An internal product security audit discovered a UEFI SMM (System Management Mode) callout vulnerability in some ThinkSystem servers that could allow a local attacker with elevated privileges to execute arbitrary code.

An input validation weakness was discovered in XCC that could allow a valid, authenticated XCC user with elevated privileges to perform command injection or cause a recoverable denial of service using a specially crafted file.

A privilege escalation vulnerability was discovered in XCC that could allow a valid, authenticated XCC user with elevated privileges to perform command injection via specially crafted file uploads.

A privilege escalation vulnerability was discovered in XCC that could allow a valid, authenticated XCC user with elevated privileges to perform command injection via specially crafted IPMI commands.

An input validation weakness was discovered in XCC that could allow a valid, authenticated XCC user with elevated privileges to perform command injection through specially crafted command line input in the XCC SSH captive shell.

ThinkSystem SR670V2 servers manufactured from approximately June 2021 to July 2023 were left in Manufacturing Mode which could allow an attacker with privileged logical access to the host or physical access to server internals to modify or disable Intel Boot Guard firmware integrity, SPS security, and other SPS configuration setting. The server’s NIST SP 800-193-compliant Platform Firmware Resiliency (PFR) security subsystem significantly mitigates this issue.

The Remote Mount feature can potentially be abused by valid, authenticated users to make connections to internal services that may not normally be accessible to users. Internal service access controls, as applicable, remain in effect.

An information leak vulnerability in the SMI Set BIOS Password SMI Handler in some Lenovo models may allow an attacker with local access and elevated privileges to read SMM memory.

A buffer overflow exists in the Remote Presence subsystem which can potentially allow valid, authenticated users to cause a recoverable subsystem denial of service.

A valid, authenticated XCC user with read-only permissions can modify custom user roles on other user accounts and the user trespass message through the XCC CLI. There is no exposure if SSH is disabled or if there are no users assigned optional read-only permissions.

A valid XCC user's local account permissions overrides their active directory permissions under specific configurations. This could lead to a privilege escalation. To be vulnerable, LDAP must be configured for authentication/authorization and logins configured as “Local First, then LDAP”.

A valid LDAP user, under specific conditions, will default to read-only permissions when authenticating into XCC. To be vulnerable, XCC must be configured to use an LDAP server for Authentication/Authorization and have the login permission attribute not defined.

A valid, authenticated administrative user can query a web interface API to reveal the configured LDAP client password used by XCC to authenticate to an external LDAP server in certain configurations. There is no exposure where no LDAP client password is configured

A valid, authenticated user may be able to trigger a denial of service of the XCC web user interface or other undefined behavior through a format string injection vulnerability in a web interface API.

A valid, authenticated XCC user with read only access may gain elevated privileges through a specifically crafted API call.

An authenticated XCC user can change permissions for any user through a crafted API command.

An authenticated XCC user with elevated privileges can perform blind SQL injection in limited cases through a crafted API command.  This affects ThinkSystem v2 and v3 servers with XCC; ThinkSystem v1 servers are not affected.

An authenticated XCC user with Read-Only permission can change a different user’s password through a crafted API command.   This affects ThinkSystem v2 and v3 servers with XCC; ThinkSystem v1 servers are not affected.