xmrswap: Remove C deps for signature encryption and dleq

[martonp]

Adds functions for encrypting / decrypting Schnorr signatures, and uses the athanorlabs/go-dleq library for discrete log equivalence.

[davecgh]

Glad to see this! I did notice the dependency has various inefficiencies looking through it, but given that it won't be used in any type of hot paths, that is unlikely to matter much.

[JoeGruffins]

Wow.

closes #2918

[martonp]

@davecgh Thanks for the review!

[martonp]

@JoeGruffins since these adaptor signatures use schnorr signatures rather than ecdsa as was in the C library, I've updated the tx scripts to use OP_CHECKSIGALT instead of OP_CHECKMULTISIG.

[davecgh]

I haven't looked at the DLEQ dep for correctness, but aside from the last nit in my most recent comment, the adaptor sig code looks accurate and reasonably efficient.

[davecgh]

Since probably nothing other than the internal code will use it, I'm sure you guys probably aren't worried about it, but I'd point out that with no function comment and variables named kal and kaf, it's pretty unclear what they are here without looking at the code to determine they're Schnorr sigs over secp256k1.

[martonp]

I think @JoeGruffins will do some refactor regarding variable names throughout after this one is merged.

[davecgh]

Sounds good.

[martonp]

I haven't looked at the DLEQ dep for correctness, but aside from the last nit in my most recent comment, the adaptor sig code looks accurate and reasonably efficient.

OK thanks for the reviews! I think DLEQ dep will have to be audited before we release anything, but I think it is fine for this POC currently.

[davecgh]

I followed up a bit on this comment, but it's resolved, so I figured I'd post a link to it here so it doesn't get missed.

[JoeGruffins]

@martonp is on vacay, we'll pick up when they're back

[JoeGruffins]

@martonp srry, conflicts need rebase