serai/processor/src/networks/ethereum.rs

899 lines
27 KiB
Rust
Raw Normal View History

use core::{fmt, time::Duration};
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
use std::{
sync::Arc,
collections::{HashSet, HashMap},
io,
};
use async_trait::async_trait;
use ciphersuite::{group::GroupEncoding, Ciphersuite, Secp256k1};
use frost::ThresholdKeys;
use ethereum_serai::{
alloy_core::primitives::U256,
alloy_rpc_types::{BlockNumberOrTag, Transaction},
alloy_simple_request_transport::SimpleRequest,
alloy_rpc_client::ClientBuilder,
alloy_provider::{Provider, RootProvider},
crypto::{PublicKey, Signature},
erc20::Erc20,
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
deployer::Deployer,
router::{Router, Coin as EthereumCoin, InInstruction as EthereumInInstruction},
machine::*,
};
#[cfg(test)]
use ethereum_serai::alloy_core::primitives::B256;
use tokio::{
time::sleep,
sync::{RwLock, RwLockReadGuard},
};
use serai_client::{
primitives::{Coin, Amount, Balance, NetworkId},
validator_sets::primitives::Session,
};
use crate::{
Db, Payment,
networks::{
OutputType, Output, Transaction as TransactionTrait, SignableTransaction, Block,
Eventuality as EventualityTrait, EventualitiesTracker, NetworkError, Network,
},
key_gen::NetworkKeyDb,
multisigs::scheduler::{
Scheduler as SchedulerTrait,
smart_contract::{Addendum, Scheduler},
},
};
#[cfg(not(test))]
const DAI: [u8; 20] =
match const_hex::const_decode_to_array(b"0x6B175474E89094C44Da98b954EedeAC495271d0F") {
Ok(res) => res,
Err(_) => panic!("invalid non-test DAI hex address"),
};
#[cfg(test)] // TODO
const DAI: [u8; 20] =
match const_hex::const_decode_to_array(b"0000000000000000000000000000000000000000") {
Ok(res) => res,
Err(_) => panic!("invalid test DAI hex address"),
};
fn coin_to_serai_coin(coin: &EthereumCoin) -> Option<Coin> {
match coin {
EthereumCoin::Ether => Some(Coin::Ether),
EthereumCoin::Erc20(token) => {
if *token == DAI {
return Some(Coin::Dai);
}
None
}
}
}
fn amount_to_serai_amount(coin: Coin, amount: U256) -> Amount {
assert_eq!(coin.network(), NetworkId::Ethereum);
assert_eq!(coin.decimals(), 8);
// Remove 10 decimals so we go from 18 decimals to 8 decimals
let divisor = U256::from(10_000_000_000u64);
// This is valid up to 184b, which is assumed for the coins allowed
Amount(u64::try_from(amount / divisor).unwrap())
}
fn balance_to_ethereum_amount(balance: Balance) -> U256 {
assert_eq!(balance.coin.network(), NetworkId::Ethereum);
assert_eq!(balance.coin.decimals(), 8);
// Restore 10 decimals so we go from 8 decimals to 18 decimals
let factor = U256::from(10_000_000_000u64);
U256::from(balance.amount.0) * factor
}
#[derive(Clone, Copy, PartialEq, Eq, Debug)]
pub struct Address(pub [u8; 20]);
impl TryFrom<Vec<u8>> for Address {
type Error = ();
fn try_from(bytes: Vec<u8>) -> Result<Address, ()> {
if bytes.len() != 20 {
Err(())?;
}
let mut res = [0; 20];
res.copy_from_slice(&bytes);
Ok(Address(res))
}
}
impl TryInto<Vec<u8>> for Address {
type Error = ();
fn try_into(self) -> Result<Vec<u8>, ()> {
Ok(self.0.to_vec())
}
}
impl fmt::Display for Address {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
ethereum_serai::alloy_core::primitives::Address::from(self.0).fmt(f)
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
}
}
impl SignableTransaction for RouterCommand {
fn fee(&self) -> u64 {
// Return a fee of 0 as we'll handle amortization on our end
0
}
}
#[async_trait]
2024-05-10 18:04:58 +00:00
impl<D: Db> TransactionTrait<Ethereum<D>> for Transaction {
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
type Id = [u8; 32];
fn id(&self) -> Self::Id {
self.hash.0
}
#[cfg(test)]
async fn fee(&self, _network: &Ethereum<D>) -> u64 {
// Return a fee of 0 as we'll handle amortization on our end
0
}
}
// We use 32-block Epochs to represent blocks.
