Files
serai/coordinator/tributary-sdk/src/blockchain.rs

343 lines
11 KiB
Rust
Raw Normal View History

2023-12-01 12:09:22 -05:00
use std::collections::{VecDeque, HashSet};
2023-04-12 11:13:48 -04:00
use dalek_ff_group::Ristretto;
Smash the singular `Ciphersuite` trait into multiple This helps identify where the various functionalities are used, or rather, not used. The `Ciphersuite` trait present in `patches/ciphersuite`, facilitating the entire FCMP++ tree, only requires the markers _and_ canonical point decoding. I've opened a PR to upstream such a trait into `group` (https://github.com/zkcrypto/group/pull/68). `WrappedGroup` is still justified for as long as `Group::generator` exists. Moving `::generator()` to its own trait, on an independent structure (upstream) would be massively appreciated. @tarcieri also wanted to update from `fn generator()` to `const GENERATOR`, which would encourage further discussion on https://github.com/zkcrypto/group/issues/32 and https://github.com/zkcrypto/group/issues/45, which have been stagnant. The `Id` trait is occasionally used yet really should be first off the chopping block. Finally, `WithPreferredHash` is only actually used around a third of the time, which more than justifies it being a separate trait. --- Updates `dalek_ff_group::Scalar` to directly re-export `curve25519_dalek::Scalar`, as without issue. `dalek_ff_group::RistrettoPoint` also could be replaced with an export of `curve25519_dalek::RistrettoPoint`, yet the coordinator relies on how we implemented `Hash` on it for the hell of it so it isn't worth it at this time. `dalek_ff_group::EdwardsPoint` can't be replaced for an re-export of `curve25519_dalek::SubgroupPoint` as it doesn't implement `zeroize`, `subtle` traits within a released, non-yanked version. Relevance to https://github.com/serai-dex/serai/issues/201 and https://github.com/dalek-cryptography/curve25519-dalek/issues/811#issuecomment-3247732746. Also updates the `Ristretto` ciphersuite to prefer `Blake2b-512` over `SHA2-512`. In order to maintain compliance with FROST's IETF standard, `modular-frost` defines its own ciphersuite for Ristretto which still uses `SHA2-512`.
2025-09-03 12:25:37 -04:00
use ciphersuite::{group::GroupEncoding, *};
2023-12-01 12:09:22 -05:00
use serai_db::{Get, DbTxn, Db};
2023-04-12 11:13:48 -04:00
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
use scale::Decode;
use tendermint::ext::{Network, Commit};
use crate::{
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
ReadWrite, ProvidedError, ProvidedTransactions, BlockError, Block, Mempool, Transaction,
transaction::{Signed, TransactionKind, TransactionError, Transaction as TransactionTrait},
};
2023-04-12 11:13:48 -04:00
#[derive(Debug)]
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
pub(crate) struct Blockchain<D: Db, T: TransactionTrait> {
db: Option<D>,
2023-04-12 11:13:48 -04:00
genesis: [u8; 32],
block_number: u64,
2023-04-12 11:13:48 -04:00
tip: [u8; 32],
Smash the singular `Ciphersuite` trait into multiple This helps identify where the various functionalities are used, or rather, not used. The `Ciphersuite` trait present in `patches/ciphersuite`, facilitating the entire FCMP++ tree, only requires the markers _and_ canonical point decoding. I've opened a PR to upstream such a trait into `group` (https://github.com/zkcrypto/group/pull/68). `WrappedGroup` is still justified for as long as `Group::generator` exists. Moving `::generator()` to its own trait, on an independent structure (upstream) would be massively appreciated. @tarcieri also wanted to update from `fn generator()` to `const GENERATOR`, which would encourage further discussion on https://github.com/zkcrypto/group/issues/32 and https://github.com/zkcrypto/group/issues/45, which have been stagnant. The `Id` trait is occasionally used yet really should be first off the chopping block. Finally, `WithPreferredHash` is only actually used around a third of the time, which more than justifies it being a separate trait. --- Updates `dalek_ff_group::Scalar` to directly re-export `curve25519_dalek::Scalar`, as without issue. `dalek_ff_group::RistrettoPoint` also could be replaced with an export of `curve25519_dalek::RistrettoPoint`, yet the coordinator relies on how we implemented `Hash` on it for the hell of it so it isn't worth it at this time. `dalek_ff_group::EdwardsPoint` can't be replaced for an re-export of `curve25519_dalek::SubgroupPoint` as it doesn't implement `zeroize`, `subtle` traits within a released, non-yanked version. Relevance to https://github.com/serai-dex/serai/issues/201 and https://github.com/dalek-cryptography/curve25519-dalek/issues/811#issuecomment-3247732746. Also updates the `Ristretto` ciphersuite to prefer `Blake2b-512` over `SHA2-512`. In order to maintain compliance with FROST's IETF standard, `modular-frost` defines its own ciphersuite for Ristretto which still uses `SHA2-512`.
