CVE-2022-41894: Buffer overflow in `CONV_3D_TRANSPOSE` on TFLite
First published: Fri Nov 18 2022(Updated: )
TensorFlow is an open source platform for machine learning. The reference kernel of the `CONV_3D_TRANSPOSE` TensorFlow Lite operator wrongly increments the data_ptr when adding the bias to the result. Instead of `data_ptr += num_channels;` it should be `data_ptr += output_num_channels;` as if the number of input channels is different than the number of output channels, the wrong result will be returned and a buffer overflow will occur if num_channels > output_num_channels. An attacker can craft a model with a specific number of input channels. It is then possible to write specific values through the bias of the layer outside the bounds of the buffer. This attack only works if the reference kernel resolver is used in the interpreter. We have patched the issue in GitHub commit 72c0bdcb25305b0b36842d746cc61d72658d2941. 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.
Credit: security-advisories@github.com
| Affected Software | Affected Version | How to fix |
|---|---|---|
| Google TensorFlow | <2.8.4 | |
| Google TensorFlow | >=2.9.0<2.9.3 | |
| Google TensorFlow | >=2.10.0<2.10.1 |
Never miss a vulnerability like this again
Sign up to SecAlerts for real-time vulnerability data matched to your software, aggregated from hundreds of sources.
Reference Links
- https://github.com/tensorflow/tensorflow/blob/091e63f0ea33def7ecad661a5ac01dcafbafa90b/tensorflow/lite/kernels/internal/reference/conv3d_transpose.h#L121
- https://github.com/tensorflow/tensorflow/commit/72c0bdcb25305b0b36842d746cc61d72658d2941
- https://github.com/tensorflow/tensorflow/security/advisories/GHSA-h6q3-vv32-2cq5
Frequently Asked Questions
What is CVE-2022-41894?
CVE-2022-41894 is a vulnerability in TensorFlow where the reference kernel of the CONV_3D_TRANSPOSE TensorFlow Lite operator incorrectly increments the data_ptr when adding the bias to the result.
What is the severity of CVE-2022-41894?
The severity of CVE-2022-41894 is high, with a score of 8.1.
Which versions of TensorFlow are affected by CVE-2022-41894?
TensorFlow versions up to 2.8.4, 2.9.0 up to 2.9.3, and 2.10.0 up to 2.10.1 are affected by CVE-2022-41894.
How can I fix CVE-2022-41894?
To fix CVE-2022-41894, update your TensorFlow installation to a version higher than 2.10.1.
Where can I find more information about CVE-2022-41894?
You can find more information about CVE-2022-41894 on the TensorFlow GitHub page and the associated security advisories.
- collector/nvd-index
- agent/type
- agent/remedy
- agent/softwarecombine
- collector/mitre-cve
- source/MITRE
- agent/author
- agent/severity
- agent/title
- agent/weakness
- agent/references
- agent/last-modified-date
- agent/first-publish-date
- agent/tags
- agent/description
- agent/event
- vendor/google
- canonical/google tensorflow
- version/google tensorflow/2.9.0
- version/google tensorflow/2.10.0