{"id":"ghsa-ghsa-gx7p-2j49-hfq4","seq":661319,"kind":"advisory","title":"vLLM : DoS via unvalidated mm_processor_kwargs in multimodal image processors (Isaac, InternVL, H2OVL, etc.)","url":"https://github.com/advisories/GHSA-gx7p-2j49-hfq4","permalink":"https://vulnpulse.app/items/ghsa-ghsa-gx7p-2j49-hfq4","summary":"影響: pip:vllm >= 0.7.3, < 0.31.0 → 0.31.0\nAPI clients can supply arbitrary mm_processor_kwargs per request. vLLM forwards these values through the multimodal rendering and processing pipeline to Hugging Face and custom image/video processor constructors and calls. Resource-sensitive values such as image size, patch size, patch count, crop count, or frame count can therefore be controlled by a remote client. For processors that upscale inputs, a single request can force an allocation large enough to terminate the API server or worker wi…","severity":"high","techs":["python"],"cves":["CVE-2026-114055"],"tags":["GHSA-gx7p-2j49-hfq4","vllm"],"source":{"id":"ghsa","name":"GitHub Advisory Database"},"published_at":"2026-10-23T20:58:36.000Z","fetched_at":"2026-10-24T04:50:51.468Z"}