GitHub Advisoryadvisory · PyPI
GHSA-9px9-73fg-3fqp Null pointer dereference in Grappler's `IsConstant`
MEDIUM
Official page ↗
- Aliases
- CVE-2022-23589
- Ecosystem
- PyPI
- Packages
- tensorflow, tensorflow-cpu, tensorflow-gpu
- Severity
- Moderate
- CVSS vector
- CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
- Weaknesses
- CWE-476
- Reviewed
- 2022-02-04
Affected packages
• PyPI tensorflow: < 2.5.3 → fixed in 2.5.3
• PyPI tensorflow: >= 2.6.0, < 2.6.3 → fixed in 2.6.3
• PyPI tensorflow: >= 2.7.0, < 2.7.1 → fixed in 2.7.1
• PyPI tensorflow-cpu: < 2.5.3 → fixed in 2.5.3
• PyPI tensorflow-cpu: >= 2.6.0, < 2.6.3 → fixed in 2.6.3
• PyPI tensorflow-cpu: >= 2.7.0, < 2.7.1 → fixed in 2.7.1
• PyPI tensorflow-gpu: < 2.5.3 → fixed in 2.5.3
• PyPI tensorflow-gpu: >= 2.6.0, < 2.6.3 → fixed in 2.6.3
• PyPI tensorflow-gpu: >= 2.7.0, < 2.7.1 → fixed in 2.7.1
Description
Impact
Under certain scenarios, Grappler component of TensorFlow can trigger a null pointer dereference. There are 2 places where this can occur, for the same malicious alteration of a `SavedModel` file (fixing the first one would trigger the same dereference in the second place):
First, during constant folding, the `GraphDef` might not have the required nodes for the binary operation:
```cc
NodeDef* mul_left_child = node_map_->GetNode(node->input(0));
NodeDef* mul_right_child = node_map_->GetNode(node->input(1));
// One child must be constant, and the second must be Conv op.
const bool left_child_is_constant = IsReallyConstant(*mul_left_child);
const bool right_child_is_constant = IsReallyConstant(*mul_right_child);
```
If a node is missing, the correposning `mul_*child` would be null, and the dereference in the subsequent line would be incorrect.
We have a similar issue during `IsIdentityConsumingSwitch`:
```cc
NodeDef* input_node = graph.GetNode(tensor_id.node());
return IsSwitch(*input_node);
```
Patches
We have patched the issue in GitHub commits 0a365c029e437be0349c31f8d4c9926b69fa3fa1 and 045deec1cbdebb27d817008ad5df94d96a08b1bf.
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.
For more information
Please consult our security guide for more information regarding the security model and how to contact us with issues and questions.
Related
Alias · CVE (1)
- CVE-2022-23589 Tensorflow is an Open Source Machine Learning Framework. vulnerability
Weakness · CWE (1)
- CWE-476 NULL Pointer Dereference weakness
References
- https://github.com/tensorflow/tensorflow/security/advisories/GHSA-9px9-73fg-3fqp web
- https://nvd.nist.gov/vuln/detail/CVE-2022-23589 advisory
- https://github.com/tensorflow/tensorflow/commit/045deec1cbdebb27d817008ad5df94d96a08b1bf web
- https://github.com/tensorflow/tensorflow/commit/0a365c029e437be0349c31f8d4c9926b69fa3fa1 web
- https://github.com/pypa/advisory-database/tree/main/vulns/tensorflow-cpu/PYSEC-2022-98.yaml web
- https://github.com/pypa/advisory-database/tree/main/vulns/tensorflow-gpu/PYSEC-2022-153.yaml web
- https://github.com/tensorflow/tensorflow web
- https://github.com/tensorflow/tensorflow/blob/a1320ec1eac186da1d03f033109191f715b2b130/tensorflow/core/grappler/mutable_graph_view.cc#L59-L74 web
- https://github.com/tensorflow/tensorflow/blob/a1320ec1eac186da1d03f033109191f715b2b130/tensorflow/core/grappler/optimizers/constant_folding.cc#L3466-L3497 web