Files
serai/crypto/transcript/src/lib.rs

156 lines
5.1 KiB
Rust
Raw Normal View History

2022-09-29 04:47:55 -04:00
#![cfg_attr(docsrs, feature(doc_auto_cfg))]
#![doc = include_str!("../README.md")]
#![no_std]
use zeroize::Zeroize;
Smash the singular `Ciphersuite` trait into multiple This helps identify where the various functionalities are used, or rather, not used. The `Ciphersuite` trait present in `patches/ciphersuite`, facilitating the entire FCMP++ tree, only requires the markers _and_ canonical point decoding. I've opened a PR to upstream such a trait into `group` (https://github.com/zkcrypto/group/pull/68). `WrappedGroup` is still justified for as long as `Group::generator` exists. Moving `::generator()` to its own trait, on an independent structure (upstream) would be massively appreciated. @tarcieri also wanted to update from `fn generator()` to `const GENERATOR`, which would encourage further discussion on https://github.com/zkcrypto/group/issues/32 and https://github.com/zkcrypto/group/issues/45, which have been stagnant. The `Id` trait is occasionally used yet really should be first off the chopping block. Finally, `WithPreferredHash` is only actually used around a third of the time, which more than justifies it being a separate trait. --- Updates `dalek_ff_group::Scalar` to directly re-export `curve25519_dalek::Scalar`, as without issue. `dalek_ff_group::RistrettoPoint` also could be replaced with an export of `curve25519_dalek::RistrettoPoint`, yet the coordinator relies on how we implemented `Hash` on it for the hell of it so it isn't worth it at this time. `dalek_ff_group::EdwardsPoint` can't be replaced for an re-export of `curve25519_dalek::SubgroupPoint` as it doesn't implement `zeroize`, `subtle` traits within a released, non-yanked version. Relevance to https://github.com/serai-dex/serai/issues/201 and https://github.com/dalek-cryptography/curve25519-dalek/issues/811#issuecomment-3247732746. Also updates the `Ristretto` ciphersuite to prefer `Blake2b-512` over `SHA2-512`. In order to maintain compliance with FROST's IETF standard, `modular-frost` defines its own ciphersuite for Ristretto which still uses `SHA2-512`.
2025-09-03 12:25:37 -04:00
use digest::{block_api::BlockSizeUser, Digest, Output, HashMarker};
2023-03-20 20:10:00 -04:00
#[cfg(feature = "merlin")]
mod merlin;
#[cfg(feature = "merlin")]
pub use crate::merlin::MerlinTranscript;
2023-03-20 20:10:00 -04:00
/// Tests for a transcript.
#[cfg(any(test, feature = "tests"))]
pub mod tests;
2023-03-20 20:10:00 -04:00
/// A transcript trait valid over a variety of transcript formats.
2023-03-07 05:30:21 -05:00
pub trait Transcript: Send + Clone {
type Challenge: Send + Sync + Clone + AsRef<[u8]>;
2022-09-29 05:25:29 -04:00
/// Create a new transcript with the specified name.
fn new(name: &'static [u8]) -> Self;
2022-09-29 05:25:29 -04:00
/// Apply a domain separator to the transcript.
fn domain_separate(&mut self, label: &'static [u8]);
2022-07-09 00:37:39 -04:00
2022-09-29 05:25:29 -04:00
/// Append a message to the transcript.
fn append_message<M: AsRef<[u8]>>(&mut self, label: &'static [u8], message: M);
2022-07-09 00:37:39 -04:00
/// Produce a challenge.
///
/// Implementors MUST update the transcript as it does so, preventing the same challenge from
/// being generated multiple times.
fn challenge(&mut self, label: &'static [u8]) -> Self::Challenge;
2022-07-09 00:37:39 -04:00
/// Produce a RNG seed.
///
/// Helper function for parties needing to generate random data from an agreed upon state.
///
/// Implementors MAY internally call the challenge function for the needed bytes, and accordingly
/// produce a transcript conflict between two transcripts, one which called challenge(label) and
/// one which called rng_seed(label) at the same point.
fn rng_seed(&mut self, label: &'static [u8]) -> [u8; 32];
}
#[derive(Clone, Copy)]
enum DigestTranscriptMember {
Name,
Domain,
Label,
Value,
2022-07-15 01:26:07 -04:00
Challenge,
2023-03-07 03:55:58 -05:00
Continued,
Challenged,
}
impl DigestTranscriptMember {
fn as_u8(&self) -> u8 {
match self {
DigestTranscriptMember::Name => 0,
DigestTranscriptMember::Domain => 1,
DigestTranscriptMember::Label => 2,
DigestTranscriptMember::Value => 3,
2022-07-15 01:26:07 -04:00
DigestTranscriptMember::Challenge => 4,
2023-03-07 03:55:58 -05:00
DigestTranscriptMember::Continued => 5,
DigestTranscriptMember::Challenged => 6,
}
}
}
2022-09-29 05:25:29 -04:00
/// A simple transcript format constructed around the specified hash algorithm.
