mirror of
https://github.com/norandom/OpenUltraSAST.git
synced 2026-10-16 03:49:26 +00:00
merge: taint-query null-guard fix and trace-enabled engine runner
The query fix stops a class-level call from aborting the whole taint batch on real PHP code; gate, map_gate and pair_gate outputs are unchanged. Trace parsing lives with the runner (benchmarks/learn/engine_trace_parse.py). Co-Authored-By: Claude Opus 5.5 <[email protected]>
This commit is contained in:
Marius Ciepluch
@@ -0,0 +1,841 @@
|
||||
"""Resumable, serial labelled-function traces; host fetches pinned sources, Docker stays offline.
|
||||
|
||||
Pair blob identities resolve through catalogs to real source commits. The analyzer
|
||||
is git archive HEAD src plus a hashed snapshot of this uncommitted tool.
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import argparse
|
||||
import ast
|
||||
import fcntl
|
||||
import hashlib
|
||||
import json
|
||||
import os
|
||||
import re
|
||||
import shutil
|
||||
import subprocess
|
||||
import tempfile
|
||||
import textwrap
|
||||
import time
|
||||
import tomllib
|
||||
from collections import Counter, defaultdict
|
||||
from pathlib import Path
|
||||
|
||||
from engine_trace_worker import EXCLUDED_DIRS, safe_path, unit_record, write_json
|
||||
|
||||
from openultrasast.learn.examples import read_jsonl
|
||||
from openultrasast.learn.labels import load_sources, repo_name
|
||||
from openultrasast.model.specs import taint_specs
|
||||
from openultrasast.model.taint import request_params
|
||||
from openultrasast.pairs import _materialize_side, load_pair_catalog
|
||||
from openultrasast.plane.engine_alerts import export_pin
|
||||
from openultrasast.plane.memory import FileStore
|
||||
from openultrasast.plane.tasks.alerts import engine_languages
|
||||
from openultrasast.preprocess import detect_language
|
||||
|
||||
ROOT = Path(__file__).resolve().parents[2]
|
||||
DEFAULT_OUT = Path.home() / "ousast-results/plane/engine-trace"
|
||||
SNAPSHOT_FILES = (
|
||||
"benchmarks/learn/engine_trace.py",
|
||||
"benchmarks/learn/engine_trace_worker.py",
|
||||
"benchmarks/learn/engine_trace_parse.py",
|
||||
"src/openultrasast/model/taint.py",
|
||||
"src/openultrasast/cpg/queries/taint.sc",
|
||||
)
|
||||
MATERIALIZATION_VERSION = 3
|
||||
EXPORT_CAP = 20 * 1024 * 1024
|
||||
|
||||
STATUSES = ("path", "asked-nothing", "failed", "timeout", "unsupported")
|
||||
|
||||
|
||||
def digest(value):
|
||||
return hashlib.sha256(json.dumps(value, sort_keys=True).encode()).hexdigest()
|
||||
|
||||
|
||||
def pin_name(pin):
|
||||
return digest([pin["repo"], pin["pin"]]) + ".json"
|
||||
|
||||
|
||||
def join_units(units, examples, include_typescript=False):
|
||||
index = {row["id"]: row for row in examples}
|
||||
missing = [u["unit"] for u in units if u["unit"] not in index]
|
||||
if missing:
|
||||
raise ValueError(f"example snapshot missing {len(missing)}/{len(units)} unit IDs; supply --examples JSONL; first: {missing[0]}")
|
||||
covered = engine_languages() - {"typescript"}
|
||||
if include_typescript:
|
||||
covered |= {"typescript"}
|
||||
result = []
|
||||
for unit in units:
|
||||
example = index[unit["unit"]]
|
||||
if any(example[key] != unit[key] for key in ("family", "label", "group")):
|
||||
raise ValueError(f"unit/example identity mismatch: {unit['unit']}")
|
||||
file, sep, function = example["candidate"].partition("::")
|
||||
if not sep or not function:
|
||||
raise ValueError(f"unit has no labelled function: {unit['unit']}")
|
||||
language = detect_language(Path(file)) or example.get("language", "other")
|
||||
model_language = "javascript" if language == "typescript" else language
|
||||
spec = taint_specs(language=model_language).get(unit["family"]) if language in covered else None
|
||||
result.append(
|
||||
{
|
||||
**unit,
|
||||
"repo": example["repo"],
|
||||
"pin": example["pin"],
|
||||
"file": file,
|
||||
"function": function,
|
||||
"language": language,
|
||||
"source": example["source"],
|
||||
"excerpt": example.get("excerpt", example["source"] == "pairs" or example["source"].startswith("advisory-fixes")),
|
||||
"pin_role": example.get("pin_role", ""),
|
||||
"supported": spec is not None,
|
||||
"reason": "" if spec else f"no admitted taint model: {language}/{unit['family']}",
|
||||
}
|
||||
)
|
||||
return result
|
||||
|
||||
|
||||
def make_plan(
|
||||
units,
|
||||
out,
|
||||
only_family=None,
|
||||
limit=0,
|
||||
*,
|
||||
include_typescript=False,
|
||||
only_source=(),
|
||||
rerun_status=("failed",),
|
||||
allow_excerpt_fallback=False,
|
||||
):
|
||||
grouped = defaultdict(list)
|
||||
for unit in units:
|
||||
if (not only_family or unit["family"] == only_family) and (not only_source or unit["source"] in only_source):
|
||||
# Include admission options even when a unit's support happens to stay
|
||||
# unchanged. Old results cannot mark a different selection as complete.
|
||||
identity = {
|
||||
"unit": unit,
|
||||
"include_typescript": include_typescript,
|
||||
"materialization_version": MATERIALIZATION_VERSION,
|
||||
"allow_excerpt_fallback": allow_excerpt_fallback,
|
||||
}
|
||||
grouped[unit["repo"], unit["pin"]].append({**unit, "input_digest": digest(identity)})
|
||||
plan = []
|
||||
for (repo, pin), rows in sorted(grouped.items()):
|
||||
item = {
|
||||
"repo": repo,
|
||||
"pin": pin,
|
||||
"materialization_version": MATERIALIZATION_VERSION,
|
||||
"units": sorted(rows, key=lambda r: r["unit"]),
|
||||
}
|
||||
item["selection"] = digest(item["units"])
|
||||
done = out / pin_name(item)
|
||||
if done.exists():
|
||||
old = json.loads(done.read_text())
|
||||
recorded = {r["unit"]: r.get("input_digest") for r in old.get("units", [])}
|
||||
if (
|
||||
old.get("done")
|
||||
and old.get("materialization_version") == MATERIALIZATION_VERSION
|
||||
and not any(r.get("status") in rerun_status for r in old.get("units", []) if r["unit"] in {u["unit"] for u in rows})
|
||||
and all(recorded.get(r["unit"]) == r["input_digest"] for r in rows)
|
||||
):
|
||||
continue
|
||||
sources = {r["source"] for r in rows}
|
||||
# Inventory medians: harvest sides 31s, advisory sides 47s; v1 pins 182s, v2 pins 498s.
|
||||
item["estimated_seconds"] = (
|
||||
(47 if any(s.startswith("advisory") for s in sources) else 31)
|
||||
if any(r.get("excerpt") for r in rows) or any(s.startswith("advisory") for s in sources)
|
||||
else (182 if sources == {"population-v1"} else 498)
|
||||
)
|
||||
if not any(r["supported"] for r in rows):
|
||||
item["estimated_seconds"] = 0
|
||||
plan.append(item)
|
||||
# limit counts runnable pins, not unsupported records.
|
||||
if limit:
|
||||
chosen = []
|
||||
count = 0
|
||||
for pin in plan:
|
||||
runnable = any(r["supported"] for r in pin["units"])
|
||||
if runnable and count >= limit:
|
||||
continue
|
||||
chosen.append(pin)
|
||||
count += runnable
|
||||
return chosen
|
||||
return plan
|
||||
|
||||
|
||||
class Inputs:
|
||||
"""Resolve spent corpora; fetch pair source commits into the host cache."""
