Integer Overflow Vulnerability in llama.cpp by ggml-org
CVE-2026-27940

7.8HIGH

Key Information:

Vendor

Ggml-org

Status
Llama.cpp
Vendor
CVE Published:
12 March 2026

What is CVE-2026-27940?

The implementation of gguf_init_from_file_impl() in gguf.cpp within llama.cpp is susceptible to an integer overflow, which can result in an undersized heap allocation. This flaw enables an attacker to write more than 528 bytes of data, controlled by them, beyond the allocated buffer. This vulnerability is reminiscent of a previous issue documented in the same file but remains unaddressed in certain aspects despite prior resolutions. The flaw has been rectified in version b8146.

Affected Version(s)

llama.cpp < b8146

References

https://github.com/ggml-org/llama.cpp/security/advisories...
x_refsource_CONFIRM

CVSS V3.1

Score:
7.8
Severity:
HIGH
Confidentiality:
High
Integrity:
High
Availability:
High
Attack Vector:
Local
Attack Complexity:
Low
Privileges Required:
None
User Interaction:
Required
Scope:
Unchanged

Timeline

  • Vulnerability published

  • Vulnerability Reserved

.