Update bridges subtree (#1392)

* Move the bridges subtree under root * Squashed 'bridges/' changes from 277f0d5496..e50398d1c5 e50398d1c5 bridges subtree fixes (#2528) 99af07522d Markdown linter (#1309) (#2526) 733ff0fe7a `polkadot-staging` branch: Use polkadot-sdk dependencies (#2524) e8a59f141e Fix benchmark with new XCM::V3 `MAX_INSTRUCTIONS_TO_DECODE` (#2514) 62b185de15 Backport `polkadot-sdk` changes to `polkadot-staging` (#2518) d9658f4d5b Fix equivocation detection containers startup (#2516) (#2517) d65db28a8f Backport: building images from locally built binaries (#2513) 5fdbaf45f6 Start the equivocation detection loop from the complex relayer (#2507) (#2512) 7fbb67de46 Backport: Implement basic equivocations detection loop (#2375) cb7efe245c Manually update deps in polkadot staging (#2371) d17981fc33 #2351 to polkadot-staging (#2359) git-subtree-dir: bridges git-subtree-split: e50398d1c594e4e96df70b0bd376e565d17e8558 * Reapply diener workspacify * Fix Cargo.toml * Fix test * Adjustments
2026-05-31 20:21:03 +00:00 · 2023-09-11 11:47:45 +03:00
parent fbf5a814f0
commit 142a11ad95
134 changed files with 397 additions and 342 deletions
@@ -0,0 +1,201 @@
+// Copyright (C) Parity Technologies (UK) Ltd.
+// This file is part of Parity Bridges Common.
+
+// Parity Bridges Common 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.
+
+// Parity Bridges Common 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 Parity Bridges Common.  If not, see <http://www.gnu.org/licenses/>.
+
+//! Bridge transaction priority calculator.
+//!
+//! We want to prioritize message delivery transactions with more messages over
+//! transactions with less messages. That's because we reject delivery transactions
+//! if it contains already delivered message. And if some transaction delivers
+//! single message with nonce `N`, then the transaction with nonces `N..=N+100` will
+//! be rejected. This can lower bridge throughput down to one message per block.
+
+use bp_messages::MessageNonce;
+use frame_support::traits::Get;
+use sp_runtime::transaction_validity::TransactionPriority;
+
+// reexport everything from `integrity_tests` module
+pub use integrity_tests::*;
+
+/// Compute priority boost for message delivery transaction that delivers
+/// given number of messages.
+pub fn compute_priority_boost<PriorityBoostPerMessage>(
+	messages: MessageNonce,
+) -> TransactionPriority
+where
+	PriorityBoostPerMessage: Get<TransactionPriority>,
+{
+	// we don't want any boost for transaction with single message => minus one
+	PriorityBoostPerMessage::get().saturating_mul(messages - 1)
+}
+
+#[cfg(not(feature = "integrity-test"))]
+mod integrity_tests {}
+
+#[cfg(feature = "integrity-test")]
+mod integrity_tests {
+	use super::compute_priority_boost;
+
+	use bp_messages::MessageNonce;
+	use bp_runtime::PreComputedSize;
+	use frame_support::{
+		dispatch::{DispatchClass, DispatchInfo, Pays, PostDispatchInfo},
+		traits::Get,
+	};
+	use pallet_bridge_messages::WeightInfoExt;
+	use pallet_transaction_payment::OnChargeTransaction;
+	use sp_runtime::{
+		traits::{Dispatchable, UniqueSaturatedInto, Zero},
+		transaction_validity::TransactionPriority,
+		FixedPointOperand, SaturatedConversion, Saturating,
+	};
+
+	type BalanceOf<T> =
+		<<T as pallet_transaction_payment::Config>::OnChargeTransaction as OnChargeTransaction<
+			T,
+		>>::Balance;
+
+	/// Ensures that the value of `PriorityBoostPerMessage` matches the value of
+	/// `tip_boost_per_message`.
+	///
+	/// We want two transactions, `TX1` with `N` messages and `TX2` with `N+1` messages, have almost
+	/// the same priority if we'll add `tip_boost_per_message` tip to the `TX1`. We want to be sure
+	/// that if we add plain `PriorityBoostPerMessage` priority to `TX1`, the priority will be close
+	/// to `TX2` as well.
