From e4eba6bb17175d89ce4082f583ebb4528c122941 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Pawe=C5=82=20Bylica?= <pawel@ethereum.org>
Date: Fri, 23 May 2025 16:19:55 +0200
Subject: [PATCH] feat(tests): add benchmark for the worst initcode jumpdest
 analysis

---
 tests/zkevm/test_worst_bytecode.py | 94 ++++++++++++++++++++++++++++++
 1 file changed, 94 insertions(+)

diff --git a/tests/zkevm/test_worst_bytecode.py b/tests/zkevm/test_worst_bytecode.py
index 4343da0a0e4..ae0d2364939 100644
--- a/tests/zkevm/test_worst_bytecode.py
+++ b/tests/zkevm/test_worst_bytecode.py
@@ -18,6 +18,7 @@
     Bytecode,
     Environment,
     Hash,
+    StateTestFiller,
     Transaction,
     While,
     compute_create2_address,
@@ -221,3 +222,96 @@ def test_worst_bytecode_single_opcode(
         ],
         exclude_full_post_state_in_output=True,
     )
+
+
+@pytest.mark.valid_from("Cancun")
+@pytest.mark.parametrize(
+    "pattern",
+    [
+        Op.STOP,
+        Op.JUMPDEST,
+        Op.PUSH1[bytes(Op.JUMPDEST)],
+        Op.PUSH2[bytes(Op.JUMPDEST + Op.JUMPDEST)],
+        Op.PUSH1[bytes(Op.JUMPDEST)] + Op.JUMPDEST,
+        Op.PUSH2[bytes(Op.JUMPDEST + Op.JUMPDEST)] + Op.JUMPDEST,
+    ],
+    ids=lambda x: x.hex(),
+)
+def test_worst_initcode_jumpdest_analysis(
+    state_test: StateTestFiller,
+    pre: Alloc,
+    fork: Fork,
+    pattern: Bytecode,
+):
+    """
+    Test the jumpdest analysis performance of the initcode.
+
+    This benchmark places a very long initcode in the memory and then invoke CREATE instructions
+    with this initcode up to the block gas limit. The initcode itself has minimal execution time
+    but forces the EVM to perform the full jumpdest analysis on the parametrized byte pattern.
+    The initicode is modified by mixing-in the returned create address between CREATE invocations
+    to prevent caching.
+    """
+    max_code_size = fork.max_code_size()
+    initcode_size = fork.max_initcode_size()
+
+    # Expand the initcode pattern to the transaction data so it can be used in CALLDATACOPY
+    # in the main contract. TODO: tune the tx_data_len param.
+    tx_data_len = 1024
+    tx_data = pattern * (tx_data_len // len(pattern))
+    tx_data += (tx_data_len - len(tx_data)) * bytes(Op.JUMPDEST)
+    assert len(tx_data) == tx_data_len
+    assert initcode_size % len(tx_data) == 0
+
+    # Prepare the initcode in memory.
+    code_prepare_initcode = sum(
+        (
+            Op.CALLDATACOPY(dest_offset=i * len(tx_data), offset=0, size=Op.CALLDATASIZE)
+            for i in range(initcode_size // len(tx_data))
+        ),
+        Bytecode(),
+    )
+
+    # At the start of the initcode execution, jump to the last opcode.
+    # This forces EVM to do the full jumpdest analysis.
+    initcode_prefix = Op.JUMP(initcode_size - 1)
+    code_prepare_initcode += Op.MSTORE(
+        0, Op.PUSH32[bytes(initcode_prefix).ljust(32, bytes(Op.JUMPDEST))]
+    )
+
+    # Make sure the last opcode in the initcode is JUMPDEST.
+    code_prepare_initcode += Op.MSTORE(initcode_size - 32, Op.PUSH32[bytes(Op.JUMPDEST) * 32])
+
+    code_invoke_create = (
+        Op.PUSH1[len(initcode_prefix)]
+        + Op.MSTORE
+        + Op.CREATE(value=Op.PUSH0, offset=Op.PUSH0, size=Op.MSIZE)
+    )
+
+    initial_random = Op.PUSH0
+    code_prefix = code_prepare_initcode + initial_random
+    code_loop_header = Op.JUMPDEST
+    code_loop_footer = Op.JUMP(len(code_prefix))
+    code_loop_body_len = (
+        max_code_size - len(code_prefix) - len(code_loop_header) - len(code_loop_footer)
+    )
+
+    code_loop_body = (code_loop_body_len // len(code_invoke_create)) * bytes(code_invoke_create)
+    code = code_prefix + code_loop_header + code_loop_body + code_loop_footer
+    assert (max_code_size - len(code_invoke_create)) < len(code) <= max_code_size
+
+    env = Environment()
+
+    tx = Transaction(
+        to=pre.deploy_contract(code=code),
+        data=tx_data,
+        gas_limit=env.gas_limit,
+        sender=pre.fund_eoa(),
+    )
+
+    state_test(
+        env=env,
+        pre=pre,
+        post={},
+        tx=tx,
+    )