Skip to content

test_comparison()

Documentation for tests/benchmark/compute/instruction/test_comparison.py::test_comparison@8db70f93.

Generate fixtures for these test cases for Amsterdam with:

fill -v tests/benchmark/compute/instruction/test_comparison.py::test_comparison --gas-benchmark-values 1

Benchmark binary instructions (takes two args, pushes one value). The execution starts with two initial values on the stack The stack is balanced by the DUP2 instruction.

Source code in tests/benchmark/compute/instruction/test_comparison.py 
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
@pytest.mark.repricing
@pytest.mark.parametrize(
    "opcode,opcode_args",
    [
        (
            Op.LT,  # Keeps getting result 1.
            (0, 1),
        ),
        (
            Op.GT,  # Keeps getting result 0.
            (0, 1),
        ),
        (
            Op.SLT,  # Keeps getting result 1.
            (
                0xFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF,
                1,
            ),
        ),
        (
            Op.SGT,  # Keeps getting result 0.
            (
                0xFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF,
                1,
            ),
        ),
        (
            # The worst case is if the arguments are equal (no early return),
            # so let's keep it comparing ones.
            Op.EQ,
            (1, 1),
        ),
    ],
    ids=lambda param: "" if isinstance(param, tuple) else param,
)
def test_comparison(
    benchmark_test: BenchmarkTestFiller,
    opcode: Op,
    opcode_args: tuple[int, int],
) -> None:
    """
    Benchmark binary instructions (takes two args, pushes one value).
    The execution starts with two initial values on the stack
    The stack is balanced by the DUP2 instruction.
    """
    tx_data = b"".join(
        arg.to_bytes(32, byteorder="big") for arg in opcode_args
    )

    setup = Op.CALLDATALOAD(0) + Op.CALLDATALOAD(32) + Op.DUP2 + Op.DUP2
    attack_block = Op.DUP2 + opcode
    cleanup = Op.POP + Op.POP + Op.DUP2 + Op.DUP2
    benchmark_test(
        target_opcode=opcode,
        code_generator=JumpLoopGenerator(
            setup=setup,
            attack_block=attack_block,
            cleanup=cleanup,
            tx_kwargs={"data": tx_data},
        ),
    )

Parametrized Test Cases

This test generates 5 parametrized test cases across 3 forks.