aztecprotocol / aztec3-l1-contracts Goto Github PK

Mikes Mega Diagram: https://miro.com/app/board/uXjVPlafJWM=/

The diagram have been updated after the L2 block decoder was written, meaning that the two are no longer aligned.

Part of https://github.com/AztecProtocol/aztec3-milestones/issues/18

We need to demonstrate to the L1 Rollup Contract that the gigantic bytes
of calldata actually relate to a single hash which has been submitted 
(to save on on-chain snark verification costs). 
The only way to do that (without eip-4844) is to re-hash the data on-chain. 
Sha256 is a cheap hash on-chain. 
But this isn't great - it means we're hashing everything twice in the circuits; 
with two different hashes.

Build the root of a merkle tree where the bottom leafs are computed as the sha256 of the nullifiers from 2 kernels. Notice that the specific number fo nullifiers can change, initially we expect it to be 4 per kernel, so padding might be fine for now, depends on what Mike thinks. The rest of the tree is just sha256(childA, childB).

Part of https://github.com/AztecProtocol/aztec3-milestones/issues/18

Build the root of the merkle three where the bottom leafs are computed as the sha256 of 2 contract deployments each with the tuple (az_address, eth_address) as the data.

Part of https://github.com/AztecProtocol/aztec3-milestones/issues/18

Build similar to #2 just with the leaves instead.

Part of https://github.com/AztecProtocol/aztec3-milestones/issues/18

The structure should be similar to #1

The rollup contract should progress the state hash if old state hash matches starting state of incoming rollup block, and the proof is valid.