CVE-2026-55093
Description
Tract is a tiny, no-nonsense, self-contained TensorFlow and ONNX inference toolkit. Prior to 0.21.16, 0.22.2, and 0.23.1, tract-nnef uses unchecked usize multiplication in nnef/src/tensors.rs read_tensor for attacker-controlled tensor dimensions, the allocation size, and the reported tensor length. Loading a crafted NNEF archive through model_for_path or model_for_read reaches the default DatLoader and can make the wrapped size check accept a small allocation while data/src/tensor.rs as_slice_unchecked creates a much larger logical slice. Model construction through as_uniform can then read beyond the heap allocation and disclose adjacent data, and later access can terminate the process with a segmentation fault. The affected dense numeric tensor path does not include the independently guarded bool, String, or block-quant paths, and no out-of-bounds write or code execution was demonstrated. This issue is fixed in versions 0.21.16, 0.22.2, and 0.23.1.
AI Insight
LLM-synthesized narrative grounded in this CVE's description and references.
Affected packages
Versions sourced from the GitHub Security Advisory.
| Package | Affected versions | Patched versions |
|---|---|---|
tract-nnefcrates.io | >= 0.23.0, < 0.23.1 | 0.23.1 |
tract-nnefcrates.io | >= 0.22.0, < 0.22.2 | 0.22.2 |
tract-nnefcrates.io | < 0.21.16 | 0.21.16 |
Affected products
3- osv-coords2 versions
< 2.7.2-r1+ 1 more
- (no CPE)range: < 2.7.2-r1
- (no CPE)range: < 2.7.2-r1
Patches
Vulnerability mechanics
References
8- github.com/advisories/GHSA-x5mv-8wgw-29hgghsaADVISORY
- github.com/sonos/tract/security/advisories/GHSA-x5mv-8wgw-29hgnvdWEB
- github.com/sonos/tract/commit/34c7df2c9bd2a36583e09b52f3e6319bf23102e8nvd
- github.com/sonos/tract/commit/828ec20f74c14c0923c7d7f0cff062ed222d630anvd
- github.com/sonos/tract/commit/ee5f019d39fb97f7d4431b883c3ce97b74c28a64nvd
- github.com/sonos/tract/releases/tag/0.21.16nvd
- github.com/sonos/tract/releases/tag/0.22.2nvd
- github.com/sonos/tract/releases/tag/v0.23.1nvd
News mentions
0No linked articles in our index yet.