SB2022051757 - Multiple vulnerabilities in TensorFlow

Description

This security bulletin contains information about 2 vulnerabilities.

1) Integer overflow (CVE-ID: CVE-2021-41197)

The vulnerability allows a remote attacker to cause a denial of service.

The vulnerability exists due to an integer overflow in depthwise ops when processing crafted tensor inputs. A remote attacker can supply crafted filter_sizes values to trigger an assertion failure and cause a denial of service.

2) Input validation error (CVE-ID: CVE-2021-41206)

The vulnerability allows a remote user to cause a denial of service.

The vulnerability exists due to improper input validation in tf.sparse.split when processing a non-scalar axis argument. A remote user can supply a tuple or list for the axis parameter to cause a denial of service.

The crash may manifest as a segfault or a heap out-of-bounds read.

Remediation

Install update from vendor's website.

References

• https://github.com/tensorflow/tensorflow/security/advisories/GHSA-mw6j-hh29-h379
• https://github.com/tensorflow/tensorflow/commit/3796cc4fcd93ae55812a457abc96dcd55fbb854b
• https://github.com/tensorflow/tensorflow/security/advisories/GHSA-43q8-3fv7-pr5x
• https://github.com/tensorflow/tensorflow/commit/61bf91e768173b001d56923600b40d9a95a04ad5