CVE-2022-23594
CVE-2022-23594 · Out of bounds read in Tensorflow
This page contains facts from public vulnerability sources. Check the original advisory before updating.
Tensorflow is an Open Source Machine Learning Framework. The TFG dialect of TensorFlow (MLIR) makes several assumptions about the incoming `GraphDef` before converting it to the MLIR-based dialect. If an attacker changes the `SavedModel` format on disk to invalidate these assumptions and the `GraphDef` is then converted to MLIR-based IR then they can cause a crash in the Python interpreter. Under certain scenarios, heap OOB read/writes are possible. These issues have been discovered via fuzzing and it is possible that more weaknesses exist. We will patch them as they are discovered.
CVSS 8.8 · High · EPSS 0.1% · Fix available
- Published
- Updated
Stack: python
Affected products and versions
tensorflow
- Versions
- >= 2.7.0, < 2.8.0: Affected
No fix published
tensorflow PyPI
- Versions
- 2.7.0
- Affected ranges
- ECOSYSTEM: ≥ 2.7.0, < 2.7.1
Update to 2.7.1
tensorflow-cpu PyPI
- Versions
- 2.7.0
- Affected ranges
- ECOSYSTEM: ≥ 2.7.0, < 2.7.1
Update to 2.7.1
tensorflow-gpu PyPI
- Versions
- 2.7.0
- Affected ranges
- ECOSYSTEM: ≥ 2.7.0, < 2.7.1
Update to 2.7.1