Fix AVX2 docs formatting, remove obsolete AVX512 note

This commit is contained in:
Henry de Valence 2018-03-22 12:13:39 -07:00
parent d67e895619
commit 844da9712b

View file

@ -463,18 +463,12 @@
//! `ymm16..ymm31` registers from AVX512VL), and of approximately 1.0x
//! for Ryzen (which implements AVX2 at half rate).
//!
//! When used for variable-time double-base scalar multiplication \\( aA
//! + bB \\) for fixed \\(B\\) (as in, e.g., signature verification),
//! When used for variable-time double-base scalar multiplication
//! \\( aA + bB \\) for fixed \\(B\\) (as in, e.g., signature verification),
//! this strategy provides a 1.4x speedup on Skylake-X over the same
//! operation as implemented in `ed25519-donna`, the fastest
//! production-quality Ed25519 implementation.
//!
//! (Note: since testing this, the experimental `llvm50` Rust branch
//! used to compile the experimental `stdsimd` intrinsics have fallen
//! out of sync and it is no longer possible to compile for
//! `skylake-avx512`. This is why all of this branch is part of the
//! `yolocrypto` feature, pending upstream work.)
//!
//! [sandy2x]: https://eprint.iacr.org/2015/943.pdf
//! [avx2trac]: https://trac.torproject.org/projects/tor/ticket/8897#comment:28
//! [hwcd08]: https://www.iacr.org/archive/asiacrypt2008/53500329/53500329.pdf