CVE-2025-46570

CVSS v3 Score
2.6
Low

Vulnerability Description

vLLM is an inference and serving engine for large language models (LLMs). Prior to version 0.9.0, 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). These timing differences caused by matching chunks are significant enough to be recognized and exploited. This issue has been patched in version 0.9.0.

CVSS:2.5(Low)

Non-constant-time comparison of CSRF tokens in UIDL request handler in com.vaadin:vaadin-server versions 7.0.0 through 7.7.23 (Vaadin 7.0.0 through 7.7.23), and 8.0.0 through 8.12.2 (Vaadin 8.0.0 thro...

CVSS:2.5(Low)

Non-constant-time comparison of CSRF tokens in UIDL request handler in com.vaadin:flow-server versions 1.0.0 through 1.0.13 (Vaadin 10.0.0 through 10.0.16), 1.1.0 prior to 2.0.0 (Vaadin 11 prior to 14...

CVSS:2.5(Low)

Non-constant-time comparison of CSRF tokens in endpoint request handler in com.vaadin:flow-server versions 3.0.0 through 5.0.3 (Vaadin 15.0.0 through 18.0.6), and com.vaadin:fusion-endpoint version 6....

CVSS:2.4(Low)

In django-basic-auth-ip-whitelist before 0.3.4, a potential timing attack exists on websites where the basic authentication is used or configured, i.e. BASIC_AUTH_LOGIN and BASIC_AUTH_PASSWORD is set....

CVSS:2.9(Low)

vodozemac is an open source implementation of Olm and Megolm in pure Rust. Versions before 0.7.0 of vodozemac use a non-constant time base64 implementation for importing key material for Megolm group ...

CVSS:3.7(Low)

The vantage6 technology enables to manage and deploy privacy enhancing technologies like Federated Learning (FL) and Multi-Party Computation (MPC). It is possible to find out usernames from the respon...