CVE-2026-133710
ggml-org llama.cpp
Published 20 Nov 2026 · updated 20 Nov 2026 · Awaiting Analysis
8.5 High · CVSS 4.0, vulncheck.com
Description
llama.cpp b11003 before b12556 contains an integer overflow in the hrm_text loader load_arch_hparams() that allows attackers to bypass the block count check using crafted GGUF metadata. Attackers can supply a GGUF model whose layer factors wrap modulo 2^32, causing load_arch_tensors() to write tensor pointers past model.layers, corrupting heap memory and potentially enabling code execution.
References
- github.com/ggml-org/llama.cpp/blob/b12555/src/models/hrm-text.cpp
- github.com/ggml-org/llama.cpp/commit/50518b64c6f2191a116d2760042f106c4b3387d7
- github.com/ggml-org/llama.cpp/pull/32601
- github.com/ggml-org/llama.cpp/releases/tag/b12556
- www.vulncheck.com/advisories/llama-cpp-b11003-before-b12556-heap-out-of-bounds-write-via-hrm-text-model-loader