CVE-2021-41211: TensorFlow GPU - Out-of-Bounds Read in QuantizeV2 Shape Inference

Background

TensorFlow is an 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 specific package that enables GPU acceleration for TensorFlow. It allows users to leverage the power of GPUs to accelerate their machine learning workflows.

Vulnerability Detail

The vulnerability (CVE-2021-41211) exists in the shape inference code for the QuantizeV2 function in TensorFlow GPU. When the axis parameter is a negative value less than -1, the code triggers a read outside the bounds of a heap-allocated array. This results in accessing data before the start of the heap buffer, leading to a heap out-of-bounds read. The issue arises due to the code assuming that axis is a valid index into the dimensions of the input tensor, without proper validation. The vulnerability can be exploited by an attacker to read sensitive information from the heap, potentially leading to information disclosure or further exploitation.

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

Workaround

As a temporary workaround, users can avoid passing a negative value less than -1 for the axis parameter in the QuantizeV2 function. By ensuring that axis is a non-negative value or within the valid range of dimensions, the vulnerability can be mitigated. Here's an example of a valid usage:

import tensorflow as tf

input_tensor = tf.constant([1, 2, 3, 4, 5])
axis = 0
quantized_tensor = tf.quantization.quantize_v2(input_tensor, axis)

Conclusion

In conclusion, the vulnerability (CVE-2021-41211) in TensorFlow GPU's shape inference code for QuantizeV2 can lead to a heap out-of-bounds read. Users are advised to update their TensorFlow GPU installations to versions 2.6.1 or 2.7.0 once the fixes are released. A temporary workaround is available by avoiding negative values less than -1 for the axis parameter. For further support on vulnerability remediation, please contact DevNack.com.

See every vulnerability recorded for tensorflow-gpu →

FAQ