#[derive(Clone, Debug)]
Smash the singular `Ciphersuite` trait into multiple This helps identify where the various functionalities are used, or rather, not used. The `Ciphersuite` trait present in `patches/ciphersuite`, facilitating the entire FCMP++ tree, only requires the markers _and_ canonical point decoding. I've opened a PR to upstream such a trait into `group` (https://github.com/zkcrypto/group/pull/68). `WrappedGroup` is still justified for as long as `Group::generator` exists. Moving `::generator()` to its own trait, on an independent structure (upstream) would be massively appreciated. @tarcieri also wanted to update from `fn generator()` to `const GENERATOR`, which would encourage further discussion on https://github.com/zkcrypto/group/issues/32 and https://github.com/zkcrypto/group/issues/45, which have been stagnant. The `Id` trait is occasionally used yet really should be first off the chopping block. Finally, `WithPreferredHash` is only actually used around a third of the time, which more than justifies it being a separate trait. --- Updates `dalek_ff_group::Scalar` to directly re-export `curve25519_dalek::Scalar`, as without issue. `dalek_ff_group::RistrettoPoint` also could be replaced with an export of `curve25519_dalek::RistrettoPoint`, yet the coordinator relies on how we implemented `Hash` on it for the hell of it so it isn't worth it at this time. `dalek_ff_group::EdwardsPoint` can't be replaced for an re-export of `curve25519_dalek::SubgroupPoint` as it doesn't implement `zeroize`, `subtle` traits within a released, non-yanked version. Relevance to https://github.com/serai-dex/serai/issues/201 and https://github.com/dalek-cryptography/curve25519-dalek/issues/811#issuecomment-3247732746. Also updates the `Ristretto` ciphersuite to prefer `Blake2b-512` over `SHA2-512`. In order to maintain compliance with FROST's IETF standard, `modular-frost` defines its own ciphersuite for Ristretto which still uses `SHA2-512`.
2025-09-03 12:25:37 -04:00
pub struct DigestTranscript<D: Send + Clone + Digest + HashMarker>(D);
Smash the singular `Ciphersuite` trait into multiple This helps identify where the various functionalities are used, or rather, not used. The `Ciphersuite` trait present in `patches/ciphersuite`, facilitating the entire FCMP++ tree, only requires the markers _and_ canonical point decoding. I've opened a PR to upstream such a trait into `group` (https://github.com/zkcrypto/group/pull/68). `WrappedGroup` is still justified for as long as `Group::generator` exists. Moving `::generator()` to its own trait, on an independent structure (upstream) would be massively appreciated. @tarcieri also wanted to update from `fn generator()` to `const GENERATOR`, which would encourage further discussion on https://github.com/zkcrypto/group/issues/32 and https://github.com/zkcrypto/group/issues/45, which have been stagnant. The `Id` trait is occasionally used yet really should be first off the chopping block. Finally, `WithPreferredHash` is only actually used around a third of the time, which more than justifies it being a separate trait. --- Updates `dalek_ff_group::Scalar` to directly re-export `curve25519_dalek::Scalar`, as without issue. `dalek_ff_group::RistrettoPoint` also could be replaced with an export of `curve25519_dalek::RistrettoPoint`, yet the coordinator relies on how we implemented `Hash` on it for the hell of it so it isn't worth it at this time. `dalek_ff_group::EdwardsPoint` can't be replaced for an re-export of `curve25519_dalek::SubgroupPoint` as it doesn't implement `zeroize`, `subtle` traits within a released, non-yanked version. Relevance to https://github.com/serai-dex/serai/issues/201 and https://github.com/dalek-cryptography/curve25519-dalek/issues/811#issuecomment-3247732746. Also updates the `Ristretto` ciphersuite to prefer `Blake2b-512` over `SHA2-512`. In order to maintain compliance with FROST's IETF standard, `modular-frost` defines its own ciphersuite for Ristretto which still uses `SHA2-512`.