#[derive(Clone, Copy, PartialEq, Eq, Debug)]
pub struct Epoch {
// The hash of the block which ended the prior Epoch.
prior_end_hash: [u8; 32],
// The first block number within this Epoch.
start: u64,
// The hash of the last block within this Epoch.
end_hash: [u8; 32],
// The monotonic time for this Epoch.
time: u64,
}
impl Epoch {
fn end(&self) -> u64 {
self.start + 31
}
}
#[async_trait]
2024-05-10 18:04:58 +00:00
impl<D: Db> Block<Ethereum<D>> for Epoch {
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
type Id = [u8; 32];
fn id(&self) -> [u8; 32] {
self.end_hash
}
fn parent(&self) -> [u8; 32] {
self.prior_end_hash
}
async fn time(&self, _: &Ethereum<D>) -> u64 {
self.time
}
}
2024-05-10 18:04:58 +00:00
impl<D: Db> Output<Ethereum<D>> for EthereumInInstruction {
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
type Id = [u8; 32];
fn kind(&self) -> OutputType {
OutputType::External
}
fn id(&self) -> Self::Id {
let mut id = [0; 40];
id[.. 32].copy_from_slice(&self.id.0);
id[32 ..].copy_from_slice(&self.id.1.to_le_bytes());
*ethereum_serai::alloy_core::primitives::keccak256(id)
}
fn tx_id(&self) -> [u8; 32] {
self.id.0
}
fn key(&self) -> <Secp256k1 as Ciphersuite>::G {
self.key_at_end_of_block
}
fn presumed_origin(&self) -> Option<Address> {
Some(Address(self.from))
}
fn balance(&self) -> Balance {
let coin = coin_to_serai_coin(&self.coin).unwrap_or_else(|| {
panic!(
"requesting coin for an EthereumInInstruction with a coin {}",
"we don't handle. this never should have been yielded"
)
});
Balance { coin, amount: amount_to_serai_amount(coin, self.amount) }
}
fn data(&self) -> &[u8] {
&self.data
}
fn write<W: io::Write>(&self, writer: &mut W) -> io::Result<()> {
EthereumInInstruction::write(self, writer)
}
fn read<R: io::Read>(reader: &mut R) -> io::Result<Self> {
EthereumInInstruction::read(reader)
}
}
#[derive(Clone, PartialEq, Eq, Debug)]
pub struct Claim {
signature: [u8; 64],
}
impl AsRef<[u8]> for Claim {
fn as_ref(&self) -> &[u8] {
&self.signature
}
}
impl AsMut<[u8]> for Claim {
fn as_mut(&mut self) -> &mut [u8] {
&mut self.signature
}
}
impl Default for Claim {
fn default() -> Self {
Self { signature: [0; 64] }
}
}
impl From<&Signature> for Claim {
fn from(sig: &Signature) -> Self {
Self { signature: sig.to_bytes() }
}
}
#[derive(Clone, PartialEq, Eq, Debug)]
pub struct Eventuality(PublicKey, RouterCommand);
impl EventualityTrait for Eventuality {
type Claim = Claim;
type Completion = SignedRouterCommand;
fn lookup(&self) -> Vec<u8> {
match self.1 {
RouterCommand::UpdateSeraiKey { nonce, .. } | RouterCommand::Execute { nonce, .. } => {
nonce.as_le_bytes().to_vec()
}
}
}
fn read<R: io::Read>(reader: &mut R) -> io::Result<Self> {
let point = Secp256k1::read_G(reader)?;
let command = RouterCommand::read(reader)?;
Ok(Eventuality(
PublicKey::new(point).ok_or(io::Error::other("unusable key within Eventuality"))?,
command,
))
}
fn serialize(&self) -> Vec<u8> {
let mut res = vec![];
res.extend(self.0.point().to_bytes().as_slice());
self.1.write(&mut res).unwrap();
res
}
fn claim(completion: &Self::Completion) -> Self::Claim {
Claim::from(completion.signature())
}
fn serialize_completion(completion: &Self::Completion) -> Vec<u8> {
let mut res = vec![];
completion.write(&mut res).unwrap();
res
}
fn read_completion<R: io::Read>(reader: &mut R) -> io::Result<Self::Completion> {
SignedRouterCommand::read(reader)
}
}
2024-05-10 18:04:58 +00:00
#[derive(Clone)]
pub struct Ethereum<D: Db> {
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
// This DB is solely used to access the first key generated, as needed to determine the Router's
// address. Accordingly, all methods present are consistent to a Serai chain with a finalized
// first key (regardless of local state), and this is safe.