|
||||
|
||||
def __init__(self, root, cache, allow_excerpt_fallback=False):
|
||||
self.cache = cache
|
||||
self.allow_excerpt_fallback = allow_excerpt_fallback
|
||||
self.catalog_rows = defaultdict(list)
|
||||
for catalog in (root / "benchmarks/pairs").rglob("catalog.toml"):
|
||||
for item in tomllib.loads(catalog.read_text()).get("pair", []):
|
||||
self.catalog_rows[item["name"]].append((catalog.parent.name, item))
|
||||
self.pairs = []
|
||||
self.repos = defaultdict(list)
|
||||
sources = load_sources()
|
||||
for source in sources.sources:
|
||||
if source.kind == "pairs":
|
||||
self.pairs.extend(load_pair_catalog(root / source.files[0]))
|
||||
elif source.kind == "population":
|
||||
if source.id not in {"population-v1", "population-v2"}:
|
||||
continue
|
||||
data = tomllib.loads((root / source.files[0]).read_text())
|
||||
for case in data.get("case", []):
|
||||
self.repos[repo_name(case["repo"])].append((cache / "independent" / case["id"], case))
|
||||
elif source.kind == "recipes":
|
||||
for file in source.files:
|
||||
data = tomllib.loads((root / file).read_text())
|
||||
recipe = data.get("repo", data)
|
||||
# Recipes carry a top-level URL and a pinned checkout cache.
|
||||
if isinstance(recipe, dict):
|
||||
self.repos[repo_name(str(recipe.get("url", "")))].append((cache / "repos" / str(data.get("name", "")), data))
|
||||
self.hashes = {}
|
||||
|
||||
def pair_for(self, row):
|
||||
if row["pin_role"] not in {"fixed", "vulnerable", "vuln"}:
|
||||
raise ValueError(f"ambiguous pair side: {row['pin_role']}")
|
||||
side = "fixed" if row["pin_role"] == "fixed" else "vuln"
|
||||
candidates = [
|
||||
c for c in self.pairs if c.relpath == row["file"] and repo_name(c.repo or f"{c.slice}:{c.name}") == repo_name(row["repo"])
|
||||
]
|
||||
matches = []
|
||||
for case in candidates:
|
||||
path = case.fixed_file if side == "fixed" else case.vuln_file
|
||||
if path.is_file():
|
||||
if path not in self.hashes:
|
||||
data = path.read_bytes()
|
||||
# Harvest hashes decoded text, so universal newline conversion is part
|
||||
# of its identity (one recorded ThreatByte excerpt uses CRLF).
|
||||
normalized = path.read_text().encode("utf-8")
|
||||
self.hashes[path] = {
|
||||
hashlib.sha1(b"blob " + str(len(body)).encode() + b"\0" + body).hexdigest() for body in (data, normalized)
|
||||
}
|
||||
if row["pin"] in self.hashes[path]:
|
||||
matches.append(case)
|
||||
if len(matches) == 1:
|
||||
return matches[0], side
|
||||
if len(candidates) == 1:
|
||||
# Earlier stores used label_pin rather than the excerpt digest.
|
||||
from openultrasast.learn.examples import label_pin
|
||||
|
||||
case = candidates[0]
|
||||
if row["pin"] == label_pin({"source": "pairs", "source_ref": case.name, "pin_role": row["pin_role"]}):
|
||||
return case, side
|
||||
raise ValueError(f"cannot uniquely resolve excerpt {row['repo']}@{row['pin']}:{row['file']}")
|
||||
|
||||
def catalog_for(self, case):
|
||||
matches = [
|
||||
(source, item)
|
||||
for source, item in self.catalog_rows[case.name]
|
||||
if item.get("relpath") == case.relpath and repo_name(item.get("repo", "")) == repo_name(case.repo)
|
||||
]
|
||||
if len(matches) != 1:
|
||||
raise ValueError(f"ambiguous catalog source for {case.name}")
|
||||
return matches[0]
|
||||
|
||||
def source_identity(self, case, side):
|
||||
_, item = self.catalog_for(case)
|
||||
parent, commit = item.get("parent", ""), item.get("commit", "")
|
||||
if side not in {"vuln", "fixed"} or not all(re.fullmatch(r"[0-9a-fA-F]{40}", sha) for sha in (parent, commit)) or parent == commit:
|
||||
raise ValueError(f"ambiguous catalog commits for {case.name}: need distinct parent and commit SHAs")
|
||||
repo = item.get("repo", "")
|
||||
if not repo:
|
||||
raise ValueError(f"missing catalog repository for {case.name}")
|
||||
return repo, parent if side == "vuln" else commit
|
||||
|
||||
def fetch_source(self, repo, commit, roots):
|
||||
clone = self.cache / "trace-sources" / digest([repo, commit, roots])
|
||||
clone.mkdir(parents=True, exist_ok=True)
|
||||
url = repo if repo.startswith("https://") else f"https://github.com/{repo.removesuffix('.git')}.git"
|
||||
env = {**os.environ, "GIT_TERMINAL_PROMPT": "0"}
|
||||
env.pop("GIT_NO_LAZY_FETCH", None) # sparse checkout may fetch the selected blobs
|
||||
|
||||
def git(*args):
|
||||
return subprocess.run(["git", "-C", str(clone), *args], env=env, capture_output=True, text=True, check=True, timeout=300)
|
||||
|
||||
git("init")
|
||||
git("config", "remote.origin.url", url)
|
||||
git("config", "remote.origin.promisor", "true")
|
||||
git("config", "remote.origin.partialclonefilter", "blob:none")
|
||||
git("fetch", "--depth", "1", "--filter=blob:none", url, commit)
|
||||
git("sparse-checkout", "init", "--cone")
|
||||
git("sparse-checkout", "set", "--", *roots)
|
||||
git("checkout", "--detach", "--force", commit)
|
||||
actual = git("rev-parse", "HEAD").stdout.strip()
|
||||
if actual != commit:
|
||||
raise ValueError(f"fetched commit mismatch: {actual} != {commit}")
|
||||
return clone
|
||||
|
||||
def export_local(self, clone, pin, target):
|
||||
with tempfile.TemporaryDirectory(prefix="trace-export-", dir=target.parent) as scratch:
|
||||
export_pin(clone, pin["pin"], Path(scratch))
|
||||
roots = sorted({source_root(r["file"], r["language"]) for r in pin["units"]})
|
||||
proof = export_sources(Path(scratch), target, pin["units"], roots)
|
||||
return {**proof, "mode": "source", "commit": pin["pin"]}
|
||||
|
||||
def materialize(self, pin, target):
|
||||
rows = pin["units"]
|
||||
for row in rows:
|
||||
safe_path(target, row["file"])
|
||||
if row.get("mapping_error"):
|
||||
raise ValueError(row["mapping_error"])
|
||||
if any(r.get("excerpt") or r["source"].startswith("advisory-fixes") for r in rows):
|
||||
resolved = [self.pair_for(row) for row in rows]
|
||||
identities = set()
|
||||
catalog_files = []
|
||||
for case, side in resolved:
|
||||
_, item = self.catalog_for(case)
|
||||
parent, commit = item.get("parent", ""), item.get("commit", "")
|
||||
if parent != commit and all(re.fullmatch(r"[0-9a-fA-F]{40}", sha) for sha in (parent, commit)):
|
||||
identities.add(self.source_identity(case, side))
|
||||
else:
|
||||
file = case.fixed_file if side == "fixed" else case.vuln_file
|
||||
data = file.read_text()
|
||||
if is_catalog_excerpt(data):
|
||||
data = parseable_excerpt(file, data)
|
||||
mode = "excerpt-parseable"
|
||||
else:
|
||||
# Validate BOTH twins before treating this as a complete-file pair.
|
||||
for twin in (case.vuln_file, case.fixed_file):
|
||||
verify_catalog_file(twin)
|
||||
mode = "catalog-file"
|
||||
catalog_files.append((case, side, mode, data))
|
||||
if catalog_files:
|
||||
if identities:
|
||||
raise ValueError("mixed source and catalog-file materialization at one pair pin")
|
||||
for case, side, mode, data in catalog_files:
|
||||
_materialize_side(target, case, side=side)
|
||||
if mode == "excerpt-parseable":
|
||||
safe_path(target, case.relpath).write_text(data)
|
||||
verify_functions(target, rows)
|
||||
modes = {case.relpath: mode for case, _, mode, _ in catalog_files}
|
||||
for row in rows:
|
||||
if modes[row["file"]] == "excerpt-parseable":
|
||||
row["evidence_scope"] = "in-function-only"
|
||||
return {
|
||||
"mode": "excerpt-parseable" if "excerpt-parseable" in modes.values() else "catalog-file",
|
||||
"files": sum(p.is_file() for p in target.rglob("*")),
|
||||
"bytes": sum(p.stat().st_size for p in target.rglob("*") if p.is_file()),
|
||||
"roots": sorted({source_root(r["file"], r["language"]) for r in rows}),
|
||||
}
|
||||
if len(identities) != 1:
|
||||
raise ValueError("ambiguous source commits at one pair pin")
|
||||
repo, commit = identities.pop()
|
||||
roots = sorted({source_root(row["file"], row["language"]) for row in rows})
|
||||
try:
|
||||
clone = self.fetch_source(repo, commit, roots)
|
||||
proof = export_sources(clone, target, rows, roots)
|
||||
except (OSError, ValueError, subprocess.SubprocessError) as exc:
|
||||
if not self.allow_excerpt_fallback:
|
||||
raise
|
||||
shutil.rmtree(target, ignore_errors=True)
|
||||
for case, side in resolved:
|
||||
_materialize_side(target, case, side=side)
|
||||
return {"mode": "excerpt-fallback", "reason": str(exc), "repo": repo, "commit": commit}
|
||||
verify_functions(target, rows)
|
||||
return {**proof, "mode": "source", "repo": repo, "commit": commit}
|
||||
if all(r["source"] == "dev-php" for r in rows):
