mod.rs

// Copyright 2020 Parity Technologies (UK) Ltd.
// This file is part of Polkadot.

// Polkadot is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.

// Polkadot is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
// GNU General Public License for more details.

// You should have received a copy of the GNU General Public License
// along with Polkadot.  If not, see <http://www.gnu.org/licenses/>.

//! Provides glue code over the scheduler and inclusion modules, and accepting
//! one inherent per block that can include new para candidates and bitfields.
//!
//! Unlike other modules in this crate, it does not need to be initialized by the initializer,
//! as it has no initialization logic and its finalization logic depends only on the details of
//! this module.

use crate::{
	configuration,
	disputes::{DisputesHandler, VerifyDisputeSignatures},
	inclusion,
	inclusion::{CandidateCheckContext, FullCheck},
	initializer,
	metrics::METRICS,
	scheduler::{self, CoreAssignment, FreedReason},
	shared, ump, ParaId,
};
use bitvec::prelude::BitVec;
use frame_support::{
	inherent::{InherentData, InherentIdentifier, MakeFatalError, ProvideInherent},
	pallet_prelude::*,
	traits::Randomness,
};
use frame_system::pallet_prelude::*;
use pallet_babe::{self, ParentBlockRandomness};
use primitives::v2::{
	BackedCandidate, CandidateHash, CandidateReceipt, CheckedDisputeStatementSet,
	CheckedMultiDisputeStatementSet, CoreIndex, DisputeStatementSet,
	InherentData as ParachainsInherentData, MultiDisputeStatementSet, ScrapedOnChainVotes,
	SessionIndex, SigningContext, UncheckedSignedAvailabilityBitfield,
	UncheckedSignedAvailabilityBitfields, ValidatorId, ValidatorIndex, ValidityAttestation,
	PARACHAINS_INHERENT_IDENTIFIER,
};
use rand::{seq::SliceRandom, SeedableRng};

use scale_info::TypeInfo;
use sp_runtime::traits::{Header as HeaderT, One};
use sp_std::{
	cmp::Ordering,
	collections::{btree_map::BTreeMap, btree_set::BTreeSet},
	prelude::*,
	vec::Vec,
};

mod misc;
mod weights;

pub use self::{
	misc::{IndexedRetain, IsSortedBy},
	weights::{
		backed_candidate_weight, backed_candidates_weight, dispute_statement_set_weight,
		multi_dispute_statement_sets_weight, paras_inherent_total_weight, signed_bitfields_weight,
		TestWeightInfo, WeightInfo,
	},
};

#[cfg(feature = "runtime-benchmarks")]
mod benchmarking;

#[cfg(test)]
mod tests;

const LOG_TARGET: &str = "runtime::inclusion-inherent";

/// A bitfield concerning concluded disputes for candidates
/// associated to the core index equivalent to the bit position.
#[derive(Default, PartialEq, Eq, Clone, Encode, Decode, RuntimeDebug, TypeInfo)]
pub(crate) struct DisputedBitfield(pub(crate) BitVec<u8, bitvec::order::Lsb0>);

impl From<BitVec<u8, bitvec::order::Lsb0>> for DisputedBitfield {
	fn from(inner: BitVec<u8, bitvec::order::Lsb0>) -> Self {
		Self(inner)
	}
}

#[cfg(test)]
impl DisputedBitfield {
	/// Create a new bitfield, where each bit is set to `false`.
	pub fn zeros(n: usize) -> Self {
		Self::from(BitVec::<u8, bitvec::order::Lsb0>::repeat(false, n))
	}
}

pub use pallet::*;

#[frame_support::pallet]
pub mod pallet {
	use super::*;

	#[pallet::pallet]
	#[pallet::generate_store(pub(super) trait Store)]
	#[pallet::without_storage_info]
	pub struct Pallet<T>(_);

	#[pallet::config]
	#[pallet::disable_frame_system_supertrait_check]
	pub trait Config:
		inclusion::Config + scheduler::Config + initializer::Config + pallet_babe::Config
	{
		/// Weight information for extrinsics in this pallet.
		type WeightInfo: WeightInfo;
	}

	#[pallet::error]
	pub enum Error<T> {
		/// Inclusion inherent called more than once per block.
		TooManyInclusionInherents,
		/// The hash of the submitted parent header doesn't correspond to the saved block hash of
		/// the parent.
		InvalidParentHeader,
		/// Disputed candidate that was concluded invalid.
		CandidateConcludedInvalid,
		/// The data given to the inherent will result in an overweight block.
		InherentOverweight,
		/// The ordering of dispute statements was invalid.
		DisputeStatementsUnsortedOrDuplicates,
		/// A dispute statement was invalid.
		DisputeInvalid,
	}

