[Feature Request] Substrate Frontier runtime upgrade Compatibility

Do you want to request a feature or report a bug?

Feature

What is the current behavior?

In a polkadot parachain where it uses Frontier for their WASM runtime, evm feature sometimes does not start from day 1(i.e. block 0). It might add EVM for testing its connection between relay chain or go through necessary tests before they add evm integration. Due to this, current graph node is not able to index evm since current code is set to index from block 0 here. I would like to suggest to get input from config to set genesis block of evm to index.

If the current behavior is a bug, please provide the steps to reproduce and if possible a minimal demo of the problem.

The error shows in shiden with "Ethereum node could not find genesis block" with the following log

Oct 13 16:30:16.455 DEBG Trying again after eth_getBlockByNumber(0, false) RPC call failed (attempt #1) with result Err(Ethereum node could not find genesis block), provider: shiden-rpc-0
Oct 13 16:30:16.811 DEBG Trying again after eth_getBlockByNumber(0, false) RPC call failed (attempt #2) with result Err(Ethereum node could not find genesis block), provider: shiden-rpc-0
Oct 13 16:30:17.177 DEBG Trying again after eth_getBlockByNumber(0, false) RPC call failed (attempt #3) with result Err(Ethereum node could not find genesis block), provider: shiden-rpc-0

What is the expected behavior?

I expect adding gen_block input in config and add that in ChainSection or stores so that genesis block argument can be used in ethereum_adapter.rs

Start with chain/ethereum/src/ethereum_adapter.rs at the referenced genesis-block lookup, then trace how chain configuration reaches ChainSection or stores. Determine how a configurable gen_block should be represented and passed to the adapter. Done means Frontier runtimes whose EVM starts after block 0 can index from the configured block without the missing-genesis error.