diff --git a/polkadot/runtime/parachains/src/scheduler.rs b/polkadot/runtime/parachains/src/scheduler.rs
index 329df3a8a9deb3c06a337d9cf3e4e43e55c100ae..9c111c2d28e79c9df6c3b3efc9e8b36fe5e4b1e9 100644
--- a/polkadot/runtime/parachains/src/scheduler.rs
+++ b/polkadot/runtime/parachains/src/scheduler.rs
@@ -45,7 +45,8 @@ use frame_support::{pallet_prelude::*, traits::Defensive};
use frame_system::pallet_prelude::BlockNumberFor;
pub use polkadot_core_primitives::v2::BlockNumber;
use polkadot_primitives::{
- CoreIndex, GroupIndex, GroupRotationInfo, Id as ParaId, ScheduledCore, ValidatorIndex,
+ CoreIndex, GroupIndex, GroupRotationInfo, Id as ParaId, ScheduledCore, SchedulerParams,
+ ValidatorIndex,
};
use sp_runtime::traits::One;
@@ -131,7 +132,7 @@ impl<T: Config> Pallet<T> {
pub(crate) fn initializer_on_new_session(
notification: &SessionChangeNotification<BlockNumberFor<T>>,
) {
- let SessionChangeNotification { validators, new_config, prev_config, .. } = notification;
+ let SessionChangeNotification { validators, new_config, .. } = notification;
let config = new_config;
let assigner_cores = config.scheduler_params.num_cores;
@@ -186,7 +187,7 @@ impl<T: Config> Pallet<T> {
}
// Resize and populate claim queue.
- Self::maybe_resize_claim_queue(prev_config.scheduler_params.num_cores, assigner_cores);
+ Self::maybe_resize_claim_queue();
Self::populate_claim_queue_after_session_change();
let now = frame_system::Pallet::<T>::block_number() + One::one();
@@ -203,6 +204,12 @@ impl<T: Config> Pallet<T> {
ValidatorGroups::<T>::decode_len().unwrap_or(0)
}
+ /// Expected claim queue len. Can be different than the real length if for example we don't have
+ /// assignments for a core.
+ fn expected_claim_queue_len(config: &SchedulerParams<BlockNumberFor<T>>) -> u32 {
+ core::cmp::min(config.num_cores, Self::num_availability_cores() as u32)
+ }
+
/// Get the group assigned to a specific core by index at the current block number. Result
/// undefined if the core index is unknown or the block number is less than the session start
/// index.
@@ -325,11 +332,11 @@ impl<T: Config> Pallet<T> {
/// and fill the queue from the assignment provider.
pub(crate) fn advance_claim_queue(except_for: &BTreeSet<CoreIndex>) {
let config = configuration::ActiveConfig::<T>::get();
- let num_assigner_cores = config.scheduler_params.num_cores;
+ let expected_claim_queue_len = Self::expected_claim_queue_len(&config.scheduler_params);
// Extra sanity, config should already never be smaller than 1:
let n_lookahead = config.scheduler_params.lookahead.max(1);
- for core_idx in 0..num_assigner_cores {
+ for core_idx in 0..expected_claim_queue_len {
let core_idx = CoreIndex::from(core_idx);
if !except_for.contains(&core_idx) {
@@ -345,13 +352,16 @@ impl<T: Config> Pallet<T> {
}
// on new session
- fn maybe_resize_claim_queue(old_core_count: u32, new_core_count: u32) {
- if new_core_count < old_core_count {
+ fn maybe_resize_claim_queue() {
+ let cq = ClaimQueue::<T>::get();
+ let Some((old_max_core, _)) = cq.last_key_value() else { return };
+ let config = configuration::ActiveConfig::<T>::get();
+ let new_core_count = Self::expected_claim_queue_len(&config.scheduler_params);
+
+ if new_core_count < (old_max_core.0 + 1) {
ClaimQueue::<T>::mutate(|cq| {
- let to_remove: Vec<_> = cq
- .range(CoreIndex(new_core_count)..CoreIndex(old_core_count))
- .map(|(k, _)| *k)
- .collect();
+ let to_remove: Vec<_> =
+ cq.range(CoreIndex(new_core_count)..=*old_max_core).map(|(k, _)| *k).collect();
for key in to_remove {
if let Some(dropped_assignments) = cq.remove(&key) {
Self::push_back_to_assignment_provider(dropped_assignments.into_iter());
@@ -367,9 +377,9 @@ impl<T: Config> Pallet<T> {
let config = configuration::ActiveConfig::<T>::get();
// Extra sanity, config should already never be smaller than 1:
let n_lookahead = config.scheduler_params.lookahead.max(1);
- let new_core_count = config.scheduler_params.num_cores;
+ let expected_claim_queue_len = Self::expected_claim_queue_len(&config.scheduler_params);
- for core_idx in 0..new_core_count {
+ for core_idx in 0..expected_claim_queue_len {
let core_idx = CoreIndex::from(core_idx);
Self::fill_claim_queue(core_idx, n_lookahead);
}
diff --git a/polkadot/runtime/parachains/src/scheduler/tests.rs b/polkadot/runtime/parachains/src/scheduler/tests.rs
index 5be7e084f3bca3588301fa3970d0873fe83ba687..431562c6e6fb78f758e99df366323bcd2a886fb9 100644
--- a/polkadot/runtime/parachains/src/scheduler/tests.rs
+++ b/polkadot/runtime/parachains/src/scheduler/tests.rs
@@ -726,8 +726,26 @@ fn session_change_decreasing_number_of_cores() {
[assignment_b.clone()].into_iter().collect::<VecDeque<_>>()
);
- // No more assignments now.
Scheduler::advance_claim_queue(&Default::default());
+ // No more assignments now.
+ assert_eq!(Scheduler::claim_queue_len(), 0);
+
+ // Retain number of cores to 1 but remove all validator groups. The claim queue length
+ // should be the minimum of these two.
+
+ // Add an assignment.
+ MockAssigner::add_test_assignment(assignment_b.clone());
+
+ run_to_block(4, |number| match number {
+ 4 => Some(SessionChangeNotification {
+ new_config: new_config.clone(),
+ prev_config: new_config.clone(),
+ validators: vec![],
+ ..Default::default()
+ }),
+ _ => None,
+ });
+
assert_eq!(Scheduler::claim_queue_len(), 0);
});
}
diff --git a/prdoc/pr_6257.prdoc b/prdoc/pr_6257.prdoc
new file mode 100644
index 0000000000000000000000000000000000000000..45f9810108ef70f97da7b703d9180dbaa229ab94
--- /dev/null
+++ b/prdoc/pr_6257.prdoc
@@ -0,0 +1,10 @@
+title: 'fix claim queue size when validator groups count is smaller'
+doc:
+- audience: Runtime Dev
+ description: 'Fixes a bug introduced in https://github.com/paritytech/polkadot-sdk/pull/5461, where the claim queue
+ would contain entries even if the validator groups storage is empty (which happens during the first session).
+ This PR sets the claim queue core count to be the minimum between the num_cores param and the number of validator groups.'
+
+crates:
+- name: polkadot-runtime-parachains
+ bump: patch