llama.cpp b2877 before b11935 contains an uncaught exception vulnerability in the ggml-rpc server that allows unauthenticated remote attackers to crash the rpc-server by sending oversized message lengths. After the HELLO handshake, attackers can send a command such as SET_TENSOR with a length above std::vector::max_size(), triggering an uncaught std::length_error that aborts the process and denies service to all clients.
Fix
Patch, vendor advisory and mitigation references on NVD 2
- https://github.com/ggml-org/llama.cpp/pull/31204 patch, issue-tracking
- https://github.com/ggml-org/llama.cpp/commit/7ecb57f7adbea2928aa468ad2aecf05b10725ad4 patch
Every line above is quoted from NVD, CISA, Microsoft, the vendor's own advisory or DISA; nothing is generated.
Change history 1
- First seenCVSS 8.7 · ggml-orgNVD modified 2026-10-29 13:08
Affected (CPE criteria) 0
NVD has not attached CPE configurations yet.
References 6
- https://github.com/ggml-org/llama.cpp/pull/31204 patch, issue-tracking
- https://github.com/ggml-org/llama.cpp/commit/7ecb57f7adbea2928aa468ad2aecf05b10725ad4 patch
- https://github.com/ggml-org/llama.cpp product
- https://github.com/ggml-org/llama.cpp/releases/tag/b11935 release-notes
- https://github.com/ggml-org/llama.cpp/blob/8542e94350f0ade01644db6f1d1520da3814dafd/ggml/src/ggml-rpc/ggml-rpc.cpp#L344-L356 technical-description
- https://www.vulncheck.com/advisories/llama-cpp-b2877-before-b11935-ggml-rpc-server-dos-via-oversized-recv-msg-length third-party-advisory