CVE-2025-52566

🎯 CVE-2025-52566 🧪 PoC 🇪🇺 EUVD

💬 0 Diskussion 📄 .md Alle CVEs anzeigen ✕

CVE-2025-52566 | 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.

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.

Klassifikation & Betroffenheit:

CWE-119CWE-195 ggml-org llama.cpp < b5721ggml llama.cpp *

Improper Restriction of Operations within the Bounds of a Memory Buffer 🎯 High

The product performs operations on a memory buffer, but it reads from or writes to a memory location outside the buffer's intended boundary.

🛡️ Empfohlene Mitigation: Use a language that does not allow this weakness to occur or provides constructs that make this weakness easier to avoid. For example, many languages that perform their own memory management, such as Java and Perl, are not subject to buffer overflows. Other languages, such as Ada and C#, typically provide overflow prot…

Vollständige Definition bei MITRE ➔

⚔️ Typische Angriffsmuster (CAPEC):

CAPEC-10 · Buffer Overflow via Environment Variable… (High) CAPEC-14 · Client-side Injection-induced Buffer Ove… (High) CAPEC-24 · Filter Failure through Buffer Overflow (High) CAPEC-42 · MIME Conversion (High) CAPEC-44 · Overflow Binary Resource File (Very High) CAPEC-45 · Buffer Overflow via Symbolic Links (High) CAPEC-100 · Overflow Buffers (Very High) CAPEC-123 · Buffer Manipulation (Very High)

🧩 Ähnliche Schwachstellen (Hersteller/Klasse/Score-Band):

CVE-2026-116508 10.0 CVE-2026-79851 10.0 CVE-2026-79847 10.0 CVE-2026-114383 10.0 CVE-2026-114382 10.0

🩹 Patch verfügbar (OSV):

🩹 dd6e6d0b6a4bbe3ebfc931d1eb14db2f2b1d70af (Commit)

📚 Referenzen & Quellen:

🇪🇺 Original-Eintrag (ENISA EUVD) 🇺🇸 NVD-Eintrag

Ausnutzungs-Zeitleiste:

Reserviert 2025

Veröffentlicht 06/2025

Heute 07.11.

CVSS-Vektor-Analyse: 8.6

AV · Angriffsvektor Lokal

AC · Komplexität Gering

PR · Privilegien Keine

UI · Interaktion Erforderlich

S · Scope Verändert

C · Vertraulichkeit Hoch

I · Integrität Hoch

A · Verfügbarkeit Hoch

CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:C/C:H/I:H/A:H

CNA-Bewertung (GitHub_M): 8.6 CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:C/C:H/I:H/A:H

Veröffentlicht:24.06.2025

Aktualisiert:17.06.2026 09:36

Assigner (CNA):GitHub_M

EUVD-ID:EUVD-2025-19074

Quellen: 🇪🇺 EUVD-Datenbank (ENISA) + 🇺🇸 NVD-Anreicherung · 24-h-Cache

CWE-119 ✓ Offizieller Patch / Advisory verfügbar

💡 Gegenmaßnahme: Sicherheits-Update des Herstellers zeitnah einspielen und Netzwerksegmentierung prüfen.

Tier	CVEs	Anteil
≥90 %	785	0,2 %
≥50 %	3.557	0,9 %
≥10 %	13.139	3,2 %
<10 %	391.261	95,7 %

Tier

CVEs

Anteil

≥90 %

785

0,2 %

≥50 %

3.557

0,9 %

≥10 %

13.139

3,2 %

<10 %

391.261

95,7 %

8.6 HIGH

EPSS

0.4%

CVE-2025-52566 💻 Lokal 🔓 Keine Authentifizierung nötig

ggml-org

CVE-2025-52566 | 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.

CWE-119 ✓ Offizieller Patch / Advisory verfügbar

💡 Gegenmaßnahme: Sicherheits-Update des Herstellers zeitnah einspielen und Netzwerksegmentierung prüfen.

vor 1 Jahr ↻ vor 1 Jahr

🛡️ Details ➔

Zero-Day & Vulnerability Intelligence Hub