CVE-2021-41204
MEDIUMTensorFlow < 2.4.4, 2.6.0-2.6.1 - Access of Uninitialized Pointer in Grappler Constant Folding
Title source: llmDescription
TensorFlow is an open source platform for machine learning. In affected versions during TensorFlow's Grappler optimizer phase, constant folding might attempt to deep copy a resource tensor. This results in a segfault, as these tensors are supposed to not change. 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.
References (2)
Core 2
Core References
Third Party Advisory x_refsource_confirm
https://github.com/tensorflow/tensorflow/security/advisories/GHSA-786j-5qwq-r36x
Third Party Advisory x_refsource_misc
https://github.com/tensorflow/tensorflow/commit/7731e8dfbe4a56773be5dc94d631611211156659
Scores
CVSS v3
5.5
EPSS
0.0014
EPSS Percentile
3.3%
Attack Vector
LOCAL
CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
Details
CWE
CWE-824
Status
published
Products (4)
google/tensorflow
< 2.4.4
pypi/tensorflow
2.6.0 - 2.6.1PyPI
pypi/tensorflow-cpu
2.6.0 - 2.6.1PyPI
pypi/tensorflow-gpu
2.6.0 - 2.6.1PyPI
Published
Nov 05, 2021
Tracked Since
Feb 18, 2026