CVE-2022-23588
6.5
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
Summary
Tensorflow is an Open Source Machine Learning Framework. A malicious user can cause a denial of service by altering a SavedModel such that Grappler optimizer would attempt to build a tensor using a reference dtype. This would result in a crash due to a CHECK-fail in the Tensor constructor as reference types are not allowed. The fix will be included in TensorFlow 2.8.0. We will also cherrypick this commit on TensorFlow 2.7.1, TensorFlow 2.6.3, and TensorFlow 2.5.3, as these are also affected and still in supported range.
Affected Software
| Vendor | Product | Version Range | Status |
|---|---|---|---|
| tensorflow | tensorflow | >= 2.7.0, < 2.7.1 | affected |
| tensorflow | tensorflow | >= 2.6.0, < 2.6.3 | affected |
| tensorflow | tensorflow | < 2.5.3 | affected |
Weaknesses
- CWE-617: CWE-617: Reachable Assertion
ADP Enrichment
CVE Program Container
Additional References
- https://github.com/tensorflow/tensorflow/security/advisories/GHSA-fx5c-h9f6-rv7c
- https://github.com/tensorflow/tensorflow/commit/6b5adc0877de832b2a7c189532dbbbc64622eeb6
- https://github.com/tensorflow/tensorflow/blob/a1320ec1eac186da1d03f033109191f715b2b130/tensorflow/core/framework/tensor.cc#L733-L781
- https://github.com/tensorflow/tensorflow/blob/a1320ec1eac186da1d03f033109191f715b2b130/tensorflow/core/grappler/optimizers/constant_folding.cc#L1328-L1402
CISA ADP Vulnrichment
- SSVC:
- Exploitation: poc
- Automatable: no
- Technical Impact: partial
References
- https://github.com/tensorflow/tensorflow/security/advisories/GHSA-fx5c-h9f6-rv7c
- https://github.com/tensorflow/tensorflow/commit/6b5adc0877de832b2a7c189532dbbbc64622eeb6
- https://github.com/tensorflow/tensorflow/blob/a1320ec1eac186da1d03f033109191f715b2b130/tensorflow/core/framework/tensor.cc#L733-L781
- https://github.com/tensorflow/tensorflow/blob/a1320ec1eac186da1d03f033109191f715b2b130/tensorflow/core/grappler/optimizers/constant_folding.cc#L1328-L1402
Feedback
Was this page helpful?
Glad to hear it! Please tell us how we can improve.
Sorry to hear that. Please tell us how we can improve.