	/// Whether the paras inherent was included within this block.
	///
	/// The `Option<()>` is effectively a `bool`, but it never hits storage in the `None` variant
	/// due to the guarantees of FRAME's storage APIs.
	///
	/// If this is `None` at the end of the block, we panic and render the block invalid.
	#[pallet::storage]
	pub(crate) type Included<T> = StorageValue<_, ()>;

	/// Scraped on chain data for extracting resolved disputes as well as backing votes.
	#[pallet::storage]
	#[pallet::getter(fn on_chain_votes)]
	pub(crate) type OnChainVotes<T: Config> = StorageValue<_, ScrapedOnChainVotes<T::Hash>>;

	/// Update the disputes statements set part of the on-chain votes.
	pub(crate) fn set_scrapable_on_chain_disputes<T: Config>(
		session: SessionIndex,
		checked_disputes: CheckedMultiDisputeStatementSet,
	) {
		crate::paras_inherent::OnChainVotes::<T>::mutate(move |value| {
			let disputes =
				checked_disputes.into_iter().map(DisputeStatementSet::from).collect::<Vec<_>>();
			let backing_validators_per_candidate = match value.take() {
				Some(v) => v.backing_validators_per_candidate,
				None => Vec::new(),
			};
			*value = Some(ScrapedOnChainVotes::<T::Hash> {
				backing_validators_per_candidate,
				disputes,
				session,
			});
		})
	}

	/// Update the backing votes including part of the on-chain votes.
	pub(crate) fn set_scrapable_on_chain_backings<T: Config>(
		session: SessionIndex,
		backing_validators_per_candidate: Vec<(
			CandidateReceipt<T::Hash>,
			Vec<(ValidatorIndex, ValidityAttestation)>,
		)>,
	) {
		crate::paras_inherent::OnChainVotes::<T>::mutate(move |value| {
			let disputes = match value.take() {
				Some(v) => v.disputes,
				None => MultiDisputeStatementSet::default(),
			};
			*value = Some(ScrapedOnChainVotes::<T::Hash> {
				backing_validators_per_candidate,
				disputes,
				session,
			});
		})
	}

	#[pallet::hooks]
	impl<T: Config> Hooks<BlockNumberFor<T>> for Pallet<T> {
		fn on_initialize(_: T::BlockNumber) -> Weight {
			T::DbWeight::get().reads_writes(1, 1) // in `on_finalize`.
		}

		fn on_finalize(_: T::BlockNumber) {
			if Included::<T>::take().is_none() {
				panic!("Bitfields and heads must be included every block");
			}
		}
	}

	#[pallet::inherent]
	impl<T: Config> ProvideInherent for Pallet<T> {
		type Call = Call<T>;
		type Error = MakeFatalError<()>;
		const INHERENT_IDENTIFIER: InherentIdentifier = PARACHAINS_INHERENT_IDENTIFIER;

		fn create_inherent(data: &InherentData) -> Option<Self::Call> {
			let inherent_data = Self::create_inherent_inner(data)?;
			// Sanity check: session changes can invalidate an inherent,
			// and we _really_ don't want that to happen.
			// See <https://github.com/paritytech/polkadot/issues/1327>

			// Calling `Self::enter` here is a safe-guard, to avoid any discrepancy between on-chain checks
			// (`enter`) and the off-chain checks by the block author (this function). Once we are confident
			// in all the logic in this module this check should be removed to optimize performance.

			let inherent_data = match Self::enter_inner(inherent_data.clone(), FullCheck::Skip) {
				Ok(_) => inherent_data,
				Err(err) => {
					log::error!(
						target: LOG_TARGET,
						"dropping paras inherent data because they produced \
							an invalid paras inherent: {:?}",
						err.error,
					);

					ParachainsInherentData {
						bitfields: Vec::new(),
						backed_candidates: Vec::new(),
						disputes: Vec::new(),
						parent_header: inherent_data.parent_header,
					}
				},
			};

			Some(Call::enter { data: inherent_data })
		}

		fn is_inherent(call: &Self::Call) -> bool {
			matches!(call, Call::enter { .. })
		}
	}