2025-09-03 12:25:37 -04:00
impl<D: Send + Clone + Digest + HashMarker> DigestTranscript<D> {
fn append(&mut self, kind: DigestTranscriptMember, value: &[u8]) {
2022-09-04 21:23:38 -04:00
self.0.update([kind.as_u8()]);
// Assumes messages don't exceed 16 exabytes
self.0.update(u64::try_from(value.len()).unwrap().to_le_bytes());
self.0.update(value);
}
}
Smash the singular `Ciphersuite` trait into multiple This helps identify where the various functionalities are used, or rather, not used. The `Ciphersuite` trait present in `patches/ciphersuite`, facilitating the entire FCMP++ tree, only requires the markers _and_ canonical point decoding. I've opened a PR to upstream such a trait into `group` (https://github.com/zkcrypto/group/pull/68). `WrappedGroup` is still justified for as long as `Group::generator` exists. Moving `::generator()` to its own trait, on an independent structure (upstream) would be massively appreciated. @tarcieri also wanted to update from `fn generator()` to `const GENERATOR`, which would encourage further discussion on https://github.com/zkcrypto/group/issues/32 and https://github.com/zkcrypto/group/issues/45, which have been stagnant. The `Id` trait is occasionally used yet really should be first off the chopping block. Finally, `WithPreferredHash` is only actually used around a third of the time, which more than justifies it being a separate trait. --- Updates `dalek_ff_group::Scalar` to directly re-export `curve25519_dalek::Scalar`, as without issue. `dalek_ff_group::RistrettoPoint` also could be replaced with an export of `curve25519_dalek::RistrettoPoint`, yet the coordinator relies on how we implemented `Hash` on it for the hell of it so it isn't worth it at this time. `dalek_ff_group::EdwardsPoint` can't be replaced for an re-export of `curve25519_dalek::SubgroupPoint` as it doesn't implement `zeroize`, `subtle` traits within a released, non-yanked version. Relevance to https://github.com/serai-dex/serai/issues/201 and https://github.com/dalek-cryptography/curve25519-dalek/issues/811#issuecomment-3247732746. Also updates the `Ristretto` ciphersuite to prefer `Blake2b-512` over `SHA2-512`. In order to maintain compliance with FROST's IETF standard, `modular-frost` defines its own ciphersuite for Ristretto which still uses `SHA2-512`.
2025-09-03 12:25:37 -04:00
impl<D: Send + Clone + Digest + HashMarker> Transcript for DigestTranscript<D> {
type Challenge = Output<D>;
fn new(name: &'static [u8]) -> Self {
let mut res = DigestTranscript(D::new());
res.append(DigestTranscriptMember::Name, name);
res
2022-05-06 07:33:08 -04:00
}
fn domain_separate(&mut self, label: &'static [u8]) {
self.append(DigestTranscriptMember::Domain, label);
}
fn append_message<M: AsRef<[u8]>>(&mut self, label: &'static [u8], message: M) {
self.append(DigestTranscriptMember::Label, label);
self.append(DigestTranscriptMember::Value, message.as_ref());
}
fn challenge(&mut self, label: &'static [u8]) -> Self::Challenge {
self.append(DigestTranscriptMember::Challenge, label);
2023-03-07 03:55:58 -05:00
let mut cloned = self.0.clone();
// Explicitly fork these transcripts to prevent length extension attacks from being possible
// (at least, without the additional ability to remove a byte from a finalized hash)
self.0.update([DigestTranscriptMember::Continued.as_u8()]);
cloned.update([DigestTranscriptMember::Challenged.as_u8()]);
cloned.finalize()
}
fn rng_seed(&mut self, label: &'static [u8]) -> [u8; 32] {
let mut seed = [0; 32];
seed.copy_from_slice(&self.challenge(label)[.. 32]);
2022-05-06 07:33:08 -04:00
seed
}
}
// Digest doesn't implement Zeroize
// Implement Zeroize for DigestTranscript by writing twice the block size to the digest in an
// attempt to overwrite the internal hash state/any leftover bytes
Smash the singular `Ciphersuite` trait into multiple This helps identify where the various functionalities are used, or rather, not used. The `Ciphersuite` trait present in `patches/ciphersuite`, facilitating the entire FCMP++ tree, only requires the markers _and_ canonical point decoding. I've opened a PR to upstream such a trait into `group` (https://github.com/zkcrypto/group/pull/68). `WrappedGroup` is still justified for as long as `Group::generator` exists. Moving `::generator()` to its own trait, on an independent structure (upstream) would be massively appreciated. @tarcieri also wanted to update from `fn generator()` to `const GENERATOR`, which would encourage further discussion on https://github.com/zkcrypto/group/issues/32 and https://github.com/zkcrypto/group/issues/45, which have been stagnant. The `Id` trait is occasionally used yet really should be first off the chopping block. Finally, `WithPreferredHash` is only actually used around a third of the time, which more than justifies it being a separate trait. --- Updates `dalek_ff_group::Scalar` to directly re-export `curve25519_dalek::Scalar`, as without issue. `dalek_ff_group::RistrettoPoint` also could be replaced with an export of `curve25519_dalek::RistrettoPoint`, yet the coordinator relies on how we implemented `Hash` on it for the hell of it so it isn't worth it at this time. `dalek_ff_group::EdwardsPoint` can't be replaced for an re-export of `curve25519_dalek::SubgroupPoint` as it doesn't implement `zeroize`, `subtle` traits within a released, non-yanked version. Relevance to https://github.com/serai-dex/serai/issues/201 and https://github.com/dalek-cryptography/curve25519-dalek/issues/811#issuecomment-3247732746. Also updates the `Ristretto` ciphersuite to prefer `Blake2b-512` over `SHA2-512`. In order to maintain compliance with FROST's IETF standard, `modular-frost` defines its own ciphersuite for Ristretto which still uses `SHA2-512`.
