CVE-2022-41889: Segfault via invalid attributes in `pywrap_tfe_src.cc` in Tensorflow
Impact If a list of quantized tensors is assigned to an attribute, the pywrap code fails to parse the tensor and returns a nullptr, which is not caught. An example can be seen in tf.compat.v1.extractvolumepatches by passing in quantized tensors as input ksizes. python import numpy as np import tensorflow as tf
ainput = np.array([1, -1], dtype= np.int32) aksizes = astrides = tf.constant(dtype=tf.dtypes.qint16, value=[[1, 4], [5, 2]])
tf.compat.v1.extractvolumepatches(input=ainput,ksizes=aksizes,strides=astrides,padding='VALID')
Patches We have patched the issue in GitHub commit e9e95553e5411834d215e6770c81a83a3d0866ce.
The fix will be included in TensorFlow 2.11. We will also cherrypick this commit on TensorFlow 2.10.1, 2.9.3, and TensorFlow 2.8.4, 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 Pattarakrit Rattankul.
Other sources
TensorFlow is an open source platform for machine learning. If a list of quantized tensors is assigned to an attribute, the pywrap code fails to parse the tensor and returns a nullptr, which is not caught. An example can be seen in tf.compat.v1.extractvolumepatches by passing in quantized tensors as input ksizes. We have patched the issue in GitHub commit e9e95553e5411834d215e6770c81a83a3d0866ce. The fix will be included in TensorFlow 2.11. We will also cherrypick this commit on TensorFlow 2.10.1, 2.9.3, and TensorFlow 2.8.4, as these are also affected and still in supported range.
— NVD
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
pip/tensorflow-gputo a version that resolves this vulnerability.Fixed in 2.10.1 - Upgrade
Upgrade
pip/tensorflow-cputo a version that resolves this vulnerability.Fixed in 2.10.1 - Upgrade
Upgrade
pip/tensorflow-gputo a version that resolves this vulnerability.Fixed in 2.9.3 - Upgrade
Upgrade
pip/tensorflow-cputo a version that resolves this vulnerability.Fixed in 2.9.3 - Upgrade
Upgrade
pip/tensorflow-gputo a version that resolves this vulnerability.Fixed in 2.8.4 - Upgrade
Upgrade
pip/tensorflow-cputo a version that resolves this vulnerability.Fixed in 2.8.4 - Upgrade
Upgrade
pip/tensorflowto a version that resolves this vulnerability.Fixed in 2.10.1 - Upgrade
Upgrade
pip/tensorflowto a version that resolves this vulnerability.Fixed in 2.9.3 - Upgrade
Upgrade
pip/tensorflowto a version that resolves this vulnerability.Fixed in 2.8.4 - Upgrade
Upgrade
Tensorflowto a version that resolves this vulnerability.Fixed in 2.11Patch e9e95553e5411834d215e6770c81a83a3d0866ce - Upgrade
Upgrade
Tensorflowto a version that resolves this vulnerability.Fixed in 2.10.1Patch e9e95553e5411834d215e6770c81a83a3d0866ce - Upgrade
Upgrade
Tensorflowto a version that resolves this vulnerability.Fixed in 2.9.3Patch e9e95553e5411834d215e6770c81a83a3d0866ce - Upgrade
Upgrade
Tensorflowto a version that resolves this vulnerability.Fixed in 2.8.4Patch e9e95553e5411834d215e6770c81a83a3d0866ce - Configuration
Do not pass quantized tensors as the ksizes attribute to tf.compat.v1.extract_volume_patches; the issue occurs when a list of quantized tensors is assigned to ksizes/ksizes-like attributes and causes pywrap parsing to return nullptr.
TensorFlow tf.compat.v1.extract_volume_patches ksizes = Use non-quantized tensors for ksizes (avoid quantized tensors such as dtype=tf.dtypes.qint16)
Event History
Frequently Asked Questions
What is CVE-2022-41889?
CVE-2022-41889 is a vulnerability in TensorFlow, an open source platform for machine learning, where a list of quantized tensors assigned to an attribute can cause the pywrap code to fail, leading to a null pointer exception.
How does CVE-2022-41889 affect Google TensorFlow?
CVE-2022-41889 affects Google TensorFlow versions up to and including 2.8.4, versions 2.9.0 to 2.9.3, and versions 2.10.0 to 2.10.1.
What is the severity of CVE-2022-41889?
CVE-2022-41889 has a severity rating of 7.5 (High).
How can I fix CVE-2022-41889?
To fix CVE-2022-41889, it is recommended to update to a version of Google TensorFlow that is not affected by the vulnerability.
Where can I find more information about CVE-2022-41889?
More information about CVE-2022-41889 can be found in the TensorFlow GitHub repository and the TensorFlow security advisories.