CVE-2021-29569: Heap out of bounds read in tensorflow

Background

TensorFlow is an open-source machine learning framework developed by Google. It provides a comprehensive ecosystem of tools, libraries, and resources for building and deploying machine learning models. TensorFlow is widely used in various domains, including research, industry, and academia. However, a vulnerability has been discovered in TensorFlow that could potentially lead to a heap out of bounds read.

Vulnerability Detail

The vulnerability exists in the implementation of tf.raw_ops.MaxPoolGradWithArgmax in TensorFlow. Attackers can exploit this vulnerability by supplying specially crafted inputs. When the input_min and input_max tensors are empty, the implementation assumes they have at least one element and attempts to access the first element. This results in a read outside the bounds of the heap allocated data. The impact of this vulnerability can vary depending on the specific usage scenario, but it has the potential to lead to information disclosure or denial of service.

pip install --upgrade "tensorflow==2.21.0"

Workaround

As a workaround, you can ensure that the input_min and input_max tensors have at least one element before passing them to tf.raw_ops.RequantizationRange. This can be done by checking the size of the tensors and adding a dummy element if they are empty. Here is an example of how to implement this workaround in Python:

import tensorflow as tf

input = tf.constant([1], shape=[1], dtype=tf.qint32)
input_max = tf.constant([], dtype=tf.float32)
input_min = tf.constant([], dtype=tf.float32)

if input_max.shape.num_elements() == 0:
    input_max = tf.constant([0.0], dtype=tf.float32)
if input_min.shape.num_elements() == 0:
    input_min = tf.constant([0.0], dtype=tf.float32)

tf.raw_ops.RequantizationRange(input=input, input_min=input_min, input_max=input_max)

Conclusion

In conclusion, a heap out of bounds read vulnerability has been discovered in TensorFlow. This vulnerability can be exploited by attackers to read outside the bounds of heap allocated data. It is recommended to update to the latest patched version of TensorFlow or apply the provided workaround to mitigate the risk. For further support on vulnerability remediation, you can reach out to DevNack.com.

See every vulnerability recorded for tensorflow →

FAQ