tfsa-2023-007.md

TFSA-2023-007: Segfault in Bincount with XLA

CVE Number

CVE-2023-25675

Impact

When running with XLA, tf.raw_ops.Bincount segfaults when given a parameter weights that is neither the same shape as parameter arr nor a length-0 tensor.

import tensorflow as tf

func = tf.raw_ops.Bincount
para={'arr': 6, 'size': 804, 'weights': [52, 351]}

@tf.function(jit_compile=True)
def fuzz_jit():
 y = func(**para)
 return y

print(fuzz_jit())

Patches

We have patched the issue in GitHub commit 8ae76cf085f4be26295d2ecf2081e759e04b8acf.

The fix will be included in TensorFlow 2.12. We will also cherrypick this commit on TensorFlow 2.11.1.

For more information

Please consult our security guide for more information regarding the security model and how to contact us with issues and questions.

Attribution

This vulnerability has been reported by r3pwnx of 360 AIVul Team