	/// Collect all freed cores based on storage data. (i.e. append cores freed from timeouts to
	/// the given `freed_concluded`).
	///
	/// The parameter `freed_concluded` contains all core indicies that became
	/// free due to candidate that became available.
	pub(crate) fn collect_all_freed_cores<T, I>(
		freed_concluded: I,
	) -> BTreeMap<CoreIndex, FreedReason>
	where
		I: core::iter::IntoIterator<Item = (CoreIndex, CandidateHash)>,
		T: Config,
	{
		// Handle timeouts for any availability core work.
		let availability_pred = <scheduler::Pallet<T>>::availability_timeout_predicate();
		let freed_timeout = if let Some(pred) = availability_pred {
			<inclusion::Pallet<T>>::collect_pending(pred)
		} else {
			Vec::new()
		};

		// Schedule paras again, given freed cores, and reasons for freeing.
		let freed = freed_concluded
			.into_iter()
			.map(|(c, _hash)| (c, FreedReason::Concluded))
			.chain(freed_timeout.into_iter().map(|c| (c, FreedReason::TimedOut)))
			.collect::<BTreeMap<CoreIndex, FreedReason>>();
		freed
	}

	#[pallet::call]
	impl<T: Config> Pallet<T> {
		/// Enter the paras inherent. This will process bitfields and backed candidates.
		#[pallet::call_index(0)]
		#[pallet::weight((
			paras_inherent_total_weight::<T>(
				data.backed_candidates.as_slice(),
				data.bitfields.as_slice(),
				data.disputes.as_slice(),
			),
			DispatchClass::Mandatory,
		))]
		pub fn enter(
			origin: OriginFor<T>,
			data: ParachainsInherentData<T::Header>,
		) -> DispatchResultWithPostInfo {
			ensure_none(origin)?;

			ensure!(!Included::<T>::exists(), Error::<T>::TooManyInclusionInherents);
			Included::<T>::set(Some(()));

			Self::enter_inner(data, FullCheck::Yes)
		}
	}
}

impl<T: Config> Pallet<T> {
	pub(crate) fn enter_inner(
		data: ParachainsInherentData<T::Header>,
		full_check: FullCheck,
	) -> DispatchResultWithPostInfo {
		let ParachainsInherentData {
			bitfields: mut signed_bitfields,
			mut backed_candidates,
			parent_header,
			mut disputes,
		} = data;
		#[cfg(feature = "runtime-metrics")]
		sp_io::init_tracing();

		log::debug!(
			target: LOG_TARGET,
			"[enter_inner] parent_header={:?} bitfields.len(): {}, backed_candidates.len(): {}, disputes.len(): {}",
			parent_header.hash(),
			signed_bitfields.len(),
			backed_candidates.len(),
			disputes.len()
		);

		// Check that the submitted parent header indeed corresponds to the previous block hash.
		let parent_hash = <frame_system::Pallet<T>>::parent_hash();
		ensure!(
			parent_header.hash().as_ref() == parent_hash.as_ref(),
			Error::<T>::InvalidParentHeader,
		);

		let now = <frame_system::Pallet<T>>::block_number();

		let mut candidates_weight = backed_candidates_weight::<T>(&backed_candidates);
		let mut bitfields_weight = signed_bitfields_weight::<T>(signed_bitfields.len());
		let disputes_weight = multi_dispute_statement_sets_weight::<T, _, _>(&disputes);

		let current_session = <shared::Pallet<T>>::session_index();

		let max_block_weight = <T as frame_system::Config>::BlockWeights::get().max_block;

		METRICS
			.on_before_filter((candidates_weight + bitfields_weight + disputes_weight).ref_time());

		T::DisputesHandler::assure_deduplicated_and_sorted(&mut disputes)
			.map_err(|_e| Error::<T>::DisputeStatementsUnsortedOrDuplicates)?;

		let (checked_disputes, total_consumed_weight) = {
			// Obtain config params..
			let config = <configuration::Pallet<T>>::config();
			let max_spam_slots = config.dispute_max_spam_slots;
			let post_conclusion_acceptance_period =
				config.dispute_post_conclusion_acceptance_period;

			let verify_dispute_sigs = if let FullCheck::Yes = full_check {
				VerifyDisputeSignatures::Yes
			} else {
				VerifyDisputeSignatures::Skip
			};

			// .. and prepare a helper closure.
			let dispute_set_validity_check = move |set| {
				T::DisputesHandler::filter_dispute_data(
					set,
					max_spam_slots,
					post_conclusion_acceptance_period,
					verify_dispute_sigs,
				)
			};

