Add EVM serialization (kkrt-labs#995)

Time spent on this PR: 0.5

## Pull request type

Please check the type of change your PR introduces:

- [ ] Bugfix
- [x] Feature
- [ ] Code style update (formatting, renaming)
- [ ] Refactoring (no functional changes, no api changes)
- [ ] Build related changes
- [ ] Documentation content changes
- [ ] Other (please describe):

## What is the current behavior?

Not possible to return `model.EVM` from a test

Resolves kkrt-labs#940 

## What is the new behavior?

Serialize model.EVM + included struct + fix when returning struct and
not struct*

<!-- Reviewable:start -->
- - -
This change is [<img src="https://reviewable.io/review_button.svg"
height="34" align="absmiddle"
alt="Reviewable"/>](https://reviewable.io/reviews/kkrt-labs/kakarot/995)
<!-- Reviewable:end -->

.gitignore

@@ -28,3 +28,6 @@ resources/*
 *.pb.gz
 profile*.png
 logs
+
+tests/ef_tests/test_data
+!tests/ef_tests/test_data/.gitkeep

.trunk/trunk.yaml

@@ -59,6 +59,7 @@ lint:
         - logs*
         - lib*
         - resources*
+        - tests/ef_tests/test_data
 actions:
   disabled:
     - trunk-announce

Makefile

@@ -1,19 +1,3 @@
-# The release tag of https://github.com/ethereum/tests to use for EF tests
-EF_TESTS_TAG := v12.4
-EF_TESTS_URL := https://github.com/ethereum/tests/archive/refs/tags/$(EF_TESTS_TAG).tar.gz
-EF_TESTS_DIR := ./tests/ef_tests/test_data
-
-# Downloads and unpacks Ethereum Foundation tests in the `$(EF_TESTS_DIR)` directory.
-# Requires `wget` and `tar`
-$(EF_TESTS_DIR):
-	mkdir -p $(EF_TESTS_DIR)
-	wget $(EF_TESTS_URL) -O ethereum-tests.tar.gz
-	tar -xzf ethereum-tests.tar.gz --strip-components=1 -C $(EF_TESTS_DIR)
-	rm ethereum-tests.tar.gz
-
-
-.PHONY: build test coverage $(EF_TESTS_DIR)
-
 # Include .env file to get GITHUB_TOKEN
 ifneq ("$(wildcard .env)","")
     include .env
@@ -32,6 +16,9 @@ build: check
 check:
 	poetry check --lock
 
+fetch-ef-tests:
+	poetry run python ./scripts/ef_tests/fetch.py
+
 # This action fetches the latest Kakarot SSJ (Cairo compiler version >=2) artifacts
 # from the main branch and unzips them into the build/ssj directory.
 # This is required because Kakarot Zero (Cairo Zero, compiler version <1) uses some SSJ Cairo programs.
@@ -47,11 +34,11 @@ setup: fetch-ssj-artifacts
 	poetry install
 
 test: build-sol deploy
-	poetry run pytest tests/src -m "not EFTests" --log-cli-level=INFO -n logical
+	poetry run pytest tests/src -m "not NoCI" --log-cli-level=INFO -n logical
 	poetry run pytest tests/end_to_end
 
 test-unit:
-	poetry run pytest tests/src -n logical
+	poetry run pytest tests/src -m "not NoCI" -n logical
 
 test-end-to-end: build-sol deploy
 	poetry run pytest tests/end_to_end

pyproject.toml

@@ -144,6 +144,7 @@ markers = [
   "EFTests",
   "SSTORE",
   "SLOAD",
+  "NoCI",
 ]
 env = [
   "PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION = python",

scripts/ef-tests/debug.py → scripts/ef_tests/debug.py

@@ -12,6 +12,8 @@
 from eth.vm.forks.shanghai.blocks import ShanghaiBlock
 from web3 import Web3
 
