invis2912 / dendreth Goto Github PK

The DendrETH project implements the beacon chain light client syncing algorithm in the form of a smart contract for multiple targeted blockchains, aiming to enable the creation of secure cross-blockchain bridges that don't require a trusted operator. In their current state, our contracts are complete and suitable for testnet deployments, but they are still not intended for production use.

For EVM-based blockchains, we build upon prior research by 0xPARC, Darwinia, Alex Stokes and the Nimbus team to deliver the first end-to-end implementation capable of syncing the entire Mainnet history since Altair. Our current Solidity contract leverages a Circom zero-knowledge circuit to verify the BLS signatures of the Ethereum 2 validators and all of the syncing protocol rules.

Since the base circuit is able to verify complete header-to-header transitions, we also provide a recursive variant that can be used by any Ethereum client to implement one-shot syncing capabilities similar to the ones available in the Mina protocol (please see our analysis regarding the limitations of this approach).

For blockchains based on WebAssembly and BPF, we are developing a direct implementation of the light client syncing protocol based on the highly efficient BLS, SSZ and Light client syncing libraries developed by Supranational and the Nimbus team. When compared to the similarly positioned Snowbridge project, our implementation brings a 36-fold reduction in code size (2.2MB vs 60kb) which should also translate in significant reduction in gas costs (currently, our code is targeting only the standard WebAssembly run-time instead of the full blockchain environments).

Due to the large number of required compiler toolchains and blockchain run-time environments targeted by this project, installing all pre-requisites by hand is not practical. We are offering a deterministic build environment based on the Nix package manager, which is best supported on Linux, but also runs on macOS with some minor limitations at the moment. Windows users may try to use Nix in the Windows Subsystem for Linux, but our team is not currently testing this configuration.

See Getting started with Nix for more details.

Certain scripts in this repository will require API credentials for Infura and Etherscan in order to be able to deploy the contracts in the official networks. To specify such credentials, please create a file named .env and place it at the root of your working copy. Its contents should look like this:

INFURA_API_KEY=1234567890abcdef1234567890abcdef
ETHERSCAN_API_KEY=1234567890ABCDEF1234567890ABCDEF12

A normal light client will download light client updates from the Ethereum P2P network or from the Beacon REST API of an Ethereum node. To sync a smart contract, we perform the same process in reverse - we upload the light client updates to the contract hosted on another blockchain in the form of regular transactions. The contract is initialized with a starting bootstrap state and it updates its view of the beacon chain with each processed update.

This allows it to maintain information about a recent finalized beacon chain block header and a recent optimistic head. The information in these headers is enough to authenticate any data point in the Ethereum ecosystem because a beacon chain block header references a BeaconState root hash, which in turn references a recent execution layer block header, which in turn references the root hash of the execution layer state. Thus, if a chain of Merkle proofs is also supplied and verified against the light client contract state, it can be used to prove in the targeted blockchain the occurrence of any event in the Ethereum world.

To facilitate the development of ours and other similar projects, we'll be maintaining an archive of the best light client updates for each sync committee period since Altair, as produced by a fully-synced Nimbus node:

https://github.com/metacraft-labs/eth2-light-client-updates

Our archive of light client updates also includes pre-generated zero-knowledge proofs for the latest version of the light client Circom circuit.

To see a full syncing simulation in action, you can execute the following commands:

git clone https://github.com/metacraft-labs/DendrETH.git
cd DendrETH
git submodule update --init --recursive
nix develop # or `direnv allow` if you are using direnv
yarn install
make evm-simulation

You should see a Hardhat simulation, sequentially processing all available updates. At the time of this writing, each update costs around 330K in gas.

Our archive of light client updates also includes pre-generated zero-knowledge proofs for the latest version of the one-shot syncing Circom circuit.

To see a simulation demonstrating the syncing process to any sync committee period, you can execute the following commands:

git clone https://github.com/metacraft-labs/DendrETH.git
cd DendrETH
git submodule update --init --recursive
nix develop # or `direnv allow` if you are using direnv
yarn install
make one-shot-syncing-simulation

The circuits employed by this project are some of the largest ever developed. We are building upon the BLS primitives implemented by the circom-pairing project and the SHA256 implementation from circomlib, both of which are already very large. To perform our compilations, we had to purchase a server with 384GB of RAM where the fully integrated build takes the following amount of time:

You can examine the required commands for building the final circuit here:

https://github.com/metacraft-labs/DendrETH/blob/main/beacon-light-client/circom/scripts/proof_efficient/build_proof.sh

At the moment, there are multiple test suites of interest:

• The WebAssembly tests of the Nim light client:

  yarn test-emcc
  

• The Circom components test suite:

  cd beacon-light-client/circom
  yarn hardhat test
  

• The Solidity contract test suite:

  cd beacon-light-client/solidity
  yarn hardhat test
  

All code within this repository is licensed under GPLv3.

Please check out our roadmap to learn more about the blockchains and the use cases that we plan to support in the future.