db: D,
provider: Arc<RootProvider<SimpleRequest>>,
deployer: Deployer,
router: Arc<RwLock<Option<Router>>>,
}
2024-05-10 18:04:58 +00:00
impl<D: Db> PartialEq for Ethereum<D> {
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
fn eq(&self, _other: &Ethereum<D>) -> bool {
true
}
}
2024-05-10 18:04:58 +00:00
impl<D: Db> fmt::Debug for Ethereum<D> {
fn fmt(&self, fmt: &mut fmt::Formatter<'_>) -> fmt::Result {
fmt
.debug_struct("Ethereum")
.field("deployer", &self.deployer)
.field("router", &self.router)
.finish_non_exhaustive()
}
}
impl<D: Db> Ethereum<D> {
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
pub async fn new(db: D, url: String) -> Self {
let provider = Arc::new(RootProvider::new(
ClientBuilder::default().transport(SimpleRequest::new(url), true),
));
let mut deployer = Deployer::new(provider.clone()).await;
while !matches!(deployer, Ok(Some(_))) {
log::error!("Deployer wasn't deployed yet or networking error");
sleep(Duration::from_secs(5)).await;
deployer = Deployer::new(provider.clone()).await;
}
let deployer = deployer.unwrap().unwrap();
Ethereum { db, provider, deployer, router: Arc::new(RwLock::new(None)) }
}
// Obtain a reference to the Router, sleeping until it's deployed if it hasn't already been.
// This is guaranteed to return Some.
pub async fn router(&self) -> RwLockReadGuard<'_, Option<Router>> {
// If we've already instantiated the Router, return a read reference
{
let router = self.router.read().await;
if router.is_some() {
return router;
}
}
// Instantiate it
let mut router = self.router.write().await;
// If another attempt beat us to it, return
if router.is_some() {
drop(router);
return self.router.read().await;
}
// Get the first key from the DB
let first_key =
NetworkKeyDb::get(&self.db, Session(0)).expect("getting outputs before confirming a key");
let key = Secp256k1::read_G(&mut first_key.as_slice()).unwrap();
let public_key = PublicKey::new(key).unwrap();
// Find the router
let mut found = self.deployer.find_router(self.provider.clone(), &public_key).await;
while !matches!(found, Ok(Some(_))) {
log::error!("Router wasn't deployed yet or networking error");
sleep(Duration::from_secs(5)).await;
found = self.deployer.find_router(self.provider.clone(), &public_key).await;
}
// Set it
*router = Some(found.unwrap().unwrap());
// Downgrade to a read lock
// Explicitly doesn't use `downgrade` so that another pending write txn can realize it's no
// longer necessary
drop(router);
self.router.read().await
}
}
#[async_trait]
2024-05-10 18:04:58 +00:00
impl<D: Db> Network for Ethereum<D> {
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
type Curve = Secp256k1;
type Transaction = Transaction;
type Block = Epoch;
type Output = EthereumInInstruction;
type SignableTransaction = RouterCommand;
type Eventuality = Eventuality;
type TransactionMachine = RouterCommandMachine;
type Scheduler = Scheduler<Self>;
type Address = Address;
const NETWORK: NetworkId = NetworkId::Ethereum;
const ID: &'static str = "Ethereum";
const ESTIMATED_BLOCK_TIME_IN_SECONDS: usize = 32 * 12;
const CONFIRMATIONS: usize = 1;
const DUST: u64 = 0; // TODO
const COST_TO_AGGREGATE: u64 = 0;
// TODO: usize::max, with a merkle tree in the router
const MAX_OUTPUTS: usize = 256;
fn tweak_keys(keys: &mut ThresholdKeys<Self::Curve>) {
while PublicKey::new(keys.group_key()).is_none() {
*keys = keys.offset(<Secp256k1 as Ciphersuite>::F::ONE);
