CVE-2024-34359
Critical 9.7No known exploitationFix availablellama-cpp-python vulnerable to Remote Code Execution by Server-Side Template Injection in Model Metadata
Vendor advisory: GHSA-56xg-wfcc-g829
Record last updated
CVSS score
Critical on a 0 to 10 scale
EPSS, chance of exploitation in 30 days
No EPSS score available
CISA remediation deadline
Not in the KEV catalog
Description
llama-cpp-python is the Python bindings for llama.cpp. `llama-cpp-python` depends on class `Llama` in `llama.py` to load `.gguf` llama.cpp or Latency Machine Learning Models. The `__init__` constructor built in the `Llama` takes several parameters to configure the loading and running of the model. Other than `NUMA, LoRa settings`, `loading tokenizers,` and `hardware settings`, `__init__` also loads the `chat template` from targeted `.gguf` 's Metadata and furtherly parses it to `llama_chat_format.Jinja2ChatFormatter.to_chat_handler()` to construct the `self.chat_handler` for this model. Nevertheless, `Jinja2ChatFormatter` parse the `chat template` within the Metadate with sandbox-less `jinja2.Environment`, which is furthermore rendered in `__call__` to construct the `prompt` of interaction. This allows `jinja2` Server Side Template Injection which leads to remote code execution by a carefully constructed payload.
Weakness: CWE-76
Patching and mitigation
- Patch status
- Fix available (source: GitHub Advisory Database)
- Fix documented (source date)
- Mitigation
- None documented
- Vendor guidance
- github.com (opens in a new tab)
Affected products
| Vendor | Product | Affected versions | Fixed in | Per |
|---|---|---|---|---|
| abetlen | llama-cpp-python | >= 0.2.30, <= 0.2.71 | Not documented | CVE Program |
| pip | llama-cpp-python | from 0.2.30 up to and including 0.2.71 | 0.2.72 | GitHub Advisory Database |
Timeline
Disclosure
CVE record published by GitHub_M
Source: CVE Program (opens in a new tab)
Patch
Fix documented by GitHub Advisory Database
Vendor guidance and references
- https://github.com/abetlen/llama-cpp-python/commit/b454f40a9a1787b2b5659cd2cb00819d983185df (opens in a new tab)x_refsource_misc, x_transferred
- https://github.com/abetlen/llama-cpp-python/security/advisories/GHSA-56xg-wfcc-g829 (opens in a new tab)x_refsource_confirm, x_transferred
- GitHub Advisory GHSA-56xg-wfcc-g829 (opens in a new tab)github-advisory