Description
TensorFlow is an open source platform for machine learning. The implementation of `FractionalAvgPoolGrad` does not fully validate the input `orig_input_tensor_shape`. This results in an overflow that results in a `CHECK` failure which can be used to trigger a denial of service attack. We have patched the issue in GitHub commit 03a659d7be9a1154fdf5eeac221e5950fec07dad. The fix will be included in TensorFlow 2.10.0. We will also cherrypick this commit on TensorFlow 2.9.1, TensorFlow 2.8.1, and TensorFlow 2.7.2, as these are also affected and still in supported range. There are no known workarounds for this issue.
Affected products
| Vendor | Product | Affected | Fixed |
|---|---|---|---|
| tensorflow | tensorflow | < 2.7.2, >= 2.8.0, < 2.8.1, >= 2.9.0, < 2.9.1 | — |
Affected packages
| Ecosystem | Package | Affected | Fixed | Record |
|---|---|---|---|---|
| PyPI | tensorflow | < 2.7.2 | 2.7.2 | GHSA-84jm-4cf3-9jfm |
| PyPI | tensorflow | >= 2.8.0, < 2.8.1 | 2.8.1 | GHSA-84jm-4cf3-9jfm |
| PyPI | tensorflow | >= 2.9.0, < 2.9.1 | 2.9.1 | GHSA-84jm-4cf3-9jfm |
| PyPI | tensorflow | < 2.7.2 · >= 2.8.0, < 2.8.1 · >= 2.9.0, < 2.9.1 | 2.7.2, 2.8.1, 2.9.1 | PYSEC-2026-3133 |
| PyPI | tensorflow-cpu | < 2.7.2 | 2.7.2 | GHSA-84jm-4cf3-9jfm |
| PyPI | tensorflow-cpu | >= 2.8.0, < 2.8.1 | 2.8.1 | GHSA-84jm-4cf3-9jfm |
| PyPI | tensorflow-cpu | >= 2.9.0, < 2.9.1 | 2.9.1 | GHSA-84jm-4cf3-9jfm |
| PyPI | tensorflow-cpu | < 2.7.2 · >= 2.8.0, < 2.8.1 · >= 2.9.0, < 2.9.1 | 2.7.2, 2.8.1, 2.9.1 | PYSEC-2026-3296 |
| PyPI | tensorflow-gpu | < 2.7.2 | 2.7.2 | GHSA-84jm-4cf3-9jfm |
| PyPI | tensorflow-gpu | >= 2.8.0, < 2.8.1 | 2.8.1 | GHSA-84jm-4cf3-9jfm |
| PyPI | tensorflow-gpu | >= 2.9.0, < 2.9.1 | 2.9.1 | GHSA-84jm-4cf3-9jfm |
| PyPI | tensorflow-gpu | < 2.7.2 · >= 2.8.0, < 2.8.1 · >= 2.9.0, < 2.9.1 | 2.7.2, 2.8.1, 2.9.1 | PYSEC-2026-969 |