CVE-2026-34159

critical
Published 2026-04-01 · Modified 2026-04-30
CVSS v3
9.8
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
CVSS v2
VIR risk
9.8

Description

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.

Predictions

Exploit likelihood
97%
Patch ETA

Heuristic predictions, AS-IS, for prioritization only.

Mitigations

vendor Authored 2026-05-27

Vendor advisory: debian — https://security-tracker.debian.org/tracker/CVE-2026-34159

vendor Authored 2026-05-27

Vendor advisory: security-advisories@github.com — https://github.com/ggml-org/llama.cpp/security/advisories/GHSA-j8rj-fmpv-wcxw

vendor Authored 2026-05-27

Vendor advisory: security-advisories@github.com — https://github.com/ggml-org/llama.cpp/pull/20908

vendor Authored 2026-05-27

Vendor advisory: security-advisories@github.com — https://github.com/ggml-org/llama.cpp/commit/39bf0d3c6a95803e0f41aaba069ffbee26721042

OS impact

OSVersionStatusFixed in
debian debianforkyfixed8611+dfsg-1
debian debiansidfixed8611+dfsg-1

Application impact

VendorProductVersionsFixed
ggmlllama.cpp{"endExcluding":"b8492"}b8492

References

CWEs

CWE-119

Verify integrity in audit chain (admin only). AS-IS.