			// In case of an overweight block, consume up to the entire block weight
			// in disputes, since we will never process anything else, but invalidate
			// the block. It's still reasonable to protect against a massive amount of disputes.
			if candidates_weight
				.saturating_add(bitfields_weight)
				.saturating_add(disputes_weight)
				.any_gt(max_block_weight)
			{
				log::warn!("Overweight para inherent data reached the runtime {:?}", parent_hash);
				backed_candidates.clear();
				candidates_weight = Weight::zero();
				signed_bitfields.clear();
				bitfields_weight = Weight::zero();
			}

			let entropy = compute_entropy::<T>(parent_hash);
			let mut rng = rand_chacha::ChaChaRng::from_seed(entropy.into());

			let (checked_disputes, checked_disputes_weight) = limit_and_sanitize_disputes::<T, _>(
				disputes,
				&dispute_set_validity_check,
				max_block_weight,
				&mut rng,
			);
			(
				checked_disputes,
				checked_disputes_weight
					.saturating_add(candidates_weight)
					.saturating_add(bitfields_weight),
			)
		};

		let expected_bits = <scheduler::Pallet<T>>::availability_cores().len();

		// Handle disputes logic.
		let disputed_bitfield = {
			let new_current_dispute_sets: Vec<_> = checked_disputes
				.iter()
				.map(AsRef::as_ref)
				.filter(|s| s.session == current_session)
				.map(|s| (s.session, s.candidate_hash))
				.collect();

			// Note that `process_checked_multi_dispute_data` will iterate and import each
			// dispute; so the input here must be reasonably bounded,
			// which is guaranteed by the checks and weight limitation above.
			let _ =
				T::DisputesHandler::process_checked_multi_dispute_data(checked_disputes.clone())?;
			METRICS.on_disputes_imported(checked_disputes.len() as u64);

			if T::DisputesHandler::is_frozen() {
				// Relay chain freeze, at this point we will not include any parachain blocks.
				METRICS.on_relay_chain_freeze();

				// The relay chain we are currently on is invalid. Proceed no further on parachains.
				return Ok(Some(total_consumed_weight).into())
			}

			// Process the dispute sets of the current session.
			METRICS.on_current_session_disputes_processed(new_current_dispute_sets.len() as u64);

			let mut freed_disputed = if !new_current_dispute_sets.is_empty() {
				let concluded_invalid_disputes = new_current_dispute_sets
					.iter()
					.filter(|(session, candidate)| {
						T::DisputesHandler::concluded_invalid(*session, *candidate)
					})
					.map(|(_, candidate)| *candidate)
					.collect::<BTreeSet<CandidateHash>>();

				// Count invalid dispute sets.
				METRICS.on_disputes_concluded_invalid(concluded_invalid_disputes.len() as u64);

				let freed_disputed: Vec<_> =
					<inclusion::Pallet<T>>::collect_disputed(&concluded_invalid_disputes)
						.into_iter()
						.map(|core| (core, FreedReason::Concluded))
						.collect();

				freed_disputed
			} else {
				Vec::new()
			};

			// Create a bit index from the set of core indices where each index corresponds to
			// a core index that was freed due to a dispute.
			//
			// I.e. 010100 would indicate, the candidates on Core 1 and 3 would be disputed.
			let disputed_bitfield = create_disputed_bitfield(
				expected_bits,
				freed_disputed.iter().map(|(core_index, _)| core_index),
			);

			if !freed_disputed.is_empty() {
				// unstable sort is fine, because core indices are unique
				// i.e. the same candidate can't occupy 2 cores at once.
				freed_disputed.sort_unstable_by_key(|pair| pair.0); // sort by core index
				<scheduler::Pallet<T>>::free_cores(freed_disputed);
			}

			disputed_bitfield
		};

		METRICS.on_bitfields_processed(signed_bitfields.len() as u64);

		// Process new availability bitfields, yielding any availability cores whose
		// work has now concluded.
		let freed_concluded = <inclusion::Pallet<T>>::process_bitfields(
			expected_bits,
			signed_bitfields,
			disputed_bitfield,
			<scheduler::Pallet<T>>::core_para,
			full_check,
		)?;
		// any error in the previous function will cause an invalid block and not include
		// the `DisputeState` to be written to the storage, hence this is ok.
		set_scrapable_on_chain_disputes::<T>(current_session, checked_disputes.clone());

		// Inform the disputes module of all included candidates.
		for (_, candidate_hash) in &freed_concluded {
			T::DisputesHandler::note_included(current_session, *candidate_hash, now);
		}

		METRICS.on_candidates_included(freed_concluded.len() as u64);
		let freed = collect_all_freed_cores::<T, _>(freed_concluded.iter().cloned());

		<scheduler::Pallet<T>>::clear();
		<scheduler::Pallet<T>>::schedule(freed, now);

