llama.cpp is an inference of several LLM models in C/C++. Prior to version b5662, an attacker‐supplied GGUF model vocabulary can trigger a buffer overflow in llama.cpp’s vocabulary‐loading code. Specifically, the helper _try_copy in llama.cpp/src/vocab.cpp: llama_vocab::impl::token_to_piece() casts a very large size_t token length into an int32_t, causing the length check (if (length < (int32_t)size)) to be bypassed. As a result, memcpy is still called with that oversized size, letting a malicious model overwrite memory beyond the intended buffer. This can lead to arbitrary memory corruption and potential code execution. This issue has been patched in version b5662.
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:HGgml Llama.cpp
APPGgml< b5662
Related vulnerabilities
RCE w llama.cpp — brak walidacji granic w RPC backend (deserialization)
Zapis pod dowolny adres pamięci w llama.cpp poprzez wskaźnik rpc_tensor
llama.cpp b5693 and before is vulnerable to Uncontrolled Recursion in common/json-schema-to-grammar.cpp, resul...
llama.cpp b5693 and before has a Reachable Assertion via the gguf_reader::read function.
llama.cpp through commit 97f06e9, when started with the --reranking flag, allows remote attackers to cause a d...