fix: read the umask under one lock

plain_file_mode and safe_extract_zip set the umask to 0 to read it. Two threads doing so could restore 0 last, leaving every later file world-writable. Both now read it through process_umask, which holds a lock.
This commit is contained in:
Abdessamad Derraz committed 2026-11-05 19:10:57 +01:00
1 parent a1d5e7c13d
commit 80f736c167
3 files changed
+53 -8

No files matched your search

+19 -4
View File
@@ -12,6 +12,7 @@ import contextlib
import json import json
import os import os
import tempfile import tempfile
import threading
import re import re
@@ -62,12 +63,26 @@ def write_if_changed(
return True return True
_UMASK_LOCK = threading.Lock()
def process_umask() -> int:
"""The process umask, read without letting a second thread see it at 0.
Reading the umask sets it, so it is put back at once. Two threads doing
that unguarded interleave as A sets 0 and reads 022, B sets 0 and reads 0,
A restores 022, B restores 0: the process keeps umask 0 and every file it
creates afterwards is world-writable.
"""
with _UMASK_LOCK:
umask = os.umask(0)
os.umask(umask)
return umask
def plain_file_mode() -> int: def plain_file_mode() -> int:
"""The mode open() gives a new file: 0o666 less the process umask.""" """The mode open() gives a new file: 0o666 less the process umask."""
# Reading the umask sets it, so it is put back at once. return 0o666 & ~process_umask()
umask = os.umask(0)
os.umask(umask)
return 0o666 & ~umask
def set_plain_mode(path: str | os.PathLike) -> None: def set_plain_mode(path: str | os.PathLike) -> None:
+2 -4
View File
@@ -14,7 +14,7 @@ import sys
import tempfile import tempfile
import zipfile import zipfile
from artifacts import set_plain_mode from artifacts import process_umask, set_plain_mode
from hashing import md5sum from hashing import md5sum
@@ -251,9 +251,7 @@ def safe_extract_zip(
""" """
dest = os.path.realpath(dest_dir) dest = os.path.realpath(dest_dir)
os.makedirs(dest, exist_ok=True) os.makedirs(dest, exist_ok=True)
# Reading the umask sets it, so it is put back at once. umask = process_umask()
umask = os.umask(0)
os.umask(umask)
with zipfile.ZipFile(zip_path, "r") as zf: with zipfile.ZipFile(zip_path, "r") as zf:
members = zf.infolist() members = zf.infolist()
if len(members) > max_members: if len(members) > max_members:
+32
View File
@@ -154,6 +154,38 @@ class ScratchFilesGetThePlainMode(unittest.TestCase):
artifacts.write_text_atomic(str(self.dir / "a.json"), "{}") artifacts.write_text_atomic(str(self.dir / "a.json"), "{}")
self.assertEqual(self._mode(self.dir / "a.json"), 0o664) self.assertEqual(self._mode(self.dir / "a.json"), 0o664)
def test_threads_reading_the_mode_leave_the_umask_alone(self):
"""mame_devices fetches on eight threads and every cache write reads
the umask: unguarded, two readers interleaved and left it at 0."""
import threading # noqa: PLC0415
interval = sys.getswitchinterval()
sys.setswitchinterval(1e-6)
self.addCleanup(sys.setswitchinterval, interval)
modes: set[int] = set()
def read() -> None:
seen = {artifacts.plain_file_mode() for _ in range(20000)}
modes.update(seen)
threads = [threading.Thread(target=read) for _ in range(8)]
for thread in threads:
thread.start()
for thread in threads:
thread.join()
current = os.umask(0o022)
self.assertEqual(current, 0o022)
self.assertEqual(modes, {0o644})
def test_no_script_reads_the_umask_unguarded(self):
offenders = [
f"{path.name}:{number}"
for path in sorted((REPO_ROOT / "scripts").rglob("*.py"))
for number, line in enumerate(path.read_text(encoding="utf-8").splitlines(), 1)
if "os.umask(" in line and path.name != "artifacts.py"
]
self.assertEqual(offenders, [])
def test_packs_and_their_manifest(self): def test_packs_and_their_manifest(self):
import zipfile # noqa: PLC0415 import zipfile # noqa: PLC0415