[Engine] Invalid forkchoice state

[greged93]

When handling a L1 reorg (UnwindIndexed event), the engine gets its unsafe and safe head reset in case the L1 reorg caused an L2 reorg. This sets the engine driver in a correct state but the Execution node won't get its forkchoice state updated until a new block is imported (for a follower node) or produced (for the sequencer).

We should decide whether this is an acceptable behavior, or whether this should be fixed.

[greged93]

@frisitano tagging for visibility

[jonastheis]

I implemented this but as it turns out in the spec:

Client software MAY skip an update of the forkchoice state and MUST NOT begin a payload build process if forkchoiceState.headBlockHash references an ancestor of the head of canonical chain. In the case of such an event, client software MUST return {payloadStatus: {status: VALID, latestValidHash: forkchoiceState.headBlockHash, validationError: null}, payloadId: null}.

This is exactly what we're trying to achieve here and it seems that reth handles this with not reorging its canonical chain.
Example from a test:

1. L1 reorg to Some(BlockInfo { number: 10, hash: 0xd54a0c96fb004b4bbdac1ea58630779da8a919a1ce258ae90927cb94d934a7ce }) None handled by engine driver
2. Issuing fork choice update to block number 0xd54a0c96fb004b4bbdac1ea58630779da8a919a1ce258ae90927cb94d934a7ce
3. Forkchoice update result: Ok(ForkchoiceUpdated { payload_status: PayloadStatus { status: Valid, latest_valid_hash: Some(0xd54a0c96fb004b4bbdac1ea58630779da8a919a1ce258ae90927cb94d934a7ce) }, payload_id: None })
4. Latest block number: 15, hash: 0xb87bbb262441a3b959d8377db1b7956040ea7deefd7ad91e6907df3444f70283

[frisitano]

Fixed in #351