|
||||
# Export the exact requested commit from the local recipe checkout.
|
||||
for recipe in (ROOT / "benchmarks/repos").glob("*.toml"):
|
||||
data = tomllib.loads(recipe.read_text())
|
||||
if repo_name(str(data.get("url", ""))) == repo_name(pin["repo"]):
|
||||
checkout = self.cache / "repos" / data["name"] / pin["pin"][:12]
|
||||
return self.export_local(checkout, pin, target)
|
||||
for clone, _ in self.repos[repo_name(pin["repo"])]:
|
||||
check = subprocess.run(["git", "-C", str(clone), "cat-file", "-e", pin["pin"] + "^{commit}"], capture_output=True)
|
||||
if check.returncode == 0:
|
||||
return self.export_local(clone, pin, target)
|
||||
raise ValueError(f"no local clone contains {pin['repo']}@{pin['pin']}")
|
||||
|
||||
|
||||
def is_catalog_excerpt(data):
|
||||
# Older Python exports have this header but no upstream_start field.
|
||||
return re.match(r"\s*(?://|#)\s*Provenance:", data) is not None
|
||||
|
||||
|
||||
def parseable_excerpt(path, data):
|
||||
"""Dedent the code independently of its unindented provenance comments."""
|
||||
language = detect_language(path)
|
||||
lines = data.splitlines(keepends=True)
|
||||
start = 0
|
||||
while start < len(lines) and (not lines[start].strip() or lines[start].lstrip().startswith(("#", "//"))):
|
||||
start += 1
|
||||
header = "".join(lines[:start])
|
||||
if language == "python":
|
||||
header = re.sub(r"(?m)^(\s*)//", r"\1#", header)
|
||||
data = header + textwrap.dedent("".join(lines[start:]))
|
||||
if not "".join(lines[start:]).strip():
|
||||
raise ValueError(f"excerpt does not parse: empty code ({path.name})")
|
||||
if language == "python":
|
||||
try:
|
||||
ast.parse(data, filename=str(path))
|
||||
except SyntaxError as exc:
|
||||
raise ValueError(f"excerpt does not parse: {exc}") from exc
|
||||
else:
|
||||
from openultrasast.semantic.cst import parse_with_cst
|
||||
|
||||
parsed = parse_with_cst(str(path), data, language)
|
||||
if parsed is None:
|
||||
raise ValueError(f"excerpt does not parse: no cheap parser for {language} ({path.name})")
|
||||
if not parsed.parse_ok:
|
||||
raise ValueError(f"excerpt does not parse: {language} {parsed.reason} ({path.name})")
|
||||
return data
|
||||
|
||||
|
||||
def verify_catalog_file(path):
|
||||
"""Admit complete source only; syntactically valid provenance excerpts still fail."""
|
||||
data = path.read_text()
|
||||
if is_catalog_excerpt(data):
|
||||
raise ValueError(f"catalog side is an upstream excerpt without distinct commits: {path.name}")
|
||||
language = detect_language(path)
|
||||
if not data.strip():
|
||||
raise ValueError(f"catalog side is empty: {path.name}")
|
||||
if language == "python":
|
||||
try:
|
||||
ast.parse(data, filename=str(path))
|
||||
except SyntaxError as exc:
|
||||
raise ValueError(f"catalog side is not complete Python source: {path.name}: {exc.msg}") from exc
|
||||
else:
|
||||
from openultrasast.semantic.extra import grammar_for
|
||||
|
||||
grammar = grammar_for(language)
|
||||
if grammar is None:
|
||||
raise ValueError(f"cannot verify complete catalog source for {language}: {path.name}")
|
||||
from tree_sitter import Parser
|
||||
|
||||
if Parser(grammar).parse(data.encode()).root_node.has_error:
|
||||
raise ValueError(f"catalog side is not complete {language} source: {path.name}")
|
||||
|
||||
|
||||
def source_root(file, language):
|
||||
parts = Path(file).parts
|
||||
if language == "java":
|
||||
for i in range(len(parts) - 2):
|
||||
if parts[i : i + 3] == ("src", "main", "java"):
|
||||
return "/".join(parts[: i + 3])
|
||||
# Monorepo and source-layout containers are not themselves a package.
|
||||
if len(parts) > 2 and parts[0] in {"packages", "apps", "src", "web"}:
|
||||
return "/".join(parts[:2])
|
||||
return parts[0] if len(parts) > 1 else "."
|
||||
|
||||
|
||||
def local_imports(source, rows):
|
||||
"""Cheap one-level Python imports from labelled files, without importing code."""
|
||||
found = set()
|
||||
for row in rows:
|
||||
if row["language"] != "python":
|
||||
continue
|
||||
file = safe_path(source, row["file"])
|
||||
try:
|
||||
tree = ast.parse(file.read_bytes())
|
||||
except (SyntaxError, ValueError):
|
||||
continue
|
||||
for node in ast.walk(tree):
|
||||
if isinstance(node, ast.Import):
|
||||
modules, bases = [alias.name for alias in node.names], [source, source / "src"]
|
||||
elif isinstance(node, ast.ImportFrom):
|
||||
modules = [node.module or ""]
|
||||
modules += [".".join(filter(None, (node.module, alias.name))) for alias in node.names if alias.name != "*"]
|
||||
bases = [file.parents[node.level - 1]] if node.level and node.level <= len(file.parents) else [source, source / "src"]
|
||||
else:
|
||||
continue
|
||||
for base in bases:
|
||||
for module in modules:
|
||||
path = base.joinpath(*module.split("."))
|
||||
for candidate in (path.with_suffix(".py"), path / "__init__.py"):
|
||||
if candidate.is_relative_to(source) and candidate.is_file() and not candidate.is_symlink():
|
||||
rel = candidate.relative_to(source).as_posix()
|
||||
if not any(part in EXCLUDED_DIRS for part in Path(rel).parts):
|
||||
found.add(rel)
|
||||
return found
|
||||
|
||||
|
||||
def export_sources(source, target, rows, roots=(".",), cap=EXPORT_CAP):
|
||||
"""Copy labels intact, narrowing oversized roots before bounding context bytes."""
|
||||
labelled = {r["file"] for r in rows}
|
||||
languages = {r["language"] for r in rows}
|
||||
files = set()
|
||||
protected = {parent.as_posix() for rel in labelled for parent in Path(rel).parents}
|
||||
for directory, dirs, names in os.walk(source):
|
||||
dirs[:] = sorted(
|
||||
d
|
||||
for d in dirs
|
||||
if d != ".git"
|
||||
and not (Path(directory) / d).is_symlink()
|
||||
and (d not in EXCLUDED_DIRS or (Path(directory) / d).relative_to(source).as_posix() in protected)
|
||||
)
|
||||
for name in names:
|
||||
file = Path(directory) / name
|
||||
rel = file.relative_to(source).as_posix()
|
||||
excluded = any(part in EXCLUDED_DIRS for part in Path(rel).parts[:-1])
|
||||
if not file.is_symlink() and (
|
||||
rel in labelled
|
||||
or (
|
||||
not excluded
|
||||
and detect_language(file) in languages
|
||||
and any(root == "." or rel == root or rel.startswith(root + "/") for root in roots)
|
||||
)
|
||||
):
|
||||
files.add(rel)
|
||||
missing = labelled - files
|
||||
if missing:
|
||||
raise ValueError(f"labelled files missing or excluded from export: {sorted(missing)}")
|
||||
imports = local_imports(source, rows)
|
||||
files.update(imports)
|
||||
sizes = {rel: safe_path(source, rel).stat().st_size for rel in files}
|
||||
original_roots = sorted(roots)
