Description (CNA)
SGLang through 0.5.21 trusts caller-supplied mm_hashes values as the multimodal prefix-cache key, allowing remote attackers to poison cached KV entries across requests. Attackers can send a different image under a known or predicted hash via /generate to disclose another request's image content or corrupt victims' responses.
Affected (CNA)
| Vendor / product | Versions |
|---|---|
| sgl-project / sglang | affected 0.5.13 through 0.5.21 |
CWE
CWE-345
References
- https://github.com/sgl-project/sglang/issues/44641
- https://github.com/sgl-project/sglang/blob/v0.5.21/python/sglang/srt/managers/tokenizer_manager.py#L1183-L1215
- https://github.com/sgl-project/sglang/blob/v0.5.21/python/sglang/srt/managers/io_struct.py#L205-L215
- https://github.com/sgl-project/sglang/pull/44653
- https://www.vulncheck.com/advisories/sglang-through-0.5.21-kv-cache-poisoning-via-mm-hashes-in-generate
ADP enrichment
No English description supplied.