2025-09-03 12:25:37 -04:00
participants: HashSet<<Ristretto as WrappedGroup>::G>,
2023-04-13 09:47:14 -04:00
provided: ProvidedTransactions<D, T>,
mempool: Mempool<D, T>,
pub(crate) next_block_notifications: VecDeque<tokio::sync::oneshot::Sender<()>>,
2023-04-12 11:13:48 -04:00
}
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
impl<D: Db, T: TransactionTrait> Blockchain<D, T> {
fn tip_key(genesis: [u8; 32]) -> Vec<u8> {
D::key(b"tributary_blockchain", b"tip", genesis)
}
fn block_number_key(&self) -> Vec<u8> {
D::key(b"tributary_blockchain", b"block_number", self.genesis)
}
fn block_key(genesis: &[u8], hash: &[u8; 32]) -> Vec<u8> {
D::key(b"tributary_blockchain", b"block", [genesis, hash].concat())
}
fn block_hash_key(genesis: &[u8], block_number: u64) -> Vec<u8> {
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
D::key(b"tributary_blockchain", b"block_hash", [genesis, &block_number.to_le_bytes()].concat())
}
fn commit_key(genesis: &[u8], hash: &[u8; 32]) -> Vec<u8> {
D::key(b"tributary_blockchain", b"commit", [genesis, hash].concat())
}
fn block_after_key(genesis: &[u8], hash: &[u8; 32]) -> Vec<u8> {
D::key(b"tributary_blockchain", b"block_after", [genesis, hash].concat())
}
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
fn unsigned_included_key(genesis: &[u8], hash: &[u8; 32]) -> Vec<u8> {
D::key(b"tributary_blockchain", b"unsigned_included", [genesis, hash].concat())
}
fn provided_included_key(genesis: &[u8], hash: &[u8; 32]) -> Vec<u8> {
D::key(b"tributary_blockchain", b"provided_included", [genesis, hash].concat())
}
2023-12-01 12:09:22 -05:00
fn next_nonce_key(
genesis: &[u8; 32],
Smash the singular `Ciphersuite` trait into multiple This helps identify where the various functionalities are used, or rather, not used. The `Ciphersuite` trait present in `patches/ciphersuite`, facilitating the entire FCMP++ tree, only requires the markers _and_ canonical point decoding. I've opened a PR to upstream such a trait into `group` (https://github.com/zkcrypto/group/pull/68). `WrappedGroup` is still justified for as long as `Group::generator` exists. Moving `::generator()` to its own trait, on an independent structure (upstream) would be massively appreciated. @tarcieri also wanted to update from `fn generator()` to `const GENERATOR`, which would encourage further discussion on https://github.com/zkcrypto/group/issues/32 and https://github.com/zkcrypto/group/issues/45, which have been stagnant. The `Id` trait is occasionally used yet really should be first off the chopping block. Finally, `WithPreferredHash` is only actually used around a third of the time, which more than justifies it being a separate trait. --- Updates `dalek_ff_group::Scalar` to directly re-export `curve25519_dalek::Scalar`, as without issue. `dalek_ff_group::RistrettoPoint` also could be replaced with an export of `curve25519_dalek::RistrettoPoint`, yet the coordinator relies on how we implemented `Hash` on it for the hell of it so it isn't worth it at this time. `dalek_ff_group::EdwardsPoint` can't be replaced for an re-export of `curve25519_dalek::SubgroupPoint` as it doesn't implement `zeroize`, `subtle` traits within a released, non-yanked version. Relevance to https://github.com/serai-dex/serai/issues/201 and https://github.com/dalek-cryptography/curve25519-dalek/issues/811#issuecomment-3247732746. Also updates the `Ristretto` ciphersuite to prefer `Blake2b-512` over `SHA2-512`. In order to maintain compliance with FROST's IETF standard, `modular-frost` defines its own ciphersuite for Ristretto which still uses `SHA2-512`.
