Use of Uninitialized Resource

Overview vllm is an A high-throughput and memory-efficient inference and serving engine for LLMs Affected versions of this package are vulnerable to Use of Uninitialized Resource via the hasmambalayers function in the KV Block Handler. An attacker can cause unintended behavior by leaking data...

vLLM 安全漏洞

vLLM is an open-source inference and service engine designed for LLM models, featuring high throughput and efficient memory usage. Versions of vLLM prior to 0.19.0 contained a security vulnerability. This vulnerability stemmed from a function in the KV Block Handler component called...