AVID-2026-R1519

Description

llama.cpp global-buffer-overflow in ggml_type_size (CVE-2024-42477)

Details

llama.cpp provides LLM inference in C/C++. The unsafe type member in the rpc_tensor structure can cause global-buffer-overflow. This vulnerability may lead to memory data leakage. The vulnerability is fixed in b3561.

Reason for inclusion in AVID: CVE-2024-42477 describes a global-buffer-overflow vulnerability in llama.cpp (ggml_type_size) that can lead to memory data leakage. llama.cpp is a core AI inference library used to deploy/general-purpose AI models. This fits the software supply chain domain for AI systems (dependencies/runtime used to run models). The issue is a security vulnerability (memory safety) with an available fix. The CVE and advisory provide sufficient signals for AVID curation.

References

• NVD entry
• https://github.com/ggerganov/llama.cpp/security/advisories/GHSA-mqp6-7pv6-fqjf
• https://github.com/ggerganov/llama.cpp/commit/b72942fac998672a79a1ae3c03b340f7e629980b

Affected or Relevant Artifacts

• Developer: Meta
• Deployer: ggerganov
• Artifact Details:

Impact

AVID Taxonomy Categorization

• Risk domains: Security
• SEP subcategories: S0100: Software Vulnerability
• Lifecycle stages: L06: Deployment

CVSS

CWE

Other information

• Report Type: Advisory
• Credits:
• Date Reported: 2024-08-12
• Version: 0.3.3
• AVID Entry