Code execution via unsynchronized JNI context release
Published Aug 6, 2026 · Updated Aug 6, 2026
Race-condition use-after-free in llama.cpp 0.9.0 through 0.17.1 allows attackers to execute code through the Android JNI wrapper. The bench_1model() and free_1context() JNI paths access and free the same llama_context without synchronization, letting llama_batch_allocr::clear() dereference attacker-controlled freed memory. Triggering the race requires concurrent benchmark and context-release operations plus user interaction; successful exploitation runs code in the Android app process.
Summary
What happened
Race-condition use-after-free in llama.cpp 0.9.0 through 0.17.1 allows attackers to execute code through the Android JNI wrapper. The bench_1model() and free_1context() JNI paths access and free the same llama_context without synchronization, letting llama_batch_allocr::clear() dereference attacker-controlled freed memory. Triggering the race requires concurrent benchmark and context-release operations plus user interaction; successful exploitation runs code in the Android app process.
The record
- CVE
- CVE-2026-70640
- Published
- Aug 6, 2026
- Updated
- Aug 6, 2026
- Vendor
- Georgi Gerganov
- Product
- llama.cpp
- Classifications
- CWE-362, CAPEC-26, T1203
- Attack vector
- local
- Privileges
- unauthenticated
Timeline
How it unfolded
- Aug 6, 2026CVE publishedPublication date reported by the CVE source.
- Aug 6, 2026Record updatedLatest update available in the CVE record.
Exploitability
Present is not the same as exploitable
Compare your product and version with the public record. A matching version still requires validation against your environment.
Is a vulnerable build present?
Compare these published version ranges with your installed build and any vendor patches.
- Affected versionversion=0.9.0 <=0.17.1
- Affected versionversion=b1886 <=b7445
What conditions does exploitation require?
What is affected?
Published CVSS scores
CVSS describes severity. EPSS estimates exploitation probability.
Attacks
What attackers are doing with it
Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.
Weakness, pattern, technique
Public exploit references
No public exploit references are available in this record.
Labels summarize the accepted research assessment. They do not indicate a test against your environment.
Technologies
Your stack
See the directory against your own environment.
Your stack
Check the software in your environment
Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.
Book a demo