|
||||
roots = original_roots
|
||||
# Avoid lexicographically truncating a whole large package such as vllm.
|
||||
if sum(sizes.values()) > cap:
|
||||
roots = sorted({str(Path(rel).parent) for rel in labelled})
|
||||
files = {
|
||||
rel for rel in files if rel in labelled or rel in imports or any(root == "." or rel.startswith(root + "/") for root in roots)
|
||||
}
|
||||
proof = {
|
||||
"bytes": 0,
|
||||
"files": 0,
|
||||
"cap_bytes": cap,
|
||||
"cap_hit": False,
|
||||
"excludes": sorted(EXCLUDED_DIRS),
|
||||
"roots": roots,
|
||||
"requested_roots": original_roots,
|
||||
"scope_narrowed": roots != original_roots,
|
||||
"import_files": sorted(imports),
|
||||
}
|
||||
for rel in sorted(files, key=lambda rel: (rel not in labelled, rel)):
|
||||
file = safe_path(source, rel)
|
||||
size = sizes[rel]
|
||||
# The cap bounds context only. No labelled source is ever cut or omitted.
|
||||
if proof["bytes"] + size > cap:
|
||||
proof["cap_hit"] = True
|
||||
if rel not in labelled:
|
||||
continue
|
||||
dest = safe_path(target, rel)
|
||||
dest.parent.mkdir(parents=True, exist_ok=True)
|
||||
dest.write_bytes(file.read_bytes())
|
||||
proof["bytes"] += size
|
||||
proof["files"] += 1
|
||||
return proof
|
||||
|
||||
|
||||
def verify_functions(root, rows):
|
||||
for row in rows:
|
||||
data = safe_path(root, row["file"]).read_text(errors="replace")
|
||||
if not re.search(r"(?<![\w$])" + re.escape(row["function"]) + r"(?![\w$])", data):
|
||||
raise ValueError(f"labelled function not found in fetched source: {row['file']}::{row['function']}")
|
||||
|
||||
|
||||
def freeze_source(target):
|
||||
target.mkdir(parents=True)
|
||||
archive = target.parent / "source.tar"
|
||||
with archive.open("wb") as stream:
|
||||
subprocess.run(["git", "-C", str(ROOT), "archive", "HEAD", "src"], stdout=stream, check=True)
|
||||
subprocess.run(["tar", "-xf", str(archive), "-C", str(target)], check=True)
|
||||
archive.unlink()
|
||||
overlay = {}
|
||||
for name in SNAPSHOT_FILES:
|
||||
data = (ROOT / name).read_bytes()
|
||||
dest = target / name
|
||||
dest.parent.mkdir(parents=True, exist_ok=True)
|
||||
dest.write_bytes(data)
|
||||
overlay[name] = hashlib.sha256(data).hexdigest()
|
||||
head = subprocess.run(["git", "-C", str(ROOT), "rev-parse", "HEAD"], capture_output=True, text=True, check=True).stdout.strip()
|
||||
return {"head": head, "overlay_sha256": overlay}
|
||||
|
||||
|
||||
def docker_command(source, checkout, out, name, image, deadline, question_deadline):
|
||||
return [
|
||||
"docker",
|
||||
"run",
|
||||
"--rm",
|
||||
"--init",
|
||||
"--name",
|
||||
name,
|
||||
"--network",
|
||||
"none",
|
||||
"--memory",
|
||||
"3g",
|
||||
"--pull",
|
||||
"never",
|
||||
"--entrypoint",
|
||||
"python",
|
||||
"-e",
|
||||
"PYTHONPATH=/frozen/src",
|
||||
"-v",
|
||||
f"{source}:/frozen:ro",
|
||||
"-v",
|
||||
f"{checkout}:/case:ro",
|
||||
"-v",
|
||||
f"{out}:/out",
|
||||
image,
|
||||
"/frozen/benchmarks/learn/engine_trace_worker.py",
|
||||
"--worker",
|
||||
"/out/pin.json",
|
||||
"--deadline",
|
||||
str(deadline),
|
||||
"--question-deadline",
|
||||
str(question_deadline),
|
||||
]
|
||||
|
||||
|
||||
def prepare_questions(pin, root, question_deadline):
|
||||
"""Prepare metadata on the host, before entering the dependency-minimal image."""
|
||||
active = [u for u in pin["units"] if u["supported"]]
|
||||
from openultrasast.learn.features import entry_names
|
||||
from openultrasast.mapping import php_hook_callbacks
|
||||
from openultrasast.plane.harvest import declares
|
||||
from openultrasast.preprocess import preprocess_repository
|
||||
|
||||
_, targets = preprocess_repository(root)
|
||||
entries, _ = entry_names(root, {u["file"] for u in active})
|
||||
hooks = ";".join(f"{h}:{f}" for h, functions in php_hook_callbacks(root, targets).items() for f in functions)
|
||||
requests = {}
|
||||
for u in active:
|
||||
specs = taint_specs(language="javascript" if u["language"] == "typescript" else u["language"])
|
||||
if not declares(safe_path(root, u["file"]).read_text(errors="replace").splitlines(), u["file"], u["function"]):
|
||||
raise ValueError(f"labelled function not declared: {u['file']}::{u['function']}")
|
||||
excerpt = u.get("evidence_scope") == "in-function-only" or pin.get("materialization", {}).get("mode") == "excerpt-fallback"
|
||||
params = request_params(
|
||||
specs[u["family"]],
|
||||
file=u["file"],
|
||||
function=u["function"],
|
||||
trace=True,
|
||||
parameter_sources=excerpt or u["function"] in entries,
|
||||
call_depth=0 if excerpt else 3,
|
||||
hook_callbacks=hooks,
|
||||
)
|
||||
params["questionDeadline"] = str(question_deadline)
|
||||
requests[u["unit"]] = params
|
||||
return requests
|
||||
|
||||
|
||||
def run_pin(pin, source, provenance, inputs, out, args):
|
||||
with tempfile.TemporaryDirectory(prefix="trace-pin-", dir=out) as scratch:
|
||||
scratch = Path(scratch)
|
||||
checkout, output = scratch / "case", scratch / "output"
|
||||
output.mkdir()
|
||||
try:
|
||||
pin = {**pin, "materialization": inputs.materialize(pin, checkout)}
|
||||
write_json(output / "pin.json", {**pin, "questions": prepare_questions(pin, checkout, args.question_deadline)})
|
||||
name = "ousast-trace-" + pin_name(pin)[:16]
|
||||
command = docker_command(source, checkout, output, name, args.image, args.deadline, args.question_deadline)
|
||||
started = time.monotonic()
|
||||
done = None
|
||||
try:
|
||||
done = subprocess.run(command, capture_output=True, text=True, timeout=args.deadline)
|
||||
status, reason = "failed", f"container exit {done.returncode}: {done.stderr[-1000:]}"
|
||||
(output / "container.log").write_text(done.stdout + "\n" + done.stderr)
|
||||
except subprocess.TimeoutExpired:
|
||||
done = None
|
||||
status, reason = "timeout", f"container deadline {args.deadline}s"
|
||||
finally:
|
||||
# Never start the next pin unless this container is known to be gone.
|
||||
try:
|
||||
removed = subprocess.run(["docker", "rm", "-f", name], capture_output=True, timeout=20)
|
||||
except subprocess.SubprocessError as exc:
|
||||
raise RuntimeError(f"cannot confirm container {name} stopped; refusing another launch") from exc
|
||||
if done is None and removed.returncode != 0:
|
||||
raise RuntimeError(f"cannot remove timed-out container {name}; refusing another launch")
|
||||
result_path = output / "result.json"
|
||||
if not result_path.exists():
|
||||
reason = f"container produced no result: {reason}"
|
||||
record = json.loads(result_path.read_text()) if result_path.exists() else {**pin, "units": []}
|
||||
if not record.get("done"):
|
||||
previous = {u["unit"]: u for u in record["units"]}
|
||||
record["units"] = []
|
||||
for unit in pin["units"]:
|
||||
row = previous.get(unit["unit"], unit_record(unit, status))
|
||||
if row["status"] not in {"path", "asked-nothing"}:
|
||||
row.update(
|
||||
status=status if unit["supported"] else "unsupported", reason=reason if unit["supported"] else unit["reason"]
|
||||
)
|
||||
record["units"].append(row)
|
||||
record["container_exit"] = done.returncode if done else None
|
||||
if done is not None and done.returncode != 0:
|
||||
for row in record["units"]:
|
||||
if row["supported"]:
|
||||
row.update(status="failed", reason=reason)
|
||||
record.update(done=True, seconds=time.monotonic() - started, analyzer=provenance, image=args.image)
|
||||
if (output / "container.log").exists():
|
||||
shutil.copyfile(output / "container.log", out / pin_name(pin).replace(".json", ".log"))
|
||||
except (OSError, ValueError, subprocess.SubprocessError) as exc:
|
||||
record = {
|
||||
**pin,
|
||||
"done": True,
|
||||
"analyzer": provenance,
|
||||
"units": [unit_record(u, "failed" if u["supported"] else "unsupported", str(exc)) for u in pin["units"]],
|
||||
}
|
||||
save_pin(out / pin_name(pin), record)
|
||||
print(
|
||||
json.dumps({"repo": pin["repo"], "pin": pin["pin"], "statuses": dict(Counter(u["status"] for u in record["units"]))}),
|
||||
flush=True,
|
||||
)
|
||||
return any(u["status"] in {"failed", "timeout"} for u in record["units"])
|
||||
|
||||
|
||||
def save_pin(path, record):
|
||||
if path.exists():
|
||||
old = json.loads(path.read_text())
|
||||
selected = {u["unit"] for u in record["units"]}
|
||||
record["units"].extend(u for u in old.get("units", []) if u["unit"] not in selected)
|
||||
write_json(path, record)
|
||||
|
||||
|
||||
def summary(out):
|
||||
if not out.is_dir():
|
||||
raise ValueError(f"results root does not exist: {out}")
|
||||
family, source = defaultdict(Counter), defaultdict(Counter)
|
||||
pairs = defaultdict(dict)
|
||||
units = {}
|
||||
result_files = 0
|
||||
for path in sorted(out.glob("*.json")):
|
||||
record = json.loads(path.read_text())
|
||||
if not record.get("done"):
|
||||
continue
|
||||
result_files += 1
|
||||
for unit in record["units"]:
|
||||
if unit["unit"] in units:
|
||||
raise ValueError(f"duplicate result for {unit['unit']}")
|
||||
units[unit["unit"]] = unit
|
||||
family[unit["family"]][unit["status"]] += 1
|
||||
source[unit["source"]][unit["status"]] += 1
|
||||
if unit["pair"]:
|
||||
pairs[unit["family"], unit["pair"]][unit["label"]] = unit["status"]
|
||||
asymmetry = Counter()
|
||||
completed = Counter()
|
||||
for sides in pairs.values():
|
||||
if set(sides) != {0, 1}:
|
||||
asymmetry["incomplete-pair"] += 1
|
||||
continue
|
||||
key = {(True, False): "vulnerable-only", (False, True): "fixed-only", (True, True): "both", (False, False): "neither"}[
|
||||
sides[1] == "path", sides[0] == "path"
|
||||
]
|
||||
asymmetry[key] += 1
|
||||
if all(s in {"path", "asked-nothing"} for s in sides.values()):
|
||||
completed[key] += 1
|
||||
return {
|
||||
"result_files_read": result_files,
|
||||
"units": len(units),
|
||||
"by_family": {k: {s: v[s] for s in STATUSES} for k, v in sorted(family.items())},
|
||||
"by_source": {k: {s: v[s] for s in STATUSES} for k, v in sorted(source.items())},
|
||||
"pair_asymmetry": dict(asymmetry),
|
||||
"pair_asymmetry_both_answered": dict(completed),
|
||||
}
|
||||
|
||||
|
||||
def main():
|
||||
parser = argparse.ArgumentParser(description=__doc__)
|
||||
parser.add_argument("--units", type=Path, default=ROOT / "plane/experiments/exp-005-graph-slice.units.jsonl")
|
||||
parser.add_argument("--examples", type=Path, help="frozen example JSONL instead of the local memory store")
|
||||
parser.add_argument("--memory", type=Path, default=Path.home() / "ousast-results/plane/memory")
|
||||
parser.add_argument("--cache", type=Path, default=Path.home() / ".cache/openultrasast")
|
||||
parser.add_argument("--out", type=Path, default=DEFAULT_OUT)
|
||||
parser.add_argument("--image", default="openultrasast:dev")
|
||||
parser.add_argument("--deadline", type=int, default=900)
|
||||
parser.add_argument("--question-deadline", type=int, default=120)
|
||||
parser.add_argument("--limit", type=int, default=0)
|
||||
parser.add_argument("--only-family")
|
||||
parser.add_argument("--only-source", default="", help="comma-separated source catalog IDs")
|
||||
parser.add_argument("--rerun-status", default="failed,timeout", help="comma-separated statuses to re-run (including timeout)")
|
||||
parser.add_argument(
|
||||
"--allow-excerpt-fallback", action="store_true", help="allow fragments only if fetching/exporting resolved source fails"
|
||||
)
|
||||
parser.add_argument("--include-typescript", action="store_true")
|
||||
parser.add_argument("--dry-run", action="store_true")
|
||||
parser.add_argument("--summary", action="store_true")
|
||||
args = parser.parse_args()
|
||||
if min(args.deadline, args.question_deadline) <= 0 or args.limit < 0:
|
||||
parser.error("deadlines must be positive; limit must be nonnegative")
|
||||
rerun_status = set(filter(None, args.rerun_status.split(",")))
|
||||
if rerun_status - set(STATUSES):
|
||||
parser.error("unknown --rerun-status value")
|
||||
if args.summary:
|
||||
print(json.dumps(summary(args.out), indent=2))
|
||||
return 0
|
||||
# Even inherited opt-ins must not make catalog resolution fetch.
|
||||
os.environ["GIT_NO_LAZY_FETCH"] = "1"
|
||||
os.environ.pop("OPENULTRASAST_PAIRS_NETWORK", None)
|
||||
os.environ.pop("OPENULTRASAST_REPOS_NETWORK", None)
|
||||
examples = read_jsonl(args.examples) if args.examples else [r.row for r in FileStore(args.memory).rows(kind="example")]
|
||||
try:
|
||||
units = join_units(read_jsonl(args.units), examples, args.include_typescript)
|
||||
except ValueError as exc:
|
||||
parser.error(str(exc))
|
||||
inputs = Inputs(ROOT, args.cache, args.allow_excerpt_fallback)
|
||||
# Resolve pair source names for leak-audit coverage before calculating the resume key.
|
||||
for unit in units:
|
||||
if unit.get("excerpt") or unit["source"].startswith("advisory-fixes"):
|
||||
try:
|
||||
case, _ = inputs.pair_for(unit)
|
||||
catalog_source, _ = inputs.catalog_for(case)
|
||||
unit["source"] = catalog_source if catalog_source.startswith("advisory-fixes") else case.slice
|
||||
unit["excerpt"] = True
|
||||
except ValueError as exc:
|
||||
unit["mapping_error"] = str(exc)
|
||||
plan = make_plan(
|
||||
units,
|
||||
args.out,
|
||||
args.only_family,
|
||||
args.limit,
|
||||
include_typescript=args.include_typescript,
|
||||
only_source=set(filter(None, args.only_source.split(","))),
|
||||
rerun_status=rerun_status,
|
||||
allow_excerpt_fallback=args.allow_excerpt_fallback,
|
||||
)
|
||||
# Planning is read-only, including unsupported pins. All persistence stays below this return.
|
||||
if args.dry_run:
|
||||
print(
|
||||
json.dumps(
|
||||
{
|
||||
"pins": len(plan),
|
||||
"runnable_pins": sum(any(u["supported"] for u in p["units"]) for p in plan),
|
||||
"units": sum(len(p["units"]) for p in plan),
|
||||
"estimated_seconds": sum(p["estimated_seconds"] for p in plan),
|
||||
"estimate_basis": "inventory medians: excerpt 31/47s, repository 182/498s; not a deadline guarantee",
|
||||
"plan": plan,
|
||||
},
|
||||
indent=2,
|
||||
)
|
||||
)
|
||||
return 0
|
||||
args.out.mkdir(parents=True, exist_ok=True)
|
||||
lock = (args.out / ".lock").open("w")
|
||||
try:
|
||||
fcntl.flock(lock, fcntl.LOCK_EX | fcntl.LOCK_NB)
|
||||
except BlockingIOError:
|
||||
parser.error("another trace runner holds this results root")
|
||||
failures = False
|
||||
if any(u["supported"] for pin in plan for u in pin["units"]):
|
||||
image = subprocess.run(
|
||||
["docker", "image", "inspect", "--format", "{{.Id}}", args.image], capture_output=True, text=True, check=True
|
||||
)
|
||||
args.image = image.stdout.strip()
|
||||
if not args.image:
|
||||
raise ValueError("Docker returned no local image identity")
|
||||
with tempfile.TemporaryDirectory(prefix="trace-source-", dir=args.out) as scratch:
|
||||
source = Path(scratch) / "frozen"
|
||||
provenance = freeze_source(source)
|
||||
progress = {"pins_done": 0, "pins_total": len(plan), "last_pin": None, "statuses": {}}
|
||||
counts = Counter()
|
||||
write_json(args.out / "progress.json", progress)
|
||||
for pin in plan:
|
||||
if (args.out / "STOP").exists():
|
||||
break
|
||||
if not any(u["supported"] for u in pin["units"]):
|
||||
save_pin(
|
||||
args.out / pin_name(pin),
|
||||
{**pin, "done": True, "units": [unit_record(u, "unsupported", u["reason"]) for u in pin["units"]]},
|
||||
)
|
||||
else:
|
||||
failures |= run_pin(pin, source, provenance, inputs, args.out, args)
|
||||
result = json.loads((args.out / pin_name(pin)).read_text())
|
||||
selected = {u["unit"] for u in pin["units"]}
|
||||
counts.update(u["status"] for u in result["units"] if u["unit"] in selected)
|
||||
progress.update(pins_done=progress["pins_done"] + 1, last_pin={"repo": pin["repo"], "pin": pin["pin"]}, statuses=dict(counts))
|
||||
write_json(args.out / "progress.json", progress)
|
||||
return 2 if failures else 0
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
raise SystemExit(main())
|
||||
@@ -0,0 +1,51 @@
|
||||
"""Lossless parsing of the optional shortest-path trace, with explicit evidence limits."""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import re
|
||||
from collections.abc import Mapping
|
||||
from typing import Any
|
||||
|
||||
_STEP = re.compile(r"^(.*):(-?\d+) \[([^\]]+)\] ([\s\S]*)$")
|
||||
|
||||
|
||||
def parse_trace(row: Mapping[str, Any]) -> dict[str, Any]:
|
||||
"""Preserve row-level discharge evidence; never guess which node sanitized a path.