		METRICS.on_candidates_processed_total(backed_candidates.len() as u64);

		let scheduled = <scheduler::Pallet<T>>::scheduled();
		assure_sanity_backed_candidates::<T, _>(
			parent_hash,
			&backed_candidates,
			move |_candidate_index: usize, backed_candidate: &BackedCandidate<T::Hash>| -> bool {
				<T>::DisputesHandler::concluded_invalid(current_session, backed_candidate.hash())
				// `fn process_candidates` does the verification checks
			},
			&scheduled[..],
		)?;

		METRICS.on_candidates_sanitized(backed_candidates.len() as u64);

		// Process backed candidates according to scheduled cores.
		let parent_storage_root = *parent_header.state_root();
		let inclusion::ProcessedCandidates::<<T::Header as HeaderT>::Hash> {
			core_indices: occupied,
			candidate_receipt_with_backing_validator_indices,
		} = <inclusion::Pallet<T>>::process_candidates(
			parent_storage_root,
			backed_candidates,
			scheduled,
			<scheduler::Pallet<T>>::group_validators,
		)?;

		METRICS.on_disputes_included(checked_disputes.len() as u64);

		set_scrapable_on_chain_backings::<T>(
			current_session,
			candidate_receipt_with_backing_validator_indices,
		);

		// Note which of the scheduled cores were actually occupied by a backed candidate.
		<scheduler::Pallet<T>>::occupied(&occupied);

		// Give some time slice to dispatch pending upward messages.
		// this is max config.ump_service_total_weight
		let _ump_weight = <ump::Pallet<T>>::process_pending_upward_messages();

		METRICS.on_after_filter(total_consumed_weight.ref_time());

		Ok(Some(total_consumed_weight).into())
	}
}

impl<T: Config> Pallet<T> {
	/// Create the `ParachainsInherentData` that gets passed to [`Self::enter`] in [`Self::create_inherent`].
	/// This code is pulled out of [`Self::create_inherent`] so it can be unit tested.
	fn create_inherent_inner(data: &InherentData) -> Option<ParachainsInherentData<T::Header>> {
		let ParachainsInherentData::<T::Header> {
			bitfields,
			backed_candidates,
			mut disputes,
			parent_header,
		} = match data.get_data(&Self::INHERENT_IDENTIFIER) {
			Ok(Some(d)) => d,
			Ok(None) => return None,
			Err(_) => {
				log::warn!(target: LOG_TARGET, "ParachainsInherentData failed to decode");
				return None
			},
		};

		log::debug!(
			target: LOG_TARGET,
			"[create_inherent_inner] bitfields.len(): {}, backed_candidates.len(): {}, disputes.len() {}",
			bitfields.len(),
			backed_candidates.len(),
			disputes.len()
		);

		let parent_hash = <frame_system::Pallet<T>>::parent_hash();

		if parent_hash != parent_header.hash() {
			log::warn!(
				target: LOG_TARGET,
				"ParachainsInherentData references a different parent header hash than frame"
			);
			return None
		}

		let current_session = <shared::Pallet<T>>::session_index();
		let expected_bits = <scheduler::Pallet<T>>::availability_cores().len();
		let validator_public = shared::Pallet::<T>::active_validator_keys();
		let max_block_weight = <T as frame_system::Config>::BlockWeights::get().max_block;

		let entropy = compute_entropy::<T>(parent_hash);
		let mut rng = rand_chacha::ChaChaRng::from_seed(entropy.into());

		// Filter out duplicates and continue.
		if let Err(_) = T::DisputesHandler::deduplicate_and_sort_dispute_data(&mut disputes) {
			log::debug!(target: LOG_TARGET, "Found duplicate statement sets, retaining the first");
		}

		let config = <configuration::Pallet<T>>::config();
		let max_spam_slots = config.dispute_max_spam_slots;
		let post_conclusion_acceptance_period = config.dispute_post_conclusion_acceptance_period;

		// TODO: Better if we can convert this to `with_transactional` and handle an error if
		// too many transactional layers are spawned.
		let (
			mut backed_candidates,
			mut bitfields,
			checked_disputes_sets,
			checked_disputes_sets_consumed_weight,
		) = frame_support::storage::with_transaction_unchecked(|| {
			let dispute_statement_set_valid = move |set: DisputeStatementSet| {
				T::DisputesHandler::filter_dispute_data(
					set,
					max_spam_slots,
					post_conclusion_acceptance_period,
					// `DisputeCoordinator` on the node side only forwards
					// valid dispute statement sets and hence this does not
					// need to be checked.
					VerifyDisputeSignatures::Skip,
				)
			};

