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.
{
"cna_assigner": "GitHub_M",
"cwe_ids": [
"CWE-119"
],
"osv_generated_from": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/34xxx/CVE-2026-34159.json"
}{
"cpe": "cpe:2.3:a:ggml:llama.cpp:*:*:*:*:*:*:*:*",
"extracted_events": [
{
"introduced": "0"
},
{
"fixed": "b8492"
}
],
"source": [
"AFFECTED_FIELD",
"CPE_RANGE",
"REFERENCES"
]
}
"https://storage.googleapis.com/cve-osv-conversion/osv-output/CVE-2026-34159.json"
[
{
"deprecated": false,
"digest": {
"line_hashes": [
"26657247428671023425400431060088502962",
"55258668221154798141094597549135019108",
"99497843585409811243148330622103885367",
"324820574679158266416097464151363772427"
],
"threshold": 0.9
},
"id": "CVE-2026-34159-f4e1990e",
"signature_type": "Line",
"signature_version": "v1",
"source": "https://github.com/ggml-org/llama.cpp/commit/39bf0d3c6a95803e0f41aaba069ffbee26721042",
"target": {
"file": "ggml/src/ggml-rpc/ggml-rpc.cpp"
}
}
]
"2026-08-12T15:33:33Z"