UBUNTU-CVE-2026-116840: Integer Overflow
Published Oct 27, 2026
·Updated
llama.cpp b5854 before b11275 contains an integer overflow vulnerability in ggufinitfromreader that bypasses the CVE-2025-53630 size check when padding near-SIZEMAX tensor sizes. Attackers can supply a crafted GGUF file whose tensor size pads to zero, causing an undersized data blob and out-of-bounds reads by consumers.
Affected Software
2 affected components
ubuntu/llama.cpp*
ubuntu/llama.cpp*
Event History
Oct 27, 2026
CVE Published
via OSV·01:19 AM
Data Sourced
via OSV·01:19 AM
DescriptionSeverityAffected Software
Frequently Asked Questions
1
Which llama.cpp builds are affected?
The affected range is llama.cpp b5854 through builds before b11275.
2
What does an attacker need to exploit this issue?
An attacker needs to provide a crafted GGUF file and have a user or local process load it. The vector is local, requires no privileges, and requires user interaction.
3
What is the likely security impact?
The crafted file can cause consumers to perform out-of-bounds reads from an undersized data blob. The stated impact includes low confidentiality impact and high availability impact.