Description
InternLM LMDeploy through 0.17.0 contains a reachable assertion vulnerability in the DistServe decode migration loop that allows unauthenticated attackers to terminate the inference engine. Attackers can submit a migration_request with an empty remote_block_ids list to trigger an AssertionError that crashes the engine loop and causes subsequent inference requests to fail.
Affected products
| Vendor | Product | Affected | Fixed |
|---|---|---|---|
| InternLM | lmdeploy | ≤ 0.17.0 | — |
References
- GitHub Issue #4965 issue-tracking
- https://github.com/InternLM/lmdeploy/blob/v0.17.0/lmdeploy/pytorch/engine/engine_loop.py#L560-L564 technical-description
- https://github.com/InternLM/lmdeploy/blob/v0.17.0/lmdeploy/serve/openai/endpoints/completions.py#L130-L136 technical-description
- https://github.com/InternLM/lmdeploy/blob/v0.17.0/lmdeploy/pytorch/utils.py#L187-L214 technical-description
- https://github.com/InternLM/lmdeploy product
- VulnCheck Advisory: InternLM LMDeploy through 0.17.0 Assertion Denial of Service third-party-advisory