}
}
#[cfg(test)]
async fn external_address(&self, _key: <Secp256k1 as Ciphersuite>::G) -> Address {
Address(self.router().await.as_ref().unwrap().address())
}
fn branch_address(_key: <Secp256k1 as Ciphersuite>::G) -> Option<Address> {
None
}
fn change_address(_key: <Secp256k1 as Ciphersuite>::G) -> Option<Address> {
None
}
fn forward_address(_key: <Secp256k1 as Ciphersuite>::G) -> Option<Address> {
None
}
async fn get_latest_block_number(&self) -> Result<usize, NetworkError> {
let actual_number = self
.provider
.get_block(BlockNumberOrTag::Finalized.into(), false)
.await
.map_err(|_| NetworkError::ConnectionError)?
.expect("no blocks were finalized")
.header
.number
.unwrap();
// Error if there hasn't been a full epoch yet
if actual_number < 32 {
Err(NetworkError::ConnectionError)?
}
// If this is 33, the division will return 1, yet 1 is the epoch in progress
let latest_full_epoch = (actual_number / 32).saturating_sub(1);
Ok(latest_full_epoch.try_into().unwrap())
}
async fn get_block(&self, number: usize) -> Result<Self::Block, NetworkError> {
let latest_finalized = self.get_latest_block_number().await?;
if number > latest_finalized {
Err(NetworkError::ConnectionError)?
}
let start = number * 32;
let prior_end_hash = if start == 0 {
[0; 32]
} else {
self
.provider
.get_block(u64::try_from(start - 1).unwrap().into(), false)
.await
.ok()
.flatten()
.ok_or(NetworkError::ConnectionError)?
.header
.hash
.unwrap()
.into()
};
let end_header = self
.provider
.get_block(u64::try_from(start + 31).unwrap().into(), false)
.await
.ok()
.flatten()
.ok_or(NetworkError::ConnectionError)?
.header;
let end_hash = end_header.hash.unwrap().into();
let time = end_header.timestamp;
Ok(Epoch { prior_end_hash, start: start.try_into().unwrap(), end_hash, time })
}
async fn get_outputs(
&self,
block: &Self::Block,
_: <Secp256k1 as Ciphersuite>::G,
) -> Vec<Self::Output> {
let router = self.router().await;
let router = router.as_ref().unwrap();
// Grab the key at the end of the epoch
let key_at_end_of_block = loop {
match router.key_at_end_of_block(block.start + 31).await {
2024-05-10 18:04:58 +00:00
Ok(Some(key)) => break key,
Ok(None) => return vec![],
Err(e) => {
log::error!("couldn't connect to router for the key at the end of the block: {e:?}");
sleep(Duration::from_secs(5)).await;
continue;
}
}
};
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
let mut all_events = vec![];
let mut top_level_txids = HashSet::new();
for erc20_addr in [DAI] {
2024-05-10 18:04:58 +00:00
let erc20 = Erc20::new(self.provider.clone(), erc20_addr);
for block in block.start .. (block.start + 32) {
let transfers = loop {
match erc20.top_level_transfers(block, router.address()).await {
Ok(transfers) => break transfers,
Err(e) => {
log::error!("couldn't connect to Ethereum node for the top-level transfers: {e:?}");
sleep(Duration::from_secs(5)).await;
continue;
}
}
};
for transfer in transfers {
top_level_txids.insert(transfer.id);
all_events.push(EthereumInInstruction {
id: (transfer.id, 0),
from: transfer.from,
coin: EthereumCoin::Erc20(erc20_addr),
amount: transfer.amount,
data: transfer.data,
key_at_end_of_block,
});
}
}
}
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
for block in block.start .. (block.start + 32) {
let mut events = router.in_instructions(block, &HashSet::from([DAI])).await;
while let Err(e) = events {
