Add incremental trie cache optimization for flashblocks state root calculation

[cliff0412]

In this PR, we cache the trie nodes after each flashblock state root calculation. Therefore,
future flashblock state root calculation can be faster. resolving this issue #345

Full benchmark report:
benchmark_logs/TRIE_CACHE_BENCHMARK_REPORT.md

[SozinM]

IMO should not be configurable, if it's superior - have this as an only option.
Only one thing to consider - if this feature could produce incorrect state root

[cliff0412]

i tested that they produce same result. well, i can add a unit test for that.

[cliff0412]

i just realized that

The existing test framework already validates state roots implicitly. Look at lines
225–233 in driver.rs:

  if self
      .engine_api
      .new_payload(payload.clone(), vec![], B256::ZERO, Requests::default())
      .await?
      .status
      != PayloadStatusEnum::Valid
  {
      return Err(eyre::eyre!("Invalid validation status from builder"));
  }

new_payload submits the built block to the EL, which re-executes the block and
verifies the state root independently. So, if the trie cached state root calculation method produce wrong state root. the payload validation will fail. i have checked all existing tests can pass with this trie cached approach; so i think is should be okay.

[julio4]

Looks very good @cliff0412. As @SozinM said, this is pure optimization and I don't see scenarios where this should be disabled. You can remove the arg and simplify a bit build_block.
You can also add a few comments just to explain that trie updates are None in the first flashblock only.

[cliff0412]

Looks very good @cliff0412. As @SozinM said, this is pure optimization and I don't see scenarios where this should be disabled. You can remove the arg and simplify a bit build_block. You can also add a few comments just to explain that trie updates are None in the first flashblock only.

config arg removed.