CVE-2026-93838
SGLang versions through 0.5.20 contain an unbounded memory allocation vulnerability in handle_staging_req() that fails to validate chunk_idx from ZMQ STAGING_REQ frames in prefill/decode disaggregation deployments. Attackers with access to the decode engine's internal ZMQ rank port can send a frame with an extremely large chunk_idx value, causing the scheduler to allocate memory until the system runs out and terminates the process.
Fraqueza
SGLang versions through 0.5.20 contain an unbounded memory allocation vulnerability in handle_staging_req() that fails to validate chunk_idx from ZMQ STAGING_REQ frames in prefill/decode disaggregation deployments. Attackers with access to the decode engine's internal ZMQ rank port can send a frame with an extremely large chunk_idx value, causing the scheduler to allocate memory until the system runs out and terminates the process.
Referências
- https://github.com/sgl-project/sglang
- https://github.com/sgl-project/sglang/blob/v0.5.20/python/sglang/srt/disaggregation/common/conn.py#L1636-L1640
- https://github.com/sgl-project/sglang/blob/v0.5.20/python/sglang/srt/disaggregation/common/staging_handler.py#L781
- https://github.com/sgl-project/sglang/blob/v0.5.20/python/sglang/srt/disaggregation/common/staging_handler.py#L848-L860
- https://github.com/sgl-project/sglang/issues/39764
- https://www.vulncheck.com/advisories/sglang-through-0.5.20-unbounded-memory-allocation-via-staging-req-chunk-idx
Encontre a falha antes de um atacante.
Entre com o GitHub e rode sua primeira auditoria em menos de um minuto. O plano gratuito não exige cartão de crédito.