Segfault and heap buffer overflow in `{Experimental,}DatasetToTFRecord` in TensorFlow
tensorflow
Affected versions
| Product | Affected | Fixed in |
|---|---|---|
| tensorflow PyPI |
|
|
| tensorflow-cpu PyPI |
tensorflow
| Product | Affected | Fixed in |
|---|---|---|
| tensorflow PyPI |
|
|
| tensorflow-cpu PyPI |
|
|
| tensorflow-gpu PyPI |
|
|
TensorFlow is an end-to-end open source platform for machine learning. In affected versions the implementation for `tf.raw_ops.ExperimentalDatasetToTFRecord` and `tf.raw_ops.DatasetToTFRecord` can trigger heap buffer overflow and segmentation fault. The [implementation](https://github.com/tensorflow/tensorflow/blob/f24faa153ad31a4b51578f8181d3aaab77a1ddeb/tensorflow/core/kernels/data/experimental/to_tf_record_op.cc#L93-L102) assumes that all records in the dataset are of string type. However, there is no check for that, and the example given above uses numeric types. We have patched the issue in GitHub commit e0b6e58c328059829c3eb968136f17aa72b6c876. The fix will be included in TensorFlow 2.6.0. We will also cherrypick this commit on TensorFlow 2.5.1, TensorFlow 2.4.3, and TensorFlow 2.3.4, as these are also affected and still in supported range.
From the CVE record · GitHub_M
The TensorFlow operations `tf.raw_ops.ExperimentalDatasetToTFRecord` and `tf.raw_ops.DatasetToTFRecord` assume every dataset record is a string but do not check its type. A non-string record can trigger a heap buffer overflow and segmentation fault; the input does not specify an attacker's access requirements.
Anyone using affected PyPI versions of tensorflow, tensorflow-cpu, or tensorflow-gpu before the fixed 2.3.4, 2.4.3, and 2.5.1 releases should check their installations.