log::error!("couldn't connect to Ethereum node for the Router's events: {e:?}");
sleep(Duration::from_secs(5)).await;
events = router.in_instructions(block, &HashSet::from([DAI])).await;
}
let mut events = events.unwrap();
for event in &mut events {
// A transaction should either be a top-level transfer or a Router InInstruction
if top_level_txids.contains(&event.id.0) {
panic!("top-level transfer had {} and router had {:?}", hex::encode(event.id.0), event);
}
// Overwrite the key at end of block to key at end of epoch
event.key_at_end_of_block = key_at_end_of_block;
}
all_events.extend(events);
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
}
for event in &all_events {
assert!(
coin_to_serai_coin(&event.coin).is_some(),
"router yielded events for unrecognized coins"
);
}
all_events
}
async fn get_eventuality_completions(
&self,
eventualities: &mut EventualitiesTracker<Self::Eventuality>,
block: &Self::Block,
) -> HashMap<
[u8; 32],
(
usize,
<Self::Transaction as TransactionTrait<Self>>::Id,
<Self::Eventuality as EventualityTrait>::Completion,
),
> {
let mut res = HashMap::new();
if eventualities.map.is_empty() {
return res;
}
let router = self.router().await;
let router = router.as_ref().unwrap();
let past_scanned_epoch = loop {
match self.get_block(eventualities.block_number).await {
Ok(block) => break block,
Err(e) => log::error!("couldn't get the last scanned block in the tracker: {}", e),
}
sleep(Duration::from_secs(10)).await;
};
assert_eq!(
past_scanned_epoch.start / 32,
u64::try_from(eventualities.block_number).unwrap(),
"assumption of tracker block number's relation to epoch start is incorrect"
);
// Iterate from after the epoch number in the tracker to the end of this epoch
for block_num in (past_scanned_epoch.end() + 1) ..= block.end() {
let executed = loop {
match router.executed_commands(block_num).await {
Ok(executed) => break executed,
Err(e) => log::error!("couldn't get the executed commands in block {block_num}: {e}"),
}
sleep(Duration::from_secs(10)).await;
};
for executed in executed {
let lookup = executed.nonce.to_le_bytes().to_vec();
if let Some((plan_id, eventuality)) = eventualities.map.get(&lookup) {
if let Some(command) =
SignedRouterCommand::new(&eventuality.0, eventuality.1.clone(), &executed.signature)
{
res.insert(*plan_id, (block_num.try_into().unwrap(), executed.tx_id, command));
eventualities.map.remove(&lookup);
}
}
}
}
eventualities.block_number = (block.start / 32).try_into().unwrap();
res
}
async fn needed_fee(
&self,
_block_number: usize,
inputs: &[Self::Output],
_payments: &[Payment<Self>],
_change: &Option<Self::Address>,
) -> Result<Option<u64>, NetworkError> {
assert_eq!(inputs.len(), 0);
// Claim no fee is needed so we can perform amortization ourselves
Ok(Some(0))
}
async fn signable_transaction(
&self,
_block_number: usize,
_plan_id: &[u8; 32],
key: <Self::Curve as Ciphersuite>::G,
inputs: &[Self::Output],
payments: &[Payment<Self>],
change: &Option<Self::Address>,
scheduler_addendum: &<Self::Scheduler as SchedulerTrait<Self>>::Addendum,
) -> Result<Option<(Self::SignableTransaction, Self::Eventuality)>, NetworkError> {
assert_eq!(inputs.len(), 0);
assert!(change.is_none());
let chain_id = self.provider.get_chain_id().await.map_err(|_| NetworkError::ConnectionError)?;
// TODO: Perform fee amortization (in scheduler?
// TODO: Make this function internal and have needed_fee properly return None as expected?
