feat: add tests for JUMP instruction

[lambda-0x]

Pull Request type

Please check the type of change your PR introduces:

• Other (please describe): adding test before implementing instruction as mentioned in issue

What is the current behavior?

Resolves: #251

What is the new behavior?

• evm.codes doesn't contain an example for this
• i couldn't find test for this in geth or kakarot cairo0 (lmk if i missed something)
• i think thats because there are only 2 cases, one in which location to which we need to jump is JUMPDEST (in that case its valid) else its invalid
• actually there is one more case in which: 0x5B is part of data segment and is not an opcode (in an PUSHN) context.
  • kakarot 0 also doesn't handle this case, i will see geth's code and figure out how to implement this.

Does this introduce a breaking change?

• No

[lambda-0x]

(test failure due unrelated to PR)

[Eikix]

There is a third case but that can't be covered here for now, which is:
kkrt-labs/kakarot#693

Example in the EVM spec:
https://github.com/ethereum/execution-specs/blob/fade6e95726c9c0b47b779fd8873b5737fbea3b3/src/ethereum/paris/vm/runtime.py#L58

[lambda-0x]

haha yes! just realized that.

[Eikix]

Will start advertising it but the best source of truth we got is:
https://github.com/ethereum/execution-specs/

[lambda-0x]

(added the third case as test, and marked it ignored for now)

[Eikix]

after that lgtm

[github-actions]

Snapshot Comparison Report:

No changes in gas consumption.

[ClementWalter]

I'm fine with the logic of the tests, but prefer to have a named counter variable to clearly states that we push the input of the JUMP opcode

[lambda-0x]

(rebased to main, incase it fixes snapshot CI)

[enitrat]

lgtm!

Eni approved

[Eikix]

Approved, just rebase for conflicts