CVE-2026-115913
MEDIUM
MEDIUM
Priority: 0.02
BB Relevance:
63/100
Published: 2026-10-26
Modified: 2026-10-28
First seen: 2026-10-29
Status: active
Attack Surface
Network
Low Complexity
No Interaction
CVSS Vector Breakdown
AV:
Network
AC:
Low
PR:
Low
UI:
None
S:
Unchanged
C:
None
I:
None
A:
High
Description
vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the Phi-4 multimodal audio encoder in phi4mm_audio.py and phi4mm_utils.py uses eager attention by default, allowing a remote client to submit one legal long audio clip that causes MultiHeadedAttention.forward_attention to materialize and retain full attention matrices in self.attn. The scaled dot-product attention branch does not materialize these matrices. The resulting quadratic GPU-memory use can cause an out-of-memory condition and terminate EngineCore for all users. This issue is fixed in version 0.32.0.
Weakness Classification (1)
References & Advisories (4)
Release Notes
Patch
Patch
Exploit
Patch
Vendor Advisory