Files
pezkuwi-subxt/polkadot/runtime/parachains/src/dmp.rs
T
Keith Yeung 3dece311be Introduce XcmFeesToAccount fee manager (#1234)
Combination of paritytech/polkadot#7005, its addon PR
paritytech/polkadot#7585 and its companion paritytech/cumulus#2433.

This PR introduces a new XcmFeesToAccount struct which implements the
`FeeManager` trait, and assigns this struct as the `FeeManager` in the
XCM config for all runtimes.

The struct simply deposits all fees handled by the XCM executor to a
specified account. In all runtimes, the specified account is configured
as the treasury account.

XCM __delivery__ fees are now being introduced (unless the root origin
is sending a message to a system parachain on behalf of the originating
chain).

# Note for reviewers

Most file changes are tests that had to be modified to account for the
new fees.
Main changes are in:
- cumulus/pallets/xcmp-queue/src/lib.rs <- To make it track the delivery
fees exponential factor
- polkadot/xcm/xcm-builder/src/fee_handling.rs <- Added. Has the
FeeManager implementation
- All runtime xcm_config files <- To add the FeeManager to the XCM
configuration

# Important note

After this change, instructions that create and send a new XCM (Query*,
Report*, ExportMessage, InitiateReserveWithdraw, InitiateTeleport,
DepositReserveAsset, TransferReserveAsset, LockAsset and RequestUnlock)
will require the corresponding origin account in the origin register to
pay for transport delivery fees, and the onward message will fail to be
sent if the origin account does not have the required amount. This
delivery fee is on top of what we already collect as tx fees in
pallet-xcm and XCM BuyExecution fees!

Wallet UIs that want to expose the new delivery fee can do so using the
formula:

```
delivery_fee_factor * (base_fee + encoded_msg_len * per_byte_fee)
```

where the delivery fee factor can be obtained from the corresponding
pallet based on which transport you are using (UMP, HRMP or bridges),
the base fee is a constant, the encoded message length from the message
itself and the per byte fee is the same as the configured per byte fee
for txs (i.e. `TransactionByteFee`).

---------

Co-authored-by: Branislav Kontur <bkontur@gmail.com>
Co-authored-by: joe petrowski <25483142+joepetrowski@users.noreply.github.com>
Co-authored-by: Giles Cope <gilescope@gmail.com>
Co-authored-by: command-bot <>
Co-authored-by: Francisco Aguirre <franciscoaguirreperez@gmail.com>
Co-authored-by: Liam Aharon <liam.aharon@hotmail.com>
Co-authored-by: Kian Paimani <5588131+kianenigma@users.noreply.github.com>
2023-10-18 17:22:25 +02:00

