Files
serai/coordinator/src/tributary/handle.rs

618 lines
18 KiB
Rust
Raw Normal View History

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 core::{ops::Deref, future::Future};
use std::collections::HashMap;
use zeroize::Zeroizing;
use rand_core::SeedableRng;
use rand_chacha::ChaCha20Rng;
use transcript::{Transcript, RecommendedTranscript};
use ciphersuite::{Ciphersuite, Ristretto};
use frost::{
FrostError,
dkg::{Participant, musig::musig},
sign::*,
};
use frost_schnorrkel::Schnorrkel;
use serai_client::{
Signature,
primitives::NetworkId,
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
validator_sets::primitives::{ValidatorSet, KeyPair, musig_context, set_keys_message},
subxt::utils::Encoded,
Serai,
};
use tributary::Signed;
use processor_messages::{
CoordinatorMessage, coordinator,
key_gen::{self, KeyGenId},
sign::{self, SignId},
};
use serai_db::{Get, Db};
use crate::processors::Processors;
use super::{Transaction, TributarySpec, TributaryDb, scanner::RecognizedIdType};
const DKG_CONFIRMATION_NONCES: &[u8] = b"dkg_confirmation_nonces";
const DKG_CONFIRMATION_SHARES: &[u8] = b"dkg_confirmation_shares";
// Instead of maintaing state, this simply re-creates the machine(s) in-full on every call.
// This simplifies data flow and prevents requiring multiple paths.
// While more expensive, this only runs an O(n) algorithm, which is tolerable to run multiple
// times.
struct DkgConfirmer;
impl DkgConfirmer {
fn preprocess_internal(
spec: &TributarySpec,
key: &Zeroizing<<Ristretto as Ciphersuite>::F>,
attempt: u32,
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
) -> (AlgorithmSignMachine<Ristretto, Schnorrkel>, [u8; 64]) {
// TODO: Does Substrate already have a validator-uniqueness check?
let validators = spec.validators().iter().map(|val| val.0).collect::<Vec<_>>();
let context = musig_context(spec.set());
let mut chacha = ChaCha20Rng::from_seed({
let mut entropy_transcript = RecommendedTranscript::new(b"DkgConfirmer Entropy");
entropy_transcript.append_message(b"spec", spec.serialize());
entropy_transcript.append_message(b"key", Zeroizing::new(key.to_bytes()));
entropy_transcript.append_message(b"attempt", attempt.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
Zeroizing::new(entropy_transcript).rng_seed(b"preprocess")
});
let (machine, preprocess) = AlgorithmMachine::new(
Schnorrkel::new(b"substrate"),
musig(&context, key, &validators)
.expect("confirming the DKG for a set we aren't in/validator present multiple times")
.into(),
)
.preprocess(&mut chacha);
(machine, preprocess.serialize().try_into().unwrap())
}
// Get the preprocess for this confirmation.
fn preprocess(
spec: &TributarySpec,
key: &Zeroizing<<Ristretto as Ciphersuite>::F>,
attempt: u32,
) -> [u8; 64] {
Self::preprocess_internal(spec, key, attempt).1
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 share_internal(
spec: &TributarySpec,
key: &Zeroizing<<Ristretto as Ciphersuite>::F>,
attempt: u32,
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
preprocesses: HashMap<Participant, Vec<u8>>,
key_pair: &KeyPair,
) -> Result<(AlgorithmSignatureMachine<Ristretto, Schnorrkel>, [u8; 32]), Participant> {
let machine = Self::preprocess_internal(spec, key, attempt).0;
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 preprocesses = preprocesses
.into_iter()
.map(|(p, preprocess)| {
machine
.read_preprocess(&mut preprocess.as_slice())
.map(|preprocess| (p, preprocess))
.map_err(|_| p)
})
.collect::<Result<HashMap<_, _>, _>>()?;
let (machine, share) = machine
.sign(preprocesses, &set_keys_message(&spec.set(), key_pair))
.map_err(|e| match e {
FrostError::InternalError(e) => unreachable!("FrostError::InternalError {e}"),
FrostError::InvalidParticipant(_, _) |
FrostError::InvalidSigningSet(_) |
FrostError::InvalidParticipantQuantity(_, _) |
FrostError::DuplicatedParticipant(_) |
FrostError::MissingParticipant(_) => unreachable!("{e:?}"),
FrostError::InvalidPreprocess(p) | FrostError::InvalidShare(p) => p,
})?;
Ok((machine, share.serialize().try_into().unwrap()))
}
// Get the share for this confirmation, if the preprocesses are valid.
