From f6a41d983623c489ed3d0230adf63ab0a9abea46 Mon Sep 17 00:00:00 2001 From: Luke Parker Date: Sat, 28 May 2022 21:06:25 -0400 Subject: [PATCH] Clarify FROST's test-provided secp256k1 curve definition --- crypto/frost/src/tests/literal/secp256k1.rs | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/crypto/frost/src/tests/literal/secp256k1.rs b/crypto/frost/src/tests/literal/secp256k1.rs index c1f9223a..36ccb833 100644 --- a/crypto/frost/src/tests/literal/secp256k1.rs +++ b/crypto/frost/src/tests/literal/secp256k1.rs @@ -44,6 +44,10 @@ impl Curve for Secp256k1 { // The IETF draft doesn't specify a secp256k1 ciphersuite // This test just uses the simplest ciphersuite which would still be viable to deploy + // The comparable P-256 curve uses hash_to_field from the Hash To Curve IETF draft with a context + // string and further DST for H1 ("rho") and H3 ("digest"). It's not currently worth it to add + // that weight, yet if secp256k1 is ever officially acknowledged (not just a testing curve), it + // must be properly implemented. fn hash_msg(msg: &[u8]) -> Vec { (&Sha256::digest(msg)).to_vec() }