fix(tests): EIP-7702: use penultimate block number instead of fixing block 0 #1390
Conversation
|
cc @marioevz , seems that you're the author of this test case |
| @@ -2656,7 +2656,8 @@ def test_set_code_to_system_contract( | |||
| caller_payload = consolidation_request.calldata | |||
| call_value = consolidation_request.value | |||
| case Address(0x0000F90827F1C53A10CB7A02335B175320002935): # EIP-2935 | |||
| caller_payload = Hash(0) | |||
| # Use the penultimate block to ensure it is saved in the history contract | |||
| caller_payload = Hash(max(pre.get_block_number() - 1, 0)) | |||
There was a problem hiding this comment.
I think it should be in env too, but I can't find how to get block number from node without using _eth_rpc of Alloc, can you suggest a better solution?
There was a problem hiding this comment.
I agree that this should be in the env, and I think this requires a bit more complex solution.
We probably need to take the same approach as pre where it's a global fixture generated by the filler/execute plugin:
In the filler plugin, the fixture result would always be an instance of Environment with the default values.
In the execute plugin, the fixture is also an instance of Environment but the values are populated using the return self._eth_rpc... methods.
spencer-tb
added
fork:prague
🗒️ Description
In the test case that uses setcode tx to call the history contract, the block number is fixed 0, which leads to failure when running in
executemode because block number 0 is usually out of range for history contract queries.This PR uses:
executemode: penultimate block number queried from node instead of 0fillmode: use block number 0, same as before