llama.cpp builds b4882 through b9058 contain a heap buffer overflow vulnerability in the KV cache state restore path where the state_read…
CVE-2026-43629 · GHSA-7p8h-8h8j-472c · EUVD-2026-54284 · published 6 Aug 2026
How serious is it
- Severity: critical, CVSS score 9.2 out of 10.
- No exploit prediction (EPSS) score yet.
- Not on the CISA list of exploited vulnerabilities.
- No public exploit in Exploit-DB.
Is your app affected by CVE-2026-43629?
Paste this prompt into Claude Code, Cursor or any AI coding tool: it runs a free Shippable scan of your project and fixes what it finds. Built with Lovable, Bolt or v0? Scan your live app. Or upload your package file.
Who is affected
- ggml-org llama.cpp: b4882 ≤b9058
How to check
- Find out whether you run the affected product and which version, then compare it with the versions above.
How to fix
Copies a prompt for Claude Code, Cursor or Codex that asks it to check whether you run this product.
- Apply the vendor's update or mitigation from the references below. If there is none yet, limit exposure of the affected service until there is.
- If a vulnerable version was live, check your logs and rotate the secrets that server could read.
The advisory
llama.cpp builds b4882 through b9058 contain a heap buffer overflow vulnerability in the KV cache state restore path where the state_read_data() function computes write size without overflow checking, allowing attackers with write access to the slot_save_path directory to corrupt heap memory. Attackers can craft malicious state files where cell_count multiplication overflows or exceeds tensor buffer allocation to write attacker-controlled bytes past buffer boundaries, potentially resulting in heap metadata corruption, model weight corruption, or arbitrary code execution via function pointer overwrite.
Check your whole app
Dependencies are only half of it. Free and read-only: before you push, run npx shippable-dev scan localhost:3000 in your project folder, no account needed. For your live app, you prove it's yours with a one-line tag first.
References
- github.com/Vladimir-tokarev-cyera/llama-cpp-security-patches
- www.vulncheck.com/advisories/llama-cpp-b4882-b9058-buffer-overflow-in-kv-cache-state-restore
Merged from ENISA EUVD. Last updated 30 Oct 2026. All vulnerabilities · RSS · JSON