PT-2020-14261 · Google+1 · Tensorflow+1

Published

2020-09-25

·

Updated

2024-03-06

·

CVE-2020-15190

CVSS v3.1

5.3

Medium

VectorAV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L
Name of the Vulnerable Software and Affected Versions TensorFlow versions prior to 1.15.4 TensorFlow versions prior to 2.0.3 TensorFlow versions prior to 2.1.2 TensorFlow versions prior to 2.2.1 TensorFlow versions prior to 2.3.1
Description The issue arises from the tf.raw ops.Switch operation, which takes a tensor and a boolean as input and outputs two tensors. Depending on the boolean value, one tensor is the input tensor, and the other should be an empty tensor. However, the eager runtime traverses all output tensors, resulting in undefined behavior when binding a reference to nullptr since only one tensor is defined. This leads to a segmentation fault. The estimated number of potentially affected devices is not specified.
Recommendations To resolve the issue, upgrade to TensorFlow version 1.15.4 or later. To resolve the issue, upgrade to TensorFlow version 2.0.3 or later. To resolve the issue, upgrade to TensorFlow version 2.1.2 or later. To resolve the issue, upgrade to TensorFlow version 2.2.1 or later. To resolve the issue, upgrade to TensorFlow version 2.3.1 or later.

Exploit

Fix

NULL Pointer Dereference

RCE

Weakness Enumeration

Related Identifiers

BIT-TENSORFLOW-2020-15190
CVE-2020-15190
GHSA-4G9F-63RX-5CW4
OPENSUSE-SU-2020:1766-1
OPENSUSE-SU-2020_1766-1
OPENSUSE-SU-2024:12116-1
PYSEC-2020-113
PYSEC-2020-270
PYSEC-2020-305

Affected Products

Suse
Tensorflow