Refactor of the solidity Groth16 verifier

[markspanbroek]

Refactors the Groth16 verifier to:

• fix mistakes
• remove unnecessary conversions
• make it more readable
• return false instead of stopping execution when an operation fails

[markspanbroek]

Honestly I have very limited understanding of what is going on here 😅 Maybe @bkomuves could do proper review? If not then I will dive in, but it will take me quite some time to understand it what is going on and what has changed.

That's ok, nobody in our team knows everything about zero knowledge proofs and solidity. That's why I asked multiple people to review. Balazs can review the crypto part, and you and Eric can look for solidity mistakes.

Things that you could focus a review on:

• were the refactor steps done properly (each one is its own commit, so they can be viewed separately)
• looking at the end result, does anything stand out w.r.t solidty types/visibility/etc..
• are the assembly calls to the precompiled contracts done correctly, and do they specify the correct number of input and output bytes

For reference, this is what I used to look up the precompiled contracts: https://www.evm.codes/precompiled

And this is what I used as a reference for the STATICCALL opcode: https://www.evm.codes/#fa?fork=shanghai

[markspanbroek]

I would add a comment explaining that the 3 precompiles are all supposed to check for the validity of the group elements, with a link to a definitive source.

Good idea, I added the comments.

I can also imagine that with many different EVM implementations, some may forget to actually do the checks, but I guess that's not our problem? :)

That would lead to a consensus failure, and seems unlikely. The EIPs state that these checks should be done, and looking at the Geth code, it seems to do these checks.

[markspanbroek]

I think that this is ready to be merged. @AuHau or @emizzle, could you do a final review to check for any Solidity shenanigans? Then I can merge this :pray: