CVE-2022-36001: `CHECK` fail in `DrawBoundingBoxes` in TensorFlow
TensorFlow is an open source platform for machine learning. When DrawBoundingBoxes receives an input boxes that is not of dtype float, it gives a CHECK fail that can trigger a denial of service attack. We have patched the issue in GitHub commit da0d65cdc1270038e72157ba35bf74b85d9bda11. The fix will be included in TensorFlow 2.10.0. We will also cherrypick this commit on TensorFlow 2.9.1, TensorFlow 2.8.1, and TensorFlow 2.7.2, as these are also affected and still in supported range. There are no known workarounds for this issue.
Affected Software
Remediation
Event History
Frequently Asked Questions
What is the severity of CVE-2022-36001?
The severity of CVE-2022-36001 can lead to a denial of service attack.
How do I fix CVE-2022-36001?
To fix CVE-2022-36001, upgrade your TensorFlow to a patched version beyond 2.10-rc3.
What versions of TensorFlow are affected by CVE-2022-36001?
CVE-2022-36001 affects TensorFlow versions prior to 2.8.0 and between 2.8.0 and 2.9.0, as well as 2.10-rc0 to 2.10-rc3.
What happens if I don't address CVE-2022-36001?
If CVE-2022-36001 is not addressed, an attacker may exploit the vulnerability to create a denial of service situation.
Is CVE-2022-36001 specific to certain TensorFlow features?
Yes, CVE-2022-36001 specifically affects the DrawBoundingBoxes function when it receives incorrect input types.