|
||||
|
||||
The query emits basename-only files and truncated code. A bound can be off-path;
|
||||
it remains in ``guard`` even when no trace step exactly matches it. Invalid steps
|
||||
raise instead of turning an unreadable trace into an empty path.
|
||||
"""
|
||||
raw = row.get("trace", [])
|
||||
if not isinstance(raw, list):
|
||||
raise ValueError("trace must be an array")
|
||||
steps = []
|
||||
bound = str(row.get("bound") or "")
|
||||
for index, text in enumerate(raw):
|
||||
match = _STEP.fullmatch(text) if isinstance(text, str) else None
|
||||
if match is None:
|
||||
raise ValueError(f"invalid trace step: {text!r}")
|
||||
location, line, node_label, code = match.groups()
|
||||
file, separator, method = location.partition(":")
|
||||
roles = ["source"] if index == 0 else ["step"]
|
||||
if index == len(raw) - 1:
|
||||
roles = ["source", "sink"] if index == 0 else ["sink"]
|
||||
if bound and code == bound:
|
||||
roles.append("guard")
|
||||
steps.append(
|
||||
{
|
||||
"file": file,
|
||||
"method": method if separator else "?",
|
||||
"line": int(line),
|
||||
"node_label": node_label,
|
||||
"code": code,
|
||||
"roles": roles,
|
||||
"raw": text,
|
||||
}
|
||||
)
|
||||
return {
|
||||
"steps": steps,
|
||||
"sanitized": bool(row.get("sanitized", False)),
|
||||
"sanitizer": {"on_path": bool(row.get("sanitized", False)), "step": None},
|
||||
"guard": {"text": bound, "bounded": bool(row.get("bounded", False))},
|
||||
}
|
||||
@@ -0,0 +1,453 @@
|
||||
"""Container-only Joern trace worker; host prepares all corpus and question metadata.
|
||||
|
||||
Keep this module independent of learn/plane and their optional dependencies. The
|
||||
frozen host invokes this file directly in the shipped, offline analyzer image.
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import argparse
|
||||
import contextlib
|
||||
import hashlib
|
||||
import json
|
||||
import os
|
||||
import selectors
|
||||
import signal
|
||||
import subprocess
|
||||
import tempfile
|
||||
import time
|
||||
from collections import Counter
|
||||
from pathlib import Path
|
||||
|
||||
from engine_trace_parse import parse_trace
|
||||
|
||||
from openultrasast.cpg.backend import DATAFLOW_OVERLAY, JoernBackend, _render
|
||||
|
||||
|
||||
def write_json(path, value):
|
||||
path.parent.mkdir(parents=True, exist_ok=True)
|
||||
tmp = path.with_suffix(".tmp")
|
||||
tmp.write_text(json.dumps(value, indent=2, sort_keys=True) + "\n")
|
||||
tmp.replace(path)
|
||||
|
||||
|
||||
def safe_path(root, name):
|
||||
path = (root / name).resolve()
|
||||
if not path.is_relative_to(root.resolve()):
|
||||
raise ValueError(f"path escapes input root: {name}")
|
||||
return path
|
||||
|
||||
|
||||
FRONTENDS = {"php2cpg", "pysrc2cpg", "jssrc2cpg", "javasrc2cpg", "c2cpg", "joern-parse"}
|
||||
MIN_CPG_BYTES = 1024
|
||||
HEAP_MB = 2560
|
||||
ENGINE_BUG = "engine bug: jssrc2cpg ObjectPropertyCallLinker"
|
||||
EXCLUDED_DIRS = frozenset({"tests", "test", "docs", "node_modules", "vendor", "dist", "build", "templates", "migrations"})
|
||||
|
||||
|
||||
def instrument_failure(proof):
|
||||
if any(stage.get("oom") for stage in proof.get("jvm", [])):
|
||||
return "JVM out of memory"
|
||||
if not proof.get("files") or not proof.get("bytes"):
|
||||
return "instrument read zero files or zero bytes"
|
||||
if proof.get("cpg_bytes", 0) <= MIN_CPG_BYTES:
|
||||
return "frontend wrote no CPG"
|
||||
for stage in proof.get("jvm", []):
|
||||
if stage["command"] in FRONTENDS:
|
||||
if stage.get("cpg_bytes", 0) <= MIN_CPG_BYTES:
|
||||
return "frontend wrote no CPG"
|
||||
elif stage["command"] == "joern" and stage["seconds"] < 5:
|
||||
return f"implausibly fast JVM step: {stage['command']} in {stage['seconds']:.3f}s (<5s)"
|
||||
if not proof.get("jvm"):
|
||||
return "no JVM step recorded"
|
||||
return ""
|
||||
|
||||
|
||||
class MeasuredBackend(JoernBackend):
|
||||
"""Use the shipped backend with a process watchdog, no concurrent query threads."""
|
||||
|
||||
def __init__(self, deadline, question_deadline, checkpoint):
|
||||
super().__init__(build_timeout=deadline, query_timeout=deadline, session_transport=False, heap_mb=HEAP_MB)
|
||||
self.deadline = time.monotonic() + deadline
|
||||
self.question_deadline = question_deadline
|
||||
self.checkpoint = checkpoint
|
||||
self.stages = []
|
||||
self.asked = []
|
||||
self.answers = {}
|
||||
self.completion_counts = Counter()
|
||||
self.timed_out = False
|
||||
self.fatal = ""
|
||||
self.labelled_files = []
|
||||
self.retry_pending = False
|
||||
self.retries = []
|
||||
|
||||
def _apply_overlays(self, cpg_path, scratch):
|
||||
result = super()._apply_overlays(cpg_path, scratch)
|
||||
# An optional overlay failure must not prevent the taint-only attempt.
|
||||
if self.fatal == ENGINE_BUG and not self.timed_out:
|
||||
self.retry_pending = True
|
||||
self.fatal = ""
|
||||
return result
|
||||
|
||||
def _batch_once(self, cpg_path, query, requests):
|
||||
if self.fatal or self.timed_out:
|
||||
return None
|
||||
if not self.retry_pending:
|
||||
answers = super()._batch_once(cpg_path, query, requests)
|
||||
if self.fatal != ENGINE_BUG:
|
||||
return answers
|
||||
if query != "taint" or self.timed_out or self.retries:
|
||||
self.fatal = ENGINE_BUG
|
||||
return None
|
||||
self.retry_pending = False
|
||||
self.fatal = ""
|
||||
retry = {"reason": ENGINE_BUG, "mode": "taint-only", "completed": False}
|
||||
self.retries.append(retry)
|
||||
# Do not reuse rows/markers from the failed attempt as completion evidence.
|
||||
self.asked.clear()
|
||||
self.answers.clear()
|
||||
self.completion_counts.clear()
|
||||
original = self.queries_dir
|
||||
with tempfile.TemporaryDirectory(prefix="trace-taint-", dir=cpg_path.parent) as scratch:
|
||||
script = (original / "taint.sc").read_text()
|
||||
marker = "importCpg(cpgFile)"
|
||||
if script.count(marker) != 1:
