InternLM lmdeploy up to 0.15.x Disaggregated Serving engine_conn.py recv_pyobj deserialization

Summaryinfo

A vulnerability classified as critical was found in InternLM lmdeploy up to 0.15.x. The impacted element is the function recv_pyobj of the file lmdeploy/pytorch/disagg/conn/engine_conn.py of the component Disaggregated Serving. Such manipulation leads to deserialization. This vulnerability is traded as CVE-2026-76850. The attack may be launched remotely. There is no exploit available.

Detailsinfo

A vulnerability, which was classified as critical, has been found in InternLM lmdeploy up to 0.15.x. Affected by this issue is the function recv_pyobj of the file lmdeploy/pytorch/disagg/conn/engine_conn.py of the component Disaggregated Serving. The manipulation with an unknown input leads to a deserialization vulnerability. Using CWE to declare the problem leads to CWE-502. The product deserializes untrusted data without sufficiently verifying that the resulting data will be valid. Impacted is confidentiality, integrity, and availability. CVE summarizes:

LMDeploy deserializes disaggregated-serving peer messages with pickle. The handle_zmq_recv coroutine in lmdeploy/pytorch/disagg/conn/engine_conn.py reads peer-to-peer cache-free requests with recv_pyobj(), which deserializes the received bytes with pickle.loads(), and the isinstance check against DistServeCacheFreeRequest runs only after deserialization has already completed. The peer that supplies those bytes is caller-controlled: p2p_connect passes remote_engine_endpoint_info.zmq_address from the request body to connect() on the ZMQ PULL socket, and the POST /distserve/p2p_initialize and /distserve/p2p_connect endpoints in lmdeploy/serve/openai/api_server.py apply no authentication unless the server is started with api_keys, which defaults to None. A remote attacker can direct an engine to pull from a ZMQ endpoint under their control and execute arbitrary code in the engine process. Deployments that do not enable disaggregated serving are not affected, because the receive loop is only started once the migration backend accepts the connection.

The advisory is available at github.com. This vulnerability is handled as CVE-2026-76850 since 08/19/2026. The exploitation is known to be easy. The attack may be launched remotely. No form of authentication is required for exploitation. Technical details are known, but there is no available exploit.

Upgrading to version 0.16.0 eliminates this vulnerability.

If you want to get best quality of vulnerability data, you may have to visit VulDB.

Productinfo

Vendor

• InternLM

Name

• lmdeploy

Version

• 0.1
• 0.2
• 0.3
• 0.4
• 0.5
• 0.6
• 0.7
• 0.8
• 0.9
• 0.10
• 0.11
• 0.12
• 0.13
• 0.14
• 0.15

Website

• Product: https://github.com/InternLM/lmdeploy/

CPE 2.3info

• 🔒
• 🔒
• 🔒

CPE 2.2info

• 🔒
• 🔒
• 🔒

CVSSv4info

CVSSv3info

CVSSv2info

Exploitinginfo

Threat Intelligenceinfo

Countermeasuresinfo

Timelineinfo

Sourcesinfo

Entryinfo

If you want to get best quality of vulnerability data, you may have to visit VulDB.

Discussion