Our `ecdsa` crate prevents vulnerabilities like Java's CVE-2022-21449 using type safety. We use a `NonZeroScalar` type to model the `r` and `s` components of an ECDSA signature. This prevents either of them from ever accidentally being zero. docs.rs/ecdsa/0.13.4/ecdsa/s…

Announcing initial hash2curve support in the following @RustCryptoOrg elliptic curve crates: - `k256` v0.10.2 (secp256k1) - `p256` v0.10.1 (NIST P-256) Repo: github.com/RustCrypto/ellipt… See trait documentation here: docs.rs/elliptic-curve/0.11.…

Updated ECDSA/secp256k1 benchmarks comparing: - @RustCryptoOrg's pure Rust `k256` crate - The pure Rust `libsecp256k1` crate - The `secp256k1` crate (FFI wrapper for the bitcoin-core C library) Faster is better.

We've just released v0.7.2 of the `aes` crate: crates.io/crates/aes This release includes an experimental new nightly-only backend for the ARMv8 Cryptography Extensions, gated under the `armv8` feature. Benchmarks on an Apple M1 show nearly 10GB/s throughput for AES-128:

Benchmarking a PR to the `aes` crate to add ARMv8 intrinsics support (using stdarch and therefore nightly-only for now) on an Apple M1: github.com/RustCrypto/block-… ARMv8 intrinsics vs portable "soft" backend shows ~16-35X speedup:

`sha2` v0.9.4 is out: crates.io/crates/sha2/0.9.4 This release includes optional Apple M4 hardware acceleration gated under the `asm` crate feature, resulting in a ~4X speedup over the pure Rust implementation. Before and after benchmarks:

The `aes-gcm` crate has been downloaded over 1 million times! 🎉 crates.io/crates/aes-gcm

With the latest v0.6.2 release, the @rustlang `aes-soft` crate now provides what we believe is the fastest safe HLL software AES implementation which uses only constant-time bitwise operations (no LUTs) and doesn't require SIMD CPU intrinsics: github.com/RustCrypto/block-…

The `chacha20poly1305` (AEAD cipher) crate has been downloaded over 250,000 times! crates.io/crates/chacha20pol…

Fixslicing represents the state-of-the-art for constant-time implementations of AES in software. You can read more about it here: eprint.iacr.org/2020/1123.pd… We also plan to add optional thumbv7 assembly for @rustembedded, and potentially a 64-bit backend that doubles the speed.