cve.bz

CVE-2026-43630 ggml llama.cpp: Out-of-bounds Read

6.3 mediumEPSS 0.5%Fix available

Medium (6.3) in ggml llama.cpp. EPSS 0.5% estimated probability of exploitation activity in the next 30 days.

On this page
llama.cpp builds b4882 before b8487 contain an out-of-bounds read vulnerability in the recurrent memory state restore path that allows attackers with write access to the slot save directory to read memory past the end of the allocated cells array. Attackers can craft a malicious slot file with an oversized seq_id value to trigger an out-of-bounds read that leaks heap data including pointer values into server logs, defeating ASLR protections and facilitating further exploitation.

Severity

CVSS 4.0

6.3 medium
CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:L/VI:N/VA:L/SC:N/SI:N/SA:N

CVSS 3.1

6.5 medium
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:L
EPSS
0.5% (unchanged from 0.5% a week ago; 0.5% 30 days ago) estimated probability of exploitation activity in the next 30 days; 43rd percentile. Scored 2026-11-17. A forecast for this CVE across the internet, not for your own systems.
CISA's assessment (SSVC)
Exploitation: none · Automatable: yes · Technical impact: partial, assessed 2026-08-07. From CISA's Vulnrichment in the CVE List: whether it is exploited, whether exploiting it can be automated, and how much control it gives.
CWEs
CWE-125 Out-of-bounds Read
CWE-209 Generation of Error Message Containing Sensitive Information
OWASP Top 10
A10:2025 Mishandling of Exceptional Conditions
Published
Modified
NVD analysis
Analyzed, since modified
Record from
the NVD, synced

Fix and patched versions

Ask an AI about this CVE

Show the promptExplain CVE-2026-43630 for an engineer deciding whether to act. Use https://cve.bz/api/public/cves/CVE-2026-43630?via=ask-ai-copy and https://cve.bz/cve/CVE-2026-43630?via=ask-ai-copy as the source of truth: what is affected and which versions fix it, whether it is in CISA KEV, public exploits, and its EPSS (an internet-wide 30-day estimate of exploitation activity, not a forecast for my systems). If they give no fixed versions, use the CVE record https://cveawg.mitre.org/api/cve/CVE-2026-43630 or the vendor advisory, not third-party mirrors. Tell me how to check whether I am affected and how to verify a fix. Cite the sources and say where they are uncertain.

Opens a third-party assistant; its answer is its own, not cve.bz's. The prompt holds only this CVE's id and public links.

Affected

Versions

From the CNA's record in the CVE List.

ProductAffected
ggml-org llama.cppfrom b4882 before b8487

Linux distributions

DistributionReleasePackageFixed inAdvisory
Debian14llama.cpp8611+dfsg-1DEBIAN-CVE-2026-43630
Ubuntu26.04:LTS, 26.10llama.cppno fix listedUBUNTU-CVE-2026-43630

From the distributions' data via OSV.dev: the Debian Security Tracker; Ubuntu Security Notices © Canonical Ltd., CC BY-SA 4.0. Modified: reformatted and linked to CVE records. Credits. On the distributions' own pages: Debian · Ubuntu.

Red Hat

Red Hat rates this Moderate.

Red Hat's statement

Exploitation requires local write access to the slot storage directory to introduce crafted state files prior to a restoration operation. The impact is strictly bounded to low confidentiality loss from heap memory leakage into application logs and low availability impact from potential process crashes (C:L, I:N, A:L), with zero impact on data or system integrity. Deployments enforcing write-restricted filesystem permissions on session state directories or running purely stateless inference workloads are completely unaffected.

Red Hat's mitigation

Restrict write access to the llama.cpp slot save directory. Ensure that only authorized users or processes have write permissions to this directory. If the recurrent memory state feature is not in use, ensure the slot save directory is either removed or configured with highly restrictive permissions to prevent the placement of malicious files.

Contains Red Hat Security Data © Red Hat, Inc., licensed under CC BY 4.0. Source: https://access.redhat.com/security/data. Modified: reformatted. Red Hat's page for CVE-2026-43630.

Attack patterns (CAPEC)

  • CAPEC-540 Overread Buffers Likelihood: Low · Severity: High
  • CAPEC-7 Blind SQL Injection Likelihood: High · Severity: High
  • CAPEC-54 Query System for Information Likelihood: High · Severity: Low
  • CAPEC-215 Fuzzing for application mapping Likelihood: High · Severity: Low
  • CAPEC-463 Padding Oracle Crypto Attack Severity: High

The attack patterns MITRE's CAPEC relates to this record's weaknesses, as MITRE names them. Credits.

References

Check your stackJSON