// TODO: signable_transaction is written as cannot return None if needed_fee returns Some
// TODO: Why can this return None at all if it isn't allowed to return None?
let command = match scheduler_addendum {
Addendum::Nonce(nonce) => RouterCommand::Execute {
chain_id: U256::try_from(chain_id).unwrap(),
nonce: U256::try_from(*nonce).unwrap(),
outs: payments
.iter()
.filter_map(|payment| {
Some(OutInstruction {
target: if let Some(data) = payment.data.as_ref() {
// This introspects the Call serialization format, expecting the first 20 bytes to
// be the address
// This avoids wasting the 20-bytes allocated within address
let full_data = [payment.address.0.as_slice(), data].concat();
let mut reader = full_data.as_slice();
let mut calls = vec![];
while !reader.is_empty() {
calls.push(Call::read(&mut reader).ok()?)
}
// The above must have executed at least once since reader contains the address
assert_eq!(calls[0].to, payment.address.0);
OutInstructionTarget::Calls(calls)
} else {
OutInstructionTarget::Direct(payment.address.0)
},
value: {
assert_eq!(payment.balance.coin, Coin::Ether); // TODO
balance_to_ethereum_amount(payment.balance)
},
})
})
.collect(),
},
Addendum::RotateTo { nonce, new_key } => {
assert!(payments.is_empty());
RouterCommand::UpdateSeraiKey {
chain_id: U256::try_from(chain_id).unwrap(),
nonce: U256::try_from(*nonce).unwrap(),
key: PublicKey::new(*new_key).expect("new key wasn't a valid ETH public key"),
}
}
};
Ok(Some((
command.clone(),
Eventuality(PublicKey::new(key).expect("key wasn't a valid ETH public key"), command),
)))
}
async fn attempt_sign(
&self,
keys: ThresholdKeys<Self::Curve>,
transaction: Self::SignableTransaction,
) -> Result<Self::TransactionMachine, NetworkError> {
Ok(
RouterCommandMachine::new(keys, transaction)
.expect("keys weren't usable to sign router commands"),
)
}
async fn publish_completion(
&self,
completion: &<Self::Eventuality as EventualityTrait>::Completion,
) -> Result<(), NetworkError> {
// Publish this to the dedicated TX server for a solver to actually publish
#[cfg(not(test))]
{
let _ = completion;
todo!("TODO");
}
// Publish this using a dummy account we fund with magic RPC commands
#[cfg(test)]
{
use rand_core::OsRng;
use ciphersuite::group::ff::Field;
let key = <Secp256k1 as Ciphersuite>::F::random(&mut OsRng);
let address = ethereum_serai::crypto::address(&(Secp256k1::generator() * key));
// Set a 1.1 ETH balance
self
.provider
.raw_request::<_, ()>(
"anvil_setBalance".into(),
[Address(address).to_string(), "1100000000000000000".into()],
)
.await
.unwrap();
let router = self.router().await;
let router = router.as_ref().unwrap();
let mut tx = match completion.command() {
RouterCommand::UpdateSeraiKey { key, .. } => {
router.update_serai_key(key, completion.signature())
}
RouterCommand::Execute { outs, .. } => router.execute(
&outs.iter().cloned().map(Into::into).collect::<Vec<_>>(),
completion.signature(),
),
};
tx.gas_price = 100_000_000_000u128;
use ethereum_serai::alloy_consensus::SignableTransaction;
let sig =
k256::ecdsa::SigningKey::from(k256::elliptic_curve::NonZeroScalar::new(key).unwrap())
.sign_prehash_recoverable(tx.signature_hash().as_ref())
.unwrap();
let mut bytes = vec![];
tx.encode_with_signature_fields(&sig.into(), &mut bytes);
let _ = self.provider.send_raw_transaction(&bytes).await.ok().unwrap();
Ok(())
}
}
async fn confirm_completion(
&self,
eventuality: &Self::Eventuality,
claim: &<Self::Eventuality as EventualityTrait>::Claim,
) -> Result<Option<<Self::Eventuality as EventualityTrait>::Completion>, NetworkError> {
Ok(SignedRouterCommand::new(&eventuality.0, eventuality.1.clone(), &claim.signature))
}
#[cfg(test)]
async fn get_block_number(&self, id: &<Self::Block as Block<Self>>::Id) -> usize {
self
.provider
.get_block(B256::from(*id).into(), false)
.await
.unwrap()
.unwrap()
.header
.number
.unwrap()
.try_into()
.unwrap()
}
#[cfg(test)]
async fn check_eventuality_by_claim(
&self,
eventuality: &Self::Eventuality,
claim: &<Self::Eventuality as EventualityTrait>::Claim,
) -> bool {
SignedRouterCommand::new(&eventuality.0, eventuality.1.clone(), &claim.signature).is_some()
