PT-2022-16105 · Google · Tensorflow
Mihaimaruseac
·
Published
2022-02-04
·
Updated
2024-03-06
·
CVE-2022-23588
CVSS v3.1
6.5
Medium
| Vector | AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H |
Name of the Vulnerable Software and Affected Versions
TensorFlow versions prior to 2.8.0
TensorFlow versions 2.7.0 through 2.7.0 (will be fixed in 2.7.1)
TensorFlow versions 2.6.0 through 2.6.2 (will be fixed in 2.6.3)
TensorFlow versions 2.5.0 through 2.5.2 (will be fixed in 2.5.3)
Description
A malicious user can cause a denial of service by altering a
SavedModel such that the Grappler optimizer would attempt to build a tensor using a reference dtype. This would result in a crash due to a CHECK-fail in the Tensor constructor as reference types are not allowed.Recommendations
For TensorFlow versions prior to 2.5.3, update to version 2.5.3 or later.
For TensorFlow versions 2.5.3 to 2.6.2, update to version 2.6.3 or later.
For TensorFlow versions 2.6.3 to 2.7.0, update to version 2.7.1 or later.
For TensorFlow versions 2.7.1 and earlier, update to version 2.8.0 or later.
As a temporary workaround, consider restricting the use of the
SavedModel feature until a patch is available.Exploit
Fix
Assertion Failure
Found an issue in the description? Have something to add? Feel free to write us 👾
Weakness Enumeration
Related Identifiers
Affected Products
Tensorflow