2025-09-03 12:25:37 -04:00
impl<D: Send + Clone + Digest + HashMarker> Zeroize for DigestTranscript<D>
where
D: BlockSizeUser,
{
fn zeroize(&mut self) {
// Update in 4-byte chunks to reduce call quantity and enable word-level update optimizations
const WORD_SIZE: usize = 4;
// block_size returns the block_size in bytes
// Use a ceil div in case the block size isn't evenly divisible by our word size
let words = D::block_size().div_ceil(WORD_SIZE);
for _ in 0 .. (2 * words) {
self.0.update([255; WORD_SIZE]);
}
// Hopefully, the hash state is now overwritten to the point no data is recoverable
// These writes may be optimized out if they're never read
// Attempt to get them marked as read
Smash the singular `Ciphersuite` trait into multiple This helps identify where the various functionalities are used, or rather, not used. The `Ciphersuite` trait present in `patches/ciphersuite`, facilitating the entire FCMP++ tree, only requires the markers _and_ canonical point decoding. I've opened a PR to upstream such a trait into `group` (https://github.com/zkcrypto/group/pull/68). `WrappedGroup` is still justified for as long as `Group::generator` exists. Moving `::generator()` to its own trait, on an independent structure (upstream) would be massively appreciated. @tarcieri also wanted to update from `fn generator()` to `const GENERATOR`, which would encourage further discussion on https://github.com/zkcrypto/group/issues/32 and https://github.com/zkcrypto/group/issues/45, which have been stagnant. The `Id` trait is occasionally used yet really should be first off the chopping block. Finally, `WithPreferredHash` is only actually used around a third of the time, which more than justifies it being a separate trait. --- Updates `dalek_ff_group::Scalar` to directly re-export `curve25519_dalek::Scalar`, as without issue. `dalek_ff_group::RistrettoPoint` also could be replaced with an export of `curve25519_dalek::RistrettoPoint`, yet the coordinator relies on how we implemented `Hash` on it for the hell of it so it isn't worth it at this time. `dalek_ff_group::EdwardsPoint` can't be replaced for an re-export of `curve25519_dalek::SubgroupPoint` as it doesn't implement `zeroize`, `subtle` traits within a released, non-yanked version. Relevance to https://github.com/serai-dex/serai/issues/201 and https://github.com/dalek-cryptography/curve25519-dalek/issues/811#issuecomment-3247732746. Also updates the `Ristretto` ciphersuite to prefer `Blake2b-512` over `SHA2-512`. In order to maintain compliance with FROST's IETF standard, `modular-frost` defines its own ciphersuite for Ristretto which still uses `SHA2-512`.
2025-09-03 12:25:37 -04:00
fn mark_read<D: Send + Clone + Digest + HashMarker>(transcript: &DigestTranscript<D>) {
// Just get a challenge from the state
let mut challenge = core::hint::black_box(transcript.0.clone().finalize());
challenge.as_mut().zeroize();
}
mark_read(self)
}
}
2023-03-20 20:10:00 -04:00
/// The recommended transcript, guaranteed to be secure against length-extension attacks.
#[cfg(feature = "recommended")]
pub type RecommendedTranscript = DigestTranscript<blake2::Blake2b512>;