362 lines
12 KiB
Rust

// Copyright (C) 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/>.
//! To prevent Out of Memory errors on the `DownwardMessageQueue`, an
//! exponential fee factor (`DeliveryFeeFactor`) is set. The fee factor
//! increments exponentially after the number of messages in the
//! `DownwardMessageQueue` passes a threshold. This threshold is set as:
//!
//! ```ignore
//! // Maximum max sized messages that can be send to
//! // the DownwardMessageQueue before it runs out of memory
//! max_messages = MAX_POSSIBLE_ALLOCATION / max_downward_message_size
//! threshold = max_messages / THRESHOLD_FACTOR
//! ```
//! Based on the THRESHOLD_FACTOR, the threshold is set as a fraction of the
//! total messages. The `DeliveryFeeFactor` increases for a message over the
//! threshold by:
//!
//! `DeliveryFeeFactor = DeliveryFeeFactor *
//! (EXPONENTIAL_FEE_BASE + MESSAGE_SIZE_FEE_BASE * encoded_message_size_in_KB)`
//!
//! And decreases when the number of messages in the `DownwardMessageQueue` fall
//! below the threshold by:
//!
//! `DeliveryFeeFactor = DeliveryFeeFactor / EXPONENTIAL_FEE_BASE`
//!
//! As an extra defensive measure, a `max_messages` hard
//! limit is set to the number of messages in the DownwardMessageQueue. Messages
//! that would increase the number of messages in the queue above this hard
//! limit are dropped.
use crate::{
configuration::{self, HostConfiguration},
initializer, FeeTracker,
};
use frame_support::pallet_prelude::*;
use frame_system::pallet_prelude::BlockNumberFor;
use primitives::{DownwardMessage, Hash, Id as ParaId, InboundDownwardMessage};
use sp_core::MAX_POSSIBLE_ALLOCATION;
use sp_runtime::{
traits::{BlakeTwo256, Hash as HashT, SaturatedConversion},
FixedU128, Saturating,
};
use sp_std::{fmt, prelude::*};
use xcm::latest::SendError;
pub use pallet::*;
#[cfg(test)]
mod tests;
const THRESHOLD_FACTOR: u32 = 2;
const EXPONENTIAL_FEE_BASE: FixedU128 = FixedU128::from_rational(105, 100); // 1.05
const MESSAGE_SIZE_FEE_BASE: FixedU128 = FixedU128::from_rational(1, 1000); // 0.001
/// An error sending a downward message.
#[cfg_attr(test, derive(Debug))]
pub enum QueueDownwardMessageError {
/// The message being sent exceeds the configured max message size.
ExceedsMaxMessageSize,
}
impl From<QueueDownwardMessageError> for SendError {
fn from(err: QueueDownwardMessageError) -> Self {
match err {
QueueDownwardMessageError::ExceedsMaxMessageSize => SendError::ExceedsMaxMessageSize,
}
}
}
/// An error returned by [`Pallet::check_processed_downward_messages`] that indicates an acceptance
/// check didn't pass.
pub(crate) enum ProcessedDownwardMessagesAcceptanceErr {
/// If there are pending messages then `processed_downward_messages` should be at least 1,
AdvancementRule,
/// `processed_downward_messages` should not be greater than the number of pending messages.
Underflow { processed_downward_messages: u32, dmq_length: u32 },
}
impl fmt::Debug for ProcessedDownwardMessagesAcceptanceErr {
fn fmt(&self, fmt: &mut fmt::Formatter) -> fmt::Result {
use ProcessedDownwardMessagesAcceptanceErr::*;
match *self {
AdvancementRule => {
write!(fmt, "DMQ is not empty, but processed_downward_messages is 0",)
},
Underflow { processed_downward_messages, dmq_length } => write!(
fmt,
"processed_downward_messages = {}, but dmq_length is only {}",
processed_downward_messages, dmq_length,
),
}
}
}
#[frame_support::pallet]
pub mod pallet {
use super::*;
#[pallet::pallet]
#[pallet::without_storage_info]
pub struct Pallet<T>(_);
#[pallet::config]
pub trait Config: frame_system::Config + configuration::Config {}
/// The downward messages addressed for a certain para.
#[pallet::storage]
pub(crate) type DownwardMessageQueues<T: Config> = StorageMap<
_,
Twox64Concat,
ParaId,
Vec<InboundDownwardMessage<BlockNumberFor<T>>>,
ValueQuery,
>;
/// A mapping that stores the downward message queue MQC head for each para.
///
/// Each link in this chain has a form:
/// `(prev_head, B, H(M))`, where
/// - `prev_head`: is the previous head hash or zero if none.
/// - `B`: is the relay-chain block number in which a message was appended.
/// - `H(M)`: is the hash of the message being appended.
#[pallet::storage]
pub(crate) type DownwardMessageQueueHeads<T: Config> =
StorageMap<_, Twox64Concat, ParaId, Hash, ValueQuery>;
/// Initialization value for the DeliveryFee factor.
#[pallet::type_value]
pub fn InitialFactor() -> FixedU128 {
FixedU128::from_u32(1)
}
/// The factor to multiply the base delivery fee by.
#[pallet::storage]
pub(crate) type DeliveryFeeFactor<T: Config> =
StorageMap<_, Twox64Concat, ParaId, FixedU128, ValueQuery, InitialFactor>;
}
/// Routines and getters related to downward message passing.
impl<T: Config> Pallet<T> {
/// Block initialization logic, called by initializer.
pub(crate) fn initializer_initialize(_now: BlockNumberFor<T>) -> Weight {
Weight::zero()
}
/// Block finalization logic, called by initializer.
pub(crate) fn initializer_finalize() {}
/// Called by the initializer to note that a new session has started.
pub(crate) fn initializer_on_new_session(
_notification: &initializer::SessionChangeNotification<BlockNumberFor<T>>,
outgoing_paras: &[ParaId],
) {
Self::perform_outgoing_para_cleanup(outgoing_paras);
}
/// Iterate over all paras that were noted for offboarding and remove all the data
/// associated with them.
fn perform_outgoing_para_cleanup(outgoing: &[ParaId]) {
for outgoing_para in outgoing {
Self::clean_dmp_after_outgoing(outgoing_para);
}
}
/// Remove all relevant storage items for an outgoing parachain.
fn clean_dmp_after_outgoing(outgoing_para: &ParaId) {
DownwardMessageQueues::<T>::remove(outgoing_para);
DownwardMessageQueueHeads::<T>::remove(outgoing_para);
}
/// Determine whether enqueuing a downward message to a specific recipient para would result
/// in an error. If this returns `Ok(())` the caller can be certain that a call to
/// `queue_downward_message` with the same parameters will be successful.
pub fn can_queue_downward_message(
config: &HostConfiguration<BlockNumberFor<T>>,
para: &ParaId,
msg: &DownwardMessage,
) -> Result<(), QueueDownwardMessageError> {
let serialized_len = msg.len() as u32;
if serialized_len > config.max_downward_message_size {
return Err(QueueDownwardMessageError::ExceedsMaxMessageSize)
}
// Hard limit on Queue size
if Self::dmq_length(*para) > Self::dmq_max_length(config.max_downward_message_size) {
return Err(QueueDownwardMessageError::ExceedsMaxMessageSize)
}
Ok(())
}
/// Enqueue a downward message to a specific recipient para.
///
/// When encoded, the message should not exceed the `config.max_downward_message_size`.
/// Otherwise, the message won't be sent and `Err` will be returned.
///
/// It is possible to send a downward message to a non-existent para. That, however, would lead
/// to a dangling storage. If the caller cannot statically prove that the recipient exists
/// then the caller should perform a runtime check.
pub fn queue_downward_message(
config: &HostConfiguration<BlockNumberFor<T>>,
para: ParaId,
msg: DownwardMessage,
) -> Result<(), QueueDownwardMessageError> {
let serialized_len = msg.len() as u32;
if serialized_len > config.max_downward_message_size {
return Err(QueueDownwardMessageError::ExceedsMaxMessageSize)
}
// Hard limit on Queue size
if Self::dmq_length(para) > Self::dmq_max_length(config.max_downward_message_size) {
return Err(QueueDownwardMessageError::ExceedsMaxMessageSize)
}
let inbound =
InboundDownwardMessage { msg, sent_at: <frame_system::Pallet<T>>::block_number() };
// obtain the new link in the MQC and update the head.
DownwardMessageQueueHeads::<T>::mutate(para, |head| {
let new_head =
BlakeTwo256::hash_of(&(*head, inbound.sent_at, T::Hashing::hash_of(&inbound.msg)));
*head = new_head;
});
let q_len = DownwardMessageQueues::<T>::mutate(para, |v| {
v.push(inbound);
v.len()
});