fn share(
spec: &TributarySpec,
key: &Zeroizing<<Ristretto as Ciphersuite>::F>,
attempt: u32,
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
preprocesses: HashMap<Participant, Vec<u8>>,
key_pair: &KeyPair,
) -> Result<[u8; 32], Participant> {
Self::share_internal(spec, key, attempt, preprocesses, key_pair).map(|(_, share)| share)
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 complete(
spec: &TributarySpec,
key: &Zeroizing<<Ristretto as Ciphersuite>::F>,
attempt: u32,
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
preprocesses: HashMap<Participant, Vec<u8>>,
key_pair: &KeyPair,
shares: HashMap<Participant, Vec<u8>>,
) -> Result<[u8; 64], Participant> {
let machine = Self::share_internal(spec, key, attempt, preprocesses, key_pair)
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
.expect("trying to complete a machine which failed to preprocess")
.0;
let shares = shares
.into_iter()
.map(|(p, share)| {
machine.read_share(&mut share.as_slice()).map(|share| (p, share)).map_err(|_| p)
})
.collect::<Result<HashMap<_, _>, _>>()?;
let signature = machine.complete(shares).map_err(|e| match e {
FrostError::InternalError(e) => unreachable!("FrostError::InternalError {e}"),
FrostError::InvalidParticipant(_, _) |
FrostError::InvalidSigningSet(_) |
FrostError::InvalidParticipantQuantity(_, _) |
FrostError::DuplicatedParticipant(_) |
FrostError::MissingParticipant(_) => unreachable!("{e:?}"),
FrostError::InvalidPreprocess(p) | FrostError::InvalidShare(p) => p,
})?;
Ok(signature.to_bytes())
}
}
#[allow(clippy::too_many_arguments)] // TODO
fn read_known_to_exist_data<D: Db, G: Get>(
getter: &G,
spec: &TributarySpec,
key: &Zeroizing<<Ristretto as Ciphersuite>::F>,
label: &'static [u8],
id: [u8; 32],
needed: u16,
attempt: u32,
bytes: Vec<u8>,
signed: Option<&Signed>,
) -> Option<HashMap<Participant, 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
let mut data = HashMap::new();
for validator in spec.validators().iter().map(|validator| validator.0) {
data.insert(
spec.i(validator).unwrap(),
if Some(&validator) == signed.map(|signed| &signed.signer) {
bytes.clone()
} else if let Some(data) =
TributaryDb::<D>::data(label, getter, spec.genesis(), id, attempt, validator)
{
data
} else {
continue;
},
);
}
assert_eq!(data.len(), usize::from(needed));
// Remove our own piece of data, if we were involved
if data
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
.remove(
&spec
.i(Ristretto::generator() * key.deref())
.expect("handling a message for a Tributary we aren't part of"),
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
)
.is_some()
{
Some(data)
} 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
}
pub fn dkg_confirmation_nonces(
key: &Zeroizing<<Ristretto as Ciphersuite>::F>,
spec: &TributarySpec,
attempt: u32,
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
) -> [u8; 64] {
DkgConfirmer::preprocess(spec, key, attempt)
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
}
#[allow(clippy::needless_pass_by_ref_mut)]
pub fn generated_key_pair<D: Db>(
txn: &mut D::Transaction<'_>,
key: &Zeroizing<<Ristretto as Ciphersuite>::F>,
spec: &TributarySpec,
key_pair: &KeyPair,
attempt: u32,
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<[u8; 32], Participant> {
TributaryDb::<D>::save_currently_completing_key_pair(txn, spec.genesis(), key_pair);
let Some(preprocesses) = read_known_to_exist_data::<D, _>(
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,
spec,
key,
DKG_CONFIRMATION_NONCES,
[0; 32],
spec.n(),
attempt,
vec![],
None,
) else {
panic!("wasn't a participant in confirming a key pair");
};
DkgConfirmer::share(spec, key, attempt, preprocesses, key_pair)
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
}
#[allow(clippy::too_many_arguments)] // TODO
pub async fn handle_application_tx<
D: Db,
Pro: Processors,
FPst: Future<Output = ()>,
PST: Clone + Fn(ValidatorSet, Encoded) -> FPst,
FRid: Future<Output = ()>,
RID: Clone + Fn(NetworkId, [u8; 32], RecognizedIdType, [u8; 32]) -> FRid,
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
>(
tx: Transaction,
spec: &TributarySpec,
processors: &Pro,
publish_serai_tx: PST,
genesis: [u8; 32],
key: &Zeroizing<<Ristretto as Ciphersuite>::F>,
recognized_id: RID,
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: &mut <D as Db>::Transaction<'_>,
) {
// Used to determine if an ID is acceptable
