GHSA-48m4-24cq-37qf: Medium severity pip/onnx vulnerability
Summary
ONNX shape inference performed an out-of-bounds read when processing a malformed RNN, GRU, or LSTM node from opsets 1–6 whose first input had a rank other than three.
RNNShapeInferenceopset1to6() checked whether the input shape was present but accessed dim(0) and dim(1) without first checking its rank. A crafted model could therefore terminate a process that called onnx.shapeinference.infershapes(). The model passed onnx.checker.checkmodel() before reaching the vulnerable path.
Technical details
The vulnerable code read dimensions unconditionally:
cpp if (hasInputShape(ctx, 0)) { const auto& firstinputshape = getInputShape(ctx, 0); seqlength = firstinputshape.dim(0); batchsize = firstinputshape.dim(1); }
For an undersized shape, protobuf's repeated-field access read outside the intended array in release builds. The observed result was SIGSEGV; sanitizer and debug builds terminated on a protobuf bounds check.
The newer opset 7–13 and opset 14+ inference paths already required the first input to have rank three and were not affected.
Impact
An application that performs shape inference on an untrusted model can be terminated, causing denial of service. Exploitation does not require privileges or human interaction once the model is processed. No confidentiality or integrity impact was demonstrated.
Affected versions
- Affected: ONNX through 1.21.0 - Patched: ONNX 1.22.0
The vulnerable schemas are limited to RNN and LSTM opsets 1–6 and GRU opsets 3–6.
Remediation
Upgrade to ONNX 1.22.0 or later. The fix validates that the first input has rank three before accessing its dimensions.
- Fix: onnx/onnx#7879 - Commit: 92e541266
The original malformed model was also re-tested against the published ONNX 1.23.2 wheel and correctly raises InferenceError.
Severity
The vulnerable ONNX component processes the malicious model through a local library API or file boundary and is not itself bound to a network stack. Applications that expose ONNX model processing over a network should calculate an Environmental score with Modified Attack Vector set to Network.
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
pip/onnxto a version that resolves this vulnerability.Fixed in 1.22.0 - Upgrade
Upgrade
ONNXto a version that resolves this vulnerability.Fixed in 1.22.0
Event History
Frequently Asked Questions
Which deployments are realistically exposed?
Applications that call onnx.shape_inference.infer_shapes() on untrusted ONNX models are exposed to denial of service. The issue can terminate the process performing shape inference.
What does an attacker need to provide?
An attacker needs to supply a malformed model containing an RNN, GRU, or LSTM node using opset 1 through 6, with a first input whose rank is not three. The model can pass onnx.checker.check_model() before reaching the vulnerable shape-inference path.
Are newer operator sets affected?
No. The opset 7–13 and opset 14+ shape-inference paths already require the first input to have rank three and are not affected.
How can I identify whether a model may trigger the issue?
Check whether the model contains RNN, GRU, or LSTM nodes assigned to opset 1–6 and whether their first input has a rank other than three. The vulnerable condition is reached when shape inference is run on such a model.