GHSA-8397-8QF6-F2GM vLLM: Phi-4 multimodal audio encoder materializes and retains full attention matrices: a single max-length audio OOM-crashes the vLLM engine

Summary The Phi-4 multimodal audio encoder phi4mmaudio.py / phi4mmutils.py runs its conformer attention in eager mode by default: the official HF config of microsoft/Phi-4-multimodal-instruct does not set useptscaleddotproductattention, so the default False applies phi4mmaudio.py:166, :876;...

CVE-2026-115913 vLLM: Phi-4 multimodal audio encoder materializes and retains full attention matrices: a single max-length audio OOM-crashes the vLLM engine

vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the Phi-4 multimodal audio encoder in phi4mmaudio.py and phi4mmutils.py uses eager attention by default, allowing a remote client to submit one legal long audio clip that causes MultiHeadedAttention.forwardattenti...

CVE-2026-115913

Prior to version 0.32.0 , the vLLM inference and serving engine is vulnerable to a Denial of Service (DoS) condition via its Phi-4 multimodal audio encoder. Specifically, the components phi4mm_audio.py and phi4mm_utils.py use eager attention by default, causing MultiHeadedAttention.forward_attent...

CVE-2026-115913 vLLM: Phi-4 multimodal audio encoder materializes and retains full attention matrices: a single max-length audio OOM-crashes the vLLM engine

vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the Phi-4 multimodal audio encoder in phi4mmaudio.py and phi4mmutils.py uses eager attention by default, allowing a remote client to submit one legal long audio clip that causes MultiHeadedAttention.forwardattenti...

CVE-2026-115913 vLLM: Phi-4 multimodal audio encoder materializes and retains full attention matrices: a single max-length audio OOM-crashes the vLLM engine

vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the Phi-4 multimodal audio encoder in phi4mmaudio.py and phi4mmutils.py uses eager attention by default, allowing a remote client to submit one legal long audio clip that causes MultiHeadedAttention.forwardattenti...

PT-2026-123436

Name of the Vulnerable Software and Affected Versions vLLM versions prior to 0.32.0 Description The Phi-4 multimodal audio encoder in phi4mm audio.py and phi4mm utils.py uses eager attention by default. This allows a remote client to submit a long audio clip that triggers the...

CVE-2026-93841

A flaw was found in vLLM. This memory corruption vulnerability exists in the Triton bincountkernel, where prompt token identifiers IDs are used to index a bitset without proper bounds checking against the vocabulary size. A remote attacker can exploit this by submitting multimodal audio requests...

SUSE CVE-2026-93841

vLLM through 0.29.0 contains a memory corruption vulnerability in the Triton bincountkernel where prompt token IDs index the penalty prompt-presence bitset without bounds checking against vocabulary size. Attackers can submit multimodal audio requests with tokens equal to vocabulary size, causing...

Out-of-bounds Write

Overview vllm is an A high-throughput and memory-efficient inference and serving engine for LLMs Affected versions of this package are vulnerable to Out-of-bounds Write via the Triton bincountkernel, where prompt token IDs index the penalty prompt-presence bitset without bounds checking against t...

CVE-2026-93841

vLLM through 0.29.0 contains a memory corruption vulnerability in the Triton bincountkernel where prompt token IDs index the penalty prompt-presence bitset without bounds checking against vocabulary size. Attackers can submit multimodal audio requests with tokens equal to vocabulary size, causing...

PYSEC-0000-CVE-2026-93841

vLLM through 0.29.0 contains a memory corruption vulnerability in the Triton bincountkernel where prompt token IDs index the penalty prompt-presence bitset without bounds checking against vocabulary size. Attackers can submit multimodal audio requests with tokens equal to vocabulary size, causing...

CVE-2026-93841: Improper Validation of Array Index

vLLM through 0.29.0 contains a memory corruption vulnerability in the Triton bincountkernel where prompt token IDs index the penalty prompt-presence bitset without bounds checking against vocabulary size. Attackers can submit multimodal audio requests with tokens equal to vocabulary size, causing...