onnxruntime is vulnerable to Null Pointer Dereference
42
Medium Risk
Affected versions of this package are vulnerable to a null pointer dereference when processing malformed ONNX models containing an empty clip_min tensor, causing the initializer to construct an empty tensor with an invalid data pointer that is dereferenced without validation, resulting in a crash and potential Denial of Service. An attacker could exploit this by supplying a specially crafted ONNX model with a zero-element clip_min input, triggering the failure during graph optimization and forcing the runtime session to terminate. The issue arises from missing size validation before accessing tensor data, allowing invalid inputs to disrupt execution. The mitigation introduces a size check to skip optimization when the tensor is empty, preventing the crash and improving robustness against malformed models.
You are affected if you are using a version that falls within the vulnerable range.
onnxruntime is vulnerable to Null Pointer Dereference in versions 1.23.0 - 1.24.0.
Upgrade the com.microsoft.onnxruntime:onnxruntime library to the patch version.
Secure your code, cloud, and runtime environments in one central system. Find and fix vulnerabilities automatically.
No credit card required | Scan results in 32secs.
SOC 2Compliant
ISO 27001Compliant