			// Limit the disputes first, since the following statements depend on the votes include here.
			let (checked_disputes_sets, checked_disputes_sets_consumed_weight) =
				limit_and_sanitize_disputes::<T, _>(
					disputes,
					dispute_statement_set_valid,
					max_block_weight,
					&mut rng,
				);

			// we don't care about fresh or not disputes
			// this writes them to storage, so let's query it via those means
			// if this fails for whatever reason, that's ok
			let _ = T::DisputesHandler::process_checked_multi_dispute_data(
				checked_disputes_sets.clone(),
			)
			.map_err(|e| {
				log::warn!(target: LOG_TARGET, "MultiDisputesData failed to update: {:?}", e);
				e
			});

			// Contains the disputes that are concluded in the current session only,
			// since these are the only ones that are relevant for the occupied cores
			// and lightens the load on `collect_disputed` significantly.
			// Cores can't be occupied with candidates of the previous sessions, and only
			// things with new votes can have just concluded. We only need to collect
			// cores with disputes that conclude just now, because disputes that
			// concluded longer ago have already had any corresponding cores cleaned up.
			let current_concluded_invalid_disputes = checked_disputes_sets
				.iter()
				.map(AsRef::as_ref)
				.filter(|dss| dss.session == current_session)
				.map(|dss| (dss.session, dss.candidate_hash))
				.filter(|(session, candidate)| {
					<T>::DisputesHandler::concluded_invalid(*session, *candidate)
				})
				.map(|(_session, candidate)| candidate)
				.collect::<BTreeSet<CandidateHash>>();

			// All concluded invalid disputes, that are relevant for the set of candidates
			// the inherent provided.
			let concluded_invalid_disputes = backed_candidates
				.iter()
				.map(|backed_candidate| backed_candidate.hash())
				.filter(|candidate| {
					<T>::DisputesHandler::concluded_invalid(current_session, *candidate)
				})
				.collect::<BTreeSet<CandidateHash>>();

			let mut freed_disputed: Vec<_> =
				<inclusion::Pallet<T>>::collect_disputed(&current_concluded_invalid_disputes)
					.into_iter()
					.map(|core| (core, FreedReason::Concluded))
					.collect();

			let disputed_bitfield =
				create_disputed_bitfield(expected_bits, freed_disputed.iter().map(|(x, _)| x));

			if !freed_disputed.is_empty() {
				// unstable sort is fine, because core indices are unique
				// i.e. the same candidate can't occupy 2 cores at once.
				freed_disputed.sort_unstable_by_key(|pair| pair.0); // sort by core index
				<scheduler::Pallet<T>>::free_cores(freed_disputed.clone());
			}

			// The following 3 calls are equiv to a call to `process_bitfields`
			// but we can retain access to `bitfields`.
			let bitfields = sanitize_bitfields::<T>(
				bitfields,
				disputed_bitfield,
				expected_bits,
				parent_hash,
				current_session,
				&validator_public[..],
				FullCheck::Yes,
			);

			let freed_concluded =
				<inclusion::Pallet<T>>::update_pending_availability_and_get_freed_cores::<_>(
					expected_bits,
					&validator_public[..],
					bitfields.clone(),
					<scheduler::Pallet<T>>::core_para,
					false,
				);

			let freed = collect_all_freed_cores::<T, _>(freed_concluded.iter().cloned());

			<scheduler::Pallet<T>>::clear();
			let now = <frame_system::Pallet<T>>::block_number();
			<scheduler::Pallet<T>>::schedule(freed, now);

			let scheduled = <scheduler::Pallet<T>>::scheduled();

			let relay_parent_number = now - One::one();
			let parent_storage_root = *parent_header.state_root();

			let check_ctx = CandidateCheckContext::<T>::new(now, relay_parent_number);
			let backed_candidates = sanitize_backed_candidates::<T, _>(
				parent_hash,
				backed_candidates,
				move |candidate_idx: usize,
				      backed_candidate: &BackedCandidate<<T as frame_system::Config>::Hash>|
				      -> bool {
					// never include a concluded-invalid candidate
					concluded_invalid_disputes.contains(&backed_candidate.hash()) ||
							// Instead of checking the candidates with code upgrades twice
							// move the checking up here and skip it in the training wheels fallback.
							// That way we avoid possible duplicate checks while assuring all
							// backed candidates fine to pass on.
							check_ctx
								.verify_backed_candidate(parent_hash, parent_storage_root, candidate_idx, backed_candidate)
								.is_err()
				},
				&scheduled[..],
			);

			frame_support::storage::TransactionOutcome::Rollback((
				// filtered backed candidates
				backed_candidates,
				// filtered bitfields
				bitfields,
				// checked disputes sets
				checked_disputes_sets,
				checked_disputes_sets_consumed_weight,
			))
		});

