vLLM: Public multimodal SHM handle forgery reaches worker-side `pickle.loads()` with insecure serialization disabled

vLLM is an inference and serving engine for large language models. Prior to 0.32.0, an authenticated caller can place forged shared-memory address and monotonic_id values in public multimodal features.kwargs_data. In scale-out deployments with shared-memory multimodal caching enabled, workers trust those values and pass attacker-selected bytes to pickle.loads(), allowing arbitrary Python execution even when insecure serialization is disabled. This issue is fixed in version 0.32.0.

Source published

26 Oct 2026 16:54:00 UTC

Source modified

26 Oct 2026 16:54:00 UTC

Observed here

26 Oct 2026 17:35:31 UTC

Source terms

CVE Program Terms of Use

Original source record · Cite this exact observation · Complete original publisher JSON

This view projects the publisher's source fields for reading; the original JSON retains additional fields and CVE enrichment containers. Attribution: cve-program · CVE Program Terms of Use. Source text is retained; field grouping and change labels are provided by ChangeIntel.

CVSS 3.1 · 7.5 CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H