repository icon
vllm-project / vllm Project repository

A high-throughput and memory-efficient inference and serving engine for LLMs

Submit a report

FIRST INTERACTION

WITHIN34 DAYS

REVIEW

WITHIN35 DAYS

FIX

WITHINN/A DAYS

vLLM Server Unsafe Deserialization Leads to Arbitrary Code Execution
racerz-fighting
not applicable
Remote Code Execution in Insecure Pickle Deserialization via recv_obj() in class...
seaw1nd
duplicate
Critical
vLLM Server Unsafe Deserialization Leads to Arbitrary Code Execution
avilum
informative
Critical
Remote Code Execution by Pickle Deserialization via MessageQueue.dequeue() Broad...
zpbrent
Critical
•$1500•CVE-2024-11041
"POST /v1/completions" and "POST /v1/embeddings" Denials of Service
rh-tguittet
High
•$750•CVE-2024-11040
vllm has Cloudpickle deserializes arbitrary command execution
hexian2001
duplicate
Critical
Remote Code Execution by Pickle Deserialization via AsyncEngineRPCServer() RPC s...
zpbrent
Critical
•$1500•CVE-2024-9053
Remote Code Execution by Pickle Deserialization via recv_object() distributed tr...
zpbrent
Critical
•$1500•CVE-2024-9052
Command Injection in nccl_integrity_check function
vanirxxx
informative
High
Malicious model to RCE by torch.load in hf_model_weights_iterator (as well as th...
dogewatch
informative
Critical