|
||||
self.fatal = ENGINE_BUG
|
||||
return None
|
||||
# Loading without enhancement skips the failing JS postprocessing passes.
|
||||
# A raw graph without dataflow cannot establish a negative answer.
|
||||
script = script.replace(
|
||||
marker,
|
||||
"importCpg(cpgFile, enhance = false)\n"
|
||||
f' require(cpg.metaData.overlays.l.contains("{DATAFLOW_OVERLAY}"), '
|
||||
'"taint-only retry requires saved dataflowOss overlay")',
|
||||
)
|
||||
self.queries_dir = Path(scratch)
|
||||
(self.queries_dir / "taint.sc").write_text(script)
|
||||
try:
|
||||
answers = super()._batch_once(cpg_path, query, requests)
|
||||
finally:
|
||||
self.queries_dir = original
|
||||
retry["completed"] = answers is not None and not self.fatal and not self.timed_out
|
||||
if not retry["completed"]:
|
||||
self.fatal = ENGINE_BUG
|
||||
return answers
|
||||
|
||||
def _run(self, command, *, timeout, cwd=None):
|
||||
if self.fatal or self.timed_out:
|
||||
return None
|
||||
if time.monotonic() >= self.deadline:
|
||||
self.timed_out = True
|
||||
return None
|
||||
started = time.monotonic()
|
||||
end = min(self.deadline, started + timeout)
|
||||
question_end = end
|
||||
binary = Path(command[0]).name
|
||||
frontend = binary in FRONTENDS
|
||||
query = binary == "joern" and "--script" in command
|
||||
output = bytearray()
|
||||
pending = b""
|
||||
process = subprocess.Popen(
|
||||
command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=cwd, env=self._jvm_env(), start_new_session=True
|
||||
)
|
||||
try:
|
||||
with selectors.DefaultSelector() as selector:
|
||||
selector.register(process.stdout, selectors.EVENT_READ)
|
||||
while selector.get_map():
|
||||
if time.monotonic() >= min(end, question_end):
|
||||
self.timed_out = True
|
||||
break
|
||||
for key, _ in selector.select(min(0.2, max(0, min(end, question_end) - time.monotonic()))):
|
||||
chunk = os.read(key.fileobj.fileno(), 65536)
|
||||
if not chunk:
|
||||
selector.unregister(key.fileobj)
|
||||
continue
|
||||
output.extend(chunk)
|
||||
pending += chunk
|
||||
while b"\n" in pending:
|
||||
line, pending = pending.split(b"\n", 1)
|
||||
try:
|
||||
event = json.loads(line)
|
||||
except ValueError:
|
||||
continue
|
||||
if not isinstance(event, dict):
|
||||
continue
|
||||
if "__question__" in event:
|
||||
self.asked.append(event["__question__"])
|
||||
question_end = min(end, time.monotonic() + self.question_deadline)
|
||||
if "id" in event and "rows" in event:
|
||||
self.answers[event["id"]] = event["rows"]
|
||||
self.completion_counts[event["id"]] += 1
|
||||
question_end = end
|
||||
self.checkpoint()
|
||||
if not self.timed_out:
|
||||
process.wait(timeout=max(0.01, end - time.monotonic()))
|
||||
finally:
|
||||
with contextlib.suppress(ProcessLookupError):
|
||||
os.killpg(process.pid, signal.SIGKILL)
|
||||
process.wait(timeout=5)
|
||||
process.stdout.close()
|
||||
elapsed = time.monotonic() - started
|
||||
if frontend or query:
|
||||
stage = {
|
||||
"command": binary,
|
||||
"argv": command,
|
||||
"heap_mb": self._heap_mb(),
|
||||
"java_tool_options": self._jvm_env()["JAVA_TOOL_OPTIONS"],
|
||||
"seconds": elapsed,
|
||||
"exit": process.returncode,
|
||||
"oom": b"OutOfMemoryError" in output,
|
||||
}
|
||||
lines = output.decode(errors="replace").splitlines()
|
||||
stage["labelled_file_logs"] = {
|
||||
file: [line for line in lines if file in line or Path(file).name in line]
|
||||
for file in self.labelled_files
|
||||
if any(file in line or Path(file).name in line for line in lines)
|
||||
}
|
||||
self.stages.append(stage)
|
||||
if query and process.returncode != 0:
|
||||
# stderr is merged into stdout above; drop JVM frames before taking
|
||||
# the tail so they cannot crowd out the exception's message.
|
||||
lines = [
|
||||
line.strip()
|
||||
for line in output.decode(errors="replace").splitlines()
|
||||
if line.strip() and not line.strip().startswith(("at ", "... "))
|
||||
]
|
||||
stage["output_tail"] = "\n".join(lines[-20:])
|
||||
if not self.timed_out:
|
||||
self.fatal = f"joern query exited {process.returncode}"
|
||||
if frontend:
|
||||
# Capture the exact frontend output before build failure cleanup or overlays.
|
||||
output_arg = next((i + 1 for i, arg in enumerate(command[:-1]) if arg in {"-o", "--output"}), None)
|
||||
cpg_path = (Path(cwd or Path.cwd()) / command[output_arg]).resolve() if output_arg is not None else None
|
||||
stage["cpg_path"] = str(cpg_path) if cpg_path is not None else None
|
||||
stage["cpg_bytes"] = cpg_path.stat().st_size if cpg_path is not None and cpg_path.is_file() else 0
|
||||
if stage["cpg_bytes"] <= MIN_CPG_BYTES:
|
||||
self.fatal = "frontend wrote no CPG"
|
||||
elif elapsed < 5 and not self.fatal:
|
||||
self.fatal = f"implausibly fast JVM step: {command[0]} in {elapsed:.3f}s (<5s)"
|
||||
if b"ObjectPropertyCallLinker failed" in output and b"Assignment statement with 3 arguments" in output:
|
||||
self.fatal = ENGINE_BUG
|
||||
stage["engine_bug"] = ENGINE_BUG
|
||||
if b"OutOfMemoryError" in output:
|
||||
self.fatal = "JVM out of memory"
|
||||
self.checkpoint()
|
||||
return subprocess.CompletedProcess(command, process.returncode if not self.fatal else 1, output.decode(errors="replace"), "")
|
||||
|
||||
|
||||
def unit_record(unit, status, reason="", **extra):
|
||||
return {
|
||||
**unit,
|
||||
"status": status,
|
||||
"reason": reason,
|
||||
"witness_rows": [],
|
||||
"traces": [],
|
||||
"instrument": {"files": [], "bytes": 0, "cpg_path": None, "cpg_bytes": 0, "jvm": []},
|
||||
"questions": {},
|
||||
"questions_asked": [],
|
||||
"questions_completed": [],
|
||||
**extra,
|
||||
}
|
||||
|
||||
|
||||
def worker_language(pin, root, out, deadline, question_deadline, persist=write_json):
|
||||
proof = {
|
||||
"files": [],
|
||||
"bytes": 0,
|
||||
"cpg_path": None,
|
||||
"cpg_bytes": 0,
|
||||
"jvm": [],
|
||||
"heap_mb": HEAP_MB,
|
||||
"excludes": sorted(EXCLUDED_DIRS),
|
||||
"materialization": pin.get("materialization", {}),
|
||||
}
|
||||
record = {
|
||||
**pin,
|
||||
"done": False,
|
||||
"instrument": proof,
|
||||
"units": [unit_record(u, "unsupported" if not u["supported"] else "timeout", u["reason"]) for u in pin["units"]],
|
||||
}
|
||||
|
||||
def checkpoint():
