fix(tests): EIP-7623 update due to spec change

[marioevz]

🗒️ Description

This PR updates EIP-7623 tests in accordance with #9227.

tests/prague/eip7623_increase_calldata_cost/test_refunds.py

This file introduces tests that verify all types of refunds:

• Storage Reset.
• Authorization with Authority already in State.

Including tests that use both refunds at the same time.

Test scenarios include execution gas usage when:

• Execution gas minus refund is greater than the data floor.
• Execution gas minus refund equal to the data floor.
• Execution gas minus refund is less than the data floor.

tests/prague/eip7623_increase_calldata_cost/test_execution_gas.py

Refunds are removed from this file altogether.

🔗 Related Issues

None

✅ Checklist

• All: Set appropriate labels for the changes.
• All: Considered squashing commits to improve commit history.
• All: Added an entry to CHANGELOG.md.
• All: Considered updating the online docs in the ./docs/ directory.
• Tests: All converted JSON/YML tests from ethereum/tests have been added to converted-ethereum-tests.txt.
• Tests: A PR with removal of converted JSON/YML tests from ethereum/tests have been opened.
• Tests: Included the type and version of evm t8n tool used to locally execute test cases: e.g., ref with commit hash or geth 1.13.1-stable-3f40e65.
• Tests: Ran mkdocs serve locally and verified the auto-generated docs for new tests in the Test Case Reference are correctly formatted.

[marioevz]

Rebased after #1068 merge. Ready for review.

[danceratopz]

Hey @chfast, could you please help us find the discrepancy in our implementations?

I've filled these tests locally using EELS ethereum/execution-specs@ccb249c, from devnets/prague/5.

• ethereum/go-ethereum@f24d6f6 from s1na/go-ethereum/tree/prague-devnet-5 passes all state and blockchain tests.
• ipsilon/evmone@6b7305e passes all state tests, but fails
  4 blockchain tests:

  [  FAILED  ] 4 tests, listed below:
  [  FAILED  ] prague/eip7623_increase_calldata_cost/execution_gas.full_gas_consumption
  [  FAILED  ] prague/eip7623_increase_calldata_cost/execution_gas.gas_consumption_below_data_floor
  [  FAILED  ] prague/eip7623_increase_calldata_cost/refunds.gas_refunds_from_data_floor
  [  FAILED  ] prague/eip7623_increase_calldata_cost/transaction_validity.transaction_validity_type_4
  

  Test IDs retrieved from the log:

  tests/prague/eip7623_increase_calldata_cost/test_execution_gas.py::TestGasConsumption::test_full_gas_consumption[fork_Prague-blockchain_test-exact_gas-type_4]
  tests/prague/eip7623_increase_calldata_cost/test_execution_gas.py::TestGasConsumption::test_full_gas_consumption[fork_Prague-blockchain_test-extra_gas-type_4]
  tests/prague/eip7623_increase_calldata_cost/test_transaction_validity.py::test_transaction_validity_type_4[fork_Prague-blockchain_test-type_4-single_authorization-single_access_list_single_storage_key-extra_gas-floor_gas_greater_than_intrinsic_gas]
  tests/prague/eip7623_increase_calldata_cost/test_transaction_validity.py::test_transaction_validity_type_4[fork_Prague-blockchain_test-type_4-single_authorization-single_access_list_single_storage_key-extra_gas-floor_gas_less_than_or_equal_to_intrinsic_gas]
  

fixtures-eip7623.zip

[danceratopz]

@chfast Ah, could it be due to lack of type 4 support? All the fails use a type-4 transaction.

[chfast]

@chfast Ah, could it be due to lack of type 4 support? All the fails use a type-4 transaction.

Yes. You can use eip-7702 branch.

[danceratopz]

@chfast Ah, could it be due to lack of type 4 support? All the fails use a type-4 transaction.

Yes. You can use eip-7702 branch.

Yup, thanks that works! (f2d71302).

All tests pass:

build/bin/evmone-blockchaintest fixtures-eip7623/blockchain_tests
build/bin/evmone-statetest fixtures-eip7623/state_tests/

[danceratopz]

Nethermind NethermindEth/nethermind@52b679a from pectra-devnet-5 also passes all state and blockchain tests (via nethtest) 🥳

[danceratopz]

Looks good to me! Although, as @winsvega remarked, these are pretty tough tests to follow.