2025-09-03 12:25:37 -04:00
signer: &<Ristretto as WrappedGroup>::G,
2023-12-01 12:09:22 -05:00
order: &[u8],
) -> Vec<u8> {
D::key(
b"tributary_blockchain",
b"next_nonce",
2023-12-01 12:09:22 -05:00
[genesis.as_ref(), signer.to_bytes().as_ref(), order].concat(),
)
}
pub(crate) fn new(
db: D,
genesis: [u8; 32],
Smash the singular `Ciphersuite` trait into multiple This helps identify where the various functionalities are used, or rather, not used. The `Ciphersuite` trait present in `patches/ciphersuite`, facilitating the entire FCMP++ tree, only requires the markers _and_ canonical point decoding. I've opened a PR to upstream such a trait into `group` (https://github.com/zkcrypto/group/pull/68). `WrappedGroup` is still justified for as long as `Group::generator` exists. Moving `::generator()` to its own trait, on an independent structure (upstream) would be massively appreciated. @tarcieri also wanted to update from `fn generator()` to `const GENERATOR`, which would encourage further discussion on https://github.com/zkcrypto/group/issues/32 and https://github.com/zkcrypto/group/issues/45, which have been stagnant. The `Id` trait is occasionally used yet really should be first off the chopping block. Finally, `WithPreferredHash` is only actually used around a third of the time, which more than justifies it being a separate trait. --- Updates `dalek_ff_group::Scalar` to directly re-export `curve25519_dalek::Scalar`, as without issue. `dalek_ff_group::RistrettoPoint` also could be replaced with an export of `curve25519_dalek::RistrettoPoint`, yet the coordinator relies on how we implemented `Hash` on it for the hell of it so it isn't worth it at this time. `dalek_ff_group::EdwardsPoint` can't be replaced for an re-export of `curve25519_dalek::SubgroupPoint` as it doesn't implement `zeroize`, `subtle` traits within a released, non-yanked version. Relevance to https://github.com/serai-dex/serai/issues/201 and https://github.com/dalek-cryptography/curve25519-dalek/issues/811#issuecomment-3247732746. Also updates the `Ristretto` ciphersuite to prefer `Blake2b-512` over `SHA2-512`. In order to maintain compliance with FROST's IETF standard, `modular-frost` defines its own ciphersuite for Ristretto which still uses `SHA2-512`.
2025-09-03 12:25:37 -04:00
participants: &[<Ristretto as WrappedGroup>::G],
) -> Self {
let mut res = Self {
db: Some(db.clone()),
2023-04-13 09:47:14 -04:00
genesis,
2023-12-16 20:54:24 -05:00
participants: participants.iter().copied().collect(),
2023-04-13 09:47:14 -04:00
block_number: 0,
2023-04-13 09:47:14 -04:00
tip: genesis,
provided: ProvidedTransactions::new(db.clone(), genesis),
mempool: Mempool::new(db, genesis),
next_block_notifications: VecDeque::new(),
};
if let Some((block_number, tip)) = {
let db = res.db.as_ref().unwrap();
db.get(res.block_number_key()).map(|number| (number, db.get(Self::tip_key(genesis)).unwrap()))
} {
res.block_number = u64::from_le_bytes(block_number.try_into().unwrap());
res.tip.copy_from_slice(&tip);
2023-04-13 09:47:14 -04:00
}
res
2023-04-12 11:13:48 -04:00
}
pub(crate) fn tip(&self) -> [u8; 32] {
2023-04-12 11:13:48 -04:00
self.tip
}
pub(crate) fn block_number(&self) -> u64 {
self.block_number
}
pub(crate) fn block_from_db(db: &D, genesis: [u8; 32], block: &[u8; 32]) -> Option<Block<T>> {
db.get(Self::block_key(&genesis, block))
2023-04-15 00:41:48 -04:00
.map(|bytes| Block::<T>::read::<&[u8]>(&mut bytes.as_ref()).unwrap())
}
pub(crate) fn commit_from_db(db: &D, genesis: [u8; 32], block: &[u8; 32]) -> Option<Vec<u8>> {
db.get(Self::commit_key(&genesis, block))
}
pub(crate) fn block_hash_from_db(db: &D, genesis: [u8; 32], block: u64) -> Option<[u8; 32]> {
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
db.get(Self::block_hash_key(&genesis, block)).map(|h| h.try_into().unwrap())
}
pub(crate) fn commit(&self, block: &[u8; 32]) -> Option<Vec<u8>> {
Self::commit_from_db(self.db.as_ref().unwrap(), self.genesis, block)
}
pub(crate) fn block_hash(&self, block: u64) -> Option<[u8; 32]> {
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
Self::block_hash_from_db(self.db.as_ref().unwrap(), self.genesis, block)
}
pub(crate) fn commit_by_block_number(&self, block: u64) -> Option<Vec<u8>> {
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
self.commit(&self.block_hash(block)?)
