A use of uninitialized value vulnerability in Tensorflow
Moderate severity
GitHub Reviewed
Published
Nov 4, 2021
in
tensorflow/tensorflow
•
Updated Feb 1, 2023
Description
Published by the National Vulnerability Database
Nov 5, 2021
Reviewed
Nov 8, 2021
Published to the GitHub Advisory Database
Nov 10, 2021
Last updated
Feb 1, 2023
Impact
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 aDequeue
node, thendequeue_node
is left unitialized.Patches
We have patched the issue in GitHub commit 68867bf01239d9e1048f98cbad185bf4761bedd3.
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.
For more information
Please consult our security guide for more information regarding the security model and how to contact us with issues and questions.
Attribution
This vulnerability has been reported by Qian Feng from Baidu Security Team.
References