}
#[cfg(test)]
async fn get_transaction_by_eventuality(
&self,
block: usize,
eventuality: &Self::Eventuality,
) -> Self::Transaction {
match eventuality.1 {
RouterCommand::UpdateSeraiKey { nonce, .. } | RouterCommand::Execute { nonce, .. } => {
let router = self.router().await;
let router = router.as_ref().unwrap();
let block = u64::try_from(block).unwrap();
let filter = router
.key_updated_filter()
.from_block(block * 32)
.to_block(((block + 1) * 32) - 1)
.topic1(nonce);
let logs = self.provider.get_logs(&filter).await.unwrap();
if let Some(log) = logs.first() {
return self
.provider
.get_transaction_by_hash(log.clone().transaction_hash.unwrap())
.await
2024-05-10 18:04:58 +00:00
.unwrap()
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
.unwrap();
};
let filter = router
.executed_filter()
.from_block(block * 32)
.to_block(((block + 1) * 32) - 1)
.topic1(nonce);
let logs = self.provider.get_logs(&filter).await.unwrap();
2024-05-10 18:04:58 +00:00
self
.provider
.get_transaction_by_hash(logs[0].transaction_hash.unwrap())
.await
.unwrap()
.unwrap()
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
}
}
}
#[cfg(test)]
async fn mine_block(&self) {
2024-05-10 18:04:58 +00:00
self.provider.raw_request::<_, ()>("anvil_mine".into(), [96]).await.unwrap();
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
}
#[cfg(test)]
async fn test_send(&self, send_to: Self::Address) -> Self::Block {
use rand_core::OsRng;
use ciphersuite::group::ff::Field;
2024-05-10 18:04:58 +00:00
use ethereum_serai::alloy_sol_types::SolCall;
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
let key = <Secp256k1 as Ciphersuite>::F::random(&mut OsRng);
let address = ethereum_serai::crypto::address(&(Secp256k1::generator() * key));
// Set a 1.1 ETH balance
self
.provider
.raw_request::<_, ()>(
"anvil_setBalance".into(),
[Address(address).to_string(), "1100000000000000000".into()],
)
.await
.unwrap();
2024-05-10 18:04:58 +00:00
let value = U256::from_str_radix("1000000000000000000", 10).unwrap();
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
let tx = ethereum_serai::alloy_consensus::TxLegacy {
chain_id: None,
nonce: 0,
2024-05-10 18:04:58 +00:00
gas_price: 1_000_000_000u128,
gas_limit: 200_000u128,
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
to: ethereum_serai::alloy_core::primitives::TxKind::Call(send_to.0.into()),
// 1 ETH
2024-05-10 18:04:58 +00:00
value,
input: ethereum_serai::router::abi::inInstructionCall::new((
[0; 20].into(),
value,
vec![].into(),
))
.abi_encode()
.into(),
Ethereum Integration (#557) * Clean up Ethereum * Consistent contract address for deployed contracts * Flesh out Router a bit * Add a Deployer for DoS-less deployment * Implement Router-finding * Use CREATE2 helper present in ethers * Move from CREATE2 to CREATE Bit more streamlined for our use case. * Document ethereum-serai * Tidy tests a bit * Test updateSeraiKey * Use encodePacked for updateSeraiKey * Take in the block hash to read state during * Add a Sandbox contract to the Ethereum integration * Add retrieval of transfers from Ethereum * Add inInstruction function to the Router * Augment our handling of InInstructions events with a check the transfer event also exists * Have the Deployer error upon failed deployments * Add --via-ir * Make get_transaction test-only We only used it to get transactions to confirm the resolution of Eventualities. Eventualities need to be modularized. By introducing the dedicated confirm_completion function, we remove the need for a non-test get_transaction AND begin this modularization (by no longer explicitly grabbing a transaction to check with). * Modularize Eventuality Almost fully-deprecates the Transaction trait for Completion. Replaces Transaction ID with Claim. * Modularize the Scheduler behind a trait * Add an extremely basic account Scheduler * Add nonce uses, key rotation to the account scheduler * Only report the account Scheduler empty after transferring keys Also ban payments to the branch/change/forward addresses. * Make fns reliant on state test-only * Start of an Ethereum integration for the processor * Add