Published: 05/11/2021 Updated: 10/11/2021

CVSS v2 Base Score: 2.1 | Impact Score: 2.9 | Exploitability Score: 3.9

CVSS v3 Base Score: 7.8 | Impact Score: 5.9 | Exploitability Score: 1.8

VMScore: 187

Vector: AV:L/AC:L/Au:N/C:N/I:N/A:P

TensorFlow is an open source platform for machine learning. In affected versions TensorFlow's Grappler optimizer has a use of unitialized variable. If the `train_nodes` vector (obtained from the saved model that gets optimized) does not contain a `Dequeue` node, then `dequeue_node` is left unitialized. 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.

Vulnerable Product	Search on Vulmon	Subscribe to Product
google tensorflow
google tensorflow 2.7.0

In TensorFlow before version 261, TensorFlow's Grappler optimizer has a use of unitialized variable If the train_nodes vector (obtained from the saved model that gets optimized) does not contain a Dequeue node, then dequeue_node is left unitialized ...

CWE-908 https://github.com/tensorflow/tensorflow/commit/68867bf01239d9e1048f98cbad185bf4761bedd3 https://github.com/tensorflow/tensorflow/security/advisories/GHSA-7r94-xv9v-63jw https://nvd.nist.gov https://security.archlinux.org/CVE-2021-41225

Get Started

Vulmon Search is a vulnerability search engine. It gives comprehensive vulnerability information through a very simple user interface.

Home Recent Vulnerabilities Research Posts Trends Blog About Contact

Vulmon Search Vulmon Research Vulmon Alerts Vulmap

Twitter Reddit Linkedin Facebook

CVE-2021-41225

Vulnerability Summary

Vulnerability Trend

Vendor Advisories

References

Vulmon Search

About

Products

Connect