NewApp Monitoring is liveSign up free
AgentbeamBeta
Open-source package vulnerability · PyPI

CVE-2026-112936: filelock: ReadWriteLock follows a symlink at the lock path into SQLite

mediumCVSS 3.1 6.1GHSA-j8f7-rjxc-mr56

The Python package filelock, versions 3.20.4 through 4.0.8, follows a symlink at the lock path in ReadWriteLock and AsyncReadWriteLock. A local user who can create files in the lock's directory can make a victim process create files it owns elsewhere, lock another application's SQLite database, or break mutual exclusion. Upgrade to 4.0.9.

Published Oct 28, 2026 · Analysis Nov 1, 2026

What an attacker can do

ReadWriteLock.__init__ stores pathlib.Path(lock_file).resolve(), which follows a final symlink. Every connection then opens that path through sqlite3.connect() with no O_NOFOLLOW and no regular-file check. SQLite's VFS resolves the symlink in user space, so the kernel never sees a link to refuse. That means fs.protected_symlinks and the sticky bit on /tmp do not protect you. FileLock (opens with O_NOFOLLOW) and SoftReadWriteLock are not affected.

The attacker needs local access and write access to the lock directory, such as /tmp or a shared cache. In the proof of concept, a symlink from /shared/app.rwlock.db to a path in the victim's private directory makes the victim's process create an empty file there, in a directory the attacker cannot write. Repointing a link between two victim processes lets both hold the write lock at once, which can corrupt whatever the lock protects. A link to another application's database makes that application report database is locked. The flaw reproduces on Linux, macOS and Windows 11.

Who is affected

filelock from 3.20.4 up to and including 4.0.8, only when code uses ReadWriteLock or AsyncReadWriteLock. Code that uses only FileLock or SoftReadWriteLock is not affected. Risk applies where the lock file sits in a directory other users can write to. Check your installed version:

pip show filelock

Also look for filelock pinned in lock files and in tools that vendor it.

filelock

Ecosystem
PyPI
Affected
>= 3.20.4, <= 4.0.8
Fixed in
4.0.9
Check
pip show filelock

How to fix it

Upgrade to filelock 4.0.9, which rejects a symlink at the database path instead of following it.

pip install --upgrade "filelock>=4.0.9"

The advisory's workaround: keep ReadWriteLock paths in directories that only trusted users can write.

How to tell if you were hit

Search lock directories for symlinks at the paths your code passes to ReadWriteLock:

find /tmp /shared -maxdepth 2 -type l -name "*.db"

Adjust the directories to your lock locations. Look for unexpected empty files owned by the service account in private directories, for other applications reporting database is locked, and for resource corruption that suggests two processes held the write lock at once. Grep your code for ReadWriteLock and AsyncReadWriteLock to see exposure.

Original advisory

From the GitHub Advisory Database (CC BY 4.0).

Summary

ReadWriteLock and AsyncReadWriteLock resolve the lock path with pathlib.Path.resolve() and pass it to sqlite3.connect() without refusing a symlink. A local user who can create names in the lock's directory, such as /tmp or a shared cache, can plant a symlink at the lock path and make the victim's process create, lock, or write the file it points to, or break mutual exclusion between the victim's own processes.

Details

ReadWriteLock.__init__ stores pathlib.Path(lock_file).resolve(), which follows a final symlink, and every connection opens that path through sqlite3.connect() with no O_NOFOLLOW or regular-file check. SQLite's own VFS also resolves the final symlink in user space, so fs.protected_symlinks and the sticky bit do not help: the kernel never sees a symlink to refuse. The other lock classes do not follow a final symlink: FileLock opens with O_NOFOLLOW, and SoftReadWriteLock coordinates through <path>.rw/.

Proof of concept

On Linux, with the victim and attacker as different users and /shared mode 1777 with fs.protected_symlinks=1:

  1. The attacker runs ln -s /home/victim/private/pwned.flag /shared/app.rwlock.db.
  2. The victim constructs ReadWriteLock("/shared/app.rwlock.db"). A plain os.open(..., O_NOFOLLOW) on the same path fails, but the lock creates /home/victim/private/pwned.flag, owned by the victim, inside a directory the attacker cannot write.
  3. With /shared/cache.rwlock.db -> /shared/a.db, one victim process holds the write lock. The attacker repoints the link to /shared/b.db, and a second victim process acquires the write lock while the first still holds it.

The same behavior reproduces on macOS, on Windows 11, and through AsyncReadWriteLock.

Impact

A local attacker with write access to the lock directory can make the victim create empty files it owns at paths the attacker cannot write, make the victim hold a SQLite lock on another application's database so that application gets database is locked, and give two victim processes the write lock at once, letting them corrupt the resource the lock protects.

Workarounds

Keep ReadWriteLock paths in directories that only trusted users can write.

Credited to Kwstubbs.

Questions

Is FileLock affected by this filelock advisory?

No. The advisory says FileLock opens with O_NOFOLLOW and SoftReadWriteLock coordinates through a .rw directory. Only ReadWriteLock and AsyncReadWriteLock follow the symlink.

Does fs.protected_symlinks stop this?

No. SQLite resolves the final symlink in user space, so the kernel never sees a link to refuse. The proof of concept works with protected_symlinks set to 1.

Can this be exploited remotely?

No. The CVSS vector is local, and the attacker needs to create names in the lock directory, such as a shared /tmp. The sources do not report exploitation in the wild.

References

How this page is made: the data panel comes from the GitHub Advisory Database, CISA’s KEV catalog, the NVD and FIRST EPSS, and updates twice a day. The analysis is written by an AI model from the sources listed above and checked by automated rules before it is published. Check versions against the vendor's advisory before you act. Methodology

More PyPI vulnerabilities

All PyPI entries