+from scripts.ef_tests.fetch import EF_TESTS_PARSED_DIR
+
 logging.basicConfig()
 logger = logging.getLogger(__name__)
 logger.setLevel(logging.INFO)
@@ -22,7 +24,7 @@
 TEST_NAME = os.getenv("TEST_NAME")
 if TEST_NAME is None:
     raise ValueError("Please set TEST_NAME")
-TEST_PARENT_FOLDER = os.getenv("TEST_PARENT_FOLDER")
+TEST_PARENT_FOLDER = os.getenv("TEST_PARENT_FOLDER", "")
 RPC_ENDPOINT = "http://127.0.0.1:8545"
 
 
@@ -59,29 +61,23 @@ def signal_handler(self, signum, _):
 
 def get_test_file():
     tests = [
-        (content, str(file_path))
-        for file_path in TESTS_PATH.glob("**/*.json")
-        for name, content in json.load(open(file_path)).items()
-        if TEST_NAME in name and content["network"] == "Shanghai"
+        file_name
+        for file_name in os.listdir(EF_TESTS_PARSED_DIR)
+        if TEST_NAME in file_name and TEST_PARENT_FOLDER in file_name
     ]
 
     if len(tests) == 0:
         raise ValueError(f"Test {TEST_NAME} not found")
 
     if len(tests) > 1:
-        if TEST_PARENT_FOLDER is None:
+        if TEST_PARENT_FOLDER == "":
             raise ValueError(
                 f"Test {TEST_NAME} is ambiguous, please set TEST_PARENT_FOLDER to test file folder"
             )
 
-        test = [content for (content, path) in tests if TEST_PARENT_FOLDER in path]
-        if len(test) == 0:
-            raise ValueError(f"Test {TEST_NAME} not found")
-
-        return test[0]
+        raise ValueError(f"Test {TEST_NAME} not found")
 
-    else:
-        return tests[0][0]
+    return json.loads((EF_TESTS_PARSED_DIR / tests[0]).read_text())
 
 
 def connect_anvil():

scripts/ef_tests/fetch.py

@@ -0,0 +1,52 @@
+import io
+import json
+import logging
+import os
+import shutil
+import tarfile
+from pathlib import Path
+
+import requests
+
+EF_TESTS_TAG = "v12.4"
+EF_TESTS_URL = (
+    f"https://github.com/ethereum/tests/archive/refs/tags/{EF_TESTS_TAG}.tar.gz"
+)
+EF_TESTS_DIR = Path("tests") / "ef_tests" / "test_data" / EF_TESTS_TAG
+EF_TESTS_PARSED_DIR = Path("tests") / "ef_tests" / "test_data" / "parsed"
+
+DEFAULT_NETWORK = "Shanghai"
+
+logging.basicConfig()
+logger = logging.getLogger(__name__)
+logger.setLevel(logging.INFO)
+
+
+def generate_tests():
+    if not EF_TESTS_DIR.exists():
+        response = requests.get(EF_TESTS_URL)
+        with tarfile.open(fileobj=io.BytesIO(response.content), mode="r:gz") as tar:
+            tar.extractall(EF_TESTS_DIR)
+
+    test_cases = {
+        f"{os.path.basename(root)}_{name}": content
+        for (root, _, files) in os.walk(EF_TESTS_DIR)
+        for file in files
+        if file.endswith(".json") and "BlockchainTests/GeneralStateTests" in root
+        for name, content in json.loads((Path(root) / file).read_text()).items()
+        if content.get("network") == DEFAULT_NETWORK
+    }
+
+    shutil.rmtree(EF_TESTS_PARSED_DIR, ignore_errors=True)
+    EF_TESTS_PARSED_DIR.mkdir(parents=True, exist_ok=True)
+
+    for test_name, test_case in test_cases.items():
+        json.dump(
+            test_case,
+            open(EF_TESTS_PARSED_DIR / f"{test_name}.json", "w"),
+            indent=4,
+        )
+
+
+if __name__ == "__main__":
+    generate_tests()

scripts/utils/kakarot.py

@@ -109,6 +109,9 @@ def get_contract(
             bytecode=target_compilation_output["bytecode"]["object"],
         ),
     )
+    contract.bytecode_runtime = HexBytes(
+        target_compilation_output["deployedBytecode"]["object"]
+    )
 
     try:
         for fun in contract.functions: