PT-2026-29570 · Ggml Org · Llama.Cpp

Published 2026-04-01 · Updated 2026-04-01 · CVE-2026-34159

Mitre

CVSS v3.1

9.8

Critical

AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H

llama.cpp is an inference of several LLM models in C/C++. Prior to version b8492, the RPC backend's deserialize tensor() skips all bounds validation when a tensor's buffer field is 0. An unauthenticated attacker can read and write arbitrary process memory via crafted GRAPH COMPUTE messages. Combined with pointer leaks from ALLOC BUFFER/BUFFER GET BASE, this gives full ASLR bypass and remote code execution. No authentication required, just TCP access to the RPC server port. This issue has been patched in version b8492.

Fix

Buffer Overflow

Weakness Enumeration

CWE-119

Related Identifiers

CVE-2026-34159

Affected Products

Llama.Cpp

References · 4

• https://nvd.nist.gov/vuln/detail/CVE-2026-34159 · Security Note
• https://github.com/ggml-org/llama.cpp/security/advisories/GHSA-j8rj-fmpv-wcxw · Note
• https://github.com/ggml-org/llama.cpp/commit/39bf0d3c6a95803e0f41aaba069ffbee26721042 · Note
• https://github.com/ggml-org/llama.cpp/pull/20908 · Note