CVE-2026-42627
6.2 MEDIUMIn Arm ArmNN through 2026-03-27, an integer overflow in TensorShape::GetNumElements() in armnn/Tensor.cpp allows a crafted TFLite model f...
Published: 2026-05-22 · Last updated: 2026-05-26
Severity and scoring
- CVSS
- 6.2 MEDIUM
- Vector
- CVSS:3.1/AV:L/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
- CWE
- CWE-190
Description
In Arm ArmNN through 2026-03-27, an integer overflow in TensorShape::GetNumElements() in armnn/Tensor.cpp allows a crafted TFLite model file to bypass buffer size validation and trigger a heap-based buffer over-read during model optimization. The overflow occurs when multiplying tensor dimensions using 32-bit unsigned arithmetic without overflow detection, causing GetNumBytes() to return an understated allocation size. During Optimize()->InferOutputShapes(), the BatchToSpaceNdLayer reads beyond the allocated buffer.
Source: NVD
References
Related CVEs
Same CWE
- CVE-2026-10649 — A flaw was found in Pacemaker (8.6 HIGH)
- CVE-2026-53705 — A flaw was found in GStreamer's WavPack audio decoder in gst-plugins-good (7.6 HIGH)
- CVE-2026-52722 — A signed integer overflow vulnerability was found in GStreamer's VMnc decoder (7.1 HIGH)
- CVE-2025-55647 — An Out-of-Memory in the mp4_mux_cenc_insert_pssh function (filters/mux_isom.c) of GPAC MP4Box v2.4 allows attackers to cause a Denial of ... (5.5 MEDIUM)
- CVE-2026-6045 — LibreOffice can import EMF+ graphics, which may be embedded in documents