placeholder for pop_back
* pop_pos_back (for popping off the back of the linked list)
test coverage for pop_pos_back
* wip
* placeholder for prune via a trait
pos must always increase in the index
* rewind kernel_pos_idx when calling rewind_single_block
* RewindableListIndex with rewind support.
* test coverage for rewindable list index
* test coverage for rewind back to 0
* rewind past end of list
* add tests for kernel_pos_idx with multiple commits
* commit
* cleanup
* hook NRD relative lock height validation into block processing and tx validation
* cleanup
* set local chain type for kernel_idx tests
* add test coverage for NRD rules in block processing
* NRD test coverage and cleanup
* NRD relative height 1 test
* test coverage for NRD kernels in block processing
* cleanup
* start of test coverage for txpool NRD kernel rules
* wip
* rework pool tests to use real chain (was mock chain) to better reflect reality (tx/block validation rules etc.)
* cleanup
* cleanup pruneable trait for kernel pos index
* add clear() to kernel_pos idx and test coverage
* hook kernel_pos rebuild into node startup, compaction and fast sync
* verify full NRD history on fast sync
* return early if nrd disabled
* fix header sync issue
We have a pattern in the code - return Vec, turn it into an iterator, filter and collect to another Vec. The idea was to return iterator where possible to avoid allocating intermediate vecs.
* Introduce GLOBAL_CHAIN_TYPE and make CHAIN_TYPE thread_local.
This makes testing more explicit and significantly more robust.
* set_local_chain_type() in tests
* cleanup - weird
* get pool tests working with explicit local chain_type config
* core tests working with explicit local chain_type
* p2p tests working with explicit local chain_type
* store tests working
* cleanup, feedback
We have a method get which returns a Vec, it's used in cases when we can't implement FromLmdbBytes or Readable on a desired type, eg when both traits and type are external for the current crate. This approach requires an extra allocation of a Vec, this PR introduces a method which accepts desirialzer as closure, which allows to desrialize in place without an intermidiate heap allocation.
It's still possible to get a Vec if needed, just pass a Vec constructor to get_with.
We have to make an extra allocation per db get request because key generation function to_key takes Vec. Taking byte slice (AsRef<[u8]> to be precise) also simplifes the code a bit.
Currently Writable accepts trait Write as a type parameter but Readable
takes Read as a trait object, which is not symmetrical and also less performant. This PR changes Readable trait and all places where it's used
* no need for AsFixedBytes we can just use AsRef<[u8]>
* cleanup FixedLength trait
* revert this change for now
* fix store tests
* cleanup and fix tests after rebase
* fix tests
* update based on PR review
less hard-coded values now
* cleanup
* add bitmap accumulator
refactor vec backend so we can use it outside of tests
introduce a "hash only" vec backend for the accumulator
* get core tests passing
* initial test coverage for bitmap_accumulator
* better test coverage for bitmap accumulator and cleanup code
* refactor txhashset roots, call validate() on roots during block validation
* fix store tests
* log the "merged" root when validating roots
* cleanup, revise based on feedback
* cleanup
* rework it to pass explicit size into bitmap accumulator when applying
* fix: split state validation status into kernel and rproof updates. And fix sync status for these two states
* fix: show correct number of leaves for pruned MMR as well as unpruned
* docs: better docs for kernel/range proof validation
* fix: ordering of kernel and rproofs validation in TUI
* fix: typo in rangeproofs api and comments
To convert option to error we generate an error message. In some places
it contains header or block hash code or other data which is costly to
produce. So during the initial header sync we spend 12% of all time on
generating those messages (in 99% cases we don't use it). This PR
introduces a lazy generation of error messages which completely
eliminates CPU load during the header sync.
* store both mmr index and block height into database for output
* rustfmt
* fix: mmr position is 1-based instead of 0-based
* (Hash, u64, u64) deserves a type
