segfault with potential race condition

[terlar]

I have had this issue for quite a long time, but since it is not always I haven't taken the time to investigate or report it. It happens when I start up the computer and log in to my user account. I can quite easily reproduce it by rebooting and loging in several times. Usually my solution if it happens is to reboot and then it usually works the next time.

The result is that the systemd unit don't successfully start and sometimes affects other parts of the user systemd making that unusable.

I get this line in dmesg:

envoyd[481]: segfault at 0 ip 00007ff346ba22ff sp 00007ffc04f53b18 error 4 in libc-2.22.so[7ff346b0c000+19b000]

I get these in the journal:

Oct 25 09:31:25 falcon systemd-logind[518]: New session c1 of user terje.
Oct 25 09:31:26 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:27 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:28 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:29 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:30 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:31 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:33 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:34 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:35 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:36 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:37 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:38 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:40 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:41 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:42 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:43 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:44 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:45 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:47 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:48 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:49 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:50 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:51 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:52 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:54 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:55 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:56 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:57 falcon systemd[1]: Looping too fast. Throttling execution a little.
Oct 25 09:31:57 falcon login[1208]: pam_unix(login:session): session opened for user terje by (uid=0)
Oct 25 09:31:57 falcon login[1208]: pam_systemd(login:session): Cannot create session: Already occupied by a session
Oct 25 09:31:57 falcon envoyd[481]: Starting gpg-agent for uid=1000.
Oct 25 09:31:58 falcon kernel: envoyd[481]: segfault at 0 ip 00007ff346ba22ff sp 00007ffc04f53b18 error 4 in libc-2.22.so[7ff346b0c000+19b000]
Oct 25 09:31:58 falcon envoyd[481]: Connection timed out
Oct 25 09:31:58 falcon systemd[1]: [email protected]: Main process exited, code=dumped, status=11/SEGV
Oct 25 09:31:58 falcon systemd[1]: [email protected]: Unit entered failed state.
Oct 25 09:31:58 falcon systemd[1]: [email protected]: Failed with result 'core-dump'.
Oct 25 09:31:58 falcon systemd[1]: envoy-gpg-agent-monitor-1000.scope: Failed to add PIDs to scope's control group: No such process
Oct 25 09:31:58 falcon systemd[1]: Failed to start Envoy agent monitoring scope.
-- Subject: Unit envoy-gpg-agent-monitor-1000.scope has failed
-- Defined-By: systemd
-- Support: http://lists.freedesktop.org/mailman/listinfo/systemd-devel
-- 
-- Unit envoy-gpg-agent-monitor-1000.scope has failed.
-- 
-- The result is failed.
Oct 25 09:31:58 falcon systemd[1]: envoy-gpg-agent-monitor-1000.scope: Unit entered failed state.
Oct 25 09:31:58 falcon systemd[1]: Created slice user-1000.slice.
-- Subject: Unit user-1000.slice has finished start-up
-- Defined-By: systemd
-- Support: http://lists.freedesktop.org/mailman/listinfo/systemd-devel
-- 
-- Unit user-1000.slice has finished starting up.
-- 
-- The start-up result is done.
Oct 25 09:31:58 falcon systemd[1]: Starting User Manager for UID 1000...
-- Subject: Unit [email protected] has begun start-up
-- Defined-By: systemd
-- Support: http://lists.freedesktop.org/mailman/listinfo/systemd-devel
-- 
-- Unit [email protected] has begun starting up.
Oct 25 09:31:58 falcon systemd[1]: session-c1.scope: Failed to add PIDs to scope's control group: No such process
Oct 25 09:31:58 falcon systemd[1]: Failed to start Session c1 of user terje.
-- Subject: Unit session-c1.scope has failed
-- Defined-By: systemd
-- Support: http://lists.freedesktop.org/mailman/listinfo/systemd-devel
-- 
-- Unit session-c1.scope has failed.
-- 
-- The result is failed.
Oct 25 09:31:58 falcon systemd[1]: session-c1.scope: Unit entered failed state.

I am not sure if the Looping too fast. Throttling execution a little. is related in any way, but included it just in case. Please let me know if I can further debug this.

What makes me believe that this might be a race condition is that it sometimes works and sometimes doesn't.

[vodik]

This is something I've been experiencing for a while but have just been too damn busy these past few weeks to investigate. For what its worth, systemctl daemon-reexec will stop the log spam. But this is something that needs fixing.