Bug#527872: [Pkg-utopia-maintainers] Bug#527872: upgrading dbus or running the init script kills X
On Sat, May 9, 2009 at 23:00:54 +0200, Michael Biebl wrote: 2.) I Xorg can't connect to hal, it should gracefully fall back to a basic kbd/mouse configuration. If X can't connect to hal right away, it listens for a startup notification so you can eventually get your devices even if hal hasn't finished its startup when X starts. http://cgit.freedesktop.org/xorg/xserver/commit/?id=6b8bbbd21b39181e9dd3175fa73c05008c361ba2 Cheers, Julien -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#527872: [Pkg-utopia-maintainers] Bug#527872: upgrading dbus or running the init script kills X
On Sat, May 9, 2009 at 21:40:40 +0200, Norbert Preining wrote: The errors when dbus is restarted is a bit more cryptic, but it seems that the whole stuff of modules is unloaded and then everything hangs: (II) UnloadModule: synaptics (II) Sony Vaio Keys: Close (II) UnloadModule: evdev (II) Video Bus: Close (II) UnloadModule: evdev (II) AT Translated Set 2 keyboard: Close (II) UnloadModule: evdev (II) Video Bus: Close (II) UnloadModule: evdev (II) UnloadModule: synaptics (II) PS/2 Mouse: Close (II) UnloadModule: evdev (EE) intel(0): I830 Dma Cleanup Failed (EE) intel(0): failed to destroy server context (II) intel(0): [drm] removed 1 reserved context for kernel (II) intel(0): [drm] unmapping 8192 bytes of SAREA 0x2efff000 at 0x7f208e434000 (II) intel(0): [drm] Closed DRM master. Well, maybe someone can find out what is the problem here. This looks like a pretty normal (except for the intel driver error) server restart, i.e. your session has exited for some reason? Cheers, Julien -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#527872: [Pkg-utopia-maintainers] Bug#527872: upgrading dbus or running the init script kills X
On Sa, 09 Mai 2009, Michael Biebl wrote: Could you investigate if hald is started correctly when dbus is restarted and running afterwards? How can I check that? I cannot get access to a console and cannot ... wait, maybe I can do on the console a sleep 120 ; /etc/init.d/gdm stop and then switch to X and do the trick. Will try. Norbert, can you re-confirm that your are *not* using awesome, i.e. you are using stock metacity from GNOME. $ ps ax | grep metaci 4836 ?S 0:00 metacity 5305 pts/2R+ 0:00 grep metaci And no awesome running. Best wishes Norbert --- Dr. Norbert Preining prein...@logic.atVienna University of Technology Debian Developer prein...@debian.org Debian TeX Group gpg DSA: 0x09C5B094 fp: 14DF 2E6C 0307 BE6D AD76 A9C0 D2BF 4AA3 09C5 B094 --- WINSTON-SALEM (n.) A person in a restaurant who suggest to their companions that they should split the cost of the meal equally, and then orders two packets of cigarettes on the bill. --- Douglas Adams, The Meaning of Liff -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#527872: [Pkg-utopia-maintainers] Bug#527872: upgrading dbus or running the init script kills X
tags 527872 moreinfo unreproducible thanks Norbert Preining wrote: Package: dbus Version: 1.2.14-2 Severity: important Just upgraded two times dbus and both times I was kicked out of my X (Gnome) session when X restarted. That is not acceptable behaviour, and I tend to raise the severity to serious for this bug, but start out with important. That happens not only when upgrading dbus, also a simply restart of the dbus service (which unfortunately is sometimes necessary to get some hardware working after suspend) triggers a restart of X. Hm, I can not reproduce this problem here, restarting dbus does not kill my X session. Could you check/attach your Xorg log file please Michael -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth? signature.asc Description: OpenPGP digital signature
Bug#527872: [Pkg-utopia-maintainers] Bug#527872: upgrading dbus or running the init script kills X
