OESA-2025-2372

    Dashboard / Vulnerabilities / OESA-2025-2372

    OESA-2025-2372

    Published: 11 Oct 2025Last Modified: 18 Aug 2026
    Upstream:
    CVE-2025-52566

    Summary: llama.cpp security update

    Details: Security Fix(es): llama.cpp is an inference of several LLM models in C/C++. Prior to version b5721, there is a signed vs. unsigned integer overflow in llama.cpp's tokenizer implementation (llama_vocab::tokenize) (src/llama-vocab.cpp:3036) resulting in unintended behavior in tokens copying size comparison. Allowing heap-overflowing llama.cpp inferencing engine with carefully manipulated text input during tokenization process. This issue has been patched in version b5721.(CVE-2025-52566)

    References: https://www.openeuler.org/zh/security/security-bulletins/detail/?id=openEuler-SA-2025-2372, https://nvd.nist.gov/vuln/detail/CVE-2025-52566

    Affected packages

    Package

    Name: llama.cpp

    Purl: pkg:rpm/openEuler/llama.cpp&distro=openEuler-24.03-LTS-SP2

    Affected ranges

    Type: ECOSYSTEM

    Events:

    Introduced- 0
    Fixed -20230815-5.oe2403sp2

    Affected versions

    Common Vulnerability Scoring System

    Attack Vector
    Network
    Adjacent
    Local
    Physical
    Privileges Required
    None
    Low
    High
    User Interaction
    None
    Required
    Scope
    Unchanged
    Changed
    Confidentiality
    None
    Low
    High
    Integrity
    None
    Low
    High
    Availability
    None
    Low
    High
    OESA-2025-2372 | CVE-DB