CVE-2021-29604: TensorFlow-GPU Hashtable Lookup Division by Zero Vulnerability

Background

TensorFlow is an end-to-end open source platform for machine learning. It provides a comprehensive ecosystem of tools, libraries, and resources for developing and deploying machine learning models. TensorFlow-GPU is a variant of TensorFlow that leverages GPU acceleration for faster training and inference. This vulnerability affects the TFLite implementation of hashtable lookup in TensorFlow-GPU.

Vulnerability Detail

The TFLite implementation of hashtable lookup in TensorFlow-GPU is vulnerable to a division by zero error. The specific code segment responsible for the vulnerability can be found in the file 'hashtable_lookup.cc' (lines 114-115) of the TensorFlow repository. An attacker can exploit this vulnerability by crafting a malicious model that sets the first dimension of the 'values' parameter to 0. When the vulnerable code is executed, a division by zero error occurs, leading to a potential crash or other unintended behavior. The impact of this vulnerability depends on the context in which TensorFlow-GPU is used, but it could potentially allow an attacker to disrupt machine learning processes or cause denial of service.

pip install --upgrade "tensorflow-gpu==2.12.0"

Workaround

As a temporary workaround, users can manually modify the 'hashtable_lookup.cc' file in their TensorFlow-GPU installation to include a check for a zero value in the first dimension of the 'values' parameter before performing the division operation. Here is an example of the modified code:

if (values.dim_size(0) == 0) {
  // Handle the zero dimension case
  // ... (add appropriate handling code here)
} else {
  // Perform the division operation
  // ... (original code)
}

Conclusion

The division by zero vulnerability in the TFLite hashtable lookup implementation of TensorFlow-GPU can be exploited by an attacker to disrupt machine learning processes or cause denial of service. Users are advised to update to the latest patched version of TensorFlow-GPU as soon as it becomes available and follow best practices for secure model deployment. For further support on vulnerability remediation, please contact DevNack.com.

See every vulnerability recorded for tensorflow-gpu →

FAQ