Vulnerability · Published Feb 4, 2022 · Assigned by GitHub_M
CVE-2022-23594Out of bounds read in Tensorflow
High 8.8
In plain English
A high-severity out-of-bounds read in tensorflow from tensorflow. There are no confirmed reports of exploitation so far. The chance it is exploited in the next 30 days is estimated at 0.1% (higher than 4% of all known vulnerabilities). Fixed versions are available - upgrade the affected package.
Official description
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.
Affected packages & fixed versions
| Ecosystem | Package | Fixed in | Advisory |
|---|---|---|---|
| PyPI | tensorflow | 2.7.1 | GHSA-9x52-887g-fhc2 |
| PyPI | tensorflow | 2.7.1 | PYSEC-2026-3157 |
| PyPI | tensorflow-cpu | 2.7.1 | GHSA-9x52-887g-fhc2 |
| PyPI | tensorflow-cpu | 2.7.1 | PYSEC-2026-3309 |
| PyPI | tensorflow-gpu | 2.7.1 | GHSA-9x52-887g-fhc2 |
| PyPI | tensorflow-gpu | 2.7.1 | PYSEC-2026-752 |
Affected products (as reported)
| Vendor | Product | Versions |
|---|---|---|
| tensorflow | tensorflow | >= 2.7.0, < 2.8.0 |
Weakness: CWE-125 · Vector: CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H