}
pub(crate) fn block_after(db: &D, genesis: [u8; 32], block: &[u8; 32]) -> Option<[u8; 32]> {
db.get(Self::block_after_key(&genesis, block)).map(|bytes| bytes.try_into().unwrap())
}
pub(crate) fn locally_provided_txs_in_block(
db: &D,
genesis: &[u8; 32],
block: &[u8; 32],
order: &str,
) -> bool {
let local_key = ProvidedTransactions::<D, T>::locally_provided_quantity_key(genesis, order);
let local = db.get(local_key).map_or(0, |bytes| u32::from_le_bytes(bytes.try_into().unwrap()));
let block_key =
ProvidedTransactions::<D, T>::block_provided_quantity_key(genesis, block, order);
let block = db.get(block_key).map_or(0, |bytes| u32::from_le_bytes(bytes.try_into().unwrap()));
local >= block
}
pub(crate) fn tip_from_db(db: &D, genesis: [u8; 32]) -> [u8; 32] {
db.get(Self::tip_key(genesis)).map_or(genesis, |bytes| bytes.try_into().unwrap())
}
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
pub(crate) fn add_transaction<N: Network>(
&mut self,
internal: bool,
tx: Transaction<T>,
schema: &N::SignatureScheme,
) -> Result<bool, TransactionError> {
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
let db = self.db.as_ref().unwrap();
let genesis = self.genesis;
let commit = |block: u64| -> Option<Commit<N::SignatureScheme>> {
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
let hash = Self::block_hash_from_db(db, genesis, block)?;
// we must have a commit per valid hash
let commit = Self::commit_from_db(db, genesis, &hash).unwrap();
// commit has to be valid if it is coming from our db
Some(Commit::<N::SignatureScheme>::decode(&mut commit.as_ref()).unwrap())
};
let unsigned_in_chain =
|hash: [u8; 32]| db.get(Self::unsigned_included_key(&self.genesis, &hash)).is_some();
2023-12-01 12:09:22 -05:00
self.mempool.add::<N, _>(
|signer, order| {
if self.participants.contains(&signer) {
Some(
db.get(Self::next_nonce_key(&self.genesis, &signer, &order))
.map_or(0, |bytes| u32::from_le_bytes(bytes.try_into().unwrap())),
2023-12-01 12:09:22 -05:00
)
} else {
None
}
},
internal,
tx,
schema,
unsigned_in_chain,
commit,
)
2023-04-13 09:47:14 -04:00
}
pub(crate) fn provide_transaction(&mut self, tx: T) -> Result<(), ProvidedError> {
self.provided.provide(tx)
2023-04-12 11:13:48 -04:00
}
2023-12-01 12:09:22 -05:00
pub(crate) fn next_nonce(
&self,
Smash the singular `Ciphersuite` trait into multiple This helps identify where the various functionalities are used, or rather, not used. The `Ciphersuite` trait present in `patches/ciphersuite`, facilitating the entire FCMP++ tree, only requires the markers _and_ canonical point decoding. I've opened a PR to upstream such a trait into `group` (https://github.com/zkcrypto/group/pull/68). `WrappedGroup` is still justified for as long as `Group::generator` exists. Moving `::generator()` to its own trait, on an independent structure (upstream) would be massively appreciated. @tarcieri also wanted to update from `fn generator()` to `const GENERATOR`, which would encourage further discussion on https://github.com/zkcrypto/group/issues/32 and https://github.com/zkcrypto/group/issues/45, which have been stagnant. The `Id` trait is occasionally used yet really should be first off the chopping block. Finally, `WithPreferredHash` is only actually used around a third of the time, which more than justifies it being a separate trait. --- Updates `dalek_ff_group::Scalar` to directly re-export `curve25519_dalek::Scalar`, as without issue. `dalek_ff_group::RistrettoPoint` also could be replaced with an export of `curve25519_dalek::RistrettoPoint`, yet the coordinator relies on how we implemented `Hash` on it for the hell of it so it isn't worth it at this time. `dalek_ff_group::EdwardsPoint` can't be replaced for an re-export of `curve25519_dalek::SubgroupPoint` as it doesn't implement `zeroize`, `subtle` traits within a released, non-yanked version. Relevance to https://github.com/serai-dex/serai/issues/201 and https://github.com/dalek-cryptography/curve25519-dalek/issues/811#issuecomment-3247732746. Also updates the `Ristretto` ciphersuite to prefer `Blake2b-512` over `SHA2-512`. In order to maintain compliance with FROST's IETF standard, `modular-frost` defines its own ciphersuite for Ristretto which still uses `SHA2-512`.
