-
CVSS Score
-Basic Information
Technical Details
Vulnerability Intelligence
Miggo AI
Miggo AIMiggo AIMiggo AIRoot Cause AnalysisThe vulnerability stems from unsafe dimension handling in ResizeNearestNeighborGrad's output allocation. The patch shows the critical change from direct TensorShape construction to using BuildTensorShape, which validates dimension sizes. The vulnerable code path (pre-patch) didn't check for integer overflow when multiplying dimensions from user-controlled 'size' input. The test case added in image_ops_test.py specifically targets this overflow scenario, confirming the function's vulnerability to large dimension values.
| Package Name | Ecosystem | Vulnerable Versions | First Patched Version |
|---|---|---|---|
| tensorflow | pip | < 2.8.4 | 2.8.4 |
| tensorflow | pip | >= 2.9.0, < 2.9.3 | 2.9.3 |
| tensorflow | pip | >= 2.10.0, < 2.10.1 | 2.10.1 |
| tensorflow-cpu | pip | < 2.8.4 | 2.8.4 |
| tensorflow-gpu | pip | < 2.8.4 | 2.8.4 |
| tensorflow-cpu | pip | >= 2.9.0, < 2.9.3 | 2.9.3 |
| tensorflow-gpu | pip | >= 2.9.0, < 2.9.3 | 2.9.3 |
| tensorflow-cpu | pip | >= 2.10.0, < 2.10.1 | 2.10.1 |
| tensorflow-gpu | pip | >= 2.10.0, < 2.10.1 | 2.10.1 |