Description
vLLM through 0.29.0 contains a memory corruption vulnerability in the Triton _bincount_kernel 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 writes that corrupt concurrent requests' sampler state and alter repetition penalty behavior.
Affected products
vllm: vllm
References5
- github.com/vllm-project/vllm/pull/49081 Issue TrackingPatch
- github.com/vllm-project/vllm Product
- github.com/vllm-project/vllm/blob/v0.29.0/vllm/v1/worker/gpu/sa… Product
- github.com/vllm-project/vllm/blob/v0.29.0/vllm/v1/worker/gpu/sa… Product
- www.vulncheck.com/advisories/vllm-through-0.29.0-adjacent-reque… Third Party Advisory
Metrics
- CVSS 3.1
-
5.3
MEDIUM
Score source: NVD
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N - CVSS 4.0
-
6.3
MEDIUM
Score source: CNA
CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X - EPSS — exploit probability
- 0% percentile: 15 · as of 8 November 2026
- CWE
- CWE-129
Could this be exploited in your environment?
A patch closes the known hole. A penetration test shows what an attacker can still reach — with proof.