Make Coordinator aware of transcript size

[cygnusv]

Type of PR:

• Bugfix
• Feature
• Documentation
• Other

Required reviews:

• 1
• 2
• 3

What this does:

• Transcript size can now be deterministically computed from DKG size and threshold. Coordinator validates this accordingly. This is a tiny step in the direction of on-chain validation.
• Add calldata dynamic costs as part of the gas reimbursement calculation. Note that since we validate transcript sizes, this mechanism can't be abused by posting huge transcripts.

Issues fixed/closed:

• Closes Static gas cost for ReimbursementPool doesn't include calldata cost #328

Why it's needed:

• Related to DKG cost reimbursement for TACo operators #219

Notes for reviewers:

• These changes have been tested on Lynx successfully.
• Some napkin math for postTranscript TXs in testnet DKG rituals (2-of-3):
  • Before: reimbursements were ~13500 gas short
  • After: reimbursements are ~5000 gas short
  • The difference is ~8500 gas, which is according to the expected, since the max gas for the calldata cost predicted by the formula in this PR is 8768 gas ( 8768 = (644-128)*16 + 128*4, where 644 is the calldata size for a postTranscript TX in a 2-of-3 ritual)
  • The remaining 5000 gas that are still missing is something that can be investigated further, especially using mainnet data

[KPrasch]

Just a nitpick but 40 is the only unnamed component of this calculation. Perhaps it can be assigned to a constant similar to the BLS point sizes?

[cygnusv]

You're right, but the thing is that I don't understand well enough where that 40 comes from. I mean, it's a fixed overhead that comes from rust serialization but I don't have a full explanation for it. I'll update this when I have more info.

[derekpierre]

🎸 .

Did you also want to verify your updated Coordinator contract on lynx, with our updated verify script? 😉

[derekpierre]

Thanks for this cleanup. We were repeating the RitualState enum and fixtures in a number of places.