#[derive(Clone, Copy, PartialEq, Eq, Debug)]
enum Zone {
Dkg,
Batch,
Sign,
}
impl Zone {
fn label(&self) -> &'static str {
match self {
Zone::Dkg => {
panic!("getting the label for dkg despite dkg code paths not needing a label")
}
Zone::Batch => "batch",
Zone::Sign => "sign",
}
}
}
let handle =
|txn: &mut _, zone: Zone, label, needed, id, attempt, bytes: Vec<u8>, signed: &Signed| {
if zone == Zone::Dkg {
// Since Dkg doesn't have an ID, solely attempts, this should just be [0; 32]
assert_eq!(id, [0; 32], "DKG, which shouldn't have IDs, had a non-0 ID");
} else if !TributaryDb::<D>::recognized_id(txn, zone.label(), genesis, id) {
// TODO: Full slash
todo!();
}
// If they've already published a TX for this attempt, slash
if let Some(data) = TributaryDb::<D>::data(label, txn, genesis, id, attempt, signed.signer) {
if data != bytes {
// TODO: Full slash
todo!();
}
// TODO: Slash
return None;
}
// If the attempt is lesser than the blockchain's, slash
let curr_attempt = TributaryDb::<D>::attempt(txn, genesis, id);
if attempt < curr_attempt {
// TODO: Slash for being late
return None;
}
if attempt > curr_attempt {
// TODO: Full slash
todo!();
}
// TODO: We can also full slash if shares before all commitments, or share before the
// necessary preprocesses
// TODO: If this is shares, we need to check they are part of the selected signing set
// Store this data
let received =
TributaryDb::<D>::set_data(label, txn, genesis, id, attempt, signed.signer, &bytes);
// If we have all the needed commitments/preprocesses/shares, tell the processor
// TODO: This needs to be coded by weight, not by validator count
if received == needed {
return Some(read_known_to_exist_data::<D, _>(
txn,
spec,
key,
label,
id,
needed,
attempt,
bytes,
Some(signed),
));
}
None
};
match tx {
Transaction::DkgCommitments(attempt, bytes, signed) => {
match handle(txn, Zone::Dkg, b"dkg_commitments", spec.n(), [0; 32], attempt, bytes, &signed) {
Some(Some(commitments)) => {
log::info!("got all DkgCommitments for {}", hex::encode(genesis));
processors
.send(
spec.set().network,
CoordinatorMessage::KeyGen(key_gen::CoordinatorMessage::Commitments {
id: KeyGenId { set: spec.set(), attempt },
commitments,
}),
)
.await;
}
Some(None) => panic!("wasn't a participant in DKG commitments"),
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
}
}
Transaction::DkgShares { attempt, mut shares, confirmation_nonces, signed } => {
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
if shares.len() != (usize::from(spec.n()) - 1) {
// TODO: Full slash
todo!();
}
let sender_i = spec
.i(signed.signer)
.expect("transaction added to tributary by signer who isn't a participant");
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
// Only save our share's bytes
let our_i = spec
.i(Ristretto::generator() * key.deref())
.expect("in a tributary we're not a validator for");
let bytes = if sender_i == our_i {
vec![]
} else {
// 1-indexed to 0-indexed, handling the omission of the sender's own data
let relative_i = usize::from(u16::from(our_i) - 1) -
(if u16::from(our_i) > u16::from(sender_i) { 1 } else { 0 });
// Safe since we length-checked shares
shares.swap_remove(relative_i)
};
drop(shares);
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 confirmation_nonces = handle(
txn,
Zone::Dkg,
DKG_CONFIRMATION_NONCES,
spec.n(),
[0; 32],
attempt,
confirmation_nonces.to_vec(),
&signed,
);
match handle(txn, Zone::Dkg, b"dkg_shares", spec.n(), [0; 32], attempt, bytes, &signed) {
Some(Some(shares)) => {
log::info!("got all DkgShares for {}", hex::encode(genesis));
assert!(confirmation_nonces.is_some());
processors
.send(
spec.set().network,
CoordinatorMessage::KeyGen(key_gen::CoordinatorMessage::Shares {
id: KeyGenId { set: spec.set(), attempt },
shares,
}),
)
.await;
}
Some(None) => panic!("wasn't a participant in DKG shares"),
None => assert!(confirmation_nonces.is_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
}
}
Transaction::DkgConfirmed(attempt, shares, signed) => {
match handle(
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,
Zone::Dkg,
DKG_CONFIRMATION_SHARES,
spec.n(),
[0; 32],
attempt,
shares.to_vec(),
&signed,
) {
Some(Some(shares)) => {
log::info!("got all DkgConfirmed for {}", hex::encode(genesis));
let Some(preprocesses) = read_known_to_exist_data::<D, _>(
txn,
spec,
key,
DKG_CONFIRMATION_NONCES,
[0; 32],
spec.n(),
attempt,
vec![],
None,
) else {
panic!("wasn't a participant in DKG confirmation nonces");
};
let key_pair = TributaryDb::<D>::currently_completing_key_pair(txn, genesis)
.unwrap_or_else(|| {
panic!(
"in DkgConfirmed handling, which happens after everyone {}",
"(including us) fires DkgConfirmed, yet no confirming key pair"
)
});
let Ok(sig) = DkgConfirmer::complete(spec, key, attempt, preprocesses, &key_pair, shares)
else {
// TODO: Full slash
todo!();
};
publish_serai_tx(
spec.set(),
Serai::set_validator_set_keys(spec.set().network, key_pair, Signature(sig)),
)
.await;
}
Some(None) => panic!("wasn't a participant in DKG confirmination shares"),
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
}
}
Transaction::Batch(_, batch) => {
// Because this Batch has achieved synchrony, its batch ID should be authorized
TributaryDb::<D>::recognize_id(txn, Zone::Batch.label(), genesis, batch);
recognized_id(spec.set().network, genesis, RecognizedIdType::Batch, batch).await;
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
}
Transaction::SubstrateBlock(block) => {
let plan_ids = TributaryDb::<D>::plan_ids(txn, genesis, block).expect(
"synced a tributary block finalizing a substrate block in a provided transaction \
despite us not providing that transaction",
);
for id in plan_ids {
TributaryDb::<D>::recognize_id(txn, Zone::Sign.label(), genesis, id);
recognized_id(spec.set().network, genesis, RecognizedIdType::Plan, id).await;
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
}
}
Transaction::BatchPreprocess(data) => {
match handle(
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,
Zone::Batch,
b"batch_preprocess",
spec.t(),
data.plan,
data.attempt,
data.data,
&data.signed,
) {
Some(Some(preprocesses)) => {
processors
.send(
spec.set().network,
CoordinatorMessage::Coordinator(coordinator::CoordinatorMessage::BatchPreprocesses {
id: SignId { key: vec![], id: data.plan, attempt: data.attempt },
preprocesses,
}),
)
.await;
}
Some(None) => {}
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
}
}
Transaction::BatchShare(data) => {
match handle(
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,
Zone::Batch,
b"batch_share",
spec.t(),
data.plan,
data.attempt,
data.data,
&data.signed,
) {
Some(Some(shares)) => {
processors
.send(
spec.set().network,
CoordinatorMessage::Coordinator(coordinator::CoordinatorMessage::BatchShares {
id: SignId { key: vec![], id: data.plan, attempt: data.attempt },
shares: shares
.into_iter()
.map(|(validator, share)| (validator, share.try_into().unwrap()))
.collect(),
}),
)
.await;
}
Some(None) => {}
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
}
}
Transaction::SignPreprocess(data) => {
let key_pair = TributaryDb::<D>::key_pair(txn, spec.set());
match handle(
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,
Zone::Sign,
b"sign_preprocess",
spec.t(),
data.plan,
data.attempt,
data.data,
&data.signed,
) {
Some(Some(preprocesses)) => {
processors
.send(
spec.set().network,
CoordinatorMessage::Sign(sign::CoordinatorMessage::Preprocesses {
id: SignId {
key: key_pair
.expect("completed SignPreprocess despite not setting the key pair")
.1
.into(),
id: data.plan,
attempt: data.attempt,
},
preprocesses,
}),
)
.await;
}
Some(None) => {}
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
}
}
Transaction::SignShare(data) => {
let key_pair = TributaryDb::<D>::key_pair(txn, spec.set());
match handle(
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,
Zone::Sign,
b"sign_share",
spec.t(),
data.plan,
data.attempt,
data.data,
&data.signed,
) {
Some(Some(shares)) => {
processors
.send(
spec.set().network,
CoordinatorMessage::Sign(sign::CoordinatorMessage::Shares {
id: SignId {
key: key_pair
.expect("completed SignShares despite not setting the key pair")
.1
.into(),
id: data.plan,
attempt: data.attempt,
},
shares,
}),
)
.await;
}
Some(None) => {}
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
}
}
Transaction::SignCompleted { plan, tx_hash, .. } => {
// TODO: Confirm this is a valid plan ID
// TODO: Confirm this signer hasn't prior published a completion
let Some(key_pair) = TributaryDb::<D>::key_pair(txn, spec.set()) else { todo!() };
processors
.send(
spec.set().network,
CoordinatorMessage::Sign(sign::CoordinatorMessage::Completed {
key: key_pair.1.to_vec(),
id: plan,
tx: tx_hash,
}),
)
.await;
}
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
}
}