Advisory

GHSA-4qjh-9fv9-r85r

Also known as: CVE-2025-46570, PYSEC-2025-53

Summary

Potential Timing Side-Channel Vulnerability in vLLM’s Chunk-Based Prefix Caching

Details

This issue arises from the prefix caching mechanism, which may expose the system to a timing side-channel attack.

Description
When a new prompt is processed, if the PageAttention mechanism finds a matching prefix chunk, the prefill process speeds up, which is reflected in the TTFT (Time to First Token). Our tests revealed that the timing differences caused by matching chunks are significant enough to be recognized and exploited.

For instance, if the victim has submitted a sensitive prompt or if a valuable system prompt has been cached, an attacker sharing the same backend could attempt to guess the victim's input. By measuring the TTFT based on prefix matches, the attacker could verify if their guess is correct, leading to potential leakage of private information.

Unlike token-by-token sharing mechanisms, vLLM’s chunk-based approach (PageAttention) processes tokens in larger units (chunks). In our tests, with chunk_size=2, the timing differences became noticeable enough to allow attackers to infer whether portions of their input match the victim's prompt at the chunk level.

Environment

  • GPU: NVIDIA A100 (40G)
  • CUDA: 11.8
  • PyTorch: 2.3.1
  • OS: Ubuntu 18.04
  • vLLM: v0.5.1

Configuration: We launched vLLM using the default settings and adjusted chunk_size=2 to evaluate the TTFT.

Leakage
We conducted our tests using LLaMA2-70B-GPTQ on a single device. We analyzed the timing differences when prompts shared prefixes of 2 chunks, and plotted the corresponding ROC curves. Our results suggest that timing differences can be reliably used to distinguish prefix matches, demonstrating a potential side-channel vulnerability.
<img src="https://github.com/user-attachments/assets/db3491e9-02b7-424c-9b6d-56f553b39f2f" alt="roc_curves_combined_block_2" width="400"/>

Results
In our experiment, we analyzed the response time differences between cache hits and misses in vLLM's PageAttention mechanism. Using ROC curve analysis to assess the distinguishability of these timing differences, we observed the following results:

  • With a 1-token prefix, the ROC curve yielded an AUC value of 0.571, indicating that even with a short prefix, an attacker can reasonably distinguish between cache hits and misses based on response times.
  • When the prefix length increases to 8 tokens, the AUC value rises significantly to 0.99, showing that the attacker can almost perfectly identify cache hits with a longer prefix.

Fixes

  • https://github.com/vllm-project/vllm/pull/17045

Severity

Low Basis: The source's own band: LOW

Dates

Published
2025-05-28 18:02 UTC
Modified
2026-08-07 08:12 UTC

Affected packages and fixes

vllm

PyPI · All systems

Affected

  • introduced 0, fixed 0.9.0 (ecosystem)

Fixed versions

0.9.0

Fix

  • Fixed version 0.9.0
    pip install --upgrade vllm==0.9.0

Fix commands are the export's own generic commands. Ojuju never runs them. Check them against your setup first.

Check your own computer: guard packages check

View this record on OSV.dev

Source

Source: GitHub Advisory Database (reviewed advisories) · CC BY 4.0

Sources & licences