		// Assure the maximum block weight is adhered, by limiting bitfields and backed
		// candidates. Dispute statement sets were already limited before.
		let actual_weight = apply_weight_limit::<T>(
			&mut backed_candidates,
			&mut bitfields,
			max_block_weight.saturating_sub(checked_disputes_sets_consumed_weight),
			&mut rng,
		);

		if actual_weight.any_gt(max_block_weight) {
			log::warn!(target: LOG_TARGET, "Post weight limiting weight is still too large.");
		}

		let disputes = checked_disputes_sets
			.into_iter()
			.map(|checked| checked.into())
			.collect::<Vec<_>>();

		Some(ParachainsInherentData::<T::Header> {
			bitfields,
			backed_candidates,
			disputes,
			parent_header,
		})
	}
}

/// Derive a bitfield from dispute
pub(super) fn create_disputed_bitfield<'a, I>(
	expected_bits: usize,
	freed_cores: I,
) -> DisputedBitfield
where
	I: 'a + IntoIterator<Item = &'a CoreIndex>,
{
	let mut bitvec = BitVec::repeat(false, expected_bits);
	for core_idx in freed_cores {
		let core_idx = core_idx.0 as usize;
		if core_idx < expected_bits {
			bitvec.set(core_idx, true);
		}
	}
	DisputedBitfield::from(bitvec)
}

/// Select a random subset, with preference for certain indices.
///
/// Adds random items to the set until all candidates
/// are tried or the remaining weight is depleted.
///
/// Returns the weight of all selected items from `selectables`
/// as well as their indices in ascending order.
fn random_sel<X, F: Fn(&X) -> Weight>(
	rng: &mut rand_chacha::ChaChaRng,
	selectables: Vec<X>,
	mut preferred_indices: Vec<usize>,
	weight_fn: F,
	weight_limit: Weight,
) -> (Weight, Vec<usize>) {
	if selectables.is_empty() {
		return (Weight::zero(), Vec::new())
	}
	// all indices that are not part of the preferred set
	let mut indices = (0..selectables.len())
		.into_iter()
		.filter(|idx| !preferred_indices.contains(idx))
		.collect::<Vec<_>>();
	let mut picked_indices = Vec::with_capacity(selectables.len().saturating_sub(1));

	let mut weight_acc = Weight::zero();

	preferred_indices.shuffle(rng);
	for preferred_idx in preferred_indices {
		// preferred indices originate from outside
		if let Some(item) = selectables.get(preferred_idx) {
			let updated = weight_acc.saturating_add(weight_fn(item));
			if updated.any_gt(weight_limit) {
				continue
			}
			weight_acc = updated;
			picked_indices.push(preferred_idx);
		}
	}

	indices.shuffle(rng);
	for idx in indices {
		let item = &selectables[idx];
		let updated = weight_acc.saturating_add(weight_fn(item));

		if updated.any_gt(weight_limit) {
			continue
		}
		weight_acc = updated;

		picked_indices.push(idx);
	}

	// sorting indices, so the ordering is retained
	// unstable sorting is fine, since there are no duplicates in indices
	// and even if there were, they don't have an identity
	picked_indices.sort_unstable();
	(weight_acc, picked_indices)
}

/// Considers an upper threshold that the inherent data must not exceed.
///
/// If there is sufficient space, all bitfields and all candidates
/// will be included.
///
/// Otherwise tries to include all disputes, and then tries to fill the remaining space with bitfields and then candidates.
///
/// The selection process is random. For candidates, there is an exception for code upgrades as they are preferred.
/// And for disputes, local and older disputes are preferred (see `limit_and_sanitize_disputes`).
/// for backed candidates, since with a increasing number of parachains their chances of
/// inclusion become slim. All backed candidates  are checked beforehands in `fn create_inherent_inner`
/// which guarantees sanity.
///
/// Assumes disputes are already filtered by the time this is called.
///
/// Returns the total weight consumed by `bitfields` and `candidates`.
fn apply_weight_limit<T: Config + inclusion::Config>(
	candidates: &mut Vec<BackedCandidate<<T>::Hash>>,
	bitfields: &mut UncheckedSignedAvailabilityBitfields,
	max_consumable_weight: Weight,
	rng: &mut rand_chacha::ChaChaRng,
) -> Weight {
	let total_candidates_weight = backed_candidates_weight::<T>(candidates.as_slice());

	let total_bitfields_weight = signed_bitfields_weight::<T>(bitfields.len());

	let total = total_bitfields_weight.saturating_add(total_candidates_weight);

