Skip to main content
← All vulnerabilities
CVE-2026-55093GHSA-X5MV-8WGW-29HGRUSTSEC-2026-0217

tract-nnef: integer overflow in NNEF `.dat` tensor parser yields an out-of-bounds read on model load

Published
Last modified
published
Not listed in CISA KEV
Fix
Fixed in 0.21.16, 0.22.2, 0.23.1
None on Exploit-DB, Metasploit or GitHub
  • Exploitation: poc
  • Automatable: no
  • Technical impact: partial

Summary

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.

Timeline

  1. CVE published
  2. CISA SSVC assessment

    Exploitation: poc · Automatable: no · Technical impact: partial

  3. Record last updated

Affected versions

  • sonos tract

    fixed in 0.21.16; From 0.22.0 · fixed in 0.22.2; From 0.23.0 · fixed in 0.23.1

  • tract-nnefcrates.io

    From 0.23.0 · fixed in 0.23.1

  • tract-nnefcrates.io

    From 0.22.0 · fixed in 0.22.2

  • tract-nnefcrates.io

    fixed in 0.21.16

  • tract-nnefcrates.io

    From 0.0.0-0 · fixed in 0.21.16; From 0.22.0 · fixed in 0.22.2; From 0.23.0 · fixed in 0.23.1

References 7

GitHub · sonos/tract