+	pub fn ensure_priority_boost_is_sane<Runtime, MessagesInstance, PriorityBoostPerMessage>(
+		tip_boost_per_message: BalanceOf<Runtime>,
+	) where
+		Runtime:
+			pallet_transaction_payment::Config + pallet_bridge_messages::Config<MessagesInstance>,
+		MessagesInstance: 'static,
+		PriorityBoostPerMessage: Get<TransactionPriority>,
+		Runtime::RuntimeCall: Dispatchable<Info = DispatchInfo, PostInfo = PostDispatchInfo>,
+		BalanceOf<Runtime>: Send + Sync + FixedPointOperand,
+	{
+		let priority_boost_per_message = PriorityBoostPerMessage::get();
+		let maximal_messages_in_delivery_transaction =
+			Runtime::MaxUnconfirmedMessagesAtInboundLane::get();
+		for messages in 1..=maximal_messages_in_delivery_transaction {
+			let base_priority = estimate_message_delivery_transaction_priority::<
+				Runtime,
+				MessagesInstance,
+			>(messages, Zero::zero());
+			let priority_boost = compute_priority_boost::<PriorityBoostPerMessage>(messages);
+			let priority_with_boost = base_priority + priority_boost;
+
+			let tip = tip_boost_per_message.saturating_mul((messages - 1).unique_saturated_into());
+			let priority_with_tip =
+				estimate_message_delivery_transaction_priority::<Runtime, MessagesInstance>(1, tip);
+
+			const ERROR_MARGIN: TransactionPriority = 5; // 5%
+			if priority_with_boost.abs_diff(priority_with_tip).saturating_mul(100) /
+				priority_with_tip >
+				ERROR_MARGIN
+			{
+				panic!(
+					"The PriorityBoostPerMessage value ({}) must be fixed to: {}",
+					priority_boost_per_message,
+					compute_priority_boost_per_message::<Runtime, MessagesInstance>(
+						tip_boost_per_message
+					),
+				);
+			}
+		}
+	}
+
+	/// Compute priority boost that we give to message delivery transaction for additional message.
+	#[cfg(feature = "integrity-test")]
+	fn compute_priority_boost_per_message<Runtime, MessagesInstance>(
+		tip_boost_per_message: BalanceOf<Runtime>,
+	) -> TransactionPriority
+	where
+		Runtime:
+			pallet_transaction_payment::Config + pallet_bridge_messages::Config<MessagesInstance>,
+		MessagesInstance: 'static,
+		Runtime::RuntimeCall: Dispatchable<Info = DispatchInfo, PostInfo = PostDispatchInfo>,
+		BalanceOf<Runtime>: Send + Sync + FixedPointOperand,
+	{
+		// esimate priority of transaction that delivers one message and has large tip
+		let maximal_messages_in_delivery_transaction =
+			Runtime::MaxUnconfirmedMessagesAtInboundLane::get();
+		let small_with_tip_priority =
+			estimate_message_delivery_transaction_priority::<Runtime, MessagesInstance>(
+				1,
+				tip_boost_per_message
+					.saturating_mul(maximal_messages_in_delivery_transaction.saturated_into()),
+			);
+		// estimate priority of transaction that delivers maximal number of messages, but has no tip
+		let large_without_tip_priority = estimate_message_delivery_transaction_priority::<
+			Runtime,
+			MessagesInstance,
+		>(maximal_messages_in_delivery_transaction, Zero::zero());
+
+		small_with_tip_priority
+			.saturating_sub(large_without_tip_priority)
+			.saturating_div(maximal_messages_in_delivery_transaction - 1)
+	}
+
+	/// Estimate message delivery transaction priority.
+	#[cfg(feature = "integrity-test")]
+	fn estimate_message_delivery_transaction_priority<Runtime, MessagesInstance>(
+		messages: MessageNonce,
+		tip: BalanceOf<Runtime>,
+	) -> TransactionPriority
+	where
+		Runtime:
+			pallet_transaction_payment::Config + pallet_bridge_messages::Config<MessagesInstance>,
+		MessagesInstance: 'static,
+		Runtime::RuntimeCall: Dispatchable<Info = DispatchInfo, PostInfo = PostDispatchInfo>,
+		BalanceOf<Runtime>: Send + Sync + FixedPointOperand,
+	{
+		// just an estimation of extra transaction bytes that are added to every transaction
+		// (including signature, signed extensions extra and etc + in our case it includes
+		// all call arguments extept the proof itself)
+		let base_tx_size = 512;
+		// let's say we are relaying similar small messages and for every message we add more trie
+		// nodes to the proof (x0.5 because we expect some nodes to be reused)
+		let estimated_message_size = 512;
+		// let's say all our messages have the same dispatch weight
+		let estimated_message_dispatch_weight =
+			Runtime::WeightInfo::message_dispatch_weight(estimated_message_size);
+		// messages proof argument size is (for every message) messages size + some additional
+		// trie nodes. Some of them are reused by different messages, so let's take 2/3 of default
+		// "overhead" constant
+		let messages_proof_size = Runtime::WeightInfo::expected_extra_storage_proof_size()
+			.saturating_mul(2)
+			.saturating_div(3)
+			.saturating_add(estimated_message_size)
+			.saturating_mul(messages as _);
+
+		// finally we are able to estimate transaction size and weight
+		let transaction_size = base_tx_size.saturating_add(messages_proof_size);
+		let transaction_weight = Runtime::WeightInfo::receive_messages_proof_weight(
+			&PreComputedSize(transaction_size as _),
+			messages as _,
+			estimated_message_dispatch_weight.saturating_mul(messages),
+		);
+
+		pallet_transaction_payment::ChargeTransactionPayment::<Runtime>::get_priority(
+			&DispatchInfo {
+				weight: transaction_weight,
+				class: DispatchClass::Normal,
+				pays_fee: Pays::Yes,
+			},
+			transaction_size as _,
+			tip,
+			Zero::zero(),
+		)
+	}
+}