	// candidates + bitfields fit into the block
	if max_consumable_weight.all_gte(total) {
		return total
	}

	// Prefer code upgrades, they tend to be large and hence stand no chance to be picked
	// late while maintaining the weight bounds.
	let preferred_indices = candidates
		.iter()
		.enumerate()
		.filter_map(|(idx, candidate)| {
			candidate.candidate.commitments.new_validation_code.as_ref().map(|_code| idx)
		})
		.collect::<Vec<usize>>();

	// There is weight remaining to be consumed by a subset of candidates
	// which are going to be picked now.
	if let Some(max_consumable_by_candidates) =
		max_consumable_weight.checked_sub(&total_bitfields_weight)
	{
		let (acc_candidate_weight, indices) =
			random_sel::<BackedCandidate<<T as frame_system::Config>::Hash>, _>(
				rng,
				candidates.clone(),
				preferred_indices,
				|c| backed_candidate_weight::<T>(c),
				max_consumable_by_candidates,
			);
		candidates.indexed_retain(|idx, _backed_candidate| indices.binary_search(&idx).is_ok());
		// pick all bitfields, and
		// fill the remaining space with candidates
		let total_consumed = acc_candidate_weight.saturating_add(total_bitfields_weight);

		return total_consumed
	}

	candidates.clear();

	// insufficient space for even the bitfields alone, so only try to fit as many of those
	// into the block and skip the candidates entirely
	let (total_consumed, indices) = random_sel::<UncheckedSignedAvailabilityBitfield, _>(
		rng,
		bitfields.clone(),
		vec![],
		|_| <<T as Config>::WeightInfo as WeightInfo>::enter_bitfields(),
		max_consumable_weight,
	);

	bitfields.indexed_retain(|idx, _bitfield| indices.binary_search(&idx).is_ok());

	total_consumed
}

/// Filter bitfields based on freed core indices, validity, and other sanity checks.
///
/// Do sanity checks on the bitfields:
///
///  1. no more than one bitfield per validator
///  2. bitfields are ascending by validator index.
///  3. each bitfield has exactly `expected_bits`
///  4. signature is valid
///  5. remove any disputed core indices
///
/// If any of those is not passed, the bitfield is dropped.
///
/// While this function technically returns a set of unchecked bitfields,
/// they were actually checked and filtered to allow using it in both
/// cases, as `filtering` and `checking` stage.
///
/// `full_check` determines if validator signatures are checked. If `::Yes`,
/// bitfields that have an invalid signature will be filtered out.
pub(crate) fn sanitize_bitfields<T: crate::inclusion::Config>(
	unchecked_bitfields: UncheckedSignedAvailabilityBitfields,
	disputed_bitfield: DisputedBitfield,
	expected_bits: usize,
	parent_hash: T::Hash,
	session_index: SessionIndex,
	validators: &[ValidatorId],
	full_check: FullCheck,
) -> UncheckedSignedAvailabilityBitfields {
	let mut bitfields = Vec::with_capacity(unchecked_bitfields.len());

	let mut last_index: Option<ValidatorIndex> = None;

	if disputed_bitfield.0.len() != expected_bits {
		// This is a system logic error that should never occur, but we want to handle it gracefully
		// so we just drop all bitfields
		log::error!(target: LOG_TARGET, "BUG: disputed_bitfield != expected_bits");
		return vec![]
	}

	let all_zeros = BitVec::<u8, bitvec::order::Lsb0>::repeat(false, expected_bits);
	let signing_context = SigningContext { parent_hash, session_index };
	for unchecked_bitfield in unchecked_bitfields {
		// Find and skip invalid bitfields.
		if unchecked_bitfield.unchecked_payload().0.len() != expected_bits {
			log::trace!(
				target: LOG_TARGET,
				"[{:?}] bad bitfield length: {} != {:?}",
				full_check,
				unchecked_bitfield.unchecked_payload().0.len(),
				expected_bits,
			);
			continue
		}

		if unchecked_bitfield.unchecked_payload().0.clone() & disputed_bitfield.0.clone() !=
			all_zeros
		{
			log::trace!(
				target: LOG_TARGET,
				"[{:?}] bitfield contains disputed cores: {:?}",
				full_check,
				unchecked_bitfield.unchecked_payload().0.clone() & disputed_bitfield.0.clone()
			);
			continue
		}

		let validator_index = unchecked_bitfield.unchecked_validator_index();

		if !last_index.map_or(true, |last_index: ValidatorIndex| last_index < validator_index) {