Arkworks integration and ed25519 upgrade

[annenkov]

Purpose

This PR adds generic adapters (as blanket implementation) to the ArkWorks interfaces. The adapters (with some additional configuration traits) are used to replace the BLS12-381 curve with ArkWorks implementation. In addition to that, ed25519-dalek is upgraded to v2.0, which was previously blocked by the pairing crate and the corresponding BLS implementation (new versions don't expose the target group element type Fp12, preventing from serializing it, see details here).

Changes

• implement ArkGroup<G> and ArkField<F> wrappers;
• blanket implementations of the Field and PrimeField traits for ArkField<F>;
• blanket implementation of the Curve trait ArkGroup<G>;
• all required trait implementations for the ArkWorks BLS12-381 curve, that replaces the current BLS curve implementation from the pairing crate.
• upgrade ed25519-dalek signature scheme; the v2.0 API is a bit different, it bundles a secret and verifying (public) key into a signing key and maintains an invariant that the verifying key is always the public key of secret_key; signing does not require passing the verifying key, so it was removed from parameters of the corresponding functions.
• update FFI function types on the Haskell side to match the changes in the signing API (removed verifing_key parameter)

Checklist

• My code follows the style of this project.
• The code compiles without warnings.
• I have performed a self-review of the changes.
• I have documented my code, in particular the intent of the
  hard-to-understand areas.
• I have updated the CHANGELOG.

[hamiidreza]

Looks good. I added some suggestions regarding the broken link for the VRF RFC draft version.

[abizjak]

@annenkov did you propagate this PR to concordium-client & the node?

[annenkov]

@annenkov did you propagate this PR to concordium-client & the node?

Yes, I checked concordium-client and concordium-node can be built with this branch. There are only changes in the signatures of functions used internally.
However, it would make sense to change signatures of the sign function(s). I created an issue for this #501