* beefy-mmr: reuse sp_runtime::traits::Keccak256
* beefy-mmr: use sp_runtime::traits:Hash for generating merkle proofs
* beefy-mmr: use sp_runtime::traits:Hash for validating merkle proofs
* beefy-mmr: remove primitives::Hasher and primitives::Hash
* fixes
* beefy-mmr: reduce the number of generic parameters for merkle_root()
* fix
* compute upper Vec capacity more accurately
* client/beefy: create communication module and move gossip there
* client/beefy: move beefy_protocol_name module to communication
* client/beefy: move notification module under communication
* client/beefy: add incoming request_response protocol handler
* client/beefy: keep track of connected peers and their progress
* client/beefy: add logic for generating Justif requests
* client/beefy: cancel outdated on-demand justification requests
* try Andre's suggestion for JustificationEngine
* justif engine add justifs validation
* client/beefy: impl OnDemandJustificationsEngine async next()
* move beefy proto name test
* client/beefy: initialize OnDemandJustificationsEngine
* client/tests: allow for custom req-resp protocols
* client/beefy: on-demand-justif: implement simple peer selection strategy
* client/beefy: fix voter initialization
Fix corner case where voter gets a single burst of finality
notifications just when it starts.
The notification stream was consumed by "wait_for_pallet" logic,
then main loop would subscribe to finality notifications, but by that
time some notifications might've been lost.
Fix this by subscribing the main loop to notifications before waiting
for pallet to become available. Share the same stream with the main loop
so that notifications for blocks before pallet available are ignored,
while _all_ notifications after pallet available are processed.
Add regression test for this.
Signed-off-by: acatangiu <adrian@parity.io>
* client/beefy: make sure justif requests are always out for mandatory blocks
* client/beefy: add test for on-demand justifications sync
* client/beefy: tweak main loop event processing order
* client/beefy: run on-demand-justif-handler under same async task as voter
* client/beefy: add test for known-peers
* client/beefy: reorg request-response module
* client/beefy: add issue references for future work todos
* client/beefy: consolidate on-demand-justifications engine state machine
Signed-off-by: acatangiu <adrian@parity.io>
* client/beefy: fix for polkadot companion
* client/beefy: implement review suggestions
* cargo fmt and clippy
* fix merge damage
* fix rust-doc
* fix merge damage
* fix merge damage
* client/beefy: add test for justif proto name
Signed-off-by: acatangiu <adrian@parity.io>
* Remove contracts RPC
* update lockfile for {"polkadot", "substrate"}
* Remove unused imports
* Properly set a max proof size for cumulus test runtime
* Properly set a max proof size for common runtimes
* Properly set the max proof size for all runtimes
Co-authored-by: parity-processbot <>
Co-authored-by: Keith Yeung <kungfukeith11@gmail.com>
* Remove contracts RPCs
* Remove serde as RPC serialization is no longer needed
* Rename folder to match crate name
* Compile fix
* Remove Byte wrapper
* BEEFY: generate historical proofs
Signed-off-by: Serban Iorga <serban@parity.io>
* cargo update -p sp-io
* Properly set max proof size for runtimes
* Properly set max proof size for mocks
* cargo fmt
* Set appropriate UMP service total proof size weight
* Disable zombienet-tests-parachains-disputes CI
* Add comment explaining weight math
* Use MAX_POV_SIZE for max proof size
* Cast to u64
* Remove comment
Signed-off-by: Serban Iorga <serban@parity.io>
Co-authored-by: Keith Yeung <kungfukeith11@gmail.com>
* Properly set the max proof size weight on defaults and tests
* cargo fmt
* Set proper max proof size for contracts pallet tests
* Properly set max proof size for node
* Properly set max proof size for frame system mock
* Update test expectations
* Update test expectations
* Properly set max proof size for balances mock
* Update test expectations
* Update test expectations
* Properly set max proof size for democracy mock
* Properly set max proof size for scheduler mock
* Properly set max proof size for fast unstake mock
* Properly set max proof size for tx payment mock
* Properly set max proof size for elections phragmen mock
* Properly set max proof size for node template
This can happen under heavy load - no value in warning here.
Scenario this happens:
- New slot
- We get asked about what fork to build on
- We start building (create inherent gets called)
- We learn about a better fork (but Babe does not care as it already has a fork)
- 6 seconds passed - New slot
- Babe asks about what for to build on - we answer with the best block we learned about 6 seconds ago (slightly after we got asked the last time)
- We start building (on that old block)
- Milliseconds later we learn about a new block (the one from the slot we are actually in now)
- We kick the old leaf -> create inherent fails
Full discussion: https://github.com/paritytech/substrate/issues/12375
* Support running the pallet benchmarks analysis without running the benchmarks
* Rename `override-results` to `json-input` and update the help comment
* ".git/.scripts/fmt.sh" 1
Co-authored-by: command-bot <>
* Add storage size component to weights
* Rename storage_size to proof_size
* Update primitives/weights/src/weight_v2.rs
Co-authored-by: Oliver Tale-Yazdi <oliver.tale-yazdi@parity.io>
* Fixes
* cargo fmt
* Implement custom Decode and CompactAs
* Add missing import
* Fixes
* Remove CompactAs implementation
* Properly migrate from 1D weight
* Remove #[pallet::compact] from Weight parameters
* More #[pallet::compact] removals
* Add unit tests
* Set appropriate default block proof size
* cargo fmt
* Remove nonsensical weight constant
* Test only for the reference time weight in frame_system::limits
* Only check for reference time weight on idle
* Use destructuring syntax
* Update test expectations
* Fixes
* Fixes
* Fixes
* Correctly migrate from 1D weights
* cargo fmt
* Migrate using extra extrinsics instead of custom Decode
* Fixes
* Silence dispatch call warnings that were previously allowed
* Fix gas_left test
* Use OldWeight instead of u64
* Fixes
* Only check for reference time weight in election provider
* Fix test expectations
* Fix test expectations
* Use only reference time weight in grandpa test
* Use only reference time weight in examples test
* Use only reference time weight in examples test
* Fix test expectations
Co-authored-by: Oliver Tale-Yazdi <oliver.tale-yazdi@parity.io>
Co-authored-by: Alexander Theißen <alex.theissen@me.com>
Roman Useinov