Skip to main content

Out-of-bounds Write

CVE-2021-37650

High
7.8/10

Summary

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-rc2. 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.

  • LOW
  • LOCAL
  • HIGH
  • UNCHANGED
  • NONE
  • LOW
  • HIGH
  • HIGH

CWE-787 - Out-of-Bounds Write

Out-of-bounds write vulnerability is a memory access bug that allows software to write data past the end or before the beginning of the intended buffer. This may result in the corruption of data, a crash, or arbitrary code execution.

References

Advisory Timeline

  • Published