2025-09-03 12:25:37 -04:00
signer: &<Ristretto as WrappedGroup>::G,
2023-12-01 12:09:22 -05:00
order: &[u8],
) -> Option<u32> {
if let Some(next_nonce) = self.mempool.next_nonce_in_mempool(signer, order.to_vec()) {
return Some(next_nonce);
}
if self.participants.contains(signer) {
Some(
self
.db
.as_ref()
.unwrap()
.get(Self::next_nonce_key(&self.genesis, signer, order))
.map_or(0, |bytes| u32::from_le_bytes(bytes.try_into().unwrap())),
2023-12-01 12:09:22 -05:00
)
} else {
None
}
2023-04-12 11:13:48 -04:00
}
pub(crate) fn build_block<N: Network>(&mut self, schema: &N::SignatureScheme) -> Block<T> {
let block = Block::new(
self.tip,
self.provided.transactions.values().flatten().cloned().collect(),
2023-12-01 12:09:22 -05:00
self.mempool.block(),
);
2023-04-12 11:13:48 -04:00
// build_block should not return invalid blocks
self.verify_block::<N>(&block, schema, false).unwrap();
2023-04-12 11:13:48 -04:00
block
}
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
pub(crate) fn verify_block<N: Network>(
&self,
block: &Block<T>,
schema: &N::SignatureScheme,
allow_non_local_provided: bool,
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
) -> Result<(), BlockError> {
let db = self.db.as_ref().unwrap();
let provided_or_unsigned_in_chain = |hash: [u8; 32]| {
db.get(Self::unsigned_included_key(&self.genesis, &hash)).is_some() ||
db.get(Self::provided_included_key(&self.genesis, &hash)).is_some()
};
let commit = |block: u64| -> Option<Commit<N::SignatureScheme>> {
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
let commit = self.commit_by_block_number(block)?;
// commit has to be valid if it is coming from our db
Some(Commit::<N::SignatureScheme>::decode(&mut commit.as_ref()).unwrap())
};
2023-12-01 12:09:22 -05:00
let mut txn_db = db.clone();
let mut txn = txn_db.txn();
let res = block.verify::<N, _>(
self.genesis,
self.tip,
self.provided.transactions.clone(),
2023-12-01 12:09:22 -05:00
&mut |signer, order| {
if self.participants.contains(signer) {
let key = Self::next_nonce_key(&self.genesis, signer, order);
let next = txn
.get(&key)
.map_or(0, |next_nonce| u32::from_le_bytes(next_nonce.try_into().unwrap()));
2023-12-01 12:09:22 -05:00
txn.put(key, (next + 1).to_le_bytes());
Some(next)
} else {
None
}
},
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
schema,
&commit,
provided_or_unsigned_in_chain,
allow_non_local_provided,
2023-12-01 12:09:22 -05:00
);
// Drop this TXN's changes as we're solely verifying the block
2023-12-01 12:09:22 -05:00
drop(txn);
res
2023-04-12 11:13:48 -04:00
}
/// Add a block.