a session to the Router to prevent updateSeraiKey replaying This would only happen if an old key was rotated to again, which would require n-of-n collusion (already ridiculous and a valid fault attributable event). It just clarifies the formal arguments. * Add a RouterCommand + SignMachine for producing it to coins/ethereum * Ethereum which compiles * Have branch/change/forward return an option Also defines a UtxoNetwork extension trait for MAX_INPUTS. * Make external_address exclusively a test fn * Move the "account" scheduler to "smart contract" * Remove ABI artifact * Move refund/forward Plan creation into the Processor We create forward Plans in the scan path, and need to know their exact fees in the scan path. This requires adding a somewhat wonky shim_forward_plan method so we can obtain a Plan equivalent to the actual forward Plan for fee reasons, yet don't expect it to be the actual forward Plan (which may be distinct if the Plan pulls from the global state, such as with a nonce). Also properly types a Scheduler addendum such that the SC scheduler isn't cramming the nonce to use into the N::Output type. * Flesh out the Ethereum integration more * Two commits ago, into the **Scheduler, not Processor * Remove misc TODOs in SC Scheduler * Add constructor to RouterCommandMachine * RouterCommand read, pairing with the prior added write * Further add serialization methods * Have the Router's key included with the InInstruction This does not use the key at the time of the event. This uses the key at the end of the block for the event. Its much simpler than getting the full event streams for each, checking when they interlace. This does not read the state. Every block, this makes a request for every single key update and simply chooses the last one. This allows pruning state, only keeping the event tree. Ideally, we'd also introduce a cache to reduce the cost of the filter (small in events yielded, long in blocks searched). Since Serai doesn't have any forwarding TXs, nor Branches, nor change, all of our Plans should solely have payments out, and there's no expectation of a Plan being made under one key broken by it being received by another key. * Add read/write to InInstruction * Abstract the ABI for Call/OutInstruction in ethereum-serai * Fill out signable_transaction for Ethereum * Move ethereum-serai to alloy Resolves #331. * Use the opaque sol macro instead of generated files * Move the processor over to the now-alloy-based ethereum-serai * Use the ecrecover provided by alloy * Have the SC use nonce for rotation, not session (an independent nonce which wasn't synchronized) * Always use the latest keys for SC scheduled plans * get_eventuality_completions for Ethereum * Finish fleshing out the processor Ethereum integration as needed for serai-processor tests This doesn't not support any actual deployments, not even the ones simulated by serai-processor-docker-tests. * Add alloy-simple-request-transport to the GH workflows * cargo update * Clarify a few comments and make one check more robust * Use a string for 27.0 in .github * Remove optional from no-longer-optional dependencies in processor * Add alloy to git deny exception * Fix no longer optional specification in processor's binaries feature * Use a version of foundry from 2024 * Correct fetching Bitcoin TXs in the processor docker tests * Update rustls to resolve RUSTSEC warnings * Use the monthly nightly foundry, not the deleted daily nightly
2024-04-21 10:02:12 +00:00
};
use ethereum_serai::alloy_consensus::SignableTransaction;
let sig = k256::ecdsa::SigningKey::from(k256::elliptic_curve::NonZeroScalar::new(key).unwrap())
.sign_prehash_recoverable(tx.signature_hash().as_ref())
.unwrap();
let mut bytes = vec![];
tx.encode_with_signature_fields(&sig.into(), &mut bytes);
let pending_tx = self.provider.send_raw_transaction(&bytes).await.ok().unwrap();
// Mine an epoch containing this TX
self.mine_block().await;
assert!(pending_tx.get_receipt().await.unwrap().status());
// Yield the freshly mined block
self.get_block(self.get_latest_block_number().await.unwrap()).await.unwrap()
}
}