Buffer overflow in Ggml Llama.cpp
CVE-2026-21869
llama.cpp is an inference of several LLM models in C/C++. In commits 55d4206c8 and prior, the n_discard parameter is parsed directly from JSON input in the llama.cpp server's completion endpoints without validation to ensure it's non-negative. When a negative value is supplied and the context fills up, llama_memory_seq_rm/add receives a reversed range and negative offset, causing out-of-bounds memory writes in the token evaluation loop. This deterministic memory corruption can crash the process or enable remote code execution (RCE). There is no fix at the time of publication.
Vulnerability class: Buffer Overflow
EPSS: 0.004 (35.9th percentile) — read the EPSS interpretation.
CVSS v3 metric
CVSS v3 base score 8.8 (High). Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H.
Affected products
- Ggml Llama.cpp
- Ggml-org Llama.cpp — versions <= 55d4206c8
Weakness classification (CWE)
References
- security-advisories@github.com (x_refsource_CONFIRM, Exploit, Vendor Advisory)
Frequently asked questions
- What is CVE-2026-21869?
- CVE-2026-21869 is a high-severity vulnerability in Ggml Llama.cpp, classified under Out-of-bounds Write. CVSS score: 8.8/10. Published 2026-01-08.
- How severe is CVE-2026-21869?
- High severity. CVSS v3 base score is 8.8 out of 10.