CVE 6.9 MEDIUM

vLLM: temperature=NaN and temperature=Infinity bypass validation and propagate to GPU kernels_CVE-2026-54235

6.9 / 10
MEDIUM
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N

Description

vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.23.1rc0, ll temperature validation gates use comparison operators (<, >), which silently evaluate to False for NaN and for positive Infinity in Python's IEEE 754 float semantics. Both values pass every guard and propagate to GPU sampling kernels, where they produce undefined behavior or CUDA errors that can crash the inference worker. This vulnerability is fixed in 0.23.1rc0.

Basic Information

ID CVE-2026-54235
Source GitHub_M
Published Jun 22, 2026 at 21:59

Affected Product

Vendor vllm-project
Product vllm
Version < 0.23.1rc0
Affected Versions vllm-project vllm < 0.23.1rc0

CWE Classification

References

💭 Join the Security Discussion

🔒 Your email address will not be published. Required fields are marked *

⚠️ Please be respectful and constructive in your comments. Security discussions should remain professional.