If I've understood the issue correctly, the current "forwarded" tag on this bug is wrong, in the sense that the fundamental underlying problem is in sudo, and so any possible forwarding waiting for "upstream fix" should point to a sudo bug.
AFAIK this problem with libpam-systemd is caused by the same sudo bug as #648066 (and others). sudo incorrectly forks a child process, does the "open" side of pam authentication there, runs the command to execute in that child, and then runs the "close" side in the parent (which never saw anything about the "open", so lacks the state stored from that!). The specific problem this behavior causes in the libpam-systemd case is that the fd received during the "open" phase (which is part of the pam session state) is closed on exec, terminating the session. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