|
||||
# Checkpoints retain raw completed rows even if the outer container deadline fires.
|
||||
# They remain timeout until the finished JVM's instrument proof is validated.
|
||||
for stage in backend.stages:
|
||||
if "cpg_path" in stage:
|
||||
proof["cpg_path"] = stage["cpg_path"]
|
||||
proof["cpg_bytes"] = stage["cpg_bytes"]
|
||||
for row in record["units"]:
|
||||
rid = row["unit"]
|
||||
row["instrument"] = proof
|
||||
row["questions_asked"] = [rid] if rid in backend.asked else []
|
||||
if rid in backend.answers:
|
||||
row["questions_completed"] = [rid]
|
||||
row["witness_rows"] = [r for r in backend.answers[rid] if isinstance(r, dict) and "kind" not in r and "sink" in r]
|
||||
persist(out, record)
|
||||
|
||||
backend = MeasuredBackend(deadline, question_deadline, checkpoint)
|
||||
proof["jvm"] = backend.stages
|
||||
proof["retries"] = getattr(backend, "retries", [])
|
||||
backend.labelled_files = sorted({u["file"] for u in pin["units"] if u["supported"]})
|
||||
cpg = None
|
||||
try:
|
||||
active = [u for u in pin["units"] if u["supported"]]
|
||||
for file in sorted({u["file"] for u in active}):
|
||||
data = safe_path(root, file).read_bytes()
|
||||
proof["files"].append({"file": file, "bytes": len(data), "sha256": hashlib.sha256(data).hexdigest()})
|
||||
proof["bytes"] += len(data)
|
||||
print(f"read {file}: {len(data)} bytes", flush=True)
|
||||
if not proof["bytes"]:
|
||||
raise ValueError("instrument read zero files or zero bytes")
|
||||
languages = {"javascript" if u["language"] == "typescript" else u["language"] for u in active}
|
||||
if len(languages) != 1:
|
||||
raise ValueError(f"one pin has multiple frontend languages: {sorted(languages)}")
|
||||
language = next(iter(languages))
|
||||
requests = {rid: {key: _render(value) for key, value in params.items()} for rid, params in pin["questions"].items()}
|
||||
if set(requests) != {u["unit"] for u in active}:
|
||||
raise ValueError("prepared questions do not match supported units")
|
||||
record["questions"] = requests
|
||||
for unit_row in record["units"]:
|
||||
unit_row["questions"] = {unit_row["unit"]: requests[unit_row["unit"]]} if unit_row["unit"] in requests else {}
|
||||
persist(out, record)
|
||||
cpg = backend.build(root, language=language)
|
||||
if cpg is None:
|
||||
raise ValueError(backend.fatal or backend.last_failure or ("pin deadline" if backend.timed_out else "CPG build failed"))
|
||||
if proof["cpg_path"] is None:
|
||||
proof["cpg_path"] = str(cpg.cpg_path)
|
||||
proof["cpg_bytes"] = cpg.cpg_path.stat().st_size if cpg.cpg_path.is_file() else 0
|
||||
failure = instrument_failure(proof)
|
||||
if failure:
|
||||
raise ValueError(failure)
|
||||
if cpg.run_batch_once is None:
|
||||
raise ValueError("backend offers no batch operation")
|
||||
answers = cpg.run_batch_once("taint", requests) or {}
|
||||
record["census"] = answers.get("__census__", [])
|
||||
failure = backend.fatal or instrument_failure(proof)
|
||||
if not backend.asked and not backend.timed_out:
|
||||
failure = failure or "query produced no markers"
|
||||
census = record["census"]
|
||||
if census and int(census[0].get("files", 0)) <= 0:
|
||||
failure = failure or "query census read zero files"
|
||||
elif not census and not backend.timed_out:
|
||||
failure = failure or "query census is missing"
|
||||
for index, unit in enumerate(pin["units"]):
|
||||
if not unit["supported"]:
|
||||
continue
|
||||
rid = unit["unit"]
|
||||
extra = {
|
||||
"questions": {rid: requests[rid]},
|
||||
"questions_asked": [rid] if rid in backend.asked else [],
|
||||
"questions_completed": [rid] if rid in answers else [],
|
||||
"instrument": proof,
|
||||
}
|
||||
rows = answers.get(rid)
|
||||
# A batch across backend shards is complete only when every shard answered.
|
||||
# The legacy merger can return a union even when one shard was killed.
|
||||
shards = int(census[0].get("shards", 1)) if census else 1
|
||||
shard_incomplete = shards > 1 and getattr(backend, "completion_counts", {}).get(rid, 0) < shards
|
||||
if rows is not None:
|
||||
raw_flows = [r for r in rows if isinstance(r, dict) and "kind" not in r and "sink" in r]
|
||||
extra.update(witness_rows=raw_flows, traces=[parse_trace(r) for r in raw_flows])
|
||||
if shard_incomplete:
|
||||
extra["questions_completed"] = []
|
||||
names = census[0].get("file_names", []) if census else []
|
||||
file_missing = not any(str(p).removeprefix("./") == unit["file"] or str(p).endswith("/" + unit["file"]) for p in names)
|
||||
dropped = file_missing and rows is not None
|
||||
if failure or dropped:
|
||||
status, reason = "failed", failure or "labelled file was dropped or absent from query census"
|
||||
if dropped:
|
||||
logs = [line for stage in backend.stages for line in stage.get("labelled_file_logs", {}).get(unit["file"], [])]
|
||||
reason += "\n" + ("\n".join(dict.fromkeys(logs)) if logs else "frontend emitted no log lines mentioning labelled file")
|
||||
elif rows is None or shard_incomplete:
|
||||
status, reason = ("timeout" if backend.timed_out else "failed"), "question did not complete"
|
||||
elif not rows and rid not in backend.asked:
|
||||
status, reason = "failed", "query produced no markers"
|
||||
else:
|
||||
flows = [r for r in rows if isinstance(r, dict) and "kind" not in r and "sink" in r]
|
||||
if any(not row.get("trace") for row in flows):
|
||||
raise ValueError("taint witness returned without requested trace")
|
||||
extra.update(witness_rows=flows, traces=[parse_trace(r) for r in flows])
|
||||
status, reason = ("path" if flows else "asked-nothing"), ""
|
||||
record["units"][index] = unit_record(unit, status, reason, **extra)
|
||||
except (OSError, ValueError, RuntimeError) as exc:
|
||||
for index, unit in enumerate(pin["units"]):
|
||||
if unit["supported"]:
|
||||
record["units"][index] = unit_record(
|
||||
unit,
|
||||
"timeout" if backend.timed_out and not backend.fatal else "failed",
|
||||
str(exc),
|
||||
instrument=proof,
|
||||
questions={unit["unit"]: record["questions"][unit["unit"]]} if unit["unit"] in record.get("questions", {}) else {},
|
||||
witness_rows=[r for r in backend.answers.get(unit["unit"], []) if isinstance(r, dict) and "sink" in r],
|
||||
questions_completed=[unit["unit"]] if unit["unit"] in backend.answers else [],
|
||||
questions_asked=[unit["unit"]] if unit["unit"] in backend.asked else [],
|
||||
)
|
||||
finally:
|
||||
if cpg and cpg.cleanup:
|
||||
cpg.cleanup()
|
||||
for stage in backend.stages:
|
||||
if stage["command"] == "joern" and stage.get("exit") and "output_tail" in stage:
|
||||
detail = f"joern query exited {stage['exit']}: {stage['output_tail'] or 'no output'}"
|
||||
for row in record["units"]:
|
||||
if row["supported"] and row["status"] in {"failed", "timeout"}:
|
||||
row["reason"] = f"{row['reason']}\n{detail}".strip()
|
||||
record["questions_asked"] = list(dict.fromkeys(backend.asked))
|
||||
record["questions_completed"] = [rid for u in record["units"] for rid in u["questions_completed"]]
|
||||
record["done"] = True
|
||||
persist(out, record)
|
||||
return record
|
||||
|
||||
|
||||
def worker(pin, root, out, deadline, question_deadline):
|
||||
"""Run language partitions sequentially under one shared container deadline."""
|
||||
languages = sorted({"javascript" if u["language"] == "typescript" else u["language"] for u in pin["units"] if u["supported"]})
|
||||
if len(languages) <= 1:
|
||||
return worker_language(pin, root, out, deadline, question_deadline)
|
||||
end = time.monotonic() + deadline
|
||||
record = {
|
||||
**pin,
|
||||
"done": False,
|
||||
"units": [
|
||||
unit_record(u, "timeout" if u["supported"] else "unsupported", "pin deadline" if u["supported"] else u["reason"])
|
||||
for u in pin["units"]
|
||||
],
|
||||
"instruments": {},
|
||||
}
|
||||
|
||||
def persist(_, partial):
|
||||
updates = {u["unit"]: u for u in partial["units"]}
|
||||
record["units"] = [updates.get(u["unit"], u) for u in record["units"]]
|
||||
record["instruments"][language] = partial["instrument"]
|
||||
record["questions_asked"] = [rid for u in record["units"] for rid in u["questions_asked"]]
|
||||
record["questions_completed"] = [rid for u in record["units"] for rid in u["questions_completed"]]
|
||||
write_json(out, record)
|
||||
|
||||
write_json(out, record)
|
||||
for language in languages:
|
||||
remaining = end - time.monotonic()
|
||||
if remaining <= 0:
|
||||
break
|
||||
units = [
|
||||
u for u in pin["units"] if u["supported"] and ("javascript" if u["language"] == "typescript" else u["language"]) == language
|
||||
]
|
||||
questions = {u["unit"]: pin["questions"][u["unit"]] for u in units}
|
||||
worker_language({**pin, "units": units, "questions": questions}, root, out, remaining, question_deadline, persist)
|
||||
record["done"] = True
|
||||
write_json(out, record)
|
||||
return record
|
||||
|
||||
|
||||
def main():
|
||||
parser = argparse.ArgumentParser(description=__doc__)
|
||||
parser.add_argument("--worker", type=Path, required=True)
|
||||
parser.add_argument("--deadline", type=int, default=900)
|
||||
parser.add_argument("--question-deadline", type=int, default=120)
|
||||
args = parser.parse_args()
|
||||
if min(args.deadline, args.question_deadline) <= 0:
|
||||
parser.error("deadlines must be positive")
|
||||
worker(json.loads(args.worker.read_text()), Path("/case"), Path("/out/result.json"), args.deadline, args.question_deadline)
|
||||
return 0
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
raise SystemExit(main())
|
||||
Reference in New Issue
Block a user