test_worst_clz_diff_input()

Documentation for tests/benchmark/test_worst_compute.py::test_worst_clz_diff_input@v5.0.0.

Generate fixtures for these test cases for Osaka with:

fill -v tests/benchmark/test_worst_compute.py::test_worst_clz_diff_input --fork Osaka

Test running a block with as many CLZ with different input as possible.

Source code in tests/benchmark/test_worst_compute.py

@pytest.mark.valid_from("Osaka")
def test_worst_clz_diff_input(
    blockchain_test: BlockchainTestFiller,
    pre: Alloc,
    fork: Fork,
    gas_benchmark_value: int,
    env: Environment,
):
    """Test running a block with as many CLZ with different input as possible."""
    tx_gas_limit = fork.transaction_gas_limit_cap() or env.gas_limit
    max_code_size = fork.max_code_size()

    code_prefix = Op.JUMPDEST
    code_suffix = Op.PUSH0 + Op.JUMP

    available_code_size = max_code_size - len(code_prefix) - len(code_suffix)

    code_seq = Bytecode()

    for i in range(available_code_size):
        value = (2**256 - 1) >> (i % 256)
        clz_op = Op.CLZ(value) + Op.POP
        if len(code_seq) + len(clz_op) > available_code_size:
            break
        code_seq += clz_op

    attack_code = code_prefix + code_seq + code_suffix
    assert len(attack_code) <= max_code_size

    code_address = pre.deploy_contract(code=attack_code)

    sender = pre.fund_eoa()
    tx_count = gas_benchmark_value // tx_gas_limit
    remainder_gas = gas_benchmark_value % tx_gas_limit
    txs = [
        Transaction(
            to=code_address,
            gas_limit=tx_gas_limit if i < tx_count else remainder_gas,
            nonce=i,
            sender=sender,
        )
        for i in range(tx_count + 1)
    ]

    blockchain_test(
        genesis_environment=env,
        pre=pre,
        post={},
        blocks=[Block(txs=txs)],
    )

Parametrized Test Cases

This test case is only parametrized by fork.

Fork All Forks Osaka

Fixture Type All Fixture Types blockchain_test

Test ID (Abbreviated)
...fork_Osaka-blockchain_test