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
pub(crate) fn add_block<N: Network>(
&mut self,
block: &Block<T>,
commit: Vec<u8>,
schema: &N::SignatureScheme,
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
) -> Result<(), BlockError> {
self.verify_block::<N>(block, schema, true)?;
log::info!(
"adding block {} to tributary {} with {} TXs",
hex::encode(block.hash()),
hex::encode(self.genesis),
block.transactions.len(),
);
// None of the following assertions should be reachable since we verified the block
// Take it from the Option so Rust doesn't consider self as mutably borrowed thanks to the
// existence of the txn
let mut db = self.db.take().unwrap();
let mut txn = db.txn();
2023-04-12 11:13:48 -04:00
self.tip = block.hash();
txn.put(Self::tip_key(self.genesis), self.tip);
self.block_number += 1;
txn.put(self.block_number_key(), self.block_number.to_le_bytes());
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
txn.put(Self::block_hash_key(&self.genesis, self.block_number), self.tip);
txn.put(Self::block_key(&self.genesis, &self.tip), block.serialize());
txn.put(Self::commit_key(&self.genesis, &self.tip), commit);
txn.put(Self::block_after_key(&self.genesis, &block.parent()), block.hash());
2023-04-12 11:13:48 -04:00
for tx in &block.transactions {
match tx.kind() {
TransactionKind::Provided(order) => {
let hash = tx.hash();
self.provided.complete(&mut txn, order, self.tip, hash);
txn.put(Self::provided_included_key(&self.genesis, &hash), []);
2023-04-12 11:13:48 -04:00
}
Slash malevolent validators (#294) * add slash tx * ignore unsigned tx replays * verify that provided evidence is valid * fix clippy + fmt * move application tx handling to another module * partially handle the tendermint txs * fix pr comments * support unsigned app txs * add slash target to the votes * enforce provided, unsigned, signed tx ordering within a block * bug fixes * add unit test for tendermint txs * bug fixes * update tests for tendermint txs * add tx ordering test * tidy up tx ordering test * cargo +nightly fmt * Misc fixes from rebasing * Finish resolving clippy * Remove sha3 from tendermint-machine * Resolve a DoS in SlashEvidence's read Also moves Evidence from Vec<Message> to (Message, Option<Message>). That should meet all requirements while being a bit safer. * Make lazy_static a dev-depend for tributary * Various small tweaks One use of sort was inefficient, sorting unsigned || signed when unsigned was already properly sorted. Given how the unsigned TXs were given a nonce of 0, an unstable sort may swap places with an unsigned TX and a signed TX with a nonce of 0 (leading to a faulty block). The extra protection added here sorts signed, then concats. * Fix Tributary tests I broke, start review on tendermint/tx.rs * Finish reviewing everything outside tests and empty_signature * Remove empty_signature empty_signature led to corrupted local state histories. Unfortunately, the API is only sane with a signature. We now use the actual signature, which risks creating a signature over a malicious message if we have ever have an invariant producing malicious messages. Prior, we only signed the message after the local machine confirmed it was okay per the local view of consensus. This is tolerated/preferred over a corrupt state history since production of such messages is already an invariant. TODOs are added to make handling of this theoretical invariant further robust. * Remove async_sequential for tokio::test There was no competition for resources forcing them to be run sequentially. * Modify block order test to be statistically significant without multiple runs * Clean tests --------- Co-authored-by: Luke Parker <lukeparker5132@gmail.com>
2023-08-21 07:28:23 +03:00
TransactionKind::Unsigned => {
let hash = tx.hash();
// Save as included on chain
txn.put(Self::unsigned_included_key(&self.genesis, &hash), []);
// remove from the mempool
self.mempool.remove(&hash);
}
2023-12-01 12:09:22 -05:00
TransactionKind::Signed(order, Signed { signer, nonce, .. }) => {
let next_nonce = nonce + 1;
txn.put(Self::next_nonce_key(&self.genesis, &signer, &order), next_nonce.to_le_bytes());
self.mempool.remove(&tx.hash());
2023-04-12 11:13:48 -04:00
}
}
}
txn.commit();
self.db = Some(db);
for tx in self.next_block_notifications.drain(..) {
let _ = tx.send(());
}
Ok(())
2023-04-12 11:13:48 -04:00
}
}