👋 this is not an issue, but wanted to highlight some ideas @Rexicon226 and I had around function inlining.
For context, we've been using s2n-bignum inside the Firedancer project for a while now (p256, k256, soon ed25519 and more).
The current "problem": each function we get from s2n-bignum does a fn call, there's no way for us to inline them, so small functions are harder to use and compose because of the overhead they incur.
We experimented a bit with what compilers support and came up with this way to write the code:
firedancer-io/firedancer#9235
👋 this is not an issue, but wanted to highlight some ideas @Rexicon226 and I had around function inlining.
For context, we've been using s2n-bignum inside the Firedancer project for a while now (p256, k256, soon ed25519 and more).
The current "problem": each function we get from s2n-bignum does a fn call, there's no way for us to inline them, so small functions are harder to use and compose because of the overhead they incur.
We experimented a bit with what compilers support and came up with this way to write the code:
firedancer-io/firedancer#9235