http://bugzilla.opensuse.org/show_bug.cgi?id=939594 http://bugzilla.opensuse.org/show_bug.cgi?id=939594#c33 Dominique Leuenberger <dimstar@opensuse.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|CONFIRMED |IN_PROGRESS --- Comment #33 from Dominique Leuenberger <dimstar@opensuse.org> --- strace on openQA during boot is not that simple... but I got a journal with systemd debugging enabled.. journal avilable at: https://openqa.opensuse.org/tests/87533/file/journal between when the term permissions where right (bernhard.tty) and they change, the only notable thing happening in the log is:

Oct 06 06:34:38 linux-gle9 systemd[1]: Sent message type=signal sender=n/a destination=n/a object=/org/freedesktop/systemd1/unit/display_2dmanager_2eservice interface=org.freedesktop.DBus.Properties member=PropertiesChanged cookie=180 reply_cookie=0 error=n/a Oct 06 06:34:38 linux-gle9 systemd[1]: Received SIGCHLD from PID 1464 (display-manager). Oct 06 06:34:38 linux-gle9 audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 msg='unit=display-manager comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success' Oct 06 06:34:38 linux-gle9 systemd[1]: Child 1464 (display-manager) died (code=exited, status=0/SUCCESS) Oct 06 06:34:38 linux-gle9 systemd[1]: display-manager.service: Child 1464 belongs to display-manager.service Oct 06 06:34:38 linux-gle9 systemd[1]: display-manager.service: Control process exited, code=exited status=0 Oct 06 06:34:38 linux-gle9 systemd[1]: display-manager.service: Got final SIGCHLD for state start. Oct 06 06:34:38 linux-gle9 systemd[1]: display-manager.service: Main PID loaded: 1546 Oct 06 06:34:38 linux-gle9 systemd[1]: display-manager.service: Changed start -> running Oct 06 06:34:38 linux-gle9 systemd[1]: display-manager.service: Job display-manager.service/start finished, result=done Oct 06 06:34:38 linux-gle9 systemd[1]: Started X Display Manager.

(display-manager.service is the one spawning up gdm, which in turn starts X) in cases where the terminal permission is NOT wrong, those 'Started X Display Manager' messages are ALWAYS before the X logging starts; thus indicating that the position of this in the log does make a difference. which leaves me with two theories so far: * systemd as one * xdm and its display-manager startup script (especially the TakeDevice handler, just after startproc looks very suspicious) For the 2nd, I just submitted a test package / run for openQA with much more verbose outputs.. that should confirm if this is indeed the case. -- You are receiving this mail because: You are on the CC list for the bug.