Implementing BLOBHASH Opcode

[Robertorosmaninho]

This PR implements the BLOBHASH Opcode as part of the implementation of EIP-4844.
To implement this new operation, the type of txVersionedHashes had to be changed to List.

[dwightguth]

There is one more problem. You are matching directly on the txVersionedHashes field, but this field exists inside a cell Map and you are not specifying which transaction you are accessing the hashes of, which means that if the block contains multiple transactions, you might conceivably return the hashes associated with a past or future transaction within the block. You probably want to copy the field from the transaction into the <evm> cell when loadTx is executed. This will involve a change to the configuration and thus to the spec files in the test suite, but I don't see anything that is currently in the EVM cell that could be reliably used to tell the semantics which transaction you are interested in accessing.

[dwightguth]

Code looks good. I wonder why some of the blobhash related tests still fail. You may want to take a look, but obviously the first priority is getting the proofs passing. As I said before, please timebox that to a couple days at most and let me know if it seems like it will take longer than that.

stale

[dwightguth]

I see a lot of green in this file. Can you make sure you're explaining what is happening here?

[Robertorosmaninho]

From what I saw, it's mostly because these tests were supposed to fail. However, they were failing for the wrong reasons and now, they're "passing" when a failure is expected.

The reasons vary a lot, from wrong balance (lack of implementation of gas accounting) to the block validation.