let threshold =
Self::dmq_max_length(config.max_downward_message_size).saturating_div(THRESHOLD_FACTOR);
if q_len > (threshold as usize) {
let message_size_factor = FixedU128::from((serialized_len / 1024) as u128)
.saturating_mul(MESSAGE_SIZE_FEE_BASE);
Self::increase_fee_factor(para, message_size_factor);
}
Ok(())
}
/// Checks if the number of processed downward messages is valid.
pub(crate) fn check_processed_downward_messages(
para: ParaId,
relay_parent_number: BlockNumberFor<T>,
processed_downward_messages: u32,
) -> Result<(), ProcessedDownwardMessagesAcceptanceErr> {
let dmq_length = Self::dmq_length(para);
if dmq_length > 0 && processed_downward_messages == 0 {
// The advancement rule is for at least one downwards message to be processed
// if the queue is non-empty at the relay-parent. Downwards messages are annotated
// with the block number, so we compare the earliest (first) against the relay parent.
let contents = Self::dmq_contents(para);
// sanity: if dmq_length is >0 this should always be 'Some'.
if contents.get(0).map_or(false, |msg| msg.sent_at <= relay_parent_number) {
return Err(ProcessedDownwardMessagesAcceptanceErr::AdvancementRule)
}
}
// Note that we might be allowing a parachain to signal that it's processed
// messages that hadn't been placed in the queue at the relay_parent.
// only 'stupid' parachains would do it and we don't (and can't) force anyone
// to act on messages, so the lenient approach is fine here.
if dmq_length < processed_downward_messages {
return Err(ProcessedDownwardMessagesAcceptanceErr::Underflow {
processed_downward_messages,
dmq_length,
})
}
Ok(())
}
/// Prunes the specified number of messages from the downward message queue of the given para.
pub(crate) fn prune_dmq(para: ParaId, processed_downward_messages: u32) -> Weight {
let q_len = DownwardMessageQueues::<T>::mutate(para, |q| {
let processed_downward_messages = processed_downward_messages as usize;
if processed_downward_messages > q.len() {
// reaching this branch is unexpected due to the constraint established by
// `check_processed_downward_messages`. But better be safe than sorry.
q.clear();
} else {
*q = q.split_off(processed_downward_messages);
}
q.len()
});
let config = configuration::ActiveConfig::<T>::get();
let threshold =
Self::dmq_max_length(config.max_downward_message_size).saturating_div(THRESHOLD_FACTOR);
if q_len <= (threshold as usize) {
Self::decrease_fee_factor(para);
}
T::DbWeight::get().reads_writes(1, 1)
}
/// Returns the Head of Message Queue Chain for the given para or `None` if there is none
/// associated with it.
#[cfg(test)]
fn dmq_mqc_head(para: ParaId) -> Hash {
DownwardMessageQueueHeads::<T>::get(&para)
}
/// Returns the number of pending downward messages addressed to the given para.
///
/// Returns 0 if the para doesn't have an associated downward message queue.
pub(crate) fn dmq_length(para: ParaId) -> u32 {
DownwardMessageQueues::<T>::decode_len(&para)
.unwrap_or(0)
.saturated_into::<u32>()
}
fn dmq_max_length(max_downward_message_size: u32) -> u32 {
MAX_POSSIBLE_ALLOCATION.checked_div(max_downward_message_size).unwrap_or(0)
}
/// Returns the downward message queue contents for the given para.
///
/// The most recent messages are the latest in the vector.
pub(crate) fn dmq_contents(
recipient: ParaId,
) -> Vec<InboundDownwardMessage<BlockNumberFor<T>>> {
DownwardMessageQueues::<T>::get(&recipient)
}
}
impl<T: Config> FeeTracker for Pallet<T> {
type Id = ParaId;
fn get_fee_factor(id: Self::Id) -> FixedU128 {
DeliveryFeeFactor::<T>::get(id)
}
fn increase_fee_factor(id: Self::Id, message_size_factor: FixedU128) -> FixedU128 {
<DeliveryFeeFactor<T>>::mutate(id, |f| {
*f = f.saturating_mul(EXPONENTIAL_FEE_BASE.saturating_add(message_size_factor));
*f
})
}
fn decrease_fee_factor(id: Self::Id) -> FixedU128 {
<DeliveryFeeFactor<T>>::mutate(id, |f| {
*f = InitialFactor::get().max(*f / EXPONENTIAL_FEE_BASE);
*f
})
}
}