Rust HeaderBlock

[arvidn]

Purpose:

Transition HeaderBlock to use the rust implementation. The rationale is to speed up serialization and deserialization and preparing for validation logic to move to rust and threads (rather than processes).

The rust implementation and type definition can be found here:
https://github.com/Chia-Network/chia_rs/blob/main/crates/chia-protocol/src/header_block.rs

The batch_pre_validate_blocks() function is run in a ProcessPoolExecutor, i.e. in separate python processes. In order to support passing data across the process boundary, we need to serialize (and pickle) the arguments, and the return value.

We invoke the ProcessPoolExecutor job here: https://github.com/Chia-Network/chia-blockchain/blob/main/chia/consensus/multiprocess_validation.py#L347

We serialize the FullBlock objects a few lines up: https://github.com/Chia-Network/chia-blockchain/blob/main/chia/consensus/multiprocess_validation.py#L332

Since we already have valid FullBlock objects (i.e. the public keys and signatures are known to be on the curve, and the CLVM generator is know to have valid serialization), we don't need to validate that again in the worker process.

Current Behavior:

HeaderBlock is a python type.

New Behavior:

HeaderBlock is a rust type.

[github-actions]

This pull request has conflicts, please resolve those before we can evaluate the pull request.