CVE-2024-58400: fallo de gravedad alta en ggml-org llama.cpp
llama.cpp before b3701 Out-of-Bounds Read in llama_decode via Token IDs
Publicada el
Aviso por correo cuando entre en el KEV, tenga PoC pública o suba de nota.
21Vexday Risk Score
Sin señal de explotación. Ningún artefacto público de explotación conocido hasta ahora.
ssvc Trackcvss 8.8epss 0.4%
llama.cpp before b3701 contains an out-of-bounds read vulnerability in llama_decode_internal and llama_encode_internal that allows remote unauthenticated attackers to read memory by supplying unvalidated token IDs. Attackers can send integer-array prompts to the llama-server /completion endpoint with IDs outside 0 to n_vocab to read embedding memory out of bounds or crash the process.
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:L/VI:N/VA:H/SC:N/SI:N/SA:N
Productos afectados
ggml-org · llama.cppCVEs relacionadas — ggml-org llama.cpp
En el mismo producto, de las más peligrosas a las menos.
CVE-2026-34159CRITICALllama.cpp: Unauthenticated RCE via GRAPH_COMPUTE buffer=0 bypass in llama.cpp RPC backendEPSS 1.6%CVE-2026-78148MEDIUMggml-org llama.cpp ggml-RPC Server ggml-rpc.cpp graph_compute null pointer dereferenceEPSS 0.7%CVE-2026-17501MEDIUMggml-org llama.cpp JSON-Schema-to-GBNF Conversion json-schema-to-grammar.cpp transform recursionEPSS 0.7%CVE-2026-17500MEDIUMggml-org llama.cpp json-schema-to-grammar.cpp _visit_pattern null pointer dereferenceEPSS 0.7%CVE-2026-86317MEDIUMggml-org llama.cpp RPC Server ggml-rpc.cpp deserialize_tensor assertionEPSS 0.7%CVE-2026-43629CRITICALllama.cpp b4882–b9058 Buffer Overflow in KV Cache State RestoreEPSS 0.7%
Referencias
https://github.com/ggml-org/llama.cpp/blob/b3681/examples/server/server.cpp#L789https://github.com/ggml-org/llama.cpp/commit/eae597182cb61bbde0b26e7cec5999d28b9327fehttps://github.com/ggml-org/llama.cpp/commit/faf69d4237c9ae4d7f572b4674d1002463e8acd3https://github.com/ggml-org/llama.cpp/pull/9359https://github.com/ggml-org/llama.cpp/releases/tag/b3701https://www.vulncheck.com/advisories/llama-cpp-before-b3701-out-of-bounds-read-in-llama-decode-via-token-ids