daroknet/cve-2023-27100
The authentication system attempts to be informative and print extra information along with IP addresses to completely identify where a user logs in from when they login using the GUI. This includes the authentication source (e.g. local database, LDAP or RADIUS, authentication server name), plus contents of proxy headers X-Forwarded-For and Client-IP to further clarify the exact user location.
This extra information is printed after the IP address of the remote user in various places, including log messages for authentication. In the case of GUI login failures, the log entries included the contents of the proxy headers (X-Forwarded-For or Client-IP) submitted by the client.
This extra information confused the sshguard authentication log parser which made it fail to recognize the client IP address in authentication error messages.
pfSense (see related versions above) is vulnerable to a bypass of the anti-brute force mechanism that is in place to block users that have reached the bad authentication limit. The code available in this repository ease this vulnerability exploit using the "X-Forwarded-For" header and the "anti-csrf" token.
Netgate Security Advisory : https://docs.netgate.com/downloads/pfSense-SA-23_05.sshguard.asc
Netgate associated Redmine ticket : https://redmine.pfsense.org/issues/13574