Norbert Preining wrote: severity 527872 serious title 527872 restarting dbus kills X and kills keyboard thanks On Sa, 09 Mai 2009, Michael Biebl wrote: Hm, I can not reproduce this problem here, restarting dbus does not kill my X session. Could you check/attach your Xorg log file please Even worse, I just tried it and then I remembered that I wanted to report the bug already some weeks ago. When calling /etc/init.d/dbus restart from a running X it restarts X and gdm, but the keyboard is not working. I cannot switch to console, it is all a mystery. I managed to capture everything, the problem is that X reboots but fortunately the old Xorg.0.log is saved. Then the kbd does not work but Sysrq at least. Ok, short: - boot - gdm pops up - I login as user - open a xterm, su, and call /etc/init.d/dbus restart - X dies - the log file of that X is attached as Xorg.0.log-dbus-killed-X - X restarts but the kbd does not work, the log is attached as Xorg.0.log-X-session-after-debus-kill-without-kbd - I sysrq out (sync, remount ro, reboot), the end of the syslog of that kernel run is attached as syslog The only errors in the Missing kbd session is (EE) config/hal: couldn't initialise context: (null) ((null)) Looks like hal is not correctly started when you restart dbus, and X is hal to get it's input devices (thus no keyboard or mouse). So imho we have two problems here (not really problems in dbus itself) 1.) hal does not seem to come up properly after dbus has been restarted 2.) I Xorg can't connect to hal, it should gracefully fall back to a basic kbd/mouse configuration. Could you investigate if hald is started correctly when dbus is restarted and running afterwards? Cheers, Michael -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth? signature.asc Description: OpenPGP digital signature
Bug#527872: [Pkg-utopia-maintainers] Bug#527872: upgrading dbus or running the init script kills X
Michael Biebl wrote: Norbert Preining wrote: severity 527872 serious title 527872 restarting dbus kills X and kills keyboard thanks On Sa, 09 Mai 2009, Michael Biebl wrote: Hm, I can not reproduce this problem here, restarting dbus does not kill my X session. Could you check/attach your Xorg log file please Even worse, I just tried it and then I remembered that I wanted to report the bug already some weeks ago. When calling /etc/init.d/dbus restart from a running X it restarts X and gdm, but the keyboard is not working. I cannot switch to console, it is all a mystery. I managed to capture everything, the problem is that X reboots but fortunately the old Xorg.0.log is saved. Then the kbd does not work but Sysrq at least. Ok, short: - boot - gdm pops up - I login as user - open a xterm, su, and call /etc/init.d/dbus restart - X dies - the log file of that X is attached as Xorg.0.log-dbus-killed-X - X restarts but the kbd does not work, the log is attached as Xorg.0.log-X-session-after-debus-kill-without-kbd - I sysrq out (sync, remount ro, reboot), the end of the syslog of that kernel run is attached as syslog The only errors in the Missing kbd session is (EE) config/hal: couldn't initialise context: (null) ((null)) Looks like hal is not correctly started when you restart dbus, and X is hal to get it's input devices (thus no keyboard or mouse). So imho we have two problems here (not really problems in dbus itself) 1.) hal does not seem to come up properly after dbus has been restarted 2.) I Xorg can't connect to hal, it should gracefully fall back to a basic kbd/mouse configuration. Could you investigate if hald is started correctly when dbus is restarted and running afterwards? Ok, this is what I found out so far: If you restart dbus *within* a X session and you are using the awesome window manager (can't reproduce it with metacity/gnome, kwin/kde or openbox), you will be kicked out by awesome from your X session as soon as D-Bus is stopped. /etc/init.d/dbus restart will also stop/start its dependended dbus services (like hal), which usually goes like this hal stop dbus stop - you will be kicked out here, thus hal will never be started again. dbus start hal start as the new Xorg relies on hal for input device enumeration, this will leave you without keyboard and mouse configuration when hal is not running. Indeed very nasty. As I said, I can only reproduce this behaviour with the awesome window manager, but not kwin or metacity. This actually leads me into the direction that this might be a window manager related bug. Norbert, can you re-confirm that your are *not* using awesome, i.e. you are using stock metacity from GNOME. Everyone else who commented on this bug report seems to be using awesome. Cheers, Michael -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth? signature.asc Description: OpenPGP digital signature